Natural resistance to liver cold ischemia-reperfusion injury associated with the hibernation phenotype

Susanne L. Lindell, Shawna L. Klahn, Timothy M. Piazza, Martin J. Mangino, Jose R. Torrealba, James H. Southard, Hannah V. Carey


The success of liver grafts is currently limited by the length of time organs are cold preserved before transplant. Novel insights to improve viability of cold-stored organs may emerge from studies with animals that naturally experience low body temperatures (Tb) for extended periods. In this study, we tested whether livers from hibernating ground squirrels tolerate cold ischemia-warm reperfusion (cold I/R) for longer times and with better quality than livers from rats or summer squirrels. Hibernators were used when torpid (Tb < 10°C) or aroused (Tb = 37°C). Livers were stored at 4°C in University of Wisconsin solution for 0–72 h and then reperfused with 37°C buffer in vitro. Lactate dehydrogenase (LDH) release after 60 min was increased 37-fold in rat livers after 72 h cold I/R but only 10-fold in summer livers and approximately three- to sixfold in torpid and aroused hibernator livers, despite twofold higher total LDH content in livers from hibernators compared with rats or summer squirrels. Reperfusion for up to 240 min had the least effect on LDH release in livers from hibernators and the greatest effect in rats. Compared with rats or summer squirrels, livers from hibernators after 72 h cold I/R showed better maintenance of mitochondrial respiration, bile production, and sinusoidal lining cell viability, as well as lower vascular resistance and Kupffer cell phagocytosis. These results demonstrate that the hibernation phenotype in ground squirrels confers superior resistance to liver cold I/R injury compared with rats and summer squirrels. Because hibernation-induced protection is not dependent on animals being in the torpid state, the mechanisms responsible for this effect may provide new strategies for liver preservation in humans.

  • transplantation
  • cold storage
  • Kupffer cells
  • sinusoidal lining cells

despite major advances in hypothermic preservation of organs before transplant, problems associated with extended cold ischemia-reperfusion (cold I/R) injury still limit the optimal use of transplantation as a therapeutic intervention for organ failure. For example, human livers can be successfully stored at 4°C for up to 12 h in University of Wisconsin (UW) solution, but graft failure is greatly increased after extended storage periods (14, 33, 34). In the rat liver transplant model, when livers are cold stored in UW for up to 24 h, 100% of the recipients survive (42); in contrast, without nutritional manipulation, none survive after receiving livers stored for 48 h (39). The damaging effects of cold I/R are evident on organ reperfusion with warmed, oxygenated blood, with damage increasing with the length of the cold ischemia period (10, 35, 40).

A variety of pharmacological strategies has been explored to improve graft function after cold I/R to extend cold storage times and provide more flexibility in organ distribution before transplant procedures (for reviews see Refs. 2, 26, 37). An alternative approach that has received little attention is to study animals that naturally experience prolonged periods at low body temperature (Tb), such as hibernators. Mammalian hibernation comprises a suite of physiological and morphological changes that facilitate survival during seasonal periods of low ambient temperatures and relative food scarcity (5). Hibernating ground squirrels maintain Tb similar to or lower than those used for human organ preservation (<10°C), and they do so for as long as 3–4 wk in extended bouts of torpor during the winter months (30). Torpor bouts are interrupted periodically during the hibernation season by the return of metabolism and Tb to near-normal levels. During arousals, blood rapidly reperfuses tissue beds, with proportionately more flow to anterior regions (heart, lungs, brain, brown adipose tissue) relative to peripheral and posterior regions including the splanchnic organs (4, 44). Because hibernator organs, such as the liver, experience minimal tissue damage during hibernation, we asked whether livers of a hibernating species, the 13-lined ground squirrel, display superior resistance to the damaging effects of cold I/R compared with a nonhibernator (rat) and whether livers are more resistance to cold I/R injury when harvested from animals during hibernation compared with the summer active season.

