FASN, SCD1 and ANXA9 gene polymorphism as genetic predictors of the fatty acid profile of sheep milk

In this study, single nucleotide polymorphisms (SNPs) in the ANXA9 (annexin 9), FASN (fatty acid synthase) and SCD1 (stearoyl-CoA desaturase 1) genes were analyzed as factors influencing fatty acid profiles in milk from Zošľachtená valaška sheep. SNP in selected genes was identified using polymerase chain reaction (PCR) and restriction fragment length polymorphism (PCR–RFLP). The long-chain fatty acids profile in sheep milk was identified by gas chromatography. Statistical analysis of the SCD1/Cfr13I polymorphism showed that the milk of the homozygous AA animals was characterized by a lower (P < 0.05) share of C4:0, C6:0, C8:0, C10:0, C12:0, C14:0 in comparison to the homozygous CC sheep. The milk of heterozygous sheep was characterized by a higher (P < 0.05) proportion of C13:0 acid compared to the milk of sheep with the homozygous AA type. A higher (P < 0.05) level of saturated fatty acids (SFA) was found in the milk of CC genotype sheep compared to the AA genotype. Our results lead to the conclusion that the greatest changes were observed for the SCD1/Cfr13I polymorphism and the least significant ones for FASN/AciI. Moreover, it is the first evidence that milk from sheep with SCD1/Cfr13I polymorphism and the homozygous AA genotype showed the most desirable fatty acids profile.

Milk and blood sampling. The material for the study was collected from 50 ewes in the similar phase of milking (25-30 days of lactation) and lactation (1st and 2nd lactation). In the lambing period, the sheep were kept in sheepfold complying with the European Union Directive (No. 116, item 778, 2010) and were fed hay ad libitum, 250 g/ewe/day of wheat middlings, and 3 kg/ewe/day of haylage. To collect milk samples from ewes, their lambs were separated overnight. The milk was collected into sterile containers and transported to the laboratory at 4 °C. Besides, peripheral blood samples were collected from the external jugular vein (EJV) of ewes into test tubes containing anticoagulant (tripotassium ethylenediaminetetraacetic acid, K 3 EDTA) for DNA isolation and immediately transported to the laboratory, and frozen at − 20 °C for subsequent analysis.

Genetic polymorphism analysis. DNA isolation was performed using the MasterPure DNA Purification
Kit for Blood, Version II (Lucigen, Middleton, WI, USA) according to the manufacturer's instructions. Animal genotyping was performed using PCR-RFLP. Three SNPs in the ANXA9 gene (intron 4, intron 5, GenBank: AY785286.1) and one in the FASN gene (exon 32, GenBank: GQ150557.1) and SCD1 (promoter region, Gen-Bank: FJ513370.1) were analyzed. Table 1 shows the location of individual SNPs and the appropriate primers, designed using the Primer3 software (http:// bioin fo. ut. ee/ prime r3-0. 4.0/), enabling the amplification of selected fragments of the analyzed genes. In the case of the reverse primer for the FASN gene, a mismatched nucleotide was introduced to create a cleavage site for the enzyme (this nucleotide is underlined in Table 1). The polymerase chain reaction (PCR) reaction was performed in a final volume of 25 μl, using 2 × PCR Master Mix (A&A Biotechnology, Gdynia, Poland),) containing 50 ng genomic DNA and 5 pmol of each primer. DNA amplification was performed using an initial denaturation at 94 °C for 5 min, followed by 30 cycles of 30 s denaturation at 94 °C each, annealing at a temperature appropriate for each gene for 30 s and extension at 72 °C for 30 s, ending with a final extension of 8 min at 72 °C (Table 1).
After the amplification of selected fragments of SCD1, FASN and ANXA9 genes, the identification of polymorphic loci in these genes was carried out using appropriately selected restriction enzymes. Individual PCR products were digested separately with restriction enzymes according to the manufacturer's recommendations.
Next, the restriction fragments were separated on agarose gels with appropriately selected agarose concentrations. The restriction enzymes and the obtained restriction fragments for individual genotypes are presented in Table 2.
Fatty acids analysis. Fats were extracted from the milk samples using the Folch method 23 . Next, the obtained fat was converted to fatty acid methyl esters using the Christopherson and Glass procedure (1969) 24 with 2 M KOH in methanol. The fatty acid profile was determined, by using a gas chromatography method (Agilent Technologies 7890A, Agilent Technologies, Santa Clara, CA, USA), with a flame ionization detector and an HP-88 capillary column designed for the separation of fatty acid methyl esters (FAMEs) (100 m length, 25 mm i.d. × 0.20 μm). The initial oven temperature was 50 °C and was increased by 3 °C/min to 220 °C. The detector and dispenser temperatures were − 270 °C and 270 respectively.
