Cholesteryl‐ester transfer protein (CETP): A Kupffer cell marker linking hepatic inflammation with atherogenic dyslipidemia?

Potential conflict of interest: Dr. Staels consults for GenFit. This work was supported by grants from the European Genomic Institute for Diabetes (ANR‐10‐LABX‐46) and the European Commission (RESOLVE, Contract FP7‐305707). J.T.H is supported by an EMBO Long Term Fellowship (ALTF 277‐2014). See Article on Page 1710 With its prevalence rising in concert with the epidemics of obesity, metabolic syndrome, and diabetes, nonalcoholic fatty liver disease (NAFLD) is rapidly becoming the most common chronic liver disease. NAFLD is a condition of the liver covering a histological spectrum from relatively benign "simple" steatosis (also called nonalcoholic fatty liver), over a more active disease phenotype, characterized by hepatocyte ballooning and lobular inflammation (nonalcoholic steatohepatitis; NASH), to fibrosis. Major challenges in NAFLD research include a better understanding of its pathophysiology to develop appropriate therapeutic strategies and the identification of diagnostic tools to replace diagnosis by invasive liver biopsies, which are not amenable for the frequent monitoring of disease progression. In this issue of Hepatology, Wang et al.1 identify hepatic Kupffer cells (KCs) as the major source of cholesteryl‐ester transfer protein (CETP), a plasma protein transported by lipoproteins, which exchanges cholesteryl‐esters for triglycerides between lipoprotein particles (Fig. 1), thus identifying it as a marker of hepatic KC levels and potential marker of NASH in humans.Figure 1: A model linking hepatic inflammation and atherogenic dyslipidemia. In healthy livers, normal KC content is reflected in low levels of plasma CETP, giving rise to high HDL cholesterol and low cholesterol content in non‐HDL particles. The present study by Wang et al. demonstrates that increased plasma CETP in obese individuals is a result of increased CETP synthesis in KCs. Elevated plasma CETP, in turn, drives the atherogenic lipid profile that is a hallmark of many NAFLD patients. Abbreviations: CE, cholesteryl‐ester; TG, triglyceride; VLDL, very‐low‐density lipoprotein.NAFLD is often considered the hepatic phenotype of metabolic syndrome, which is characterized by a proatherogenic lipoprotein profile consisting of high triglycerides and remnant lipoprotein particles and low high‐density lipoprotein (HDL) cholesterol. CETP is a glycoprotein that plays an important role in the metabolism of HDL as well as apolipoprotein (apo)B‐containing lipoproteins. Plasma CETP levels are inversely correlated with HDL cholesterol, and elevated levels are associated with increased risk for cardiovascular disease (CVD).2 Interestingly, CETP expression varies greatly among mammal species. Mice and rats, which are highly resistant to atherosclerosis, lack CETP naturally and therefore carry the majority of plasma cholesterol in HDL. Conversely, rabbits, monkeys, and humans all express CETP, are low‐density lipoprotein (LDL) mammals and, consequently, prone to develop CVD owing to atherosclerosis. These observations provided a rationale for the development of specific CETP inhibitors to treat CVD. Although much attention has focused on the potential of CETP as a therapeutic target for CVD, the relative contributions of the various sites of CETP synthesis are not well studied. As with many apolipoproteins, its expression is highest in the liver, but peripheral tissues such as the small intestine, spleen, and adipose tissue may also contribute to its plasma levels.3 In the liver, there is discordant evidence supporting hepatocytes or nonparenchymal cells as the primary source of CETP synthesis. The original characterization of the human CETP gene by northern blotting analysis found a signal in perfused hepatocytes,3 whereas studies in cynomolgus monkeys suggest a nonparenchymal origin.4 To facilitate molecular studies in rodents, a humanized transgenic mouse was developed that expresses a human CETP minigene under the control of its natural flanking regions,5 and much of what we know about the regulation of CETP gene expression comes from this model. For instance, hepatic CETP has been shown to be strongly induced by cholesterol, likely through activation of the nuclear receptor liver X receptor, in transgenic mice6 and monkeys.7 In the current study, the investigators used a combination of human cohorts and animal models to identify hepatic KCs as the major site of CETP production. First, the investigators found hepatic CETP expression to be >10‐fold higher in liver than in either subcutaneous or visceral adipose tissue of obese patients