Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

The association of gene polymorphisms with milk production and mastitis resistance phenotypic traits in dairy cattle

The association of gene polymorphisms with milk production and mastitis resistance phenotypic... Ann. Anim. Sci., Vol. 23, No. 2 (2023) 419–430 DOI: 10.2478/aoas-2022-0091 The associa Tion of gene polymorphisms wiTh milk producTion and masTiTis resisT ance phenoTypic TraiTs in dairy ca TTle* 1♦ 1,2 3 4 Emilia Bagnicka , Paulina Brzozowska , Kacper Żukowski , Renata Grochowska Institute of Genetics and Animal Biotechnology Polish Academy of Sciences, Postępu 36A, 05-552 Jastrzębiec, Poland Faculty of Dietetics, Warsaw College of Engineering and Health, Bitwy Warszawskiej 1920 r. 18, 02-366 Warszawa, Poland National Research Institute of Animal Production, 32-083 Balice n. Kraków, Poland Institute of Agricultural and Food Economics – National Research Institute, Świętokrzyska 20, 00-002 Warszawa, Poland Corresponding author: e.bagnicka@igbzpan.pl abstract The aim of this study was to evaluate the association between gene polymorphisms (snps) and mastitis indicators and their relationship with milk production profitability in dairy herd. A functional analysis was also performed of five genes containing the studied SNPs and those located close by. DNA was isolated from the hair bulb of 320 dairy cows kept in three herds and SNP-microarray analysis was performed. The data on 299 cows was subjected to final statistical analysis using AI-REML method with one-trait repeatability test-day animal model and pedigree information using the DMU4 package. Five from 35 SNPs significantly associated with mastitis indicators or production traits and located within a gene or no more than 500,000 nucleotides from the gene were selected for the functional and economic analysis. A questionnaire was also developed to collect associated economic data of 219 cows from three herds, such as the value of milk production and direct costs incurred over three years; this allowed the gross margin, direct profitability index and direct costs incurred to produce one liter of milk to be determined, among others. None of the five studied SNPs were related to protein content. The rs110785912(T/A), found near CXCR4, and rs136813430(T/C), located in the TLR4 gene exon, were associated with lnSCC, while rs110455063(C/G), located near IGF-I, was associated with milk yield, fat and total solid contents. rs109421300(T/C), associated with fat/ protein content ratio, as well as fat and total solid content, is located in the DGAT1 gene intron. rs41587003(A/C), located in the DLG2 gene intron, was associated with lactose content. The economic analysis revealed differences between the variants of the three tested SNPs. The T/C variant of the rs136813430(T/C) SNP was characterized by the highest gross margin, the highest direct profitability index and the lowest costs incurred to produce 1 liter of milk. The T/A variant of rs110785912(T/A) was related to low lnscc and was charac - terized by the highest direct profitability index. In turn, the C/C variant of the rs41587003(A/C) was related to the lowest level of lactose and the highest costs of milk production. It appears that rs136813430(T/C) may be the most promising of the tested snps for increasing the profitability of milk production. To our knowledge, it is the first effort to assess directly a correlation between the DNA polymorphism and economic output of a dairy enterprise. Key words: SNP marker, intron, exon, association, mastitis indicator, dairy economics, milk production profitability Mastitis is still one of the major diseases affecting breeding progress is rather slow, as genetic variability di- dairy cattle, with negative effects on milk production rectly related to resistance accounts for only a small part (Davies et al., 2009). Despite the wealth of research per- of the total variance; however, this value is significant. formed on the physiological and cellular processes tak- The other trait considered as additive mastitis indicator ing place in the mammary gland in response to infection, is lactose content (Bagnicka et al., 2016; Andrei et al., knowledge of many of the defense mechanisms remains 2009). incomplete (Sordillo, 2005). Mastitis is negatively cor- Inflammation is a complex trait controlled by many related with milk yield, with high-yielding cows being genes. As such, it is difficult to develop an appropriate more susceptible (Carlen et al., 2005). selection strategy to increase immunity. Such programs Most genetic studies of mastitis indicators to date are complicated by the lack of information on the ge- have focused on the number of somatic cells (SCC) in netic basis of animal immunity, the complexity of the milk. The heritability coefficient (h ) for SCC was esti- functional interaction between host and pathogen, which mated from 0.05 to 0.20 (Ptak et al., 2007; Sender et al., may trigger a range of different immune responses, and 2013). With such a low to moderate degree of heritability, the influence of the environment (Marogna et al., 2010). __________ *This study was supported by the Leading National Research Centre Scientific Consortium’s “Healthy Animal – Safe Food” initiative under the Ministry of Science and Higher Education decision No. 05-1/KNOW2/2015, grant KNOW2015/CB/PRO1/52, to Emilia Bagnicka. The pub- lication costs were co-financed by the subsidy of the Ministry of Agriculture and Rural Development (DŻW.pp.862.1.2023). 420 E. Bagnicka et al. material and methods The search for genes related to the susceptibility/resist- ance of dairy cows to udder inflammation and production animals and farms traits has recently been supported by the development of The first step of the research was carried out on 320 genome-wide association studies (GWAS) using pheno- dairy cows from three herds. All were kept in free stall typic data and data on the genotypes of the studied indi- barns with constant access to enclosures. The farms dif- viduals. One of the most popular methods of genotyping fered in terms of the number of cows and feeding system, concerns the use of dense sets of SNP (Single Nucleotide resulting from diversified area of arable land, including Polymorphism) microarrays. permanent grassland. The mean herd yield was 9,393 lit- Taking advantage of genomics for breeding purposes ers of milk per lactation, with mean fat and protein con- raises the question of whether the use of selection meth- tents of 4.18% and 3.49%, respectively. Cows were fed in ods based on SNPs influencing the immunity of the cow the total mixed ration (TMR) system, balanced in accord- mammary gland is economically justified, i.e., whether ance with the Institut national de la recherche agronom- this approach will reduce losses related to udder diseas- ique (INRA) standards and adapted to Polish conditions es and thus increase the efficiency of dairy cows, and, (Brzóska et al., 2014), with constant access to water. The consequently, reduce milk production costs in a herd. cows were milked twice a day with a mechanical milking The future of livestock production is not about produc- system (DeLaval, Tumba, Sweden). All herds were as- tion intensification at all costs, which requires increas- sessed for dairy and breeding performance under routine ing expenditure to account for more limited resources performance control. Phenotypic data on the daily milk and environmental pollution. A more desirable goal is to yield and the component contents was obtained from the minimize production costs thanks to the use of inter alia Polish Federation of Cattle Breeders and Dairy Farmers, new genomic technologies, which enable the selection based on which a database with 8090 observations was of individuals with specific genetic potential in terms of prepared. Information on daily milk yield, protein, fat, production and health. dry matter, lactose contents and SCC were gathered. The Most of the research on udder health to date has fo- yield of milk components and fat/protein ratio were also cused on estimating the costs of preventing inflamma- calculated. The descriptive statistic of all studied traits is tion, losses due to mastitis, and effectively managing presented in Table S1. The pedigree database contained a herd to reduce such losses (Hagnestam-Nielsen and information on 7,035 animals, along with their birth Østergaard, 2009). The use of new technologies based year, and 2,057 sires and 4,634 dams with 64 sires hav- on SNPs has increased the efficiency of selecting eco- ing more than 10 daughters (range between 10 and 61). nomically significant traits. Hence, a number of studies However, the genotyped cows were descendants of 153 have been carried out in cattle to identify relationships sires (54, 40 and 59 sires in each herd respectively) and between gene polymorphisms and various production 263 dams. Ten sires of genotyped cows had daughters in and functional traits (Lü et al., 2011; Yuan et al., 2013). three or two herds. Previous studies have only indicated the potential of economic benefits related to higher meat and milk pro- Sampling and gene polymorphism identification duction or lower SCC in milk; however, these traits DNA was isolated from approximately 100 hair fol- do not guarantee greater milk production profitability, licles from each cow using a kit for isolating DNA from which is determined by the production volume and the blood and tissues (Macherey-Nagel, Düren, Germany). costs incurred. None of these studies have examined the The procedure was performed according to the manu- profitability of using individual SNPs in cattle selection facturer’s protocol, with the following modification: the based on real economic data like gross margin, direct hair root cells were disintegrated by dipping four times profitability index and costs incurred to produce 1 liter in liquid nitrogen, and then thawing in a water bath at of milk. Therefore, this paper attempts to fill this gap by 37°C and increasing the amount of added proteinase analyzing the economic effects of SNPs in genes related from 25 µl to 50 µl. Just before performing the microar- to milk production and health based on information re- ray analysis, quantitative DNA analysis was performed garding the production value and direct costs obtained using intercalating dyes capable of specifically binding to from farms. double-stranded DNA (Quant-iT PicoGreen dsDNA kit, It was hypothesized that the use of gene polymor- Thermo Fisher Scientific). Fluorescence measurement phisms related to udder inflammation in dairy cows was performed using a Fluoroskan Ascent FL appara- would increase milk production profitability; there- tus (Thermo Fisher Scientific, Waltham, Massachusetts, fore, the aim of this study was to evaluate the associa- USA) according to the manufacturer’s protocol. Eight tion between impact of five selected polymorphisms samples with concentrations below 50 ng DNA/µl were (SNPs) and mastitis indicators or production traits as excluded from the further analysis. well as to investigate the impact of selected identi- Microarray analysis was performed on 120 ng of DNA fied SNPs on the profitability of milk production in of 312 cows from each run using the BeadChip Bovine a dairy herd. A functional analysis was also made of 50k v3 kit. The results were interpreted using the Genom- the genes that include the studied SNPs and genes lo- eStudio 1.9.4 software (Illumina, San Diego, CA, USA). cated nearby. Marker SNPs vs. mastitis and dairy cattle farm profitability 421 As the DNA was isolated from hair follicles, no per- the analysis to estimate the differences between variants mission was needed from the Local Ethics Committee for of particular SNPs. Post hoc test with the Bonferroni cor- biological material collection. rection was used. Five from 35 SNPs significantly asso- ciated with mastitis indicators or then with production association study traits and located within a gene or no more than 500,000 The next step in establishing the association between nucleotides from the gene were selected for the function- individual SNPs and selected mastitis indicators and milk al and economic analysis. traits was to conduct a GWAS-based basic bioinformatic Before statistical analysis, the total SCC were trans- analysis using the Plink.2 program (Chang et al., 2015) formed to natural logarithm values (lnSCC). χ2 test was for more than 50K SNPs; a linear regression model was applied to determine whether SNP genotypes frequency used for this purpose, taking into account the fixed effect and Hardy–Weinberg equilibrium held. of SNP and population structure among individuals as co- variate. Genotypes with low frequency and SNPs which Functional analysis of genes with analyzed SNPs were monomorphic in the studied group were excluded Seven genes associated with studied SNPs were sub- from the analysis. As a result, the number of SNPs test- jected to functional analysis using the Kyoto Encyclope- ed was reduced down to 46K. All phenotypic traits were dia of Genes and Genomes (KEGG) pathway database averaged using an interquartile distribution: IQR(x) = (Kanehisa at al., 2016) and Gene Ontology Annotation quantile(x. 3/4) - quantile(x. 1/4). In order to check the dif- (GO) (Ashburner et al., 2000). ferences in the environment between the studied herds, the multidimensional scaling analysis (MDA) was performed. economic analysis Genotyping rating was 0.98. All outliers, i.e. information To assess the value of the SNP polymorphisms for on 13 cows, were removed from further analysis. selection purposes, a questionnaire was developed to col- Following the above, a preliminary analysis of the lect the actual data on milk production of 219 cows and associations between SNPs and mastitis indicators and the direct costs incurred over three calendar years, i.e., milk traits was conducted using the PROC MIXED SAS 2015, 2016 and 2017 in three herds. Altogether, 384 re- package (SAS/STAT 2002–2012) with a mixed model cords were analyzed. The analysis was based on calcula- using information on production of 299 genotyped cows. tions in accordance with the AGROKOSZTY methodol- All SNPs which were found to be associated with ud- ogy (Methodology, 1999; Skarżyńska, 2012). der health in the previous step of the analysis (lnSCC– The research covered revenues, i.e., the value of po- a natural logarithm from SCC, lactose content) and with tential commercial production (assuming that the sales production traits (milk yield, protein and fat content or volume is equal to the production volume) and direct yield) were taken into account. Altogether, 52 SNPs were costs. The gross margin was adopted as the basic meas- selected, of which only 35 were significantly associated ure of the assessment of the obtained economic effects. with the studied traits in this analysis of variance. Finally, The calculations do not take into account support with these 35 SNPs were analyzed by AI-REML to obtain so- subsidies under the Common Agricultural Policy. In lutions for fixed effects, using the DMU4 package (Mad- calculating the gross margin, only the value of the main sen and Jensen, 2013). A one-trait repeatability test-day product, i.e., milk, traded on the market was considered. animal model was applied. The annual value of milk production of each genotyped Phenotypic data was distributed among 21 calving cow was calculated based on the monthly milk yield of years from 1997 to 2017 – each year constituted a sep- the genotyped cows, and the mean monthly selling prices arate class and calving season: 1 – Winter (December, of milk for a given farm. January, February), 2 – Spring (March, April, May), 3 Due to the inability to record information on the in- – Summer (June, July, August), and 4 – Autumn (Sep- curred direct costs per genotyped cow, the gross margin tember, October, November). Daily milking data were was calculated based on the direct costs collected for the collected across 1996–2017. Finally, 150 classes of inter- entire dairy cow herd, and this value was converted to action were created between herd, calving year and sea- a single cow in the herd (Skarżyńska, 2012). For this pur- son (HYS) and 537 classes of interaction between herd, pose, information