The preferred alleles or genes, associated with productive traits or survival traits, as a response to artificial selection or adaptation to the surrounding environment, are targeted by the selection process (artificial or natural), causing their frequencies to be elevated compared to those under norm or non-selection pressure status. As a result, when comparing the genomes of various breeds, it is possible that the genome architecture has changed in some specific regions that collectively manifested in subsequent generations as signals. Such signals, also known as selection signatures (SS), may be recognized using the variance in linkage disequilibrium. According to several investigators [13, 17, 18, 6], the varLD technique is optimal for detecting the SS over long or short genomic distances. To the best of our knowledge, this is the first research on Bahraini sheep breeds to be published, particularly at the genetic level of Tomtom and Nuaimi in comparison to other European and Asian Breeds. The obtained results indicated that the decays of LD are much higher in those breeds who has selection signature compared to the others. Our results revealed the regions of the highest varLD scores (within the autosome chromosomes containing the SS) to be distributed within 1000 kb in distance.
Principal components
The results obtained clearly identified the high distance observed between the Asian Mouflon and European Mouflon in addition to Tomtom breeds that were clearly separated from each other and from Nuaimi, Afshari and Zel breeds. Previously Ciani et al., [19] reported the high distance between Asian Mouflon and European Mouflon breeds that did not affect the structure of the PCA when removing the European Mouflon from the analysis of 256 sheep breeds genomes that indicated the low contribution of the European Mouflon breed to the genomic pool of Asian Mouflon (totally separated). The Asian Mouflon was reported to be in a separate cluster when compared to East Asian, Middle Eastern, European, and African populations and was of the lowest LD (r2) when was compared to the Improved or modern breeds of sheep [20].
Highest selections signatures and the genes located near those SS detected in breeds contrasts
The Tomtom/Zel breeds comparison revealed the strongest signal that concurred with high LD on chromosome 6 of Tomtom breed as compared to Zel. At chromosomal level, similar findings were reported by Taghizadeh et al., [21] when investigating the selection signatures in zel and other sheep breeds and Zhang et al., [22] who found several SNPs located on the 14 chromosomes (Chr-1, chr2, chr-3, chr-4, chr-6, chr-7, chr-8, chr-9, chr-10, chr12, chr-14, chr-18, chr-21 and chr25) to affect some developmental traits in farm animals. In sheep, chromosome 6 was reported to inhabit the highest significant selection signature at genomic position of 69.39 Mb [23] others [24, 25] also reported chr-6 genes to play a substantial role in selection process. For instance, in our study, FBXL5 gene that was detected near the observed selection signatures found in chr-6 is involved in the cellular maintenance through its key function in the intestinal iron sensitivity [26]. While the detected mutations in TAPT1 gene interrupt ciliogenesis and resulted in a congenital syndrome in addition to several clinical and phenotypical signs that may observed such as skeleton hypo mineralization, fractures in the intra-uterine, and numerous abnormal developments of the lungs, brain and kidney [27]. PROM1 is a protein-coding gene, which plays a role in cell apoptosis and its proliferations in addition to differentiation. Comparable findings (to ours), Zhang et al., [22] reported PROM1, genes as candidate gene to body weight of Mongolia cashmere goat. The same authors also show a SNP located near LDB2 gene - on Chr-6- was indicating it as candidate gene signal for stimulating the Mongolian Cashmere goat body weights. In addition to that a gene located within the same chromosome (6) called FBXL5 was found to affect body weight development. Previously LDB2 gene was claimed to affect body weight [28], furthermore Wei et al., [29] reported it to be strongly associated with carcass and some growth traits of chickens. The Transmembrane Anterior Posterior Transformation 1 (TPAT1) gene that identified near the selection signature in our study was designated as biological regulator and was correlated to the residual concentrate intake of the Italian Brown Swiss cattle [30].
Nuaimi/Tomtom comparison revealed UGT8 genes (chr-6) to have a key role in selection and development. Previously Smallwood et al., [31] and Halbleib and Nelson [32] demonstrated UGT8 gene to associate with the endocrine and the nervous system’s function and development. Kim et al., [33] detected selection signature near this gene when studying sheep and goat genome in arid environments. Similar to our findings, Zhao et al., [34] specified selection signatures located on chromosome- 24 to be associated with the selection process conducted for different purposes (naturally or artificially).
