TY - JOUR
AU - García, Andrés, J
AB - Abstract Tumor-initiating cells (TICs), a subpopulation of cancerous cells with high tumorigenic potential and stem-cell-like properties, drive tumor progression and are resistant to conventional therapies. Identification and isolation of TICs are limited by their low frequency and lack of robust markers. Here, we characterize the heterogeneous adhesive properties of a panel of human and murine cancer cells and demonstrate differences in adhesion strength among cells, which exhibit TIC properties and those that do not. These differences in adhesion strength were exploited to rapidly (~10 min) and efficiently isolate cancerous cells with increased tumorigenic potential in a label-free manner by use of a microfluidic technology. Isolated murine and human cancer cells gave rise to larger tumors with increased growth rate and higher frequency in both immunocompetent and immunocompromised mice, respectively. This rapid and label-free TIC isolation technology has the potential to be a valuable tool for facilitating research into TIC biology and the development of more efficient diagnostics and cancer therapies. cell adhesion, cancer, tumor-initiating cells, hydrodynamic, microfluidic Insight box Identification and isolation of tumor-initiating cells (TICs), a subpopulation of cancerous cells with high tumorigenic potential and stem-cell-like properties, are limited by their low frequency and lack of robust markers. We demonstrate differences in adhesive signatures among murine and human cells, which exhibit TIC properties and those that do not. These differences in adhesion strength were exploited to efficiently isolate cancerous cells with increased tumorigenic potential in a label-free manner using a microfluidic technology. Isolated murine and human cancer cells gave rise to larger tumors with increased growth rate and higher frequency in immunocompetent and immunocompromised mice. This technology has the potential to facilitate research into TIC biology and the development of more effective diagnostics and therapies. INTRODUCTION Tumors are heterogeneous tissues, which contain a subpopulation of cancer cells with high tumorigenic potential and stemness properties [1–8]. This subpopulation of cells has been referred to as “tumor-initiating cells” (TICs), “cancer stem cells” (CSCs), or “cancer stem-like cells” (CSLCs). Although there is debate whether TICs isolated from cancer patients are ‘bona fide’ stem cells [9], accumulating evidence indicates that these malignant cells are exceedingly tumorigenic and resistant to treatment with conventional chemotherapies and radiation [7, 10]. As such, TICs likely persist after chemo/radiotherapy and result in multi-drug resistance as well as disease relapse [7, 10]. TIC-enriched populations have been identified in established cell lines and patient samples using discrete surface markers (e.g. CD44hi/CD24lo, CD133+, ALDH+, and ESA+) and their ability to generate tumor spheres and xenograft tumors [1–7, 11–14]. However, TIC sub-populations from various sources differ greatly in their surface marker expression profile, and, to date, there is no universal marker profile to identify TICs [7, 10, 12]. Furthermore, there is no evidence that any combination of surface markers isolates TICs to a high degree of purity [7]. The inability to effectively purify TIC sub-populations is a profound impediment to characterizing the biology of these cells with precision as well as analyzing patient samples for effective diagnosis or prognosis [7, 10]. Therefore, there is a significant and unmet need for unbiased, efficient, label-free technologies for the identification and purification of cancer cell populations, including TIC sub-populations, from heterogeneous cultures and tumors. Flow cytometric and magnetic-based selection methods have been developed to enrich for TICs, including surface antigen expression profiles [15–19], aldehyde dehydrogenase (ALDH) [20–22] or reactive oxygen species levels [23], and flow cytometric side population analysis [24]. A key drawback of these label-based methods is the inherent variability among different cancer types, even within tumor samples and cell lines of the same cancer type, causing markers that work for some tissues to be suboptimal for others. Furthermore, these methods rely on defined markers and reagents (e.g. antibodies), are time consuming, require specialized equipment, and are difficult to scale-up. In contrast, label-free biophysical methods based on differential cell adhesion or stiffness have been explored to isolate cancer cells [25]. For example, increased cell elasticity is correlated to increased migratory behavior in cancer cells [26–29] and activation of the EMT program [30]; both of these biological phenomena correlate with TIC phenotype [31]. Selection for highly elastic breast cancer cells showed enrichment for the CD44high/CD24low TIC phenotype and increased mammosphere formation [32]. However, cell elasticity may not be a robust selection marker as direct measurements of stiffness using atomic force microscopy showed that TICs from other sources are softer than non-TIC cells [33]. Selection of cancer cells from non-cancerous cells based on differential adhesion due to dysregulated adhesive mechanisms has also been reported [34–36]. These approaches exploit differences in binding rate or capture efficiency, which are time-, cell type-, and extracellular matrix-dependent. Here, we present a label-free strategy for the separation of cancer cell populations from cell cultures and tumor explants based on hydrodynamic detachment forces using the μSHEAR (micro-Stem cell High Efficiency Adhesion-based Recovery) microfluidic technology [37]. Advantages of the μSHEAR technology include its label-free and unbiased basis, facile operation and scale-up, high performance in terms of purification efficiency, yield and cell viability, flexibility in performing negative or positive selection, and microfluidics format that allows for integration with other analytical platforms such as single-cell surface marker profiling and genomic and proteomic technologies. RESULTS Human breast cancer cell lines exhibit diverse adhesion strengths The adhesion strength, defined as the force required to detach a cell from a surface, of various human breast cancer cell lines (details provided in Supplementary Table 1) as well as non-cancerous immortalized mammary cells was measured using the spinning disk assay [37–39]. In this assay, cells adhering to an extracellular matrix-coated glass coverslip are exposed to a linear range of hydrodynamic forces (Fig. 1A). Cell detachment can then be quantified at specific radial locations and fitted to a sigmoid to obtain τ50, the shear stress at which 50% of the cells detach (Fig. 1B). Fibronectin (FN) was chosen as the ECM protein for coating because it plays a crucial role in breast cancer invasiveness and migration [40]. Non-cancerous immortalized mammary cells (hTERT-HME1) had significantly higher adhesion strength to FN (997 ± 68 dyne/cm2) compared to all cancerous cell lines examined (P < 0.0001) (Fig. 1C). Mean adhesive strength values varied significantly among breast cancer cell lines, ranging from 60 to 700 dyne/cm2 (Supplementary Table 2). Figure 1 Open in new tabDownload slide Cell adhesion strength to FN for human breast cancer cell lines. (A) Schematic of the spinning disk assay. Cells adhere to a FN-coated circular coverslip, which are spun at a predetermined speed (ω). The rotation applies shear stress (τ) proportional to the radial position (r) within the cover slip and cell detachment can be quantified at predetermined radial positions. (B) Representative detachment profiles for MDA-MB-231, MDA-MB-453, and hTERT-HME1 cell lines. (C) Adhesion strength measurements for the hTERT-HME1 immortalized mammary cell line as well as a panel of breast cancer cell lines (One-way ANOVA ****P < 0.0001, detailed statistical comparisons available in Supplementary Table 2, each symbol represents an independent sample). Figure 1 Open in new tabDownload slide Cell adhesion strength to FN for human breast cancer cell lines. (A) Schematic of the spinning disk assay. Cells adhere to a FN-coated circular coverslip, which are spun at a predetermined speed (ω). The rotation applies shear stress (τ) proportional to the radial position (r) within the cover slip and cell detachment can be quantified at predetermined radial positions. (B) Representative detachment profiles for MDA-MB-231, MDA-MB-453, and hTERT-HME1 cell lines. (C) Adhesion strength measurements for the hTERT-HME1 immortalized mammary cell line as well as a panel of breast cancer cell lines (One-way ANOVA ****P < 0.0001, detailed statistical comparisons available in Supplementary Table 2, each symbol represents an independent sample). The adhesion strength of a cell to its ECM is dependent on its morphology/size, integrin receptor expression levels, ECM ligand density, integrin-ECM bond numbers and distribution, and the association between ligated integrins and cytoskeletal elements [41–42]. We evaluated integrin expression levels for a subset of these human breast cancer and non-cancerous immortalized mammary cells to determine whether integrin expression levels correlated with adhesion strength (Fig. S1). This flow cytometry analysis demonstrated significant differences in expression of particular integrin subunits among these cell lines. However, the differences in integrin subunit expression do not account for the differences in adhesion strength among these cell lines. The adhesion strength of MDA-MB-231, MDA-MB-453, and MCF7 breast cancer cells was also assessed using the μSHEAR technology [37] (Fig. 2A). In this microfluidic device, hydrodynamic shear forces proportional to the fluid flow rate are applied to adherent cells [37]. Cells adhering to FN were exposed to predetermined shear forces for 10 min, with different devices being exposed to different amounts of shear force, while cell detachment was visualized with a microscope. In this fashion, a profile for adherent cells vs. surface stress was generated for each cell line (Fig. 2B). The detachment profiles clearly demonstrate lower adhesion strength values for the cancer cell lines compared to the non-cancerous hTERT-HME1 and a distribution in adhesion strength for each cell line with a fraction of cells exhibiting high adhesion strength. Figure 2 Open in new tabDownload slide μSHEAR microfluidic-based separation of cell mixtures. (A) Schematic of the μSHEAR microfluidic device. Cells are cultured in FN-coated microfluidic channels. Defined shear-stress levels are applied to enrich for cell populations based on their adhesion properties. (B) Detachment profiles for MDA-MB-231, MDA-MB-453, and MCF7 human breast cancer lines and the hTERT-HME1 immortalized mammary cell line as assessed by μSHEAR. For each microfluidic device, a pre-determined amount of shear stress was applied and cell detachment quantified by use of a microscope with motorized stage (results from four independent experiments shown). (C) Enrichment of hTERT-HME1 cells from mixed cultures with MDA-MB-231 breast cancer cells by μSHEAR, achieving hTERT-HME1 purities of 90–95% (Two-way ANOVA *P < 0.05, ****P < 0.0001). Figure 2 Open in new tabDownload slide μSHEAR microfluidic-based separation of cell mixtures. (A) Schematic of the μSHEAR microfluidic device. Cells are cultured in FN-coated microfluidic channels. Defined shear-stress levels are applied to enrich for cell populations based on their adhesion properties. (B) Detachment profiles for MDA-MB-231, MDA-MB-453, and MCF7 human breast cancer lines and the hTERT-HME1 immortalized mammary cell line as assessed by μSHEAR. For each microfluidic device, a pre-determined amount of shear stress was applied and cell detachment quantified by use of a microscope with motorized stage (results from four independent experiments shown). (C) Enrichment of hTERT-HME1 cells from mixed cultures with MDA-MB-231 breast cancer cells by μSHEAR, achieving hTERT-HME1 purities of 90–95% (Two-way ANOVA *P < 0.05, ****P < 0.0001). The ability to separate cancerous cells from non-cancerous hTERT-HME1 cells based on differences in adhesion strength using the μSHEAR technology was explored. The MDA-MB-231 cell line was chosen for this assay since it was among the cell lines, which exhibited the largest adhesion difference relative to the hTERT-HME1 cells. The cells lines were fluorescently labeled with different dyes, mixed, and cultured in μSHEAR devices. In a similar fashion to the previously described μSHEAR experiments, the cell mixture was exposed to a predetermined shear stress of 440 dynes/cm2 and the detachment of cells was visualized with a microscope. After 10 min, MDA-MB-231 constituted on average <10% of the cell population remaining in the device, down from 45 to 85% in the initial mixture of cells, achieving more than a 10-fold reduction in cancer cells within the mixture of cells (Fig. 2C). Notably, the final purity of the cell population was independent from the initial purity. These results demonstrate that the differences in adhesion strength between these cell lines can be exploited to enrich for one cell line. Murine breast cancer cell line E0771 exhibits an adhesion strength signature distinct from non-cancerous mammary cells Cell function is highly dependent on the adhesive interactions between cells and their extracellular environment [43–44], and dysregulation of cell adhesion can cause a range of diseases, including cancer [45]. The aforementioned spinning disk results suggest that cancerous cells have adhesion strength signatures, which differ significantly from non-cancerous cells. To further test this, the spinning disk assay was used to measure the adhesion strength of E0771 murine breast cancer