Divergent adaptation to highland and tropical environments in Bolivian Creole cattle

Abstract

Bolivian Creole cattle populations evolved under low levels of breeding management and, during more than 500 years of natural selection, became adapted to various environments such as the contrasting highland and subtropical environments. Recently, highland Creole cattle were crossbred with Holstein to improve dairy production. The aim of this research was to evaluate the divergent adaptation through selection footprints of Bolivian Creole cattle from Andean highland and tropical lowlands, and to evaluate the effect of Holstein introgression in highland Creole. For this purpose, 130 Creole cattle (75 highland, 55 lowland) and 88 Holstein were genotyped using a microarray. The database was used to determine population structure and admixture and detect selection sweeps using F ST , Rsb, XP-EHH and ROH. Ancestry inference suggested that selection peaks were not due to Holstein introgression. The NCBI database was used to retrieve genes from the common regions and then perform gene ontology analysis. The most prominent selection peaks were on BTA20 and BTA23 and included the PRLR (slick phenotype) and Class I and IIa BoLA genes. Other windows contained candidate genes for hypoxia ( ANXA2 , NDUFA4L2 ), angiogenesis, immune response ( IL7R , IL6ST , IL31RA , C6 , C7, STAT6 , NKG2A , IRAK4 , KLR, CLEC ), oxidative stress ( GSTA, HSD17B6 ) and morphological traits ( PLAG1, CHCHD7 , CAP2, ARL15) . GO analysis revealed enrichment terms and pathways related to immune response, glutathione and retinol metabolism and reported QTLs for coat characteristics, immune response, and tick resistance. The results suggest the complex mechanism in the adaptation of Bolivian Creole cattle to the contrasting highland and subtropical environments.

Full text

Page 1/25
Divergent adaptation to highland and tropical
environments in Bolivian Creole cattle
Guillermo Giovambattista
National University of La Plata
Olivia Marcuzzi
National University of La Plata
Paulo Alvarez Cecco
National University of La Plata
Leonidas Olivera
National University of La Plata
Juan Pereira Rico
Universidad Autónoma Gabriel René Moreno
Francisco Calcaterra
National University of La Plata
Ariel Loza Vega
Universidad Autónoma Gabriel René Moreno
Pilar Peral Garcia
National University of La Plata
María Fernandez
National University of La Plata
Andres Rogberg Muñoz
University of Buenos Aires
Article
Keywords: selection footprints, heat stress, hypoxia, slick, BoLA, SNP
Posted Date: July 3rd, 2024
DOI: https://doi.org/10.21203/rs.3.rs-4492487/v1
License:   This work is licensed under a Creative Commons Attribution 4.0 International License. 
Read Full License

Page 2/25
Additional Declarations: There is no duality of interest

Page 3/25
Abstract
Bolivian Creole cattle populations evolved under low levels of breeding management and, during more
than 500 years of natural selection, became adapted to various environments such as the contrasting
highland and subtropical environments. Recently, highland Creole cattle were crossbred with Holstein to
improve dairy production. The aim of this research was to evaluate the divergent adaptation through
selection footprints of Bolivian Creole cattle from Andean highland and tropical lowlands, and to
evaluate the effect of Holstein introgression in highland Creole. For this purpose, 130 Creole cattle (75
highland, 55 lowland) and 88 Holstein were genotyped using a microarray. The database was used to
determine population structure and admixture and detect selection sweeps using FST, Rsb, XP-EHH and
ROH. Ancestry inference suggested that selection peaks were not due to Holstein introgression. The
NCBI database was used to retrieve genes from the common regions and then perform gene ontology
analysis. The most prominent selection peaks were on BTA20 and BTA23 and included the
PRLR
(slick
phenotype) and
Class I
and
IIa BoLA
genes. Other windows contained candidate genes for hypoxia
(
ANXA2
,
NDUFA4L2
), angiogenesis, immune response (
IL7R
,
IL6ST
,
IL31RA
,
C6
,
C7,
STAT6
,
NKG2A
,
IRAK4
,
KLR, CLEC
), oxidative stress (
GSTA, HSD17B6
) and morphological traits (
PLAG1, CHCHD7
,
CAP2,
ARL15)
. GO analysis revealed enrichment terms and pathways related to immune response, glutathione
and retinol metabolism and reported QTLs for coat characteristics, immune response, and tick
resistance. The results suggest the complex mechanism in the adaptation of Bolivian Creole cattle to the
contrasting highland and subtropical environments.
INTRODUCTION
In 1493, Spanish conquerors brought Creole cattle to the American continent (Primo, 1992). They were
initially introduced to the Caribbean islands and then transported to South and North America during the
rst half of the 16th century. The routes to South America included journeys from the Caribbean to the
northern coast (actual Colombia and Venezuela), as well as routes from Central America via the Pacic
coast to the Alto Perú, then extending through the todays Bolivia and Paraguay, and further south to the
pampa and Patagonia (Wilkins, 1984; Primo, 1992; Felius, 1995). Concurrently, Portuguese cattle were
directly shipped to the actual Brazilian territory (De Alba, 1978; Primo, 1992). These animals quickly
adapted to the continent’s diverse environmental conditions, leading to an exponential increase in their
population. Nowadays, local Creole cattle breeds persist in nearly all American countries
(http://www.ansi.okstate.edu/breeds/cattle/). In Bolivia, several Creole cattle populations have adapted
to multiple environments such as seasonal oodplains, dry forests, tropical plains, temperate valleys,
and highland plains. Highland Bolivian Creoles are predominantly located in the western region of the
country, spanning the Departments of Oruro, Potosi and part of La Paz, Cochabamba, and Chuquisaca.
These animals are raised at altitudes ranging from 2,500 to 4,200 metres, characterised by a cool and
dry environment. In contrast, lowland Creoles inhabit the eastern region, which includes tropical
savannahs and subtropical dry forests within the Departments of El Beni, Santa Cruz, and parts of La

