Genomic Analysis of Sarda Sheep Raised at Diverse Temperatures Highlights Several Genes Involved in Adaptations to the Environment and Heat Stress Response Authors: Giustino Gaspa, Alberto Cesarani, Alfredo Pauciullo, Ilaria Peana, Nicolò P. P. Macciotta Published: 2024 ABSTRACT Simple SummaryIn extensive breeding systems, environmental conditions strongly influence animal behavior and production. During the process of evolution, animals tended to adapt their morphology and physiology to environmental conditions, leaving genomic imprints. This adaptation can be traced in the animals’ genomes, relating environmental features to genome-wide differentiation metrics. In this study, using Sarda sheep living at different temperatures as a case study, we compared their genomes to highlight traces of thermal tolerance and adaptation. 1. Introduction Livestock production is affected by complex traits that can be modeled according to several factors (e.g., genetics, climate, management, and nutrition). It is well known that livestock production and health are affected by environmental conditions. Indeed, environmental adaptation is an important evolutionary feature among animals [1,2]. Thanks to their homoeothermic feature, mammals are able to cope with most extreme environments, distinguished by different temperatures [3]. The process of evolution causes changes in animal genomes, exploiting variations in their morphology (e.g., body size, skin, wool, and fat thickness) and physiological traits (e.g., fat metabolism) in order to adapt to different climatic conditions. Although livestock have adapted to live and produce in different environmental conditions [4], some factors can still have detrimental effects on production, growth, or fitness traits. The response of animals to variations in climatic factors, such as increases in temperature, may induce heat stress [5,6]. The negative effect of temperature, measured using the temperature–humidity index (THI), on productive or reproductive traits, has been the object of several studies in both cattle [7,8,9,10] and small ruminants [11,12,13]. Thermic stress may potentially worsen—even in temperate zones—due to the emerging local modifications of weather associated with climate change [14]. Sheep raised under adverse meteorological conditions face heat stress, which can alter their behaviors. Moreover, physiological changes due to harsh environments can negatively impact reproduction, milk production, and composition [15,16]. Climate change represents a major challenge that governments, research institutions, and agribusiness stakeholders must face in the very near future. In this respect, it is of crucial importance to identify genotypes that enable the production of food under changing environmental conditions. To date, landscape genetics approaches have investigated genome changes driven by environmental conditions, analyzing the effects of temperature, rainfall, altitude, and other environmental features on the genotype distribution using an integrative framework that combines landscape ecology, population genetics, and molecular markers [17,18]. This approach aims to understand how environmental factors modify gene flow and animal genotypes [19] and predict the spread of diseases (or disease vectors) and harmful species [20]. Burrow [21] pointed out that breeding schemes must account for the environment in which animals are raised. The large availability of genomic data has allowed for comparison of divergent breeds and the identification of selection signatures for artificial selection [22] or environmental adaptation [23,24]. Moreover, by studying indigenous breeds living in hot and arid environments, complex gene networks, which mediate the capabilities of sheep and goats to face environmental gradients, have been hypothesized [25,26]. In our view, sheep breeds from southern Europe may represent an interesting model by which to study the effects of environmental adaptation using SNP data, due to the low impact of artificial selection and predominantly extensive farming system [27]. To minimize confounding factors (e.g., the difference in demographic history that occurs when a joint analysis of multiple breeds is carried out) as much as possible, in our study, only one breed was investigated with the aim of detecting the genomic signatures of temperature gradients. The Sarda breed is autochthonous in Sardinia but is also raised in Central and Southern Italy. Sarda sheep account for about 80% of Italian dairy sheep and 43% of