Candidate signatures of positive selection for environmental adaptation in indigenous African cattle: A review

Environmental adaptation traits of indigenous African cattle are increasingly being investigated to respond to the need for sustainable livestock production in the context of unpredictable climatic changes. Several studies have highlighted genomic regions under positive selection probably associated with adaptation to environmental challenges (e.g. heat stress, trypanosomiasis, tick and tick-borne diseases). However, little attention has focused on pinpointing the candidate causative variant(s) controlling the traits. This review compiled information from 22 studies on signatures of positive selection in indigenous African cattle breeds to identify regions under positive selection. We highlight some key candidate genome regions and genes of relevance to the challenges of living in extreme environments (high temperature, high altitude, high infectious disease prevalence). They include candidate genes involved in biological pathways relating to innate and adaptive immunity (e.g. BoLAs , SPAG11 , IL1RL2 and GFI1B ), heat stress (e.g. HSPs , SOD1 and PRLH ) and hypoxia responses (e.g. BDNF and INPP4A ). Notably, the highest numbers of candidate regions are found on BTA3, BTA5 and BTA7. They overlap with genes playing roles in several biological functions and pathways. These include but are not limited to growth and feed intake, cell stability, protein stability and sweat gland development. This review may further guide targeted genome studies aiming to assess the importance of candidate causative mutations, within regulatory and protein-coding genome regions, to further understand the biological mechanisms underlying African cattle's unique adaption.

Following the development of agricultural societies, human migrations and trading networks, taurine and zebu cattle dispersed worldwide, as evidenced by zooarchaeological, iconographic and molecular genetic studies (Edwards et al., 2007;Kantanen et al., 2009;Pérez-Pardal et al., 2010).These global dispersals brought taurine and zebu cattle into contact with related bovid species such as yak (Bos grunniens), gaur (Bos gaurus) and banteng (Bos javanicus), with several episodes of hybridisation and introgression now documented (e.g.Wu et al., 2018).Also, introgressions from local aurochs probably contributed to the genetic ancestry of modern cattle, as has recently been shown for taurine cattle in the Levant (e.g.Verdugo et al., 2019) and North Africa (Ginja et al., 2023).All these evolutionary events shaped today's lineage diversity of worldwide domestic cattle.
The domestication and dispersion processes of cattle led to anthropological and natural selection pressures, which followed the development of farmers and pastoral societies worldwide (Diamond, 2002;Flint & Woolliams, 2008).These events are the roots of environmental adaptation, production variation and phenotypic diversity of modern cattle (e.g.Felius et al., 2014).Likewise, these selection events imprinted the genome's adaptive diversity and its architecture (e.g.Kim et al., 2020).The phenotypic and genetic diversity in present-day worldwide cattle breeds represents a unique genetic resource for successful and sustainable breeding improvement programmes.
With its diverse environmental conditions and its complex history of biological and agricultural introductions, the African continent is considered one of the main reservoirs of cattle diversity (Kim et al., 2020;Kim, Hanotte, et al., 2017;Mwai et al., 2015).Indigenous African cattle genomes have now been extensively studied by researchers aiming to understand the genetic basis of their unique diversity, environmental adaptation, and to a lesser extent, production traits.More particularly, several studies have sought to unravel the genomic footprints of positive selection (Bahbahani et al., 2017;Bahbahani, Afana, et al., 2018;Bovine HapMap Consortium, 2009;Kim et al., 2020;Kim, Hanotte, et al., 2017;Mekonnen et al., 2019;Taye et al., 2018;Tijjani et al., 2019Tijjani et al., , 2022)).
Here, after briefly summarising the history of the genome diversity of indigenous African cattle, we review the findings on signatures of positive selection.We intend to provide a comprehensive overview of the main candidate positive signatures of selection reported in African taurine, African zebu and their crossbreds.In addition, the review aims to guide further genome studies, paving the way for the identification of candidate causative mutations (e.g.single nucleotides polymorphisms (SNPs) and structural variants (SVs)) within regulatory and coding regions to understand further the biological mechanisms underlying the unique environmental adaptation of indigenous African cattle.
The review was compiled from 22 articles, retrieved and selected from known databases (ScienceDirect and PubMed), following the Joanna Briggs Institute protocol for scoping reviews (Peters et al., 2020).These databases were accessed on 16 December 2022.

DOM E ST ICAT ION, A DM I XT U R E A N D I N T ROGR E S SIONS H AV E SH A PED T H E GE NOM E S OF A F R ICA N CAT T L E
The diverse ecological zones in Africa, ranging from the Sahelian arid zones to the sub-humid and humid tropical forests, and from afro-alpine altitudes to ones below sea-level, have witnessed, throughout history, cattle pastoralists' strategic mobility and settlements coupled with socio-cultural and trading activities (Figure 1, Table S1).Accompanying these early movements, the humpless taurine were the first cattle to reach and disperse within the continent, according to archaeological and molecular evidence (Blench & MacDonald, 1999;Epstein & Mason, 1984;Hanotte et al., 2002).
The first unambiguous presence of domestic cattle in the continent dates from the end of the seventh millennium BC in the South of Egypt (Gautier, 1987;Gifford-Gonzalez & Hanotte, 2011;Grigson, 1991;Lesur, 2022).Predating Asian zebu arrival, these taurine cattle dispersed across a wide geographic area that covered East, Central and West Africa, with southern movements probably following the desertification of the Sahara around the third millennium BC (Felius et al., 2014;Lesur, 2022).These taurine cattle are today mainly found in the tsetse-infested belt in the whole of the West Africa subregion, with some populations now representing the only remaining non-zebu admixed taurine cattle on the continent.Uniquely, these populations express a unique heritable tolerance to trypanosome infection, a parasitic flagellate protozoon transmitted by the tsetse fly Glossina sp.(Mattioli et al., 2000;Murray et al., 1984), responsible for causing trypanosomiasis.Trypanotolerance attributes have been reported in both longhorn and shorthorn African taurine cattle, including the N'Dama (longhorn taurine) and Muturu (shorthorn taurine) (Freeman et al., 2004;Hanotte et al., 2002;Tijjani et al., 2019).