Page 4/25
Paz, Cochabamba, and Chuquisaca, situated at altitudes below 500 metres above sea level
(https://senamhi.gob.bo).
The adaptation process to a particular environment requires the development of specic physiological
and morphological characteristics. For example, high altitude environmental conditions could include
low oxygen levels, cold temperatures, high UV exposure and limited food availability (Friedrich and
Wiener, 2020). In contrast, the adaptation to warm tropical and subtropical environments primarily
requires heat tolerance, where changes in coat features and vascularization of the dermis are benecial.
Additionally, the development of resistance to parasites, especially ticks, and other infectious diseases is
crucial (Barendse, 2017). Adaptation is the result of evolutionary and ontogenetic events that contribute
to genetic changes over generations.
Selection is one of the most important genetic processes that cause changes in specic genomic
regions, consequently creates unique genetic patterns or footprints known as selection signatures
(Mignon-Grasteau
et al.
, 2005; Saravanan
et al.
, 2020; Falchi
et al
., 2023). Identifying these signals can be
useful for determining genes and benecial mutations that occur in a given population, including those
adapted to different environments (Zhao
et al
., 2015). With the emerging era of genomics and the advent
of high-density SNP arrays, next-generation sequencing (NGS) technologies, and bioinformatics tools,
various methods have been developed to detect regions subject to selection in multiple species. These
include within-population approaches based on linkage disequilibrium (LD; iHs, rEHH, and LDD), site
frequency spectrum (Tajima’s D and Fay and Wu’s H), and reduced local variability (runs of homozygosity
[ROH]), as well as between-population statistical methods such as single-site differentiation (FST) and
differentiation based on haplotypes (XP-EHH) (Qanbari
et al
., 2010; Gautier and Vitalis, 2012; Saravanan
et al
., 2020).
In recent years, there has been increased interest in identifying selection signatures for high-altitude and
tropical adaptation in livestock species. The Qinghai-Tibet and Bolivian Altiplano Plateausare among the
most extreme high-altitude regions in the world, making them idealmodels for studying high-altitude
adaptation in a diverse range of native species, including humans (Yi
et al
., 2010; Peng
et al
., 2011; Xu
et
al
., 2011; Lorenzo
et al
., 2014), domestic animals (Qiu
et al.
, 2012; Li
et al
., 2014; Wang
et al.
, 2015; Ma
et
al.
, 2019b), and wildlife (Cai
et al.
, 2013; Ge
et al.
, 2013). Conversely, selection signatures have also been
extensively studied in cattle bred in tropical environments in Africa (Tijjani
et al.
, 2022; Kambal
et al.
,
2023), Asia (Nayak
et al.
, 2024) and South America (Maiorano
et al.
, 2018). Over the last ve centuries, it
is expected that Bolivian cattle residing at high and low altitude have developed physiological strategies
and morphological features to adapt to the harsh conditions of the Altiplano Plateaus and tropical
environment, respectively.
Furthermore, highland Bolivian Creole cattle are traditionally utilised by local communities for
subsistence farming, primarily in milk and cheese production. In recent decades, Holstein cattle from
Argentina and Uruguay have been introduced into the Bolivian highlands to improve dairy production.
However, this European breed exhibits low adaptability to altitude and a high mortality rate, often

Page 5/25
attributed to high altitude pulmonary hypertension (HAPH) resulting from chronic exposure to hypoxia or
other stress factors prevalent at high altitude (Wang
et al.
, 2018). Consequently, local community oral
communications suggest the crossbreeding between Creole and Holstein resulted in cattle adapted to
high altitude with increased production.
The objectives of the present study were: (1) to identify selection signatures in the genome of Bolivian
Creole cattle considering their divergent adaptation to highland and tropical lowland environments. For
this purpose, a microarray dataset was analysed using a population differentiation method (FST), two LD-
based methods (Rsb and XP-EHH), and a local variability reduction method (ROH); and (2) to evaluate
whether the percentage of Holstein ancestry in regions under selection, differs from the average
genome-wide estimated proportion.
MATERIALS AND METHODS
Animal samples
Hair and blood samples were collected from 130 Bolivian Creole cattle from Altiplano highlands and
tropical lowlands regions. Highland Bolivian Creoles (HBC; n = 75) were sampled at three locations in the
departments of La Paz, Oruro, and Cochabamba at an altitude of 3,700 - 4,000 metres above sea level
(Table 1 and Fig. 1). Environmental conditions include average annual precipitation of 300 - 400 mm,
concentrated in summer (between January and March), and a media temperature of 8°C, ranging from
0°C to 20°C (https://senamhi.gob.bo). Lowland Bolivian Creoles (LBC; n = 55) samples were collected
from three different sites in the vast plain of the department of Santa Cruz (Table 1 and Fig. 1). These
animals live in subtropical conditions at an altitude of 200 - 500 metres above sea level, with an average
annual rainfall of 800 - 1000 mm and a mean temperature of 25.4°C, ranging from 16°C to 32°C
(https://senamhi.gob.bo). In addition, DNA samples from Holstein (Ho; n = 88) and Zebu breeds (ZEB;
Brahman, n = 45 and Nellore, n = 4) were included. The Institutional Committee on Care and Use of
Experimental Animals (CICUAL) from the School of Veterinary Sciences of the National University of La
Plata (Buenos Aires, Argentina) reviewed and approved all animal procedures (89-1-18T CICUAL).
Table 1. Sampling sites of Highland Bolivian Creole (HBC) and Lowland Bolivian Creole (LBC).
Population Sampling site n Altitude (MSL)
HBC San Pedro de Totora, Oruro Department 52 4,000
 Bolivar, Cochabamba Department 14 3,735
 Omasuyos, La Paz Department 9 3,840
LBC Palmar Tapera, Santa Cruz Department 29 230 - 380
 Chiquitos, Santa Cruz Department 16 416
 Obispo Santistevan, Santa Cruz Department 10 450

Page 6/25
DNA extraction and genotyping
Genomic DNA was extracted from hair samples using the DNeasy Blood & Tissue Kit (Qiagen, Hilden,
Germany) and from blood using theWizard Genomic DNA Purication kit (Promega, MA, USA). Samples
were genotyped in a GeneTitanTM platform (Applied Biosystems™, CA, USA) using the microarrays
AxiomTM Bos1 Genotyping Array r3 (Applied Biosystems™) containing 648,855 SNPs and the custom
array ArBos1 containing 58,088 SNPs. A single genotype matrix was constructed using the merge
function in PLINK v1.9 software (Purcell
et al
., 2007), resulting in a nal database of 48,360 common
SNPs. Raw data were processed using AxiomTM Analysis Suite software 4.0 (Applied Biosystems™),
setting sample and SNPs call rates at ≥ 97%. Datasets were exported in .PED and .MAP le format for
further analysis. The SNPs positions were assigned according to the bovine genome assembly UMD 3.1.
Population structure and relationships
Population structure and the degree of admixture were determined in HBC, LBC, Ho, and ZEB. SNPs were
ltered using the --indep 50 5 2 command in PLINK v1.9 resulting in a set of 21,834 unlinked genetic
markers. Considering the historical data of a possible introgression of foreign genetics in Bolivian Creole
cattle, the admixture with Ho and ZEB breeds was tested by a Bayesian clustering-model implemented in
fastSTRUCTURE (Raj
et al
., 2014). The ChooseK algorithm indicated that the model complexity that
maximises marginal likelihood was K2, and the model components used to explain structure in data was
K3. The graphical representation of the results was performed using Distruct v.2.3 (Chhatre, 2018).
Additionally, a principal component analysis (PCA) was performed to assess the divergence of
individuals from each population using the function --pca in PLINK v1.9. The results were visualised
using the R library ggplot2.
Inference of local ancestry
To infer Ho local ancestry within the HBC genome, haplotype phasing was rst performed using
SHAPEIT5 software (Hofmeister
et al
., 2023). Then, Flare (Browning
et al
., 2023) was used to determine
ancestry origin (Creole cattle or Ho) for every SNP in each HBC individual using the default parameters
with the exception of em=false. Local ancestry was estimated at chromosome and whole-genome
levels. 
Selection footprint analysis
To detect candidate regions for high altitude and tropical lowland adaptation, selection footprints were
analysed using four methods: the xation index statistic (FST; based on gene frequencies), Rsb and XP-
EHH (based on haplotype extension), and ROH (a local variability reduction method; Sabeti
et al.
, 2002,
2007; Voight
et al.
, 2006; Saravanan
et al
., 2020).
The average FST was calculated using the --fst command in PLINK v1.9, and the pairwise FST was
estimated using the OutFLANK R package for LBC, HBC, and Ho (Whitlock and Lotterhos, 2015). This