undergoing bariatric surgery. The investigators then assessed the cell type(s) responsible for CETP messenger RNA (mRNA) synthesis showing strong correlation of CETP mRNA with mRNA of MARCO, a KC marker, but not with the parenchymal cell marker, albumin. Using in situ hybridization, the investigators demonstrated a high degree of colocalization of CETP mRNA and CD68 mRNA in liver biopsies from both healthy and obese patients. Finally, they show that plasma CETP positively correlates with the number of CD68+ cells in liver. Taken together, these data make a strong case that the liver is the most important contributor to plasma CETP levels in the context of obesity, and that it is most abundantly expressed in CD68+ KCs. To provide experimental evidence, the investigators made use of the human CETP transgenic mouse described above crossed in the APOE*3‐Leiden background. Treatment with clodronate liposomes, which eliminates hepatic KCs, strongly reduced CETP expression in Western diet–fed mice. Moreover, treatment with niacin and fenofibrate, which improves the atherogenic dyslipidemia in metabolic syndrome patients as well as in the studied animal model, resulted in a concomitant reduction in hepatic KC content and CETP expression. One important caveat of studies in this animal model is whether the CETP minigene contains all regulatory regions allowing faithful reproduction of the human CETP expression pattern. Whereas the transgene includes the natural 5' and 3' flanking regions, most introns are removed, potentially impacting regulatory mechanisms that control CETP expression.5 Additionally, the concomitant reductions in KC and CETP levels observed subsequent to fenofibrate and niacin treatment could be secondary to their action on peroxisome proliferator‐activated receptors (PPARs) and apoB lipoprotein metabolism. Still, given that clodronate reduces hepatic macrophages through chemically induced apoptosis, direct reduction in KC numbers leads to reduced hepatic CETP mRNA and plasma CETP protein in this model. The finding that plasma CETP levels correlate strongly with hepatic CETP mRNA, which is primarily produced by the KCs, begs the question: Can plasma CETP be a useful marker for hepatic KC content and activity in the diagnosis of NASH? Although further studies are obviously required, the answer may not be so straightforward. The investigators show that induction of an immune‐inflammatory response in the CETP transgenic mice with the Bacillus Calmette‐Guérin (BCG) vaccine increases hepatic mRNA and plasma CETP levels, suggesting that CETP levels not only correlate with content, but also with the activity state of the KCs. However, previous studies have demonstrated that treatment of these mice with lipopolysaccharide (LPS) rather reduces plasma CETP, despite increasing the hepatic KC population.8 Thus depending on the pathophysiological stimulus, CETP expression and levels may vary positively or negatively with disease activity. What could be the mechanism of such differential regulation? In this study, the investigators consider CETP as a constitutively expressed marker of KCs. However, there are numerous stimuli that alter CETP mRNA levels appreciably, including cholesterol feeding5 and farnesoid X receptor agonists.9 Furthermore, it is known that BCG and LPS signal through different inflammatory pathways, possibly leading to opposite effects on CETP levels despite increased KC amounts. Moreover, KCs, as with all macrophages, may display distinct phenotypes, from the proinflammatory M1 to the anti‐inflammatory, profibrotic M2 polarized macrophage phenotype.10 Currently, there are no data connecting KC polarization status to CETP mRNA or plasma levels, an important missing link to evaluate the potential of CETP as a marker for NASH. Further studies are clearly necessary to determine whether and how CETP expression is regulated in KCs in the context of chronic liver diseases. CVD is the most common cause of mortality among NAFLD patients. Though the connection between metabolic syndrome, NAFLD and CVD is obvious, the molecular mechanisms are not yet well explained. The findings that plasma CETP levels are reflective of hepatic KC content and activity may also provide a mechanism linking NAFLD to atherogenic dyslipidemia of metabolic syndrome. The results of the current study are, in many aspects, exciting and pertinent, with currently ongoing phase III clinical studies assessing the impact of the CETP inhibitors, anacetrapib and evacetrapib, on CVD and the dual PPARα/δ agonist, elafibranor, on NASH, closely linked pathologies, now also linked through CETP. Author names in bold designate shared co‐first authorship.