AIM: To investigate the effects of mammalian target of rapamycin (mTOR) inhibition on liver regeneration and autophagy in a surgical resection model.

METHODS: C57BL/6 mice were subjected to a 70% partial hepatectomy (PH) and treated intraperitoneally every 24 h with a combination of the mTOR inhibitor rapamycin (2.5 mg/kg per day) and the steroid dexamethasone (2.0 mg/kg per day) in phosphate buffered saline (PBS) or with PBS alone as vehicle control. In the immunosuppressant group, part of the group was treated subcutaneously 4 h prior to and 24 h after PH with a combination of human recombinant interleukin 6 (IL-6; 500 μg/kg per day) and hepatocyte growth factor (HGF; 100 μg/kg per day) in PBS. Animals were sacrificed 2, 3 or 5 d after PH and liver tissue and blood were collected for further analysis. Immunohistochemical staining for 5-Bromo-2’-deoxyuridine (BrdU) was used to quantify hepatocyte proliferation. Western blotting was used to detect hepatic microtubule-associated protein 1 light chain 3 (LC3)-II protein expression as a marker for autophagy. Hepatic gene expression levels of proliferation-, inflammation- and angiogenesis-related genes were examined by real-time reverse transcription-polymerase chain reaction and serum bilirubin and transaminase levels were analyzed at the clinical chemical core facility of the Erasmus MC-University Medical Center.

RESULTS: mTOR inhibition significantly suppressed regeneration, shown by decreased hepatocyte proliferation (2% vs 12% BrdU positive hepatocyte nuclei at day 2, P < 0.01; 0.8% vs 1.4% at day 5, P = 0.02) and liver weight reconstitution (63% vs 76% of initial total liver weight at day 3, P = 0.04), and furthermore increased serum transaminase levels (aspartate aminotransferase 641 U/L vs 185 U/L at day 2, P = 0.02). Expression of the autophagy marker LC3-II, which was reduced during normal liver regeneration, increased after mTOR inhibition (46% increase at day 2, P = 0.04). Hepatic gene expression showed an increased inflammation-related response [tumor necrosis factor (TNF)-α 3.2-fold upregulation at day 2, P = 0.03; IL-1Ra 6.0-fold upregulation at day 2 and 42.3-fold upregulation at day 5, P < 0.01] and a reduced expression of cell cycle progression and angiogenesis-related factors (HGF 40% reduction at day 2; vascular endothelial growth factor receptor 2 50% reduction at days 2 and 5; angiopoietin 1 60% reduction at day 2, all P≤ 0.01). Treatment with the regeneration stimulating cytokine IL-6 and growth factor HGF could overcome the inhibitory effect on liver weight (75% of initial total liver weight at day 3, P = 0.02 vs immunosuppression alone and P = 0.90 vs controls) and partially reversed gene expression changes caused by rapamycin (TNF-α and IL-1Ra levels at day 2 were restored to control levels). However, no significant changes in hepatocyte proliferation, serum injury markers or autophagy were found.

CONCLUSION: mTOR inhibition severely impairs liver regeneration and increases autophagy after PH. These effects are partly reversed by stimulation of the IL-6 and HGF pathways.

The liver has the remarkable ability to regenerate in order to compensate for lost or damaged liver tissue after injury and thereby restore liver function and maintain homeostasis. This process is ultimately required after living-donor liver transplantation, in which a small-for-size graft is subjected to ischemia and reperfusion injury and transplanted into a recipient with urgent metabolic needs. In this situation, both loss of a substantial part of the initial liver mass as well as oxidative stress after reperfusion are central mechanisms of hepatic injury[1,2].

Liver resection triggers release of the cytokines tumor necrosis factor (TNF) and interleukin 6 (IL-6), crucial priming factors for the initiation of hepatocyte proliferation by activation of the janus activated kinases/signal transducer and activator of transcription (JAK/STAT) pathway[3-5]. This priming phase stimulates resting hepatocytes to enter the G₁ phase of the cell cycle. Simultaneously, growth factors including hepatocyte growth factor (HGF), contribute to the passage of hepatocytes from the G₁ into the S phase by activating the phosphoinositide-3 kinase (PI3K)/Akt signal transduction pathway[6-8]. PI3K/Akt interacts with the mammalian target of rapamycin (mTOR), involved in the control of protein synthesis, cell size and proliferation[9,10]. Both cascades lead to activation of a variety of signaling pathways, including upregulation of several downstream cyclins like cyclin D1, which is associated with the G₁-S phase transition of hepatocytes[3,4,6,11,12].