Page 7/25
script calculates the divergence in individual SNP allele frequencies between pairs of populations using
the Lewontin-Krakauer FST method with an accurate null model. Signicance was determined by
examining the right tail of the null model distribution, with a p-value of <0.01 considered signicant.
Genomic window sizes were determined based on regions containing signicant SNPs separated by no
more than 0.25 Mb. For extreme SNPs, an additional 0.25 Mb was included on the external side of the
window. The FST values were visualised in a Manhattan plot using the
qqman
package in RStudio v4.3.3.
For LD-based analysis, the obtained haplotypes using SHAPEIT5 were used to compute Rsb (Voight
et
al.
, 2006) and XP-EHH (Sabeti
et al
., 2007) using the
rehh
R package (Gautier and Vitalis, 2012). Rsb and
XP-EHH statistics were designed to detect regions with high levels of haplotype homozygosity over an
unexpectedly long distance (relative to neutral expectations) and measure the amount of extended
haplotype across populations. The 1% top value was set as the threshold to select the potential SNPs
under selection. Genomic window sizes were determined based on regions containing signicant SNPs
separated by no more than 0.25 Mb. For extreme SNPs, an additional 0.25 Mb was added to the external
side of the window.
A R-script developed by Gorssen
et al
. (2021) was used to detect ROH in HBC, LBC, and Ho. ROH
incidence was calculated as the percentage of animals with a SNP within a ROH segment for a given
population. The minimal threshold for detection of ROH islands was set to 30%, meaning that a ROH had
to be present in at least 30% of the animals of each population to be included in a ROH island
(Supplementary Table S1). Results were visualised in Manhattan plots using the
qqman
R package.
SNPs located within the ROH island were ltered using the --from-bp --to-bp command in PLINK 1.9 and
the linkage disequilibrium (LD) between the SNPs included in detected windows was estimated with the
r2 parameter and visualised using the Haploview 4.2 software (Barrett
et al.
, 2005).
Gene ontology
The regions identied by each selection footprint methodology were compared to determine the
overlapping windows, and then the positions were converted to the ARS-UCD 1.2 reference genome
assembly using the Lift Genome Annotations from UCSC (https://genome.ucsc.edu/cgi-bin/hgLiftOver).
Genes and QTLs included in the common windows were retrieved from the National Center for
Biotechnology Information (NCBI, https://www.ncbi.nlm.nih.gov), FAANGMINE
(https://faangmine.rnet.missouri.edu) and Cattle QTL Database https://www.animalgenome.org/cgi-
bin/QTLdb/index). Functional analysis was performed using DAVID (https://david.ncifcrf.gov) to detect
biological enrichment pathways and terms. In addition, the individual functional signicance of some
genes was reviewed based on literature.
RESULTS
Population structure and relationships

Page 8/25
To evaluate the genetic composition within the studied Creole populations and the gene introgression of
foreign breeds (Ho and ZEB), two analyses were performed. The results from the cluster analysis using
fastSTRUCTURE are presented in Fig. 2a. When considering K = 2, the Taurine/Zebuine genetic
component was evidenced, with negligible levels of ZEB component in Bolivian Creole cattle, probably
due to the origin of Nelore and Brahman from Creole dams and/or some Zebuine introgression into
Creole. The ZEB estimated average percentage was 3.08% (±4.75%) for HBC and 8.76% (±3.82%) for
LBC. In K = 3, this ZEB inuence was indistinguishable. Instead, two Taurine components were observed,
one common to HBC and LBC and the other corresponding to Ho. Noteworthy, the HBC showed an
average Ho introgression of 31.26% (±12.15%), ranging from 6.33% to 50.31%.
The PCA analysis showed that PC1 accounted for48.39% of the variance and clearly differentiated the
ZEB and the three Taurine populations. PC2 explained 10.97% of the genetic variance and separated the
individuals from LBC, HBC, and Ho (Fig. 2b). In particular, HBC appeared between Ho and LBC, which
could be related with a possible introgression of Ho genes into this population, in agreement with the
cluster analysis results. PCA without ZEB animals endorsed this result (Supplementary Fig. S1). Most
individuals from each population differentiated with both components suggesting the presence of
enough population structure necessary to perform further selection footprint analysis.
Inference of local ancestry analysis
The inference of local ancestry for each SNP in HBC showed an average Ho introgression value of 0.357,
which was similar to fastSTRUCTURE results. This analysis also showed uneven percentages of Ho
origin among chromosomes, with values ranging from 0.14 (BTA2) to 0.63 (BTA20). This disparate
distribution was also observed within chromosomes, with enriched regions (>0.5) in BTA20, BTA15 and
BTA13, among others (Fig. S2). Few of the regions enriched with Ho ancestry coincide with peaks found
in the HBC-LBC comparison, particularly the one on BTA20. Notably, this region (34-44 Mb) contains
milk-related QTLs, such as milk yield, milk fat and protein percentages, and milk acid content, all of which
were found in Holstein cattle (https://www.animalgenome.org/cgi-bin/QTLdb/index).
Footprint analysis based on loci genetic divergence
The rst approach to detect evidence of selection was performed via genetic differentiation (FST)
between pairs of populations: HBC vs LBC, HBC vs Ho and LBC vs Ho. These analyses consider that a
statistically signicant FST value at a locus, compared to the genomic average, could indicate evidence
of selection. The FST estimations for the HBC-LBC, HBC-Ho, and LBC-Ho comparisons resulted in
average weighted values of 0.033, 0.049 and 0.085 respectively. These results reinforce those from PCA
and fastSTRUCTURE and were expected considering the genetic distances between the breeds and
historical data.
The results for the pairwise FST index are presented in Fig. 3a-c and Supplementary Table S2.
Considering the p-value right tail threshold of 0.01, this analysis identied 1,081 SNPs for HBC-LBC, 915
SNPs for HBC-Ho and 695 SNPs for LBC-Ho. These corresponded to FST values ≥ 0.161, 0.248, and

Page 9/25
0.418, respectively. In all three comparisons, signicant SNPs were found across all chromosomes,
encompassing a large number of genes (9,373 for HBC-LBC, 7,993 for HBC-Ho and 6,713 for LBC-Ho). As
shown in Fig. 3, the most signicant peak was observed for HBC-LBC in BTA 20 which also corresponds
to the largest SNP window. Remarkably, this region included the candidate gene proposed for slick
phenotype (Littlejohn
et al.
, 2014).
Footprint detection by haplotype-based analysis
Rsb and XP-EHH LD-based methods were also used to detect selection footprints. Results for the six
pairwise comparisons are presented in Fig. 4 and 5 and Supplementary Table S3. In the HBC vs LBC
comparison, the Rsb method identied 444 SNPs distributed along 16 chromosomes, while the XP-EHH
method detected 480 markers distributed across 18 chromosomes using a 1% threshold. The most
signicant peaks, observed in BTA20 and BTA23, were consistent across both Rsb and XP-EHH
methods. The region on BTA20 included the gene previously reported for slick hair phenotype (PRLR,
Littlejohn
et al.
, 2014), while BTA23 included the major histocompatibility complex genes harbouring
BoLA-DQA, BoLA-DQB and BoLA-DRB. Three prominent peaks in BTA2, BTA5 and BTA10 were also
observed in this comparison using XP-EHH. For HBC vs Ho, Rsb showed 446 SNPs distributed in 17
chromosomes and the most prominent peaks were located in BTA20 and BTA23, followed by several
discrete peaks in BTA2, BTA5, BTA9, BTA10, BTA16 and BTA26. The XP-EHH method detected 446 SNPs
in 19 chromosomes; the most prominent peaks were located in BTA20 and BTA23, as observed in Rsb,
and also in BTA5 and BTA26. For LBC vs Ho comparison, Rsb showed 441 SNPs distributed in 28
chromosomes, and the signicant peaks were spotted in BTA20 and BTA23, followed by several discrete
peaks in multiple chromosomes. The XP-EHH method detected 441 SNPs in 23 chromosomes, being the
most prominent peaks in BTA20 and BTA23 and other discrete peaks in several chromosomes. These
results were expected given the larger pairwise genetic distance.
Footprint detection by ROH analysis
The analysis, conducted using an R script developed by Gorssen
et al
. (2021), revealed predominantly
low levels of ROH incidence across populations (<30% of the analysed individuals of each population),
with the exception of a notable ROH island observed in LBC on chromosome 20, around 36-40 Mb. This
island encompasses 106 SNPs and was found in 19 out of 47 animals (40.43%). Within this window, 47
genes were annotated in the ARS-UCD 1.2 cattle genome reference, including the Slick candidate gene
(Fig. 6b and Table S4). Noteworthy, this ROH corresponded to the selection footprints observed when
LBC and HBC were compared using Rsb and XP-EHH interpopulation methods. Neither HBC nor Ho
presented ROH islands above the dened threshold (Fig. 6a, c and Table S4). The LD analysis of this LBC
window showed 27 blocks that included multiple haplotypes with high levels of LD among them. As
expected, this window exhibited low LD values in HBC except for the region between 38.41 and 38.60
Mb. Interestingly, this short LD block was also observed in Ho (Fig. 7a-c).
Common regions between methods

