Association Between SAH, an Acyl-CoA Synthetase Gene, and Hypertriglyceridemia, Obesity, and Hypertension
Background— The SA gene (SAH) has been isolated by differential screening from a genetically hypertensive rat strain as a candidate gene that may contribute to hypertension. Recently, the SA protein has been reported to be highly homologous to bovine xenobiotic–metabolizing medium-chain fatty acid:CoA ligase.
Methods and Results— To clarify the pathophysiological significance of SAH, we searched for polymorphisms of human SAH and performed association studies using a large cohort (4000 subjects) representing the general population in Japan. We found 2 polymorphisms in the promoter region and single-nucleotide polymorphisms in introns 5, 7, and 12 and exon 8. One of the variants, an A/G polymorphism in intron 12, just 7 bp upstream from exon 13, strongly affected plasma triglyceride, plasma cholesterol, body mass index (BMI), waist-to-hip ratio (W/H), and blood pressure status. The effect of this genotype on blood pressure seems to be conveyed through its effects on BMI and W/H. Transient expression of the SA protein in mammalian cells confirmed that it is expressed in mitochondria and has medium-chain fatty acid:CoA ligase activity. The A/G polymorphism was found to be associated with the expression level of SA mRNA in peripheral mononuclear cells in vivo.
Conclusions— The G allele of SAH was found to be associated with multiple risk factors, including hypertriglyceridemia, hypercholesterolemia, obesity, and hypertension. This observation should open a new area for future research in multiple-risk-factor syndromes.
Received September 26, 2001; revision received October 26, 2001; accepted October 29, 2001.
Interactions between genetic and environmental factors are thought to play important roles in the pathogenesis of common diseases. The use of association studies in large epidemiological cohorts with a large number of single-nucleotide polymorphisms throughout a single gene or throughout the genome is a new strategy for identifying genes that contribute to common diseases.1–3 In the present study, we applied this strategy to the SA gene (SAH) to examine whether it influences blood pressure.
SAH was isolated by differential screening from a genetically hypertensive rat strain, the spontaneously hypertensive rat.4 The expression of SAH in the kidneys of the spontaneously hypertensive rat is markedly higher than that in kidneys of a normotensive control strain, the Wistar-Kyoto rat. SAH is expressed mainly in proximal tubules and hepatocytes.5 Recently, SA protein has been reported to be significantly homologous to bovine xenobiotic–metabolizing medium-chain fatty acid (MCFA):CoA ligase.6 Analyses of several F2 rat cohorts7,8 and the establishment of several congenic rat strains9,10 have confirmed that the SA gene locus contributes to blood pressure regulation in the rat. Thus, SAH is a candidate gene for human essential hypertension.
Several small-scale association studies, however, have given conflicting results regarding whether SAH contributes to hypertension in humans.11,12 To clarify this issue, we thoroughly searched for polymorphisms of human SAH and performed association studies using a large cohort (4000 subjects) representing the general population in Japan.
The selection criteria and design of the Suita Study have been described previously.13 The genotype of SAH was determined in 4039 subjects (written informed consent was obtained).
The characteristics of the subjects analyzed in the present study are summarized in Table 1 according to the A/G genotype in intron 12. Hypertension was defined as systolic blood pressure ≥140 mm Hg, diastolic blood pressure ≥90 mm Hg, or the current use of antihypertensive medication. Total cholesterol and triglyceride levels were determined by enzymatic methods using kits (L-TC Wako, Wako Pure Chemical, and Clinimate TG-2, Daiichi Chemicals).
Genomic DNA from 32 subjects was used as a template for sequence analyses. The promoter region (up to −2.1 kb), exons 1 to 14, and their flanking regions were sequenced. The primer sequences will be provided on request. The polymorphisms were determined by use of the TaqMan system (PE Applied Biosystems) (Table 2).
Assessment of the Expression Level of SAH mRNA
The expression level of SAH mRNA was assessed by a competitive reverse transcription–polymerase chain reaction (RT-PCR) method.14 The cRNA, which lacks the region between nucleotide 1064 and 1074 (GenBank accession D16350), was synthesized. RNA was extracted as previously described14 from peripheral mononuclear cells purified by a Ficoll density gradient. Peripheral mononuclear cells were obtained from healthy doctors who understood the significance of this study (written informed consent was obtained). Total RNA (1 μg) combined with the deletion-mutated cRNA was reverse-transcribed, and the resultant cDNA mixture was amplified by primers covering the region between nucleotides 968 and 1111. The length of the PCR product from native mRNA was 144 bp, and that from the deletion-mutated cRNA was 133 bp. The expression level of mRNA is given as the ratio of the 144-bp PCR fragment to the 133-bp fragment.