We used the isolated perfused liver (IPL) model to assess viability of control or cold-stored livers. We hypothesized that ground squirrel livers, whether from summer-active or hibernating squirrels, would exhibit less damage after extended cold I/R compared with rats, and furthermore that livers from torpid ground squirrels would be more resistant to cold I/R damage than would livers from summer squirrels. We also studied a smaller group of ground squirrels while they were euthermic (Tb = ∼37°C) during the hibernation season. Those experiments allowed us to determine whether organ temperature at harvest was responsible for resistance to cold I/R injury in hibernators (19).


Experimental animals.

The UW Institutional Animal Care and Use Committee approved all procedures for animal maintenance and euthanasia. Inbred male Brown-Norway rats (100–200 g) had free access to food (standard rat chow) and water before harvesting of livers. Thirteen-lined ground squirrels (Spermophilus tridecemlineatus; 100–200 g) of both sexes were collected in the vicinity of Madison, Wisconsin. Active squirrels had free access to water and food (Purina Rodent Chow no. 5001, supplemented with sunflower seeds) and were held in rooms maintained at 22°C and a 12:12-h light-dark cycle. Animals were given ivermectin shortly after capture to eliminate nematode parasites. Summer active squirrels had been in captivity for at least 6 wk and were fed ad libitum until they were used from June through early August. In September-October, squirrels were placed in a temperature-controlled room maintained at 5°C in constant darkness. Food and water were removed from a squirrel's cage after it began regular bouts of torpor. Livers were harvested from squirrels in the summer (Tb = ∼37°C) or after at least 4 wk of hibernation, when in deep torpor (Tb = 5–8°C) or when fully aroused (Tb = 37°C).

Liver harvest/preservation.

Anesthesia was induced with an isoflurane/O2 mixture using a vaporizer (Surgivet/Anesco). Tb was monitored with a rectal probe. After laparotomy, animals were heparinized (1,000 U/kg iv) and the common bile duct was cannulated with polyethylene tubing (PE-10) for bile collection. For ground squirrels, the gall bladder was ligated and removed. The portal vein was cannulated, and the livers were flushed with 15 ml 4°C UW solution, removed, and either used immediately (0 h preservation “control”) or cold stored in UW solution for 24–96 h (27, 28).


Isolated liver perfusion was performed by recirculating 50 ml of Krebs-Henseleit buffer (KHB) through the portal vein, with a roller pump set to deliver a constant flow of 2.5 ml·g−1·min. This flow rate was sufficient to support normal metabolism. A pressure transducer was connected close to the inlet of the portal vein, and flow and pressure were recorded on a Powerlab/4SP (ADInstruments, Castlehill, Australia) using Chart 4 software to calculate vascular resistance (VR). The perfusate (KHB) was equilibrated with O2-CO2 mixture (95:5%), which resulted in an arterial Po2 of 400–500 mmHg and a venous Po2 of ∼200 mmHg. The temperature was kept constant at 37°C with a heat exchanger. After the specified storage periods, the livers were weighed (to adjust the flow rate) and flushed with KHB (1 ml/g) at room temperature, before reperfusion (IPL). Livers were reperfused for 60 or 240 min. Perfusate samples were collected every 30 min for blood gas analyses and release of lactate dehydrogenase (LDH). The LDH concentration was measured spectrophotometrically using a commercial kit (340-LD; Sigma, St Louis, MO) and expressed per gram wet liver mass (U·l−1·g) (27, 28). Bile was collected at 30-min increments, weighed, and expressed per gram wet liver mass (mg/g).

Hepatocyte isolation and measurement of total LDH.

Hepatocytes were isolated from fresh (no storage) livers from rat, torpid, or summer squirrels. Briefly, the livers were digested with collagenase (Type 2, 250 U/ml, Worthington Biochemical, Freehold NJ), and hepatocytes were isolated by centrifugation and snap-frozen in liquid nitrogen (22). Hepatocytes were thawed at room temperature, sonicated in distilled water, and LDH concentration was measured, as described in IPL. Protein concentration was analyzed using a biuret reaction (17), and LDH was expressed as units per gram protein.

Trypan blue uptake.