Page 10/25
When comparing Rsb-XP-EHH, 16, 29 and 27 windows were found in HBC vs LBC, HBC vs Ho, and LBC vs
Ho, respectively (Supplementary Table S3). For HBC vs LBC, the genomic regions were distributed in 10
chromosomes, comprising 23.67 Mb of the total genome, and contained 292 genes and 2,814 reported
QTLs (Supplementary Table S3). The main signals were observed in BTA20 and BTA23 while other
chromosomes (BTA 2, 5, and 10) also contained signicant peaks (Fig. 4a and 5a). Remarkably, this was
in agreement with the higher FST values for HBC vs LBC and the homozygosity island observed for LBC in
BTA20 (Fig. 3b and 7b). For HBC vs Ho, the regions spanned a total of 23.05 Mb distributed in 13
chromosomes and included 559 genes and 3,904 reported QTLs (Supplementary Table S3). The main
peaks were observed in BTA23, followed by BTA20, 2, 5, 11, 13, 16, and 26 (Fig. 4b and 5b). Finally, the
analysis of LBC vs Ho evidenced signicant SNP windows in 16 chromosomes and covered 21.99 Mb.
These regions included 274 genes and 4,898 reported QTLs with the main peaks observed in BTA20, 23,
26, and 5 (Fig. 4c, 5c and Supplementary Table S3). In these last comparisons (HBC vs Ho and LBC vs
Ho) the FST results did not show a clear peak although there were signicant values across all
chromosomes (Fig. 3b, c).
The Venn diagram detailed the regions shared between the three comparisons analyses (Supplementary
Fig. S3). Although most peaks were observed in the same chromosomes, there was no common window
including the three of them. Two, four and ten windows were shared by two comparisons
(Supplementary Fig. S3), which could indicate footprints related to highland and subtropical adaptation,
and the difference in historical origin and selection process between Holstein and Creole cattle. As
expected, the region containing the Slick candidate gene (PRLR) was detected when LBC was compared
to the other two populations, which could explain their adaptation to warm climate.
Functional analysis
The functional analysis was performed using the genes located in the common regions for HBC and LBC
comparison. Most of the detected signicant enriched KEGG pathways were related to immune response
and to autoimmune and infectious diseases, including antigen processing and presentation, Th1 and Th2
cell differentiation, Th17 cell differentiation, among others. Also, pathways related to glutathione and
retinol metabolisms were observed (Table S4). Furthermore, fourteen signicant terms for biological
processes related to immune system function and metabolism (glutathione, retinol, and steroid; Table
S4) were observed.
DISCUSSION
Creole cattle were brought to the American continent by Spanish conquerors in 1493. They were initially
introduced to tropical environments in the Caribbean islands and then spread across the continent
(Primo, 1992). Towards the south, Lima (Peru) was the main focus of dispersal and they crossed to
western Bolivia, a region characterised for having one of the three main plateaus across the world, the
Andean Plateau, located at ~3,000 m above the sea level (Primo, 1992; Friedrich and Wiener, 2020).
These high-altitude environments have lower barometric pressure and ~40% lower atmospheric oxygen

Page 11/25
pressure than the sea level, colder temperatures, and increased UV radiation. As a result of long-term
evolution, Creole cattle from the highland developed several strategies and mechanisms to evolve
variations in the cardio-respiratory system, metabolic pathways, and morphological traits in order to
adapt to their local environment at high altitude (Li
et al.
, 2021). During the XVI century, the migration
continued from the west highland to the tropical and subtropical plains in eastern Bolivia. This region
required the adaptation to new environmental conditions and the development of resistance to specic
diseases (e.g. tick, babesiosis, anaplasmosis) and heat stress, and changes in phenotypic features like
coat, as shorter and thinner hair is benecial in warm climates (da Silva
et al
., 2003; Bayssa
et al
., 2021).
Adaptive evolutionary mechanisms may involve changes in one or few loci with major effects or in
multiple loci with small effects (Storz
et al
., 2010; Friedrich and Wiener, 2020). These genotypic, and
consequently phenotypic, variations were essential to allow them to survive and develop in these harsh
habitats (Storz
et al.
, 2010; Aguirre-Rofrio
et al.
, 2019; Friedrich and Wiener, 2020; Rojas-Espinosa
et al.
,
2023; Alvarez-Cecco
et al.
, 2024).
To assess selection footprints, it is essential to propose a correct hypothesis and select the appropriate
population to validate it. For this reason, rstly the population structure and purity of HBC and LBC were
evaluated. The fastSTRUCTURE and PCA clearly differentiate both populations. The cluster analysis
showed 3-8% of common ancestry between zebu breeds (Nelore and Bahman) and Bolivian Creole
populations. This could be a consequence of a Zebu introgression on Creole cattle and/or the
foundational origin of those Zebu breeds that include absorption of Creole dams. Lirón
et al
. (2006)
reported Zebu introgression using autosomal (4-8%) and holandric (≈10%) microsatellites in Bolivian
Creole cattle. Additionally, HBC presented high levels of Ho introgression which could be related to the
introduction of Holstein animals into the Highland region according to oral history. This crossbreeding
strategy was carried out in order to improve milk production since pure Holstein individuals exhibited
high mortality rates in this highland environment. Considering this context, HBC-Ho and LBC-Ho
comparisons were added in the current footprint analysis to identify regions under selection due to Ho
introgression.
The analyses used for selection footprints and gene ontology were interpreted together in order to nd
common regions and reveal the potential mechanisms involved in divergent adaptation to highland and
subtropical environments. The most prominent common region was located in the centre of BTA20. This
region contains the candidate gene responsible for the slick-hair phenotype often found in Creole cattle
breeds (Olson
et al.
, 2003; Porto-Neto
et al.
, 2018; Sosa
et al.,
2021; Sosa
et al.
, 2022). Littlejohn
et al
.
(2014) rst reported a mutation in the prolactin receptor (PRLR) responsible for the slick-hair phenotype
in purebred Senepol caused by a single base deletion [20:39136558 GC > G] in exon 10 that introduces a
stop codon and results in a truncated protein. Additional studies in South American Creole cattle breeds
from warm environments found that individuals with slick-hair phenotype were discordant with this
reported frameshift mutation. Instead, three nonsense variants leading to stop codons and a SNP which
produced a synonymous mutation were found, all in the same region of the PRLR sequence encoding the
cytoplasmic portion of the protein receptor (Porto-Neto
et al.
, 2018). All these collectively termed slick
mutations generate truncated proteins with nearly identical effects on the protein function. Although the

