In many regions of the world, climate change is likely to lead to higher average temperatures, humidity, hotter daily maximum and more frequent heat waves, which could increase HS for livestock. The cattle population contributes to and is affected by the projected increase in global warming. As global demand for food products of animal origin is expected to rise by 70% by 2050, this demand must be met in a manner that has a negligible impact on the environment by improving the genetic quality of livestock using cutting-edge technologies and techniques (1). Temperature and humidity levels that exceed certain comfort zones lead to worsening the environmental conditions for dairy cattle and sub-tropical areas. Thus, dairy cows are more prone to environmental HS because of the rigorous selection for high milk yield combined with the high metabolic heat produced from the fermentation of additional dry matter during lactation (2, 3). As the high-producing dairy cows experience HS which negatively affects milk production, fertility, health and welfare traits (4–9), there is an urgent need to mitigate the effects of HS on dairy cows to improve the profitability and sustainability of the dairy production system.

The impact of HS in the dairy industry can be mitigated through a combination of different approaches including physical adjustments of the environment, enhanced nutrition, and management practices (10, 11). However, these alternatives are usually not cost-effective and tend to be unfeasible in the long term, especially in extensive or semi-intensive production systems (12). Additionally, these strategies might not work well in pasture-based dairy production systems where cows are exposed to solar radiation for a large part of their time while grazing (13). One strategy to permanently resolve the problem is genetic selection for improved heat tolerance combined with better management practices. It is more important to consider that the success of genetic selection of economically important traits depends on its heritability and additive genetic variation. However, physiological indicators of HS traits, including respiration rate (RR), rectal temperature (RT), and drooling score (DS) are low heritability traits (12, 14, 15) governed by a large number of quantitative trait loci (QTL), each with a small effect, so the possibility of rapid genetic progress through traditional genetic selection is low. Therefore, a better understanding of the genetic architecture of a trait, encompassing the underlying candidate causal variants with their corresponding effects and the trait variation that can be explained by genetic variation related to HS would be essential to derive genomic prediction and achieve rapid genetic progress for heat tolerance trait in dairy cattle. This prompts the need to identify genomic variants associated with heat tolerance phenotype, thus contributing an additional feature in the headway of genomic selection (GS) of heat-resistant dairy animals (16).

Recent developments in high-throughput sequencing techniques and emerging genomic data have opened up the opportunity to decipher the genomic variants affecting economically important traits in cattle. For a handful of economic traits, QTL mapping, transcriptome profiling and GWAS were conducted under HS conditions to gain better insight into the genetic basis for HS response in cattle. Several recent GWAS and transcriptome studies have been conducted to uncover candidate genes and causal variants associated with various indicators of heat tolerance traits in a wide range of dairy cattle breeds (11, 12, 14, 15, 17–24). In addition to GWAS, selection signatures can be used to associate candidate genes or variants subject to selection with traits of interest that can subsequently be used in GS (25). There are significant distinctive genetic traces or prints left behind in the genome that were subjected to selection called selection signature (26). Recently, genomic data provided additional possibilities to identify these footprints or genomic areas of selection linked to adaptation and productivity in cattle (27, 28). There has been an upsurge interest in identifying and exploring these genomic signatures of selection using high-density SNPs and revealed candidate genomic regions associated with environmental adaptation and thermotolerance in cattle (29–31). Since genome-wide significant single nucleotide polymorphism become affordable, GS has been tested and recognized by the dairy industry as a successful and easy-to-implement alternative (1). Nevertheless, a comprehensive review of the genetic basis of HS and milk production response in dairy cattle is scarce in the literature. Therefore, this review set out to undertake a comprehensive search for candidate genomic regions and/or genes and selection signatures associated with HS response in dairy cattle with the potential for simultaneous improvement of heat resistant and milk yield traits in dairy cows. Because of its importance for future breeding programs, the main mechanisms by which thermal stress affects essential immune system functions and influences animal health and welfare are also discussed. In addition, breeding strategies and recent approaches to mitigate HS in dairy cows were also discussed in this review.

The effects of HS on dairy cows are profound, and contribute significantly to lower overall milk production, fertility, and impaired health and welfare (4–8). In particular, high-yielding dairy cows are more susceptible to the stressful effects of HS (13). It is noteworthy that continuous selection for high milk yield in dairy cows, combined with the increased metabolic heat produced by the fermentation of extra dry matter during lactation (2, 3), compromises the maintenance of homeothermy under HS conditions. For instance, when lactating dairy cows are exposed to high ambient temperature and high relative humidity for prolonged periods, their ability to dissipate heat gained through metabolic process and from the environment is reduced, making them susceptible to HS (32, 33). To reduce their heat load, the cow reduces her feed intake and consequently milk production (34). Despite the fact that lactating cows are more susceptible to heat stress, HS exposure during the dry period affects subsequent lactation and has long-lasting effects on the progeny (35, 36). In order to assess the effects of HS in lactating and dry cows, particularly the simultaneous effect of temperature and relative humidity, the temperature-humidity index (THI) is widely used (34). On the other hand, changes in production traits, especially milk yield, have often been used to determine the THI threshold for HS (6). It was worth mentioning that, increasing THI linearly decreases milk production in lactating dairy cows (6, 37, 38). In terms of losses, HS causes production losses of 600–900 kg of milk per cow/lactation (39). Routine selection of dairy cows for higher milk yield may result in poor response of dairy cows to HS, due to the antagonistic genetic relationship between production level and specific ability to respond to HS (11, 37, 40). This clearly indicates the urgent need to develop sustainable strategies to reduce the detrimental effects of HS on milk yield in dairy cattle.

