Stearoyl-CoA desaturase-1 (SCD1) has gained much interest as a future drug target to treat fatty liver and its consequences. However, there are few and inconsistent human data about expression and activity of this important enzyme. We investigated activity and expression of SCD1 and their relationships with liver fat (LF) content in human liver samples. Fifty subjects undergoing liver surgery were studied. SCD1 activity was estimated from the ratio of oleate (C18:1) to stearate (C18:0) within lipid subfractions. Furthermore, SCD1 mRNA expression and LF content were measured. Similarly to previous studies, we observed a strong positive correlation between LF content and the C18:1/C18:0 ratio in the combined fatty acid (FA) fractions (r = 0.96, P < 0.0001), which could be interpreted as higher SCD1 activity with increasing LF. However, hepatic SCD1 mRNA expression did not correlate with LF (r = 0.16, P = 0.13). To solve these conflicting data, we analyzed the FA composition of hepatic lipid subfractions. With increasing LF content the amount of FAs from the triglyceride (TG) fraction increased (r = 0.96, P < 0.0001), whereas the FAs from the phospholipid (PL) fraction remained unchanged (r = −0.17, P = 0.19). Of these two major lipid fractions, the C18:1/C18:0 ratio in TG was 16-fold higher than in PL. Supporting the SCD1 mRNA expression data, the C18:1/C18:0 ratio of the TG or PL fraction did not correlate with LF (r = 0.26, P = 0.12 and r = 0.08, P = 0.29). We provide novel information that SCD1 activity and mRNA expression appear not to be elevated in subjects with high LF content. We suggest that the FA composition of lipid subclasses, rather than of mixed lipids, should be analyzed to estimate SCD1 activity.
- fatty liver
- stearoyl-coenzyme A desaturase-1
- fatty acids
nonalcoholic fatty liver disease (NAFLD) is the main cause of elevated serum liver enzymes among the general population, and the wide spectrum of liver damage ranges from simple steatosis to steatohepatitis, advanced fibrosis, and cirrhosis with end-stage liver disease (1). NAFLD also predicts the development of the metabolic syndrome and is associated with the development of atherosclerosis, insulin resistance, and type 2 diabetes (1, 23, 31, 32, 34, 36). Mechanisms involved in the process of hepatic fat accumulation include increased supply of fatty acids (FAs) from dietary fat and FA release from adipocytes to the liver, primarily impaired hepatic lipid oxidation (9, 15, 18, 22, 32), increased intake of carbohydrates (16), and proinflammatory signaling (3, 37). Furthermore, hepatic de novo lipogenesis (15, 18) that is predominantly regulated by sterol regulatory binding protein-1c (35) contributes to hepatic fat accumulation. After the assembly of triglycerides, very-low-density lipoprotein (VLDL) particles transport FAs from the liver to adipose tissue and other sites (35).
The enzyme stearoyl-CoA desaturase-1 (SCD1), which is predominantly expressed in the liver, plays an important role in this pathway. SCD1 catalyzes the synthesis of monounsaturated long-chain FAs from saturated fatty acyl-CoAs. The preferred substrates for SCD1 are palmitoyl- (C16:0) and stearoyl-CoA (C18:0), which are converted to palmitoleoyl- (C16:1,n-7) and oleoyl-CoA (C18:1,n-9), respectively. These FAs in the ester form are the major constituents of VLDL-triglycerides (VLDL-TGs) (10, 25).
SCD1-deficient mice had decreased liver fat content (6, 8), suggesting that the lack of SCD1 protected mice from fat accumulation in the liver. Consequently, SCD1 became a drug target to treat fatty liver (7). However, these animals were lean (27) and on a low-fat high-carbohydrate diet developed metabolic disorders as severe hypercholesterolemia, large decrease in high-density lipoprotein cholesterol (HDL-c) levels, and lower VLDL production (11). These results are supported by two studies (4, 30) and imply that liver-specific downregulation of SCD1 may impair VLDL assembly and thus may abolish the capacity of the liver to clear intrahepatic TG. Acute liver-specific downregulation of SCD1 by antisense oligonucleotides resulted in an increase, but not a decrease, in liver fat in rodents (13). In contrast, there was no increase in liver fat content caused by chronic SCD1 deletion in liver-specific knockout mice (24).
