Cell-compatible isotonic freezing media enabled by thermo-responsive osmolyte-adsorption/exclusion polymer matrices (2025)

Isotonic polyZI solutions as freezing media

PolyZIs were synthesized by introducing a vinyl group into an imidazolium/carboxylate-type ZI and an imidazolium/sulfonate-type ZI, which have demonstrated good performance as low-molecular ZIs¹⁷. The carboxylate-type polyZI was readily soluble in ultrapure water, while the sulfonate-type polyZI was not (Supplementary Fig.1). Therefore, we used the carboxylate-type polyZI as the cryoprotectant in all subsequent experiments.

Two types of cells [mouse normal fibroblasts (mNF) and human kidney cells (BOSC)] were cryopreserved using the polyZI. Low-molecular ZIs generally result in moderate and low viability after the cryopreservation of mNF and BOSC, respectively¹⁷. For comparison, the imidazolium/carboxylate-type ZI monomer (the structure: Supplementary Fig.1), which is the starting material of polyZI, was used as the low-molecular ZI. The relative number of living cells after freeze/thaw was used as an indicator in this study; the living cell numbers obtained after using the polyZI solutions were normalized to that obtained after using a commercial freezing medium (Culture Sure Freezing Medium, containing DMSO). The relative number of living cells after the freeze/thaw was confirmed to be low when a 10% (w/v) ZI monomer aqueous solution (aq.) was used (mNF: 0.46, BOSC: 0.10). Moreover, the relative number of living cells was unexpectedly low when a 10% (w/v) polyZI aq. was used (mNF: 0.02, BOSC: 0.2; Supplementary Fig.2). This was caused by the low osmotic pressure of the polyZI aq. Osmotic pressure is determined by the molarity of solutes. Due to the low molarity of the polyZI aq., which was caused by the high-molecular-weight, it was speculated that the osmotic pressure was lower than the isotonic pressure and the cells became swollen. The osmotic pressure of the 10% (w/v) polyZI aq. was only 5 mOsm, which is extremely hypotonic (cf., culture medium: ~300 mOsm. 10% (w/v) ZI monomer aq.: 860 mOsm; Supplementary Table1).

NaCl was added to adjust the osmotic pressure to isotonic. As expected, the addition of 0.1% to 2.0% (w/v) NaCl improved the viability of the cells (Fig.2). In particular, the highest relative number of living cells was observed at 1.0% (w/v) NaCl, which is approximately isotonic with saline and the medium (cf., 10% (w/v) polyZI/1.0% (w/v) NaCl aq.: 289 mOsm; phosphate-buffered saline (PBS): 280 mOsm; culture medium: 328 mOsm). Thus, we discovered that the optimal condition for a freezing medium would be isotonic osmolarity.

It has been reported that the polyZI, especially a carboxylate-type polyZI, exhibits poor cryopreservation performance²⁶. In the present study, we could not conclude that the difference between the previously reported PolyZI and the polyZI used in this study was significant. Nevertheless, it has been shown that the polyZI structure is a necessary condition of good cryoprotectants, but not sufficient.

To confirm that the isotonic polyZI/1.0% (w/v) NaCl aq. was not toxic to cells, the dead cell ratio was measured after a 60 min immersion in the solution at room temperature (Supplementary Fig.3). We found that the dead cell ratio in the polyZI/1.0% (w/v) NaCl aq. was comparable to that in the medium. Specifically, the dead cell ratio in the polyZI/2.0% (w/v) NaCl aq. was lower than that in 2.0% (w/v) NaCl aq., indicating that osmotic shock was assumed to be alleviated, as carboxylated ɛ-poly-L-lysine has demonstrated previously¹⁴.

Behaviour of polyZI in NaCl aqueous solutions and its relevance to the cryoprotecting effect

An understanding of the behaviour of polyZI in solutions is important for the clarification of the cryoprotecting mechanism. As mentioned previously, 10% (w/v) polyZI aq. without NaCl had an osmotic pressure of 5 mOsm, which was very low. This strongly indicates that polyZI exists in a specific state in water. For example, the osmotic pressures of the 10% (w/v) sodium polyacrylate (anionic polymer) and polyvinyl alcohol (electrically neutral polymer) aqueous solutions were 59 mOsm and 76 mOsm, respectively (Supplementary Table2). The molecular weight of polyZI, sodium polyacrylate, and polyvinyl alcohol were 86.3, 221.4, and 69.7 kg/mol, respectively, confirming that the extremely low osmotic pressure of polyZI was not due to the molar concentration. The osmotic pressure evaluated in this study was based on the depression of vapor pressure. Considering these facts, the low osmotic pressure of polyZI aq. translates to the high remaining water activity, suggesting that the polyZI does not interact strongly with water molecules. In other words, the polyZI aggregates through intramolecular charge-mediated interactions. This suggestion is supported by the 0% unfrozen water of polyZI aq. detected using differential scanning calorimetry (DSC). More specifically, approximately 10 nm- and 100 nm-long aggregates were observed by dynamic light scattering (DLS) (Fig.3), confirming that polyZI aggregates at the molecular level in water.

