Applying macromolecular crowding to enhance ...

2 downloads 0 Views 3MB Size Report
Prerequisites for collagen matrix deposition in vitro arenotoptimal . ...... accumulation of type V collagen fibrils in accordance with cell aggregation,. Journal of ...
ADR-12105; No of Pages 14 Advanced Drug Delivery Reviews xxx (2011) xxx–xxx

Contents lists available at ScienceDirect

Advanced Drug Delivery Reviews j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / a d d r

Applying macromolecular crowding to enhance extracellular matrix deposition and its remodeling in vitro for tissue engineering and cell-based therapies☆ Clarice Chen a,c, Felicia Loe a,c,1, Anna Blocki a,c, Yanxian Peng a, Michael Raghunath a,b,⁎ a b c

Division of Bioengineering, Faculty of Engineering, National University of Singapore, Singapore Department of Biochemistry, Yong Loo Lin School of Medicine, National University of Singapore, Singapore NUS Graduate School for Engineering and Integrative Sciences (NGS), National University of Singapore, Singapore

a r t i c l e

i n f o

Article history: Received 15 December 2010 Accepted 2 March 2011 Available online xxxx Keywords: Microenvironment Excluded volume effect Stem cells Differentiated cells Matrix maturation Collagen assembly

a b s t r a c t With the advent of multicellular organisms, the exterior of the cells evolved dramatically from highly aqueous surroundings into an extracellular matrix and space crowded with macromolecules. Cell-based therapies require removal of cells from their crowded physiological context and propagating them in dilute culture medium to attain therapeutically relevant numbers whilst preserving their phenotype. However, bereft of their microenvironment, cells under perform and lose functionality. Major efforts currently aim to modify cell culture surfaces and build three dimensional scaffolds to improve this situation. We discuss here alternative strategies that enable cells to re-create their own microenvironment in vitro, using carbohydrate-based macromolecules as culture media additives that create an excluded volume effect at defined fraction volume occupancies. This biophysical approach dramatically enhances extracellular matrix deposition by differentiated cells and stem cells, and boosts progenitor cell differentiation and proliferation. We begin to understand how well cells really can perform ex vivo if given the chance. © 2011 Elsevier B.V. All rights reserved.

Contents 1. 2.

3.

4.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . What is macromolecular crowding (MMC) and how does it work? . . . . . . . . . 2.1. MMC is an ancient biological principle . . . . . . . . . . . . . . . . . . 2.2. MMC and the excluded volume effect . . . . . . . . . . . . . . . . . . 2.3. Determining the FVO ψ . . . . . . . . . . . . . . . . . . . . . . . . . 2.4. Tailoring crowded systems . . . . . . . . . . . . . . . . . . . . . . . . 2.5. Mono- versus mixed macromolecular crowding . . . . . . . . . . . . . . 2.6. Why MMC is not the same as a simple volume reduction of culture medium Applications of MMC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Contemporary biochemical assays and cell culture are not macromolecularly 3.2. Prerequisites for collagen matrix deposition in vitro are not optimal . . . . 3.3. The charge of crowders influences ECM deposition speed and morphology . 3.4. The pleiotropic effects of MMC in ECM formation . . . . . . . . . . . . . 3.4.1. MMC accelerates procollagen conversion by PCP . . . . . . . . . 3.4.2. MMC accelerates supramolecular assembly . . . . . . . . . . . . 3.4.3. MMC accelerates ECM stabilization . . . . . . . . . . . . . . . 3.4.4. Increased remodeling of ECM under MMC . . . . . . . . . . . . The role of MMC in building microenvironments for stem cell culture . . . . . . 4.1. Using MMC-constructed decellularised matrices . . . . . . . . . . . . . . 4.2. Growing stem cells directly under MMC . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . crowded . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

☆ This review is part of the Advanced Drug Delivery Reviews theme issue on "From Tissue Engineering To Regenerative Medicine – The Potential And The Pitfalls”. ⁎ Corresponding author at: Division of Bioengineering, Faculty of Engineering & Dept of Biochemistry, Yong Loo Lin School of Medicine, National University of Singapore, Singapore. Tel.: + 65 6516 5307(DSO building); fax: + 65 6776 5322(DSO building). E-mail address: [email protected] (M. Raghunath). URL: http://www.tissuemodulation.com (M. Raghunath). 1 Present address: Singapore MIT Alliance for Research and Technology, National University of Singapore, Singapore. 0169-409X/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.addr.2011.03.003

Please cite this article as: C. Chen, et al., Applying macromolecular crowding to enhance extracellular matrix deposition and its remodeling in vitro for tissue engineering and..., Adv. Drug Deliv. Rev. (2011), doi:10.1016/j.addr.2011.03.003

2

C. Chen et al. / Advanced Drug Delivery Reviews xxx (2011) xxx–xxx

5. Where and when does MMC not work? . . . . 6. The future of applied macromolecular crowding 7. Conclusion . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

1. Introduction The human body is comprised of approximately 411 different cell types [1], and they reside in or migrate through microenvironments comprising a variety of extracellular matrix (ECM) [2]. Cell-based therapies hinge on one particular platform technology: ex vivo cell culture. This technology is based on the sudden separation of cells from their tissue microenvironment (harvest, mechanical and enzymatic destruction) and placing them on plastic in the presence of huge amounts of salt water, additives, and fetal calf serum, still an ill-defined compound but difficult to replace. Cell culture has come a long way from glass flasks to modern cell culture polystyrene that has been treated with gas plasma to make it more cell-friendly. Huge efforts in industry have been made to ensure quality and reproducibility of the hardware, as well as cell culture sera and media compositions. These efforts are reflected in textbooks that specifically deal with establishing cell cultures [3]. So far, this technology has served us well, but it has become forgotten over time that what we have in cell culture flasks and in many bioreactors is pathological: the disproportion of aqueous medium to cell mass, the substitution of an organic support by plastic, and the lack of macromolecular crowdedness as such, a hallmark of the microenvironment of cells in metazoans. In fact, if such a disproportion of fluid to cell would occur in the human body this would incur immediate medical attention, yet biologists all over the world are complacent to grow cells under conditions that in the clinical world would be addressed as oedema or effusion. It is therefore small wonder that cell-based therapy has now reached a glass ceiling in its attempts to create larger three dimensional structures. One reason is the limited diffusion of oxygen and nutrients into the tissue in structures thicker than 1–2 mm [4–8], which is a mass transfer problem in vitro, and a microvascularisation issue after implantation. The other limitation, which we propose to remove by applying macromolecular crowding (MMC), is set by the ability to create structures coherent enough to be manipulatable. Scaffold-free approaches involving cell printing and cell sheet technology are both dependent on the presence of sufficient ECM to stabilise the structures [9,10], but unfortunately the deposition of ECM in vitro with uncrowded conditions is an inefficient process. Stem cell based therapies are currently hitting a roadblock because ex vivo propagation of stem and progenitor cells on tissue culture polystyrene results in decaying proliferation and differentiation capacities [11,12]. A recent technology feature summarised the worldwide efforts by materials scientists and chemical engineers to produce microenvironments for stem cells [13]. On one hand it is intriguing to witness efforts to supercede some 700 million years of evolution and materials testing, on the other hand this testifies the glaring disregard for the capacity of cells themselves to manufacture their ECM. In this review, we share possible ways to control and augment this cellular capacity by re-introducing macromolecular crowding into culture systems, in order to create a well-developed ECM that can provide cellular cohesion and tissue strength to aid in tissue engineering, and building microenvironments for stem cell work both for basic research and for therapeutic applications. 2. What is macromolecular crowding (MMC) and how does it work? 2.1. MMC is an ancient biological principle This review does not intend to give a comprehensive overview of crowding theory, but rather aims at explaining the principle of

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

0 0 0 0 0

macromolecular crowding (MMC) to the practitioner. Our focus is the effect of MMC on the formation of ECM, and we intend to highlight the basics of this ancient principle of life to scientists in regenerative medicine. Possible applications include improving cell performance in culture and preserving the phenotype of precious progenitor cells to facilitate cell-based therapy. With this disclaimer in mind and the intention to bridge the current gap between theory and application, we would like to begin with the statement that all living systems are highly crowded [14]. This is true of the interiors of cells, whether bacterial, animal, or plant, and the exterior of most cells of multicellular organisms. The crowding element is derived from macromolecules such as proteins, carbohydrates, lipids and nucleic acids, that form macromolecular complexes and supramolecular assemblies such as cellular organelles and membranes [15]. Although up to 40% of the cytoplasmic volume can be occupied by macromolecules [16,17], the usual range lies between 20 and 30% [18]. Notably, the volume-occupying macromolecules also tend to have a net fixed electrical surface charge. The question arises whether crowding is just an incidental occurrence during evolution or was necessary for the origin of life. Interestingly, the earliest life forms (viruses, archaea and prokaryotes) have been found to have crowded structural and functional units. The total concentration of protein and RNA inside bacteria like E. coli is in the range of 300–400 g/L. As we go up the tree of evolution, crowding persists as a highly conserved property of higher organisms [19]. 2.2. MMC and the excluded volume effect Macromolecular crowding functions by way of the excludedvolume effect (EVE) and is often referred to as the “volume of a solution that is excluded to the center of mass of a probe particle by the presence of one or more background particles in the medium” [19]. Fractional volume occupancy Ψ (FVO) denotes the fraction of the total volume occupied by macromolecules (Fig. 1). Thermodynamically, volume exclusion lowers the configurational and conformational freedom (entropy) leading to elevated basal free energy of the reactant macromolecules and a number of downstream effects [20]. These may be identified as (1) folding of biopolymers (e.g. proteins and nucleic acids) into native states optimal for function [21], (2)