Although HS affects milk yield in dairy cows, more attention should be paid to how HS affects milk composition. For example, previous studies reported a significant correlation between increasing somatic cell count (SCC) in milk and increasing THI in dairy cattle (41–43). In the context of HS, consideration of the environmental sensitivity of fatty acids in milk for genome analysis and gene annotation is a novel approach, where HS exacerbates changes in milk fatty acid profiles including stearic acid (C18:0), polyunsaturated fatty acids (PUFA), saturated fatty acids (SFA) and unsaturated fatty acids (UFA) during early lactation in high-yielding cows (18, 44). It is argued that exposure to HS in a high THI environment appears to inhibit milk fat and protein synthesis in lactating dairy cows, which in turn directly affects the synthesis of energy-corrected milk yields (45). More fundamentally, heat exposure appears to impair the capacity of the mammary gland to synthesize milk proteins by down-regulating the expression of key milk protein genes such as β-casein and butyrophilin (46). Additionally, HS negatively affects the synthesis of milk proteins by decreasing the transcription of metabolism-related genes and increasing inflammation-related genes (47). Heat stress could also alter the triacylglycerol profile of milk, which is characterized by a decrease in triacylglycerol groups with predominantly short-chain FAs to medium-chain FAs and a concomitant increase in groups with predominantly long-chain FAs (48). In addition, the authors demonstrated that HS significantly decreases the content of some polar lipid classes, especially lysophosphatidylcholine, which appears to be a lipid marker of HS in dairy cows.

Genetic factors in cattle have been recognized as a main feature that can explain part of the differences in response to HS in dairy cows. For example, Otto et al. (15) performed a GWAS in Gir × Holstein F2 experimental population to identify genetic markers responsible for genetic variation in response to HS and used the breed of origin of alleles (BOA) approach to evaluate the origin of marker alleles of candidate genes. The authors found that most animals that responded better to the effects of HS had 2 alleles of the Holstein breed, while a high proportion of heat-stressed animals had 2 alleles of the Gir breed. This indicates Holstein breed alleles could be related to a more complex response to HS effects, which could be explained by the fact that Holstein animals are more affected by HS than Gir animals, and thus having more complex genetic mechanisms to defend the body from the harmful effects of HS. Therefore, revealing the origin of marker alleles of candidate genes for heat tolerance in dairy cattle populations can help in understanding the genetic variation of the trait, which is subsequently used to estimate breed-specific SNP effects to improve genomic prediction for heat tolerance and production traits in dairy cattle.

Most economically important traits in farm animals are associated with polygenes located at QTLs that are widely distributed throughout the genome. With the availability of high-throughput sequencing technologies, detection of a number of high-impact QTL in dairy is possible, suggesting genomic regions and genes responsible for significant differences in the yield and composition under HS conditions. In addition, a wide range of genomic approaches have been employed to identify genetic variants linked to physiological indicators of response to HS. However, the heat tolerance trait is highly polygenic and influenced by multiple variants, each with small effects on the phenotype, suggesting that the trait may be more suitable for genomic selection tools such as those currently used in the Australian dairy industry (49, 50), than for methods that exploiting few QTLs with major effects. In QTLs, it is difficult to find a few markers that explain a large proportion of genetic variance, but when found, they can revolutionize the selection process of dairy cows tolerant to HS. For example, the slick-hair gene is clearly linked to heat tolerance in dairy cows as a result of a single gene effect (51), i.e., improved ability to dissipate heat (52). Moreover, the integration of omics information could help to reveal nodes of the HS control network and eventually find a panel of markers that can be used in the selection of heat-tolerant dairy animals with higher productivity. Overall, heat tolerance in dairy cows is a complicated phenomenon that calls for the combination of phenotypes and omics information to provide accurate tools for selective breeding without compromising productivity.