Page 12/25
specic mechanism by which these mutations alter prolactin signalling is not known, they appear to
enhance the inhibition of hair growth caused by the prolactin (Sosa
et al.
, 2022). Therefore, the slick-
haired animals are characterised by a short sleek hair coat and lower follicle density. This feature
confers superior ability to thermoregulate under heat stress conditions, through heat loss from skin
convection and conduction (Olson
et al.
, 2003; Porto-Neto
et al.,
2018; Florez-Murillo
et al.,
2021; Sosa
et
al.
, 2021). This selection sweep was detected when LBC was compared to HBC and Ho, but was not
observed in HBC-Ho. In LBC, this chromosomal region exhibited higher LD and lower haplotypic diversity
than HBC, evidencing the positive selection of slick-hair phenotype in Creole cattle from eastern Bolivia.
Moreover, considering the reported QTLs within this region, the matched peaks of selection footprints
and Ho ancestry may indicate selection towards dairy production while maintaining the long hair
necessary to adapt to western highland Bolivia. Other candidate genes found in this region were
SLC45A2
,
HSPB3,
and
DNAJC21
.
SLC45A2
is involved in the melanin synthesis and is associated with
skin and coat pigmentation variation in several species (Mariat
et al.
, 2003; Soejima
et al.
, 2007; Dooley
et al.
, 2013; Wang
et al
., 2016; Bâlteanu
et al.
, 2021). Ding
et al.
(2022) found a relationship between the
different alleles of
SLC45A2
and heat tolerance in indigenous Chinese cattle. Meanwhile,
HSPB3
and
DNAJC21
are heat shock proteins (HSPs) and have been related to thermotolerance through
association, selection footprint and transcriptomic studies (Otto
et al.
, 2019; Lemal
et al.
, 2023; Wang
et
al.
, 2024; Alvarez-Cecco
et al.
, 2024). In agreement, QTLs for coat texture and hair length were reported
in these regions (https://www.animalgenome.org/cgi-bin/QTLdb/index). These results support the
divergent adaptation related with the local environment of Bolivian Creole cattle, while short slick-hair is
benecial for heat loss through skin convection and conduction in lowland subtropical environments,
long hair is desirable for highland Creole cattle which are exposed to colder temperatures.
Considering environmental microorganisms and adaptation, livestock in high or lowland areas are
exposed to different pathogens. It has been reported that the diversity and composition of the skin
microbiome, which is associated with animal’s health, is different when comparing high and lowland
adapted individuals (Zeng
et al.
, 2017; Sun
et al.
, 2019; Ma
et al
., 2019a). In this sense, it was expected
that the second main peak was detected in the Bovine Lymphocyte Antigen (BoLA) system located in
BTA23. This selection sweep included Class I (BoLA
Class I
,
BoLA-NC1
,
TRIM
,
JSP.1
) and Class IIa (e.g.,
BoLA-DQA
,
BoLA-DQB
,
BoLA-DRB3
) genes and QTLs related to immune response, infectious diseases,
and parasite load. Previous works have demonstrated association between different resistance to
specic diseases and the allelic variability of BoLA genes. Particularly,
BoLA-DRB3
alleles were widely
associated with infectious diseases, such as virus-induced lymphoma and proviral load in bovine
leukemia virus (BLV) infection, somatic cell count in milk in mastitis, endo and ectoparasites, immune
response traits, response to vaccination and production traits (e.g., milk yield). Furthermore,
BoLA-DQA1
was associated with proviral load in BLV infection and mastitis (Reviewed in Takeshima and Aida, 2008;
Aida
et al.
, 2015). As mentioned above, this observed peak on BTA23 could be due to the differential
exposure to pathogens. Cattle from tropical regions exhibit high resistance to infestation by
ectoparasites (e.g., ticks; Ortega
et al.
, 2023) and endoparasites (e.g., anaplasma, babesia; Casa
et al
.,
2023) while highland Creole cattle are more susceptible, particularly evidenced when they are moved to

Page 13/25
tropical plains. Additionally, one of the main features of the BoLA region is the gene copy number
variation. Qiu
et al
. (2012), studying the high altitude adapted Yak species, proposed that the presence of
multicopy genes plays a relevant role in the divergent genetic architecture of adaptation, particularly in
the functional categories ‘olfactory sensation’ and ‘host defence immunity’. Remarkably, this region on
BTA23 contained several genes that belong to the olfactory receptor family (
OR
genes) and reported
QTLs related to immune response, tick resistance and disease susceptibility
(https://www.animalgenome.org/cgi-bin/QTLdb/index). Other peaks included genes involved in immune
response, located on BTA20 (
IL7R
,
IL6ST
,
IL31RA
,
C6
,
C7, OSMR, LIFR
), and BTA5 (
STAT6
,
NKG2A
, IRAK4,
KLR
and
CLEC
genes). GO analysis evidenced signicant enriched KEGG pathways and biological
processes related to adaptive immunity and immune response to multiple diseases.These ndings
support the hypothesis that Bolivian Creole cattle from highland and tropical environments present
divergent adaptation in response to the differential exposure to pathogens.
The adaptation to highland or tropical habitats also involves metabolic pathways and morphological
traits. Reduced oxygen availability is the primary stressor of high altitude conditions and restricts the
correct functioning of respiratory and cardiovascular systems (Ivy & Scott, 2015). Moreover, chronic
hypoxia increases the production of reactive oxygen species (ROS; Wang
et al.
, 2024). The reduction of
the overall metabolic rate and modications of the oxygen cascade and haematological system, such as
red blood cell count and amount of haemoglobin, are necessary to cope with low oxygen levels
(Hochachka
et al
., 1996; Weber, 2007;
Stortz et al
., 2010). In addition, smaller body size decreases the
energy demands (Friedrich and Wiener, 2020). Experimental works have demonstrated that hypoxia
upregulates the expression of
ANXA2
and
NDUFA4L2
genes, found in BTA10 and BTA5 respectively,
through the direct action of hypoxia-inducible factor-1 (
HIF-1;
Huang
et al.
, 2011; Liu
et al.
, 2021).
Remarkably, the HIF gene family was extensively associated with altitude adaptation in livestock species
and other mammals (Reviewed in Friedrich and Wiener, 2020). It has also been shown that
ANXA2
belongs to a common pathway relevant to brin homeostasis and angiogenesis, while
NDUFA4L2
plays a
key role in the development of pulmonary artery hypertension (PAH; Jacovina
et al.
, 2009; Huang
et al.
,
2011; Hajjar, 2015; Liu
et al.
, 2021). In addition, a candidate gene for haematological parameters,
CPLANE1,
and two vascular endothelial growth factor (VEGF) genes,
NRP1
and
NRP2
, were identied
(Oh
et al.
, 2002; Alghamdi
et al.
, 2020; Yang
et al.
, 2024). While angiogenesis helps to increase blood
ow and oxygen supply under low oxygen conditions in high altitude, this physiological process of
growing new blood vessels in the skin improves the heat dissipation in tropical environments. The
enrichment analysis also evidenced pathways and terms related to glutathione (
GSTA1-5
in BTA23) and
retinol metabolism (
HSD17B6
,
RDH16
and
SDR9C7
in BTA5), which could be indicative of an oxidative
and heat stress response. Previous works evidenced the selection of antioxidase-related genes in
hypoxia-tolerant mammals (Wang
et al.
, 2024). While
GST
gene family plays an important role in cellular
detoxication to reduce the damage caused by ROS, HSD17B6 catalyses the oxido-reduction of different
molecules (Deng
et al.
, 2024). Regarding retinol metabolism, heat stress can decrease vitamin levels,
including retinol, retinoyl β-glucuronide and biotin which have anti-oxidative properties removing ROS
(Yang
et al.
, 2022). Finally, in agreement with previous reports about body size and adaptation to

