Hepatic Stearoyl-CoA Desaturase (SCD)-1 Activity and Diacylglycerol but Not Ceramide Concentrations Are Increased in the Nonalcoholic Human Fatty Liver

The subjects (all Caucasians and Finns) were recruited from patients undergoing laparoscopic gastric bypass surgery based on the following inclusion criteria: 1) age 18–65 years; 2) no known acute or chronic disease except for obesity or type 2 diabetes based on history, physical examination, and standard laboratory tests (blood counts, serum creatinine, thyroid-stimulating hormone, and electrolyte concentrations), and electrocardiogram; and 3) alcohol consumption <20 g/day. The nature and potential risks of the study were explained to all subjects before obtaining their written informed consent. The patients had not lost weight before surgery (mean weight change over 10 months was −1.1 kg). The study protocol was approved by the ethics committee of the Helsinki University Central Hospital.

On the morning before surgery, blood samples were taken after an overnight fast for measurement of A1C, fasting serum insulin and C-peptide, liver enzymes, serum triglyceride, and LDL and HDL cholesterol concentrations. Body weight was recorded to the nearest kilogram using a calibrated weighting scale with subjects barefoot and wearing light indoor clothing. Wedge biopsies of the liver were taken at surgery. Approximately one-half of the liver sample was sent to the pathologist for routine histopathological assessment, while the rest was immediately frozen and stored in liquid nitrogen. The fat content of the liver biopsy specimens (percent of hepatocytes with macrovesicular and microvesicular steatosis) was determined by an experienced liver pathologist (J.A.) in a blinded fashion (23). The percent of macrovesicular steatosis was used as the liver fat percent. One of the patients had mild (grade 1) necroinflammatory and fibrotic changes, and two of the patients had mild (grades 1 and 2) fibrotic changes.

Before analyzing the lipid composition of the liver, frozen samples (5 mg) were mixed with an IS mixture containing 0.5–1 μg/sample of phosphatidylcholine (PC) (17:0/0:0), ceramide (Cer) (d18:1/17:0), PC (17:0/17:0), phosphatidylethanolamine (PE) (17:0/17:0), triglycerides (TGs) (17:0/17:0/17:0), and 200 μl chloroform:methanol (2:1). The tissues were homogenized with grinding balls in a mixer mill at 25 Hz for 2 min, and 50 μl of 0.9% NaCl was added. The samples were vortexed for 2 min, and after 30 min standing, they were centrifuged at 10,000 rpm for 3 min. The labeled lipid standard mixture was added into the separated lipid extracts (1 μg/sample) before UPLC-MS analysis.

Lipid extracts were analyzed on a Waters Q-Tof Premier mass spectrometer combined with an Acquity UPLC. The column (at 50°C) was an Acquity UPLC bridged ethyl hybrid C18 1 × 50 mm with 1.7-μm particles. The solvent system included 1) water (1% 1 mol/l NH₄Ac, 0.1% HCOOH) and 2) acetonitrile/isopropanol (5:2, 1% 1 mol/l NH₄Ac, 0.1% HCOOH). The gradient started from 65% A/35% B, reached 100% B in 6 min, and remained there for the next 7 min. There was a 5-min reequilibration step before the next run. The flow rate was 0.200 ml/min, and the injected amount was 1.0 μl. Reserpine was used as the lock spray reference compound. Lipid profiling was carried out using ESI⁺ mode, and the data were collected at a mass range of 300–1,200 m/z, with a scan duration of 0.2 s. The data were processed by using MZmine software (version 6.0 [24]) and the lipid identification was based on an internal spectral library (25). The relative amounts of all the identified lipids were quantified by calibrating with corresponding class-specific internal standards if available. Sphingomyelins and diacylglycerols were calibrated with PC (17:0/17:0) as an internal standard, while all lysophospholipids were normalized using lysoPC (17:0). Characteristics of the analytical method have been described in detail in the supplement of our previous study (26). The applied platform affords broad screening of multiple lipid classes, including triacylglycerols, cholesterol esters, and major phospholipids, from total lipid extracts within a single sample run. Due to use of ESI⁺ mode in MS analysis, the platform is not optimal for detection of some negatively charged phospholipids such as phosphatidylserines and phosphatidylinositols. The method does not cover glycosphingolipids or the low–molecular weight reactive lipids such as eicosanoids and free fatty acids.