Since HS suppresses the immune and endocrine system, exacerbate health and metabolic disorders (53, 54) and ultimately affect the performance of dairy cattle, understanding the underlying mechanism by which HS modulates the immune response of cows is crucial and must be considered in future breeding strategies. It is important to remember that, Lengi et al. (55) observed significantly lower concentrations of granulocytes in milk from heat-stressed dairy cows. Neutrophils, the most common type of granulocyte found in SCCs, are recruited by chemo-attractants to infection sites such as interleukin-8 (IL-8), including the mammary gland, where they phagocytose and destroy pathogens (56). During HS, the mammary gland’s ability to respond to infection is compromised, as evidenced by a decline in the concentration of viable granulocytes there. Heat stress possesses a significant negative impact on both arms of the immune system, the innate and the acquired immunity. Exposure to HS causes a variety of changes in the physiology of the body, including activation of the hypothalamic–pituitary–adrenal (HPA) axis and secretion of glucocorticoids (7, 57). Immune cells have glucocorticoid receptors which allow glucocorticoids to stimulate immune response during acute stress, however, the functions of the immune system get impaired when the exposure to stress is elongated (chronic stress) (58). The major immune cells that contribute to innate immunity are phagocytes (i.e., neutrophils, monocytes, macrophages), cytokines and inflammatory mediators-producing cells (i.e., neutrophils, macrophages, mast cells, natural killer cells). The acquired (adaptive) immune response is composed of T cells and B cells that contribute to cell-mediated immunity and humoral immunity, respectively (59, 60). Although the innate and acquired immune cells are distinct from each other, their interplay is evident through their specific roles in initiating and regulating immune responses (Figure 1).

Of the cells of the innate immune system, neutrophils are the first line of defense, and they kill and eliminate invading microorganisms by phagocytosis, degranulation, and release of neutrophil extracellular traps (61). Heat stress impairs the potential of bovine neutrophils for phagocytosis and reactive oxygen species (ROS) production under in vitro and in vivo conditions (41, 53, 62). This indicates that HS impaired the pathogen recognition ability of neutrophils and suppressed inflammatory immune response to pathogens. Mononuclear phagocytes, including blood monocytes and their derived tissue macrophages, are another important player in the innate defense against invading pathogens. They are divided into functional subsets based on their roles during the inflammatory response. Classically activated M1 monocytes/macrophages are pro-inflammatory cells with a higher potential for antimicrobial functions, while alternatively activated M2 monocytes/macrophages have an anti-inflammatory phenotype with a special role during the resolution of inflammation (63, 64). Catozzi et al. (65) exposed bovine monocytes to high temperatures and investigated the expression profile of genes responsible for monocyte/macrophage polarization. They reported a polarization of monocytes from a classical M1 to a non-classical M2 phenotype which is associated with suppressed innate immune response and could favor the humoral immune response. An M2-like macrophage phenotype, which was identified by the high expression of the M2 cell marker CD163, was also found in the intestinal lymphoid tissue of dairy cows under HS (66). Antigen presentation is an important immune process to translate information obtained by antigen-presenting cells (APCs) into specific actions of the desired T-cell subsets. Although there is scanty information on the effect of thermal stress on antigen processing and presentation in bovines, several studies on humans and mice reported that heat shock impairs the presentation of exogenous antigens by MHC II and suppresses co-stimulatory functions in APCs such as macrophages, dendritic cells, or B cells (67).

The desired cellular or humoral immune responses are triggered based on a repertoire of pro and anti-inflammatory cytokines mainly produced by CD4+ T helper cells (68). Polarized T helper 1 cells (Th1) activate cellular immunity and enhance antimicrobial responses against intracellular pathogens by secreting type 1 pro-inflammatory cytokines such as tumor necrosis factor-α (TNF-α), interferon-γ (IFN-γ) and interleukin-12 (IL-12). Polarized T helper cells (Th2), on the other hand, secrete type 2 cytokines (like IL-4, IL-5, IL-10 and TGF-β) that play a regulatory role during the immune response (IL-10 and TGF- TGF-β) favorite allergic reactions (IL-4 and IL-5), and stimulate B cells for the production of special antibody isotypes (IL-4 and IL-13) to fight extracellular bacteria and helminths (69). Heat stress stimulates the secretion of stress hormones, the expression of heat shock proteins (HSPs) and alters the cytokine profile, which in turn alters the Th1:Th2 balance. Several studies have shown that thermal stress suppresses the secretion of Th1 cytokines (cellular immunity) and stimulates the secretion of Th2 cytokines (humoral immunity), which impairs cell-mediated immunity, decreases lymphocyte proliferation, and increases the susceptibility of heat-stressed animals to various infectious diseases (7, 54). However, the switch to Th2 cytokines does not ensure an adequate humoral immune response (Figure 1). On the contrary, IgG response to an innocuous antigen and to booster immunizations was significantly lower in heat-stressed compared to cooled cows during the dry period and after parturition (53). This is also confirmed by the reduced number of CD21-positive B cells in cows exposed to HS compared to normal cows (70). In addition, Hu et al. (71) reported that chronic HS significantly impairs the immune response to foot and mouth disease (FMD) vaccination by decreasing IgG2a levels and suppressing cytotoxic T-cell response, T-cell proliferation and IFN-γ expression in both CD4+ and CD8+ cells.