Besides being a key regulator of cell growth and proliferation, mTOR was recently identified to play an important role in the control of autophagy[13-15]. Autophagy is an evolutionarily conserved lysosomal degradation pathway that plays an important protective role in case of cellular injury by mediating the elimination of damaged cellular components[13]. In non-hepatic cells, autophagy has not only been implicated as a survival response, but also as a mediator of cell death during stress conditions[16,17]. Autophagy might therefore play a role in liver regeneration, though this has not been thoroughly studied. This is of special interest to the field of liver transplantation as mTOR inhibition, in combination with a short course of steroids, is an attractive alternative for current calcineurin inhibitor based immunosuppressive strategies. Calcineurin inhibitors are neurotoxic, associated with a 20% incidence of chronic kidney dysfunction and carry a cumulative risk for de novo malignancy of up to 55% at 15 years after liver transplantation[18-22]. mTOR inhibitors like rapamycin therefore represent an important immunosuppressive option, especially in patients with calcineurin inhibitor-induced neurotoxicity, poor renal function and possibly also in patients with hepatocellular carcinoma. However, in the initial phase after liver transplantation, the mTOR inhibitor rapamycin is rarely used, since it is reported to delay liver regeneration[23-25].

Rapamycin inhibits mTOR complex 1 (mTORC1) by complex formation with FK506 binding protein 12, thereby acting on its downstream messengers and abrogating translation initiation and protein synthesis, which results in cell cycle arrest at the G₁ to S phase[23-25]. Cyclin D1 as well as p21 are shown to be important downstream messengers of the rapamycin-mediated cell cycle arrest[26-28]. The exact underlying cellular and molecular mechanisms by which mTOR inhibition attenuates liver regeneration and the interplay between mTOR inhibition and autophagy in liver regeneration needs to be further characterized.

Both after kidney as well as deceased liver transplantation, mTOR inhibition in combination with steroids has proven an efficient immunosuppressive strategy. Addition of an mTOR inhibitor to steroid treatment might therefore also show beneficial effects after living-donor liver transplantation, especially in patients with compromised renal function. Aim of this study is to investigate the effects of mTOR inhibition, in combination with the steroid dexamethasone, on liver regeneration and autophagy in a surgical resection model and in particular its involvement in IL-6 and HGF stimulated pathways. Besides mimicking the post-transplant treatment strategy, this combination of immunosuppressants also allowed more specific investigation of the effects of exogenous IL-6 and HGF, since steroids are multi-potent inhibitors of endogenous production of pro-inflammatory cytokines like TNF and IL-6[29]. Effects on body and liver weight, hepatocyte proliferation, autophagy and hepatic function and injury were analyzed at specific time points after surgery in a 70% partial hepatectomy (PH) model in mice.

Current immunosuppressive strategies in the first period after liver transplantation mostly involve treatment with steroids in combination with mycophenolic acid, IL-2 receptor antagonists or calcineurin inhibitors[31]. These regimes are however associated with chronic renal failure, with an incidence of up to 20% kidney dysfunction over time[18]. The mTOR inhibitor and immunosuppressant rapamycin, in contrast to the calcineurin inhibitors tacrolimus and cyclosporin, does not cause nephrotoxicity and is suggested to be a good alternative in transplant patients with deteriorating renal function[32-34].

Recently, mTOR inhibition has gained wide interest in the treatment of cancer[35,36]. Therefore, also in patients transplanted for hepatocellular carcinoma, mTOR inhibitors are an attractive alternative with reported inhibitory effects on tumor growth and recurrence[37-40]. However, mTOR is a key regulator of cell growth and proliferation and its inhibition is reported to have detrimental effects on liver regeneration[23-25]. There may however be a more intricate relation as mTOR also regulates metabolism and inhibition of mTOR may preserve energy supplies for the remaining hepatocytes after liver resection to keep up metabolic function. This is supported by a recent publication showing excellent results in patients treated de novo with rapamycin after living-donor liver transplantation as well as data from animal experiments showing no increase in mortality with rapamycin treatment, even after a 90% liver resection and despite inhibited hepatocyte proliferation[41,42].

Additionally, mTOR has been implicated to be of paramount importance in the control of autophagy, a general term for pathways in which cytoplasmic material, including soluble macromolecules and organelles, are delivered to lysosomes for degradation[13,43-45]. Autophagy is thought to have evolved as a stress response mechanism that allows organisms to survive during harsh conditions, probably by regulating energy homeostasis[16]. Early histomorphologic studies showed a decrease in autophagic bodies of up to 98% at day 1 after PH[46-48]. This can support the hypothesis that the inhibition of intracellular autophagic degradation in regenerating liver has its biochemical equivalent, i.e., inhibited protein catabolism, and is interpreted as an important and adequate mechanism to shift from the physiological steady state to compensatory growth of the liver after PH. Degli Esposti et al[49] showed the presence of autophagy in 21% of good functioning human liver grafts 2 h after reperfusion, without differences between normal and steatotic livers. Ischemic preconditioning in this study increased autophagy only in steatotic livers, which appeared to have a protective effect on post-operative function. Wang et al[50] showed that autophagy is essential for hepatocyte resistance to oxidant stress and that loss of macroautophagy led to overactivation of the c-Jun N-terminal kinase signaling pathway that induced cell death. Therefore we studied the interplay between liver regeneration, mTOR inhibition and autophagy in a transplant-related 70% PH model. In accordance with the findings of others, we found a significant decrease in proliferating hepatocytes from 12% to 2% after mTOR inhibition, with concomitant decreases in hepatic gene expression of the cell cycle genes cyclin D1 and PCNA[25,42,51]. This was furthermore accompanied by increased serum transaminases, suggesting increased liver injury.

