1. Introduction

It is well known that cell mechanics plays an important role in various cellular functions [1–4], and in many disease-related mutations, especially for carcinogenesis [5]. Besides, it was also demonstrated that cell's mechanical properties can be used as an intrinsic and reliable marker to distinguish between different cell types [6,7]. Cells’ cytoskeleton, the fundamental support and basic structure for cell mechanics, has a complex and dynamic framework mainly composed of actin filaments, microtubules and intermediate filaments, which continuously assemble and disassemble according to cell status. This dynamics can be modified by drugs, as well as by environment alterations, both internal to the cell or external [8]. In particular, the aforementioned processes of living cells are strongly temperature dependent because the relevant chemical reactions, the protein functionality and the constitutive materials are inherently temperature sensitive. Therefore, cellular mechanical properties can be affected by temperature changes. Different techniques for cell-mechanics study, ranging from micropipette aspiration to atomic force microscopy and optical stretching, have been extensively studied and summarized in literature [9]. Thermo-rheological experiments have been recently reported in literature [10], and optical stretchers have been exploited to analyze cell deformability as a function of environmental temperature, allowing to obtain detailed data about temperature-sensitive transient receptor potential vanilloid (TRPV2) and thermal unbinding of transient cross-links [11]. In this study, we investigate the temperature impact on cell mechanics by focusing on long-term temperature effect in a wide range of temperatures (~0 – 35 °C), especially in the low-temperature range. To evaluate the temperature-effect on cell mechanics, we used two integrated optofluidic microchips fabricated by 3D femtosecond laser micromachining [12], combined with an additional temperature control system: in the first chip (an optical stretcher, OS in the following) we exploit optical forces to induce cell deformation, whereas in the second one (a constriction-chip, CC in the following) we test cell mechanics in terms of their ability to passively squeeze through a small constriction.

2. Material and method

2.1 Optical stretcher

The OS, shown in Fig. 1 and already described in [13], was mounted on a phase contrast microscope equipped with a CCD camera. The input tube was connected to a vial containing the cell suspension, while the output one was immersed in water to avoid any possible pressure turbulence caused by droplet detachment or droplet size-variation during the experiments. The cell flow rate inside the microchip was regulated by an external micro-pump with high-precision pressure control (~1mbar). A CW Yb-doped fiber laser (P_MAX = 10 W @ 1070 nm) was used as light source: the optical power was evenly split into two fibers and then coupled to the two waveguides of the chip through a single-mode optical fiber. The entire system was controlled using a custom LabVIEW program. The cell stretching experiment was based on the following procedure: low laser power (~25mW, P_trap) was first emitted by each waveguide to capture a single cell [14]; the laser power was then raised to the “stretching power” (~1.2 W per side) for 5 s, and then reduced to P_trap for 5 s more before releasing the cell. A video was recorded, at 15 fps, during the whole procedure.

 

Fig. 1 a) image of the optical stretcher chip; b) microchannel with two facing waveguides (not completely in focus). Yellow dashed line indicates the trapping/stretching region; c) and d) represent the same cell trapped and stretched respectively. Green contours are cell borders identified by the recognition algorithm.

Download Full Size | PDF

Thanks to the high imaging quality of the chip, and to the use of an edge detection algorithm [15], it was possible to recognize the contour of every measured cell, in each frame, with subpixel accuracy, as shown for example in Fig. 1(c)-1(d), thus allowing to extract the cellular deformation of each measured cell. By indicating with x_orig (y_orig) the initial cell dimension along the optical beams (and cell flowing) direction, the cell radius “RMS” was calculated using Eq. (1). Similarly, by indicating with x_max (y_min) the cell dimensions in the maximum deformation condition (see Fig. 1(d)), the maximum ellipticity variation (Var parameter) was evaluated using Eq. (2), where Corr is a correction factor taking into account the impact of different cell size on optical forces distribution [15,16].

