Nuclear control of inflammation and fibrosis in nonalcoholic steatohepatitis: therapeutic potential of dual peroxisome proliferator-activated receptor alpha/delta agonism


  • Pablo Quintero Ph.D.,

    1. Departamento de Gastroenterología, Pontificia Universidad Católica de Chile, Santiago, Chile
    Search for more papers by this author
  • Marco Arrese M.D.

    Corresponding author
    1. Departamento de Gastroenterología, Pontificia Universidad Católica de Chile, Santiago, Chile
    • Address reprint requests to: Marco Arrese, M.D., Department of Gastroenterology, Pontificia Universidad Católica de Chile, Marcoleta 367, Santiago 833-0024, Chile. E-mail:; fax: 56-2-6397780.

    Search for more papers by this author

  • See Article on Page 1941

  • Potential conflict of interest: Nothing to report.


collagen type 1, alpha 1




liver function tests


methionine choline deficient


metabolic syndrome


nonalcoholic fatty liver disease


nonalcoholic steatohepatitis


peroxisome proliferator-activated receptor


type 2 diabetes mellitus


transforming growth factor beta


tissue inhibitor of metalloproteinase.

Nonalcoholic fatty liver disease (NAFLD) is currently regarded as the most common liver disease worldwide, affecting 25%-30% of the general population.[1] The spectrum of this entity spans from simple hepatic steatosis to nonalcoholic steatohepatitis (NASH), which can result in progressing fibrosis, liver cirrhosis, and, ultimately, hepatocellular carcinoma in a significant proportion of patients.[2] Thus, NAFLD is associated with an increased liver-related morbidity and mortality and is emerging as a leading cause of liver transplantation.[3] In addition, patients with NAFLD exhibit an increased risk of developing both type 2 diabetes mellitus (T2DM) and cardiovascular disease.[4] For these reasons, timely and effective treatment of patients with NAFLD, and particularly those with NASH, is indicated to prevent metabolic consequences and eventually hamper the development of liver cirrhosis. However, current treatment options are limited to lifestyle changes, which are effective but difficult to achieve because of adherence issues. Although many pharmacological agents have been proposed to treat patients with NAFLD/NASH, the only drugs tested to date in large, randomized, controlled trials are pioglitazone and vitamin E, which have shown efficacy for treatment of NASH.[1] However, their therapeutic value is limited and several safety concerns have been raised recently. Therefore, the development of novel, pathophysiologically targeted, safe, and effective therapies is urgently needed.

In this issue of HEPATOLOGY, Staels et al.[5] report promising preclinical data on the effects of a dual peroxisome proliferator-activated receptor (PPAR)-α/δ agonist (GFT505) in rodent models of NAFLD/NASH and hepatic fibrosis, along with some clinical data on the effects of the compound on liver function tests (LFTs) in humans. Before getting into details of their work, a few words on the role of PPARs in NAFLD/NASH are in order.

PPARs are lipid-activated nuclear receptors highly conserved in mammals that, upon activation by the appropriate ligand, control complex networks of target genes involved in a myriad of processes, including energy homeostasis, inflammatory response, and lipid and carbohydrate metabolism.[6] Receptors of this family form heterodimers with the nuclear retinoid X receptor and are divided in three subtypes, each encoded by a different gene: PPAR-α (NRC11 1); PPAR-δ (NRC2, also named β/δ); and PPAR-γ (NRC3). Though PPAR-α and PPAR-γ have a relatively restricted tissue expression, being predominantly expressed in hepatocytes and adipocytes, respectively, PPAR-δ exhibits a more ubiquitous expression with particularly high abundance in muscle tissue and macrophages. Activation of different PPARs represents an important pharmacological target because of the multifaceted metabolic effects on lipid and carbohydrate metabolism and their effects on innate immunity and inflammatory responses. Specifically, and among other effects in different tissues or cell types, activation of PPAR-α stimulates mitochondrial and peroxisomal fatty acid β-oxidation, activation of PPAR-γ determines insulin sensitization and enhances glucose metabolism, and activation of PPAR-δ reduces triglyceride accumulation, protects against lipotoxicity caused by ectopic lipid deposition, and regulates hepatic insulin resistance. All three PPAR isotypes exhibit anti-inflammatory effects.[7] Therefore, modulation of the activation of these transcription factors, which are misregulated in NAFLD/NASH,[8] is perfectly suited as a therapeutic approach to control inflammatory and metabolic signaling in NAFLD/NASH, as has been previously delineated.[9] Available data indicate that PPAR-α activation with synthetic ligands (fibrates) is able to abolish steatosis and reduce fatty liver in rodents, but has limited effects in humans.[10] On the other hand, PPAR-γ ligands (thiazolidinediones) have demonstrated to be effective in reducing liver fat content, decreasing serum levels of aminotransferases, and also ameliorating steatosis, inflammation, and even fibrosis in patients with NAFLD/NASH.[1] However, drugs in this class are associated with undesirable side effects, such as fluid retention and decreased bone mass, and some concerns regarding long-term safety have recently emerged. In particular, data indicate that rosiglitazone may increase the risk for cardiovascular events, and pioglitazone possibly increases the risk of bladder cancer.[11] Finally, because PPAR-δ activation reduces fat burden in liver cells and modulates hepatic inflammation and fibrosis in animal models,[12, 13] targeting this receptor could be of benefit for patients with NAFLD. Clinical studies with PPAR-δ agonists in moderately obese men, patients meeting diagnostic criteria for metabolic syndrome (MetS), or patients with dyslipidemia, most of them likely suffering from NAFLD, are promising in this regard,[13] but available data are limited.