cells as well as cells isolated from the mammary fat pad of C57BL/6 mice (Fig. 3), from which the E0771 line was originally derived [46–47]. Consistent with our observations for human cells, non-cancerous mouse mammary fat pad cells have significantly higher adhesion strength to FN (289 ± 92 dyne/cm2) compared to cancerous E0771 cells (60.7 ± 21 dyne/cm2). Figure 3 Open in new tabDownload slide Adhesion strength for E07771 murine cancer and fat pad cells. Adhesion strength of E0771-GFP+ breast cancer cells and fat pad cells isolated from C57BL/6 female mice to FN (Student’s t-test ****P < 0.0001). Figure 3 Open in new tabDownload slide Adhesion strength for E07771 murine cancer and fat pad cells. Adhesion strength of E0771-GFP+ breast cancer cells and fat pad cells isolated from C57BL/6 female mice to FN (Student’s t-test ****P < 0.0001). Murine cancer cell sub-populations with different tumorigenic potentials can be isolated by adhesive force The adhesion strength analyses of human and murine cancer cell lines suggest the presence of sub-populations of cells with differential adhesive properties. As such, whether different adhesive subpopulations of cells within an established cell line have different tumorigenic potentials was explored. We chose to study the E0771 breast cancer cell line [46–47], as this murine line allows evaluation of tumor formation in immunocompetent C57BL/6 mice. E0771 cells were transduced with a lentivirus for eGFP to generate eGFP-expressing E0771 (E0771-GFP+) cells for tracking transplanted cells; eGFP expression did not influence the adhesion strength of this cell line (Fig. S2). A titration assay was performed with a range of E0771-GFP+ cell doses (125–10,000 cells) injected into the mammary pad of C57BL/6 female mice, and tumor size was determined by caliper measurements of the diameter of the tumor mass at defined time points (Fig. S3). In general, tumor latency and tumor size increased with increasing cell dose. A dose of 250 cells was selected for subsequent experiments as this dose resulted in small-to-medium sized tumors detectable at 7 days post-implantation. Based on adhesion data collected using the spinning disk technology (Fig. 3A), μSHEAR-based separation of adhesive and detached fractions of E0771-GFP+ cells was performed for three different shear-stress levels: 110 dynes/cm2, 220 dynes/cm2, and 330 dynes/cm2. FN was chosen as the coating ECM protein because it plays a crucial role in breast cancer invasiveness and migration [40]. Furthermore, mammary epithelial cell interactions with FN have been shown to induce EMT [48], which in turn is correlated with TIC properties. E0771-GFP+ cells that detached upon application of hydrodynamic shear forces (detached fraction) and cells that remained attached in the device (attached fraction) were collected. In addition, cells in the device that were not exposed to hydrodynamic shear were collected to serve as controls. The collected cell fractions were counted and 250 cells from each fraction were implanted into the mammary pad of C57BL/6 female mice (Fig. 4A). Tumor sizes were measured at 9, 11, 13, 15, and 17 days post-implantation. For cells exposed to 110 dynes/cm2, there was a small but significant decrease in tumor size for the attached cell fraction compared to the control group at 15 and 17 days post-implantation (Fig. 4B). Remarkably, the detached cell fraction isolated at 220 dynes/cm2 produced significantly larger tumors at 15 and 17 days post-implantation compared to the control unsorted E0771 cells (Fig. 4C). Furthermore, the attached fraction yielded significantly smaller tumors than both the detached fraction and the control group. No differences in tumor size were detected among any of the groups exposed to 330 dynes/cm2 (Fig. 4D). Figure 2E shows tumor sizes at the completion of the experiment on day 17 for both attached and detached cell fractions for all shear-stress isolation conditions, along with average volume for control unsorted E0771-GFP+ cells. The cell fraction detached at 220 dyne/cm2 generated larger tumors compared to cell fractions detached at 110 and 330 dyne/cm2 as well as the attached fraction exposed to 220 dyne/cm2. Furthermore, the tumors generated by the detached cell fraction at 220 dyne/cm2 were 50% larger than those formed by unsorted E0771-GFP+ cells. In addition to increased tumor size, the detached cell fraction at 220 dyne/cm2 generated tumors faster compared to other groups as indicated by the percentage of mice with detectable tumors as a function of time (Fig. 4F). Histological staining of explanted tumors confirms the growth and size of the implanted E0771-GFP+ cells (Fig. S4). Taken together, these results demonstrate that sub-populations of cells in the E0771 murine cell line with different tumorigenic potential that can be isolated by differences in adhesive force to FN. Figure 4 Open in new tabDownload slide μSHEAR-mediated enrichment of TIC-like murine breast cancer cells. (A) Schematic of experimental setup. E0771-GFP+ cells were cultured in FN-coated μSHEAR devices and separated based on pre-determined shear-stress values (110, 220, and 330 dyne/cm2). Adherent and detached cell fractions as well as unsorted cells were injected into mice and tumor formation monitored. (B–D) Tumor growth curves for detached and attached cell fractions and unsorted cells after applying (B) 110 dyne/cm2, (C) 220 dyne/cm2, or (D) 330 dynes/cm2. (E) Tumor volumes for recovered cell fractions and unsorted controls. Results were analyzed using two-way ANOVA (mean ± standard error of the mean. $: detached fraction vs. attached fraction, #: detached fraction vs. unsorted control, &: attached fraction vs. unsorted control. One symbol P < 0.05; two symbols P < 0.01; three symbols P < 0.001; four symbols P < 0.0001). (F) Percentage of mice with detectable tumors over time post implantation. Figure 4 Open in new tabDownload slide μSHEAR-mediated enrichment of TIC-like murine breast cancer cells. (A) Schematic of experimental setup. E0771-GFP+ cells were cultured in FN-coated μSHEAR devices and separated based on pre-determined shear-stress values (110, 220, and 330 dyne/cm2). Adherent and detached cell fractions as well as unsorted cells were injected into mice and tumor formation monitored. (B–D) Tumor growth curves for detached and attached cell fractions and unsorted cells after applying (B) 110 dyne/cm2, (C) 220 dyne/cm2, or (D) 330 dynes/cm2. (E) Tumor volumes for recovered cell fractions and unsorted controls. Results were analyzed using two-way ANOVA (mean ± standard error of the mean. $: detached fraction vs. attached fraction, #: detached fraction vs. unsorted control, &: attached fraction vs. unsorted control. One symbol P < 0.05; two symbols P < 0.01; three symbols P < 0.001; four symbols P < 0.0001). (F) Percentage of mice with detectable tumors over time post implantation. Human colon cancer cells separated by