Fig. 1. A simplified representation of the generation of EVE through the presence of macromolecules. The schematic reflects the situation that a test molecule (red) encounters in a given volume element (box). The crowders (black) occupy about 30% of this volume, the fractional volume occupancy (FVO, ψ) therefore is 30% (v/v). While FVO can be calculated, the additional unavailable volumes represent a challenge to compute as several factors such as electrostatic repulsion and hydration shell need to be considered. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Please cite this article as: C. Chen, et al., Applying macromolecular crowding to enhance extracellular matrix deposition and its remodeling in vitro for tissue engineering and..., Adv. Drug Deliv. Rev. (2011), doi:10.1016/j.addr.2011.03.003

C. Chen et al. / Advanced Drug Delivery Reviews xxx (2011) xxx–xxx

stronger macromolecular transition complexes with longer half-lives (eg. enzyme–substrate) leading to more product, and (3) buffering effect of crowded environments on biological function under conditions of adverse pH, temperature or ionic strength [22]. Biological reactions performed in dilute in vitro conditions proceed at significantly tardier rates than their in vivo counterparts [23]. However, MMC and EVE have not been appreciated in the biological domain and applications are yet to be fully developed and implemented [24]. For our intents and purposes, we define a macromolecule within the realm of EVE in biological systems to have a molecular weight (MW) in the range of 50 to 500 kDa, with a hydrodynamic radius RH between 5 and 50 nm. Structurally and chemically, the macromolecule could be derived from any representative of the carbohydrate/ protein/nucleic acid/lipid family. However, the crowding behaviour is entirely dependent on the physical property of the macromolecule. The density or crowdedness of macromolecules is determined by the approximation of a solute content. For the interior of cells, concentrations of macromolecules have been determined to be in a range from 50 to 400 mg/mL [25]. This concentration refers to the cumulative concentration of all species of macromolecules present, rather than the concentration of any single macromolecule. The presence of these macromolecules means that they significantly occupy space in the given medium, and biophysicists refer to such medium as “crowded” or “confining” [25]. Biological macromolecules such as enzymes and proteins function in highly crowded environments. Crowding and EVE are thus inevitable requirements for life and typical of the intra and extracellular milieu. Cell culture is a typical example, where cells anchored to a culture plate find themselves bathed in an ocean of medium that is hardly representative of its in vivo microenvironment. MMC can be mimicked experimentally by adding inert synthetic or natural macromolecules to a system in vitro [24]. As a result of the mutual impenetrability of solute molecules (Pauli Exclusion Principle), steric repulsions are generated and this affects the activity of a solute — which is highly dependent on its steric freedom or the volume that is available to the centre of the mass of each molecular species. This is the most fundamental of all interactions between macromolecules in solution and is always present at finite concentration, independent of the magnitude of additional electrostatic or hydrophobic interactions. It is conceivable that the addition of the inert crowders to occupy a significant volume fraction in the medium will place constraints on the active factors present in the microenvironment, thus driving associations and reactions [24,25].

3

Table 1 Hydrodynamic radii of differently charged macromolecules as determined by Dynamic Light Scattering in Hank's Balanced Salt Solution.

Net charge

DxS

DxS

PSS

Dextran

500 kDa

10 kDa

200 kDa

670 kDa 400 kDa 70 kDa

Negative Negative Negative Neutral

Hydro. radius (nm) 47

b1

22

21

Fc

Fc

Neutral

Neutral

8

4

but polar-hydrophilic molecules such as Ficolls and polyvinylpyrrolidone), creates large hydration shells around the mass of the macromolecule. This might also involve water-structuring which needs further investigation so that the degree of hydration can be accurately identified [29]. Our measurements show that within each charge group, the RH is correlated with the MW. However, although we expected an RH of 18 nm for the negatively charged crowder poly (sodium 4-styrene sulfonate) (PSS) 200 kDa, we found an RH of 22 nm. A combined theoretical and empirical estimation revealed a 3fold higher surface charge density of PSS in comparison to dextran sulfate 500 kDa (DxS) (ζ potential, [30]). This implies that even though a molecule may be smaller in terms of MW, its EVE might be comparable to a larger molecule due to the combined electrostatic and steric effects. This also applies to neutral crowders, whereby higher concentrations of neutral crowders need to be added to culture media in comparison to that of negatively charged ones to attain comparable volume exclusion. In fact, the empirically determined working concentrations of DxS in our hands range from 50 to 100 μg/mL, while the neutral Ficolls have to be used in the milligram range. However, the fact that 200 times more Ficoll molecules are needed when their RH is only 3.5 times smaller than those of DxS suggests that RH alone cannot fully account for these differences and that further considerations apply (see below). In essence, the FVO is a parameter that can be calculated from the RH and is important for an approximate assessment (Table 2). When charge comes into consideration, it is important to remember that electrostatic repulsion and hydration shells will augment the effect, resulting in FVO possibly under reporting the crowding power in a chosen system in relation to target test molecules or processes to be influenced by MMC. The total excluded volume is difficult to compute, as the unavailable space is difficult to estimate. The FVO ψ, however, can be determined once the RH of the crowding molecules and their numbers are known. Based on the assumption that the molecule in question is spherical, the volume of a single crowding molecule can be

2.3. Determining the FVO ψ One important concept in crowding theory is that macromolecular crowding is most effective if the crowder (better, the volume it occupies) and the test molecule (better, the volume it occupies) are of similar size [26,27]. Therefore to select any crowder for a biological reaction, it becomes essential to choose similar sized crowders as the reactant molecules (enzymes/substrates). Hence, prior knowledge of the hydrodynamic radii of the crowders and the substrate molecules become critical and herein lies the importance of our biophysical approach to select suitable crowders. To characterise and predict the crowding power of a molecule, we determined the respective hydrodynamic radius (RH) of a variety of crowding molecules by dynamic light scattering (Table 1). The RH of a crowding molecule describes its effective size in a physiological and aqueous environment. The charge (net or total negative charge for anionic macromolecules and partial charge for neutral but polar macromolecules) has a major influence on the RH in physiological salt solutions and a pH of 7.4. Surface charge can lead to electrostatic repulsion (of like-charged test molecules) and thus further enhance the basal volume exclusion due to steric repulsion, at least 2-fold [28]. In addition, the surface charge, whether total or partial (for neutral

Table 2 Calculations for estimating the factional volume occupancy. Albumin is used as an example. Calculate the volume of each albumin molecule considered as a sphere with a radius of 4 nm (4e-9 m) Volume of a sphere = 4/3 πr3 = 4/3 π (4e-9)3 = 268 × 10-27 cubic meter → 1 Calculate the number of albumin molecules in 80 mg ▪ From literature, we know that the MW of albumin is ~ 69,000 Da. ▪ That means, 69,000 grams of albumin contain Avogadro number of molecules. ▪ From this, we then calculate number of molecules of albumin present in 80 mg as follows: (80 × 10-3) × (6.023 × 1023) ÷ 69,000 = 69.8 × 1016 → 2 Calculate the Fraction volume Occupancy in 1 ml Now, we calculate the volume occupied by these number of albumin molecules in 1 ml by multiplying 1 & 2 and then express in percentage to estimate the fraction volume occupancy: 268 × 10-27 × 69.8 × 1016 = 18% (v/v)

Please cite this article as: C. Chen, et al., Applying macromolecular crowding to enhance extracellular matrix deposition and its remodeling in vitro for tissue engineering and..., Adv. Drug Deliv. Rev. (2011), doi:10.1016/j.addr.2011.03.003

4

C. Chen et al. / Advanced Drug Delivery Reviews xxx (2011) xxx–xxx

Table 3 Fractional volume occupancy (FVO) exerted by crowders/crowder mixtures at the most effective concentrations (determined empirically) in terms of accelerated collagen deposition. Macromolecular Molecular Hydrodynamic Effective Fractional vol. crowders weight (kDa) radius (nm) concentration occupancy Ψ Ficoll 70 Ficoll 400 Dextran sulfate Polystyrene4-sulfonate Albumin

70 400 500 200

4 8 46 22

69

4

37.5 mg/mL 25 mg/mL 100 μg/mL 50 μg/mL

~ 17% combined 5.2% 0.7%

80 mg/mL

18%

determined. Multiplied by the amount of crowding molecules (to be calculated via molecular weight and concentration used), the total volume that the crowders occupy can be computed. We had empirically titrated the most efficacious working concentrations of the negatively charged crowders [30]. When we went back to calculate ψ, we were surprised to find them relatively small in comparison to the neutral Ficolls (Table 3). The current explanation for this discrepancy can be put forth as follows: For negatively charged crowders, the electrostatic forces amplifying steric exclusion effects are dependent on the charge distribution on the surface of the macromolecule, and the smaller the surface area, the greater the charge density. Hence, far smaller concentrations of these crowders are empirically needed to exert similar exclusion effects as their neutral counterparts. However, further research needs to determine a generic formula to surmount this quantitative challenge. Additionally, the effect of negatively charged macromolecular crowders on structuring water as kosmotropic agents remains to be determined. Interestingly, the contribution of water is not considered for calculating volume exclusion [31], although macromolecules can interact indirectly via water molecules at much lower concentrations than those needed for direct interactions. Thus crowding effects may be larger if hydrodynamic interactions are also considered [32] and this might explain the strong effects seen with DxS and PSS despite smaller calculable ψ based on RH and molarity alone. 2.4. Tailoring crowded systems For our work on the stem cell microenvironment of human bone marrow-derived mesenchymal stem cells (hMSCs), we used the