After various preservation times (0–72 h) and 60 min of reperfusion (IPL), some livers were infused with KHB containing 200 μM trypan blue for 7 min and fixed in 2% glutaraldehyde /2% paraformaldehyde buffer (1, 23). Liver sections were stained with eosin. Trypan blue was identified in the nuclei of damaged sinusoidal lining cells (SLC) and counted manually (60×) in eight different fields of eight different sections (a total of 64 fields/liver). Values were expressed as mean number of SLC per field (SLC/field) (1, 36).

Colloidal carbon uptake.

Uptake of colloidal carbon (CC) was used as a measure of Kupffer cell phagocytosis. CC (black India ink, Pelikan 17) was prepared as described previously (27). The density of CC was analyzed by overnight desiccation (mg/ml). After various storage times (0–72 h), the livers were reperfused in oxygenated KHB at 37°C in a nonrecirculating system (initial equilibration period). After 30 min, the reperfusion medium was switched to KHB containing CC (yielding A623 of ∼2.0 units) to assess Kupffer cell phagocytotic activity, and the livers were reperfused for an additional 30 min. During this time, perfusate samples were collected every 5 min to determine uptake of CC by comparing the absorbance at 623 nm of the perfusate going into the liver to the perfusate leaving the liver. The rate of CC uptake was expressed as micrograms per minute times grams liver tissue (11, 27).

Mitochondrial respiration.

Mitochondria were isolated from 2.5 g liver tissue after cold I/R by standard techniques using differential centrifugation (25, 43). Protein concentration of the final preparation was measured using the Biuret method (17). The rate of oxygen consumption at 37°C was measured using a modification of techniques described in Ref. 31. Briefly, after 5 mM succinate was added, state 3 respiration was initiated by adding ADP to a final concentration of 250 μM. State 4 respiration was calculated by averaging the steady-state rates of oxygen consumption measured before ADP addition and after its depletion. The ratio of state 3 to state 4 respiration represented the respiratory control ratio (RCR), which was used as a measure of the functional integrity of mitochondria isolated from each group.


Data are expressed as means ± SE and are presented as original values in the figures. Repeated-measures ANOVA was used to analyze the data in Figs. 2 and 4. The data in Fig. 2 were first log-transformed because of unequal variances among the groups. All other data sets were analyzed by one-way ANOVA or t-tests. If an ANOVA F-value was significant, then post hoc comparisons among groups were analyzed using the least-square difference method (SAS GLM). A P value ≤0.05 was considered statistically significant.


LDH release and bile production.

The release of LDH into the perfusate buffer over a 60-min period was used as a marker of hepatocellular damage after cold I/R. LDH release from rat livers subject to 72 h cold I/R was 37-fold higher than in control (0 h) livers (Fig. 1A). In summer squirrel livers, LDH concentration increased significantly after 24 h cold I/R, and after 72 h was ∼10-fold higher compared with control livers. In torpid livers, LDH release increased ∼2.5-fold after 72 h cold I/R. LDH release in summer livers was significantly greater than in torpid after 72 h cold I/R (Fig. 1A). The difference in LDH release after cold I/R between torpid and summer livers was not due to a difference in the total amount of liver LDH, because LDH concentration was approximately twofold greater in hepatocytes isolated from fresh torpid livers (1865.7 ± 99.1 U/g, n = 11) compared with fresh summer livers (895.1 ± 109.5 U/g, n = 9; P < 0.05). LDH content of rat hepatocytes was similar to summer livers (868.9 ± 41.4 U/g, n = 10). In livers harvested from aroused hibernators, 72 h cold I/R increased LDH release approximately sixfold over values in control livers. LDH release after 72 h cold I/R in aroused livers was similar to that in torpid livers but significantly less than in summer livers (Fig. 1A).

Fig. 1.

Lactate dehydrogenase (LDH) release (A) and bile production (B) after 0, 24, or 72 h cold storage and 60 min warm reperfusion with Krebs-Henseleit buffer. Values are means ± SE for 9–12 animals per group except for aroused hibernators, which are 6 per group (24 h storage time not studied). For a given species/activity group, means with different letters are significantly different (P ≤ 0.05). For A, 72 h storage time, summer LDH was significantly different from torpid or aroused squirrels (*P ≤ 0.05). For B, 0 h storage time, torpid was significantly different from summer or aroused squirrels (#P ≤ 0.05).

Bile production during the 60-min warm reperfusion period was used as a marker of hepatocellular function (3). In rats, increasing cold storage time led to progressively lower bile production (Fig. 1B), such that bile production after 72 h cold I/R was reduced by 89% from control levels. The amount of bile produced was also significantly reduced after 72 h cold I/R in summer and aroused hibernators, but the effects were proportionately less (81% reduction in summer and 67% in aroused). There were no significant differences in bile production among the three storage times for torpid livers. An unexpected observation was the difference among the three squirrel groups in bile production in control livers (0 h). Bile production in torpid control livers was significantly less than in summer, but levels in aroused hibernators were similar to those in summer livers and significantly greater than in torpid livers (Fig. 1B). The amount of bile produced from aroused hibernator livers after 72 h cold I/R tended to be higher than in summer or torpid livers, although the differences were not significant (Fig. 1B).

We next determined the effect of 72 h cold I/R on liver function during an extensive 240-min reperfusion period. Repeated-measures ANOVA indicated significant effects of animal group, storage time, and reperfusion time, as well as an interaction effect of storage time and group (i.e., the change in LDH release over the reperfusion period is dependent on both storage time and animal group). In control livers from all groups, LDH levels remained low up to 180 min of reperfusion and then showed a modest, nonsignificant increase over the next hour (Fig. 2). Cold I/R (72 h) had a significant effect on LDH release from all livers, with the most pronounced effect observed in the rats (Fig. 2). The curves for LDH release after 72 h cold I/R in livers from summer squirrels and rats were not significantly different (due to the relatively high variability around mean values in these groups), but both were greater than in the torpid squirrels. Thus livers from torpid ground squirrels are most resistant to extensive reperfusion after cold storage followed by summer squirrels and rats.

Fig. 2.

LDH release during prolonged reperfusion in rat and squirrel livers after 0 or 72 h cold I/R. Values are means ± SE. n = 4 Animals per species/storage group except for summer squirrel livers stored for 0 h, for which n = 5. LDH release curves were significantly different by repeated-measures ANOVA based on log transformed data (*P ≤ 0.05).


Infusion of trypan blue was used to assess cell damage induced by cold I/R in rat and ground squirrel livers. In all groups, very few parenchymal cells contained dye (not shown). The trypan blue-positive cells consisted mostly of sinusoidal lining cells (endothelial and Kupffer cells). In rat livers, the number of trypan blue-positive SLC increased significantly after 24 cold I/R compared with control livers, and a further increase occurred after 72 h cold I/R (Fig. 3). In summer livers, there were more trypan blue-positive SLC after 72 h cold I/R compared with control or 24 h cold I/R (Fig. 3). No differences were observed in numbers of trypan blue-positive cells in torpid livers at any storage/reperfusion time.

Fig. 3.

Effect of cold ischemia-reperfusion (I/R) on trypan blue uptake in rat and squirrel livers. Values are means ± SE numbers of positively stained sinusoidal lining cells (SLC) per field. Samples sizes are (per storage group) 6 rats, 5 torpid hibernators, and 5 summer squirrels. Within each species/activity group, values with different letters are significantly different (P ≤ 0.05). Summer was significantly different from torpid for 72 h (*P ≤ 0.05).


