- Top of page
- MATERIALS AND METHODS
15-Deoxy-Δ12,14-prostaglandin J2 (15d-PGJ2) is an endogenous peroxisome proliferator-activated receptor γ (PPARγ) agonist that suppresses progressive matrix deposition; however, little is known about the effects of 15d-PGJ2 on human peritoneal mesothelial cells (HPMCs). We investigated the following: (i) the expression of PPARγ; (ii) the effect of 15d-PGJ2 on angiotensin II (Ang II)-induced fibronectin (FN) expression and secretion; (iii) the effect of 15d-PGJ2 (with or without Ang II and with or without the specific PPARγ antagonist GW9662) and pioglitazone, a synthetic PPARγ agonist, on hepatocyte growth factor (HGF) expression and secretion; (iv) the effect of HGF on Ang II-induced FN expression and secretion; (v) the expression of c-Met (a specific HGF receptor) and its phospho-signal; and (vi) the involvement of HGF in the effect produced by 15d-PGJ2 using selective c-Met inhibitor PHA-665752. The presence of PPARγ was detected by western blot analysis. 15d-PGJ2 inhibited Ang II-induced FN expression and increased HGF expression, even in the presence of Ang II. This effect of HGF expression was completely prevented by co-treatment with GW9662. Additionally, upregulation of HGF secretion induced by 15d-PGJ2 and HGF production induced by pioglitazone was revealed. We demonstrated the presence of c-Met, and presented evidence that HGF inhibits Ang II-induced FN expression and activates phosphorylation of c-Met, which is blocked by PHA-665752; 15d-PGJ2 also activated c-Met phosphorylation. Furthermore, PHA-665752 attenuates the inhibitory effects of 15d-PGJ2 on FN secretion. These findings suggest that 15d-PGJ2 has a novel and potent antifibrotic effect in HPMC and this action is likely mediated by HGF.
Peritoneal dialysis (PD) is a widely used renal replacement therapy. Peritoneal fibrosis (PF), characterized by submesothelial fibrosis and a thickening of the basement membrane, is an inevitable occurrence in PD patients over time (1). In turn, PF can lead to the development of encapsulating peritoneal sclerosis (EPS), a potentially fatal complication of long-term PD treatment. EPS can be avoided with the prevention of PF; however, the biological mechanisms underlying PF remain unclear and an effective treatment has not been identified.
Peroxisome proliferator-activated receptor γ (PPARγ) is a ligand-dependent transcription factor belonging to the nuclear hormone receptor superfamily (2). The most potent natural agonist for PPARγ is 15-deoxy-Δ12,14-prostaglandin J2 (15d-PGJ2), an endogenous prostaglandin D2 metabolite (3). A variety of synthetic thiazolidinediones, such as pioglitazone, rosiglitazone, and troglitazone, are also classified as PPARγ agonists and are used as insulin-sensitizing agents in patients with type 2 diabetes. Recent findings indicate that these agents play a role in the suppression of progressive fibrosis (4–6); however, there is scant data regarding the role of PPARγ agonists in human peritoneal mesothelial cells (HPMCs).
Accumulating evidence indicates that HPMCs may produce various cytokines (7) and extracellular matrix (ECM) proteins (8). Treatment of HPMCs with angiotensin II (Ang II) induces the expression of fibronectin (FN) (9), a key component of the ECM that serves as a scaffold for the deposition of other proteins. Therefore, this study focused on the effects of 15d-PGJ2 on FN expression in HPMCs.
Hepatocyte growth factor (HGF) was first purified as a potent mitogen for mature hepatocytes (10). Studies from numerous laboratories have shown that HGF is an endogenous antifibrotic factor that is capable of ameliorating fibrotic lesions and preserving organ function. Although two recent reports have suggested a beneficial effect of HGF upon HPMCs (11,12), the role of HGF in PF has not been thoroughly investigated. While the induction of HGF expression by PPARγ agonists has been reported (4,13,14), there are few reports on the production of HGF by HPMCs, and the association between PPARγ agonists and HGF is unclear.
