Bacterial Cell Printing - Springer Link

Cell and Organ Printing

Bradley R. Ringeisen · Barry J. Spargo · Peter K. Wu Editors

Cell and Organ Printing

123

Editors Bradley R. Ringeisen Naval Research Laboratory (NRL) Div. Chemistry Code 6115 4555 Overlook Ave. SW. Washington DC 20375 USA [email protected]

Dr. Barry J. Spargo Naval Research Laboratory (NRL) Chemical Dynamics and Diagnostics Branch Code 6110 4555 Overlook Ave. SW. Washington DC 20375 USA [email protected]

Dr. Peter K. Wu Department of Physics Southern Oregon University 1250 Siskiyou Blvd Ashland, OR 97520 USA [email protected]

ISBN 978-90-481-9144-4 e-ISBN 978-90-481-9145-1 DOI 10.1007/978-90-481-9145-1 Springer Dordrecht Heidelberg London New York Library of Congress Control Number: 2010933518 © Springer Science+Business Media B.V. 2010 No part of this work may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise, without written permission from the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work. Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)

Foreword

Fast forward into the future of medicine. . .not long enough for it to be in the realm of science fiction, perhaps 25–50 years. 2060 is far enough away to dream but short enough that scientists can contribute relevant research today. Will it be possible to grow replacement organs? Could a doctor who is treating someone with advanced heart disease or liver failure simply make a new heart or liver rather than waiting on a lengthy transplant list? This technology is most definitely not available now, with the exception of simple and thin tissues such as bladder membranes or skin grafts. Even these monumental accomplishments of modern medicine are not without their limitations – state-ofthe-art skin grafting procedures usually involve removing skin from other parts of the body (a limited supply), and tissue engineered bladders, while a vast improvement over the use of intestinal tissue, have hurdles before complete in vivo function is achieved. However, these successes, in combination with the growing size of organ transplant waiting lists, highlight the desire and need for organ replacement therapy to become a reality. From a monetary standpoint, tissue engineering and regenerative medicine research are now mainstream disciplines that receive nearly $200 million a year in funding from the United States alone (as of 2001; www.tissueengineering.gov). The sources vary but include the National Institutes of Health, National Science Foundation, the Defense Advanced Research Projects Agency and the Food and Drug Administration. Add this to the thriving private sector R&D budget that is estimated at $3.5+ billion per year [1], and you have a distinct economic signal pointing towards the importance of achieving organ replacement therapy. There are two major hurdles to overcome before achieving this lofty goal. First, a regenerative and sustainable source of cells that will not be rejected by the host’s immune system must be found. This issue is complex and is not addressed by the subject matter found in this book. However, there are several very recent reviews on the subject that point towards novel and/or autologous sources of stem cells for regenerative medicine applications [2–5]. The second obstacle to engineering replacement organs is how to organize specialized cells into three dimensional (3D) tissues. This research usually is focused on one of two approaches: growing cell-seeded scaffolds in vitro or recruitment of cells from the body onto and into molded/fabricated scaffolds post-transplantation. v

vi

Foreword

Unfortunately, both of these methods often result in homogeneous tissues that resemble meat-flavored Jell-OTM more than natural tissues or organs.1 I’m exaggerating here to prove a point – traditional tissue engineering approaches focus on either (a) letting the body do the hard work (vascularization, forming of microscopic cell-cell junctions, controlling cell signaling) or (b) letting materials engineers create intricate cell-free scaffolds. In the later case, the scaffolds usually do not take into account the ultimate goal of generating a heterogeneous tissue made up of multiple cell types, different types of extracellular matrices and interconnected 3D vasculature. Rather, they are usually made from a homogeneous porous material. There has been amazing progress in the sophistication and architecture of tissue engineering scaffolds, including time-release of growth factors and nano-modifications. However, ultimately those scaffolds are only as good as the structure and placement of cells inside them, which usually correlates to homogeneous and uniform distribution, e.g. Jell-O! As for cell organization and growth factor recruitment in vivo, I can’t think of a better tool to use than the body. It’s an impressive machine. However, there are limits to what the body can do when it comes to infiltrating a foreign object (scaffold), self-organizing multiple cell types, generating layers and patterns of different extracellular matrices, etc. Ultimately, tissue engineered organs, no matter how they are created, will most likely utilize some of the body’s natural ability to recruit and differentiate cells, but I am skeptical whether the body can take a polymer or inorganic scaffold that has been randomly seeded with cells and mold it into a complete tissue or organ. A small group of scientists about ten years ago decided that there was a third, and possibly better, way to create a tissue or organ – PRINT IT! Just as a group of contractors would build a house from the bottom up – masons using brick and mortar for the foundation, carpenters using wood for the frame, plumbers using piping to bring clean water to and remove waste from the house – one could potentially use a 3D printer to build tissues from the bottom up, including along the way all the proper components of tissue (different cell types, extracellular matrix, growth factors, vascular network). In this manner, engineered tissues could be transformed from homogeneous cell-flooded scaffold plugs into heterogeneous constructs that have all the right parts in the right place in the right shape and size. This cell or organ printing concept was outlined first in a research article published in the year 2000 in Biotechnology and Bioengineering [6]. These visionaries used a laser pulse to trap and push individual cells through a nozzle and out onto a glass slide, organizing them in straight lines. Shortly after this publication, several other researchers (including myself and my co-Editors) used laser induced forward transfer techniques or desktop ink jet printers to further demonstrate that living cells could be patterned into hydrogel tissue scaffolds. Since these early publications in

1 I have to confess that this term is not my own! Prof. Robert Dennis (University of North Carolina)

coined the term several years ago in a personal communication with me. We were discussing the pros and cons of current tissue engineering practices in the context of cell and organ printing.

Foreword

vii

Number of Cell/Organ Printing Publications per Year

Number of Publications

30 25 20 15 10 5 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 Publication Year

Fig. 1 Tracking the number of cell/organ printing publications over the past decade. 2009 data is for a half year. Via Scopus Search, provided by Elsevier B.V.

2000–2003, the concept of cell and organ printing has blossomed into a rich discipline, including researchers from across the globe. Figure 1 shows how the field of cell and organ printing has grown over the years. Now with 125 total publications, averaging over 20 a year for the last 4 years, researchers have thoroughly demonstrated that living cells can be printed via a number of actuations including electrospray, extrusion via micropens and ejection through photothermal, thermal or optical mechanisms. These are some of the reasons why Prof. Wu, Dr. Spargo, and I decided to edit a book on this subject. The topic has come of age, and it is time to categorize the progress made to date. We have included research that uses printing technology to deposit or guide cells for tissue engineering applications, but for completeness, we have also included chapters describing bacteria printing, biomolecular printing that could be used to build growth or recruitment factors into scaffolds, tissue microdissection as well as live cell printing. The breadth of approaches includes 3D freeform fabrication, ink jet, laser guidance, and modified laser direct write techniques. We hope that this book is not the final word but the first word, defining how these tools have been used to take the first steps towards the ultimate goal of creating heterogeneous tissue mimics. Only time will tell whether cell printers will truly become organ printers, but I think that the technologies described in this book hold promise to achieve what the field of tissue engineering has striven for over the last two decades – to create something more than meat-flavored Jell-OTM ! Washington, DC

Bradley R. Ringeisen

viii

Foreword

References 1. Lysaght M, Nguy A, Sullivan K (1998) An economic survey of the emerging tissue engineering industry. Tissue Eng 4:231 2. Burba I, Devanna P, Pesce M (2010, Jan) When cells become a drug. Endothelial progenitor cells for cardiovascular therapy: aims and reality. Recent Pat Cardiovasc Drug Discov 5(1):1–10 3. Quante M, Wang TC (2009, Dec) Stem cells in gastroenterology and hepatology. Nat Rev Gastroenterol Hepatol 6(12):724–737. Epub 2009 Nov 3 4. O‘Malley J, Woltjen K, Kaji K (2009, Oct) New strategies to generate induced pluripotent stem cells. Curr Opin Biotechnol 20(5):516–521 5. Harris DT (2009) Non-haematological uses of cord blood stem cells. Br J Haematol 147:177–184 6. Odde DJ, Renn MJ (2000) Laser-guided direct writing of living cells. Biotechnol Bioeng 67:312–318

Contents

Part I

Biological Freeform Fabrication

1 3D On-Demand Bioprinting for the Creation of Engineered Tissues Seung-Schik Yoo and Samuel Polio Part II

Ink Jet Approaches

2 Reconstruction of Biological Three-Dimensional Tissues: Bioprinting and Biofabrication Using Inkjet Technology . . . . . . Makoto Nakamura 3 Piezoelectric Inkjet Printing of Cells and Biomaterials . . . . . . . Rachel Saunders, Julie Gough, and Brian Derby Part III

3

23 35

Modified Laser Induced Forward Transfer (LIFT) Approaches

4 Laser-Induced Forward Transfer: A Laser-Based Technique for Biomolecules Printing . . . . . . . . . . . . . . . . . P. Serra, M. Duocastella, J.M. Fernández-Pradas, and J.L. Morenza 5 Biological Laser Printing (BioLP) for High Resolution Cell Deposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bradley R. Ringeisen, C.M. Othon, Xingjia Wu, D.B. Krizman, M.M. Darfler, J.J. Anders and P.K. Wu 6 High-Throughput Biological Laser Printing: Droplet Ejection Mechanism, Integration of a Dedicated Workstation, and Bioprinting of Cells and Biomaterials . . . . . . Fabien Guillemot, Bertrand Guillotin, Sylvain Catros, Agnès Souquet, Candice Mezel, Virginie Keriquel, Ludovic Hallo, Jean-Christophe Fricain, and Joëlle Amedee 7 Absorbing-Film Assisted Laser Induced Forward Transfer of Sensitive Biological Subjects . . . . . . . . . . . . . . . . . . . . B. Hopp, T. Smausz, and A. Nógrádi

53

81

95

115

ix

x

Contents

Part IV

Laser Guidance Approaches

8 Laser Guidance-Based Cell Micropatterning . . . . . . . . . . . . . Zhen Ma, Russell K. Pirlo, Julie X. Yun, Xiang Peng, Xiaocong Yuan, and Bruce Z. Gao Part V

Self Organization and Biological Guidance

9 What Should We Print? Emerging Principles to Rationally Design Tissues Prone to Self-Organization . . . . . . . . . . . . . . N.C. Rivron, J. Rouwkema, R. Truckenmüller, and C.A. van Blitterswijk 10

Biological Guidance . . . . . . . . . . . . . . . . . . . . . . . . . . Jan-Thorsten Schantz and Harvey Chim

11

Patterning Cells on Complex Curved Surface by Precision Spraying of Polymers . . . . . . . . . . . . . . . . . . . . . . . . . . Mauris N. DeSilva

12

Fabrication of Growth Factor Array Using an Inkjet Printer . . . Kohei Watanabe, Tomoyo Fujiyama, Rina Mitsutake, Masaya Watanabe, Yukiko Tazaki, Takeshi Miyazaki, and Ryoichi Matsuda

Part VI 13

137

163

173

187 203

3-Dimensional Scaffold Cell Printing

3D-Fiber Deposition for Tissue Engineering and Organ Printing Applications . . . . . . . . . . . . . . . . . . . . . . . . . . N.E. Fedorovich, L. Moroni, J. Malda, J. Alblas, C.A. van Blitterswijk, and W.J.A. Dhert

225

Part VII Printing Bacteria 14

Bacterial Cell Printing . . . . . . . . . . . . . . . . . . . . . . . . . Bradley R. Ringeisen, Lisa A. Fitzgerald, Stephen E. Lizewski, Justin C. Biffinger, and Peter K. Wu

243

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

257

Contributors

J. Alblas Department of Orthopaedics, University Medical Center Utrecht, 3508 GA, Utrecht, The Netherlands, [email protected] Joëlle Amedee INSERM, U577, Bordeaux, F-33076 France; University of Victor Segalen Bordeaux 2, Bordeaux, F-33076 France, [email protected] J.J. Anders Uniformed Services University for the Health Sciences, Bethesda, MD, USA, [email protected] Justin C. Biffinger Department of Chemistry, Naval Research Laboratory, Washington, DC 20375-5342, USA, [email protected] Sylvain Catros INSERM, U577, Bordeaux, F-33076 France; University of Victor Segalen Bordeaux 2, Bordeaux, F-33076 France, [email protected] Harvey Chim Department of Plastic Surgery, Case Western Reserve University, Cleveland, OH, USA, [email protected] M.M. Darfler Expression Pathology, Inc., Rockville, MD, USA, [email protected] Brian Derby School of Materials, Materials Science Centre, University of Manchester, Manchester, UK, [email protected] Mauris N. DeSilva General Dynamics Information Technology, Fairfax, VA, USA; Naval Medical Research Unit San Antonio, San Antonio, TX, USA, [email protected] W.J.A. Dhert Department of Orthopaedics, University Medical Center Utrecht, 3508 GA, Utrecht, The Netherlands; Faculty of Veterinary Medicine, Utrecht University, 3508 TD, Utrecht, The Netherlands, [email protected] M. Duocastella Departament de Física Aplicada i Òptica, Universitat de Barcelona, Martí i Franquès 1, 08028 Barcelona, Spain, [email protected] N.E. Fedorovich Department of Orthopaedics, University Medical Center Utrecht, 3508 GA, Utrecht, The Netherlands, [email protected]

xi

xii

Contributors

J.M. Fernández-Pradas Departament de Física Aplicada i Òptica, Universitat de Barcelona, Martí i Franquès 1, 08028 Barcelona, Spain, [email protected] Lisa A. Fitzgerald Department of Chemistry, Naval Research Laboratory, Washington, DC 20375-5342, USA, [email protected] Jean-Christophe Fricain INSERM, U577, Bordeaux, F-33076 France; University of Victor Segalen Bordeaux 2, Bordeaux, F-33076 France, [email protected] Tomoyo Fujiyama Department of Life Sciences, The University of Tokyo, Tokyo 153-8902, Japan, [email protected] Bruce Z. Gao Department of Bioengineering and COMSET, Clemson University, Clemson, SC, USA, [email protected] Julie Gough School of Materials, Materials Science Centre, University of Manchester, Manchester, UK, [email protected] Fabien Guillemot INSERM, U577, Bordeaux, F-33076 France; University of Victor Segalen Bordeaux 2, Bordeaux, F-33076 France, [email protected] Bertrand Guillotin INSERM, U577, Bordeaux, F-33076 France; University of Victor Segalen Bordeaux 2, Bordeaux, F-33076 France, [email protected] Ludovic Hallo Centre Lasers Intenses et Applications, UMR 5107 CEA V CNRS – Université Bordeaux 1, 33405 Cedex, France, [email protected] B. Hopp Research Group on Laser Physics, Hungarian Academy of Sciences and University of Szeged, Dóm tér 9, H-6720 Szeged, Hungary, [email protected] Virginie Keriquel INSERM, U577, Bordeaux, F-33076 France; University of Victor Segalen Bordeaux 2, Bordeaux, F-33076 France, [email protected] D.B. Krizman Expression Pathology, Inc., Rockville, MD, USA, [email protected] Stephen E. Lizewski Department of Chemistry, Naval Research Laboratory, Washington, DC 20375-5342, USA, [email protected] Zhen Ma Department of Bioengineering and COMSET, Clemson University, Clemson, SC, USA, [email protected] J. Malda Department of Orthopaedics, University Medical Center Utrecht, 3508 GA, Utrecht, The Netherlands; Institute of Health and Biomedical Innovation, Queensland University of Technology, Brisbane, QLD 4001, Australia, [email protected]

Contributors

xiii

Ryoichi Matsuda Department of Life Sciences, The University of Tokyo, Tokyo 153-8902, Japan, [email protected] Candice Mezel Centre Lasers Intenses et Applications, UMR 5107 CEA V CNRS – Université Bordeaux 1, 33405 Cedex, France, [email protected] Rina Mitsutake Department of Life Sciences, The University of Tokyo, Tokyo 153-8902, Japan, [email protected] Takeshi Miyazaki Canon Inc., Tokyo 146-8501, Japan, [email protected] J.L. Morenza Departament de Física Aplicada i Òptica, Universitat de Barcelona, Martí i Franquès 1, 08028 Barcelona, Spain, [email protected] L. Moroni Institute for BioMedical Technology (BMTI), University of Twente, 7500 AE, Enschede, The Netherlands, [email protected] Makoto Nakamura Graduate School of Science and Engineering for Research, University of Toyama, Toyama, Japan, [email protected] A. Nógrádi Department of Ophthalmology, University of Szeged, Korányi fasor 10–11, H-6720 Szeged, Hungary, [email protected] C.M. Othon Physics Department, Wesleyan University, Wesleyan Station, Middletown, CT 06459, USA, [email protected] Xiang Peng Institute of Optoelectronics, Shenzhen University, Shenzhen, Guang Dong, PR China, [email protected] Russell K. Pirlo Department of Bioengineering and COMSET, Clemson University, Clemson, SC, USA, [email protected] Samuel Polio Department of Biomedical Engineering, Boston University, Boston, MA, USA, [email protected] Bradley R. Ringeisen Department of Chemistry, Naval Research Laboratory, Washington, DC 20375-5342, USA, [email protected] N.C. Rivron MIRA Institute for Biomedical Technology and Technical Medicine, University of Twente, 7500 AE Enschede, The Netherlands, [email protected] J. Rouwkema MIRA Institute for Biomedical Technology and Technical Medicine, University of Twente, 7500 AE Enschede, The Netherlands, [email protected] Rachel Saunders School of Materials, Materials Science Centre, University of Manchester, Manchester, UK, [email protected]

xiv

Contributors

Jan-Thorsten Schantz Department for Plastic and Hand Surgery, Klinikum rechts der Isar Technische Universität München, 81675 München, Germany, [email protected] P. Serra Departament de Física Aplicada i Òptica, Universitat de Barcelona, Martí i Franquès 1, 08028 Barcelona, Spain, [email protected] T. Smausz Department of Optics and Quantum Electronics, University of Szeged, Dóm tér 9, H-6720 Szeged, Hungary, [email protected] Agnès Souquet INSERM, U577, Bordeaux, F-33076 France; University of Victor Segalen Bordeaux 2, Bordeaux, F-33076 France, [email protected] Yukiko Tazaki Department of Life Sciences, The University of Tokyo, Tokyo 153-8902, Japan, [email protected] R. Truckenmüller MIRA Institute for Biomedical Technology and Technical Medicine, University of Twente, 7500 AE Enschede, The Netherlands, [email protected] C.A. van Blitterswijk Institute for BioMedical Technology (BMTI), University of Twente, 7500 AE, Enschede, The Netherlands, [email protected] Kohei Watanabe Department of Life Sciences, The University of Tokyo, Tokyo 153-8902, Japan; Canon Inc., Tokyo 146-8501, Japan, [email protected] Masaya Watanabe Department of Life Sciences, The University of Tokyo, Tokyo 153-8902, Japan; Canon Inc., Tokyo 146-8501, Japan, [email protected] Xingjia Wu Uniformed Services University for the Health Sciences, Bethesda, MD, USA, [email protected] Peter K. Wu Department of Physics, Southern Oregon University, Ashland, OR 97520, USA, [email protected] Seung-Schik Yoo Department of Radiology, Harvard Medical School, Brigham and Women’s Hospital, Boston, MA, USA, [email protected] Xiaocong Yuan Key Laboratory of Optoelectronic Information Science and Technology, Institute of Modern Optics, Ministry of Education of China, Nankai University, Tianjin, PR China, [email protected] Julie X. Yun Department of Bioengineering and COMSET, Clemson University, Clemson, SC, USA, [email protected]

Part I

Biological Freeform Fabrication

Chapter 1

3D On-Demand Bioprinting for the Creation of Engineered Tissues Seung-Schik Yoo and Samuel Polio

Abstract Three-dimensional freeform fabrication, a technique which capitalizes on the ability to print various biological materials and cells along with various tissue scaffold materials, is gaining popularity in tissue engineering due to its potential role in the creation of biomimetic tissues and organs. The flexibility to design and create various 3D cell-scaffold composites gives direct bioprinting a significant advantage over conventional lithography-based approaches in tissue engineering. In this chapter, we present our computer-assisted 3D biological printer, which allows dispensing of various types of hydrogel-based scaffold materials and cells, as well as the techniques to construct multi-layered cell-hydrogel composites. The strategies to generate hydrogel channels and to embed hydrogel matrix to time-release watersoluble factors are introduced together with several production examples using adult mammalian cells and stem cells for the on-demand composition of artificial tissues.

List of Abbreviations 3D ALP DMEM ECM ELA FB FF GUI KC MeHA MEMS PDMS

Three-dimensional Alkaline phosphatase Dulbecco’s modified Eagle’s medium Extracellular matrix Embryonic like-adult (stem cells) Fibroblasts Freeform fabrication Graphical-user-interface Keratinocytes Methacrylated hyaluronic acid Micro-electro-mechanical systems Polydimethylsiloxane

S.-S. Yoo (B) Department of Radiology, Harvard Medical School, Brigham and Women’s Hospital, Boston, MA, USA e-mail: [email protected] B.R. Ringeisen et al. (eds.), Cell and Organ Printing, C Springer Science+Business Media B.V. 2010 DOI 10.1007/978-90-481-9145-1_1,

3

4

RP TED TTL UV VEGF

S.-S. Yoo and S. Polio

Rapid prototyping Thermoelectric device Transistor-transistor logic Ultraviolet Vascular endothelial growth factor

1.1 Introduction In spite of advances in organ transplantation and the understanding of mechanisms behind the injury and pathology of human organs, the technology for tissue/organ regeneration and repair is still in its infancy. Recently, there has been a growing interest in tissue engineering whereby the artificially-engineered tissue products made of naturally-driven or synthetic biomaterials are used to replace damaged or defective tissues with strategic placements of cells exhibiting normal functional potency. Various tissue engineering techniques can be employed to prepare three-dimensional (3D) cellular environments that mimic the natural physiological/geometrical conditions of major organ systems [1]. The construction of such a cellular assembly will also serve as a useful biomedical research tool to help understand the pathology of specific diseases. Therefore, it may open potential clinical applications such as drug-testing/screening platforms, control of stem cell differentiation/proliferation, and, ultimately, the regeneration of human organs [2, 3]. Also, a stand-alone mini artificial organ system, based on integration of a bioreactor to promote the culture and growth of an engineered tissue, has been sought after [4]. By constructing an artificial tissue from an assembly of cells and biomaterials, tissue engineering techniques can offer 3D cellular environments that mimic the natural physiological/geometrical architecture of a biological tissue. Stratified layers of skin, lobular structures found in liver/endocrine organs, and neural tissue architecture in brain cortices are a few examples of such 3D structures. Traditionally, the cells-of-interest are embedded in or coated on biocompatible materials that serve as temporary scaffolds to support their integration into the existing physiological environment [1]. Volumetric solid scaffolds with built-in porosity have been assembled to provide the cells with adequate perfusion as well as surface-attachment [5]. However, culturing cells on/within porous solid 3D scaffolds often does not provide the appropriate tissue architecture, not to mention the one found in a soft tissue; thus making them unsuitable for engineering complex and complete organs-of-interest. Synthetic and naturally-driven hydrogels have gained popularity as candidate scaffold materials for non-skeletal tissue engineering due to their ability to facilitate the transport of oxygen through diffusion and to integrate readily into the surrounding extracellular matrix (ECM). Controllable dissociation/biodegradation of hydrogels in physiological environments also makes hydrogels useful materials in tissue engineering [6]. With these advantages over solid scaffold materials, hydrogels have been used to provide the necessary 3D cell culture environments for chondrocytes [7] and hepatocytes [8]. With the advancement of micro-electro-mechanical systems (MEMS) technology and microfluidics, mammalian cells can be patterned and cultured in 3D

1

3D On-Demand Bioprinting for the Creation of Engineered Tissues

5

[9, 10]. Although these soft-lithographic approaches can reliably generate biological patterns with accuracy on the cell-level, it needs sophisticated production and co-alignment of individual lithographic masks to assemble a multi-layered cell structure, rendering the high-throughput, rapid construction of a tissue with heterogeneous cells difficult to achieve. Recently, a new breed of techniques has emerged to enable direct injection/printing of cells along with a hydrogel-based scaffold material in an attempt to construct tissues and organs in 3D [11]. Freeform fabrication (FF), also known as rapid-prototyping (RP), has been applied to create an engineered tissue by dispensing and patterning cells and other tissue constructs as liquid droplets, in a layer-by-layer fashion, onto the desired spatial locations without the use of lithography [12–14]. The techniques, referred to as ‘direct cell printing’ or ‘bioprinting’, are based on several different methods of cell dispensing. These include cell dispensing using ink-jet [11] or laser-printing technology [15, 16]. Ink-jet cell printing uses a modified commercial version of bubble jet or piezoelectric printer to dispense cells on the hydrogel material to form a layer of cell-hydrogel composites. The laser-printing technique is based on focusing a high-energy laser pulse onto a spot above the cellladen gel and subsequent dispensing of the cells underneath the ‘evaporated’ spot. We have developed a bioprinting platform, using miniature electro-mechanical microvalves as dispensers, to construct 3D cell-hydrogel tissue composites. The printer, which can accommodate up to 8 different printing materials (loaded in the ‘bio cartridge’), provided the opportunity to create several important features of tissue architecture, all in 3D. The features include; (1) multi-cellular multi-layered 3D tissue architecture, (2) fluidic channels for perfusion and for lumen-formation, and (3) a scaffold which integrates water-soluble growth factors/cytokine and releases them over time. In this chapter, we introduce the hardware and software architecture of the bio-printer, which support the printing of various classes of hydrogel materials (chemically, enzymatically, thermally, photo-crosslinkable) and mammalian cells (including stem cells). The flexibility of the bio-printing was demonstrated through examples of (1) multi-layered multi-cell printing directly onto non-planar surface; (2) constructing fluidic channels within the printed hydrogel, and; (3) printing fibrin gel to release growth factors to the cell containing collagen scaffold. The further potential utility in regenerative medicine is introduced via bio-printing of embryonic-like adult (ELA) stem cells to create an osseous tissue.

1.2 Biological Printer 1.2.1 Hardware Architecture of the Biological Printer The 3D bioprinter is operated with four dispensing/printing channels (print >4 different liquid-state compositions in a single batch), based on rapid gating of fluidic paths under pneumatic pressure. The main advantage of using a pneumaticallydriven electromechanical valve is that various types of liquid materials with

6


viscosities up to 200 Pa·s can be dispensed. In addition, the individual droplet size, depending on the viscosity of the material, can be controlled by adjusting the pressure and valve gating time. Based on Bernoulli’s principle, the speed of the ejected droplet is controlled by regulating pressure to the syringe. Low ejection velocity (typically less than 3 m/s) and low operational pneumatic pressure (on the order of 1–3 psi) promote high cell viability. The general operating principle of the printing mechanism is explained as follows. Cell suspensions, un-crosslinked hydrogel precursors, or growth factors, all in liquid form, are placed in 5 or 10 mL disposable syringes, i.e., biological printing cartridges. The air pressure for dispensing was provided by a pressurized air tank (HEPA filtered), and the pressure to each syringe was controlled by a digital pressure regulator (ITV-2010; SMC, Japan). The tube from the syringe outlet was linked to a set of electromechanical microvalves (SMLD; Fritz Gyger AG, ThunGwatt, Switzerland, 150 μm nozzle diameters) in which the fluidic pathway from the syringe was gated. The gating signal was provided by using a standard TTL (Transistor-transistor logic) pulse (pulse width > 175 μs). The maximum duty cycle allowed dispensing at a rate of at least 1,000 Hz. The overall schematic of the printing hardware is shown in Fig. 1.1. The printer consists of an array of four microvalves, the printing head, mounted onto a horizontal (x–y) robotic stage (Newmarksystems, CA), which controlled the timing and location of dispensing of printable materials. The target substrate was mounted to another robotic stage that moved along the vertical direction. The entire device was housed in a laminar flow hood. A high-speed camera (Pixelink PL-A741;

Fig. 1.1 Picture of the modular tissue printing platform; (1) 4 syringes as ‘cartridges’ to load cell suspensions and hydrogel precursors; (2) an array of 4-channel dispensers; (3) horizontal stage; (4) vertical stage; (5) target platform; (6) target substrate; (7) camera, (8) stage heater, vertical stage heater; (9) independent heated/cooled dispenser

1


7

Ottawa, Canada) was used to measure the droplet size while two video cameras (UBV-49; Logitech, CA and MW200; MobiTechPlus Inc, Korea) monitored the stage movement as well as construction of the printed tissue. One of the dispenser units as well as the vertical stage were temperature controlled (operating temperature between 5 and 40◦ C) by a solid-state thermoelectric device (TED; TE Technology, Traverse City, MI). To accommodate ultraviolet (UV)-crosslinkable hydrogel such as methacrylated hyaluronic acid (MeHA) [6], the printer was made to be UV-resistive and had an adapter for the light-source. Syringes and tubing used in this experiment were disposable, while all the hardware parts were designed in detachable modules for easy assembly and modification. The distance between the dispenser nozzle and target substrate is detected by an ultrasonic range finder (SRF04; Devantech, Norfolk, UK) mounted on the dispenser array and controlled by adjusting the location of vertical stage.

1.2.2 Software Interface and Hardware Implementation We have developed comprehensive software with a user-friendly graphical-userinterface (GUI) to operate the bio-printer (schematic shown in Fig. 1.2). To provide the maximal flexibility for generating printing patterns with different spatial

Fig. 1.2 Schematics of implementation of the 3D bioprinter. Input images can be chosen from a variety of sources including CAD files or 3D radiological images. In-house software generated dispensing coordinates/vectors as well as the printing sequence whereby the user controlled the dispensing resolutions and gradients through the graphic-user-interface (GUI). The printing information, after conversion to the robot controller language, was fed to the printer. The volume of droplet was adjusted independently by controlling the pneumatic pressure to the fluid paths or the opening duration of the valve-based dispenser

8


resolutions and gradients, the software consisted of mainly two parts – (a) coordinate generator and (b) printer controller. First, the coordinate generator was implemented using the MATLAB computation environment (Mathworks, Natick, MA). To be compatible with various image sources such as radiological data, 3D layer-by-layer image sets, or 2D picture images, the image sources can be loaded into the software as a grayscale bitmap image, and then droplet dispensing coordinates are sampled from the loaded images based on the user-defined printing resolution. The number of voxels/pixels of the loaded image was made to exceed the user-defined spatial resolution using image-editing software. Most generic image-editing software, for example, Photoshop (Adobe), can be used to create the images. The printing path can be defined in either sequential line printing or boundary-printing followed by sequential filling, which has been adopted for many types of commercial plotters. We also added an algorithm to minimize the printing path (thus reducing the printing time) by clustering and sorting the printing coordinates. The optimized coordinates and printing sequence were then used as input data for the printer controller. The distance between each dispensing point (which determine the printing resolution) and with the desired printing dimension, were user-definable. The printer controller software, developed using Visual Basic (Microsoft), generates the control codes for robotic stages (Galil Motion Control, Inc., Rocklin, CA) and droplet dispensing. The robot movements can also be controlled manually. Printing progress was visualized as dots on the printed coordinate in the GUI in real-time for users.

1.2.3 Flexibility in Bioprinting Using Electromechanical Valves We have printed more than 30 human and animal cell lines (e.g. chondrocytes, C17.2, PC6, AML12, C166, HL1, HeLa, neural precursor cells, mesenchymal stem cells, and mouse embryonic stem cells), and tested the cell viability after printing using a commercially available live/dead assay kit (calcein AM/ ethidium homodimer-1, Invitrogen, MA). There was no significant difference in cell viability (all > 95%) compared to manually plated cells, suggesting that our cell printing technique can be generally applied to most of cell types. Since the valve (with nozzle diameter of 150 μm) and controlling mechanism allow for dispensing viscous liquid materials, along with an independent temperature control for the dispenser itself, a wide range of hydrogel materials have been successfully printed and crosslinked. For example, temperature-sensitive gelatin and agarose are heated and completely liquefied, and are maintained in liquid form using a temperature-controlled unit at 40◦ C (Fig. 1.1 (9)) for dispensing. Thermo-reversible hydrogel, such as PEG triblock materials (liquid-gel transition temperature of 37◦ C) [17], have been successfully printed by cooling the dispenser to 5◦ C. Photo-crosslinkable hydrogel precursor, MeHA (2% weight concentration in 1x PBS) which was mixed with photoinitiator (Iragacure 2959; 1.5% weight), has been printed and crosslinked by a UV source in the 270–290 nm bands. Enzymatically-crosslinked fibrin gel was formed by printing both fibrinogen and thrombin solution onto the same location in the scaffold. One

1


9

of the most widely-used hydrogel materials in tissue engineering, collagen, which is pH-sensitive (liquid in acidic environment and gelled in neural condition), was also printed using our novel method to create multi-layered cell hydrogel composites based on the nebulization of a sodium bicarbonate (NaHCO3 ) solution. Chemicallycrosslinkable alginate gel was constructed by printing sodium alginate solution (1wt%) and crosslinked using nebulized calcium chloride (CaCl2 ) solution. The concentration of any desired material should be optimized before printing since high concentrations would make the hydrogel precursor too viscous to be printed.

1.3 Method of Constructing Multi-layered Cell-Hydrogel Composites To construct multi-layered cell-hydrogel composites in 3D, collagen or sodium alginate solution, the deposited solution must be crosslinked to form a hydrogel layer before printing any subsequent layers. One of the hurdles preventing multi-layered construction of a hydrogel-cell composite is that printed hydrogel precursors (before gelation) are prone to being distorted during the application of the cross-linking material [18]. The dispensing of low-viscosity hydrogel precursors and crosslinking agents onto the same location does not generate the desired printing pattern since two liquid drops, when placed in close proximity, immediately merge together due to surface tension. Conventionally, either the hydrogel precursor or crosslinker can be patterned onto a target substrate and then submerged into the appropriate crosslinker solution (Fig. 1.3) [11, 13]. These methods, however, carry the risk of washing off the printed product during the crosslinking process. In addition, they require a separate container (illustrated in Fig. 1.3) to prepare a leveled surface of hydrogel or

Fig. 1.3 Illustration of schematics used in conventional bio-printing. This method is prone to distortion of hydrogel. In addition, the constructed cell-containing hydrogel is not readily integrated into a separate cell culture platform such as a bioreactor

10


crosslinking materials. Therefore, one needs to detach the printed hydrogel block from the vertical stage for further tissue culture.

1.3.1 Technique Enabling Multi-layered Construction of Cell-Hydrogel Composites We developed a novel method [18] to enable the construction of multi-layered 3D hydrogel composites. The overall schematic of the process is illustrated in Fig. 1.4. The substrate surface (either a Petri dish or polydimethylsiloxane-PDMS) is first coated with a crosslinking agent using nebulization via an ultrasonic transducer (SU-1051 W, Sunpentown, CA). The uncrosslinked hydrogel precursor layer is then printed on the coated surface and crosslinked to form a gel. During this process, the generation of ultra-fine mists with droplets less than 2 μm in diameter is crucial to crosslink the dispensed collagen precursors without macroscopically distorting the printing morphology. The droplets of cell suspension in culture media are then dispensed on the partially-crosslinked hydrogel layer to embed cells in the hydrogel. Nebulized crosslinking solution is applied to crosslink the remainder of the hydrogel layer. The crosslinker coating on the top surface serves as the crosslinking material for the next layer to be printed. The process is repeated to construct multiple layers of cell/hydrogel composites in 3-D. It is important to note that this technique allows the tissue layers to be printed directly onto non-planar surfaces without the need for a separate container to house crosslinking materials. This capability is important to expedite the fabrication process whereby the tissue/organ can be printed directly onto a PDMS-based bioreactor enclosure.

Fig. 1.4 The schematic of multi-layered composition of hydrogels and cells using administration of the nebulized crosslinking agent on the printed hydrogel precursors

1.3.2 3D Cell-Hydrogel Scaffold Construction The bioprinter-based 3D hydrogel fabrication method was first applied in creating multi-layered engineered tissue composites consisting of human skin fibroblasts (FB) and keratinocytes (KC), which mimicked skin layers [18]. To demonstrate the ability to print and culture multi-layered cell-hydrogel composites on a non-planar

1


11

surface such as for skin wound-repair, the tissue composites were printed directly on PDMS-based target substrates with 3D surface contours. Primary adult human dermal FB and primary adult human epidermal KC (ScienCell Laboratory, Carlsband, CA) were cultured in standard conditions of 37◦ C and 5% CO2 . FB media was supplemented with 2% fetal bovine serum and 1% FB growth supplements while 1% KC growth supplements were added to KC media. 1% penicillin-streptomycin was also supplemented to both culture media. Using the method described to enable construction of multi-layer cell-collagen composites, a total of 10 layers of cell-containing collagen (Rat tail origin type-I) were sequentially printed in planar square pattern on a 60 mm tissue culture dish. The surface of the tissue culture dish or PDMS surface was coated with nebulized NaHCO3 solution, and the un-crosslinked collagen precursor was printed on the coated surface and crosslinked to form a gel due to the pH change. FB and KC layers were located in the second and the eighth layer of collagen hydrogel (counted from the bottom layer), respectively. Five layers of collagen were sandwiched between the layers of FB and KC to demonstrate the presence of spatially-distinctive cell layers. After testing different printing resolutions (from 200 to 900 μm, with a step of 100 μm), the printing resolution of 300 μm was selected for subsequent 3D printing experiments. The distance between the nozzle and the target substrate was maintained at 5 mm. Figure 1.5 shows side-projected confocal microscopy images of printed, multilayered artificial skin tissue at day 4 of culture after immunostaining. The clear distinctive layers of FB and KC were visible. Figure 1.5a (β-tublin labeling) illustrates the both bottom and upper cell layers contain β-tublin. The FB layer, approximately 100 μm below the surface of the culture media, shows an extensive tree-like morphology, which is common in a 3D culture environment (Fig. 1.5b).

Fig. 1.5 Cell images after multi-layered printing of FB and KC on the tissue culture dish. Volume rendered immunofluorescent images of multi-layered printing of KC and FB and its projection of (a) keratin-containing KC layer and (b) β-tublin-containing KC and FB. The inter layer distance was approximately 75 μm. Bar=150 μm (Adapted from Biomaterials [18])

12


The technique was also applied to pattern rat embryonic neural cell lines (neurons and astrocytes) in a 3-D collagen block with high viability [19]. Rat tail origin type I collagen was also used as a hydrogel precursor. First, the precursor was diluted to 1.12 mg/mL with 0.02 N acetic acid solution (CH3 COOH) and 1X PBS (volume ratio of 1:1:2) and was kept in an ice bath. The dilution factor was determined from the collagen density that showed the neurite outgrowth among three different densities of collagen (3). Rat embryonic day 18 neurons (BrainBits LLC, Springfield, IL) at concentration of 3.0 × 106 cells/mL were suspended in Neurobasal Media (Gibco, San Diego, CA) with 2% B27 Supplement (Gibco), 0.5 mM glutamine (Gibco), and 25 μM glutamate (Gibco). Prior to printing, the cells were stained with CellTrackerTM CMDiI (Invitrogen) according to the manufacturer’s instructions in order to visualize the cells and replated. After ensuring cell viability after staining, the cells were trypsinized and loaded into a bio-cartridge. Using the multi-layered printing mechanism illustrated in Fig. 1.4, the process was repeated to construct multiple layers of collagen and cells. Each collagen precursor layer was printed to cover a 10 × 10 mm2 square area using the inter-dispensing distance (spatial resolution) of 600 μm. The droplets of collagen precursor were printed at 1.8 psi with a valve opening time of 500 μs, and the neurons were printed at 1.2 psi with a valve opening time of 500 μs. After printing, the neural cell-collagen composites were cultured at 37◦ C and 5% CO2 in Neurobasal media with 2% B27 supplement, 0.5 mM glutamine, and 25 μM glutamate. Half of the media was replaced with fresh media without glutamate every 3 or 4 days. Figure 1.6 shows the constructed cell-collagen composites with 8 layers of collagen, which has neurons embedded within the 2nd, 4th, and 6th layers. Between the cell layers, one collagen layer was inserted to secure the cross patterning of cells. After 7 days of culture, neurons were visualized.

Fig. 1.6 3D rendering of a printed, multi-layered 3D collagen hydrogel scaffold with cross pattern of DiI stained rat embryonic neurons

1


13

1.4 Creation of Hydrogel Channel During the culturing of cell-containing tissue composites, the scaffold hydrogel should be adequately perfused for the reception of the growth factors, nutrients, and oxygen as well as for the removal of cellular waste. Several approaches have been developed to achieve this goal. These include constructing PDMS-based capillary networks, culturing cells on a permeable membrane, and embedding microfluidic channels directly onto or within the hydrogel materials. We printed fluidic channels embedded in a multi-layered hydrogel scaffold using on-demand FF in combination with a sacrificial strategy where the channel cavity is first filled with temporary materials that are subsequently removed. The principle behind the construction of fluidic channels is based on Golden’s recent work [20] whereby complicated channels are constructed within hydrogels by the lithographic patterning of gelatin as a sacrificial element. Since the developed bio-printer can print various hydrogel materials in 3-D, the adaptation from this lithographic approach is straightforward. Collagen hydrogel precursor (which may be replaced by fibrin or alginate; to be determined from the optimization) and gelatin (Porcine skin Type A) are used for the construction of the hydrogel scaffold with fluidic channels. Collagen hydrogel precursor diluted to 2.23 mg/mL with 1x PBS is primarily used to construct the hydrogel block which contains the channel structure. Gelatin, 7% by weight, in distilled water was prepared in advance and kept at 40◦ C to liquefy before printing and then loaded into a heated dispenser unit. The schematic shown in Fig. 1.7 illustrates an example of constructing 5 layers of collagen with patterned sacrificial gelatin channels. First, the surface of the tissue culture dish was coated with a nebulized crosslinking solution. An initial layer of collagen was printed to fill a planar area. Nebulized crosslinking solution was applied on the top surface to crosslink the remainder of the printed collagen bed. In the next layer, collagen was patterned while leaving space for the gelatin patterns. After crosslinking the collagen pattern, gelatin was printed within the groove and

Fig. 1.7 Printing scheme for the generation of multi-layered collagen hydrogel scaffold containing two fluidic channels made of gelatin. The gelatin was removed after heating the block, and the channel was filled with colored beads for visualization (right panel)

14


subsequently cooled to room temperature (20◦ C). An additional layer of collagen was printed on top of the hydrogel layer containing the gelatin patterns to seal the channel space. The process was repeated again to print different shapes. After the construction of the 3-D collagen scaffold, the structure was subsequently heated to 37◦ C (by heating target substrates) so that liquefied gelatin in the channels could be carefully removed by perfusion with warmed culture media using a syringe needle. Upon removing the sacrificial gelatin channels, the fluidic channel served as a conduit for media perfusion. At the dispensing conditions of the gelatin (pressure: 6 psi; valve opening duration: 450 μs; printing resolution: 700 μm), a channel width of approximately 400 μm was achieved. Once the printing was completed, the hydrogel structure was kept in an incubator to liquefy the gelatin. A blunt syringe needle was then connected to a syringe pump (NE-1000, New Era Pump Systems, Wantagh, NY) and FB culture media was perfused into the channel lumen at a rate of 4.0 μL/min through the needle. Based on the examination of lumen pressure of the hydrogel channel, the 10 mm-long channel in a collagen structure resisted up to 103.4 mmHg (= 2 psi), which is approximately the average blood pressure in an artery (normally 80~120 mmHg). FBs embedded in the thick hydrogel scaffold showed sustained high viability in locations nearby the channel, while cells embedded in the scaffold at the same depth without the channel showed reduced viability. Our approach demonstrated that 3D bioprinting alone could allow the construction of artificial tissue and the embedded fluidic channels.

1.5 Vascular Endothelial Growth Factor (VEGF) Releasing Fibrin in Collagen Gel The controlled time release of growth factors and cytokines into a tissue engineered scaffold matrix embedded with cells is critical for the development of functional tissues. Growth factors, such as vascular endothelial growth factor (VEGF), can help to promote cell migration, differentiation, viability, growth, or other cellular responses [21], especially when such scaffolds are unable to exploit growth factors present within the body to sustain cellular development [1]. However, the constant infusion of growth factor at the site of implantation is not feasible. Also, a difficulty in using growth factors is that many of them are water soluble, thus they may be removed rapidly from the scaffold and the surrounding environment in a hydrogel scaffold relatively quickly [22]. Thus, the incorporation of a scaffold which releases soluble growth factors would be ideal as it would deliver the factors locally to the site of need over an extended period of time. We successfully incorporated VEGF into hybrid fibrin/collagen gel scaffold, which was able to guide murine neural stem cells from the C17.2 lineage to a target as the VEGF was released over a period of 3 days. To create the fibrin pattern by printing, two separate solutions are prepared in liquid phase. One solution contained fibrinogen (fraction 1, type IV from bovine plasma; Sigma-Aldrich, St. Louis,

1


15

MO) at 62.8 mg/mL and aprotinin (Sigma-Aldrich) at 132 U/mL in PBS. A second solution containing thrombin at 133.2 NIH U/mL, 4.76 μg/μL of heparin, and calcium chloride at 11.8 mg/mL (all acquired from Sigma-Aldrich) was prepared. The VEGF stock solution was diluted with distilled water to 0.1 mg/mL and then used to dilute each of these printing solutions in a 1:1 (by volume) ratio prior to printing. Printing procedure was carried out as follows as shown in Fig. 1.8a. First, 10 μL of each solution were loaded into separate channels of the bio-printer. The solutions were printed at 5 psi and a valve opening time of 500 μs onto a Petri dish in a circular pattern (5 mm in diameter). Subsequently, the thrombin and VEGF containing solution was printed directly on top of the first pattern at 3 psi and a valve opening time of 500 μs, resulting in the formation of a gel. After approximately 2 min, nebulized NaHCO3 solution was applied and a square collagen (1.16 mg/mL) pattern was printed. C17.2 cells were printed adjacent to the fibrin gel in a rectangular pattern at a concentration of 1 × 106 cells/mL, pressure of 1.1 psi, and with a valve opening time of 500 μs. Prior to printing, the cells were maintained at 5% CO2 at 37◦ C and cultured using Dulbecco’s modified Eagle’s medium (DMEM) with 4 mm L-glutamine and 25 mm of D-glucose. The media was further supplemented with 10% fetal bovine serum, 5% horse serum, 100 U/mL penicillin, and 100 μg/mL streptomycin. After printing the cells, the surface of the layer of collagen and cells was nebulized with sodium bicarbonate and crosslinked. The entire cell-hydrogel composites were cultured in 100% humidity without the media. A small amount (100 μL) of serum-free DMEM was placed on top of the scaffold since submersing the entire scaffold in the media would have resulted in the accelerated release of the growth factor. The cells were observed over a period of 3 days (day 0 being the initial printing day and day 3 being the 3rd day of incubation) through compiling images of the cell region up to 1 mm away from the fibrin gel. As shown in Fig. 1.8b, the cells grew and migrated towards the source of the VEGF, which was bound to the fibrin gel. Cells in the control samples without the VEGF did not display such behavior as the cells tended to shrink and die (data not shown). This experiment demonstrated

Fig. 1.8 (a) Schematic of fibrin scaffold with VEGF, and (b) Cell migration in scaffold at day 0 and day 3. These two images show the same area of the collagen scaffold with VEGF at two time points. Cells can be seen moving from the region near the black line on the left to the rightmost line, which represents the edge of the fibrin gel border

16


that fibrin printed inside of collagen using the bio-printer was able to release growth factors in a controlled manner. With further experimental refinement, the kinematics of the release can be analyzed more thoroughly for a fuller understanding of how the growth factor would be released.