protein concentration in the physiological hMSC microenvironment as a basis, which ranges from 20.6 g/L to 80 g/L. As a main representative of the blood-rich microenvironment in the bone marrow, we postulated albumin as the main representative. We determined via dynamic light scattering, the hydrodynamic radius of albumin in a physiological buffer (Hanks buffered salt solution) to be around 4 nm [30]. Based on these calculations, the fraction volume occupancy due to albumin turns out to be ~18% at physiological conditions (Table 2). We determined experimentally, that the Fc cocktail composed of Ficoll 70 kDa (Fc70) at 37.5 mg/mL and Ficoll 400 kDa (Fc400) at 25 mg/mL gave optimal results in terms of ECM deposition and in molecular biology assays. This coincides with respective FVOs of 8.7% and 8%. This cocktail would have around 17% total FVO, which is similar to estimates of in vivo conditions (Table 3). We have used this mixture successfully in ECM production [33] and for our stem cell work (Section 4). 2.5. Mono- versus mixed macromolecular crowding We originally assessed the efficiency of crowders to drive collagen deposition [30,34]. Using mono crowding, we found that individually, both neutral macromolecules Fc70 and Fc400 at 50 mg/mL were unable to increase collagen deposition. This was attributed to their neutral charge and relatively small hydrodynamic radius. However, inspired by work on mixed macromolecular crowding [21], we combined Fc70 and Fc400 (in individually reduced concentrations) and discovered that this mixture gave a surprisingly substantial ECM deposition after 5 to 7 days [33]. Apparently, in a cocktail of macromolecules, it is possible for one macromolecule species to have crowding effects on another species, and in such a way, improve the efficacy of MMC overall [21]. 2.6. Why MMC is not the same as a simple volume reduction of culture medium It stands to reason that MMC and the resulting EVE basically creates pockets of upconcentrated reactants. Accordingly, it would appear that similar effects could be achieved by simply growing cells in smaller volumes of culture medium. However, both theoretical considerations and practical testing demonstrate that this is not the case. In one particular set of experiments with WI-38 fibroblasts, illustrated in Fig. 2, we found that 500 μl DxS-crowded medium gives 17 times more

Fig. 2. Reducing the culture medium volume does not enhance collagen deposition to the same extent as MMC. (a) fibroblast cultures were grown under non-crowded conditions with reduced volumes of medium and showed at best, a 2.2-fold collagen deposition using 200 μl of medium. In contrast, DxS-crowded fibroblasts showed a much more enhanced collagen deposition at 17.3-fold. When medium volume was reduced to 100 μl and 50 μl, the wells dried out, resulting in cell death. (b) thermodynamic activities in 500 μl of DxScrowded medium far exceed that of 200 μl and 100 μl of uncrowded fibroblast cultures (reproduced with permission from Peng and Raghunath [43]).

Please cite this article as: C. Chen, et al., Applying macromolecular crowding to enhance extracellular matrix deposition and its remodeling in vitro for tissue engineering and..., Adv. Drug Deliv. Rev. (2011), doi:10.1016/j.addr.2011.03.003

C. Chen et al. / Advanced Drug Delivery Reviews xxx (2011) xxx–xxx

collagen I deposition that the non-crowded 500 μl medium, but still eight times more than a reduced volume of 200 μl of non-crowded culture medium. Regardless of using 500 μl or 200 μl of uncrowded culture medium, the total volume will equal the available volume and so the thermodynamic activity, by definition, will equal 1. So, there is a concentration effect to a limited extent in smaller medium volumes. Now, if we do not reduce the culture medium volume, but add for example 100 μg/mL DxS (FVO ψ = 5.2%) [34], the available volume decreases to one-tenth of total volume, thus increasing the thermodynamic activity to 10 and above. Hence, MMC works much more efficiently than mere volume reduction, not to mention that larger volumes of culture medium provide much more nutrition and oxygenation to cells, while preventing their dehydration. 3. Applications of MMC 3.1. Contemporary biochemical assays and cell culture are not macromolecularly crowded It is unfortunate that the notation of biochemical reactions (enzymatic catalysis, supramolecular aggregation, and receptor ligand interactions) usually does not take into account the size of the molecules involved and the environment they take place in. Until recently, crowding was still considered as a “non-ideality” [35] factor that is best ignored. Zimmerman and Minton [19] observed that crowding effects need to be taken into account when attempting to relate biochemical and biophysical observations made in vitro to physiological processes observed in vivo, and also surmised that crowding is an important parameter to be exploited in vitro. Both experimental and theoretical work has demonstrated substantial (order-of-magnitude) effects of crowding on a broad range of biochemical, biophysical and physiological processes in vitro, including nucleic acid and protein conformation and stability, protein– protein and protein–DNA association equilibria and kinetics (such as protein crystallization, protein fiber formation and bundling for example), and catalytic activity of enzymes and cell volume regulation [15,19,26,36]. Nevertheless, typical in vitro methods irrespective of the particular setting (tube, cuvette, ELISA plate, cell culture dish or flask, PCR) are done in aqueous solutions with concentrations of macromolecules ranging from 1 to 10 mg/mL [24]. Under these conditions, crowding effects are negligible. Turning our attention to cell culture, we find that cells are typically cultured on plain or thinly coated tissue culture plastic and are submerged in large volumes of non-crowded aqueous medium. This situation is far from physiological, as the cells have been derived from a microenvironment comprised of dense arrays of ECM macromolecules. In addition, the extracellular fluid (composed of interstitial fluid and blood plasma) contains proteinaceous macromolecules. Collectively these macromolecules occupy a range from approximately 7% to 40% of the total fluid volume which translates into a high protein concentration ranging from 20.6 g/L to 80 g/L [18]. Therefore, the extracellular fluid is termed ‘volume-occupied’ or ‘crowded’ rather than concentrated, since no single macrosolute species is concentrated [19]. While blood plasma has solute concentrations of around 80 mg/mL [24], typical cell culture employs supplements of 5%–20% fetal bovine serum resulting in solute concentration of 4–16 mg/mL. Obviously, current culture conditions do not provide a crowded environment and this raises the question of how this situation can be remedied. 3.2. Prerequisites for collagen matrix deposition in vitro are not optimal When adequately provided with ascorbic acid (which is not necessarily the case in publications in the tissue engineering field and cell biological assays), fibrogenic cells secrete substantial amounts of fibrillar and non-fibrillar procollagens [37]. However, before procollagen I can participate in supramolecular assembly, it has to be proteolytically cleaved to collagen I by bone morphogenetic protein

5

1/procollagen C-proteinase (PCP), which in turn is regulated by an allosteric binding molecule, procollagen C-endopeptidase enhancer (PCOLCE). It is still not widely known; that the activity of PCP is low under current non-crowded cell culture conditions. As a result, only small amounts of insoluble collagen matrix are formed, while most of the (water soluble) procollagen is lost during culture medium changes and in bioreactor outflows [30,34]. Twenty five years ago Bateman et al. [38] observed that the addition of polyethylene glycol, dextran T-40 or polyvinylpyrrolidone to fibroblast cultures lead to a complete association of collagen with the cell layer, while no procollagen/collagen was present in the culture medium. Studying the activity of PCP, Hojima et. al. [39] found that both the presence of 500 kDa DxS and polyethylene glycol increased the respective activities of purified procollagen C-proteinase and purified procollagen N-proteinase in vitro in a cell-free system. After these observations, only one publication followed making use of DxS crowding to study matrix deposition from fibroblasts harbouring a collagen I mutation [40]. The system was not used since then until we revisited it over a decade later to characterise it in greater detail (Table 4) [30,34,41]. 3.3. The charge of crowders influences ECM deposition speed and morphology We originally aimed for speed [30,34] and therefore developed a system for fast and efficient collagen deposition using negatively charged crowders with large hydrodynamic radii and subsequent higher EVE. In fact, we could routinely deposit more collagen I in the form of aggregates/granules in 48 h compared with 6–8 weeks in static systems. We used this system in our screening tool for antifibrotic drugs [33,42]. However in this short time window, neutral crowders like Ficoll or neutral dextran as monocrowders appeared to be ineffective. This impression changed with the application of mixed MMC. The mixture of Fc70 and Fc400 also substantially enhanced ECM deposition in a time window of 5–7 days [33,43] but in a reticular pattern, that promises to be more cohesive in cell culture (Fig. 3). 3.4. The pleiotropic effects of MMC in ECM formation We now present experimental evidence that MMC works on several levels of collagen and ECM deposition, namely (1) extracellular proteolytic trimming (2) driving supramolecular assembly and (3) crosslinking. 3.4.1. MMC accelerates procollagen conversion by PCP The first mode of action on collagen deposition was originally suggested to be the accelerated conversion of procollagen I to collagen I [38,39]. We directly confirmed this by Western blot showing an Table 4 Summary of ECM components deposited using 500 kDa DxS for 46 h by WI-38 fibroblasts. Method Immunocytochemistry

MALDI-TOF

Collagen I Collagen IV Collagen V Biglycan Laminin 1 TGase 2 Fibronectin-1 Fibrillin-1 LTBP-1 Fibulin,1, Fibulin 2 Tenascin R Heparin sulfate Decorin

Collagen α2(I) Collagen α1(III) Collagen α1(V) Collagen α2(VI), α3(VI) Collagen α1(XII) TGase 2 Fibronectin-1 Fibrillin-1 Vitronectin Elastin microfibril interfacer 1 Tenascin C IGFBP7 Thrombospondin-1

Please cite this article as: C. Chen, et al., Applying macromolecular crowding to enhance extracellular matrix deposition and its remodeling in vitro for tissue engineering and..., Adv. Drug Deliv. Rev. (2011), doi:10.1016/j.addr.2011.03.003

