DNA Polymorphisms in SREBF 1 and FASN Genes Affect Fatty Acid Composition in Korean Cattle ( Hanwoo )

Sterol regulatory element binding factor 1 (SREBF1) and fatty acid synthase (FASN) genes play an important role in the biosynthesis of fatty acids and cholesterol, and in lipid metabolism. This study used polymorphisms in the intron 5 of bovine SREBF1 and in the thioesterase (TE) domain of FASN genes to evaluate their associations with beef fatty acid composition. A previously identified 84-bp indel (L: insertion/long type and S: deletion/short type) of the SREBF1 gene in Korean cattle had significant associations with the concentration of stearic (C18:0), linoleic (C18:2) and polyunsaturated fatty acids (PUFA). The stearic acid concentration was 6.30% lower in the SS than the LL genotype (p<0.05), but the linoleic and PUFA contents were 11.06% and 12.20% higher in SS compared to LL (p<0.05). Based on the sequence analysis, five single nucleotide polymorphisms (SNPs) g.17924G>A, g.18043C>T, g.18440G>A, g.18529G>A and g.18663C>T in the TE domain of the FASN gene were identified among the different cattle breeds studied. Among these, only g.17924 G>A and g.18663C>T SNPs were segregating in the Hanwoo population. The g.17924G>A SNP is a non-synonymous mutation (thr2264ala) and was significantly associated with the contents of palmitic (C16:0) and oleic acid (C18:1). The oleic acid concentration was 3.18% and 2.79% higher in Hanwoo with the GG genotype than the AA and AG genotypes, respectively (p<0.05), whereas the GG genotype had 3.8% and 4.01% lower palmitic acid than in those cattle with genotype AA and AG, respectively (p<0.05). Tissue expression data showed that SREBFI and FASN genes were expressed in a variety of tissues though they were expressed preferentially in different muscle tissues. In conclusion, the 84-bp indel of SREBF1 and g.17924G>A SNP of the FASN gene can be used as DNA markers to select Hanwoo breeding stock for fatty acid composition. (


INTRODUCTION
In the beef industry, the amount and distribution of fat within muscle are directly associated with quality and value of the meat.Fatty acid composition of beef is of great interest because of its implications for human health.Saturated fatty acids (SFA) are the main fatty acids found in meat and about 80% of the fatty acid in beef is composed of palmitic, stearic and oleic acid and the remaining 20% is distributed among 30 different fatty acids (Whetsell et al., 2003).Among the SFA, lauric, myristic and palmitic acid are considered to have the most harmful cardiovascular effects (Mozaffarian et al., 2005), whereas stearic acid is not detrimental to human health (Bonanome and Grundy, 1988).On the contrary, monounsaturated (MUFA) and polyunsaturated (PUFA) fatty acids increase hepatic low density lipoprotein (LDL) receptor activity, thereby decreasing the circulating concentration of LDL-cholesterol (Woollett et al., 1992).In addition to marbling, the fatty acid composition is the main factor defining meat quality parameters such as texture and taste and can be improved by an increased ratio of MUFA to SFA (Yang et al., 1999).
During the past few decades, advances in molecular genetics have lead to the identification of genes or markers affecting meat quality traits.DNA polymorphisms in some candidate genes have been reported for their association with beef quality traits.Polymorphisms in μ-calpain, lysyl oxidase and calpastatin genes are associated with beef tenderness trait (Casas et al., 2006).The genes encoding leptin (Shin and Chung, 2007a); DGAT1 (Thaller et al., 2003); TG (Shin and Chung, 2007b) and GH1 (Barendse et al., 2006) have been associated with beef marbling.Recent studies have reported associations between fatty acid composition of beef and polymorphisms in candidate genes like SREBF1 (Hoashi et al., 2007) and FASN (Zhang et al., 2008).
SREBF1 is a member of the basic helix-loop-helixleucine zipper family of transcription factors that is involved in adipocyte differentiation, biosynthesis of cholesterol and fatty acids (Brown and Goldstein, 1997) as well as playing an essential role in energy homeostasis.SREBF1 is predominantly expressed in adipose tissue and liver (Kim and Spiegelman, 1996).Its expression has been observed in human and mouse muscles (Shimomura et al., 1997) and in a wide variety of chicken tissues (Assaf et al., 2003).FASN is a multifunctional enzyme complex that regulates de novo biosynthesis of long chain fatty acids (Roy et al., 2001).Bovine FASN gene has been mapped on BTA19 where several QTL affecting beef fatty acid composition (Zhang et al., 2008), adipose fat and milk fat content (Roy et al., 2006;Morris et al., 2007) were found.Whole genome shotgun sequence from NCBI shows that SREBF1 is also located on the same chromosome (http://www.ncbi.nlm.nih.gov/projects/mapview).The four exons (39 to 42) in the FASN complex that encode for the TE domain are responsible for fatty acid synthesis, mainly C16:0, by hydrolyzing the acyl-S-phosphopantetheine thioester.Therefore, the TE domain determines the product chain length of FASN and variation in the TE domain among individuals would be a candidate for heritable differences in fatty acid composition (Zhang et al., 2008).Bovine FASN expression was higher in brain, testis and adipose tissue than in liver and heart (Roy et al., 2005).Knockout or transgenic mice experiments demonstrated that SREBP transcription factors play key roles in the regulation of FASN transcription (Shimano et al., 1999).
An 84-bp indel in intron 5 of the SREBF1 gene had significant association with MUFA content in Japanese Black cattle (Hoashi et al., 2007) and polymorphisms in the TE domain of the FASN gene affect fatty acid composition in Angus cattle (Zhang et al., 2008).Kim et al. (2002) reported that SFA (C16:0 and C18:0) and MUFA (C16:1 and C18:1) comprised nearly 90% of Hanwoo muscle lipid and C18:1 concentration was nearly 90% of the total MUFA.However, to date, no information on SREBF1 and FASN gene polymorphisms in Hanwoo and their corresponding effects on fatty acid composition are available.Therefore, the objectives of this study were to investigate mRNA expression profiles of SREBF1 and FASN genes and to evaluate SNPs within SREBF1 and FASN genes for their association with fatty acid composition in Hanwoo.