The aggregated polyZI is unlikely to exhibit cryoprotective effects; therefore, the addition of NaCl is assumed to loosen aggregation based on the electrostatic interaction. As expected, the addition of 1.0% (w/v) NaCl reduced the minimum size of polyZI from 10 nm to 3 nm, confirming that aggregation was loosened (Fig.3). On the other hand, the addition of NaCl also increased the maximum size of polyZI from 100 nm to 120 nm. This may have been caused by the pre-treatment of the sample; the samples were passed through filters with a pore size of 3 μm, and aggregates larger than 120 nm included in the sample without NaCl were most likely removed. However, this requires more detailed consideration. These results do not conflict with those obtained by ¹H nuclear magnetic resonance (NMR) spectroscopy. As the NaCl concentration increased, the polyZI-derived peaks became sharper. Thus, the addition of NaCl is assumed to loosen the aggregation and increase the molecular motion (Supplementary Fig.4). This loosening effect depends on the NaCl concentration (Supplementary Fig.5).

These results motivated us to investigate whether the loosening effect was caused specifically by NaCl. Different types of additives were added to the solution to achieve the same osmolarity as 1.0% (w/v) NaCl; sucrose, an electrically neutral sugar, and a ZI monomer and trimethylglycine (TMG), which are zwitterions. We found that these additives did not loosen aggregation, at least, although clear results were not obtained. We then hypothesized that free ions may play a role in the loosening effect; therefore, a large organic ion (an ionic liquid, 1-ethyl-3-methylimidazolium acetate) was used as an additive. However, no significant change in the aggregation state was observed; except for a slight decrease in the minimum size of the small aggregates observed at approximately 10 nm. These results indicate that free small ions, specifically NaCl, are important for the loosening of polyZI aggregation.

As mentioned above, the aggregated polyZI scarcely interacted with water (osmotic pressure: 5 mOsm, proportion of unfrozen water: 0%) when not mixed with NaCl. The interaction between polyZI and water and NaCl in NaCl aq. was investigated. We found that the addition of 10% (w/v) polyZI to 1.0% (w/v) NaCl aq. decreased the osmotic pressure from 327 mOsm to 289 mOsm, and the proportion of unfrozen water in the solution decreased from 11% to 4% (Supplementary Table1). These results suggest an increase in water activity and indicate that the NaCl interacting with water was stripped off by the polyZI; suggesting that the polyZI/NaCl interaction is stronger than that of water/NaCl. These results suggest that NaCl was adsorbed onto the polyZI, thereby changing its structure. This strong interaction with NaCl was unique to polyZI and was not observed in the other polymers (Supplementary Table2).

We further investigated whether the shift of the aggregation state and interaction of polyZI plays an essential role in the cryoprotecting effect. Cells were cryopreserved using solutions containing the other additives at the same osmolality as 1.0% (w/v) NaCl. The relative number of living cells was low when sucrose (neutral solute) and the ZI monomer and TMG (zwitterionic solutes) were used (Supplementary Fig.6). When an ionic liquid (organic and free ionic solute) was used, the relative number of living BOSC cells after cryopreservation was moderate, whereas that of the mNF was low. These findings are in agreement with the above-mentioned state of loosened aggregation, and it was confirmed that aggregation loosening by adding NaCl plays an essential role in the cryoprotective effect, based not only on an osmotic pressure adjustment but also on a structure/interaction shift.

Cryoprotective mechanism 1: inhibition of ice crystal formation

The primary cryoprotective mechanism of common cryoprotectants is associated with the inhibition of intracellular and extracellular ice crystal formation. However, the polyZI/NaCl aq. does not inhibit intracellular ice crystal formation, because it is isotonic and intracellular dehydration does not occur (Supplementary Fig.7).

Therefore, the inhibition of extracellular ice crystal formation was considered. The proportion of antifreeze water in the polyZI/NaCl aq. was only 4% (Supplementary Table1), which was lower than those of other low-molecular ZI solutions, DMSO solutions, and a commercial freezing medium (10–36%)¹⁷. This also supports the notion that the suppression of extracellular ice crystal formation is not a critical factor.

Ice recrystallisation inhibition is also a factor to consider^27,28. However, we found that polyZI did not inhibit ice recrystallisation (Supplementary Figs.8 and 9).