hydrodynamic forces have higher tumor formation properties We next examined the adhesion strength characteristics for CA2 human sporadic TICs established from primary human colonic cancer resections [49–50]. These cells contain a large population of cells (20–30%) with high levels of ALDH activity (Fig. 5A), which has been shown to be a reliable marker for TIC activity in various cell lines, including the CA2 cell line [49–50]. CA2 cells were cultured and expanded in suspension culture as previously described [50]. CA2 cells were plated onto FN- or Matrigel (MT)-coated surfaces for 4 days, stained for ALDH activity, and sorted into ALDHLO and ALDHHI fractions. Transition to adhesion culture on MT- or FN-coated surfaces for 4 days did not alter the frequency of ALDHHI TIC cells in the population, although a reduction in the fraction of ALDHHI cells was observed at 7 days of adhesion culture (Fig. S5). The adhesion strength for ALDHLO and ALDHHI fractions for cells in 4-day adhesion culture was then measured using the spinning disk assay. ALDHHI cells showed higher adhesion strength to MT-coated surfaces compared to ALDHLO fraction. For FN-coated surfaces, the overall levels of adhesion strength were significantly lower than those observed for MT (Fig. 5B). Figure 5 Open in new tabDownload slide μSHEAR-mediated enrichment of human cancer CA2 cells with increased tumorigenic potential. (A) ALDH activity levels in CA2 cells measured using the ALDEFLUOR assay. (B) CA2 cells were sorted based on their ALDH activity levels and the adhesion strength properties of ALDHHI and ALDHLO cell populations to FN- and MT-coated surfaces were assessed using the spinning disk assay. For MT, ALDHHI exhibited a significantly higher adhesion strength than ALDHLO cells. (C) Tumor volume for μSHEAR-isolated adherent and detached fractions and control CA cells implanted in NSG mice. Results were analyzed using two-way ANOVA (mean ± standard error of the mean; $: attached fraction vs. unsorted control, &: attached fraction vs. suspension control, #: detached fraction vs. attached fraction. One symbol P < 0.05; two symbols P < 0.01; three symbols P < 0.001; four symbols P < 0.0001). Immunostaining for (D) EpCAM, (E) ALDH1, and (F) E-cadherin in tissue sections. Figure 5 Open in new tabDownload slide μSHEAR-mediated enrichment of human cancer CA2 cells with increased tumorigenic potential. (A) ALDH activity levels in CA2 cells measured using the ALDEFLUOR assay. (B) CA2 cells were sorted based on their ALDH activity levels and the adhesion strength properties of ALDHHI and ALDHLO cell populations to FN- and MT-coated surfaces were assessed using the spinning disk assay. For MT, ALDHHI exhibited a significantly higher adhesion strength than ALDHLO cells. (C) Tumor volume for μSHEAR-isolated adherent and detached fractions and control CA cells implanted in NSG mice. Results were analyzed using two-way ANOVA (mean ± standard error of the mean; $: attached fraction vs. unsorted control, &: attached fraction vs. suspension control, #: detached fraction vs. attached fraction. One symbol P < 0.05; two symbols P < 0.01; three symbols P < 0.001; four symbols P < 0.0001). Immunostaining for (D) EpCAM, (E) ALDH1, and (F) E-cadherin in tissue sections. We selected MT-coated surfaces for subsequent experiments because this ECM resulted in significant differences in adhesion strength between ALDHHI and ALDHLO. Furthermore, a basement membrane-rich matrix such as Matrigel is a physiologically relevant matrix for this epithelial cell line. μSHEAR-based separation of the cells at 47.4 dynes/cm2 was performed, and detached and attached fractions were collected. This shear-stress value was selected because it led to the detachment of ~50% of the CA2 cells cultured within the μSHEAR devices. Cells cultured in the microfluidic device but not exposed to hydrodynamic forces were used as unsorted controls. In addition, cells maintained in suspension culture and not plated on MT-coated supports were used as an additional control. Detached and attached cell fractions, unsorted and suspension cells (100 cells) were injected into the subcutaneous space in the flank of NSG immunocompromised mice. Tumor sizes were measured every other day starting 23 days post-implantation by use of calipers, based on previous tumor growth data [49]. The attached fraction of CA2 cells formed ~3-fold larger tumors compared to the detached fraction and unsorted cells (Fig. 5C). Furthermore, the attached cell fraction generated tumors in 100% (5/5) of mice, whereas the detached cell fraction only formed tumors in 60% (3/5) of mice. Immunostaining of explanted tissue demonstrate positive staining for human EpCAM, ALDH1, and E-cadherin (Fig. 5D-F), confirming that the implanted CA2 cells gave rise to these tumors. Collectively, these results demonstrate that a sub-population of human colon cancer cells with higher tumor formation properties can be separated by hydrodynamic forces. DISCUSSION The discovery of TICs and characterization of the roles that they play in tumor progression, metastasis, and relapse have revolutionized the field’s approach to cancer and its treatment [51–53]. A better understanding of the importance of TICs has led to the development of novel therapeutic agents, which are currently being evaluated in clinical and preclinical studies [54]. Most of these novel therapies could be used in conjunction with traditional therapies, such as chemotherapy, to target both non-TIC and the TIC subpopulations of cells within tumors [54]. Nevertheless, the efficient and reproducible purification of TIC population remains a hurdle for further advancements in the field. In this study, we examined differences in adhesive forces between non-cancerous and cancerous cells and found significant differences among these populations. We also applied a microfluidic method to enrich for mouse and human cancer cells with the ability to form larger tumors with increased frequency and faster growth rates, all of which are hallmarks of TIC phenotype. Cell adhesion impacts tumorigenesis via various mechanisms, and it is likely that the dominant mechanism(s) depends on many factors including tumor type/origin, microenvironment, and patient. As such, we do not expect that a simple parameter (adhesion strength) would serve as a ‘universal’ marker of tumorigenesis, so the goal of this study was not to correlate adhesion strength to tumorigenesis. Instead, the objective of this study was to test whether tumorigenic cells could be enriched based on differences in cell-adhesion strength regardless of the underlying mechanisms driving tumorigenesis. We show that breast cancer cells (several human lines, murine E0771 line) exhibit lower adhesion strength to FN compared to normal human mammary epithelial and murine fat pad cells. We also demonstrate differences in adhesion strength to Matrigel among ALDHHI and ALDHLO sub-populations