Page 14/25
different environments, four candidate genes related to height and carcass conformation were spotted in
the sweep selection regions including
PLAG1, CHCHD7
,
CAP2
and
ARL15
(Pureld
et al.
, 2019;
Ghoreishifar
et al.
, 2020; Zhang
et al.
, 2022; Zhao
et al.
, 2022).
In conclusion, the ndings in the present work suggest that the divergent adaptation of Bolivian Creole
cattle populations involves multiple and complex mechanisms. This includes changes in coat features
and other morphological traits, immune response, and metabolic processes such as hypoxia and stress
response. The inference ancestry analysis evidenced an uneven distribution of Ho introgression in the
HBC genome. Except for BTA20, the enrichment regions did not match the selection footprints.
Therefore, these sweeps could be a consequence of divergent adaptation of Bolivian Creole cattle to
highland and tropical lowland environments.
Declarations
ACKNOWLEDGEMENTS
The authors thank the Centro de Investigación Agrícola Tropical (CIAT, Santa Cruz, Bolivia) and Centro de
Ecología Aplicada Simón I. Patiño (Santa Cruz, Bolivia) for providing us with the bovine samples. This
study was funded by the National Council for Scientic and Technical Investigations (CONICET,
Argentina, Grant PUE-2016 N° 22920160100004CO), the National Fund for Scientic and Technological
Investigation (FONCYT-ANPCyT, Argentina, Grant N° PICT-2016-3033), National University of La Plata,
Argentina (Grant V247) and the Fondo Argentino de Cooperación Sur-Sur y Triangular (FO.AR; Grant
6560).
AUTHOR CONTRIBUTION STATEMENT
Conceived and designed the work: MEF, ARM, and GG. Sample collection and data acquisition: JAPR,
ALV, and GG. Analysed the data: OM, PAC, LHO, and FC. Contributed to reagents/materials/analysis tools
acquisition: GG, JAPR, and PPG. Drafted or revised the manuscript: OM, ARM, and GG. Approved the nal
version: OM, PAC, LHO, JAPR, FC, ALV, PPG, MEF, ARM, and GG. 
CONFLICT OF INTEREST
The authors declare no competing interests.
DATA ARCHIVING
The genomic data used in the present study are available at the Open Science Framework platform
(https://osf.io/cs726; DOI 10.17605/OSF.IO/CS726).
RESEARCH ETHICS STATEMENT
All animal procedures were reviewed and approved by the Institutional Committee on Care and Use of
Experimental Animals (CICUAL) from the School of Veterinary Sciences of the National University of La

Page 15/25
Plata (Buenos Aires, Argentina; protocols 89-1-18T, 41.2.14T).
References
1. Aguirre-Riofrio EL, Abad-Guamán RM, Uchuari-Pauta ML (2019) Morphometric evaluation of
phenotypic groups of creole cattle of southern Ecuador. Diversity 11:221
2. Aida Y, Takeshima S, Baldwin CL, Kaushik AK (2015) Bovine immunogenetics. In: The genetics of
cattle. ed. Garrick DJ, Ruvinsky A, Surrey, United Kingdom, pp 153-191.
3. Alghamdi AAA, Benwell CJ, Atkinson SJ, Lambert J, Johnson RT, Robinson SD (2020) NRP2 as an
emerging angiogenic player; promoting endothelial cell adhesion and migration by regulating
recycling of α5 integrin. Front Cell Dev Biol 8.
4. Alvarez-Cecco P, Balbi M, Bonamy M, Rogberg Muñoz A, Olivera H, Giovambattista G, et al. (2024)
Skin transcriptome analysis in Brangus cattle under heat stress. J Therm Biol 10;121:103852.
5. Bâlteanu VA, Cardoso TF, Amills M, Luigi-Sierra MG, Egerszegi I, Anton I, et al. (2021) Red and blond
Mangalitza pigs display a signature of divergent directional selection in the SLC45A2 gene. Anim
Genet 52(1): 66–77
. Barendse W (2017) Climate adaptation of tropical cattle. Annu Rev Anim Biosci 5(1):133–150
7. Barrett JC, Fry B, Maller J, Daly MJ (2005) Haploview: analysis and visualization of LD and haplotype
maps. Bioinformatics 21(2):263-5
. Bayssa M, Yigrem S, Betsha S, Tolera A (2021) Production, reproduction and some adaptation
characteristics of Boran cattle breed under changing climate: A systematic review and meta-
analysis. PloS One 16(5):e0244836.
9. Browning SR, Waples RK, Browning BL (2023) Fast, accurate local ancestry inference with FLARE.
Am J Hum Genet 110(2):326-335
10. Cai Q, Qian X, Lang Y, Luo Y, Xu J, Pan S, et al. (2013) The genome sequence of the ground tit
Pseudopodoces humilis provides insights into its adaptation to high altitude. Genome Biol 14:R29
11. Casa MS, Giovambattista G, Fonteque GV, Miguel EL, Vogel CIG, Miletti LC, et al. (2023) Identication
of Anaplasma marginale, Babesia bovis and Babesia bigemina resistance alleles in Crioulo Lageano
cattle using PCR-SBT and BoLA-DRB3 gene sequencing. Front Vet Sci 10:1256928
12. Chhatre VE (2018) Distruct v2.3, A modied cluster membership plotting script. Available at
http://distruct2.popgen.org
13. da Silva RG, La Scala N, Tonhati H (2003) Radiative properties of the skin and haircoat of cattle and
other animals. Trans ASAE 46(3)
14. De Alba J (1978) Progress in the selection of the Latin American diary Criollo. World Anim Rev FAO
28:26–30
15. Deng T, Liang M, Du L, Li K, Li J, Qian L, et al. (2024) Transcriptome Analysis of Compensatory
Growth and Meat Quality Alteration after Varied Restricted Feeding Conditions in Beef Cattle. Int J
Mol Sci. 25(5):2704