1.6 Synovium-Deriven Embryonic Like-Adult Stem Cell Printing for Artificial Bone Recently, a new line of stem cells called embryonic–like adult (ELA) stem cells (Parcell Laboratroy, MA) has been discovered from the synovial fluid of patients with rheumatoid/osteoarthritis or damaged cartilage in the knees [23]. Also known as RASF-MSCs, the ELA has superb self-renewal potency with multi lineage differentiation capability with several key genetic compositions found in embryonic stem cells [23]. For these reasons, ELA cells show great potential in regenerative medicine and tissue engineering. We were motivated to print the ELA stem cells, which have initial morphological characteristics of a mesenchymal stem cell, into the collagen scaffold and further differentiate ELA-collagen composite into osseous tissue. The ability to create artificial bone tissue, which can be tailored to fit the shape of a desired bone through bio-printing, may help to treat damages incurred by diseases such as osteoarthritis, and may replace demineralized bone grafts for skeletal damages. We were motivated to construct osseous tissue based on 3D printing, and implanted the tissue in an animal (rat) for testing of initial immunological compatibility. The ELA cells, printed in passage 3 (duration 200 μs and pressure of 2 psi), were prepared in 1 × 106 cells/mL concentration, and underwent a live/dead viability test (Invitrogen, MA) to examine if the printing process affected the ELA cell’s viability. Printed cells had a survival rate of 98%±0.8, which was indistinguishable from the manually plated cells (99%±0.3; n=5; p>0.05). Upon testing viability, the ELA cells were printed in a ‘dog biscuit’-shaped scaffold (dimensioned 5 × 10 mm2 ), consisting of 3 collagen layers. Two sets of identical patterns, one undergoing osteogenic differentiation and the other without the differentiation as control condition, were printed. Once the printing process was complete, the entire cell-collagen composites were incubated in mesenchymal stem cell basal medium (MesenPro RS; Invitrogen) for a period of 2 weeks. During this expansion period, the media was changed every 3 days, and the cells were observed under the microscope. At the end of this expansion period, one plate was differentiated into osteocytes by supplying osteocyte differential basal medium for 2-weeks. The control plate was supplied with the mesenchymal stem cell basal medium. At the end of this differentiation period, the plates were stained with BCIP/NBT (5-bromo-4-chloro-3-indolyl phosphate/ nitro blue tetrazolium, Sigma Aldrich), which labels the presence of alkaline phosphatase (ALP). ALP is an enzyme that is present in bone tissue, so its presence would indicate that the ELA cells have differentiated into the desired tissue type. Figure 1.9 shows the printed ELA cells in the collagen scaffold under brightfield microscope. After day 18 (Fig. 1.9b), cells reached sufficient confluency before

1


17

Fig. 1.9 (a) ELA cell printed in collagen, (b) the end of expansion period at day 18, and (c) at the end of osteogenic differentiation at day 32. Scale bar=100 μm

differentiation with osteocyte media. Differentiation of the ELA cells into osteocytes can be seen on day 32 (Fig. 1.9c). Prior to BCIP/NBT staining, both plates showed similar printing morphology (Fig. 1.10a, b). Using the stain, only the ELA cells under the osteogenic differentiation showed dark purple colorations on the scaffold (Fig. 1.10c) while the control plate showed no staining, and thus, an absence of ALP (Fig. 1.10d). Although there are obscured cell margins due to the cells outgrowing the scaffold (and thus adhering to the plate surface), the original collagen pattern (marked in dotted line) was relatively well-preserved. We further examined the possibility of transplanting the cultured artificial bone tissue into a bone defect created on a rat skull (Sprag Dawey rat, male, 310 g). First, ten layers of ELA cells in the collagen scaffold (using the same printing parameters as above) were printed to prepare a 10 mm-diameter scaffold disk. After 2 weeks of expansion and 3 weeks of ostegenic differentiation, an artificial bone tissue was prepared. A burr hole (approximately 5 mm diameter) was drilled on to the rat’s skull to create a bone defect. The prepared bone structure, which was flexible and pliable, was subsequently transplanted to seal the burr hole (Fig. 1.10f). Further study is being conducted to study the mechanical and biological properties of transplanted bone structure for the potential use in bone transplantation.

Fig. 1.10 (a) Printed ELA-cell in collagen scaffold after expansion (2 weeks) and differentiation (3 weeks), (b) control plate without differentiation, (c) BCIP/NBT stain of experimental plate and (d) control plate, (e) zoomed figure of BCIP/NBT stain of experimental plate with printed morphology marked in dotted line, and (f) the site of transplantation onto the rat skull bone

18


1.7 Conclusions The on-demand construction of a lab-on-a-chip style miniature organ, housed in a multi-chamber bioreactor (‘organ-on-a-chip’) as created by bioprinting, will enable the rapid screening and testing of responses from organs-of-interest against external chemical stressors/insults. Since various environmental factors of different chemical origins can be rapidly tested, high-throughput screening of toxicological profiles or drug efficiency. We also believe that the developed techniques and hardware will have an extended impact on stem cell research and regenerative medicine. The proliferation and differentiation of stem cells is known to be dependent on the surrounding microcellular environment, as well as on cell-to-cell interactions. One example of such dependencies can be found from the effect of feeder cells for lineage induction of embryonic stem cell lines [24]. On-demand 3D bioprinting techniques can strategically place stem cells in a user-definable cellular environment with the aim of targeted differentiation of desired cell lines for potential application in regenerative medicine. The 3D bio-printer can also pattern different types of cells in sophisticated spatial geometries with gradients of cell density. This ability will provide a unique opportunity for studying physiological mechanisms such as cell signaling properties in vitro. An example of such a study includes the placement of neural cells on to an array of electrodes to study neural network processing. In summary, the bioprinting technology for creating artificial tissue structures in 3D will promote a wide spectrum of immediate and future applications, serving a broad range of research and industrial communities.

References 1. Griffith LG, Naughton G (2002) Tissue engineering–current challenges and expanding opportunities. Science 295(5557):1009–1014 2. Langer R, Vacanti JP (1993) Tissue engineering. Science 260(5110):920–926 3. Orive G, Anitua E, Pedraz JL et al (2009) Biomaterials for promoting brain protection, repair and regeneration. Nat Rev Neurosci 10(9):682–692 4. Martin I, Smith T, Wendt D (2009) Bioreactor-based roadmap for the translation of tissue engineering strategies into clinical products. Trends Biotechnol 27(9):495–502 5. Hutmacher DW, Sittinger M, Risbud MV (2004) Scaffold-based tissue engineering: rationale for computer-aided design and solid free-form fabrication systems. Trends Biotechnol 22(7):354–362 6. Yeh J, Ling Y, Karp JM et al (2006) Micromolding of shape-controlled, harvestable cell-laden hydrogels. Biomaterials 27(31):5391–5398 7. Hwang NS, Varghese S, Elisseeff J (2007) Cartilage tissue engineering: directed differentiation of embryonic stem cells in three-dimensional hydrogel culture. Methods Mol Biol 407:351–373 8. Underhill GH, Chen AA, Albrecht DR et al (2007) Assessment of hepatocellular function within PEG hydrogels. Biomaterials 28(2):256–270 9. Mata A, Kim EJ, Boehm CA et al (2009) A three-dimensional scaffold with precise microarchitecture and surface micro-textures. Biomaterials 30(27):4610–4617 10. Whitesides GM, Ostuni E, Takayama S et al (2001) Soft lithography in biology and biochemistry. Annu Rev Biomed Eng 3:335–373

1


19

11. Boland T, Xu T, Damon B et al (2006) Application of inkjet printing to tissue engineering. Biotechnol J 1(9):910–917 12. Mironov V (2003) Printing technology to produce living tissue. Expert Opin Biol Ther 3(5):701–704 13. Sachlos E, Czernuszka JT (2003) Making tissue engineering scaffolds work. Review on the application of solid freeform fabrication. Eur Cell Mater 5:29–40 14. Yeong WY, Chua CK, Leong KF et al (2004) Rapid prototyping in tissue engineering: challenges and potential. Trends Biotechnol 22(12):643–652 15. Barron JA, Wu P, Ladouceur HD et al (2004) Biological laser printing: a novel technique for creating heterogeneous 3-dimensional cell patterns. Biomed Microdevices 6(2):139–147 16. Ringeisen BR, Kim H, Barron JA et al (2004) Laser printing of pluripotent embryonal carcinoma cells. Tissue Eng 10(3–4):483–491 17. Hwang MJ, Suh JM, Bae YH et al (2005) Caprolactonic poloxamer analog: PEG-PCL-PEG. Biomacromolecules 6(2):885–890 18. Lee W, Debasitis JC, Lee VK et al (2009a) Multi-layered culture of human skin fibroblasts and keratinocytes through three-dimensional freeform fabrication. Biomaterials 30(8):1587–1595 19. Lee W, Pinckney J, Lee V et al (2009b) Three-dimensional bioprinting of rat embryonic neural cells. Neuroreport 20(8):798–803 20. Golden AP, Tien J (2007) Fabrication of microfluidic hydrogels using molded gelatin as a sacrificial element. Lab Chip 7(6):720–725 21. Schmidt NO, Przylecki W, Yang W et al (2005) Brain tumor tropism of transplanted human neural stem cells is induced by vascular endothelial growth factor. Neoplasia 7(6):623–629 22. Willerth SM, Rader A, Sakiyama-Elbert SE et al (2008) The effect of controlled growth factor delivery on embryonic stem cell differentiation inside fibrin scaffolds. Stem Cell Res 1(3):205–218 23. Fan J, Varshney RR, Ren L et al (2009) Synovium-derived mesenchymal stem cells: a new cell source for musculoskeletal regeneration. Tissue Eng Part B Rev 15(1):75–86 24. Dravid G, Hammond H, Cheng L (2006) Culture of human embryonic stem cells on human and mouse feeder cells. Methods Mol Biol 331:91–104

Part II

Ink Jet Approaches

Chapter 2

Reconstruction of Biological Three-Dimensional Tissues: Bioprinting and Biofabrication Using Inkjet Technology Makoto Nakamura

Abstract We have developed three-dimensional (3D) fabrication technology using direct cell printing to overcome several intrinsic problems of tissue engineering and regenerative medicine. Initial experiments using inkjet technology showed good feasibility of direct printing of living cells, however, problems such as drying and ink bleeding persisted. Using gel precursor and gel reactant such as sodium alginate solution and calcium chloride solution, we succeeded in developing a gelation technique capable of printing living cells by inkjet safely and effectively. This technique prevents cells from drying and maintaining printed position even in wet conditions. In addition, semi-solid hydrogel enables us to fabricate 3D structures using the individual inkjet droplet as the base unit of fabrication. Our ‘3D Bioprinter’ is capable of fabricating 3D hydrogel structures such as fibers, sheets, lattices and tube structures with only hydrogel and with hydrogel and living cells together. Using multi-nozzles, 3D hydrogel structures with different types of living cells were successfully fabricated. In this chapter, we present the development of 3D biofabrication by direct cell printing using an inkjet and hydrogel. We will also discuss the present state of 3D tissue engineering and potential issues for future developments using our approach in the context of development of artificial organs.

2.1 Introduction ‘Bioprinting’ and ‘Biofabrication’ represent the procedures of printing and fabrication using biological or bio-functional materials. The materials used in Bioprinting and Biofabrication are called Bio-inks, which include living cells, extracellular matrices, proteins and growth factors and other biological and bio-functional materials [1–5]. The results of this research have direct application in bio-medical, life science, and basic and clinical medicine. M. Nakamura (B) Graduate School of Science and Engineering for Research, University of Toyama, Toyama, Japan e-mail: [email protected]

B.R. Ringeisen et al. (eds.), Cell and Organ Printing, C Springer Science+Business Media B.V. 2010 DOI 10.1007/978-90-481-9145-1_2,

23

24

M. Nakamura

Inkjet printing technology has been developed to print high resolution printing with photographic quality, providing the capacity to eject very small droplets of several colored inks on demand. Recently, this technology has also been used to print various biomaterials, such as nucleic acid, proteins [6], growth factors [7, 8], and biological cells [1–3, 5, 9–15]. Applications of inkjet bioprinting include biochips, diagnosis for the tailor made medicine based on genetic information and the development of scientific research tools. Cell printing is also useful for cell chips. Moreover, it has the great potential to solve and break through some of the intrinsic problems in present tissue engineering.

2.2 Tissue Engineering and the Issues in Tissue Engineering Tissue engineering is one of the important technologies of the twenty-first century. Its goal is to produce functional cell, tissue, and organ to repair, replace or enhance biological function that has been lost by disease and injury. It is also one of the most promising approaches to solve the problems of the shortage of suitable organs for transplantation. Since Vacanti and Langer demonstrated the concept of tissue engineering, a scaffold based approach has been the standard method to produce engineered three-dimensional (3D) tissues and organs [16–18]. Scaffolds are the temporal architectural 3D structures for cell adhesion and cell growth and are fabricated by engineering procedures according to the morphological designs. They are usually made of biodegradable materials such as poly-glycolic acid and poly-lactic acid. Cells are seeded onto the scaffolds and cultured in vitro in advance before implantation. After cell adhesion, the cell-adhered scaffolds are implanted into the recipients. After implantation, biodegradable materials are degraded and finally only the implanted cells remain and form functional tissues in vivo. This is the strategy and the scenario of the scaffold based approach. However, only a few simple tissues with simple structures and simple components have been successfully fabricated, despite the great hopes and the initial favorable results. Based on the histological considerations, the tissues of complex organs have four important characteristics. These characteristics are shown in Fig. 2.1; (1) they are 3D structures, (2) they have the characteristic micro-structures required to fulfill the particular function of the organ, (3) they are composed of multiple types of cells and extra-cellular matrices and (4) they have a complex vascular network to sustain the cells in the organs. No tissue capable of complex physiological function has been artificially fabricated. Scaffold based approaches to tissue engineering have several problems and limitations. Chief among these is the inability to adequately control [11, 19]: (i) cell distribution in the 3D structures, especially deep below the surface of the scaffold; (ii) the positioning of multiple cell types; (iii) composition of the scaffold at specific locations; (iv) local concentration of growth factors; (v) induction of blood

2

Reconstruction of Biological Three-Dimensional Tissues

25

Fig. 2.1 Characteristics of the tissues of physiological functional organs

capillaries; (vi) enhancement of the target organ cells at specific locations, and (vii) biodegradation of the scaffold material. In the scaffold based approach, cells are seeded onto acellular scaffolds. Therefore, the cell distribution and cell composition inside of the scaffold cannot be controlled at all. And the morphogenesis of essential tissue architecture especially microstructures is completely dependent upon the cells alone. Therefore, it is very difficult to fabricate physiologically functional tissues which have special micro-3D structures using the scaffold based approach. To realize the goal of tissue engineering, a new approach must be developed which does not have the aforementioned problems.

2.3 Rationale for Direct Cell Printing in Tissue Engineering Tissue engineering needs effective technology to arrange individual cells directly onto the targeted 3D spatial position with a spatial resolution required by micro structures. To achieve this, we need to develop some micro cell handling technology and the 3D fabrication technology using a layer by layer procedure. Using Computer Aided Design (CAD) and Computer Aided Manufacturing (CAM) technologies, we developed ‘tissue CAD/CAM methods’. This concept is shown in Fig. 2.2. A similar technology by applying CAD, CAM, and Computer Aided Engineering (CAE) technology to tissue engineering was demonstrated and developed by Sun et al. called Computer Aided Tissue Engineering (CATE) [20]. Inkjet printing technology has been developed to eject microscopic droplets of several colored inks on demand. High resolution printing with photographic quality is achieved using inkjet technology. We believe this technology is capable of handling individual cells and print them with a resolution suitable for biological tissue.

26

M. Nakamura

Fig. 2.2 Next technology to scaffold based approach

The sizes of the printed dots by inkjet printers are almost the same size as a cell. Initial experiments of cell printing have shown good feasibility in direct printing of living cells. Inkjet technology is useful for direct seeding of individual cells and has many advantages for tissue engineering [9, 11]. The advantages in inkjet for tissue engineering are demonstrated in Table 2.1. However, several problems were identified in direct cell printing with inkjet, such as drying, ink bleeding in wet conditions, and how to make 3D structures. Inkjet droplets are so small that they dry immediately. Immediate drying is useful for picture printing, but is harmful for living cells. If cells are printed onto wet substrates to prevent drying, the printed cells spread out and lose print resolution. This is called bleeding. 3D fabrication is impossible by merely printing cell suspensions which have no mechanical strength. The gelation technique of inkjet printing was developed to solve these problems. Two different types of gel solution, gel precursor and gel reactant, were used. It is well known that an aqueous sodium alginate solution forms a hydrogel when it comes into contact with Ca2+ ions. Alginate hydrogel is one of the biocompatible hydrogels [11–16]. When 0.8–1% of sodium alginate solution was ejected by an inkjet head into 2% calcium chloride solution, we found the ejected individual droplets form micro gel beads. We also found micro gel beads fused and form fibers when a line was printed. Alginate hydrogels provide both structural strength for 3D structures and an aqueous environment for cells. They are thus ideal for 3D fabrication while maintaining cell viability. Using the gelation technique, constructing 3D structures with accurate positioning of living cells within the structure becomes possible by inkjet.

2


27

Table 2.1 Characteristics of inkjet and advantages for tissue fabrication Characteristics of Inkjet technology

Advantages for tissue fabrication

(1) High resolution printing • Fine control of ejecting poiunts • Pico-liter sized ink droplets

High resolution fabrication • Fabrication of fine products with high resolution • Pico-liter sized ink droplets Fabrication of composite products with different cells, materials, and growth factors Rapid fabrication, Fabrication of large-sized products Easy to apply to computer aided biofabrication Printable onto gels, aqueous solution, cells, or directly onto the targets wounds during surgical operation Usability of reactive gel material and reactive two materials. Preventive effects for friction or contact damages Printing biological materials cells proteins, DNAs, biopolymers, humoral factors, drugs, nanomaterials

(2) Color printing with different multi-colored inks (3) High speed printing with kHz ejection per one nozzle (4) Established connection to computer (5) Printability to several subjects, papers, solid mass, disk, dishes, gels and aqueous solution (6) Non-contact printing

(7) Printability of several inks; aqueous inks, pigment inks, suspension of several materials, reactive solution

In author’s technology of inkjet 3D biofabrication based on cell printing (8) Ejectability of living cells without Direct cell printing, direct cell handling and direct significant damages cell positioning (9) Gelation of inkjet droplets Non-blotting printing in the aqueous subjects 3D fabrication into aqueous solution Preventation from drying, Fabrication of fragile products 3D digital fabrication 3D digital biofabrication

2.4 Development of a 3D Bioprinter and 2D and 3D Biofabrication by Inkjet Cell Printing We developed an inkjet printer, ‘3D Bioprinter’, using the gelation technique. This work was supported by Kanagawa Academy of Science and Technology, Japan. Figure 2.3a shows the first 3D Bioprinter prototype [11–16]. We designed and developed this printer to be the most suitable one for 3D fabrication using living cells. We used two types of inkjet heads, piezo and static electricity actuated [21], both of which can eject living cells safely without significant heat. Inkjet heads can be moved along the X, Y and Z axis and are controlled by a computer. Ejecting frequency and the head motion speed can be adjusted depending on the structure to be fabricated. Using this 3D Bioprinter, we manufactured several 2D and 3D structures, such as fibers, sheets, and the lamination of 2D sheets and 3D lattices, using layer by layer printing with hydrogel only and with hydrogel and living cells together [12–16]. Figure 2.3b shows lines of printed cells. Cells could be printed without spreading

28

M. Nakamura

Fig. 2.3 3D bioprinter (a) and 2D printing and 3D fabrication by direct cell printing (b–d). HeLa cells stained by red fluorescent dye were printed with linear pattern onto calcium chloride solution (b). 3D tube structure was fabricated with alginate hydrogel, which contains living HeLa cells stained red fluorescent dye (c, d) [11]

even in the aqueous substrates. This was because the printed cells were supported by hydrogel fibers. Each fiber is about 40 μm thick. 3D tube structures can also be fabricated (Figs. 2.3c, d, and 2.4). To create a tubular structure, the inkjet head is moved along a circle while dispensing ink droplets continuously. 3D tube structures with different diameters were successfully fabricated in a CaCl2 solution. The fragile hydrogel structures maintain their shapes well when fabricated in the aqueous CaCl2 solution. The aqueous environment also prevents cell drying and improves cell viability. To manufacture heterogeneous 3D biological tissues with several different types of cells the 3D bioprinter is modified to use multi-nozzle inkjets. Using the new 3D bioprinter, several 3D hydrogel structures were successfully fabricated with different ink materials. Shown in Fig. 2.5 is 3D lamination of sheets and tubes with different colored hydrogels. Double walled tube structures could be fabricated with different colored hydrogels by alternately changing ink nozzles and circle diameters, respectively (Fig. 2.6b). And finally, we succeeded in fabricating double walled tubes with vascular endothelial cells inside and smooth muscle cells outside by cell printing, which mimicks the histology of blood vessels (Fig. 2.6a, c–e) [22]. In this

2


29

Fig. 2.4 Fabrication of 3D tube structures by 3D bioprinter. Cell suspension with sodium alginate was ejected into Calcium Chloride solution with inkjet head moving at the circular motion. Fabrication of 3D tube with colored hydrogel (b). Fabrication of cell containing tube by cell printing (c) and produced 3D tube containing living cells (d)

case, we used the endothelial cells stained green with cell tracker green as one ink, and the smooth muscle cells stained red with cell tracker red as another ink. 3D gel/cell structures can be successfully fabricated by laminating layers of printed sheets using inkjet. We have fabricated thick 3D laminated gel sheets more than 7 mm in thickness, and long gel tubes longer than 18 cm using our 3D Bioprinter. Based on these results, we have shown that it is feasible for large scale and complicated 3D structures to be fabricated by inkjet 3D cell printing procedures, using multiple cell types.

2.5 Perspectives on 3D Biofabrication by Inkjet Cell Printing We have been successful in fabricating several 3D structures with different types of cells and hydrogels using our 3D Bioprinter. This technique using inkjet cell printing eliminates some of the serious problems of tissue engineering. This is a new approach in which 3D complex tissues can be built up by direct positioning of individual cells. We demonstrated the feasibility of inkjet cell printing for 3D

30

M. Nakamura

Fig. 2.5 3D Fabrication by different hydrogels. 3D lamination of different colored 2D sheets (a, b), and 3D fabrication of a striped tube by different colored inks

biofabrication by gelation of inkjet droplets. This is a first step towards engineering a complex tissue, and there are still many problems and issues to be overcome (Fig. 2.7). 3D fabrication of complex tissues using soft, fragile and wet materials like hydrogel requires more study. Fabrication using gelation of inkjet droplets is a unique approach, where the process, ink materials, and jetting parameters need to be optimized. The fabricated products using the 3D Bioprinter are only hydrogel structures mixed with cells. Even though the cells are positioned in the correct location, no tissues have been formed. We must learn how to coerce the cells to form tissues so eventually they will be able to perform their intended function in the body. This process is called Bio-processing [23]. It is very important to develop the technologies to control not only the spatial positioning of the cells but also maintain their viability, promote their interaction to develop into tissues, and eventually initiate the functions of the tissue. It is important to develop how to use extracellular matrices and growth factors and develop effective ones for bio-processing. Alginate hydrogel is a good material for 3D fabrication, but not good for cell growth. Thus, more suitable materials for cell growth and tissue development should be developed. Blood vessel network formation is a big issue, too, and we have started to design and fabricate 3D structures with sufficient perfusion systems [24].

2


31

Fig. 2.6 Fabrication of double walled 3D tube structures. Schema of the histology of an artery and the fabrication of double walled tubes (a). Double walled tube fabricated with two different colored gels (b), and double walled tube arranged with vascular endothelial cells and smooth muscle cells (c–e) [22]

Fig. 2.7 Perspective concept of 3D biofabrication towards tissue and organ engineering

To study the 3D structures we created, we need to observe and evaluate their morphology and cellular viability, as well as physiological function development before the fabricated products can be clinically used. But there are only a few instruments to observe such large 3D tissues and to evaluate their physiological function non-invasively. We need to develop such supporting technologies for 3D tissue engineering, too. Tissue engineering is a multi-disciplinary field and technologies from different fields are also needed.

32

M. Nakamura

2.6 Conclusion In conclusion, the age of the research and development of 3D biofabrication, or 3D tissue engineering has begun. Direct cell printing and the layer by layer fabrication approach are very promising. There are numerous hurdles to be overcome at present, but it is because we are just at the initial steps. The author believes there is the potential for humans and science to overcome them and achieve successfully the fabrication of physiological tissues in the future. Acknowledgments The author would like to thank all of the students of the Nakamura Laboratory in the University of Toyama, all of the members of the Bioprinting project of Kanagawa Academy of Science and Technology in Japan, and also all the parties involved in this project. The author would like to thank several developers of Japanese inkjet machines for their kind support and suggestions in regard to the ink jet technology. This work was supported by the Kanagawa Academy of Science and Technology, and Grantsin-aid for Scientific Research (#17300146, #18880042, and #18760518) from the Japan Society for the Promotion of Science.

References 1. Wilson WC, Boland T (2003) Cell and organ printing 1: protein and cell printers. Anat Rec 272A:491–496 2. Boland T, Mironov V, Gutowska A, Roth EA, Markwald RR (2003) Cell and organ printing 2: fusion of cell aggregates in three-dimensional gels. Anat Rec 272A:497–502 3. Mironov V, Boland T, Trusk T, Forgacs G, Markwald RR (2003) Organ printing: computeraided jet-based 3D tissue engineering. Trends Biotechnol 21:157–161 4. Jakab K, Neagu A, Mironov V, Markwald RR, Forgacs G (2004) Engineering biological structures of prescribed shape using self-assembling multicellular systems. PNAS 101:2864–2869 5. Boland T, Xu T, Damon B, Cui X (2006) Application of inkjet printing to tissue engineering. Biotechnol J 1(9):910–917 6. Roda A, Guardigli M, Russo C, Pasini P, Baraldini M (2000) Protein microdeposition using a conventional ink-jet printer. BioTechniques 28:492–496 7. Watanabe K, Miyazaki T, Matsuda R (2003) Fabrication of growth factor array by color inkjet printer. Zool Sci 20:429–434 8. Campbell PG, Miller ED, Fisher GW, Walker LM, Weiss LE (2005) Engineered spatial patterns of FGF-2 immobilized on fibrin direct cell organization. Biomaterials 26:6762–6770 9. Nakamura M, Kobayashi A, Takagi F, Watanabe A, Hiruma Y, Ohuchi K, Iwasaki Y, Horie M, Morita I, Takatani S (2005) Biocompatible inkjet printing technique for designed seeding of individual living cells. Tissue Eng 11:1658–1666 10. Xu T, Jin J, Gregory C, Hickman JJ, Boland T (2005) Inkjet printing of viable mammalian cells. Biomaterials 26:93–99 11. Nakamura M (2006) Bioprinting-challenge in building biological tissues and organs. J Clin Exp Med (Igaku no Ayumi) 218:139–144 (in Japanese) 12. Nakamura M, Nishiyama Y, Henmi C, Yamaguchi K, Mochizuki S, Takiura K, Nakagawa H (2006) Inkjet bioprinting as an effective tool for tissue fabrication. Proceeding of digital fabrication 2006, international conference on digital fabrication, Denver, pp 89–92 13. Nishiyama Y, Nakamura M, Henmi C, Yamaguchi K, Mochizuki S, Nakagawa H, Takiura K (2007) Fabrication of 3D cell supporting structures with multi-materials using the bio-printer. Proceeding of 2007 international manufacturing science and engineering conference, MSEC 2007 conference papers, ASME, Atlanta, MSEC2007-31064

2


33

14. Nakamura M, Nishiyama Y, Henmi C, Iwanaga S, Nakagawa H, Yamaguchi K, Akita K, Mochizuki S, Takiura K (2008) Ink jet three-dimensional digital fabrication for biological tissue manufacturing: analysis of alginate microgel beads produced by ink jet droplets for three dimensional tissue fabrication. J Imaging Sci Tech 52(6):060201–060201-6 15. Nishiyama Y, Nakamura M, Henmi C, Yamaguchi K, Mochizuki S, Nakagawa H, Takiura K (2009) Development of three-dimensional bio-printer: construction of cell supporting structures using hydrogel and state-of-the-art inkjet technology. J Biomech Eng 131(3):035001-1 16. Langer R, Vacanti JP (1993) Tissue engineering. Science 260:920 17. Kaihara S, Vacanti JP (1999) Tissue engineering: toward new solutions for transplantation and reconstructive surgery. Arch Surg 134:1184 18. Hutmacher DW, Sittinger M, Risbud MV (2004) Scaffold-based tissue engineering: rationale for computer-aided design and solid free-form fabrication systems. Trends Biotechnol 7: 354–362 19. Background on Tissue Engineering Bone tissue engineering initiative, Carnegie-Mellon University, Pittsburgh. http://www.cs.cmu.edu/People/tissue/tutorial.html 20. Sun W, Lal P (2004) Recent development on computer aided tissue engineering: overview, scope and challenges. Biotechnol Appl Biochem 39:29–47 21. Kamisuki S, Hagata T, Tezuka C, Fujii M, Atobe M (1998) A low power small, electrostatically-driven commercial inkjet head. Proceedings IEEE the 11th annual international workshop on micro electro-mechanical systems (MEMS’98), pp 63–68. 22. Calvert P (2007) Printing cells. Science 318:208–209 23. Matsuda T (2003) At the conference in Tokyo Medical & Dental University, Tokyo 24. Takiura K, Doi A, Yamaguchi K, Akita K, Nishiyama Y, Henmi C, Nakamura M (2008) Alginate gel honeycomb structures fabricated with the bioprinter, digital fabrication, 9 Sept 2008, Technical programe and Proceedings, pp 501–503

Chapter 3

Piezoelectric Inkjet Printing of Cells and Biomaterials Rachel Saunders, Julie Gough, and Brian Derby

Abstract Tissue engineering is a rapidly expanding field which aims to repair damaged tissue using a more regenerative approach. Technology has advanced to provide complex scaffolds with controlled architecture and porosity however problems with incorporating cells into the scaffold structure still persist. Standard cell seeding techniques can result in a poor cell seeding density, pore occlusion and is limited with regards to cell penetration, scaffold size and cell placement. Drop-ondemand inkjet printing is a fabrication technique which is capable of depositing materials layer-by-layer to form complex constructs. This technique has the potential to be used as a tool for the deposition of living cells. If this could be achieved then the simultaneous deposition of multiple cell types and scaffold matrix could yield a reality whereby human tissue could be fabricated with a precision not only applicable to scaffold architecture but also to the placement of multiple cell types. This work presents an insight into the effect of printing parameters on cell viability, deposition characteristics and explores methods of immobilization with an aim to achieve three-dimensions.

3.1 Introduction Biological printing involves the deposition of both mammalian cells and proteins, with current research aiming to characterise and determine the effects of the printing parameters on viability and establish the feasibility of the technique. This chapter expands on our previous viability study and look at the effect of printing on several mammalian cell types. The effect of piezoelectric drop-on-demand inkjet printing parameters on the viability of bovine chondrocytes and human osteoblasts is presented.

R. Saunders (B) School of Materials, Materials Science Centre, University of Manchester, Manchester, UK e-mail: [email protected]


35

36

R. Saunders et al.

The feasibility of utilising drop-on-demand (DOD) inkjet printing as a tool for patterning and depositing a variety of functional and biological materials has already been established [1, 2]. Research within this field is scattered not only with regards to the cell types used but also the methods of printer actuation [3–7]. The ability to pattern cells has considerable advantages in the optimisation of scaffold seeding. Currently cells are seeded into a scaffold, either through dynamic or static techniques, as a secondary processing step after scaffold fabrication. These techniques often result in occlusion of the scaffold pores, limited penetration depth and a nonuniform distribution of cells [8]. The ability to pattern cells and scaffold matrix material at the same time would be a unique approach allowing for greater control over cell placement and concentrations of cells whilst allowing multiple cell types to be used and located separately. There are two main modes of actuation that are used in DOD printers: thermal, and piezoelectric. The authors have previously reported a cell survival rate of > 90% for HT1080 fibrosarcoma cells printed using a piezoelectric inkjet [5]. A schematic of the piezoelectric DOD printer used in this and previous studies in shown in Fig. 3.1. The piezoelectric DOD printer consists of a cylindrical piezoelectric actuator surrounding a glass capillary nozzle with a diameter of 60 μm. Actuation of the piezoelectric transducer is via an applied voltage pulse in the form of a repeated electrical waveform which causes expansion and contraction resulting in the emission of a drop at a drop generation frequency around 1 kHz. The mechanics of droplet ejection have been discussed in detail in an earlier publication [9]. Briefly, drops are ejected by the generation of a pressure pulse in a fluid filled chamber behind the printing orifice and in a piezoelectric printer this is chiefly controlled by the magnitude of the applied voltage, the actuating pulse duration and frequency. In an earlier report we showed that by changing the actuating voltage it is possible to exercise some control over the velocity of an ejected drop, with the drop velocity, and hence the stresses experienced by cells in suspension either during drop formation or drop impact, increasing with increasing actuation voltage [5]. This study found that cell survival post-printing decreased slightly with increasing actuation voltage but at low printing voltages the cell survival rate was indistinguishable from that of a non-printed control specimen. The rise time of the voltage pulse

Fig. 3.1 Schematic of a piezoelectric drop on demand mode

3

Piezoelectric Inkjet Printing of Cells and Biomaterials

37

(which might be expected to influence the stress rate experienced by a cell in suspension) had no significant effect on cell survival statistics. In addition a study of cell activity post-printing found no evidence for any influence of printing conditions on cell activity. There has been significant published work on the deposition of cell suspensions using thermal inkjet printers. Thermal printers also function through the propagation of pressure pulses in a liquid but in this case the pulse is generated through the vaporisation of a small volume of liquid in a chamber behind the printing orifice. The expansion of the consequent bubble during local heating, and its rapid collapse when the thermal pulse is removed, generates the pressure variation required for printing. The temperatures involved in thermal actuation can locally exceed 300◦ C [1], although only a very small fraction of the liquid is vaporized in a thermal printer. Xu et al. employed a modified thermal printer for the deposition of Chinese hamster ovary cells reporting a > 90% cell viability after printing [7]. This is comparable to the > 90% survival rate reported previously by the authors with HT0180 fibrosarcoma cells using piezoelectric printing; however, Xu also reports that the average number of cells lysed during the formulation of the ‘bioink’ averaged 15% This reduces the cell viability for the overall procedure to approximately 75% of the original sample, which is a substantially lower survival rate than found in our previous study [5]. However, given the low survival of the unprinted control in Xu’s work, the most likely difference between these two studies is not the mode of operation of the printer but the preparation of the cell suspension and the fluid medium used. Xu used phosphate buffered saline (PBS) as the media [7], whereas our earlier work used a more complex cell culture media including foetal calf serum [5]. The brief analysis of the differences between our earlier work [5] and that of Xu et al. [7] illustrates the difficulties in comparing the response of different cell types to deposition and patterning using inkjet printing when the work is carried out in different laboratories and using different experimental procedures, protocols and equipment.

3.2 Biological Printing The characterisation of cells post-printing poses several problems as methods have to take into consideration the variance in cell numbers that can occur as part of the printing process. The following section discusses cell distribution and viability of mammalian cells printed using a piezoelectric DOD printer. Printing experiments were carried out using a single-jet stationary piezoelectric printhead, Microjet MJ-AB-01 (Microfab Inc., Plano, TX, USA) as previously described in our earlier work [5]. Bovine chondrocytes and human osteoblast cell suspensions were directly printed from a 60 μm diameter jet at a frequency of 10 kHz directly onto a well plate surface (Costar). The piezoelectric excitation pulse voltage used was in the range of 40–80 V. In order to reduce possible variability the actuating pulse width was adjusted to ensure resonance as previously detailed [9].

38

R. Saunders et al.

3.2.1 Cell Patterning and Distribution It was only possible to produce stable inkjet printing conditions with an actuation voltage of 40 V or greater. Circular patterns of chondrocyte suspensions printed at this voltage on a glass microscope slide are displayed in Fig. 3.2. Under these conditions, the spherical drops generated at the printer have a diameter of 80 μm and these spread to a diameter on the substrate of 220 μm, which is equivalent to a sessile drop of contact angle approximately 14◦ . In Fig. 3.1a the diameter of each spread drop is smaller than the interdrop spacing of 314 μm and thus a circular array of isolated drops results. These individual drops are highly irregular and this is probably caused by a combination of two factors. First, at 40 V actuation, we are at the lower limit of printability and thus there is less consistency between individual drops. Second, it is well known that in inkjet printing the initial drop shape is not spherical but instead forms as a teardrop with a long extended tail [10]. Normally, surface tension will pull the tail into the head of the drop to form a sphere before impact. However, if the tail is long it may detach to form a smaller satellite drop or if the printhead is close to the substrate, a non spherical drop may impact the surface, which coupled with a translational velocity can also lead to satellite drop impacts distant from the main drop. We hypothesise that one of these behaviour leads to the irregular, non-circular spread drop foot prints and the presence of smaller satellite drops in Fig. 3.2. Drops printed at higher voltages (Fig. 3.3) do not show these irregular features to the same extent. In Fig. 3.2b, c it is clearly seen that when the interdrop spacing is smaller than the spread drop diameter, drop coalescence leads to a circular feature. The stability of circular and linear features that are formed by spread drop coalescence is an important feature that allows complex structures to be fabricated by inkjet printing. It might be expected that surface energy minimisation will lead to coalesced drops forming a larger circular footprint sessile drop but this neglects the phenomenon of contact line pinning where even a limited amount of solvent evaporation can lead to precipitation of solids at the liquid/substrate/air triple line, effectively pinning it and limiting the geometry of coalescence [11, 12]. It is possible to predict the shape of

Fig. 3.2 Circular patterns with a diameter of 1 mm generated from printed chondrocyte suspensions. Individual isolated drops spread to a contact diameter of approximately 220 μm. All images printed at 10 kHz with 40 V actuation pulse. Angular drop spacing along the circumference is: (a) π/5 or 310 μm, (b) π/10 or 160 μm, (c) π/15 or 100 μm

3


39

Fig. 3.3 Isolated drops of chondrocyte suspension printed for cell distribution statistical analysis. Lower image is of a dried drop showing precipitation from the media around two cells

the coalesced droplets assuming contact line pinning and static equilibrium of the surface forces [13]. Figure 3.3 also shows that the distribution of cells per printed drops is not homogenous which is consistent with concerns expressed in previous work where cell suspensions were prone to settling over periods of time leading to a systematic variation in the number of cells per drop [5]. In order to characterise the distribution of cells in drops, patterns of individual, well separated drops of the cell suspension were printed on a transparent substrate (Fig. 3.3) and the number of cells within each drop counted using optical microscopy. Figures 3.4 and 3.5 shows a frequency analysis of the number of cells per drop for two experiments with chondrocyte (95 printed drops), and osteoblast (400 printed drops) suspensions. Superimposed upon the experimental data in each case is a representation of the predicted cell distribution frequency, if it were to follow a Poisson distribution. A Poisson distribution describes the cell count frequency if the cells suspended in the printhead are randomly distributed through the volume and if the printed drop selects a random number of cells from the suspension. In order to assess whether the data represented in Figs. 3.4 and 3.5 describes a random selection of the number of cells per drop we apply an appropriate significance test. A characteristic of the Poisson distribution is that the distribution has an identical value of mean and variance. The mean and variance of the frequency distribution data for the chondrocyte and osteoblast suspensions is presented in Table 3.1.

40

R. Saunders et al.

Fig. 3.4 Frequency analysis of the distribution of cells per drop in a series of isolated printed cell suspensions: data from 95 printed drops of a chondrocyte suspension

Fig. 3.5 Frequency analysis of the distribution of cells per drop in a series of isolated printed cell suspensions: data from 400 printed drops of an osteoblast suspension

This ratio is compared with the χ 2 distribution for the degrees of freedom of the experiment (in this case with 6 data intervals we have 5◦ of freedom) to obtain the appropriate probability (P-value) for the null hypothesis, that the distribution is indeed random. For the chondrocyte suspension the appropriate probability is 0.946 and for the osteoblasts 0.855. In both cases these are significantly greater than the normally accepted threshold of 0.05 required to reject the null hypothesis and thus

3


41

Table 3.1 Summary of data for the cell distribution experiments pertaining to both the printed chondrocytes and osteoblasts

Chondrocytes Osteoblasts

Mean

Variance

Variance/mean

P(χ 2 )

1.12 0.77

1.30 1.50

1.19 1.96

0.946 0.855

we conclude that there is no evidence that the cell distribution follows any regular pattern. It is important to realise that this statistical analysis gives no information as to whether the cell distribution can be described by any particular statistical function. A Poisson distribution describes the probability of occurrence of events with a relatively low chance and can be used to describe very low cell concentrations in a drop when they occur randomly. This is a reasonable assumption in our case because the volume of a cell is so small as to be negligible compared to the volume of a drop. Although a distinct statistical certainty that the cell distribution per drop follows this distribution is not returned by the data we cannot reject the null hypothesis, that the cell distribution is random. The analysis of the distribution of cells per drop is significant because it indicates that there is as yet no control over the number of cells per printed drop. This has wide reaching implications with regards to characterisation of printed cell viability

3.2.2 Cell Viability The analysis and characterisation of cell viability post-processing is a complex process. The variation in cell numbers between samples necessitates data normalisation and the employment of several key steps to enable useful results. Statistical analysis is crucial with regards to isolating the effect of printing parameters. Determination of cell population through metabolic assays is commonplace. However any assay which is fatal to the sample cell population is unsuitable for the analysis of printed cells as the assumption is made that each sample contains comparable initial cell populations. This is an invalid assumption as the cells deposited per drop can be influenced by factors such as sedimentation, printing time, and drop mass (which in turn is linked to the pulse amplitude). It is therefore impossible to attribute ay cell loss over time purely to damage caused by the printing process. In order to circumvent this non-toxic metabolic assays must be utilised and the data normalised to each individual cell sample. To quantitatively establish the influence of printing parameters on the viability of deposited cells a live-dead assay was performed. Calcein AM is permeable to cell membranes and fluoresces green when it binds to esterase enzymes within a live cell. Live cells are impermeable to ethidium homodimer-1 which fluoresces red when bound to the DNA of cells with damaged membranes. The live-dead assay was carried out within 3 h of printing to enable accurate cell counting. A representative image of a stained sample of printed osteoblasts is presented in Fig. 3.6.