Italian ovine stock [28,29]. These animals are well adapted to different types of farming systems, ranging from extensive to semi-intensive pasture-based farming. Sarda sheep are characterized by their small–medium size with white fleece and polledness in both sexes. They are generally selected for their milk quantity and resistance to scrapie [30]. The genetic merit of their breeding stock varies according to the production system in operation, due to limited ram exchanges and large variability among geographic zones. Their milk is fully transformed into cheese, with a production of 60,000 tons/year [31]. The aim of this study was to identify traces of environmental adaptation in the Sarda sheep genome by comparing two groups of animals. The Sarda population provides a good example of sheep that have adapted to the Mediterranean climate, and the test groups were defined according to the historical average temperatures of the zones where they are raised: one group living in the hot areas and another in the cold areas of the second-largest Mediterranean island. 2. Materials and Methods 2.1. Animal and Genotypic Data A total of 825 Sarda breed dairy ewes were genotyped using the Infinium Ovine SNP50 v1 BeadChip (Illumina Inc., San Diego, CA, USA), containing 54,241 markers. Quality control was carried out using PLINK v1.9 [32] on both subjects and SNPs. No animals were discarded, according to a call rate < 0.95. Subsequently, SNPs were filtered based on the call rate (>0.975), deviation from the Hardy–Weinberg equilibrium (p-value < 0.00001), and minor allele frequency (>0.02), and, ultimately, 45,947 markers were used. The ewes, farmed in 46 different flocks (Supplementary Figure S1), were allocated into two groups—“cold environment” (CE, 395 ewes) and “hot environment” (HE, 430 ewes)—according to the average maximum temperature over a 20-year period recorded by a network of 39 out of the 60 weather stations operated by the regional department of the environmental protection agency (ARPAS, http://www.sardegnaambiente.it/arpas/ accessed on 10 July 2024) of Sardinia. A neighborhood criterion was adopted for matching the genotypic record to CE or HE class: 50 km was the maximum distance from the flocks that allowed us to consider them belonging to a particular weather station. The breaking temperature was fixed at 21 °C. The animals were selected from large numbers of flocks in order to be largely representative of herd book variability. The 825 dairy ewes were the offspring of 499 rams, with 2.1 ± 1.7 ewes served by each ram on average. The Sarda samples were plotted against the first two principal coordinates of the multidimensional scaling of genotype data using the --mds-plot flag of PLINK v1.9 [32] (see Supplementary Figure S2). 2.2. FST Evaluation Wright’s fixation index (FST) was used following the metrics proposed by Nei [33] and implemented via an in-house Python script. Subsequently, a locally weighted scatterplot smoothing regression (LOWESS) was applied using PROC LOESS in the SAS software 9.2 (SAS Institute, Cary, NC, USA) according to [34] (the smoothing parameters are reported in Table S1). The smoothing parameters were chosen in such a way that each window interval always included 20 SNPs. The chromosome-wide smoothing parameters were computed as 20/number of SNPs. The control chart of PROC SHEWHARD in SAS was used to identify outlier markers that exceeded three standard deviations from the chromosomic mean value, according to Sorbolini et al. [35] (see Supplementary Files for full scripts). The Manhattan plot of LOESS-predicted FST was prepared using the qqman package (v. 0.1.9) in R (v. 4.3.1). 2.3. Gene Annotation Analysis Annotated genes in genomic regions corresponding to the FST outlier signals associated with thermal gradients were downloaded from the National Center for Biotechnology Information database (www.ncbi.nlm.nih.gov, accessed on 15 July 2024) using the Ovis_aries_rambouillet.ARS-UI_Ramb_v2.0.112.gtf assembly file and retrieved with the GALLO package (v. 1.5) in R [36] via the functions import_gff_gtf() and find_genes_qtls_around_markers(). An interval of 250 kb upstream and downstream from each significant marker was considered. Additionally, from the gene list, genes previously associated with environmental adaptation were sought from public databases. Gene enrichment analysis was conducted using the R package gprofiler2 (v 0.2.3) [37], with homo sapiens set as the reference model. A co-expression pattern network was also analyzed (https://string-db.org/ accessed on 10 September 2024). 