6

C. Chen et al. / Advanced Drug Delivery Reviews xxx (2011) xxx–xxx

collagen was now present in the cell layer. As we had no evidence for a change in amount of PCP, we inferred that its activity must have changed [34]. Mixed MMC using neutral crowders showed a significant increase of ECM deposition compared with controls, but not as dramatic as with negatively charged DxS. The slower deposition rate of collagen I seen with mixed MMC, correlates well with a slower conversion rate of procollagen I to collagen I and the presence of intermediate cleavage products before being finally trimmed to collagen (Fig. 4). We also found evidence for a second proteolytic event that was enhanced, namely the cleavage of the allosteric regulator PCOLCE of PCP into two fragments. One fragment represented the active form of PCOLCE and the second was a C-terminal fragment of 16 kDa [34], which was shown to inhibit matrix metalloproteinase-2 [44]. More work needs to be done to elucidate whether MMC increases the binding of enzyme to substrate, or the binding of PCOLCE to PCP, and the role of the 16 kDa PCOLCE fragment, but it is most likely that MMC influences all of these processes. We have found an increased deposition of other collagens and ECM components across a range of cell types that benefit from MMC, as well as fibronectin that does not require proteolytic trimming, but depends on the extracellular regulation of disulfide exchange to form a matrix [45]. Further work needs to be done to evaluate the effects of MMC on other matrix systems like microfibrilelastin complexes.

Fig. 3. Enhanced collagen I and fibronectin deposition under different MMC modalities in WI-38 fibroblast monolayer cultures. (a) densitometric analysis of SDS-PAGE gels of pepsin treated cell layers. After 6 days, MMC enhanced the deposition of collagen I markedly with the highest yield of deposited collagen I using the negatively charged crowder, DxS. Mixed Fc-crowding using two different Ficoll species also spurred collagen I deposition. (b) representative immunocytochemistry images of deposited collagen I and fibronectin shows a reticular pattern of deposited proteins using mixed Fc-crowding and a granular pattern using DxS. Immunocytochemical images from Peng & Raghunath [43] with permission.

increase in procollagen conversion in the presence of DxS using antibodies against the procollagen C-propeptide and the central triple helical portion of the collagen molecule [33,34]. Control cultures contained only procollagen in the medium, with little collagen detected in the cell layer (Fig. 4). In the presence of DxS, procollagen in the medium was not detectable, but procollagen processed to

3.4.2. MMC accelerates supramolecular assembly As a second mode of MMC action, we propose the enhanced supramolecular assembly of collagen triple helices to collagen fibers. Supramolecular assemblies have been studied in the past under crowding. For example, MMC results in drastically increased rates of aggregation and fibrillation of human α-synuclein. This is of particular medical importance because the aggregation of α-synuclein has been found to be closely linked to the development of Parkinson's disease. An excellent review [23] recorded several experiments in which the oligomerization of actin [46], spectrin [47], tubulin [48] and FtsZ-GDP [49] were all augmented under crowded conditions, as was the selfassociation of fibrinogen [48]. In order to isolate the process of supramolecular assembly, we used fluorescently labelled telocollagen I so that no enzymatic processing was necessary to promote supramolecular assembly. When added to fibroblast cultures, we noticed an increased incorporation of the labelled collagen populations into the growing ECM under crowding (Fig. 5a). In a cell-free system, turbidimetry analyses of collagen gel formation under

Fig. 4. Enhanced cleavage of the procollagen C-propeptide under MMC. Medium samples (pooled triplicates) were resolved on a 3–8% gradient NuPAGE gel, transferred to a nitrocellulose membrane and immunoblotted for procollagen C-propeptide which even after removal from the procollagen triple helix remains a trimer under non-reduced conditions (therefore termed C3). Control (Ctr) samples were compared with crowded DxS, mixed Fc-crowding (Fc) samples after 2 days and 6 days of culture. For both time points the non crowded controls (ctr) show that the majority of procollagen population remains unconverted, while the DxS crowded samples show a clear shift of the equilibrium towards the cleaved procollagen molecules, resulting in collagen α bands (not visible in this blot, but putative position indicated by grey arrow) and removed C3 peptides, still migrating as a trimer. Fc crowding shows an intermediary state of cleavage between controls and DxS crowded samples. The differences between treatments are more evident after a longer culture time of 6 days. (Original data). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Please cite this article as: C. Chen, et al., Applying macromolecular crowding to enhance extracellular matrix deposition and its remodeling in vitro for tissue engineering and..., Adv. Drug Deliv. Rev. (2011), doi:10.1016/j.addr.2011.03.003

C. Chen et al. / Advanced Drug Delivery Reviews xxx (2011) xxx–xxx

7

Fig. 5. Enhanced supramolecular assembly of collagen I under MMC. (a) 50 μg/mL FITC-labelled atelocollagen I was added to fibroblast cell cultures and allowed to incubate for 24 h. Addition of mixed Fc-crowders enhanced the incorporation of labeled collagen into the ECM of the living cell layer. (b–c) spontaneous collagen I supramolecular assembly under cellfree conditions using a gel forming assay. (b) phase-contrast microscopy images of fibril formation in a 1.5 mg/mL collagen I solution at various time points. Collagen I fibrils formed faster and were thicker under mixed Fc-crowding. 200 × magnification. (c) a turbidity assay was performed to quantify the rate of collagen I fibril formation and absorbance was read at 313 nm. Upper panel: the turbidity read-out system is dependent on the concentration of the collagen I solution. This was performed under non-crowded conditions. Middle panel: addition of mixed Fc-crowders to 2.4 mg/mL collagen I solution resulted in an increased rate and amount of fibril formation. Triplicates of each condition were measured. Lower panel: addition of mixed Fc-crowders to 1.5 mg/mL collagen I solution also caused an increase in the rate of fibril formation and amount of fibrils. Triplicates of each condition were measured. (Original data).

crowding suggest a faster supramolecular assembly, in particular, under suboptimal conditions of gel formation (Fig. 5b and c). These data strongly indicate that MMC drives supramolecular assembly of collagen. Fig. 6 suggests that supramolecular assembly might also be accelerated between other ECM components and their ligands, such as fibronectin and latent-transforming growth factor β-binding protein 1. However, more work is needed to distinguish supramolecular assembly effects from those on enzymatic processes in greater detail. 3.4.3. MMC accelerates ECM stabilization The next level of matrix enhancement or accelerated matrix maturation would be the stabilisation of the matrices by enzymatic and non-enzymatic crosslinking to protect it from proteolytic attack. Looking into enzymatic processes, we studied the effects of lysyl oxidase activity, that gives rise to covalent crosslinks after deamidation of lysine and hydroxylysine residues of collagen α-chains (reviewed in [50]). Densitometric analyses of SDS-PAGE gels of pepsin extracted collagen matrices produced by fibroblasts under MMC, indicated an increased ratio of β bands versus α bands suggesting an increased lysyl oxidase-mediated crosslinking of the deposited collagen. The presence of these β bands was diminished by the introduction of β-aminopropionitrile, a lysyl oxidase inhibitor, into the culture medium [30,33]. Interestingly, earlier work had suggested the use of recombinant lysyl oxidase to stabilise engineered tissue [51] or to introduce copper ions or nanoparticles to stimulate lysysl oxidase activity [52,53]. Our data suggest that matrix-producing

cells release sufficient lysyl oxidase into the peri/extra cellular space in vitro, and that supplementation with a recombinant enzyme might not be necessary if cell cultures are kept under MMC (Fig. 7). Another interesting enzyme is transglutaminase 2 (TGase2), a matrix crosslinking enzyme (reviewed in [54–56]) that targets several matrix proteins like fibronectin [45], collagen VII [55], latent transforming growth factor binding protein 1 [57,58] and collagen I (Paul et al., unpublished). Using an activity assay [59], we found that ECM deposited in vitro always carries associated TGase2 activity. In relation to a higher amount of matrix deposited under MMC, we observed an increase of endogenous TGase2 activity incorporated into the ECM (Fig. 7). This points to an increased presence of enzyme along with its activity within the deposited matrix and suggests an increased stabilisation of ECM. 3.4.4. Increased remodeling of ECM under MMC Finally, remodeling of ECM occurs through tightly regulated matrix metalloproteinase (MMP) activity and the picture gets more complex when tissue inhibitors of MMPs (TIMPs) and related molecules sequestered in the ECM, on cell surfaces and present in fetal calf serum are taken into consideration. We have made the initial observation that a TIMP-like fragment of PCOLCE is generated under MMC but further studies need to demonstrate tighter interactions of TIMPs and MMPs under MMC. Indirect evidence for a support of remodeling comes from in vitro studies of matrix remodeling under adipogenic differentiation of human mesenchymal stem cells (hMSCs). hMSCs

Please cite this article as: C. Chen, et al., Applying macromolecular crowding to enhance extracellular matrix deposition and its remodeling in vitro for tissue engineering and..., Adv. Drug Deliv. Rev. (2011), doi:10.1016/j.addr.2011.03.003

8

C. Chen et al. / Advanced Drug Delivery Reviews xxx (2011) xxx–xxx

Fig. 6. Accelerated ECM deposition under MMC in different cell types. (a) collagen II deposited by porcine chondrocytes in the absence and presence (b) of mixed Fc-crowding. (c) collagen IV deposited by murine transformed astrocytes (Neu7) in the absence and presence (d) of mixed Fc-crowding. (e) deposition of latent transforming growth factor β binding protein 1 in WI-38 fibroblasts in the absence and (f) presence of mixed Fc-crowding. (g) fibulin-2 deposition in WI-38 fibroblasts in the absence and presence (h) of Ficoll crowding. (i) De novo sythesised fibronectin deposition by bone marrow derived MSCs in culture medium containing 0% serum 16 h after seeding and (j) under mixed Fc-crowding. (k) decellularised keratinocyte matrix stained for collagen IV after detergent removal of cells and (l) after 2 days of crowding with poly(sodium 4-styrene sulfonate). (Original data).

