| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,1
* Department of Pediatrics,
Department of Molecular Biology and Pharmacology, and
Department of Cellular Biology and Physiology, Washington University School of Medicine, St. Louis, Missouri, USA
1Correspondence: Department of Pediatrics, Washington University School of Medicine, 660 S. Euclid Ave., Box 8208, St. Louis, MO 63110, USA. E-mail: rudnick_d{at}kids.wustl.edu
SPECIFIC AIMS
On the basis of recent studies showing that liver regeneration is impaired in a number of animal models of fatty liver disease, and others describing significant hepatic fat accumulation in a liver-specific peroxisome proliferator activated receptor gamma (PPAR
) overexpressing mouse model, we hypothesized that PPAR activity is likely to be regulated during normal liver regeneration and that disruption of such regulation could impair the regenerative response. Therefore, the specific aims of this research were to determine the effects of PPAR-activating thiazolidinediones on liver regeneration using the rodent partial hepatectomy model.
PRINCIPAL FINDINGS
1. Thiazolidinediones inhibit hepatocellular proliferation and cyclin expression during liver regeneration with efficacies corresponding to their potencies of PPAR activation
To begin to characterize the effect of thiazolidinediones on liver regeneration, hepatocellular proliferation was determined in C57BL/6J mice treated with troglitazone, pioglitazone, or rosiglitazone or with vehicle control 36 h after partial hepatectomy. The 36-hour time point corresponds to peak proliferation in wild-type (WT), untreated mice. This analysis showed that hepatocellular proliferation is significantly reduced by rosiglitazone [6±3% hepatocellular bromodeoxyuridine (BrdU) incorporation at 36 h after partial hepatectomy], moderately reduced by pioglitazone (15±6%) and not at all suppressed by troglitazone (33±15%) as compared to vehicle control (30±6%, Fig. 1
A, B, *P<0.02 for rosiglitazone vs. vehicle). More extensive analyses of hepatocellular proliferation from 24 to 48 h after partial hepatectomy also showed greater suppression of proliferation by rosiglitazone than by pioglitazone (Fig. 1C
, *P<0.04 for vehicle vs. pioglitazone or rosiglitazone; **P<0.02 for vehicle vs. rosiglitazone). However, by 72 h after partial hepatectomy, hepatocellular proliferation in rosiglitazone-treated animals is comparable to that seen in control animals. Taken together with studies establishing that rosiglitazone is a more potent PPAR activator than pioglitazone, and pioglitazone is more potent than troglitazone, these data indicate that thiazolidinediones impair hepatocellular proliferation following partial hepatectomy with potencies corresponding to their relative efficacies of PPAR activation.
|
To provide further evidence for the suppressive action of thiazolidinediones on liver regeneration, the effects of these drugs on hepatic cyclin expression and mitotic progression following partial hepatectomy were determined. Cyclin D1 regulates G1-S phase cell cycle progression and its hepatic expression is induced by 24 h after partial hepatectomy during normal liver regeneration. Rosiglitazone treatment significantly reduces the expression of cyclin D1 mRNA (*P <0.04) and protein (**P <0.01) at this time point. Pioglitazone results in intermediate suppression of both gene (P =0.06) and protein (*P <0.02) expression, while troglitazone does not suppress the expression of either. These data are consistent with the relative effects of troglitazone, pioglitazone, and rosiglitazone on hepatocellular proliferation. Cyclin B1 regulates G2-M phase cell cycle progression and its expression is induced by 48 h after partial hepatectomy. Both pioglitazone and rosiglitazone administration resulted in reduced expression of cyclin B1 mRNA and protein at this time point; however, only the rosiglitazone effect on cyclin B1 protein expression reached statistical significance (*P <0.04). Finally, analysis of mitotic body frequency at 48 h after partial hepatectomy, which is the time point corresponding to peak mitotic frequency in untreated C57BL/6J mice, showed greater reduction by rosiglitazone (2±1 mitoses per high-powered field) than by pioglitazone (25±14) compared to vehicle control (40±4). Taken together with the data reported above, these observations support the conclusion that thiazolidinediones impair liver regeneration with efficacies corresponding to PPAR-activating potential.
2. Rosiglitazone does not effect early TNF
-IL6-STAT3 signaling during liver regeneration
To begin to identify the mechanistic basis for the inhibitory activity of thiazolidinediones on liver regeneration, the effects of the most potent of these drugs, rosiglitazone, on signaling events known to be important for normal liver regeneration were investigated. Activation of the TNF-IL6-STAT3 signaling pathway was examined first. This analysis showed that plasma TNF and IL6 levels were comparably induced over the initial 6 h following partial hepatectomy in rosiglitazone-treated and control mice, indicating that the inhibitory effects of these drugs on liver regeneration does not result from suppression of activation of this cytokine signaling cascade. Moreover, by 12 h after surgery, plasma TNF and IL6 levels were significantly greater in rosiglitazone- compared to vehicle-treated animals (*P<0.05). Similarly, determination of hepatic STAT3 activation demonstrated comparable induction of phosphorylated STAT3 in rosiglitazone- and vehicle-treated mouse liver over the first 6 h after partial hepatectomy and significantly increased hepatic levels of activated STAT3 in rosiglitazone-treated mice at 12 and 24 h after surgery (*P <0.04). On the basis of these observations, the effects of rosiglitazone on expression of suppressor-of-cytokine-signaling (SOCS)-3 mRNA, which is a negative regulator of activated STAT3 induced during liver regeneration, was determined. The result showed that SOCS3 gene expression in regenerating liver is not suppressed by rosiglitazone administration, indicating that rosiglitazone augmentation of hepatic STAT3 activation does not result from impaired SOCS3 induction.