42

R. Saunders et al.

a

b

c

d

Fig. 3.6 Fluorescence micrograph of osteoblasts stained with calcein AM and ethidium homodimer-1 printed at a range of voltages with a constant signal rise time of 3 μs: (a) pipetted control, (b) printed at 40 V, (c) printed at 60 V, and (d) printed at 80 V

Histograms of the data obtained from experiments on osteoblast and chondrocyte suspensions over a range of printing parameters with an actuation signal amplitude of 40–80 V and rise times of 3–6 μs are displayed in Fig. 3.7. In all cases the cell survival rate is greater than 95% and to investigate this further requires appropriate statistical analysis. Single factor ANalysis Of VAriance (ANOVA) analysis on the transformed osteoblast data identified a significant difference within the data between the printed samples and the control (F(9.91) = 9.21 with P = 9.07 × 10–10 (F(Pcrit5% ) = 1.99)). Similar behaviour is seen with the transformed chondrocyte data, ANOVA analysis returning F(9.91) = 4.85 with P = 2.75 × 10–5 (F(Pcrit5% ) = 1.99). In order to further investigate the cause of the differences and separate the influence of actuation voltage and signal rise time, double factor ANOVA analysis was performed on both sets of data without the control sample and the results set out in Table 3.2. It can be clearly seen that the actuation signal voltage (amplitude) has a significant influence on cell survival statistics, while the signal rise time has no statistically significant effect. These data and the results of the analysis are broadly similar to that reported in our earlier study of the inkjet printing of fibrosarcoma derived cells [5], except

3


43

Fig. 3.7 Survival statistics using a live/dead assay after printing cell suspensions with a range of actuation voltages and signal rise times compared with a pipetted control: (a) osteoblasts, (b) chondrocytes. The error bars indicate the standard error Survival statistics using a live/dead assay after printing cell suspensions with a range of actuation voltages and signal rise times compared with a pipetted control: (a) osteoblasts, (b) chondrocytes. The error bars indicate the standard error of the mean from 10 replicates

44

R. Saunders et al.

Table 3.2 Double factor ANOVA results for printed cell data exploring the influence of printer actuation voltage and signal rise time on cell survival, showing calculated F statistic, critical value of F for 5% probability of null hypothesis, and computed probability, P, of the null hypothesis

Osteoblasts Waveform amplitude Rise time Interaction Chondrocytes Waveform amplitude Rise time Interaction

F(2.91) or F(4.91)

F(P 0.05 )

P

5.49 0.03 4.79

3.10 3.10 2.47

5.6 × 10–3 0.97 0.0015

4.84 0.42 7.21

3.11 3.11 2.48

0.010 0.66 4.3 × 10–5

that in the earlier study there was no significant interaction between rise time and amplitude. It is instructive to compare our data graphically for osteoblasts and chondrocytes with that from the previous study of fibroblast-like cells. In Fig. 3.8 we plot the data for cell survival for all three cell types as a function of actuator voltage with the 3 rise times; we also show a linear regression to each dataset. Figure 3.8 shows the data from the fibrosarcoma cells. There is a clear and consistent trend with the cell survival rate decreasing with increasing actuator voltage (and hence mechanical stress) and there is little difference observed for the three signal rise times. Indeed the regression slope for the three rise times appears to be effectively constant. The data for osteoblasts and chondrocytes in Fig. 3.8b, c show a much more varied response with no consistent trend seen with the actuation amplitude. This scatter in behaviour may explain why the ANOVA analysis appears to indicate an interaction between the rise time and the voltage, even though there is no statistical evidence for the rise time influencing cell survival (Table 3.2). It is more likely that the few printing parameters investigated (only three voltages) and the relatively large variance of the data has not allowed us to capture the full response of the cell survival rate to actuating voltage. However, we note that for the three cell types the cell survival statistics are of little difference from that of the unprinted control for all printing conditions investigated. Initial cell survival should not be the only factor when considering deposition techniques. Functionality and long-term viability is required for the fabrication of tissue engineered constructs. The Alamar Blue metabolic assay was established in previous work as the most appropriate means of measuring long term viability because it is non-toxic to the cell population, thus allowing for continuous monitoring of the proliferation of printed samples over time. At present it is impossible to guarantee a precise and reproducible cell concentration for each printed sample; therefore a normalisation procedure was introduced which normalised the data from each time point to the original 4 h measurement for each sample eliminating the more obvious influences of initial cell number on the observed trends. Figure 3.9 presents histograms showing the normalized cell proliferation data obtained from printed osteoblast and chondrocyte cell suspensions after time in

3


Fig. 3.8 Cell survival rates of the different cell types plotted as a function of printhead actuation voltage for the three signal rise times: (a) fibrosarcoma cells, (b) osteoblasts (c) chondrocytes

a

b

c

45

46

R. Saunders et al.

Fig. 3.9 Normalised Alamar Blue assay data for cell viability in culture as a function of time after printing for a range of inkjet printing parameters and a pipetted control, error bars show the standard error of the mean from 10 replicates: (a) osteoblasts, (b) chondrocytes

3


47

Table 3.3 Single factor ANOVA of cell proliferation data and the control sample showing calculated F-statistic data and computed probability of null hypothesis at 4 different time points; critical value of F(9, 90) = 1.99 for P = 0.05 and F(7.63) = 2.16 Osteoblasts

Chondrocytes

Hours after printing

F(9.90)

P

F(7.63)

P

24 48 72 96

3.32 7.45 5.91 20.55

1.5 × 10–03 4.6 × 10–08 1.9 × 10–06 2.4 × 10–18

1.39 7.13 2.37 5.59

0.20 9.8 × 10–8 0.02 4.2 × 10–6

culture up to 96 h. As discussed previously the data for each individual printing run has been normalised using the 4 h data point as a reference. This reduces the variation introduced by the number of cells dispensed in each printing sequence. The osteoblasts (Fig. 3.9a) show a consistent trend of cell proliferation and increasing cell number up to 96 h after printing, with cell numbers in the printed specimens either comparable to or exceeding that of the control sample. The data for the chondrocytes (Fig. 3.9b) shows qualitatively similar behaviour but the apparent difference between the printed samples and the controls are not so marked for each time step. ANOVA statistical analysis of the data is summarised in Table 3.3. The osteoblast cells shows a consistent significant difference within these groups of tests for all time points. The chondrocyte data shows a more complex behaviour with the 4 h time point showing no significant variation but those at longer time intervals showing variation below the P = 0.05 threshold. Double factor ANOVA analysis of the data from the individual time points for both cell types is presented in Table 3.4. Osteoblasts cells show a significant influence of actuation voltage on cell proliferation behaviour and no evidence for a consistent influence of rise time. Chondrocyte behaviour is not so clear, with no equivocal evidence for a significant effect of actuator voltage at all time points and no evidence for any effect of rise time. In this respect the chondrocyte data is most similar to our earlier observations on the behaviour of fibrosarcoma cells [4], which found no evidence for any significant influence of voltage or rise time on cell proliferation monitored using Alamar Blue. The mechanism for this difference in proliferation behaviour is not clear. The osteoblast data shows a consistent increase in cell number for all time points between printed samples and the control. This could possibly be explained if the control sample had a higher initial cell seeding concentration than the printed samples. This would cause a reduction in the rate of proliferation over time due to issues such as space and competition for nutrients. We carried out ANOVA on the osteoblast data without the control sample and again found significant statistical differences within the group with the 96 h time point revealing the greatest statistical significance with F (8.91) = 13.22 with P = 3.1 × 10–13 (F (Pcrit5% ) = 1.99). It would however be reasonable to assume that the physical influence of printing condition is minimal after 96 h as the cell populations would have renewed through

F(2.91)

6.61 12.92 10.66 40.40

24 48 72 96

F(2.91)

3.70 7.54 2.28 9.15

P

2.2×10–03 1.4×10–05 7.8×10–05 6.8×10–13 0.029 9.9× 10–04 0.11 2.6× 10–04

P 0.76 0.81 3.98 1.67

F(4.91)

Interaction

0.55 0.52 5.3×10–03 0.16

P 1.33 15.94 0.36 11.73

F(2.61)

0.27 2.5×10–06 0.7 4.7×10–05

P

Voltage

Rise time

Voltage

Hours after printing

Chondrocytes

Osteoblasts

2.45 0.45 1.47 0.42

F(2.61)

Rise time

0.095 0.64 0.24 0.66

P

1.13 8.19 4.73 6.52

F(4.61)

Interaction

0.35 2.2×10–05 2.1×10–03 1.9×10–04

P

Table 3.4 Two factor ANOVA results for the cell viability (proliferation) data obtained from osteoblast and chondrocyte suspensions, the influence of actuation voltage and rise time is explored

48 R. Saunders et al.

3


49

division decreasing the affected cell fraction. For example a cell which is mechanically damaged by the printing process would be expected to die as a result; if the cell is capable of division then the daughter cells would not inherit the full extent of the damage if any. It is possible, however, that the cellular response to the mechanical stress on printing influences intercell communication and this could result in a retained influence of printing conditions over time.

3.3 Matrix Printing The printing of biological materials cannot succeed alone and is therefore linked to the incorporation of printed matrix materials to achieve a useful three dimensional construct. The ability to print and control the gelation of a scaffold material has many advantages and can help overcome key obstacles preventing the progression of biological printing such as hydration and pattern stability. Matrix printing can also be exploited to form precise chambers in which individual cells can be trapped and monitored to increase our understanding of cell-cell communication A popular material of choice for biological printing are hydrogels as these offers the aqueous environment required for cell survival and a phase change enabling it to form a three-dimensional structure post printing. Examples of such gels include sodium alginate and fibrin-thrombin combination as well as synthetic choices such as pluronics. The material choice is dictated by several factors such as cytocompatibility, printability, gel strength and the ease of phase change; the resultant material choice is often a compromise dictated by logistics and final application.

3.4 Conclusion This chapter reported the successful deposition of mammalian cells using a piezoelectric DOD printer. However there is still progress to be made in standardising a suitable protocol which can encompass all variable factors. Until this becomes a reality it is impossible to assume that all cells will remain unaffected by the printing parameters; as highlighted by the osteoblast proliferation results. Long term studies into the overall cell phenotype and behaviour post printing are also required. The second aspect of the bioprinting lies in the formulation of a suitable ‘ink’ which can offer compatibility both with biological components and the printing apparatus.

References 1. Calvert P (2001) Inkjet printing for materials and devices. Chem Mater 13(10):3299–3305 2. Derby B (2008) Bioprinting: inkjet printing proteins and hybrid cell-containing materials and structures. J Mater Chem 18(47):5717–5721

50

R. Saunders et al.

3. Boland T, Tao X, Damon BJ, Manley B, Kesari P, Jalota S, Bhaduri S (2007) Drop-on-demand printing of cells and materials for designer tissue constructs. Mater Sci Eng C Biomimetic Supramol Syst 27(3):372–376 4. Nakamura M, Kobayashi A, Takagi F, Watanabe A, Hiruma Y, Ohuchi K, Iwasaki Y, Horie M, Morita I, Takatani S (2005) Biocompatible inkjet printing technique for designed seeding of individual living cells. Tissue Eng 11(11–12):1658–1666 5. Saunders RE, Gough JE, Derby B (2008) Delivery of human fibroblast cells by piezoelectric drop-on-demand inkjet printing. Biomaterials 29(2):193–203 6. Xu T, Gregory CA, Molnar P, Cui X, Jalota S, Bhaduri SB, Boland T (2006) Viability and electrophysiology of neural cell structures generated by the inkjet printing method. Biomaterials 27(19):3580–3588 7. Xu T, Jin J, Gregory C, Hickman JJ, Boland T (2005) Inkjet printing of viable mammalian cells. Biomaterials 26(1):93–99 8. Wendt D, Marsano A, Jakob M, Heberer M, Martin I (2003) Oscillating perfusion of cell suspensions through three-dimensional scaffolds enhances cell seeding efficiency and uniformity. Biotechnol Bioeng 84(2):205–214 9. Reis N, Ainsley C, Derby B (2005) Ink-jet delivery of particle suspensions by piezoelectric droplet ejectors. J Appl Phys 97(9):94903 10. Martin GD, Hoath SD, Hutchings IM (2008) Inkjet printing – the physics of manipulating liquid jets and drops. J Phys Conf Ser 105:012001 11. Deegan RD, Bakajin O, Dupont TF, Huber G, Nagel SR, Witten TA (1997) Capillary flow as the cause of ring stains from dried liquid drops. Nature 389(6653):827–829 12. Deegan RD, Bakajin O, Dupont TF, Huber G, Nagel SR, Witten TA (2000) Contact line deposits in an evaporating drop. Phys Rev E 62(1):756–765 13. Smith PJ, Shin DY, Stringer JE, Derby B, Reis N (2006) Direct ink-jet printing and low temperature conversion of conductive silver patterns. J Mater Sci 41(13):4153–4158

Part III

Modified Laser Induced Forward Transfer (LIFT) Approaches

Chapter 4

Laser-Induced Forward Transfer: A Laser-Based Technique for Biomolecules Printing P. Serra, M. Duocastella, J.M. Fernández-Pradas, and J.L. Morenza

Abstract The high focusing power of lasers makes them adequate for micropatterning applications. Laser-induced forward transfer (LIFT) is a direct-writing technique allowing the deposition of tiny amounts of material from a donor thin film to a solid substrate through the action of a pulsed laser beam. Although LIFT was originally developed to operate with solid films, it has been demonstrated that deposition is also possible from liquid films. In this case the material is directly ejected in the liquid state from the film and transferred to the receptor substrate, where it deposits in the form of a microdroplet. The relative translation of the film-substrate system respect to the laser beam enables the formation of two-dimensional patterns. This makes LIFT adequate for biomolecule printing: microdroplets of biological solutions can be transferred onto solid substrates to produce patterns of immobilized biomolecules. In this chapter a review on the LIFT technique for biomolecule printing is carried out, from its origins to the most recent developments. The characteristics and performances of the technique are described in detail, with special attention to the diverse possible modes of operation and transfer mechanisms. It is also shown that significant benefits in terms of resolution, speed, contamination, and sample consumption can be obtained through LIFT when compared to other more conventional direct-writing approaches. Finally, the feasibility of the technique for biomolecule printing is demonstrated through examples of successful deposition of a large set of different biomolecules.

4.1 Introduction The production of micron-scale patterns of biomolecules constitutes a major demand of several new fields in the biomedical area. Thus, the emergence of genomics and the recent advances in gene therapy are difficult to conceive without the invention of DNA microarrays [1–3]. The appearing of such devices has made P. Serra (B) Departament de Física Aplicada i Òptica, Universitat de Barcelona, Martí i Franquès 1, 08028, Barcelona, Spain e-mail: [email protected] B.R. Ringeisen et al. (eds.), Cell and Organ Printing, C Springer Science+Business Media B.V. 2010 DOI 10.1007/978-90-481-9145-1_4,

53

54

P. Serra et al.

possible the parallel and rapid detection and identification of DNA sequences, which has resulted in an unprecedented increase in the capability of acquiring and processing genetic information. In an analogous way, protein microarrays play a similar role in the expanding fields of proteomics, clinical diagnostics and drug discovery [4–6]. In addition, miniaturization is a major concern in biosensors fabrication: it is essential for multianalyte detection, the possibility of implanting biosensors for in-situ diagnostic and control of pathologies, or biosensors integration in lab-on-a-chip systems [7–10]. Finally, another large field of application of biomolecule micropatterning is tissue engineering and cellular studies [11–13]. Patterns of biomolecules can act as templates for cells, allowing the analyses of cell adhesion and growth, cell-cell interactions, or cell clustering. In fact, biomolecule micropatterning constitutes one of the main indirect strategies for cell printing. The need for generating micropatterns of biomolecules has prompted much research on microfabrication technologies, either through the development of novel techniques, or through the adaptation of previously existing ones to the new materials. Two completely different approaches are possible to achieve the objective of micropatterning: pattern-transfer and direct-writing. Pattern-transfer is a term which includes all the techniques in which the designed pattern is previously prepared in a mask or a mold, and then transferred as a whole on the substrate where the definitive pattern will be produced [14]. Photolithography is possibly the best known micropatterning technique which corresponds to the pattern-transfer class. It is a perfect example of a technique being adapted from a different area, in this case microelectronics. In photolithography, feature resolutions down to 1–2 μm can be routinely obtained [15, 16]. Like most pattern-transfer techniques, it is well suited for large scale production, thanks to its parallel fabrication character. However, photolithography is expensive and time consuming. Another widely used pattern-transfer technique, especially for cell printing applications is microcontact printing [17, 15, 18]. This technique, with a resolution similar to that of photolithography, is more flexible and cost-effective, which has made it very popular in biological research. In contrast to pattern-transfer, direct-writing involves techniques in which the different features of a pattern are sequentially produced [19]. This serial character allows a very rapid transition from the design of the pattern to its realization, since in this case the time-consuming and expensive process of mask or mold preparation is avoided. Consequently, direct-writing techniques are usually not the best option for mass production, but appear to be very convenient in customized fabrication and, therefore, in research applications. The most common direct-writing technique is inkjet printing [20, 21]. This is another example of adaptation to biomolecule printing of a previously existing technique, in this case from the graphics art industry. Inkjet printing is inexpensive, flexible, and fast, especially when compared to pattern-transfer techniques. However, typical resolutions are not so good: most conventional printers present resolutions of several tens of microns [22, 15]. Another serious drawback of inkjet printing is clogging of print heads. This problem becomes critical when good resolution is required, as droplet dimensions are determined by the nozzle diameter [23, 24]. Another common direct-writing technique is

4

Laser-Induced Forward Transfer

55

pin-microspotting [25]. Although flexible and cheap, it is not as fast as inkjet printing, and achievable resolutions are similar and even worse [22]. It is worth mentioning that in the field of cell printing, direct-writing techniques present the additional attractive feature of allowing the direct printing of cells [26]. In the field of direct-writing techniques, laser-induced forward transfer appears to be an interesting alternative which enables many of the drawbacks associated with most conventional technologies to be overcome. In this chapter a detailed survey of the characteristics and performances of the technique is carried out, reviewing the different possible modes of operation and transfer mechanisms, and paying special attention to the use of the technique in the area of biomolecules printing.

4.2 The Laser-Induced Forward Transfer Technique 4.2.1 Overview Laser-induced forward transfer (LIFT) is an additive laser-direct writing technique which allows the fabrication of micrometric patterns of diverse materials on solid substrates. Its principle of operation is schematized in Fig. 4.1, and can be described as follows: a solid thin film of the material to be written (donor film) is previously deposited onto a substrate transparent to the laser radiation (donor substrate), and located parallel to and facing the substrate where the pattern will be created (acceptor substrate). In a first step, a pulsed laser is used to induce the transfer of a small fraction of material from the donor film onto the acceptor substrate. The laser beam is focused at the interface between the donor film and the donor substrate, and when a laser pulse is fired, this propagates through the transparent donor substrate, and is finally absorbed at the mentioned interface. The energy transfer from the laser pulse to the donor film results in the ejection of a portion of material from this film towards the acceptor substrate, where the portion lands, leading to the formation of a pixel. This step is then followed by the translation of the donor-acceptor system up to the position where the next pixel must be formed, and deposition proceeds again. In this way any two-dimensional pattern can be written through repetition of the described process as many times as pixels composing the pattern. Alternatively to donor-acceptor translation, patterning can be also achieved through scanning of the laser beam along the donor film. LIFT was reported in such an early date as 1971, in a patent by A.D. Brisbane [27] in which the principle of operation of the technique is described in all detail. In that document it is proposed that LIFT could be used for the formation of Pt interconnects on silicon, presumably for microelectronics applications. Surprisingly, the first published paper appeared much later, in 1986 [28]. In that classical paper, J. Bohandy et al. describe the deposition of Cu lines of about 50 μm width on silicon and silica substrates using ArF excimer laser radiation. The same research group also succeeded depositing Cu and Al features on fused silica [29] by means of irradiation with the second harmonic of a Nd:YAG laser. Although the first published experiments were all carried out under high vacuum conditions, it was

56

P. Serra et al.

Fig. 4.1 LIFT setup scheme: A pulsed laser beam (1) is focused by means of an optical system (2) through a transparent substrate (3) onto a donor thin film (4). As a result, material from the film is transferred (5) on the acceptor substrate (6)

soon demonstrated that LIFT could successfully operate at atmospheric pressure, which considerably simplifies the experimental procedure [30]. This triggered an extensive research on LIFT which lasts to the present day and that has resulted in the deposition of a large variety of inorganic materials: mainly pure metals, like Au [30], Ti [31], W [32], Cr [33], Ni [34], Al [35], and Zn [36], but also simple oxides: Al2 O3 [37], In2 O3 [38, 39], V2 O5 [40], and even high-temperature superconductors [41]. All those results served to prove the feasibility of LIFT for micropatterning inorganic materials, and helped determine the parameters affecting its degree of spatial resolution. Some of them are related to the laser beam: its dimensions on the thin film (mainly determined by the focusing power of the optical system), the beam intensity distribution, and the laser pulse duration. Small beam dimensions and short pulse durations result in smaller amounts of transferred material. Indeed, the irradiated area determines the region where the energy delivered by the laser beam is absorbed, and the duration of the pulse limits the degree of spreading of this energy away the irradiated area due to heat conduction. In addition, the nature of the material to be transferred also contributes to the final resolution limit of the technique: the effect of the laser beam on the target material is very dependent on its optical and thermal properties. Strongly related with this, another factor which also influences the degree of spatial resolution is the thickness of the donor thin film [42]. Finally, there is an important contribution of the geometrical configuration of the system to the resolution of the technique: the separation distance between the thin film and

4


57

Fig. 4.2 (a) Scheme of transfer mechanism via partial vapor formation and propulsion of the molten film. (b) Optical microscopy images of Cr dots deposited on glass by femtosecond laser microdeposition (adapted from [33])

the acceptor substrate is in general a crucial factor in this sense, since it determines in some extent the degree of spreading of the deposited material on the acceptor substrate [43]. In the classical model proposed to describe LIFT [42], the material ablation process takes place as follows: the laser pulse is absorbed at the film-donor substrate interface, resulting in heating of that area, and in the further generation of a melt front that propagates through the film until it reaches the free surface. Meanwhile, the material in the interface attains its boiling point, leading to the formation of a high-pressure vapor bubble that expands, and propels the molten material in front of it towards the acceptor substrate. This scheme, illustrated in Fig. 4.2a, clearly implies that both transfer and deposition occur in the liquid state, resulting from melting of the solid film, and that the mechanism of transfer is the laser-generated vapor bubble. The microscopy images in Fig. 4.2b show LIFT transferred Cr dots displaying some splashing; this indicates that transfer indeed took place in molten phase. The described transfer mechanism is therefore adequate for the deposition of simple inorganic materials, which can melt and re-solidify without significant alteration of their properties. However, it is completely inappropriate for the transfer of biological materials, like biomolecules or cells, which would irreversibly decompose in the instance previously described. Therefore, in the pursuit of biological materials deposition through LIFT, it is required to move away from the classical scheme, where both melting and vaporization of the transferred material take place, and search for alternative methods that enable such problematic issues to be overcome.

4.2.2 Alternative Approaches The first strategy followed to overcome the irreversible decomposition of sensitive materials during LIFT was the use of a sacrificial layer between the donor substrate and the donor thin film [44] [45]. In such case, the laser radiation is absorbed in the intermediate layer, typically a metal, which vaporizes in the same way as shown in

58

P. Serra et al.

the scheme of Fig. 4.2a; during vapor expansion, the material on top of the vapor cloud is pushed away, causing the release of a portion of the thin film of interest, which is transferred to the acceptor substrate. Since this transferred material is not vaporized or melted, it can be deposited without physical or chemical transformation and, in this way, the transfer of fragile materials appears to be possible. Recently, this approach has been followed to successfully deposit different polymers; the metal absorbing layer has been substituted by an intermediate layer of triazene, a polymer which easily decomposes into gaseous fragments. Thus, XeCl excimer laser radiation has been used to deposit polymers like gelatine, methylcellulose [46] or aryltriazene polymers [47]. Although this strategy has been mainly used in organic electronics applications, it is worth mentioning that it can be successfully applied to the deposition of biological materials as well. In fact, the first feasibility test for the use of an intermediate layer of decomposable triazene consisted in the deposition of mammalian cells without harm [48]. Another alternative to classical LIFT for the deposition of fragile materials consists in the transfer of films composed of a powder of the material of interest embedded in a matrix, typically an organic binder strongly absorbent to the laser radiation. The organic matrix is then preferentially vaporized under the action of the laser pulse, leading to the release of the undamaged powder which deposits onto the acceptor substrate. According to this transfer mechanism, which relies on the same principle as the previously existing technique matrix-assisted pulsed laser evaporation (MAPLE), this direct-write version was named MAPLE-DW [49, 50]. The technique has been used to successfully deposit powders of diverse materials, like Ag, BaTiO3 , SrTiO3 and Y3 Fe5 O12 , through excimer laser irradiation [51]. According to the given definition, MAPLE-DW would not be appropriate for transferring biological materials, since these are not usually found in powder form. However, it has to be noted that the acronym MAPLE-DW has been often used in the literature to designate the transfer of any multicomponent film, in spite of the actual transfer mechanism [52]. As a consequence, the term has also been employed to designate biomolecules and cells deposition in instances which do not exactly fit into the given definition. Nevertheless, such generic use tends to be abandoned by most authors in order to avoid further confusion.

4.3 Liquids Printing 4.3.1 LIFT of Liquid Films The strategies described in the previous section indeed allow circumventing the problematic issues of classical LIFT associated with the potential thermal decomposition of the material to be transferred. However, they do not appear to be the optimum choice in the case of biomolecules printing, as most focus on the use of solid films. Biological systems are usually provided and handled in solution, and this state furnishes them with the required mobility which promotes their binding to the acceptor substrate once printed. Luckily, LIFT can be also carried out with

4


59

liquid films, and in this case there is no need to completely melt or vaporize the fraction of material which is transferred to the acceptor substrate. This makes possible the deposition of very fragile materials, as it was first demonstrated with inorganic materials for microelectronic applications [53]. In this case, transfer takes place directly from liquid donor films. The material to be transferred is then suspended or dissolved in a liquid to form an ink, which is spread on the donor substrate, and LIFT is carried out from the resulting liquid film. In these conditions the effect of the laser pulse, absorbed at the solid-liquid interface between the donor substrate and the film, results in the direct vaporization of a small fraction of the ink and, therefore, in the formation of a high-pressure vapor bubble that expands, propelling the remaining liquid in front of it (Fig. 4.3). The material to be deposited, suspended or dissolved in the ink, is dragged by the propelled solvent and transferred to the acceptor substrate, where it finally lands. In this scheme the solvent acts as a transport vector for the material of interest. In the LIFT of liquid films, the transfer mechanism is similar to that of Fig. 4.2a. The main difference is that in the case of inks there is no need to melt the film, and thus the transferred material is not overheated, which prevents it from decomposition. Furthermore, boiling temperatures of normal liquid solvents are in general much lower than those of solids; in consequence, the amount of energy delivered by the laser pulse to generate the vapor bubble is lower than in classical LIFT, and therefore it would be also lower the amount of heat transferred to the surrounding material. All these considerations lead to the conclusion that the LIFT of liquid films is a smart process where transfer occurs in a way gentle enough to allow the deposition of biological materials without harm. This has been extensively proved through the successful transfer of a wide variety of biomolecules and cells, as is illustrated in Table 4.1. Under the appropriate irradiation conditions, liquid deposition through LIFT takes place in the form of circular and uniform microdroplets (Fig. 4.4), which turns LIFT into a true microprinting technique. In fact, when applied to transfer from liquid films, LIFT operates in a way completely analogous to inkjet printing, the most extended microprinting technique. However, LIFT possesses some unique advantages over ink jet techniques. In contrast to inkjet printing, where the range of printable viscosities is always narrow [24], there are no important restrictions in LIFT concerning the composition and rheology of the materials that can be

Fig. 4.3 Scheme of transfer mechanism for LIFT of liquid films via vapor bubble formation and pushing of the liquid film

Nitrocellulose coated glass slide Matrigel coated slides, silinated glass slides Nitrocellulose coated glass slide

Glass slides Poly-l-lysine coated slide

ArF (193 nm, 20 ns) ArF (193 nm, 20 ns)

ArF (193 nm, 20 ns) Nd:YAG (266 nm, 3–5 ns), KrF (248 nm, 2.5 ns)

KrF (248 nm, 2.5 ns)

KrF (248 nm, 500 fs) Nd:YAG (355 nm, 10 ns)

Polyphenol oxidase (PPO)

Bovine serum albumin (BSA)

Glass slides Poly-l-lysine coated slide Nitrocellulose coated glass slides

Nd:YAG (355 nm, 10 ns) KrF (248 nm, 2.5 ns)

KrF (248 nm, 500 fs) Nd:YAG (355 nm, 10 ns)

KrF (248 nm, 15 ns–500 fs)

Treponema pallidum 17 kDa antigen Alkaline phosphatase

Glutathione t–transferase (GST) Rabbit antibody immunoglobulin G (IgG) Enzyme horseradish peroxidase

Polystyrene plate, quartz plate, digene silanated slide Nylon-coated glass slide Nitrocellulose coated glass slide

ArF (193 nm, 20 ns)

Anti-BSA

Pt microelectrode Polystyrene plate, quartz plate, digene silanated slide

Poly-l-lysine coated slide Poly-l-lysine coated slide Glass slides

Nd:YAG (355 nm, 10 ns) Nd:YAG (355 nm, 10 ns) KrF (248 nm, 500 fs)

Salmon sperm DNA Human cDNA Lambda bacteriophage DNA

Substrate

Laser

Biomolecule or cell type

Table 4.1 Biomolecules and cells deposited through LIFT

Gold (10 nm)

Titanium (60 nm) Gold (100 nm), titanium (75 nm), titanium oxide (85 nm) – Titanium (60 nm)

Gelatin

– Gold (35 nm), titanium (75 nm), titanium oxide (85 nm) Gold (100 nm), titanium (75 nm), titanium oxide (85 nm) – Titanium (60 nm)

– Gelatin

Titanium (60 nm) Titanium (50 nm) –

Absorbing film

[68]

[60] [66, 67, 59]

[65] [22]

[49, 50]

[60] [64]

[22]

[62] [63]

[61] [49, 50]

[54] [55–59] [60]

References

60 P. Serra et al.

KrF (248 nm, 15 ns–500 fs)

ArF (193 nm, 30 ns) KrF (248 nm, 15 ns) KrF (248 nm, 15 ns–500 fs)

KrF (248 nm, 15 ns–500 fs)

KrF (248 nm, 15 ns–500 fs) KrF (248 nm, 15 ns–500 fs)

Photobiotin

Streptavidin Peptides (from the adenovirus fiber) Avidin

Titin

Biotin

Matrigel coated slides

KrF (248 nm, 2.5 ns)

ArF (193 nm, 20 ns)

Matrigel coated slides

Nd:YAG (266 nm, 3–5 ns), KrF (248 nm, 2.5 ns)

–

– Titanium, titanium oxide (75–85 nm) Gold (35 nm), titanium (75 nm), titanium oxide (85 nm) Gold (35 nm), titanium (75 nm), titanium oxide (85 nm)

Gelatin

Gold (10 nm) –

Cromium (40 nm)

Gold (10 nm) Gold (10 nm)

Gold (10 nm)

– Gold (10 nm) Gold (10 nm)

Gold (10 nm)

Absorbing film

[75, 49]

[77]

[63]

[75] [76]

[49, 50]

[70] [61]

[74]

[68] [73]

[73]

[71] [72] [73]

[69, 70]

References


Rat cardiac cells

Matrigel coated slides, silinated glass slides

ArF (193 nm, 20 ns) Nd:YAG (266 nm, 5 ns)

ArF (193 nm, 20 ns)

Human osteosarcoma cells

Glass slides, nitrocellulose coated glass slides, agarose and ORMOCER treated glass slides Glass slides Gold coated glass slides Glass slides and nitrocellulose coated glass slides Glass slides and nitrocellulose coated glass slides Nitrocellulose coated glass slides Glass slides and nitrocellulose coated glass slides Si and Low Temperature Oxide on Si (LTO/Si) ORMOCER coated glass slides Quartz plate

Substrate

Polystyrene plate, quartz plate, digene silanated slide Matrigel coated slides Matrigel coated slides

KrF (248 nm, 15 ns) ArF (193 nm, 20 ns)

Amyloid peptide Chinese hamster ovaries

Nd:YAG (266 nm, 4 ns)

Laser


Table 4.1 (continued)

4 61

KrF (248 nm, 30 ns)

KrF (248 nm, 30 ns)

KrF (248 nm, 30 ns)

KrF (248 nm, 2.5 ns)

ArF (193 nm, 10 ns)

Nd:YVO4 (355 nm, 15 ns) ArF (193 nm, 20 ns) ArF (193 nm, 20 ns)

ArF (193 nm, 20 ns) ArF (193 nm, 20 ns) ArF (193 nm, 20 ns) Nd:YAG (266 nm, 5 ns)

ArF (193 nm, 20 ns)

Rat Schwann cells

Pig lens epithelial cells

Astrogial cells

Bovine aortic endothelial cells

Rat neuroblasts cells

Bovine embryonic stem cells Human dermal fibroblasts Bovine pulmonary artery endothelial cells Human breast cancer cells Rat neural stem cells Mouse myoblasts Rat olfactory ensheathing cells

Escherichia coli

ArF (193 nm, 20 ns)

KrF (248 nm, 30 ns)

Prostate tissue

Fungus (trichoderma conidia)

KrF (248 nm, 2.5 ns)

Laser


Polystyrene plate, quartz plate, digene silanated slide Cell media coated plate

Si glass slides, nutrient agar culture plates Agar-coated glass slide, solid LB Petri dish, microtiter Plate

Matrigel coated Petri dish Matrigel coated Petri dish Matrigel coated Petri dish Cell culture chamber coated with poly(l-lysine) and filled with medium, Matrigel coated slides

Glass plates coated with a thin wet gelatin layer Glass plates coated with a thin wet gelatin layer Glass plates coated with a thin wet gelatin layer Matrigel coated slides, cell media coated slides Basement membrane (Matrigel) coated slides Matrigel coated Petri dish Matrigel coated Petri dish Matrigel coated Petri dish

Substrate

Table 4.1 (continued)

Silver (50 nm)

Gelatin

Gold (35 nm), titanium (75 nm)

–

[85]

[49, 50]

[84]

[83]

[81] [81] [81] [82]

[80] [81] [81]

Polyimide (4 μm) – – – – – Titanium, titanium oxide (30–40 nm)

[48]

[79]

[78]

[78]

[78]

References

Metal or metal oxide (10–100 nm) –

Silver (100 nm)

Silver (100 nm)

Silver (100 nm)

Absorbing film

62 P. Serra et al.

4


63

Fig. 4.4 Optical microscopy image of an array of droplets transferred from a liquid film of a solution containing water and glycerol at 50% v/v

printed [52]. On the other hand, the degree of spatial resolution is potentially higher than that of inkjet printing and other printing techniques, like pin-microspotting, and since the laser is a non-contact tool, contamination can be highly minimized. Finally, an important issue must be highlighted: LIFT is a nozzle-free technique. Its degree of resolution is achieved through focusing the laser beam on the donor film, which allows overcoming the clogging and drying problems characteristic of inkjet printing that compromise the performance of printing heads [23].

4.3.2 LIFT with an Absorbing Layer Most biological solutions are transparent to a wide range of common laser wavelengths, like those corresponding to solid state Nd:YAG lasers (355, 532 and 1,064 nm). This fact would make it nearly impossible to carry out LIFT of biomolecules with these wavelengths, since the laser radiation would not be absorbed in the donor film. Thus, the first biomolecule transfer experiments were performed with excimer lasers, mainly ArF lasers, which 193 nm radiation is well absorbed by almost any solution [49, 62, 61]. However, this poses several problems from a practical point of view that would make preferable the use of solid state lasers. Excimer lasers are gas lasers, less stable than solid state lasers, and with higher maintenance requirements. In addition, wavelengths as short as 193 nm require expensive optics with a relatively short life-time. Finally, there exist very small and inexpensive Nd:YAG lasers with a power high enough for LIFT applications, and which can be easily implemented in a desktop printing system, something very difficult to foresee with excimer lasers. This transparency issue can be overcome by using an absorbing film between the donor substrate and the liquid film, a strategy identical to the one used with the sacrificial layer described in Section 4.2.2. The laser pulse is then absorbed in the

64

P. Serra et al.

intermediate layer, leading to its vaporization and resulting in the formation of a vapor bubble in a similar way as in the LIFT of liquid films with no absorbing layer. As expected, the vapor bubble propels some liquid in front of it, so that the final result is the same as in the previous instances (Fig. 4.5). Several radiation absorbing materials have been proven successful for such a purpose; among them, metals like Ti, Au, Ag, and polymers like triazene (Table 4.1). In the particular case of metallic absorbing layers, which is the most extended one, the refined technique has received different names, depending on the authors. Thus, it is sometimes referred as BioLP, an acronym which stands for biological laser printing [63, 84], and sometimes as AFA-LIFT, which stands for absorbing film assisted laser induced forward transfer [85, 78]. It has to be emphasized that any damage effect to the biological material due to contamination by material ablated from the absorbing film has never been observed. In fact, for very thin metallic layers (tens of nm) no evidence of metal has been found in the deposited spots [54], and only with thicker films (hundreds of nm) a few residues can be detected [86]. It must be taken into account that the thickness of the metallic absorbing layer is always much lower than that of the donor liquid film, typically of a few tens of μm. In addition to allowing operation with practically any laser wavelength, the use of the intermediate absorbing layer presents a supplementary beneficial effect: it shields the biomaterial to be transferred from radiation exposure, an exposure which always constitutes a risk, especially under UV irradiation. Calculations of the optical penetration depth inside the absorbing layer have indeed revealed that for thicknesses of a few tens of nm, more than 99% of the incident laser radiation is absorbed in the layer [63]. The absorbing layer acts as an energy conversion device, converting the radiant laser energy via absorption and conduction into heat. Further calculations demonstrate that, in addition to this, the amount of heat subsequently

Fig. 4.5 (a) Scheme of transfer mechanism for the LIFT of liquid films by using an absorbing layer. (b) Optical microscopy images of transferred droplets (bottom) and corresponding ablated spots on a 50 nm Ti absorbing layer (top)

4


65

flowing into the liquid donor film is negligible. In consequence, the major part of the energy delivered to the absorbing layer is used to ablate this layer, creating the already described vapor bubble, with minimal thermal effect on the biomolecules. Bubble expansion leads to energy transfer to the liquid in the form of kinetic energy, which results in material ejection. An alternative strategy in the use of absorbing layers has been recently suggested: transfer through thick polymer films [80]. In this case the absorbing layer is composed of a polymer (e.g. polyimide) and has a thickness similar to that of the donor liquid film. Under these circumstances, material transfer takes place in a quite different way. The laser pulse is absorbed at the interface with the transparent donor substrate, giving place to the vaporization of a small fraction of the absorbing layer and, therefore, to the formation of a vapor bubble inside it. The pressure exerted by the expanding bubble locally deforms the polymer layer, which pushes the liquid material towards the acceptor substrate. In this transfer mechanism, the absorbing layer is not removed, but only plastically deformed; this makes the interaction between absorbing layer and donor film purely mechanical. Very recent experiments comparing both absorbing layers, thin metallic and thick polymeric, have revealed that although metallic layers in general do not lead to any damage to the transferred materials, in the case of exceptionally fragile materials the thick polymeric layer strategy can be advantageous [87]. Finally, it is interesting to note that transfer without absorbing layer removal, as here described, is also possible with metallic films under low intensity laser irradiation [88].

4.4 Experimental Setup and Methods The central element in a LIFT setup is the laser source. Among the great variety of pulsed lasers adequate for LIFT, short pulse lasers (pulse durations between fs and ns) are always the best choice, being wavelength selection determined by the optical characteristics of the material to be transferred. As mentioned in Section 4.3.2, Nd:YAG lasers are a very interesting option, with a broad range of wavelengths available: 1,064 nm (infrared), 532 nm (visible), 355, 266 and 213 nm (ultraviolet). Most commercially competitive lasers provide pulses with durations of a few nanoseconds. However, nanosecond lasers will probably be displaced by femtosecond lasers in the near future. The unique properties of femtosecond radiation make this kind of lasers the perfect candidate for LIFT applications: the extremely short pulse duration allows minimizing thermal effects, and promotes non-linear absorption in treated materials, which makes such absorption wavelength independent. In addition to the laser source, an optical system is another essential element in a LIFT setup: it allows focusing the laser beam in order to achieve the desired spatial resolution. Usually, a simple microscope objective can meet such requirement: commercially available objectives can easily attain beam diameters between 1 and 10 μm on the sample. Although the focusing system is the only optical element essential to the setup, many other accessory optical devices are normally found in

66

P. Serra et al.

a LIFT configuration: mirrors, lenses, diaphragms, beam splitters, and windows are used to shape, attenuate or conduct the beam along different paths in order to optimize its characteristics and better control the transfer process (for example, through in situ analysis of the pulse energy and beam intensity distribution). It is also desirable a CCD camera coaxial with the microscope axis to provide in situ control of the process. Finally, the translation stage is the last essential device in a LIFT setup. It makes possible the translation of the donor/acceptor system respect to the laser beam. The stage must provide translation steps smaller than the laser beam diameter in order to allow achieving good definition of the produced patterns. Although the sample translation is two dimensional, three dimensional XYZ stages are better suited, since they make possible the translation of the donor/acceptor substrate system along the laser beam dimension in a controlled way, thus providing a means of accurate focusing and defocusing of the laser beam on the sample. As mentioned in Section 4.2.1, galvanometric mirrors or other beam deflection systems can be used alternatively to the translation stage in order to deflect the beam following a predefined pattern on the sample. Liquid film preparation is a peculiar step in the LIFT process that cannot be found in any other printing technique. There exist a wide variety of methods that allow preparing rather uniform liquid films with thicknesses between 1 and 100 μm. Roller and blade coaters are probably the most popular, and can be perfectly used in biomolecule printing through LIFT [54, 65]. These methods require sample volumes of a few tens of microliter, volumes that are similar to those required by most conventional printing techniques [22]. An alternative possibility which provides with excellent results is the preparation of shallow wells on the donor substrate. These wells can have lateral dimensions in the order of several millimeters and depths of a few tens of microns, and can be made of photoresist or simply tape [77]. A precise sample volume is collected with a pipette and poured on the bottom of the well. One of the most interesting aspects concerning the use of wells is that they can accommodate sample volumes as small as 0.5 μL, which appears to be especially convenient when working with scarce and expensive samples, something which cannot be achieved with other techniques. The preparation of uniform films requires good wettability of the liquid to the absorbing layer; thus, the liquid can be spread on the layer surface in an easy way. Luckily, most metals are quite hydrophilic, and proteins usually act as good surfactants [65]. In the case of other biomolecules printing, like DNA, wettability can be improved through the addition of surfactants to the solution [54]. Finally, the last issue that has to be considered in relation with film preparation is the problem of evaporation: liquid films tend to evaporate quickly. Evaporation can be prevented through the addition of glycerol to the biological solution. However, care must be taken with protein solutions, since glycerol often acts as a protein denaturing agent: a compromise has to be reached between evaporation rate and glycerol concentration. Donor substrates must be transparent to the employed laser radiation. For visible, infrared, and near ultraviolet radiation, glass or even some polymers can be perfectly used: simple microscope slides are a usual option. Wavelengths below

4


67

350 nm require the use of fused silica substrates. Concerning the preparation of the absorbing layer, in the case of metals, simple coating techniques like evaporation or sputtering allow easily obtaining rather uniform and adherent films with thickness between 10 nm and 1 μm. In the case of polymeric absorbing layers, spin coating is in general the best option. Regarding acceptor substrates, conventional immobilization solutions are perfectly adequate [89, 90]. Thus, treated glass slides, like those coated with poly-L-lysine, amine, aldehyde or epoxy, can be used to adsorb or covalently bind biomolecules to the glass surface. These substrates are transparent to visible radiation, and appear to be very convenient for fluorescence detection. On the other hand, glass slides coated with porous membranes, like nylon or nitrocellulose, are alternative solutions; in these cases the biomolecules are immobilized through entrapment in the pores of the membrane. Such substrates are mostly used in chromogenic detection. One relevant issue from the experimental point of view is the separation between donor film and acceptor substrate. As it has already been pointed out in Section 4.2.1, in classical LIFT that distance is a crucial parameter which strongly influences the resolution of the technique. In order to avoid material spreading, very short separations should be required, typically of a few tens of microns. In the case of LIFT of liquid films, this would represent a serious problem in the design of a printing machine: the minimum contact between donor film and acceptor substrate would result in the complete wetting of the acceptor substrate and, consequently, in its irreversible loss. To avoid this, the design should grant a constant separation between donor film and acceptor substrate, with a very small degree of tolerance, which is not easy to achieve along the whole printing area of a microscope slide, for example. Luckily, the situation is not critical in truth, since it has been found that in the case of liquids there exists a wide range of separation distances allowing the printing of uniform droplets [73, 64, 91]. Figure 4.6 illustrates how this range extends over

Fig. 4.6 Optical microscopy image of a microarray of droplets transferred at different donoracceptor separation distances. The separation distance increases from left to right and is indicated on the top of each column in μm

68

P. Serra et al.

a few millimeters, and how the morphology and dimensions of the printed droplets are rather independent of separation between donor film and acceptor substrate. This aspect is a direct consequence of the particular transfer mechanism that takes place during the LIFT of liquids, and which is described in detail in Section 4.5.2.

4.5 Physics of the Transfer Process 4.5.1 Droplets Deposition The possibility of depositing microdroplets was already pointed out in the first work on protein printing through LIFT [62]. That pioneering work set the basis for future studies which would soon demonstrate that indeed LIFT could constitute a realistic alternative technique for biomolecules printing, allowing uniform deposition of circular microdroplets with high reproducibility and control [22, 54, 65, 58], and with better resolution than that obtained with more conventional techniques [64]. The influence of the different laser parameters on the characteristics of the LIFT deposited microdroplets has been analyzed in detail [22, 88, 58]. These studies have revealed that the size of the droplets depends on both laser pulse energy and laser focusing conditions: the droplet diameter increases with laser pulse energy for fixed laser beam dimensions, and decreases with focusing [88, 66, 67]. In fact, for each laser focusing conditions three different energy ranges can be identified (Fig. 4.7). Below a certain energy threshold no transfer occurs. At higher pulse energies transfer takes place in the form of well-defined circular droplets; the width of this energy window depends on the laser focusing conditions, and decreases with beam diameter. For energies above this window, irregular droplets, with satellites and splashing are obtained. Furthermore, it is also interesting to note that droplet dimensions also decrease with liquid film thickness; thus, through the use of very thin donor films and strong focusing, droplets with diameters below 10 μm have been successfully printed [64]. Their corresponding volume is about 10 fL, well below the minimum

Fig. 4.7 Optical microscopy image of a microarray of droplets transferred at increasing laser pulse energies (from left to right)

4


69

volumes attainable through conventional printing techniques [22]. It can be concluded, then, that good control on the microdroplets dimensions can be exerted through the simultaneous variation of both laser pulse energy and laser focusing conditions, and that excellent degrees of resolution can be achieved. The described analysis of the morphology of the LIFT deposited droplets [88] has also provided information on the physical mechanisms underlying the transfer process. The study has revealed that the droplet volume presents a linear dependence with laser fluence (energy per unit area on the sample), as shown in Fig. 4.8. Similar behaviors are typical of droplets generated by different kinds of inkjet printers: the droplet volume presents a linear relationship with a technological parameter directly related to the energy delivered to the liquid [92, 93]. The observed dependence in the case of LIFT reveals that liquid ejection is activated by two different thresholds. A primary fluence threshold (Fo in Fig. 4.8), independent of the laser focusing conditions, is the responsible for the generation of the vapor bubble inside the liquid film (see Section 4.3.1). However, bubble formation is not a sufficient condition for droplet deposition. In fact, the plots in Fig. 4.8 reveal that there is no material deposition unless a fluence value F1 >Fo , different for each different focusing conditions, is overcome. This value constitutes the secondary fluence threshold. For deposition to take place, the energy delivered to the bubble has to be high enough to allow the displaced liquid contacting the acceptor substrate (this aspect will be clarified in Section 4.5.2). The finding of the parameter Fo provides with additional information

Fig. 4.8 Plot of the transferred droplet volume versus laser pulse fluence for different beam dimensions: () ωx = 95 μm, ωy = 64 μm; () ωx = 73 μm, ωy = 47 μm; (•) ωx = 51 μm, ωy = 36 μm; () ωx = 31 μm, ωy = 25 μm; (—) Linear fit. Optical microscope images of representative series of droplets are shown on the right

70

P. Serra et al.

on the transfer process. This parameter, alongside the thickness of the donor film, determines the amount of material released per laser pulse. Indeed, it is shown that there is a good correspondence between the droplets volume and the volume of the cylinder which base has the lateral dimensions of the vapor bubble (region where the laser pulse local fluence overcomes Fo ) and which height is equal to the thickness of the liquid film [88, 67]. As a rule of thumb, the droplet volume is approximately given by the volume of the described cylinder.