in CA2 human colon cancer TICs. For both E0711 and CA2 cells, we enriched for TIC fractions by exploiting differences in adhesion strength. The less adherent E0711 cell fraction and more adherent CA2 fraction were more tumorigenic than the unsorted parental population. We note important distinctions between these two different cancer lines. First, E0771 are murine breast cancer cells, whereas CA2 are human colon cancer cells. The tumor source, type, and origin (including species) could impact the adhesivity of the tumorigenic cells. Second, the adhesion strength for these lines was characterized for different matrices—FN was used for E0771 cells, while Matrigel was used for CA2 cells. We note that the adhesion strength to FN of the tumorigenic ALDHHI CA2 cells is lower than ALDHLO CA2 cells. Our study demonstrates the ability to enrich for tumorigenic cell fractions based on differences in adhesion strength using a facile and label-free methodology. This work provides the foundation for expanded and systematic analyses of patient-derived tumor cells based on tumor type, origin, genetics, and patient history. For instance, direct of comparison of the adhesion strength and tumorigenecity for patient-derived cells from tumors of similar origin, for example breast or colon tissue, would provide a platform to evaluate mechanisms connecting cell-matrix adhesion to tumorigenesis. Furthermore, evaluation of tumorigenicity for cell fractions separated by this adhesion strength-based technology for patient-derived tumors of similar origin (e.g. breast, colon, and lung) could be implemented to assess whether a cell fraction with a distinctive adhesion strength signature provides a tool for diagnostic purposes. ALDHs are a family of enzymes, which play a role in the metabolism of aldehydes [55]. Prior work has established a correlation between high levels of ALDH and stemness and succeeded in isolating hematopoietic stem cells based on ALDH activity [56–58]. More recently, high ALDH levels have been associated with other stem cell types as well as TICs [20–21, 59–61]. Here, we used a human colorectal cancer cell line, CA2, to study the relationship between ALDH expression and adhesion strength, showing significantly higher adhesion strength values in cells with high ALDH activity. Moreover, using the μSHEAR microfluidic technology, we enriched for CA2 cells that, when implanted into mice, formed larger tumors with higher frequency. Microfluidic cell isolation methods have emerged as powerful platforms for cell enrichment, particularly for cancer cells (as reviewed in [25, 62–63]). These systems employ differential surface marker expression [64–65] or differences in cell mechanical properties (e.g. elasticity) [66–68]. Selection of cancer cells from non-cancerous cells based on differential adhesion has also been reported [34–36]. The label-free technology μSHEAR technology used in our study exploits differences in steady state detachment force, which are scalable, robust, and easy to implement. Our strategy for enriching for more tumorigenic cell populations could facilitate research into TIC biology and the development of more efficient cancer therapies. In the long term, the μSHEAR technology could enable TIC detection in diagnostic settings, allowing physicians to better determine optimal therapeutic regimens. METHODS Cell culture All human breast cancer cell lines as well as the hTERT-HME1 cell line were acquired from ATCC and cultured according to their protocols. All cell lines were sub-cultured at 70–80% confluency. E0771 cells were cultured in RPMI 1640 supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin at 37 °C, 5% CO2. The cell line CA2 established from primary human colonic biopsies was cultured in suspension in ultra-low-attachment 6-well plate (Corning) at a density of 100 000 to 150 000 cells per well as previously described [50]. Briefly, CA2 cells were suspended in a serum-free medium containing DMEM/F12 (Life Technologies), 6 mg/mL glucose (Sigma), 10 nmol/L progesterone (Sigma), glutamine (Life Technologies), 10 μg/mL insulin (Sigma), 13 μg/mL transferrin (Sigma), 15 nmol/L sodium selenite (Sigma), 4 mg/mL bovine serum albumin (BSA; Sigma), 50 μmol/L putrescine (Sigma), and 15 mmol/L HEPES (Life Technologies). Cells were supplemented with 10 ng/mL fibroblast growth factor and 20 ng/mL epidermal growth factor (Sigma) every other day. Cells were sub-cultured every 4–6 days. For adhesion studies, single CA2 cells were plated in tissue culture treated six wells coated either with FN (10 μg/mL in phosphate buffered saline solution – PBS, 30 min) and blocked with bovine serum albumin (1% BSA in PBS, 30 min) or coated with Matrigel™ hESC-Qualified Matrix according to the manufacturers guidelines (60 min) and cultured for 4 days. ALDH staining and sorting of CA2 cells CA2 cells were trypsinized and stained for ALDH activity using the ALDEFLUOR assay (StemCell Technologies) according to the manufacturer’s guidelines. Propidium iodine (ThermoFisher) was used as a dead cell stain at a 1:1000 dilution. Cells were sorted using a BD Aria Cell Sorter. CA2 TICs were isolated and propagated under IRBs held at the University of Michigan, University of Florida and the Cleveland Clinic. STR analyses confirm cell uniqueness (Duke University DNA Analysis Facility). Integrin expression profiling MDA-MB231, MDA-MB453, MCF7, and hTERT-HME1 cells were thawed, plated, and expanded to reach ~80% confluency. Cells were incubated in primary antibody (see Table S3 for list and dilutions) for 1 h on ice in staining buffer (2% FBS). Cells were washed twice using the staining buffer and incubated in secondary antibody for 50 min on ice in the staining buffer. Integrin expression measured by flow cytometry (BD Accuri C6). GFP lentiviral transduction of E0771 cells Stably transduced E0771-eGFP+ cells were generated by transducing E0771 mouse melanoma cells with LV-CMV-GFP lentivirus and sorting for GFP+ cells on a BD Aria sorter. Briefly, cells were seeded at 60% confluency in six-wells and allowed to attach overnight. The cultures were then incubated in lentivirus at several multiplicity of infections (MOIs) ranging from 10 to 50. Four days post-infection, the cells were trypsinized and the GFP expression levels assessed in a flow cytometer. The MOI that gave the best efficiency of infection, MOI = 50, was selected for further use. Two rounds of sorting were performed in order to achieve a >99% GFP+ population of cells by use of the BD Aria sorter. Isolation of fat pad cells The tissue in the abdominal mammary glands of female C57BL/6 mice was excised and digested with collagenase D (0.375 U/ml) and hyalurodinase (125 U/ml) for 2–3 h at 37 °C with mechanical agitation every 30 min, 0.25% trypsin-EDTA (5 min, 37 °C), red blood cell lysis buffer (5 min, room temperature), and DNAse (10 min, 0.1 mg/ml, room temp). The mixture of cells was filtered through