The other intercontinental cattle movements in Africa stemmed from the Indian subcontinent.It was mainly a male-mediated introduction and dispersal of animals of Asian zebu origin.It probably involved several waves of introduction and/or migration to and within the continent.These influxes were tied to the sea-routed long-distance trading activity between the Indian subcontinent and East Africa, through the Gulf of Oman and the Strait of Bab-el-Mandeb in the Horn of Africa.The earliest evidence of humped cattle, a painting in Karnak (Egypt), dates from around 1500 BC (Epstein, 1971).The Asian zebu may have reached the African continent in large numbers in the late seventh century with the development of Swahili civilisation and societies, which were closely embedded in the Indian Ocean trading networks.It led to crossing with the already present native taurine cattle, forming the hybrid cervico-thoracic humped sanga cattle.Interestingly, a recent genetic study identified a major taurine × indicine cattle admixture event from ca. 750-1050 years ago (Kim et al., 2020).These hybrid cattle became the commonest populations in the eastern and southern regions of the African continent.Around the end of the nineteenth century, the devastating rinderpest pandemic wiped out more than two-and-a-half million heads of African cattle (Epstein, 1971).Cattle predominantly of Asian zebu background then dispersed throughout the continent.They may be the origin of today so-called African zenga cattle (zebu × sanga crosses) and African zebu (Mwai et al., 2015;Rege, 1999).
Following their complex history, indigenous African cattle breeds today represent a unique mosaic of genetic diversity.At the autosomal DNA level, genomic mosaicism has been found in most indigenous African cattle with different proportions of the two domestic subspecies (zebu and taurine) (Kim et al., 2020;Kim, Hanotte, et al., 2017).A large proportion of bulls found in East Africa and along the Sahelian belt of West Africa carry a male Y indicine chromosome (Bradley et al., 1994;Hanotte et al., 2000).A dominant zebu genome influence is now visible in nearly all indigenous African cattle populations with the exception of taurine populations living deep within the tsetse belt in West Africa.Also, all African cattle have retained their maternal mitochondrial taurine ancestry and it has recently been shown that the ancestral incompatibility between the mitochondrial DNA and the autosomal genome has contributed to the shaping of the autosomal genome of African cattle (Kwon et al., 2022;Ward et al., 2022).
It has been proposed that the African aurochs B. primigenius africanus might have been domesticated on the continent (Grigson, 1991).So far, no ancient or modern archaeological or genetic evidence supports  S1 for climate class abbreviations; Beck et al., 2018;Kottek et al., 2006).These maps were generated with ArcGIS v.10.7.1 (ESRI, 2019).
such an event (Pitt et al., 2019), although it remains possible that local introgression with wild African aurochs may have shaped the genome of at least some African cattle populations.Indeed, two recent studies support local aurochs' introgression, in ancient domestic cattle from the southern Levant (Verdugo et al., 2019), a geographical area close to the taurine entry point on the African continent, and in modern Tunisia (Ginja et al., 2023).

T H E CON T R I BU T ION OF GE NOM E -W I DE POLY MOR PH I SM ST U DI E S TO T H E C U R R E N T K NOW L E DGE OF I N DIGE NOUS A F R ICA N CAT T L E DI V E R SI T Y
Following early work using mitochondrial and microsatellite DNA markers (Bradley et al., 1996;Hanotte et al., 2002), African cattle genomes have been extensively studied using genotypic and whole genome resequencing information.These studies have included population-level assessments of genetic diversity, admixture and demography dynamics, to reveal clues on the origin and history of the populations.These studies have also paved the way for the identification of signatures of selection in the autosomal genome, to identify not only candidate genome regions but also putative genes and variants that, ultimately, might be associated with economically important, as well as environmental adaptive traits (Bahbahani et al., 2015(Bahbahani et al., , 2017;;Bahbahani, Salim, et al., 2018;Kim et al., 2020;Tijjani et al., 2019).
In all these studies, most African humped cattle were found to possess a proportion of taurine and zebu ancestries, in agreement with previous microsatellite findings (e.g.Hanotte et al., 2002).Likewise, a high proportion of zebu background was observed in the genomes of African cattle populations in the Horn of Africa, including the trypanotolerant Ethiopian Sheko.Accordingly, the latter should now be reclassified as a crossbred taurine × zebu rather than as taurine (Rege et al., 1996).
Moving to West Africa, a close relationship was found between the zebu breeds in Cameroon and Nigeria, with a higher proportion of zebu genetic background in these populations compared with the zebu introgressed West African taurine (Ibeagha-Awemu et al., 2019).Recent studies have also explored maternal and/or paternal DNA markers to trace the demography of African cattle (Bradley et al., 1994;Kambal et al., 2021;Kwon et al., 2022;Mauki et al., 2021;Pérez-Pardal et al., 2010;Ward et al., 2022).All these studies support the fact that the genetic backgrounds of different ancestral origins, admixed and selected over time in different proportions according to the geography and environment of the continent, are at the roots of the success of African cattle pastoralism.
Until now, all genome-wide re-sequencing studies of African cattle populations have relied on European taurine reference genome assemblies, making them susceptible to reference-biased variant calling.However, recently, de novo genome assemblies have been generated for Asian indicine (Low et al., 2020) and for two African cattle breeds (N'Dama and Ankole) (Talenti et al., 2022).These latter two were integrated with genomic data from worldwide cattle breeds, leading to the identification of 116.1 Mb of sequence not represented in the reference Hereford based genome assembly, ARS-UCD1.2(Talenti et al., 2022).In the future, it may be expected that the availability of more de novo African cattle genomes and their integration into a pangenome will provide a more complete and comprehensive picture of African cattle diversity (Talenti et al., 2022).

