The acceleration of technological advancements in the field of genomics, followed by the assembly of reference genome sequences of different species, has been greatest the contribution to the identification of genes associated with complex traits 15–17. Furthermore, the genetic background of economically important livestock traits such as disease resistance, milk yield, meat yield, and hematologic parameters have been revealed by associating phenotype and genotype data with a linear mixed model based analysis, which is the commonly used approach in GWAS 10,12,18–20. In this study, genomic heritability estimates for 19 blood parameters were obtained in Akkaraman sheep, which is a widespread indigenous sheep breed in Turkey. Genome-wide association analyses for each trait was implemented, which led to the suggestion of 19 specific genes and 3 unidentified locations (Table 3).
(FIGURE 2)
There are highly variable results from the few studies conducted on blood parameters in sheep. 9,21–23. For instance, a study conducted on Santa Inês sheep found the mean values for RBC, PLT, WBC, HCT, and HGB traits to be 9.97, 383.8, 10.96, 29.05, and 8.43 respectively 22. In the current study, these values were found to be similar for HGB (9.51) and WBC (9.68); however, for PLT (455.0) and HCT (23.25), the values were found to be higher. Finally, the mean value for RBC (5.43) was found to be lower. In another study conducted on Alpine Merino sheep, the means for MCH, MCHC, and RWD_CV were found to be 13.24, 37.74, and 0.39 which are all substantially lower than the means found in the present study [MCH (24.16), MCHC (50.86) and RWD_CV (14.30)]. The possible differences for the results can be attributed to different environmental (i.e., climate, ecotype, altitude), physiological, age, and health-related conditions as well as different genetic backgrounds of the breeds.
(FIGURE 3)
In the present study, low to moderate (from 0.00 to 0.55) heritability estimates were found for blood parameters (Table 2). Studies conducted on the estimation of heritabilities for blood parameters of sheep are quite scarce. In a study of Santa Inês sheep, heritability estimates using the genomic relationship matrix were found to be 0.18, 0.17, 0.22, 0.20, and 0.16 for the RBC, PLT, WBC, HCT, and HGB traits respectively 24. Results in the current study were similar for RBC (0.20), PLT (0.13), HCT (0.21), and HGB (0.12), but higher for WBC (0.55). Again, for Santa Inês sheep, the heritability estimation for HCT (0.18) was found to be compatible with our current study 25. Additive genetic variance of the traits and, accordingly, heritability is affected by many natural factors such as selection, mutation, genetic drift, and genetic migration as well as artificial selection 26. Populations exposed to these natural and artificial factors for different periods of time and in various manners can be expected to have differing heritability for the same trait. However, the heritability estimates of many traits in our study were quite similar with the results of previous studies, albeit these studies contained different breeds. The concordant moderate heritability estimates of many blood parameters in different sheep breeds might lead one to suggest that there has not been any systematic selection applied to these parameters which might alter the sharing of genetic variance in the phenotypic characteristics of these animals. Thus, the genetic improvement of these parameters by selection strategies is quite possible.
As a result of the GWAS conducted in this study, a total of 20 SNPs were found to be significant for 15 different blood parameters, with 7 being genome-wide and 13 being chromosome-wide. No significant SNPs were found for HGB, RDW_SD, MON and PDW traits. The current study agrees with those prior, which show that the number of SNPs in peaks obtained for complex traits in sheep is low during GWAS analyses 9,10,12,19,27. The main reasons for this are suggested to be the relatively short linkage disequilibrium in the sheep genome and the highly polygenic structure of measured traits 28,29.
Erythrocytes are an important cell type as they directly affect performance and growth as well as the health status of animals 30. The present GWAS determined that 6 erythrocyte-related traits were significantly associated with 8 SNPs, which were made up of 2 genome-wide and 6 chromosome-wide SNPs on OAR chromosome 1, 2, 5, 18 and 21 (Table 3). Importantly, among those, SNPs with pleotropic effects were observed (rs424122039, rs402657016, and rs418188401) for RBC, HCT, MCH, MCHC, and RDW_CW traits. The top associated SNP regarding all measured CBC traits was observed for MCHC (p = 4.83e-11), which was rs424122039 on OAR chromosome 18 (Figure 1 and Table 3). Furthermore, two other significant SNPs for MCH and HCT (rs424122039, rs402657016) were found to be adjacent to that defined for MCHC. Importantly, these SNPs were within an approximately ~52 Kb distance and discovered in the MYADM-likegene family on OAR chromosome 18. Interestingly, the association between the MYADM-like gene family and variable RBC parameters in sheep has been previously reported by Gonzalez et al. (2013)10. In the multi-breed GWAS, the discovered region and rs424122039 SNP were specifically found to be associated with MCHC of three sheep breeds namely Columbia, Polypay, and Rambouillet. The region carrying the MYADM-like gene family has been reported to be conserved throughout many mammalian species (i.e., cow, pig, goat) 31,32. Protein products synthesized by the MYADM gene family have been reported to be involved in membrane formation and regulatory processes in myeloid cell lines 33,34. In addition to these functions, the increase of MYADM gene expression levels in pluripotent cells is involved in the completion of erythropoiesis 34. The genomic region containing the significant SNPs encompasses a nearly 800Kb window and was found to be a gene-rich region mainly populated by Myeloid-associated Differentiation Marker-like (MYADM-like) protein coding genes, pseudogenes, and other genes with a high number of SNP variants (Supplementary Figure 2). This same region was also found to be significant in selection signature analyses during the ovine HapMap project, which was carried out with 74 sheep breeds from many parts of the world 28 and by Abied et. al. with thin- and fat-tailed sheep breeds from Sudan and China 35. The extreme historical selection pressure of the genomic region emphasizes that this region provides superiority in the adaptive and survival capacity of sheep. Although, the top SNP and the region of interest has been previously associated with the MCHC parameter 10, this is the first study discovering the association between rs424122039 and other red blood cell traits (i.e., MCH and HCT) in sheep.