a 40 μm filter twice, resuspended in the relevant volume of media, and used for the adhesion experiments. Spinning disk assay Circular glass cover slips (25 mm diameter) were clean with ethanol, coated with FN (10 μg/mL) for 30 min, and blocked with 1% BSA for 30 min. Cells were seeded onto coverslips and cultured overnight at concentrations of 50,000–200,000 cells/ml depending on the cell line in order to achieve 40–50% confluency. After 24 h, the coverslips were spun for 5 min in PBS, thus applying a range of forces to the cells proportional to the cell’s radial position in the cover slip. The cells were fixed with 4% paraformaldehyde for 15 min, permeabilized in 0.05% Triton-X100 for 40 min, stained with DAPI for 30 min, washed three times with PBS, and mounted into slides for imaging. For E0771-GFP+ experiments, the cells were also stained for GFP for 2 h with a 1:100 dilution of the primary antibody (Abcam #ab290) and for 1 h with a 1:200 dilution of an anti-rabbit 488 secondary antibody with three washes (1% BSA) after incubation with each antibody. The number of cells at defined radial positions was quantified by use of a fluorescence microscope with a mechanical stage. A MATLAB program was used to fit the data into sigmoidal and calculate the τ50 (force required to detach 50% of the cells) [42, 69]. Microfluidic device fabrication Polydimethylsiloxane (PDMS) (Sylgard 184, Dow Corning) microfluidic devices were fabricated as reported earlier using a negative photoresist (SU-8 2050, 50-μm thickness, MicroChem) and UV photolithography [37]. Patterned negative molds were then exposed to vapor-phase tridecafluoro-1,1,2,2-tetrahydrooctyl-1-trichlorosilane (United Chemical Technologies) in a vacuum desiccator to prevent adhesion of PDMS. A 5-mm-thick layer of degassed PDMS mixture (10:1) was cast onto the mold and cured at 110 °C for 20 h. Cast PDMS devices were peeled off and then punctured for inlet–outlet holes and bonded to glass coverslips by exposure to oxygen plasma for 15 s and allowed to bond at 110 °C for 30 min. μSHEAR adhesion experiments Devices were first washed with ethanol, washed with PBS, either coated with FN (10 μg/mL in PBS, 45 min) and blocked (1% BSA, 45 min) or coated with Corning Matrigel™ hESC-Qualified Matrix according to the manufacturers guidelines (90 min), and washed with complete media. Cells were then introduced at a concentration of 4–10 × 10 cells/mL [6] and cultured at 37 °C, 5% CO2 overnight. Predetermined amounts of force were applied to the cells for a 10 min period by flowing PBS at well-defined flow rates controlled by a Harvard Instruments Pico Plus Elite (#70–4506) syringe pump. Detached cells were collected and treated with trypsin in suspension, and attached cells were trypsinized and collected. Trypsin in both fractions was inactivated by use of serum-containing media. For detachment experiments, cells were labeled with CellTracker™ Red CMTPX Dye (ThermoFisher Scientific #C34552) within the devices according to manufacturer’s guidelines prior to force application. Devices were loaded into a Nikon TE300 microscope equipped with a Ludl motorized stage, Spot-RT camera, and Image Pro analysis system and the number of GFP+ and/or CellTrackerRED+ cells was quantified along eight specific locations at predetermined time points. E0771 tumor generation and size measurement eGFP-E0771 cells were injected (125–10 000 cells/injection for tumor titration assays; 250 cells/injection in tumor growth assays in 20 μL of 1:1 Matrigel to saline solution) through the inguinal mammary gland into the abdominal mammary gland of female C57BL/6 mice to establish tumors. Tumor diameter sizes were measured in 3D at days 9,11,13,15, and 17 using digital calipers. CA2 tumor generation and size measurement CA2 cells were injected (100 cells/injection in 100 μL of 1:1 Matrigel to CA2 media) into the subcutaneous flank of female NSG mice to establish tumors. Tumor diameter sizes were measured in 3D every other day staring 23 days post implantation using digital calipers. Statistics Data points were plotted using Prism (GraphPad) with a horizontal line representing the mean and vertical error bars indicating the standard error. Statistical significance (P < 0.05) was determined using either, Student’s t-test, one way ANOVA with Tukey post hoc test or two way ANOVA with Tukey post hoc test as noted with Prism (GraphPad). Sigmoidal non-linear regression was also used in some parts of the analysis (GraphPad). Funding This work was supported by the National Cancer Institute of the National Institutes of Health under Award Number R21 CA202849 (A.J.G., S.N.T.). Conflict of interest This work forms the basis of a patent application filed by the Georgia Tech Research Corp listing E.A.C., S.N.T., and A.J.G. as inventors. References 1. Al-Hajj M , Wicha MS, Benito-Hernandez A et al. Prospective identification of tumorigenic breast cancer cells . Proc Natl Acad Sci U S A 2003 ; 100 : 3983 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat 2. Singh SK , Hawkins C, Clarke ID et al. Identification of human brain tumour initiating cells . Nature 2004 ; 432 : 396 – 401 . Google Scholar Crossref Search ADS PubMed WorldCat 3. O'Brien CA , Pollett A, Gallinger S et al. A human colon cancer cell capable of initiating tumour growth in immunodeficient mice . Nature 2007 ; 445 : 106 – 10 . Google Scholar Crossref Search ADS PubMed WorldCat 4. Kim MP , Fleming JB, Wang H et al. ALDH activity selectively defines an enhanced tumor-initiating cell population relative to CD133 expression in human pancreatic adenocarcinoma . PLoS One 2011 ; 6 :e20636. OpenURL Placeholder Text WorldCat 5. Li C , Heidt DG, Dalerba P et al. Identification of pancreatic cancer stem cells . Cancer Res 2007 ; 67 : 1030 – 7 . Google Scholar Crossref Search ADS PubMed WorldCat 6. Dalerba P , Dylla SJ, Park IK et al. Phenotypic characterization of human colorectal cancer stem cells . Proc Natl Acad Sci U S A 2007 ; 104 : 10158 – 63 . Google Scholar Crossref Search ADS PubMed WorldCat 7. Meacham CE , Morrison SJ. Tumour heterogeneity and cancer cell plasticity . Nature 2013 ; 501 : 328 – 37 . Google Scholar Crossref Search ADS PubMed WorldCat 8. Chaffer CL , Brueckmann I, Scheel C et al. Normal and neoplastic nonstem cells can spontaneously convert to a stem-like state . Proc Natl Acad Sci U S A 2011 ; 108 : 7950 – 5 . Google Scholar Crossref Search ADS PubMed WorldCat 9. Kreso A , Dick JE. Evolution of the cancer stem cell model . Cell Stem Cell 2014 ; 14 : 275 – 91 . Google Scholar Crossref Search ADS PubMed WorldCat 10. Marjanovic ND , Weinberg RA, Chaffer CL. Cell plasticity and heterogeneity in cancer . Clin Chem 2013 ; 59 : 168 – 79 . doi: 10.1373/clinchem.2012.184655 . Google Scholar Crossref Search ADS PubMed WorldCat Crossref 11. Azzam DJ , Zhao D, Sun J et al. Triple negative breast cancer initiating cell subsets differ in functional and molecular characteristics