Nanowarming of vitrified pancreatic islets as a cryopreservation technology for transplantation

Abstract Biobanking of pancreatic islets for transplantation could solve the shortage of donors, and cryopreservation of vitrified islets is a possible approach. However, a technological barrier is rewarming of large volumes both uniformly and rapidly to prevent ice formation due to devitrification. Here, we describe successful recovery of islets from the vitrified state using a volumetric rewarming technology called “nanowarming,” which is inductive heating of magnetic nanoparticles under an alternating magnetic field. Convective warming using a 37°C water bath as the gold standard for rewarming of vitrified samples resulted in a decrease in the viability of mouse islets in large volumes (>1 ml) owing to devitrification caused by slow warming. Nanowarming showed uniform and rapid rewarming of vitrified islets in large volumes. The viability of nanowarmed islets was significantly improved and islets transplanted into streptozotocin‐induced diabetic mice successfully lowered serum glucose. The results suggest that nanowarming will lead to a breakthrough in biobanking of islets for transplantation.


| INTRODUCTION
Transplantation of organs and tissues saves millions of lives and improves the quality of life each year globally. However, the global shortage of organs and tissues for transplantation has been recognized as one of the greatest crises of biomedicine. 1 A major limitation of transplantation is the ischemic injury in organs and tissues during preservation. The timeframe in which organs and tissues remain viable and functional under cold storage (generally on ice) is considerably short. 2 Pancreases are used for transplants up to only 8-12 h, 3 which limits transport of pancreases from donors to potential recipients at geographically remote locations.
A promising approach to biobanking of organs and tissues is icefree cryopreservation by vitrification using rapid cooling that exceeds the critical cooling rate (CCR; the rate necessary to avoid ice formation during cooling) to an extremely low temperature (≤130 C) below the glass transition. 4 Cryopreservation by vitrification was originally developed to preserve cells and small/thin tissues such as embryos and ovaries in small volumes (generally hundreds of microliters) of cryoprotective agent (CPA). 5,6 While the CCR is relatively easy to achieve by direct immersion in liquid nitrogen (À196 C), the main barrier of cryopreservation is to achieve rapid and uniform rewarming that exceeds the critical warming rate (CWR; the rate necessary to avoid ice formation due to devitrification during rewarming), which is an order of magnitude higher than the CCR. 4 The conventional method of rewarming is convective warming by immersing vitrified CPA in a water bath at 37 C until temperature at the center of CPA reaches 0 C, to avoid cellular or tissue damage caused by overheating and CPA toxicity, which leads to a low warming rate at the center of a large volume of CPA with large diameters. Additionally, high-temperature gradients between the center and edge of the large volume of CPA cause thermal stress strain that leads to cracking during rewarming. 7 Nanowarming is a cryopreservation technology for volumetric warming of vitrified CPAs. [8][9][10][11] Superparamagnetic nanoparticles generate heat under an alternating magnetic field by Brownian and Néelian relaxation. 12,13 In general, magnetite (Fe 3 O 4 ) with a size of around 10 nm and magnetic field applicators that generate an alternating magnetic field with an order of tens of kA m À1 at several 100 kHz have been used. 8 15 In 2017, the second applied nanowarming to vitrified fibroblasts, vessel segments and heart valve tissues. 8 Our group reported that vitrified human-induced pluripotent stem cells were successfully rewarmed by nanowarming. 9 In these cases, cells and tissues were vitrified by immersion in an mCPA. More recently, a silica-coated magnetic nanoparticle was reported to have sufficient dispersion stability for perfusion of rat kidneys. 11 Chiu-Lam et al.
developed magnetic nanoparticles coated with a dense, covalently grafted brush of polyethylene glycol (PEG) for perfusion of rat hearts, and reported whole-heart cryopreservation and nanowarming, 10 which demonstrated the proof-of-concept of nanowarming for biobanking and transplantation. Although these reports showed recovery of cell viability and tissue histology, the functionality of the nanowarmed tissues and organs, which is the most important factor for transplants, remains to be demonstrated. Islet transplantation has been an established therapy as a potential curative treatment for patients with insulin-dependent diabetes and severe hypoglycemia. 16 In addition to the shortage of donors, a major problem in the clinical transplantation of islets is the low yield of islets that can be obtained from one donor pancreas. Islets must be collected from different donors and preserved until sufficient islets are obtained.
To overcome these issues, cryopreservation of islets will offer the potential to biobank islets for transplantation on demand. 17 19 and human 20 islets in 1 ml CPAs were successfully rewarmed by convective warming. The mouse islets showed insulin secretion and the capability to lower blood glucose to normal levels after transplantation into streptozotocin (STZ)-induced diabetic mice. 19 In the current study, we apply nanowarming as a volumetric rewarming technology to islets cryopreservation for biobanking and clinical transplantation. Specifically, cell viability, glucose-stimulated insulin release, and the capability to lower blood glucose levels after transplantation were investigated to demonstrate the feasibility of nanowarming for clinical islet transplantation.