VR (pressure/flow) was monitored over a 60 min reperfusion period in control and 72 h cold I/R livers. Repeated-measures ANOVA indicated significant effects of animal group and reperfusion time and an interaction effect between these two variables. Although there was no significant effect of storage time for any of the animal groups, there was a trend for increased VR after 72 h cold I/R in the rat, summer squirrel, and torpid livers (aroused livers showed the reverse trend; Fig. 4A). For clarity, the combined data for control and 72 h cold I/R for each group are shown in Fig. 4B. Initial VRs were similar in rat and summer livers, and both were significantly greater than initial VRs in torpid livers (Fig. 4B). VRs were similar in rat and summer squirrel livers at all reperfusion times. VR did not change in summer livers over the reperfusion period and was only modestly affected in the rats. In contrast, VRs of torpid livers began to fall shortly after the start of reperfusion and thereafter were significantly lower than initial values (Fig. 4B). Torpid VRs were lower than the corresponding VRs in summer and rat livers at all reperfusion times. In aroused hibernator livers, initial VRs were similar to those in summer squirrel and rat livers, but they decreased immediately after the start of reperfusion and subsequently reached values that were significantly lower than the corresponding values in rats or summer squirrels but similar to values in torpid livers.

Fig. 4.

Effect of reperfusion on vascular resistance (VR) of rat and squirrels livers after 0 or 72 h cold ischemia. A: mean VR ± SE (mmHg·min·ml−1·g) for control (•) and 72 h cold I/R (▴) livers. Lines are rats (---), summer (…), torpid (___), aroused hibernators (__··__··) for 6–8 animals per group. There was no significant effect of storage time in any of the groups. B: mean VR ± SE for combined control and 72 h cold I/R values in rat (•), summer (▴), torpid (▪), and aroused hibernators (□); n = 12–16 animals per group. *Significantly different from VR at 1 min within a group. #Torpid significantly different from 1 min VR in rat, summer, or aroused (P ≤ 0.05). All values in torpid and aroused livers were significantly less than in corresponding rat and summer livers from 5 to 60 min.

Kupffer cell phagocytosis.

We examined how cold I/R affects Kupffer cell phagocytosis in squirrel and rat livers using the uptake of CC. The uptake in rat and summer squirrel livers was increased by 72 h cold I/R but was unchanged in torpid squirrel livers (Fig. 5). CC uptake was significantly greater in summer vs. torpid squirrel livers after 72 h cold I/R (Fig. 5).

Fig. 5.

Effect of cold I/R on Kupffer cell phagocytosis in rat and squirrel livers. Values are means ± SE uptake of colloidal carbon by perfused livers. Samples sizes are (per storage group) 6 rats, 5 torpid hibernators, and 5 summer squirrels. Within each species/activity group, values with different letters are significantly different (P ≤ 0.05). Summer significantly different from torpid for 72 h (*P ≤ 0.05).

Mitochondrial respiration.

RCR measured after 60 min reperfusion was used as an indicator of mitochondrial function in control and cold-stored livers. Compared with control rat livers, RCR was significantly reduced in rat livers after 24 and 72 h cold I/R (Fig. 6). In summer and torpid livers, RCR also decreased after 72 h cold I/R compared with control livers, but RCR was still significantly higher in torpid compared with summer livers after 72 h cold I/R (Fig. 6).

Fig. 6.

Effect of cold I/R on mitochondrial respiratory control ratios in rat and squirrel livers. Values are means ± SE ratios of state 3 to state 4 respiration. Sample sizes are (per storage group) 6 rats, 5 torpid hibernators, and 5 summer squirrels. Within each species/activity group, values with different letters are significantly different. Summer was significantly different from torpid for 72 h (*P ≤ 0.05).


The development of hypothermic preservation techniques revolutionized the field of organ transplantation and significantly increased the availability and quality of grafts, including the liver (38). Despite these advances, liver preservation times are still limited and preservation injury is responsible for a significant number of graft failures that necessitate retransplantation and maximum medical effort to achieve adequate patient survival (16). The quality of human liver grafts after transplant correlates well with the length of time the organ is exposed to cold ischemia after harvest (14, 33, 34). Although extended periods at low temperatures can damage cells, it is the reperfusion with warm (37°C), oxygenated blood after cold ischemia that is most responsible for organ damage and eventual graft failure (10, 35, 40). Thus new approaches to improve preservation techniques that minimize damage due to cold ischemic storage and reperfusion are needed to increase the number and quality of donor livers.