4.5.2 Liquid Ejection and Transfer The studies described in Section 4.5.1 provide indirect information on the LIFT process, like the existence of the two transfer thresholds. However, the characterization of the deposited material alone is not enough to build a complete description of liquid ejection and transfer during LIFT. Such description has been provided by time-resolved imaging, a technique which allows visualizing liquid evolution during the whole transfer process. Stop-action images of the liquid dynamics are obtained through coupling a fast CCD imaging system to a microscope objective, and synchronizing it with laser pulse firing. Time-resolved imaging was already used to characterize material transfer during the LIFT of solid films [94, 95, 34]. The first analyses carried out on the LIFT of liquid films are previous to the application of the technique to biomolecules printing [96, 97]. In those analyses, a dense inorganic paste with a rheology very different from that of typical biological solutions was transferred. The turbulent liquid dynamics revealed by the images could in no way lead to the formation of welldefined circular microdroplets. However, those images already displayed one of the main characteristics of liquid ejection during LIFT: the jetting behavior [97]. Further imaging analyses carried out with biological solutions showed the formation of a short liquid jet which broke into individual droplets, in a similar way to liquid ejection from the nozzle of a printing head during inkjet printing [22]. Those results, alongside the resemblance between the LIFT deposited microdroplets and those obtained with inkjet printing, led to the interpretation that liquid transfer took place in the form of individual droplets which traveled in free flight until they landed on the acceptor substrate [22, 88]. However, recent time-resolved imaging experiments have changed that view, and revealed that liquid transfer follows a rather different dynamics during LIFT [98]. Time-resolved images of liquid ejection and transfer during LIFT at different delay times after the laser pulse are shown in Fig. 4.9 [98]. The images display a liquid dynamics which proceeds through the formation of surprisingly long jets that propagate in free space in a very stable way. Such dynamics can be described as follows: ablation of the absorbing layer by the laser pulse results in the formation of a vapor bubble inside the liquid film at the interface with the transparent donor substrate. The bubble, initially at a very high pressure, expands and displaces some liquid like an inflating balloon. During the asymmetric expansion of a bubble close to the free surface of a liquid, an overpressure is generated in the pole [99, 100].

4


71

Fig. 4.9 Images of the liquid ejection process obtained at different time delays with respect to the laser pulse. Integration time is 100 ns. The coated bottom surface of the donor substrate is seen as dark background on the top of the images. This surface acts as a mirror producing a reflected image upwards. A dotted line has been added for reference at the liquid film surface position (adapted from [98])

Such overpressure displaces the liquid in the pole towards the region where the pressure is lower, that is, the free space, leading to the generation of a jet. As expansion proceeds, the bubble inner pressure decreases, until it can no longer overcome the force due to external pressure plus liquid surface tension, which results in bubble collapse. During and after bubble collapse, the jet keeps on advancing at constant velocity while its diameter progressively decreases. Finally, Plateau-Rayleigh instabilities develop in the liquid surface, and it results in jet breakup. Typical liquid propagation speeds range between 10 and 100 m/s [98, 96]. The most surprising aspect of the observed dynamics is the perfect stability of the jets and their extremely high aspect ratio: if it is assumed that the jet maintains its constant velocity after its tip has left the imaged field, aspect ratios of about 800 are estimated [98]. According to these results, droplet deposition must proceed through contact of the liquid jet with the acceptor substrate instead of through individual flying droplets.

4.6 Biomolecules Printing 4.6.1 DNA Printing The main interest of DNA printing is the development of biosensors allowing parallel and rapid detection and identification of genes and gene mutations for genomics and biomedical diagnostic applications (see Section 4.1). Such objectives are usually pursued through the fabrication of DNA microarrays, which thanks to the high degrees of integration currently achievable allow processing high volumes of

72

P. Serra et al.

information in a single device. In this frame, LIFT, with its potential advantages over more conventional printing techniques, can play a relevant role in the preparation of DNA microarrays and other DNA-based miniaturized devices. The LIFT of liquid films was first applied to DNA printing in 2003 [54]. In that very preliminary work, microdroplets of a solution containing double-stranded DNA of salmon sperm were printed onto poly-L-lysine coated glass, and then stained with ethidium bromide (EtBr) and submitted to a fluorescence test. The test revealed fluorescence signal only in locations corresponding to the printed pixels, so that EtBr correctly stained the printed DNA. However, that experiment did not involve DNA hybridization and therefore, did not prove that LIFT was feasible for microarray printing. Nevertheless, that simple study constituted the proof-of-principle for future works which would extensively prove the suitability of the technique for the desired purpose. The feasibility of LIFT for DNA microarrays printing was first demonstrated through the preparation of a functional microarray which allowed discriminating single DNA strands corresponding to two different human genes [56–59]. The results of that experiment are illustrated in Fig. 4.10. A microarray of droplets (Fig. 4.10a) was prepared with three different solutions: droplets in rows 1, 4, and 7 contained a single strand DNA of the v-ets avian erythroblastosis virus E26 oncogene homolog 2 (ETS2, 2,205 base pairs long), droplets in rows 2, 5, and 8 corresponded to the buffer solution alone (negative control), and droplets in rows 3, 6, and 9 contained a single strand DNA of the mitogen-activated protein kinase 3 gene (MAPK3, 525 base pairs long). After deposition, the microarray was submitted to a standard hybridization protocol with the complementary strands of the transferred DNA, each one differently tagged: ETS2 with Cy3 and MAPK3 with Cy5. The fluorescence image recorded after hybridization is presented in Fig. 4.10b. This result demonstrates that hybridization occurred only where DNA was deposited (no fluorescence signal was recorded in the columns corresponding to the negative control), that there was enough signal to be easily detected with a

Fig. 4.10 (a) Image of a microarray of droplets of three different solutions: ETS2 DNA (1, 4, 7), negative control (2, 5, 8) and MAPK3 DNA (3, 6, 9). (b) Corresponding fluorescence image after hybridization with Cy3 abd Cy5 tagged complementary strands

4


73

conventional fluorescence test, and that the prepared microarray was very specific in gene discrimination. Furthermore, it must be noted that the performance of LIFT-prepared DNA microarrays is comparable to that of those prepared with other conventional techniques. Indeed, quantitative analyses where LIFT was confronted with pinmicrospotting proved that both techniques provided equivalent results [55]. DNA spots of identical characteristics (same solutions, concentrations and spot diameters) were printed on the same slides through both LIFT and pin-microspotting, and submitted to hybridization with their perfect complementary strands tagged with fluorophores (Cy3 and Cy5). Measurements of fluorescence intensity corresponding to specific and non-specific hybridization revealed very similar values for both techniques. The functionality of LIFT for microarrays printing was demonstrated to be analogous to that of the most extended conventional techniques. DNA printing through LIFT has also been carried out from solid films [101, 60]. In this case, the use of femtosecond laser radiation is claimed to be a requirement for successful transfer: biomolecules are submitted to direct irradiation, and only ultrafast pulses can grant minimal thermal damage. However, no hybridization tests which proved the viability of the technique for microarrays printing have ever been reported.

4.6.2 Protein Printing In a similar way to DNA, protein printing also represents a promising avenue of research in a broad field of applications, from the preparation of protein microarrays for proteomics and drug discovery to the fabrication of miniaturized biosensors for clinical diagnostics, food and environment control, or defense. And, among them, it is particularly interesting to mention protein patterning for cell-growth studies in tissue engineering (see Section 4.1). Once again, LIFT appears as an attractive direct-writing technique which could fulfill the objectives envisioned in these applications. The first reported protein being printed through a LIFT-based technique was an enzyme, polyphenol oxidase [61]. In fact, MAPLE-DW was used to deposit a mixture of banana tissue (which contains polyphenol oxidase) with mineral oil and graphite powder onto previously fabricated metal electrodes (shown in Fig. 4.11a). The functionality of the enzyme after transfer was tested through electrochemical characterization of the deposited material in the presence of catechol; the test revealed that polyphenol oxidase preserved its enzymatic activity after transfer (Fig. 4.11b). The ambiguous use of the term MAPLE-DW pointed out in Section 4.2.2 does not allow knowing whether that experiment really corresponded to LIFT of liquid films. Bovine serum albumin (BSA) was the first reported protein unambiguously being printed through the LIFT of liquid films [62]. Furthermore, that work presents the additional interest of reporting the first biomolecules microarray ever printed with the technique. ArF laser radiation was used to prepare microarrays of both biotinylated BSA and anti-BSA on nitrocellulose-coated glass.

74

P. Serra et al.

Fig. 4.11 (a) Banana tissue based electrochemical biosensor showing Pt electrodes with deposited PPO/graphite/mineral oil composite on one of the electrodes. (b) Current vs. catechol concentration at a bias of –2.0 V (from [61])

The positive results of the fluorescence tests against Cy5-tagged streptavidin and Cy5-tagged BSA revealed that proteins were functional in the selective recognition of their complementary molecule after deposition, which proved the feasibility of LIFT for protein printing. Figure 4.12 shows an example of immunoassay on LIFT printed BSA. The large variety of existing proteins has resulted in an extensive number of different protein types being printed through LIFT, from simple peptides [72] to titin, the largest known protein [73]. Table 4.1 illustrates well such variety. LIFT has been used to deposit enzymes, like alkaline phosphatase [22], glutathione S-transferase [60], or horseradish peroxidase [68], antibodies, like immunoglobulins [66, 67, 59], and antigens, like the Treponema pallidum 17 kDa antigen [65]. This last case, with application in the fabrication of syphilis diagnostic kits, presents the interesting feature that the printed microarray (deposited on nylon), was submitted to a chromogenic test against serum containing both specific and non-specific antibodies, revealing antigen-antibody binding only in the specific case. Finally, it has to be also noted that biological molecules other than proteins have also been successfully printed through the LIFT of liquid films: biotin [74, 73], avidin [73], and streptavidin [71]. The use of femtosecond lasers for printing proteins through LIFT has received some attention [68,69,73]. In fact, a comparative study between femtosecond and

4


75

Fig. 4.12 (a) A 2,000-spot BSA microarray after immunoassay with anti-BSA (imaged with fluorescent detection; 266 spots shown). (b) Higher magnification of one portion of the microarray (from [22])

nanosecond radiation [68] revealed that the former case was clearly advantageous for protein printing. In that work, LIFT with both radiations was employed for depositing droplets of a solution containing horseradish peroxidase (HRP). Figure 4.13 displays some of those deposits after submission to chromogenic test. Quantitative analyses carried out through the combination of Bradford and enzymatic assays allowed determining the percentage of active HRP after transfer.

Fig. 4.13 Horseradish peroxidase spots deposited with femtosecond and nanosecond laser pulses after chromogenic assay (from [73])

76

P. Serra et al.

Results indicated that up to 78% of deposited biomolecules were functional in the case of femtosecond radiation, while this percentage dropped to 54% in the case of nanosecond radiation, revealing the significant advantage of using ultrafast radiation. This, in addition to the excellent features pointed out in Section 4.4, make femtosecond lasers a promising tool for biomolecules printing through LIFT.

4.7 Summary The application of the LIFT technique to material patterning from liquid films has proved successful for biomolecule printing. At the appropriate laser pulse energy and focusing conditions, material deposition takes place in the form of circular and uniform microdroplets, with minimum diameters of the order of a few microns. Material transfer proceeds through the formation of liquid jets displaying an extraordinary aspect-ratio; these propagate at constant velocity until they contact the acceptor substrate, where microdroplets are formed. Its characteristic mode of operation, as well as its particular transfer mechanism, provide LIFT with the unique properties which allow it to compete with other conventional printing techniques in terms of resolution, speed, contamination, and sample consumption. The controlled deposition of such diverse biological molecules as DNA, proteins, peptides and vitamins, without compromising their activity, proves the feasibility of LIFT for biomolecule printing, and definitely makes it an interesting tool for cell printing applications. Acknowledgments Acknowledgment is given for support by MCI of the Spanish Government (Projects MAT2007-62357 and CSD2008-00023), and Fondo Europeo de Desarrollo Regional (FEDER).

References 1. Heller MJ (2002) DNA microarray technology: devices, systems, and applications. Annu Rev Biomed Eng 4:129–153 2. Schulze A, Downward J (2001) Navigating gene expression using microarrays – a technology review. Nat Cell Biol 3: E190–E195 3. Stoughton RB (2005) Applications of DNA microarrays in biology. Annu Rev Biomed Eng 74:53–82 4. MacBeath G (2002) Protein microarrays and proteomics. Nat Genet 32:526–532 5. Templin MF, Stoll D, Schrenk M et al (2002) Protein microarray technology. Trends Biotechnol 20:160–166 6. Templin MF, Stoll D, Schwenk JM et al (2003) Protein microarrays: promising tools for proteomic research. Proteomics 3:2155–2166 7. Murphy L (2006) Biosensors and bioelectrochemistry. Curr Opin Chem Biol 10:177–184 8. Vo-Dinh T, Cullum B (2000) Biosensors and biochips: advances in biological and medical diagnostics. Fresenius J Anal Chem 366 :540–551 9. Wang J (2008) Electrochemical glucose biosensors. Chem Rev 108:814–826 10. Wilson GS, Gifford R (2005) Biosensors for real-time in vivo measurements. Biosens Bioelectron 20:2388–2403 11. Blawas AS, Reichert WM (1998) Protein patterning. Biomaterials 19:595–609

4


77

12. Ito Y (1999) Surface micropatterning to regulate cell functions. Biomaterials 20:2333–2342 13. Yap FL, Zhang Y (2007) Protein and cell micropatterning and its integration with micro/nanoparticles assembly. Biosens Bioelectron 22:775–788 14. Nagel DJ (2002) Technologies for micrometer and nanometer pattern and material transfer. In: Piqué A, Chrisey DB (eds) Direct-write technologies for rapid prototyping applications. Academic Press, San Diego 15. Falconet D, Csucs G, Grandin HM et al (2006) Surface engineering approaches to micropattern surfaces for cell-based assays. Biomaterials 27:3044–3063 16. Fink J, Thery M, Axioune A et al (2007) Comparative study and improvement of current cell micro-patterning techniques. Lab Chip 7:672–680 17. Csucs G, Michel R, Lussi JW et al (2000) Microcontact printing of novel co-polymers in combination with proteins for cell-biological applications. Biomaterials 24:1713–1720 18. Zhang S, Yan L, Altman M et al (1999) Biological surface engineering: a simple system for cell pattern formation. Biomaterials 20:1213–1220 19. Chrisey DB, Piqué A (2002) Introduction to direct-write technologies for rapid prototyping. In: Chrisey DB, Piqué A (eds) Direct-write technologies for rapid prototyping applications. Academic Press, San Diego 20. Roth EA, Xu T, Das M et al (2004) Inkjet printing for high-throughput cell patterning. Biomaterials 25:3707–3715 21. Turcu F, Tratsk-Nitz K, Thanos S et al (2003) Ink-jet printing for micropattern generation of laminin for neuronal adhesion. J Neurosci Methods 131:141–148 22. Barron JA, Young HD, Dlott DD et al (2005a) Printing of protein microarrays via a capillaryfree fluid jetting mechanism. Proteomics 5:4138–4144 23. Calvert P (2001) Inkjet printing for materials and devices. Chem Mat 13:3299–3305 24. Sirringhaus H, Shimoda T (2003) Inkjet printing of functional materials. MRS Bull 28: 802–807 25. Ito Y, Nogawa M (2003) Preparation of a protein micro-array using a photo-reactive polymer for a cell-adhesion assay. Biomaterials 24:3021–3026 26. Ringeisen BR, Othon CM, Barron JA et al (2006) Jet-based methods to print living cells. Biotechnol J 1:930–948 27. Brisbane AD (1971) Pattern deposit by laser. United States Patent 3,560,258 28. Bohandy J, Kim BF, Adrian FJ (1986) Metal deposition from a supported metal film using an excimer laser. J Appl Phys 60:1538–1539 29. Bohandy J, Kim BF, Adrian FJ et al (1988) Metal-deposition at 532 nm using a laser transfer technique. J Appl Phys 63:1158–1162 30. Schultze V, Wagner M (1991) Laser-induced forward transfer of aluminum. Appl Surf Sci 52:303–309 31. Kántor Z, Tóth Z, Szörényi T (1992) Laser induced forward transfer: the effect of supportfilm interface and film-to-substrate distance on transfer. Appl Phys A 54:170–175 32. Kántor Z, Szörényi T (1995) Dynamics of long-pulse laser transfer of micrometer-sized metal patterns as followed by time-resolved measurements of reflectivity and transmittance. J Appl Phys 78:2775–2781 33. Zergioti I, Mailis S, Vainos NA et al (1998a) Microdeposition of metals by femtosecond excimer laser. Appl Surf Sci 127–129:601–605 34. Sano T, Yamada H, Nakayama T et al (2002) Experimental investigation of laser induced forward transfer process of metal thin films. Appl Surf Sci 186:221–226 35. Willis DA, Grosu V (2005) Microdroplet deposition by laser-induced forward transfer. Appl Phys Lett 86:244103/1-8 36. Claeyssens F, Klini A, Mourka A et al (2007) Laser patterning of Zn for ZnO nanostructure growth: comparison between laser induced forward transfer in air and in vacuum. Thin Solid Films 515:8529–8533 37. Greer JA, Parker TE (1988) Laser-induced forward transfer of metal oxides to trim the frequency of surface acoustic wave resonator devices. SPIE Proc 998:113–125

78

P. Serra et al.

38. Papakonstantinou P, Vainos NA, Fotakis C (1999) Microfabrication by UV femtosecond laser ablation of Pt, Cr and indium oxide thin films. Appl Surf Sci 151:159–170 39. Zergioti I, Mailis S, Vainos NA et al (1998b) Microdeposition of metal and oxide structures using ultrashort laser pulses. Appl Phys A 66:579–582 40. Chakraborty S, Sakata H, Yokoyama E et al (2007) Laser-induced forward transfer technique for maskless patterning of amorphous V2O5 thin film. Appl Surf Sci 254:638–643 41. Fogarassy E, Fuchs C, Kergerve F et al (1989) Laser-induced forward transfer of high-Tc YBaCuO and BiSrCaCuO superconducting thin films. J Appl Phys 66:457–459 42. Adrian FJ, Bohandy J, Kim BF et al (1987) A study of the mechanism of metal-deposition by the laser-induced forward transfer process. J Vac Sci Technol B 5:1490–1494 43. Tamada H, Sano T, Nakayama T et al (2002) Optimization of laser-induced forward transfer process of metal thin films. Appl Surf Sci 197–198:411–415 44. Tolbert WA, Lee IYS, Doxtader MM et al (1993a) High-speed color imaging by laserablation transfer with a dynamic release layer- fundamental mechanisms. J Imaging Sci Tech 37:411–422 45. Tolbert WA, Lee IYS, Wen X et al (1993b) Laser-ablation transfer imaging using picosecond optical pulses – ultra-high-speed, lower threshold and high-resolution. J Imaging Sci Tech 37:485–489 46. Fardel R, Nagel M, Nüesch FA et al (2007a) Laser forward transfer using a sacrificial layer: influence of the material properties. Appl Surf Sci 254:1322–1326 47. Fardel R, Feurer P, Lippert T et al (2007b) Laser ablation of aryltriazene photopolymer films: effects of polymer structure on ablation properties. Appl Surf Sci 254:1332–1337 48. Doraiswamy A, Narayan RJ, Lippert T et al (2006) Excimer laser forward transfer of mammalian cells using a novel triazene absorbing layer. Appl Surf Sci 252:4743–4747 49. Chrisey DB, Piqué A, Mc-Gill RA et al (2003) Laser deposition of polymer and biomaterial films. Chem Rev 103:553–576 50. Wu PK, Ringeisen BR, Krizman DB et al (2003) Laser transfer of biomaterials: matrixassisted pulsed laser evaporation (MAPLE) and MAPLE direct write. Rev Sci Inst 74: 2546–2557 51. Zhang C, Liu D, Mathews SA et al (2003) Laser direct-write and its application in low temperature Co-fired ceramic (LTCC) technology. Microelectron Eng 70:41–49 52. Arnold CB, Serra P, Piqué A (2007) Laser direct-write techniques for printing of complex materials. MRS Bull 32:23–31 53. Piqué A, Fitz-Gerald J, Chrisey DB et al (2000) Direct writing of electronic materials using a new laser assisted transfer/annealing technique. Proc SPIE 3933:105–112 54. Fernández-Pradas JM, Colina M, Serra P et al (2004a) Laser-induced forward transfer of biomolecules. Thin Solid Films 453–454:27–30 55. Colina M, Serra P, Fernández-Pradas JM et al (2005) DNA deposition through laser induced forward transfer. Biosens Bioelectron 20:1638–1642 56. Fernández-Pradas JM, Serra P, Colina M et al (2004b) Production of biomolecule microarrays through laser induced forward transfer. Proc SPIE 5662:582–585 57. Fernández-Pradas JM, Duocastella M, Colina M et al (2007) Production of miniaturized biosensors through laser-induced forward transfer. Proc SPIE 6592:65920R1–8 58. Serra P, Colina M, Fernández-Pradas JM et al (2004b) Preparation of functional DNA microarrays through laser-induced forward transfer. Appl Phys Lett 85:1639–1641 59. Serra P, Fernández-Pradas JM, Colina M et al (2006) Laser-induced forward transfer: a direct-writing technique for biosensors preparation. JLMN 1:236–242 60. Zergioti I, Karaiskou A, Papazoglou DG et al (2005) Femtosecond laser microprinting of biomaterials. Appl Phys Lett 86:163902/1-3 61. Wu PK, Ringeisen BR, Callahan J et al (2001) The deposition, structure, pattern deposition, and activity of biomaterial thin-films by matrix-assisted pulsed-laser evaporation (MAPLE) and MAPLE direct write. Thin Solid Films 398–399:607–614 62. Ringeisen BR, Wu PK, Kim H et al (2002a) Picoliter-scale protein microarrays by laser direct write. Biotechnol Prog 18:1126–1129

4


79

63. Barron JA, Wu P, Ladouceur HD et al (2004a) Biological laser printing: a novel technique for creating heterogeneous 3-dimensional cell patterns. Biomed Microdevices 6:139–147 64. Duocastella M, Colina M, Fernández-Pradas JM et al (2007) Study of the laser-induced forward transfer of liquids for laser bioprinting. Appl Surf Sci 253:7855–7859 65. Serra P, Fernández-Pradas JM, Berthet FX et al (2004a) Laser direct writing of biomolecule microarrays. Appl Phys A 79:949–952 66. Duocastella M, Colina M, Fernández-Pradas JM et al (2008a) Laser-induced forward transfer of liquids for miniaturized biosensors preparation. JLMN 3:1–4 67. Duocastella M, Fernández-Pradas JM, Domínguez J et al (2008b) Printing biological solutions through laser-induced forward transfer. Appl Phys A 93:941–945 68. Dinca V, Farsari M, Kafetzopoulos D et al (2008b) Patterning parameters for biomolecules microarrays constructed with nanosecond and femtosecond UV lasers. Thin Solid Films 516:6504–6511 69. Dinca V, Ranella A, Popescu A et al (2007b) Parameters optimization for biological molecules patterning using 248-nm ultrafast lasers. Appl Surf Sci 254:1154–1158 70. Dinca V, Kasotakis E, Mourka A et al (2008c) Fabrication of amyloid peptide micro-arrays using laser-induced forward transfer and avidin-biotin mediated assembly. Phys Stat Sol C 5:3576–3579 71. Dinu CZ, Dinca V, Howard J et al (2007) Printing technologies for fabrication of bioactive and regular microarrays of streptavidin. Appl Surf Sci 253:8119 72. Dinca V, Kasotakis E, Catherine J et al (2007a) Development of peptide-based patterns by laser transfer. Appl Surf Sci 254:1160–1163 73. Dinca V, Ranilla A, Farsari M et al (2008a) Quantification of the activity of biomolecules in microarrays obtained by direct laser transfer. Biomed Microdevices 10:719–725 74. Boutopoulos C, Andreakou P, Kafetzopoulos D et al (2008) Direct laser printing of biotin microarrays on low temperature oxide on Si substrates. Phys Stat Sol A 205:2505–2508 75. Barron JA, Ringeisen BR, Kim H et al (2004c) Application of laser printing to mammalian cells. Thin Solid Films 453–454:383–387 76. Barron JA, Spargo BJ, Ringeisen BR (2004d) Biological laser printing of three dimensional cellular structures. Appl Phys A 79:1027–1030 77. Barron JA, Krizman DB, Ringeisen BR (2005b) Laser printing of single cells: statistical analysis, cell viability, and stress. Ann Biomed Eng 33:121–130 78. Hopp B, Smausz T, Kresz N et al (2005) Survival and proliferative ability of various living cell types after laser-induced forward transfer. Tissue Eng 11:1817–1823 79. Chen CY, Barron JA, Ringeisen BR (2006) Cell patterning without chemical surface modification: cell-cell interactions between printed bovine aortic endothelial cells (BAEC) on a homogeneous cell-adherent hydrogel. Appl Surf Sci 252:8641–8645 80. Kattamis NT, Purnick PE, Weiss R et al (2007) Thick film laser induced forward transfer for deposition of thermally and mechanically sensitive materials. Appl Phys Lett 91:171120/1-3 81. Schiele NR, Koppes RA, Corr DT et al (2008) Laser direct writing of combinatorial libraries of idealized cellular constructs: biomedical applications. Appl Surf Sci 255:5444–5447 82. Othon CM, Wu X, Anders JJ et al (2008) Single-cell printing to form three-dimensional lines of olfactory ensheathing cells. Biomed Mater 3:034101/1-6 83. Ringeisen BR, Chrisey DB, Piqué A et al (2002b) Generation of mesoscopic patterns of viable Escherichia coli by ambient laser transfer. Biomaterials 23:161–166 84. Barron JA, Rosen R, Jones-Meehan J et al (2004b) Biological laser printing of genetically modified Escherichia coli for biosensor applications. Biosens Bioelectron 20:246–252 85. Hopp B, Smausz T, Antal Z et al (2004) Absorbing film assisted laser induced forward transfer of fungi (Trichoderma conidia). J Appl Phys 96:3478–3488. 86. Smausz T, Hopp B, Kecskeméti G et al (2006) Study on metal microparticle content of the material transferred with absorbing film assisted laser induced forward transfer when using silver absorbing layer. Appl Surf Sci 252:4738–4742 87. Kattamis NT, McDaniel ND, Bernhard S et al (2009) Laser direct write printing of sensitive and robust light emitting organic molecules. Appl Phys Lett 94:103306/1-3

80

P. Serra et al.

88. Colina M, Duocastella M, Fernández-Pradas JM et al (2006) Laser-induced forward transfer of liquids: study of the droplet ejection process. J Appl Phys 99:084909/1-7 89. Bickerstaff GF (1997) Immobilization of enzimes and cells: some practical considerations. In: Bickerstaff GF (ed) Immobilization of enzimes and cells. Humana Press Inc, Tottowa 90. Heise C, Bier FF (2006) Immobilization of DNA on microarrays. In: Wittmann C (ed) Immobilization of DNA on chips II. Springer, Berlin 91. Serra P, Duocastella M, Fernández-Pradas JM et al (2009) Liquids microprinting through laser-induced forward transfer. App Surf Sci 255:5342–5345 92. Asai A, Hara T, Endo I (1987) One-dimensional model of bubble growth and liquid flow in bubble jet printers. Jpn J Appl Phys 26:1794–1801 93. Reis N, Ainsley CA, Derby B (2005) Ink-jet delivery of particle suspensions by piezoelectric droplet ejectors. J Appl Phys 97:094903/1-5 94. Nakata Y, Okada T (1999) Time-resolved microscopic imaging of the laser-induced forward transfer process. Appl Phys A 69:S275–S278 95. Papazoglou DG, Karaiskou A, Zergioti I et al (2002) Shadowgraphic imaging of the sub-ps laser-induced forward transfer process. Appl Phys Lett 81:1594–1596 96. Young D, Auyeung RCY, Piqué A et al (2001) Time-resolved optical microscopy of a laserbased forward transfer process. Appl Phys Lett 78:3169–3171 97. Young D, Auyeung RCY, Piqué A et al (2002) Plume and jetting regimes in a laser based forward transfer process as observed by time-resolved optical microscopy. Appl Surf Sci 197–198:181–187 98. Duocastella M, Fernández-Pradas JM, Serra P et al (2008c) Jet formation in the laser forward transfer of liquids. Appl Phys A 93:453–456 99. Blake JR, Gibson DC (1981) Growth and collapse of a vapor cavity near a free surface. J Fluid Mech 111:123–140 100. Brujan EA, Nahen K, Schmidt P et al (2001) Dynamics of laser-induced cavitation bubbles near an elastic boundary. J Fluid Mech 433:251–281 101. Karaiskou A, Zergioti I, Fotakis C et al (2003) Microfabrication of biomaterials by the sub-ps laser-induced forward transfer process. Appl Surf Sci 208–209:245–249

Chapter 5

Biological Laser Printing (BioLP) for High Resolution Cell Deposition Bradley R. Ringeisen, C.M. Othon, Xingjia Wu, D.B. Krizman, M.M. Darfler, J.J. Anders, and P.K. Wu

Abstract Biological laser printing, or BioLP, is a modified laser induced forward transfer (LIFT) technique that has the demonstrated ability to print cells from living cultures and paraffin-embedded fixed tissue sections. Detailed studies have been published that demonstrate the energy conversion layer used by BioLP to absorb incident laser energy and promote forward transfer of biological materials prevents damage to the printed cells. Additionally, this layer helps maintain reproducible transfer conditions so that large arrays and patterns of cells can be precisely generated down to the single cell level. BioLP’s nozzle-free print head allows both liquids and solids to be printed with micron-scale resolution and is unique to modified LIFT technologies. This chapter will describe recent experiments that applied BioLP to two distinct applications of cell printing: regenerative medicine and tissue microdissection. Specifically, three dimensional patterns of olfactory ensheathing cells (OECs) were created and the application of these scaffolds to spinal cord repair will be discussed. High resolution stem and branch patterns of human umbilical vein endothelial cells (HUVECs) have also been created by BioLP. These patterns show controlled differentiation and lumen formation along the printed pattern. Co-culture printing experiments were also performed where the natural vascular structure of endothelial and smooth muscle cell contacts are explored as well as interactions between OECs and rat cortical neurons. We will also discuss the use of BioLP as a tissue microdissection tool. We show examples of prostate tissue dissection where single malignant cells are removed from a prepared tissue section.

5.1 Introduction Ten years ago the concept of printing patterns of living cells was just beginning to be explored experimentally. There were a few direct write techniques being used for micron-scale pattern formation including modified laser induced forward transfer B.R. Ringeisen (B) Department of Chemistry, Naval Research Laboratory, Washington, DC 20375-5342, USA e-mail: [email protected]


81

82

B.R. Ringeisen et al.

(LIFT) techniques [1], laser guidance techniques [2], and extrusion pens [3]. Most of the experiments using these techniques focused on printing non-biological inks such as conductive and resistive pastes or metal films. As these techniques progressed for electronics applications, it seemed logical to pursue other research areas including biological printing [4]. The first technique to attempt live cell depositions was laser guidance, which was used to print embryonic chick spinal cord cells oneby-one down onto a glass slide [5]. This ground-breaking experiment was the first direct write deposition of live cells and provided proof that cells could be organized directly onto a surface without lithographic surface chemical modification to promote selective adhesion. Shortly after these initial experiments, many other techniques followed suit and performed live cell printing [6–10]. Due to its gentle and nozzle-free printing mechanism, modified LIFT techniques began to attempt deposition of biological materials. The first to do so was a technique referred to as matrix assisted pulsed laser evaporation direct write, or MAPLE DW, that printed live E. coli bacteria and mouse pluripotent embryonal carcinoma cells in separate experiments [9, 11]. This technique was later refined specifically for biological deposition and is now referred to as biological laser printing, or BioLP. BioLP has demonstrated the ability to print living cells with near 100% viability and no phenotypic damage [12–14]. Cells have been printed into three dimensional patterns and at the single cell level. Multiple cell types have also been printed in adjacent patterns where cell-cell communication and/or cell migration can be studied. This chapter will detail recent cell printing experiments performed with BioLP. We will first discuss the use of olfactory ensheathing cells (OECs) to create three dimensional scaffolds for spinal cord therapy. We will then describe experiments using human umbilical vein endothelial cells (HUVECs) and smooth muscle cells (HUVSMCs) to mimic the natural mammalian vascular structure in printed scaffolds. Finally, we will discuss how BioLP, due to its unique orifice-free design, can be used for rapid and precise tissue microdissection of paraffin-embedded tissue sections. The versatility of BioLP to perform such diverse biological deposition experiments lends promise for the tool to be used in a wide variety of scientific and commercial applications.

5.2 Experimental Method 5.2.1 Cell Culture and Bioinks OECs and rat cortical neurons were harvested and purified following a previously published technique with some modifications [15]. OECs were centrifuged and re-suspended in a complete medium (DMEM-F12 supplemented with 10% fetal bovine serum, 1% glutamine, 2% penicillin-streptomycin and 1% gentamicin) to yield a suspension containing 1 × 105 cells/ml. The cell suspension was seeded into an uncoated flask for 24 h. The supernatant was transferred into an uncoated

5

Biological Laser Printing (BioLP) for High Resolution Cell Deposition

83

flask for 48 h (37◦ C and 5% CO2 ). During this period, all other cell types in the cell suspension attach to the uncoated plastic flask, while the OECs remain in the supernatant. Using this purification process, the OEC suspension was ~93% pure. One adult rat yielded approximately 200,000 OECs. The OECs were cultured in suspension (up to 2 days) until printing. HUVEC were purchased from Allcells, LLC (Emeryville, California) and cultured according to the protocol provided. HUSMC were purchased from ATCC and cultured according to the protocol provided by the supplier. In both cases, per ml of media is fortified with 1–2 μg of Fungizone (Invitrogen) and 0.01 mL penicillinstreptomycin (Gibco). Cells used for printing were in their 2–4 passages (supplier states that up to 6 passages are possible prior to losing functionality). They were harvested by first detaching the cells from the bottom of the culturing flask using Trypson-EDTA. The procedure is described in the first part of the subculturing protocol provided by the cell vender. The cell solution is then centrifuged, the supernatant removed, and the pellet resuspended in media resulting in a cell density range of 105 –107 cells/mL. This cell suspension, the cell bioink, is directly printed without additives. To print cells by BioLP, cell ink was spread evenly across the surface of the target on top of the TiO2 film (~5 μL/cm2 ). The thickness of the layer of cell solution is ~10–50 μm, depending on the ink used and the application technique. Substrates are glass microscope slides coated with 100–200 μm thick MatrigelTM (BD Bioscience). Printed cells were allowed to acclimate in the MatrigelTM for up to 5 min before media was added. Printed droplets were approximately 50 μm in diameter and were deposited 50–150 μm from each other along defined structures. The laser control parameters and the concentration of the cell ink can be changed to control how many cells are printed per droplet, ranging from one cell up to as many as tens of cells.

5.2.2 BioLP Apparatus As shown in Fig. 5.1, BioLP uses laser pulses (MPB Technologies PSX-100 Excimer Laser, 248 nm, 2.5 ns FWHM, rep rate=0.1–100 Hz, Emax =5 mJ) focused onto a support to print small aliquots of cell solution to a receiving substrate [12]. The target is an optically transparent quartz plate coated on one side with a thin (10–100 nm) layer of TiO2 that acts as a laser absorption layer. The opposing side of the metal layer is coated with a 10–100 μm thick layer of cell ink and media which makes up the bioink (Section 5.2.1 above). When a UV laser pulse is focused at the target-absorption layer interface, the metal layer absorbs nearly all the incident laser energy, converting the photon energy into thermal and/or mechanical energy. This converted energy results in a droplet of bioink being pushed away from the target towards the substrate. Studies have shown that the printed droplet undergoes little to no heating and is transferred in the same phase as it existed on the target support [16].

84


Fig. 5.1 Schematic of the BioLP apparatus for printing cells

Previous studies used Comet assays to determine that even when greater than 90% of the incident laser energy penetrates through the target to the cells, the celllaser interaction does not induce significant single or double strand breaks in the intracellular DNA [11]. Adjusting the laser pulse energy and spot diameter controls the diameter (30–300 μm) and volume (hundreds of fL to several nL) of the deposited aliquot [16]. The current rate of transfer is variable up to 100 spots per second, but could be upwards to thousands of spots per second with faster repetition rate lasers. The BioLP resolution is derived from the spot diameter of the focused laser, thereby overcoming some of the limitations of ink-jet and micropen cell printing approaches, namely the clogging (viscosity, cell agglomeration, etc.) of print heads or capillaries used by these techniques to achieve micron-scale resolution.

5.2.3 Tissue Prepration For tissue microdissection, 10 μm thick tissue sections were cut from a fomalin fixed paraffin embedded (FFPE) whole-mount prostate tissue block, placed on the coated slides, and heated for 60 min at 58◦ C. Paraffin was removed by treatment in SubX organic solvent (Surgipath Medical Industries, Richmond, IL) twice for 5 min; this was followed by tissue rehydration through multiple, graded ethanol solutions and distilled water. Tissue was counterstained with Mayer’s hematoxylin, dehydrated through graded ethanol solutions, and air-dried. Tissue was rehydrated prior to microdissection with 50% glycerol in water for 5 min.

5


85

Slides were placed upside down below the 10X objective of the BioLP apparatus and visualized to locate cellular regions with specific histological features. By scanning the laser and receiving substrate via computer-aided design/computeraided machining (CAD/CAM), microdissection was automatically performed by directing laser pulses over the previously defined and mapped cellular regions to achieve complete transfer of cells within the selected area into a 1.5-ml low binding microcentrifuge receiving tube.

5.2.4 Fluorescent Imaging Live/Dead Assay: Cells were tagged with a live/dead viability/cytotoxicity assay from Molecular Probes (L-3224). Cell viability was assessed with 16 μM calcein AM solution and cellular damage was found via use of a 32 μM ethidium homodimer-1 (EthD-1) solution. The concentration of the assay was adjusted to optimize imaging within the 3-D basement membrane scaffold. MatrigelTM scaffolds were rinsed with 1X Phosphate Buffer Solution for 10 min to remove excess phenol red. Approximately 50 μL of the live/dead assay was pipetted onto the surface of the scaffold. The scaffold was returned to the incubator for 20 min to enhance uptake of the assay solution by the cells. R 488 carboxylic acid diacetate, succinimidyl ester Cell Labeling: Oregon Green (Invitrogen Corp., Carlsbad, CA) was used for long-term cell tracking. Cells were suspended in phosphate-buffered saline (PBS) with 10 μM Oregon Green and incubated for 30 min at room temperature with continuous shaking. Cell Tracker Orange was used to label rat cortical neurons. Labeled cells were centrifuged, rinsed once with PBS, centrifuged again and then resuspended in complete medium.

5.3 Live Cell Printing 5.3.1 Printing Olfactory Ensheathing Cells (OECs) and Rat Cortical Neurons Droplets of OEC and rat coritical neuron bioinks were printed to separate 75 μm thick layers of Matrigel coated onto glass slides. The Matrigel serves several purposes during the cell pritning process. First, it acts as a cushion to less the cell impact on the receiving substrate. Secondly, it hydrates the cells and prevents the extremely small volume droplet from evaporating on contact with the receiving substrate. Lastly, it has a broad range of growth factors, including collagen and lamanin, that help the printed cells adhere to the recieving substate and begin differentiation. Figure 5.2 shows two fluorescent micrographs. Panel (a) is a live/dead assay (green = live, orange = dead) of printed OECs that were incubated for 72 h prior to staining. This image demonstrates that not only are the printed OECs viable,

86


Fig. 5.2 Live dead stain of (a) rat OECs and (b) rat cortical neurons into Matrigel layers by BioLP. OECs show characteristic psuedopod extension while neurons show neurite outgrowth. Grayscale image shows fluorescence from a live/dead stain (green = live, orange = dead). The color image shows all green, indicating all cells in the image were viable (for colors see online version)

they also retained their ability to differentiate and perform cell-cell communication. OECs have the ability to communicate over long distances [17], resulting in extended psueodopods up to hundreds of microns long that act to guide neurite growth in vivo. Figure 5.2a shows one such OEC extension that stretches for over 500 μm in the direction of another printed OEC. Figure 5.2b shows printed rat cortical neurons on a separate layer of Matrigel. The printed neurons show 100% viabiilty and have retained the ability to extend neurite processes. By printing different concentrations of neurons to the Matrigel layer, we found that at low printed cell densities, neuron growth is limited and neurite extension is eliminated (data not shown). At higher printed cell densities, we found that neuron growth and neurite extension were encouraged. We hypothesize that at higher cell densities, cell signaling molecules and growth factors were more concentrated, enabling enhanced growth and neurite extension. Cell printing enables different cell densities to be printed to homogeneous ‘pro-growth’ surfaces with high levels of growth factors. Without cell printing, the only method of controlling cell density on surfaces involves patterning surfaces with pro-growth and anti-growth molecules. These lithographically patterned inhomogeneous surfaces do not allow cell-cell communication and differentiated growth in the same manner that is observed on a homogeneous pro-growth surface such as Matrigel. Therefore, cell printing allows true cell-cell interactions to be observed rather than observing cell-surface interactions, as is often the case with lithographically patterned surfaces.

5.3.2 Co-culture Printing of OECs and Rat Cortical Neurons Recent studies have shown that injected OECs promote axon re-growth and functional repair in the injured spinal cord of rats [18, 19]. BioLP has been used to create 3D scaffolds of OECs that were implanted into the spinal cord of rats [14]. These

5


87

scaffolds were created by repeatedly printing lines of OECs in a layer-by-layer fashion until a near-mm thick hydrogel scaffold was built. The hope was that the lines of OECs would better guide axon regeneration and promote healing of the spinal cord injuries. Even though these experiments represented the first implantation of a cell printed scaffold, the results showed no improvement over randomly injected OECs. In addition to creating implantable scaffolds, BioLP can also be used to learn more about the interactions between OECs and neurons via in vitro studies. Figure 5.3 shows six micrographs of an OEC/neuron co-culture created by printing parallel lines of both cell types onto a hydrogel layer. This figure demonstrates BioLP’s high resolution cell printing capabilities, creating several millimeter long parallel lines of two cell types while maintaining single cell resolution along the length of the lines. Panel (a) shows the phase contrast image of two parallel cell lines immediately after printing. The larger circles in panel (a) are indentations in the hydrogel due to the impact of the droplet with the surface. As is demonstrated by the green fluorescence shown in panel (b), several OECs make up the line on the right hand side of the image. Panel (c) shows orange fluorescence emitted from several neurons that were printed in a line on the left hand side of the image. The fluorescent tags used in these studies are carried in the live cells, so we were able to monitor the growth and interaction of both cell types with time. Figure 5.3 (d–e) are phase contrast and fluorescent images of the same hydrogel layer after 48 h of incubation. We find that one OEC has stretched out toward the line of neurons. We expected to find the OECs interacting more strongly with each other rather than with the parallel line of neurons, as is demonstrated in

Fig. 5.3 (a) Phase contrast micrograph of co-culture printed rat cortical cells and OECs immediately after printing. (b) UV fluorescent micrograph showing Oregon green stained OECs. Grayscale image is shown, but bright fluorescence indicates OECs in panels (b) and (f) and neurons in panels (c) and (f). (c) UV fluorescent micrograph showing neurons with Cell Tracker orange. (d–f) Same culture imaged after 48 h with an OEC stretching out towards the printed neurons (for colors see online version)

88


Fig. 5.2a. However, we found that the OECs repeatedly grew towards the neurons. Unfortunately, the density of neurons is too low to see significant neurite outgrowth. Future studies need to be performed that utilize more growth factors or higher cell densities to achieve sufficient neurite outgrowth to observe further neuron/OEC interaction. The experiment shown in Fig. 5.3 does highlight the unique ability of cell printing to deposit patterns of different cell types on a pro-growth surface. Matrigel enables the cells to grow in any and all directions, so the cells patterned in close proximity can signal and grow with each other free of cell-surface interactions. In the experiment shown in Fig. 5.3, the OECs appear to growth preferentially towards the neurons, perhaps sensing the presence of those cells and altering their migration and/or growth due to the presence of the co-culture.

