Lipid tethering of breast tumor cells enables real-time imaging of free-floating cell dynamics and drug response

Abstract

Free-floating tumor cells located in the blood of cancer patients, known as circulating tumor cells (CTCs), have become key targets for studying metastasis. However, effective strategies to study the free-floating behavior of tumor cells in vitro have been a major barrier limiting the understanding of the functional properties of CTCs. Upon extracellular-matrix (ECM) detachment, breast tumor cells form tubulin-based protrusions known as microtentacles (McTNs) that play a role in the aggregation and re-attachment of tumor cells to increase their metastatic efficiency. In this study, we have designed a strategy to spatially immobilize ECM-detached tumor cells while maintaining their free-floating character. We use polyelectrolyte multilayers deposited on microfluidic substrates to prevent tumor cell adhesion and the addition of lipid moieties to tether tumor cells to these surfaces through interactions with the cell membranes. This coating remains optically clear, allowing capture of high-resolution images and videos of McTNs on viable free-floating cells. In addition, we show that tethering allows for the real-time analysis of McTN dynamics on individual tumor cells and in response to tubulin-targeting drugs. The ability to image detached tumor cells can vastly enhance our understanding of CTCs under conditions that better recapitulate the microenvironments they encounter during metastasis.

Keywords: breast cancer; circulating tumor cells; microfluidics; microtentacles; polyelectrolyte multilayers.

Figure 1. PEMs form a cytophobic layer allowing McTN visualization on microfluidic devices

A. Schematic depicting coating of microfluidic slide with PEMs to maintain free-floating behavior of tumor cells. B. PEM coatings increase in thickness (left axis, black) with the number of bilayers but maintain optical clarity (right axis, gray). Data (mean ± SEM) correspond to samples in triplicate. C. Maximum intensity z-projection of MDA-MB-436 cells on PMA₄/PAAm₄ surfaces showing McTNs (arrows). Scale bar = 10μm. Percent of D. MDA-MB-436 and E. MCF-7 cells (mean ± SEM) remaining on surfaces after washing uncoated slides or slides coated with 1, 4, or 8 PMA/PAAm bilayers. Data represents mean values of three independent experiments.

Figure 2. Modification of PEMs with lipid tethers

A. Schematic depicting how lipid-terminated PEMs promote interaction with tumor cell membranes. B. Film thickness and C. optical clarity (mean ± SEM) after addition of DOTAP (blue) and DOPC [red]. Lipids promote growth of film while maintaining an optically-clear substrate for imaging. Data correspond to the mean of samples prepared in triplicate with three measurements per surface.

Figure 3. DOTAP tethers breast cancer cells

Percent cell retention of A. MDA-MB-436 and B. MCF-7 cells plated on microfluidic slides coated with PMA₄/PAAm₄ bilayers alone or with DOTAP. Percent cell retention of C. MDA-MB-436 and D. MCF-7 cells plated on microfluidic slides coated with PMA₄/PAAm₄ bilayers alone or with DOPC. The remaining cells after each wash was quantified with CellProfiler and normalized to the initial cell number. Data represents mean of triplicate independent experiments (mean ± SEM). E. Representative images of MDA-MB-436 cells at time 0 and after 3 subsequent washes on PMA₄/PAAm₄ bilayers with no tether and with PEM-DOTAP or PEM-DOPC tethers at 4x magnification. Scale bar = 200μm. *P<0.05 **P<0.01.

Figure 4. Lipid tethering retains free-floating characteristics of breast tumor cells and does not affect cell viability

A. McTN quantification of MDA-MB-436 and MCF-7 cells suspended on a low-attach plate, microfluidic slides with PEM-no tether, and microfluidic slides with PEM-DOTAP tether. Data represents blinded quantification of McTN frequency from three independent experiments with 100 cells counted for each (mean ± SEM). B. Representative images of McTNs (arrows) on MDA-MB-436 cells seeded on PEM-no tether and PEM-DOTAP tether microfluidic slides at 40x magnification. Scale bar = 10μm. Viability of C. MDA-MB-436 and D. MCF-7 cells calculated at 0 and 6 hrs after seeding on microfluidic slides with PEM-DOTAP tether. Data represents mean cell viability from three independent experiments (mean ± SEM). E. Representative images show viability of MDA-MB-436 cells tethered by DOTAP for 6 hrs. Phase contrast, live (green fluorescence), and dead (red fluorescence) images taken at 4x magnification. Scale bar = 200μm.

Figure 5. Lipid tethering allows for real-time McTN imaging in response to drug treatment and minimizes effects of drift

A. McTN (arrows) imaging of MDA-MB-436 cells seeded on microfluidic slides with PEM-no tether (i-iii) and PEM-DOTAP tether (iv-vi). Representative 1μm slice (i and iv), maximum z-projection of 5 slices at one time point (ii and v), and maximum t-projection after 20 frames (iii and vi) are shown at 60x magnification. B. McTN (arrow) imaging of MDA-MB-436 cells seeded on microfluidic slides with PEM-DOTAP tether after treatment with 5μM colchicine for 15 mins and 1μg/ml paclitaxel for 120 mins. Maximum intensity z-projections of five 1μm slices at one time point are shown at 60x magnification. Complete time-lapse movies are available in Fig. S7. Scale bar = 10μM. C. Manual quantification of the average number of McTNs per cell (mean ± SEM) in MDA-MB-436 cells treated with colchicine and paclitaxel. **P<0.01 ***P<0.001.