3. Results 3.1. FST Outliers In Figure 1, smoothed FST values predicted with LOWESS are reported for each of the 27 ovine chromosomes, highlighting the upper signals: 623 SNPs resulted as outliers in the FST values and were grouped into 97 selection signatures based on the CC approach (Supplementary Table S1). OAR1 was the chromosome with the highest number of outliers (83), while OAR12 showed the lowest number of SNPs (5). The smoothed FST values ranged from 0. 0005 to 0.01. On average, the smoothed FST values were lower than non-smoothed ones and lower than those retrievable from the literature on sheep breed comparison [38]. 3.2. Positional Candidate Genes From the outlier signals detected via FST analysis (Supplementary Table S2), 280 positional annotate genes in 27 chromosomes were retrieved using the GALLO software package (v. 1.5) within a 500 kb-interval from the peak SNPs. Annotation details are reported in Supplementary Table S3. Around the declared significant markers, 51 genes in 17/27 sheep chromosomes have been previously associated with several phenotypes in domestic ruminants using different approaches (e.g., thermal tolerance, thermal stress, adaptation to extreme conditions, milk yield and fat composition, body fat composition, and fatty acid metabolism). Detailed descriptions of retrieved genes, their roles, and a full bibliography are all provided in Supplementary Table S4. Table 1 shows the most meaningful selection signatures in OAR1, OAR2, OAR3, OAR4, OAR5, OAR6, OAR8, OAR11, OAR15, OAR16, OAR20, and OAR22. A total of 23 genes found in this study were previously putatively related to heat stress in ruminant GWASs for heat stress (FCGR1A, MRPL9, SERPINE2, MDH1 UGP2, RIPK1, SERPINB9), in candidate gene studies (HSPB3, SOCS3), or were differently expressed genes (DEGs) in RNA-seq experiments with animals subjected to heat stress condition (TDRKH, DNER, UGP2, MYO1G, SORL1, SERPINB1, HIF1AN). Besides the genes directly involved with heat stress, Table 1 also highlights genes that were found in the literature as selection signatures of environmental adaptation, defined using climatic or other environmental variables (NMUR1, PDE6D, COPS7B, STK17A, MFAP3, COMMD8, SCFD2, SNX3, RNF121). The other 32 genes highlighted in our study have been previously associated with phenotypes that might be indirect indicators of climatic adaptation. In particular, 25 genes were previously found to be associated with physiological traits that regulate metabolism, the mobilization of energy, and increased metabolic heat body: body weight (PHGDH, SGCB, and TPM1), skeletal muscle development in sheep and cattle (SCAMP1, HIF1AN), fatness and fat deposition (RORC, AGPAT2, ABCD2, MFAP32, YTHDC1, RNF121, SIRT3, SCD, RNF121, LYRM1), intramuscular fatty acid composition (AGPAT2, ABCD2, MFAP3, SCD), and fat-type tail in sheep (COPS7B, PDE6D, and USPL1). They may be indirectly linked to heat tolerance. They are also involved in cell molecular mechanisms related to wool bulb regression and regeneration in sheep (S100A11), ultraviolet radiation resistance (EDEM1), and immune response (FNDC3B, PELI1) (Table 2). Finally, 11 genes were found to be associated with productive traits, such as amount of milk (HERC3, SCFD2, CHUK, EEF2K), milk fat (SPATA16, GALNT10, SCD, SEC31B, ZP2), or immune response (PELI1, HERC3) (Table 3). Eight positional candidate genes highlighted in our study (MFAP3, RNF121, HIF1AN, PELI1, PDE6D, COPS7B, SCFD2, SCD) were linked to more than one phenotypic or environmental trait and identified in more than one species (recurrent genes in Table 1, Table 2 and Table 3). Gene enrichment analysis was conducted in order to profile the 51 genes reported in Table 1, Table 2 and Table 3, highlighting a significant intersection among these genes relative to GO molecular functions, cellular components, and binding proteins (Table 4 and Table S5). The GO:0019771 (high-affinity IgG receptor activity) was related to immune response (FCGR1A). GO:0030060 and GO:0030060 (MDH1 and PHGDH genes; malate dehydrogenase activity) exert an important metabolic role in energy production. GO related to the inhibition of molecular function or enzyme activity (GO:0004857, GO:0140678) presented the same gene intersection (SERPINE2, PDE6D, SOCS3, SORL1, SERPINB9, SEPINB1) as the GO for peptidase inhibitors and regulations (GO:0004866, GO:0030414), which relate to interaction among the SERPINE2, SORL1, SERPINB9, and SERPINB1 genes. These are protease inhibitors that may protect cells from damage by inhibiting excessive proteolysis during inflammation conditions [98]. The GO-associated cellular components were mainly related to endo- and extra-cellular vesicle trafficking; in particular, GO:0031410 (cytoplasmic vesicle) and GO:0097708 (intracellular vesicle) presented a significant intersection for 11 genes (S100A11, SERPINE2, UGP2, MDH1, MYO1G, SNX3, SORL1, PTPRJ, SERPINB9, SERPINB1, PHGDH). Moreover, the STRING protein–protein interaction network, based on co-expression and co-occurrence in databases or experiments, is shown in Supplementary Figure S3. 