| Volume limitation of convective warming in rewarming of vitrified islets
The commonly used CPA solution VS55, which has a CCR of 2.5 C min À1 , a CWR of 50 C min À1 , and a glass transition temperature of À123 C, 21 was used in this study. Variously sized glass vials (Table S1) that contained VS55 (1-30 ml) were submersed in liquid nitrogen and the temperature at the center of the vials was monitored over time (Figure 1a). Temperature-time plots showed smooth curved lines during cooling for 1 and 8 ml, whereas inflection points (e.g., at around À50 C), which might suggest ice formation, were observed for 20 and 30 ml. However, all samples had a transparent glassy appearance at around À196 C, which indicated that VS55 was successfully vitrified at volumes smaller than 30 ml (16 mm radius). Table 1 shows the cooling rates and all samples (1,8,20, and 30 ml) achieved a CCR of 2.5 C min À1 by cooling in liquid nitrogen.
During the rewarming process (Figure 1b), convective warming resulted in devitrification of VS55, which was evidenced by ice formation in the sample larger than 20 ml (14 mm radius) by direct visual inspection (inset photograph of Figure 1b). During rewarming of 30-ml samples, two typical changes of temperature in the time course were observed around À100 to À80 C and À50 to À30 C, which indicated the recrystallization temperature (À82 C) and the melting temperature (À38 C) of VS55. 21 Warming rates of samples with 1-, 8-, 20-, and 30-ml volumes are shown in Table 1. Only the warming rate in 1 ml (5 mm radius) well exceeded the CWR, which indicated that convective warming using a water bath at 37 C was too slow to achieve the CWR in large-volume CPAs (≥several milliliters).
Next, we measured the cell viability of vitrified islets after convective warming (Figure 1c) using the Sytox green/Hoechst 33342 assay. 9 The cell viability of fresh islets isolated from mice by the collagenase digestion method was 77.7% ± 10.8%. For islet cryopreservation, 0.2 ml CPA in a 2-ml polypropylene cryovial was used as a control (gold standard). The cell viability after rewarming by convective warming was 60.9% ± 7.2%, which was approximately 80% of the fresh islets. In the case of a small volume with 1 ml (warming rate, 163.2 ± 5.7 C min À1 [-

| Characterization of warming rates by mCPAmediated nanowarming
Using mCPA, nanowarming rates were evaluated in an alternating magnetic field ( Figure 2). Magnetite nanoparticles of 10 nm in size, which were assumed to be superparamagnetic, 12,13 were well dispersed in VS55, and 1 ml vitrified mCPA (magnetite concentrations: 0, 1, 3, 5, and 10 mg ml À1 ) with liquid nitrogen was exposed to an alternating magnetic field (power outputs: 0, 2.5, 5, and 10 kW; frequencies: 108, 163, and 208 kHz). Temperature-time plots showed an inflection point at around À45 C during rewarming, especially under control conditions (magnetite concentrations: 0 mg ml À1 or power outputs: 0 kW), which may suggest ice formation. The warming rates increased with the magnetite concentration (Figure 2a During rewarming by convective warming, cracks were observed in 8 and 20 ml CPA ( Figure S1) because their ΔT max was well beyond the stress-to-fracture temperature limit of VS55 glass (38 C). 11 However, nanowarming of mCPA showed no obvious temperature gradient between the center and edge of samples ( Figure 3b) or obvious cracks in the samples. Importantly, the nanowarming rates were independent of the sample volume ( Figure 3b). These results indicate that nanowarming is a scalable technique that enables uniform and rapid rewarming. Figure 4a shows the scheme of nanowarming islets. Pancreatic islets were isolated from mice, precultured for 2 days, 22 vitrified in mCPA by submersion in liquid nitrogen, and nanowarmed by an alternating magnetic field for rewarming. To load and remove VS55, a slow stepwise dilution protocol 23 at a low temperature (on ice) was used to minimize osmotic stress in the islets ( Figure S2). Nanowarmed islets were applied to a cell strainer to remove magnetite nanoparticles prior to the slow stepwise dilution protocol. After the filtration and dilution, magnetite nanoparticles were not detected in nanowarmed islets.  The cooling and warming rates at the center of vials were calculated by the slopes between À38 C (the melting temperature of VS55) and À123 C (the glass transition temperature of VS55). Data are the mean and SD of three independent experiments.