3. Rosiglitazone suppresses Skp2 mRNA expression but does not effect p21Waf1/Cip1 or p27Kip1 protein levels during liver regeneration
Troglitazone has been reported to suppress cellular proliferation in culture via a mechanism involving inhibition of mRNA expression of the SCF ubiquitin-ligase complex Skp2 and subsequent accumulation of specific cell-cycle inhibitory targets of Skp2-dependent degradation including p21Waf1/Cip1 and p27Kip1. Therefore, the effects of thiazolidinedione administration on Skp2 gene expression and p21Waf1/Cip1 and p27Kip1 protein accumulation during liver regeneration were investigated. The effects of rosiglitazone were investigated first. This analysis showed that hepatic Skp2 mRNA expression, which is maximally induced between 24 and 48 h after partial hepatectomy, is markedly suppressed by rosiglitazone administration (*P <0.03); however, under these conditions and during this timeframe, steady state protein levels of p21Waf1/Cip1 and p27Kip1 are not augmented. The level of p21Waf1/Cip1 protein is increased in rosiglitazone-treated mice at 72 h after surgery (*P <0.01); however, this effect is too late to account for the earlier effects of rosiglitazone on hepatocellular proliferation. Neither troglitazone nor pioglitazone has any effect on either Skp2 mRNA expression or p21 or p27 protein expression at 36 h after partial hepatectomy (which is the time point corresponding to peak Skp2 mRNA induction during normal liver regeneration). Together, these data indicate that the effects of rosiglitazone on hepatocellular proliferation during liver regeneration do not result from impaired degradation of the cell cycle inhibitors p21Waf1/Cip1 and p27Kip1 but leave open the possibility that thiazolidinedione effects are mediated by suppression of some other Skp2-dependent activity such as targeted degradation of other cell-cycle regulators.
CONCLUSIONS AND SIGNIFICANCE
The studies reported here show that the thiazolidinedione rosiglitazone impairs liver regeneration. This drug is known to elicit many of its effects through binding and activation of the nuclear steroid hormone receptor PPAR. Although PPAR-independent effects of the thiazolidinediones have also been described, the observations reported here showing that the relative inhibitory activities of rosiglitazone, pioglitazone, and troglitazone on hepatocellular proliferation and cyclin expression following partial hepatectomy correspond precisely with the relative PPAR-activating potencies of these drugs and support the conclusion that the inhibitory effect of rosiglitazone on liver regeneration is mediated through its interaction with PPAR. Thus, these data implicate, for the first time, PPAR as an important candidate regulator of the hepatic regenerative response.
The data reported here do not distinguish between the possibilities that the inhibitory effects of rosiglitazone on hepatic regeneration depend on drug-interactions with hepatic vs. extrahepatic PPAR. Indeed, under normal physiological conditions, levels of hepatic PPAR expression appear to be low, and the functional importance of such expression has not been clearly elucidated. However, the observation that hepatic PPAR expression is increased in genetically and environmentally induced models of fatty liver disease suggests that hepatic PPAR expression can be an important determinant of liver-specific physiology and pathophysiology.
Our data indicate that the mechanisms responsible for the inhibitory effects of rosiglitazone on liver regeneration do not depend on suppression of early cytokine signaling. Indeed, plasma levels of TNF and IL6 and hepatic levels of activated STAT3 were unchanged over the first 6 h after hepatectomy and were augmented 12–24 h after hepatectomy in rosiglitazone- vs. vehicle-treated animals. These data would seem to conflict with published observations showing that thiazolidinediones can suppress lipopolysaccharide (LPS)-stimulated TNF production; however, this apparent contradiction can be reconciled by a recent report indicating that proinflammatory cytokine production during liver regeneration is independent of LPS signaling. Furthermore, our data, taken together with published data indicating that IL6 supplementation can suppress hepatocellular proliferation in WT mice subjected to partial hepatectomy, raises the possibility that the impaired regenerative response in rosiglitazone-treated animals may even be the result of increased cytokine signaling. Rosiglitazone does suppress the early activation but not later changes in the activity of hepatic p38 MAP kinase during liver regeneration; however, the functional significance of regulated p38 MAP kinase activity during the regenerative response has not yet been clearly elucidated. The data presented here also indicate that the impaired regenerative response seen in rosiglitazone-treated mice does not result from disruption of regulation of C/EBP gene expression, or impaired degradation of the cell cycle inhibitors p21Waf1/Cip1 or p27Kip1. Our data showing that thiazolidinediones inhibit cyclin D1 expression raise the possibility, as suggested by analyses in cell culture, that direct PPAR-dependent inhibition of cyclin expression may be responsible for the effects of rosiglitazone on liver regeneration.
Finally, our data have important potential implications with respect to the hepatotoxicity associated with clinical use of thiazolidinediones. Thiazolidinediones were developed as antidiabetic, insulin-sensitizing drugs. Troglitazone, which was the first of these drugs used to treat patients with insulin-resistant diabetes mellitus, was associated with the development of idiosyncratic acute liver failure, and therefore withdrawn from clinical use. Hepatotoxicity has subsequently been reported in patients taking pioglitazone and rosiglitazone, raising the possibility that thiazolidinediones exert class-specific hepatotoxicity through their effects on PPAR activity. Although the mechanistic basis for such toxicity remains entirely unknown, the studies described here raise the possibility that thiazolidinedione-mediated hepatotoxicity results from PPAR-dependent inhibition of hepatic regeneration, and provide support for current clinical practices in which these drugs are avoided or used judiciously in patients with known or suspected liver disease.
|
FOOTNOTES
To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.06-6511fje
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