Sampling and phenotypic data
The genomic DNA and meat samples were collected from the National Institute of Animal Science (NIAS), Suwon, Korea.DNA samples were obtained from 90 Hanwoo individuals having fatty acid measurement data and these animals were reared under a progeny testing program at NIAS.They were the half-sib progenies of 22 sires and the number of progeny varied from 2 to 7 per sire.Their feeding condition, concentration and forage intake and fattening period were controlled.Genomic DNA samples of Limousine, Angus, Simmental, Brahman and Red Chittagong (Bangladeshi native) cattle were obtained from the University of Adelaide, Australia and Bangladesh Agricultural University, Bangladesh and used to determine allele frequencies of SREBF1 and FASN gene polymorphisms.For cDNA synthesis, three different types of skeletal muscles; longissimus muscle (LM), round muscle (RM) and lean meat of short rib (SRM) and four organ tissue (heart, spleen, liver and lung) samples were also collected within 30 min of slaughter.The samples were then cut into small pieces and immediately transferred into liquid nitrogen and stored at -70°C until RNA preparation.Fatty acid components were measured from muscle tissues by a direct trans-esterification method according to Lepage and Roy (1986) using a gas chromatograph (GC; Hewlett Packard, model 5972).

SNP discovery and genotyping
Primers were constructed based on the sequences obtained from GenBank using Primer 3 software (Rozen and Skaletsky, 2000) to detect the 84 bp indel in intron 5 of SREBF1 (NC_007317) and TE domain sequence of FASN (AF285607) gene.The oligonucleotide sequences, annealing temperatures and PCR product sizes are presented in Table 1.Eighteen animals from five different cattle breeds were used for initial polymorphism screening in the FASN gene.The PCR amplification was performed using a GeneAmp 2700 (Applied Biosystems, USA) thermal cycler in a 25 μl reaction volume containing 50 ng of genomic DNA, 1×PCR gold buffer (50 mM KCl, 10 mM Tris-HCl, pH 8.3), 1.5 mM MgCl 2, 200 μM dNTPs, 0.4 pmol of each primer and 1 U Taq polymerase (Ampli Tag Gold TM , Applied Biosystems, USA).The thermal cycling conditions were as follows: 94°C for 10 min for initial denaturation, followed by 35 cycles of 30 sec at 94°C, 30 sec at 59-60°C, 30 sec at 72°C and a final extension step at 72°C for 10 min.For the SREBF1 gene, PCR products were cloned using the pGEM ® -T easy vector systems (Promega, USA).Recombinant plasmid DNAs were purified using a QIAprep miniprep (Qiagen, USA).For the FASN gene, PCR products were purified using an Accuprep ® PCR purification kit (Bioneer, Korea) and sequencing was performed on a 3100 automated DNA sequencer (Applied Biosystems, USA).The SNPs were discovered by comparing sequences of 18 individuals using Chromas ver.2.01 (www.technelysium.com.au);ClustalW (http://www.ebi.ac.uk/tools/clustalw);MEGA ver.4.0 (Tamura et al., 2007) and GENSCAN software (Burge and Karlin, 1997).The SNP genotyping was carried out in a total reaction volume of 20 μl using PCR-RFLP technique.The resulting fragments were analyzed in 3% agarose gel with ethidium bromide in 1×TBE buffer.An additional primer pair was designed for genotyping of the indel in intron 5 (Table 1).