deposit a matrix that is rich in collagen I with an admixture of collagen IV, and this process is enhanced by MMC (data not shown). Under adipogenic induction using routine protocols, remodeling occurs whereby the collagen I matrix becomes scanty and the collagen IV matrix becomes abundant and is rearranged to form cocoons to enclose the newly differentiating adipocytes [60, in preparation]. Under MMC, this remodeling appears pronounced and the enhanced presence of collagen IV in the matrix is reflected by a downregulation of collagen IV mRNA, a clear sign of matrix sensing and negative feedback (Fig. 8). It is conceivable that the interactions between ECM, TIMPS and MMPs are strongly influenced by MMC, and more work needs to be done to unravel the pathways that are most influenced by it. In summary, MMC works on ECM formation and maintenance in vitro on multiple levels that are all based on protein–protein interactions (Fig. 9). 4. The role of MMC in building microenvironments for stem cell culture 4.1. Using MMC-constructed decellularised matrices Stem cells are proposed to reside in a microenvironment or “niche” composed of soluble factors, cellular components and solid ECM components. In cooperation, these components help to support stem cell self-renewal, multipotentiality via cell fate determination and preconditioning of progeny daughter cells, and also facilitates differentiation in response to appropriate signals [60–63]. Conventionally, ex vivo expansion of hMSCs is done on tissue culture polystyrene (TCPS), which is often accompanied by a progressive decrease in proliferative rate before reaching a complete stop after 22–23 cell doublings [11]. This limits the generation of therapeutically

relevant cell numbers in vitro. In addition, hMSCs lose multipotentiality with propagation on TCPS. They spontaneously commit to a particular cell lineage resulting in the production of a heterogeneous differentiated progeny with diminished self-renewal capacity, rather than the production of identical daughter stem cells [61,64]. Again, this curbs the usefulness of the cells for regenerative medicine purposes. The loss of stem cell properties of hMSCs propagated on TCPS strongly indicates that the marrow-like microenvironment plays a crucial role in the maintenance of MSC self-renewal. The ability to engineer a specialized microenvironment or niche mimicking the bone marrow ECM to support self-renewal and enhance the proliferation rate of hMSCs while maintaining multipotentiality during extensive ex vivo expansion has recently come into the spotlight, and will be of great benefit for tissue engineering applications involving hMSCs. Work done by Lai et al. promoted the proliferation and maintenance of hMSCs by culture on decellularised marrow stomal cell-derived ECM [65]. Sun et al. demonstrated the capability of ECM created by younger MSCs to rejuvenate older MSCs by reducing reactive oxygen species levels, enriching cultures for cells exhibiting higher levels of telomerase activity, and enhancing bone forming capacity [66]. Acellular human ECM-based substrates generated by human fibroblasts and decellularised, were also used to improve the propagation of human embryonic and induced pluripotent stem cells over 15 passages without the need for conditioned medium and with reduced bFGF supplementation [67]. We used MMC to enhance ECM creation by fibroblasts, decellularised these matrices by detergent treatment and used these matrices to support the propagation of hMSCs. We observed increased proliferation rates (shorter doubling time) of hMSCs cultured on ECM formed under MMC (Fig. 10a), while replication on plastic decreased

Please cite this article as: C. Chen, et al., Applying macromolecular crowding to enhance extracellular matrix deposition and its remodeling in vitro for tissue engineering and..., Adv. Drug Deliv. Rev. (2011), doi:10.1016/j.addr.2011.03.003

C. Chen et al. / Advanced Drug Delivery Reviews xxx (2011) xxx–xxx

9

Fig. 7. Enhancing effects of MMC on crosslinking events in vitro. (a) DxS- and mixed Fc-crowding lead to an enhanced β:α band ratio in SDS-PAGE gels suggesting enhanced lysyloxidase mediated activity. (b) a histochemical transglutaminase assay was conducted to visualise endogenous enzymatic activity. The irreversible incorporation of biotinylated cadaverine into the matrix was visualised using DTAF-labelled streptavidin and reveals increased TGase2 activity within the granular ECM deposits achieved under DxS crowding and the reticular deposits under Fc crowding. (Original data).

Fig. 8. Remodeling of the ECM during adipogenic differentiation. Naïve MSCs (passage 4) were induced to undergo adipogenesis for 3 weeks in the absence (− MMC) or presence (+MMC) of mixed Fc-crowding (mMMC). (a) immunostaining reveals stronger deposition of collagen IV under crowded conditions, and the restructuring of the collagen IV matrix under adipogenic differentiation into a honeycomb appearance. Under mMMC, the restructuring is still visible but more material is deposited. (b) real-time PCR analysis of the expression of ECM marker, collagen IV, reveals a marked down regulation of specific mRNA suggesting a negative feed-back loop (n = 3; error bars are ± s.d.; * P b 0.05). (Original data).

Please cite this article as: C. Chen, et al., Applying macromolecular crowding to enhance extracellular matrix deposition and its remodeling in vitro for tissue engineering and..., Adv. Drug Deliv. Rev. (2011), doi:10.1016/j.addr.2011.03.003

10

C. Chen et al. / Advanced Drug Delivery Reviews xxx (2011) xxx–xxx

Fig. 9. Schematic representation of key events in collagen matrix deposition that are either known or proposed to be enhanced by MMC. As data on remodeling are currently scarce the suggested events in this diagram await experimental verification.

Fig. 10. Propagation of hMSCs on WI-38 fibroblast matrix deposited under DxS crowding retains their proliferative and adipogenic potential after long-term ex vivo culture. (a) hMSCs were propagated on either TCPS or matrix for 60 days. Plates were fixed and stained with DAPI every 7 days then assessed for cell numbers via adherent cytometry. (b) Brightfield images at 200 × magnification of hMSCs propagated for 5 passages on either TCPS or Matrix. (c) hMSCs were propagated on either TCPS or matrix for 28 days, then reseeded on TCPS and adipogenically induced. Induced monolayers were fixed and co-stained with DAPI and Nile red and assessed via adherent cytometry for area of stained lipid droplets. (d) Nile red stained monolayers of induced cultures (20 × magnification). (Original data).

Please cite this article as: C. Chen, et al., Applying macromolecular crowding to enhance extracellular matrix deposition and its remodeling in vitro for tissue engineering and..., Adv. Drug Deliv. Rev. (2011), doi:10.1016/j.addr.2011.03.003

C. Chen et al. / Advanced Drug Delivery Reviews xxx (2011) xxx–xxx

(longer doubling time). By day 45, hMSCs on decellularised matrices had reached passage 7 while those on TCPS were one passage behind. In addition, hMSCs propagated on matrix displayed a more spindleshaped morphology indicative of a preserved stemness profile compared to those on TCPS which had a more flattened and spreadout morphology (Fig. 10b). Adipogenesis of hMSCs that had been serially propagated on matrix for 4 weeks resulted in 23% of the population differentiating compared with 12.5% of those propagated on TCPS (Fig. 10c), and this was visually confirmed by the presence of more lipid droplets in cell monolayers induced on decellularised matrices (Fig. 10d). A similar application of MMC was used to rapidly create microenvironments for human embryonic stem cells (hESCs) in our laboratory. Typical culture of hESCs requires the co-culture of feeder cells or a commercial gel matrix, matrigel [68,69]. The presence of feeder cells would require their subsequent removal, thereby adding additional preparatory steps. Matrigel, on the other hand, is derived from mouse sarcoma tissue [70], which carries a xenogenic threat [71]. We created bioassembled matrices by applying MMC on fibroblasts to enhance matrix deposition, followed by decellularisation with detergent lysis. These matrices are of human origin, and created under low serum conditions, reducing xenogenic exposure of hESCs. The decellularised matrices supported more population doublings (42.5% increase) of hESCs compared to the matrigel control in the same period of time (Fig. 11a). Besides the increase in proliferation, hESCs maintained pluripotent morphologies (Fig. 11b) and pluripotent marker expression (SSEA-3 and SSEA-4) comparable to controls. The hESCs were also able to form teratomas in SCID mice, hence demonstrating retention of differentiation capacity, while retaining normal karyotypes. Thus, our bioassembled matrices were able to retain desirable pluripotent states of hESCs while increasing the attainable number of population doublings, highlighting the crucial role that MMC plays during matrix deposition (Peng et al., in preparation). In contrast to work done by Abraham et al., which required 6–8 days of culture past 100% confluency to obtain deposited ECM [67,72], the creation of matrices under MMC required 72 h with subconfluent cell culture conditions. Additionally, matrices that were made under non-crowded conditions within the same time frame were unable to sustain the pluripotent proliferation of hESCs, highlighting the crucial role that MMC plays during matrix deposition. Besides improving the propagation of stem cells in culture, ECM also plays a role in stem cell differentiation. Decellularised in vitro generated mineralized ECM [73] and bone-specific ECM derived from an osteogenic cell line [74], were shown to direct the osteogenic differentiation of hMSCs and murine ESCs, respectively. Similarly, we

