| COMMENTARIES |
Cholesterol Synthesis Inhibition Elicits an Integrated Molecular Response in Human Livers Including Decreased ACAT2
Mats Eriksson, M.D., Ph.D. and Paolo Parini, M.D., Ph.D., Center for Endocrinology, Metabolism & Diabetes, Department of Medicine and Clinical Chemistry, Department of Laboratory Medicine, Karolinska University Hospital Huddinge, S- 141 86 Stockholm, Sweden
Inhibition of the enzyme hydroxyl-methyl-glutaryl coenzyme A (HMG-CoA) reductase, which leads to a reduced cholesterol synthesis, represents a major breakthrough in modern medicine. By cholesterol synthesis inhibition (ChSI), plasma levels of atherogenic LDL particles can be substantially reduced resulting in lower cardiovascular morbidity and mortality, in patients with and without manifest disease [1]. The lipid lowering effects of this class of drugs (statins) are generally ascribed to the compensatory increase in hepatic LDL receptor expression, which results from the activation of the transcription factor sterol regulatory element binding protein 2 (SREBP-2) that follows the reduced intracellular cholesterol [2]. At higher degrees of HMG-CoA reductase inhibition, the secretion VLDL cholesterol and triglycerides are also reduced, whereas HDL cholesterol levels may tend to be higher, unchanged, or somewhat lowered depending on the administered statin [3,4]. The mechanism(s) behind the latter changes are less well characterized, and their relationship to the more positive clinical effects of high-dose ChSI on cardiovascular disease has been debated.
We have recently demonstrated the presence of a specific enzyme catalyzing the esterification of cholesterol in human hepatocytes: acyl-Coenzyme A:cholesterol acyltransferase 2 (ACAT2) [5]. On the basis of animal studies we have postulated that this enzyme – in contrast to ACAT1 which is the major cholesterol esterifying enzyme in cells other than hepatocytes and enterocytes – is involved in the secretion of cholesteryl esters in VLDL from the liver [6]. In order to further characterize the function of ACAT2 and to identify how different degrees of cholesterol synthesis inhibition affect human hepatic cholesterol metabolism, we have now performed a detailed analysis of the changes in hepatic cholesterol metabolism induced by low and high degrees of ChSI in a group of Swedish normocholesterolemic gallstone patients enrolled in a randomized, placebo-controlled study [7]. Thirty-seven patients were randomized to placebo (Placebo), or 20 mg/d fluvastatin (Low-ChSI), or to 80 mg/d atorvastatin (High-ChSI) for a treatment period of 4 weeks.
Plasma total cholesterol and the lathosterol/cholesterol ratio, an indirect measurement of whole body cholesterol synthesis, were determined in order to verify that different degrees of ChSI were reached. As seen in most studies, a tendency to a reduction in total cholesterol was observed with Placebo whereas Low-ChSI and High-ChSI resulted in 18% (p < 0.05) and 44% (p < 0.001) reductions, respectively. Similarly, a nonsignificant reduction in the lathosterol/cholesterol ratio was observed with Placebo, whereas Low-ChSI and High-ChSI induced significant reductions of 42% (p < 0.05) and 70% (p < 0.001), respectively. Hence, three different levels of cholesterol synthesis were clearly apparent after treatment in the three experimental groups. Analysis of LDL receptor protein expression in pooled hepatic membranes showed an increase which was inversely related to the degree of ChSI. Measurement of the LDL receptor gene expression showed a significant (2.7-fold) induction of the mRNA levels only in the High-ChSI group (p < 0.005). To further verify the expected transcriptional effects induced by the treatments, hepatic mRNA levels of HMG-CoA reductase were determined. Corresponding to the LDL receptor changes, HMG-CoA reductase mRNA was induced in the High-ChSI group (p < 0.05) whereas no significant change was observed in the Low-ChSI group. Accordingly, the expression of SREBP-2 mRNA showed a trend towards an increase that was related to ChSI. Similar results were also observed for protein convertase subtilisin kexin (PCSK)-9.
Separation of plasma lipoproteins by size exclusion chromatography demonstrated a reduction in the LDL cholesterol concentration that was inversely related to the degree of LDL receptor induction. Low-ChSI and High-ChSI showed 23% (p < 0.01) and 60% (p < 0.001) reductions in plasma LDL-cholesterol from baseline, respectively. Similarly, the magnitude of VLDL cholesterol reduction was related to the degree of ChSI. VLDL cholesterol was reduced by 19% in the Low-ChSI (p < 0.001) and by 55% (p < 0.001) in the High-ChSI group. A significant decrease in HDL cholesterol (-25%; p < 0.01) was also observed in the High-ChSI group.
