Genes with major effects are more important than genes with minor effects in the variation observed among individuals, and such genes are considered candidate genes. One of these genes, the leptin gene, has a pleiotropic effect on the regulation of feed consumption, energy metabolism, body weight, growth, carcass composition, milk and fertility traits, as well as on the immune system in cattle (Taniguchi et al., Reference Taniguchi, Itoh, Yamada and Sasaki2002), and is considered a candidate gene. Several studies have been conducted to determine the association of the leptin gene with yield-related traits in cattle (Liefers et al., Reference Liefers, Te Pas, Veerkamp and Van Der Lende2002; Aytekin, Reference Aytekin2011). The associations between Sau3AI polymorphism of leptin gene and milk yield (Liefers et al., Reference Liefers, Te Pas, Veerkamp and Van Der Lende2002; Ghazanfari et al., Reference Ghazanfari, Nassiry and Moussavi2006; Metin Kiyici et al., Reference Metin Kiyici, Arslan, Akyuz, Kaliber, Aksel and Çinar2019), milk components (Aytekin, Reference Aytekin2011; Maletić et al., Reference Maletić, Paprikić, Lazarević, Hodžić, Davidović, Stanišić and Stanimirović2019; Metin Kiyici et al., Reference Metin Kiyici, Akyüz, Kaliber, Arslan, Aksel and Çinar2020) and health status (Ferchichi et al., Reference Ferchichi, Jemmali, Amiri, Ben Gara and Rekik2018) in various breeds were previously investigated. The sum of the single nucleotide polymorphism variants on which the candidate genes are based (additive genetic effects) results in the breeding value of the animals. However, the high cost of using current molecular analyzes in animal husbandry limits their use or the number of genes examined. We hypothesized that there will be a difference in the population regarding the Sau3AI polymorphism of the leptin gene and this difference might impact some milk production traits.
To the best of our knowledge from the literature, there is no study on the relationship between Sau3AI polymorphism and milking time (MT), 100 or 200 d of lactation milk yield (LMY100, LMY200) and dry period (DP) in Holstein Friesian dairy cattle. The aim of this study was to investigate the Sau3AI polymorphism of the leptin gene and the effect of this polymorphism and several macro-environmental factors on some milk performance traits in Holstein Friesian dairy cattle.
Materials and methods
Animals, sampling and yield recording
Whole blood samples were taken from 212 Holstein Friesian dairy cattle reared on a private dairy farm in Kırşehir province of Türkiye. No ethical approval was required as the samples were taken by a practicing veterinarian for routine management purposes (brucellosis testing) and subsequently made available for the research. Milk production traits obtained from the Afifarm herd management program (Ver 5.4.2) were lactation milk yield as well as average daily, peak, 100, 200 and 305d milk yield (LMY, ADMY, PMY, LMY100, LMY200 and LMY305; kg), activity (A; step/day), conductivity (C; mS/cm), milking time (MT; min), days in milk (DIM; day), days to peak milk yield (DPMY; day), and length of dry period (DP; day). Further details as well as details of blood collection, DNA extraction and PCR-RFLP method are specified in the online Supplementary File materials and methods.
Statistical analysis
The PopGene Version 1.32 software (Yeh et al., Reference Yeh, Yang, Boyle, Ye and Mao1997) was used for statistical analysis of allele and genotype frequencies and heterozygosity (Nei, Reference Nei1973) of the gene region. The chi-square (χ2) test was performed to determine whether the population was in Hardy–Weinberg equilibrium (Düzgüneş et al., Reference Düzgüneş, Kesici and Gürbüz1983). Since the number of animals in the first lactation is high, the analysis of the data is not suitable for the repeated measurement of mixed model, therefore we used the General Linear Model (GLM) to evaluate the effects of genotype and environmental factors. Details of the model are given in the online Supplementary File. Tukey's multiple comparison test was performed to assess the differences between means that were significant as a consequence of the analysis of variance. Minitab v16.1.1 software package (Minitab, 2010) was used for statistical analysis. The multiple trait derivative free restricted maximum likelihood (MTDFREML) package was used to estimate variance components and genetic parameters using the best linear unbiased prediction (BLUP) method (Boldman et al., Reference Boldman, Kriese, Van Vleck, Van Tassell and Kachman1995). Genotype was not included in the statistical model used to estimate variance components and heritability. Also, factors such as year, parity, and covariance vary in the models according to the different traits. In the estimation of variance components, heritability, and EBVs, the statistically significant factors as a result of the analysis of variance were included in the model (detailed in online Supplementary File and Table S1). The variance components and heritability were estimated using the Restricted Maximum Likelihood (REML) technique in the MTDFREML program. The BLUP technique was used to determine the EBVs in the MTDFREML package program and the breeding value of each animal was calculated as described in the online Supplementary File. A one-way analysis of variance (ANOVA) was performed to examine the variation between genotypes in terms of EBVs. Tukey's multiple comparison test was used to compare the means of genotypes whose effect was found to be significant by analysis of variance.