The expression construct for human SAH was purchased from Invitrogen (GeneStorm expression-ready human clones). SA cDNA is expressed under the control of a cytomegalovirus (CMV) promoter. COS1 cells were transiently transfected by this expression vector by LipofectAmine Plus Reagent (Gibco-BRL). The transfected cells were suspended in buffer A (50 mmol/L Tris-HCl [pH 8.0], 1 mmol/L EDTA, 1 mmol/L dithiothreitol, and 10% glycerol) containing a protease inhibitor cocktail (Sigma) and disrupted by sonication. The resulting supernatant was used for enzymatic assay for acyl-CoA synthetase. Acyl-CoA synthetase activity for octanoic acid and palmitoic acid was assayed according to the method of Vessey and Hu.15
To determine the cellular localization of the SA protein, the cDNA encoding human SAH with the C-terminal Myc tag was subcloned into pCI mammalian expression vector (Promega) and expressed in HeLa cells. To detect the SA protein and mitochondria simultaneously, the cells were incubated with anti–Myc-tag rabbit polyclonal antibody (MBL) and anti–human mitochondria mouse monoclonal antibody (Chemicon). The cells were then double-stained with Alexa Fluor 488–labeled anti-rabbit IgG and Alexa Fluor 568-labeled anti-mouse IgG (Molecular Probes).
To explore the significance of an A/G polymorphism in intron 12, we constructed a minigene that included the region between exons 11 and 14 (3′-untranslated region) under the control of the CMV promoter (pcDNA 3.1 as a vector). The correctly spliced mature transcript (M transcript) was detected as a PCR product of 204 bp by use of exon 12 and 14 primers. The unspliced transcript was detected as a PCR product of 229 bp with intron 12 and exon 13/14 primers. The ratio of M transcript to the unspliced RNA was expressed as the ratio of 209-bp to 229-bp PCR product. To assess the expression levels of M transcript of the A and G alleles, pRL-CMV vector (Promega), in which Renilla luciferase is under the CMV promoter, was included in the transfection mixture as an internal standard. The expression level of M transcript was assessed by the ratio of the PCR product from M transcript (204 bp) to that from luciferase (283 bp).
To explore the regulatory effects of an insertion/deletion polymorphism in the promoter region, we constructed SAH promoter/luciferase fusion genes. The polymorphisms were an insertion/deletion (alleles I and D) at −1037 and a G(119952)A polymorphism at −407. The transcription initiation site was determined by 5′-RACE, and the major site was numbered +1. The haplotypes determined were D/G, D/A, I/A, and I/G. The promoter region between −2052 and +253 was subcloned into pGL2-Basic (Promega), which does not contain any promoter sequence. Transfection was performed in MDCK cells with PRL-CMV vector (Promega) as an internal standard. Photinus and Renilla luciferase activities were measured with a kit (PG-DUAL-SP, Toyo Ink, Co).
A more detailed description of materials and methods, including primer sequences, will be provided on request.
Values are expressed as mean±SEM or mean±SD. All statistical analyses were performed with the JMP and StatView statistical packages (SAS Institute Inc). Multiple linear regression and multiple logistic analyses were performed with other covariates (sex and age). Residuals of blood pressure values, waist-to-hip ratio (W/H), triglyceride, and cholesterol were calculated by adjustment for sex and age. Differences in numerical data among the groups were analyzed by 1-way/2-way ANOVA and the unpaired t test. Differences in frequencies and the degree of linkage disequilibrium were tested by contingency table analysis.
Polymorphisms found in SAH are summarized in Table 3. The genotypes in introns 5 and 7 were in complete linkage disequilibrium with the exon 8 polymorphism in the 96 subjects analyzed. Therefore, these 2 polymorphisms were not tested. The pairwise linkage disequilibrium of the I/D with the intron 12 polymorphism is shown in Table 4.
We determined the genotypes of the promoter I/D, promoter G/A, exon 8 G/C, and intron 12 A/G polymorphisms in the entire study population. Table 1 shows the characteristics of the study population according to the intron 12 polymorphism. Because there were only 4 GG genotypes, the GG and AG genotypes were combined into one group.
The intron 12 polymorphism significantly affected body mass index (BMI), W/H, percentage of antihypertensive treatment, systolic and diastolic blood pressure, heart rate, fasting blood glucose, and triglyceride (Table 1). After adjustment for age and sex, the intron 12 polymorphism significantly affected BMI, W/H, triglyceride, and systolic and diastolic blood pressure (Table 1). The D allele in the promoter tended to affect the triglyceride level. Multiple regression analysis indicated that the triglyceride level was determined by age (P<0.0001), sex (P<0.0001), and the genotype of the I/D polymorphism (II+ID=1, DD=2, P=0.0654). None of the other polymorphisms of SAH significantly affected these phenotypic variables (data not shown). Multiple logistic analysis, in which age, sex, and the intron 12 polymorphism were included as independent variables, indicated that the odds ratio of the AG+GG genotype for the presence of hypertension was 1.49 (95% CI 1.05 to 2.10, P=0.0235).