In summary, polyZI does not suppress intracellular or extracellular ice crystal formation even when freezing and thawing; thus, another factor, namely the formation of polymer matrices, is assumed to be its primary cryoprotective mechanism.

Cryoprotective mechanism 2: polymer matrices formation

In the present study, we sought to visualize the extracellular polyZI matrix. Initially, we attempted to synthesize polyZI with a fluorescent chromophore; however, the synthesis could not be achieved. Then we observed the cell behavior in detail and found that the cells in polyZI/NaCl aq. adhered to each other and formed aggregates (Fig.4). This finding indirectly supports the assumption that polyZI forms matrices around the cells. It has been reported that a low-molecular ZI strongly interacts with cell membranes through an electrostatic interaction¹⁶. Since the polyZI does not interact strongly with water, the polyZI was expanded by the addition of NaCl and then strongly interacts with the cell membrane^29,30 and forms matrices.

Molecular dynamics (MD) simulation supports the formation of polyZI matrix. ZI dimer and octamer were simulated. The ZI dimers and octamers partially covered the cell membrane (Fig.5 and Supplementary Fig.11). More specifically, we confirmed that polyZI strongly interacts with the cell membrane through ionic interactions between the amino group of dioleoylphosphatidylethanolamine (DOPE) of the cell membrane and the anion of polyZI (Supplementary Figs.11–27, specifically Supplementary Figs.12 and 13). This polyZI/cell membrane interaction is in equilibrium and repeats adsorption/desorption quickly. On the other hand, we observed that there are ZI octamers adsorbing onto the cell membrane for a long time. It was slightly embedded into the cell membrane by hydrophobic interactions between the oligomer backbone and 1,2-dioleoyl-sn-glycero-3-phospho-rac-(1-glycerol) (DOPG) (Supplementary Figs.11, 20 and 21). Since the polyZI used in the experiment is 132-mer, it is assumed to adsorb onto the cell membrane for a longer period, and a large area of the cell membrane may be covered. Unfortunately, MD simulations cannot handle the long-chain polymers, like polyZI used in the experiments, due to the complexity of the initial structure determination and the computational cost.

In addition, ZI octamer aggregation was also observed in this simulation due to interaction between oligomer backbones (Supplementary Fig.11). This agrees with the DLS results (see Fig.3).

Differential scanning calorimetry (DSC) also supports the interaction between polyZI and cell membrane (Supplementary Fig.28). The liposomes were stabilized by the addition of polyZI, which shifted the melting point of the liposomes to higher temperatures (Fig.6). In addition, the melting point peak was broadened, meaning that the range of the melting point was widened. This is consistent with the MD simulation that polyZI partially covers the liposome rather than uniformly covers the entire cell membrane. From the data shown above, polyZI matrix forms around cells. Here we would conclude that the critical cryoprotective mechanism of polyZI/NaCl aq. is considered to be the polyZI matrix, inhibiting the influx of extracellular ice nuclei.

Matsumura et al. speculated that the matrices of carboxylated ɛ-poly-L-lysine, which is a polyampholyte, trap NaCl, thereby mitigating osmotic shock¹⁴. As shown in Supplementary Fig.3, the polyZI reduced the dead cell ratio in hypertonic 2.0% (w/v) NaCl aq., which suggests that the polyZI matrix has the same function under hypertonic conditions.

The time course of osmotic pressure sensed by cells was estimated from the cell size (namely the volume change through cell dehydration) in the 10% (w/v) polyZI/1.0% (w/v) NaCl aq. to investigate the function of the matrix in detail. We found that the cell sizes were similarat room temperature, regardless of the addition of polyZI (Supplementary Fig.7). When the temperature was lowered at −1 °C/min, the cells were gently dehydrated in the polyZI/NaCl aq. but not in the NaCl aq. (Fig.7). Thus, dehydration was affected by temperature rather than soaking time (see Supplementary Fig.7). The temperature-dependent dehydration indicates that the polyZI matrix adsorbed and locally concentrated NaCl at subzero temperatures.

In addition to preventing the influx of extracellular ice nuclei, temperature-dependent dehydration may contribute to the inhibition of intracellular ice crystal formation. Here we discuss this point, referring to cryopreservation using low-molecular ZIs. Cell dehydration is the main cryoprotective mechanism of low-molecular ZIs and the cell volume in low-molecular ZI aq. becomes 40% or less at 0 °C compared to that in a culture medium at room temperature¹⁷. The relative volume at 0 °C and −13 °C in polyZI/NaCl aq. was 81% and 66%, respectively, indicating that the dehydration by polyZI/NaCl alone may not be sufficient but may help the inhibition of intracellular ice crystal formation together with the polyZI matrices. More interestingly, the volume returned to its original state when the temperature increased (Supplementary Fig.10), suggesting that adsorption/desorption was reversible and the osmotic pressure returned to normal during thawing. The cryoprotective mechanism of polyZI is not only strengthening the polymer matrix formation but also the reversible adsorption/desorption, which is partially different from that reported for polyampholytes.