GE NOM IC FOOT PR I N T S OF NAT U R A L SE L EC T ION
Several studies have attempted to decipher the impact of environmental selection on the genomes of African cattle (Bahbahani et al., 2017;Hanotte et al., 2003;Kim, Hanotte, et al., 2017;Noyes et al., 2011;Orenge et al., 2011;Taye et al., 2018).Such information would be useful in designing appropriate and effective selective breeding programmes aiming to improve productivity, while preserving the environmental sustainability of cattle production systems.Natural selection events have imprinted genomic regions with distinctive variation at the genotype and phenotype levels (Oleksyk et al., 2010).Different strategies may be used to detect such environmental signatures of selection.
One of the first investigations was an on-station study at the International Livestock Research Institute, Kenya, more than two decades ago.It aimed to unravel the genetic mechanism of trypanotolerance in F2 crosses between trypanotolerant N'Dama and trypanosusceptible Kenyan Boran (Hanotte et al., 2003).It included linkage mapping analysis of microsatellite markers to identify quantitative trait loci (QTL) associated with recorded animal phenotypes.Such a study required a large number of animals with recorded phenotypes and pedigree information.Such experiments are challenging to design and expensive to implement even under controlled (onstation) conditions.
New strategies have therefore been designed, taking advantage of whole genome sequencing information on a small number of individuals with geographic coordinates of sampling as the entry-point proxies of environmental selection pressures.These strategies aim to detect the genome signature of positive selection following a selective sweep or a hitchhiking effect on a beneficial genome variant (SNPs or SVs).These will affect upstream and downstream chromosomal sites, which may rise in frequency and/or approach fixation over generations (Fay & Wu, 2000;Oleksyk et al., 2010;Saravanan et al., 2020).These alleles may arise from previously unselected background variants, de novo variants, or variants that may have been introduced following an admixture and/or introgression event (Peter et al., 2012; see review by Saravanan et al., 2020).Statistical models that measure allele frequency deviations, diversity loss or changes in haplotype structures enable the detection of signatures of positive selection (Qanbari & Simianer, 2014;Utsunomiya et al., 2015).
Several statistical tests may be combined to increase the detection power and reduce the likelihood of false positive results (Randhawa et al., 2016;Utsunomiya et al., 2015).The findings may also be further supported by complementary approaches such as genomewide association studies and transcriptome analysis (e.g.Aliloo et al., 2020;Hanotte et al., 2003;Noyes et al., 2011).A significant challenge associated with the detection of positive signatures of selection is the presence of demographic factors (e.g.migration, admixture, fine-scale population structure, population expansion or bottleneck).These may result in genome variation patterns (e.g.changes in frequencies of haplotypes or SNPs) similar to those arising from positive selection (Haasl & Payseur, 2016;Nielsen et al., 2005;Weigand & Leese, 2018).
Many candidate regions under positive selection, often overlapping with genes or coding regions, have been reported in indigenous African cattle breeds.In most studies, comparisons with different cattle populations (African and non-African breeds) as well as within and between breeds were performed (Table S2).For regions with annotated genes, researchers have sought to get insights into the biological significance of the candidate genes by investigating their functions, including using functional enrichment and gene ontology analyses.The highest numbers of genomic regions under positive selection were found on BTA3, BTA5 and BTA7.These regions intersect with candidate genes having roles in several biological functions and pathways, including but not limited to growth and feed intake, cell stability, protein stability and sweat gland development (Tables 2 & S3).Table 3 summarises the information about the number of studied breeds, methods used for genome-wide variation scanning, the bovine genome of reference, statistical tests and the number of informative SNPs.
In the following sections, we highlight the genomic regions reported by these studies and the biological function of the candidate genes for three important adaptive traits in indigenous African cattle: heat tolerance, parasitic disease resistance and adaptation to high altitude (e.g.hypoxia, cold temperature, UV radiation).

Heat tolerance
Besides the deleterious impact of heat stress on cattle productivity, indirect effects stem from the impact on water availability and forage quantity and quality.The ability to dissipate excess heat, regulate body temperature and preserve water are key survival characteristics, with B. taurus indicus cattle being better adapted to these challenges at cellular and physiological levels than their taurine counterparts (Beatty et al., 2006;Hansen, 2004).Cattle may control their body temperatures through sensible mechanisms, including conduction, convection and radiation, and through latent mechanisms, such as sweating and panting (Dos Santos et al., 2021;Hansen, 2004).Shade seeking, drinking and reducing feed intake are behavioural strategies to maintain thermal equilibrium when an upper critical temperature is exceeded (Herbut et al., 2018), while body surface area, coat pigmentation and hair thickness and length are also of high importance (Hansen, 1990(Hansen, , 2004)).
Several studies have reported candidate genes related to African cattle heat stress adaptation.The detection of selective sweeps overlapping heat shock proteins (HSPs) and their transcription factors (HSFs) has been linked to cattle responses to stressful environmental heat conditions.These proteins, such as HSPA4, HSPA9, HSPB9 and HSF5 may act to protect the cell and prevent protein denaturation during exposure to heat stress (Basiricò et al., 2011;Kim, Hanotte, et al., 2017).In addition, genes that encode chaperone proteins (DNAJC8, DNAJC18), which contribute to the conformational folding and unfolding of heat shock proteins (HSP70) (Kampinga & Craig, 2010), are found within a candidate genome region of low heterozygosity in African cattle (Bahbahani et al., 2017).
It is also known that heat stress induces the downregulation of expression of antioxidant enzymes, leading to oxidative stress (Chauhan et al., 2021).Comparative genomic analysis between African and commercial breeds (European and Asian) has revealed several genes involved in the control of oxidative stress, e.g.superoxide dismutase 1 (SOD1) and glutathione peroxidase 7 (GPX7) (e.g.Taye, Lee, Caetano-Anolles, et al., 2017).The SOD1 and GPX7 proteins have crucial antioxidant enzymatic activities protecting the cell from cytotoxicity and damage that may occur owing to the excessive formation of intra-and extracellular reactive oxygen species (ROS) during heat stress (Belhadj Slimen et al., 2016;Kim, Hanotte, et al., 2017).