Data on hepatic expression and activity of SCD1 in humans have been very limited until now. In agreement with the animal data about effects of liver-specific downregulation of SCD1 (13), we recently found that the hepatic SCD1 activity estimated from the C18:1 n-9/C18:0 ratio in the serum VLDL-TG that are assembled in the liver, was negatively associated with liver fat content in obese humans (33). In contrast, two recent studies found increased hepatic SCD1 activity estimated from the C18:1,n-9/C18:0 ratio in liver biopsies of subjects with NAFLD and in chronic hepatitis C patients with fatty liver (2, 19); however, hepatic SCD1 mRNA expression was not examined in these studies. We hypothesize that the different results may be due to the different analytic strategies used. To address this important issue and advance the understanding of the relationship between SCD1 activity and liver fat content, in the present study we chose a similar approach and measured liver fat content and hepatic SCD1 mRNA expression in liver biopsies of 50 individuals. Additionally, because the analysis of FA ratios from unfractionated lipids may generate data that are very difficult to interpret (17), we measured and analyzed the hepatic FA composition within five lipid subfractions in liver biopsies of 25 individuals.
SUBJECTS AND METHODS
Data from 50 Caucasians (19 F/31 M, age 64 ± 12 yr, BMI 25 ± 4 kg/m2; 25 of whom had extensive measurements of hepatic FA composition in 5 lipid subfractions; 8 F/17 M, age 66 ± 9 yr, BMI 25 ± 4 kg/m2) undergoing liver surgery for hepatic carcinoma or solitary liver metastasis, were included in the present study (Department of General, Visceral and Transplant Surgery, University of Tübingen). Patients were fasted overnight prior to collection of the liver biopsies. The patient's routine laboratory test results (means ± SE) on the day prior to collection of the biopsies were hemoglobin 14 ± 1.7 g/dl, white blood cell count 7,376 ± 1,968/μl, platelets 304 ± 120 × 1,000/μl, urea nitrogen 29 ± 8 mg/dl, total protein 7.6 ± 0.4 g/dl, bilirubin 0.7 ± 0.3 mg/dl, prothrombin time (international normalized ratio, INR) 1.0 ± 0.08, aspartate aminotransferase 41 ± 29 U/l, alanine aminotransferase 38 ± 24 U/l, creatinine 0.9 ± 0.2 mg/dl, and C-reactive protein 1.1 ± 1.5 mg/dl. Subjects tested negative for viral hepatitis and had no liver cirrhosis. Informed written consent was obtained from all participants, and the local medical ethics committee approved the protocol. Liver samples were taken from normal, nondiseased tissue during surgery, immediately frozen in liquid nitrogen, and stored at −80°C.
Determination of FA pattern and liver TG content.
Tissue samples were homogenized in PBS containing 1% Triton X-100 with a TissueLyser (Qiagen, Hilden, Germany). To determine the liver fat content, TG were quantified in the homogenate with the ADVIA 1650 clinical chemistry analyzer (Siemens Healthcare Diagnostics, Eschborn, Germany) and calculated as milligrams per 100 milligrams of tissue. Using thin-layer chromatography (TLC), the liver homogenate was then separated into five subfractions containing phospholipids (PL), diacylglycerol, free fatty acids (FFA), TG, and cholesteryl esters. Transesterification and quantification by gas chromatography with a flame ionization detector was performed as previously described (33). In detail, the samples were cleared from protein by use of 2-propanol, n-heptane, and 2 mol/l phosphoric acid. Toluol, methanol, and water were added, and after centrifugation at 4,000 rpm (8,175 g) the upper phase was dried under nitrogen. The lipids were solved in CHCl3-CH3OH and applied to a silica gel chromatography plate (Merk, Darmstadt, Germany). PL, diacylglycerol, FFA, TG, and cholesterol esters were separated using n-hexane, diethyl ether, and acetic acid as a solvent. To identify the localization of the fractions, a pooled control plasma was also separated on each plate and lipid fractions were visualized by 2,7-dichlor-fluorescein under ultraviolet light. The fractions were scraped off the TLC plate, transferred to screw-capped vials, and dissolved in a methanol-toluol mixture (4:1 vol/vol) containing cis-13,16,19-docosatrienoic acid as an internal standard. Transesterification was performed by incubation with acetyl chloride at 100°C for 1 h. The cold sample was neutralized with K2CO3, and the upper phase was concentrated to 80 μl under nitrogen. The FA methyl esters were quantified using gas chromatography with a flame ionization detector, as previously described (28).
Determination of hepatic SCD1 mRNA expression by quantitative real-time PCR.