Frozen liver tissues (5 mg) were spiked with 40–60 μg standards and homogenized in capped 2-ml microtubes (Sarstedt) at −20°C with chloroform:methanol (2:1; 400–600 μl) by using zirconium oxide balls in a mixer mill (25 Hz for 5 min). A total of 100–150 μl NaCl (0.9%) was added by vortexing, and after 1 h extraction time, the lower layer was separated by centrifuging at 10,000 rpm for 5 min. A total of 100–200 μl aliquots from the extracts were evaporated into dryness and used for fatty acid analyses.

The evaporation residues were redissolved into 700 μl petroleum ether (boiling point 40–60°C) by vortexing. Sodium methoxide (250 μl; 0.5 mol/l NaOMe in MeOH) and a couple of boiling stones were added, and the mixture was boiled at 45°C for 5 min (27). The samples were acidified with 500 μl of 15% NaHSO₄, 100 μl petroleum ether was added, and the samples were centrifuged in 2 ml microtubes at 10,000 rpm for 5 min. The petroleum ether layers were separated into gas chromatography vials, evaporated, and redissolved into 100 μl hexane.

Gas chromatographic analysis of fatty acids was performed using the Agilent 5890 series II gas chromatography, equipped with a 25-m FFAP column (0.32 mm inner diameter). The injection volume was 2 μl and split ratio was 1:23. Helium was used as carrier gas and the oven temperature program was from 70°C to 240°C at 7°C/min. The injector and detector (frame ionization detector) temperatures were 260°C and 300°C, respectively.

Plasma glucose concentrations were measured in duplicate with the glucose oxidase method using Beckman Glucose Analyzer II (Beckman Instruments, Fullerton, CA) (28). Serum free insulin concentrations were measured with the Auto-DELFIA kit (Wallac, Turku, Finland) and C-peptide concentrations by radioimmunoassay (29). Serum HDL cholesterol and triacylglycerol concentrations were measured with the enzymatic kits from Roche Diagnostics using an autoanalyzer (Roche Diagnostics Hitachi, Hitachi, Tokyo, Japan). The concentration of LDL cholesterol was calculated using the Friedewald formula (30). Serum aspartate transaminase (AST), alanine transaminase (ALT), and γ glutamyltransferase (γGT) activities were determined, as recommended by the European Committee for Clinical Laboratory Standards.

Desaturase and elongase activities were estimated from product-to-precursor ratios of the percentages of individual fatty acids according to the following equations: SCD1 = (18:1/18:0), Δ6 desaturase = (18:3n − 6/18:2n − 6), Δ5 desaturase = (20:4/20:3), and elongase = (18:0/16:0) (31). Since SCD1 preferentially converts 18:0 to 18:1 rather than 16:0 to 16:1 (19), the 18:1/18:0 index was used to estimate SCD1 activity. The lipogenic index was determined from the 16:0/18:2n-6 ratio (32).

Data are shown as means ± SE or, for non–normally distributed data, as median followed by the 25th and 75th percentiles. Correlation analyses were performed using Spearman's nonparametric rank correlation coefficient. The unpaired Student's t test was used to compare the NAFL⁻ and NAFL⁺ groups. Calculations were made using GraphPad Prism version 4.00 for Windows (GraphPad Software, San Diego, CA), SysStat Statistical Package (SysStat version 10; SysStat, Evanston, IL), and SPSS 16.0 for Windows (SPSS, Chicago, IL). P < 0.05 was considered statistically significant.

NAFL⁺ and NAFL⁻ groups were comparable with respect to age, sex, body weight, BMI, fasting serum insulin, fasting serum C-peptide, A1C, fasting serum triacylglycerols, fasting serum LDL cholesterol, and serum γGT concentrations. Fasting serum HDL cholesterol concentrations were lower and serum ALT and serum AST concentrations were higher in the NAFL⁺ than in the NAFL⁻ group (Table 1).

