GPR35: mechanisms of action and agonism as a potential therapeutic strategy for non-alcoholic fatty liver diseases

One of the consequences of the so called 'obesity epidemic' is that a variety of diseases and conditions linked to it have increased substantially in recent years. These include a spectrum of conditions initiated by the accumulation of excess amounts of fat in the liver that are not linked directly to over-use of alcohol. Such non-alcoholic fatty liver disease (NAFLD) encompasses a range of states from the initial build-up of fat in liver cellss (steatosis), to the additional development of inflammation (steatohepatitis) and liver scarring (fibrosis) that, in extreme cases, can potentially result in liver cirrhosis. NAFLD and non-alcoholic steatohepatitis (NASH) now affect some 3-12% of adults in the U.K. Although these conditions can be addressed by diet restriction there are currently no approved medicines to assist this process or that can be used when patients find this challenging to achieve and maintain. Recently when studying a cell type that is used as an experimental model of human liver cells (hepatocytes) we found that a drug called lodoxamide, that is used clinically in eyedrops, was able to prevent fat accumulation in these cells. Some years ago in different studies we had shown that lodoxamide was able to activate a cell surface receptor called GPR35, but only effectively when the receptor was the human version and not the version found in mice. We went on to show that when we genetically removed GPR35 from the hepatocyte-like cells lodoxamide no longer had this effect, suggesting that GPR35 must be involved. Mice are often used to study possible effects of drugs and medicines before deciding if they are suitable for trials in humans. However, because lodoxamide does not work effectively on mouse GPR35, we have modified mice so they have a human form of this receptor instead. Using liver cells taken from these so called "humanised" (transgenic) mice we were again able to show that lodoxamide was very effective in preventing fat accumulation.
We now plan to build on this work to try to gain further confidence that activating GPR35 in humans might be an effective strategy to treat fatty liver diseases. There is, however, much to do to provide such confidence. Firstly, at the moment, although we see how effective lodoxamide is, we do not know what molecular mechanisms are triggered that link GPR35 activation to the final beneficial outcomes. We will study this in both the hepatocyte cell models and in liver cells taken from the humanised mice. As part of this we will create another transgenic mouse with a different type of human GPR35 receptor that will directly help us to discover the key mechanisms. In addition, the gene for GPR35 in humans is more complex than the equivalent one in mice. This results in humans having an additional form of GPR35 that mice do not have. We will therefore also explore the details of this additional human specific "splice variant" and how it responds to lodoxamide and other chemical compounds that can activate GPR35. Finally, but potentially of greatest significance, we will take these GPR35 transgenic mice and maintain them either on a 'high-fat diet' designed to make them gain weight and accumulate fat in the liver or a modified high-fat diet that in addition to promoting fat accumulation also causes liver inflammation and fibrosis similar to human NASH. We will assess whether treating mice maintained on such diets with lodoxamide, is able to limit or prevent disease development when we subsequently analyse the liver, blood and other tissues for pathology at the end of the experiments. These studies will provide both basic knowledge on how GPR35 functions at a molecular level and will allow us to assess if activation of GPR35 in these humanised mice provides sufficient evidence to begin to move towards possible clinical studies using either lodoxamide or other GPR35 activating drugs.

Studies performed on both wild type and genome-edited human hepatocyte HepG2 cells and hepatocytes isolated from transgenic knock-in mice in which we replaced the poorly characterised G protein-coupled receptor GPR35 with the human orthologue, with an additional HA-epitope tag sequence to assist immunological detection, have indicated the ability of activation of this receptor to suppress lipid accumulation. This is consistent with activation of GPR35 being a potential strategy to limit or reverse non-alcoholic fatty liver disease (NAFLD) in humans. To further assess this potential, we will now explore intracellular mechanisms responsible for the GPR35-dependent effects. Such studies will include measures of the reconstitution of cellular functions when various forms and mutants of human GPR35 splice variants are (re)-introduced in GPR35 knock-out HepG2 cells and the development and assessment of hepatocyte function in a new, to-be-made transgenic mouse line, in which GPR35 will be replaced by a phosphorylation-deficient mutant of human GPR35. In concert with such detailed underpinning studies, we will directly assess the effects of GPR35 activation in both a standard high-fat diet model that will mimic aspects of early NAFLD and also a choline-deficient, L-amino-acid-defined, high-fat diet (CDAHFD) that is an established pre-clinical dietary model recapitulating the more severe histopathological characteristics of the progressive NAFLD subphenotype non-alcoholic steatohepatitis (NASH). Such studies will be performed across the group of humanised GPR35 transgenic mouse lines to assess the pharmacological characteristics of GPR35 agonists that might be most appropriate to move towards clinical studies.