| Nanowarming of pancreatic islets
Also, we counted the number of nanowarmed islets after the filtration and dilution, and 81% ± 14% (n = 3) of islets were recovered from approximately 800 islets in 20 ml CPA. The morphology of islets after nanowarming was similar to that of freshly isolated islets, whereas cell-cell gaps were clearly observed in islets after convective warming ( Figure 4b). Moreover, the Sytox green/Hoechst 33342 assay revealed that nanowarmed islets were highly viable, while dead cells were observed in vitrified islets after rewarming by convective warming ( Figure S3). Also, nanowarmed islets showed both insulin and glucagon immunofluorescence staining similar to freshly isolated islets ( Figure S4). Nanowarming of islets in 20 ml mCPA markedly improved cell viability (83.3% ± 21.0%) compared with convective warming (1.6% ± 2.7%), which was comparable to the 0.2-ml sample (gold standard) (78.3% ± 9.3%) ( Figure 4c). As a control, we also conducted a cryopreservation of islets with slow cooling (1 C min À1 ) in 20 ml RPMI-1640 medium supplemented with 10% DMSO. After rewarming by convective warming, cell viability of slow cooling was 8.8% ± 4.7%, which was significantly lower than that of fast cooling + nanowarming ( Figure S5). As a mitochondrial marker, we examined ATP levels in islets after rewarming ( Figure S6) and found that nanowarming maintained ATP content in islets better than convective warming (0.94 ± 0.14 vs. 0.06 ± 0.01, Figure  nonfasting blood glucose of ≥300 mg dl À1 were established by STZ administration and untreated diabetic mice showed hyperglycemia for 30 days. Transplants of 400 freshly isolated islets lowered the blood glucose level to the normal level (≤200 mg dl À1 ). The nanowarmed islets (400 picked islets after nanowarming) successfully lowered blood glucose levels, whereas islets rewarmed by convective warming (400 picked islets after convective warming) did not achieve euglycemia. Figure 5b shows the blood glucose levels at 30 days after islet transplantation. The blood glucose level of mice treated with nanowarmed islets was comparable to that of mice treated with freshly isolated islets and was significantly lower than that of mice treated with islets rewarmed by convective warming. At 4 weeks after transplantation, an intraperitoneal glucose tolerance test (IPGTT) was performed to control blood glucose levels (Figure 5a,c), which was likely to be caused by damage due to cryopreservation using VS55. In the present study, 400 of nanowarmed islets were needed to lower the blood glucose levels in diabetic mice (Figure 5a,b), whereas researchers reported that freshly isolated 250 islets could lower the blood glucose levels. 24 The relative islet potency following cryopreservation was thus estimated to be approximately 0.6 (250/400 for freshly isolated islets/nanowarmed islets), which was consistent with that based on the SI (2.4/4.1 for freshly isolated islets/nanowarmed islets) in GSIS ( Figure 4e). That is, 1.6 of nanowarmed islets would be needed to replace one freshly isolated islet. Recently, Marquez-Curtis et al.
reported a single-cell transcriptome study to identify differentially expressed genes between fresh and cryopreserved human islet cells. 27 Further study including transcriptome analysis is needed to elucidate the mechanism of reduced functionality in islets nanowarmed with VS55.
In addition to intracellular ice formation that leads to cell death, intercellular ice formation impairs the integrity of multicellular tissues, and ice-free cryopreservation by vitrification is crucial for longterm biobanking of tissues and organs for transplantation. A mixture of cryoprotectants instead of one cryoprotectant in a slow-freezing method solidifies at extremely low temperatures (glass transition temperature of À120 to À130 C) without ice formation. As a model  Islets isolated from the mouse pancreas were precultured for 2 days and then added to mCPA (5 mg magnetite ml À1 VS55) in glass vials that were directly immersed in liquid nitrogen. Vitrified samples were then rewarmed by alternating magnetic field exposure of 10 kW at 208 kHz. (b) Morphology of islets. Bright-field micrographs (top) and cross-section (H&E staining) (bottom) of the islet. Left, freshly isolated islets; center, vitrified islets after rewarming by convective warming; right, vitrified islets after rewarming by nanowarming. Scale bars, 100 μm. (c) Effects of nanowarming on cell viability of islets. After rewarming vitrified islets (5 islets ml À1 ), cell viability was assayed using a cell viability imaging kit based on Hoechst 33342 for live cells and SYTOX green nucleic acid stain for dead cells. White columns, convective warming (water bath at 37 C); black column, nanowarming. Data are expressed as the mean ± SD of three independent experiments. Welch's t-test was performed to compare the difference between the two groups. *p < 0.05. (d) Glucose challenge of islets. Mouse islets (20 islets per well) were subsequently incubated for 1 h in HKRB that contained 3 mM (white columns) and 20 mM (black columns) glucose. Insulin secretion (ng/islet/h) is expressed as the mean ± SD of triplicates. (e) Stimulation index. Data are expressed as the mean ± SD of three independent experiments. chemically modified magnetite nanoparticles in CPA, such as interactions among nanoparticles, CPA, and target tissues/organs, remain to be investigated to develop new mCPAs specific for nanowarming.