5.3.3 Human Umbilical Vein Endothelial Cells (HUVECs) Endothelial cells (ECs) are a good model cell for studying angiogenesis, or the creation of vascular networks. When seeded into 3D hydrogel scaffolds, these cells have the ability to wrap around and extend towards each other, forming interconnected networks of lumens [20, 21]. The cell clusters, stretched cells, and cell chains form a polygonal network that span the hydrogel matrix. The size of the clusters and cell chains are of similar size and the network is isotropic. This outcome should be expected because the hydrogel provides no symmetry breaking features and the cells are seeded without a preset pattern. High resolution cell printing could be used to create ordered networks of ECs, utilizing the ability to pattern cells on hydrogels to create basic stem and branch structures of vessels (100 μm’s to mm-scale) and microcapillaries (2 mm long) composed of printed HUVECs with several branches (~1 mm) extending away from the main stem. The micrograph was taken with a phase contrast microscope 24 h after printing was performed. The image shows that the HUVECs have stretched out and connected with the neighboring deposited cells. The deposited clusters of HUVECs (from each printed droplet, approximately 50 μm between each droplet) have connected autonomously to form complete lumina. At the crutch of the branch/stem when the distance between the stem and

5


89

Fig. 5.4 Phase contrast micrograph of a printed branch/stem structure of HUVECs in Matrigel

branch is shortest, the HUVEC stretched and connected the two parts forming secondary lumina. A second printed stem and branch HUVEC structure is shown in Fig. 5.5. This image is shown at higher magnification so that the individual clusters of printed cells can be seen more clearly. Additionally, Fig. 5.5 more clearly shows the innate ability of HUVECs to communicate with one another, linking up the printed stem and branch to form a complete network. The stem and branch structure grew from the printed pattern of cells, but the secondary lumina (connecting the stem and branch) are formed by the HUVEC’s inherent growth tendency to connect with other HUVEC. This structure is potentially the initial formation of a lumen network. Another interesting observation from Figs. 5.4 and 5.5 is that growth of secondary lumen from a well developed stem or branch lumina tends to be normal to the developed lumen. Most of the secondary lumina shown in these figures are normal to the stem or branch they developed from.

5.3.4 Co-culture Printing of HUVECs and Human Umbilical Vein Smooth Muscle Cells (HUVSMCs) In natural tissue, the HUVECs are the lining of the umbilical veins and they are encased in HUVSMCs. The two types of cells must provide a symbiotic relationship with each other to maintain the structure of a vein. From the results shown in Figs. 5.4 and 5.5, we know that lumina can be formed from printed HUVEC after

90


Fig. 5.5 Magnified view of lumen formation by HUVECs along the printed pattern

1 day of differentiation. However, these interconnected structures deconstruct after several days in culture, leaving cells that look more similar to 2D cultures than 3D (i.e., lumen formation degrades and cells begin proliferating rather than extended in a defined pattern/lumen). We wanted to explore whether the introduction of HUVSMCs altered the long term growth characteristics of the HUVEC branch/stem structures. Therefore, we deposited the same branch/stem pattern using a HUVSMC bioink. This smooth muscle cell pattern was deposited directly on top of a HUVEC branch/stem structure that was printed 24 h earlier. Figure 5.6 shows the result of the result of printing HUVSMCs on top of HUVECs 9 days post printing. Contrary to when HUVSMCs are seeded alone into a

Fig. 5.6 Co-culture printing of HUVECs (line) and HUVSMCs (stretched cell along the line)

5


91

hydrogel (contractile state), the HUVSMCs that are printed in close proximity to the HUVECs are stretched out in an elongated, proliferative state. One such HUVSMC is shown in Fig. 5.6, highlighted by an arrow and text. This HUVSMC is elongated in the direction of the lumen-connected HUVEC pattern, indicating that the growth, differentiation and directionality of the smooth muscle cell are influenced by the presence of the HUVEC structure. Because it is impossible to unambiguously distinguish HUVSMCs from HUVECs without immunocytochemical staining, we have no information as to the relative concentration of the two cells types. However, the shape of the two cell types are distinct, especially when the HUVSMCs are in the proliferative state (stretched long), as is shown in this co-culture printing experiment. Further evidence (data not shown), indicate that the presence of the HUVSMC upholds the integrity of the HUVECs branch/stem structure, enabling the lumen structure to be maintained for much longer times than when the ECs are patterned alone [22].

5.4 Tissue Microdissection Due to the nozzle-free nature of the BioLP mechanism, nearly all types of biological materials can be printed, including solid-phase tissue sections. Standard histopathological tissue sections are usually 5–10 μms thick, making it possible for a laser pulse to directly extract cells from the tissue section. Various staining protocols are used to visually (optically) distinguish between different cell types. The optical nature of this differentiation is one reason why BioLP is perfectly suited to perform tissue microdissection. During cell printing, BioLP uses a camera to continuously monitor the bioink on the target support. For tissue microdissection, this camera can be used to locate the cells to be dissected (removed) from the tissue section. BioLP was used to remove malignant cells from a stained tissue section of prostate tissue. In the example shown in Fig. 5.7, a 2 × 2 mm swatch of tissue was completely removed (panel (b)). At higher magnification, there is evidence that the energy conversion layer (laser absorption layer) has been removed to initiate the transfer of solid-phase tissue. The spot size of the laser was approximately 20 μms, enabling near single-cell resolution for microdissection. Further automation and computer-control is being developed for tissue microdissection applications by Expression Pathology, Inc. (Rockville, MD), where computer recognition software is being used to discern between stroma and malignant cells. Additionally, computer software is also being used to enable real-time marking of cells for single cell isolation and removal. Past studies have also shown that cells removed via BioLP can be further analyzed for protein content [23]. Therefore, the laser processing does not damage the molecular components of the dissected cells. BioLP is a useful tool that goes well beyond traditional ink jet or other nozzlebased micropen cell printers in its ability to deposit non-traditional biological matrices including solid-phase materials.

92


Fig. 5.7 An example of human prostate tissue microdissection by BioLP. Images from left to right: stained tissue sample, 2 mm square microdissection, and magnified view of laser tissue dissection

5.5 Summary BioLP is a versatile cell printer that is capable of reaching the ultimate limit by positioning single cells in defined patterns onto pro-growth surfaces or removing single cells from solid-phase matrices. This chapter summarized recent high resolution experiments where several relevant cell types for tissue engineering were used to create unique biomimetic patterns in hydorgel layers. Single cell lines of OECs were printed in order to study cell-cell communication, growth and migration under controlled co-culture environments. Additionally, HUVECs and HUVSMCs were patterned into branch/stem structures to study cell-cell interaction and growth behavior in human vascular formation. Cell printing and BioLP in particular have the ability to precisely place defined numbers of different cells in close proximity to one another in pro-growth, 3D materials such as hydrogels with high growth factor concentrations. The experiments laid out here may lead to the ability to print thick scaffolds with pre-laid living fluidic networks. Networks such as these could enable nutrient delivery and waste effluent for thick engineered tissue or organs. Finally, due to BioLP’s unique flat printhead, we showed that solid tissue sections can be microdissected at the single cell level. We believe BioLP holds a unique place in cell printing due to its ability to print from both liquid and solid inks while placing cells at both high density and with single cell precision.

References 1. Chrisey D, Pique A, McGill R et al (2003) Laser deposition of polymer and biomaterial films. Chem Rev 103:553–576 2. Renn MJ, Chrisey DB, Gamota DR, Helvajian H, Taylor DP (eds) (2000) Direct-write technologies for rapid prototyping applications. Materials Research Society, San Francisco, pp 107–114 3. King BH, Dimos D, Yang P et al (1999) Direct-write fabrication of integrated, multilayer ceramic components. J Electroceram 3:173–178

5


93

4. Odde DJ, Renn MJ (1999) Laser-guided direct writing for applications in biotechnology. Trends Biotechnol 17:385–389 5. Odde DJ, Renn MJ (2000) Laser-guided direct writing of living cells. Biotechnol Bioeng 67:312–318 6. Mironov V, Boland T, Trusk T et al (2003) Organ printing: computer-aided jet-based 3d tissue engineering. Trends Biotechnol 21:157–161 7. Roth EA, Xu T, Das M et al (2004) Ink-jet printing for high-throughput cell patterning. Biomaterials 25:3707–3715 8. Wilson WC Jr, Boland T (2003) Cell and organ printing 1: protein and cell printers. Anat Rec A Discov Mol Cell Evol Biol 272:491–496 9. Ringeisen BR, Chrisey DB, Pique A et al (2002) Generation of mesoscopic patterns of viable escherichia coli by ambient laser transfer. Biomaterials 23:161–166 10. Ringeisen B, Othon C, Barron J et al (2006) Jet-based methods to print living cells. Biotechnol J 1:930–948 11. Ringeisen BR, Kim H, Barron JA et al (2004) Laser printing of pluripotent embryonal carcinoma cells. Tissue Eng 10:483–491 12. Barron JA, Wu P, Ladouceur HD et al (2004) Biological laser printing: a novel technique for creating heterogeneous 3-dimensional cell patterns. Biomed Microdevices 6:139–147 13. Chen CY, Barron JA, Ringeisen BR (2006) Cell patterning without chemical surface modification: cell-cell interacftions between bovine aortic endothelial cells (baec) on a homogeneous cell-adherent hydrogel. Appl Surf Sci 252:8641–8645 14. Othon CM, Wu X, Anders JJ et al (2008) Single-cell printing to form three-dimensional lines of olfactory ensheathing cells. Biomed Mater 3:034101 15. Nash HH, Borke RC, Anders JJ (2001) New method of purification for establishing primary cultures of ensheathing cells fom the adult olfactory bulb. Glia 34:81–87 16. Barron J, Young H, Dlott D et al (2005) Printing of protein microarrays via a capillary-free fluid jetting mechanism. Proteomics 5:4138–4144 17. Kocsis JD, Lankford KL, Sasaki M et al (2009) Unique in vivo properties of olfactory ensheathing cells that may contribute to neural repair and protection following spinal cord injury. Neurosci Lett 456:137–142 18. Chen H, Zheng X, Sheng W et al (2009) Transplantation of low-power laser-irradiated olfactory ensheathing cells to promote repair of spinal cord injury in rats. Neural Regen Res 4:171–177 19. Teng X, Nagata I, Li HP et al (2008) Regeneration of nigrostriatal dopaminergic axons after transplantation of olfactory ensheathing cells and fibroblasts prevents fibrotic scar formation at the lesion site. J Neurosci Res 86:3140–3150 20. Vernon RB, Sage EH (1995) Between molecules and morphology: extracellular matrix and creation of vascular form. Am J Pathol 147:873–883 21. Cascone I, Giraudo E, Caccavari F et al (2003) Temporal and spatial modulation of rho gtpases during in vitro formation of capillary vascular network: adherens junctions and myosin light chain as targets of rac1 and rhoa. J Biol Chem 278:50702–50713 22. Wu PK, Ringeisen BR (2010) BioLP printing and development of HUVEC and HUVSMC branch/stem structure on hydrogel layers. Biofabrication 2:014111 23. Hood BL, Darfler MM, Guiel TG et al (2005) Proteomic analysis of formalin-fixed prostate cancer tissue. Mol Cell Proteomics 4:1741–1753

Chapter 6

High-Throughput Biological Laser Printing: Droplet Ejection Mechanism, Integration of a Dedicated Workstation, and Bioprinting of Cells and Biomaterials Fabien Guillemot, Bertrand Guillotin, Sylvain Catros, Agnès Souquet, Candice Mezel, Virginie Keriquel, Ludovic Hallo, Jean-Christophe Fricain, and Joëlle Amedee

Abstract High-Throughput Biological Laser Printing (HT BioLP) requires taking into account spatio-temporal proximity of laser pulses (that means pulse-to-pulse distance and laser pulse frequency). The droplet ejection mechanism is indeed governed by vapor bubble dynamics (bubble growth and collapsing) and it is thus related to both the condition of laser irradiation and the rheological properties of the liquid film (viscosity, surface tension). We present a rapid prototyping workstation which has been designed for HT BioLP applications. It is equipped with an infra-red pulsed laser (pulse duration = 30 ns, wavelength = 1,064 nm, f = 1–100 kHz), galvanometric mirrors (scanning speed up to 2,000 mm/s), micrometric translation stages (x, y, z) and a dedicated software. Then, after describing experimental conditions leading to the high resolution printing (including cell density, laser parameters, etc.) of biological components, we present some typical multi-component and 3D printings achieved using this workstation. Finally, considering different criteria (speed, inoquity, etc.) the potentiality of HT BioLP is discussed as an alternative technology in Tissue Engineering applications.

6.1 Introduction In parallel to scaffold-based approaches involving cell seeding of porous structures [1], some authors have suggested three-dimensional biological structures can be built from the bottom up by the technology of bioprinting: the automated, computeraided printing of cells, cell aggregates and biomaterials [2–4]. Towards this end, commercially available ink-jet printers have been successfully redesigned [5] or new ones built [6, 7] to pattern biological assemblies according to a computer-aided F. Guillemot (B) INSERM, U577, Bordeaux, F-33076 France; University of Victor Segalen Bordeaux 2, Bordeaux, F-33076 France e-mail: [email protected] B.R. Ringeisen et al. (eds.), Cell and Organ Printing, C Springer Science+Business Media B.V. 2010 DOI 10.1007/978-90-481-9145-1_6,

95

96

F. Guillemot et al.

design template. Pressure-operated mechanical extruders have been also developed to handle living cells and cell aggregates [3]. Parallel to these methods, laser-assisted printing technologies have emerged as alternative methods for the assembly and micropatterning of biomaterials and cells. Laser-guided direct writing (LGDW) is a technique capable of trapping multiple cells in a laser beam and deposit them as a steady stream on arbitrary non-absorbing surfaces [8, 9]. Biological laser printing is based on the laser-induced forward-transfer (LIFT) technique in which a pulsed laser is used to induce the transfer of material from a source film [10]. If the material is not embedded in a laser light-absorbing matrix, a thin sacrificial absorbing layer is necessary. Accordingly, processes have been called Matrix Assisted Pulsed Laser Evaporation – Direct Write (MAPLE-DW) [11] or Absorbing Film Assisted – LIFT (AFA-LIFT) [12], respectively. Since BioLP has proven more effective with the aid of the intermediate light-absorbing layer, MAPLE-DW has been abandoned. Thus, under suitable irradiation conditions, and for liquids presenting a wide range of rheologies, the material can be deposited in the form of well-defined circular droplets with a high degree of spatial resolution [13, 14].

6.2 Droplet Ejection Mechanism 6.2.1 Principle A typical LIFT experimental set up is generally composed of three elements: a pulsed laser source, a target coated with the material to be printed, and a receiving substrate (see Fig. 6.1). The target is a three layer component: a support, which is transparent to the laser radiation wavelength, coated with a thin absorbing layer, coated itself with a transfer layer, named bioink, that contains the elements to be printed (biomaterials, cells, biomolecules, etc.). As described elsewhere [10, 15, 16], the generation of microdroplets by LIFT proceeds through 6 consecutive steps: (i) (ii) (iii) (iv) (v) (vi)

The laser energy is first deposited into the skin depth of the absorbing layer; The absorbing layer is then heated in its skin depth; inducing heating of a very thin film of the bioink near the absorbing layer; a vapor bubble is generated depending on the laser irradiation intensity; the vapor bubble expands, and; triggers the bioink – air interface deformation.

It has been shown that depending on the laser energy, three LIFT regimes could be isolated: the sub-threshold, the jetting and the plume regimes [10]. In addition, it was found that the volume of deposited material depends linearly on the laser pulse energy, and that a minimum threshold energy has to be overcome for droplet ejection to occur [15]. As described in Fig. 6.1, microdroplet ejection is dependent on multiple parameters. In this chapter, the main ones will be discussed with a view to high-throughput applications.

6

High-Throughput Biological Laser Printing

97

Fig. 6.1 A typical LIFT experimental set up is generally composed of three elements: a pulsed laser source, a target coated with the material to be transferred, and a receiving substrate. The target is a three layers component: a support, transparent to the laser radiation wavelength, coated with a thin absorbing layer, coated itself with a transfer layer, named bioink, that contains the elements to be printed (biomaterials, cells, biomolecules,. . .)

6.2.2 The Absorbing Layer To determine which metal to use for the absorbing layer and its thickness, the metal absorption coefficient and its skin depth have to be considered. The absorbing layer would be chosen to have a high absorption coefficient and a thickness slightly higher than the metal skin depth. The absorption coefficient A is defined by the Fresnel relations (Equation 1). The skin depth depends on the laser wavelength and the imaginary part of the complex refractive index of the metal (Equation 2): 4n (n + 1)2 + κ 2 λ δ= 2π κ

A(λ) = 1 − R =

(1) (2)

where R the reflection coefficient, n and κ the real and the imaginary part of the complex refractive index respectively, δ the skin depth and λ the laser wavelength. Table 6.1 presents titanium and gold optical characteristics at room temperature in the UV and IR range. In the IR range, a titanium absorbing layer is a more efficient transducer layer than a gold absorbing layer because of its higher absorption coefficient (45% in titanium instead of 0.9% in gold). The titanium skin depth at 1,064 nm is 50 nm. In the UV range, a titanium or a gold coating can be used since the absorption coefficients are similar (55.7 and 65.5%, respectively).

98

F. Guillemot et al.

Table 6.1 Comparison of optical characteristics of titanium and gold at room temperature at 1,064 and 355 nm

Titanium Gold

λ (nm)

n

κ

δ (nm)

A (%)

R (%)

1, 064 355 1, 064 355

3.35 1.3 0.1 0.58

3.30 2.01 6.54 1.73

50 28 25 32

45 55.7 0.9 65.5

55 44.3 99.1 34.5

6.2.3 Bubble Dynamics: The Rayleigh – Plesset Equation When the laser pulse is focused at the silica – absorbing layer interface, the laser energy is deposited into the titanium skin depth. By thermal conduction the rest of the titanium layer and first molecular layers of the bioink, located near the absorbing layer, are heated, producing a bioink vapor bubble and a fast expansion at the bioink-air interface. The growth phase of water bubbles in the bulk of a liquid has been studied in details (see [17] for instance). It has been shown that the growing – collapsing process could be described by the Rayleigh – Plesset equation, which states [18]: kPl 3 RR¨ + R˙ 2 = 2 ρl

R0 R

3γ −

2σl Pl − Pv R˙ − 4υ − ρl R ρl R

(3)

where R the vapor bubble radius, ρ l the liquid density, Pl the liquid pressure, ν the kinematic viscosity, σ l the surface tension, γ the ratio of specific heats, Pv the saturated pressure in the bubble and k the ratio of gas pressure in the bubble, Pg , to the liquid pressure Pl . This equation is the expression of the vapor bubble radius evolution versus time. It mainly depends on the liquid kinematic viscosity and surface tension. The evolution of a water bubble radius versus time is plotted in Fig. 6.2 for the following parameters: ρ = 1,000 kg·m–3 , ν = ν0 = 1.05×10–6 m2 /s, σ l = 72 mN.m–1 , k = 50,000, R0 = 1.26 μm, γ = 1.4, Pl = 1.105 Pa and Pv = 2,330 Pa (for more details, see [19]) and for kinematic viscosity values corresponding to different hydrogel solutions (at 10◦ C): v1 = 10 × 10−6

m2 (0.1% Alginate + 30% Glycerol) ; s

v2 = 40 × 10−6

m2 (0.5% Alginate + 30% Glycerol) ; s

v3 = 108 × 10−6

m2 (1% Alginate + 30% Glycerol) ; s

6


99

Fig. 6.2 Radius (in μm) evolution vs. time (in μs) of a vapor bubble for different viscosities (ν0 = 1.10–6 m2 /s/ν1 = 10.10–6 m2 /s/ν2 = 40.10–6 m2 /s/ν3 = 108.10–6 m2 /s)

The higher the kinematic viscosity of the liquid, the lesser the bubble radius oscillates, the smaller the maximum vapor bubble radius, and the faster the maximum radius is reached (Fig. 6.2). By adding a surfactant such as Sodium Dodecyl Sulfate, the surface tension of the hydrogel decreases from 72 to 65 mN/m, which seems to have no effect on the bubble radius evolution. Consequently, the bubble dynamics seems to be governed mainly by the kinematic viscosity of the solution.

6.2.4 Jet Formation: A Complex Threshold Mechanism The Rayleigh-Plesset equation describes bubble dynamics in an infinite volume. Because the size of the vapor bubble is not negligible compared to the bioink thickness, the interactions of the bubble with the free surface have to be taken into account. Towards this end, Pearson et al. and Robinson et al. [20, 21] have demonstrated, (i) when the bubble reaches its maximum diameter, it begins to collapse

100

F. Guillemot et al.

under the external pressure strengths, and (ii) a jet may be formed according to the dimensionless standoff distance (Equation 4), which is the ratio between the distance h (distance between the initial vapor bubble centroid and the free surface) and the maximum bubble radius, Rmax . =

h Rmax

(4)

In Equation 4, the maximum bubble radius, Rmax , depends on laser energy, through the k ratio in the Rayleigh-Plesset equation, and viscosity of the bioink (see Fig. 6.2), while h is related to the thickness of the bioink film. Experimentally, by printing different mixtures of alginate and glycerol, for a given laser energy, we have verified that the ratio depends on the viscosity ν and the thickness h of the bioink film (Fig. 6.3). With a bioink made of 1% of alginate and 30% of glycerol and a bioink thickness of 20 μm, droplets were successfully printed onto the substrate (Fig. 6.3a). When the bioink thickness was increased up to 100 μm, the substrate remained clear of any droplets, whereas bubbles were observed onto the target (Fig. 6.3b). If the bioink viscosity is increased (3% w/v alginate and 30% Glycerol) and the bioink thickness is decreased (20 μm), bubbles are also observed onto the target (Fig. 6.3c). Consequently, the three regimes generally observed in LIFT experiments (sub-threshold, jetting and plume regimes [10]) are not solely the result of laser irradiation conditions but, also, of rheological properties and film thickness of the bioink. In other words, jetting is not simply occurring on the basis of an energy threshold mechanism [15] but rather on the basis of a complex G(E, h, ν) threshold mechanism. Over a given laser energy for which a vapor bubble is formed at

Fig. 6.3 (a) Droplets printed onto the substrate obtained with a mixture of 1% alginate and 30% glycerol and a bioink thickness of 20 μm. (b) Bubbles on the target obtained within a mixture of 1% alginate and 30% glycerol and a bioink thickness of 100 μm. (c) Bubbles on the target obtained with a mixture of 3% alginate and 30% glycerol and a bioink thickness of 20 μm. The white spots correspond to the laser impacts and the black ones to the bubbles

6


101

Fig. 6.4 The three LIFT regimes: (a) The subthreshold regime: (E, h, ν) > 1 , (b) the jetting regime: < 2 , (c) the plume regime: 1 > > 2

the absorbing layer – bioink interface, the three above-mentioned regimes can be distinguished (Fig. 6.4): (i) If is higher than a threshold value 1 : the droplet ejection cannot occur, unless the substrate is close to the target. These are the conditions to observe the sub-threshold regime (Fig. 6.4a). (ii) If is lower than a threshold value 2 : the bubble expands until it bursts, giving rise to the plume regime (Fig. 6.4b). (iii) If . is ranged between 1 and 2 : the bubble expands, then collapses and finally a jet is formed (Fig. 6.4c). Integration of a High-Throughput BioLP workstation.

6.3 Integration of a High-Throughput BioLP Workstation 6.3.1 Terms of Reference To design a High-Throughput Biological Laser Printer, various pulsed lasers were evaluated for their suitability with living cells and biomaterials as well as for rapid prototyping applications. Major requirements considered were: (i) the wavelength λ should not induce alteration of biological materials, due to the potential denaturation of DNA by UV lighting, near infra-red lasers were preferable to UV lasers. (ii) the pulse duration τ and the repetition rate f must be considered with the purpose of high throughput processes. (iii) the beam quality, including divergence (q), spatial mode, and pulse-to-pulse stability (ptp) has to be taken into account to ensure the reproducibility, stability and high resolution of the system. A laser-based workstation dedicated to tissue engineering applications should be designed with the purpose of executing various tasks rather than solely bioprinting.

102

F. Guillemot et al.

Consequently, the mean laser power should be high enough to perform additional processes such as micro-machining, sintering, polymerization.

6.3.2 Design With regard to these criteria, a solid Nd:YAG crystal laser (Navigator I, Newport Spectra Physics) was selected with the following specifications (λ = 1,064 nm, τ = 30 ns, f =1–100 kHz, q = 3.4 mrad, TEM00 , ptp < 1,5% rms, P = 7 W). A 5-axe positioning system has been integrated to the workstation (NovaLase, S.A., Canéjan, France) with the purpose of printing multi-color patterns and building 3D biostructures. The substrate is held with a (x, y, z) motorized micrometric translation stage whose resolution is 1 μm for (x, y) axis and 5 μm for the z axis. In order to achieve multi-color printing, a high resolution (1◦ angular resolution) motorized carousel with a loading capacity of 5 different ribbons has been designed (Fig. 6.5b). Substrate positioning system and carousel are held on the same vertical axis in order to vary focusing conditions without changing the gap distance between each other. Droplet generation from the ribbon surface is performed by steering the laser beam by means of a high speed scanning system composed of two galvanometric mirrors (SCANgine 14, ScanLab), with a scanning speed reaching 2,000 mm/s, and a large field optical F-theta lens (S4LFT, Sill Optics, France) (F = 58 mm).

Fig. 6.5 (a) View of the high-throughput biological laser printer; (b) high resolution positioning system placed below the carousel holder with a loading capacity of 5 different ribbons

6


103

Focal setting in the ribbon and (x, y, z) substrate positioning are assisted by a CCD camera through the optical scanning system. The integration of the laser and R softall above-mentioned components has been made possible using Solidworks ware. Substrate positioning, carousel driving, video observation and pattern designs are monitored with one dedicated software. Finally, this CAD/CAM workstation (Fig. 6.5a) has been placed in a cell culture room in order to allow cell printing experiments.

6.3.3 High-Throughput Printing Parameters In HT BioLP technique, the spatio-temporal pulse proximity has to be taken into account in order to obtain a high printing resolution and reproducible results. In other words, both the distance and the time between laser spots have to be considered. Regarding the distance between 2 spots, contiguous laser spots should not overlap each other to avoid an alteration of the energy conversed by the absorbing layer. The recovery area of two discs is given by the following formula: √ Arec = 2R arcsin 2

4R2 − D2 2R

−

D 2 4R − D2 2

(5)

with

D=

ν f

(6)

which leads to the covering rate τrec : τrec =

Arec π R2

× 100

(7)

where R is the laser impact radius, D the distance between two laser impacts, v the velocity of the scanner mirrors and f the laser repetition rate. Depending on Arec , laser energy might be partially absorbed by the metal layer, which may alter droplet ejection. Regarding time, the characteristic time is more related to the bubble lifetime than to the pulse duration. Indeed, as shown above, bubble growth and collapsing are much slower events (~10 μs) than the energy deposit (~10 ns) and are, in addition, strongly sensitive to the viscosity of the bioink (Fig. 6.2). Figure 6.6 is an example of coalescent bubbles due to spatio-temporal pulse proximity for a viscous bioink and different scanning velocities (laser repetition rate is 5 kHz). Hence, laser repetition rate (or alternatively scanning velocity) has to be taken into account to avoid coalescence of vapor bubbles within the thickness of the bioink.

104

F. Guillemot et al.

Fig. 6.6 Coalescence of successive vapor bubbles observed within viscous solutions and low scanning velocity (from the top to the bottom of the picture the scanning velocity are 150, 120, 100, 70 and 50 mm/s)

6.4 Results and Discussion 6.4.1 Cell Printing 6.4.1.1 Ink Composition Considerations HT BioLP requires cells to be in a liquid suspension prior to being printed onto the substrate. However, extracellular matrix is critical for cell homeostasis in vivo. Also, in order to print a 3D material containing cells, the liquid ink should gel post printing onto the substrate. With respect to the layer-by-layer 3D building strategy, the gelling process is necessary (i) to stabilize the printed 2D pattern, (ii) to support subsequent ink layer. In addition, the gelling should not be harmful to the cells. According to these terms of reference, the cell containing ink was supplemented with 1% (w/v) alginate (Protanal 10/60, FMC Biopolymer) hydrogel as a preliminary approach to mimic extracellular matrix. The viscosity of different inks is shown in Table 6.2. It is noteworthy that 30% (v/v) glycerol added to a 1% (w/v) alginate

Table 6.2 Viscosity of culture media supplemented with alginate, glycerol, and cells. Glycerol is added to the ink to prevent evaporation. The cell model used in this experiment is the rabbit carcinoma cell line B16, at a density of 40.106 cell/mL. Experiment was carried on at 10◦ C Solution in culture media (DMEM)

Vicosity (Pa.s)

Alginate 0.5% (w/v); glycerol 30% (v/v) Alginate 1% (w/v) Alginate 1% (w/v); glycerol 30% (v/v) Alginate 1% (w/v); 40.106 cell/ml

0.05 0.1 0.11 0.12

6


105

ink accounts for 10% of viscosity, while 40.106 cells/mL accounts for 20% of the viscosity. Other viscous solutions suitable for BioLP have been described by Othon et al. [22]. 6.4.1.2 Viscosity and Laser Energy Deposit: Two Parameters that Influence Resolution Printing resolution depends on the ink viscosity and the laser energy deposit. Laser energy deposit can be modulated by tuning laser power or diaphragm aperture stop. Table 6.3 shows that the ejected droplet diameter onto the substrate is correlated to viscosity and laser power. The higher the viscosity and/or the lesser the energy deposit, the smaller the droplet diameter. It is possible to achieve similar resolution, i.e. similar droplet size, with a 0.1% (w/v) alginate ink printed with a laser pulse energy of 6 μJ (droplet size: (49 ± 3.5) μm, n = 15), and with a 1% alginate (w/v) ink printed with a pulse energy of 12 μJ (droplet size: (51 ± 4.2) μm, n = 15). A wide range of extracellular matrices, characterized by as many different viscosities can thus be printed at a similar resolution. Table 6.3 Diameter (in μm) of the ejected droplet onto the substrate, depending on alginate concentration in the ink, and laser power. The different inks were composed of mQ water supplemented with 30% (v/v) glycerol, with varying concentration of alginate. c: coalescence of contiguous droplets onto the substrate. n.t.: no transfer of the ink onto the substrate Energy (μJ) alginate (w/v) (%)

4.5

6

7.5

9

0.1 0.5 1.00

49 ± 4 (n=15) 38 ± 3 (n=15) n.t.

69 ± 4 (n=15) 55 ± 5 (n=15) 48 ± 4 (n=15)

C 64 ± 5 (n=15) 46 ± 3 (n=15)

c 62 ± 6 (n=15) 51 ± 4 (n=15)

6.4.1.3 High Cell Density Printing in High Spatial Resolution HT BioLP nozzle free set up precludes the cell printing process from clogging issues. Thus, inks loaded with cell densities similar to densities observed in living tissue can be used. The effect of the cell suspension viscosity on printing resolution was addressed. Cell suspensions supplemented with or without 1% (w/v) alginate were compared (Fig. 6.7). Cells were printed at a density of 50 million cells/mL, according to the pattern of the Olympic flag. Results show splashing [15] occurs in absence of alginate (Fig. 6.7a). In presence of 1% (w/v) alginate, splashing is virtually absent (Fig. 6.7b). One percent alginate increased the solution viscosity up to 0.1 Pa.s. With such a viscosity, splashing of the ink onto the substrate is reduced, and thus the resolution of the printed pattern is increased. As shown in Table 6.2, presence of cells into the ink impacts its viscosity. Laser parameters should then be adapted depending on the desired pattern to print. Different laser parameters were tested on two cell densities: 50 million cells/mL and 100 million cells/mL (Fig. 6.8). If droplet diameter is large enough, two contiguous

106

F. Guillemot et al.

Fig. 6.7 Cellularized 2D pattern resolution according to viscosity. Fifty million cells per ml were suspended in DMEM supplemented with 10% glycerol (a) plus 1% (w/v) alginate (b) Satellite droplets (splashing) are virtually absent when 1% (w/v) alginate is added to the ink

Fig. 6.8 Cell printing resolution according to the cell density in the ink (DMEM, 1% (w/v) alginate, 5% glycerol), the diaphragm aperture stop (D) and the laser energy deposit (P). Magnification 25×

printed droplets may coalesce, thus drawing a continuous line of cells. As a consequence, the higher the laser energy, the wider the printed cell line. Decreasing laser energy would then allow us to print cells one-by-one, next to each other. However, because HT BioLP is a LIFT based and nozzle free device, the number of cells in each ejected droplet is statistic. If cell density is too low on the ribbon, the ejected droplet of ink may not contain any cell (Fig. 6.8, upper left panel, diaphragm aperture stop D = 5 mm). To overcome this problem, at least two strategies can be proposed. (1) Increased laser energy leads to the ejection of bigger droplets. As a result, cells are more likely to be dragged off by draining/capillary effect (Fig. 6.8, upper left panel, D = 18–11 mm, and lower left panel, D = 11 mm). (2) Cell density can be increased up to the point cells are touching each other at the surface of the ribbon, i.e. 100 million cells/mL. In such a case, the viscosity of the ink is increased, and a higher laser energy deposit is necessary for printing (Fig. 6.8, upper right panel, D = 5 mm). A third strategy would implement a ‘aim and shoot’ technology into the HT BioLP device. In conclusion, HT BioLP is suitable to print versatile patterns such as cell clusters, cell banners and cell lines according to desired pattern (Fig. 6.9). Also,

6


107

Fig. 6.9 Basic cell patterns printed by HT BioLP. Single drop of ink containing varying number of cells can be printed at desired coordinates (left panel, magnification 200×). If drops of ink are printed close enough to each other onto the substrate, they may coalesce and form continuous line (upper right panel, magnification 25×). By adjusting energy deposit with suitable viscosity/cell density of the ink, cells could be printed one by one, at a cell-level resolution (lower right panel, magnification 25×)

cell-level resolution of cell organization is achievable by this LIFT based method to print cells in high throughput conditions (5 kHz). Optimization is possible in terms of higher throughput and higher resolution.

6.4.2 2-Dimensional Printing of Nano-Hydroxyapatite For bone tissue engineering applications of HT BioLP, a ‘bio-mimetic’ approach was chosen: it consisted in the printing of the individual components of bones. This tissue is composed of nano-sized hydroxyapatite crystals embedded in a protein matrix mainly composed of type-1 collagen [23]. In order to prepare the printable mineral phase of this tissue, a nanohydroxyapatite slurry (nHA) was synthesized by wet chemical precipitation [24] Briefly, a 0.6 M phosphoric acid solution was added dropwise into a 1 M calcium hydroxide solution and was kept at ambient temperature and pH 9 during 24 h under constant stirring. Then, 30% Glycerol (v/v) was added to the solution to improve the viscosity of the slurry and to prevent vaporizing of the material [25]. A thin film (30 μm) of the solution was coated on the ribbon with a ‘doctor blade’ device. The HT BioLP parameters were first adjusted to print a matrix of droplets; the scanner speed was 300 mm/s, and the energy per pulse was comprised between 6 and 12 μJ. This lead to the formation of droplets ranging from 40 to 100 μm wide, depending on the laser energy [15], i.e., the higher the laser energy the larger droplets (Fig. 6.10), and a threshold appeared at 12 μJ after which droplets were not well defined and lead to the formation of splashing. In this experiment, it appeared that the nanometric size of the hydroxyapatite crystals and their homogeneous size distribution in solution played an important role in the printing resolution.

6.4.3 3-Dimensional Printing of Nano Sized Hydroxyapatite For tissue engineering applications and building of composite structures for tissue repair, three-dimensional printing is admitted to be a critical issue [26]; towards this

108

F. Guillemot et al.

Fig. 6.10 nHA droplets printed with several laser energies. The higher energy, the larger droplets

Fig. 6.11 Three Dimensional printing of 15 superimposed layers of nHA observed by Scanning Electron Microscopy parallel to the receiving substrate. Left image: low magnification (x130) of the printed material. Right image: higher magnification (x360) of the ‘step’ (50 μm high) created by the stacked layers of nHA

end, the bioprinting device should be able to print stacks of material. A nHA solution was printed in several stacks to obtain a three dimensional structure (Fig. 6.11). Fifteen (stacked) layers of n-HA can be observed by scanning electron microscopy. This 3D structure has an average thickness of 45 μm. In this case, the material was highly compacted after the printing. In order to build a 3D porous structure, some sacrificial material layer [27] could be printed concomitantly with nHA to produce the desired porosity into the material.

6.4.4 Multicolor Printing Since biological tissues are composed of multiple components in close interactions with each other (cells of different types, extracellular matrix, etc. . .), not only 3 dimensional structures but also multiple ‘colors’ of biological materials should be considered in a tissue engineering product. Figure 6.12 illustrates the multicolor printing capability of the HT BioLP. Two distinct suspensions of cells (human

6


109

Fig. 6.12 Sequential two color cell printing in 2D. Human endothelial cell line (Ea hy 926) were loaded with 4 mM of Calcein-AM (b) or with fluorescent Dil-LDL (c) prior to the experiment. The two cell suspensions (60.106 cells/mL) in DMEM, supplemented with 1% (w/v) alginate were then printed according to a pattern of concentric circles (a). The two cell suspensions were each blade coated onto two separate ribbons. Both ribbons were placed onto the wheel distributor of the HT BioLP. Calcein loaded cell suspension was used to print the inner circle (1.2 mm diameter). Dil-LDL loaded cell suspension was used to print the outer circle (1.6 mm diameter). The overlay of Fig. 6.3b, c shows the two circles partially overlap due to the coalescence of the two circles

endothelial cell lineage Eahy926, 60 million cells /ml) were sequentially printed in 2 dimensions according to a pattern of two concentric circles (Fig. 6.12a). The inner circle was printed first using a suspension of cells loaded with green fluorescent calcein (Fig. 6.12b). Secondly, the outer circle was printed using a suspension of cells loaded with red fluorescent Dil-LDL (Fig. 6.12c). Differential staining of the cells revealed the two circles overlapped, which was likely due to bioink coalescence between the two circles. This experiment shows it is possible to print cells in close contact to each other, with a high cell concentration, according to a desired spatial organization. The printing resolution achievable by the HT BioLP is consistent with the study of cell-to-cell, or cell-to-material interaction as well.

6.4.5 In Vivo Printing To the best of our knowledge, all previous bioprinting studies involve in vitro experiments. We have performed some preliminary assays for in vivo printing [28]. More precisely, the purpose of our study was to fill a critical sized calvarial bone defect in mouse by printing a hydroxyapatite slurry (as prepared in Section 6.3.2). Briefly, the defects were performed bilaterally in mouse calvaria under general anesthesia, leaving the dura mera exposed. Then, the animals were placed inside the HT BioLP workstation. A specific mouse holder was designed in order to print the bioink onto the mouse dura mera instead of the quartz substrate depicted in Fig. 6.1. Then,

110

F. Guillemot et al.

Fig. 6.13 In Vivo printing of nHA. Histological sections of mouse calvarial defects after 1 week. (a) nHA aggregates are present in close contact to the dura mera (arrows). (b) Higher magnification of nHA aggregates. (DM: Dura Mera; FT: Fibrous Tissue)

30 nHA layers were printed inside one defect. The contro-lateral defect was used as negative control. Three groups of 10 animals each were studied. Animals were sacrificed after 1 week, 1 month or 3 weeks, respectively. The histological results have shown that the material was present in the test defects of all the groups (Fig. 6.13). However, bone repair was inconstant. As a conclusion, we show in vivo bioprinting is possible. Future experiment in this model should improve the mechanical and biological properties of the printed material.

6.4.6 Discussion Hypothesizing the innocuity of bioprinting technologies in regards to cell fate and biomaterial properties, we propose the definition of a bioprinting efficiency coefficient (BEC) which could also convey the bio-manufacturing rate. The dimensional equation of bio-manufacturing rate could be written as: L 1 L3 = × L3 × T T L speed × volume BEC = resolution

BEC =

(8) (9)

In Equation (9), speed, volume and resolution could be defined as the bioprinter writing speed (m/s), the typical volume of cells or biomaterials deposited per droplet (μm3 ) and the printing resolution (μm), respectively. The higher the BEC, the more efficient the bioprinting process for building 3D structures. Obviously, the high resolution criterion, which is required to precisely position cells, growth factors and other biomolecules, counterbalance the volume criterion. To address this latter issue, building 3D bio-structures would proceed from a combination of processes and technologies interacting with bioprinting and not only through the 3D deposit of microdroplets of cells and biomaterials [26]. Towards this

6


111

end, some authors have introduced the principle of bioprinting on biopapers [4]. More generally, additional technologies such as fused deposition modeling [29], aerosols [30], electrospinning [31] as well as optically driven methods (such as photo-polymerizing, micro-machining, laser catapulting [32], LGDW [8],. . .) would thus be combined with bioprinting when the micro-architecture and the cell microenvironment of some tissue parts or layers do not require a precise positioning of components. Consequently, a coefficient expressing the integrability or the interacting capacity of the bioprinting method has to be introduced in equation [9], leading to: BEC = integrability ×

speed × volume resolution

(10)

Finally, as regards BEC rate and because a high-throughput biological laser printer could be easily combined with other laser-based processes, Tissue Engineering Assisted by Laser should become more important in the near future. This capacity of combining different optical monitored processes is indeed, in our opinion, one of the major advantages of laser printing vs. traditional bioprinting methods involving either thermal or piezoelectric ink-jetting as well as pin depositing. In the future, such additional optical processes would include, but not be limited to, micro-machining, cutting, photo-polymerizing, sintering, etc.

6.5 Conclusions and Perspectives In this chapter, we have shown that High-Throughput Biological Laser Printing (HTBioLP) requires to take into account spatio-temporal proximity of laser pulses as well as the rheological properties of the bioink since the droplet ejection mechanism is mainly governed by vapor bubble dynamics. Hence, jetting thresholds are discussed as regards the dimensionless parameter which is the ratio between the initial bubble centroid and the free surface. After the development of a highthroughput biological laser workstation, we have also shown its potentiality for depositing a wide range of biological components, all of which are required for tissue engineering: biopolymers, nano-sized particles of hydroxyapatite as well as human endothelial cells. Then, after describing experimental conditions leading to the high resolution printing of biological components, we present some typical multi-component and 3D printings achieved using this workstation. These results lead us to emphasize the criteria that are required for building 3D structures through bioprinting process: the writing speed, volume fraction of deposited materials, process resolution and its capacity to be combined with other tissue engineering methods. Acknowledgments The authors would like to thank GIS ‘Advanced Materials in Aquitaine’ and Région Aquitaine for financial support. In addition, the authors would like to thank Reine Bareille and Murielle Rémy for their help in cell culture experiments.