The effects of the intron 12 polymorphism on various phenotypes were more evident in subjects <60 years old (Table 5). After adjustment for age and sex, the intron 12 polymorphism significantly affected BMI, W/H, triglyceride, cholesterol, and systolic and diastolic blood pressure in this younger subpopulation (Table 5). Multiple logistic analysis, in which age, sex, and the intron 12 polymorphism were included as independent variables, indicated that the odds ratio of the AG+GG genotype for the presence of hypertension was 2.01 (95% CI 1.19 to 3.35, P=0.0081).
Function of the SA Protein
The SA protein has been reported to be highly homologous to a bovine xenobiotic–metabolizing MCFA:CoA ligase.6 COS1 cells transfected by pcDNA3.1/GS-human SA had significantly higher acyl-CoA synthetase activity for octanoic acid than those transfected with pcDNA3.1/GS, which confirmed that human SA protein had MCFA:CoA ligase activity (Figure 1).
Anti-mitochondria antibody showed a spaghetti-like staining pattern in HeLa cells. Anti-Myc-tag antibody showed a similar pattern in the transfected cells. We concluded that human SA protein is associated with mitochondria (data not shown).
Functional Significance of the Intron 12 A/G and Promoter I/D Polymorphisms
We assessed the expression levels of SA mRNA according to the genotype of intron 12 A/G and the promoter I/D polymorphisms (Figure 2). The expression level of SA mRNA in mononuclear cells in subjects with the AG genotype (all DD genotype in the promoter) (n=4) was ≈4 times higher than that in subjects with the AA genotype (1 II, 4 ID, and 3 DD genotypes in the promoter) (n=8) (P=0.0002). The I/D polymorphism did not appear to have any significant effects on the SA mRNA level (data not shown).
The above observation suggested that the intron 12 A/G polymorphism might influence the expression level of SA mRNA. Because the A/G polymorphism is in the polypyrimidine tract of the intron and this tract has been suggested to influence splicing,16 we examined the effects of this polymorphism on splicing efficiency, which might then affect the mRNA expression level.
A minigene that contained exons 11 to 14 under the control of the CMV promoter was constructed. The expression level of the correctly spliced M transcript compared with the control RNA level (Renilla luciferase RNA) was significantly higher in the G allele than in the A allele (Figure 3). The ratio of M transcript to the unspliced transcript was higher in the G allele than in the A allele (Figure 3A). This experiment indicates that this single nucleotide change significantly affected the expression level through affecting splicing of this intron under these experimental conditions. It was recently reported that an intronic sequence variation affects the mRNA expression level in WNK1, which causes pseudohypoaldosteronism type II.17
The functional significance of I/D polymorphism of SAH was assessed by a transient transfection assay with MDCK cells. Two-way ANOVA indicated that the I/D polymorphism, but not the G(119952)A polymorphism, significantly affected promoter activity. The promoter activity of the D allele was about twice that of the I allele in MDCK cells. Although this I/D polymorphism did not affect the mRNA level in peripheral mononuclear cells (see above), this polymorphism may have functional significance in other relevant tissues, such as kidney, liver, and adipose tissues.
In the present study, we found that human SA gene product has acyl-CoA synthetase activity for MCFA and that the A/G polymorphism in intron 12 affects BMI, W/H, triglyceride, cholesterol, and blood pressure status. Especially in subjects <60 years old, this genotype strongly affects triglyceride, cholesterol, and blood pressure. The clustering of multiple risk factors, including obesity, hypertension, dyslipidemia, and hyperuricemia, is known as syndrome X or insulin-resistance syndrome.18,19 These syndromes are likely to be heterogeneous, and SAH, an acyl-CoA synthetase, seems to contribute to a subset of these syndromes.
MCFAs are abundant in milk, coconut oil, and various semisynthetic oils.20,21 Activation of MCFA takes place mostly in the mitochondrial matrix by acyl-CoA synthetases for MCFA, and most of the MCFA incorporated into hepatocytes is subject to β-oxidation.20,21 Some of the acetyl-CoA produced during MCFA oxidation is directed toward ketone body production, and the rest is directed to de novo synthesis of long-chain fatty acids (LCFAs), which are then incorporated into triglyceride or other complex lipids.20,21
There are ≥5 genes in the SA gene family.22 Two (KS1 and KS2) are highly homologous to the entire SA protein (Figure 4). KS1 is localized ≈210 kb upstream of SAH. These acyl-CoA synthetases may be functionally linked to specific metabolic pathways, as observed in acyl-CoA synthetases for LCFA in the liver.23 A higher expression of the SA protein might lead to higher de novo synthesis of LCFA from MCFA, because the G allele was associated with a higher plasma triglyceride level. Preferential deposition of MCFA to triglyceride might lead to a higher accumulation of peripheral triglyceride (visceral obesity) in the long term. The association of the A/G polymorphism with plasma triglyceride was more evident in subjects <60 years old, which may indicate that the primary intermediate is hypertriglyceridemia. Further investigation is needed to clarify the precise mechanisms by which SAH influences the phenotypes observed in the present study.