2.2 Constriction chip

The CC (see Fig. 2) has two inlets, connected to two vials, and two outlets with tubing dipped in water solution as for the OS. The core of the chip is similar to a cell sorter [17], but additionally the light output by a waveguide is used to apply a controlled force on selected cells, as indicated in the figure, sending them to the constriction. The operating principle of the CC is described in the following: one of the two inlets (inlet 2) is fed with a pure buffer solution and the other one (inlet 1) with the same buffer containing cells. By balancing the pressures of the two channels, a stable laminar flow is obtained in the central channel so that all of the cells coming from inlet 1 exit through outlet 1. When a cell must be tested, the laser source is turned on, sending the selected cell into the other stream (in the “lower part” of the central channel in Fig. 2), and making the selected cell to flow towards the constriction branch. When the sorted cell reaches the constriction, it stops flowing and partially clogs the channel. A slow pressure ramp (1 mbar/s) is then applied to inlet 2, so as to increase the force pushing the cell. The pressure required by each cell to passively squeeze through the constriction is defined as the “passing pressure” of that cell. It should be noted that, differently from conventional constrictions, the constriction length in our chip is very short (< 10μm), hence the experiment results are more related to the “cell entrance” into the constriction than to the cell passage through it. Due to friction and adhesion effects more deformable cells need a higher pressure to squeeze through the constriction, as discussed in [18].

 

Fig. 2 microscope image of constriction chip under bright field, locally enlarged constriction part has dimension of 8 × 12 μm². Scare bar: 100μm

Download Full Size | PDF

2.3 Cell sample and temperature control

Highly metastatic human melanoma (A375MC2) [19] cell sample was obtained from the Cancer Research Center (Howard Hughes Medical Institute, MIT), cultured in Petri dishes and grown in Dulbecco’s Modified Eagle’s Medium supplemented with 10% Fetal Bovine Serum, penicillin (0.1 mg/ml), streptomicin (100 U/ml), 0.2 mM glutamine and 1X non-essential aminoacid (all by Euroclone). Cells were maintained into an incubator at 37°C in a humidified atmosphere at 5% CO₂. For each measurement, 10⁶ cells were plated in Petri dishes and detached about 24 h later by trypsinization. Cells were then suspended (2 × 10⁵ cells/ml) in phosphate buffered saline (PBS) solution for each measurement.

Cells were tested at five temperatures: ~0°C, 5°C, 15°C, 25°C and 35°C, obtained by immersing the two chips, and a large part of the tubes, in a temperature controlled water bath. Thanks to a temperature sensor, positioned close to the measurement region, and by regularly flushing water at the desired temperature, we achieved a relatively stable setup temperature, with a maximum deviation from the desired value of ± 1°C, which is sufficient to avoid overlapping of the temperature intervals. To keep cells at the desired temperature, and to give to cells’ internal structures a relatively long time to re-organize, we kept the samples at the target temperature for 60 min before each measurement session.

3. Results and discussion

Given the variability of cell properties, > 60 cells were analyzed by OS and CC at every temperature. Measurements at ~0°C were performed using the OS only, to evaluate the impact of microtubule depolymerization, which was confirmed by fluorescent microscopy.

3.1 Cellular size and ellipticity

A first analysis of the cytoskeleton response to different temperatures was carried out by analyzing the change of cell size and ellipticity, defined as the ratio between horizontal and vertical dimension of the trapped cell. Cell size distribution at the different temperatures is reported in Fig. 3(a) while in Fig. 3(b) we show the measured cell ellipticity. Both data are presented as standard box plots, where the central squares indicate the measurement mean value, the central line for the median value, the box for interquartile range, the whiskers for 1%–99% and the external symbols for maximum/minimum values. It is evident that cell size is not significantly affected by temperature; however it is worth noticing that by increasing the temperature a rounder cell shape is generally observed (i.e. the ratio between x and y is closer to 1), offering a preliminary suggestion that the cell cytoskeleton could become softer when temperature is increased, thus more easily reaching the minimum-energy configuration.

 

Fig. 3 A375 MC2 cell size distribution a) and x/y ratio b) at different temperatures. Measurements were performed while trapping the cell by using the OS. Box width is not related to temperature variations.