11

have found that decellularised adipocyte matrix directs the adipocyte differentiation of hMSCs without the addition of induction supplements (Loe et al., in preparation). Remarkably, matrices created under MMC were the most efficacious. Hence, MMC can also be applied to rapidly create cell-specific matrices capable of orchestrating differentiation. 4.2. Growing stem cells directly under MMC The successful propagation of hMSCs on human biossembled fibroblast matrices suggested that MSCs might do very well on ECM they have deposited themselves. Using mixed macromolecular crowding with Ficoll, we predicted that hMSCs would build their own microenvironment in the undifferentiated state and rebuild and remodel it during differentiation. In fact, we observed that undifferentiated hMSCs under MMC built a matrix richer in fibronectin (compare Fig. 6i and j) and collagen I (data not shown), and also proliferated faster (unpublished observation). When induced into differentiation, we not only observed lineage specific remodeling of the ECM (compare Fig. 8) but also a faster differentiation speed and greater yield (percentage of cells differentiating) when osteogenesis and in particular, adipogenesis were induced (Fig. 12). 5. Where and when does MMC not work? From the point of microenvironment building, MMC is primarily dependent on the synthetic capacity of the cells. For collagen biosynthesis, an important cofactor like ascorbic acid is a crucial prerequisite to achieve full post-translational modifications of prolyl and lysyl residues resulting in thermostability, secretion and extracellular crosslinking [33]. Cells that do not produce much ECM like hESCs for example, cannot be induced to build a microenvironment even under MMC, and therefore need to rely on a preformed microenvironment. As we are working in the constraints of currently available cell culture plastic materials and their surface properties, we have observed that negatively charged crowders do not work well with cells that primarily do not adhere firmly, such as endothelial cells, for example, even if the cells were allowed to attach for 16 h prior to crowding. These cells tend to detach quickly in the presence of DxS, but adhere well even when seeded in the presence of neutral Ficoll. We assume a competition of charges during the adhesion process when negatively charged crowders are present. The addition of negatively charged DxS to porcine MSCs was observed to induce partial differentiation into chondrocytes and osteoblasts [75]. In our hands and with human cells, however, we saw variable results and

Fig. 11. hESCs grow better on a human bioassembled matrix made under MMC. hESC plated on WI-38 matrix created under DxS MMC showed stable colonies with preserved proliferative effects and retained pluripotent characteristics. (a) matrix-propagated hESCs proliferated at a faster rate than control hESCs cultured on Matrigel. (b) matrix propagated hESCs exhibited pluripotent morphology, showing that in spite of the increased proliferation, bioassembled matrix was able to maintain hESCs in a pluripotent state. (Original data).

Please cite this article as: C. Chen, et al., Applying macromolecular crowding to enhance extracellular matrix deposition and its remodeling in vitro for tissue engineering and..., Adv. Drug Deliv. Rev. (2011), doi:10.1016/j.addr.2011.03.003

12

C. Chen et al. / Advanced Drug Delivery Reviews xxx (2011) xxx–xxx

Fig. 12. Differentiation of human bone marrow-derived MSCs is enhanced under mixed Fc-crowding (mMMC). (a) passage 4 Lonza hMSCs were induced adipogenically on tissue culture polystyrene. Under mMMC, cells exhibited a higher content of lipid droplets. (b) when subjected to a standard osteogenic induction cocktail, hMSCs from the same passage and batch differentiated faster and more efficiently into osteoblasts as demonstrated here using Alizarin Red as calcification indicator. (Original data). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

mostly growth inhibition in the presence of DxS, but excellent proliferation and differentiation responses under neutral MMC (Fig. 12). However, we have not been able to direct monolyers of hMSCs into chondrogenic differentiation (as gauged by collagen II deposition), although primary chondrocyte cultures do very well under Ficoll crowding (Fig. 6). As we have used relatively late passages (Lonza, passage 4 and later) the picture might change considerably by using freshly isolated hMSCs. It also should be noted that DxS exerts growth factor binding effects [76]. This and similarly negatively charged crowders may therefore disturb the equilibrium of growth factors. Thus, negatively charged crowders cannot be viewed as chemically inert in the strict sense and this may or may not be beneficial for the respective culture system they are used in. Therefore, using neutrally charged macromolecular crowders might be preferable. Again, only the empirical approach can determine which crowders work best in a given cell culture system. 6. The future of applied macromolecular crowding Obviously, MMC is an ancient biological principle that has enabled multicellular life, as we know it. Perplexingly, this principle has been largely ignored in the daily operational life of biochemists, cell biologists, and tissue engineers. While a limited group of scientists has been studying the effects of MMC from several angles, our group has set out to translate the existing body of knowledge on MMC into practice. In general, the potency of MMC has not been fully fathomed, neither from the modelling side nor from its applications in cell biology. For example, mixed MMC offers a vast amount of permutations that can combine two, three and more macromolecular

crowders of disparate charge and hydrodynamic radii. As computation and mathematics get extremely challenging to describe events under mixed MMC, the cell biologist and the biochemist must go ahead with empirical work until modelling work has caught up. The increased performance of cell culture under these conditions is highly rewarding, and it was the existing literature on mixed macromolecular crowding [21] that inspired us to explore applications in the realm of DNA hybridisation [77] and stem cell work with highly interesting results (in preparation). One of the current challenges faced by tissue engineers, is that tissue constructs need to be larger and more complex. We envision that macromolecular crowding can enhance both static and bioreactor settings by inducing cells to coat polymeric biomaterials with their own matrix. Organ printing is an emerging field, and it has enabled the formation of capillary-like tubes from pellets of endothelial cells [10]. The biggest obstacle to this technique is keeping the printed cell constructs stable while they produce their own ECM. Although printing into hydrogels promises to generate more stable cell constructs, an intrinsic ECM must be produced to finally stabilise the construct. Hydrogels therefore will present the next challenge for applied MMC. Despite their high water content that would offer ample room for crowding, mesh sizes and diffusibility of crowders come into play in these confined spaces, and far lower concentrations might apply. The application of MMC to create ECM-rich cell sheets that can be lifted off from thermosensitive polymer coatings is certainly very appealing. The cell sheet technology, pioneered by Teruo Okano and coworkers is dependent on a basic ECM containing fibronectin and collagen V that lend cohesion to a hyperconfluent mono/oligolayer of cell sheets [78,79], so that it lifts off in toto when the temperature is lowered. In our hands, however, the extensively

Please cite this article as: C. Chen, et al., Applying macromolecular crowding to enhance extracellular matrix deposition and its remodeling in vitro for tissue engineering and..., Adv. Drug Deliv. Rev. (2011), doi:10.1016/j.addr.2011.03.003

C. Chen et al. / Advanced Drug Delivery Reviews xxx (2011) xxx–xxx

deposited and matured ECM appears to prevent a temperaturesensitive separation from commercially available materials, as the adherence of the cells to their substrate is too strong. We look forward to further developments in thermosensitive materials and their improved grafting densities on TCPS that will finally allow a combination of these two technologies. There is successful scaffoldfree tissue engineering work out there, for example on small blood vessels [80] that might benefit from crowding, and it will truly be a challenge to apply MMC in industrial settings to create more robust three-dimensional tissue constructs. As we have not found significant increases in viscosity of culture media containing milligram amounts of Ficoll, MMC would lend itself to flow-based culture systems. We have published preliminary data on bioreactor systems [34] but are convinced that more experienced groups can take MMC to greater heights in macro and possibly, microfluidics systems. Looking beyond cell culture, we have shown that PCR and related techniques benefit from MMC [77,81], but the potential of MMC in antibody-antigen interactions has not been fathomed, and neither have ligand-receptor interactions. 7. Conclusion Macromolecular crowding as a biotechnology feature is coming of age and will transform from an arcane subject once pursued by a relatively select group of scientists, to a larger audience in translational biology and medicine. This is reflected by the creation of a new subgroup “Biopolymer Biophysics in vivo” within the Biophysical Society (http://www.biophysics.org). We look forward to continued reciprocal dialogue between modelling, prediction, highly selective and precise measurements and the empirical day-to-day work in complex biochemical and biological systems. Clearly, by observing nature and emulating biological principles like macromolecular crowding, we will, in many cases, see a highly rewarding increased cell performance. Acknowledgements We would like to apologise to all, whose work on MMC were not cited in this review due to space constraints. We would like to encourage scientists that find MMC appealing for their system to join us in the BPS subgroup and to get into a dialogue with us. We owe particular thanks to Dr Karthik Harve Subramhanya (NUS) for critical reading and editing of the manuscript. Keratinocyte data were contributed by Ms Paula Benny, porcine chondrocyte data by Mr. Christian Leicht, transglutaminase data by Mr Pradeep Paul. We thank Ms Stella Chee for excellent technical assistance. The authors acknowledge support by the NUS Tissue Engineering Programme (Life Science Institute NUS), the Faculty Research Committee grants from the National Medical Research Council of Singapore (NMRC grant CPG/003/2004), the Faculty of Engineering, Office of Research FRC grants R-397-000-017-112 and R-397-000-081-112 and a startup grant from the Office of Life Science (R-397-000-604-712) and the Provost of NUS (R-397-000-604-101), the NUS-Baden-Wuerttemberg grant (R-397-000-080-646) and the Singapore MIT Alliance for Research Technology (SMART) Innovation Centre. References [1] M.K. Vickaryous, B.K. Hall, Human cell type diversity, evolution, development, and classification with special reference to cells derived from the neural crest, Biological reviews of the Cambridge Philosophical Society 81 (2006) 425–455. [2] B. Young, J.S. Lowe, A. Stevens, J.S. Lowe, Wheater's functional histology: a text and colour atlas, 4th edition ed., Churchill Livingstone, Sydney, 2000. [3] R.I. Freshney, G. Vunjak-Novakovic, Culture of cells for tissue engineering, in: R. Ian Freshney (Ed.), Basic Principles of Cell Culture, John Wiley and Sons, USA, 2006.

