Long-term treatment with olanzapine has been reported to induce NAFLD in population with schizophrenia, 26 whereas the precise underlying mechanism remains unclear. This study is the first to comprehensively investigate the role of PCSK9 in olanzapine-induced NAFLD, with a particular focus on SREBP-1c-PCSK9-NPC1L1 and -genes related to lipid metabolism. As in our previous study,7 we successfully established a mouse model of olanzapine-related NAFLD and performed histological examinations and liver function tests. In addition, we created cell models that were analyzed using lipid droplet staining and cell activity analysis. Since only mild fibrosis was observed, even in the high-dose olanzapine group at week 8, olanzapine treatment appears to mainly cause hepatic steatosis rather than fibrosis, meaning that interventions targeting olanzapine metabolism-related side effects should act against steatosis.
Since its discovery in 2003, the critical role of PCSK9 in atherosclerotic cardiovascular disease via the LDLR-dependent pathway has been clarified;27 however, it can also exert more complex effects on lipid metabolism, such as in NAFLD. A small number of studies have reported controversial findings regarding whether PCSK9 exerts a protective or detrimental effect in NAFLD. In one study of PCSK9 deficient mice, amplified fatty acid uptake and triglyceride and lipid droplet accumulation were observed, suggesting that PCSK9 deficiency may promote liver steatosis.28 Conversely, a human study of 201 patients reported that circulating PCSK9 levels correlated with hepatic fat accumulation independently of other NAFLD metabolic risk factors.10 Another study demonstrated that the Chinese medicine berberine could reduce PCSK9 mRNA and protein levels and sequentially ameliorate hepatic steatosis.29 Here, we found that olanzapine increased PCSK9 mRNA and protein levels in mouse livers and cell lines in a time- and dose-dependent manner while increasing the number and size of lipid droplets. Notably, PCSK9 overexpression and rhPCSK9 intervention led to hepatic steatosis, whereas PCSK9 knockdown reversed this effect and ameliorated lipid accumulation, as determined by ORO staining. Thus, our results demonstrate that high intrahepatic PCSK9 levels play an essential role in lipid storage in the liver.
Lipid transporter receptors on the surface of hepatocytes play key roles in intra- and extra-cellular lipid metabolism, and various studies have suggested that PCSK9 might regulate LDLR and its closest family members (LRP1, VLDLR, and ApoER2), CD36, and NPC1L1.12–15 We found that both olanzapine and rhPCSK9 treatment significantly increased NPC1L1 mRNA and protein levels, consistent with previous findings in intestinal epithelial cells.14,20 NPC1L1 is known to interact with intestinal phytosterol and cholesterol transporters, which are critical for intestinal cholesterol absorption and systemic cholesterol homeostasis.30,31 Therefore, specific antibodies targeting NPC1L1 represent a promising strategy for preventing high-fat diet-induced fatty liver32 and hepatic NPC1L1-exacerbated diet-induced steatosis, which is accompanied by decreased hepatic VLDL secretion.33 Since PPARα activation can downregulate NPC1L1 expression34 and we observed a decrease in PPARα gene expressions, PCSK9 may upregulate NPC1L1 protein expression by decreasing PPARα levels. Thus, our findings further elucidate the role of NPC1L1 in NAFLD.
We also observed remarkable reductions in LRP1, VLDLR, and ApoER2 levels in olanzapine-treated mice and cell lines, consistent with previous studies suggesting that the increased PCSK9 expression acted on these three receptors at the post-translational level. Surprisingly, LDLR and CD36 levels decreased in the rhPCSK9 intervention group, but increased under olanzapine administration even with increased PCSK9. It should be noted that LDLR and CD36 gene expressions were markedly upregulated by olanzapine treatment, suggesting that olanzapine can regulate LDLR and CD36 at the transcriptional level, independently of its post-translational effects via the PCSK9 pathway. Inhibitors of protein kinase B (Akt) increase LDLR expression through extracellular signal-regulated kinase (ERK)-dependent stabilization of LDLR mRNA 35,36, and proliferative and neuroprotective effects of olanzapine in vitro appear to be mediated by activation of PI3K/Akt and ERK pathways in PC12 cells37. In addition, it has been reported that the hepatic mTOR signaling pathway is enhanced in olanzapine-induced dyslipidemia38 and plays a crucial role in regulating lipid biosynthesis and lipogenesis via peroxisome proliferator-activated receptor gamma (PPARγ),39 a key regulator of CD36.40–42 This could help explain the conflicting evidence regarding changes in LDLR and CD36 levels in this study. Furthermore, although high PCSK9 levels are strongly associated with elevated plasma LDL-C levels due to their effects on LDLR degradation, we found that olanzapine-induced dyslipidemia was mainly characterized by hypertriglyceridemia, not hyper-LDL-C, likely due to increased LDLR levels driven by the mechanisms stated above.