To identify novel protective mechanisms that reduce the deleterious effects of extended cold storage, we turned to mammalian hibernators. The hibernation season is characterized by prolonged bouts of torpor, when body temperature falls to a few degrees above ambient temperatures (typically 0–10°C) and metabolic rates are reduced to <4% of active levels (30). Dramatic physiological changes accompany torpor-arousal cycles, including profound changes in cardiac output, ventilation, and tissue perfusion. These characteristics have led to increasing use of hibernators as models for enhanced resistance to stress and trauma states in nonhibernating species (5, 7, 12, 13). Despite the obvious parallels between hibernation and organ preservation, few studies have examined whether mammalian hibernators display enhanced abilities to withstand the deleterious effects of prolonged cold ischemia compared with summer euthermic individuals of the same species or with nonhibernating species. Churchill et al. (8, 9) studied livers from euthermic and hibernating Columbian ground squirrels that were fresh or preserved at 4°C for up to 72 h and found no differences between euthermic and torpid ground squirrels in hepatic glycolytic enzyme activities or tissue adenylate levels after cold preservation; the fall in adenylate levels in livers over a 24-h period of cold ischemia was similar in hibernating and euthermic animals (9). Anaerobic glycolysis was also similar in the two groups, as indicated by increasing lactate levels over the ischemic period. However, livers in those studies were studied directly after cold storage and were not reperfused after the ischemic storage period, when damage is most likely to occur. In a preliminary report, Green (18) observed that kidneys from torpid ground squirrels are more tolerant to 72 h of hypothermic storage (as assessed by animal survival after organ transplantation) compared with kidneys taken from euthermic squirrels, rats, or rabbits. In the present study, we compared the effects of cold ischemia followed by warm reperfusion in livers from rats and ground squirrels. Our results suggest that livers of hibernating ground squirrels not only have superior cold-storage properties compared with rats, but there is a seasonal shift from summer to hibernating phenotypes that increases liver resistance to cold I/R damage.

We used the release of LDH in liver perfusates in the IPL model as a marker for hepatocellular damage. The results showed that rat livers released increasingly more LDH during the reperfusion period as cold-storage time increased from 0 to 72 h. Release of LDH in summer squirrel livers also increased significantly after 72 h cold I/R, although to a lesser extent than in rat livers, and the increase in torpid livers was very modest. This remarkable ability of hibernator livers to display minimal damage after extended cold I/R was confirmed in preliminary experiments with torpid livers that were stored for 96 h cold I/R. LDH release in those livers was similar to that after 72 h cold I/R (data not shown). Additional evidence that livers from hibernators show superior resistance to cold I/R damage was provided by the experiments in which livers were reperfused for up to 4 h (Fig. 2). Compared with control livers, 72 h cold I/R in all animal groups increased the release of LDH, but there was a significant interaction effect of storage time with animal group. This resulted in a more pronounced effect in the rat and summer squirrel livers compared with the torpid livers. In fact, LDH release curves from rat and summer livers subject to 72 cold I/R were not significantly different, and both were higher than in torpid livers.

It is possible that the superior resistance of torpid livers to cold I/R injury was due to the low Tb and metabolism of the animals when livers were harvested. Numerous studies have demonstrated a protective effect of hypothermia on tissue viability after I/R injury (19). This does not appear to be the case, because LDH release from livers of hibernators used while euthermic during the hibernation season was similar to that in torpid livers and significantly lower than in summer livers. It could also be argued that the lower LDH release from torpid livers compared with summer was due to lower total amounts of LDH in hibernator livers; however, LDH activity in hepatocytes isolated from torpid livers was significantly higher, not lower, than in summer squirrel (and rat) livers. This finding underscores the greater protection displayed by hepatocytes of hibernating squirrels, because despite having higher total LDH levels, they release significantly less enzyme after cold I/R than do their summer counterparts or in the nonhibernator species.

