|Home | About | Journals | Submit | Contact Us | Français|
The inhibitory effect of dextrose-supplementation on liver regeneration was first described more than 4 decades ago. Nevertheless, the molecular mechanisms responsible for this observation have not yet been elucidated. We investigated these mechanisms using the partial hepatectomy model in mice given standard or 10% dextrose (D10)-supplemented drinking water. The results showed that D10-treated mice exhibited significantly reduced hepatic regeneration compared to controls, as assessed by hepatocellular BrdU incorporation and mitotic frequency. D10 supplementation did not suppress activation of HGF, induction of TGFα expression or TNFα-IL6 cytokine signaling, p42/44 ERK activation, immediate early gene expression, or expression of C/EBPβ, but did augment expression of the mito-inhibitory factors C/EBPα, p21Waf1/Cip1, and p27Kip1. In addition, FoxM1 expression, which is required for normal liver regeneration, was suppressed by D10 treatment. Finally, D10 did not suppress either FoxM1 expression or hepatocellular proliferation in p21 null mice subjected to partial hepatectomy, establishing the functional significance of these events in mediating the effects of D10 on liver regeneration.
These data show that the inhibitory effect of dextrose-supplementation on liver regeneration is associated with increased expression of C/EBPα, p21, and p27, and decreased expression of FoxM1, and that D10-mediated inhibition of liver regeneration is abrogated in p21 deficient animals. Our observations are consistent with a model in which hepatic sufficiency is defined by homeostasis between the energy-generating capacity of the liver and the energy demands of the body mass, with liver regeneration initiated when the functional liver mass is no longer sufficient to meet such demand.
The liver has remarkable regenerative potential, which permits recovery from functional deficits induced following hepatic injury (1–3). Partial hepatectomy in rodents has been the most extensively used experimental model for investigating the molecular, cellular, and physiologic mechanisms that control this highly regulated response (4). Analyses using this system have led to the identification of a number of signals that are regulated during and necessary for normal liver regeneration. For example, the early hepatic regenerative response is characterized by initiation of Wnt (5–7), growth factor- (8–12), and cytokine-dependent (13–15) signaling, induction of p42/44 extracellular signal regulated kinase (ERK) activity (16) and activation of transcription factors including β-catenin, NFκB, and STAT3 (17–19). These events direct an immediate-early gene expression program (20) culminating in hepatocellular re-entry into and progression through the cell cycle. Ultimately, this leads to restoration of normal hepatic mass. Despite this knowledge, an integrated understanding of the precise mechanistic regulation of the hepatic regenerative response remains incomplete. Indeed, the nature and identities of the most proximal and distal signals that initiate and terminate hepatic regeneration are still largely unknown.
Liver mass is maintained in health or recovered by regeneration following injury in precise proportion to body mass (21). This well known observation suggests that the signals that initiate and terminate the hepatic regenerative response might be coupled to systemic demands on hepatic function. Consistent with this idea, previous studies have shown that rodents become hypoglycemic following partial hepatectomy, and that either intravenous or enteral dextrose-supplementation markedly suppresses the hepatic regenerative response (22–26). Although these observations were first made more than four decades ago, neither their functional significance nor mechanistic basis has yet been elucidated. In this manuscript, we describe our analyses of the molecular mechanisms responsible for dextrose-mediated inhibition of liver regeneration.
Male, 2–3 month old wildtype C57Bl/6J and Cdkn1a (p21)-null mice (B6;129S2-Cdkn1atm1Tyj/J; Jackson Laboratory, Bar Harbor, ME) were maintained on 12 h dark-light cycles with ad libitum access to standard rodent chow and water until 60 hours before surgery. At that time, experimental mice were provided ad libitum access to sterile-filtered 10% dextrose (D10) in drinking water while control animals were given unsupplemented sterile water. Access to chow was continued in both groups, and D10 and unsupplemented water were changed daily. Mice were subjected to partial hepatectomy using standard methodology (27–30): Mice were sedated with inhaled Isoflurane (VEDCO, Inc., St. Joseph, MO) via anesthesia vaporizer, then subjected to mid-ventral laparotomy with exposure, ligation and resection of the left and median hepatic lobes, and closure of the peritoneal and skin wounds. At serial times after surgery, animals were sacrificed by inhaled CO2 and plasma and liver tissue were harvested. Blood glucose was determined by standard glucometric analysis (Ascensia Contour, Bayer Healthcare, Tarrytown, NY) immediately prior to sacrifice. Five to 15 animals were examined at each time point for each genotype and treatment group. All experiments were approved by the Animal Studies Committee of Washington University and conducted in accordance with institutional guidelines and the criteria outlined in the “Guide for Care and Use of Laboratory Animals” (NIH publication 86-23).