Total RNA extraction and complementary DNA (cDNA) synthesis
A 250 mg sample of each tissue was ground in liquid nitrogen and total RNA was extracted from LM, RM, SRM, heart, liver, spleen and lung using RNeasy fibrous tissue and RNeasy midi kit (Qiagen, CA, USA).Following RNA isolation, the concentration and purity of the extracted RNA were quantified by spectrophotometer (Ultrospec 3100 pro , Biochrom Ltd., England) at 260/280 nm.Single-stranded cDNA was synthesized from 1 μg of total RNA by incubation at 42°C for 1 h using a reverse transcription system (Promega, USA).The reaction was performed by PTC-200 programmable thermocycler (MJ research, USA) in a 20 μl mixture consisting of 15 U avian myeloblastosis virus (AMV) reverse transcriptase, 20 U of RNasin ribonuclease inhibitor, 25 ng random primers, 5 mM MgCl 2 , 1 mM deoxynucleoside triphosphate (dNTP) mix and 1X reverse transcription buffer (10 mM Tris-HCl at pH 8.8), 50 mM KCl and 0.1% Triton ® X-100).The reaction was terminated by heating at 95°C for 5 min and rapidly cooling on ice.The reaction mixture was diluted 10 times and 2 μl of the diluted cDNA sample was used as a template for the mRNA expression experiment.

mRNA expression analysis by semi-quantitative real time PCR
Amplification of SREBF1 and FASN gene cDNA was carried out by semi-quantitative real time PCR (Bioneer corp., Korea) for mRNA expression analysis.The PCR mixture contained 0.5 U HS Prime Taq DNA polymerase, 2.5 mM MgCl 2 , 1.0 mM dNTPs, 1× SYBR green fluorescent dye, 1× buffer (10 mM Tris-HCl; pH 8.0, 50 mM KCl, 0.25 mM EDTA, 0.05 mM DTT, 0.25% Tween 20, 0.25% Nonidet P-40 and 25% glycerol) and 1.0 pmol of each primer in a 20 μl reaction volume.The amplification parameters consisted of initial denaturation at 94°C for 5 min, followed by 40 cycles at 94°C for 30 sec, 58-60°C for 30 sec and 72°C for 30 sec and a final extension at 72°C for 1 min.Each reaction was carried out in triplicate and each expression analysis was also performed in two independent experiments.For normalization of expression data, β-actin gene was used as an internal control.The normalized target gene expression level for each tissue was expressed in comparison to heart, chosen to be the reference tissue.The relative gene expression level was calculated by the ΔΔCt method (Livak and Schmittgen, 2001) using exicycler 3 software (Bioneer, Korea).

Statistical analysis
The genotype and allele frequencies were calculated according to Falconer and Mackay (1996).A Chi-square test was performed to test the Hardy-Weinberg equilibrium (HWE) at each locus by comparing expected and observed genotype frequencies (Falconer and Mackay, 1996).Associations between individual SNP and fatty acid measurements were evaluated using the mixed model of SAS for windows 9.1.3.The linear model used was as follows: Where Y ijk is the observation of the fatty acid measurement traits; μ is the overall mean, YS i is the effect of i th year and season of calving, G j is the fixed effect of j th genotype, b is the regression co-efficient for slaughter age in days, D is the slaughter age in days and e ijk is the random residual effect.In addition, sire is considered as a random effect.