13

[4] S.J. Bryant, K.S. Anseth, The effects of scaffold thickness on tissue engineered cartilage in photocrosslinked poly(ethylene oxide) hydrogels, Biomaterials 22 (2001) 619–626. [5] B.S. Kim, A.J. Putnam, T.J. Kulik, D.J. Mooney, Optimizing seeding and culture methods to engineer smooth muscle tissue on biodegradable polymer matrices, Biotechnology and Bioengineering 57 (1998) 46–54. [6] I. Martin, G. Vunjak-Novakovic, J. Yang, R. Langer, L.E. Freed, Mammalian chondrocytes expanded in the presence of fibroblast growth factor 2 maintain the ability to differentiate and regenerate three-dimensional cartilaginous tissue, Experimental Cell Research 253 (1999) 681–688. [7] B. Obradovic, R.L. Carrier, G. Vunjak-Novakovic, L.E. Freed, Gas exchange is essential for bioreactor cultivation of tissue engineered cartilage, Biotechnology and Bioengineering 63 (1999) 197–205. [8] C.W. Patrick Jr., P.B. Chauvin, J. Hobley, G.P. Reece, Preadipocyte seeded PLGA scaffolds for adipose tissue engineering, Tissue Engineering 5 (1999) 139–151. [9] B.P. Chan, K.W. Leong, Scaffolding in tissue engineering: general approaches and tissue-specific considerations, European Spine Journal 17 (Suppl. 4) (2008) 467–479. [10] V. Mironov, R.P. Visconti, V. Kasyanov, G. Forgacs, C.J. Drake, R.R. Markwald, Organ printing: tissue spheroids as building blocks, Biomaterials 30 (2009) 2164–2174. [11] A. Banfi, A. Muraglia, B. Dozin, M. Mastrogiacomo, R. Cancedda, R. Quarto, Proliferation kinetics and differentiation potential of ex vivo expanded human bone marrow stromal cells: implications for their use in cell therapy, Experimental Hematology 28 (2000) 707–715. [12] J.R. Mauney, V. Volloch, D.L. Kaplan, Matrix-mediated retention of adipogenic differentiation potential by human adult bone marrow-derived mesenchymal stem cells during ex vivo expansion, Biomaterials 26 (2005) 6167–6175. [13] N. Blow, High-throughput screening: designer screens, Nature Methods 6 (2009) 105–108. [14] A.B. Fulton, How crowded is the cytoplasm? Cell 30 (1982) 345–347. [15] A.P. Minton, Confinement as a determinant of macromolecular structure and reactivity. II. Effects of weakly attractive interactions between confined macrosolutes and confining structures, Biophysical Journal 68 (1995) 1311–1322. [16] R.J. Ellis, A.P. Minton, Cell biology: join the crowd, Nature 425 (2003) 27–28. [17] C. Ebel, G. Zaccai, Crowding in extremophiles: linkage between solvation and weak protein–protein interactions, stability and dynamics, provides insight into molecular adaptation, Journal of Molecular Recognition 17 (2004) 382–389. [18] R.J. Ellis, Macromolecular crowding: obvious but underappreciated, Trends in Biochemical Sciences 26 (2001) 597–604. [19] S.B. Zimmerman, A.P. Minton, Macromolecular crowding: biochemical, biophysical, and physiological consequences, Annual Review of Biophysics and Biomolecular Structure 22 (1993) 27–65. [20] D. Hall, A.P. Minton, Macromolecular crowding: qualitative and semiquantitative successes, quantitative challenges, Biochimica et Biophysica Acta 1649 (2003) 127–139. [21] M.S. Cheung, D. Klimov, D. Thirumalai, Molecular crowding enhances native state stability and refolding rates of globular proteins, Proceedings of the National Academy of Sciences of the United States of America 102 (2005) 4753–4758. [22] R. Goobes, N. Kahana, O. Cohen, A. Minsky, Metabolic buffering exerted by macromolecular crowding on DNA–DNA interactions: origin and physiological significance, Biochemistry 42 (2003) 2431–2440. [23] A.P. Minton, Models for excluded volume interaction between an unfolded protein and rigid macromolecular cosolutes: macromolecular crowding and protein stability revisited, Biophysical Journal 88 (2005) 971–985. [24] R.J. Ellis, Macromolecular crowding: an important but neglected aspect of the intracellular environment, Current Opinion in Structural Biology 11 (2001) 114–119. [25] A.P. Minton, The influence of macromolecular crowding and macromolecular confinement on biochemical reactions in physiological media, The Journal of Biological Chemistry 276 (2001) 10577–10580. [26] A.P. Minton, Macromolecular crowding and molecular recognition, Journal of Molecular Recognition 6 (1993) 211–214. [27] N. Tokuriki, M. Kinjo, S. Negi, M. Hoshino, Y. Goto, I. Urabe, T. Yomo, Protein folding by the effects of macromolecular crowding, Protein Science 13 (2004) 125–133. [28] C.C. Gyenge, O. Tenstad, H. Wiig, In vivo determination of steric and electrostatic exclusion of albumin in rat skin and skeletal muscle, The Journal of Physiology 552 (2003) 907–916. [29] J.M. Zheng, G.H. Pollack, Long-range forces extending from polymer-gel surfaces, Physical Review 68 (2003) 031408. [30] R.R. Lareu, I. Arsianti, H.K. Subramhanya, P. Yanxian, M. Raghunath, In vitro enhancement of collagen matrix formation and crosslinking for applications in tissue engineering: a preliminary study, Tissue Engineering 13 (2007) 385–391. [31] O.G. Berg, The influence of macromolecular crowding on thermodynamic activity: solubility and dimerization constants for spherical and dumbbell-shaped molecules in a hard-sphere mixture, Biopolymers 30 (1990) 1027–1037. [32] H.J. Guttman, C.F. Anderson, M.T. Record Jr., Analyses of thermodynamic data for concentrated hemoglobin solutions using scaled particle theory: implications for a simple two-state model of water in thermodynamic analyses of crowding in vitro and in vivo, Biophysical Journal 68 (1995) 835–846. [33] C.Z. Chen, Y.X. Peng, Z.B. Wang, P.V. Fish, J.L. Kaar, R.R. Koepsel, A.J. Russell, R.R. Lareu, M. Raghunath, The Scar-in-a-Jar: studying potential antifibrotic compounds from the epigenetic to extracellular level in a single well, British Journal of Pharmacology 158 (2009) 1196–1209. [34] R.R. Lareu, K.H. Subramhanya, Y. Peng, P. Benny, C. Chen, Z. Wang, R. Rajagopalan, M. Raghunath, Collagen matrix deposition is dramatically enhanced in vitro when crowded with charged macromolecules: the biological relevance of the excluded volume effect, FEBS Letters 581 (2007) 2709–2714.

Please cite this article as: C. Chen, et al., Applying macromolecular crowding to enhance extracellular matrix deposition and its remodeling in vitro for tissue engineering and..., Adv. Drug Deliv. Rev. (2011), doi:10.1016/j.addr.2011.03.003