To analyse experimental chromatograms a comparative analysis of the retention times of the fatty acid methyl ester standards (Sigma-Aldrich) was performed using the ChemStation software (Agilent Technologies, USA).
The desaturase index was also calculated of fatty acids as a ratio of unsaturated cis-9 to unsaturated + saturated cis-9 for different FAs 25,27 as follows.
Statistical analysis. The frequencies of genotypes and alleles and the Hardy-Weinberg equilibrium for individual SNPs were calculated using the POPGENE software 27 , the effective number of alleles (Ne) was evaluated according to Kimura and Crow (1964) 28 and expected heterozygosity (He) and the polymorphism information content (PIC) were evaluated according to Nei's methods 29 .
The statistical analysis of the influence of selected SNPs on the fatty acid profile in sheep milk was carried out in the Statistica 13.1 program (StatSoft Poland, Krakow, Poland). The results of the study were statistically analyzed using one-way ANOVA followed by a multiple comparisons Tukey Post-Hoc Test. Pearson correlation coefficient (r) with a two-tailed test of significance was conducted to examine the relationship between certain parameters.
The statistical analysis was performed using the following model: where yij-analysed trait, µ-overall mean, ai-the effect of genotype on trait value, and eij-the effect of random error.
The correlation coefficients have been calculated with the following formula:

Results
The analysis of the obtained genotyping results suggested that for almost all polymorphic loci tested, the presence of all three possible genotypes was identified; only in the FASN gene, two out of three possible genotypes were identified. Table 3 shows the frequencies of genotypes and alleles and the expected heterozygosity, the effective number of alleles, PIC, and the value of χ2 calculated based on genotyping results. The expected heterozygosity for almost all examined loci was quite similar and ranged from 0.449 to 0.497; the only exception was the FASN polymorphism where two out of three genotypes were identified. In the case of the effective number of alleles, similar values were observed for most of the analyzed SNPs in the range of 1.814-1.987, except for the SCD1 genes, where this value was 1.150. The PIC value for all tested SNPs was similar and ranged from 0.311 to 0.373, that according to the classification of Botstein et al. (1980) 30 indicates an average polymorphism (0.25 < PIC value < 0.5). The next analyzed parameter was the Hardy-Weinberg equilibrium value (HWE). The distribution of genotypes consistent with the HWE was at P > 0.05. The analysis of the results collected in Table 3 shows that the ANXA9/HinfI polymorphism was found to be incompatible with the Hardy-Weinberg equilibrium. We also assessed the effects of SNPs within the fatty acid synthase gene (FASN), however, we detected only two genotypes (CC, CT) out of 3 FASN/AciI polymorphisms (without TT). Due to this, we did not calculate HWE for FASN/AciI.
The profile of individual saturated fatty acids in sheep milk in relation to individual genotypes of the studied polymorphisms of the SCD1 and FASN genes is presented in Table 4. Statistical analysis for the SCD1/Cfr13I polymorphism showed that the milk of individuals with the homozygous AA genotype was characterized by a lower (P < 0.05) share of butanoic acid (C4:0), hexanoic acid (C6:0), octanoic acid (C8:0), decanoic acid (C10:0), dodecane (C12:0), tetradecanoate (C14:0) than the homozygous CC sheep. On the other hand, the milk of sheep with the heterozygous genotype was characterized by a higher (P < 0.05) share of tridecanoic acid (C13:0) as compared to the milk of sheep with the AA genotype. A higher (P < 0.05) total level of saturated fatty acids (SFA) was found in the milk of the homozygous CC sheep compared to those of the homozygous AA genotype. Table 5 presents the share of unsaturated fatty acids in sheep milk in relation to the studied polymorphisms of the SCD1 and FASN genes. The analysis of the SCD1/Cfr13I polymorphism revealed that the proportion of (Z)-11-eicosenoic acid (C20:1) in milk was lower (P < 0.05) in the homozygous AA sheep in relation to the heterozygous and homozygous CC individuals. Higher (P < 0.05) levels of MUFA were noted in the homozygous AA individuals compared to the homozygous CC individuals. The relationship for all cis-5,8,11,14,17eicosapentaenoic acid (C20:5n3) was reverse: in the milk of the homozygous AA sheep the level of this acid was lower (P < 0.05) compared to individuals with the CC genotype. The polymorphism in the FASN gene will not statistically affect the fatty acid profile in sheep milk. For the remaining saturated and saturated fatty acids in milk, no statistical effect of polymorphisms for the SCD1 gene was observed. Polymorphism in the FASN, SDC and ANXA9 genes did not affect the desaturase index in sheep milk (Tables 5 and 7).