Vitrification allows cryopreservation of tissues and organs indefinitely in theory, which may overcome the geographical problem of transplantation. In the present study, similar to the previous study on cryopreservation of human induced pluripotent stem cells, 9 we conducted a trial in which mouse islets (20 ml mCPA) were vitrified in Kyoto University (Kyoto, Japan) and transported to Nagoya University (Aichi, Japan) (>100 km) at an ultralow temperature (≤À150 C) using a dry shipping container (cat# DR-2DS, Cryo One, Osaka, Japan). As a result, the cell viability of islets vitrified in Kyoto University and nanowarmed in Nagoya University was comparable to that of islets both vitrified and nanowarmed in Nagoya University. Vitrification is thus robust and stable, and the major factors required for successful vitrification of tissues and organs are the cooling rate and storage temperature. For example, VS55 needs the cooling rate to exceed a CCR of 2.5 C min À1 and the storage temperature to be below the glass transition temperature of À123 C.
While vitrification has been practically used for cryopreservation of embryonic stem cells 36 and embryos 5 in CPA with small volumes, rewarming of large-volume samples has remained challenging, which is related to the dual needs for the warming rate and uniformity. For example, warming rates faster than a CWR of 50 C min À1 and thermal  gradients of ≤38 C are required for successful rewarming of VS55 to avoid damage caused by ice formation and cracking, respectively. 11,21 To achieve successful rewarming, nanowarming is advantageous over the conventional rewarming technique by convective warming. The nanowarming rates depended on the magnetic nanoparticle concentration (Figure 2a), power output (Figure 2b), and frequency ( Figure 2c). The quantity of heat generation by magnetic nanoparticles is proportional to the frequency and the square of the magnetic field strength. 13 The mechanisms of magnetic heating of superparamagnetic nanoparticles are Néel relaxation and Brownian relaxation. Néel relaxation is derived from reorientation of the magnetic moment in the same direction as the applied magnetic field with each field oscillation. 37 Brownian relaxation is caused by the friction that arises from the rotation of the particle in the carrier liquid. 38 Because CPA is highly viscous, the dominant mechanism of heat dissipation may be Néel relaxation. Further study is needed to elucidate the mechanism of heat dissipation of magnetic nanoparticles in CPA.
In order to minimize toxicity from CPA exposure and osmotic shock, CPA needs to be carefully loaded and removed from biologic systems. In the present study, a slow stepwise dilution protocol 23 was used to mitigate osmotic damage in the islets ( Figure S2). By combining with filtration of nanowarmed islets, magnetite nanoparticles were not detected in nanowarmed islets after the slow stepwise dilution.
Researchers have reported image-guided characterization of mCPA distribution in organs/tissues by microcomputed tomography imaging, 8,11 magnetic resonance imaging, 8 and magnetic particle imaging. 10 These imaging technologies will be a potent tool to better guide the loading and unloading process.
In the present study, we succeeded in nanowarming 20 ml mCPA that contained 800 islets. Clinically, in the Edmonton protocol, at least 10,000 islet equivalent (IEQ) per 1 kg of body weight is required for a transplant. A human pancreas has $1 Â 10 6 islets, but successful islet isolation is reported to be 1 Â 10 5 to 3.5 Â 10 5 IEQ per donor, 39 which indicates that two or three donors are required for a transplant.
For example, if a 60 kg patient needs 6 Â 10 5 IEQ, a system capable of cryopreserving 2 Â 10 5 IEQ from three donors each is required.
Taylor and Baicu proposed a vitrification protocol in which the islet concentration in CPA is 500 IEQ ml À1 . 17 Therefore, the CPA volume for 2 Â 10 5 IEQ is estimated to be several hundred milliliters and scale-up of the heating coil is necessary for clinical application. In the present study, a coil with a diameter of 70 mm and a height of 80 mm was used for nanowarming. We previously prototyped a coil with a diameter of 300 mm for clinical cancer hyperthermia. However, in cryopreservation, samples with a large diameter will not vitrify because of the slow cooling rate even with the use of liquid nitrogen.
For VS55, a rough estimation suggests that samples with a diameter of <50 mm can achieve the CCR (2.5 C min À1 ). However, during rewarming, we demonstrated that nanowarming resulted in uniform warming and the warming rates were independent of the sample volume (Figure 3b). For scale-up, a long (in height) cylindrical coil has been considered to be a possible design for a clinical applicator of an alternating magnetic field.