was obtained on the costs of feed from year and month of milking (hym). Parity was grouped outside the farm, of own feed from potential commercial into four classes, with the fourth containing lactations and non-commodity products, and of veterinary meas- higher than the third. The final statistical classification ures and services (including semen and insemination) fitted HYS, parity, and SNPs along with the Legendre and specialist costs (e.g., hoof correction, preparation of polynomial of order 5 regression on standardized DIM animals for exhibitions). nested within parity as fixed effects and genetic additive, Thus, milk production profitability for the entire herd permanent environment, hym, and residual effects as ran- of dairy cows in individual farms was determined based on dom (File S1). Each analysis included the relationship milk production value, calculated as the product of the aver- between one or two SNPs and the tested features at a time age milk yield per cow in a year and the mean annual milk due to computational limitation which is the shortage of price for a given herd, and direct costs, estimated based on the study as not all epistatic influences are considered in direct costs incurred in the entire herd in a year, per cow. 422 E. Bagnicka et al. The following economic categories were considered: alleles is presented in Table 2. Studied SNPs were asso- gross margin (GM) = PV – DC ciated with traits at least at P<0.05, except rs110785912 where: PV – production value (PLN); DC – direct which was associated with lnSCC at P = 0.0763. For costs (PLN) rs110785912(A/T) and rs110455063(C/G), a very high direct profitability index (DP) = (PV / DC) × 100 frequency of one allele was reported. The χ test results where: PV – production value (PLN), DC – direct indicate that the observed frequency of genotypes does costs (PLN) not differ from that expected one (Table 2). direct costs incurred to produce 1 liter of milk The first studied SNP, rs110785912(T/A), also known (MDC) = DC / MP as ARS-BFGL-NGS-50482, associated with lnSCC in where: DC – direct costs (PLN), MP – milk produc- our study, is located in the intron of a gene marked as tion (liters) LOC1049123 (NCBI-1). The function of the gene is un- To determine the significance of differences between known. The SNP identified 447,880 nucleotides from the tested SNP variants, an analysis of variance with the C-X-C chemokine receptor type 4 (CXCR4) gene a single-factor model was performed for each economic (Table 3). A/A and A/T genotypes were associated with category; in each calculation, a single SNP and the par- lower lnSCC than T/T (Table S2). The lnSCC herit- ity over the three studied years were used as fixed ef- ability coefficient estimated in the tested population was fects. The calculations were performed using the PROC 0.13, with the investigated trait demonstrating high co- MIXED SAS package (SAS/STAT 2002–2012). efficient of variance (CV=34.5%) which reflects a total variability. SNP rs110455063(C/G) known also as ARS-BF- results GL-NGS-15787, was located in the intron of gene with unknown function and marked as LOC104972477 gene polymorphism analysis (NCBI-2) and, simultaneously, was located 34,194>(5’) The multidimensional scaling analysis (MDA), nucleotides from the insulin-like growth factor I (IGF-I) which presents the environmental differences between gene. Its frequency was 0.228 for G/G, 0.759 for C/G, the studied herds and Manhattan plots as the result of and 0.013 C/C (Table 2). It was found to be associated GWAS analysis for some studied traits, is presented with milk yield, fat and total solid content. G/G variant in supplementary file (Figure S1 a and S1 b, A–E). carriers had the highest milk production while the highest After three-step selection, based on the present find- fat content was found in the milk of cows with the C/C ings SNPs: rs110785912(T/A), rs110455063(C/G), genotype, and the lowest in heterozygotes. The highest rs109421300(T/C), rs41587003(A/C), and rs136813430 dry matter content was observed in the milk of heterozy- (T/C) were selected for the functional and economic gous cows and the lowest content in cows with C/C. The analysis. total variability of the milk yield expressed as CV was The outcomes of the association analysis of selected high as 60%, while fat was 22.5% and dry matter content SNPs with functional and production traits of dairy cows only 8.37%. The heritability coefficients were 0.38, 0.17 are shown in Table 1, while the distribution of genes and and 0.20, respectively (Table S2). Table 1. Relationship of the studied SNPs with the functional and production characteristics of dairy cows Trait Chromosome rs number P-value** production traits Milk yield 5 rs110455063 0.021 Fat content 5 rs110455063 <0.0001 14 rs109421300 Dry matter content 5 rs110455063 0.0004 14 rs109421300 <0.0001 health mammary gland indicators lnSCC 2 rs110785912 0.0763 8 rs136813430 <0.0001 Lactose content 29 rs41587003 0.02701 **P-value after Bonferroni correction. lnSCC – natural logarithm of the number of somatic cells in milk (SCC). Marker SNPs vs. mastitis and dairy cattle farm profitability 423 Table 2. Distribution of genotypes and alleles of analyzed SNPs Alleles’ frequency/ χ test value/ SNP N Distribution of genotype in the studied population Genotypes’ frequency P-value rs110785912(A/T) 299 (A/A)=243 (AT)=55 (T/T)=1 A=0.905; T=0.095 1.33 /0.813|0.184|0.003| 0.40 rs110455063(C/G) 299 (G/G)=227 (CG)=68 (C/C)= 4 G=0.873; C=0.127 0.19 /0.228|0.759|0.013 0.83 rs136813430(T/C) 294 (C/C)=112 (T/C)=131 (T/T)=51 C=0.604; T=0.396 1.39 /0.380|0.450|0.170 0.24 rs109421300(T/C) 299 (A/A)= 140 (A/G)=137 5(G/G)=22 A=0.697; G=0.303 1.2 /0.468|0.458|0.074 0.14 rs41587003(A/C) 299 (A/A)=101 (AC)= 141 (C/C)=57 A=0.694; C=0.306 0.39/ /0.410|0.570|0.020 0.53 N – number of genotyped cows , SNP = single nucleotide polymorphism. Table 3. Location of the five SNPs related to the analyzed traits in the bovine genome and position in the gene/relative to the nearest gene SNP location SNP’s position in Symbol of the Gene location in the in the gene/in Chr. SNP (nucleotides) Trait the chromosome closest gene chromosome relation to the closest gene 8 rs136813430(T/C) lnSCC 108829443 TLR4 108828899–108839913 Exon 29 rs41587003(A/C) lactose content 10172197 DLG2 9997510–12157167 Intron 2 rs110785912(A/T) lnSCC 61134245 LOC1049123 61096534–61148308 Intron CXCR4 61250082–61253877 115837>(5’) 5 rs110455063(C/G) daily milk yield, fat content, 61134245 LOC104972477 66131228–66164165 Intron dry matter content IGF1 66191602–66264083 34194>(5’) 14 rs109421300(T/C) fat to protein ratio, fat content, 1801116 DGAT1 1795425–1804838 Intron dry matter content GPIHBP1 1348153–1350911 450205<(3’) CYP11B1 1550747–1558425 242691<(3’) Chr. = chromosome; allele: A (adenine), C (cytosyne), G (guanine), T (thymine), lnSCC – a natural logarithm of somatic cell count. TLR4 – Toll-like receptor 4; DLG2 – Disks large homolog 2; CXCR4 – C-X-C chemokine receptor type 4; IGF1 – Insulin-like growth factor I; DGAT1 – Diacylglycerol O-acyltransferase 1; GPIHBP1 – Glycosylphosphatidylinositol-anchored high density lipoprotein-binding protein 1; CYP11B1 – Cytochrome P450 11B1. The rs109421300(T/C), alias ARS-BFGL-NGS-4939, the highest value was associated with the A/A variant and associated with fat/protein content ratio, fat and total sol- the lowest with C/C. The heritability of lactose content id content and fatty acid profile (Table 1), is located in the was estimated at 0.16, and the total variability (CV) of intron of the diacylglycerol O-acyltransferase 1 (DGAT1) this trait was 9.23% (Table S2). gene (Table 3); it is also sited 450,205<(3’) nucleotides Finally, SNP rs136813430(T/C) (BTA-82770-no-rs), from the glycosylphosphatidylinositol-anchored high associated with lnSCC (Table S2), was located in the density lipoprotein-binding protein 1 (GPIHBP1) gene exon of the toll-like receptor 4 (TLR4) gene (Table 3). and 242,691>(3’) from the cytochrome P450 11B1 (mi- The milk of cows with the C/T genotype contained more tochondrial) (CYP11B1) gene. The lowest fat content was somatic cells than the other variants of genotypes. found in animals with the T/T variant, and the highest None of the analyzed SNPs were directly related to with the C/C genotype. The highest dry matter content protein content or yield. The indirect relationship ex- was found in the milk of cows with the C/C genotype pressed as fat to protein ratio was mentioned above. and the lowest in milk of cows with T/T. The total vari- ability of the fat to protein ratio was found to be average functional analysis (CV=20.8%), and the h of this trait was 0.14 (Table S2). Seven genes with known functions (TLR4 – Toll-like SNP rs41587003(A/C) (UA-IFASA-7512) was lo- receptor 4; DLG2 – Disks large homolog 2; CXCR4 – cated in the intron of the disks large homolog 2 (DLG2) C-X-C chemokine receptor type 4; IGF1 – Insulin-like gene (Table 3) and was associated with lactose content: growth factor I; DGAT1 – Diacylglycerol O-acyltrans- 424 E. Bagnicka et al. ferase 1; GPIHBP1 – Glycosylphosphatidylinositol-an- sis revealed that IGF-I is involved in immune processes chored high density lipoprotein-binding protein 1; CY- (response to stress, immune system process, signal trans- P11B1 – Cytochrome P450 11B1) were associated with duction) and growth and development, cell differentia- studied SNPs and subjected to functional analysis. tion, cell population proliferation, and anatomical struc- In silico GO annotation analysis revealed that the ture development (Figure 1, Table S3). protein products of two from five genes i.e., TLR4 and In the DGAT1 gene intron is located SNP No. CXCR4, whose polymorphisms were associated with rs109421300(T/C). Protein product of DGAT1 gene lnSCC, were involved in several common biological pro- is involved not only in the lipid metabolism, sulphur cesses connected with immune system such as homeo- metabolism or biosynthesis process, but also in the static process, immune system process, signal transduc- process of homeostasis and the immune system (Figu- tion, or response to stress (Figure 1). However, they were re 1). also involved in cell differentiation, cell morphogenesis, Rs109421300(T/C) SNP is located 242,691<(3’) cell motility, or cell death. Thus, further association stud- from the CYP11B1 gene and 450,205<(3’) from the GPI- ies are needed to check the relationships with other dis- HBP1 gene. GO annotation analysis indicated that the eases and productivity of the cows. The GO number and protein produced by GPIHBP1 gene is involved in lipid definitions were gathered in Table S3 (supplementary metabolic processes, transmembrane transport, cellular files) based on QuickGO database (term/GO:0002376). nitrogen compound metabolic process, or biosynthetic The next SNP – rs41587003(A/C), associated with lac- process, and also in homeostatic process (Figure 1). tose content, is localized in the intron of DLG2 gene which No hits for GPIHBP1 gene were found in KEGG analy- is required for perception of chronic pain through N-me- sis. As in silico GO analysis showed, CYP11B1 is in- thyl-D-aspartate (NMDA) receptor signaling pathway. The volved in lipid metabolic, circulatory system, homeo- DLG2 gene encodes a member of the membrane-associated static, and immune system processes (Figure 1) while guanylate kinase (MAGUK) family. The protein product of according to KEGG analysis i.a. in metabolic pathways DLG2 gene is involved in processes such as biosynthetic, (Table S4). protein transport, cellular protein modification, cytoskeleton- Two of the studied SNPs, rs110785912(A/T) and dependent intra-cellular transport, cell-cell signaling, cell ad- rs110455063(C/G), are located within LOC1049123 and hesion and cellular nitrogen compound metabolic (Figure 1, LOC104972477 loci, respectively, thus no functional Table S4). analysis is possible. These SNPs were associated with In turn, SNP rs110455063(C/G) is located 34,194 nt lnSCC – rs110785912 (A/T), and dry matter and fat con- 5’ from the IGF-I gene. In silico GO annotation analy- tents – rs110455063(C/G). Figure 1. Common biological processes names in which protein products of the studied genes are involved – in silico GO annotation analysis Marker SNPs vs. mastitis and dairy cattle farm profitability 425 The studied genes are involved in many physiological discussion pathways. The protein products of IGF-I and TLR4 genes are involved in 35 and 28 pathways, respectively with animal sample some of them common for both genes such as proteo- Despite the rather small sample size, as the asso- glycans in cancer (bta05205), HIF-1 signaling pathway ciation analysis was conducted using 299 animals with (bta04066), and PI3K-Akt signaling pathway (bta04151). approx. 8,000 phenotypic records, our obtained values IGF-I and CXCR4 are involved in two common path- for the studied SNPs appear generally consistent with ways (endocytosis pathway – bta04144 and pathways in previous studies. Moreover, the standard errors (SE) of cancer – bta05200), while CYP11B1 and DGAT1 in met- solutions for genotypes were low which also proves the abolic pathway (bta01100). Most of pathways in which low bias of the obtained results. With an average of two studied genes are involved are connected with various daughters per sire and only 12% of the dams having more diseases, however, some of them are involved in regula- than one daughter, our sample of 299 can be considered tion of cellular functions such as transcription, transla- representative of the population for a polymorphism tion, proliferation, growth, and others. study because it is not dominated by a small number of large families (Oprządek et al., 2015). economic analysis The heritability coefficient values obtained for milk The value of the studied polymorphisms for selec- yield was within the range of estimates obtained in the tion purposes was determined by economic analysis Polish dairy cows’ population (h =0.20–0.40) (Aerts et (Table S5). An analysis of variance was performed al., 2021; Rzewuska and Strabel, 2013). CV for milk to determine the significance of the differences be- yield was very high (60%) maybe because of different tween the variants of the analyzed SNPs with regard lactation duration. to their obtained economic effects (Table S6). Statis- A high CV for fat content was obtained in the present tically significant differences were found between the study (22.5%) and this is in line with values obtained by allelic variants of SNP rs136813430(T/C) for all ana- previous studies (Krag et al., 2013). The estimated h value lyzed economic categories. No such differences were (0.17) was slightly lower than the coefficients obtained in found for rs110455063(C/G) or rs109421300(T/C) the Polish active population (0.25–0.28) for the first three for any of the analyzed categories. However, for SNP parities using test-day model (Rzewuska and Strabel, 2013). rs110785912(T/A), significant differences were found The heritability coefficient for lactose content was low- between homozygous AA and heterozygous TA for er (0.16) than the values reported previously (0.26–0.34) the direct profitability index. Moreover, and for SNP in the Polish dairy cattle population (Rzewuska and Stra- rs41587003(A/C) significant differences were found bel, 2013), while the low total variability was obtained in between homozygous C/C and T/C regarding the pro- the present study (CV=9.23%). Therefore, this association duction costs of one liter of milk (Table S3). analysis for the lactose content with the tested SNP should Based on the obtained results, SNP rs136813430(T/C) be repeated on a much greater population sample. seems to be the most useful SNP for selection purposes, The heritability coefficient of lnSCC was 0.13 in the as it was found to have a significant influence on each studied group of animals. Again, despite the small sam- of the three economic categories, with statistically sig- ple size and high total variability of lnSCC (CV=34.5%), nificant differences observed between its variants. Cows this value did not differ from those obtained in the Pol- with the T/C genotype demonstrated the highest gross ish populations of dairy cows (0.13–0.16) (Sender et al., margin and direct profitability