Hibernator livers also retained a greater proportion of bile production after cold I/R compared with livers from summer squirrels or rats. After 72 h cold I/R, bile production was reduced to 33 and 45% of control levels in aroused and torpid livers, respectively. In contrast, bile production was reduced to 11 and 19% of control levels in rats and summer squirrels, respectively. It should be noted that although cold I/R did not significantly affect bile production in torpid livers, control values in that group were lower than in summer or aroused squirrels. The reason for low bile production in torpid livers is not entirely clear, because bile production in control livers harvested from aroused hibernators was significantly greater than in torpid livers and was similar to that in summer squirrels. Thus nutritional factors alone cannot account for the low bile production in torpid livers (because both torpid and aroused hibernators had not eaten for several weeks). It is worth noting that although bile production from aroused livers was significantly reduced after 72 h cold I/R, it tended to be higher than in summer or torpid livers. This suggests that not only are hepatocytes from aroused hibernators more resistant to damage after extended cold I/R compared with summer animals (as indicated by LDH release), they are better able to maintain bile production after cold I/R.

Protection of hibernator livers from damage after cold I/R was also evident from mitochondrial RCRs, which were used as a measure of mitochondrial functional integrity. Although RCRs fell significantly in all animal groups from 0 to 72 h cold I/R, the reduction was least severe in the hibernator livers. In addition, RCRs in hibernator livers after 72 h cold I/R were significantly higher than in summer livers, despite similar control values in the two groups. The relative preservation of mitochondrial function in hibernators after cold I/R is consistent with the lower degree of hepatocellular damage in this group, as indicated by LDH release into perfusates.

The damaging effect of cold I/R on SLC integrity has been well documented (reviewed in Refs. 2 and 37). Cold ischemia induces detachment of SLCs and, in particular, sinusoidal endothelial cells (SECs) from the sinusoidal surface. Although they remain alive during the ischemic period, SECs die rapidly on reperfusion, and the extent of SEC damage is correlated with length of cold ischemia (15). As expected, in our study, 24 h cold I/R significantly increased trypan blue uptake by rat SLCs, indicating loss of SEC membrane integrity, and the effect was further evident after 72 h cold I/R. A similar but less dramatic effect was observed after 72 h cold I/R in summer squirrel livers. The lack of significant effect of cold I/R on trypan blue uptake in SLC of torpid livers suggests that maintenance of SEC integrity may be a key mechanism responsible for the enhanced resistance to hepatocellular damage after cold I/R in torpid livers. Because SEC death leads to congestion and constriction of sinusoids and impaired hepatic perfusion, the enhanced preservation of these cells in torpid livers after cold I/R likely results in better maintenance of the hepatic microcirculation when organs are reperfused with blood after the storage period. A downstream effect of SEC detachment from the sinusoidal lining after cold storage is thought to be activation of matrix metalloproteinases (MMP) including MMP-9 and MMP-2, which subsequently contribute to liver preservation injury (41). Interestingly, we found, using our IPL model, that cold I/R did not affect release of either MMP in rat livers and actually increased the release of MMP-9 from torpid livers (6).

Postischemic microvascular perfusion injury is one of the hallmarks of hepatic dysfunction after I/R, with longer ischemic episodes (cold or warm) associated with increased microcirculatory dysfunction (37). To determine whether the effect of cold I/R in rat and ground squirrel livers involved microvascular dysfunction, we monitored VR changes during IPL experiments. Initial VR values were similar in livers from summer squirrels and rats, and there was minimal change in VR over time for either group. In contrast, livers harvested from torpid hibernators had significantly lower initial VR values than both summer or rat livers, and VR fell shortly after the start of reperfusion. The lower VR of torpid livers was due to reduced perfusion pressure, because perfusate flow rate was held constant during the reperfusion period. Perfusate oxygen consumption, however, was noted not to be flow limiting in these experiments (not shown), because the venous Po2 did not drop below 200 mmHg in any group. Interesting, livers harvested from aroused hibernators had initial VRs similar to those in the rats and summer squirrels, but their response to reperfusion then mimicked that of torpid livers, with an immediate fall in VR that reached values comparable with those in torpid livers by 60 min of reperfusion. Thus intrinsic mechanisms present in the liver when squirrels assume the hibernating phenotype may be critical in maintaining the hepatic microcirculation after cold I/R.