An additional non-enzymatic antioxidant defence mechanism is suggested to be driven endogenously by glutathione and ascorbic acids (e.g.vitamins C and E).A decline in their concentration has been reported in  (Pandey et al., 2014) during hot seasons, and a reduction in the free radical concentration is recorded in Asian buffalo supplemented with ascorbic acid (Kumar et al., 2011).It supports their attenuation effect on the consequences of oxidative stress, a result that is consistent with the positive signature of selection regions overlapping two candidate Solute Carrier Family 23 genes (SLC23A1 and SLC23A2) (Taye, Lee, Caetano-Anolles, et al., 2017).These genes encode proteins that are active in bulk vitamin C transport, supporting its delivery and absorption in different tissues.
An important heat adaptation biological pathway is the prolactin one.The multi-function nature of the prolactin hormone (PRL) was linked physiologically to homeothermic regulation with the discovery of a dominant mutation in hairy and heat-intolerant New Zealand dairy cattle (Davis et al., 2017;Littlejohn et al., 2014).Moreover, the same study found a mutation in exon 10 of the prolactin receptor (PRLR) gene associated with the heat-tolerant slick-coat phenotype in the taurine Senepol (Littlejohn et al., 2014).
The predicted alteration caused by SLICK variant(s) on the tertiary structure of prolactin receptor is expected to affect its interaction with the prolactin hormone, explaining the observed defect in lactation and sweat gland dysfunction (Flórez Murillo et al., 2021;Littlejohn et al., 2014;Porto-Neto et al., 2018).A recent study in the Holstein breed aimed to evaluate whether  (Sosa et al., 2022).It included histological, immunohistochemical and differential gene expression analyses.The mutation was not found to be associated with the proportion of sweat glands in the skin, but with an increase in the amount of the immunoreactive forkhead transcription factor A1 (FOXA1) protein.The FOXA1 protein serves to regulate the expression of a set of genes involved in the secretory function of sweat glands (Cui et al., 2012;Sosa et al., 2022).Interestingly, we found in African indicine a signature of selection (XP-EHH) on BTA3 in a region intersecting with the prolactin-releasing hormone gene (PRLH), which harbours a highly conserved missense variant at exon 2 (Kim, Hanotte, et al., 2017;Tijjani et al., 2022).This variant might be involved in the regulation of the expression and stimulation of the prolactin hormone (PRL) and its receptor (PRLR) (Kim, Hanotte, et al., 2017).The superior heat tolerance of zebu in comparison with taurine cattle has been attributed to the large number and size of the sweat glands close to the skin surface (Collier & Gebremedhin, 2015;Hansen, 2004;Nay & Hayman, 1956;Pan, 1963).With other skin phenotypes, e.g.shorter hair and light-coloured coats, it may explain the enhanced capability of zebu cattle to dissipate heat.
Other candidate selected regions reported in African cattle include genes involved in osmoregulation (e.g.BHMT2, ITPR2, CFTR and SCNN1D) (Taye, Lee, Caetano-Anolles, et al., 2017).The BHMT2 has a secondary beneficial effect on maintaining water balance under hyperosmotic conditions that may occur during heat stress, thus reducing the body temperature through the amount of sweat production (Del Vesco et al., 2015).Moreover, the signatures of selection at the ITPR2 and ITPRIP gene regions support more effective sweating in African zebu.Both ITPR2 and ITPRIP proteins have a role in regulating the intracellular movements of calcium ions in the sweat glands (Cheruiyot et al., 2021;Klar et al., 2014;Taye, Lee, Caetano-Anolles, et al., 2017).Similarly, proteins encoded by the GFTR and SCNN1D genes are anion channels that facilitate the transport of chlorides and sodium absorption, respectively (Lee et al., 2007;McDonagh et al., 2015).
Cattle with a light coat colour will have a reduced light absorption and an excellent reflecting ability to solar radiation, contributing to body heat temperature regulation in tropical environments (Hansen, 2004).As in many mammals, the coat colour phenotype is a complex quantitative trait expressed by the interaction between several genes controlling the ratio of two major classes of melanin (eumelanin (black-brown pigment) and phaeomelanin (red-yellow pigment); Hubbard et al., 2010) transporter (SLC45A2) and Ras oncogene family genes (RAB17, RAB37) were detected in African cattle breeds.These genes play a direct role in melanogenesis (Cieslak et al., 2011;Dlamini et al., 2022;Taye et al., 2018;Taye, Lee, Caetano-Anolles, et al., 2017;Taye, Lee, Jeon, et al., 2017).A mutation in the MLPH gene in dogs, cats and rabbits is thought to disrupt the transportation of melanosomes to the outer edges of melanocytes (Philipp et al., 2005); it prevents the transfer of melanosomes into the keratinocytes, thus resulting in hairs of lighter colour (Demars et al., 2018;Ishida et al., 2006).Moreover, variants detected at SLC45A2 have been linked to the cream colour phenotype in horse and plumage colour variation in chickens (Gunnarsson et al., 2007;Mariat et al., 2003).The MC1R gene, which has essential roles in skin pigmentation, is present in a signature of positive selection region on BTA18 in Nguni cattle (Makina et al., 2014).Another region putatively associated with heat tolerance is found on BTA5.It harbours the homeobox genes, HOXC12 and HOXC13, which have roles in determining skin thickness and the number of hair follicles (Makina et al., 2014;Taye, Lee, Caetano-Anolles, et al., 2017;Tijjani et al., 2022).Besides environmental selection, it may also be argued that these footprints of selection involving coat colour and coat structural properties might have been directly influenced by aesthetic and social-cultural preferences.
