Association between THRSP Gene Polymorphism and Fatty Acid Composition in Milk of Dairy Cows

Simple Summary The purpose of this study was to detect polymorphism in thyroid hormone-inducible hepatic protein gene (THRSP) and analyze its influence on the fatty acid composition of milk in Jersey and Holstein-Friesian cattle. One single nucleotide polymorphism (SNP) was detected and determined in 224 cows. It was demonstrated that the analyzed variant had a significant influence on several fatty acids content in milk. Obtained results could be applied in breeding programs for improving the quality of milk. Abstract Thyroid hormone-inducible hepatic protein is involved in the de novo synthesis of fatty acids in the lactating mammary gland. Different variants of the gene that encodes this protein may be associated with its different activity. The primary aim of this study was to find polymorphism in the THRSP gene and estimate the relationship between individual genotypes and fatty acid composition in milk. Investigations were carried out on 224 cows represented by two breeds—Jersey (n = 80) and Polish Holstein-Friesian (n = 144). Polymorphism in THRSP was detected by Sanger sequencing; however, genotypes were determined by the PCR-RFLP method. It was shown that the analyzed variant had a significant (p < 0.05) influence on palmitic and stearic fatty acids as well as on fatty acids with a chain length of 14, 16, and 6–16 in Jersey breed and on caproic, palmitic, myristoleic, and palmitoleic fatty acids in H-F. Obtained results indicated that analyzed SNP in bovine THRSP gene (rs42714482) may be considered as a potential marker for fatty acid composition in milk


Introduction
Milk fat is a source of fatty acids in the human diet, which can be beneficial for health or can be associated with the risk of some diseases [1]. The fatty acid profile can be modified by many factors; among them, environmental and genetic seems to be most important. Whereas environmental conditions can be managed by farmers if the genetic background is more complex to handle. First, we need to know which genes and their variants are correlated with fatty acid composition in milk. Therefore, it is important to analyze candidate genes for this trait, which can be typed based on the physiological role of the encoded protein, QTL mapping or GWAS.
Thyroid hormone-inducible hepatic protein gene (THRSP) encodes Spot14 (S14) protein, which is associated with regulation of the de novo fatty acid synthesis in the liver, adipose tissue, and lactating mammary gland [2]. It was shown that overexpression of THRSP in bovine mammary epithelial cells increased triacylglycerol levels and enhanced the expression of following lipogenic genes: fatty acid synthase (FAS), peroxisome proliferator-activated receptor γ (PPARγ) and sterol regulatory element-binding protein 1 (SREBP1) [3]. THRSP gene polymorphisms were investigated in many domestic animals. Analysis indicated that the THRSP gene in chicken is duplicated and 2 forms are present-THRSPα and THRSPβ. In both paralogs, insertion-deletion (indel) polymorphism was found. It was shown that THRSPα variants were correlated with the deposition of abdominal fat in chickens [4]. In another study, one SNP was detected in chicken THRSPα which together with earlier analyzed indel were associated with growth and body composition traits [5]. Similarly, detected SNP in the goat THRSP gene was studied in relation to growth traits. It was shown that different genotypes were correlated with body weight and chest girth in the Boer goat breed [6]. In pigs, however, SNP found in 5 proximal regulating region of THRSP was associated with average backfat thickness, average daily weight gain, and loin-eye area [7]. Also in cows, polymorphism in the coding region of this gene was analyzed in relation to meat and carcass traits, as well as fatty acid composition in meat. It was shown that different THRSP genotypes were correlated with water holding capacity and meat tenderness in Qinchuan cattle and with unsaturated, monounsaturated, and few individual fatty acids in the meat of Hanwoo cattle [8,9]. In dairy cattle-Italian H-F, THRSP variants were associated with some milk production traits [10]. Bovine THRSP gene is located on chromosome 29 (Bta29) and consists of 2 exons. The length of the transcript is 1398 bp; however, the protein is composed of 148 aa [11,12]. Although the role of THRSP in the synthesis of fatty acids during lactation, polymorphism in the gene that encodes this protein was not studied in relation to the fatty acid composition of milk. Therefore this study aimed to detect polymorphism in the THRSP gene of Jersey and Polish Holstein-Friesian cattle and to perform association analysis for milk fatty acids profile.