Regulatory T cells (Tregs) secrete anti-inflammatory cytokines such as TGF-β and IL-10, which maintain host immune homeostasis by balancing Th1:Th2 activity and are characterized by the expression of the IL-2 receptor CD25 and the transcription factor forkhead box P3 (FOXP3) (72). Although the effects of HS on Tregs in cattle are unknown, studies in mice have shown that the number of Tregs decreases and their immunosuppressive capacity is impaired in response to HS exposure (73, 74). Therefore, maintaining the optimal function of Tregs to balance the activity of inflammatory cytokines and the Th1:Th2 ratio is essential for maintaining homeostasis and preventing inflammatory diseases in cattle under HS conditions. The mechanism by which HS modulates the immune response and renders dairy cows susceptible to invading pathogens explains the higher incidence of mastitis, metritis, respiratory disease, displaced abomasum, retained foetal membranes and lameness during HS (5). Of interest, Alhussien and Dang (41) have reported genetic variability in resilience to HS in Indian native breeds of tropical environments. The mammary defense mechanism and milk production were less affected in Gir and Tharparkar cows as compared to Sahiwal cows under HS conditions (41). This highlights that there is variation in heat resilience among Zebu breeds indigenous to warm climates that should be considered in breeding programs. Especially during the late-gestation and prepartum period, HS of the dairy cow not only impaired their post-partial immune function resulting in higher susceptibility to uterine infections (75) but also carried over negative effects on the passive transfer of colostral IgG to the newborn dairy calves and compromised their immune function (76).

It is evident that substantial genetic variation that explains part of the variation in response to heat tolerance exists among animals (52, 77). One way to determine the nature and extent of genetic variation for heat tolerance is to genetically assess HS indicator traits in dairy cattle. Numerous research on genetic evaluation of heat tolerance in dairy cattle has been based on analysis of performances under HS conditions. With substantial advancement in the area of quantitative genetics, phenomics and genomics, studies on genetic and genomic evaluation expanded to a wide range of HS indicator traits that optimally includes several other indicators of thermoregulation targeted for heat tolerance improvement in dairy cattle breeds (14, 18, 44, 50). Besides indicator traits for HS, progress has also been made in the development of resilience indicators traits for dairy cattle breeding (78–81), which encompasses multiple traits and could be considered in the breeding goals to accelerate genetic progress and enhance farm profitability (2).

In an attempt to evaluate the genetic effect of HS and genetic tolerance for milk production and its components in dairy cattle, genetic parameters were estimated for several indicators of HS phenotype and milk production response as well as resilience indicators traits in dairy cattle (12, 50, 79, 80, 82–86). Alternatively, several HS indicators, for example, RT, RR, and DS are common physiological indicators used to identify heat-tolerant animals within a breed (13, 87). Another way to assess heat tolerance is to monitor changes in milk production traits under warm environmental conditions (88–91). Therefore, these HS indicator traits are often considered when aiming to genetically evaluate HS response in animals. In this context, the heritability estimates for RT, RR, and DS are low, with heritability estimates ranging from 0.03 (DS) to 0.06 (RT) in the Holstein-Friesian cattle population (14). However, Dikmen et al. (87) reported a higher estimate of heritability (0.17) for RT in the Holstein cattle population. Considering an average THI threshold of 62 for milk production traits, heritability estimates increased with increased THI above its threshold of 62 for milk yield (0.20–0.23) and protein yield (0.14–0.16) and remained firm for fat yield (0.17) in Holstein cattle (83). With a THI of 76, Sungkhapreecha et al. (89) reported heritability estimates varied from moderate to low, with values of 0.344 for milk yield, 0.087 for somatic cell score and 0.061 for milk fat to protein ration in tropical dairy cattle breeds. A reduction in additive genetic variance in milk yield was observed with increasing THI levels in purebred Zebu cattle (90). For milk fatty acid profiles, heritability estimates were higher for UFA (UFA, MUFA, and PUFA), while lower for SFA traits (SFA, C16:0, and C18:0) under the THI of 10 degrees over its threshold of 68 than an average THI threshold of 68 in Brazilian Holstein cattle (82). Similar findings by Hammami et al. (91) observed higher heritability estimates for PUFA and C18:1 at high THI levels in Belgium Holstein cows. In addition, greater additive genetic variances were estimated for C18:0, PUFA, and UFA uder HS conditions compared to thermo- neutral conditions during the first lactation period, reflecting environmental sensitivity during early lactation in high-yielding Holstein dairy cows (18). This genetic variability for milk fatty acid profiles in relation to THI impedes motivation to consider these traits in the breeding objective for genetic selection to leverage heat tolerance in dairy cattle, thus, response to selection would be expected (Figure 2). Regardless of the effects of the lactation stage, Bohlouli et al. (44) estimated the greatest additive genetic variances for C18:0, MUFA, PUFA, and UFA at high THI than under temperate climate conditions, whereas fat yield, palmitic acid (C16:0), and SFA decreased with increasing THI. Nevertheless, Bohlouli et al. (44) identified C18:0, MUFA, PUFA, and UFA as climate-sensitive traits due to large genetic variance at the extreme ends of the THI scale, indicating the future possibility of improving thermotolerance in dairy cattle through genetic selection. Most importantly, larger genetic variation was observed for C18:1, indicating the greatest sensitivity to HS conditions in a tropical climate (91). With respect to genetic trends, the genetic components of HS have had negative effects on milk production and quality traits of dairy cattle over the years (37, 90). Despite the low heritability estimates for physiological indicator traits, incorporation of these traits in the genetic evaluation programs and consider these traits in the selection index helps to achieve optimum genetic progress for heat tolerance, while maintaining milk production traits in dairy cattle.

