Ultra-High Throughput Enrichment of Viable Circulating Tumor Cells

3 downloads 0 Views 2MB Size Report
rare circulating tumor cells (CTCs) from the peripheral blood of cancer patients potentially provide ... SKBR3 cell line was also used for DNA FISH. The cells.
Ultra-High Throughput Enrichment of Viable Circulating Tumor Cells B.L. Khoo2#, M.E. Warkiani1#, Guofeng Guan1, 3, Daniel Shao-Weng Tan4,5#, Alvin S.T. Lim8, WanTeck Lim4, Yoon Sim Yap4, Soo Chin Lee10, Ross A. Soo10, Jongyoon Han1,11*, Chwee Teck Lim1,2,3,6* 1

BioSystems and Micromechanics (BioSyM) IRG, Singapore-MIT Alliance for Research and Technology (SMART) Centre, Singapore 2 Mechanobiology Institute, National University of Singapore, Singapore 3 Department of Bioengineering, National University of Singapore, Singapore 4 Department of Medical Oncology, National Cancer Centre Singapore, Singapore 5 Cancer Therapeutics Research Laboratory, National Cancer Centre Singapore, Singapore 6 Clearbridge BioMedics Pte Ltd, Singapore 7 Sequenom Inc, California, USA 8 Department of Pathology, Singapore General Hospital 9 Departments of Hematology-Oncology and Pharmacology, National University Hospital, Singapore 10 Department of Hematology-Oncology, National University Hospital, Singapore 11 Department of Electrical Engineering and Computer Science, Department of Biological Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts, USA # These authors contributed equally to this paper. Abstract — Detection, enumeration and characterization of rare circulating tumor cells (CTCs) from the peripheral blood of cancer patients potentially provide critical insights into tumor biology and is promising for cancer diagnosis and prognosis. Here, we present a novel multiplexed spiral microfluidic device for ultra-high throughput, label-free enrichment of CTCs from clinically relevant blood volumes. The fast processing time of the technique (7.5 mL blood in < 5 min) and high sensitivity of the device lends itself to a broad range of potential genomic and transcriptomic applications. The method can specifically separate and preserve all fractions of blood (i.e., plasma, CTCs and PBMC) for diverse downstream analysis. CTCs were detected from 100% (10/10) of blood samples collected from patients with advanced stage metastatic breast (1256 CTC/ml) or lung cancer (30-153 CTC/ml). Cancer cells were characterized with immunostaining and fluorescence in situ hybridization (FISH) (HER2/neu). Retrieved cells were unlabelled and hence more viable for propagation and other informative analysis such as the next generation sequencing (NGS) to guide treatment and individualized patient care. Keywords— Circulating tumor cells (CTCs), Cell separation, Cancer detection, Inertial microfluidics, Point-of-care diagnostics

I. INTRODUCTION Circulating tumor cells (CTCs) are rare cancer cells of solid tumor origin found in peripheral blood of cancer patients, shed from either primary or secondary tumors. Isolating CTCs potentially provide a convenient source for disease and prognosis indicators, as well as the characterization of tumor phenotype/genotype [1]. CTC characterization will provide immense opportunities in the field of novel and personalized oncology therapeutics [2].

Despite recent technological advancements, detection and recovery of CTCs remain technically challenging due to their rarity in blood and heterogeneity [3]. Various approaches has been explored until now for identification and enrichment of these rare cells [3,4] such as affinity based approaches [5] or through use of micro-fabricated membranes [6,7,8]. Recently, high-throughput label-free cell sorting using inertial microfluidics inside curvilinear microchannels has been reported [9]. By altering microchannel cross sections, we lately have demonstrated the development of a biochip with trapezoidal cross-section [10] for ultra-fast enrichment of CTCs with extremely high sensitivity (near 100% detection rate) and specificity (1 CTC for every 100 WBC). Here, we demonstrate the multiplexed version of this chip which can critically process large volume of blood samples in a short period of time. All blood components can be fractionated (i.e., plasma, CTCs and PBMC) under high purity for various biomarker analysis studies or sensitive molecular assays such as qRT-PCR. Characterization of the enriched CTCs were demonstrated with immunofluorescence assays (Pan-cytokeratin+/CD45-) and HER2/neu fluorescence in-situ hybridization (FISH). The device is suited to process larger blood volumes (20 mL in ~ 15 min), a growing need for obtaining large number of CTCs for multiple downstream tests. II. MATERIALS

AND METHODS

A. Ethnics statement Healthy blood samples and clinical blood specimens were acquired from volunteers or patients with advanced stage metastatic breast cancer or non-small cell lung cancer

J. Goh (ed.), The 15th International Conference on Biomedical Engineering, IFMBE Proceedings 43, DOI: 10.1007/978-3-319-02913-9_1, © Springer International Publishing Switzerland 2014

1

2

B.L. Khoo et al.