TNFα and IL6 levels were determined using the Bio-Plex multiplex bead-based assay (Bio-Rad, Hercules, CA), and insulin levels were measured using a commercially available enzyme-linked immunoassay (Linco, St. Charles, MO), each according to manufacturer’s instructions.
Total RNA was analyzed for expression of specific genes of interest using real-time reverse-transcriptase polymerase-chain-reaction (RT-PCR) as described in Supplementary Material and previously (28). (Target-specific forward and reverse primers listed in Supplementary Table 1.)
Data were analyzed using SigmaPlot and SigmaStat software (SPSS, Chicago, IL). Unpaired Student’s t-test for pair-wise comparisons and ANOVA for multiple groups were use to compare hepatocellular BrdU incorporation, mitotic body frequency, liver weight, and mRNA and protein expression levels between experimental groups, with significance (alpha) set at 0.05. Data are reported as mean ± standard error.
To investigate the functional significance of the hypoglycemia that occurs following partial hepatectomy and the mechanistic basis for the inhibitory effect of dextrose-supplementation on liver regeneration (22–26), we first characterized the kinetics of the murine hypoglycemic response to partial hepatectomy. The results showed that mice developed significant hypoglycemia detectable by 3 hours, persisting through 72 hours, and resolved by 7 days after surgery (Supplementary Figure 1). Next in order to characterize the inhibitory effect of dextrose supplementation on liver regeneration, the hepatocellular proliferative response to partial hepatectomy was compared between mice offered 10% dextrose-supplemented water (D10) and those given unsupplemented water. Water intake in D10-treated mice was ~4-fold higher (17±0.6 mL/animal/day) than in controls (4.7±0.4, *p<0.001). Analysis of the regenerative response showed significantly reduced hepatocellular BrdU incorporation in D10-treated mice 36 hours after partial hepatectomy (Figures 1A, B, p<0.02), which is the timepoint corresponding to peak proliferation during liver regeneration in wildtype mice (Figure 1B and (27)). Hepatocellular mitoses were also reduced in these animals at 48 hours after surgery (Figures 1C, D, *p<0.04), the time of peak mitotic progression during normal liver regeneration (Figures 1C, D, (29)). Liver:body weight 72 hours after surgery was ~10% lower in D10-treated animals (3.3±0.3%) compared to controls (3.7±0.1), however this difference did not reach statistical significance (p=0.22). The effect of D10-supplementation on hypoglycemia after partial liver resection was also analyzed, with results showing increased plasma glucose in D10-supplemented animals at 36 hours after partial hepatectomy (Figure 1E, p<0.003) but not at other time-points from 0–72 hours after surgery. Insulin levels were also increased at this timepoint (Figure 1F), however this difference was not significant (p=0.06). These results are consistent with prior analysis of dextrose-mediated inhibition of liver regeneration (25), and suggest that dextrose-supplementation must influence glucose flux into liver, extrahepatic tissues, or both during the regenerative response.
To begin to investigate the mechanisms responsible for D10-mediated inhibition of liver regeneration, the influence of dextrose on growth factor-dependent signaling was investigated. This analysis showed that expression and activation of hepatocyte growth factor (HGF), which promotes normal liver regeneration via activation of c-Met (8–11), were comparable in control and D10-treated animals subjected to partial hepatectomy (Figures 2A–C). Furthermore, p42/44 ERK activation, which is dependent on c-Met signaling during liver regeneration (11), was not disrupted by D10-supplementation (Figure 2D). Finally, expression of the epidermal growth factor receptor (EGF-R) ligand transforming growth factor α (TGFα), which increases in liver following partial hepatectomy (31), was comparable in regenerating liver from D10-treated and control animals (Figures 2E, F). These data show that D10 does not inhibit activation of HGF or expression of TGFα in regenerating liver.
Hepatic β-catenin expression is required for normal liver regeneration (6;7) and can augment EGF-R signaling (32). Canonical activation of this pathway requires Wnt-dependent inhibition (by phosphorylation) of glycogen synthase kinase 3β (GSK3β), which otherwise phosphorylates and promotes the degradation of β-catenin (33). GSK3β also phosphorylates and inhibits glycogen synthase (34). Thus, GSK3β regulates pathways involved in liver regeneration and glucose homeostasis. Therefore, the influence of D10-supplementation on GSK3β phosphorylation and β-catenin-dependent gene expression was examined. The results showed that D10 did not suppress hepatic GSK3β phosphorylation. In fact, higher levels of phosphorylated GSK3β were seen in livers from D10-treated versus control animals (Figure 3A, B; *p<0.01). Similarly, D10 did not impair transcriptional induction of many β-catenin targets (7;33), including Axin2 (Figure 3C), c-Fos, c-Myc (Supplementary Figure 3A, C) and Cyclin D (Figure 5A). Thus, D10 does not disrupt GSK3β-dependent regulation of β-catenin-dependent gene expression during liver regeneration.