4. Discussion Response to heat stress and heat tolerance are generally treated as quantitative traits [58] as they are assumed to have a polygenic background. The determinism of heat tolerance, measured according to the THI, is genetically negatively correlated with production traits—at least in dairy cattle—and improving this feature may lead to detrimental effects on other economic traits [99]. The knowledge of genetic variants associated with heat tolerance may be a first step towards dissecting the genetic architecture of such a complex trait. Several studies investigating the associations between thermal stress and genotypes [100,101,102] have identified genes likely to affect this trait. In our investigation, the use of an FST-based approach allowed us to suggest positional candidate genes close to outlier SNP markers. Many of these genes have been already associated with the response to thermal stress (Table 1) or environmental adaptation (Table 2 and Table 3) in the literature. 4.1. FST Outliers and Response to Thermal Stress In our study, the predicted FST values were evaluated in terms of a chromosome-wide distribution, rather than as absolute FST values. This is because they were the product of a local regression that emphasizes larger values when the neighboring SNPs also presented high values (probable outlier signals), whereas the same signal was regressed toward the window’s mean if the majority of neighboring FST had lower values [35]. Most of the signals were present in the first three sheep chromosomes, reflecting the larger size of the chromosomes. In the following, the most meaningful selection signatures are discussed according to the results presented in Table 1. In OAR1, we found seven selection signatures (Supplementary Table S2). Two nearby FST signals (98.7–101.1 Mb) corresponded to the Fragment of IgG Receptor Iα (FCGR1A) and Mitochondrial Ribosomal Protein L9 (MRPL9) genes, which have been found to be previously associated with heat stress response in Zebu cattle [39]. Mehla et al. [39] have pointed out that gene expression patterns diverged between animals exposed to heat stress and groups of animals raised under optimal climate conditions. The FCGR1A gene was observed to be suppressed at 4 h after heat stress but was induced at 24 h and 48 h into heat recovery. Ten selection signatures were detected on OAR2, corresponding to 23 genes (Supplementary Tables S3 and S4). Four FST signals between 22.6 and 23.3 Mb were further investigated, where we retrieved the SERPINE2, NMUR1, COPS7B, and PDE6D genes. Dikmen et al. [41] found that SERPINE2 explained 3.0% of the variance in the sweating rate. This gene, which has been reported to be associated with sweating rate and thermoregulation in dairy cows during heat stress, produces a proteinase-inhibiting thrombin and urokinase-type plasminogen activator [41]. Furthermore, it is involved in most of the molecular processes linked to the inhibition of peptidase (Table 4, GO:0004857, GO:0140678, GO:0004866, GO:0061135, GO:0051248). The inhibition of thrombin can induce changes in epithelial cells from human eccrine sweat glands [103]. The Neuromedin U Receptor 1 (NMUR1) gene was found to be a signature of positive selection for environmental adaptation in sheep [23]. The NMUR1 gene encodes for a receptor of NMU involved in appetite regulation [104], which is generally suppressed during heat stress. The COPS7B and PDE6D genes were suggested in our study as positional candidate genes related to adaptation to temperature. These findings are in agreement with Gouveia et al. [42] for COPS7B in sheep, whereas no indication was reported for PDE6D [42]. Despite there being no evidence directly linking PDE6D to heat stress in ruminants, regarding its involvement in stress responses and the regulation of cellular processes, Gouveia et al. [42] proposed PDE6D as a positional candidate gene that differentiates