A number of studies have assessed the impact of high temperatures on milk production in African dairy cattle (Ekine-Dzivenu et al., 2020;Moreki & Tsopito, 2013;Nesamvuni et al., 2012), with a decline in milk production correlating to the increase in the temperature-humidity index.It has been hypothesised that heat tolerant cattle breeds will better regulate their metabolism and their energy expenditure which will in turn impact milk production (Hansen, 2004).Footprints of selection in African cattle include genes involved in metabolism and feed intake pathways (e.g.FTO, IGF-1, ATRN, PRKCZ, ACLY, GHR and NFATC2) (Bahbahani et al., 2015;Taye, Lee, Caetano-Anolles, et al., 2017).Contributing to the regulation of metabolic heat production, these genes may represent an adaptive response of African cattle to high-temperature environments (Taye, Lee, Caetano-Anolles, et al., 2017;Tijjani et al., 2022).Of particular interest are the candidate IGF-1, ACLY and PRKCZ proteins, which are involved in insulin metabolism, carbohydrate regulation and lipid metabolic processes that are essential for the body's energy balance (Bahbahani et al., 2015;Taye, Lee, Caetano-Anolles, et al., 2017;Tijjani et al., 2022).Also, IGF-1 and GHR are good examples of genes with pleiotropic effects that may influence the expression of several phenotypes.A large body of investigations has shown the association of these genes with stature, reproduction and milk production in cattle (see, Cheruiyot et al., 2021;Ge et al., 2001;Gobikrushanth et al., 2018), which may explain the variations in body size seen in zebu cattle across different environments (DAGRIS, 2007).

Trypanosomiasis tolerance
Owing to its estimated economic cost, trypanosomiasis has been ranked among the top 10 global cattle diseases impacting African cattle production (Perry et al., 2002).The disease symptoms include anaemia, weight loss and a high level of blood parasitaemia.Besides its impact on productivity, it may lead to infertility and calf abortion and even be fatal if not treated.Within the geographic distribution of its biological vector, the tsetse fly Glossina sp., we do find a few trypanosomiasis tolerant breeds of African taurine descent (e.g.N'Dama, Muturu, Baoule and Lagune).As such, they remain healthy and productive compared with breeds of indicine or European background (Berthier et al., 2015;Chandler, 1958;Murray et al., 1984).
Several studies have now attempted to unravel the genetic control of trypanotolerance.The pioneering work of Hanotte et al. (2003) identified 10 major QTL on several chromosomes (e.g.BTA7, BTA13, BTA17 and BTA20) linked to anaemia, parasitaemia and body weight phenotypes.Interestingly, a few of the 'tolerant' alleles originated from the Kenyan Boran.Subsequent studies have identified lists of candidate genes intersecting with these QTL as well as differentially expressed genes at various time points following trypanosomiasis infection (Noyes et al., 2011;O'Gorman et al., 2009).
Recently, positive signature of selection analysis has been performed in West African taurine breeds to identify candidate genetic controls and genes linked to trypanotolerance.A strong signal with a unique haplotype pattern was found in a candidate regulatory genomic region upstream of CARD11 gene (Kim et al., 2020;Kim, Ka, et al., 2017).The CARD11 is involved in the activation of protein signalling complexes for T and B lymphocyte development (Pomerantz et al., 2002).This candidate regulatory region could therefore modulate the expression of CARD11 and contribute to African taurine tolerance and defence against trypanosomiasis infection (Kim et al., 2020;Kim, Ka, et al., 2017).Other candidate adaptive selection footprints for trypanotolerance included well-known genes associated with anaemia (e.g.STOM) and inflammatory (e.g.IL1RL2) responses (Kim, Ka, et al., 2017).The IL1RL2 could be controlling the initial parasitaemia peak by interacting with pro-inflammatory mediators and generating a cell signalling cascade that contributes to pathogen clearance.
Muturu is a trypanotolerant indigenous African taurine breed, which is found within the humid and subhumid agro-ecologies of Nigeria.Tijjani et al. (2019) identified strong signals of selection on BTA23 in both Muturu and N'Dama, overlapping several candidate genes of the bovine MHC Classes I and II.They also include genes linked to innate immune response (Tripartite Motif genes -TRIMs).Moreover, a possible association was hypothesised between a shared fixed non-synonymous (missense) variant at exon 3 on the BOLA-DQA2 gene (rs208515389) and the trypanotolerance attribute of Muturu and N'Dama.Furthermore, candidate selected regions were uncovered in Muturu at BTA11 and 24 in comparison with the N'Dama.These regions intersect with the Obg-Like ATPase 1 (NPC1) and Growth Factor Independent 1B (GFI1B) genes, which both encode proteins that are essential for the development of erythrocytes and megakaryocytes (Saleque et al., 2002).Moreover, the genome-wide iHS score in Muturu supports a strong signature of selection on BTA23 which overlaps with several unannotated protein-coding genes and will require further functional investigations (Tijjani et al., 2019).
In South-West Ethiopia, around the Bench Maji zone, where a high prevalence of trypanosomiasis is recorded (Gebre et al., 2022), the Sheko cattle are well known for their better trypanotolerance compared with other local neighbouring breeds (e.g.Horro, Abigar) (Lemecha et al., 2006).Following iHS, Rsb and CLR analyses, Mekonnen et al. (2019) reported several candidate genes (e.g.MIGA1, SPAG11, ERN1 and CAPG) in the Sheko which might be associated with trypanotolerance attributes, such as anaemia, immune response and neurological dysfunction.Here, the sperm-associated antigen 11 (SPAG11) has an important role in immune response in cattle (Avellar et al., 2007).Indeed, differential expression of the gene transcript produces isoforms encoding defensin-like peptides, which form multimeric complexes as a defence against pathogens.These isoforms were also found to have roles in recruiting T cells and dendritic cells (Yang et al., 1999).
Besides its link to anaemia, Hypoxia-Inducible Factor-1 (HIF-1) is involved in the control of the activation and functions of immune cells (Bhandari & Nizet, 2014;Palazon et al., 2014).It has a protective effect, boosted during oxygen depletion by the expression of hypoxia-inducible factors that promote immunity profiles.Indeed, low oxygen tension renders the hypoxia-inducible factor (HIF) subunits stable and able to form transcriptional complexes that, in turn, regulate an array of genes involved in innate immune cell function and pathogen clearance (Hammond et al., 2020;Schaffer & Taylor, 2015).Therefore, it is possible that two hypoxia-associated genes (ERN1 and CAPG) found within footprints of selection in the Sheko are linked to the control of trypanosome infection in this breed (Mekonnen et al., 2019).Other relevant overrepresented immunity-related biological pathways in the Sheko include T-cell chemotaxis and cell-cell adhesion, interleukin-3 (IL-3) signalling and cyclin-dependent kinase (Cdk2) pathways (Bahbahani et al., 2015;Mekonnen et al., 2019).