index compared to other 2013; Rzewuska and Strabel, 2013). variants, and the lowest costs incurred to produce 1 liter For dry matter content, both heritability (0.20) and of milk. variation coefficient (CV=8.37%) did not differ substan- SNP rs110785912(T/A), related to lnSCC in milk, tially from previous studies in Polish population (Yazgan had a significant influence on the direct profitability in- et al., 2010). dex with significant differences observed between cows Similarly, for fat:protein ratio, the h value (0.14) and with the A/A and T/A genotypes. The T/A cows demon- the total variability (CV=20.8%) did not differ substan- strated the most favorable profitability index and the low- tially from previous values, i.e. h between 0.18 and 0.40 est somatic cell content. (Rzewuska and Strabel, 2013) in a study in a big popula- In the case of SNP rs41587003(A/C), significant dif- tion of more than 19,000 Polish HF dairy cows (159,044 ferences were found between the C/C and T/C variants lactation records). regarding the production costs of 1 liter of milk. No stud- Hence, it can be seen that the heritability and total ies have so far examined the functional features of this variation coefficients obtained in the studied group of gene; however, our present findings suggest that this SNP animals did not differ actually from previous studies on is associated with lactose content in milk: variants with a Polish HF population. Moreover, as the genotyped cows higher milk lactose content demonstrated lower costs of were descendants of 153 sires and 263 dams, the probe producing one liter of milk. The lowest costs of produc- from population was not overwhelmed by one or two ing 1 liter of milk were observed for the T/C heterozy- families. As such, our findings SNPs can be regarded as gotes and the highest for the C/C homozygotes. representative, despite the small sample size. 426 E. Bagnicka et al. analysis of gene polymorphisms CXCR4 gene protein product on the processes of the im- Commercially available SNP microarrays are em- mune system taking place in the cells of the udder tissues ployed in GWAS. However, a large majority of SNPs do and leukocytes of the milk alveoli, is extremely impor- not correlate with breeders’ interest-piqued traits. As a re- tant from the point of view of breeding work towards the sult, it is crucial to research all known SNPs and choose selection of cows resistant to inflammation of the udder. for GWAS only those that significantly affect the traits Further association studies are needed to check the that are taken into account in the breeding goal in a par- relationships with other diseases and productivity of the ticular population. Gene containing such SNPs in their cows, especially that both TLR4 and CXCR4 are in- structure can become the candidate for main genes or volved in biological processes connected with both im- SNP is considered as a marker that indicates the quality mune system and cell differentiation and death. All of and value of the genes linked to it. these procedures are necessary for an effective udder The rs136813430(T/C) SNP, associated in our own to function. One of the SNP is localized in the intron of work with lnSCC and identified in the TLR4 gene exon DLG2 gene. As a heterodimer formed with a related fam- (ENSBTAG00000006240) (chromosome 8) is highly ily member, protein product of the DLG2 gene may inter- polymorphic: a total of 36 SNPs have been identified act at postsynaptic sites to form a multimeric scaffold for in 14 breeds of cattle. TLR4 is believed to initiate non- the clustering of receptors, ion channels, and associated specific and adaptive immune responses that defend the signaling proteins. To our knowledge, the function of body against pathogens by inducing the overexpression DLG2 has not been studied in cattle till now. However, in of the pro-inflammatory cytokines IL-1, L-6 and IL-8 humans, it has a wide range of cellular functions includ- involved in non-specific immunity (Wang et al., 2007). ing those connected with nervous system, Parkinson’s Although the identified mutation located in the exon is disease or some kinds of cancers (Zhuang et al., 2019; synonymous, it might be that the polymorphic variant Shao et al., 2019). It interacts with the cytoplasmic tail of is responsible for intensifying the immune response of NMDA receptor subunits and with inward rectifying po- the udder against invasion by increasing production of tassium channels (UniProt Consortium, 2019, Q15700). pro-inflammatory cytokines. This may hasten the annihi - Until now, it has not been connected with lactogenesis. lation of the pathogen without initiating the specific re- Thus, further study is needed to identify its function in sponse and recruitment of leukocytes from the blood into bovine and its connection with lactose content, especially the udder. Therefore, SNP rs136813430(T/C) appears to that DLG2 is involved in many processes including bio- be a promising candidate in the selection of cows for re- synthetic, protein transport, or cellular protein modifica - sistance to mastitis, as demonstrated by the significant tion. role played by the TLR4 gene protein product during in- IGF-I gene (ENSBTAG00000011082) (chromosome flammatory processes of the udder, and our own findings 5 in bovine) protein product is also called somatomedin regarding the relationship between the SNP and lnSCC. C (Meuwissen et al., 2002). IGF-I mediates many biolog- So far, this SNP has not been indicated as a marker of ical processes, increasing glucose absorption, regulating mastitis or milk composition. cellular differentiation, or proliferation, inhibiting apop- Located close to the CXCR4 (ENSBTAG 0000001060) tosis, or increasing lipid synthesis (De la Rosa Reyna et gene (chromosome 2), the rs110785912(A/T) SNP is al., 2010). IGF-I is released from the liver in response also associated with lnSCC. The gene is a member of to growth hormone (GH) that controls growth and lac- the chemokine receptor family (Busillo and Benovic et tation, thus IGF-I is thought that controlling lactation al., 2007), and is involved in many developmental pro- through production of milk ingredients (Lucy, 2008). It cesses, as well as in many disease processes caused by is also produced locally in a tissue-specific manner. IGF- viruses, such as bovine non-cytopathic viral diarrhoea I can be used to monitor udder health by measuring its virus (BVDV) (Weiner et al., 2012). CXCR4 expres- concentration in milk (Liebe and Shams, 1998). The au- sion was found to be elevated in the secretory epithe- thors’ own research showed a relationship between SNP lial tissue in cow udder quarters following infection by rs110455063(C/G) located 34,194 nt 5’ from the IGF-I coagulase-positive and coagulase-negative staphylo- gene with the dry matter, and fat contents, and daily milk cocci compared to pathogen-free tissues (Kościuszuk yield. So far, in the association analyses, this SNP has not et al., 2017). The CXCR4 receptor is believed to play been associated with any trait of milk, so it is not possible a key role in lymphocyte transport and directing them to compare our results with the results of other authors. to the lymph nodes (Busillo and Benovic, 2007). So far, It is proposed to include the SNP found in IGF-I gene in the rs110785912(A/T) SNP has not been studied by other MAS or genomic selection to improve cow productivity, authors in association analyses. Based on our own re- but it is necessary to conduct further studies on greater sults and literature data on the significant involvement populations of dairy cattle. of CXCR4 in the activation of the immune system, this DGAT1 gene (ENSBTAG00000026356) (chromo- SNP has been proposed as a marker of udder health to some 14) codes for a microsomal enzyme, using diacylg- be used in Marked Assisted Selection (MAS). Confirma- lycerol and fatty acetyl coenzyme A as substrates to cata- tion of the close relationship between the tested SNP and lyze the final stage of triacylglycerol synthesis (Cases et SCC in milk, probably through the direct influence of the al., 1998). It influences fat metabolism and its yield and Marker SNPs vs. mastitis and dairy cattle farm profitability 427 percentage in milk. Many polymorphic loci have been the different population size or structure. It is recom- found in this gene and most of them are located in the in- mended that any population used in such a study should trons, promoter and untranslated regions (UTR) of exons. be subjected to GWAS testing. SNP analyzed in our study was located within this gene The analyzed SNPs were located within genes or clos- at position 1801,116 and was related to the fat content, er than around 500,000 nucleotides (nt) from the near- dry matter content and fat to protein ratio. Meredith et est gene. As a long distance between the marker and the al. (2013) showed a relationship between the studied SNP gene reduces the probability of inheriting the analyzed and the milk fat yield. Jiang et al. (2019) found a strong SNP with the gene, such distant loci cannot be reli- antagonistic pleiotropy between fat yield and milk and able markers of the traits coded by an individual gene. protein yield for this SNP; the C allele was responsible for Simulations (Kruglyak, 1999) and empirical analyses this extremely antagonistic pleiotropy between positive fat based on human data (Dunning et al., 2000) suggest yield and negative milk and protein yield, while the T al- that linkage disequilibrium (LD) extends only a few lele had antagonistic pleiotropy between negative fat yield kilobases (kb) around common SNPs; however, other and positive milk and protein yield. Chen et al. (2015) studies indicate that this distance can be greater than showed its association with SCC during inflammation 100 kb (Abecasisi et al., 2001) or even beyond 1 Mb caused by E. coli and S. uberis. Many studies have found (Taillon-Miller et al., 2000). Nevertheless, it is sug- that the region that includes the DGAT1 gene has a large gested that due to genetic drift, admixture, selection impact on milk fat content and on several other production and smaller effective population sizes, which reduce traits including milk yield, percentage of fat, and percent- allelic heterogeneity, livestock demonstrate greater age of protein (Ashwell et al., 2004; Maxa et al., 2012). LD than humans, and LD can extend over several hun- The fat:protein ratio was 1.15 for heterozygous cows: it dred base pairs (McRae et al., 2002). should be highlighted that it is the best ratio between these It is possible that the analyzed SNPs which exist traits, because it determines the best texture and taste for outside any gene sequences, may be found in enhancer cheese curd. DGAT1 can be considered a candidate gene (expression enhancer) or suppressor regions – they can for the fat content and fatty acid profile of milk. permanently turn off expression through an epigenetic si- The GPIHBP1 (ENSBTAG00000049125) (chromo- lencing mechanism; as such, they have a strong influence some 14) encodes a protein belonging to the lymphocyte on gene transcription (activation) and expression. More- antigen 6 (Ly6) family. GPIHBP1 plays a key role in the over, the promoter region, a section of genomic DNA transport and localization of lipoprotein lipase (LPL) upstream of the transcription start site (TSS) of the gene synthesized by myocytes and adipocytes and creates commonly referred to as the +1 position, contains regu- a platform for lipolysis in endothelial cells (Yang et al., latory information for transcription initiation. Enhancer- 2017). In our research rs109421300(T/C) SNP was relat- promoter interactions precisely control the spatiotempo- ed to the dry matter and fat contents and the fat to protein ral timing of the activation of particular genes (Danino et ratio. Meredith et al. (2013) showed that this SNP was re- al., 2015). The enhancers are cis-regulatory elements first lated to the yield of milk fat. Fang and Pausch (2019) also identified in the SV40 virus genome; these can influence proposed the inclusion of this GPIHBP1 gene for use in expression from a distance of even 1 mega base (1 Mb) selection to improve milk yield. As this SNP is located in and may be located within the target gene, or upstream or close distance to the GPIHBP1 gene (450,205 nt towards downstream from it. They show high specificity regard- 3’), so it is likely to be inherited along with it. ing cell type and response to stimuli. Complex regulatory SNP rs109421300(T/C) is closely located to CY- genes often have multiple enhancers, with each being re- P11B1 gene which encodes a member of the cytochrome sponsible for expression in a given cell type or situation. P450 superfamily. Proteins catalyze many reactions re- In addition, a single enhancer can act on many genes at lated to drug metabolism and the synthesis of choles- once. The genes whose expression is necessary for the terol or steroids and other lipids. This protein localizes function of the cells in each situation may not have any to the mitochondrial inner membrane and is involved in enhancers or silencers. Enhancers and silencers often the conversion of progesterone to cortisol in the adrenal cooperate to maintain the balance of tissue-specific pro- cortex. CYP11B1 affects cortisol production, androgen moter activity. Most SNPs have been found in noncoding function, and ultimately the proliferation of mammary regions containing enhancers (Xia and Wei, 2019). gland cells (Brettes and Mathelin, 2008). As the probability of inheriting a SNP with a gene is Two SNPs, rs110785912(A/T) and rs110455063(C/G), reduced when a long distance exists between marker and located within LOC1049123 and LOC104972477 loci functional gene, loci located at a large distance cannot be in the NCBI database are marked as genes of unknown reliable markers of the features encoded by a given gene, function. These SNPs were associated with lnSC Cor dry particularly when assessing the predicted breeding value matter, and fat contents. However, it was impossible to of offspring. However, as genomic selection allows indi- compare our results with the results of other authors, be- viduals with high breeding value to be selected based on cause there is no available literature information. SNP markers, all information on SNPs and their associa- Our results for some of the analyzed SNPs differ from tion with production traits is useful in the breeding work. previous studies. These differences are probably due to Genomic information allows genetically ideal animals to 428 E. Bagnicka et al. be identified at a very young age, with greater accuracy superfamily (MAGUK). This is a huge protein complex, than estimates based on the average genetic value of the with more than 1,000 members, including scaffold or parents. Further studies should examine the functions of cytoskeletal proteins, receptors and signaling enzymes. the protein products of studied genes; this would provide These proteins are a part of the postsynaptic protein scaf- a better understanding of their participation in metabolic fold of excitatory synapses, and contain various domains pathways and cellular processes associated with the im- (e.g., PDZ, GK, SH3) that enable them to bind to many mune system of cows and their relationship with produc- of the proteins present in synapses (Zhu et al., 2016). tion traits. Although no studies have examined the functional fea- tures of protein encoded by this gene, our findings indi- economic analysis cate that the SNP lying in this gene appears to be related SNP rs136813430(T/C) was found to have prom- to lactose content in milk. Lactose level is also an indi- ise for assessing potential milk production profitability. cator of udder health: its decrease often indicates in- Cows with the T/C genotype were characterized by the flammation of the udder, resulting in loss of production highest gross margin and direct profitability index, as and additional costs related to treating cows (Bruckmaier well as the lowest costs incurred to produce one liter of et al., 2004). According to the conducted research, vari- milk. This SNP probably influences the number of so- ants with a higher milk lactose content were character- matic cells (SCC) possibly