The G allele in intron 12 corresponds predominantly to the D allele in the promoter (Table 4). The estimated haplotype frequencies for IA, IG, DA, and DG are 0.294, <0.01, 0.686, and 0.019. The residuals of plasma triglyceride after adjustment for age and sex (mean±SEM, mmol/L) were −0.047±0.021 (IA/IA), −0.031±0.025 (IA/DA), +0.028±0.022 (DA/DA), and +0.115±0.092 (DG/DA+DG/DG). The profound effects of the intron 12 polymorphism on various phenotypes may be due to additive or synergistic effects of the D allele in the promoter.
The amino acid change (K359N) in exon 8 may not have functional significance, because the other members of this family have Q at this position, which is homologous to N (Table 5). Moreover, this amino acid residue is outside the motifs of this family.24
The intron 12 A/G polymorphism affected blood pressure status. Multiple logistic analyses that included BMI and W/H in addition to the genotype, age, and sex as independent variables downplayed the importance of the genotype as a predictor (P=0.106 in the total population and P=0.0436 in subjects <60 years old). Therefore, the substantial effects of the genotype on blood pressure seem to be conveyed through its effects on BMI and W/H.
In conclusion, human SA, acyl-CoA synthetase for medium-chain fatty acid, contributes to multiple risk factors, including obesity, hypertriglyceridemia, hypercholesterolemia, and hypertension. This intriguing observation opens a new area for future research in multiple-risk-factor syndromes.
Iwai N, Inagami T. Isolation of preferentially expressed genes in the kidneys of hypertensive rats. Hypertension. 1991; 17: 161–169.
Yang T, Hassan SA, Singh I, et al. SA gene expression in the proximal tubule of normotensive and hypertensive rats. Hypertension. 1996; 27 (pt 2): 541–551.
Harris EL, Dene H, Rapp JP. SA gene and blood pressure cosegregation using Dahl salt-sensitive rats. Am J Hypertens. 1993; 6: 330–334.
Iwai N, Tsujita Y, Kinoshita M. Isolation of a chromosome 1 region that contributes to high blood pressure and salt sensitivity. Hypertension. 1998; 32: 636–638.
Frantz SA, Kaiser M, Gardiner SM, et al. Successful isolation of a rat chromosome 1 blood pressure quantitative trait locus in reciprocal congenic strains. Hypertension. 1998; 32: 639–646.
Iwai N, Ohmichi N, Hanai K, et al. Human SA gene locus as a candidate locus for essential hypertension. Hypertension. 1994; 23: 375–380.
Nabika T, Bonnardeaux A, James M, et al. Evaluation of the SA locus in human hypertension. Hypertension. 1995; 25: 6–13.
Iwai N, Baba S, Mannami T, et al. Association of sodium channel γ-subunit promoter variant with blood pressure. Hypertension. 2001; 38: 86–89.
Iwai N, Shimoike H, Kinoshita M. Cardiac renin-angiotensin system in the hypertrophied heart. Circulation. 1995; 92: 2690–2696.
Roscigno RF, Weiner M, Garcia-Blanco M. A mutational analysis of the polypyrimidine tract of introns: effects of sequence differences in pyrimidine tracts on splicing. J Biol Chem. 1993; 268: 11222–11229.
Wilson FH, Disse-Nicodeme S, Choate KA, et al. Human hypertension caused by mutations in WNK kinases. Science. 2001; 293: 1107–1111.
Reaven GM. Banting lecture 1988. Role of insulin resistance in human disease. Diabetes. 1988; 37: 1595–1607.
Bach AC, Ingenbleek Y, Frey A. The usefulness of dietary medium-chain triglyceride in body weight control: fact or fantasy? J Lipid Res. 1996; 37: 708–726.
Muoio DM, Lewin TM, Wiedmer P, Coleman RA. Acyl-CoAs are functionally channeled in liver: potential role of acyl-CoA synthetase. Am J Physiol. 2000; 279: E1366–E1373.
Steinberg SJ, Morgenthaler J, Heinzer AK, et al. Very long-chain acyl-CoA synthetases: human bubblegum represents a new family of proteins capable of activating very long-chain fatty acid. J Biol Chem. 2000; 275: 35162–35169.