The ability of an animal to respond to HS can be assessed by physiological indicators of HS traits, such as RT, RR, and DS, which increase when the animal is exposed to a warm environment (13, 87). Another way to assess heat tolerance is to monitor changes in milk production traits or analyze milk yield trait alternations by HS (45, 88). These physiological indicators of HS traits shown to be heritable (12, 14, 15, 49, 87), indicate that genetic gain can be made through selection. Notably, the GxE work by Cheruiyot et al. (77) reported the existence of substantial genetic variation for heat tolerance in dairy cattle, which can provide a better insight into the genetic selection of thermo-tolerant dairy cows. However, heat tolerance is a complex trait, governed by a myriad of adaptative responses (behavioral, physiological, cellular, etc.) that tend to be polygenic, and as a result, many genetic variants, each with small effects contribute to the phenotype. In these cases, the heat tolerance trait has been associated with a number of QTL regions and candidate genes (14, 15, 18, 92). This polygenicity of a trait can be a challenge in achieving rapid genetic progress for heat tolerance in dairy cattle, particularly in cases when the heritability of a trait is low. To avoid these problems, genome-wide DNA markers should be used to capture the effects of many genetic variants. In this regard, the application of genomic selection is ideal and is increasingly used to accelerate genetic progress, by increasing the accuracy of selection and shorting generation intervals.

Identifying genome-wide markers and pathways underpinning the genetic mechanisms that increase thermotolerance in dairy cattle is at the infancy stage, but it is gaining momentum considering the increasing challenges imposed by climate change, as shown by the recent increase in published studies (2, 11, 12, 14, 18, 92–96). Genome-wide association studies are most commonly used to test tens if not hundreds of thousands of genomic variants to find those statistically associated with heat tolerance traits in cattle. Although GWAS have led to the identification of trait-associated genomic variants, the researchers have conducted validation of GWAS to prioritize the variants most likely to be causal to thermo-tolerance and to provide several biological insights that can be leveraged to minimize HS while maintaining milk production in dairy cattle using a large dataset (2, 12, 14, 93). As such, biological validation of candidate genes for RT as HS indicator in Holstein cattle observed that QTL regions on BTA3, 4, 8, 13, 14, 17 and 29 are perhaps the most significant regional association signals with heat tolerance in dairy cattle (12). A similar study by Luo et al. (14) revealed that the genomic regions on BTA3, 6, 8, 12, 14, 21 and 24 were significantly associated with HS response in Holstein dairy cows. Several other previous studies confirmed the significant effects of QTL regions on BTA6 and 17 for RT during HS in Holstein cattle (15, 16). Luo et al. (94) reported the presence of a genomic region on BTA14 strongly associated with thermotolerance in Australian Holstein cattle as part of a conditional GWAS strategy. Of interest, Sigdel et al. (11) performed whole-genome mapping to identify candidate genes associated with milk production traits under HS conditions in US Holstein cows and demonstrated a significant association between BTA5, 14, and 15 and milk production. Interestingly, Bohlouli et al. (18) claimed that the highest number of significant SNP markers on BTA14 were significantly associated with HS response of milk fatty acids profiles (e.g., PUFA) in the first and third lactation stages under thermoneutral and HS conditions in dairy cows. This provided stronger evidence, suggesting that QTL effects in BTA14 altered the concurrent improvement of milk production and heat tolerance traits in dairy cows. Previously, Dikmen et al. (24) identified QTL regions on BTA4, 5, 16, 24, and 26 correlated with rectal temperature for heat-stressed Holstein animals using the one-step genomic BLUP approach. Similarly, regions significantly affecting RT in lactating Holstein cows were detected in BTA4, 6, and 24 (16). In addition, the region on BTA22 (95) and BTA26 (96) was reported to have a significant association with heat tolerance in dairy cows. According to the results of the literature search, the candidate genomic regions associated with heat tolerance traits in dairy cows are located on BTA3, 5, 14, and 16, respectively. Despite the fact that BTA14 carries important genes for traits of economic importance, the relationship between BTA14 and milk-related traits is well described in the literature. For example, the significant SNP markers on BTA14 located within or near the potential candidate genes DGAT1, PPP1R16A, FOXH1, and PLEC had the greatest effects on fatty acid across lactations under different conditions of HS in dairy cows (18). Therefore, the specific candidate genes on BTA3, BTA5, and BTA14 may hold promise for simultaneously improving heat tolerance and milk-related traits in dairy cattle. Overall, these QTL regions shed new light on candidate genes that are potentially associated with heat tolerance trait architecture, while maintaining milk production traits in dairy cattle. In summary, with the emerging genomic data, a growing list of candidate genes is being uncovered for genetic improvement to enhance aspects of heat tolerance, while maintaining milk production traits in dairy cattle (Table 1).