due to the role played by the ized by a lower level of production costs per liter of milk. protein encoded by the TLR4 gene, in which the SNP is In this analysis, the lowest production costs for one liter of located. milk were recorded for T/C heterozygotes and the highest Many studies have confirmed the importance of Toll- for C/C homozygotes. Therefore, there is a certain contra- like receptors (TLRs), including TLR4, in the first line of diction between the selection and economic goals. defense against invading pathogens, where they are be- To summarize, the present study examines the influ- lieved to initiate the innate immune response (Kawai and ence of selected SNP polymorphisms occurring in genes Akira, 2005; De Schepper et al., 2008). Concerning the related to udder health and milk production on the profit- lower SCC content, the T/C variant may bestow higher ability of milk production. The analysis was compared resistance to inflammation of the udder. Therefore, it is with data concerning production value and direct costs an interesting candidate for the purposes of breeders. which was obtained from commercial farms. An analy- Information about the variants of SNP sis of variance revealed significant differences between rs136813430(T/C) may therefore prove useful when allelic variants in the case of three tested SNPs. The selecting cows with a specific genotype in programs in- heterozygous (T/C) variant of SNP rs136813430(T/C) tended to regulate the content of fatty acids in milk; prop- was associated with a low content of lnSCC; it was also er selection will guarantee not only an appropriate fatty characterized by the highest gross margin, the highest di- acid content, but also improve economic profitability. rect profitability index and the lowest costs incurred to In the case of SNP rs110785912(T/A) associated with produce one liter of milk (P = 0.01). This therefore ap- lnSCC in milk, a significant correlation was found with pears to be the most promising of the tested SNPs for se- the direct profitability index. The SNP lies in proximity lection purposes. The T/A variant of rs110785912(T/A) to the CXCR4 gene, the product of which is involved in demonstrated low lnSCC content in milk and the high- the immune system of cattle (Weiner et al., 2012). It is est direct profitability index, while the C/C variant of therefore possible that the area within the CXCR4 gene, rs41587003(A/C) demonstrated lowest lactose level and and thus also in the rs110785912(T/A) locus, may be the highest costs of producing one liter of milk. The ob- involved in the functioning of the immune system of tained results indicate that there is an economic justifica - cows; however, this requires additional research. The tion for including SNP variants located within, or close most favorable genotype in terms of the profitability in- to, genes involved in the immune system and milk pro- dex turned out to be the heterozygous T/A variant, which duction in cattle selection. was associated with the lowest content of somatic cells. As SCC is an indirect indicator of udder health, the low conclusions cell content may mean that cows with this genotype are While our cow sample size may seem somewhat small less susceptible to mastitis than those with other gene for this kind of study the bias can mainly regard the vari- variants. This variant therefore offers promise in selec- ance component magnitudes of the phenotypic outcome tion processes intended to increase the resistance to in- and SNP alleles’ frequencies, as the entire population flammation while ensuring high milk production profit- variation could have been narrowed in the sample. We ability. have, however, discussed this issue with the conclusion On the other hand, regarding SNP rs41587003(A/C), that our sample is suitable for the undertaken research. significant relationships were found between the C/C and On the other hand, the SNP-phenotypic output associa- T/C variants with regard to the cost of producing one liter tion analysis is much less dependent on the sample size of milk. The studied SNP is in the intron of the DGL2 (if at all, given our sample size). gene called also Postsynaptic Density Protein PSD-93. Hence, the above let us recommend the It encodes a protein belonging to the guanylate kinase rs136813430(T/C) SNP, located within the TLR4 gene, Marker SNPs vs. mastitis and dairy cattle farm profitability 429 (2015). Second-generation PLINK: rising to the challenge of larg- as a candidate gene for explaining the variance of SCC er and richer datasets. Gigascience, 4: 7. in milk, as indicative of susceptibility to mastitis, but Chen X., Cheng Z., Zhang S., Werling D., Wathes D.C. (2015). Com- also for explaining the variance of the gross margin and bining genome wide association studies and differential gene ex- the direct dairy profitability index. The latter is, to our pression data analyses identifies candidate genes affecting mastitis caused by two different pathogens in the dairy cow. Open J. Anim. knowledge, the first effort to assess directly a correlation Sci., 5: 358–393. between the DNA polymorphism and economic output of Danino Y.M., Even D., Ideses D., Juven-Gershon T. (2015). The core a dairy enterprise. promoter: At the heart of gene expression. BBA-Gene Regul. Mech., 1849: 1116–1131. Davies G., Genini S., Bishop S.C., Giuffra E. (2009). An assessment of opportunities to dissect host genetic variation in resistance to acknowledgements infectious diseases in livestock. Animal, 3: 415–436. The authors thank prof. Marek Łukaszewicz for in- De la Rosa Reyna X.F., Montoya H.M., Castrellón V.V., Rincón valuable help in improving the quality of this work. The A.M.S., Bracamonte M.P., Vera W.A. (2010). Polymorphisms in authors gratefully acknowledge the Polish Federation of the IGF1 gene and their effect on growth traits in Mexican beef cattle. Gen. Mol. Res., 9: 875–883. Cattle Breeders and Dairy Farmers for phenotypic data De Schepper S., De Ketelaere A., Bannerman D.D., Paape M.J., Peel- on the daily milk yield and the component contents and man L., Burvenich C. (2008). The toll-like receptor-4 (TLR-4) pedigree file studied dairy cows. pathway and its possible role in the pathogenesis of Escherichia coli mastitis in dairy cattle. Vet. Res., 39: 5. Dunning A.M., Durocher F., Healey C.S., Teare M.D., McBride S.E., Carlomagno F., Xu C-F., Dawson E., Rhodes S., Ueda S., Lai E., references Luben R.N.., Van Rensburg E.J., Mannermaa A., Kataja V., Renn- art G., Dunham I., Purvis I., Easton D., Ponder B.A.J. (2000). The extent of linkage disequilibrium in four populations with distinct Abecasis G.R., Noguchi E., Heinzmann A., Traherne J.A., Bhattacha- ryya S., Leaves N.I., Anderson G.G., Zhang Y., Lench N.L., Carey demographic histories. Am. J. Hum. Genet., 67: 1544–1554. A., Cardon L.R., Moffatt M.F. Cookson W.O.C. (2001). Extent Fang Z.H., Pausch H. (2019). Multi-trait meta-analyses reveal 25 quantitative trait loci for economically important traits in Brown and distribution of linkage disequilibrium in three genomic re- gions. Am. J. Hum. Genet., 68: 191–197. Swiss cattle. BMC Genomics, 20: 695. Aerts J., Piwczyński D., Ghiasi H., Sitkowska B., Kolenda M., Önder Hagnestam-Nielsen C., Østergaard S. (2009). Economic impact of H. (2021). Genetic parameters estimation of milking traits in Pol- clinical mastitis in a dairy herd assessed by stochastic simula- tion using different methods to model yield losses. Animal, 3: ish Holstein-Friesians based on automatic milking system data. Animals, 11: 1943. 315–328. Andrei S., Pintea A., Bunea A., Groza I., Bogdan L., Ciupe S., Matei Jiang J., Ma L., Prakapenka D., Van Raden P.M., Cole J.B., Da Y. (2019). A large-scale genome-wide association study in U.S. Hol- S., Crainic D. (2009). Non-enzymatic antioxidants concentration and lipids peroxidation level in milk from cows with subclinical stein cattle. Front. Genet., 10: 412. mastitis. Bull. Univ. Agric. Sci. Vet. Med. Cluj. Napoca Vet. Med., Kanehisa M., Sato Y., Kawashima M., Furumichi M., Tanabe M. 66. (2016). KEGG as a reference resource for gene and protein an- notation. Nucleic Acids Res., 44: D457–D462. Ashburner M., Ball C.A., Blake J.A., Botstein D., Butler H., Cherry J.M., Davis A.P., Dolinski K., Dwight S.S., Eppig J.T., Harris Kawai T., Akira S. (2005). Pathogen recognition with toll-like recep- M.A., Hill D.P., Issel-Tarver L., Kasarskis A., Lewis S., Matese tors. Curr. Opin. Immunol., 17: 338–344. J.C., Richardson J.E., Ringwald M., Rubin G.N., Sherlock G. Kosciuczuk E.M., Lisowski P., Jarczak J., Majewska A., Rzewuska M., Zwierzchowski L., Bagnicka E. (2017). Transcriptome profil - (2000). Gene ontology: tool for the unification of biology. Nat. Genet., 25: 25–29. ing of Staphylococci-infected cow mammary gland parenchyma. Ashwell M.S., Heyen D.W., Sonstegard T.S., Van Tassel C.P., Da P.M., BMC Vet. Res., 13: 1–12. Krag K., Poulsen N.A., Larsen M.K., Larsen L.B., Jans L.L., Buiten- Ron M., Weller J.I. (2004). Detection of quantitative trait loci af- fecting milk production, health, and reproductive traits in Holstein huis B. (2013). Genetic parameters for milk fatty acids in Danish cattle. J. Dairy Sci., 87: 468–475. Holstein cattle based on SNP markers using a Bayesian approach. Bagnicka E., Łukaszewicz M., Ådnøy T. (2016). Genetic parameters BMC Genet., 14: 1471–2156. Kruglyak L. (1999). Prospects for whole-genome linkage disequilibrium of somatic cell score and lactose content in goat’s milk. J. Anim. Feed Sci., 25: 210–215. mapping of common disease genes. Nature Genet., 22: 139–144. Brettes J.P., Mathelin C. (2008). Effect dual DeVuyst s androgens sur Liebe A., Shams D. (1998). Growth factors in milk: interrelationships with somatic cell count. J. Dairy Res., 65: 93–100. la glande mammaire. Bull. Cancer, 95: 495–502. Bruckmaier R.M., Ontsouka C.E., Blum J.W. (2004). Fractionized Lü A., Hu X., Chen H., Dong Y., Zhang Y., Wang X. (2011). Novel milk composition in dairy cows with subclinical mastitis. Vet. SNPs of the bovine PRLR gene associated with milk production Med. Czech, 49: 283–290. traits. Biochem. Genet., 49: 177–189. Lucy M. (2008). Functional differences in the growth hormone and Brzóska F., Kowalski Z.M., Osięgłowski S., Strzetelski J. (2014). IZ PIB-INRA Standard of ruminants’ feeding: nutrient value of insulin-like growth factor axis in cattle and pigs: Implications for French and domestic fodders for ruminants (in Polish), Strzetelski post-partum nutrition and reproduction. Reprod. Domest. Anim., 43: 31–39. J. (ed.). Foundation IZ PIB Patronus Animalium, Kraków, Poland. Busillo J.M., Benovic J.L. (2007). Regulation of CXCR4 signaling. Madsen P., Jensen J. (2013). A user’s guide to DMU – a package for Biochim. Biophys. Acta, 1768: 952–963. analysing multivariate mixed models. Version 6; release 5.2. Carlen E., Strandberg E., Roth A. (2005). Genetic parameters for clini- Marogna G., Rolesu S., Lollai S., Tola S., Leori G. (2010). Clinical findings in sheep farms affected by recurrent bacterial mastitis. cal mastitis, somatic cell score, and production in the first three lactations of Swedish Holstein cows. J. Dairy Sci., 87: 3062–3070. Small Rumin. Res., 88: 119–125. Cases S., Smith S.J., Zheng Y.W., Myyers H.M., Lear S.R., Sande E., Maxa J., Neuditschko M., Russ I., Förster M., Medugorac I. (2012). Novak S., Collins C., Welch C.B., Lusis A.J., Erickson S.K. Fa- Genome-wide association mapping of milk production traits in Braunvieh cattle. J. Dairy Sci., 92: 5357–5364. rese R.V. (1998). Identification of a gene encoding an acyl CoA: diacylglycerol acyltransferase, a key enzyme in triacylglycerol McRae A.F., McEwan J.C., Dodds K.G., Wilson T., Crawford A.M., synthesis. PNAS USA, 95: 13018–13023. Slate J. (2002). Linkage disequilibrium in domestic sheep. Genet- ics, 160: 1113–1122. Chang C.C., Chow C.C., Tellier L.C., Vattikuti S., Purcell S.M., Lee J.J. 430 E. Bagnicka et al. Meredith B.K., Berry D.P., Kearney F., Finlay E.K., Fahey A.G., Brad- Sordillo L.M. (2005). Factors affecting mammary gland immunity and ley D.G., Lynn D.J. (2013). A genome-wide association study for mastitis susceptibility. Livest. Prod. Sci., 98: 89–99. somatic cell score using the Illumina high-density bovine bead- Taillon-Miller P., Bauer-Sardiña I., Saccone N.L., Putzel J., Laitinen chip identifies several novel QTL potentially related to mastitis T., Cao A., Kere J., Pilia G., Rice J.P., Kwok P-Y. (2000). Juxta- susceptibility. Front. Genet., 4: 229. posed regions of extensive and minimal linkage disequilibrium in Methodology of calculating the gross margin for agricultural produc- human Xq25 and Xq28. Nat. Genet., 25: 324–328. tion activity (in accordance with the standards of the European UniProt Consortium. (2019). UniProt: a worldwide hub of protein Union (in Polish). Institute of Agricultural and Food Economics- knowledge. Nucleic acids research, 47: D506–D515. https://www. National Research Institute (IERiGŻ), Warszawa, 1999. uniprot.org/uniprot/Q15700, access, December 15, 2021. Meuwissen T.H.E., Karlsen A., Lien S., Olsaker I., Goddard M.E. Wang Y., Chen T., Han C., He D., Liu H., An H., Cai Z., Cao X. (2007). (2002). Fine mapping of a quantitative trait locus for twinning Lysosome-associated small Rab GTPase Rab7b negatively regu- rate using combined linkage and linkage disequilibrium mapping. lates TLR4 signaling in macrophages by promoting lysosomal Genetics, 161: 373–379. degradation of TLR4. Blood, 110: 962–971. NCBI-1, National Center for Biotechnology Information, https://www. Weiner C.M., Smirnova N., Webb B.T., Van Campen H., Hansen T.R. ncbi.nlm.nih.gov, access May 2019. (2012). Interferon stimulated genes, CXCR4 and immune cell re- NCBI-2, National Center for Biotechnology Information, https://www. sponses in peripheral blood mononuclear cells infected with bo- ncbi.nlm.nih.gov/genome/?term=LOC104972477, access Decem- vine viral diarrhea virus. Res. Vet. Sci., 93: 1081–1088. ber 14, 2021. Xia J.H., Wei G.H. (2019). Enhancer dysfunction in 3D genome and Oprządek J., Sender G., Pawlik A., Łukaszewicz M. (2015). Locus disease. Cells, 8: 1281. BoLA-DRB3 is just an ordinary site of the polygene when ex- Yang J., Liu X., Wang D., Ning C., Wang H., Zhang Q., Jiang L. plaining genetic variance of somatic cell count and milk yield. J. (2017). Functional validation of GPIHBP1 and identification of Dairy Res., 82: 449–452. a functional mutation in GPIHBP1 for milk fat traits in dairy cat- Ptak E., Brzozowski P., Jagusiak W., Zdziarski K. (2007). Genetic pa- tle. Sci. Rep., 7: 8546. rameters for somatic cell score for Polish Black-and-White cattle Yazgan K., Makulska J., Węglarz A., Ptak E., Gierdziewicz M. (2010). estimated with a random regression model. J. Anim. Feed. Sci., Genetic relationship between milk dry matter and other milk traits 16: 357–369. in extended lactations of Polish Holstein cows. Czech J. Anim. Rzewuska K., Strabel T. (2013). Genetic parameters for milk urea Sci., 55: 91–104. concentration and milk traits in Polish Holstein-Friesian cows. J. Yuan Z., Li J., Li J., Gao X., Xu S. (2013). SNPs identification and its Appl. Genet., 54: 473–482. correlation analysis with milk somatic cell score in bovine MBL1 Sender G., Korwin-Kossakowska A., Pawlik A., Hammed K., gene. Mol. Biol. Rep., 40: 7–12. Oprządek J. (2013). Genetic basis of mastitis resistance in dairy Zhu J., Shang Y., Zhang M. (2016). Mechanistic basis of MAGUK- cattle – a review. Ann. Anim. Sci., 13: 663–673. organized complexes in synaptic development and signalling. Nat. Shao Y.W., Wood G.A., Lu J., Tang Q.L., Liu J., Molyneux S., Chen Y., Rev. Neurosci., 17: 209–223. Fang H., Adissu H., McKee T., Waterhouse P., Khokha R. (2019). Zhuang R.J., Bai X.X., Liu W. (2019). MicroRNA-23a depletion pro- Cross-species genomics identifies DLG2 as a tumor suppressor in motes apoptosis of ovarian cancer stem cell and inhibits cell mi- osteosarcoma. Oncogene, 38: 291–298. gration by targeting DLG2. Cancer Biol. Ther., 20: 897–911. Skarżyńska A. (2012). The impact of the milk yield of cows on the profitability of milk production (in Polish). Zagadnienia Eko - Received: 10 IV 2022 nomiki Rolnej (Probl. Agricult. Econom.), 330: 90–111. Accepted: 29 XI 2022 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Annals of Animal Science de Gruyter