112

F. Guillemot et al.

References 1. Hutmacher DW (Déc 2000) Scaffolds in tissue engineering bone and cartilage. Biomaterials 21(24):2529–2543 2. Mironov V, Boland T, Trusk T, Forgacs G, Markwald RR (Avr 2003) Organ printing: computer-aided jet-based 3D tissue engineering. Trends Biotechnol 21(4):157–161 3. Jakab K, Norotte C, Damon B, Marga F, Neagu A, Besch-Williford CL, Kachurin A, Church KH, Park H, Mironov V, Markwald R, Vunjak-Novakovic G, Forgacs G (Mar 2008) Tissue engineering by self-assembly of cells printed into topologically defined structures. Tissue Eng A 14(3):413–421 4. Mironov V, Visconti RP, Kasyanov V, Forgacs G, Drake CJ, Markwald RR (Avr 2009) Organ printing: tissue spheroids as building blocks. Biomaterials 30(12):2164–2174 5. Boland T, Xu T, Damon B, Cui X (Sep 2006) Application of inkjet printing to tissue engineering. Biotechnol J 1(9):910–917 6. Nakamura M, Kobayashi A, Takagi F, Watanabe A, Hiruma Y, Ohuchi K, Iwasaki Y, Horie M, Morita I, Takatani S (13 Jan 2006) Biocompatible inkjet printing technique for designed seeding of individual living cells. [Internet] [cité 17 Sep 2008]. Available from http://www.liebertonline.com/doi/abs/10.1089/ten.2005.11.1658 7. Saunders RE, Gough JE, Derby B (Jan 2008) Delivery of human fibroblast cells by piezoelectric drop-on-demand inkjet printing. Biomaterials 29(2):193–203 8. Nahmias Y, Schwartz RE, Verfaillie CM, Odde DJ (2005) Laser-guided direct writing for three-dimensional tissue engineering. Biotechnol Bioeng 92(2):129–136 9. Nahmias Y, Odde DJ (21 Déc 2006) Micropatterning of living cells by laser-guided direct writing: application to fabrication of hepatic–endothelial sinusoid-like structures. [Internet] [cité 17 Sep 2008]. Available from http://www. nature.com/nprot/journal/v1/n5/pdf/nprot.2006.386.pdf 10. Young D, Auyeung RCY, Piqué A, Chrisey DB, Dlott DD (Sep 2002) Plume and jetting regimes in a laser based forward transfer process as observed by time-resolved optical microscopy. Appl Surf Sci 30:197–198, 181–187 11. Barron JA, Ringeisen BR, Kim H, Spargo BJ, Chrisey DB (Avr 2004) Application of laser printing to mammalian cells. Thin Solid Films 453–454:383–387 12. Hopp B, Smausz T, Kresz N, Barna N, Bor Z, Kolozsvári L, Chrisey DB, Szabó A, Nógrádi A (13 Jan 2006) Survival and proliferative ability of various living cell types after laser-induced forward transfer. [Internet] [cité 17 Sep 2008]. Available from http://www.liebertonline.com/doi/abs/10.1089/ten.2005.11.1817 13. Barron JA, Wu P, Ladouceur HD, Ringeisen BR (June 2004) Biological laser printing: a novel technique for creating heterogeneous 3-dimensional cell patterns. Biomed Microdevices 6(2):139–147 14. Barron JA, Krizman DB, Ringeisen BR (Fév 2005) Laser printing of single cells: statistical analysis, cell viability, and stress. Ann Biomed Eng 33(2):121–130 15. Duocastella M, Colina M, Fernández-Pradas J, Serra P, Morenza J (July 2007) Study of the laser-induced forward transfer of liquids for laser bioprinting. Appl Surf Sci 253(19): 7855–7859 16. Duocastella M, Fernández-Pradas J, Serra P, Morenza J (1 Nov 2008) Jet formation in the laser forward transfer of liquids. Appl Phys A Mater Sci Process 93(2):453–456 17. Brennen CE (1 Jan 1995) Cavitation and bubble dynamics. [Internet] [cité 23 Jul 2009]. Available from http://caltechbook.library.caltech.edu/1/ 18. Xiu-Mei L, Jie H, Jian L, Xiao-Wu N (2008) Growth and collapse of laser-induced bubbles in glycerol–water mixtures. Chinese Phys B 17(7):2574–2579 19. Mézel C, Hallo L, Souquet A, Guillemot F, Breil J (2009) Self-consistent modeling of a LIFT process in the nanosecond laser pulse regime. Phys Plasma 16(12):123112-12 20. Pearson A, Cox E, Blake JR, Otto SR (Avr 2004) Bubble interactions near a free surface. Eng Anal Bound Elem 28(4):295–313

6


113

21. Robinson PB, Blake JR, Kodama T, Shima A, Tomita Y (15 Juin 2001) Interaction of cavitation bubbles with a free surface. J Appl Phys 89(12):8225–8237 22. Othon CM, Wu X, Anders JJ, Ringeisen BR (2008) Single-cell printing to form threedimensional lines of olfactory ensheathing cells. Biomed Mater 3(3):034101 23. Cowin S, Author, Telega J, Reviewer (0 Juillet 2003) Bone mechanics handbook, 2nd edn. – Appl Mech Rev 56(4):B61–B63 24. Afshar A, Ghorbani M, Ehsani N, Saeri MR, Sorrell CC (Mai 2003) Some important factors in the wet precipitation process of hydroxyapatite. Mater Des 24(3):197–202 25. Keriquel V, Guillemot F, Arnault I, Guillotin B, Miraux S, Amédée J, Fricain J-C, Catros S (2010) In vivo bioprinting for computer- and robotic-assisted medical intervention: preliminary study in mice. Biofabrication 2(1):014101. doi: 10.1088/1758-5082/2/1/014101 26. Boland T, Tao X, Damon BJ, Manley B, Kesari P, Jalota S, Bhaduri S (Avr 2007) Dropon-demand printing of cells and materials for designer tissue constructs. Mater Sci Eng C 27(3):372–376 27. Lee W, Debasitis JC, Lee VK, Lee J, Fischer K, Edminster K, Park J, Yoo S (March 2009) Multi-layered culture of human skin fibroblasts and keratinocytes through three-dimensional freeform fabrication. Biomaterials 30(8):1587–1595 28. Kériquel V, Guillotin B, Arnault I, Miraux S, Amédée J, Guillemot F, Fricain J, Catros S (2010) In vivo high-throughput biological laser printing of nano-hydroxyapatite in mice calvarial defects. Biofabrication 2(1):014101 29. Zein I, Hutmacher DW, Tan KC, Teoh SH (Fév 2002) Fused deposition modeling of novel scaffold architectures for tissue engineering applications. Biomaterials 23(4):1169–1185 30. Gagné L, Rivera G, Laroche G (Nov 2006) Micropatterning with aerosols: application for biomaterials. Biomaterials 27(31):5430–5439 31. Prabhakaran MP, Venugopal J, Chyan TT, Hai LB, Chan CK, Tang ALY, Ramakrishna S (25 July 2008) Electrospun biocomposite nanofibrous scaffolds for neural tissue engineering. Tissue Eng A [Internet] [cité 30 Sep 2008]. Available from http://www. ncbi.nlm.nih.gov/pubmed/18657027 32. Vogel A, Lorenz K, Horneffer V, Huettmann G, von Smolinski D, Gebert A (31 Aoû 2007) Mechanisms of laser-induced dissection and transport of histologic specimens. Biophys J biophysj.106.102277

Chapter 7

Absorbing-Film Assisted Laser Induced Forward Transfer of Sensitive Biological Subjects B. Hopp, T. Smausz, and A. Nógrádi

Abstract The traditional Laser Induced Forward Transfer technique – originally developed for transfer of inorganic thin films from a transparent holder to a facing substrate – has been modified to allow the controlled transfer of sensitive biomaterials and cells. In our experimental arrangement the biological substrates (biomaterials, cells and conidia) are spread onto the surface of a thin metal layer coated fused silica plate. According to the name of the developed method – Absorbing-Film Assisted Laser Induced Forward Transfer (AFA-LIFT) – the target biological layer is protected from the photonic and thermal effects of the laser irradiation by the metal film which absorbs the laser energy and converts it to kinetic energy resulting in the ejection of the biological structures. The biomaterials were transferred either from dry (conidia) or wet (living cells) environment. After the process the substrates were placed in culture medium and the state of the transferred materials was monitored with optical microscopy. In case of Trichoderma longibrachiatum conidia 20 h incubation time after the transfer the highest germination ratio (number of the germinated conidia divided by the estimated number of transferred conidia) was around 75% reached at 355 mJ/cm2 laser fluence. Freshly isolated cells (rat Schwann and astroglial and pig lens epithelial cells) were transferred from solution, spread on the holder in 140–160 μm thick layer to substrate covered with a wet gelatin layer. The trypan blue dye exclusion test showed that 80–85% of the transferred cells still remained intact after transfer. The initially round-shaped cells 24 h after the transfer started to proliferate and differentiate. After 1 week the cells formed a monolayer with no signs of cellular degeneration. The dynamics of the process was studied with a fast photographic arrangement using a 1 ns pulse-length dye laser beam as a probe pulse. In case of ‘dry transfer’ the velocity of the front of the ejected conidia plume was 1,150 m/s, while the lower limit of the estimated initial acceleration was 109 ×g. For the ‘wet transfer’

B. Hopp (B) Research Group on Laser Physics, Hungarian Academy of Sciences and University of Szeged, Dóm tér 9, H-6720, Szeged, Hungary e-mail: [email protected]


115

116

B. Hopp et al.

the velocity of the liquid jet containing the cells was ‘only’ 122 m/s and the acceleration was above 107 ×g. This means that the studied conidia and cells can tolerate extremely high acceleration at the beginning of the ejection and the deceleration during the impact to the acceptor plate without significant damage. The absorbing metal film missing from the irradiated area and numerical thermal calculations both indicate that the metal is evaporated at the fused silica-metal interface by the laser beam and the expanding gas layer initiates the ejection of the biological structures. In order to avoid the possible thermal damages of the cells caused by nanosecond laser irradiation, cell transfer using femtosecond excimer laser was also studied. However, femtosecond AFA-LIFT proved to be less advantageous for cell transfer, probably due to the extreme mechanical forces produced by the short laser pulse.

7.1 Introduction Laser Induced Forward Transfer (LIFT) is a laser-based direct write method, which was originally developed for the controlled transfer of absorbing metal films from a transparent holder (donor) to a substrate (acceptor). The two plates are placed parallel to each other with a gap of up to few hundred micrometers between them [1–6]. This setup also constituted the basics of Matrix Assisted Pulsed Laser Evaporation Direct Write (MAPLE-DW), designed to fabricate mesoscopic electronic devices from composite materials, which has also been successfully applied for cell and biomaterial patterning [7–10]. The scenario behind the development of the Absorbing Film Assisted LIFT (AFA-LIFT) was to transfer biological materials placed on an absorbing film-coated holder where structure and viability can be preserved. In the case of living cells the use of nutrients or culture medium is required; however unlike MAPLE-DW, this medium does not have the role of absorbing matrix material. Using thin layers with high absorptivity at the applied laser wavelength the photonic and thermal damages exerted upon the living organism or biomaterial can be diminished. The role of this absorbing layer is somewhat similar to that of a dynamic release layer [11], with a difference that in the latter case the material to be transferred is released and propelled away from the dynamic release layer.

7.2 The Experimental Setup A 50–100 nm thick metal layer (mainly silver) deposited by vacuum evaporation onto the fused silica holder served as the absorbing and carrying layer. Since the penetration depth (10–20 nm) of the applied KrF excimer laser beam (λ=248 nm, FWHM=30 ns or 500 fs depending on the applied laser type) is much smaller than the film thickness the biomaterials placed on it are protected from damaging effects of the UV laser irradiation. The applied biological organisms were conidia in dry state and various types of living cells in culture medium spread

7

Absorbing-Film Assisted Laser Induced Forward Transfer

117

in thickness of 140–160 μm over the metal thin film. The distance between the acceptor plate and the holder was in the range of 600–1,000 μm. The surface of the substrate was determined by the aim of the actual experiment: pure glass, agar culture medium, (dry or wet) gelatin- and Matrigel-coated glass were used. A relatively homogeneous part of the laser beam was selected and focused onto the absorber layer through the transparent holder (fused silica plate) onto a 250–300 μm diameter spot, the applied laser fluence for the successful transfer was in the range of 200–350 mJ/cm2 . Each target area was exposed to a single UV pulse. The experiments were carried out at room temperature in air. The pulse energy was controlled by an attenuator and measured shot by shot. The donor-acceptor arrangement was illuminated by a halogen lamp and observed by a camera attached to a microscope. The above setup made the control of the AFA-LIFT process possible (Fig. 7.1). For better understanding of the transfer process time-resolved studies were performed using a laser-based photographic arrangement. For this purpose the basic AFA-LIFT setup was completed with an illumination-camera system. The ejected material was illuminated parallel to the fused silica plate by a nitrogen laser pumped Coumarin 153 dye laser beam (λ = 453 nm, FWHM = 1 ns) electronically delayed relative to the pulse of the excimer laser (Fig. 7.2). The shadow graph of the material

Fig. 7.1 Experimental arrangement for absorbing film assisted laser induced forward transfer of biological organisms

118

B. Hopp et al.

Fig. 7.2 Laser-based photographic setup for time-resolved study of the material transfer processes during AFA-LIFT

ejection was observed and recorded by a microscope–video camera–monitor system. The time-delay between the UV and probe laser pulses was varied in 0 and 100 μs range. In some cases the acceptor plate was removed to allow undisturbed observation of the material ejection process.

7.3 Laser Transfer of Fungus Conidia For the first series of our AFA-LIFT experiments fungus conidia were chosen because of their high tolerance against the environmental effects. Species of the Trichoderma genus are common soilborne filamentous fungi with teleomorphs belonging to the Hypocreales order of the Ascomycota division [12]. Several isolates have outstanding ability to produce extracellular degrading enzymes, the phenomenon which corresponds to their ecological role in the decomposition of plant litters [13]. With the extracellular enzymes, Trichoderma strains are able to attack other fungi, based on their agricultural application as biological controlling agents acting against plant-pathogenic fungi [14, 15]. Although, Trichoderma strains are mainly saprophytic soil organisms, in some cases they can cause opportunistic infections in immunocompromised humans ranging from localized infections to disseminated fatal diseases [16]. The Trichoderma longibrachiatum strain CECT 20106 derived from the Spanish Type Culture Collection (CECT) was maintained on yeast extract agar culture medium (1.0% glucose, 0.5% yeast extract, 1.0% potassium dihydrogen phosphate, 2% agar in distilled water, all w/v). As donor species of the transfer experiments, conidia of the Trichoderma strain were harvested from a 10 days old culture, suspended in distilled water, spread at a 2,530±455/mm2 surface density on a quartz plate coated with 50 nm thick silver film, and then dried in air. The acceptor glass plate carrying the dried agar culture medium faced the donor slide with an approx.

7


119

1 mm gap between them, the size of the spot irradiated with the nanosecond excimer laser was 0.08 mm2 [17]. After the laser transfer was performed at fluences in 35–2,600 mJ/cm2 range, the acceptor plates were incubated for 20 h in humid 25◦ C environment and then observed with an optical microscope. Corresponding to the spatial positions of the illuminated donor spots, several islands of germinated Trichoderma conidia appeared on the culture medium. The germ density in these pixels showed a strong dependence on the applied fluence, as can be seen in Fig. 7.3. Most of the germinated conidia are located within the area corresponding to the irradiated spot on the donor

Fig. 7.3 Optical microscopic images of germinated Trichoderma conidia islands produced by AFA-LIFT at different fluences, after 20 h incubation. The last picture shows the donor surface after irradiation with a single pulse of 224 mJ/cm2 fluence [17]

120

B. Hopp et al.

layer, while some are spread in a larger area. Smaller spreading area could probably be obtained by decreasing the distance between the donor and the acceptor surfaces. Transferred silver particulates could not be identified by optical microscope. The last picture in the panel shows an area of the donor layer irradiated by a single pulse of 225 mJ/cm2 fluence; the small dark spots are the conidia. For estimating the efficiency of the transfer process the area populated by the germinated conidia was measured by image analysis software. As Fig. 7.4 shows, the increase of fluence up to 355 mJ/cm2 resulted in a steep increase of the area occupied by the fungi, probably due to the increased numbers of biologically active transferred conidia. Further increase of the fluence resulted in a sudden drop of the efficiency to an approximately constant level. The likely cause of this phenomenon at high fluences is that the high intensity laser pulses not only transfer the conidia, but may simultaneously cause thermal and/or mechanical damage reducing their viability. To estimate the number of transferred and germinated conidia the average area of several separated germs was determined and the covered areas were divided by this value. This is an underestimation because we could not take into consideration the overlaps on the printed areas. The germination ratio was defined as the ratio of the number of germinated conidia and the average conidia number on the laser irradiated donor area. This value was found to be higher than 75% at 355 mJ/cm2 fluence (Fig. 7.4). Using the germination ratio both the efficacy of the transfer and the survival rate of the conidia exposed to effects of laser irradiation can be summarized.

Fig. 7.4 Area populated by the germinated conidia and the germination ratio (the number of the germinated conidia divided by the number of conidia on the irradiated area of the donor layer) after the AFA-LIFT process as the function of the applied laser fluence [17]

7


121

In order to obtain more insight into the mechanism of the AFA-LIFT procedure the spatial and temporal characteristics of the temperature profiles in the absorbing silver film were calculated. This was done by computing the thermal transport equation numerically, taking into account the temperature dependence of the thermophysical data and the phase changes. Only the solution of the one-dimensional heat flow equation was required since the lateral dimensions of the heated material were several orders of magnitude greater than that of film thickness. The reflectivity of the fused silica plate–silver layer system was measured to be 36% at 248 nm. Since the optical absorption depth (14 nm) is approximately a quarter of the silver layer thickness (50 nm) it could be argued that the energy of the UV laser pulse was totally absorbed in the metal film acting as a heat plate. The thermal conductivity of the silver is relatively high, thus the temperature distribution within the whole layer is approximately uniform. The contacting fused silica substrate is heated up by the heat flow originating from the metal layer. The heat conduction of the surrounding air was also considered. The investigated fluences were 35 and 355 mJ/cm2 . It was found that the maximum temperatures of the silver layer were 447◦ C (reached at 50 ns after the onset of the laser pulse) and 2,210◦ C (at 33 ns) when working with 35 and 355 mJ/cm2 pulses, respectively. This means that two different cases can be distinguished during our experiments. In the first case the absorbing film does not melt and evaporate, while in the second case both melting and evaporation occurs. Nevertheless, it was observed that the whole irradiated layer was removed in both cases. We believe that as the degassing temperature of the fused silica is much lower than the melting point of the silver, during the excimer laser illumination a gas layer formed at the film-substrate interface exerting a positive pressure [18, 19], resulting in a blowing off effect of the metal film and the harvested conidia (35 mJ/cm2 ). Application of 355 mJ/cm2 of fluence ensured overall melting and evaporation. Therefore, the group of the ejected/evaporated silver atoms obtained approximately 1.23·10−5 J thermal energy. According to our assumption the silver vapor and microdroplets formed a diffuse matrix carrying away the conidia. A more detailed characterization of the transfer process could be reached by a time-resolved study [20]. The shape of conidia can be approximated by a rotation ellipsoid having 4.5 and 2.2 μm average length and diameter, respectively. Since these dimensions are smaller than the 6 μm spatial resolution of our imaging system individual ejected conidia could not be observed. Therefore, the donor plate was covered with conidia of high density (Fig. 7.5) as compared to the previous experiments to ensure a good observation during the transfer. The applied laser fluence was 355 mJ/cm2 , where the highest transfer efficacy was obtained. Figure 7.6 shows snapshots of the ejected material at different delays in the case of the irradiated conidia–silver–fused silica sample. The pictures recorded without excimer laser pulses are subtracted as background. It can be seen that material ejection started at around 100 ns and finished after 10 μs. The propagation of the plume closely follows the movement of the shockwave front. In the last picture the conidia aggregates can be identified. In order to differentiate the silver particles and conidia during the transfer process this experiment was repeated using silver-coated quartz plate without conidia. As

122

B. Hopp et al.

Fig. 7.5 Optical microscopic image of the donor plate. Conidia of the Trichoderma strain were spread on silver-coated surface of quartz plate [20]

Fig. 7.6 Snapshots of the ejected conidia plume at different time delays observed by fast- photographic arrangement during AFA-LIFT of Trichoderma longibrachiatum conidia. The applied fluence was 355 mJ/cm2 [20]

shown in Fig. 7.7, in this case material emission was not detected, only a weakly discernible shock wave referred to the presence of ejected silver. Optical microscopy showed no evidence for deposited silver particles and/or droplets on the acceptor surface. Therefore we believe that the irradiated silver layer (ca. 4.2·10−8 ×g material per laser pulse) was completely evaporated. However, it is possible that deposition

7


123

Fig. 7.7 Pure silver-coated quartz plate was also irradiated by excimer pulses. Significant material ejection was not observable, only a weakly discernible shock wave referred to the ejected silver [20]

of condensed silver microdroplets possessing a diameter less than 700 nm (the resolution of our optical microscope) occurred. The distance between the front of the emitted material plume and the donor surface was measured for the applied delays (Fig. 7.8). The linear relationship evident in this figure indicates a constant travel velocity of 1,150 m/s during the transfer, which is about three times the sound velocity. Accordingly, the biological samples

Fig. 7.8 Distance between the front of ejected plume and the donor surface as a function of the time delay. The average ejection velocity (1,150 m/s) was determined on the basis of the slope of the line fitted to the measured data [20]

124

B. Hopp et al.

have to tolerate without significant damage the extremely high acceleration at the beginning of the ejection and the deceleration during the impact at the acceptor surface. On the basis of our recordings (assuming that the duration of the acceleration does not exceed 100 ns) and the calculated maximum velocity, the estimated average acceleration of the ejected conidia is approximately 109 ×g. This acceleration will expose the conidia to unprecedented levels of mechanical shear. Several space research studies proved that some types of bacteria and spores can endure extreme conditions (very high/low pressure and temperature, acceleration, UV irradiation, etc). In these experiments the extreme high acceleration values were realized using approaches such as explosion, shock wave, ultracentrifuge, gun, etc and the maximum acceleration achieved was in the range of 105 –106 ×g [21–24]. A seemingly inconsistent result of our investigation is that the studied Trichoderma conidia experienced little damage during the transfer because they were alive and successfully reproduced after deposition. It is difficult to give a feasible explanation for their survival, but it can be argued that the high survival rates were likely due to: (a) the extremely short duration (from 0 to ~100 ns) of acceleration compared to other experiments; (b) the small dimensions (~μm) of the conidia, and; (c) the uniform force applied during desorption. It is also widely accepted that the conidia are very resistant to physical and chemical insults. The AFA-LIFT transfer process can be divided into four consecutive elements: (a) the majority of the UV photons are absorbed by the silver layer causing localized electronic excitation (Fig. 7.9a, b) The sudden evaporation of metal (it expands dramatically producing a fast shock wave above the sample in air). The collective action of gentle collisions between the evaporated silver and the conidia on the surface will accelerate them within a few tens of nanoseconds (Fig. 7.9b). The average thermal velocity for the evaporated silver atoms was calculated to be in the same order of magnitude with the plume front velocity; (c) The conidia are then transferred across the 1 mm gap with supersonic velocity following the shock wave front (Fig. 7.9c, d) The transferred conidia impact to the acceptor surface and deposit on it (Fig. 7.9d). The culture medium on the substrate glass plate has two important roles in the procedure: it supplies the deposited conidia with nutrient and more importantly it reduces the impact shock through its elasticity. The estimated maximum impact energy (the whole kinetic energy at the substrate surface before the collision) of a conidia is very low, approx. 10−8 J.

Fig. 7.9 The transfer process according to our experimental results and hypothesis [20]

7


125

7.4 Laser Transfer of Living Cells Although laser-based cell printing techniques such as MAPLE-DW [7–9], LIFT [25] and biological laser printing (BioLP) [26, 27] have been used for transfer of biomaterials or highly resistant cells such as cell lines or tumor cells, the intriguing questions raised, whether more vulnerable, freshly isolated primary cells (a) may also be successfully transferred and (b) the transferred cells survive, differentiate, and maintain their phenotype for longer periods (several weeks) of time after laserinduced transfer. Therefore the aim of the following studies was to investigate the survival and in vitro development of the phenotype of various freshly isolated cell types after laser transfer and to improve our understanding of the material ejection processes during AFA-LIFT [28]. Three types of purified and cultured cells were used in the AFA-LIFT experiment: rat Schwann and astroglial cells and pig lens epithelial cells. Schwann cells were isolated from adult sciatic nerves by 0.2% collagenase and 0.2% trypsin treatment for 20 min. The isolated cells were plated onto Petri dishes previously covered with isogeneic rat serum. The cells were cultured in HEPES-buffered Dulbecco’s modified Eagle’s medium (HDMEM; Sigma, Budapest, Hungary, for details of cell culture see [28]). Astroglial cells were obtained from newborn rat brain hemispheres by homogenizing the hemispheres with culture medium. The supernatant was carefully removed and the remaining pellet was plated onto Petri dishes and cultured in HDMEM containing 10% fetal calf serum. Lens epithelial cells were collected from pig lens capsules. The cornea was carefully removed with the scleral rim, the lens was removed, and the anterior lens capsule was prepared. After 0.2% trypsin digestion for 20 min the cells were centrifuged, plated onto Petri dishes, and cultured in HDMEM containing 10% fetal calf serum. The cells were left to grow for 10–14 days or until they formed a monolayer and then were harvested. The donor slide was a fused silica plate covered with 100 nm thick silver film. Solution containing cultured cells in a 2–6·106 cell/mL concentration was spread in 140–160 μm thickness. The acceptor glass plate coated with a thin wet gelatin layer was placed facing the donor plate with a 0.6-mm gap between them. The area irradiated by the nanosecond KrF excimer laser beam was approximately 0.07 mm2 and the applied fluence was 360 mJ/cm2 . The vitality and survival of the cells were evaluated by 0.5% trypan blue staining before and after laser-pulsed cell transfer. The acceptor plates carrying the transferred cells were immediately processed for cell culture and the cells were investigated and photographed every day. The survival, differentiation, and morphological phenotype of the various types of transferred cells were investigated. The trypan blue dye exclusion test showed that 98–99% of each cell type was alive before cell transfer and that 80–85% of the transferred cells still remained intact after transfer. The cells remained rounded and grouped on the day of transfer (Fig. 7.10a0–c0), but had already started to proliferate and differentiate 1 day after cell transfer (Fig. 7.10a1–c1). By 3 days only well-differentiated cells with long processes and intact cell body and nucleus were observed (Fig. 7.10a3– c3). Although some degenerated cells and debris were seen, the rest of the cultured cells appeared intact. One week after cell transfer the cultured cells formed a

126

B. Hopp et al.

Fig. 7.10 Optical microscopic images of the transferred [Schwann (a), astroglial (b), and lens epithelial (c)] cells at various incubation times (0, 1, 3, and 14 days) and those of the control cell cultures (a contr.–c contr.) [28]

monolayer and this intact layer of cells was still present 2 weeks after laser-pulsed cells transfer (Fig. 7.10a14–c14). No degenerating cells were observed at any of these later time points. Control cell cultures, that is, the same density but larger amounts of cells plated at the time of cell transfer and kept in culture under the same conditions, grew faster and formed a monolayer by 3 days after plating.

7


127

The difference between transferred and control cells was, however, hardly visible by 2 weeks in the case of lens epithelial cells whereas transferred Schwann cells and astroglial cells showed a lower density than control cells (Fig. 7.10a contr.– c contr.). Dynamics of the cell transfer was also studied with the above-mentioned fast-photographic arrangement. Figure 7.11 shows images of the material ejection taken at different time delays. The emission started at around 500 ns, the liquid jet hits the acceptor surface about 4 μs after the excimer laser pulse and then a part of the jet was reflected from the glass plate. At this time point there is a fast liquid stream directed from the donor toward the acceptor plate. After a few tens of microseconds this stream slowed down and stopped. Afterwards slow backflow and collapse were observed. The connection between the donor and acceptor plates completely disappeared at about 1 ms and a liquid droplet remained on the acceptor surface. The distance between the peak of the jets and the original liquid surface was measured and plotted as the function of the time delay. A linear relation was found between these values (Fig. 7.12). The average ejection velocity obtained by linear fitting was 122 m/s, about an order of magnitude less than in the case of conidia transfer (1,150 m/s) in dry environment.

Fig. 7.11 Liquid jet ejection process at different delays recorded by the photographic arrangement. At about 1 μs a liquid jet is ejected from the solution layer. This reaches the acceptor surface at 4 μs and then a part of the jet is reflected from it. The connection between the donor and acceptor plates completely discontinues at about 1 ms and a liquid droplet remains on the acceptor surface [28]

128

B. Hopp et al.

Fig. 7.12 Distance between the peak of the ejected liquid jet and the donor surface as a function of the time delay. The average ejection velocity (122 m/s) was determined on the basis of the slope of the line fitted to the measured data [28]

Assuming that the acceleration period of the jets and droplets containing the cells is less than 1 μs (when the constant velocity is already reached, according to Fig. 7.12), the estimated minimum average acceleration of the ejected cells is approximately 107 ×g. This is two orders of magnitude lower than the value determined in the case of conidia transfer. Ringeisen et al., in their laser printing procedure, estimated a deceleration of transferred cells during the impact against the acceptor surface of 106 ×g [27]. This means that the majority of the studied cells tolerate the extremely high acceleration both at the beginning of the ejection and the deceleration during the impact to the acceptor plate without significant damage. It is thought that there are two critical points regarding the survival of the cells during the transfer: thermal effects of the laser irradiation and impact to the acceptor surface. Some of the cells being in direct contact with the boiling and evaporating metal absorbing layer may be exposed to intensive heat-effect resulting in irreversible damage of the cells. Moreover, those cells that survived the transfer process and adhered to the acceptor plate originated from the slow part of the stream phase in the ejected jet and still had enough impact energy to avoid backdrift. As a sketch in Fig. 7.13 about the main steps of the transfer shows, few cells in and around the peak of the jet perished during impact on the gelatin layer (Fig. 7.13c). Most of the cells survived the ejection in the middle of the jet; however, several cells could not adhere to the acceptor plate and these drifted back to the donor surface (Fig. 7.13e, f). The arrows in Fig. 7.13 indicate the momentary direction of the stream velocity.

7


129

Fig. 7.13 Draft scheme of the liquid jet ejection process during AFA-LIFT of cells [28]

7.5 Femtosecond AFA-LIFT of Living Cells Although several successful cell transfer experiments were carried out in our laboratory, the survival rate of the transferred cells never exceeded 80–85%. It is thought that this effect is partly due to the thermal damage of the sensitive biological samples generated by the nanosecond laser irradiation. A possible solution of this problem can be the use of femtosecond lasers for the transfer procedure. Ultrashort laser pulses reportedly reduce the indirect thermal effects, thus allowing the effective transfer of sensitive biomaterials at high spatial resolution for micro-fabricating patterns. Zergioti et al. successfully used femtosecond laser pulses (500 fs, 248 nm) to print DNA molecules. The dynamics of the process was investigated by stroboscopic Schlieren imaging [29]. They provided evidence that femtosecond laser-based direct transfer of DNA onto various substrates without the assistance of any transferring matrix is a clean, one step process not limited to oligomer structures [30]. The aim of our following study was to investigate whether the femtosecond AFA-LIFT method: (a) is suitable for controlled transfer of living cells onto various acceptor surfaces, and; (b) by reducing the damaging thermal effects it is able to increase the survival rate and thus to widen the range of target cell types. Our investigations focused on the usefulness of various absorbing layer and acceptor coating materials and the reaction of different cell types to the transfer procedure. Absorbing layer materials included metals, such as silver, gold, chromium and copper, and biomaterials (gelatine, polyhydroxy-butyrate and Matrigel, an extracellular matrix analogue). The layer thickness was 50 nm for metals and several micrometers for biomaterials according to their higher UV light penetration depths. The following cell types were used for transfer: retinoblastoma cell line (WERI-1), neuroblastoma cell line (SH-SY5Y), primary astroglial cells isolated from forebrain of neonatal rat pups, osteoblastoma cell line (Saos-2) and myeloma cell line (K562). The reason for applying several types of primary and tumor cell lines was to study

130

B. Hopp et al.

the behavior and resistance of the various cell types to femtosecond laser transfer under specific circumstances. The acceptor surface was one of the followings: pure glass, (dry or wet) gelatin-coated glass, Matrigel-coated glass. The distance between donor and acceptor plates was approximately 600 μm. The irradiating laser fluence on a round-shaped 250 μm diameter spot was 225 mJ/cm2 , the lowest value where reproducible material transfer could be achieved, while we could prevent the development of great mechanical forces. After the transfer procedure the cells were kept in culture and their survival, proliferation, growth and differentiation were checked every day. Concerning the suitability of the absorbing layers, all the applied metals were successfully used for transfer. When cells were transferred to untreated or dry gelatin-coated glass surfaces the survival rate of any cell type was very low (1,000 Hz) compared to LGDW. Methods involve the use of optical methods may affect the chromosome stability and should be addressed before implementing such methods in cell patterning and tissue engineering. In addition, the effect of increased temperatures should be addressed to ensure that these methods will not have a negative impact on biologically active materials. In addition to these methods, a photoimmobilization technique has been used for direct writing of patterns using a heterobifunctional photolinker [30, 31]. Although, this method allow patterning gradients of cell adhesive molecules in a direct write fashion the complex chemistry involved has limited the use of this method. In this chapter we describe a simpler method that uses CAD driven direct write technique to achieve cell patterns in two steps [32]. In the first step, a substrate and the patterning polymer will be chosen according to the following criteria, and the polymer will be patterned using aerosol-based precision spraying system as shown schematically in Figs. 11.2, 11.3, and 11.4: (1) Positive Patterning: non-adhesive substrate is patterned with an adhesive biopolymer (2) Negative Patterning: an adhesive substrate is patterned with a non-adhesive polymer In the second step, patterned substrates are used for plating cells (media exchange may be necessary for some cell types). Thus, the use of a direct writing of a polymer provides a simplified and flexible approach to pattern cells. This method eliminated the use of a masks, complex surface chemistry, and expensive intermediate tools. Deposited polymer patterns can be changed easily by writing new CAD programs. Additionally, the cost associated with cell patterning can be reduced due the low

Fig. 11.2 A diagram showing the polymer processing for patterning substrates

11

Patterning Cells on Complex Curved Surface by Precision Spraying of Polymers

191

Fig. 11.3 Polymer solution is ultrasonically atomized to form small aerosol droplets approximately 1–5 μm in diameter with a 107 mm–3 droplet density. The output mist is collimated and combined with annular sheath gas to direct it onto the substrate mounted on a translating stage through a 150 micron-sized orifice with a 2.5·10–4 mm3 /s deposition rate. After patterning, substrates are sterilized using a 70% ethanol solution, air dried in a standard tissue culture hood, and rehydrated with PBS prior to seeding of cells

Fig. 11.4 A schematic representation of (1) positive patterning and (2) negative patterning to obtain cell patterns. Positive – in which an inhibitory substrate such as PDMS was patterned with a focus beam of an aerosolized bio-active adhesive materials such as laminin or polyethylenimine (PEI). Negative – in which an adhesive substrate was patterned with a focus beam of an aerosolized bio-inert inhibitory material such as polytetrafluoroethylene (PTFE) or PDMS

192

M.N. DeSilva

number of required hard tools, reduced processing time and the greater design capability and flexibility (reduced product development cost). Furthermore, this method provides an enhanced cell patterning approach on complex geometrical surfaces and on a variety of material combinations and ranges.

11.2 Experimental Methods 11.2.1 Precision Spraying of Polymers There has been a new class of CAD driven Direct Write Techniques introduced in the recent years as an attractive technology to pattern substrates for the use in the electronics manufacturing and assembling industry. In the Direct Write technique, one can deposit a material of interest on a substrate to generate patterns to build features in electronic devices using computer programs. This chapter will discuss one of the attractive Direct Write Techniques developed by Renn and co-workers, called Maskless Mesoscale Materials Deposition [33], which we called ‘Precision Spraying Technology.’ In this method the patterning polymer is subjected to an aerodynamic focusing for depositing polymer precisely and accurately. A schematic representation of the process is shown in Figs. 11.2, and 11.3. Following are the three main parts in the patterning process: (1) First part of the process is to aerosolize the polymer solution using an ultrasonic or a pneumatic atomizer. Here, micron size aerosol droplets are formed with size ranging approximately 1–5 μm in diameter. The droplet density is approximately 107 /mm3 droplets. (2) The second part of the process is guiding the aerosolized polymer solution towards the deposition head by collimating the droplet mist using a user controlled gas flow. (3) The third part of the process is to combine the aerosol with a user controlled annular sheath gas at the deposition heat to direct the aerosol onto the substrate through an orifice of a focusing nozzle. The size of the orifice and deposition rate will determine the feature size of the lines patterned using this Direct Write technique. In this chapter we will discuss results obtained using a 150 μm diameter orifice with a 2.5 × 10–4 mm3 /s deposition rate. A vector based tool path operated by CAD-computer software, computer controlled 3D motion of stage and a computer controlled shuttering system that controls spraying of the polymer on substrate allows the user to pattern polymer on complex surfaces. The atomization of the polymer needs to be optimized depending on the physical properties of the polymer solution. In this chapter, we will discuss results obtained by ultrasonic aerosolization of the liquid to pattern bio-active polymers (i.e. PEI and laminin) and ultrasonic or pneumatic aerosolization of the polymer solutions to pattern hydrophobic, bio-inert polymers (i.e. PTFE and PDMS).

11


193

11.2.2 Polymer Solutions 11.2.2.1 Bio-Active Polymers We will show results obtained with two different bio-active polymers. Polyethylenimine (PEI) aqueous solution: Embronic chick forebrain neurons were cultured by precision spraying of a 10 mg/ml of 50% (w/v) (Sigma) PEI solution as a cell adhesive molecule. Natural mouse laminin: LLCPK1-α epithelial cells and NIH 3T3 fibroblasts were cultured by precision spraying of 0.02:1 (v:v) ratio of 1 mg/ml laminin in 50 mM Tris HCl and 0.15 M NaCl buffer (Invitrogen Corporation):de-ionized water as a cell adhesive molecule. 11.2.2.2 Polydimethylsiloxane (PDMS) We used Sylguard 184 Silicone Elastomer (Dow Corning, Midland, MI) with a 10:1 (w:w) ratio of elastomer monomer:curing agent as described by the manufacturer. The two components were thoroughly mixed and allowed to mix well for approximately 30 min at room temperature. After spraying of PDMS on substrates, they were stored at room temperature under to cure for at least 5 days. 11.2.2.3 Polytetrafluoroethylene (PTFE) We used commercially available Teflon FEP 121A fluoropolymer resin (Dupont), also known as poly(tetrafluoroethylene) (PTFE) to spray on substrates. Post processing of the patterned substrates was done according to the manufacturer’s protocols. Briefly, the three step curing process of PTFE patterns involved: (1) an initially heating to 107◦ C for removing water; (2) secondary heating step to 250◦ C for removing wetting agents, and; (3) a final heating step to 265◦ C. Then the substrates were allowed to cool down to room temperature. This is approximately a 15 min curing process.

11.2.3 Cell Culturing 11.2.3.1 Embryonic Chick Forebrain Cell Culture Embryonic (E7-E8) chick forebrain cells were cultured as described in Baldi et al. [34]. Briefly, cerebral hemispheres were removed from E7-E8 chick embryos. After incubating tissues at 37◦ C in 0.25% trypsin solution (Gibco) for 15 min, tissues were resuspended in media (F12 media supplemented with 10% fetal bovine serum, 100 units/ml penicillin, 100 μg/ml streptomycin, 250 ng/ml amphotericin B, and 20 mM HEPES, adjusted to pH 7.3, all components from Gibco). The cells in tissues were dissociated by triturating 5 times through a 21-gauge syringe needle. Patterned substrates were used to culture dissociated cells, and media was exchanged approximately after 30 min.

194

M.N. DeSilva

11.2.3.2 LLCPK1-α Epithelial Cell Culture The LLCPK1-α epithelial cells were cultured as described by Rusan et al. [35]. Briefly, patterned substrates were used to culture cells in growth media containing Opti-MEM GlutaMAX, HEPES buffer and sodium bicarbonate (Gibco Invitrogen Corp) with 5% FBS and 1% antibiotic/antimycotic solution (penicillin, streptomycin, and amphotericin B). Media was exchanged approximately after 24 h of plating, and every 1–2 days thereafter. 11.2.3.3 NIH 3T3 Fibroblast Cell Culture NIH 3T3 fibroblast cells (American Type Culture Collection; Manassas, VA) were subcultured according to ATCC standards in an incubator maintained at 37◦ C and 5% CO2. Patterned substrates were used to culture cells in media containing Dulbecco’s Modified Eagle’s Medium (DMEM) with 4.5 mg/ml glucose containing 10% fetal bovine serum (Gibco), 100 U/ml penicillin/streptomycin (Gibco), 1.5 mg/ml sodium bicarbonate, 584 μg/ml L-glutamine, and 110 μg/ml sodium pyruvate (Cellgro Inc.). Media was exchanged approximately after 24 h of plating, every 2–3 days thereafter.

11.2.4 Imaging 11.2.4.1 Light Microscopy A Nikon Eclipse TE200 inverted light microscope was used to observe cells. A 10X, 0.25NA phase contrast objective with a 0.52 NA long working distance condenser was used for verifying pattern formation. Digital images were collected with a Princeton Instruments MicroMax 782-Y cooled CCD camera coupled to Metamorph image processing software. Fluorescent images were acquired using a 100 W mercury arc lamp equipped with a Uniblitz shutter and FITC filter cubes (Chroma Tech., #C 41001). 11.2.4.2 Scanning Electron Microscopy In order to visualize cells using a scanning electron microscope, cells were fixed at room temperature according to the following protocol. The cell culture was prerinsed with Ca++ - and Mg++ - free PBS buffer solution three times to remove excess protein. Cells were fixed using 2% glutaraldehyde (Electron Microscopy Sciences, Hatfield, PA) in 0.1 M cacodylate buffer for 2.5 h. After rinsing the sample with 0.1 M cacodylate buffer twice, the post fixation was done using 1% osmium tetroxide (Electron Microscopy Sciences, Hatfield, PA) in 0.1 M cacodylate buffer on ice for 30 min. Subsequently, the sample was rinsed again two times with 0.1 M cacodylate buffer. Then, the sample was dehydrated sequentially by immersing it in the following order: twice with 50% ethanol for 10 min each, 70% ethanol for 10 min each, twice with 80% ethanol for 5 min each, twice with 90% ethanol for 5 min each, twice with 100% ethanol for 5 min each, and dry absolute ethanol

11


195

for 5 min. Samples were then subjected to critical point drying using a Tousimis Critical Point Dryer – Samdri 780A (Rockville, MD). Samples were then immediately coated with 60% gold and 40% palladium using Denton Vacuum System (DV 502A High Vacuum Evaporator; Moorestown, NJ). Samples were observed using a cold field emission scanning electron microscope (S-4700, Hitachi High Technologies America, Pleasanton, CA).

11.2.5 Positive Patterning of Cells In positive patterning, bioactive polymer patterns were obtained on hydrophobic substrates such as PDMS. Sylguard 184 Silicone Elastomer (Dow Corning, Midland, MI) with a 10:1 (w:w) ratio of elastomer monomer:curing agent were mixed and allowed to sit at room temperature for 30 min. Approximately, 250 mg of the mixture was poured into a 35 mm petri dish. A thin PDMS film of approximately 35 μm was obtained by spinning the dish at 3,000 rpm for 30 s. Again, this coating was allowed to cure for approximately 5 days at room temperature. Patterned substrates were sterilized using a 70% ethanol solution, and air-dried in a standard tissue culture hood. Prior to plating cells, substrates with bioactive polymers were rehydrated with PBS solution. Figure 11.5(a) shows a fluorescence microscope

Fig. 11.5 (a) A fluorescence image showing the patterning of FITC-laminin on PDMS substrate. (b) through (d): Images of cell patterns obtained via positive patterning. Scale Bar 100 μm. This figure was originally published in De Silva et al., Biotechnology and BioEngineering, 5; 93(5): 919–927

196

M.N. DeSilva

image of FITC-laminin patterned on PDMS substrate. Cell patterns for LLCPK1-α epithelial cells and NIH 3T3 fibroblast cells can be obtained in serum containing media within 24 h with a simple media exchange, and embryonic chick forebrain cell patterns were obtained within 30 min of plating as seen in Fig. 11.5(b)–(d). These results are consistent with previous studies of cell patterning via microcontact printing of PDMS substrates [36].

11.2.6 Negative Patterning of Cells In negative patterning, hydrophobic polymers such as PDMS and PTFE were patterned on relatively higher hydrophilic substrates. After patterning, substrates were cured as described in polymer solutions sections. A 70% ethanol solution was used to sterilize the substrates followed by an air-drying step. Prior to cell culture, substrates were rehydrated with PBS solution. PTFE and PDMS patterned glass substrates generate water droplet patterns due to differences in the hydrophility of the substrates and patterned materials as seen in Fig. 11.6(a). NIH 3T3 fibroblast cell patterns obtained using negative patterning method is shown in Fig. 11.6(b)–(d).

Fig. 11.6 (a) A water pattern formed by negative patterning (b) through (d) shows images of cell patterns obtained via negative patterning. This figure was originally published in De Silva et al., Biotechnology and BioEngineering, 5; 93(5):919–927

11


197

These results support the recent report that discussed preferential protein adsorption and cell adhesion on intermediate to high surface tension surfaces [37].

11.2.7 Patterning Cells on Complex Curved Surfaces 11.2.7.1 Complex Curved Surfaces Polymer solutions were patterned on pulled glass micropipettes until it formed a fine point. We used 1 mm O.D. thin wall borosilicate glass capillaries (World Precision Instruments, Inc., Sarasota, FL) and a Model P-30 vertical pipette puller (Sutter Instrument Company, Novato, CA). Then, we made 10 mm size pieces by cutting the pulled micropipettes that included the fine tip. These pieces were glued on to a 15 mm X 15 mm cover slip. After patterning step was completed, a 70% ethanol solution was used to sterilize the substrates followed by an airdrying step. Prior to cell culture, substrates were rehydrated with PBS solution. NIH 3T3 fibroblast cell patterns obtained on this type of curved surface are shown in Fig. 11.7(a) and (b).

Fig. 11.7 (a) and (b) show an example of cell pattern obtained on complex curve. This figure was originally published in De Silva et al., Biotechnology and BioEngineering, 5; 93(5):919–927

11.3 Summary In this chapter we discussed the micropatterning of animal cells using a direct write technique to pattern polymers, and a subsequent cell-plating step. This enabled the entire patterning process to be achieved in two steps. Cellular patterns can be obtained either by positive patterning where an inhibitory substrate is patterned with a bio-active adhesive material or negative patterning where an adhesive substrate is patterned with a bio-inert inhibitory material. Additionally, this method

198

M.N. DeSilva

can be used to pattern cells on complex curved surfaces, such as a tapered glass micropipette. The use of precision spraying of polymers provides a relatively simple method to rapidly micropattern cells. We envision this method to be used as a low-cost mass production of patterned substrates and for rapid prototyping of cell patterns. Interestingly, the precision spraying provided a method obtain cellular patterns on complex curved surfaces that is simpler than previously used microcontact printing method, where the curved substrate needs to be rolled on an inked patterned stamp followed by a wet chemical etch [38]. The ability to pattern irregular surfaces with PS makes it an attractive approach that the complex microcontact printing method.

11.4 Applications, Limitations and Future Directions 11.4.1 Application The precision spraying method described here could be used for: (1) mass production of biologically active surfaces; (2) fabrication of medical devices; (3) facilitation of cell biological research, and; (4) fabrication of biosensors in rapid drug screening or toxicity testing. In addition, the method will facilitate the construction of tissue-based biosensors for bioterrorism and biological warfare defense applications. Furthermore, although this is a simple and flexible rapid prototyping method for achieving cell patterns, the current limitations of the PS method did not allow us to obtain patterns with small feature sizes (i.e. < 10 μm), and so PS is currently not ideally suited for the study of single cell-cell interactions. Nevertheless, we envision that the ability to achieve patterns with sufficient features for performing basic biology research and for biotechnological applications using this simple, flexible rapid prototype approach will allow PS to become a useful tool for biologists and bioengineers.