The analysis of the share of short-and long-chain saturated fatty acids with respect to the individual genotypes of the studied polymorphisms in the ANXA9 gene is presented in Table 6. Statistical analysis for the ANXA9/HinfI polymorphism showed that the milk of individuals with the homozygous CC genotype was characterized by a higher (P < 0.05) content of pentadecanoic acid (C15:0) and eicosanoic acid (C20:0) than the milk of heterozygous sheep. In contrast, the milk of sheep with the CG genotype was characterized by a lower (P < 0.05) proportion of octadecanoic acid (C18:0) compared to the milk of sheep with the homozygous GG genotype. For the remaining saturated fatty acids in milk, no statistical effect of ANXA9 polymorphisms was observed.      Table 7. shows the contribution of UFA in Zošľachtená valaška milk with respect to individual genotypes of the ANXA9 polymorphisms examined in this study. Analysis of the ANXA9/NlaIII polymorphism showed that the share of (all-Z) -5,8,11,14-eicosatetraenoic acid (20:4n6) in milk was lower (P < 0.01) in the sheep with the homozygous AA genotype in relation to individuals with heterozygous and homozygous GG genotypes. In the case of the ANXA9/HinfI polymorphism, it was shown that the milk of sheep with the homozygous GG genotype was characterized by a higher proportion of trans-octadecenoic acid (C18:1n9t) compared to the homozygous CC (P < 0.01) and heterozygous (P < 0.05) sheep. Also, a higher (P < 0.05) share of CLA was found in the homozygous GG individuals compared to heterozygous ones. For the ANXA9/Tru1I polymorphism, it was found that animals with the homozygous AA genotype were characterized by a lower (P < 0.05) proportion of C18:1n9t acid in milk compared to heterozygous sheep and higher (P < 0.05) compared to the homozygous CC animals. The polymorphism in the ANXA9 gene did not statistically affect the share of other unsaturated fatty acids in sheep milk.
Positive genetic correlations were observed between MUFA and UFA and SCD1/Cfr13I, although negative correlations for C20:5n3 was observed (Table 9). We also showed a positive correlation between C20:4n6 and ANXA9/NlaIII.

Discussion
Milk fat and fatty acid (FA) profiles affect milk and cheese quality. Some previous studies 13,31,32 , shown some genes putatively affecting the milk fat and FA content included FASN and the SCD1 genes. FAS catalyzes the de novo synthesis of small-to medium-chain FAs (Crisà et al., 2010), while the SCD1 gene affects the milk FA profile and is involved in the synthesis of monounsaturated FA by introducing a double bond in the delta-9 position of C14:0, C16:0 and C18:0 13,32 . However, the role of ANXA9 in lactation is unknown 33 .
Sheep milk is characterized by high variability in the level of fatty acids [34][35][36] . The level of saturated fatty acids (SFA) ranges from 49.43 to 82.9 g/100 g fat. On the other hand, the share of MUFA and polyunsaturated fatty acids (PUFA) is slightly lower (MUFA: 11.95-45.26 g/100 g fat, PUFA 2.79-12.24 g/100 g fat) 35 . The results obtained in the author's research correspond with the literature data.
Because the biohydrogenation process takes place in the rumen content, nutrition is the main factor determining the fatty acid profile in the milk of ruminants 14 . However, the level of fatty acids in sheep milk shows a low to moderate genetic variation for individual acids (ranging from 0.01 to 0.47) 35 . Therefore, nutrition and genetics play a key role in modulating the composition of milk fat 34 . Moreover, due to the crucial role of the FASN and SCD1 genes in lipid metabolism and/or their location in chromosomal regions associated with milk content and quality, SNP polymorphisms within these genes allow partially explain the variation of FA composition in sheep milk.