A few shared trypanotolerance adaptive footprints of positive selection have also been found between African taurine and some African humped cattle breeds (e.g.Kenana and Borgou) (Bahbahani, Salim, et al., 2018;Gautier et al., 2009).Considering that all African humped cattle are admixed with African taurine, it is possible that the relevant ancestral taurine alleles were favoured during past admixture events.

Tick resistance
The complexity of tick-borne disease challenges in Africa includes the ability of diverse tick species to transmit a wide range of pathogenic organisms (e.g.theileriosis, anaplasmosis and babesiosis).They represent major impediments to cattle health and production, especially in the East and Southern African regions (Phiri et al., 2010).The first study aiming to understand the genetic control of tolerance to tick infestation used a protein biomarker (serum amylase) (Ashton et al., 1968).Resistance to ticks and tick-borne diseases have been reported in African cattle breeds (e.g.Boran (African zebu), Ankole and Nguni (sanga), N'Dama (African taurine) and the composite Afrikaner) (Ali & de Castro, 1993;Mapholi, 2015;Mwai et al., 2015;Wambura et al., 1998), with the cellular immune response probably playing an important role (Mattioli et al., 2000;Mwangi et al., 1998).Machado et al.'s (2010) study suggests that such adaptation may also be season specific.
Under the assumption that tick and tick-borne disease challenges were already present in the centre(s) of origin of zebu cattle on the Asian continent, the identification of footprints of positive selection in the genome may be expected (e.g.Kim et al., 2020).In a study that included four South African cattle breeds (Afrikaner, Nguni, Bonsmara and Drakensberger), several genes belonging to the keratin family (e.g.KRT222, KRT24, KRT25, KRT26 and KRT27) were found in Nguni within a candidate selected region on BTA19 (19: 42,896,897,840) (Makina et al., 2014).The coat's skin thickness and structure may play a role here.Collagen genes, such as COL12A1 and COL8A1, were also found to be under selection in some African breeds (e.g.Kenyan Boran, Kenana, and Ogaden), which were reported to be resistant to some tick-borne diseases (Taye et al., 2018).A positive linear relationship was found between the level of coat smoothness of Nguni cattle and tick count (Marufu et al., 2011).These studies support the idea that skin structural properties may represent an important physical barrier to tick infestation and their pathogenic agents.In addition, coat colour and, henceforth, melanogenesis-associated candidate genes, may also be involved (Taye, Lee, Caetano-Anolles, et al., 2017;Verissimo et al., 2002).Indeed, it is possible that breeds with lighter-coloured hair would be less susceptible to tick infestations, as ticks would be more easily seen and picked up by predators (e.g.oxpeckers) or animals self-grooming.
Another relevant epidermis-related trait in African cattle is the control of the secretion of a lubricated and acidic film of sebum on the skin, supposedly unfavourable to tick attachment (Kongsuwan et al., 2010).The MC5R candidate gene found at a homozygous sweep on BTA24 in African cattle is here of interest (Taye et al., 2018).This gene regulates the sebaceous glands secretion of a complex mixture of lipids, called sebum, onto the hair follicles, contributing to the skin integrity as well as the inflammatory response following the attachment of ticks (Zhang et al., 2011).
Indeed, following the biting of an ectoparasite, several inflammatory mediators are released at high concentrations, e.g.histamine, which has been shown to influence tick salivation, feeding and attachment through the enhanced recruitment of pro-inflammatory cells into biting sites (Falcone et al., 2001).Also, PRG3 and ATRN are present within a candidate selected genome region in African indicine breeds (Taye et al., 2018).PRG3 stimulates the release of histamine, one of the important determinants of tick rejection (Falcone et al., 2001).Besides, many other immune homeostasis-related genes have been reported in African humped cattle breeds that plausibly may be linked to their adaptive responses to a wide range of infectious agents transmitted by tick bites (Kim, Hanotte, et al., 2017;Makina et al., 2014;Mattioli et al., 2000;Taye et al., 2018).In a study on East African zebu breeds, Tijjani (2019) found that several shared candidate selective sweeps overlapped with tick resistance QTL.These include a region on BTA2 that overlaps a dry-season-specific tick-resistant QTL (Machado et al., 2010).

Adaptation to high altitude
In high-altitude African environments, cattle breeds face major environmental challenges such as cold temperatures, low oxygen partial pressure, high UV solar radiation and limited feed resources (Friedrich & Wiener, 2020).Despite these challenges, we do find indigenous African cattle breeds living at high altitudes, typically above 2500 m above sea level (m.a.s.l.), e.g.Bale, Simien and Choke cattle from Ethiopia (Alkorta-Aranburu et al., 2012;Rege & Tawah, 1999;Terefe et al., 2022).Uniquely adapted to the high-altitude environments, they support the livelihoods of local human communities.Edea et al. (2013Edea et al. ( , 2014) ) and Terefe et al. (2022) have reported candidate positively selected regions and genes enriched in functional categories (gene ontology terms and pathways) associated with high-altitude stressors, from the analysis of DNA SNP-Chip and whole-genome sequencing data, respectively.
Several studies have compared livestock populations living at high (>2500 m.a.s.l.) and low (<1500 m.a.s.l.) altitudes in Ethiopia to understand how natural selection might have shaped the genomes of cattle living at high altitudes (Edea et al., 2013(Edea et al., , 2014;;Terefe et al., 2022;Wuletaw et al., 2011).Using a genetic differentiationbased statistical test (F ST ), differentiated loci were found between highland-and lowland-dwelling cattle breeds.These overlap with genes involved in the responses to hypoxia (Edea et al., 2014;Terefe et al., 2022).For example, the brain-derived neurotrophic factor (BDNF) gene promotes changes in pulmonary artery structure and function, which may be linked to tolerance to hypoxia and ischemia (Chen et al., 2012).In the study by Edea et al. (2014) the gene ontology analysis of the candidate genes shows that the majority of these have metabolic functions (e.g.ATP2A3, MYO18B, and INPP4A).A reduction in body metabolic rate might be an adaptive mechanism to the high-altitude environment characterised by a lower oxygen concentration (Riek et al., 2019).In the Sheko, HIF was one of the overrepresented pathways.It is linked to the production of red blood cells (erythropoiesis) and haemoglobin, which may overcome the reduced oxygen concentration at high altitudes (Mekonnen et al., 2019). However, Wuletaw et al. (2011) failed to find any differences in haematological parameters (red blood cell counts, haemoglobin and haematocrit values) between Ethiopian Simien cattle living above 2700 m.a.s.l. and other indigenous Ethiopian breeds living at lower altitudes, including at sea level.They therefore proposed that the indigenous cattle from the Simien mountain area are adapted to high altitudes by largely eliminating the hypoxic pulmonary vasoconstrictor response (Wuletaw et al., 2011).