| Study design
The aim of this study was to evaluate the efficacy of nanowarming to rewarm islets after vitrification compared with conventional rewarming by convective warming. We first conducted experiments using the conventional method by rapid cooling in liquid nitrogen and rewarming by convective warming in a 37 C water bath to demonstrate the size limitation caused by CWR in rewarming using 1-30 ml CPAs. The glass vials used in this study are shown in Table S1. The tube was then centrifuged at 440 g for 10 min without brake.

| Nanowarming
Nanowarming of vitrified samples was performed as reported previously. 9 Briefly, an alternating magnetic field (31.9 kA m À1 ) was cre-  25 The serum glucose of the mice was measured every 2 or 3 days throughout the 4-week observation period.
An IPGTT was performed after overnight fasting at 4 weeks after transplantation. The mice were fasted for 16 h and then a glucose solution (1 g in 10 ml kg À1 body weight) was injected intraperitoneally. Blood glucose levels were measured before (À15 min) and every 15 min for 120 min after injection. The AUC of IPGTT was also calculated.

| Statistical analysis
Welch's t-test was used to compare differences between two groups. Comparisons between multiple groups were performed by Welch's t-test with Bonferroni's correction. Differences were considered statistically significant at p < 0.05.

| CONCLUSIONS
In the present study, we have demonstrated that nanowarmed islets function to release insulin in response to glucose stimulation and successfully lower the blood glucose levels of STZ-induced diabetic mice after transplantation. Before entering clinical application, a preclinical study using human islets from deceased donors should be conducted. Jutte et al.
reported that human islets appeared to withstand vitrification very well compared with mouse islets, 20 which suggests that the results obtained from mouse islets in the present study are promising. Taken together, our results suggest that ice-free cryopreservation by nanowarming will lead to breakthroughs in biobanking of islets for transplantation.

CONFLICT OF INTEREST
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.