Activation of TNFα-IL6-STAT3 signaling, which regulates liver regeneration (13–15;35), was examined next. This analysis showed that plasma TNFα and IL6 levels were comparably induced (Supplementary Figure 2A, B) and hepatic STAT3 was similarly phosphorylated (Supplementary Figure 2C, D) after partial hepatectomy in D10-treated and control animals. Thus, D10-supplementation does not prevent cytokine signaling during liver regeneration.
Next, the immediate early gene response to partial hepatectomy was investigated. As noted above, induction of hepatic mRNA expression of c-Fos and c-Myc and that of c-Jun, which are characteristic of this response (20;36), were not suppressed by D10 (Supplementary Figures 3A–C). In contrast, expression of PEPCK, which regulates gluconeogenesis and is also part of the immediate early gene response, was suppressed by D10-supplementation prior to partial hepatectomy; however, its subsequent induction was not affected (Supplementary Figure 3D, #x0002A;p<0.04). Taken together, these data show that D10 does not cause global disruption of hepatic immediate early gene expression following partial hepatectomy.
The expression patterns of the CCAAT/Enhancer Binding Proteins (C/EBPs) α and β are precisely regulated during liver regeneration, with C/EBPα levels declining and C/EBPβ levels increasing over the initial 24 hours following partial hepatectomy ((37) and Figure 4A–F). Several observations suggest that such regulation is important for normal hepatic regeneration: For example, C/EBPβ null mice exhibit impaired liver regeneration (38) and C/EBPα is mito-inhibitory in many cell types and tissues (39). To further investigate the basis for D10-mediated inhibition of liver regeneration, the effect of dextrose on hepatic expression of these factors was evaluated. The results showed that induction of hepatic C/EBPβ mRNA expression was comparable in D10-treated and control animals (Figure 4A). Furthermore, D10 did not inhibit induction of either the liver enriched activator protein (LAP) or liver enriched inhibitory protein (LIP). LIP and LAP are alternative C/EBPβ translational products reported to differentially modulate liver regeneration (40). Our analysis showed modestly increased hepatic expression of LIP in D10 treated animals during early liver regeneration, although this difference was not statistically significant (Figures 4B–C). In contrast, hepatic C/EBPα expression was dysregulated in D10-supplemented animals, with dextrose-treatment associated with failure to suppress C/EBPα mRNA expression 0–24 hours after partial hepatectomy (Figure 4D) and increased expression of the transcriptionally active 42 kDa C/EBPα protein isoform at 0 and 24 hours after surgery (Figure 4E–F, **p<0.02 versus water). These data raise the possibility that disrupted regulation of C/EBPα expression may contribute to the inhibitory effect of D10 on liver regeneration.
The early signaling events that characterize the hepatic regenerative response, including those described above and others, culminate in cyclin-dependent hepatocyte re-entry into and progression through the cell cycle. As part of this process, hepatocellular cyclin D1 and cyclin E expression are induced during G1 and mediate progression into S phase (41). Subsequently, cyclin B expression is induced and promotes progression through G2 into the mitotic phase. To further characterize the inhibitory effect of D10 on liver regeneration, the influence of D10 on hepatic cyclin mRNA expression was examined. The results showed that cyclin D1 and cyclin E mRNA expression were not suppressed by D10 (Figure 5A, B). In fact, cyclin D1 was more highly expressed after partial hepatectomy in D10-treated than in control mice, although this difference did not reach statistical significance (Figure 5A, p=0.3). In contrast, cyclin B1 expression was significantly decreased in D10-treated mice (Figure 5C, *p<0.05). Taken together, these data indicate that D10-supplementation inhibits liver regeneration downstream of cyclin D1 expression but prior to initiation of hepatocellular DNA synthesis, thus raising the possibility that the inhibitory effect of D10 is mediated by cyclin dependent kinase inhibitors. To investigate this possibility, the effect of D10-supplementation on expression of p21Waf1/Cip1 and p27Kip1 following partial hepatectomy was evaluated. The results showed that both mRNA and protein expression of p21 (Figure 6A–C; *p<0.001; **p<0.03) and p27 (Figure 6D–F; *p<0.001; **p<0.01) were increased by D10-supplementation during early regeneration, consistent with a potential functional role for these cell cycle inhibitors in mediating the inhibitory effect of D10 on liver regeneration.