The association of gene polymorphisms with milk production and mastitis resistance phenotypic traits in dairy cattle

Download PDF
12 pages

Loading next page...
 
/lp/de-gruyter/the-association-of-gene-polymorphisms-with-milk-production-and-JXjqkqFjss

References (40)

Publisher
de Gruyter
Copyright
© 2023 Emilia Bagnicka et al., published by Sciendo
ISSN
1642-3402
eISSN
2300-8733
DOI
10.2478/aoas-2022-0091
Publisher site
See Article on Publisher Site

Abstract

Ann. Anim. Sci., Vol. 23, No. 2 (2023) 419–430 DOI: 10.2478/aoas-2022-0091 The associa Tion of gene polymorphisms wiTh milk producTion and masTiTis resisT ance phenoTypic TraiTs in dairy ca TTle* 1♦ 1,2 3 4 Emilia Bagnicka , Paulina Brzozowska , Kacper Żukowski , Renata Grochowska Institute of Genetics and Animal Biotechnology Polish Academy of Sciences, Postępu 36A, 05-552 Jastrzębiec, Poland Faculty of Dietetics, Warsaw College of Engineering and Health, Bitwy Warszawskiej 1920 r. 18, 02-366 Warszawa, Poland National Research Institute of Animal Production, 32-083 Balice n. Kraków, Poland Institute of Agricultural and Food Economics – National Research Institute, Świętokrzyska 20, 00-002 Warszawa, Poland Corresponding author: e.bagnicka@igbzpan.pl abstract The aim of this study was to evaluate the association between gene polymorphisms (snps) and mastitis indicators and their relationship with milk production profitability in dairy herd. A functional analysis was also performed of five genes containing the studied SNPs and those located close by. DNA was isolated from the hair bulb of 320 dairy cows kept in three herds and SNP-microarray analysis was performed. The data on 299 cows was subjected to final statistical analysis using AI-REML method with one-trait repeatability test-day animal model and pedigree information using the DMU4 package. Five from 35 SNPs significantly associated with mastitis indicators or production traits and located within a gene or no more than 500,000 nucleotides from the gene were selected for the functional and economic analysis. A questionnaire was also developed to collect associated economic data of 219 cows from three herds, such as the value of milk production and direct costs incurred over three years; this allowed the gross margin, direct profitability index and direct costs incurred to produce one liter of milk to be determined, among others. None of the five studied SNPs were related to protein content. The rs110785912(T/A), found near CXCR4, and rs136813430(T/C), located in the TLR4 gene exon, were associated with lnSCC, while rs110455063(C/G), located near IGF-I, was associated with milk yield, fat and total solid contents. rs109421300(T/C), associated with fat/ protein content ratio, as well as fat and total solid content, is located in the DGAT1 gene intron. rs41587003(A/C), located in the DLG2 gene intron, was associated with lactose content. The economic analysis revealed differences between the variants of the three tested SNPs. The T/C variant of the rs136813430(T/C) SNP was characterized by the highest gross margin, the highest direct profitability index and the lowest costs incurred to produce 1 liter of milk. The T/A variant of rs110785912(T/A) was related to low lnscc and was charac - terized by the highest direct profitability index. In turn, the C/C variant of the rs41587003(A/C) was related to the lowest level of lactose and the highest costs of milk production. It appears that rs136813430(T/C) may be the most promising of the tested snps for increasing the profitability of milk production. To our knowledge, it is the first effort to assess directly a correlation between the DNA polymorphism and economic output of a dairy enterprise. Key words: SNP marker, intron, exon, association, mastitis indicator, dairy economics, milk production profitability Mastitis is still one of the major diseases affecting breeding progress is rather slow, as genetic variability di- dairy cattle, with negative effects on milk production rectly related to resistance accounts for only a small part (Davies et al., 2009). Despite the wealth of research per- of the total variance; however, this value is significant. formed on the physiological and cellular processes tak- The other trait considered as additive mastitis indicator ing place in the mammary gland in response to infection, is lactose content (Bagnicka et al., 2016; Andrei et al., knowledge of many of the defense mechanisms remains 2009). incomplete (Sordillo, 2005). Mastitis is negatively cor- Inflammation is a complex trait controlled by many related with milk yield, with high-yielding cows being genes. As such, it is difficult to develop an appropriate more susceptible (Carlen et al., 2005). selection strategy to increase immunity. Such programs Most genetic studies of mastitis indicators to date are complicated by the lack of information on the ge- have focused on the number of somatic cells (SCC) in netic basis of animal immunity, the complexity of the milk. The heritability coefficient (h ) for SCC was esti- functional interaction between host and pathogen, which mated from 0.05 to 0.20 (Ptak et al., 2007; Sender et al., may trigger a range of different immune responses, and 2013). With such a low to moderate degree of heritability, the influence of the environment (Marogna et al., 2010). __________ *This study was supported by the Leading National Research Centre Scientific Consortium’s “Healthy Animal – Safe Food” initiative under the Ministry of Science and Higher Education decision No. 05-1/KNOW2/2015, grant KNOW2015/CB/PRO1/52, to Emilia Bagnicka. The pub- lication costs were co-financed by the subsidy of the Ministry of Agriculture and Rural Development (DŻW.pp.862.1.2023). 420 E. Bagnicka et al. material and methods The search for genes related to the susceptibility/resist- ance of dairy cows to udder inflammation and production animals and farms traits has recently been supported by the development of The first step of the research was carried out on 320 genome-wide association studies (GWAS) using pheno- dairy cows from three herds. All were kept in free stall typic data and data on the genotypes of the studied indi- barns with constant access to enclosures. The farms dif- viduals. One of the most popular methods of genotyping fered in terms of the number of cows and feeding system, concerns the use of dense sets of SNP (Single Nucleotide resulting from diversified area of arable land, including Polymorphism) microarrays. permanent grassland. The mean herd yield was 9,393 lit- Taking advantage of genomics for breeding purposes ers of milk per lactation, with mean fat and protein con- raises the question of whether the use of selection meth- tents of 4.18% and 3.49%, respectively. Cows were fed in ods based on SNPs influencing the immunity of the cow the total mixed ration (TMR) system, balanced in accord- mammary gland is economically justified, i.e., whether ance with the Institut national de la recherche agronom- this approach will reduce losses related to udder diseas- ique (INRA) standards and adapted to Polish conditions es and thus increase the efficiency of dairy cows, and, (Brzóska et al., 2014), with constant access to water. The consequently, reduce milk production costs in a herd. cows were milked twice a day with a mechanical milking The future of livestock production is not about produc- system (DeLaval, Tumba, Sweden). All herds were as- tion intensification at all costs, which requires increas- sessed for dairy and breeding performance under routine ing expenditure to account for more limited resources performance control. Phenotypic data on the daily milk and environmental pollution. A more desirable goal is to yield and the component contents was obtained from the minimize production costs thanks to the use of inter alia Polish Federation of Cattle Breeders and Dairy Farmers, new genomic technologies, which enable the selection based on which a database with 8090 observations was of individuals with specific genetic potential in terms of prepared. Information on daily milk yield, protein, fat, production and health. dry matter, lactose contents and SCC were gathered. The Most of the research on udder health to date has fo- yield of milk components and fat/protein ratio were also cused on estimating the costs of preventing inflamma- calculated. The descriptive statistic of all studied traits is tion, losses due to mastitis, and effectively managing presented in Table S1. The pedigree database contained a herd to reduce such losses (Hagnestam-Nielsen and information on 7,035 animals, along with their birth Østergaard, 2009). The use of new technologies based year, and 2,057 sires and 4,634 dams with 64 sires hav- on SNPs has increased the efficiency of selecting eco- ing more than 10 daughters (range between 10 and 61). nomically significant traits. Hence, a number of studies However, the genotyped cows were descendants of 153 have been carried out in cattle to identify relationships sires (54, 40 and 59 sires in each herd respectively) and between gene polymorphisms and various production 263 dams. Ten sires of genotyped cows had daughters in and functional traits (Lü et al., 2011; Yuan et al., 2013). three or two herds. Previous studies have only indicated the potential of economic benefits related to higher meat and milk pro- Sampling and gene polymorphism identification duction or lower SCC in milk; however, these traits DNA was isolated from approximately 100 hair fol- do not guarantee greater milk production profitability, licles from each cow using a kit for isolating DNA from which is determined by the production volume and the blood and tissues (Macherey-Nagel, Düren, Germany). costs incurred. None of these studies have examined the The procedure was performed according to the manu- profitability of using individual SNPs in cattle selection facturer’s protocol, with the following modification: the based on real economic data like gross margin, direct hair root cells were disintegrated by dipping four times profitability index and costs incurred to produce 1 liter in liquid nitrogen, and then thawing in a water bath at of milk. Therefore, this paper attempts to fill this gap by 37°C and increasing the amount of added proteinase analyzing the economic effects of SNPs in genes related from 25 µl to 50 µl. Just before performing the microar- to milk production and health based on information re- ray analysis, quantitative DNA analysis was performed garding the production value and direct costs obtained using intercalating dyes capable of specifically binding to from farms. double-stranded DNA (Quant-iT PicoGreen dsDNA kit, It was hypothesized that the use of gene polymor- Thermo Fisher Scientific). Fluorescence measurement phisms related to udder inflammation in dairy cows was performed using a Fluoroskan Ascent FL appara- would increase milk production profitability; there- tus (Thermo Fisher Scientific, Waltham, Massachusetts, fore, the aim of this study was to evaluate the associa- USA) according to the manufacturer’s protocol. Eight tion between impact of five selected polymorphisms samples with concentrations below 50 ng DNA/µl were (SNPs) and mastitis indicators or production traits as excluded from the further analysis. well as to investigate the impact of selected identi- Microarray analysis was performed on 120 ng of DNA fied SNPs on the profitability of milk production in of 312 cows from each run using the BeadChip Bovine a dairy herd. A functional analysis was also made of 50k v3 kit. The results were interpreted using the Genom- the genes that include the studied SNPs and genes lo- eStudio 1.9.4 software (Illumina, San Diego, CA, USA). cated nearby. Marker SNPs vs. mastitis and dairy cattle farm profitability 421 As the DNA was isolated from hair follicles, no per- the analysis to estimate the differences between variants mission was needed from the Local Ethics Committee for of particular SNPs. Post hoc test with the Bonferroni cor- biological material collection. rection was used. Five from 35 SNPs significantly asso- ciated with mastitis indicators or then with production association study traits and located within a gene or no more than 500,000 The next step in establishing the association between nucleotides from the gene were selected for the function- individual SNPs and selected mastitis indicators and milk al and economic analysis. traits was to conduct a GWAS-based basic bioinformatic Before statistical analysis, the total SCC were trans- analysis using the Plink.2 program (Chang et al., 2015) formed to natural logarithm values (lnSCC). χ2 test was for more than 50K SNPs; a linear regression model was applied to determine whether SNP genotypes frequency used for this purpose, taking into account the fixed effect and Hardy–Weinberg equilibrium held. of SNP and population structure among individuals as co- variate. Genotypes with low frequency and SNPs which Functional analysis of genes with analyzed SNPs were monomorphic in the studied group were excluded Seven genes associated with studied SNPs were sub- from the analysis. As a result, the number of SNPs test- jected to functional analysis using the Kyoto Encyclope- ed was reduced down to 46K. All phenotypic traits were dia of Genes and Genomes (KEGG) pathway database averaged using an interquartile distribution: IQR(x) = (Kanehisa at al., 2016) and Gene Ontology Annotation quantile(x. 3/4) - quantile(x. 1/4). In order to check the dif- (GO) (Ashburner et al., 2000). ferences in the environment between the studied herds, the multidimensional scaling analysis (MDA) was performed. economic analysis Genotyping rating was 0.98. All outliers, i.e. information To assess the value of the SNP polymorphisms for on 13 cows, were removed from further analysis. selection purposes, a questionnaire was developed to col- Following the above, a preliminary analysis of the lect the actual data on milk production of 219 cows and associations between SNPs and mastitis indicators and the direct costs incurred over three calendar years, i.e., milk traits was conducted using the PROC MIXED SAS 2015, 2016 and 2017 in three herds. Altogether, 384 re- package (SAS/STAT 2002–2012) with a mixed model cords were analyzed. The analysis was based on calcula- using information on production of 299 genotyped cows. tions in accordance with the AGROKOSZTY methodol- All SNPs which were found to be associated with ud- ogy (Methodology, 1999; Skarżyńska, 2012). der health in the previous step of the analysis (lnSCC– The research covered revenues, i.e., the value of po- a natural logarithm from SCC, lactose content) and with tential commercial production (assuming that the sales production traits (milk yield, protein and fat content or volume is equal to the production volume) and direct yield) were taken into account. Altogether, 52 SNPs were costs. The gross margin was adopted as the basic meas- selected, of which only 35 were significantly associated ure of the assessment of the obtained economic effects. with the studied traits in this analysis of variance. Finally, The calculations do not take into account support with these 35 SNPs were analyzed by AI-REML to obtain so- subsidies under the Common Agricultural Policy. In lutions for fixed effects, using the DMU4 package (Mad- calculating the gross margin, only the value of the main sen and Jensen, 2013). A one-trait repeatability test-day product, i.e., milk, traded on the market was considered. animal model was applied. The annual value of milk production of each genotyped Phenotypic data was distributed among 21 calving cow was calculated based on the monthly milk yield of years from 1997 to 2017 – each year constituted a sep- the genotyped cows, and the mean monthly selling prices arate class and calving season: 1 – Winter (December, of milk for a given farm. January, February), 2 – Spring (March, April, May), 3 Due to the inability to record information on the in- – Summer (June, July, August), and 4 – Autumn (Sep- curred direct costs per genotyped cow, the gross margin tember, October, November). Daily milking data were was calculated based on the direct costs collected for the collected across 1996–2017. Finally, 150 classes of inter- entire dairy cow herd, and this value was converted to action were created between herd, calving year and sea- a single cow in the