A highly active mineral-based ice nucleating agent supports in situ cell cryopreservation in a high throughput format

Abstract

Cryopreservation of biological matter in microlitre scale volumes of liquid would be useful for a range of applications. At present, it is challenging because small volumes of water tend to supercool, and deep supercooling is known to lead to poor post-thaw cell viability. Here, we show that a mineral ice nucleator can almost eliminate supercooling in 100 µl liquid volumes during cryopreservation. This strategy of eliminating supercooling greatly enhances cell viability relative to cryopreservation protocols with uncontrolled ice nucleation. Using infrared thermography, we demonstrate a direct relationship between the extent of supercooling and post-thaw cell viability. Using a mineral nucleator delivery system, we open the door to the routine cryopreservation of mammalian cells in multiwell plates for applications such as high throughput toxicology testing of pharmaceutical products and regenerative medicine.

Keywords: cell cryopreservation; heterogeneous ice nucleation; ice nucleation; multiwell plates; supercooling; toxicology screening.

Figure 1.

(a) Cartoon of IR-NIPI instrument with example IR images of 96-well plate and well ice nucleation events used to obtain all the data in this figure. (b) Demonstration of thermal history of well liquid and determination of ice nucleation temperature (T_nuc). Numbers 1–4 correspond to IR images of well at particular points of the thermal history, depicted on the right edge of the figure.

Figure 2.

Schematics for (a) HepG2 plate cryopreservation protocol and (b,c) IceStart arrays. Note that IceStart arrays were only used for the high-throughput trials in Results section.

Figure 3.

Boxplot showing IR-NIPI results of droplet freezing assays for 50 µl droplets of suspensions of several ice nucleating materials in water at different concentrations. Boxes depict 25th–75th percentiles, bars depict absolute ranges and yellow stars depict median temperatures of droplet freezing.

Figure 4.

(a–e) Thermal history traces of droplet freezing assays shown in figure 3 at concentrations of 1% (w/w). (f–i) Detailed thermal history of LDH1 runs at all concentrations illustrating the apparent elimination of supercooling. Numbers indicate droplet temperature (°C) on x-axis and time into freezing run (s) on y-axis.

Figure 5.

Number of active sites per unit mass as a function of temperature (n_m(T)) of LDH1 and other strong ice-nucleators. Data points derived using droplet freezing assays in either 1 µl or 50 µl droplets. Fits for each material were plotted apart from birch pollen which is a parameterization from Pummer et al. [29]. No fit was plotted for AgI due to the spread of data points.

Figure 6.

Results for a 96-well plate with HepG2 culture cryopreserved with IR-NIPI instrument showing relationship between T_nuc and post-thaw cell survival. (a) Correlation plot of individual well T_nuc and post-thaw cell survival values; (b) spatial visualization of plate with post-thaw cell survival data; (c) spatial visualization of plate with well T_nuc data; (d) thermal history of wells with detail for wells spiked with LDH1. Data for wells with pure water droplets, used for temperature calibration of the IR camera, are also shown.

Figure 7.

Results for high-throughput trials of monolayers of HepG2 cells cryopreserved in 96-well plate with and without controlled IN. Viability assays performed immediately post-thaw (a–c) and 5 days post-thaw (d–f). Example results for individual plates showing layout of IN method and results are shown in (c) and (f) with numbers denoting thousands of cells and colour strength indicating post-thaw viability (s.d. = standard deviation). Boxplots and ANOVA results for individual freezing runs are shown in (a) and (d). Overall post-thaw survival rates averaged over all eight runs are shown in (b) and (e). Note that manual nucleation was not used in the 5-day trial as this technique could not be executed in sterile conditions. Error bars indicate standard error of mean; bars and asterisk (*) indicate significance between groups was p < 0.05.

Figure 8.

Microscope images of HepG2 cell monolayers after neutral red staining illustrating differences in cell morphology dependent on IN control. Low magnification images of whole wells on left of (c–f) illustrate condition of monolayer (well diameter is 6.4 mm for indication of scale). Higher magnification images on right of (c–f) illustrate condition of individual cells. Scale bar represents 200 µm.