The Forkhead Box transcription factor, FoxM1, is essential for normal hepatic regeneration (42) and is known to suppress hepatic expression of p21 during the regenerative response (43). Therefore the effect of D10 on hepatic FoxM1 expression during liver regeneration was investigated. The results showed that induction of FoxM1 mRNA expression after partial hepatectomy, which normally peaks 36–48 hours after surgery ((42) and Figure 7), was significantly suppressed by dextrose-administration (Figure 7, *p<0.04). Thus, D10 may inhibit hepatic regeneration by suppressing the induction of FoxM1 expression.
The data described above indicate that D10-supplementation is associated with increased expression of the mito-inhibitory factors C/EBPα, p21, and p27, and decreased expression of an essential promoter of liver regeneration, FoxM1, in regenerating liver. To address the possibility that elevation of p21 expression may mediate subsequent inhibitory effects of D10 on hepatocellular cell cycle progression, the effects of dextrose on liver regeneration in p21 null mice were investigated. In this experiment, regenerating liver was harvested 36 hours after partial hepatectomy, which is the timepoint corresponding to peak hepatocellular proliferation in p21 knockout mice (44). In contrast to wildtype mice, hepatocellular proliferation was not suppressed by D10-supplementation in these animals (Figure 8A, B). In addition, neither FoxM1 nor cyclin B expression was inhibited by D10-treatment in p21 null mice (Figure 8C, D). In fact, expression of each these genes was ~2-fold higher in D10-treated compared to control animals, although these differences were not statistically significant (p=0.2 for cyclin B; p=0.1 for FoxM1). Taken together, these data establish that p21 is required for the inhibitory effects of D10 on FoxM1 expression and hepatocellular proliferation following partial hepatectomy.
The studies reported here elucidate the mechanisms that contribute to dextrose-mediated inhibition of hepatic regeneration, demonstrating D10-dependent increases in expression of the mito-inhibitory transcription factor C/EBPα and the cell cycle inhibitors p21 and p27, and decreased expression of FoxM1 following partial hepatectomy. These data provide in vivo evidence suggesting that early suppression of hepatic C/EBPα expression after partial hepatectomy, which is a well-known characteristic of normal liver regeneration (37), is required for a competent regenerative response. FoxM1 expression is known to suppress p21 expression during and also to be essential for normal liver regeneration (42;43). The data reported here showing that D10-mediated inhibition of FoxM1 expression and hepatic regeneration is abrogated in p21 deficient mice indicate that p21 must also negatively regulate FoxM1 expression.
The mechanistic perturbations identified in the studies reported here are remarkably similar to those described in association with the impaired hepatic regenerative response seen in older rodents. Indeed, sustained C/EBPα expression (45), increased expression of p21 (45), and decreased FoxM1 expression (42;46) are each associated with the diminished regenerative capacity of the aged rodent liver. Taken together, these observations suggest that similar perturbations in metabolic regulation following partial hepatectomy in D10-supplemented and old animals may account for the impaired regenerative response observed in each of these settings. Our data may also have broader mechanistic implications. Taken together with the central role of the liver as the principle intermediary between dietary nutrient uptake and extrahepatic energy consumption (47), the observations reported here suggest a model in which functional hepatic sufficiency is defined by homeostasis between the energy-generating capacity of the liver and the energy demands of the body mass, with liver regeneration initiated when the functional liver mass is insufficient to meet such demands. This model is consistent with a recent report suggesting that rapid, marked loss of hepatic ATP after partial hepatectomy contributes to the signals that initiate liver regeneration (48).
Finally, these data have potential clinical implications for hospitalized patients with acute liver injury. It is common practice in such patients to provide an intravenous dextrose infusion while monitoring serum glucose. This practice is intended to prevent morbidity associated with hypoglycemia as a result of compromised hepatic function. The data reported here highlight the possibility that this activity may have unanticipated effects on the ability of the liver to recover in these settings, and suggest that studies examining the relationship between glycemic control and recovery from acute failure, partial resection, and even transplantation of small for size grafts may lead to improved management and outcomes in these settings.
We are grateful to Drs. Paul Hruz, Lou Muglia, Eyal Shteyer, and Phillip Tarr for helpful discussions regarding these studies.
Financial Support: These studies were supported by NIH (DK068219) and CDHNF/TAP (to DAR), the Digestive Disease Research Core Center (NIH P30 DK52574), a fellowship from the American Liver Foundation (to AW), and by Institutional Training Grant T32-HD07409 (to YT).