DI SC US SION
We provided here an overview of the candidate genomic regions under positive selection in African cattle breeds, with a focus on environmental adaptation.Studies including African crossbred cattle with exotic breeds (e.g.Cheruiyot et al., 2018;Van der Nest et al., 2021) and/or studies addressing non-environmental adaptative traits (Nanaei et al., 2020) were excluded.Henceforth, the review was compiled from 22 studies selected using three main criteria: (i) genome-wide scanning for positive selection; (ii) genome-wide analysis using genome sequences or SNP-Chip data; and (iii) inclusion of indigenous African cattle (taurine, zebu, sanga, zenga).
Several candidate regions that were either breed specific or shared between breeds were reported in each study.The challenge is therefore to identify true signals from false positive ones, e.g.following population demographic history.Here, several approaches may be envisaged.For example, replication studies may be performed with cattle populations of different evolutionary histories but living under the same environmental challenges (e.g.Bahbahani et al., 2017).The results may also be overlapped with genome-wide association studies and transcriptome analysis studies (e.g.Aliloo et al., 2020) as well as association mapping followed by targeted differential candidate gene expression analysis (e.g.Noyes et al., 2011 andO'Gorman et al., 2009 (cattle), Wragg et al., 2012Wragg et al., , 2013 (chicken) (chicken)).Equally challenging are the definition and recording of the phenotypes shaped by natural selection as well as the identification of the key environmental selection parameters at the roots of the selection pressures.Ultimately, the identification of the causative polymorphisms followed by gene editing approaches will demonstrate the relevance of the candidate selected regions and their polymorphisms.While such information is still largely missing for African indigenous cattle, it may be expected that the search for candidate causative polymorphisms will increasingly be the focus of future studies, paving the way to genome editing in African cattle.
Environmental challenges experienced by African cattle populations are complex.The definition of an agro-ecology includes many environmental parameters such as temperature, humidity, rainfall patterns and intensity, vegetation cover, etc.These environmental parameters will be influenced by the landscape topography, which may add other selection pressures (e.g.altitude).Understanding the key natural selection pressures and their interactions is an essential step in unravelling the genetic control of local adaptation.Ecological niche modelling has been applied for the identification of the main environmental selection pressures on indigenous chicken populations of Ethiopia (Kebede et al., 2021;Vallejo-Trujillo et al., 2022).A landscape genomics approach has also been used to understand environmental adaptation in indigenous Ethiopian sheep populations (Wiener et al., 2021).Applying the same approaches to identify the key environmental stressors in African cattle may help identify the most relevant candidate positively selected regions.Interestingly, in African indigenous chicken populations, some candidate positively selected regions overlap with many genes (Gheyas et al., 2021).
Here, a single mutational event (SNP or SV) (Kebede et al., 2021;Vallejo-Trujillo et al., 2022) might influence several molecular pathways and phenotypes.Supergenes have been reported in butterflies (Komata et al., 2022), as well as in wild birds (Jeong et al., 2022), where they have been shown to influence morphology, plumages and behaviour.There is also now increasing evidence that SVs may be involved in ecotype differentiation of wild species (Faria et al., 2019).Still, no supergene(s) has been reported for environmental adaptation in livestock species, including cattle.
The mosaic genome of African cattle is now well established (Kim et al., 2020), with two main ancestral backgrounds (taurine, indicine) and several waves of introductions and dispersion within and across the continent.Shaped by human and natural selection, admixed taurine × indicine African cattle are now thriving on the continent.Each genome background would have brought its own set of environmental adaptations from its centres of origin.For example, there is little doubt that the successful introgression of African indicine background into African taurine populations was favoured by the indicine adaptation to the dryland environments of the Indian subcontinent (e.g.Kim et al., 2020).However, hybridisation comes at a cost and beyond F1 unselected hybrids may have reduced fitness (Adavoudi & Pilot, 2021).Of relevance here, two recent studies have reported in African hybrid cattle populations signatures of positive selection in genome regions including genes involved in nuclear-mitochondrial DNA interactions and reproduction (Kwon et al., 2022;Ward et al., 2022).
Importantly, two cattle populations with different evolutionary histories but exposed to the same environmental challenge could respond differently to the same selection pressures.So, we should not expect the same signature of positive selection to be present in these two populations.A good example is the response to heat stress, which may include, besides adaptive cellular responses, behavioural ones (Pereira et al., 2014).Even within the same molecular pathway, the mutational response may be different (e.g.prolactin pathway) (Littlejohn et al., 2014;Sosa et al., 2022).Signals of positive selection may also include genes with pleiotropic effects, such as those involved in various metabolic pathways (e.g.Bahbahani et al., 2015).Accordingly, different selection pressures may be acting on the same genomic region and associated genes, which might explain the overlapping between candidate regions in various commercial breeds (such as Holstein and Jersey), in response to artificial selection for productivity, and in indigenous African cattle in response to natural selection.