(NSCLC) respectively. This study was approved by our institutional review board and local ethics committee informed and written consent was obtained from all patients. B. Microfluidic chip fabrication A master mold was specifically designed with SolidWorks software, with an 8-loop spiral microchannel with one inlet and two outlets with increasing radius (8-24 mm). Channel width trapezoid cross-section is 600 ȝm and the inner/outer height ratio was optimized at 80/130 ȝm. Microfluidic chips were fabricated with standard soft-lithography techniques as described elsewhere [10] using polydimethylsiloxane (PDMS). The multiplexed device was obtained by the bonding of three separate chips using oxygen plasma treatment and manual alignment.

Cells were stained for 30 min on ice after fixation with 4% paraformaldehyde (Sigma, USA, 10mins) and permeabilisation (0.1% Triton 100X, Thermo Scientific, USA). Cells positive for pan-CK, Hoechst and negative for CD45 with characteristic cancer cells morphology (i.e., high nucleus to cytoplasm ratio) were identified as CTCs (Fig. 2A). Cells stained positively for CD45 and Hoechst and negatively for pan-cytokeratin are determined as WBC. G. Cell viability assay Isolated CTCs were seeded onto polylysine coated 2D cell culture substrates and bright field images of cell attachment to substrate and subsequent spreading were taken. Cells were stained with propidium iodide (PI) and imaged to determine percentage of cell viability after lysis and processing.

C. Cell culture Two commercially available green fluorescent protein (GFP) –tagged human cancer cell lines, namely breast adenocarcinoma (MCF-7 and MDA-MB-231) (HTB-22TM, ATCC, USA), were used as controls in placement of CTCs. SKBR3 cell line was also used for DNA FISH. The cells were maintained as described elsewhere [10].

H. DNA fluorescence in-situ hybridization (FISH) Fluorescence in situ hybridization (FISH) was performed on HER2/neu amplified SKBR3, HER2/neu nonamplified MDA-MB-231 cells lines and isolated CTCs according to the manufacturer’s protocol [10]. III. RESULTS

D. Blood sample preparation For all experiments unless otherwise mentioned, whole blood were lysed with RBC lysis buffer (G-Bioscience, USA) for 5 min at room temperature with continuous gentle mixing. Cell pellet was obtained under centrifugation at 1000g for 5 min. The cell pellet was resuspended accordingly with PBS buffer to desired concentrations.

A. Spiral biochip design and characterisation Figure 1 illustrates the blood processing procedures in spiral microfluidic channels via hydrodynamic separation. The asymmetrical trapezoid cross-sections promotes the generation of Dean vortex cores near the outer wall of larger channel depth.

E. Device characterization In all experiments, the spiral biochip was mounted on an inverted phase contrast microscope (Olympus IX71) connected to a high speed CCD camera (Phantom v9, Vision Research Inc., USA). Cell suspension was transferred to a 10 mL syringe and processed at 1700 μL/min using a syringe pump connected to the microchannel with Tygon® tubing. F. Immunostaining Immunostaining with fluorescein isothiocyanate (FITC) conjugated pan-cytokeratin (CK) (1:100, MiltenyiBiotec Asia Pacific, Singapore), allophycocyanin (APC) conjugated CD45 marker (1:100, MiltenyiBiotec Asia Pacific, Singapore) and Hoechst was used for CTC identification.

Fig. 1 (A) Optical image of a multiplexed spiral chip. (B) Schematic representation of blood processing workflow of CTC enrichment using the trapezoidal cross section spiral microfluidic chip; CTCs are focused near the inner wall due to the combination of hydrodynamic forces while WBCs are trapped inside the core of dean vortex formed closer to the outer wall. (C) Various downstream assays which can be done on enriched CTCs.

IFMBE Proceedings Vol. 43

Ultra-High Throughput Enrichment of Viable Circulating Tumor Cells

A significant 104-fold of CTCs could be enriched with respect to red blood cells (RBCs) depleted nucleated cell fraction. Cell concentration of 0.5× concentration (< 3×106 WBCs/mL) was utilized after optimization, and 7.5ml of sample can be processed in a short period of 6 mins via a single layered chip. This is arguably the highest throughput rate to date by a CTC enrichment device. The processing speed is further reduced by multiplexing the biochips.

3

MB-231) were used as controls. The HER2/CEP17 (chromosome enumeration probe) was determined. Amplified HER2 expression means the HER2/CEP17 ratio in single nuclei is > 2.

B. Recovery rate and cell viability using cancer cell lines Two cell lines (i.e., MCF-7 and MDA-MB-231) were utilised to characterise CTC isolation and recovery rate of the biochip. A clinically relevant cell count (~ 500 cells) was seeded into 7.5 mL of whole blood obtained from healthy donors. A high recovery rate was obtained (~85% for MCF7 and 87% for MDA-MB-231 cell lines respectively). Isolated cells were re-plated onto conventional 2-D substrates to demonstrate viability via attachment. C. Clinical validation 7.5 mL of blood from 5 healthy samples (control), 5 metastatic breast cancer samples and 5 NSCLC samples were tested for device validation. CTCs were detected in 10/10 (100%) patient samples with counts ranging from 12-56 CTCs/mL (breast samples) and 30-153 CTCs/mL (NSCLC samples) (Fig. 2B). A distinct threshold of pan-CK+ detected to determine positivity of CTCs was established at