The HGF protein binds to its high-affinity receptor c- Met tyrosine kinase, leading to phosphorylation of multiple serine and tyrosine residue sites. Studies have been advanced, particularly in oncogenes (15), that suggest that c-Met would be an important target for cancer therapy; however, the evidence is limited for the c-Met/HGF pathway in HPMCs. Therefore, the goal of this study was to evaluate the antifibrotic effect of 15d-PGJ2, an endogenous PPARγ agonist, in HPMCs through the c-Met/HGF pathway.
- Top of page
- MATERIALS AND METHODS
The goal of this study was to evaluate the antifibrotic effects of 15d-PGJ2 in HPMCs through the c-Met/HGF pathway. We demonstrated that 15d-PGJ2 reduces the Ang II-induced expression of FN and stimulates HGF production and secretion, and that HGF inhibits Ang II-induced FN expression. We also detected c-Met in HPMC lysates and found that its phosphorylation is induced by HGF and 15d-PGJ2. Furthermore, we found that HGF might contribute to the effects of 15d-PGJ2 using PHA-665752.
Ang II is a potent vasoactive peptide that also plays a regulatory role in the fibrotic process (17). The fibrotic effects of Ang II are related to the increased synthesis of ECM proteins. Increased levels of Ang II levels have been found in the supernatants of glucose-supplemented cultures of HPMCs (18), and an elevation of Ang II is evident in the PD effluent of patients with infectious peritonitis (9).
PPARγ agonists have garnered attention for their potential antifibrotic activity. Guo et al. reported that 15d-PGJ2 and pioglitazone have an inhibitory effect on transforming growth factor-β1 (TGF-β1)-induced FN expression in human mesangial cells (5), and pioglitazone, rosiglitazone and troglitazone inhibit Ang II-induced hypertrophy of neonatal rat cardiac myocytes (6). In this study, the expression of PPARγ protein was confirmed in HPMCs. Next, the inhibitory effect of 15d-PGJ2 on Ang II-induced FN expression was verified. The preincubation time of 15d-PGJ2 was set at 24 h for the production and secretion of HGF. The mRNA expression and protein levels were analyzed at 6 h and 24 h, respectively, based on data from a previous study (9). Since it appears that the cumulative production of FN contributes to the initial histopathological response observed in the development of PF, our results suggest that 15d-PGJ2 may have therapeutic potential as an antifibrotic agent in patients with PD. A recent study showed that rosiglitazone improves the structural and functional alterations of the peritoneum in an experimental rat model of EPS (19).
On the other hand, it is noteworthy that the PPARγ agonist pioglitazone provides limited benefit in the treatment of established liver fibrosis in rats (20). Pioglitazone arrested the progression of hepatic fibrosis only if the drug was introduced early enough in the course of disease progression (20). These data suggest that if PPARγ agonists are to be of therapeutic benefit in patients with liver disease, then treatment should be undertaken as early as possible in the course of the disease. Early treatment is also likely to be optimal in patients undergoing PD whose peritoneum is exposed to long-term bioincompatible dialysis fluids, high glucose concentrations, or refractory or repeated episodes of infectious peritonitis.