Frozen tissue was homogenized in a TissueLyser, and RNA was extracted using the RNeasy Tissue Kit (Qiagen) according to the manufacturer's instructions. Reverse transcription of total RNA quantitative PCR of SCD1 and β-actin was performed on the LightCycler system (Roche, Mannheim, Germany) with SYBR Green as described previously (28).
Data are given as means ± SD. Linear regression analysis was performed to compare activity indexes between lipid fractions. The statistical software package JMP 4.0 (SAS Institute, Cary, NC) was used, SCD1 activity was adjusted for BMI, and a P value of ≤0.05 was considered statistically significant.
In the liver samples, liver fat content displayed a large variability (0.3–17.6%), allowing us to study relationships over a broad range of liver fat content. First, we investigated whether the hepatic C18:1/C18:0 ratio in the combined FA fractions correlated with liver fat content. We found a very strong positive correlation between the C18:1/C18:0 ratio in the combined FA fractions and liver fat content (Fig. 1A). In concordance with earlier reports, this would indicate that SCD1 activity linearly increases with increasing amount of liver fat. When we used the C16:1,n-7/C16:0 ratio as an estimate of hepatic SCD1 activity, a similar yet weaker relationship with liver fat content was found (Fig. 1B).
Next, we investigated the relationship of hepatic SCD1 mRNA expression normalized to β-actin with liver fat content, as previously suggested (19). SCD1 is a short-lived enzyme that is primarily regulated on the level of mRNA expression. Surprisingly, we found no correlation between SCD1 mRNA expression and liver fat content, even in our extended group of 50 relatively lean subjects (Fig. 2).
To clarify these conflicting data from mRNA expression and hepatic FA ratios, we further analyzed the FA composition of five hepatic lipid fractions that were separated by TLC. The main origin of hepatic FAs are the TG and phospholipid (PL) fractions, which together accounted for 88 ± 6% (mean ± SD) of the total FAs in the liver (Fig. 3A). As the FA composition strongly differs between lipid species, we next investigated the C18:1/C18:0 ratio within these five lipid fractions. Interestingly, this ratio largely differed among the analyzed lipid fractions (Fig. 3B). The TG fraction contains significantly less stearate and more oleate than the PL fraction and therefore had a 16-fold higher C18:1/C18:0 ratio in our samples (Fig. 3B). We then investigated whether the contribution of individual lipid fractions to the total hepatic FAs is dependent on liver fat content. We observed that with increasing liver fat content the amount of FAs derived from the TG fraction increased, whereas the FAs from the PL fraction remained unchanged (Fig. 3C).
As the TG fraction has a 16-fold higher C18:1/C18:0 ratio than the PL fraction, this explains that the C18:1/C18:0 ratio in total hepatic FAs is mainly dependent on the liver fat content.
On the basis of these findings, we then investigated the relationships between the liver fat content and the C18:1/C18:0 ratios of the TG and PL fractions. Importantly, the strong correlation that was found when the total FAs were analyzed (Fig. 1A), was not apparent anymore (Fig. 4, A and B). Similar results were found between the C16:1/C16:0 ratios with liver fat content (Fig. 4, C and D).
On the basis of several indirect findings from animals indicating that increased SCD1 activity induces hepatic fat accumulation, this enzyme has become a future drug target to treat fatty liver and its metabolic consequences (25). However, recent animal data about liver-specific silencing of SCD1 suggested that impairment of SCD1 expression and activity is accompanied by hepatic steatosis (13). Investigating the relationship between the C18:1/C18:0 ratio in the serum VLDL-TG with liver fat content measured by 1H magnetic resonance spectroscopy (1H-MRS), we then provided the first human data supporting this hypothesis (33).
In two recent human studies, increased hepatic SCD1 activity, estimated from the C18:1/C18:0 ratio of the total hepatic FA, was observed in liver biopsies of subjects with fatty liver (2, 19), a finding suggesting that increased SCD1 activity in the liver is involved in the process of hepatic lipid accumulation. In the present study we addressed this controversial issue in more detail and additionally measured SCD1 mRNA expression and analyzed the FA composition in 5 hepatic lipid fractions.