Results
Sau3AI polymorphism on intron 2 region of leptin gene in Holstein Friesian dairy cattle
The PCR and restriction products of a 422 bp region of the leptin gene region are shown in the online Supplementary Fig. S1. The frequencies of A and B alleles and AA, AB, and BB genotypes for the leptin gene polymorphism were determined to be 0.8821 and 0.1179, and 0.764, 0.236 and 0.000, respectively. χ2 analysis showed that the leptin gene polymorphism was at the Hardy–Weinberg equilibrium (P > 0.05). The results concerning allele and genotype frequencies and heterozygosity values are given in the online Supplementary Table S2.
Association analysis, variance components and breeding values
The effects of the Sau3AI polymorphism of the leptin gene and some environmental factors on milk production traits as determined by the GLM analysis are shown in Table 1. In addition, the least squares mean and regression coefficients of the environmental factors affecting the milk production traits are detailed in the online Supplementary Tables S3–S6. There were no significant effects of the Sau3AI polymorphism of the leptin gene on any of the measured traits (Table 1).
Table 1. Associations between Leptin/Sau3AI gene and milk yield traits in Holstein Friesian dairy cattle
NS, non significance; N, number of animals; 
$\bar{X}$, least mean squares; 
$S_{\bar{X}}$, standard error; R 2, adjusted determination coefficient; LMY, lactation milk yield; A, activity; C, conductivity; MT, milking time; LMY305, 305 d of lactation milk yield; LMY200, 200 d of lactation milk yield; LMY100, 100 d of lactation milk yield; DIM, days in milk; PMY, peak milk yield; DPMY, days to peak milk yield; ADMY, average daily milk yield; DP, dry period; FCA, first calving age; SP, service period; CPMY, conductivity of peak milk yield; MTPMY, milking time of peak milk yield; APMY, activity of peak milk yield; CDPMY, conductivity of days to peak milk yield; MTDPMY, milking time of days to peak milk yield; ADPMY, activity of days to peak milk yield; CI, calving interval.
*Statistical significance level (*P < 0.05, **P < 0.01).
The modeled and estimated variance components and heritability are shown in the online Supplementary Table S7. The −2 log L values used to evaluate whether the heritability is proper for selection program showed variation between −1300.13 (DIM) and 1483.56 (DPMY) among all traits. Regarding heritability, it was determined that the direct heritability (h a2) of different traits varied between 0.03 ± 0.283 (for DP) and 0.50 ± 0.183 (for C), whereas the same figures for maternal heritability (h m2) were 0.01 ± 0.001 (for LMY200) and 0.25 ± 0.031 (for ADMY). For all milk yield traits except LMY305 and LMY200, the correlation between additive genetic effect and maternal effect was found to be high and positive. In production characteristics, it was shown that PMY (63.22%), DP (77.38%) and DPMY (79.94%) were most affected by environmental conditions (online Supplementary Table S7). The estimated breeding values (EBVs) of milk yield features in accordance with the genotype are shown in Table 2. The accuracy of the estimated breeding values (EBV) varied between 56% and 84% when evaluated in terms of genotype. In terms of EBVs, no significant relation between genotypes and milk production traits existed except for MT, for which EBVs of cows with AA and AB genotypes were estimated as 0.0526 ± 0.3688 and −0.0682 ± 0.2593, respectively, a significant difference (P < 0.05; Table 2). Based on this difference obtained from standardized data, it can be said that the B allele shortens the milking time of the cows in the present study.
Table 2. Estimated breeding values of milk yield traits according to genotypes
N, number of animals; EBVs, estimated breeding values; 
$\bar{X}$, means of EBVs; r IH, accuracy; 
$S_{\bar{x}}$, standard error; S x, standard deviation.
a,bP < 0.05.