The hepatic activity and gene expression of ACAT2 was related to the degree of ChSI. When compared to the controls, patients in the high-ChSI group had a 50% reduction in microsomal ACAT2 activity, while those in the Low-ChSI group only had a minor decrease. The decrease in ACAT2 activity in the High-ChSI group was paralleled by a decrease in ACAT2 protein expression. Measurements of ACAT2 mRNA levels also showed a significant decrease. No effects were observed for the microsomal activity or for the mRNA expression of ACAT1. Analysis of the gene expression of the apolipoproteins involved in VLDL secretion revealed a significant decrease in apo E mRNA in the High ChSI group (-34%; p < 0.05), whereas no effects on apo B mRNA abundance were observed upon ChSI. We also determined the apo B and apo E content in the VLDL before and during the different treatments. Interestingly, similar decreases (~ 30%) in apo B and apo E were observed both in Low-ChSI and High-ChSI groups. No effects on the mRNA expression of the microsomal triglyceride transfer protein (MTP) were observed in response to ChSI.
As mentioned above, HDL cholesterol levels were slightly reduced by High-ChSI treatment. This was independent of changes in apo A-I (% difference from baseline; Control, 0.70 ± 2.81; Low-ChSI, 21.6 ± 12.2; High-ChSI, 3.74 ± 6.94). Since animal experiments indicate that the hepatic expression of the HDL receptors, scavenger receptor class B type I (SR-BI), may partly regulate plasma HDL cholesterol levels, we measured the protein expression of its human counterpart, CLA-I. Unexpectedly, Western blot analysis did not show any change of this protein in response to ChSI nor was there any change in its mRNA levels. Other hepatic factors involved in the formation of plasma HDL, such as apo A-I, ABCA1, and CETP were not influenced by ChSI, at least not at the mRNA level.
We also assessed the effect of ChSI on biliary lipid composition in gallbladder bile. The absolute concentrations of all biliary lipids (cholesterol, bile acids, and phospholipids) were reduced by High-ChSI treatment. This was associated with a reduction of the relative proportion of cholesterol (-40%; p < 0.05), resulting in a decreased saturation of gallbladder bile with cholesterol (-38%; p < 0.05). The mRNA expression of the biliary export pumps for cholesterol, ABCG5 and ABCG8, were not influenced by ChSI treatment neither was the protein expression of ABCG8. No effect of ChSI was seen on the composition of individual bile acids or on bile acid production assayed by measurement of the plasma levels of C4/cholesterol. The plasma plant sterols, campesterol and sitosterol, were increased during High-ChSI treatment. Since the plant sterol/cholesterol ratio generally reflects intestinal absorption, this might be taken as an indication of increased absorption of dietary cholesterol. However, since the change in plant sterol ratios was inversely correlated to molar % cholesterol in bile (R = -0.44 for campesterol and R = -0.40 for sitosterol, respectively; both p < 0.05), it may be more plausible to ascribe this relative change as reflection of the reduced secretion of biliary cholesterol.
In conclusion, cholesterol synthesis inhibition by statins resulted in increasing induction of HMG-CoA reductase, LDL receptors, and a parallel reduction of ACAT2 and apo E expression. These changes occurred together with reduced plasma levels of VLDL and LDL cholesterol. The slight lowering of HDL seen in patients receiving atorvastatin 80 mg/day could not be correlated to changes in hepatic HDL-receptor CLA-I (SR-BI) or in ABCA1 expression. Finally, statin treatment resulted in reduced biliary cholesterol levels, further indicating a decreased turnover of hepatic cholesterol.
Apart from the known HMG-CoA reductase inhibition, statin treatment also reduces ACAT2 activity in human liver and this effect (a new pleiotropic effect of statins), in combination with a reduced apo E expression, may contribute to the lowering of VLDL cholesterol seen in addition to the decrease in LDL. Hepatic ACAT2 plays a key role in the synthesis and composition of apo B-containing lipoproteins, as demonstrated in ACAT2-deficient mice, in which the hepatic CE secretion into Apo B-containing lipoprotein is reduced [8]. In several mouse models, ACAT2 has been shown to be “pro-atherogenic” and its depletion leads to a drastic reduction of atherosclerotic lesions (for review see [6]). This reduction occurs despite the plasma levels of apo B, showing that not only the number of apo B-containing lipoproteins, but also the amount of ACAT2-derived cholesteryl esters present in the core of these lipoproteins seem to be a critical factor. Therefore, the observation of a lower microsomal activity of ACAT2 following ChSI in the liver of Swedish gallstone patients may result in additional, and LDL-receptor-independent, positive effects on cardiovascular disease.
References