The concentrations (nmol/mg liver tissue) of glycerophosphatidic acid, ether-linked phosphatidylcholines, lysophosphatidylethanolamine, ether-linked lysophosphatidylethanolamine, and di- and triacylglycerols were higher the NAFL⁺ than the NAFL⁻ group (Table 2). The concentrations of ceramides, sphingomyelins, phosphatidylcholines, phosphatidylethanolamines, and ether-linked phosphatidylethanolamines were comparable between the groups. The results remained unchanged if the lipid data were normalized to total phospholipid concentrations (data not shown). Total hepatic lipid concentration was positively related to liver fat content (r = 0.53, P = 0.036). The livers of the NAFL⁺ group contained proportionally more esterified and free oleate (18:1n-9) and less esterified and free stearate (18:0) and long polyunsaturated fatty acids (Table 3) than the NAFL− group. The concentration of oleic (18:1n-9) fatty acid was higher in the NAFL⁺ group than in the NAFL⁻ group (0.32 ± 0.04 vs. 0.18 ± 0.03 nmol/mg tissue, P = 0.007).

Correlation coefficients between each of the 136 individual TGs (expressed relative to total hepatic TG) and histological liver fat content were calculated. These correlation coefficients were then plotted against the number of double bonds in the respective TGs. An inverse relationship was observed (r = −0.57, P < 0.0001, Fig. 1), implying that the TGs that were positively associated with liver fat content had only a few double bonds, whereas those TGs negatively related to liver fat content had many double bonds.

The SCD1 activity index, as estimated from the 18:1n-9–to–18:0 fatty acid ratio in the liver was 1.6-fold higher in the NAFL⁺ (5.7 ± 0.4) than in the NAFL⁻ (3.7 ± 0.4, P = 0.004) group. Liver fat percent was positively related to hepatic SCD1 activity index (r = 0.70, P = 0.003, Fig. 2). The activities of Δ5 and Δ6 desaturases in the liver were comparable between the groups (data not shown). The elongase activity index (18:0/16:0) was lower in the NAFL⁺ (0.23 ± 0.02) than in the NAFL⁻ (0.33 ± 0.02, P = 0.005) group and was inversely related to the percent liver fat (r = −0.78, P = 0.0004, Fig. 2). The hepatic lipogenic index (16:0/18:2n-6) was positively related to liver fat content (r = 0.54, P = 0.030, Fig. 2) but unrelated to hepatic SCD1 activity index (r = 0.28, NS).

The proportional amounts of long polyunsaturated fatty acids, such as 20:3n-6 and 22:6n-3, were strongly inversely related to SCD1 (r = −0.94, P < 0.0001, and r = −0.82, P < 0.0001) and positively to elongase (r = 0.83, P < 0.0001, and r = 0.86, P < 0.0001) activity indexes in the liver.

The hepatic concentrations of diacylglycerols were positively related to those of TGs (r = 0.58, P = 0.018) and to liver fat percent (r = 0.62, P = 0.0097, Fig. 3). Hepatic SCD1 activity index was positively related to the concentration of diacylglycerols (r = 0.71, P = 0.002) and TGs (r = 0.66, P = 0.005) but unrelated to those of ceramides (r = 0.07, NS) in the liver. The subjects (n = 3) with mild inflammatory or fibrotic changes fell on the same regression lines as others (data not shown).

Studies characterizing the human fatty liver are few because of methodological and ethical limitations to sample human liver tissue. In the present study, we analyzed the lipidome of the human liver of subjects with either normal or increased liver fat content due to nonalcoholic causes. We found that hepatic concentrations of diacylglycerols but not ceramides increase with increasing liver fat content. In addition, SCD1 activity index, as estimated from the product-to-precursor ratio, and lipogenic activities were increased, whereas long polyunsaturated fatty acids were depleted in fatty livers.

The final step in triacylglycerol synthesis is catalyzed by diacylglycerol acyltransferases (DGATs), which produce triacylglycerols from diacylglycerols (7). By definition, the concentrations of triacylglycerols in the human fatty liver are increased (1,10,33,34), which may explain the increase in their direct precursors, diacylglycerols (35). Consistent with this, the hepatic concentrations of diacylglycerols were positively related to those of triacylglycerols. In addition, we found hepatic diacylglycerol concentrations to be directly related to liver fat content, as determined by histology. Diacylglycerols are well-known allosteric activators of PKC, an enzyme which has been linked to insulin resistance in a variety of rodent models (36,37) as well as in the human liver (36). In mice, liver-specific overexpression of DGATs results in accumulation of triacylglycerols in the liver, whereas the concentrations of diacylglycerols and the activity of PKC remain unchanged (38). These mice do not exhibit any signs of hepatic insulin resistance (38). On the other hand, suppression of DGAT2 in mice decreases diacylglycerol concentrations and PKC activation and increases hepatic insulin sensitivity (39). These data together with the present findings would imply that diacylglycerols may contribute to hepatic insulin resistance, which is tightly related to liver fat content in humans (2,3,14).