Identification of sequence polymorphisms and genotyping
In the current study, SREBF1 and FASN genes were chosen based on their biological functions to evaluate association between polymorphisms and beef fatty acid composition.We identified one 84-bp indel and five SNPs in the SREBF1 and FASN genes, respectively.The details of all polymorphisms, mutation type and corresponding restriction enzymes are shown in Table 3.The 84-bp length polymorphism in intron 5 of the SREBF1 gene was detected directly from amplified products.Genotyping was carried out based on the different length of PCR products (Figure 1) and detected as S (deletion/short type, 363 bp) and L (insertion/long type, 447 bp) types.The genotype frequencies were 0.52 (LL type), 0.40 (LS type) and 0.08 (SS type), and estimated allele frequencies were 0.72 and 0.28 for L and S allele, respectively.We also investigated a total of 188 cattle from 5 different breeds to determine minor allele frequency and the estimated S allele frequencies were 0.22, 0.00, 0.00, 0.05 and 0.01 for Limousine (n = 36), Angus (n = 34), Simmental (n = 49), Brahman (n = 20) and Red Chittagong cattle (n = 49), respectively (data not shown).
Three sets of primers were used to amplify the TE domain of the FASN gene and the sequencing results revealed five nucleotide substitutions in different cattle breeds (Table 3); three in coding exons (g.17924 G>A, g.18440G>A and g.18663C>T) and two in introns (g.18043C>T and g.18529G>A).Among the identified SNPs, g.17924G>A and g.18440G>A were nonsynonymous type which replaced amino acid from alanine (GCC) to threonine (ACC), and glutamic acid (GAG) to lysine (AAG), respectively.In our study, a Bos indicus specific novel mutation 18440G>A was identified and was found only in Brahman and Red Chittagong cattle (Figure 2B).The SNP was genotyped using restriction enzyme Hpy188III, with A allele resulting in 247/222,175 and 85, 47 bp bands of the 554 bp amplicon and the G allele resulting in 247, 222 and 85 bp bands.The mutant A allele frequency was 0.93 (20) and 0.84 (46), respectively, in Brahman and Red Chittagong cattle, whereas this allele frequency was 0 in 112 Bos taurus individuals genotyped from Hanwoo, Angus, Hereford, Simmental and Shorthorn breeds.This missense mutation might play a role for beef fatty acid content differentiation between zebu and taurine cattle, and could be used as a marker with further investigation for breed discrimination.However, we  identified only g.17924 G>A and g.18663C>T polymorphisms in the TE domain of Hanwoo cattle, whereas the later polymorphism had no restriction site.The genotypes of g.17924G>A SNP were determined by the PCR-RFLP method using MscI restriction enzyme (Figure 2A).The G allele represented two fragments of 362 and 262 bp, while the A allele showed three fragments of 262, 195 and 167 bp.The frequencies of GG, GA and AA genotypes were 0.73, 0.22 and 0.04, respectively.The G and A allele frequencies were 0.84 and 0.16, respectively (Table 4).The genotype frequencies were in agreement with HWE for each polymorphism (p>0.05).

Association analysis with fatty acid composition
Among the identified polymorphisms, the 84 bp length polymorphisms in SREBF1 and g.17924G>A SNP in the FASN gene were investigated for association with beef fatty acid composition.The fatty acid concentration profiles for SREBF1 and FASN genes are shown in Table 5 and 6, respectively.The SREBF1 genotypes were associated with the concentration of stearic (C18:0), linoleic (C18:2) and PUFA in Hanwoo (Table 5).The C18:2 and PUFA contents were 11.06% and 12.20% higher in the SS genotype as compared to LL (p<0.05), while the concentration of C18:0 was 6.30% lower in SS than LL genotype (p<0.05).Moreover, no significant differences were observed between SS and LS genotypes for C18:0, C18:2 and PUFA contents.The γ-linoleic acid content also tended to be greater (p = 0.08) in the SS genotype than in LL and LS genotypes.However, no significant association was found between the SREBF1 genotypes and total SFA, UFA and MUFA content.In the FASN gene, significant association was observed between g.17924G>A SNP genotypes and the concentrations of palmitic (C16:0) and oleic acid (C18:1) (Table 6).The C18:1 concentration was 3.18% and 2.79% higher in Hanwoo with GG genotype than in AA and AG genotypes, respectively (p<0.05).In addition, cattle with genotype GG had 3.81% and 4.01% lower C16:0 than in those with genotype AA and AG, respectively (p<0.05).Total UFA tended to be greater (p = 0.08) and a lower trend for SFA concentrations was also observed in the GG genotype than in AG and AA genotypes (p = 0.08).However, no significant association was observed between g.17924A>G genotypes and other considered fatty acid concentrations.