Discussion
In marker-assisted selection research it is critical to investigate the association between candidate genes and yield characteristics in order to improve the effectiveness of selection. There have been numerous studies of leptin gene polymorphism in dairy cattle, and there is no real consensus regarding outcomes. Trakovická et al. (Reference Trakovická, Moravčíková and Kasarda2013) reported that the milk yield of Slovak Spotted and Slovak Pinzgau cows with the leptin AA genotype was higher than cattle with the AB and BB genotypes, whereas Al-Janabi et al. (Reference Al-Janabi, Al-Rawi and Al-Anbari2018) determined that the milk yield of Holstein Friesian cattle with the AB genotype was higher than the AA genotype and Maletic et al. (Reference Maletić, Paprikić, Lazarević, Hodžić, Davidović, Stanišić and Stanimirović2019) stated that the association between milk yield and genotypes in Busha and Busha × Podolian crossbreed cattle was insignificant. Our results differ from the studies of Trakovická et al. (Reference Trakovická, Moravčíková and Kasarda2013) and Al-Janabi et al. (Reference Al-Janabi, Al-Rawi and Al-Anbari2018) but are similar to the results of Maletić et al. (Reference Maletić, Paprikić, Lazarević, Hodžić, Davidović, Stanišić and Stanimirović2019). According to Ferchichi et al. (Reference Ferchichi, Jemmali, Amiri, Ben Gara and Rekik2018), Holstein Friesian cattle with the leptin AB genotype had fewer lameness problems than cattle with the AA and BB genotypes. Unlike the current study, Metin Kiyici et al. (Reference Metin Kiyici, Akyüz, Kaliber, Arslan, Aksel and Çinar2020) noted that conductivity in Holstein Friesian cattle with the BB genotype was significantly higher than in animals with the AA and AB genotypes and Moussavi et al. (Reference Moussavi, Ahouei, Nassiry and Javadmanesh2006) reported that Holstein Friesian cattle with the AB genotype had a higher LMY305. Despite these various differences, Aytekin (Reference Aytekin2011) claimed that the B allele can be used in selection to increase milk yield due to a statistically non-significant trend between the Sau3AI polymorphism and the LMY305 in Brown Swiss cattle. In Holstein Friesian cows raised in Türkiye, Metin Kiyici et al. (Reference Metin Kiyici, Arslan, Akyuz, Kaliber, Aksel and Çinar2019) observed a significantly lower LMY305 in leptin BB genotype than either AA or AB (which were similar). Ghazanfari et al. (Reference Ghazanfari, Nassiry and Moussavi2006) noted that the first 60 and 100 d of milk yield of animals with the AB genotype was higher than with the AA and BB genotypes in Brown Swiss cattle, and suggested that the AA genotype had a more optimal DIM. Al-Janabi et al. (Reference Al-Janabi, Al-Rawi and Al-Anbari2018) found the opposite: DIM in Holstein Friesian cattle with AB genotype was higher than in AA genotypes whilst the AB genotype had an earlier and longer period of PMY than the AA. In an earlier study, Liefers et al. (Reference Liefers, Te Pas, Veerkamp and Van Der Lende2002) showed that animals with the AB genotype produced more milk (1.23–1.32 kg per day) than the AA genotype, which is different with present study. Under commercial farming conditions, PMY and DPMY vary greatly according to care and feeding, while DP depends on herd management practices, so these differences are perhaps not surprising. Nevertheless, our own findings of lack of significant differences do not agree with any previous reports, but the variation between different studies is noteworthy.
Considering the difficulty of family selection in selection programs for milk yield traits, it was concluded that mass selection would be more appropriate for traits with a heritability of more than 0.25 (A, C, LMY305, LMY200, DIM, and PMY) while family selection would be more proper for traits with a heritability less than 0.25 (LMY, MT, LMY100, ADMY, DPMY, and DP). In a study by Javanmard et al. (Reference Javanmard, Khaledi, Asadzadeh and Solimanifarjam2010), an insignificant association between the EBVs for Holstein Friesian bulls' milk yield and their leptin genotypes (AA and AB) was determined, and this result agrees with the current study. The only significant association that we observed related to milking time. Vierbauch et al. (Reference Vierbauch, Peinhopf-Petz and Wittek2021) reported that over-milking caused a 30% change in front teat morphology, but did not cause a significant change in the rear teats. To decrease the risk of mastitis due to MT in Holstein Friesian dairy cattle, the AB genotype could be selected without compromising other milk yield traits.
In conclusion, with the exception of a small effect on milking time, we have failed to identify significant differences in milk yield traits as a consequence of leptin gene polymorphism in Holstein Friesian dairy cattle.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0022029923000717
Acknowledgments
The farm where the study was conducted and completed Ph.D. thesis are also included in the Supplementary File.