herd (Skarżyńska, 2012). For this pur- son (HYS) and 537 classes of interaction between herd, pose, information was obtained on the costs of feed from year and month of milking (hym). Parity was grouped outside the farm, of own feed from potential commercial into four classes, with the fourth containing lactations and non-commodity products, and of veterinary meas- higher than the third. The final statistical classification ures and services (including semen and insemination) fitted HYS, parity, and SNPs along with the Legendre and specialist costs (e.g., hoof correction, preparation of polynomial of order 5 regression on standardized DIM animals for exhibitions). nested within parity as fixed effects and genetic additive, Thus, milk production profitability for the entire herd permanent environment, hym, and residual effects as ran- of dairy cows in individual farms was determined based on dom (File S1). Each analysis included the relationship milk production value, calculated as the product of the aver- between one or two SNPs and the tested features at a time age milk yield per cow in a year and the mean annual milk due to computational limitation which is the shortage of price for a given herd, and direct costs, estimated based on the study as not all epistatic influences are considered in direct costs incurred in the entire herd in a year, per cow. 422 E. Bagnicka et al. The following economic categories were considered: alleles is presented in Table 2. Studied SNPs were asso- gross margin (GM) = PV – DC ciated with traits at least at P<0.05, except rs110785912 where: PV – production value (PLN); DC – direct which was associated with lnSCC at P = 0.0763. For costs (PLN) rs110785912(A/T) and rs110455063(C/G), a very high direct profitability index (DP) = (PV / DC) × 100 frequency of one allele was reported. The χ test results where: PV – production value (PLN), DC – direct indicate that the observed frequency of genotypes does costs (PLN) not differ from that expected one (Table 2). direct costs incurred to produce 1 liter of milk The first studied SNP, rs110785912(T/A), also known (MDC) = DC / MP as ARS-BFGL-NGS-50482, associated with lnSCC in where: DC – direct costs (PLN), MP – milk produc- our study, is located in the intron of a gene marked as tion (liters) LOC1049123 (NCBI-1). The function of the gene is un- To determine the significance of differences between known. The SNP identified 447,880 nucleotides from the tested SNP variants, an analysis of variance with the C-X-C chemokine receptor type 4 (CXCR4) gene a single-factor model was performed for each economic (Table 3). A/A and A/T genotypes were associated with category; in each calculation, a single SNP and the par- lower lnSCC than T/T (Table S2). The lnSCC herit- ity over the three studied years were used as fixed ef- ability coefficient estimated in the tested population was fects. The calculations were performed using the PROC 0.13, with the investigated trait demonstrating high co- MIXED SAS package (SAS/STAT 2002–2012). efficient of variance (CV=34.5%) which reflects a total variability. SNP rs110455063(C/G) known also as ARS-BF- results GL-NGS-15787, was located in the intron of gene with unknown function and marked as LOC104972477 gene polymorphism analysis (NCBI-2) and, simultaneously, was located 34,194>(5’) The multidimensional scaling analysis (MDA), nucleotides from the insulin-like growth factor I (IGF-I) which presents the environmental differences between gene. Its frequency was 0.228 for G/G, 0.759 for C/G, the studied herds and Manhattan plots as the result of and 0.013 C/C (Table 2). It was found to be associated GWAS analysis for some studied traits, is presented with milk yield, fat and total solid content. G/G variant in supplementary file (Figure S1 a and S1 b, A–E). carriers had the highest milk production while the highest After three-step selection, based on the present find- fat content was found in the milk of cows with the C/C ings SNPs: rs110785912(T/A), rs110455063(C/G), genotype, and the lowest in heterozygotes. The highest rs109421300(T/C), rs41587003(A/C), and rs136813430 dry matter content was observed in the milk of heterozy- (T/C) were selected for the functional and economic gous cows and the lowest content in cows with C/C. The analysis. total variability of the milk yield expressed as CV was The outcomes of the association analysis of selected high as 60%, while fat was 22.5% and dry matter content SNPs with functional and production traits of dairy cows only 8.37%. The heritability coefficients were 0.38, 0.17 are shown in Table 1, while the distribution of genes and and 0.20, respectively (Table S2). Table 1. Relationship of the studied SNPs with the functional and production characteristics of dairy cows Trait Chromosome rs number P-value** production traits Milk yield 5 rs110455063 0.021 Fat content 5 rs110455063 <0.0001 14 rs109421300 Dry matter content 5 rs110455063 0.0004 14 rs109421300 <0.0001 health mammary gland indicators lnSCC 2 rs110785912 0.0763 8 rs136813430 <0.0001 Lactose content 29 rs41587003 0.02701 **P-value after Bonferroni correction. lnSCC – natural logarithm of the number of somatic cells in milk (SCC). Marker SNPs vs. mastitis and dairy cattle farm profitability 423 Table 2. Distribution of genotypes and alleles of analyzed SNPs Alleles’ frequency/ χ test value/ SNP N Distribution of genotype in the studied population Genotypes’ frequency P-value rs110785912(A/T) 299 (A/A)=243 (AT)=55 (T/T)=1 A=0.905; T=0.095 1.33 /0.813|0.184|0.003| 0.40 rs110455063(C/G) 299 (G/G)=227 (CG)=68 (C/C)= 4 G=0.873; C=0.127 0.19 /0.228|0.759|0.013 0.83 rs136813430(T/C) 294 (C/C)=112 (T/C)=131 (T/T)=51 C=0.604; T=0.396 1.39 /0.380|0.450|0.170 0.24 rs109421300(T/C) 299 (A/A)= 140 (A/G)=137 5(G/G)=22 A=0.697; G=0.303 1.2 /0.468|0.458|0.074 0.14 rs41587003(A/C) 299 (A/A)=101 (AC)= 141 (C/C)=57 A=0.694; C=0.306 0.39/ /0.410|0.570|0.020 0.53 N – number of genotyped cows , SNP = single nucleotide polymorphism. Table 3. Location of the five SNPs related to the analyzed traits in the bovine genome and position in the gene/relative to the nearest gene SNP location SNP’s position in Symbol of the Gene location in the in the gene/in Chr. SNP (nucleotides) Trait the chromosome closest gene chromosome relation to the closest gene 8 rs136813430(T/C) lnSCC 108829443 TLR4 108828899–108839913 Exon 29 rs41587003(A/C) lactose content 10172197 DLG2 9997510–12157167 Intron 2 rs110785912(A/T) lnSCC 61134245 LOC1049123 61096534–61148308 Intron CXCR4 61250082–61253877 115837>(5’) 5 rs110455063(C/G) daily milk yield, fat content, 61134245 LOC104972477 66131228–66164165 Intron dry matter content IGF1 66191602–66264083 34194>(5’) 14 rs109421300(T/C) fat to protein ratio, fat content, 1801116 DGAT1 1795425–1804838 Intron dry matter content GPIHBP1 1348153–1350911 450205<(3’) CYP11B1 1550747–1558425 242691<(3’) Chr. = chromosome; allele: A (adenine), C (cytosyne), G (guanine), T (thymine), lnSCC – a natural logarithm of somatic cell count. TLR4 – Toll-like receptor 4; DLG2 – Disks large homolog 2; CXCR4 – C-X-C chemokine receptor type 4; IGF1 – Insulin-like growth factor I; DGAT1 – Diacylglycerol O-acyltransferase 1; GPIHBP1 – Glycosylphosphatidylinositol-anchored high density lipoprotein-binding protein 1; CYP11B1 – Cytochrome P450 11B1. The rs109421300(T/C), alias ARS-BFGL-NGS-4939, the highest value was associated with the A/A variant and associated with fat/protein content ratio, fat and total sol- the lowest with C/C. The heritability of lactose content id content and fatty acid profile (Table 1), is located in the was estimated at 0.16, and the total variability (CV) of intron of the diacylglycerol O-acyltransferase 1 (DGAT1) this trait was 9.23% (Table S2). gene (Table 3); it is also sited 450,205<(3’) nucleotides Finally, SNP rs136813430(T/C) (BTA-82770-no-rs), from the glycosylphosphatidylinositol-anchored high associated with lnSCC (Table S2), was located in the density lipoprotein-binding protein 1 (GPIHBP1) gene exon of the toll-like receptor 4 (TLR4) gene (Table 3). and 242,691>(3’) from the cytochrome P450 11B1 (mi- The milk of cows with the C/T genotype contained more tochondrial) (CYP11B1) gene. The lowest fat content was somatic cells than the other variants of genotypes. found in animals with the T/T variant, and the highest None of the analyzed SNPs were directly related to with the C/C genotype. The highest dry matter content protein content or yield. The indirect relationship ex- was found in the milk of cows with the C/C genotype pressed as fat to protein ratio was mentioned above. and the lowest in milk of cows with T/T. The total vari- ability of the fat to protein ratio was found to be average functional analysis (CV=20.8%), and the h of this trait was 0.14 (Table S2). Seven genes with known functions (TLR4 – Toll-like SNP rs41587003(A/C) (UA-IFASA-7512) was lo- receptor 4; DLG2 – Disks large homolog 2; CXCR4 – cated in the intron of the disks large homolog 2 (DLG2) C-X-C chemokine receptor type 4; IGF1 – Insulin-like gene (Table 3) and was associated with lactose content: growth factor I; DGAT1 – Diacylglycerol O-acyltrans- 424 E. Bagnicka et al. ferase 1; GPIHBP1 – Glycosylphosphatidylinositol-an- sis revealed that IGF-I is involved in immune processes chored high density lipoprotein-binding protein 1; CY- (response to stress, immune system process, signal trans- P11B1 – Cytochrome P450 11B1) were associated with duction) and growth and development, cell differentia- studied SNPs and subjected to functional analysis. tion, cell population proliferation, and anatomical struc- In silico GO annotation analysis revealed that the ture development (Figure 1, Table S3). protein products of two from five genes i.e., TLR4 and In the DGAT1 gene intron is located SNP No. CXCR4, whose polymorphisms were associated with rs109421300(T/C). Protein product of DGAT1 gene lnSCC, were involved in several common biological pro- is involved not only in the lipid metabolism, sulphur cesses connected with immune system such as homeo- metabolism or biosynthesis process, but also in the static process, immune system process, signal transduc- process of homeostasis and the immune system (Figu- tion, or response to stress (Figure 1). However, they were re 1). also involved in cell differentiation, cell morphogenesis, Rs109421300(T/C) SNP is located 242,691<(3’) cell motility, or cell death. Thus, further association stud- from the CYP11B1 gene and 450,205<(3’) from the GPI- ies are needed to check the relationships with other dis- HBP1 gene. GO annotation analysis indicated that the eases and productivity of the cows. The GO number and protein produced by GPIHBP1 gene is involved in lipid definitions were gathered in Table S3 (supplementary metabolic processes, transmembrane transport, cellular files) based on QuickGO database (term/GO:0002376). nitrogen compound metabolic process, or biosynthetic The next SNP – rs41587003(A/C), associated with lac- process, and also in homeostatic process (Figure 1). tose content, is localized in the intron of DLG2 gene which No hits for GPIHBP1 gene were found in KEGG analy- is required for perception of chronic pain through N-me- sis. As in silico GO analysis showed, CYP11B1 is in- thyl-D-aspartate (NMDA) receptor signaling pathway. The volved in lipid metabolic, circulatory system, homeo- DLG2 gene encodes a member of the membrane-associated static, and immune system processes (Figure 1) while guanylate kinase (MAGUK) family. The protein product of according to KEGG analysis i.a. in metabolic pathways DLG2 gene is involved in processes such as biosynthetic, (Table S4). protein transport, cellular protein modification, cytoskeleton- Two of the studied SNPs, rs110785912(A/T) and dependent intra-cellular transport, cell-cell signaling, cell ad- rs110455063(C/G), are located within LOC1049123 and hesion and cellular nitrogen compound metabolic (Figure 1, LOC104972477 loci, respectively, thus no functional Table S4). analysis is possible. These SNPs were associated with In turn, SNP rs110455063(C/G) is located 34,194 nt lnSCC – rs110785912 (A/T), and dry matter and fat con- 5’ from the IGF-I gene. In silico GO annotation analy- tents – rs110455063(C/G). Figure 1. Common biological processes names in which protein products of the studied genes are involved – in silico GO annotation analysis Marker SNPs vs. mastitis and dairy cattle farm profitability 425 The studied genes are involved in many physiological discussion pathways. The protein products of IGF-I and TLR4 genes are involved in 35 and 28 pathways, respectively with animal sample some of them common for both genes such as proteo- Despite the rather small sample size, as the asso- glycans in cancer (bta05205), HIF-1 signaling pathway ciation analysis was conducted using 299 animals with (bta04066), and PI3K-Akt signaling pathway (bta04151). approx. 8,000 phenotypic records, our obtained values IGF-I and CXCR4 are involved in two common path- for the studied SNPs appear generally consistent with ways (endocytosis pathway – bta04144 and pathways in previous studies. Moreover, the standard errors (SE) of cancer – bta05200), while CYP11B1 and DGAT1 in met- solutions for genotypes were low which also proves the abolic pathway (bta01100). Most of pathways in which low bias of the obtained results. With an average of two studied genes are involved are connected with various daughters per sire and only 12% of the dams having more diseases, however, some of them are involved in regula- than one daughter, our sample of 299 can be considered tion of cellular functions such as transcription, transla- representative of the population for a polymorphism tion, proliferation, growth, and others. study because it is not dominated by a small number of large families (Oprządek et al., 2015). economic analysis The heritability coefficient values obtained for milk The value of the studied polymorphisms for selec- yield was within the range of estimates obtained in the tion purposes was determined by economic analysis Polish dairy cows’ population (h =0.20–0.40) (Aerts et (Table S5). An analysis of variance was performed al., 2021; Rzewuska and Strabel, 2013). CV for milk to determine the significance of the differences be- yield was very high (60%) maybe because of different tween the variants of the analyzed SNPs with regard lactation duration. to their obtained economic effects (Table S6). Statis- A high CV for fat content was obtained in the present tically significant differences were found between the study (22.5%) and this is in line with values obtained by allelic variants of SNP rs136813430(T/C) for all ana- previous studies (Krag et al., 2013). The estimated h value lyzed economic categories. No such differences were (0.17) was slightly lower than the coefficients obtained in found for rs110455063(C/G) or rs109421300(T/C) the Polish active population (0.25–0.28) for the first three for any of the analyzed categories. However, for SNP parities using test-day model (Rzewuska and Strabel, 2013). rs110785912(T/A), significant differences were found The heritability coefficient for lactose content was low- between homozygous AA and heterozygous TA for er (0.16) than the values reported previously (0.26–0.34) the direct profitability index. Moreover, and for SNP in the Polish dairy cattle population (Rzewuska and Stra- rs41587003(A/C) significant differences were found bel, 2013), while the low total variability was obtained in between homozygous C/C and T/C regarding the pro- the present study (CV=9.23%). Therefore, this association duction costs of one liter of milk (Table S3). analysis for the lactose content with the tested SNP should Based on the obtained results, SNP rs136813430(T/C) be repeated on a much greater population sample. seems to be the most useful SNP for selection purposes, The heritability coefficient of lnSCC was 0.13 in the as it was found to have a significant influence on each studied group of animals. Again, despite the small sam- of the three economic categories, with statistically sig- ple size and high total variability of lnSCC (CV=34.5%), nificant differences observed between its variants. Cows this value did not differ from those obtained in the Pol- with the T/C genotype demonstrated the highest gross ish populations of dairy cows (0.13–0.16) (Sender et al., margin