Page 16/25
1. Ding C, Ma J, Yan H, Meng Y, Qi X, Qu K, et al. (2022) Distribution of a missense mutation
(rs525805167) within the SLC45A2 gene associated with climatic conditions in Chinese cattle.
Gene, 835:146643
17. Dooley CM, Schwarz H, Mueller KP, Mongera A, Konantz M, Neuhauss SCF, et al. (2013) Slc45a2 and
V‐ATPase are regulators of melanosomal pH homeostasis in zebrash, providing a mechanism for
human pigment evolution and disease. Pigment Cell Melanoma Res 26(2):205–217
1. Falchi L, Cesarani A, Mastrangelo S, Senczuk G, Portolano B, Pilla F, et al. (2023) Analysis of runs of
homozygosity of cattle living in different climate zones. J Anim Sci 101: skad061
19. Felius M (1995) Cattle Breeds: An encyclopedia. Misset, Doetinchem, the Netherlands
20. Florez-Murillo JM, Landaeta-Hernández AJ, Kim ES, Bostrom JR, Larson SA, Pérez O’Brien, et al.
(2021) Three novel nonsense mutations of prolactin receptor found in heat‐tolerant Bos taurus
breeds of the Caribbean Basin. Anim Genet 52(1):132–134
21. Friedrich J, Wiener P (2020) Selection signatures for high-altitude adaptation in ruminants. Anim
Genet 51(2): 157–165
22. Gautier M, Vitalis R (2012) rehh: an R package to detect footprints of selection in genome-wide SNP
data from haplotype structure. Bioinformatics 28:1176-1177
23. Ge GL, Cai Q, Shen YY, San A, Ma L, Zhang Y (2013) Draft genome sequence of the Tibetan antelope.
Nat Commun 4:1858
24. Ghoreishifar SM, Eriksson S, Johansson AM, Khansed M, Moghaddaszadeh-Ahrabi S, Parna N, et
al. (2020) Signatures of selection reveal candidate genes involved in economic traits and cold
acclimation in ve Swedish cattle breeds. Genet Sel Evol 52:52
25. Gorssen W, Meyermans R, Janssens S, Buys N (2021) A publicly available repository of ROH islands
reveals signatures of selection in different livestock and pet species. Genet Sel Evol 53(2)
2. Hajjar KA (2015) The biology of annexin A2: From vascular brinolysis to innate immunity. Trans Am
Clin Climatol Assoc 126:144–155
27. Hochachka PW, Buck LT, Doll CJ, Land SC (1996) Unifying theory of hypoxia tolerance:
molecular/metabolic defense and rescue mechanisms for surviving oxygen lack. Proc Natl Acad Sci
93:9493–8
2. Hofmeister RJ, Ribeiro DM, Rubinacci S, Delaneau O (2023) Accurate rare variant phasing of whole-
genome and whole-exome sequencing data in the UK Biobank. Nat Genet 55:1243–1249
29. Huang B, Deora AB, He KL, Chen K, Sui G, Jacovina AT, et al. (2011) Hypoxia-inducible factor-1 drives
annexin A2 system-mediated perivascular brin clearance in oxygen-induced retinopathy in mice.
Blood 118(10): 2918–2929
30. Ivy CM, Scott GR (2015) Control of breathing and the circulation in high-altitude mammals and birds.
Comp Biochem Physiol A Mol Integr Physiol 186:66–74
31. Jacovina AT, Deora AB, Ling Q, Broekman MJ, Almeida D, Greenberg CB, et al. (2009) Homocysteine
inhibits neoangiogenesis in mice through blockade of annexin A2–dependent brinolysis. J Clin

Page 17/25
Invest 119(11):3384-94
32. Kambal S, Tijjani A, Ibrahim SAE, Ahmed MKA, Mwacharo JM, Hanotte O (2023) Candidate
signatures of positive selection for environmental adaptation in indigenous African cattle: A review.
Anim Genet 54(6):689–708
33. Lemal P, May K, König S, Schroyen M, Gengler N (2023) Invited review: From heat stress to disease—
Immune response and candidate genes involved in cattle thermotolerance. J Dairy Sci 106(7):4471-
4488
34. Li Y, Wu DD, Boyko AR, Wang GD, Wu SF, Irwin DM, et al. (2014) Population variation revealed high-
altitude adaptation of Tibetan mastiffs. Mol Biol Evol 31:1200–1205
35. Li F, Qiao Z, Duan Q, Nevo E (2021) Adaptation of mammals to hypoxia. Anim Models Exp Medi
4(4):311–318
3. Liron JP, Peral-García P, Giovambattista G (2006) Genetic characterization of Argentine and Bolivian
Creole cattle breeds assessed through microsatellites. J Hered 97(4):331-339
37. Littlejohn MD, Henty KM, Tiplady K, Johnson T, Harland C, Lopdell T, et al. (2014) Functionally
reciprocal mutations of the prolactin signalling pathway dene hairy and slick cattle. Nat Commun
5(1):5861
3. Liu Y, Nie X, Zhu J, Wang T, Li Y, Wang Q, et al. (2021) NDUFA4L2 in smooth muscle promotes
vascular remodeling in hypoxic pulmonary arterial hypertension. J Cell Mol Med 25(2):1221–1237
39. Lorenzo FR, Huff C, Myllymäki M, Olenchock B, Swierczek S, Tashi T, et al. (2014) A genetic
mechanism for Tibetan high-altitude adaptation. Nat Genet 46:951–956
40. Ma Y, Ma S, Chang L, Wang H, Ga Q, Ma L, et al. (2019a) Gut microbiota adaptation to high altitude in
indigenous animals. Biochem Biophys Res Commun 516(1):120–126
41. Ma YF, Han XM, Huang CP, Zhong L, Adeola AC, Irwin DM, et al. (2019b) Population Genomics
Analysis Revealed origin and High-altitude Adaptation of tibetan pigs. Sci Rep 9:11463
42. Maiorano AM, Lourenco DL, Tsuruta S, Ospina AMT, Stafuzza NB, Masuda Y, et al. (2018) Assessing
genetic architecture and signatures of selection of dual purpose Gir cattle populations using
genomic information. PloS One 13(8):e0200694
43. Mariat D, Taourit S, Guérin G (2003) A mutation in the MATP gene causes the cream coat colour in
the horse. Genet Sel Evol 35(1):119–133
44. Mignon-Grasteau S, Boissy A, Bouix J, Faure JM, Fisher AD, Hinch GN, et al. (2005) Genetics of
adaptation and domestication in livestock. Livestock Prod Sci 93(1): 3–14
45. Nayak SS, Panigrahi M, Rajawat D, Ghildiyal K, Sharma A, Jain K, et al. (2024) Deciphering climate
resilience in Indian cattle breeds by selection signature analyses. Trop Anim Health Prod 56(2):46
4. Oh H, Takagi H, Otani A, Koyama S, Kemmochi S, Uemura A, et al. (2002) Selective induction of
neuropilin-1 by vascular endothelial growth factor (VEGF): A mechanism contributing to VEGF-
induced angiogenesis. Proc Natl Acad Sci USA 99(1):383–388