Animals
The experiment covered 224 cows that belong to Jersey (n = 80) and Polish Holstein-Friesian (n = 144) breeds. The first group was reared in a tie-stall barn in Greater Poland Voivodeship; however, the second was reared in a free-stall barn in West Pomeranian Voivodeship. Feeding and management of animals on both farms were very similar. Cows were fed by use of a total mixed ration (TMR) diet that contains corn silage, grass haylage, alfalfa silage, straw, solvent-extracted soybean meal, as well as minerals and vitamins. The nutritional composition of feeds applied in the experiment is presented in Table 1. Blood samples were collected into tubes containing K 3 EDTA during a routine veterinary check-up. Milk samples were collected in lactations 1-4 during trial milking performed by the Polish Federation of Cattle Breeders and Dairy Farmers. The mean day of lactation for Holstein-Friesian cows was 195; however, for Jersey, it was 151.

Polymorphism Analysis
DNA was isolated from peripheral blood by use of MasterPure™ DNA Purification Kit for Blood Version II (Epicentre Biotechnologies, Madison, WI, USA). Following primers pair, covering bovine THRSP exon 1 was designed using Primer3 software [13]: F 5 -GCTGTGTTGACCTACTGGC-3 , R 5 -CGGCCACCATTACCTTTCCT3 . Primers were designed based on the ENSBTAG00000011666 sequence [10]. PCR cycling was as follows: initial denaturation at 94 • C/5 min, 35 cycles of 94 • C/30 s, 61 • C/45 s, 72 • C/30 s, and final extension at 72 • C/5 min. PCR was performed in a total volume of 15 µL that contains 50-80 ng of genomic DNA, 1.5 mM MgCl2, 0.3 mM of dNTP mix, 12 pmol of each primer, and 0.35 U of Taq polymerase (EURx, Gdansk, Poland). The presence of specific amplicons (598 bp) was confirmed in 1.5% agarose gel with Perfect™ 100-1000 bp DNA Ladder (EURx, Gdansk, Poland). Sequencing of amplicons was performed by an external service (Genomed, Warsaw, Poland). PCR-RFLP method was applied to determine detected THRSP gene variants (rs42714482) [14]. The same pair of primers and conditions as mentioned above were used in PCR. A total of 10 µL of obtained amplicons were digested by BstC8I enzyme (SibEnzyme, Novosibirsk, Russia) in 55 • C at least 3 h. Restriction fragments were separation in 4% agarose gels with a 50 bp DNA Ladder (Genoplast, Rokocin, Poland).

Milk Composition Analysis
To avoid a period of negative energy balance, milk samples were collected after the 90th day of lactation during morning milking, according to PN-EN ISO 5555:2002 standard. Next samples were transported to the laboratory and stored at −20 • C until further analysis. Lipids were extracted from milk by use of chloroform and methanol mixture in a 2:1 ratio. Next fatty acids were converted into methyl esters using boron trifluoride according to PN-EN ISO 12966-2:2011 standard. Analysis of fatty acids methyl esters was performed using gas chromatography mass-spectroscopy method (GC-MC) in agreement with PN-EN ISO 5508:1996 standard. Fatty acids were identified based on their relative retention time in relation to retention times of standard (SupelcoTM 37 Component FAME Mix, Sigma-Aldrich, Saint Louis, MI, USA). Following fatty acids were analyzed: , UFA/SFA ratio, atherogenic index (AI) and thrombogenic index (TI). The ∆9 desaturase and atherogenic/thrombogenic indices were calculated using the formulas proposed by Lock and Garnsworthy [15] and Ulbricht and Southgate [16].

Statistical Analysis
Statistical analysis was conducted using R packages [17]. Pedigree data were arranged in Pedigree Viewer, ver. 6.5 (University of New England, Armidale, Australia) [18]. An additive relationship matrix based on a three-generation pedigree was generated using the kinship2 R package (Rochester, MI, USA) [19]. To estimate associations between individual THRSP genotypes and analyzed traits following mixed linear model was estimated and applied using the lmekin function from the coxme R package (Rochester, MI, USA) [20]: where: Y-the value of the analyzed trait, µ-overall mean, G-fixed effect of THRSP genotype, LS-fixed effect of lactation number and season, β1A-regression coefficient for the age of cow, β2DLC-regression coefficient for the day of lactation when milk was collected, α-random polygenic effect taking into consideration pedigree relationships, e-the random error. Statistical significance of results was indicated below a p-value of 0.05 (p < 0.05).

Results
Sequencing of bovine THRSP exon 1 allowed detection of one missense SNP located in position 193 of the transcript (152 in cds) ( Figure 1). This C/T substitution led to alanine to valine exchange in position 51 of the protein.
where: Y-the value of the analyzed trait, µ-overall mean, G-fixed effect of THRSP genotype, LS-fixed effect of lactation number and season, β1A-regression coefficient for the age of cow, β2DLC-regression coefficient for the day of lactation when milk was collected, α-random polygenic effect taking into consideration pedigree relationships, e-the random error. Statistical significance of results was indicated below a p-value of 0.05 (p < 0.05).