The polyZI has triple functions and is suitable for cryopreservation: (1) firm polyZI matrices inhibit the inflow of extracellular ice nuclei, (2) polyZI floating in the solution absorbs NaCl to alleviate osmotic shock at room temperature, (3) the extracellular polyZI matrices concentrate NaCl to dehydrate and suppress intracellular ice crystal formation at subzero temperatures. To date, the functions (2) and (3) have not been reported in polyampholytes. More specifically, to the best of our knowledge, the function (3) has not been previously reported, while the function (2) has only been reported at different and specific conditions: after freezing and thus not at room temperature¹⁴. Although it is not clear whether these functions are specifically exhibited in polyZI only, and not all polyampholytes, the polyZI is assumed to adsorb NaCl more strongly than polyampholytes due to its charge. If the polyZI structure is a key factor, we speculate that the more specific key factor is: the polyZI is an aprotic polymer imparting constant charges, while carboxylated ɛ-poly-L-lysine is a protic polymer exhibiting an equilibrium of neutral/charged states. We believe that these unique functions based on the unique characteristics play a key role in the cryopreservation mechanism under isotonic conditions.

Cell proliferation after cryopreservation in 10% (w/v) polyZI/1.0% (w/v) NaCl aq

To investigate cell proliferation after cryopreservation, the cryopreserved cells were cultured for several days, and cells were counted. Although the number of living cells in polyZI/NaCl aq. immediately after thawing was not significantly different from that in the commercial freezing medium, the proliferation of cells cryopreserved with polyZI/NaCl aq. was significantly slower (Fig.8). We observed the cells in detail and found that cell adhesion to the dishes was less when using polyZI/NaCl aq. The polyZI matrix could play a role in the inhibition of adhesion. In this experiment, the cryoprotectant (100 μL) was washed once with a culture medium (1 mL). These results suggest that this washing process was not appropriate for the removal of the polyZI matrix.

Hence, more effective washing solvents were sought: medium, water, 0.5% (w/v) NaCl aq., 2% (w/v) NaCl aq., 3.1% (w/v) ZI monomer aq., 6.4% (w/v) sucrose aq. and 6.4% (w/v) sucrose/0.5% (w/v) NaCl aq. The 3.1% (w/v) ZI monomer aq. and the 6.4% (w/v) sucrose aq. were equimolar to the 1.0% (w/v) NaCl aq. We did not examine washing with the 1.0% (w/v) NaCl aq., because it has similar nature to the culture medium. The sucrose aq. resulted in a high living cell number after freezing, thawing, and culturing (Fig.9). Water and the NaCl aqs. were not effective, whereas the ZI monomer aq. was only effective for BOSC. Since NaCl specifically affects the polyZI aggregation states (see Fig.3), neutral or zwitterionic solutions are probably better than NaCl solutions. In this way, washing with sucrose solution made it easier to remove polyZI and cells proliferated, but the complete removal was not achieved at this primitive stage. Since polyZI is a new CPA, we believe that with further study we will find a way to remove it.

We also assumed that since the polyZI was attached to cells, and therefore, trypsinization may detach it from the cells. However, no positive effect was observed (see Fig.9).

Cryopreservation of freeze-vulnerable cells with the isotonic polyZI/NaCl aq

Freeze-vulnerable cells (OVMANA and K562) were subjected to cryopreservation using the polyZI/NaCl aq. When K562 cells were cryopreserved using the polyZI/NaCl aq., the relative number of living cells immediately after thawing tends to be higher than that when using a commercial medium, suggesting the importance of isotonic cryopreservation (Fig.10). In contrast, the number of living K562 cells cultured for 5 days after washing with 6.4% (w/v) sucrose aq. was approximately the same as that when a commercial medium was used. Similar results were obtained for OVMANA cells (Fig.10). The polyZI/NaCl aq. was effective for freeze-vulnerable cells presumably due to avoiding the osmotic shock but the removal of the polyZI matrices is still challenging.

Because there is currently no way to completely remove the polyZI matrices, the positive effects of the isotonic freezing medium and the negative effects of adhesion inhibition may counteract each other. Nevertheless, commercial freezing mediums have been optimized for decades, and the polyZI freezing medium still has significant room; however, we do believe that it has substantial potential because the simple mixture composed of only polyZI/NaCl/water provides the same level of cryoprotective performance as the commercial medium. In this study, we designed an isotonic cryopreservation agent that does not induce osmotic stress in cells; however, further studies should be conducted to improve the washing method.