The identification of key candidate genes responsible for variation in thermoregulation could inevitably help to improve the efficiency of selective breeding, especially for traits with low heritability in dairy cattle breeding programs. In addition, investigation of the physiological systems regulated by genes involved in thermoregulation is an area that requires further study, as understanding the biological control will aid to prioritize candidate genes for genomic selection strategies (105). A considerable number of GWAS and transcriptome studies have been investigated and identified several candidate causal variants potentially linked with milk production and various physiological indicators of HS in dairy cattle (11, 12, 14, 15, 17–19, 24, 92–94, 96, 106, 107). Recent work has shown that the SNPs rs8193046, rs43410971, and rs382039214, within the TLR4, GRM8, and SMAD3 genes, respectively, appear to be significantly associated with 305-day milk yield, RT and RR in Holstein dairy cows kept under heat-stress condition (94). Candidate genes SPAG17, FAM107B, TSNARE1, RALYL and PHRF1 identified in Holstein cattle have been reported to be associated with RT, with FAM107B and PHRF1 genes validated by functional analysis based on gene expression during thermal stress in a peripheral blood mononuclear cell model (12). The previous study by Luo et al. (12) evaluated physiological indicators of HS in Holstein dairy cattle and reported that TSNARE1, perhaps the potential gene, is significantly associated with RT. Luo et al. (14) also discovered seven major candidate genes (PMAIP1, SBK1, TMEM33, GATB, CHORDC1, RTN4IP1 and BTBD7) associated with physiological indicators of HS (RT, RS and DS) in Holstein cattle by weighted single-step genome-wide analysis studies (WssGWAS). In the transcriptome study by Czech et al. (106), the RAB39B gene was found to be significantly associated with RT, DS and RS under HS conditions in dairy cattle. In addition, Diaz et al. (107) reported five genes: E2F8, GATAD2B, BHLHE41, FBXO44, and RAB39B which were significantly associated with HS in cattle. Dikmen et al. (16) reported two candidate genes (ATPA1A and HSP70A) associated with RT and RR during HS in lactating Holsteins. Previously, Dikmen et al. (24) also carried out GWAS for RT of lactating Holstein cows under HS conditions to find potential SNPs that could function as QTL for the same trait. They reported QTL markers that either contained or were in close proximity to specific functional genes such as U1 spliceosomal RNA, NCAD, SNORA19, RFWD2, SCARNA3, SLCO1C1, PDE3A, KBTBD2, LSM5 and GOT1. Garner et al. (22) have already revealed BDKRB1 and SNORA19 as potential candidate genes associated with HS in dairy cows. A whole-genome association mapping study by Sigdel et al. (11) revealed that candidate genes HSF1, MAPK8IP1, and CDKN1B are directly involved in several cellular responses to HS in lactating dairy cows, for example activation of HSP (PEX16, HSF1, EEF1D, and VPS28), reduction of oxidative stress, modulation of apoptosis process (MAPK81P1, CREB3L1), DNA maintenance (TONSL), and thermotolerance (CRY2). The UCN3 gene that is engaged in the genetic regulation of stress tolerance and oxidative stress in Holstein Friesian was described by (108).

A handful of studies have highlighted the impact of HS on the fatty acid profiles of milk from dairy cattle and identified genes potentially associated with response to HS. Recently, Bohlouli et al. (18) identified the candidate genes, AMFR for HS response of PUFA, ADGRB1, DENND3, DUSP16, EFR3A, EMP1, ENSBTAG00000003838, EPS8, MGP, PIK3C2G, STYK1, TMEM71, GSG1, SMARCE1, CCDC57, and FASN for HS response of SFA, ENSBTAG00000048091, PAEP, and EPPK1 for HS response of UFA in dairy cattle and thus reported as a potential biomarker for heat tolerant animals. In another study looking at the candidate genes for coat color phenotypes of HS in cattle, Bahbahani et al. (98) indicated that PMEL is a strong candidate gene associated with eumelanin synthesis and thus could control coat color in cattle. In addition, the ERBB3 and MYO1A genes have been proposed as the most likely candidate genes for coat color phenotypes in locally adapted tropical cattle breeds (31). It is indicated that the differences between Mongolian cattle (Bos-taurus) and Minnan cattle (Bos-indicus) resulting from DVL2 mutations could alter the normal transcription and expression of the DVL2 gene and also affect hair growth, allowing it to acclimatize hot-climate of southern China and cold climate of northern China (109).

The changes in cows’ response to HS can also be detected by the expression or activation of specific biological markers, particularly heat shock transcription factor (HSF), heat shock proteins (HSP70, HSP90 and HSP27), TLR2/4 and IL2/6 (110). In particular, an evolutionarily conserved transcription factor known as heat shock transcription factor one (HSF1) binds the promoter regions of HSPs to control their stress-inducible synthesis in response to the environment (40). Heat shock proteins (HSPs) are highly conserved stress proteins that have been proposed as promising biomarkers of HS, of which HSP70 has been shown to be upregulated in dairy cows’ mammary epithelial cells during HS to increase mammary gland thermotolerance (22). According to Liu et al. (111), heat-tolerant Chinese Holstein dairy cows had significantly higher plasma HSP (HSP70 and HSP90) and cortisol levels compared with non-heat-tolerant dairy cows. Potential candidate genes such as HSPA1B, HSPA1L, HSPA12A, GRXCR1, FKBP4, and IL-6 were reported to have a correlation with thermotolerance in tropically adapted dairy cattle breeds (102). The HSPA9 and HSPB9 genes, as well as two other members of the DNAJ family (DNAJC8 and DNAJC18), have been linked to HS (97). Interestingly, highly significant causal mutation candidates in the HSF1gene for heat tolerance have been identified in Holstein cattle in Australia (23), and the United States (11), including MGST1 (51). Transcriptome analysis by Liu et al. (111) identified the OAS2, MX2, IFIT5, and TGFB2 genes associated with heat tolerance and involved in the immune effector process in dairy cows, with the TGFB2 gene being part of the MAPK signaling pathway.