First, estimating the hepatic SCD1 activity index from the C18:1/C18:0 and C16:1/C16:0 ratios of the total hepatic FA as previously described (2, 19), we confirmed the strong positive relationship of these measures with liver fat content in our human liver biopsy samples. However, when we subsequently measured hepatic SCD1 mRNA expression, we could not detect a positive correlation. This supports an important study by Westerbacka et al. (39), which also did not find increased hepatic SCD1 mRNA expression in humans with fatty liver. We then investigated why such large differences in estimating the relationship between hepatic SCD1 activity and liver fat content exist in the same samples and analyzed five hepatic lipid subfractions and their contribution to total hepatic FA. We found clear signs of FA partitioning and a characteristic FA composition, particularly between the two major hepatic lipid species TG and PL. The C18:1/C18:0 ratio was 16-fold and the C16:1/C16:0 ratio 3-fold higher in TG than in PL. Similarly strong differences between these ratios have been reported in other human lipids as well as in rodent liver (5, 14). In agreement with a previous report (29), the amount of FA from the TG fraction increased with increasing liver fat content, whereas the FAs from the PL fraction remained unchanged. Therefore, the C16:1/C16:0 and C18:1/C18:0 ratios in total hepatic FAs are predominantly influenced by the amount of TG but not the FA ratios in the individual fractions. For example, the subjects with the highest liver fat content also has the highest ratio, because in this person the highest TG content is present in the liver that has high C16:1/C16:0 and C18:1/C18:0 ratios. Thus, the positive relationship of the C18:1/C18:0 and C16:1/C16:0 ratios with liver fat content, which was found when the indexes were estimated from total hepatic FAs (Fig. 1, A and B and in Refs. 2 and 19), may merely reflect an increase in FAs derived from TG in subjects with high liver fat content. This concept is illustrated in the schematic Fig. 5.
As the difference between the PL and TG fraction is much stronger for the C18:1/C18:0 ratio (16-fold) than for the C16:1/C16:0 ratio (3-fold) this explains why the correlation between liver fat content and the unfractionated C18:1/C18:0 ratio (Fig. 1A) is much stronger than for the C16:1/C16:0 ratio (Fig. 1B). Defective VLDL secretion, a widely proposed pathophysiological mechanism in fatty liver disease, alone would be sufficient to cause such an increase of the hepatic TG fraction (12, 20, 26). The analysis of total hepatic FA in samples with different TG content may therefore easily be overinterpreted as a positive relationship between liver fat and SCD1 activity, a phenomenon that has also been suggested for plasma lipids (17). Within the TG and PL fractions, no clear relationship between the C18:1/C18:0 and C16:1/C16:0 ratios and liver fat content was found. We consider these data of great importance regarding the ongoing efforts that aim to silence SCD1 in humans with fatty liver.
In our previous study (33), when no liver biopsies were available, we even found a negative relationship between the SCD1 activity index and liver fat content measured by 1H-MRS. However, this relationship was apparent in obese but not in lean individuals. This suggests that SCD1 activity becomes important in the regulation of liver fat under increasing load of adiposity, a hypothesis that is supported by other studies (21, 38). This may also explain why we did not detect this negative relationship in our mainly lean subjects included in the present study.
A clear limitation of the study is the lacking metabolic characterization of the patients and the fact that they suffered from malignancies. Although these were not advanced disease stages, we cannot exclude the possibility that different results might have been obtained with severely obese individuals. However, sufficient liver tissue samples to perform mRNA expression and detailed lipid analysis, as performed in this study, are rarely available. Furthermore, obtaining liver biopsy samples from healthy individuals for research purposes is ethically not acceptable. Furthermore, the sample size of the study limits the power to detect weaker associations.
In conclusion, we used state-of-the-art measurements to get an insight into the role of hepatic SCD1 activity in lipid metabolism in humans. With this approach we provide important novel information that SCD1 activity and mRNA expression appear not to be elevated in subjects with high liver fat content. Furthermore, we have strong evidence supporting that the FA composition of lipid subclasses, rather than of mixed lipids should be analyzed to estimate SCD1 activity and avoid misinterpretations.
The study was supported by grants from the Deutsche Forschungsgemeinschaft (KFO 114 and a Heisenberg Grant to N. Stefan, STE 1096/1-1) and in part by a grant from the German Federal Ministry of Education and Research to the German Center for Diabetes Research. These studies were also supported by a grant from the Deutsche Forschungsgemeinschaft (KFO 114 and GRK 1302/1) and by a grant from the Ministry of Education, Youth and Sport of the Czech Republic (MSM0021627502).
No conflicts of interest are reported by the authors.
We thank all the study participants for their cooperation and Mareike Walenta and Iris Mertens for excellent technical assistance.
- Copyright © 2011 the American Physiological Society