In the case of polymorphism in the SCD1 and ANXA9 genes, the analyzed SNPs are mapped in the noncoding parts, i.e. the promoter region and introns, respectively. Mutations in regulatory regions such as promoters or enhancers can disrupt or create new transcription factor binding sites and cause changes in transcription regulation. Mutations in untranslated regions may affect the regulation of translation or modify the microRNA Table 8. Correlation coefficients between the polymorphism in genes and the content of individual saturated fatty acids (SFA) in sheep of milk. SCFA short-chain fatty acids, LCFA long-chain fatty acids, SFA saturated fatty acids; (g/100 g of fat). *Significance at P < 0.05 was marked by an asterisk. www.nature.com/scientificreports/ binding sites and thus affect mRNA stability 37 . The polymorphism analyzed in the FASN gene is an exon mapped change and it is a synonymous mutation. In our work, it was not investigated whether the analyzed mutations in the FASN gene directly influenced the expression of this gene. In contrast, it is generally believed that the amino acid sequence of proteins determines the expression, folding, and function of a protein, while mutations that alter the basic structure of a protein may affect these properties 38 . In general, silent mutations can modify all phases of the gene expression process, resulting in the amplification or reduction of proteins concentration. Therefore, while most silent mutations do not alter protein functionality, they can dramatically alter protein abundance 37 . In many cases, the effect of the SNP polymorphism rather regulate gene expression than changing the amino acid sequence as was shown by Knutsen et al. in bovine 39 . Intronic mutations in the ANXA9 gene are not located at sites involved in splicing and are not conserved. In the case of polymorphism in the promoter region of the SCD1 gene, Garciá-Fernández et al. (2009) 40 , based on the characteristics of the promoter of the bovine SCD1 gene, report that the 31 C > A polymorphism is located between two conserved regions that include the critical binding sites of the transcription factor. Importantly, the second conserved promoter region is the critical region for the expression of the SCD-transcriptional enhancer element (STE). This region, which is 109 bp downstream of the polymorphism understudy, contains the sterol response element-binding protein (SREBP) and PUFA response element and plays a key role in the inhibitory effect of CLA and oleic acid on SCD transcription 40 .
Stearic-CoA desaturase, also known as Δ9-desaturase, is an enzyme [EC 1.14.19.1] associated with an endoplasmic reticulum (ER) that catalyzes the formation of monounsaturated fatty acids (MUFAs) from de novosynthesized or food-supplied saturated fatty acids (SFAs) 41 . It is a rate-limiting enzyme in monounsaturated fatty acid (MUFA) biosynthesis that introduces a cis double bond between carbons 9 and 10 in the saturated fatty acid spectrum, with a preference for C16: 0 and C18: 0. This enzyme plays a key role in lipid metabolism and the maintenance of membrane fluidity, based on the physiological importance of the ratio between saturated and monounsaturated fatty acids 40 44 . However, it is the T allele in FASN that is responsible, at least in part, for a higher level of fatty acids in milk 13. In our research, the FASN polymorphism affected no changes in the fatty acid composition in milk, which may be related to the low share of the T allele in the sheep genotype. The study performed on sheep Table 9. Correlation coefficients between the polymorphism in genes and the content of individual unsaturated fatty acids (UFA) in the milk of sheep. MUFA monounsaturated fatty acids, PUFA Polyunsaturated fatty acids; (g/100 g of fat). *Significance at P < 0.05 was marked by an asterisk.  (FASN) contributes to influencing the medium-and long-chain FAs (C5:0 and C15:0), and a tendency was also observed for association with C18:1n9c and C18:3n6 45 . Studies by other authors have shown a link between polymorphisms (other than those studied in the present work) of the SCD1 gene and the composition of fatty acids in ruminant products, including sheep milk. They found that the analyzed SNPs in SCD1 significantly influenced the level of C16:1 acid, C18:1 trans-11 acid, the content of SFA and MUFA in sheep milk 46 . Moreover, Gu et al. (2019) 47 analyzed the polymorphism in the promoter region of the SCD1 gene (g.133A > C) showed that the presence of the C allele results in higher levels of MUFA and lower levels of SFA. In our study we found, that the milk of the homozygous CC sheep contains a higher level of C20:1 and C20:5n3 fatty acids than the milk of sheep with AA genotype.
In the case of SFA and MUFA, a higher share of MUFA and a lower share of SFA were obtained in the milk of the homozygous AA sheep. The presence of features linked to polymorphism is related, among others, to the specificity of population and animal species 48 , which may explain the obtained results.
There are only a few studies in the available literature describing the effect of SCD1 46 and ANXA9 polymorphisms on the composition of sheep milk, including the fatty acid profile. The polymorphism in the ANXA9 gene affects the yield of milk fat in cows 33 and the level of fat in sheep milk 8 . Our research revealed the effect of SNP in the ANXA9 gene on the share of the following acids: C15:0; C18:0; C20:0 and C18:19t; C20:4n6; and CLA.