Results
Sequencing of bovine THRSP exon 1 allowed detection of one missense SNP located in position 193 of the transcript (152 in cds) ( Figure 1). This C/T substitution led to alanine to valine exchange in position 51 of the protein.   where: Y-the value of the analyzed trait, µ-overall mean, G-fixed effect of THRSP genotype, LS-fixed effect of lactation number and season, β1A-regression coefficient for the age of cow, β2DLC-regression coefficient for the day of lactation when milk was collected, α-random polygenic effect taking into consideration pedigree relationships, e-the random error. Statistical significance of results was indicated below a p-value of 0.05 (p < 0.05).

Results
Sequencing of bovine THRSP exon 1 allowed detection of one missense SNP located in position 193 of the transcript (152 in cds) ( Figure 1). This C/T substitution led to alanine to valine exchange in position 51 of the protein.   The distribution of THRSP genotypes and alleles with Hardy-Weinberg equilibrium (HWE) in both analyzed breeds of cows are presented in Table 2. It was shown that TT genotype was most frequent in Jersey (0.33) group; however, CC was most frequent in Polish H-F (0.46). In the case of alleles, reverse tendency was observed where T was the major allele in Jersey cows (0.58), while C in H-F (0.68). Analysis showed that the distribution of genotypes was in agreement with the HWE expectation. Frequencies of individual genotypes were significantly different between Jersey and Polish Holstein-Friesian breeds (χ 2 = 29.442, p < 0.01) Fatty acid composition of milk with some indexes in relation to THRSP variants is shown in Tables 3 and 4. In Jersey cows, statistically significant differences (p < 0.05) were found between individual genotypes and following milk fatty acids: palmitic, stearic as well as following indexes: fatty acids with a chain length of 14, 16, and 6-16. In Polish H-F cows, associations (p < 0.05) were found for caproic, palmitic, myristoleic, and palmitoleic fatty acids in milk. In both breeds, only one trait was common as correlated with THRSP genotypes-palmitic fatty acid. Jersey cows with TT genotypes were characterized by the highest value; however, Polish H-F with the same variant had the lowest.  , SFA-saturated fatty acids, UFA-unsaturated fatty acids, MUFAmonounsaturated fatty acids, PUFA-polyunsaturated fatty acids, AI-atherogenic index, TI-thrombogenic index; a,b -means with different superscripts in the same row differ statistically significant at p < 0.05, n.s.-not significant.

