Single-cell adhesion strength and contact density drops in the M phase of cancer cells

Abstract

The high throughput, cost effective and sensitive quantification of cell adhesion strength at the single-cell level is still a challenging task. The adhesion force between tissue cells and their environment is crucial in all multicellular organisms. Integrins transmit force between the intracellular cytoskeleton and the extracellular matrix. This force is not only a mechanical interaction but a way of signal transduction as well. For instance, adhesion-dependent cells switch to an apoptotic mode in the lack of adhesion forces. Adhesion of tumor cells is a potential therapeutic target, as it is actively modulated during tissue invasion and cell release to the bloodstream resulting in metastasis. We investigated the integrin-mediated adhesion between cancer cells and their RGD (Arg-Gly-Asp) motif displaying biomimetic substratum using the HeLa cell line transfected by the Fucci fluorescent cell cycle reporter construct. We employed a computer-controlled micropipette and a high spatial resolution label-free resonant waveguide grating-based optical sensor calibrated to adhesion force and energy at the single-cell level. We found that the overall adhesion strength of single cancer cells is approximately constant in all phases except the mitotic (M) phase with a significantly lower adhesion. Single-cell evanescent field based biosensor measurements revealed that at the mitotic phase the cell material mass per unit area inside the cell-substratum contact zone is significantly less, too. Importantly, the weaker mitotic adhesion is not simply a direct consequence of the measured smaller contact area. Our results highlight these differences in the mitotic reticular adhesions and confirm that cell adhesion is a promising target of selective cancer drugs as the vast majority of normal, differentiated tissue cells do not enter the M phase and do not divide.

Figure 1

Schematic representation of the adhesion strength measurements on single cells using a computer-controlled micropipette and the high-resolution RWG biosensor (Single-Cell RWG)^,. (A) Region of interest (ROI) of RGD-displaying Petri dish surface containing HeLa Fucci cells is scanned, then cells are automatically detected and selected for measurement. The developed device automatically adjusted the vacuum in a syringe connected to a micropipette with 70 micron opening, positioned the micropipette above the targeted cell and opened the fluidic valve. Adhesion characteristics of cells were evaluated by calculating the ratio of still adhering cells after the application of subsequent suction force steps on hundreds of cells. (B) Incident light is coupled into the biosensor chip through the waveguide grating and penetrates into a 150 nm depth into the sample (adhering cell) on the chip surface in the form of an evanescent field (red shadow above the waveguide surface). Thus field is ideal to monitor the cell-substratum contact zone. Biochemical or cellular events, such as cell adhesion change the local refractive index inside the evanescent field, resulting in a wavelength shift of the reflected light. Due to the large spatial resolution, individual cells are clearly visible on the recorded wavelength shift map. The wavelength shift is sensitive to nanometer scale changes in the cell adhesion contacts (perpendicular to the sensor surface). Such tiny variations are not resolvable by traditional optical microscopy.

Figure 2

(A) Ratio of adherent HeLa Fucci cells on PLL-g-PEG-RGD surface at different vacuum values, as measured by the computer-controlled micropipette. * indicates significant difference between the ratio of different cell phases. (B) Phase contrast image of a small area of the Petri dish. The shortest path along all cells selected for measurement was calculated by the CellSorter software and projected onto the image in yellow. (C) Expression of the Fucci fluorescent marker set along the cell cycle. The red reporter is attached to the CDT1 protein, expressed early in the cell cycle, while the green reporter is attached to Geminin protein, which is expressed later in the cell cycle. Both reporters are expressed for a short time resulting in yellow color. Neither proteins are expressed during mitosis, making the cells invisible (colorless) in fluorescent images. Scale bar represents 100 µm. (D) Correlation between the cell color and the average cell area measured. Bar chart results are from 4 different micropipette experiments, where n = 88 for red cells, n = 90 for green cells, n = 95 for yellow cells and n = 32 for colorless cells. * indicates significant difference between the areas of cells of different colors.

Figure 3

(A) Cell adhesion strength of HeLa Fucci cells on PLL-g-PEG-RGD surface as a function of cell color and morphology measured with the computer-controlled micropipette. Aggregated cell adhesion data shown in Fig. 2A were separated on the basis of cell morphology. As expected, round cells showed a lower affinity to adhere to the surface than flat cells did, regardless of cell color. (Flattened colorless cells were very rare, thus we could not collect enough data to obtain their adhesion curve.) (B) Determination of the average cell area as a function of cell color and cell morphology. All charts show results from 4 different micropipette experiments, and n = 31 flattened red cells, n = 57 rounded red cells, n = 56 flattened green cells, n = 34 rounded green cells, n = 57 flattened yellow cells, n = 38 rounded yellow cells, n = 32 rounded colorless cells. (C) Cells were divided into eight groups based on their color and morphology. From top to bottom: rounded red cell, flattened red cell, rounded green cell, flattened green cell, rounded yellow cell, flattened yellow cell, rounded colorless cell. Scale bars represent 10 µm.

Figure 4

Comprehensive representation of the biosensor data. (A) Top row: Phase contrast and fluorescent composite images of typical red, yellow, green, and colorless cells (respectively). Scale bar represents 25 µm. Bottom row: Composite images of the respective top row cells overlayed from the biosensor wavelength shift map and the phase contrast image at t = 10 min. (B) Near-stationary time dependence of maximal biosensor pixel values (WS) for representative cells depicted in Fig. 4A. (C) Schematic representation of the robotic fluidic force microscopy (FluidFM BOT) measurement process. A special hollow FluidFM cantilever capable of whole cell force-spectroscopy was used to calibrate biosensor signals for adhesion force and energy conversion as described in. (D) Time-dependence of IWS and deduced force/energy signals of single cells in different phases of the cell cycle. (E) Correlation between cell color and IWS, adhesion energy, and adhesion strength. A significant difference was found between colorless cells with respect to green and yellow cells. Graphs show the average of n = 7 red cells, n = 11 green cells, n = 8 colorless cells, and n = 11 yellow cells ± SEM.

Figure 5

Biosensor data of the recorded single-cell signals at a single pixel underneath of each cell. (A) The local biosensor signals of individual cells (maximal WS pixel value corresponding to a given cell) followed a lognormal distribution (solid line) in all cell-cycle states. (B) Normalized data showed significant differences between the various phases in a decreasing order of red, green, and yellow cells, with a distinctly less signal for the colorless cells. Mean ± SEM is shown. (n = 325 red cells, n = 197 green cells, n = 30 colorless cells, and n = 102 yellow cells).