Specific biological pathways associated with nervous systems functions (interaction between neuroactive ligands and receptors, and glutamatergic synapses) and metabolism (citrate or Krebs cycle) are important factors in elucidating the mechanisms of heat tolerance in dairy cattle, warranting extensive follow-up of functional investigations (2). Cheruiyot et al. (93) attempted to conduct a GWAS study using several approaches, including conditional analyses for heat resistance employing a large sample size and genotype data set collected from dairy cows. They showed that the ITPR1, ITPR2, and GRIA4 genes are related to the neuronal system and the NPFFR2, CALCR and GHR genes are related to neuroactive ligand-receptor interaction functions, which may be important for metabolic homeostasis in lactating dairy cows during thermal stress, with GHR, NPFFR2 and CALCR providing the strongest evidence. In addition, Otto et al. (15) reported that the putative candidate genes, including LIF, OSM, TXNRD2, and DGCR8, undergo changes in biological processes in response to the effects of HS in crossbred dairy cows. On the contrary, the putative candidate genes KIFC2, VPS13B, USP3, and SCD have been reported to be involved in heat tolerance in lactating dairy cows, with the SCD gene encoding a fatty acid metabolism enzyme and possibly required for metabolic homeostasis during HS in mammals (93).

Further, to see the specific biological functions, the lists of candidate genes for heat tolerance across different studies were subjected to GO functional annotation and KEGG pathway enrichment analysis. PANTHER¹ and DVID² data bases with cut-off point p < 0.05 and Benjamini-Hochberg false discovery rate (FDR) < 0.05 were used. In terms of biological processes, genes related to HS response as well as biological regulation, metabolic processes, or immune responses were found to be the most represented (Figure 3A), consistent with the proposed candidate genes and their ontology (40). In terms of molecular function, genes for catalytic activity and binding were the most represented (Figure 3B). Interestingly, estrogen signaling pathway is the top enriched pathway (p < 0.001), comprising of 10 candidate genes (FKBP4, HSPA1L, ITPR1, ITPR2, KRT24, KRT25, KRT26, KRT27, KRT28, PFR) involved in HS response in dairy cattle (Figure 3C). A total of 4 genes were enriched with legionellosis pathway (HSPA1L, HSF1, 1 L6, TLR4). The renin secretion pathway was enriched with 4 candidate genes (AQP1, ITPR1, ITPR2, PDE3A). It was also observed that the enrichment of genes in the neuroactive ligand-receptor interaction pathway (CALCR, GRIA4, GRM8, GHR, MC5R, NPFFR2, PRLH, UCN3), PI3K-Akt signaling pathway (CDKN1B, GHR, IL6, MTOR, OSM, SGK1, TLR4), HIF-1 signaling pathway (CDKN1B, IL6, MTOR and TLR4) and JAK–STAT signaling pathway (LIF, GHR, IL6, MTOR and OSM) with other pathways and stress responses (Figure 3C). In line with these pathways (Figure 3C), the previous studies found that the candidate genes for HS response enriched HSF1mediated heat shock response, estrogen signaling pathway, HIF-1 signaling pathway, PI3K-Akt signaling pathway, MAPK singling pathway and immune response pathways in dairy cattle (93, 112, 113).

Selection pressure leaves distinct footprints in the areas of the genome that are exposed to selection called selection signature (25, 26, 114). When a derived allele or variant with fitness progress occurs in a population with positive selection, this results in neighboring related alleles being carried further alongside the selected variant (25). This indicates positions that are close to regions under selection are mostly affected by background selection and genetic hitchhiking (115). Thus, selection signature detection is used to identify these footprints/signs of selection (116), by which candidate genes related to a particular phenotype can be targeted. Understanding how selection acts on livestock genomes and detection of selection signature provides a better insight into the progress of artificial selection, which may especially benefit the optimization of breeding programs to improve animal resilience and other traits of economic importance (20, 25, 117). With the recent development and prevalent of high-throughput sequencing technologies along with various powerful statistical methods, there is an increased interest in genome-wide detection of selection signature, providing insights on the mechanisms of natural and/or artificial selection and uncovering candidate genomic regions under selection related to adaptation and climate resilience in cattle (20, 29–31, 100, 101, 118).