Download Full Size | PDF

3.2 Optical deformation and passage through constrictions

In OS, cells under optical force show a creep compliance behavior [14,20]. The dependence of the maximum ellipticity variation (Var, see Eq. (2) on temperature is reported in Fig. 4(a), showing that higher values of Var are observed at higher temperatures, in agreement with results recently reported in literature [10,11]. In order to evaluate the significance of the differences, Mann-Whitney U-test was performed as the population was non-normal according to the Kolmogorov–Smirnov test. The obtained results on optical deformability showed a statistically significant difference (p < 10⁻³) between all the five temperatures.

 

Fig. 4 Optical deformation a) and passing pressure b) of A375 MC2 at different temperatures

Download Full Size | PDF

As known from literature [10,21,22], laser power can induce a sudden increase of the cell temperature, making it difficult to compare the deformation values obtained at different optical power levels. As in our case the same optical power is applied to all OS measurements, we expect an increase of about 30°C in the irradiated region for all the samples, provided that absorption coefficient changes due to the different temperatures are negligible. Consequently, the cells’ temperature-difference between the samples is maintained, thus not affecting the analysis of our results. Data reported in Fig. 4(a) also show that the increase of optical deformability is stronger at higher temperatures: a possible explanation for this effect is that more components of cytoskeleton and faster processes are involved at higher temperature; another possibility is that at higher temperature the cytosol viscosity decreases, allowing for smoother and faster cells’ shape changes [11], and coherently with the ellipticity data reported in Fig. 3(b). It is worth discussing the results obtained by measuring cells kept at ~0 °C, as microtubules, considered as rigid cytoskeletal component [23], are known to depolymerize at low temperatures [24,25]. Even if a higher optical deformability of cells could be intuitively expected, we observed a slight decrease of Var, with respect to the value obtained at 5°C, thus confirming that microtubules do not have a major impact on cell compliance measurements performed by OS, as suggested in [26].

Cell ability to pass through a microchannel with a cross-section significantly smaller than the cell itself attracted much attention [1,27], as this feature is particularly evident in metastatic cells, thus suggesting its use for cells’ diagnostics, or for new drugs testing. Figure 4(b) reports the experimental results obtained with the CC chip by applying the procedure described in section 2.2. The results show that the passing pressure of A375MC2 gradually increases with temperature: this means that cell passive-squeezing through the constriction is hindered by the high temperature, as a higher pressure (i.e. a stronger force) is required to push the cell through the CC. These results are in good agreement with those already reported in [18] about mechanical properties of cells undergoing different drugs treatments: softer cells (in that case modified by drug-treatments) required higher pressures to be squeezed through the constriction. This is due to the fact that more deformable cells create a larger contact area between the cell and the constriction entrance, hence increasing the overall friction and requiring a higher passing pressure. Regarding the results reported in Fig. 4(b) statistically significant differences were obtained comparing the different samples, always by exploiting the Mann-Whitney U-test, with the only exclusion of the 15°C-25°C comparison.

It is interesting to underline that, not being affected by any laser-induced “temperature-shift”, the CC measurements at high temperature confirm the deformability trends observed by OS. In addition the low-temperature CC measurements allow investigating cells’ properties in a significantly different range of cell-temperatures. As a final remark regarding the cell viability, it is worth noting that in the case of stretching experiments performed on cells initially at 35°C, the impact of temperature on their viability has to be carefully considered [21], since during the measurement the cell is expected to exceed the damage-threshold temperature. Conversely, if low-temperature investigations are conducted, the effect of 5-s heating on viability is generally negligible, as discussed in [7].

4. Conclusion

By using two temperature controlled optofluidic microchips we analyzed the changes of metastatic cells’ mechanical properties induced by temperature variations. It is observed that cell optical deformability is strongly increased by raising cell temperature. Additionally, cellular ability to passively-squeeze through a constriction was investigated, and the obtained results prove that as temperature is increased a higher pressure is required for cell passage. As this behavior was observed for softer cells, these results strongly support those obtained by using the optical stretcher. Overall, results demonstrate that cell mechanical properties are strongly dependent on temperature, and that as the temperature is increased the cell appears much softer. Although cellular mechanics is a very complex process, this study shows that even moderate temperature variations can induce a clear modification of cells’ properties (e.g. + 1% optical deformation for a 2°C temperature increase), thus highlighting the importance of a proper temperature control, both in experimental setups for cell mechanics evaluation, and to carry out comparisons between different samples.