Expression analysis of SREBF1 and FASN genes
The mRNA expression profiles of bovine SREBF1 and FASN genes were investigated using semi-quantitative real time PCR (qPCR) from LM, RM, SRM, heart, liver, spleen and lung samples.The oligonucleotides used in qPCR are shown in Table 2. Results of expression patterns in seven different tissues revealed that SREBF1 and FASN genes were ubiquitously expressed.SREBF1 gene was highly expressed in different muscle tissues, while moderate expression was observed in heart and liver, and lower expression in spleen and lung (Figure 3A).For the FASN gene, the highest mRNA expression was found in longissimus muscle, with lower expression in round muscle, short rib muscle, heart, liver and lung, and expression was lowest in spleen (Figure 3B).

DISCUSSION
SREBF1 belongs to the family of basic helix-loop-helixleucine zipper transcription factors that bind to the sterol regulatory element DNA sequence TCACNCCAC and acts as a key regulatory element for fatty acid biosynthesis and cholesterol homeostasis (Felder et al., 2005).Like other mammals, there were no splice variants detected in the bovine SREBF1 gene and it has the most similarity with the human SREBF1a sequence (Hoashi et al., 2007).Human SREBF1a is a more potential transcriptional activator than the other isoform SREBF1c, due to its longer NH2-terminal trans-activation domain (Shimano et al., 1997).The 84 bp indel in Hanwoo illustrated three different genotypes which are in accordance with previous results reported by Hoashi et al. (2007) in Japanese Black cattle.Furthermore, this length polymorphism was only found in the Limousine breed among the five different breeds studied suggesting that it might be a breed specific phenomenon and may not be segregating widely in different cattle populations.Our study demonstrated that the length polymorphism in the SREBF1 gene is significantly associated with the content of several beef fatty acids.The SS genotype contributes to higher unsaturated fatty acid but lower saturated fatty acid content in meat than the LL and LS genotypes.This agrees with the previous findings of Hoashi et al. (2007).They found that the SS type contributed to 1.3% higher MUFA proportion in intramuscular fat (p<0.05).Two silent mutations in the SREBF1 gene were significantly associated (p<0.05) with intramuscular fat content in pigs (Chen et al., 2008) which supports our present findings.Recent studies showed that SNPs in the intron region had significant association with marbling score (Cheong et al., 2007), carcass and meat quality traits (Di Stasio et al., 2005;Shin and Chung, 2007c).Although the 84 bp indel of SREBF1 gene is not located in the coding sequence, it is possible that this length polymorphism may affect mRNA expression level/translation efficiency and thereby indirectly contributes to fat quality characteristics (Kennes et al., 2001).
FASN is an enzymatic system composed of a 272 kDa multifunctional protein that is involved in the synthesis of fatty acids and plays a central role in de novo lipogenesis in mammals.Recently, a significant QTL for fatty acid composition was mapped on BTA19 and a candidate gene in this region, FASN, was identified (Roy et al., 2001;Morris et al., 2007).In the current study, we focused on the polymorphisms in the TE domain of the FASN gene and five SNPs were identified.Among them, two SNPs g.17924G>A and g.18663C>T were reported previously by Zhang et al. (2008) in Angus cattle and the former one was also reported in Jersey and Limousine cattle breeds by Morris et al. (2007).We independently identified 3 other SNPs g.18043C>T, g.18440G>A and g.18529G>A in our study.We found that SNP g.17924G>A was significantly associated with C16:0 and C18:1 which accords with the previous findings of Morris et al. (2007) and Zhang et al. (2008).They reported that MUFA and SFA contents had significant association with g.17924G>A polymorphism and the GG genotype contained higher amount of UFA and relatively lower SFA content than AG and AA genotypes.In another investigation, polymorphisms g.763G>C and g.16009A>G associated with milk fat content in a Holstein-Friesian population was reported by Roy et al. (2006) which supports our study.It is mentioned here that the favorable GG genotype frequency was only 13% in the Angus population (Zhang et al., 2008) whereas the estimated frequency for this genotype was 73% in the Hanwoo population.This might have occurred naturally in the Hanwoo population through long term selective breeding for other associated traits.The product of mammalian FASN is mainly C16:0, with minor amounts of C14:0 and it also contributes to the contents of other fatty acids.In the human, the TE domain has a hydrophobic groove that constitutes the fatty acyl substrate binding site (Chakravarty et al., 2004) with high specific activity towards C16-acyl ACP, but not C14-acyl ACP.The amino acid substitution in the TE domain predicted by the SNP g.17924A>G may influence the structure of the substrate binding site and consequently affect the specific activity of the TE domain (Zhang et al., 2008).We observed a higher mRNA expression in three different muscles for both SREBF1 and FASN genes which indicates the potential influence of these genes on cattle muscle compared with the other tissues studied.In the human, SREBF1 isoforms were expressed in all tissues studied, but higher expression was found in adipose tissue, brain, liver and testis (Felder et al., 2005).SREBF1 is predominantly expressed in liver and adipose tissue of the mouse (Kim and Spiegelman, 1996).The SREBF1 was expressed preferentially in the liver and uropygial gland in chicken compared with other tissues examined, adipose tissue, lung, kidney, intestine, muscle, brain and testis (Assaf et al., 2003).This discrepancy might be due to physiological differences between the two species.Roy et al. (2005) reported that FASN expression was higher in bovine brain, testis and adipose tissue while lower expression was observed in liver and heart.However, we did not include adipose tissue samples in our experiment.In our results, FASN gene had the highest expression in LM among the three different muscles which supports that this gene may contribute remarkably to adipogenesis in LM.
In conclusion, we identified an 84-bp indel in SREBF1 and two SNPs in the FASN gene in Korean Hanwoo cattle.We found that the polymorphisms in these genes had significant association with composition of several fatty acids of beef.The exact molecular and physiological mechanisms underlying the association of SNP with fatty acid content reported in this study are unknown.The use of these polymorphisms as genetic markers can be a useful tool for selection of animals and improve meat quality in Korean cattle.