Discussion
THRSP is a nuclear protein that can regulate lipogenesis [21]. It was found that THRSP may regulate milk fat synthesis by directly affecting the activity of some classical lipogenic enzymes [3]. Recently, Salcedo-Tacuma et al. [22] indicated THRSP as an inhibitor of lipid synthesis in adipose tissue of periparturient Holstein cows.
THRSP gene is involved in de novo fatty acid synthesis. The fatty acids that are synthesized de novo belong to short-chain and medium-chain length acids, from C4 to C14 and also some C16. The C18 fatty acids and some C16, however, arise from the plasma lipids [23]. In our study, we analyzed fatty acids from C6 to C18. We decided to include C18 fatty acids because an earlier report showed a significant decrease in C18:0 upon THRSP overexpression in goat mammary epithelial cells [24].
THRSP gene in Polish H-F and Jersey breeds in relation to milk fatty acids composition was analyzed for the first time. In a previous study, Oh et al. [9] selected 8 SNPs in the bovine THRSP gene from GenBank [13] for analysis in Korean cattle (Hanwoo). Only two of them, namely g.78 G > A and g.184 C > T were polymorphic in this breed. Comparison of sequences from Ensembl [11] and GenBank [14] showed that the g.184 C > T variant is the same as detected in our work (29:17994763 C/T, rs42714482). Analyzed polymorphism causes amino acid substitution in position 51 of the encoded transcription factor, which probably may modify its interaction with regulated genes and, thus, differentiate processes connected with fat milk synthesis, including the composition of individual fatty acids in milk. We found that the frequency of individual genotypes in Korean cattle (CC-0.17, CT-0.46, TT-0.37) was very similar to those observed in our study in Jersey breed. Evolutionary analysis of SNPs located in genes of Hanwoo cattle in relation to data from other cattle breeds, i.e., Jersey, Simmental, Angus, and Holstein showed that the Korean breed was distinctly separated from the other four breeds. Further SNPs analysis, however, showed that the THRSP gene is not classified as Hanwoo-specific, which may reflect similar genetic parameters in Korean and Jersey breeds [25]. When comparing the frequency of THRSP genotypes in Italian Holstein-Friesian cattle (CC-0.48, CT-0.42, TT-0.10), the same was observed in our work for Polish Holstein-Friesian breed [10].
As mentioned earlier, the THRSP gene was not investigated in relation to milk fatty acids. It was studied, however in relation to carcass traits and fatty acid composition of muscle fat in Korean Cattle, as well as to health and milk production traits in Italian Holstein cattle. It was shown that g.184 C > T SNP is significantly correlated with myristic, palmitic, myristoleic, oleic, linoleic, linolenic fatty acids in m. longissimus dorsi, as well as with saturated fatty acids, monounsaturated fatty acids, and monounsaturated/saturated ratio [9]. In our study, we found an association between THRSP genotypes and palmitic fatty acid in milk of Jersey and Polish H-F cows and myristoleic fatty acid in HF cows. In the muscle of Hanwoo and the milk of Jersey cattle, the highest content of palmitic fatty acid was found for the TT genotype. In the Polish H-F breed, however, a reverse tendency was observed, because the TT genotype was correlated with its lowest value. Similarly, TT genotype was favorable for myristoleic fatty acid in the muscle of Hanwoo cows but in the milk of Polish Holstein-Friesian cows, it was disadvantageous. These differences may result from fatty acids content in milk of particular cattle breeds. A significant breed effect on the content, fatty acid profile and atherogenic or thrombogenic properties of milk fat was reported by Sobotka et al. [26]. The concentrations of long-chain saturated fatty acids were significantly higher in the milk fat of Jersey cows than in the H-F breed. The milk of H-F cows had lower fat content but provided more health benefits than the milk of Jersey. It was also confirmed by the analysis of Jersey, H-F, and three other breeds reared in the Netherlands. Breed differences were found for individual fatty acids, among them for palmitic and myristoleic [27]. We cannot also exclude that observed differences in association tendency may be linked with different rearing conditions for both breeds. We found also an association of TT genotypes with higher content of caproic and myristoleic fatty acid in Polish H-F and with lower stearic fatty acid in Jersey, but this relationship was not observed in Hanwoo cattle. Analysis of THRSP gene in Italian Holstein cows also showed a significant association for milk yield, milk fat yield, milk protein yield and productivity, functionality, and type index, with allele T being favorable for these traits [10]. In our study, we did not find any relationships between individual THRSP genotypes and milk yield, fat yield, and fat content both in Polish Holstein-Friesian and Jersey breeds. Similarly, THRSP genotypes were not found to influence the atherogenic and thrombogenic indexes. As mentioned earlier, we also noticed significant associations for stearic, caproic, palmitoleic fatty acids and fatty acids with a chain length of 14, 16, and 6-16. Stearic acid is common in nature, both in animal and vegetable organisms; however, its level is usually higher in animal than vegetable fat [28]. In milk, it arises from the plasma lipids, but interestingly, the highest expression of THRSP is correlated with its decrease in goat mammary epithelium [24]. In the Jersey breed, TT genotype was associated with the lowest level of stearic acid in milk. Caproic, caprylic, and capric fatty acids are known as the reason for the specific aroma of goat and sheep milk [29]. Analysis of caproic acid amount in the milk of different species showed its highest level in goat milk, lowest in sheep; however, in cow milk, it was slightly higher than in sheep [30]. In Polish Holstein-Friesian cows, the TT variant of THRSP related to the highest content of this fatty acid. Palmitoleic acid shows anti-inflammatory and antidiabetic activity and is produced by desaturation of palmitic acid [31]. In Polish H-F animals we found that TT genotype is correlated with the lowest amount of palmitoleic, as well as earlier mentioned palmitic fatty acid, which can reflect their metabolic relationships. In the case of confirmed associations of THRSP polymorphism and fatty acids with a chain length of 14, 16, and 6-16 in Jersey cattle, we obtained unclear results. The highest value of the first trait was found for the CC genotype; however, the second and third traits were found for the TT genotype.

Conclusions
The conducted experiment showed that THRSP polymorphism (rs42714482) is associated with milk fatty acid composition in Jersey and Polish Holstein-Friesian cattle. It covers palmitic, stearic caproic, myristoleic, and palmitoleic fatty acids, as well as fatty acids with a chain length of 14, 16, and 6-16 in a particular breed. Only palmitic acid content was the common trait for both breeds, however, with opposite tendency, which may reflect breed differences or/and feeding and housing conditions. Analyzed SNP in bovine THRSP gene could be taken into consideration as a potential marker for fatty acid composition in milk.

Institutional Review Board Statement:
The bioethics committee approval of this study is not necessary because the experiment was carried out on biological material. This material was derived from breeding farms and was collected during a routine veterinary check-up.