It is thus of interest that PPARγ binds to the putative PPAR responsive element in the promoter region of the HGF gene, with PPARγ ligand stimulation leading to increased HGF gene transcription, mRNA expression, and protein secretion (4,13). In addition, previous studies have reported that there are not only PPARγ-dependent but also PPARγ-independent pathways involved in the mechanism of 15d-PGJ2 action (21). We report here for the first time that 15d-PGJ2 increases HGF mRNA expression and HGF production through PPARγ-dependent pathways in HPMCs. The rapid increase in HGF mRNA expression was consistent with a recent observation demonstrating the peak level of HGF expression by human mesangial cells at 3 h following 15d-PGJ2 stimulation (4). As HGF inhibits Ang II-induced FN, we first thought that Ang II might inhibit HGF. In other words, we thought that HGF and FN would show opposite actions. It is notable that Ang II had little influence on HGF production and that 15d-PGJ2 was able to induce HGF production in the presence of Ang II. It is also important that 15d-PGJ2 extracellularly secreted HGF. This results in activation of c-Met and plays a role in the antifibrotic action. Furthermore, the finding that pioglitazone upregulates HGF production will be clinically valuable, because this agent is already in use.
In vivo studies have shown that HGF has the potentially therapeutic effect of suppressing ECM overproduction (22). In PD, Nakamura et al. demonstrated that HGF expression is substantially increased in the damaged peritoneal tissues of PD patients in areas without severe structural changes (23), while HGF appears to have a beneficial effect in HPMCs (11,12) and in a rat model treated by peritoneal dialysate (24). In this study, treatment with rhHGF completely suppressed Ang II-induced FN expression with no evidence of cell injury. These results support a model where HGF acts as an intermediate effector mediating the antifibrotic actions of 15d-PGJ2.
Since PPARγ agonists upregulated HGF expression and HGF inhibited FN expression, we investigated the role of c-Met activated by HGF. Numerous studies have demonstrated that activation of c-Met, which results in the binding and phosphorylation of adaptor proteins and the subsequent activation of signal transducers, is involved in cell growth, cell survival, angiogenesis, cell motility, and migration. In this study, HGF dramatically induced phosphorylation of c-MET at both pY1230/1234/1235 and pY1003. Consistent with a previous study (4), we also found that 15d-PGJ2 activates c-Met phosphorylation, indicating that HGF levels secreted by 15d-PGJ2 were high enough for activation. PHA-665752, a selective small molecule inhibitor of c-Met, produced a dose-dependent reduction of these tyrosine phosphorylations. This is one of the first studies to show that HGF activates phosphorylation of c-Met in HPMCs, and that this activation can be blocked by PHA-665752.
Finally, this study investigated whether the antifibrotic effect of 15d-PGJ2 is mediated by HGF. Our findings suggest the possibility that HGF contributes to the effect of 15d-PGJ2. Although the mechanism of this action must be considered, there are few studies on the signaling pathways between Ang II and FN. One possibility is that the antifibrotic effect may contribute to inhibiting TGF-β1. TGF-β1 is the major profibrotic factor and is thus likely to be involved in the accumulation of ECM, such as FN, laminin, and collagen. Indeed, the beneficial action of HGF against TGF-β1 was shown in a previous report (11). We examined whether 15d-PGJ2 reduces Ang II-induced TGF-β1 mRNA and protein expression, resulting in a significant decrease at 24 h (data not shown); however, we could not find a significant difference in less than 24 h, which seems to be a little too late to accumulate FN. Alternatively, TGF-β1-independent signal transduction pathways, such as protein kinase C, protein tyrosine kinases, mitogen-activating protein kinases (MAPK), extracellular signal-regulated kinase (ERK), c-Jun amino terminal kinase (JNK) and p38 MAPK, may be involved (17). We previously reported that Ang II-induced FN expression is mediated by the activation of ERK1/2 and p38 MAPK (5), but not JNK (9). Guo et al. reported that 15d-PGJ2 significantly inhibits TGF-β1-induced phosphorylation of ERK1/2 and p38 MAPK. Likewise, the antifibrotic effects of HGF via ERK1/2 have been noted (25), but there are no published accounts regarding p38 MAPK. The contribution of p38 MAPK to the action of HGF has apparently gone unnoticed. Altogether, it is assumed that 15d-PGJ2-induced HGF might not regulate the activation of p38 MAPK, which leads to our result that the inhibitory effect of PHA-665752 was insufficient.