Ceramides are sphingolipids that appear to mediate saturated fat–induced insulin resistance (6). Lipidomic analyses of livers of ob/ob mice have shown a strong association between hepatic ceramide content and the degree of steatosis (25). The relationships between hepatic ceramide concentrations and inflammatory changes were not analyzed in the latter study (25). In humans, we have previously found adipose tissue to be inflamed and contain more ceramides in subjects with high liver fat content compared with weight-matched subjects with low liver fat content without inflammation in adipose tissue (40). In the present study, we did not find hepatic ceramide concentrations to be related to liver fat content in human livers lacking significant inflammatory and fibrotic changes. These data do not exclude the possibility that ceramides contribute to hepatic insulin resistance in nonalcoholic steatohepatitis. Also, the present negative findings regarding ceramides need to be confirmed in larger groups of patients.

The concentrations of both diacylglycerols and ceramides in myocytes have been shown to increase as a consequence of silencing of SCD1 by siRNA (41). This raises the possibility that metabolism of these bioactive lipids may be regulated by SCD1. In the present study, we found the SCD1 activity index to be positively related to diacylglycerol but not ceramide concentrations. This is consistent with monounsaturated fatty acids, products of SCD1, being necessary for normal rates of synthesis of triacylglycerols (42,43). Hepatic RNA levels and activity of SCD1 are increased in obese (22) and lipoatrophic (44) mice with hepatic steatosis. In humans, SCD1 activity is considerable higher in skeletal muscle of obese compared with lean subjects (45). The present data suggest that, in the human liver, SCD1 activity increases with increasing liver fat content. Direct measurements of SCD1 activity would be helpful in this respect but could not be performed because of limited sample size.

It has recently been shown that short-term high-carbohydrate compared with high-fat feeding activates hepatic SCD1, as estimated from the composition of VLDL-TG fatty acids and de novo lipogenesis in healthy subjects with normal serum ALT concentrations (46). De novo lipogenesis is significantly increased in subjects with NAFLD compared with healthy controls (12,15,16), possibly as a consequence of an increase in sterol regulatory element–binding protein 1c (47–49), a key transcriptional activator of lipogenic genes, including SCD1 (50). In the present study, hepatic lipogenic index, estimated from the 16:0/18:2n-6 ratio (32), was positively related to liver fat content but not to SCD1 activity index. In contrast, the proportional amounts of long polyunsaturated fatty acids in the liver were strongly inversely related to hepatic SCD1 activity index and liver fat content. This is consistent with the ability of polyunsaturated fatty acids to suppress expression of both sterol regulatory element–binding protein 1c (51,52) and SCD1 (53) and activate genes involved in hepatic fatty acid oxidation (31). Subjects with NAFLD have been shown to consume less polyunsaturated fat than normal control subjects in some (54,55) but not all (56) cross-sectional studies.

Taken together, these data would suggest that depletion of long polyunsaturated fatty acids derived from essential fatty acids characterizes hepatic fat accumulation. These changes are associated with increases in both the hepatic lipogenic and SCD1 activity indexes. A recent pilot study suggested that prolonged supplementation of n-3 polyunsaturated fatty acids reduces liver fat in subjects with NAFLD (57), suggesting that polyunsaturated fatty acids could be beneficial in treating NAFLD.

This study was supported by research grants from the Academy of Finland (to H.Y.-J.), the Sigrid Juselius Foundation (to H.Y.-J.), the Finnish Diabetes Research Foundation (to A.K.), and the Biomedicum Helsinki Foundation (to A.K.). This work is part of the project “Hepatic and adipose tissue and functions in the metabolic syndrome” (www.hepadip.org), which is supported by the European Commission as an Integrated Project under the 6th Framework Programme (contract LSHM-CT-2005-018734) (to H.Y.-J. and M.O.).

This study also received grant support from Novo Nordisk Foundation (to H.Y.-J.). No other potential conflicts of interest relevant to this article were reported.

We gratefully acknowledge Mia Urjansson, Katja Tuominen, and Laxman Yetukuri for excellent technical assistance and the volunteers for their help.