The fall in VR on reperfusion in the torpid and aroused hibernators suggests the existence of one or more mediators that promote flow through the sinusoids and is active in the blood-free IPL preparation. One potential mediator is nitric oxide (NO) released by the action of one or more isoforms of NO synthase (NOS). These include the constitutive, endothelial form of NOS (eNOS), which is an important hepatic vasodilator, and the inducible form (iNOS), which has been implicated in ischemic preconditioning in liver and other organs (20, 32, 37). We reported recently that livers of ground squirrels during the hibernation season express higher levels of immunoreactive eNOS and iNOS compared with summer animals (6). Increased hepatic NOS expression in hibernators could potentially contribute to the fall in VR on reperfusion we observed in torpid and aroused vs. summer squirrel livers. However, in preliminary experiments, there was no effect of the NOS inhibitor nitro-l-arginine methyl ester on VR response curves in torpid or aroused livers, either in control or cold I/R conditions (data not shown). Further studies are needed to identify the mechanisms responsible for the regulation of VR during hibernation.

Although the activation of Kupffer cells was not measured directly in our study, we used phagocytosis of carbon particles as a marker for altered Kupffer cell function after cold I/R. It is thought that initial stimulation of Kupffer cells leads to enhanced phagocytosis with minimal secretory activity, whereas continued stimulation (e.g., prolonged cold preservation/reperfusion) increases the release of free radicals, cytokines, and other inflammatory mediators (21). The release of toxic mediators from activated Kupffer cells is a well-described early event in the hepatocellular damage and microcirculatory failure that occurs after cold or warm I/R (2, 24, 26, 37). We observed a significant increase in phagocytosis after 72-h cold I/R in livers of rats and summer squirrels but not in the hibernators. This observation supports the idea that the hibernation phenotype is associated with a change in the normal signaling cascade by which Kupffer cells sense and respond to the hepatic environment after extended cold I/R. Assuming that enhanced phagocytosis reflects the early activation of Kupffer cells (29) in the IPL model, the absence of an effect of cold I/R on carbon uptake in the hibernators may reflect a suppression of Kupffer cell function that contributes to the better preservation characteristics of their livers relative to those of summer squirrels or rats. Further investigation is needed to understand how the hibernation phenotype affects Kupffer cell function in normal and ischemic states.

In conclusion, this study demonstrates that the hibernation phenotype in ground squirrels confers superior resistance to liver cold I/R injury compared with a nonhibernating species (rat) or with livers from ground squirrels studied during the summer (nonhibernating) season. Our findings suggest that within the same species, there is a physiological switch from a less-resistant (summer) phenotype to a super-resistant (hibernating) phenotype. Hibernation-induced protection from extended cold I/R was evident for hepatocellular integrity (LDH release and mitochondrial RCR), bile production, SEC viability, VR, and Kupffer cell activation. Furthermore, we found that the protective hibernation phenotype was expressed not only in livers harvested from torpid squirrels, but also from those used at euthermic Tb (∼37°C). Our findings therefore provide evidence for a preconditioning effect that is associated with the preparation and/or maintenance of the hibernation phenotype that is independent of the hypothermia and hypometabolism of the torpid state (5). Further elucidation of the mechanism(s) that underlie the increased resistance to cold I/R damage associated with the hibernation phenotype, and strategies to translate this information to organ preservation, could have a profound effect on the quality, duration, and availability of human organs for transplantation. These insights should also be more generally applicable to development of novel protection and preconditioning strategies to minimize damage due to regional I/R, whole body hemorrhagic shock and other trauma states.


This study was supported by Defense Advanced Research Projects Agency Contract N66001–02-C-8054 (approved for public release, distribution unlimited).


The authors thank Courtney Fleck and Michael Grahn for assistance with animals and Murray Clayton for assistance with statistical analysis.


  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


View Abstract