and in gamma-secretase inhibitor drug responses . EMBO Mol Med 2013 ; 5 : 1502 – 22 . doi: 10.1002/emmm.201302558 . Google Scholar Crossref Search ADS PubMed WorldCat Crossref 12. Hwang-Verslues WW , Kuo WH, Chang PH et al. Multiple lineages of human breast cancer stem/progenitor cells identified by profiling with stem cell markers . PLoS One 2009 ; 4 :e8377. OpenURL Placeholder Text WorldCat 13. Ponti D , Costa A, Zaffaroni N et al. Isolation and in vitro propagation of tumorigenic breast cancer cells with stem/progenitor cell properties . Cancer Res 2005 ; 65 : 5506 – 11 . Google Scholar Crossref Search ADS PubMed WorldCat 14. Charafe-Jauffret E , Ginestier C, Bertucci F et al. ALDH1-positive cancer stem cells predict engraftment of primary breast tumors and are governed by a common stem cell program . Cancer Res 2013 ; 73 : 7290 – 300 . Google Scholar Crossref Search ADS PubMed WorldCat 15. Wright MH , Calcagno AM, Salcido CD et al. Brca1 breast tumors contain distinct CD44+/CD24-and CD133+ cells with cancer stem cell characteristics . Breast Cancer Res 10 : R10 . Crossref Search ADS PubMed WorldCat 16. Vassilopoulos A , Chisholm C, Lahusen T et al. A critical role of CD29 and CD49f in mediating metastasis for cancer-initiating cells isolated from a Brca1-associated mouse model of breast cancer . Oncogene 2014 ; 33 : 5477 – 82 . Google Scholar Crossref Search ADS PubMed WorldCat 17. Curley MD , Therrien VA, Cummings CL et al. CD133 expression defines a tumor initiating cell population in primary human ovarian cancer . Stem Cells 2009 ; 27 : 2875 – 83 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 18. Son MJ , Woolard K, Nam D-H et al. SSEA-1 is an enrichment marker for tumor-initiating cells in human glioblastoma . Cell Stem Cell 2009 ; 4 : 440 – 52 . Google Scholar Crossref Search ADS PubMed WorldCat 19. Adhikari AS , Agarwal N, Wood BM et al. CD117 and Stro-1 identify osteosarcoma tumor-initiating cells associated with metastasis and drug resistance . Cancer Res 2010 ; 70 : 4602 – 12 . Google Scholar Crossref Search ADS PubMed WorldCat 20. Xu X , Chai S, Wang P et al. Aldehyde dehydrogenases and cancer stem cells . Cancer Lett 2015 ; 369 : 50 – 7 . Google Scholar Crossref Search ADS PubMed WorldCat 21. Shenoy A , Butterworth E, Huang EH. ALDH as a marker for enriching tumorigenic human colonic stem cells . Methods Mol Biol 2012 ; 916 : 373 – 85 . Google Scholar Crossref Search ADS PubMed WorldCat 22. Ginestier C , Hur MH, Charafe-Jauffret E et al. ALDH1 is a marker of normal and malignant human mammary stem cells and a predictor of poor clinical outcome . Cell Stem Cell 2007 ; 1 : 555 – 67 . Google Scholar Crossref Search ADS PubMed WorldCat 23. Shi X , Zhang Y, Zheng J et al. Reactive oxygen species in cancer stem cells . Antioxid Redox Signal 2012 ; 16 : 1215 – 28 . Google Scholar Crossref Search ADS PubMed WorldCat 24. Hadnagy A , Gaboury L, Beaulieu R et al. SP analysis may be used to identify cancer stem cell populations . Exp Cell Res 2006 ; 312 : 3701 – 10 . Google Scholar Crossref Search ADS PubMed WorldCat 25. Cermeño EA , García AJ. Tumor-initiating cells: emerging biophysical methods of isolation . Curr Stem Cell Rep 2016 ; 2 : 21 – 32 . Google Scholar Crossref Search ADS PubMed WorldCat 26. Xu W , Mezencev R, Kim B et al. Cell stiffness is a biomarker of the metastatic potential of ovarian cancer cells . PloS One 2012 ; 7 :e46609. OpenURL Placeholder Text WorldCat 27. V. Swaminathan , K. Mythreye, E. T. O'Brien , A. Berchuck , G. C. Blobe , R. Superfine, Mechanical stiffness grades metastatic potential in patient tumor cells and in cancer cell lines . Cancer Res 2011 , 71 . 5075 – 5080 . Google Scholar Crossref Search ADS PubMed WorldCat 28. Kumar S , Weaver VM. Mechanics, malignancy, and metastasis: the force journey of a tumor cell . Cancer Metastasis Rev 2009 ; 28 : 113 – 27 . doi: 10.1007/s10555-008-9173-4 . Google Scholar Crossref Search ADS PubMed WorldCat Crossref 29. Katira P , Bonnecaze RT, Zaman MH. Modeling the mechanics of cancer: effect of changes in cellular and extra-cellular mechanical properties . Front Oncol 2013 ; 3 : 145 . doi: 10.3389/fonc.2013.00145 . Google Scholar Crossref Search ADS PubMed WorldCat Crossref 30. Zhang W , Kai K, Ueno NT et al. A brief review of the biophysical hallmarks of metastatic cancer cells . Cancer Hallmarks 2013 ; 1 : 59 . Google Scholar Crossref Search ADS PubMed WorldCat 31. Yang J , Weinberg RA. Epithelial-mesenchymal transition: at the crossroads of development and tumor metastasis . Dev Cell 2008 ; 14 : 818 – 29 . Google Scholar Crossref Search ADS PubMed WorldCat 32. Zhang W , Kai K, Choi DS et al. Microfluidics separation reveals the stem-cell–like deformability of tumor-initiating cells . Proc Natl Acad Sci USA 2012 ; 109 : 18707 – 12 . Google Scholar Crossref Search ADS PubMed WorldCat 33. Babahosseini H , Ketene AN, Schmelz EM et al. Biomechanical profile of cancer stem-like/tumor-initiating cells derived from a progressive ovarian cancer model . Nanomedicine 2014 ; 10 : 1013 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat 34. Kwon KW , Choi SS, Lee SH et al. Label-free, microfluidic separation and enrichment of human breast cancer cells by adhesion difference . Lab Chip 2007 ; 7 : 1461 – 8 . doi: 10.1039/b710054j . Google Scholar Crossref Search ADS PubMed WorldCat Crossref 35. Zhang Y , Wu M, Han X et al. High-throughput, label-free isolation of cancer stem cells on the basis of cell adhesion capacity . Angew Chem Int Ed 2015 ; 54 : 10838 – 42 . Google Scholar Crossref Search ADS WorldCat 36. Bansal N , Davis S, Tereshchenko I et al. Enrichment of human prostate cancer cells with tumor initiating properties in mouse and zebrafish xenografts by differential adhesion . Prostate 2014 ; 74 : 187 – 200 . Google Scholar Crossref Search ADS PubMed WorldCat 37. Singh A , Suri S, Lee T et al. Adhesion strength-based, label-free isolation of human pluripotent stem cells . Nat Methods 2013 ; 10 : 438 – 44 . doi: 10.1038/nmeth.2437 . Google Scholar Crossref Search ADS PubMed WorldCat Crossref 38. García AJ , Ducheyne P, Boettiger D. Quantification of cell adhesion using a spinning disk device and application to surface-reactive materials . Biomaterials 1997 ; 18 : 1091 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat 39. Zhou DW , García AJ.. Measurement systems for cell adhesive forces . J Biomech Eng 2015 ; 137 : 020908. OpenURL Placeholder Text WorldCat 40. Li CL , Yang D, Cao X et al. Fibronectin induces epithelial-mesenchymal transition in human breast cancer MCF-7 cells via activation of calpain . Oncol Lett 2017 ; 13 : 3889 – 95 . Google Scholar Crossref Search ADS PubMed WorldCat 41. García AJ , Huber F, Boettiger D. Force required to break alpha5beta1 integrin-fibronectin bonds in intact adherent cells is sensitive to integrin activation state . J Biol Chem 1998 ; 273 : 10988 – 93 . Google Scholar Crossref Search ADS PubMed WorldCat 