and direct profitability index compared to other 2013; Rzewuska and Strabel, 2013). variants, and the lowest costs incurred to produce 1 liter For dry matter content, both heritability (0.20) and of milk. variation coefficient (CV=8.37%) did not differ substan- SNP rs110785912(T/A), related to lnSCC in milk, tially from previous studies in Polish population (Yazgan had a significant influence on the direct profitability in- et al., 2010). dex with significant differences observed between cows Similarly, for fat:protein ratio, the h value (0.14) and with the A/A and T/A genotypes. The T/A cows demon- the total variability (CV=20.8%) did not differ substan- strated the most favorable profitability index and the low- tially from previous values, i.e. h between 0.18 and 0.40 est somatic cell content. (Rzewuska and Strabel, 2013) in a study in a big popula- In the case of SNP rs41587003(A/C), significant dif- tion of more than 19,000 Polish HF dairy cows (159,044 ferences were found between the C/C and T/C variants lactation records). regarding the production costs of 1 liter of milk. No stud- Hence, it can be seen that the heritability and total ies have so far examined the functional features of this variation coefficients obtained in the studied group of gene; however, our present findings suggest that this SNP animals did not differ actually from previous studies on is associated with lactose content in milk: variants with a Polish HF population. Moreover, as the genotyped cows higher milk lactose content demonstrated lower costs of were descendants of 153 sires and 263 dams, the probe producing one liter of milk. The lowest costs of produc- from population was not overwhelmed by one or two ing 1 liter of milk were observed for the T/C heterozy- families. As such, our findings SNPs can be regarded as gotes and the highest for the C/C homozygotes. representative, despite the small sample size. 426 E. Bagnicka et al. analysis of gene polymorphisms CXCR4 gene protein product on the processes of the im- Commercially available SNP microarrays are em- mune system taking place in the cells of the udder tissues ployed in GWAS. However, a large majority of SNPs do and leukocytes of the milk alveoli, is extremely impor- not correlate with breeders’ interest-piqued traits. As a re- tant from the point of view of breeding work towards the sult, it is crucial to research all known SNPs and choose selection of cows resistant to inflammation of the udder. for GWAS only those that significantly affect the traits Further association studies are needed to check the that are taken into account in the breeding goal in a par- relationships with other diseases and productivity of the ticular population. Gene containing such SNPs in their cows, especially that both TLR4 and CXCR4 are in- structure can become the candidate for main genes or volved in biological processes connected with both im- SNP is considered as a marker that indicates the quality mune system and cell differentiation and death. All of and value of the genes linked to it. these procedures are necessary for an effective udder The rs136813430(T/C) SNP, associated in our own to function. One of the SNP is localized in the intron of work with lnSCC and identified in the TLR4 gene exon DLG2 gene. As a heterodimer formed with a related fam- (ENSBTAG00000006240) (chromosome 8) is highly ily member, protein product of the DLG2 gene may inter- polymorphic: a total of 36 SNPs have been identified act at postsynaptic sites to form a multimeric scaffold for in 14 breeds of cattle. TLR4 is believed to initiate non- the clustering of receptors, ion channels, and associated specific and adaptive immune responses that defend the signaling proteins. To our knowledge, the function of body against pathogens by inducing the overexpression DLG2 has not been studied in cattle till now. However, in of the pro-inflammatory cytokines IL-1, L-6 and IL-8 humans, it has a wide range of cellular functions includ- involved in non-specific immunity (Wang et al., 2007). ing those connected with nervous system, Parkinson’s Although the identified mutation located in the exon is disease or some kinds of cancers (Zhuang et al., 2019; synonymous, it might be that the polymorphic variant Shao et al., 2019). It interacts with the cytoplasmic tail of is responsible for intensifying the immune response of NMDA receptor subunits and with inward rectifying po- the udder against invasion by increasing production of tassium channels (UniProt Consortium, 2019, Q15700). pro-inflammatory cytokines. This may hasten the annihi - Until now, it has not been connected with lactogenesis. lation of the pathogen without initiating the specific re- Thus, further study is needed to identify its function in sponse and recruitment of leukocytes from the blood into bovine and its connection with lactose content, especially the udder. Therefore, SNP rs136813430(T/C) appears to that DLG2 is involved in many processes including bio- be a promising candidate in the selection of cows for re- synthetic, protein transport, or cellular protein modifica - sistance to mastitis, as demonstrated by the significant tion. role played by the TLR4 gene protein product during in- IGF-I gene (ENSBTAG00000011082) (chromosome flammatory processes of the udder, and our own findings 5 in bovine) protein product is also called somatomedin regarding the relationship between the SNP and lnSCC. C (Meuwissen et al., 2002). IGF-I mediates many biolog- So far, this SNP has not been indicated as a marker of ical processes, increasing glucose absorption, regulating mastitis or milk composition. cellular differentiation, or proliferation, inhibiting apop- Located close to the CXCR4 (ENSBTAG 0000001060) tosis, or increasing lipid synthesis (De la Rosa Reyna et gene (chromosome 2), the rs110785912(A/T) SNP is al., 2010). IGF-I is released from the liver in response also associated with lnSCC. The gene is a member of to growth hormone (GH) that controls growth and lac- the chemokine receptor family (Busillo and Benovic et tation, thus IGF-I is thought that controlling lactation al., 2007), and is involved in many developmental pro- through production of milk ingredients (Lucy, 2008). It cesses, as well as in many disease processes caused by is also produced locally in a tissue-specific manner. IGF- viruses, such as bovine non-cytopathic viral diarrhoea I can be used to monitor udder health by measuring its virus (BVDV) (Weiner et al., 2012). CXCR4 expres- concentration in milk (Liebe and Shams, 1998). The au- sion was found to be elevated in the secretory epithe- thors’ own research showed a relationship between SNP lial tissue in cow udder quarters following infection by rs110455063(C/G) located 34,194 nt 5’ from the IGF-I coagulase-positive and coagulase-negative staphylo- gene with the dry matter, and fat contents, and daily milk cocci compared to pathogen-free tissues (Kościuszuk yield. So far, in the association analyses, this SNP has not et al., 2017). The CXCR4 receptor is believed to play been associated with any trait of milk, so it is not possible a key role in lymphocyte transport and directing them to compare our results with the results of other authors. to the lymph nodes (Busillo and Benovic, 2007). So far, It is proposed to include the SNP found in IGF-I gene in the rs110785912(A/T) SNP has not been studied by other MAS or genomic selection to improve cow productivity, authors in association analyses. Based on our own re- but it is necessary to conduct further studies on greater sults and literature data on the significant involvement populations of dairy cattle. of CXCR4 in the activation of the immune system, this DGAT1 gene (ENSBTAG00000026356) (chromo- SNP has been proposed as a marker of udder health to some 14) codes for a microsomal enzyme, using diacylg- be used in Marked Assisted Selection (MAS). Confirma- lycerol and fatty acetyl coenzyme A as substrates to cata- tion of the close relationship between the tested SNP and lyze the final stage of triacylglycerol synthesis (Cases et SCC in milk, probably through the direct influence of the al., 1998). It influences fat metabolism and its yield and Marker SNPs vs. mastitis and dairy cattle farm profitability 427 percentage in milk. Many polymorphic loci have been the different population size or structure. It is recom- found in this gene and most of them are located in the in- mended that any population used in such a study should trons, promoter and untranslated regions (UTR) of exons. be subjected to GWAS testing. SNP analyzed in our study was located within this gene The analyzed SNPs were located within genes or clos- at position 1801,116 and was related to the fat content, er than around 500,000 nucleotides (nt) from the near- dry matter content and fat to protein ratio. Meredith et est gene. As a long distance between the marker and the al. (2013) showed a relationship between the studied SNP gene reduces the probability of inheriting the analyzed and the milk fat yield. Jiang et al. (2019) found a strong SNP with the gene, such distant loci cannot be reli- antagonistic pleiotropy between fat yield and milk and able markers of the traits coded by an individual gene. protein yield for this SNP; the C allele was responsible for Simulations (Kruglyak, 1999) and empirical analyses this extremely antagonistic pleiotropy between positive fat based on human data (Dunning et al., 2000) suggest yield and negative milk and protein yield, while the T al- that linkage disequilibrium (LD) extends only a few lele had antagonistic pleiotropy between negative fat yield kilobases (kb) around common SNPs; however, other and positive milk and protein yield. Chen et al. (2015) studies indicate that this distance can be greater than showed its association with SCC during inflammation 100 kb (Abecasisi et al., 2001) or even beyond 1 Mb caused by E. coli and S. uberis. Many studies have found (Taillon-Miller et al., 2000). Nevertheless, it is sug- that the region that includes the DGAT1 gene has a large gested that due to genetic drift, admixture, selection impact on milk fat content and on several other production and smaller effective population sizes, which reduce traits including milk yield, percentage of fat, and percent- allelic heterogeneity, livestock demonstrate greater age of protein (Ashwell et al., 2004; Maxa et al., 2012). LD than humans, and LD can extend over several hun- The fat:protein ratio was 1.15 for heterozygous cows: it dred base pairs (McRae et al., 2002). should be highlighted that it is the best ratio between these It is possible that the analyzed SNPs which exist traits, because it determines the best texture and taste for outside any gene sequences, may be found in enhancer cheese curd. DGAT1 can be considered a candidate gene (expression enhancer) or suppressor regions – they can for the fat content and fatty acid profile of milk. permanently turn off expression through an epigenetic si- The GPIHBP1 (ENSBTAG00000049125) (chromo- lencing mechanism; as such, they have a strong influence some 14) encodes a protein belonging to the lymphocyte on gene transcription (activation) and expression. More- antigen 6 (Ly6) family. GPIHBP1 plays a key role in the over, the promoter region, a section of genomic DNA transport and localization of lipoprotein lipase (LPL) upstream of the transcription start site (TSS) of the gene synthesized by myocytes and adipocytes and creates commonly referred to as the +1 position, contains regu- a platform for lipolysis in endothelial cells (Yang et al., latory information for transcription initiation. Enhancer- 2017). In our research rs109421300(T/C) SNP was relat- promoter interactions precisely control the spatiotempo- ed to the dry matter and fat contents and the fat to protein ral timing of the activation of particular genes (Danino et ratio. Meredith et al. (2013) showed that this SNP was re- al., 2015). The enhancers are cis-regulatory elements first lated to the yield of milk fat. Fang and Pausch (2019) also identified in the SV40 virus genome; these can influence proposed the inclusion of this GPIHBP1 gene for use in expression from a distance of even 1 mega base (1 Mb) selection to improve milk yield. As this SNP is located in and may be located within the target gene, or upstream or close distance to the GPIHBP1 gene (450,205 nt towards downstream from it. They show high specificity regard- 3’), so it is likely to be inherited along with it. ing cell type and response to stimuli. Complex regulatory SNP rs109421300(T/C) is closely located to CY- genes often have multiple enhancers, with each being re- P11B1 gene which encodes a member of the cytochrome sponsible for expression in a given cell type or situation. P450 superfamily. Proteins catalyze many reactions re- In addition, a single enhancer can act on many genes at lated to drug metabolism and the synthesis of choles- once. The genes whose expression is necessary for the terol or steroids and other lipids. This protein localizes function of the cells in each situation may not have any to the mitochondrial inner membrane and is involved in enhancers or silencers. Enhancers and silencers often the conversion of progesterone to cortisol in the adrenal cooperate to maintain the balance of tissue-specific pro- cortex. CYP11B1 affects cortisol production, androgen moter activity. Most SNPs have been found in noncoding function, and ultimately the proliferation of mammary regions containing enhancers (Xia and Wei, 2019). gland cells (Brettes and Mathelin, 2008). As the probability of inheriting a SNP with a gene is Two SNPs, rs110785912(A/T) and rs110455063(C/G), reduced when a long distance exists between marker and located within LOC1049123 and LOC104972477 loci functional gene, loci located at a large distance cannot be in the NCBI database are marked as genes of unknown reliable markers of the features encoded by a given gene, function. These SNPs were associated with lnSC Cor dry particularly when assessing the predicted breeding value matter, and fat contents. However, it was impossible to of offspring. However, as genomic selection allows indi- compare our results with the results of other authors, be- viduals with high breeding value to be selected based on cause there is no available literature information. SNP markers, all information on SNPs and their associa- Our results for some of the analyzed SNPs differ from tion with production traits is useful in the breeding work. previous studies. These differences are probably due to Genomic information allows genetically ideal animals to 428 E. Bagnicka et al. be identified at a very young age, with greater accuracy superfamily (MAGUK). This is a huge protein complex, than estimates based on the average genetic value of the with more than 1,000 members, including scaffold or parents. Further studies should examine the functions of cytoskeletal proteins, receptors and signaling enzymes. the protein products of studied genes; this would provide These proteins are a part of the postsynaptic protein scaf- a better understanding of their participation in metabolic fold of excitatory synapses, and contain various domains pathways and cellular processes associated with the im- (e.g., PDZ, GK, SH3) that enable them to bind to many mune system of cows and their relationship with produc- of the proteins present in synapses (Zhu et al., 2016). tion traits. Although no studies have examined the functional fea- tures of protein encoded by this gene, our findings indi- economic analysis cate that the SNP lying in this gene appears to be related SNP rs136813430(T/C) was found to have prom- to lactose content in milk. Lactose level is also an indi- ise for assessing potential milk production profitability. cator of udder health: its decrease often indicates in- Cows with the T/C genotype were characterized by the flammation of the udder, resulting in loss of production highest gross margin and direct profitability index, as and additional costs related to treating cows (Bruckmaier well as the lowest costs incurred to produce one liter of et al., 2004). According to the conducted research, vari- milk. This SNP probably influences the number of so- ants with a higher milk lactose content were character- matic cells (SCC) possibly due to the role played by the ized by a lower level of