11.4.2 Limitations of Precision Spraying One limitation of precision spraying is its low edge resolution (~10–15 μm), which is the result of over spraying of the deposited material. Although, over spraying can be controlled by adjusting the gap between the deposition nozzle and the substrate and flow rates, it will be difficult to completely overcome the limitation as the precision spraying is somewhat dependent on the intrinsic properties of the aerosolized and substrate materials. Another drawback of this method is the difficulty of achieving uniform surface coverage due to the mismatch in surface energies of the deposited and substrate materials. Nevertheless, in positive patterning, cells preferred to adhere on the patterned regions avoiding bare PDMS, and in negative patterning, the number of cells

11


199

adhered on the non-permissive regions was minimal, and SEM analysis showed that cells in the non-permissive regions were poorly adherent and tended to round up off the surface. We also found that aerosolizing PDMS was more challenging than aerosolizing PTFE. This difficulty was due to the viscosity difference between the two materials. The viscosity of PDMS is higher than that of PTFE, and so it was difficult to use ultrasonic waves for PDMS atomization. A pneumatic atomization technique was used to form a PDMS aerosol while heating the PDMS solution to around 60◦ C via a heat gun. Although heating PDMS helped decrease the viscosity and promote the aerosolization process, the elevated temperature caused it to start curing during the deposition process, and thus limited the lifetime of the PDMS feed solution. Another drawback, again due to the higher viscosity of PDMS, was that aerosolized PDMS tended to clog the deposition head and reduce the efficiency of patterning since frequent cleaning was required. Despite the advantages, this method has limitations that include restrictions in edge definition due to over spraying and non-uniform adsorption of deposited bioactive polymers due to the mismatch in surface free energies. The former problem may be at least partially solved by optimizing sheath gas flow rate and process gas flow rate for reducing the over spraying. Also, amphiphiles may provide better interfaces to minimize surface energy mismatch.

11.5 Future Directions Precision Spraying of polymers provide an attractive method to pattern polymers. However, this method may need further optimization for improving edge resolution and obtaining uniform surface coverage of patterned material. One can envision the integration of multiple deposition heads and/or multiple patterning materials to extend the capabilities for patterning complex biological tissue structures. The adaptation of this technology to deposit live cells on complex biological structures will provide a method to obtain desired tissue structures and will provide a very attractive tissue engineering capabilities for the medical community. Furthermore, we can investigate the possibility to use this technology for patterning electronic components and biological components simultaneously for the development of biological sensors.

11.6 Disclaimer The opinions expressed in this article are the private views of the author and should not be construed as reflecting official policies of the U.S. Navy, Department of Defense, or the U.S. Government.

200

M.N. DeSilva

11.7 Copyright Statement The authors are military service members or employees of the U.S. Government and/or its contractors. This work was prepared as part of my official duties. Title 17 U.S.C. § 105 provides that ‘Copyright protection under this title is not available for any work of the United States Government.’ Title 17 U.S.C. § 101 defines a U.S. Government work as a work prepared by a military service member or employee of the U.S. Government as part of that person’s official duties. Acknowledgments I would like to thank Dr. David J. Odde (University of Minnesota), Dr. Michael J. Renn (Optomec Design Co.) and Mr. Jason Paulsen (Optomec Design Co.) for their collaboration in this study. Also, I thank Abhinav Arneja, Patricia Wadsworth, and Andrew Bicek for their advice in NIH 3T3 fibroblast cell culture and LLCPK1-α epithelial cell culture, and Chris Frethem and Alice Ressler at the Electron Microscopy Lab at the University of Minnesota for their technical support. Funding for this project was partially provided by NSF-BITS Grant No. EIA0130875 to D.J.O. and through the Microfabricated Neural Networks Interest Group of the Biomedical Engineering Institute at University of Minnesota. This work was originally published in Biotechnology and BioEngineering, 5; 93(5):919–927. The support to format this work into a book chapter was provided under a U.S. Army contract W911QY-08-D-0017-0005 awarded to General Dynamics Information Technology. The writing of this chapter was supported under a U.S. Army contract W911QY-08-D-0017-0005 awarded to General Dynamics Information Technology.

References 1. Letourneau PC (1975) Possible roles for cell-substratum adhesion in neuronal morphogenesis. Dev Biol 44:77–91 2. Hammarback JA, Palm SL, Furcht LT, Letourneau PC (1985) Guidance of neurite outgrowth by pathways of substratum adsorbed laminin. J Neurosci Res 13:213–220 3. Corey JM, Wheeler BC, Brewer GJ (1991) Compliance of Hippocampal neurons to patterned substrate networks. J Neurosci Res 30:300–307 4. Ma W, Liu Q-Y, Jung D, Manos P, Pancrazio JJ, Ascaffner AE, Barker JL, Stenger DA (1998) Central neuronal synapse formation on micropatterned surfaces. Dev Brain Res 111:231–243 5. Corey JM, Wheeler BC, Brewer GJ (1996) Micrometer resolution silane-based patterning of hipocampal neurons:critical variables in photoresist and laser ablation process for substrate fabrication. IEEE Transact Biomed Eng 43:944–954 6. Kleinfeld D, Kahler H, Hockberger PE (1988) Controlled outgrowth of dissociated neurons on patterned substrates. J Neurosci 8:4098–4120 7. Bhatia SN, Yarmush ML, Toner M (1997) Controlling cell interactions by micropatterning in co-cultures: Hepatocytes and 3T3 fibroblasts. J Biomed Mater Res 34:189–199 8. Clark P, Britland S, Connolly P (1993) Growth cone guidance and neuron morphology on micropatterned laminin surfaces. J Cell Sci 105:203–212 9. Weiss P (1945) Experiments on cell and axon orientation in vitro: The role of colloidal exudates in tissue organization. J Exp Zool 100:53–386 10. Clark P, Connolly P, Curtis ASG, Dow JAT, Wilkinson CDW (1991) Cell guidance by ultrafine topographyin vitro. J Cell Sci 99:73–77 11. Webb A, Clark P, Skepper J, Compston A, Wood A (1995) Guidance of oligodendrocytes and their progenitors by substratum topography. J Cell Sci 108:2747–2760 12. Singhvi R, Kumar A, Lopez GP, Stephanopulos GN, Wang DIC, Whitesides GM, Ingber DE (1994) Engineering cell shape and function. Science 264:696–698 13. Chang JC, Brewer GJ, Wheeler BC (2003) A modified microstamping technique enhances polylysine transfer and neuronal cell patterning. Biomaterials 24:2863–2870

11


201

14. Soekarno A, Lom B, Hockberger PE (1993) Pathfinding by neuroblastoma cells is directed by preferential adhesion to positively charged surfaces. Neuroimage 1:129–144 15. Brewer GJ, Deshmane S, Ponnusamy E (1998) Precocious axons and improved survival of rat hippocampal neurons on lysine-alanine polymer substrates. J Neurosci Methods 85:13–20 16. Tai HC, Buettner HM (1998) Neurite outgrowth and growth cone morphology on micropatterned surfaces. Biotechnol Prog 14:364–370 17. Wheeler BC, Corey JM, Brewer GJ, Branch DW (1999) Microcontact printing for precise control of nerve growth in culture. J Biomech Eng 121:73–78 18. Branch DW, Wheeler BC, Brewer GJ, Leckband DE (2000) Long-term maintenance of patterns of Hippocampal pyramidal cells on substrates of polyethylene glycol and microstamped polylysine. IEEE Transact Biomed Eng 47:290–300 19. Irimia D, Karlsson JOM (2003) Development of cell patterning technique using poly(ethylene glycol) Disilane. Biomed Microdevices 5:185–194 20. Gopalan SM, Flaim C, Bhatia SN, Hoshijima M, Knoell R, Chien KR, Omens JH, McCulloch AD (2003) Anisotropic stretch-induced hypertrophy in neonatal ventricular myocytes micropatterned on deformable elastomers. Biotechnol Bioeng 81:578–587 21. Odde DJ, Renn MJ (1999) Laser-guided direct writing for applications in biotechnology. Trends Biotechnol 17:385–389 22. Nahmias YK, Gao BZ, Odde DJ (2004) Dimensionless parameters for the design of optical traps and laser guidance system. Appl Opt 43:3999–4006 23. Marquez GJ, Renn MJ, Miller WD (2002) Aerosol-based direct-write of biological materials for biomedical applications. Mat Res Soc Symp Proc 698:343–349 24. Barron JA, Wu P, Ladouceur HD, Ringeisen BR (2004) Biological laser printing: A novel technique for creating heterogeneous 3-dimensional cell patterns. Biomed Microdevices 6:139–147 25. Wu PK, Ringeisen BR, Krizman DB, Frondoza CG, Brooks M, Bubb DM, Auyeung RCY, Pique A, Sparge B, McGill RA, Chrisey DB (2003) Laser transfer of biomaterials: matrix assisted pulsed laser evaporation (MAPLE) and MAPLE direct write. Rev Sci Instrum 74:2546–2557 26. Chrisey DB, Pique RA, McGill RA, Horwitz JS, Ringeisen BR, Bubb DM, Wu PK (2003) Laser deposition of polymer and biomaterial films. Chem Rev 103:553–576 27. Nahmias YK, Arneja A, Renn MJ, Odde DJ (2005) Cell patterning on biological gels. Tissue Eng 11(5–6):701–708 28. Roth EA, Xu T, Das M, Gregory C, Hickman JJ, Boland T (2004) Inkjet printing for highthroughput cell patterning. Biomaterials 25:3707–3715 29. Calvert P (2001) Inkjet printing for materials and devices. Chem Mater 13:3299–3305 30. Hypolite CL, McLernon TL, Adams DN, Chapman KE, Herbert CB, Huang CC, Distefano MD, Hu W-S (1997) Formation of microscale gradients of protein using heterobifunctional photolinker. Bioconjug Chem 8:658–663 31. Herbert CB, McLernon TL, Hypolite CL, Adams DN, Pikus L, Huang CC, Fields GB, Letourneau PC, Distefano MD, Hu W-S (1997) Micropatterning gradients and controlling surface densities of photoactivable biomolecules on self-assembled monolayers of oligo(ethylene glycol) alkanethiolates. Chem Biol 4:731–737 32. De Silva MN, Paulsen J, Renn MJ, Odde DJ (2005) Two-step cell patterning on planar and complex surface by precision spraying of polymers. Biotechnol Bioeng 5; 93(5):919–927 33. Marquez GJ, Renn MJ, Miller WD (2002) Aerosol-based direct-write of biological materials for biomedical applications. Mat Res Soc Symp Proc 698:343–349 34. Baldi A, Fass JN, De Silva MN, Odde DJ, Ziaie B (2003) A micro-tool for mechanical manipulation of in vitro cell arrays. Biomed Microdevices 5:291–295 35. Rusan NM, Fagerstrom CJ, Yvon AM, Wadsworth P (2001) Cell cycle-dependent changes in microtubule dynamics in living cells expressing green fluorescent protein-alpha tubulin. Mol Biol Cell 12:971–980

202

M.N. DeSilva

36. De Silva MN, Desai RA, Odde DJ (2004) Micro-Patterning of animal cells on PDMS substrates in the presence of serum without use of adhesion inhibitors. Biomed Microdevices 6:219–222 37. Tan JL, Liu W, Nelson CM, Srivatsan R, Chen CS (2004) Simple approach to micropattern cells on common culture substrates by tuning substrate wettability. Tissue Eng 10:865–872 38. Jackman RJ, Wilbur JL, Whitesides GM (1995) Fabrication of submicrometer features on curved substrates by microcontact printing. Science 269:664–666

Chapter 12

Fabrication of Growth Factor Array Using an Inkjet Printer Kohei Watanabe, Tomoyo Fujiyama, Rina Mitsutake, Masaya Watanabe, Yukiko Tazaki, Takeshi Miyazaki, and Ryoichi Matsuda

Abstract Although multiple growth factors influence the fate of cells in vivo, it is technically difficult to reproduce similar condition in vitro. To overcome this problem, we have developed growth factor array, a system to study compound effects of multiple growth factors fabricated with a commercial color inkjet printer. By replacing color inks to 2–4 growth factors and printing them on the tissue culture substratum, we prepared growth factor arrays. Culturing cells on the array, we studied the compound effects of growth factors during myogenic and/or osteogenic differentiation of C2C12 myoblast and mesenchymal stem cells in a single culture dish. The cells grown on the array exhibited various levels of differentiation depending on the dose and the combination of growth factors. Since inkjet printer is capable to manipulate several colors simultaneously, this method is suitable for multivariate analyses of growth factors. This method may provide a powerful tool for regenerative medicine, especially for stem cell research on the control of cell-fate determination and differentiation.

12.1 Introduction In living organisms, cells are constantly under the influence of extrinsic environmental factors such as growth factors, hormones, and extracellular matrices. Proliferation, differentiation, migration, and cell death are controlled by a complicated combination of stimuli from these factors. The mechanisms of these factors in the developmental processes of multi-cellular organisms have not yet been clarified completely, potentially due to the enormous number of factors that have to be examined in order to construct the in vivo conditions in vitro. Conventional studies

K. Watanabe (B) Department of Life Sciences, The University of Tokyo, Tokyo 153-8902, Japan; Canon Inc., Tokyo 146-8501, Japan e-mail: [email protected]


203

204

K. Watanabe et al.

with simple experimental systems considering one or two factors give insufficient analytical results. Researches are actively pursuing methods to convert these stem cells into specific cell types to supplement damaged tissues. For example, it has been reported that marrow stromal cells can differentiate to adipocytes, chondroblasts, osteoblasts, cardiac muscle cells, hepatocytes, and neurons under certain culture conditions [1, 2, 3, 4]. Furthermore, muscle satellite cells, which are usually in quiescent state and activated for muscle regeneration process, can also differentiate to adipocytes, osteocytes, and chondrocytes by treating them with certain growth factors. Undifferentiated mouse embryonic stem (ES) cells have been found to express FGF receptor FGFR1, -R2, and -R3 [5]. The expression of IGF-1 receptors in C3H10T1/2 fibroblast is stimulated by BMP-2 [47]. As exemplified above, multiple receptors are expressed in one cell, suggesting that the stimulations from multiple factors may lead to novel effects which were not yet identified. Avila et al. [6] reported that neural differentiation of PC12 pheochromocytoma cell line was promoted by the presence of nicotine and nerve growth factor (NGF). Zebboudj et al. [7] reported that coexistence of matrix GLA protein and bone morphogenetic protein-2 (BMP-2) promotes calcification of blood vessels. Myogenic differentiation of C2C12 myoblast cell line was enhanced by IGF-I but inhibited by FGF-2 [8]. Furthermore, C2C12 muscle cells exhibit osteogenic differentiation by BMP-2 and IGF-I [9, 10]. All these reports suggest that experiments using a combination of multiple factors, not a single factor, at a wide range of concentrations is necessary. It has been shown that no single growth factor is sufficient for deciding the differentiation fate of ES cells. Schuldiner et al. [11] examined eight different growth factors for human ES cells and reported that none of these directed differentiation to only one cell type, but, rather, altered the relative populations of a specific cell type. Loeser et al. [12] reported that coexistence of insulin-like growth factor-I (IGF-I) and bone morphogenetic protein 7 (BMP-7) stimulates the growth of chondrocytes. Although activin A induces various cell types in amphibian embryonic cells, it is not sufficient to induce phenotypes in murine ES cells [13, 14]. Therefore, the tools for analyzing the compound effect of multiple factors will make it possible to find the optimum conditions for cell differentiation efficiently. In recent years, computer-aided tissue engineering technology called ‘bioprinting’ has emerged [15]. Inkjet technology has drawn attention as a powerful tool for biological research [16]. Inkjet technology is capable of ejecting solutions evenly in any place or area, enabling easy preparation of computergenerated patterns on the substratum. Since inkjet printers can eject tiny droplets at the picoliter scale, it is possible to easily control the amount of deposited substance by changing the number of ejections at any one place. Exploiting this characteristic, studies in DNA microarray preparation [17, 18, 19], protein handling [20, 21], and patterning of cells [22–24] have been reported. Furthermore, inkjet technology can be also used for analyzing the compound effects of multiple factors. We have previously proposed the concept of ‘Growth Factor Array’, a novel cell-based analysis tool. Growth factor arrays are fabricated with a conventional color inkjet

12

Fabrication of Growth Factor Array Using an Inkjet Printer

205

printer by replacing color inks to solutions containing growth factors and printing them on the tissue culture substratum [25]. Since growth factor arrays make it possible to analyze compound effects of multiple growth factors on single culture dish, it can be used to analyze the optimal conditions for differentiation of stem cells. This idea has been adopted by other research groups [26, 27]. In this chapter, we present our work on growth factor array fabrication, giving several experimental examples to prove the concept of the growth factor array. Growth factor arrays composed of 2–4 growth factors were fabricated with several approaches for growth factor immobilization onto the substratum.

12.2 Materials and Methods 12.2.1 Materials Fibroblast Growth Factor-2 (FGF-2), Insulin-like Growth Factor-I (IGF-I), Bone Morphogenetic Protein-2 (BMP-2), Epidermal Growth Factor (EGF), Plateletderived growth factor-BB (PDGF-BB) were purchased from R&D Systems Inc. (Minnesota, USA). [125 I] IGF-I and horseradish peroxidase conjugated anti-mouse IgG antibody were purchased from Amersham Biosciences Corp (New Jersey, USA). Alexa Fluoro 488 conjugated anti-rabbit IgG antibody, Alexa Fluoro 680 conjugated anti-mouse IgG antibody, and TOTO-3 were purchased from Molecular Probes, Inc. (Oregon, USA). Anti-myogenin polyclonal antibody was purchased from Santa Cruz Biotechnology. MF20 hybridoma was purchased from Developmental Studies Hybridoma Bank (Univ. Glowa), and its culture supernatant was used as the anti-mouse striated muscle type myocin heavy chain (MyHC) antibody. Anti-alkaline phosphatase monoclonal antibody was purchased from Biogenesis Ltd. (Poole, UK). IRDye 800 conjugated anti-mouse IgG antibody was purchased from Rockland Immunochemicals, Inc. (Pennsylvania, USA). Dulbecco’s Modified Eagle’s Medium (DMEM) -high glucose type, fetal bovine serum (FBS), Antibiotic-Antimicotic, trypsin-EDTA, bovine insulin, and holotransferrin were purchased from Invitrogen Corp. (California, USA). Bovine fibronectin was purchased from Itoham Foods Inc. (Hyogo, Japan). Other reagents were purchased from SIGMA-Aldrich, unless otherwise noted. Inkjet printers used in this research were BJ F850, PIXUS 950i, Pixus iP8600 (all from Canon, Tokyo Japan). They were used with some modification, as described in the following sections. Mouse myoblast cell line C2C12 was kindly provided from Dr. Yoichi Nabeshima of National Center of Neurology and Psychiatry, Japan. For C2C12 cell culture, DMEM containing 20% FBS was used as growth medium. As the differentiation medium, DMEM containing 10 μg/mL insulin, 5 μg/mL transferrin, 5 nM sodium selenite, 1 mg/mL bovine serum albumin (ITS medium) or DMEM containing 2% FBS was used.

206

K. Watanabe et al.

12.2.2 Growth Factor Array Using Photoreactive Growth Factors Growth factor arrays using photoreactive growth factors were prepared as described below. Photoreactive IGF-I, FGF-2, [125 I] IGF-I were prepared by introducing a phenyl azide group to growth factors, according to the methods described in previous works [28, 29]. The cover of the inkjet printer BJ F850 was removed, and the print head was washed with distilled water. Polystyrene substratum was placed 1 mm beneath the print head. Photoreactive growth factors were injected into the print head, and the growth factors were printed onto the substratum. To increase the density of the growth factor in a certain area, the number of printed droplets in that area was multiplied. After the printing, substratum was UV-irradiated at a distance of 10 cm for 20 s using a mercury lamp (Zeiss HBO 50, 50 W) to immobilize the growth factor on the substratum. Residual growth factors were removed by washing with PBS. With this method, a bifactor growth factor array with 16 areas representing combinations of IGF-I and FGF-2 at various concentrations was prepared. To examine the fidelity of printing and the efficiency of growth factor immobilization at each stage of this process, photoreactive [125 I] IGF-I was prepared and evaluated by measuring the γ-ray intensity. For growth factor array analysis, C2C12 myoblasts were inoculated at 50 cells/mm2 in growth medium, and after 24 h the medium was replaced with ITSmedium as differentiation medium. After a further 24 h of culture, cells were fixed with methanol, and immunostained with anti-myogenin antibody as a marker of myogenic differentiation. Anti-rabbit IgG-Alexa Fluor 488-labeled antibody was used as the secondary antibody. Nuclear DNA was also stained with Hoechst 33258. For comparison with the effect of growth factors in soluble state, combinations of IGF-I and FGF-2 in soluble state were examined at concentrations of 0, 2, 8 and 20 ng/mL, resulting in 16 combinations.

12.2.3 Growth Factor Array with Surface Activated Substratum Substrata with activated surface to immobilize growth factors were prepared by treating the surface of the substrata with tresyl activated dextran according to the previous method [30]. Growth factors dissolved in carbonate buffer (pH 7.2) containing 0.3% glycerin were printed with inkjet printer Pixus 950i on the surface activated substratum. The substratum was incubated at 4◦ C for 16 h in moist conditions. Residual growth factors were removed by washing with PBS, and the remaining active groups were blocked with 0.1% gelatin or fibronectin solution. By this method, growth factor arrays composed of 3 growth factors (IGF-I, FGF-2, BMP-2) (Fig. 12.1) and 4 growth factors (IGF-I, FGF-2, BMP-2, PDGF) were prepared. The efficiency of growth factor immobilization on the surface activated substratum was extrapolated from the measured efficiency of immobilization of BSA. For growth factor array analysis, C2C12 cells were cultured in growth medium for 24 h followed by the medium replacement with the differentiation medium containing 2% FBS. After 4 days cells were fixed with 10% formalin, blocked for

12


207

Fig. 12.1 Preparation of growth factor array using surface activated substratum. (a) Immobilization of growth factors by TAD-treated substratum. (b) Printed patterns of 3 growth factors on the surface activated substratum (pg/mm2 ). (c) Growth factor array with surface activated substratum placed in a culture dish (scale bar: 1 cm)

1 h, and immunostained with either monoclonal anti-ALP antibody as osteogenic marker or anti-MyHC monoclonal antibody as myogenic marker, and further stained with Anti-mouse IgG IRDye 800-labeled antibody. Cells were also stained with TOTO-3 as the indicator of the cell number. The fluorescent intensity from TOTO-3 and differentiation markers were measured with the fluorescent image analyzer Odyssey (LI-COR).

12.2.4 Growth Factor Array in Liquid System In order to compare the effects of growth factors on the surface activated substratum with those in the soluble state, growth factor arrays in liquid system were prepared as described below. Solutions of growth factors were dissolved in distilled water containing 5% glycerin. The solutions were printed with Pixus 950i in the pattern that matches the wells of 96-well plate. After printing, 200 μL of differentiation

208

K. Watanabe et al.

Fig. 12.2 Preparation of growth factor array in liquid system. (a) Schematic diagram of preparation of growth factor array in liquid system using paraffin sheets and 96-well plate. (b) Printed pattern of 3 growth factors on the paraffin sheets (ng/mL)

medium was injected in the wells of 96-well plates and the sheets were touched tightly to the 96-well plate, so that the printed areas corresponded to the wells. The plate was inverted to dissolve the growth factors in the medium, according to the previous method [31] (Fig. 12.2).

12.2.5 Growth Factor Array with Slow-Release System Growth factor arrays with slow-release system were prepared as follows. Photoreactive gelatin was prepared by introducing a phenyl azide group to gelatin molecule in the same manner with the photoreactive growth factors. Photoreactive gelatin solution with and without growth factors dissolved in PBS were infused in the print head of inkjet printer iP8600. First, photoreactive gelatin without growth factors was printed on the substratum and air-dried. Then, solution containing growth factors were printed and air-dried. After printing, substratum was

12


209

UV-irradiated at an intensity of 20 mJ/cm2 using a UV-crosslinker, and washed with distilled water 3 times. The total amount of photoreactive gelatin printed in each area was controlled such that the same amount was deposited in all areas. In order to avoid the attachment of cells outside of the printed areas, substrata were coated with graft copolymer of poly-L lysine and polyethylene glycol (PLL-g-PEG) synthesized according to the previous work [32]. To examine the slow-releasing effect of growth factors from the substratum, 96-well plate was incubated with the mixture of photoreactive gelatin and EGF then UV-irradiated at 0.2, 2, 20 and 200 mJ/cm2 . After the UV-irradiation, the wells were incubated with 200 μL of PBS containing 0.1% BSA at 37◦ C. The supernatant was sampled and the amount of EGF in the supernatant was determined by ELISA. To examine the feasibility of growth factor array with slow-release system, C2C12 cells in growth medium were suspended in differentiation medium and cultured at 2×104 cells/cm2 . After 4 days of culture, the differentiation analysis was performed in the same manner with that of growth factor array with surface activated substratum. MSCs were also cultured with growth factor arrays in a slow-releasing system as follows. MSCs isolated from the femur of a Wister rat (3 weeks, male) were cultured on the growth factor array with slow-release system at 1×104 cells/cm2 in alphaMEM containing 15% FBS, 10–8 M dexamethasone, 10 mM β-glycerophosphate, 50 μg/mL ascorbic acid for 14 days [33]. Medium was changed every 4 days. After 14 days, cells were fixed with 10% formaldehyde and evaluated the bone differentiation by immunostaining with anti-ALP antibody.

12.3 Results 12.3.1 Growth Factor Array with Photoreactive Growth Factors In order to examine the accuracy of growth factor printing with the inkjet printer and the immobilization rate of photoreactive growth factors, photoreactive [125 I] IGF-I was prepared. From the measured intensity of γ-ray from the printed [125 I] IGF-I as well as that remaining on the substratum after washing, it turned out that the printing of solutions is accurate and stable, but the immobilization efficiency decreased as the amount of protein increased. Growth factor arrays consisted of 16 combinations of IGF-I and FGF-2 were prepared. The quantity of immobilized IGF-I and FGF-2 was calculated from the immobilization rate obtained from the previous result are 0, 21, 64, and 149 pg/mm2 and 0, 41, 79, 175 pg/mm2 , respectively. C2C12 cells were cultured on the growth factor array for 48 hr in growth medium, followed by another 24 hr in differentiation medium. In order to compare the effect of growth factors immobilized on the array with those in a soluble state, C2C12 cells were cultured under 16 different conditions corresponding to combinations of four different concentrations (0, 2, 8, 20 ng/mL) of IGF-I and FGF-2. For both experiments, the onset of myogenin

210

K. Watanabe et al.

Fig. 12.3 Effect of IGF-I and FGF-2 on myogenin expression. (a) C2C12 cells were cultured with soluble IGF-I (0, 2, 8, 20 ng/mL) in combination with soluble FGF-2 (0(), 2(), 8(), 20(×) ng/mL). (b) C2C12 cells were cultured on growth factor arrays containing 16 sections consisting of IGF-I (0, 21, 64, 149 pg/mm2 ) and FGF-2 (0(), 41(), 79(), 175(×) pg/mm2 )

expression was examined. When C2C12 cells were cultured on the growth factor arrays, the myogenin expression ratio varied among the different combinations of IGF-I and FGF-2 (Fig. 12.3). However, the myogenin expression pattern was different from that of the soluble growth factors. For soluble growth factors, the effect of IGF-I was remarkably high at 8 ng/mL and approximately 20 ng/mL of IGF-I appeared to be close to saturation. On the growth factor arrays, the ratio of the myogenin expression increased with 21 pg/mm2 of IGF-I, though the higher quantity of IGF-I barely increased its stimulation. In contrast, the suppressive effect of FGF-2 was very clear for soluble FGF-2. However, the effect of the FGF-2 immobilized on the growth factor array seemed to be weaker. Only the concentration of 175 pg/mm2 FGF-2 lowered the myogenin expression ratio.

12.3.2 Growth Factor Array with Surface Activated Substratum 12.3.2.1 Growth Factor Array with 3 Growth Factors The immobilization efficiency of surface activated substratum was evaluated with BSA by ELIZA. The concentrations of growth factors used in this experiment did not exceed the immobilization capacity of activated surface. Hence, the concentrations of growth factors immobilized on the surface activated substratum were estimated from the printed quantity. C2C12 cells were cultured on growth factor arrays fabricated with surface activated substratum. These arrays consisted of combinations of printed FGF-2, IGF-I and BMP-2 (0, 22, 61.8, 200 pg/mm2 , respectively) with 64 areas in all. Cell growth

12


211

or differentiation was significantly promoted on the areas where growth factors were immobilized. To analyze the effects of growth factor combinations for myogenic differentiation, the relative fluorescent intensity of MyHC was obtained by standardizing with fluorescent intensity of TOTO-3 (Fig. 12.3). Without BMP-2, MyHC expression increased with IGF-I dose, but this effect was attenuated by FGF-2. However, when BMP-2 co-existed in the area, especially in high concentrations, the attenuation effect by FGF-2 was unstable. To analyze the effects of growth factor combinations for osteogenic differentiation, the relative fluorescent intensity of ALP was obtained by standardizing fluorescent intensity of TOTO-3, but the relative fluorescent intensity of ALP was weak independent of the presence of BMP-2 (Data not shown). 12.3.2.2 Growth Factor Analysis in Liquid System with 3 Growth Factors Growth factor arrays in liquid system were prepared to compare with the growth factor arrays fabricated with surface activated substratum. First, 3 growth factors, IGF-I (0, 4, and 40 ng/mL), FGF-2 (0, 4 and 40 ng/mL) and BMP-2 (0, 10 and 100 ng/mL) were used to prepare growth factor array in liquid system, resulting in 27 combinations (Fig. 12.2). After culturing C2C12 myoblast with this growth factor array in liquid system, TOTO-3, MyHC, and ALP were immunostained, and the fluorescent intensity was compared in the same way shown before. Relative fluorescent intensity of MyHC standardized by TOTO-3 fluorescent intensity is shown in Fig. 12.5. Without BMP-2, relative fluorescent intensity of MyHC increased as the concentration of IGF-I increase and this effect was attenuated by FGF-2. When BMP-2 coexists, expression of MyHC was suppressed and the relative fluorescent intensity decreased dose dependently. At high concentration (100 ng/mL) of BMP-2, the activity of IGF-I to promote differentiation to myoblasts was almost completely suppressed. 12.3.2.3 Evaluation of Immobilized BMP-2 Since the effect of immobilized BMP-2 was not clear in previous experiments, the effect of BMP-2 at higher concentrations was evaluated independently. BMP-2 was

Fig. 12.4 Myogenic differentiation of C2C12 myoblast cultured on the growth factor array fabricated with surface activated substratum. The relative fluorescent intensity of MyHC was obtained by standardizing with fluorescent intensity of TOTO-3

212

K. Watanabe et al.

Fig. 12.5 Myogenic differentiation on growth factor array in liquid system consisted of 3 growth factors; IGF-I, FGF-2, and BMP-2. Relative fluorescent intensity of MyHC increased as the concentration of IGF-I increase and this effect was attenuated by FGF-2. When BMP-2 coexists, expression of MyHC was suppressed and the relative fluorescent intensity decreased with dosage

Fig. 12.6 The effect of myogenic differentiation by soluble BMP-2(a) and immobilized BMP-2 (b) and the effect of osteogenic differentiation by soluble BMP-2 (c) and immobilized BMP-2 (d)

immobilized on the surface-activated substratum and C2C12 was cultured on that substratum. In this experiments, higher concentration of BMP-2 works inhibitory for MyHC expression and enhanced ALP expression dose dependently, which correspond to the effect of soluble BMP-2 (Fig. 12.6). In this experiment, the highest concentration immobilized on the substratum was 4132 pg/mm2 , corresponding to 536 ng/ml in the soluble state, which is an enormously high concentration compared to the normal experimental condition in vivo or in vitro. At 1542 pg/mm2 , little effect for inhibition for myogenic differentiation was observed, whereas bone differentiation was close to the effect of soluble BMP-2 at 33 ng/ml. These results indicate that the activity of BMP-2 was lowered or changed through the printing and/or immobilization process to the substratum. 12.3.2.4 Growth Factor Array with 4 Growth Factors Growth factor arrays composed of 4 growth factors were also prepared with surface activated substratum (Fig. 12.7a). This growth factor array consisted of 81 areas with combinations of EGF, FGF-2, IGF-I and PDGF, and the compound effects of these factors were examined. These proteins are known to promote the growth of C2C12 and FGF-2 is inhibitory for myogenic differentiation dose-dependently. C2C12

12


213

Fig. 12.7 Analysis of 4 growth factors. (a) The schematic diagram of growth factor array with 4 growth factors (EGF, PDGF-BB, FGF-2, IGF-I) fabricated with surface activated substratum and that in liquid system. (b) The effect of myogenic differentiation analyzed on the growth factor array with 4 growth factors fabricated with surface activated substratum. The concentrations of growth factors are as follows. EGF: 0, 67, 200 pg/mm2 , PDGF-BB: 0, 100, 300 pg/mm2 , FGF-2: 0, 67, 200 pg/mm2 , IGF-I: 0, 133, 400 pg/mm2 . (c) Myogenic differentiation on growth factor array in liquid system. The concentrations of growth factors are as follows. EGF: 0, 15, 44 ng/mL, PDGF-BB: 0, 22, 66 ng/mL, FGF-2: 0, 15, 44 ng/mL, IGF-I: 0, 30, 88 ng/mL

myoblasts were cultured on this growth factor array for 8 h in growth medium, then 72 hr in differentiation medium. After the culture, cell growth and myogenic differentiation was analyzed by staining TOTO-3 and MyHC antibody, respectively (Fig. 12.7b). The results show that cell growth was promoted by IGF-I, EGF, and FGF-2 but not by PDGF-BB. The compound effect for cell growth was similar to the sum of the effect of individual growth factors. Myogenic differentiation was inhibited by EGF but not by FGF-2 and PDGF-BB. 12.3.2.5 Growth Factor Array in Liquid System with 4 Growth Factors Growth factor arrays in a liquid system composed of 4 soluble growth factors were also prepared in the same method. The growth factor pattern is the same as growth factor array in solid system (Fig. 12.7a). As a result, EGF, FGF-2, IGF-1 and PDGF-BB were promotive for growth of C2C12. FGF-2, EGF and PDGF-BB were inhibitory for myogensis (Fig. 12.7c). The inhibitory effect of FGF-2 and EGF was higher than that of immobilized growth factors.

214

K. Watanabe et al.

12.3.3 Growth Factor Array with Slow Release System 12.3.3.1 Evaluation of Slow Release of Growth Factors To evaluate the slow-release of growth factors, EGF was retained on the substrate with photoreactive gelatin with different photo-crosslink conditions. After immobilization of growth factors, the quantity of EGF released to PBS was measured with ELIZA (Fig. 12.8a). EGF release was detected within 24 h at all the conditions. However, when UV-irradiation condition was low (0.2 mJ/cm2 ) or high (200 mJ/cm2 ), the amount of EGF in PBS was constant while EGF amount increased at intermediate condition of UV irradiation (2 and 20 mJ/cm2 ). After 7 days, EGF on the substratum was immunostained with anti-EGF antibody (Fig. 12.8b). Little EGF was detected with the substratum irradiated with 0.2 and 2 mJ/cm2 UV, while retained EGF was still detected at the substratum irradiated with 20 and 200 mJ/cm2 . According to this result, the remaining experiments were operated with a 20 mJ/cm2 UV-irradiation.

Fig. 12.8 Evaluation of slow-release of growth factors. (a) The quantity of released growth factor from the substratum was measured by detecting the released EGF in PBS by ELIZA. (b) Remaining EGF on the substratum after 7 days incubation in PBS detected by immunostaining the substratum with EGF antibody

12.3.3.2 Culture of C2C12 Myoblast with Growth Factor Array in Slow Release System C2C12 myoblast was cultured on the growth factor arrays with slow release system consisting of IGF-I and BMP-2. After the culture, C2C12 was immunostained with TOTO-3 and either with anti-MyHC antibody as a myogenic marker or anti-ALP antibody as an osteogenic marker (Fig. 12.9). Cell growth was not significantly different among each area. However, myogenic differentiation was suppressed by BMP-2 in a dose-dependent manner. At the highest concentration of BMP-2 (450 pg/mm2 ), even the effect of myogenic differentiation by IGF-I was completely suppressed. On the contrary, osteogenic markers were not significantly different among areas (Data not shown).

12


215

A a

BMP-2

IGF-I b

c

Bar=2mm Green : TOTO-3 Red : Myosin Havy Chain

Fig. 12.9 Effect of BMP-2 and IGF-I on C2C12 culture in growth factor array with slow-release system. (a) C2C12 myoblast was cultured for 4 days on growth factor array in slow-release system. a: Merge, b: Fluorescence image of TOTO-3 staining. c: Fluorescence image indicating myogenic differentiation (scale bar: 2 mm). (b) Cells were immunostained with anti-MyHC antibody. Relative intensity of MF20 was measured. At the highest concentration of BMP-2(450 pg/mm2 ), even the effect of myogenic differentiation by IGF-I was completely suppressed

12.3.3.3 MSC Culture on Growth Factor Array with Slow Release System MSCs were cultured on the growth factor arrays with slow release system used in the previous experiment. MSCs were cultured for 14 days, and after the culture, cells were immunostained with osteogenic marker anti-ALP antibody (Fig. 12.10).

A a

BMP-2

IGF-I b

c

Bar=2mm Green㧦TOTO-3 Red㧦alkaline phosphatase (ALP)

Fig. 12.10 MSC culture on growth factor array in slow-release system consisted of BMP-2 and IGF-I. (a) Fluorescent image of MSC cultured on growth factor array. a: Merge, b: Fluorescence image of TOTO-3 staining. C: Fluorescence image indicating osteogenic differentiation (bar: 2 mm). (b) ALP expression of MSC on growth factor array with slow release system (0, 45, 150, 450 pg/mm2 of growth factors were combined) was measured by immunostaining with antiALP antibody. In the areas where BMP-2 was printed at concentrations of 150 and 450 pg/mm2 , osteogenic differentiation was promoted depended on IGF-I concentration

216

K. Watanabe et al.

In this culture, the osteogenic differentiation was not promoted by BMP-2, but when IGF-I coexisted, osteogenic differentiation was promoted, depending on IGF-I concentration.

12.4 Discussion 12.4.1 Growth Factor Array Using Photoreactive Growth Factors Immobilization efficiency of photoreactive growth factors was examined using [125 I] IGF-I. From this experiment, it was shown that the growth factors were printed with high accuracy. The immobilization efficiency decreased as the quantity of protein on the substratum increased. According to previously published work, growth factors immobilized with a phenyl-azido arm may form multi-layers on the substrata [28, 34]. This result suggests that when the quantity of the printed growth factors increases, these growth factor may form multi-layers, preventing a direct reaction with the substratum, and some of the protein may be washed away by rinsing with PBS. To study the feasibility of the growth factor array, we first fabricated the arrays consisting of IGF-I and FGF-2 and cultured C2C12 myoblasts. We also confirmed the effect of soluble IGF-I and FGF-2 for myogenic differentiation. Several laboratories have shown that soluble IGF-I, at concentrations up to 20 ng/mL, promotes myogenic differentiation in a dose-dependent manner [35]. On the other hand, it has also been reported that FGF-2 inhibits expression of the myogenic transcription factor MyoD [8, 9, 36]. In the present work, FGF-2 inhibited the myogenin expression induced by IGF-I in a dose-dependent manner. Interestingly, soluble FGF-2 did not completely inhibit the myogenin expression. Approximately 5% of the cells remained myogenin-positive. According to the previous work, cultured muscle cells synthesize substantial amounts of IGF-I, which may cause the basal myogenin expression [37]. Compared to that of soluble growth factors, the effect of immobilized growth factors were lower. There are two possibilities for this discrepancy. The first possibility is a weakened activity of photoreactive growth factors from introducing the azido-phenyl group. In this method, the azido-phenyl group is supposed to be introduced to the amino groups of the N-terminus or to lysine residues. This chemical modification may cause conformational changes that can reduce the interaction activity with receptors. Another possibility is the restricted flexibility of immobilized growth factors due to the immobilization to the substratum via short arm, which may interfere the interaction with the receptor. Elongating the length of arm may improve the activity of immobilized growth factors.

12.4.2 Growth Factor Array Using Surface Activated Substratum Although the use of photoreactive growth factors was one good way to immobilize growth factors on the substratum, a photoreactive arm needs to be introduced to

12


217

each of the growth factors. To make the fabrication process simpler, immobilization process without modifying growth factors will be necessary. We have prepared the surface activated substratum by treating polystyrene substratum with tresyl activated dextran and we have made the growth factor array with 3 and 4 growth factors. From the results of growth factor arrays using 3 factors, it was confirmed that myogenic differentiation of C2C12 myoblast was promoted by IGF-I, and suppressed by FGF-2, which is similar to the result from growth factor array with photoreactive growth factors. In both cases, promotion of myogenic differentiation by higher concentration of IGF-I was independent of the concentrations of FGF-2 and BMP-2. Ito [38] reported that immobilized insulin bound to a receptor stimulates the cells continuously, resulting in a stronger effect than in dissolved insulin. In this experiment, the effect of IGF-I might be enhanced in the same manner. BMP-2, on the other hand, neither suppressed myogenic differentiation nor promoted osteogenic differentiation, when the immobilized amount was relatively small, close to the concentrations of other factors. However, when higher concentration of immobilized BMP-2 (4,132 pg/mm2 ) was used, osteogenic differentiation was promoted. The density of 4,132 pg/mm2 in the immobilized state corresponds to 536 ng/mL in a soluble state, which is an extremely high concentration. At a concentration of 1,542 pg/mm2 in immobilized system, suppression of myogenic differentiation was not significant, but promotion of osteogenic differentiation was observed at about the same extent as in the liquid system at a concentration of 33 ng/mL. This suggests that the activity of BMP-2 was not simply decreased by immobilization but rather changed. Growth factor arrays composed of four factors (EGF, FGF-2, IGF-1, and PDGF-BB) were also examined. These factors were expected to exert synergic or combinatorial effects, because the downstream signal transduction pathways of these factors are slightly different. However, such effect was not detected. PDGFBB showed lower activity compare to the native effect in liquid system. EGF and FGF-2, on the contrary, showed higher activity in the immobilized state. PDGF-BB at high concentration, however, showed suppression effects on growth and myogenic differentiation and also in the immobilized system, indicating that it is active in the immobilized system, though the activity is low. The low activity of PDGF-BB may be due to its binding mechanism to the receptors. PDGF-BB binds to the receptor as a dimer. It is known that one subunit of PDGF-BB binds to two receptors, whereas one receptor binds to two subunits. Because of this complicated binding, the affinity with the receptor may be lowered in immobilized system. As for EGF, it has been confirmed by previous work that the activity of EGF is higher in immobilized systems than in a liquid system, in agreement with this work [38]. The possible reason for elevated activity in the previously reported immobilized system is that the down regulation of the ligands does not occur in immobilized system, and thus activation of the receptors continues for a long period of time. In some cases, the effect of growth factors in immobilized and liquid systems are different. For example, heparin-binding EGF-like growth factor (HB-EGF), which belongs to the EGF family, acts in two forms in vivo as secretory form (paracrine) and transmembrane form (juxtacrine) which exert different effects. Growth of hepatocyte strain EPI70.7, for instance, is promoted by juxtacrine of HB-EGF [39].

218

K. Watanabe et al.

With paracrine, on the other hand, promotion of growth was not detected at all, although the ligand was considered to be bounded to the receptor. Therefore, it is likely that both the actions of soluble and immobilized systems have different functions in vivo. In fact, in this work, the activity of BMP-2 seemed to have changed in the immobilized system. At some concentrations, suppression of myogenic differentiation was weaker, and osteogenic differentiation was stronger, in comparison with those in soluble state. Many growth factors are down regulated by endocytosis after binding to the receptor. EGF and TGF-β underwent endocytosis, activating signal transduction pathways downstream in endosome [40]. Furthermore, EGF, like other growth factors, activates multiple signal transduction pathways downstream by binding to the receptors, and the degree of activation is different depending on the signal transduction either across the cell membrane or from inside the endosome [41, 42]. For immobilized growth factors, the signal is continuously transmitted on the cell membrane, whereas both paracrine and juxtacrine signals are transmitted in soluble growth factors. Considering this fact, the effect of growth factors in immobilized and liquid systems were different at least in terms of the signal transduction pathway.

12.4.3 Growth Factor Array in Slow Release System As discussed above, immobilizing growth factors on the substratum may change the activity of growth factors. To examine the effect of growth factors more natively, we further constructed the slow-release system for growth factor arrays with photoreactive gelatin as the retaining material. In this slow release system, growth factors are retained to the substratum but not tightly immobilized. According to the result from retaining growth factors on the substratum using photoreactive gelatin, the slow release depends on the intensity of UV-irradiation. When the intensity of UV-irradiation was low (0.2 mJ/cm2 ), most of the EGF was lost after the first wash. On the other hand, when the UV intensity was high (200 mJ/cm2 ), EGF was released in the supernatant for the first 24 h, but much less was released thereafter. Ito et al. reported that release of erythropoietin immobilized by photoreactive gelatin was not observed by treating with UV-irradiation at 160 mJ/cm2 . This implies that growth factors are firmly immobilized to the substratum by photoreactive gelatin with high UV intensity. On the contrary, when the UV intensity was 2 or 20 mJ/cm2 , slow release continued for 1 week. At these intensities, it is possible to minimize the loss of EGF in the first washing, yet keep the ability to release EGF later on. Culture of C2C12 myoblasts in the slow release system resulted in effects similar to that in liquid system. The possible concern about the immobilization of growth factors using photoreactive growth factors or surface activated substratum is the decrease in activity and changes in action of the growth factors. Slow-release systems by photoreactive gelatin may solve these problems. However, the increased effect of growth factors in a slow release system, as reported by Bhang, was not

12


219

observed in our experiments [43]. It may be because the culture duration was too short (4 days) in our experiment. For the growth factor arrays in the slow-release system consisting of BMP-2 and IGF-I, suppression of myogenic differentiation for C2C12 myoblast was confirmed with BMP-2 at the concentration of 450 pg/mm2 , independent of the concentration of IGF-I. We have also prepared the slow-release system in a 96-well plate (data not shown). Compared with the results of slow release culture on a 96-well plate, the amount of BMP-2 per unit area was larger in arrays but suppression effect of myogenic differentiation was lower in the growth factor array. This is most likely due to the released BMP-2 remaining in the 96-well plate, whereas in the growth factor array, it flows out to culture medium, resulting in the decreased effect of BMP-2 on the BMP-2 retained area. Promotion of osteogenic differentiation was not observed as we had expected, neither by BMP-2 alone nor by BMP-2 and IGF-I together. Katagiri et al. stated that, in culture of C2C12 myoblast, BMP-2 suppresses the myogenic differentiation at 100 ng/mL and the promotion of ALP expression can be observed at 300 ng/mL or higher [44]. One possible reason for observing the osteogenic differentiation but not the myogenic differentiation is that the BMP-2 concentration was high enough to suppress myogenic differentiation but not high enough as to promote osteogenic differentiation. The previous work showed that osteogenic differentiation of MSC is not promoted by BMP-2 alone but promoted by BMP-2 and FGF-2 together [45]. In other works, it has been reported that IGF-I induces expression of BMP-2, and BMP2 and IGF-I together promote osteogenic differentiation of C3H10T1/2 cells more effectively than BMP-2 alone [46, 47]. In this work, it was confirmed that BMP-2 and IGF-I together exert compound effect on osteogenic differentiation of MSC.