Other SNPs associated with the erythrocyte-related traits were within genes Sodium Voltage-Gated Channel Alpha Subunit 7 (SCN7A) and Sodium Voltage-Gated Channel Alpha Subunit 9 (SCN9A), which were found to have pleiotropic effects associated with RBC and RDW_CV traits. This SCN gene family encodes for voltage-gated sodium channel proteins which are mainly involved in ion transport (GO:0006811), especially sodium ion transport (GO:0006814), in mammalian cells. Likewise, the Histone Deacetylase 4 (HDAC4) gene was found nearby to rs423050506 SNP which was significant for total RBC counts. This gene encodes a protein which belongs to class II of the histone deacetylase family. As is well known, histones play a critical role in transcriptional regulation, and accordingly, cell cycle progression, and developmental events. While mature red blood cells of many mammalian species lack DNA, it is possible that HDAC4 plays an important role during RBC development.
Three genome-wide and four chromosome-wide significant SNPs were associated with leukocyte related traits on OAR chromosome 1, 2, 3, 4, 13, 19 and 20 (Figure 2 and Table 3). The SNP rs423900300 was found to be associated with total WBC count and is located within the Integrin Subunit Alpha 9 (ITGA9) gene on OAR chromosome 19. This gene encodes an alpha integrin protein which mediates cell-cell and cell-matrix adhesion. Involved biological processes include cell adhesion (GO:0007155), integrin-mediated signaling pathway (GO:0007229), extracellular matrix organization (GO:0030198), and neutrophil chemotaxis (GO:0030593). In a study of breast cancer in humans, the expression of ITGA9 was downregulated and lost in the majority of breast cancer cell line 36. On the other hand, SNP rs419009933 within LOC101117887 on chromosome 20 showed pleotropic effect on NEU and LYM traits. Interestingly, this SNP had a negative effect on NEU with 4.35 % of additive genetic variance, whereas it had positive effect on LYM accounting for3.07 % of additive genetic variance (Table 3). Additionally, the Alpha-1,6-Mannosylglycoprotein 6-Beta-N-Acetylglucosaminyltransferase (MGAT5) gene was found to be related to the neutrophil to lymphocyte ratio (NEU/LYM). MGAT5 encodes a protein belonging to the glycosyltransferase family and acts in biological processes such as the positive regulation of cell migration (GO:0030335) and the positive regulation of the STAT signaling pathway (GO:1904894). Previously, Bahaie et al. determined that deficiency of MGAT5 cause increased neutrophilic inflammation in allergen challenged mice 37. This gene also was found to be associated with T-cell glycosylation and plasma composition of the IgG glycome in ulcerative colitis in humans 38.
SNP rs429955018 on OAR chromosome 1 was associated with EOS with relatively high additive genetic variance (18%). This SNP was located within the Solute Carrier Family 19 Member 1 (SLC19A1) gene. This gene is involved in many biological processes such as folic acid transport (GO:0015884) and vitamin metabolic processes (GO:0006766). It is well known that folic acid and other vitamins have a critical role in the production of white blood cells, particularly granulocytes 39. The ~750Kbp long region flanked by SNP rs428421565 and rs424356478, were also associated with the EOS trait, where this region, on OAR chromosome 3 contains many genes (i.e., OSBPL8, ZDHHC17, CSRP2, NAV3, E2F7). Finally, SNP rs417370910, present within the Amphiphysin (AMPH) gene on OAR chromosome 4, was associated with BAS (Figure 4). AMPH is mainly involved in endocytosis (GO:0006897). Amphiphysin is one of the adhesion proteins, and used in many cells that perform endocytosis, which is a process regarding immune system cells (i.e., basophiles, macrophages) that is important in destroying the antigens (i.e., soluble antigens) 40.
Within the platelet traits studied, there were two genome-wide and three chromosome-wide significant SNPs detected for PLT, MPV and PCT on chromosome 2 and 25 (Figure 3 and Table 3). SNP rs419857573 was within undefined region LOC105608461 and had pleotropic effects on PLT and PCT, explaining 2.73 % and 7.58 % of the additive genetic variance, respectively (Table 3). Besides that, rs424459233 within the Methionine Sulfoxide Reductase A (MSRA) gene on OAR chromosome 2 was associated with MPV with 5.24 % of additive genetic variance.
The results of the present GWAS revealed many genes (i.e., SCN7A, SCN9A, HDAC4, MYADM-like, ITGA9, MGAT5, SLC19A1, SLC35F3 and AMPH) which are important to basic biological functions such as cellular transport, histone deacetylation, cell differentiation, erythropoiesis, and endocytosis 33,34,36–38,40. Some of the identified SNPs (i.e., rs418188401, rs402657016, rs424122039, rs419009933 and rs419857573) were responsible for pleotropic effects on WBC, RDW_CV, HCT, MCH, MCHC, NEU, LYM, PLT, and PCT measured traits. Furthermore, SNP rs424122039 was found to be associated with MCH, MCHC, and HCT, supporting prior work which linked MCHC measurements to the same gene-rich MYADM-like region10. Previous studies have also shown that this region is under historical selection pressure 28,35. In this context, the MYADM-like gene was suggested by Gonzalez et. al., (2013) as a potential candidate gene associated with erythrocyte related traits (i.e., MCHC); moreover, the results of our current study confirmed this hypothesis. Our current study is one of the limited number of studies aimed at understanding the genetic components underlying blood parameters in sheep. The current study and future studies of its kind will shed light on the genetic region that effect measured blood traits and how to use these associations to further selection programs.