Rupertus et al[40] recently described that rapamycin had no detrimental effects on liver regeneration, yet in their study hepatocyte proliferation was not actually measured, but only estimated from wet liver weight at 12 d after hepatectomy. In our experiment, wet liver weight after mTOR inhibition was still lower at day 5 after liver resection. In the study of Dahmen et al[42] BrdU incorporation decreased from 17% to less than 1% at 2 d after 90% hepatectomy, without effects on survival. In the study of Palmes et al[25] the same effects were found, with decreased gene expression levels of TNF-α, HGF and TGF-β at day 2 after a 70% liver resection. Interestingly, in our series, we found a significant upregulation of TNF-α, downregulation of HGF, but no significant changes in IL-6 and TGF-β gene expression.

Similar to the Palmes study, gene expression of the angiogenic factors VEGF-R2 and Ang-1 was downregulated in our experiments. Inhibition of angiogenesis is suggested to be one of the most relevant mechanisms by which tumor growth and recurrence is inhibited[39,40].

In our study, mTOR inhibition furthermore resulted in a profound upregulation of IL-1Ra gene expression, which was not reported before. IL-1Ra is an anti-inflammatory cytokine, reported to be released in response to both surgical as well as toxic liver injury and to have a protective effect after CCl4-induced toxic liver injury[52-54].

We investigated whether the inhibition in hepatocyte proliferation could be overcome by kick-starting the priming phase of liver regeneration by pre-resection administration of IL-6 and HGF, both described to stimulate liver regeneration, especially in combined treatment[55-57]. It appeared that treatment with exogenous IL-6 and HGF partly reversed the negative effects of rapamycin by restoring TNF-α and IL-1Ra gene expression to control levels, significantly increasing gene expression of Cyclin D1 and PCNA and normalizing liver weight reconstitution. However, no significant increase in hepatocyte proliferation was found and serum transaminases were even further elevated, suggesting increased hepatocyte damage. This is in line with the findings of Haga et al[9], who found in their model of LPdk1KO mice that the PI3K/PDK1/Akt/mTOR pathway was regulated independent of the IL-6/JAK/STAT3 pathway. An alternative explanation for the increase in liver weight could be cellular hypertrophy cq. edema, which is supported by the decreased number of hepatocytes per HPF in this treatment group.

For the first time, we describe that mTOR inhibition also significantly increased hepatic autophagy during liver regeneration after PH. Earlier, Kondomerkos et al[58] showed that mTOR inhibition by rapamycin increased autophagy in the liver and heart of newborn animals. This effect may compensate for the decreased hepatocyte proliferation, as increased autophagy ameliorates oxidative stress and saves cellular energy.

Finally, the ongoing loss of body weight in mice treated with rapamycin is noteworthy. Similar effects of rapamycin on body weight have previously been reported by DiJoseph et al[59] and Zafar et al[60]. The role of mTOR in metabolism is complicated; it has been described that chemical inhibitors of glycolysis and mitochondrial function suppress mTORC1 activity, indicating that mTORC1 senses cellular energy[35]. This is crucial, because mTORC1-driven growth processes consume a large fraction of cellular energy and thus could be deleterious to starving cells. The mTORC1 pathway indirectly senses low ATP by a mechanism that is centred on the AMP-activated protein kinase[61]. During starvation, mTOR must be downregulated to avoid energy expenditure in absence of nutrients. Therefore pharmacological inhibition of mTORC1 could lead to a defective energy sensing system, mimicking starvation. On the other hand, rapamycin, as mTORC1 inhibitor, may protect the regenerating liver through this mechanism by slowing down the anabolic processes and saving energy and this may account for the fact that animals survive, despite seriously hampered liver regeneration.

In summary, this study investigated the role of mTOR in liver regeneration in vivo and more specific in IL-6 and HGF stimulated signaling pathways. mTOR inhibition resulted in inhibited liver regeneration and increased hepatic autophagy. Although exogenously administered IL-6 and HGF could overcome the rapamycin-induced inhibited reconstitution of liver mass and furthermore upregulated gene expression of factors known to be downstream of mTOR, no significant beneficial effects on body weight, hepatocyte proliferation, autophagy or markers of liver injury were seen. To interpret these data on mTOR inhibition in relation to the clinical setting of living-donor liver transplantation, it is important to realize that the model used is limiting in that it is purely a liver regeneration model without ischemia and reperfusion injury or alloreactivity. However, from these results, the use of mTOR inhibitors in the early post-transplant setting can currently not be recommended, despite their recently reported beneficial effects on cancer development and kidney function.

The authors would like to thank Dr. Wendy van Veelen, Jasper Deuring and Fred Bonthuis for technical support.