42. Gallant ND , Michael KE, García AJ. Cell adhesion strengthening: contributions of adhesive area, integrin binding, and focal adhesion assembly . Mol Biol Cell 2005 ; 16 : 4329 – 40 . Google Scholar Crossref Search ADS PubMed WorldCat 43. Hynes RO . Integrins: versatility, modulation, and signaling in cell adhesion . Cell. 1992 ; 69 : 11 – 25 . Google Scholar Crossref Search ADS PubMed WorldCat 44. Hoffman BD , Grashoff C, Schwartz MA. Dynamic molecular processes mediate cellular mechanotransduction . Nature 2011 ; 475 : 316 – 23 . Google Scholar Crossref Search ADS PubMed WorldCat 45. Desgrosellier JS , Cheresh DA. Integrins in cancer: biological implications and therapeutic opportunities . Nat Rev Cancer 2010 ; 10 : 9 – 22 . Google Scholar Crossref Search ADS PubMed WorldCat 46. Sirotnak FM , DeGraw JI, Schmid FA et al. New folate analogs of the 10-deaza-aminopterin series. Further evidence for markedly increased antitumor efficacy compared with methotrexate in ascitic and solid murine tumor models . Cancer Chemother Pharmacol 1984 ; 12 : 26 – 30 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 47. Sugiura K , Stock CCJC. Studies in a tumor spectrum. I. Comparison of the action of methylbis (2-chloroethyl) amine and 3-bis (2-chloroethyl) aminomethyl-4-methoxymethyl-5-hydroxy-6-methylpyridine on the growth of a variety of mouse and rat tumors . Cancer 1952 ; 5 : 382 – 402 . Google Scholar Crossref Search ADS PubMed WorldCat 48. Park J , Schwarzbauer JE. Mammary epithelial cell interactions with fibronectin stimulate epithelial-mesenchymal transition . Oncogene 2014 ; 33 : 1649 – 57 . Google Scholar Crossref Search ADS PubMed WorldCat 49. Signs SA , Fisher RC, Tran U et al. Stromal miR-20a controls paracrine CXCL8 secretion in colitis and colon cancer . Oncotarget 2018 ; 9 : 13048 . Google Scholar Crossref Search ADS PubMed WorldCat 50. Carpentino JE , Hynes MJ, Appelman HD et al. Aldehyde dehydrogenase-expressing colon stem cells contribute to tumorigenesis in the transition from colitis to cancer . Cancer Res 2009 ; 69 : 8208 – 15 . Google Scholar Crossref Search ADS PubMed WorldCat 51. Mannello F . Understanding breast cancer stem cell heterogeneity: time to move on to a new research paradigm . BMC Med 2013 ; 11 : 169 . Google Scholar Crossref Search ADS PubMed WorldCat 52. Huff CA , Matsui WH, Smith BD et al. Strategies to eliminate cancer stem cells: Clinical implications . Eur J Cancer 2006 ; 42 : 1293 – 7 . Google Scholar Crossref Search ADS PubMed WorldCat 53. Jones RJ , Matsui W. Cancer stem cells: from bench to bedside . Biol Blood Marrow Transpl 2007 ; 13 : 47 – 52 . Google Scholar Crossref Search ADS WorldCat 54. Dragu DL , Necula LG, Bleotu C et al. Therapies targeting cancer stem cells: Current trends and future challenges . World J Stem Cells 2015 ; 7 : 1185 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 55. Koppaka V , Thompson DC, Chen Y et al. Aldehyde dehydrogenase inhibitors: a comprehensive review of the pharmacology, mechanism of action, substrate specificity, and clinical application . Pharmacol Rev 2012 ; 64 : 520 – 39 . Google Scholar Crossref Search ADS PubMed WorldCat 56. Storms RW , Trujillo AP, Springer JB et al. Isolation of primitive human hematopoietic progenitors on the basis of aldehyde dehydrogenase activity . Proc Natl Acad Sci USA 1999 ; 96 : 9118 – 23 . Google Scholar Crossref Search ADS PubMed WorldCat 57. Storms RW , Green PD, Safford KM et al. Distinct hematopoietic progenitor compartments are delineated by the expression of aldehyde dehydrogenase and CD34 . Blood 2005 ; 106 : 95 – 102 . Google Scholar Crossref Search ADS PubMed WorldCat 58. Kastan MB , Schlaffer E, Russo J-E et al. Direct demonstration of elevated aldehyde dehydrogenase in human hematopoietic progenitor cells . Blood 1990 ; 75 : 1947 – 50 . Google Scholar Crossref Search ADS PubMed WorldCat 59. Januchowski R , Wojtowicz K, Zabel M. The role of aldehyde dehydrogenase (ALDH) in cancer drug resistance . Biomed Pharmacother 2013 ; 67 : 669 – 80 . Google Scholar Crossref Search ADS PubMed WorldCat 60. Huang EH , Hynes MJ, Zhang T et al. Aldehyde dehydrogenase 1 is a marker for normal and malignant human colonic stem cells (SC) and tracks SC overpopulation during colon tumorigenesis . Cancer Res 2009 ; 69 : 3382 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat 61. Yang L , Ren Y, Yu X et al. ALDH1A1 defines invasive cancer stem-like cells and predicts poor prognosis in patients with esophageal squamous cell carcinoma . Modern Pathol 2014 ; 27 : 775 – 83 . Google Scholar Crossref Search ADS WorldCat 62. Didar TF , Tabrizian M. Adhesion based detection, sorting and enrichment of cells in microfluidic lab-on-Chip devices . Lab Chip 2010 ; 10 : 3043 – 53 . Google Scholar Crossref Search ADS PubMed WorldCat 63. Shields CW , Reyes CD, López GP. Microfluidic cell sorting: a review of the advances in the separation of cells from debulking to rare cell isolation . Lab Chip 2015 ; 15 : 1230 – 49 . Google Scholar Crossref Search ADS PubMed WorldCat 64. Han W , Allio BA, Foster DG et al. Nanoparticle coatings for enhanced capture of flowing cells in microtubes . ACS Nano 2010 ; 4 : 174 – 80 . Google Scholar Crossref Search ADS PubMed WorldCat 65. Tasadduq B , McFarland B, Islam M et al. Continuous sorting of cells based on differential P Selectin glycoprotein ligand expression using molecular adhesion . Anal Chem 2017 ; 89 : 11545 – 51 . Google Scholar Crossref Search ADS PubMed WorldCat 66. Islam M , Brink H, Blanche S et al. Microfluidic sorting of cells by viability based on differences in cell stiffness . Sci Rep 2017 ; 7 : 1997 . Google Scholar Crossref Search ADS PubMed WorldCat 67. Islam M , Mezencev R, McFarland B et al. Microfluidic cell sorting by stiffness to examine heterogenic responses of cancer cells to chemotherapy . Cell Death Dis 2018 ; 9 : 239 . doi: 10.1038/s41419-018-0266-x . Google Scholar Crossref Search ADS PubMed WorldCat Crossref 68. Mohammadalipour A , Burdick MM, Tees DFJ. Deformability of breast cancer cells in correlation with surface markers and cell rolling . FASEB J 2018 ; 32 : 1806 – 17 . doi: 10.1096/fj.201700762R . Google Scholar Crossref Search ADS PubMed WorldCat Crossref 69. García AJ , Ducheyne P, Boettiger D. The effect of surface reaction stage on fibronectin-mediated adhesion of osteoblast-like cells to bioactive glass . J Biomed Mater Res 1998 ; 40 : 48 – 56 . Google Scholar Crossref Search ADS PubMed WorldCat © The Author(s) 2020. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permission@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Hydrodynamic shear-based purification of cancer cells with enhanced tumorigenic potential
JF - Integrative Biology
DO - 10.1093/intbio/zyz038
DA - 2020-02-22
UR - https://www.deepdyve.com/lp/oxford-university-press/hydrodynamic-shear-based-purification-of-cancer-cells-with-enhanced-0ZOedrvWbf
SP - 1
VL - 12
IS - 1
DP - DeepDyve
ER -