production costs per liter of milk. protein encoded by the TLR4 gene, in which the SNP is In this analysis, the lowest production costs for one liter of located. milk were recorded for T/C heterozygotes and the highest Many studies have confirmed the importance of Toll- for C/C homozygotes. Therefore, there is a certain contra- like receptors (TLRs), including TLR4, in the first line of diction between the selection and economic goals. defense against invading pathogens, where they are be- To summarize, the present study examines the influ- lieved to initiate the innate immune response (Kawai and ence of selected SNP polymorphisms occurring in genes Akira, 2005; De Schepper et al., 2008). Concerning the related to udder health and milk production on the profit- lower SCC content, the T/C variant may bestow higher ability of milk production. The analysis was compared resistance to inflammation of the udder. Therefore, it is with data concerning production value and direct costs an interesting candidate for the purposes of breeders. which was obtained from commercial farms. An analy- Information about the variants of SNP sis of variance revealed significant differences between rs136813430(T/C) may therefore prove useful when allelic variants in the case of three tested SNPs. The selecting cows with a specific genotype in programs in- heterozygous (T/C) variant of SNP rs136813430(T/C) tended to regulate the content of fatty acids in milk; prop- was associated with a low content of lnSCC; it was also er selection will guarantee not only an appropriate fatty characterized by the highest gross margin, the highest di- acid content, but also improve economic profitability. rect profitability index and the lowest costs incurred to In the case of SNP rs110785912(T/A) associated with produce one liter of milk (P = 0.01). This therefore ap- lnSCC in milk, a significant correlation was found with pears to be the most promising of the tested SNPs for se- the direct profitability index. The SNP lies in proximity lection purposes. The T/A variant of rs110785912(T/A) to the CXCR4 gene, the product of which is involved in demonstrated low lnSCC content in milk and the high- the immune system of cattle (Weiner et al., 2012). It is est direct profitability index, while the C/C variant of therefore possible that the area within the CXCR4 gene, rs41587003(A/C) demonstrated lowest lactose level and and thus also in the rs110785912(T/A) locus, may be the highest costs of producing one liter of milk. The ob- involved in the functioning of the immune system of tained results indicate that there is an economic justifica - cows; however, this requires additional research. The tion for including SNP variants located within, or close most favorable genotype in terms of the profitability in- to, genes involved in the immune system and milk pro- dex turned out to be the heterozygous T/A variant, which duction in cattle selection. was associated with the lowest content of somatic cells. As SCC is an indirect indicator of udder health, the low conclusions cell content may mean that cows with this genotype are While our cow sample size may seem somewhat small less susceptible to mastitis than those with other gene for this kind of study the bias can mainly regard the vari- variants. This variant therefore offers promise in selec- ance component magnitudes of the phenotypic outcome tion processes intended to increase the resistance to in- and SNP alleles’ frequencies, as the entire population flammation while ensuring high milk production profit- variation could have been narrowed in the sample. We ability. have, however, discussed this issue with the conclusion On the other hand, regarding SNP rs41587003(A/C), that our sample is suitable for the undertaken research. significant relationships were found between the C/C and On the other hand, the SNP-phenotypic output associa- T/C variants with regard to the cost of producing one liter tion analysis is much less dependent on the sample size of milk. The studied SNP is in the intron of the DGL2 (if at all, given our sample size). gene called also Postsynaptic Density Protein PSD-93. Hence, the above let us recommend the It encodes a protein belonging to the guanylate kinase rs136813430(T/C) SNP, located within the TLR4 gene, Marker SNPs vs. mastitis and dairy cattle farm profitability 429 (2015). Second-generation PLINK: rising to the challenge of larg- as a candidate gene for explaining the variance of SCC er and richer datasets. Gigascience, 4: 7. in milk, as indicative of susceptibility to mastitis, but Chen X., Cheng Z., Zhang S., Werling D., Wathes D.C. (2015). Com- also for explaining the variance of the gross margin and bining genome wide association studies and differential gene ex- the direct dairy profitability index. The latter is, to our pression data analyses identifies candidate genes affecting mastitis caused by two different pathogens in the dairy cow. Open J. Anim. knowledge, the first effort to assess directly a correlation Sci., 5: 358–393. between the DNA polymorphism and economic output of Danino Y.M., Even D., Ideses D., Juven-Gershon T. (2015). The core a dairy enterprise. promoter: At the heart of gene expression. BBA-Gene Regul. Mech., 1849: 1116–1131. Davies G., Genini S., Bishop S.C., Giuffra E. (2009). An assessment of opportunities to dissect host genetic variation in resistance to acknowledgements infectious diseases in livestock. Animal, 3: 415–436. The authors thank prof. Marek Łukaszewicz for in- De la Rosa Reyna X.F., Montoya H.M., Castrellón V.V., Rincón valuable help in improving the quality of this work. The A.M.S., Bracamonte M.P., Vera W.A. (2010). Polymorphisms in authors gratefully acknowledge the Polish Federation of the IGF1 gene and their effect on growth traits in Mexican beef cattle. Gen. Mol. Res., 9: 875–883. Cattle Breeders and Dairy Farmers for phenotypic data De Schepper S., De Ketelaere A., Bannerman D.D., Paape M.J., Peel- on the daily milk yield and the component contents and man L., Burvenich C. (2008). The toll-like receptor-4 (TLR-4) pedigree file studied dairy cows. pathway and its possible role in the pathogenesis of Escherichia coli mastitis in dairy cattle. Vet. Res., 39: 5. Dunning A.M., Durocher F., Healey C.S., Teare M.D., McBride S.E., Carlomagno F., Xu C-F., Dawson E., Rhodes S., Ueda S., Lai E., references Luben R.N.., Van Rensburg E.J., Mannermaa A., Kataja V., Renn- art G., Dunham I., Purvis I., Easton D., Ponder B.A.J. (2000). The extent of linkage disequilibrium in four populations with distinct Abecasis G.R., Noguchi E., Heinzmann A., Traherne J.A., Bhattacha- ryya S., Leaves N.I., Anderson G.G., Zhang Y., Lench N.L., Carey demographic histories. Am. J. Hum. Genet., 67: 1544–1554. A., Cardon L.R., Moffatt M.F. Cookson W.O.C. (2001). Extent Fang Z.H., Pausch H. (2019). Multi-trait meta-analyses reveal 25 quantitative trait loci for economically important traits in Brown and distribution of linkage disequilibrium in three genomic re- gions. Am. J. Hum. Genet., 68: 191–197. Swiss cattle. BMC Genomics, 20: 695. Aerts J., Piwczyński D., Ghiasi H., Sitkowska B., Kolenda M., Önder Hagnestam-Nielsen C., Østergaard S. (2009). Economic impact of H. (2021). Genetic parameters estimation of milking traits in Pol- clinical mastitis in a dairy herd assessed by stochastic simula- tion using different methods to model yield losses. Animal, 3: ish Holstein-Friesians based on automatic milking system data. Animals, 11: 1943. 315–328. Andrei S., Pintea A., Bunea A., Groza I., Bogdan L., Ciupe S., Matei Jiang J., Ma L., Prakapenka D., Van Raden P.M., Cole J.B., Da Y. (2019). A large-scale genome-wide association study in U.S. Hol- S., Crainic D. (2009). Non-enzymatic antioxidants concentration and lipids peroxidation level in milk from cows with subclinical stein cattle. Front. Genet., 10: 412. mastitis. Bull. Univ. Agric. Sci. Vet. Med. Cluj. Napoca Vet. Med., Kanehisa M., Sato Y., Kawashima M., Furumichi M., Tanabe M. 66. (2016). KEGG as a reference resource for gene and protein an- notation. Nucleic Acids Res., 44: D457–D462. Ashburner M., Ball C.A., Blake J.A., Botstein D., Butler H., Cherry J.M., Davis A.P., Dolinski K., Dwight S.S., Eppig J.T., Harris Kawai T., Akira S. (2005). Pathogen recognition with toll-like recep- M.A., Hill D.P., Issel-Tarver L., Kasarskis A., Lewis S., Matese tors. Curr. Opin. Immunol., 17: 338–344. J.C., Richardson J.E., Ringwald M., Rubin G.N., Sherlock G. Kosciuczuk E.M., Lisowski P., Jarczak J., Majewska A., Rzewuska M., Zwierzchowski L., Bagnicka E. (2017). Transcriptome profil - (2000). Gene ontology: tool for the unification of biology. Nat. Genet., 25: 25–29. ing of Staphylococci-infected cow mammary gland parenchyma. Ashwell M.S., Heyen D.W., Sonstegard T.S., Van Tassel C.P., Da P.M., BMC Vet. Res., 13: 1–12. Krag K., Poulsen N.A., Larsen M.K., Larsen L.B., Jans L.L., Buiten- Ron M., Weller J.I. (2004). Detection of quantitative trait loci af- fecting milk production, health, and reproductive traits in Holstein huis B. (2013). Genetic parameters for milk fatty acids in Danish cattle. J. Dairy Sci., 87: 468–475. Holstein cattle based on SNP markers using a Bayesian approach. Bagnicka E., Łukaszewicz M., Ådnøy T. (2016). Genetic parameters BMC Genet., 14: 1471–2156. Kruglyak L. (1999). Prospects for whole-genome linkage disequilibrium of somatic cell score and lactose content in goat’s milk. J. Anim. Feed Sci., 25: 210–215. mapping of common disease genes. Nature Genet., 22: 139–144. Brettes J.P., Mathelin C. (2008). Effect dual DeVuyst s androgens sur Liebe A., Shams D. (1998). Growth factors in milk: interrelationships with somatic cell count. J. Dairy Res., 65: 93–100. la glande mammaire. Bull. Cancer, 95: 495–502. Bruckmaier R.M., Ontsouka C.E., Blum J.W. (2004). Fractionized Lü A., Hu X., Chen H., Dong Y., Zhang Y., Wang X. (2011). Novel milk composition in dairy cows with subclinical mastitis. Vet. SNPs of the bovine PRLR gene associated with milk production Med. Czech, 49: 283–290. traits. Biochem. Genet., 49: 177–189. Lucy M. (2008). Functional differences in the growth hormone and Brzóska F., Kowalski Z.M., Osięgłowski S., Strzetelski J. (2014). IZ PIB-INRA Standard of ruminants’ feeding: nutrient value of insulin-like growth factor axis in cattle and pigs: Implications for French and domestic fodders for ruminants (in Polish), Strzetelski post-partum nutrition and reproduction. Reprod. Domest. Anim., 43: 31–39. J. (ed.). Foundation IZ PIB Patronus Animalium, Kraków, Poland. Busillo J.M., Benovic J.L. (2007). Regulation of CXCR4 signaling. Madsen P., Jensen J. (2013). A user’s guide to DMU – a package for Biochim. Biophys. Acta, 1768: 952–963. analysing multivariate mixed models. Version 6; release 5.2. Carlen E., Strandberg E., Roth A. (2005). Genetic parameters for clini- Marogna G., Rolesu S., Lollai S., Tola S., Leori G. (2010). Clinical findings in sheep farms affected by recurrent bacterial mastitis. cal mastitis, somatic cell score, and production in the first three lactations of Swedish Holstein cows. J. Dairy Sci., 87: 3062–3070. Small Rumin. Res., 88: 119–125. Cases S., Smith S.J., Zheng Y.W., Myyers H.M., Lear S.R., Sande E., Maxa J., Neuditschko M., Russ I., Förster M., Medugorac I. (2012). Novak S., Collins C., Welch C.B., Lusis A.J., Erickson S.K. Fa- Genome-wide association mapping of milk production traits in Braunvieh cattle. J. Dairy Sci., 92: 5357–5364. rese R.V. (1998). Identification of a gene encoding an acyl CoA: diacylglycerol acyltransferase, a key enzyme in triacylglycerol McRae A.F., McEwan J.C., Dodds K.G., Wilson T., Crawford A.M., synthesis. PNAS USA, 95: 13018–13023. Slate J. (2002). Linkage disequilibrium in domestic sheep. Genet- ics, 160: 1113–1122. Chang C.C., Chow C.C., Tellier L.C., Vattikuti S., Purcell S.M., Lee J.J. 430 E. Bagnicka et al. Meredith B.K., Berry D.P., Kearney F., Finlay E.K., Fahey A.G., Brad- Sordillo L.M. (2005). Factors affecting mammary gland immunity and ley D.G., Lynn D.J. (2013). A genome-wide association study for mastitis susceptibility. Livest. Prod. Sci., 98: 89–99. somatic cell score using the Illumina high-density bovine bead- Taillon-Miller P., Bauer-Sardiña I., Saccone N.L., Putzel J., Laitinen chip identifies several novel QTL potentially related to mastitis T., Cao A., Kere J., Pilia G., Rice J.P., Kwok P-Y. (2000). Juxta- susceptibility. Front. Genet., 4: 229. posed regions of extensive and minimal linkage disequilibrium in Methodology of calculating the gross margin for agricultural produc- human Xq25 and Xq28. Nat. Genet., 25: 324–328. tion activity (in accordance with the standards of the European UniProt Consortium. (2019). UniProt: a worldwide hub of protein Union (in Polish). Institute of Agricultural and Food Economics- knowledge. Nucleic acids research, 47: D506–D515. https://www. National Research Institute (IERiGŻ), Warszawa, 1999. uniprot.org/uniprot/Q15700, access, December 15, 2021. Meuwissen T.H.E., Karlsen A., Lien S., Olsaker I., Goddard M.E. Wang Y., Chen T., Han C., He D., Liu H., An H., Cai Z., Cao X. (2007). (2002). Fine mapping of a quantitative trait locus for twinning Lysosome-associated small Rab GTPase Rab7b negatively regu- rate using combined linkage and linkage disequilibrium mapping. lates TLR4 signaling in macrophages by promoting lysosomal Genetics, 161: 373–379. degradation of TLR4. Blood, 110: 962–971. NCBI-1, National Center for Biotechnology Information, https://www. Weiner C.M., Smirnova N., Webb B.T., Van Campen H., Hansen T.R. ncbi.nlm.nih.gov, access May 2019. (2012). Interferon stimulated genes, CXCR4 and immune cell re- NCBI-2, National Center for Biotechnology Information, https://www. sponses in peripheral blood mononuclear cells infected with bo- ncbi.nlm.nih.gov/genome/?term=LOC104972477, access Decem- vine viral diarrhea virus. Res. Vet. Sci., 93: 1081–1088. ber 14, 2021. Xia J.H., Wei G.H. (2019). Enhancer dysfunction in 3D genome and Oprządek J., Sender G., Pawlik A., Łukaszewicz M. (2015). Locus disease. Cells, 8: 1281. BoLA-DRB3 is just an ordinary site of the polygene when ex- Yang J., Liu X., Wang D., Ning C., Wang H., Zhang Q., Jiang L. plaining genetic variance of somatic cell count and milk yield. J. (2017). Functional validation of GPIHBP1 and identification of Dairy Res., 82: 449–452. a functional mutation in GPIHBP1 for milk fat traits in dairy cat- Ptak E., Brzozowski P., Jagusiak W., Zdziarski K. (2007). Genetic pa- tle. Sci. Rep., 7: 8546. rameters for somatic cell score for Polish Black-and-White cattle Yazgan K., Makulska J., Węglarz A., Ptak E., Gierdziewicz M. (2010). estimated with a random regression model. J. Anim. Feed. Sci., Genetic relationship between milk dry matter and other milk traits 16: 357–369. in extended lactations of Polish Holstein cows. Czech J. Anim. Rzewuska K., Strabel T. (2013). Genetic parameters for milk urea Sci., 55: 91–104. concentration and milk traits in Polish Holstein-Friesian cows. J. Yuan Z., Li J., Li J., Gao X., Xu S. (2013). SNPs identification and its Appl. Genet., 54: 473–482. correlation analysis with milk somatic cell score in bovine MBL1 Sender G., Korwin-Kossakowska A., Pawlik A., Hammed K., gene. Mol. Biol. Rep., 40: 7–12. Oprządek J. (2013). Genetic basis of mastitis resistance in dairy Zhu J., Shang Y., Zhang M. (2016). Mechanistic basis of MAGUK- cattle – a review. Ann. Anim. Sci., 13: 663–673. organized complexes in synaptic development and signalling. Nat. Shao Y.W., Wood G.A., Lu J., Tang Q.L., Liu J., Molyneux S., Chen Y., Rev. Neurosci., 17: 209–223. Fang H., Adissu H., McKee T., Waterhouse P., Khokha R. (2019). Zhuang R.J., Bai X.X., Liu W. (2019). MicroRNA-23a depletion pro- Cross-species genomics identifies DLG2 as a tumor suppressor in motes apoptosis of ovarian cancer stem cell and inhibits cell mi- osteosarcoma. Oncogene, 38: 291–298. gration by targeting DLG2. Cancer Biol. Ther., 20: 897–911. Skarżyńska A. (2012). The impact of the milk yield of cows on the profitability of milk production (in Polish). Zagadnienia Eko - Received: 10 IV 2022 nomiki Rolnej (Probl. Agricult. Econom.), 330: 90–111. Accepted: 29 XI 2022

Journal

Annals of Animal Science – de Gruyter

Published: Apr 1, 2023

Keywords: SNP marker; intron; exon; association; mastitis indicator; dairy economics; milk production profitability

There are no references for this article.