Brazilian locally adapted sheep breeds. In OAR3, six selection signatures were identified (Supplementary Table S2). In the region 44.1–44.5 Mb, five genes were observed as potential candidates (PEL1, VPS54, UGP2, WDPCP, MDH; see Table S2). Among the genes reported in Table 1, MDH1 (Malate Dehydrogenase 1) is an important metabolic enzyme that catalyzes reactions in the tricarboxylic acid cycle pathway, thus playing a role in energy production [43]. MDH1, together with UDP-Glucose Pyrophosphorylase 2 (UGP2) and Dimethylglycine Dehydrogenase (DMGHD), were hypothesized as candidate genes for heat tolerance by Cheruiyot et al. [43]; they also were found to be overexpressed in liver tissue in an experiment relating heat stress conditions with physiological parameters in dairy cows [45]. These three genes are over-represented in both the present study and the existing literature, as the citrate cycle/tricarboxylic acid cycle pathway is central to mitochondrion energetics and might help to prevent cellular damage during heat stress. A total of 54 SNPs resulted as outliers for FST, defining four selection signatures in OAR4 (28 genes; see Supplementary Tables S1 and S2); in particular, in the range between 78.7 and 79.7 Mb, we retrieved two genes: Serine/Threonine Kinase 17a (STK17A) and Myosin IG (MYO1G) (Table 1). The STK17A gene has been suggested by Yang et al. [50] as being positively selected for high-altitude and arid environment adaptation in Tibetan sheep, given its role as a regulator of cellular reactive oxygen species (ROS)—an important functional activity in the pathway of Hypoxia-inducible Factor1 (HIF1). More interestingly, the subunit 1-α of H1F (HAF1AN, located in OAR 22: 21.1–21.2 Mb, Table 1) retrieved in our study has been reported to be overexpressed in heat-stressed dairy cattle [105]. The plasmatic HIF-1α of heat-stressed cows—indicative of the risk of oxidative stress—showed a higher level under heat stress conditions [105]. The second gene, MYO1G, was upregulated in the peripheral blood leukocytes of Zebu cattle exposed to heat stress [46]. In addition, [106] has reported MYO1G as a positional candidate in a selection signature emerging from a comparison among African cattle breeds. The MYO1G gene is associated with immune response and host defense, playing a specialized role in immune cells (i.e., T cells and leukocytes), where it regulates cell migration and immune synapse formation [107]. The MYO1G gene is part of the unconventional myosin family, a group of proteins involved in intracellular transport, cell motility, and cytoskeletal organization. Our enrichment analysis collocated MYO1G in several gene intersections for GO related to vesicle, extracellular vesicle, and extracellular exosome (GO:0031982, GO:1903561, GO:0043230, GO:0070062, GO:0070062; Table 4). Unlike conventional myosins, which participate in muscle contraction, unconventional myosins such as MYO1G play roles in diverse cellular functions (particularly related to membrane dynamics), and have been associated with actin cytoskeleton remodeling and plasma membrane functions (e.g., cytoskeletal integrity, cell motility and migration, and membrane repair), as well as the regulation of cell stiffness in B-lymphocytes [107]. This latter function is necessary for the recovery of membrane proteins and the prevention of cellular damage. Indeed, heat stress could damage proteins at the cell surface, and MYO1G might aid in their internalization and processing via the endosomal system. This is vital for cell survival, as it prevents the accumulation of damaged proteins on the membrane. Heat stress often initiates an immune response, and MYO1G could play a role in ensuring that immune cells maintain their ability to move and respond to stress-induced inflammation. Sheep use various mechanisms to adapt to heat stress, such as increasing their respiration rate and changing their body temperature [6]. The MYO1G gene, being involved in cellular transport and cytoskeletal interactions, may potentially play a role in maintaining cellular integrity under heat stress conditions [108]. OAR5 presented four selection signatures and five positional candidate genes, as reported in Supplementary Tables S1 and S3 (NMUR2, FAM114A2, MFAP3, GALNT10, HAND1). Although only Microfibril Associated Protein 3 (MFAP3) has been significantly enriched under thermal tolerance conditions in local Chinese cattle [48], NMUR2 is the second receptor of the aforementioned NMU, which has been observed in brain sites that are