Efforts to develop new agents that simultaneously combine the beneficial effects of agonizing different PPARs (dual PPAR-α/γ, -α/δ, or -γ/δ agonists or even panagonists α/δ/γ) have been made.[14] Indeed, these multimodal drugs represent an attractive class of agents with therapeutic potential for T2DM, MetS, dyslipidemia, and, likely, NAFLD/NASH. Several dual PPAR-α/γ agonists have been tested in recent years, but a number of safety concerns raised questions about their clinical applications. PPAR-α/δ agonist have been developed more recently, with GFT505 being a first-in-class agent.

The work by Staels et al.[5] is the first preclinical study assessing the efficacy of the dual PPAR-α/δ agonist, GFT505, in mouse models of NAFLD/NASH. The investigators first explored the pharmacokinetics of the compound in rats, showing that GFT505 undergoes extensive enterohepatic cycling. This is interesting because it implies that the drug acts mainly in the liver with limited effects in peripheral organs and potential safety implications. For the evaluation of the modulatory effects of GFT505 on hepatic inflammatory parameters and fibrogenesis in experimental NASH, researchers chose to use models with active inflammation (methionine-choline-deficient [MCD]-diet-fed mice and hApoE2-KI mice fed a western diet) and mild-to-moderate fibrosis (db/db mice fed with the MCD diet). The researchers also attempted to dissect the relative contributions of PPAR-α and PPAR-δ agonism to the hepatoprotective actions of GFT505 by using hApoE2 knock-in/PPAR-α knockout (KO) mice. Further exploration of the antifibrotic effects of GFT505 in a more intense fibrotic model, such as CCl4-intoxicated rats, was also carried out. Collectively, data show that GFT505 significantly attenuated steatosis, inflammation, and fibrosis in the models used. The modulatory effects of GFT-505 correlated with reduced hepatic gene expression of proinflammatory (interleukin-1 β, tumor necrosis factor α, and the macrophage marker F4/80) and profibrotic (transforming growth factor β, tissue inhibitor of metalloproteinase 2, collagen type I, α 1 and collagen type I, α 2) genes. Indeed, the researchers should be commended for their extensive work in trying to assess the hepatoprotective actions of GFT505 as well as to evaluate the individual contributions of PPAR-α and PPAR-δ agonism to the observed effects. The latter is important because GFT505 has greater selectivity for PPAR-α than for the PPAR-δ isoform. In fact, in those experiments involving hApoE2-KI/PPARα KO mice, GFT505 exhibited a potent antisteatotic effect and antifibrotic activity likely related to specific activation of PPAR-δ. With regard to the latter, it is worth mentioning that some discrepant data have been published on the antifibrotic effect of PPAR-δ agonists. In fact, whereas Iwaisako et al.,[12] in agreement with the current data, reported antifibrotic effects of the PPAR-δ agonist, KD3010, another group published that another PPAR-δ agonist (GW501516) stimulates proliferation of hepatic stellate cells and actually promotes liver fibrosis.[15] This indicates that PPAR-δ agonists may differ significantly in their hepatoprotective and antifibrotic effects, which may relate to differences in PPAR specificity, tissue distribution, potency, and metabolism of the agonists.

The experimental models used in studies such as the one commented on above are always a matter of debate, considering that there is not an “ideal” model of NASH. In fact, feeding the MCD diet has complex metabolic consequences and does not necessarily recapitulate the pathophysiological features of NAFLD/NASH in humans. Also, the hApoE2-KI mouse is more a model of mixed dyslipidemia and atherosclerosis, rather than one of NASH. In this regard, it would have been informative to test GFT505 in a more “metabolic” model, such as the one induced by feeding mice with a high long-chain trans-fat solid diet and high-fructose corn syrup.[16] Certainly, there are a number of issues to consider when translating mice work into humans that can only be solved by human trials. In this regard, the researchers provide preliminary data from a combined analysis of four phase II clinical studies carried out in patients with MetS. Results indicate that GFT505 positively influenced LFTs in this patient population. Because altered LFTs in this context are likely related to NAFLD/NASH, this information suggests that GFT505 is of potential benefit in this entity. Moreover, a very recent report from the same group publishing the article in comment showed that GFT505 also improves hepatic and peripheral insulin sensitivity in abdominally obese subjects,[17] giving more support to the potential benefit of the drug in the treatment of NAFLD. A randomized, controlled trial specifically designed to assess the efficacy and safety of GFT505 in NASH patients is underway ( Identifier: NCT01694849) to confirm this contention.

In conclusion, preclinical testing of PPAR-α/δ agonist GFT505 is encouraging because of its multifaceted actions (Fig. 1), and if its efficacy is confirmed, we could count it as an effective liver-targeted drug for the treatment of NAFLD/NASH in the near future. Therefore, confirmatory human data are eagerly awaited with the hope of not witnessing the disappointing fate of similar agents that, in spite of showing beneficial effects in experimental models, only modestly influence human disease or are associated with severe unwanted effects.

Figure 1.

Therapeutic potential of the dual PPAR-α/δ agonism in NAFLD. GFT505, a first-in-class PPAR-α/δ agonist, decreased hepatic steatosis and reduced the expression of proinflammatory and profibrotic genes, attenuating steatosis, inflammation, and fibrosis in rodent models of steatosis, steatohepatitis, and hepatic fibrosis. If confirmed in humans and no safety issues arise, this class of compounds could represent a therapeutic advance for NAFLD/NASH management.


  1. 1

    NRC1, 2, and 3 are the official denomination of PPARs suggested by the Nuclear Receptor Nomenclature Committee, which is based on the evolution of the conserved DNA- and ligand-binding domains (for details, see: