The role of antifreeze glycopeptides (AFGP) and polyvinyl alcohol/polyglycerol (X/Z-1000) cocktails as ice modulators during partial freezing of rat livers

The current liver organ shortage has pushed the field of liver transplantation to develop new methods to prolong the preservation time of livers from the current clinical standard of static cold storage. Our approach, termed partial freezing, aims to induce a thermodynamically stable frozen state at deeper storage temperatures (−10°C to −15°C) than can be achieved with supercooling, while simultaneously maintaining a sufficient unfrozen fraction to limit dehydration and ice damage. This research first demonstrated that partially frozen glycerol treated rat livers were functionally similar after thawing from either −10 or −15°C with respect to subnormothermic machine perfusion metrics and histology. Next, we assessed the effect of adding either of two ice modulators, antifreeze glycoprotein (AFGP) and a polyvinyl alcohol/polyglycerol combination (X/Z-1000), on the viability and structural integrity of partially frozen rat livers compared to glycerol-only control livers. Results showed that AFGP livers had high levels of ATP and the least edema but suffered from significant endothelial cell damage. X/Z-1000 livers had the highest levels of ATP and energy charge (EC) but also demonstrated endothelial damage and post-thaw edema. Glycerol-only control livers exhibited the least DNA damage on Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining but also had the lowest levels of ATP and EC. Further research is necessary to optimize the ideal ice modulator cocktail for our partial-freezing protocol. Modifications to cryoprotective agent (CPA) combinations, as well as improvements to machine perfusion CPA loading and unloading, can help improve the viability of these partially frozen organs.


143
AFGP has been shown to inhibit both ice recrystallization and ice growth below T M (the 144 thermodynamic freezing point). These glycopeptides inhibit ice growth by attaching to multiple 145 faces of ice crystals [20][21][22][23]. AFGPs have also been shown to raise the homogenous ice 146 nucleation temperature (T H ) by organizing water into a more ice-like state [24]. However, since 147 the temperature range in our partial freezing protocol is well above T H [25], the issues at hand 148 involve the role of AFGP in ice shaping and ice recrystallization inhibition. Although AFGP can 149 shape ice into damaging spicules [26], this effect may be outweighed under our storage 150 conditions by the ice growth and recrystallization inhibitory effects of AFGP. X-1000 is a 2 151 kilodalton (kDa) polyvinyl alcohol [27] that contains 20% vinyl acetate, which improves the 152 solubility and ice-inhibiting effects of X-1000 presumably by preventing self-association 153 between X-1000 chains. Polyvinyl alcohol is known to inhibit ice recrystallization [28,29]. Z-154 1000 is a polyglycerol that inhibits heterogeneous ice nucleation [30], and together X/Z-1000 has 155 been shown to protect rat hearts during supercooling [31-33] and is functional from 0°C to 156 temperatures below -120°C [16].

157
In the present study, whole rat livers were frozen for up to 5 days at high subzero 158 temperatures (-10°C to -15°C) by combining glycerol and ice nucleating agents (INAs) with the 159 use of subnormothermic machine perfusion (SNMP) at 21°C. Further, two ice modulators, 160 antifreeze glycoprotein (AFGP) and a polyvinyl alcohol/polyglycerol combination (X/Z-1000), 161 were tested. Livers frozen with the inclusion of either AFGP or X/Z-1000 were compared to the 162 control group (with glycerol as the main CPA) with primary outcomes being perfusion metrics, 163 ATP, energy charge (EC), weight gain, and histology. We call this protocol "partial freezing" 164 since it induces a thermodynamically stable frozen state at deeper storage temperatures (as low 165 as -15°C) than can be achieved with supercooling, while simultaneously maintaining a sufficient 166 unfrozen fraction to limit dehydration and ice damage. 178 (5) partial freezing of rat liver, (6) thawing of rat liver, (7) unloading of CPAs during HMP, and 179 (8) functional recovery of frozen rat livers during SNMP.

180
Within this protocol, we first compared partially frozen livers at -10°C (n=4 livers) and -181 15°C (n=9 livers) with 12% glycerol. Upon finding minimal differences between these two 182 groups, we combined them as a control and compared them to livers partially frozen with 0.5 183 mg/ml (0.05% w/v) of AFGP (n = 4 livers) or 0.1% X-1000/0.2% Z-1000 (total, 0.3% w/v; n = 4 184 livers) ice modulating agents added to the preservation solution. After freezing, liver viability on 185 SNMP was compared between the 12% glycerol control group and the two ice modulated 186 groups. 214 glycerol, 4000 U/l of heparin, 24 mg/l of dexamethasone, 25 mg/ml of hydrocortisone, 40,000 215 ug/l of penicillin, 40,000 U/l of streptomycin, and sodium bicarbonate as needed to maintain a 216 physiological pH) (figure 1, step 3; Table S1). HMP was continued for 1 hour to ensure 217 complete equilibration of solution throughout the liver parenchyma. In the case of livers frozen with glycerol only, the chiller temperature was pre-cooled to 232 either -10°C or -15°C, and the liver was stored at one of these temperatures for 1 to 5 days.
233 Livers frozen with either of the two ice modulation candidates were stored at -15°C.

234
After partial freezing, livers were thawed

262
Rat liver weight was measured directly after procurement, prior to freezing, after 263 thawing, and after viability testing. Weight gain was calculated as the percentage increase at the 264 end of recovery compared to the liver weight after procurement. Vascular resistance was 265 calculated by dividing the perfusion pressure in the PV by the flow rate per gram of liver using 266 the weight of the liver after procurement as the reference standard weight. Oxygen consumption 267 rates were calculated as (pO 2 in -pO 2 out )*F/W where pO 2 in and pO 2 out were the oxygen contents 268 per ml of inflowing and outflowing perfusate, respectively, and the difference between them 269 multiplied by the perfusion rate (F, in ml/min) provided the total oxygen uptake per minute. This     513 The authors declare that the data supporting the findings of this study are available within 514 the paper and its supplementary information files. Any additional data, if needed, will be 515 provided upon request.