Physiological role of hepatic NPC1L1 in human cholesterol and lipoprotein metabolism: New perspectives and open questions

The importance of the cholesterol transporter Niemann-Pick C1 like protein 1 (NPC1L1)in the small intestine for efficient absorption of cholesterol is well documented and its physiological importance is clear (1). Similarly, the drug ezetimibe (Zetia™), which blocks function of this transporter, reduces plasma cholesterol, primarily LDL, by dramatically reducing intestinal absorption of dietary and biliary cholesterol (2). There are controversies about the exact mechanism by which NPC1L1 functions: whether the protein is on the apical cell membrane or intracellular vesicles or both (3^–5), and whether ezetimibe directly blocks NPC1L1 or interrupts its association with other proteins in the sterol absorption and transport pathway (6^–8). Nevertheless, it is now accepted that this protein is the key player in sterol absorption, although other transporters may play ancillary roles (9, 10).

NPC1L1 is also abundant in human liver, although its physiological role with regard to hepatic cholesterol and lipoprotein metabolism is less clear. This lack of information reflects, in large part, species differences: mice, the most widely used animal model for NPC1L1 studies as well as cholesterol metabolism, do not express NPC1L1 in liver. Transgenic mice with liver-specific overexpression (20 times) of human NPC1L1 were found to have reduced biliary cholesterol secretion (11). In this issue of the Journal of Lipid Research, Kurano et al. (12) bring new light to the topic of hepatic NPC1L1 and also raise some intriguing questions regarding specific hepatic cholesterol pools and their differential metabolic utilization for specific lipoproteins.

In these studies, Kurano et al. (12) used adenoviral gene transfer to transiently express human NPC1L1 in mouse hepatocytes at levels approximately similar to those seen in human liver. Many of their data corroborate the findings of Temel et al. (11) who analyzed transgenic mice with chronic hepatic expression of human NPC1L1. Both groups demonstrated a 5 to 10-fold decrease in biliary cholesterol that could be “corrected” by ezetimibe treatment (11). This result shows one more parallel between enterocyte and hepatocyte function with regards to sterol and lipoprotein metabolism: sterol is recovered apically and secreted basolaterally as lipoproteins. Thus, teleologically, NPC1L1 functions to conserve cholesterol for the organism. Clinically, the data from these studies indicate, contrary to the general dogma, that ezetimibe reduces LDL not only by blocking cholesterol absorption in the intestine but also by increasing its hepatic biliary disposal and thereby increasing reverse cholesterol transport. A potential corollary to increased biliary cholesterol is that ezetimibe would increase the risk for gallstones. However, while low NPC1L1 expression was found in one cohort of gallstone disease patients (13), ezetimibe has not been found to increase this risk in the SHARP trial (14) and there is some limited evidence that it may decrease risk due to the overriding increase in cholesterol excretion (15).

As expected from decreased biliary output, Kurano et al. (12) showed that intrahepatic cholesterol increased with NPC1L1 expression, while LDL receptor and nuclear SREBP2 decreased. In keeping with these changes, plasma LDL cholesterol and apoB levels increased while remnant clearance rate decreased. Curiously, hepatic NPC1L1 expression caused a decrease in VLDL triglyceride content although VLDL cholesterol was not changed. Kurano et al. (12) present data suggesting a mechanism involving decreased MTTP and FASN expression due to decreased FoxO1 levels. Also curious, these effects were minimally changed by ezetimibe treatment. However, ezetimibe has been reported to increase MTTP expression in patients with nonalcoholic steatohepatitis and has been suggested as a possible therapy (16).