important for appetite regulation [104]. Interestingly, outlier FST signals were retrieved for OAR6 (seven outlier signals and 19 genes), OAR8 (two outliers, seven genes), OAR11 (two outliers, 12 genes), OAR15 (three outliers, 10 genes), OAR16 (three outliers, five genes), OAR20 (one outlier, nine genes), and OAR22 (two outliers, 22 genes). The outlier SNP markers and positional candidates retrieved in these genomic regions are reported in Supplementary Tables S2 and S3. The region between 66.9 and 69.6 Mb, where a gene related to immune response was found (the COMM Domain Containing 8 (COMMD8), is of particular interest. Although there is limited evidence regarding COMMD8 in the context of heat stress, this gene has been reported, using a landscape genomic approach [49], to be a positional environmental outlier. Indeed, members of the COMMD family are involved in regulating Nuclear Factor k Chain Transcription in B Cells (NF-κB) signaling and copper homeostasis [109], both of which might be influenced by thermal stress. The proper regulation of NF-κB is important for managing the inflammatory response and protecting against heat-stress-induced cell apoptosis [110]. The FST signals in OAR11 (28.9–29.2) were close to the SOCS3 gene map (Suppressor of Cytokine Signaling 3). SOCS3 is involved in regulating cytokine signaling pathways, which are a crucial cellular stress response mechanism [111]. Heat stress induces an inflammatory response, and SOCS3 helps to modulate this response by negatively regulating pro-inflammatory cytokines such as IL-6 and TNF-α [112]. In bovines, SOCS3 downregulated GH-dependent transcriptional activation in response to heat stress [52]. Although the same authors [52] pointed out that heat stress alone was partially responsible for the SOCS3-related negative modulation of hepatic GH receptor abundance, a reduction in GH levels during thermal stress has also been observed by other authors [6]. In OAR15, the Sortilin-related Receptor 1 (SORL1) gene highlighted in our study has also been shown to be upregulated in the peripheral blood leukocytes of Zebu exposed to heat stress [46]. One of the three FST outlier signals in OAR16 (24.8 Mb) mapped close to the positional candidate Heat Shock Protein Family B (small) Member 3 gene (HSPB3). HSPB3 has been associated with the heat stress response [56]. These authors compared miRNAs levels and target genes in two Holstein cattle groups: one heat-stressed and the other not. The HSPB3 gene was significantly associated with stress exposure. Moreover, other heat-response miRNA showed differences in the induction of their expression levels between buffalo and native cattle [112]. There is no direct link between the HSPB3 gene and heat stress in ovines; however, heat-shock proteins (HSPs) are well known to play a critical role in cellular responses to stress, including thermal stress, across different species [46]. HSPB3 expression in the muscles of chickens raised in lowland areas varied according to the time of the day, in contrast to highland chickens [113]. These proteins are key players in protecting cells from the damaging effects of heat as they ensure proper protein folding and prevent aggregation. Although HSPB3 has not been highlighted in the same way for ovine heat stress, it could potentially play a similar protective role in muscle or other tissues as it does in other animals. The more commonly studied HSP70 and HSP90 genes have been associated with heat adaptation in domestic ungulates [114]. In OAR20, one selection signature and nine genes were retrieved from a public database (Supplementary Tables S1 and S2). Three interesting positional candidate genes, RIPK1, SERPINB9 and SERPINB1, located at 49.6–49.9 Mb, are involved in immune response and will be jointly discussed due to their roles in the negative regulation of the protein metabolic process (Table 4) and because they have been reported to be related to heat stress responses in ruminants. The Receptor Interacting Serine/Threonine Kinase 1 (RIPK1) gene is central to cell survival, apoptosis, and necrosis. Under heat stress, cells undergo apoptosis as part of a protective mechanism for removing damaged cells and preventing inflammation due to necrotic cell death [115]. The RIPK1 gene has been proposed as a positional candidate that would explain beef cattle body temperature during heat stress [57]. The Serpin Family B Member 9 (SERPINB9) gene has also been proposed for this function. These genes code for proteins involved in inflammation