Cell Systems, Volume 5 Supplemental Information High-Throughput Microfluidic Labyrinth for the Label-free Isolation of Circulating Tumor Cells Eric Lin, Lianette Rivera-Báez, Shamileh Fouladdel, Hyeun Joong Yoon, Stephanie Guthrie, Jacob Wieger, Yadwinder Deol, Evan Keller, Vaibhav Sahai, Diane M. Simeone, Monika L. Burness, Ebrahim Azizi, Max S. Wicha, and Sunitha Nagrath
Figure S1
Figure S2
Figure S3
Figure S4
Table S1 Sample number P.1 P.2 P.3 P.4 P.5 P.6 P.7 P.8 P.9 P.10 P.11 P.12 P.13 P.14 P.15 type Staining CTC/mL Stage CK-19 18 ATDC 2 Unknown ZEB1 5 CK-19 20 ATDC 0 Unknown ZEB1 5 CK-19 23 ATDC 0 ZEB1 20 CK-19 32 Locally ATDC 0 advanced/borderline ZEB1 0 resectable CK-19 12 ATDC 0 Borderline resectable ZEB1 25 CK-19 17 ATDC 2 ZEB1 0 CK-19 55 Borderline ATDC 12 resectable/metastatic ZEB1 30 CK-19 12 Borderline ATDC 7 resectable/metastatic ZEB1 32 CK-19 10 ATDC 0 ZEB1 63 CK-19 32 ATDC 0 ZEB1 10 CK-19 12 ATDC 4 ZEB1 2 CK-19 17 Borderline ATDC 10 resectable/metastatic ZEB1 19 CK-19 33 ZEB1 23 CK-19 19 ZEB1 15 Locally advanced CK-19 26 ZEB1 19 Borderline resectable
P.16 P.17 P.18 P.19 P.20 CK-19 19 ZEB1 6 CK-19 51 ZEB1 39 CK-19 30 ZEB1 23 CK-19 15 ZEB1 19 CK-19 4 ZEB1 8 Unknown Locally advanced
Table S2 Sample Number Type CTC/mL Stage Tumor type ER/PR/HER2 B1 Breast 0.0 ER/PR+/HER2- B2 Breast 0 ER/PR+/HER2- B3 Breast 3.3 ER/PR+/HER2+ B4 Breast 6.3 ER/PR+/HER2- B5 Breast 8.0 ER/PR+/HER2- B6 Breast 8.0 ER/PR+/HER2- B7 Breast 12.5 ER/PR+/HER2- B8 Breast 12.5 ER/PR+HER2- B9 Breast 31.3 ER/PR+/HER2- B10 Breast 0 ER/PR+HER2- B11 Breast 0 ER/PR/HER2- B12 Breast 0.7 ER/PR+/HER2+ B13 Breast 1.4 ER/PR+/HER2- B14 Breast 1.4 ER+/PR-/HER2+ B15 Breast 1.4 ER/PR+/HER2- B16 Breast 1.4 ER/PR+/HER2- B17 Breast 1.4 ER/PR+HER2+ B18 Breast 2.1 ER/PR+/HER2- B19 Breast 2.1 ER/PR+/HER2- B20 Breast 2.1 ER/PR+/HER2- B21 Breast 2.1 ER/PR+/HER2- B22 Breast 2.8 ER+/PR-/HER2- B23 Breast 2.8 ER/PR+/HER2+ B24 Breast 2.8 ER/PR+/HER2- B25 Breast 2.8 ER/PR+/HER2- B26 Breast 2.8 ER/PR+/HER2- B27 Breast 3.5 ER+/PR-/HER2- B28 Breast 3.5 ER/PR+/HER2- B29 Breast 4.2 ER/PR+/HER2- B30 Breast 4.2 ER/PR+/HER2- B31 Breast 4.2 ER+/PR-/HER2- B32 Breast 4.9 ER/PR+/HER2- B33 Breast 5.6 ER/PR+/HER2-
B34 Breast 5.6 ER/PR+/HER2+ B35 Breast 6.3 ER/PR+/HER2- B36 Breast 6.3 ER/PR+/HER2- B37 Breast 7 ER/PR+/HER2- B38 Breast 7 ER/PR-/HER2+ B39 Breast 7 ER+/PR-/HER2- B40 Breast 7.7 ER/PR-/HER2+ B41 Breast 7.7 ER/PR/HER2- B42 Breast 7.7 ER/PR-HER2+ B43 Breast 8.4 ER/PR/HER2- B44 Breast 9.1 ER/PR+/HER2- B45 Breast 9.8 ER/PR+/HER2- B46 Breast 12.6 ER/PR+/HER2- B47 Breast 13.3 ER/PR+/HER2- B48 Breast 16.8 ER/PR+/HER2- B49 Breast 21.7 ER/PR+/HER2- B50 Breast 11.7 ER/PR+/HER2+ B51 Breast 10.3 ER/PR+/HER2- B52 Breast 11.2 ER/PR+/HER2- B53 Breast 7.5 ER/PR-/HER2+ B54 Breast 9.3 ER/PR+/HER2- B55 Breast 6.6 ER/PR+/HER2- B56 Breast 9.8 ER/PR+/HER2-
Table S3 GAPDH CD24 KRT5 CDH3 CTNNB1 IL6 HPRT1 CD44 KRT7 TGFb1 WNT2 IL6R RAB7A ALDH1a1 KRT8 TGFbR1 Twist1 IL6STP1 EpCAM ALDH1a3 AR TP53 SNAI1 CXCL8 Vimentin PROM1 KRT18 PTEN SNAI2 CXCR1 ERBB2 CCND1 KRT19 PI3K ZEB1 CXCR4 EGFR ANXA3 CDH1 AKT1 ZEB2 MTOR GSK3B GATA3 CDH2 AKT3 EZH2 NFKB1 ESR1 ABCG2 ID1 NOTCH1 MCL1 MKI67 PGR ABCB1 ID2 NOTCH2 BAX PCNA TAZ NESTIN MET NOTCH3 BCL2 CD3D YAP1 LIN28A BRCA1 NUMB TSPAN6 ITGAM SOX2 ITGB3 MUC1 HES1 TM4SF1 CD14 UXT ITGA6 STAP2 HEY2 TMEM57 MS4A1 SOCS3 FBXW7 AMOTL2 DLL1 NANOG PTPRC TP63 MMP9 TNKS1BP1 JAG2 POU5F1 MCAM
Table S4 CTC Category EMT Definition (based on marker expression) CD44 + and CD24 low/- MET ALDH1a3 + Dual ALDH1a3 +, CD24 low/- and CD44 + CK 2 CK + markers (CK 7, 8, 18, 19) WBC 1 Blood marker (PTPRC, ITGAM, MS4A1, and CD3D)
Supplementary Figure captions Figure S1. Related to Figure 1. Physics of the Labyrinth. a, Particle on a channel is focused by lift forces, which includes the force from shear gradient and from wall effect. When the two forces are balanced, the particles reach their equilibrium position. b, In a curved channel, the fluid close to the inner wall will be drawn away from the center of curvature by centrifugal force, causing two vortices to appear at top and bottom, accordingly. c, Particles will be focused to several equilibrium positions in microfluidic channels. For square channels, there will be 4 equilibrium positions and for rectangular channels, there will be 2. d, Desired focusing position occurs when ratio for the larger particle is greater than