Beyond the receptor-dependent pathway, multiple factors are involved in hepatic fat accumulation, including enhanced de novo lipid synthesis, uptake, and cholesterol biosynthesis, and decreased fatty acid oxidation and secretion.43 In our model of olanzapine-induced NAFLD, we observed an increase in the expression of genes related to lipid synthesis (ACC, FAS, SCD1, DGAT1, and ACL), uptake (FATP1), and cholesterol synthesis (HMGCR, HMGCS, ACSS2, FDPS, and CYP51A1), whereas the levels of genes related to lipid oxidation (SCAD and PPARα) decreased. These data suggested a preliminary mechanism via which olanzapine affects hepatic steatosis. The next, we successfully found factors directly targeted by PCSK9, including FAS, SCD1, ACL, FATP1, HMGCR, HMGCS, CYP51A1, SCAD, and PPARα, using rhPCSK9 and PCSK9 siRNA intervention experiments. Despite the fact that olanzapine and rhPCSK9 reduced the gene expression of SCAD and PPARα, their levels were unaffected by PCSK9 siRNA transfection during olanzapine treatment. This could be due to another factor in addition to the decreased effect of PCSK9 against SCAD and PPARα. In other words, olanzapine not only downregulated SCAD and PPARα levels through PCSK9, but also upregulated them via unknown factors that should be explored in further studies. Taken together, PCSK9 drives hepatic lipid accumulation under olanzapine by promoting de novo lipogenesis, lipid uptake, cholesterol biosynthesis, and repressing lipid oxidation.
Having demonstrated that olanzapine upregulates PCSK9 and drives hepatocyte steatosis via receptor-dependent and non-receptor-dependent pathways, we sought to understand how olanzapine upregulated PCSK9 levels. The SREBP2 is the main transcription factor that regulates the expression of PCSK9. Under low cholesterol conditions, the N-terminal domain of SREBP-2 is released from the membrane bound
full-length SREBP-2, travels to the nucleus, and activates target PCSK9 gene transcription.44 In addition, SREBP1c is an important transcription factor that mediates lipogenic gene expression and has been reported to facilitate the activation and expression of its downstream genes that regulate fatty acid synthase in cultured cells and rat liver.5,45−46 Recently, a luciferase assay in HCV genomic replicon cells showed that SREBP-1c upregulated PCSK9 promoter activity,47 while other studies have found that PCSK9 is driven by SREBP-1c under special conditions.48,49 Surprisingly, we did not observe a significant change in SREBP-2 levels when treated with olanzapine and the SREBP-2 pathway could not explain the olanzapine-induced upregulation of PCSK9. But olanzapine dose-dependently increased SREBP-1c levels, with SREBP-1c suppression or overexpression accordingly altering PCSK9 levels. Thus, SREBP-1c appears to regulates PCSK9 in olanzapine-induced NAFLD.
Previous studies have demonstrated that NAFLD is not driven by a single mechanism, but instead involves a variety of pathways.50–52 Consequently, factors that were changed by olanzapine treatment but not by rhPCSK9 or siPCSK9 intervention, such as genes related to de novo lipogenesis (ACC and DGAT1) and cholesterol biosynthesis (ACSS2, and FDPS), may also participate in olanzapine-related NAFLD, just not via the PCSK9 pathway. In addition, the roles of LDLR and CD36 in NAFLD should be further examined. It is generally recognized that cholesterol absorption by LDLR and LCFA uptake via CD36 exacerbate lipid accumulation, causing NAFLD. However, current evidence is inconsistent. For LDLR, an NAFLD model has been constructed using LDLR deficient mice in combination with other interventions,53,54 and few studies have directly focused on LDLR in NAFLD. Some clinical and animal studies have shown that CD36 levels are significantly higher in fatty livers 55–58 and that plasma soluble CD36 is associated with fatty liver and its severity,59–62 yet CD36 deficiency has also been shown to exacerbate hepatic steatosis.63–65 For us, when considering individual cells and the liver, lipid uptake via LDLR and CD36 undoubtedly increase lipid accumulation; however, in the whole body LDLR and CD36 deficiency not only affects lipid uptake in the liver but also other complex processes. For example, under CD36-deficient conditions muscle tissue cannot store enough fatty acids, which may facilitate lipolysis in adipose tissue and increase plasma free fatty acid levels, thereby promoting NAFLD. This hypothesis should be investigated in future research. Just in our olanzapine-induced NAFLD model but not LDLR- or CD36-deficient mice, we think the upregulated LDLR and CD36 appears to increase the lipid accumulation.
In summary, this study is the first to systematically examine the mechanisms underlying olanzapine-induced NAFLD via the PCSK9 pathway (Fig. 7). Notably, olanzapine drives NAFLD by upregulating SREBP-1c, and thereby increasing PCSK9 levels, resulting in elevated NPC1L1 levels and changes in the expression of genes related to lipid metabolism, including increased lipid synthesis (FAS, SCD1, and ACL), uptake (FATP1), and cholesterol synthesis (HMGCR, HMGCS, and CYP51A1), and decreased lipid oxidation (SCAD and PPARα). Beyond this PCSK9-dependent pathway, elevated de novo lipogenesis enzymes (ACC and DGAT1) and cholesterol biosynthesis (ACSS2, and FDPS) also act as stimulatory factors in olanzapine-induced NAFLD. Furthermore, increased LDLR and CD36 levels could be associated with olanzapine-related hepatic steatosis and require further study.