and cellular death in response to external stresses that cause tissue damage. Heat stress often triggers oxidative stress, protein misfolding, metabolic alterations, and immune system activation, making genes associated with these processes especially important for adaptation to heat stress [116]. In other mammalian models, under heat shock conditions, RIPK1 activity helps to balance the decision between cell survival and death [117]. Serpin Family B Member 1 (SERPINB1) is a protease inhibitor that protects cells from proteolytic damage during stress, which has been found to be downregulated in Zebu cattle exposed to heat stress [46]. The same author explained the reduction in the expression of immune response genes as a consequence of reduced metabolic activity and feed intake during heat stress [46]. 4.2. Fst Outliers and Environmental Adaptation Climate adaptation also has consequences for different livestock features, such as body size and composition, lipid metabolism, and changes in the amount and quality of milk production. Livestock body size is a complex phenotype, affected by several different features, which has undergone severe transformation during evolution: adaptations to the environment are strongly associated with body dimensions in mammals, humans included [3,118,119,120]. In Table 2, the genes found in this study, which have been associated in the existing literature with body size, amount of fat, and lipid metabolism, are reported. HAF1AN was identified as a transcriptional regulator (either in ovine and bovine contexts) in a comparative gene expression study on skeletal muscle longissimus thoracis muscle [82], but has also been associated with heat stress, due to its synergistic activation of the expression of HSP family members in dairy cattle [105]. As far as wool is concerned, the S100 Calcium Binding Protein A11 (S100A11) gene was mapped in a selection signature highlighted in this study. This gene has been reported to be associated with wool characteristics [61]. These authors observed the overexpression of this gene in the second stage of the life cycle of hair bulbs; the S100A11 gene was also associated with the inhibition of keratinocyte apoptosis [121]. EDEM1 is an intriguing gene for environmental adaptation. Indeed, EDEM1 is associated with lifespan in Drosophila melanogaster and Caenorhabditis elegans [122]. EDEM1 levels were reduced in fibroblasts obtained from a dwarf mouse model of longevity. These fibroblasts were more resistant to cell death from stressors such as UV light [76] and have been linked to the youthfulness of skin in humans [75]. Pellino E3 Ubiquitin Protein Ligase 1 (PELI1) and HECT and RLD Domain Containing E3 Ubiquitin Protein Ligase 3 (HERC3) are candidate genes associated with immune responses in animals, regulating the inflammatory response initiated by NF-κB [123]. Despite there being no direct evidence of the effects of these genes on thermal stress, prolonged exposure to high temperatures has negative consequences on the immune response in animals. Furthermore, Gupta et al. [124] have reported that heat stress causes a decrease in white blood cell count. Additionally, in the literature, the Protein Tyrosine Phosphatase, Receptor Type J (PTPRJ) gene has been associated with high-altitude adaptation in Cashmere breed goats [73]. As previously mentioned, the tail type has a strong association with climate adaptation, especially in sheep [72]. Taking a FST-based approach, Yuan et al. [72] observed a significant signal of the differentiation between thin- and fat-tail sheep near the Ubiquitin Specific Peptidase Like 1 (USPL1) gene, where the latter is known to be an adaptative trait. COP9S7B was suggested in this article to be a positional gene that relates to temperature adaptation in sheep, in agreement with [42], but COP9S7B has also been highlighted as a selection signature that differentiates fat-tail from non-fat-tail sheep [62]. For many genes found in our study (RORC, AGPAT2, AGCD2, MFAP3, YTHDC1, RNF121, SIRT3, SCD), genetic associations with body fat composition [59,63,64,65,66,69,74,77] and intramuscular fatty acid metabolism have been reported [78,79,80,81] (Table 2 and Supplementary Table S4); other interesting genes have been associated with milk fat in dairy ruminants (GALNT10 [85], CHUK [90], SCFD2 [51,88,89], SEC31B [94] ZP2 [95,96], SCD [79,81,91,93] and SPATA16 [84], Table 3). Hormones secreted during stressful situations (e.g., epinephrine) have been observed to induce lipolysis and to increase the amount of NEFAs, and similar mechanisms could be hypothesized for thermal stress. However, ref. [52] has reported that fat mobilization in cows under stress was not as intense as expected. 