Another striking and unexpected result of hepatic NPC1L1 expression reported by both groups is the secretion of significant amounts of large apoE-rich HDL particles. Kurano et al. show that these particles are essentially devoid of apoA-I as well as apoB, and that their secretion is substantially reduced by ezetimibe treatment of the mice. They also present data showing increased apoE secretion by HepG2 cells when incubated with cholesterol-bile salt micelles and that this is also blocked by ezetimibe. The suggestion is that cholesterol recovered from bile is transported to a compartment that specifically utilizes apoE for HDL secretion. Contrary to some current models that showed the involvement of ABCA1 in HDL secretion (17), ABCA1 is not necessarily part of this secretory pathway because its levels were not changed by expression of NPC1L1. Interestingly, the secretion of apoE-rich HDL appears to be NPC1L1 dependent because cholesterol accumulation by an alternative mechanism (lanosterol synthase overexpression) did not result in secretion of these particles. Also, there may be a requirement for apoE to mobilize this pool as NPC1L1 expression in apoE-knockout mice did not increase HDL levels. However, the latter mice have severe metabolic abnormalities that may have obscured the effect.

The model generally described by Kurano et al. (12) as well as others (11) involves recovery of biliary cholesterol by hepatocytes expressing NPC1L1. However, it is also plausible that the presence of NPC1L1 in an endosomal recycling compartment associated with the apical membrane actually diverts bile-targeted cholesterol from being secreted and sequesters it into a unique pool. Such a mechanism would be consistent with current data as well as several previous studies that describe NPC1L1 as being a critical protein not only for cholesterol uptake but also for intracellular shuttling of vesicular cholesterol (3^–5, 18). In either case, these results highlight the field's incomplete understanding of hepatic NPC1L1 function, as well as intracellular cholesterol traffic and metabolism, and highlight the need for additional detailed studies into how cholesterol is compartmentalized in cells, especially hepatocytes, and how these different pools are utilized.

The clinical ramifications of the current and previous data from mice with hepatic NPC1L1 expression are not immediately apparent with regard to lipoprotein metabolism. Although ezetimibe lowered HDL somewhat in hamsters and apoE knockout mice (19, 20), the drug has not been linked with significantly decreased HDL in humans as the current mouse data might suggest. Rather, clinical studies have shown the opposite in most cases, with reduced triglycerides, as well as LDL, upon Zetia™ treatment and essentially no change in HDL (14, 21, 22). However, it is important to note that the mice in the current study were analyzed while on a chow diet with ∼10% of calories from fat, and some effects might be obscured in clinical studies that typically involve subjects with hyperlipidemia and/or metabolic disease who are consuming diets with higher quantities of fat. Further, mice lack CETP, which also profoundly affects lipoprotein metabolism in humans and most mammalian models other than mice and rats (23). An additional possibility is that transient expression of NPC1L1 has very different effects than chronic expression because compensatory mechanisms may take longer than 5 days to normalize metabolism.

Nonetheless, these new studies reveal that hepatic NPC1L1 has important functions in the liver, and perhaps elsewhere, that remain poorly understood and merit further investigation. Cellular cholesterol levels can affect, directly or indirectly, several pathways related to fat and glucose metabolism and energy storage (24, 25), as well as lipid and lipoprotein metabolism. Thus, NPC1L1 and ezetimibe may affect the size and/or utilization of particular cholesterol pools that could have important clinical ramifications, especially in the context of metabolic disease characterized by obesity and insulin resistance. Lack of NPC1L1 reduces fatty liver disease in mice and in ezetimibe-treated humans (26, 27). Nomura et al. (28) showed that ezetimibe can ameliorate insulin resistance in Zucker obese rats and provided data indicating that the effects were due to alterations in hepatic glucose metabolism. Labonté et al. (29) showed that NPC1L1 knockout mice and ezetimibe-treated mice are also resistant to diet-induced obesity and remain insulin sensitive even when hepatic NPC1L1 is not present. The mechanism for this effect was not determined, and decreased fat absorption or increased basal metabolism may be plausible contributors to this phenomenon. Importantly, improved insulin sensitivity has also been reported in several small clinical trials with ezetimibe (30, 31). The use of mice that demonstrate hepatic NPC1L1 expression will allow the distinction to be made between hepatic and intestinal contributions and will provide a more appropriate animal model for studies of the effects of ezetimibe and other cholesterol modulating therapies on the prevalent metabolic diseases of hepatosteatosis, gallstone disease, and insulin resistance.