the ratio for small particle, and the ratios are not too close to 1 nor too greater than 1. At such position, the particles can be well separated. e. Hypothesis for sharp corners in Labyrinth. Strong Dean vortices at corners would help the focusing of small particles. Inertial lift forces acting on smaller particles are much weaker compared to that on larger particles (F Z is proportional to the 4th power of particle size). Therefore, smaller particles (e.g. RBCs, WBCs) are generally difficult to focus. For instance, the large particles (red and blue) are easily focused thus they are already at the focusing position while passing through the sharp corners. The smaller particles (green and yellow), on the other hand, are not focused yet at the corner. The focusing of the green particle is enhanced by the strong Dean vortex at the corner and thus the green particle gets closer to its focusing position after passing through the corner. The focusing of yellow particle will be enhanced by the other corners in opposite direction. f. Optimized design of Labyrinth g. SEM image of Labyrinth s outlet h. SEM image of corners and turns in Labyrinth. Scale bar 1mm. Figure S2. Related to Figure 2. Optimization of Labyrinth using cell lines. Flow stabilization measurements for a. MCF-7 breast cancer cell line and b. WBCs. c. PANC1 cell line recovery measured at different at flow rates ranging from 1-3 ml/min. d. WBC removal measured at different at flow rates ranging from 1-3 ml/min. e. Cell Recovery for low number (10cells/mL) of cell spiked in blood. f. WBC distribution across four outlets in Labyrinth after flow stabilization Figure S3. Related to Figure 3. Patient sample processing. a. Gallery of CTCs recovered from breast cancer patients. b. Scatter plot of CK and CD45 fluorescent intensity levels from patient samples for CTC identification. As shown in the plot, the upper left quarter contains CTCs and the lower right quarter contains the WBCs. After a patient sample was processed through Labyrinth and fluorescently stained with DAPI, CD45, and CK, a typical WBC from the slide was selected as reference for its CD45 and CK fluorescent intensity levels. A cell could be identified as a WBC by having similar CD45 and CK expressions comparing to the reference cell. A cell could be identified as a CTC by having at least 50% lower CD45 and 50% higher CK expressions comparing to the reference cell. Cells with both CD45 and CK significant expressions with respect to the reference cell would be considered as double positive cells. Cells with DAPI positive but much lower expressions in CD45 and CK when compared to the reference cell would be considered as double negative cells. c. CTCs/mL of blood recovered from breast cancer patients using whole blood with single and double Labyrinth. d. WBC/mL of blood remained after processing whole blood (without any RBC removal/pre-processing) from breast cancer patients using both single and double Labyrinth. e. Single vs Double Labyrinth CTC counts (per ml of whole blood) in breast cancer patients using pretreated blood for RBC removal. f. Comparison between total number of CTCs/mL in pancreatic cancer patients stained positive for CK19 vs total number of CTCs/mL in whole blood that stained positive for either epithelial (CK19) or for any of the two EMT markers used (ZEB1 and ATDC). g. Somatic mutations for pancreatic cancer related genes observed in 5 patient samples (C.1-C.5), Healthy Control (HC) and No Template Control (NTC)
Figure S4. Related to Figure 4. Single cell multiplex gene expression profile of isolated CTCs. Blood samples from cancer patients and healthy controls were processed by Labyrinth as described in the methods. Heat map analysis of the gene expression levels of 47 single cells including 31 patient derived CTCs from 6 different patients, 8 patient derived contaminating WBCs (Pts-WBCs) and 8 WBCs from healthy controls (HC-WBCs). The genes are grouped based on the genotypic characterization as described in Table S4. Supplementary Table captions Table S1. Related to Figure 3. Quantification of CTCs from pancreatic cancer patients Table S2. Related to Figure 3. Quantification of CTCs from breast cancer patients Table S3. Related to Figure 4. List of 96 genes studied in single CTCs Table S4. Related to Figure 4. CTC categories for marker-based classification