In Nuaimi/Zel breeds comparison the selection signatures suggested the pyridoxal dependent decarboxylase domain containing 1 (PDXDC1, chr-24) gene to be a candidate gene that may act in the developmental process when comparing the two breeds. Zheng et al., [35] concluded a strong correlation between PDXDC1 gene and the amino acid expression in goat fetal fibroblast. Moreover, Rrn3 gene is an important mammalian RNA polymerase formation that acting as an initiation factor for the RNA polymerase transcription [36]. The MPV17 gene has potential effects on the development process as demonstrated by Uusimmaa et al., [37] who find a relationship between the hepato cerbral maternal DNA depletion syndrome with a mutation detected in the MPV17 gene. Furthermore, in humans, mutations in MPV17 were linked to recessive autosomal adult-onest leukoencephalopathy and neuropathy in addition to changes in the mtDNA sequence (deletion) of the skeletal muscle. The NTN1 gene was reported to be in a region associated with sheep’s body size (33, 38]. Another gene (NDE1 gene; located on Chr-24) was demonstrated by Alkuraya et al., [39] to has a substantial role in the cortical and cerebral neurogenesis in Homo sapiens. MARF1 gene that detected to harbor selection signature on chr24 was demonstrated to be required for the completion of oogenesis in mice [40, 41]. Chromosome 2 has also contributed to the developmental process amid selection through IKZF2 gene and is believed to act as a regulatory transcriptional factor for the maturation of the outer hair cell [42]. IKZF2 gene was proposed to play a regulatory function to T cells in relation to human leukemia [43], while in bovine this function was not approved by de Souza et al., [44]. The BMERB1 gene expression was observed in many tissues but it was highly abundant in the testis, hypothalamus, whole embryo, mammary gland, and the brain stem, that may guide to have a key role in reproductive activities of the sheep [45]
The Afshari breed comparisons made with Nuaimi and Tomtom, revealed chromosome 22 of high importance since the selection signatures located within it. Adeniyi et al., [46] indicated chromosome 22 to localize many selection signatures when studying Balusha ovine breeds/strains. BAG3 gene is one of the genes found on chromosome 22, which was demonstrated by Paten et al., [47] to be expressed highly during lactation in the mammary glands. It was suggested to be used as a marker for assisting the selection for improving milk persistency as it was claimed to be involved in conserving the secretory milk cells and the overall milk productivity. ATE1 gene was identified in Nellore Cattle to host a selection signature that associated with milk and growth rate [48]. By applying the hapFLK method for detecting selection signature comparable results were also obtained by Yuan et al., [49] who reported AET1 and WDR11 genes to be among the genes that found harboring important selection signatures detected in the Chinese indigenous sheep breeds. DLC1 (Chr-26) gene was indicated among the genes that signaling against parasite infections, as its expression was activated following studying long non-coding RNAs in adult sheep challenged for nematode infection [50]. Furthermore, in esophageal cancer cells; dropping down of the DLC1 gene expression will motivate the glycolysis of those cells and its overall growth [51]. Chromosome 20 also found to have sequence differences and differed in its structure between Nuaimi/Asian Mouflon breeds. One of its genes (CDKAL1) variants (rs4712524 and rs4712523) differed in their frequency and were associated with human type two diabetes mellitus [52].
The average temperature in Bahrain is higher than that in the other non-Gulf countries, which may cause heat stress that is difficult to adapt to, forcing the animal's body to resist stress through adaptation by increasing the gene expression of some genes. Perhaps some genes stop expressing while others increase, favoring some genotypes desired and resisted in such circumstances.
In addition to that, the obtained results and genomic regions that contains the proposed selection signatures may have explained by that both artificial and natural selection should have left their imprints on specific genomic regions. As a result, identifying genomic regions with such selection signatures may aid in clarifying the molecular processes of selection along with identifying likely contender genes of interest [34]. Furthermore, the specific surrounding conditions prevailed in Bahrain such as high ambient temperature1 (may go up to 48.7 oC) along with high relative humidity (real feel can be as high as 55 oC), in addition to the preference of the sheep owners that focused on specific phenotypic characteristics may have led to an elevation in the frequencies of the loci that host the selection signatures. Similarly, and by using Transcriptomic analysis (RNA-sequencing), Lu, et al., [53] discovered that some genes (near to some selection signature detected in our study) were differentially expressed in heat-stressed conditions, including LDB2, TAPT1, FBXL7, GRK5, WDR86, PLPP4, C24H16orf72, DLC1, KIF114, GRK5, RGS10, LOC105614340 and PLPP4 which may explain partially our findings.