4.3. Heat Stress and Immune Response Many of the genes listed in Table 1, Table 2 and Table 3 are involved in cellular stress responses, including protein folding (HSPB3), immune regulation (MYOG1, SORL1, SOCS3, SERPINE2, CHUK, HERC3, PELI1), metabolic adaptation (SIRT3, MDH1), and apoptosis control (STK17A, RIPK1, SERPINB1, SERPINB9). These genes, both individually and collectively, might contribute to the ability of sheep to mitigate the detrimental effects of heat stress. They ensure that cells maintain homeostasis by controlling inflammation, promoting survival pathways, and ensuring proper protein handling during periods of thermal stress [47,52,53,54]. The small network reported in Figure 2, derived from the larger protein–protein interaction network composed of genes retrieved in this study (Supplementary Figure S2), includes the key regulators of inflammatory responses, apoptosis, NF-κB signaling, and protease inhibition, all of which are crucial for cells to survive and adapt under heat stress conditions. Heat stress responses can be related to different molecular mechanisms, such as the inflammatory response or the activation of cell death pathways. The SOCS3 gene helps to regulate cytokine activity, preventing excessive inflammation [110], whereas proteins such as RIPK1 mediate cell death decisions to ensure the damaged cells are cleared without excessive necrosis, which would lead to inflammation [115]. On the other hand, molecular mechanisms related to survival pathways and protective action might be associated with heat stress responses [117]. NF-κB pathway members such as CHUK and COMMD8 activate survival pathways (inducing the expression of heat shock proteins and other protective molecules) [109], while SERPINB1 prevents protease-induced damage during stress response, making it a key player in cellular homeostasis under heat stress [57]. Conserved Helix Ubiquitous Kinase (CHUK) is part of the NF-κB signaling pathway, which plays a role in responding to stress stimuli; it has also been associated with mammary gland health in dairy sheep [90]. Heat stress can activate NF-κB signaling, leading to the expression of HSP and inflammatory mediators. This helps cells to cope with the accumulation of damaged proteins and other stress-induced injuries. In cattle, the NF-κB pathway was activated under heat stress, enhancing the expression of protective proteins such as HSP, which refold damaged proteins and prevent protein aggregation [125]. PELI proteins (PELI1/PELI2) exert functions in the immune signaling pathways, especially in the activation of toll-like receptors (TLR) and interleukin-1 (IL-1) pathways [126]. These pathways are important for the innate immune response, which is activated during heat stress. PELI proteins help to propagate signals that lead to the activation of NF-κB which, in turn, induces the expression of heat shock proteins and other stress-response genes [47]. 5. Conclusions Heat stress is a well-known problem on dairy farms, due to its negative impacts on milk production. The financial loss related to this stress in the American dairy industry has been estimated at USD 900 million per year (dairybusiness.com, accessed on 10 September 2024). Losses due to climate change and the associated increase in maximum temperatures are expected to be further exacerbated in the future. This problem is particularly acute in subtropical and hot regions where, most of the time, animals are raised in semi-extensive or extensive conditions; in the Mediterranean area, for example, ewes spend most of their time outside, especially in the hot season. In these environments, heat stress leads to decline in both the amount and quality of milk. Our study investigated climate-mediated selection signatures using genome-wide differentiation metrics computed between two groups of dairy sheep reared in the same climatic zone but in areas characterized by different maximum temperatures. The genes highlighted in this study have been previously associated with heat tolerance or adaptation traits in the existing literature. This study suggests the suitability of a simple analytical approach for retrieving positional candidate genes associated with heat stress and other adaptation traits, ultimately arguing for the viability of this model in studying climate adaptation in animals.