In cattle, genome-wide SNPs analysis has identified several candidate genome regions on BTA5, 6, 8, 23, and 26 under positive selection signature, containing interesting genes associated with heat tolerance in tropically adapted dairy cattle breeds (102). Moreover, the candidate region on BTA21 contains genes associated with heat tolerance suggested to be under selection in Shanghai Holstein cattle (100). A previous study focusing on the genomic signatures of divergent selection using high-density SNPs in taurine and Zebu cattle populations, already identified candidate genes associated with thermotolerance, including HSP70, HSF1, CMPK1, NPM1, and GCN2 in Zebu cattle (31). Moreover, Freitas et al. (20) identified a positively selected candidate gene (PLA2G2A) underlying lipid metabolism in cattle in warm environments. Several other candidate variants and genes that may be responsible for differences in heat tolerance in particular environments and production systems have been discovered in potentially selected regions, including ITGA9, ACAT2, and PLAC8 in dairy cattle (93). Saravanan et al. (102) documented genomic regions encoding candidate genes HSPA1L, HSPA1B, DNAJB4, GRXCR1, OLA1, SP9, and HSPA12A as selection signatures regulating heat tolerance in indigenous dairy cattle breeds. In addition, Liu et al. (100) identified a number of potentially and positively selected novel genes, such as NDUFB3, RGS3, UBD, DIS3L2, NRXN2, PEX14, SPTLC2, AQP1, and PTPN9, that are associated with adaptation to tropical humidity and harsh environmental conditions in Holstein cattle. Moreover, a previous finding in African cattle reported candidate genes SLC9A4, PLCB1, FTO, ITPR2 (101), and SOD1 (95) that have responded strongly to selection for heat tolerance. In addition, the HSPA4 gene involved in protecting cells from heat damage and preventing protein denaturation showed selection success for heat tolerance in several cattle breeds in Africa (95) and China (109). Candidate genes studied, such as AQP5, RAD50, and RETREG1, have been shown to regulate acclimatization to heat in Russian cattle breeds (118).

Due to the increasing frequency and intensity of changes in temperature worldwide, identifying and developing dairy cattle resistant to HS is critical for breeding dairy herds that are better suited to future climatic challenges. In addition to known strategies to reduce the detrimental effects of HS, such as physical environmental modifications and improved feeding practices, the need for genetic progress, which includes selection (genetic and genomic) for heat tolerance with high milk yield is a current state of genetic research that could lead to a long-term solution to the problem (Figure 4). Also, these strategies might not be effective in pasture-based dairy production systems where dairy cows are exposed to solar radiation for much of their time while grazing (13). Several research attempts have been made to find breeding solutions for HS, which is already a feature of dairy cattle breeding programs in many regions. Breeding strategies for dairy cows through the genetic development of heat-tolerant breeds from genetic and genomic perspectives, targeted genome editing, and other options including epigenetic modifications are discussed in this section.

In this review, we provide a comprehensive overview of the genomic regions and candidate genes associated with heat tolerance that can expand our knowledge to the genetic mechanisms of HS, which could open new avenues for mitigating the impacts of global warming on dairy cows. Many dairy farmers use a variety of strategies to keep their cows cool, such as shade, ventilation, cooling with water, drinking water, cooling in the barn, fans, sprinklers, etc. However, these strategies are quite costly to be of practical use in extensive dairy production systems under smallholder conditions, especially in hot, less developed and agriculture-dependent countries. As the effect of global warming intensifies, even locally adapted dairy breeds could be affected, and future dairy farmers will suffer the greatest economic losses. This will cause farmers to switch from dairy production to other sectors, such as beef production, as dairy cows are more affected by HS than beef cattle. Therefore, breeding dairy cows with greater heat tolerance ability, perhaps through genetic means, and the associated support in these vulnerable regions for food security must be addressed. While identifying heat-tolerant dairy cattle is challenging due to the complex phenomenon of HS and the antagonism between heat tolerance and production traits, the use of genome-wide information could be a way to unravel the genetic mechanisms of heat tolerance in dairy cows, that can be accustomed to the selection programs. Progress has been made in using genomic information to identify candidate genes and causative variants associated with heat tolerance and milk production traits in dairy cattle, that can be used for marker-assisted selection, genomic selection and gene editing programs. The literature shows that heat tolerance is a complex trait influenced by many genes in the genome, with the specific genes HSF1, SPAG17, DNAJC8, HSPA9 and DNAJC18 appears to have been reported in more than one independent studies. Further studies of these genes could help to understand the genetic mechanisms of adaptation to HS, which may help to simultaneously improve heat tolerance and production traits in dairy cattle. In addition, optimal breeding strategies for the genetic development of heat-tolerant dairy cows provide long-term solutions to HS effects, that are essential for addressing the dual challenge of increasing dairy production to feed the increasing human population, while addressing the impacts of global warming. Overall, this review could serve as a valuable resource material for the dairy breeding industry aimed at increasing heat tolerance, while maintaining milk production traits.

DW conceived, conducted the literature search, performed visualization, and wrote the first draft. JH, GM, and BS provided critical appraisal and suggestions. MA involved in conceptualization, writing – review and editing. All authors contributed to the evaluation, editing, and approval of the final version of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.