While the focus of this review was the detection of positive selection signatures, environmental selection may also involve balancing selection.In this regard, African human populations adapted to the environmental challenge of malaria (Amambua-Ngwa et al., 2012) are a relevant example.Contrastingly, the imprints of balancing selection on the genomes of African cattle are yet to be documented.However, considering the complexity of environmental challenges, with geographic areas on the African continent witnessing seasonal extreme and unpredictable weather patterns (e.g.dry and wet seasons of varied intensity) as well as several livestock parasitic diseases, balancing and/or polygenic selection might be expected.For example, it is possible that balancing selection may have favoured genes enhancing innate immunity while polygenic selection may have acted to stabilise the mosaic taurine/indicine proportions in the genomes of African cattle following the twin challenges of climate and disease adaptation.
Last but not least, the availability of a cattle pangenome will be particularly helpful to unravel the complexity of African cattle genome diversity.This is particularly the case considering the legacy of multiple introductions and admixture on the continent of cattle populations of taurine and indicine origin.Associated with long-read sequencing (such as, Oxford Nanopore Technology and Pacific Biosciences) information, a cattle pangenome will allow complex structural variants that are difficult to call from short-read paired-end sequencing to be unravelled in fine detail.It may also offer the opportunity to discover 'African adaptation supergene(s)', for which the discovery is constrained by the limitations of shortread sequencing and today's bioinformatics tools, as well as the European taurine bias of the bovine genome of reference.

CONC LUSION
Climatic changes are impacting the pastoral and smallholder livestock production systems of the African continent.A key question is whether livestock populations across the continent possess the resilience and tolerance to deal with future environmental extremes and changing ecological conditions.This requires efforts to characterise today's breeds and to understand their current environmental adaptations and their genetic mechanisms.Domestic cattle have successfully conquered the African continent since time immemorial.However, they are now witnessing for the first time, human-driven rapid environmental changes at an unprecedented rate.There is now an urgent need to fully characterise in finer detail at genome level the uniqueness of African cattle adaptations to inform future sustainable genetic improvement programmes and new breeding strategies for African cattle populations.Furthermore, modelling and simulation-based research that takes into account several adaptive scenarios and incorporating time-series data are needed to get deeper insights into the past and future events that have shaped or may shape the genomes of modern-day African cattle.Only then will we be able to respond to the present and future needs of the African cattle industry and the consumers of livestock products and by-products.

CON F L IC T OF I N T E R E ST STAT E M E N T
The authors declare no conflict of interests.

F
Maps of African environmental and ecological conditions faced by indigenous African cattle.(a) Annual climatic trends (mean temperature and precipitation) based on WorldClim 2.1 data (1970-2000) with a spatial resolution of 10 m, and elevation in metres (Fick & Hijmans, 2017).(b) Continental location of indigenous African breed groups reported in this review (African taurine, African zebu, sanga, zenga), pinpointed on the Köppen-Geiger climate map (1980-2016) (see Table

13652052, 0 ,
Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/age.13353by EBMG ACCESS -ETHIOPIA, Wiley Online Library on [29/10/2023].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Conceptualization: S.K., O.H. and A.T. Data curation: S.K. and S.A.E.I.Writing -original draft preparation: S.K. and O.H. Review & editing: O.H., J.M.M., MK.A.A., A.T. and S.K.AC K NOW L E DGM E N T SWe acknowledge the support received from the International Livestock Research Institute, Ethiopia, the Organization for Women in Science for the Developing World, the Swedish International Development Cooperation Agency and the Faculty of Animal Production at the University of Khartoum.This work is included in the framework of Sumaya Kambal's PhD project which is funded by the above-mentioned institutes.F U N DI NG I N FOR M AT IONThis work was funded in part by the International Livestock Research Institute LiveGene programme, supported by the CGIAR Research Program on Livestock (Livestock CRP), which is funded by contributors to the CGIAR Trust Fund (http://www.cgiar.org/about-us/ our-funde rs/).We acknowledge the support of the Bill & Melinda Gates Foundation, UK aid from the UK Foreign, Commonwealth, and Development Office (FCDO) under the auspices of the Centre for Tropical Livestock Genetics and Health, established jointly by the University of Edinburgh, Scotland's Rural College, and the International Livestock Research Institute (grant agreement OPP1127286).The findings and conclusions contained within this paper are those of the authors and do not necessarily reflect the positions or policies of the Bill & Melinda Gates Foundation nor the UK Government.Open-access publishing cost was supported by the University of Nottingham (United Kingdom).
13652052, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/age.13353by EBMG ACCESS -ETHIOPIA, Wiley Online Library on [29/10/2023].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License List of publicly available whole genome sequence data of indigenous African cattle.
T A B L E 1 a Based on the geographical area inhabited by groups: EAZ, East African zebu; WAT, West African taurine; AAD, East African admixed which refers to both sanga and zenga groups.b Kenyan East African shorthorn zebu.c Sheko cattle were originally grouped as a Humpless Shorthorn Taurine.13652052, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/age.13353by EBMG ACCESS -ETHIOPIA, Wiley Online Library on [29/10/2023].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 13652052, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/age.13353by EBMG ACCESS -ETHIOPIA, Wiley Online Library on [29/10/2023].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License goats , 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/age.13353by EBMG ACCESS -ETHIOPIA, Wiley Online Library on [29/10/2023].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Summary of studies that used positive signature of selection approaches to identify candidate genes probably associated with environmental adaptation in indigenous African cattle.Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/age.13353by EBMG ACCESS -ETHIOPIA, Wiley Online Library on [29/10/2023].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License cattle inheriting the SLICK1 mutation in the PRLR gene (rs517047387; Littlejohn et al., 2014) have greater capacity for sweating (i.e.sweat gland abundance and function) than wild-type cattle a Based on the Gene Ontology (GO) repository.13652052T A B L E 3 . It is noteworthy that signatures of selection spanning melanophilin (MLPH), melanocyte stimulating hormone receptor (MC1R), membrane-associated Continued 13652052, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/age.13353by EBMG ACCESS -ETHIOPIA, Wiley Online Library on [29/10/2023].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Abbreviations: GBS, genotype-by-sequencing; WGS, whole-genome sequencing.aExcluding the breeds used as reference and outgroup.See TableS2for the names of studied and reference breeds.T A B L E 3 13652052, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/age.13353by EBMG ACCESS -ETHIOPIA, Wiley Online Library on [29/10/2023].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License