Quantitative phase microscopy for non-invasive live cell population monitoring

Abstract

We present here a label-free development based on preexisting Quantitative Phase Imaging (QPI) that allows non-invasive live monitoring of both individual cells and cell populations. Growth, death, effect of toxic compounds are quantified under visible light with a standard inverted microscope. We show that considering the global biomass of a cell population is a more robust and accurate method to assess its growth parameters in comparison to compiling individually segmented cells. This is especially true for confluent conditions. This method expands the use of light microscopy in answering biological questions concerning live cell populations even at high density. In contrast to labeling or lysis of cells this method does not alter the cells and could be useful in high-throughput screening and toxicity studies.

Figure 1

(A) 4-day time-lapse imaging of a single HEK cell. Dot-plot of the dry mass feature at every time point (15 min interval). The cell dry mass increases to a maximum value, and suddenly drops to a roughly half minimum value, corresponding to cell mitosis. It then slowly increases again until mitosis, and so on. (B) Detail of the cell being monitored (turquoise-colored cell) at lapses 6–7, when mitosis of the adjacent cell (burgundy-colored) results in a brief artifact in the segmentation of the monitored cell, resulting in the sudden but transient drop of dry mass seen at the beginning of the time-lapse. The "missing" dry mass is in the burgundy area circled by the thin white line. The picture field width is approximately 150 µm.

Figure 2

(A) Image a: Cos cells, segmented in b. The width of the picture is approximately 200 µm. Image c: HEK cells, segmented in d. The pictures are approximately 300 µm in width. For both Cos and HEK cells, some segmented areas encompass more than a single cell, as illustrated in the drawing of the 2 ellipses. However, for both cell types, all the cells present in the fields are within the segmented areas. (B) Bar graph representation of the ratio "integrated dry mass /integrated nuclear fluorescence intensity" for 4 different cell types at different plating concentrations. The error bars are the standard deviations calculated from the average of the Summed Dry mass/Total fluorescence ratios of all matching interferograms/fluorescent cell images analyzed in each of the 19 conditions presented in this experiment (4 cell types and 5 plating densities, except REF: 4 plating densities). There were from 12 to 24 independent interferograms/fluorescent images analyzed for each of the 19 conditions presented. Number of cells analyzed are shown in Supplementary Information, Table S1. (C) 2 examples of REF cell incomplete segmentation. a' and b' are contrasted version of a and b, respectively, where we can see that the cell periphery is barely detectable (circled with red dashed line). The picture field width is approximately 300 µm.

Figure 3

92.5 h-live monitoring of HEK populations starting on day 1 after addition of increasing staurosporine (STS) concentrations. Each line on these plots corresponds to the monitoring of one field encompassing multiple cells together with its linear fit (same color), and each STS concentration is monitored on 9 fields. (A) 4 nM STS, (B) 20 nM STS, (C) 100 nM STS, (D) 500 nM STS.

Figure 4

(A) Average of the 9 fields monitored for the 4 STS concentrations depicted in Fig. 3, with the same color code. Green: 4 nM STS, yellow: 20 nM STS, orange: 100 nM STS, red: 500 nM STS. (B)Same as (A), but using Surface features instead of Dry mass.

Figure 5

6-day live monitoring of REF populations in the presence of increasing staurosporine (STS) concentrations. Green: 0 nM STS; orange: 4 nM STS; red: 20 nM STS. (A) Each line on these plots corresponds to the monitoring of one field encompassing multiple cells together with its linear fit (same color), and each STS concentration was monitored on 2 fields with initial seeding of the 24-plate with 20,000 cells. (B) Average of the 2 fields of the 3 conditions monitored in (A), with the same color code. Green: 0 nM STS, orange: 4 nM STS; red: 20 nM STS.