Figure 1 .
Figure 1.Genotyping of SREBF1 84-bp indel polymorphism in intron 5.The arrows show the DNA fragment size (bp) for L and S genotypes which were fractionated by 2% agarose gel.

Figure 2 .
Figure 2. Genotyping of the g.17924G>A (A) and g.18440G>A (B) polymorphisms in the FASN gene.SNP positions are shown as shaded letters in chromatograms.The arrowheads show the size of DNA fragment (bp) and M is the 100 bp size marker.

Figure 3 .
Figure 3. Analysis of mRNA expression patterns in different tissues using quantitative real time PCR for SREBF1 (A) and FASN (B) genes.The numbers on the Y axis indicate fold differences.The obtained gene expression values of each gene were normalized to the expression level of β-actin as controls for the same sample.Each reaction was carried out in triplicate and experiments were repeated at least two times.Bars indicate standard deviation.

Table 1 .
Primer sequences, annealing temperatures and sizes of PCR products for SERBF1 and FASN genes

Table 2 .
Oligonucleotide sequence information and PCR conditions for quantitative real time PCR

Table 3 .
Polymorphisms detected in SREBF1 and FASN genes from five cattle breeds a Represent position of each SNP relative to the published cattle sequence as it appears in GenBank.

Table 4 .
Genotype and allele frequencies of SREBF1 and FASN genes in Hanwoo Falconer and Mackay (1996)um test for each locus according toFalconer and Mackay (1996).N denotes total number of animals investigated and values in the parentheses indicate the number of observation in each genotype.

Table 5 .
Effect of 84-bp indel in intron 5 of SREBF1 gene on fatty acid composition in Hanwoo 1

Table 6 .
Effect of g.17924A>G mutation of FASN gene on beef fatty acid composition in Hanwoo 1