Furthermore, Carta et al. (2008) 49 performed a genome-wide scan for loci associated with FA composition in sheep's milk, where the most significant QTL that affects the fatty acid composition (chromosome-wise thresholds) associated with FASN were detected on OAR11 (C14:0 and C16:0) and OAR6 (MUFA). Our analysis of QTL located next to the studied genes by using Sheep QTLdb (https:// www. anima lgeno me. org/ cgi-bin/ QTLdb/ index) showed additional QTL on OAR11 (cis-10-C17:1; QTL 14,223; and C10:0; QTL 14,220) 50 . Analysis performed for SCD1 showed QTL on OAR22 that affect C4:0 (QTL:13,868), C16:1 (QTL:13,869;), C18:1 (QTL:13,870), C18:2(n-6) (QTL:13,871), and MUFA (QTL:13,877) content in milk. The QTL related to the ANXA9 gene and FA composition in milk has not been demonstrated in sheep. However, Calvo et al. suggested that ANXA9 may be a candidate gene for milk production traits, as well as for SCC in cattle, concerning its localisation in the proximity of QTL for these traits 16 . This was also confirmed by Martinez et al. (2010) in a Spanish Holstein-Friesian population 33 . Furthermore, Kulig et al. (2015) have shown that there are associations between the ANXA9 polymorphism and SCC in the milk of Jersey cows 51 .
In the human diet, foods with a low n-6/n-3 share, ranging from 1:1 to 4:1, are desirable 52 . Maintaining a sufficiently low n-6/n-3 ratio in the diet has a positive effect on the cognitive functions of the body and reduces the risk of depression 53 . Additionally, the high consumption rate of (n-3)/(n-6) PUFAs reduces the risk of neoplastic diseases and inflammations 9,52,54 .
In the case of the SCD1/Cfr13I polymorphism, an increase in C20:5n3 was observed in milk of the sheep with homozygous CC genotypes, without changes in the level of n-6 acids, which may cause an increase in the value of (n-3)/(n-6). In the case of ANXA9/NlaIII polymorphism analysis for AA homozygous animals, a decrease in the C20:4n6 acid level was found, which can also be considered a favourable phenomenon. Supplementation of CLA in the diet of animals and humans improves the metabolism of glucose and lipids in the body 55 . In our research, an increase in the level of CLA in the milk of sheep with the GG ANXA9/HinfI genotype was observed.
According to studies by other authors, long-chain unsaturated fatty acids are important in the human diet 52,54,55 . Replacing SFA with PUFA in the human diet is beneficial for the cardiovascular system 56 . In the diet, saturated acids, such as C18:0, C14:0 and C12:0, adversely affect the increase of plasma lipoproteins 57 . On the other hand, the consumption of saturated fatty acids with a carbon chain from C12 to C16 increases the level of plasma LDL-C 56,58 . In our research, an increase in the level of C18:0 acid was noted in the ANXA9/ HinfI polymorphism in the homozygous GG individuals, while in the homozygous AA sheep, in the case of the SCD1/Cfr13I polymorphism, a decrease in the level of C12:0, C13:0, C14:0 acids and an increase in MUFA level in milk was observed. Consuming MUFA has a positive effect on humans, reduces total cholesterol and LDL fraction in the blood 59. The share of individual fatty acids in the human diet determines both physical and mental health. The present study determined the relationship between the studied polymorphisms and the fatty acid profile in milk and the results may be used in the selection program in sheep flocks: the choice of an appropriate genotype variant will ensure the desired fatty acid profile in milk.

Conclusions
The research was aimed at searching for the best genetic marker influencing the fatty acid profile in sheep milk. These preliminary findings show that the greatest changes were observed in sheep with SCD1/Cfr13I polymorphism and the least significant ones in FASN/AciI. Milk obtained from the homozygous AA sheep (SCD1/ Cfr13I) had the best fatty acid profile. Slight changes in the milk of sheep were found for polymorphisms in the ANXA9 gene. For the ANXA9/NlaIII polymorphism (AA genotype animals) a favourable reduction in the level of C20:4n6 acid in the milk was noted. Changes were also found for the ANXA9/HinfI polymorphism (homozygous GG individuals). The milk of these sheep was characterized by an increase in the level of C18:1n9t and CLA acids, which is beneficial; unfortunately, an undesirable increase in the proportion of C18:0 acid was also found. Because presented results indicate the association of the analyzed genotypes with the fatty acid profile in Zošľachtená valaška, a larger prospective study will be continued to explore the effect of SNPs, also in the other genes showing effects on sheep milk. Moreover, results of our study can be useful for breeders, especially that Zošľachtená valaška sheep are included in the breeding program and sheep with homozygous AA (SCD1/Cfr13I) could be used during the breed improvement program for genetic selection if no detrimental trait is associated with this genotype.