14

C. Chen et al. / Advanced Drug Delivery Reviews xxx (2011) xxx–xxx

[35] G. Ralston, Effects of crowding in protein solutions, Journal of Chemical Education 10 (1990) 857–860. [36] A.P. Minton, How can biochemical reactions within cells differ from those in test tubes? Journal of Cell Science 119 (2006) 2863–2869. [37] S. Murad, D. Grove, K.A. Lindberg, G. Reynolds, A. Sivarajah, S.R. Pinnell, Regulation of collagen synthesis by ascorbic acid, Proceedings of the National Academy of Sciences of the United States of America 78 (1981) 2879–2882. [38] J.F. Bateman, W.G. Cole, J.J. Pillow, J.A. Ramshaw, Induction of procollagen processing in fibroblast cultures by neutral polymers, The Journal of Biological Chemistry 261 (1986) 4198–4203. [39] Y. Hojima, B. Behta, A.M. Romanic, D.J. Prockop, Cleavage of type I procollagen by C- and N-proteinases is more rapid if the substrate is aggregated with dextran sulfate or polyethylene glycol, Analytical Biochemistry 223 (1994) 173–180. [40] M. Valli, F. Zolezzi, M. Mottes, F. Antoniazzi, F. Stanzial, R. Tenni, P. Pignatti, G. Cetta, Gly85 to Val substitution in pro alpha 1(I) chain causes mild osteogenesis imperfecta and introduces a susceptibility to protease digestion, European Journal of Biochemistry FEBS 217 (1993) 77–82. [41] K.S. Harve, M. Raghunath, R.R. Lareu, R. Rajagopalan, Macromolecular crowding in biological systems: dynamic light scattering (DLS) to quantify the excluded volume effect (EVE), Biophysical Reviews and Letters 1 (2006) 317–325. [42] Z. Wang, C. Chen, S.N. Finger, S. Kwajah, M. Jung, H. Schwarz, N. Swanson, F.F. Lareu, M. Raghunath, Suberoylanilide hydroxamic acid: a potential epigenetic therapeutic agent for lung fibrosis? European Respiratory Journal 34 (2009) 145–155. [43] Y. Peng, M. Raghunath, Learning from Nature: emulating macromolecular crowding to drive extracellular matrix enhancement for the creation of connective tissue, in: D. Eberli (Ed.), Tissue Engineering, In Teh, Vukovar, Croatia, 2010, pp. 103–118. [44] J.D. Mott, C.L. Thomas, M.T. Rosenbach, K. Takahara, D.S. Greenspan, M.J. Banda, Post-translational proteolytic processing of procollagen C-terminal proteinase enhancer releases a metalloproteinase inhibitor, The Journal of Biological Chemistry 275 (2000) 1384–1390. [45] D.F. Mosher, F.J. Fogerty, M.A. Chernousov, E.L. Barry, Assembly of fibronectin into extracellular matrix, Annals of the New York Academy of Sciences 614 (1991) 167–180. [46] R.A. Lindner, G.B. Ralston, Macromolecular crowding: effects on actin polymerisation, Biophysical Chemistry 66 (1997) 57–66. [47] R. Lindner, G. Ralston, Effects of dextran on the self-association of human spectrin, Biophysical Chemistry 57 (1995) 15–25. [48] G. Rivas, J.A. Fernandez, A.P. Minton, Direct observation of the self-association of dilute proteins in the presence of inert macromolecules at high concentration via tracer sedimentation equilibrium: theory, experiment, and biological significance, Biochemistry 38 (1999) 9379–9388. [49] G. Rivas, J.A. Fernandez, A.P. Minton, Direct observation of the enhancement of noncooperative protein self-assembly by macromolecular crowding: indefinite linear self-association of bacterial cell division protein FtsZ, Proceedings of the National Academy of Sciences of the United States of America 98 (2001) 3150–3155. [50] C.M. Kielty, M.E. Grant, The collagen family: structure, assembly, and organization in the extracellular matrix, in: P.M. Royce, B. Steinmann (Eds.), Connective Tissue and its Heritable disorders, Wiley-Liss, New York, 2002, pp. 159–222. [51] W.M. Elbjeirami, E.O. Yonter, B.C. Starcher, J.L. West, Enhancing mechanical properties of tissue-engineered constructs via lysyl oxidase crosslinking activity, Journal of Biomedical Materials Research 66 (2003) 513–521. [52] S.L. Dahl, R.B. Rucker, L.E. Niklason, Effects of copper and cross-linking on the extracellular matrix of tissue-engineered arteries, Cell Transplantation 14 (2005) 861–868. [53] C.R. Kothapalli, A. Ramamurthi, Copper nanoparticle cues for biomimetic cellular assembly of crosslinked elastin fibers, Acta Biomaterialia 5 (2009) 541–553. [54] M. Griffin, R. Casadio, C.M. Bergamini, Transglutaminases: nature's biological glues, The Biochemical Journal 368 (2002) 377–396. [55] M. Raghunath, B. Hopfner, D. Aeschlimann, U. Luthi, M. Meuli, S. Altermatt, R. Gobet, L. Bruckner-Tuderman, B. Steinmann, Cross-linking of the dermoepidermal junction of skin regenerating from keratinocyte autografts. Anchoring fibrils are a target for tissue transglutaminase, The Journal of Clinical Investigation 98 (1996) 1174–1184. [56] S.T. Khew, P.P. Panengad, M. Raghunath, Y.W. Tong, Characterization of amine donor and acceptor sites for tissue type transglutaminase using a sequence from the C-terminus of human fibrillin-1 and the N-terminus of osteonectin, Biomaterials 31 (2010) 4600–4608. [57] M. Raghunath, C. Unsold, U. Kubitscheck, L. Bruckner-Tuderman, R. Peters, M. Meuli, The cutaneous microfibrillar apparatus contains latent transforming growth factor-beta binding protein-1 (LTBP-1) and is a repository for latent TGF-beta1, The Journal of Investigative Dermatology 111 (1998) 559–564.

[58] E. Verderio, C. Gaudry, S. Gross, C. Smith, S. Downes, M. Griffin, Regulation of cell surface tissue transglutaminase: effects on matrix storage of latent transforming growth factor-beta binding protein-1, Journal of Histochemistry & Cytochemistry 47 (1999) 1417–1432. [59] M. Raghunath, H.C. Hennies, F. Velten, V. Wiebe, P.M. Steinert, A. Reis, H. Traupe, A novel in situ method for the detection of deficient transglutaminase activity in the skin, Archives of Dermatological Research 290 (1998) 621–627. [60] R. Schofield, The relationship between the spleen colony-forming cell and the haemopoietic stem cell, Blood Cells 4 (1978) 7–25. [61] C.M. Digirolamo, D. Stokes, D. Colter, D.G. Phinney, R. Class, D.J. Prockop, Propagation and senescence of human marrow stromal cells in culture: a simple colony-forming assay identifies samples with the greatest potential to propagate and differentiate, British Journal of Haematology 107 (1999) 275–281. [62] K.A. Moore, I.R. Lemischka, Stem cells and their niches, Science (New York, N.Y.) 311 (2006) 1880–1885. [63] C.A. Gregory, J. Ylostalo, D.J. Prockop, Adult bone marrow stem/progenitor cells (MSCs) are preconditioned by microenvironmental “niches” in culture: a twostage hypothesis for regulation of MSC fate, Science Signalling: The Signal Transduction Knowledge Environment 2005 (2005) e37. [64] D. Baksh, L. Song, R.S. Tuan, Adult mesenchymal stem cells: characterization, differentiation, and application in cell and gene therapy, Journal of Cellular and Molecular Medicine 8 (2004) 301–316. [65] Y. Lai, Y. Sun, C.M. Skinner, E.L. Son, Z. Lu, R.S. Tuan, R.L. Jilka, J. Ling, X.D. Chen, Reconstitution of marrow-derived extracellular matrix ex vivo: a robust culture system for expanding large-scale highly functional human mesenchymal stem cells, Stem Cells and Development 19 (2010) 1095–1107. [66] Y. Sun, W. Li, Z. Lu, R. Chen, J. Ling, Q. Ran, R.L. Jilka, X.D. Chen, Rescuing replication and osteogenesis of aged mesenchymal stem cells by exposure to a young extracellular matrix, Faseb Journal (2011). [67] S. Abraham, S.D. Sheridan, B. Miller, R.R. Rao, Stable propagation of human embryonic and induced pluripotent stem cells on decellularized human substrates, Biotechnology Progress 26 (2010) 1126–1134. [68] J.A. Thomson, J. Itskovitz-Eldor, S.S. Shapiro, M.A. Waknitz, J.J. Swiergiel, V.S. Marshall, J.M. Jones, Embryonic stem cell lines derived from human blastocysts, Science (New York, N.Y.) 282 (1998) 1145–1147. [69] C. Xu, M.S. Inokuma, J. Denham, K. Golds, P. Kundu, J.D. Gold, M.K. Carpenter, Feeder-free growth of undifferentiated human embryonic stem cells, Nature Biotechnology 19 (2001) 971–974. [70] H.K. Kleinman, M.L. McGarvey, J.R. Hassell, V.L. Star, F.B. Cannon, G.W. Laurie, G.R. Martin, Basement membrane complexes with biological activity, Biochemistry 25 (1986) 312–318. [71] M.J. Martin, A. Muotri, F. Gage, A. Varki, Human embryonic stem cells express an immunogenic nonhuman sialic acid, Nature Medicine 11 (2005) 228–232. [72] S. Abraham, M.J. Riggs, K. Nelson, V. Lee, R.R. Rao, Characterization of human fibroblast-derived extracellular matrix components for human pluripotent stem cell propagation, Acta Biomaterialia 6 (2010) 4622–4633. [73] R.A. Thibault, L. Scott Baggett, A.G. Mikos, F.K. Kasper, Osteogenic differentiation of mesenchymal stem cells on pregenerated extracellular matrix scaffolds in the absence of osteogenic cell culture supplements, Tissue Engineering Part A 16 (2010) 431–440. [74] N.D. Evans, E. Gentleman, X. Chen, C.J. Roberts, J.M. Polak, M.M. Stevens, Extracellular matrix-mediated osteogenic differentiation of murine embryonic stem cells, Biomaterials 31 (2010) 3244–3252. [75] B.S. Noble, V. Dean, N. Loveridge, B.M. Thomson, Dextran sulfate promotes the rapid aggregation of porcine bone-marrow stromal cells, Bone 17 (1995) 375–382. [76] T. Kajio, K. Kawahara, K. Kato, Stabilization of basic fibroblast growth factor with dextran sulfate, FEBS Letters 306 (1992) 243–246. [77] K.S. Harve, R. Lareu, R. Rajagopalan, M. Raghunath, Understanding how the crowded interior of cells stabilizes DNA/DNA and DNA/RNA hybrids-in silico predictions and in vitro evidence, Nucleic Acids Research 38 (2009) 172–181. [78] T. Kihara, Y. Takemura, Y. Imamura, K. Mizuno, T. Hayashi, Reconstituted type V collagen fibrils as cementing materials in the formation of cell clumps in culture, Cell and Tissue Research 318 (2004) 343–352. [79] T. Kihara, Y. Imamura, Y. Takemura, K. Mizuno, E. Adachi, T. Hayashi, Intercellular accumulation of type V collagen fibrils in accordance with cell aggregation, Journal of Biochemistry 144 (2008) 625–633. [80] N. L'Heureux, L. Germain, R. Labbe, F.A. Auger, In vitro construction of a human blood vessel from cultured vascular cells: a morphologic study, Journal of Vascular Surgery 17 (1993) 499–509. [81] R.R. Lareu, K.S. Harve, M. Raghunath, Emulating a crowded intracellular environment in vitro dramatically improves RT-PCR performance, Biochemical and Biophysical Research Communications 363 (2007) 171–177.

Please cite this article as: C. Chen, et al., Applying macromolecular crowding to enhance extracellular matrix deposition and its remodeling in vitro for tissue engineering and..., Adv. Drug Deliv. Rev. (2011), doi:10.1016/j.addr.2011.03.003