12.5 Conclusion We have developed techniques to analyze compound effects of growth factors in various states, including immobilized, soluble, and slow release. Inkjet printer technology was highly effective for changing the density and the combinations of immobilized growth factors. Chen and Ito reported the patterning of EGF with concentration gradient using photo lithography by changing the intervals of thin lines so that the number of growth factor molecules touching the cells change [48]. An inkjet printer, however, allows much easier preparation of concentration gradient by simply multiply the printing times. Also, in vitro control of cell patterning such as muscle cell alignment can be achieved easily. By combining various immobilization techniques, inkjet technology makes it possible to construct the conditions similar to in vivo in various states such as soluble and immobilized. By further combining the patterning and various cytokines, growth factor arrays will be a powerful and essential tool for multivariable analysis for tissue engineering and regenerative medicine, especially for the optimization of differentiation conditions for pluripotent cells to differentiate to various types of cells.

220

K. Watanabe et al.

References 1. Deng W, Obrocka M, Fischer I, Prockop DJ (2001) In vitro differentiation of human marrow stromal cells into early progenitors of neural cells by conditions that increase intracellular cyclic AMP. Biochem Biophys Res Commun 282:148–152 2. Petersen BE, Bowen WC, Patrene KD, Mars WM, Sullivan AK, Murase N, Boggs SS, Greenberger JS, Goff JP (1999) Bone marrow as a potential source of hepatic oval cells. Science 284:1168–1170 3. Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R, Mosca JD, Moorman MA, Simonetti DW, Craig S, Marshak DR (1999) Multilineage potential of adult human mesenchymal stem cells. Science 284:143–147 4. Prockop DJ (1997) Marrow stromal cells as stem cells for nonhematopoietic tissues. Science 276:71–74 5. Mummery CL, van Rooyen M, Bracke M, van den Eijnden-van Raaij J, van Zoelen EJ, Alitalo K (1993) Biochem Biophys Res Commun 191(1):188–195 6. Avila AM, Davila Garcia MI, Ascarrunz VS, Xiao Y, Kellar KJ (2003) Differential regulation of nicotinic acetylcholine receptors in PC12 cells by nicotine and nerve growth factor. Mol Pharmacol 64:974–986 7. Zebboudj AF, Shin V, Bostrom K (2003) Matrix GLA protein and BMP-2 regulate osteoinduction in calcifying vascular cells. J Cell Biochem 90(4):756–765 8. Florini JR, Magri KA (1989) Effects of growth factors on myogenic differentiation. Am J Physiol 256:C701–C711 9. Florini JR, Ewton DZ, Magri KA (1991) Hormones, growth factors, and myogenic dfferentiation. Annu Rev Physiol 53:201–216 10. Florini JR, Ewton DZ, Coolican SA (1996) Growth hormone and the insulin-like growth factor system in myogenesis. Endocr Rev 16:481–517 11. Schuldiner M, Yanuka O, Itskovitz-Eldor J, Melton DA, Benvenisty N (2000) Effects of eight growth factors on the differentiation of cells derived from human embryonic stem cells. Proc Natl Acad Sci USA 97(21):11307–11312 12. Loeser RF, Pacione CA, Chubinskaya S (2003) The combination of insulin-like growth factor 1 and osteogenic protein 1 promotes increased survival of and matrix synthesis by normal and osteoarthritic human articular chondrocytes. Arthritis Rheum 48:2188–2196 13. Ma R, Latif R, Davies T (2009) Thyrotropin-independent induction of thyroid endoderm from embryonic stem cells by Activin A. Endocrinology 150:1970–1975 14. Sulzbacher S, Schoroeder IS, Truong TT, Wobus AM (2009) ActivinA-induced differentiation of embryonic stem cells into endoderm and pancreatic progenitors – The influence of differentiation factors and culture conditions. Stem Cell Rev 5:159–173 15. Mironov V, Reis N, Derby B (2006) Review: bioprinting: a beginning. Tissue Eng 12(4): 631–634 16. Alper J (2004) Bioengineering. Biology and the inkjets. Science 305(5692):1895 17. Allain LR, Askari M, Stokes DL, Vo-Dinh T (2001) Microarray sampling-platform fabrication using bubble-jet technology for a biochip system. Fresenius J Anal Chem 371(2):146–150 18. Goldmann T, Gonzalez JS (2000) DNA-printing: utilization of a standard inkjet printer for the transfer of nucleic acids to solid supports. J Biochem Biophys Methods 42:105–110 19. Okamoto T, Suzuki T, Yamamoto N (2000) Microarray fabrication with covalent attachment of DNA using Bubble Jet technology. Nature Biotech 18:438–441 20. Nilsson S, Lager C, Laurell T, Birnbaum S (1995) Thin-Layer immunoaffinity chromatography with bar code quantitation of C-reactive protein. Anal Chem 67:3051–3056 21. Roda A, Guardigli M, Russo C, Pasini P, Baraldini M (2000) Protein microdeposition using a conventional ink-jet printer. Bio Techniques 28:492–496 22. Klebe RJ (1988) Cytoscribing: a method for micropositioning cells and the construction of two- and three-dimensional synthetic tissues. Exp Cell Res 179:362–373 23. Sanjana NE, Fuller SB (2004) A fast flexible ink-jet printing method for patterning dissociated neurons in culture. J Neurosci Methods 136(2):151–163

12


221

24. Xu T, Gregory CA, Molnar P, Cui X, Jalota S, Bhaduri SB, Boland T (2006) Viability and electrophysiology of neural cell structures generated by the inkjet printing method. Biomaterials 27(19):3580–3588 25. Watanabe K, Miyazaki T, Matsuda R (2003) Growth factor array fabrication using a color ink jet printer. Zool Sci 20:429–434 26. Campbell PG, Weiss LE (2007) Tissue engineering with the aid of inkjet printers. Expert Opin Biol Ther 7(8):1123–1127 27. Phillippi JA, Miller E, Weiss L, Huard J, Waggoner A, Campbell P (2008) Microenvironments engineered by inkjet bioprinting spatially direct adult stem cells toward muscle- and bone-like subpopulations. Stem Cells 26:127–134 28. Ito Y, Chen G, Imanishi Y (1996) Photoimmobilization of insulin onto polystyrene dishes for protein-free cell culture. Biotechnol Prog 12:700–702 29. Matsuda T, Sugawara T (1995) Photochemical protein fixation on polymer surfaces via derivatized phenyl azido group. Langmuir 11:2272–2276 30. Gregorius K, Mouritsen S, Elsner HI (1995) Hydrocoating: a new method for coupling biomolecules to solid phases. J Immunol Methods 181:65–73 31. Piwowarczyk W, Matsuda R (1990) A large scale dot blot ELISA using the 96-well culture plate. J Immunol Methods 132(1):147–149 32. VandeVondele S, Voros J, Hubbell JA (2003) RGD-Grafted Poly-L-lysine-graft-(polyethylene glycol) copolymers block non-specific protein adsorption while promoting cell adhesion. Biotechnol Bioeng 82(7):784–790 33. Lennon DP, Haynesworth SE, Dennis JE, Caplan AI (1995) A chemically defined medium supports in vitro proliferation and maintains the osteochondral potential of rat marrow-derived mesenchymal stemcells. Exp Cell Res 219:211–222 34. Ito Y, Li JS, Takahashi T, Imanishi Y, Okabayashi Y, Kido Y, Kasuga M (1997) Enhancement of the mitogenic effect by artificial juxtacrine stimulation using immobilized EGF. J Biochem 121:514–520 35. Milasincic DJ, Calera MR, Farmer SR, Pilch PF (1996) Stimulation of C2C12 myoblast growth by basic fibroblast growth factor and insulin-like growth factor 1 can occur via mitogen-activated protein kinase-dependent and -independent pathways. Mol Cell Biol 16(11):5964–5973 36. Vaidya TB, Rhodes SJ, Taparowsky EJ, Konieczny SF (1989) Fibroblast growth factor and transforming growth factor beta repress transcription of the myogenic regulatory gene MyoD1. Mol Cell Biol 9(8):3576–3579 37. Tollefsen SE, Sadow JL, Rotwein P (1989) Coordinate expression of insulin-like growth factor II and its receptor during muscle differentiation. Proc Natl Acad Sci USA 86:1543–1547 38. Ito Y (1998) Tissue engineering by immobilized growth factors. Mater Sci Eng C 6(4): 267–274 39. Higashiyama S, Iwamoto R, Goishi K, Raab G, Taniguchi N, Klagsbrun M, Mekada E (1995) The membrane protein CD9/DRAP 27 potentiates the juxtacrine growth factor activity of the membrane-anchored heparin-binding EGF-like growth factor. J Cell Biol 128:929–938 40. Le Roy C, Wrana JL (2005) Clathrin- and non-clathrin-mediated endocytic regulation of cell signalling. Nat Rev Mol Cell Biol 6:112–126 41. Vieira AV, Lamaze C, Schmid SL (1996) Control of EGF receptor signaling by clathrinmediated endocytosis. Science 274:2086–2089 42. Wang Y, Pennock S, Chen X, Wang Z (2002) Endosomal signaling of epidermal growth factor receptor stimulates signal transduction pathways leading to cell survival. Mol Cell Biol 22:7279–7290 43. Bhang SH, Jeon O, Choi CY, Kwon YH, Kim BS (2007) Controlled release of nerve growth factor from fibrin gel. J Biomed Mater Res A 80(4):998–1002 44. Katagiri T, Yamaguchi A, Komaki M, Abe E, Takahashi N, Ikeda T, Rosen V, Wozney JM, Fujisawa-Sehara A, Suda T (1994) Bone morphogenetic protein-2 converts the differentiation pathway of C2C12 myoblasts into the osteoblast lineage. J Cell Biol 27:1755–1766

222

K. Watanabe et al.

45. Hanada K, Dennis JE, Caplan AI (1997) Stimulatory effects of basic fibroblast growth factor and bone morphogenetic protein-2 on osteogenic differentiation of rat bone marrow-derived mesenchymal stem cells. J Bone Miner Res 12:1606–1614 46. Li H, Bartold PM, Zhang CZ, Clarkson RW, Young WG, Waters MJ (1998) Growth hormone and insulin-like growth factor I induce bone morphogenetic proteins 2 and 4: a mediator role in bone and tooth formation? Endocrinology 139:3855–3862 47. Raiche AT, Puleo DA (2004) In vitro effects of combined and sequential delivery of two bone growth factors. Biomaterials 25(4):677–685 48. Chen G, Ito Y (2001) Gradient micropattern immobilization of EGF to investigate the effect of artificial juxtacrine stimulation. Biomaterials 22:2453–2457

Part VI

3-Dimensional Scaffold Cell Printing

Chapter 13

3D-Fiber Deposition for Tissue Engineering and Organ Printing Applications N.E. Fedorovich, L. Moroni, J. Malda, J. Alblas, C.A. van Blitterswijk, and W.J.A. Dhert

Abstract Scaffold manufacturing technologies are moving towards systems that combine a precise control over scaffold 3-dimensional (3D) architecture with incorporation of cells in the fabrication process. A rapid prototyping technology, termed 3D-fiber deposition, generates 3D scaffolds that can satisfy these requirements, while maintaining a completely open porosity that can reduce nutrient diffusion limitations. This extrusion-based technique is effective in fabrication of porous thermoplastic scaffolds that function as instructive templates for seeded cells and can also be used to build viable tissue equivalents by layered deposition of cell-laden hydrogels. Therefore, this technology could accelerate and improve the assembly and functionality of tissue-engineered constructs.

13.1 Background Rapid Prototyping (RP) techniques are powerful tools to fabricate scaffolds for tissue engineering applications. These model-based technologies build porous 3-dimensional (3D) scaffolds in a controlled manner layer-by-layer, through material deposition on a stage [1, 2]. During fabrication, the layers are deposited as interpenetrating fiber networks (materials and voids), resulting in 3D structures that are per definition porous, interconnected scaffolds. A controlled, completely accessible pore network of the scaffolds ensures adequate oxygen and nutrient supply, and supports tissue ingrowth upon in vivo implantation [3, 4]. Such scaffolds can

N.E. Fedorovich (B) Department of Orthopaedics, University Medical Center Utrecht, 3508 GA, Utrecht, The Netherlands e-mail: [email protected] L. Moroni (B) Institute for BioMedical Technology (BMTI), University of Twente, 7500 AE, Enschede, The Netherlands e-mail: [email protected]


225

226

N.E. Fedorovich et al.

be built with a customized external shape and internal morphology by computeraided design and computer-aided manufacturing (CAD/CAM) technologies. This results in 3D matrices with controlled mechanical properties matching those of the tissue to be replaced. Furthermore, acquisition of computer tomography or magnetic resonance imaging datasets, used as input data, enables fabrication of anatomically shaped scaffolds. This flexibility and versatility in creating scaffolds presents the opportunity to use rapid prototyping devices to generate improved scaffolds for tissue engineering and to study the effects of different structural phenomena on tissue regeneration [5]. Recently, additive manufacturing by rapid prototyping has been adapted for deposition of cell-laden matrices, with the aim of creating living tissue equivalents [6]. This approach, termed organ- or tissue printing, enables defined deposition of multiple cell populations, potentially mimicking natural cell arrangement. The rationale is that design strategies that more closely mimic the anatomical organization of cells, matrix and bioactive molecules can help to unravel the mechanisms underlying cell-cell and cell-matrix interactions and could contribute to engineering functional grafts for tissue regeneration. Among RP technologies, 3D-fiber deposition (3DF) has been used recently for skeletal tissue engineering, while offering solutions also for a wide range of potential applications. 3DF is an extrusion-based technology able to process melted thermoplastic polymers [7] or (cell-laden) hydrogels [8, 9] (Fig. 13.1). When using thermoplastic polymers, 3DF works similarly to fused deposition modeling (FDM): considering it as an XYZ robotic unit, polymers are put in a stainless steel syringe and heated through a thermosetting cartridge unit. When the melted phase is attained, nitrogen pressure is applied to the syringe through a pressurized cap to extrude the polymers on a stage as a fiber, obtained layer-by-layer from a specific computer-aided design (CAD) model. Extrusion pressure and deposition speed are typically optimized, depending on the polymer viscosity and on the extrusion needle dimensions. Fabrication of hydrogel scaffolds proceeds analogously, and was first reported by Landers et al. [10]. Hydrogel materials are inserted in disposable plastic syringes and extruded, either volumetrically or pressuredriven, on a stationary stage. A temperature-regulating water-filled metal jacket surrounds the plastic syringe enabling hydrogel deposition at physiological temperatures, followed either by reactive plotting (polymer solution is deposited in a crosslinker solution and forms a hydrogel) or by cross-linking of the scaffold after extrusion. The architecture of resulting 3D scaffolds is characterized by parameters such as the fiber diameter (dependent on the nozzle diameter, the deposition speed, and the extrusion pressure), the spacing between fibers in the same layer, the layer thickness and the angle of fiber deposition in different layers (Fig. 13.2). In this chapter, we present some of the most successful attempts where 3DF has been used to fabricate 3D scaffolds for tissue engineering and organ printing applications. Future perspectives on developments and clinical translation of 3DF and scaffold-based tissue engineering strategies are also discussed.

13

3D-Fiber Deposition for Tissue Engineering and Organ Printing Applications

227

Fig. 13.1 3D-fiber deposition principle; schematic drawing of the components (a); main components include a positional control unit linked to a computer containing CAD software, a thermostatically controlled heating jacket, a volumetric- or pneumatic dispensing unit with a syringe containing melted polymer or (cell-laden) hydrogel and a nozzle; adapted from Woodfield et al. [7] with permission of the author; fabrication of thermoplastic (b) and hydrogel scaffolds (c)

13.2 Deposition of Polymers As thermoplastic polymers, poly(ε-caprolactone) (PCL), poly(lactic acid) (PLA), poly(glycolic acid) (PGA), poly[poly(ethylene oxide) terephthalate-co-(butylene) terephtalate] (PEOT/PBT), and starch-based polymers and co-polymers have been most abundantly used for the fabrication of 3D scaffolds for hard and soft tissue engineering applications [11–14]. For example, the Hutmacher group successfully demonstrated fabrication of PCL-based 3D scaffolds supporting bone and cartilage

228


Fig. 13.2 Models of (a) solid and (b) hollow fiber 3D scaffolds and SEM pictures showing different (c–f) pore shapes and (g–j) layer configurations. Fiber deposition angle can be changed (g) every single layer or (h–j) every two or more layers, thus resulting in a different pore height. Fibers with the same orientation can be deposited (h) in the same position or (i) displaced to create a ladder structure facilitating cell attachment; scaffolds’ pore network is characterized by fiber diameter (d1), fiber spacing (d2), and layer thickness (d3); scale bars: (C, G, I, J) 1 mm; (e, f) 2 mm; (d, h) 500 μm; reproduced from Moroni et al. [52] with permission of the publisher

tissue formation both in vitro and in vivo [15–17]. In addition, Rucker et al. [4] and Laschke et al. [18, 19] showed that PLGA-based 3DF scaffolds alone or in combination with growth factor containing gels sustain microvasculature ingrowth enhancing biological performance of the tissue construct. Our group has been focusing on PEOT/PBT copolymers as a model polymer to create 3DF scaffolds that not only can mimic the mechanical properties of targeted tissues to regenerate [7, 20], but can also display smart surface properties to enhance cell-material interactions and tissue development [14, 21]. Specifically, by varying scaffolds’ architectural and structural parameters like total porosity, pore size and shape, and fiber dimension and thickness (Fig. 13.2), it was possible to generate a mathematical model able to predict the mechanical properties of the manufactured 3DF scaffolds depending on specific PEOT/PBT co-polymer compositions [22]. Furthermore, biphasic scaffolds with a shell-core fiber architecture were fabricated by exploiting viscous encapsulation of polymeric blends, where a PEOT/PBT copolymer with adequate mechanical properties to mimic articular cartilage comprised the core fiber and a PEOT/PBT co-polymer favorable to cartilage regeneration comprised the shell coating of the core fiber [14]. The same principle could also be applied to immiscible polymeric blends where the core polymer is selectively removed in a solvent-non-solvent bath to create 3D hollow-fibrous scaffolds that may find applications in neural and vascular regeneration [23]. A further effort towards designing

13

3D-Fiber Deposition for Tissue Engineering and Organ Printing Applications

229

and fabricating hierarchical scaffolds for complex tissue regeneration was aimed towards the integration of different RP platforms, namely 3DF and stereolithography, to generate polymeric-ceramic assembled hybrid constructs for osteochondral regeneration. When implanted for a month in mice these constructs show regeneration of cartilaginous and bone tissues in the scaffolds’ compartment optimized for these tissues [24]. The Reis group has further elaborated on architectural organization with a similar approach to produce hierarchically organized starch-based 3DF scaffolds and simple effective tubular constructs for bone and spine injury treatment, respectively [11, 25]. In a recent development, extrusion-based tools commonly include multi-dispensing feature that makes it possible to create scaffolds with different biomaterials. Such scaffolds can display different physicochemical properties within the same construct [26]. The structural and architectural characteristics of 3DF polymer scaffolds can be varied to obtain constructs with tailored physicochemical and mechanical properties and with tunable porosity (Fig. 13.3). A completely interconnected porous network of these scaffolds results in improved nutrient diffusion and cell viability in the central areas of tissue engineered constructs compared to conventional porogenand mold-based 3D scaffolds [3, 27, 28]. As previously mentioned, anatomically

Fig. 13.3 3D scaffolds seeded with cells showing proliferation (a; actin staining (phalloidin), with the courtesy of Dr. Gawlitta) and viability of multipotent stromal cells (MSCs) (b; live/dead assay) and chondrocytes (c; live/dead assay; scale bar=500 μm); d–i: anatomical scaffolds can be fabricated from μCT or MRI datasets showing high fidelity between the model (d–f) and the manufactured structure (g–i); reproduced from Moroni et al. [2] with permission of the publisher

230


shaped scaffolds can be fabricated from computed tomography or magnetic resonance imaging datasets offering the possibility of customized solutions for specific clinical cases, such as the reconstruction of a tracheal segment [29]. 3D scaffolds with femoral and tibial plateau articular shape implanted in rabbit knees successfully support cartilage regeneration and restore joint functionality [30]. When anatomical scaffolds are used, improved cartilage tissue regeneration can be expected from computational modeling analysis [22]. These predicted results were confirmed by in vitro studies [29]. Although 3DF techniques demonstrated to considerably improve the quality of tissue-engineered constructs [13, 15, 31–33], the high temperatures involved during fabrication of melted polymers remain still critical and limit the possibilities for direct incorporation of biological factors to enhance the scaffolds’ bioactivity. A solution could be envisioned if not only metallic [34] or ceramic pastes [35, 36], but also polymeric pastes that can be processed at room temperature will be developed. Alternatively, surface modification techniques could be used to functionalize the fibers and allow grafting of the bioactive agents at specific sites [37, 38]. Yet, hydrogels remain so far the only biomaterial class able to support biological factor incorporation during fabrication. We have explored different hydrogels as biomaterials to encapsulate cells for tissue regeneration and organ printing [8, 39]. These results will be discussed in more detail in the following sections.

13.3 Deposition of Hydrogels Dispensing techniques such as 3DF can process a broad range of materials including hydrogels. These water-swollen polymer networks maintain a distinct 3D mesh structure by virtue of crosslinks that are either covalent bonds or physical interactions [40, 41]. Hydrogel matrices are typically extruded by 3DF at pressures of 1.5–4 bars through tapered conical needles to minimize thixotropic (i.e. becoming less viscous upon extrusion due to shear-thinning) behavior. Hydrogels vary greatly in their ease of processing with 3DF due to differences in gelation rate [8]. If the hydrogel formation is too slow (sec-min), the printed fibers and the resultant scaffolds will sag and lose their 3D shape. If the gel formation is too fast (mm, making individual spot resolutions irrelevant. However, if bioprospecting is to be truly realized, high throughput printing and identification methods will be necessary, making small spot resolution a necessary requirement.

14.4.2 Potential Applications There are several potential applications of bacterial cell printing, and we will describe several below. Carbon nanotubes (CNTs) are an avenue to fabricate thin films composed of materials with stronger and faster carrier mobilities and smaller thickness. These characteristics are essential for use in microelectronic and nanoelectronic applications. However, one obstacle in the fabrication of single walled carbon nanotubes (SWCNT) by the HiPco (High-pressure CO dissociation) process is the amount of iron impurities contained in the CNTs due to the use of iron as a catalyst in the process. One group at the California State University Long Beach (T. Mangir et al.) has utilized bacterial inkjet printing as a way to remove the iron impurities from CNTs without affecting the condition and/or structure. Iron is a main nutrient for bacteria to produce nitrogen. Therefore, the biological cleaning process takes advantage of this process by growing the bacterium in a starving environment (lack of iron nutrients) and then subjecting the bacterium to a source of iron (i.e. an environment containing only the raw CNT which has the iron impurities). The absorption of iron begins when the bacterium reaches the contaminated CNT, thus removing the iron impurities. After the removal of iron from the CNTs, the next step is to separate the individual SWCNTs from the bacterium. The process is achieved by using inkjet technology as well where the material is converted from a wet environment (bacterial culture) back to a dry environment. There are many complications which arise when clinical bacteriologists try to cultivate a bacterium isolated from a human infection under standard laboratory conditions. For example, they often loose certain exhibited characteristics which are seen in human growth (i.e. polymeric ultrastructures). Since printing of bacteria has been shown to be a gentle method which allows for a precise positioning of cells, this

254


method would allow observations of an individual bacterium from a human isolate without extensive culturing or processing. Furthermore, printing of bacteria could be done in a three dimensional spacing with could allow for mechanical probing by AFM of a single bacteria. The identification of microbial populations within environmental samples is no longer a novel microbiological technique. However, the ability to rapidly isolate culturable microbes from an environmental sample, which could contain tens or hundreds of different types of microorganisms, is an untapped frontier. One traditional microbiological method for the isolation of single colonies from mixed cultures incorporates the plating of serial dilutions [36–38]. By varying the type of nutrient plate used for isolation, unwanted populations can be eliminated while the growth of desired species can be enhanced [39, 40]. By its nature, dilution can only be used to isolate major components in environmental samples and is not suited to isolate low population density species. As demonstrated by the examples shown above, cell printers can isolate droplets containing a single bacterium. Operating under these conditions, a cell printer could be used to isolate single species of microorganisms from infinitely complex and diverse samples. By forming arrays of single cells, cell printers could be used to separate and isolate unknown species from complex environmental samples in a high throughput manner, avoiding the pitfalls of traditional dilution and plating experiments, e.g. laborous and exclude low population species from discovery. High-density bacterial colony arrays are routinely used for the construction of genomic and expression libraries. These arrays are used for high-throughput screening of thousands of bacteria to investigate gene expression, identify specific DNA sequences or to search for differentially expressed genes. There is an increasing need for suitable methods to create bacterial patterns to build bacterial biosensors for environmental monitoring, detection of toxicological contamination and applications in the area of public defense. Bacterial cell printing could also be used to form cell density gradients. These gradient could be valuable in pharmacological screening, where toxicity and drug effectiveness assays require testing on varying densities of cells.

14.5 Summary Several technologies have now shown that it is possible to print viable and functional bacteria. Stretching this capability to the single cell limit raises interesting possiblities for bioprospecting and high throughput assay for pharmacuetical, environmental and medical applications. Due to the gentle depositions and fast printing capabilities, all the demonstrated printing techniques could find niche applications in different areas. The field of bacterial printing is nearly a decade old, but recent research has shown the growing potential in a number of interesting basic and applied fields.

14

Bacterial Cell Printing

255

References 1. Ringeisen BR, Othon CM, Barron JA et al (2006) Jet-based methods to print living cells. Biotechnol J 1:930–948 2. Mironov V, Boland T, Trusk T et al (2003) Organ printing: computer-aided jet-based 3d tissue engineering. Trends Biotechnol 21:157–161 3. Boland T, Xu T, Damon B et al (2006) Application of inkjet printing to tissue engineering. Biotechnol J 1:1910–1917 4. Hood BL, Darfler MM, Guiel TG et al (2005) Proteomic analysis of formalin-fixed prostate cancer tissue. Mol Cell Proteomics 4:1741–1753 5. Barron JA, Krizman David B, Ringeisen Bradley R (2005) Laser printing of single cells: statistical analysis, cell viability, and stress. Ann Biomed Eng 33:121–130 6. Mongkoldhumrongkul N, Best S, Aarons E et al (2009) Bio-electrospraying whole human blood: analysing cellular viability at a molecular level. J Tissue Eng Regen Med 3:562–566 7. Clarke JDW, Jayasinghe SN (2008) Bio-electrosprayed multicellular zebrafish embryos are viable and develop normally. Biomed Mater 3:4 8. Ringeisen BR, Chrisey DB, Pique A et al (2002) Generation of mesoscopic patterns of viable escherichia coli by ambient laser transfer. Biomaterials 23:161–166 9. Barron JA, Rosen R, Jones-Meehan J et al (2004) Biological laser printing of genetically modified escherichia coli for biosensor applications. Biosensors Bioelectron 20:246–252 10. Wang X, Yan Y, Pan Y et al (2006) Generation of three-dimensional hepatocyte/gelatin structures with rapid prototyping system. Tissue Eng 12:83–90 11. Yan Y, Wang X, Pan Y et al (2005) Fabrication of viable tissue-engineered constructs with 3d cell-assembly technique. Biomaterials 26:5864–5871 12. Xu T, Gregory CA, Molnar P et al (2006) Viability and electrophysiology of neural cell structures generated by the inkjet printing method. Biomaterials 27:3580–3588 13. Jayasinghe S, Townsend-Nicholson A (2006) Bio-electrosprays: the next generation of electrified jets. Biotechnol J 1:1018–1022 14. Townsend-Nicholson A, Jayasinghe S (2006) Cell electrospinning: a unique biotechnique for encapsulating living organisms for generating active biological microthreads/scaffolds. Biomacromolecules 7:3364–3369 15. Barron JA, Wu P, Ladouceur HD et al (2004) Biological laser printing: a novel technique for creating heterogeneous 3-dimensional cell patterns. Biomed Microdevices 6:139–147 16. Othon CM, Wu X, Anders JJ et al (2008) Single-cell printing to form three-dimensional lines of olfactory ensheathing cells. Biomed Mater 3:034101 17. Ringeisen BR, Kim H, Barron JA et al (2004) Laser printing of pluripotent embryonal carcinoma cells. Tissue Eng 10:483–491 18. Hopp B, Smausz T, Kresz N et al (2005) Survival and proliferative ability of various living cell types after laser-induced forward transfer. Tissue Eng 11:1817–1823 19. Chen CY, Barron JA, Ringeisen BR (2006) Cell patterning without chemical surface modification: cell–cell interacftions between bovine aortic endothelial cells (baec) on a homogeneous cell-adherent hydrogel. Appl Surf Sci 252:8641–8645 20. Roth EA, Xu T, Das M et al (2004) Ink-jet printing for high-throughput cell patterning. Biomaterials 25:3707–3715 21. Xu T, Petridou S, Lee EH et al (2004) Construction of high-density bacterial colony arrays and patterns by the ink-jet method. Biotechnol Bioeng 85:29–33 22. Ringeisen BR, Lizewski SE, Fitzgerald LA, Biffinger JC, Knight CL, Crookes-Goodson WJ, Wu PK (2010) Single cell isolation of bacteria from microbial fuel cells and potomac river sediment. Electroanalysis 22(7–8):875–882 23. Pardo L, Boland T (2003) Characterization of patterned self-assembled monolayers and protein arrays generated by teh ink-jet method. Langmuir 19:1462–1466

256


24. Merrin J, Leibler S, Chuang J (2007) Printing multistrain bacterial patterns with a piezoelectric inkjet printer. PLoS ONE 2:e663 25. Eagles PAM, Qureshi AN, Jayasinghe SN (2006) Electrohydrodynamic jetting of mouse neuronal cells. Biochem J 394:375–378 26. Jayasinghe SN, Qureshi AN, Eagles PAM (2006) Electrohydrodynamic jet processing: an advanced electric-field-driven jetting phenomenon for processing living cells. Small 2: 216–219 27. Kim JH, Hwang J, Jung HI (2009) Direct pattern formation of bacterial cells using micro-droplets generated by electrohydrodynamic forces. Microfluid Nanofluid doi: 10.1007/s10404-009-0441-6 28. Barron JA, Wu P, Ladouceur HD et al (2004) Biological laser printing: a novel technique for creating heterogeneous 3-dimensional cell patterns. Biomed Microdevices 6:139–147 29. Saunders R, Gough J, Derby B (2005) Ink jet printing of mammalian primary cells for tissue engineering applications. Mater Res Soc Symp Proc 845:57–62 30. Sumerel J, Lewis J, Doraiswamy A et al (2006) Pizoelectric ink jet processing of materials for medical and biological applications. Biotechnol J 1:976–987 31. Barron J, Young H, Dlott D et al (2005) Printing of protein microarrays via a capillary-free fluid jetting mechanism. Proteomics 5:4138–4144 32. Bechor O, Smulski DR, Van Dyk TK et al (2002) Recombinant microorganisms as environmental biosensors: pollutants detection by escherichia coli bearing fabaâ C2 ::Lux fusions. J Biotechnol 94:125–132 33. Ringeisen BR, Henderson E, Wu PK et al (2006) High power density from a miniature microbial fuel cell using shewanella oneidensis dsp10. Environ Sci Technol 40:2629–2634 34. Bond DR, Holmes DE, Tender LM et al (2002) Electrode-reducing microorganisms that harvest energy from marine sediments. Science 295:483–485 35. Tender LM, Reimers CE, Stecher HA et al (2002) Harnessing microbially generated power on the seafloor. Nat Biotechnol 20:821–825 36. Kim H, Choo YJ, Song J et al (2007) Marinobacterium litorale sp nov in the order oceanospirillales. Int J Syst Evol Microbiol 57:1659–1662 37. Landa BB, de Werd HAE, Gardener BBM et al (2002) Comparison of three methods for monitoring populations of different genotypes of 2,4-diacetylphloroglucinol-producing pseudomonas fluorescens in the rhizosphere. Phytopathology 92:129–137 38. Nye KJ, Turner T, Coleman DJ et al (2001) A comparison of the isolation rates of salmonella and thermophilic campylobacter species after direct inoculation of media with a dilute faecal suspension and undiluted faecal material. J Med Microbiol 50:659–662 39. Takehara T, Kuniyasu K, Mori M et al (2003) Use of a nitrate-nonutilizing mutant and selective media to examine population dynamics of fusarium oxysporum f. Sp spinaciae in soil. Phytopathology 93:1173–1181 40. Lee JK, Jung DW, Yoon K et al (2006) Effect of diluent salt concentration and ph on the enumeration of vibrio parahaemolyticus by direct plating on selective agar. Food Sci Biotechnol 15:866–870

Index

A Absorbing film assisted laser induced forward transfer (AFA-LIFT), 64, 96, 115–132 Absorbing layer/UV absorbing layer, 58, 63–67, 70, 96–98, 101, 116, 128–131 Agar, 62, 117–118, 246–252 Astroglial cell, 125, 127, 129–130 B Bacteria printing, 243, 251 Bioactive hydrogels, 166–167 Biodegradation, 4, 25 Bio-inks, 23 Biological guidance, 163–170, 173–184, 187–200, 203–219 Biological laser printing (BioLP), 64, 81–92, 95–111, 125, 189, 246, 248–253 Biomolecule printing, 53–54, 76 Biopaper, 111 Bioprinting, 3–18, 23–32, 95–111, 204 Bioprinting efficiency coefficient (BEC), 110–111 Bone Morphogenetic Protein-2 (BMP-2), 204–206, 210–212, 214–219 Bone/artificial bone, 16–17, 107, 109–110, 176–179, 181–183, 204–205, 209, 212, 222, 229, 231, 233–234 Bubble dynamics, 98–99, 111 C Calcium chloride (CaCl2 ), 9, 15, 26, 28–29 Cardiomyocytes, 137, 153 C2C12 myoblasts, 204, 206, 211, 214–219 Cell-cell interactions/communication, 49, 54, 82, 86, 138, 152, 169, 198 CellChipTM , 138

Cell guidance, 147–148, 177–180 Cell patterning, 38–41, 137–138, 140, 146–148, 150, 154, 190, 192, 196 Cell printing, 5, 8, 16–17, 24–31, 54–55, 82, 84–91, 103–107, 109, 125, 225–236, 243–254 Cell/tissue based sensors, 188 Cellular aggregation, 167 Cell viability, 6, 8, 12, 26, 28, 35, 37, 41–49, 147, 229, 252–253 Chondrocytes, 4, 8, 35, 37–48, 168–169, 204, 229, 234 Co-culture, 86–91 Collagen, 5, 9–17, 107, 125, 166–168, 230, 233 Computer aided design (CAD), 7, 25, 85, 103, 181, 189–190, 192, 226–227, 229, 246 Computer aided engineering (CAE), 25 Computer aided manufacturing (CAM), 25, 85, 103, 189, 226, 246 Cross linking, 9, 26 Curved surfaces, 187–200 Cytokine, 5, 14, 177–181 D 3D bioprinter, 5, 7, 27–30 3D fiber deposition (3DF), 225–236 Direct write technique, 81, 190, 192, 197 DNA microarrays, 53, 71–73, 204 Drop-on-demand (DOD), 35–36 Droplet resolution, 7, 12, 25–27, 54, 63, 67–69, 83, 87–88, 96, 105–106, 110, 189, 244, 247–248, 252 volume, 69, 70, 248 Drug testing/screening, 4 Dry printing, 26, 63

B.R. Ringeisen et al. (eds.), Cell and Organ Printing, C Springer Science+Business Media B.V. 2010 DOI 10.1007/978-90-481-9145-1,

257

258 E Electrochemically active bacteria (EAB), 250 Electrohydrodynamic (EHD) jetting, 244–248, 252–253 Electromechanical valves, 5, 8–9 Embryonic-like adult (ELA) stem cells, 5, 16 Endothelial cells, 28–29, 31, 62, 82, 88–89, 109, 164–165, 168–169, 178, 184, 234 Engineered tissues, 3–18, 169, 181 Epidermal Growth Factor (EGF), 167, 205, 209, 212–214, 217–218 Epithelial cells, 62, 125, 127, 168, 193–194, 196 Escherichia coli (E. coli), 62, 82, 244, 246–249 Excimer laser, 55, 58, 63, 83, 116–117, 119, 121, 125, 127, 130 Extracellular matrix (ECM), 4, 104, 108, 129, 138, 169, 176, 188, 231, 235 Extrusion-based printing, 226, 229 F Femtosecond (fs) laser, 57, 65, 73–75, 129–132 Fibroblast Growth Factor-2 (FGF-2), 204–206, 209–213, 216–217, 219 Fibroblasts, 10, 44, 62, 88, 169, 193–194, 196–197, 204–205 Freeform fabrication, 3–18 Fresnel relations, 97 Frontal-parietal calvarial bone flap, 181 Fungus printing, 62 G Gel, 5, 8–11, 14–16, 26–27, 29, 31, 49, 104, 167–168, 228, 230–235 Gelation, 9, 26–27, 30, 49, 166, 230–231 Gel precursor, 26 Generalized Lorenz Mie Theory (GLMT), 143–146, 148 Growth factor, 5–6, 13–16, 23–24, 27, 30, 85–86, 88, 110, 150, 164, 167–168, 178, 203–219, 228 array, 203–219 H Hepatocytes, 4, 204, 217 High throughput biological laser printing (HT BioLP), 95–111 Human umbilical vein, 88–89 Hydrogel, 4–15, 26–30, 49, 87–88, 91–92, 98–99, 104, 166–168, 225–227, 230–236, 246 Hydroxyapatite, 107–109, 111, 181

Index I Indicator plate, 250–251 Ink jet, 5, 23–32, 35–49, 59, 84, 91, 95, 111, 138, 189, 244–248, 252–253 Inks bio liquid, 248 neuron, 85 solid, 92 color, 24–25, 27, 30, 203, 205 pigment, 27 Inorganic materials, 56–57, 59 Insulin-like Growth Factor-I (IGF-I), 204–206, 209–217, 219 In vivo experiment, 86, 180, 183, 212 In vivo printing, 109–110 J Jet formation, 99–101 K Keratinocytes, 10 L Laminated sheets, 29 Laminin, 189, 191–193, 195–196 Laser guidance cell micropatterning, 148–152 Laser-guided direct write (LG DW), 96, 111, 140, 189–190 Laser induced forward transfer (LIFT), 53–76, 81–92, 95–111, 115–132, 244, 246, 248, 252 Liquid jet, 70–71, 116, 127–129 Lithographic approaches, 5 Live/dead assay, 8, 43, 85, 229 Lumens, 5, 14, 88–91, 169 M Mammalian cells astroglial cell; 125, 127, 129–130 cardiomyocytes, 137, 153 chondrocytes; 4, 8, 35, 37–48, 168–169, 204, 229, 234 endothelial cells; 28–29, 31, 62, 82, 88–89, 109, 164–165, 168–169, 178, 184, 234 epithelial cells; 62, 125, 127, 168, 193–194, 196 hepatocytes; 4, 204, 217 neurons; 12, 82, 85–88, 140, 152–154, 177, 193, 204 olfactory ensheathing cells; 62, 82, 85–86 osteoblasts; 35, 37, 39–48, 129, 178, 204 Schwann cells; 62, 125, 127

Index Maskless mesoscale material deposition technique (M3D), 187 MatrigelTM , 60–62, 83, 85–86, 88–89, 117, 129–130, 169, 230, 233 Matrix assisted pulsed laser evaporation direct write (MAPLE DW), 58, 73, 82, 96, 116, 125, 138, 189, 246 Mesenchymal stem cells (MSCs), 8, 16, 168, 176, 178, 183, 209, 215, 219, 229, 233 Microbial fuel cell (MFC), 250–251 Microelectrode array (MEA), 150, 154 Micro-electro-mechanical systems (MEMS), 4 Microfabricated tissue, 163 Microfluidics, 4, 13, 139 Microneedle delivery, 178–180 Micropatterns/micropatterning, 54, 56, 96, 137–155, 166, 197–198 Micro-structure, 24–25, 154 Microvalves, 5–6 Mie coefficients, 144 Multicolor printing, 108–109 Multi-layer scaffolds, 12–13 Multiple cell types, 24, 29, 36, 82, 138 Multipotentency, 229 Myogenic differentiation, 204, 206, 211–219 N Nano-hydroxyapatite, 107 Nanosecond (ns) laser, 65, 75, 116, 129–130 Nd:YAG laser, 55, 63, 65 Neuronal network, 137, 153–154 Neurons, 12, 82, 85–88, 140, 152–154, 177, 193, 204 Nozzle-free technique, 63 Number of cells per drop, 39, 247 O Olfactory ensheathing cells, 62, 82, 85–86 Optical force, 138–146, 148, 153, 189 Optical/laser tweezers, 139–140 Organ printing, 225–236 Osseous tissue, 5, 16, 177 Osteoblast, 35, 37, 39–48, 129, 178, 204 P Peptide printing, 74 Photolithography, 54, 188–189 Photopolymerizable hydrogel, 167, 232 Photoreactive/photoactive growth factors, 206, 208–210, 216–218 Piezoelectric ink jet, 111, 245, 247, 252–253 Platelet-derived growth factor-BB (PDGF-BB), 205, 213, 217

259 Pluripotency, 176 Pneumatic pressure, 5–7 Poisson statistics, 39, 41 Polydimethylsiloxane (PDMS), 10–11, 13, 166–167, 187–189, 191–193, 195–196, 198–199 Polyethylenimine (PEI), 187, 191–193 Poly(glycolic acid) (PGA), 227 Poly(lactic acid) (PLA), 227 Poly (lactic/glycolic) acid (PLGA), 228 Polymeric scaffolds, 177–178, 181 Poly[poly(ethylene oxide) terephthalate-co-(butylene) terephtalate] (PEOT/PBT), 227–228 Polytetrafluoroethylene (PTFE), 187, 191, 193 Poly(ε-caprolactone) (PCL), 178, 180–183, 227 Porosity, 4, 108, 228–229, 231, 233 Precision spraying (PS), 187–200 Protein printing, 68, 73–76 Proteins, 23–24, 27, 54, 66, 68, 72–76, 91, 107, 174, 176, 178, 194, 197, 204–205, 209, 212, 216, 230, 234, 247–248 R Rapid prototyping (RP), 5, 101, 189, 198, 225–226, 229 Rayleigh–Plesset equation, 98–99 Receiving substrate agar; 62, 117–118, 246–252 collagen, 5, 9–17, 107, 125, 166–168, 230, 233 gel; 5, 8–11, 14–16, 26–27, 29, 31, 49, 104, 167–168, 228, 230–235 hydrogel; 4–15, 26–30, 49, 87–88, 91–92, 98–99, 104, 166–168, 225–227, 230–236, 246 MatrigelTM ; 60–62, 83, 85–86, 88–89, 117, 129–130, 169, 230, 233 Recruitment (of cells), 174 Regenerative medicine, 5, 16, 18, 174, 182, 188, 243 Resolution, 7–8, 11–12, 14, 24–27, 53–56, 63, 65, 67–69, 81–92, 95–96, 101–103, 105–107, 109–111, 121, 123, 129, 150, 152, 154, 189, 198–199, 231, 244, 246–248, 252–253 Ribbon, 102–103, 106–107, 109 S Schwann cells, 62, 125, 127 Self organization of cells, 163–170, 173–184, 187–200, 203–219

260 Semi-solid hydrogel, 23 Shewanella oneidensis MR–1, 250 Single cell printing, 244 Skin layers, 10 Sodium alginate, 9, 26, 29, 49 Solid printing, 27, 58, 91 Stem cell differentiation, 4 Stem cell migration, 174–177 Stem cell niche, 176–178 Stem cells, 5, 8, 14, 16–17, 62, 168, 174–178, 183, 204–205 Substratum/photoactive substratum, 204–214, 216–218 Syringe, 6–7, 14, 149, 151, 178, 193, 226–227, 232, 245–246 T Thermal ink jet, 244, 247, 253 Thermoplastic scaffolds, 225, 227 Three dimensional printing (3D Printing), 11, 16, 107–108, 111 Time-release, 14

Index Tissue architecture, 4–5, 25 assembly, 163, 170 engineering, 4, 9, 16, 24–27, 29, 31, 54, 73, 88, 101, 107–108, 164, 174, 177–179, 181–183, 188, 190, 199, 225–236, 244 microdissection, 82, 84, 91–92, 243 self organization, 163–165, 167–169 Trichoderma longibrachiatum conidia, 115, 122 Tube structures, 28–29, 31 V Vascular endothelial growth factor (VEGF), 14–16, 164–165, 169, 178, 180–181 Vascular guidance, 177, 181–182 Vascularization, 88, 177–178, 181–183, 233–234 Vascular network, 24, 88, 169, 181–182, 184 Vessels, 28, 30, 88, 164–165, 177–178, 180–181, 183–184, 204