Page 18/25
47. Olson TA, Lucena C, Chase CC, Hammond AC (2003) Evidence of a major gene inuencing hair
length and heat tolerance in Bos taurus cattle. J Anim Sci 81(1):80–90
4. Ortega MF, Giovambattista G, Cutullé C, Santos DD, Nava S, Bonamy M, et al. (2023) Phenotypic
evaluation of genetic resistance to the tick Rhipicephalus (Boophilus) microplus in Argentine Creole
cattle. Ticks and Tick-Borne Dis 14(6):102223
49. Otto PI, Guimarães SEF, Verardo LL, Azevedo ALS, Vandenplas J, Sevillano CA, et al. (2019) Genome-
wide association studies for heat stress response in Bos taurus × Bos indicus crossbred cattle. J
Dairy Sci 102(9):8148-8158
50. Peng Y, Yang Z, Zhang H, Cui C, Qi X, Luo X, et al. (2011) Genetic variations in Tibetan populations
and high-altitude adaptation at the Himalayas. Mol Biol Evol 28:1075-1081
51. Porto-Neto LR, Bickhart DM, Landaeta-Hernandez AJ, Utsunomiya YT, Pagan M, Jimenez E, et al.
(2018) Convergent evolution of slick coat in cattle through truncation mutations in the prolactin
receptor. Front Genet 9:57
52. Primo AT (1992) El ganado bovino ibérico en las Américas 500 años después. Arch Zootec 41:421–
432
53. Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira M, Bender D, et al. (2007) PLINK: A Tool Set for
Whole-Genome 288 Association and Population-Based Linkage Analyses. Am J Hum Genet
81(3):559–575
54. Pureld DC, Evans RD, Berry DP (2019) Rearmation of known major genes and the identication of
novel candidate genes associated with carcass-related metrics based on whole genome sequence
within a large multi-breed cattle population. BMC Genomics 20:720
55. Qanbari S, Pimentel ECG, Tetens J, Thaller G, Lichtner P, Shari AR, et al. (2010) A genome-wide scan
for signatures of recent selection in Holstein cattle. Anim Genet 41(4):377–389
5. Qiu Q, Zhang G, Ma T, Qian W, Wang J, Ye Z, et al. (2012) The yak genome and adaptation to life at
high altitude. Nat Genet 44:946–949
57. Raj A, Stephens M, Pritchard JK (2014) fastSTRUCTURE: variational inference of population
structure in large SNP data sets. Genetics 197(2):573–589
5. Rojas-Espinoza R, Macedo R, Suaña A, Delgado A, Manrique YP, Rodríguez H, et al. (2023)
Phenotypic characterization of Creole cattle in the Andean highlands using bio-morphometric
measures and zoometric indices. Animals 13:1843
59. Sabeti PC, Reich DE, Higgins JM, Levine HZP, Richter DJ, Schaffner SF, et al. (2002) Detecting recent
positive selection in the human genome from haplotype structure. Nature 419:832–837
0. Sabeti PC, Varilly P, Fry B, Lohmueller J, Hostetter E, Cotsapas C, et al. (2007) Genome-wide
detection and characterization of positive selection in human populations. Nature 449:913–918
1. Saravanan KA, Panigrahi M, Kumar H, Bhushan B, Dutt T, Mishra BP (2020) Selection signatures in
livestock genome: A review of concepts, approaches and applications. Livestock Science
241:104257

Page 19/25
2. Soejima M, Koda Y (2007) Population differences of two coding SNPs in pigmentation-related genes
SLC24A5 and SLC45A2. Int J Legal Med 121:36-39
3. Sosa F, Carmickle AT, Jiménez-Cabán E, Ortega MS, Dikmen S, Negrón-Pérez V, et al. (2021)
Inheritance of the SLICK1 allele of PRLR in cattle. Anim Genet 52(6):887-890
4. Sosa F, Santos JE, Rae DO, Larson CC, Macchietto M, Abrahante JE, et al. (2022) Effects of the
SLICK1 mutation in PRLR on regulation of core body temperature and global gene expression in liver
in cattle. Animal 16(5):100523
5. Storz JF, Scott GR, Cheviron ZA (2010) Phenotypic plasticity and genetic adaptation to high-altitude
hypoxia in vertebrates. J Expl Biol 213(Pt 24):4125-36
. Sun G, Zhang H, Wei Q, Zhao C, Yang X, Wu X, et al. (2019) Comparative analyses of fecal microbiota
in European Mouon (Ovis orientalis musimon) and Blue Sheep (Pseudois nayaur) living at low or
high altitudes. Front Microbiol 10:1735
7. Takeshima SN, Aida Y (2008) Structure, function and disease susceptibility of the bovine major
histocompatibility complex. Anim Sci J 77:138-150
. Tijjani A, Salim B, da Silva MVB, Eltahir HA, Musa TH, Marshall K, et al. (2022) Genomic signatures
for drylands adaptation at gene-rich regions in African zebu cattle. Genomics 114(4):110423
9. Voight BF, Kudaravalli S, Wen X, Pritchard JK (2006) A map of recent positive selection in the human
genome. PLoS Biol 4(3):e72
70. Wang MS, Li Y, Peng MS, Zhong L, Zong-Ji W, Li QY, et al. (2015) Genomic analyses reveal potential
independent adaptation to high altitude in Tibetan chickens. Mol Biol Evol 32:1880-1889
71. Wang H, Xue L, Li Y, Zhao B, Chen T, Liu Y, et al. (2016) Distribution and expression of SLC45A2 in the
skin of sheep with different coat colors. Folia Histochem Cyto 54(3):143-150
72. Wang S, Azarfar A, Wang Y, Cao Z, Shi Q, Li S (2018) Hematological and vasodilator characteristics
for high altitude acclimatization in Holstein heifers ascended to high altitude. Asian-Australas J
Anim Sci
73. Wang L, Yan X, Wu H, Wang F, Zhong Z, Zheng G, et al. (2024) Selection Signal Analysis Reveals
Hainan Yellow Cattle Are Being Selectively Bred for Heat Tolerance. Animals 14:775
74. Weber RE (2007) High-altitude adaptations in vertebrate hemoglobins. Respir Physiol Neurobiol
158:132-42
75. Whitlock MC, Lotterhos KE (2015) Reliable Detection of Loci Responsible for Local Adaptation:
Inference of a Null Model through Trimming the Distribution of F(ST). Am Nat 186(Suppl 1):S24-36
7. Wilkins JV (1984) Criollo Cattle of the Americas. Animal Genetic Resources Information No. 1. FAO,
Rome, Italy
77. Xu S, Li S, Yang Y, Tan J, Lou H, Jin W, et al. (2011) A genome-wide search for signals of high-altitude
adaptation in Tibetans. Mol Biol Evol 28:1003–1011
7. Yang S, Liu J, Gu Z, Liu P, Lan Q (2022) Physiological and Metabolic Adaptation to Heat Stress at
Different Altitudes in Yaks. Metabolites 12(11):1082

Page 20/25
79. Yang T, Luo H, Lou W, Chang Y, Brito LF, Zhang H, et al. (2024) Genetic background of hematological
parameters in Holstein cattle based on genome-wide association and RNA sequencing analyses. J
Dairy Sci. 2024. Published online doi:10.3168/jds.2023-24345
0. Yi X, Liang Y, Huerta-Sanchez E, Jin X, Cuo ZXP, Pool JE, et al. (2010) Sequencing of 50 human
exomes reveals adaptation to high altitude. Science 329:75-78
1. Zeng B, Zhao J, Guo W, Zhang S, Hua Y, Tang J, et al. (2017) High-altitude living shapes the skin
microbiome in humans and pigs. Front Microbiol 8:1929
2. Zhang Z, Yang P, He P, Xu J, Lyu S, Liu X, et al. (2022) Distribution and association study of PLAG1
gene between copy number variation and Chinese cattle populations. Anim Biotechnol 33(2):273–
278
3. Zhao F, McParland S, Kearney F, Du L, Berry DP (2015) Detection of selection signatures in dairy and
beef cattle using high-density genomic information. Genet Sel Evol 47(1):49
4. Zhao L, Li F, Yuan L, Zhang X, Zhang D, Li X, et al. (2022) Expression of ovine CTNNA3 and CAP2
genes and their association with growth traits. Gene 807:145949
Figures

Page 21/25
Figure 1
Sampling sites of Highland Bolivian Creole (HBC; red) and Lowland Bolivian Creole (LBC; yellow).
Figure 2
a. Cluster analysis (K2-K3) and b. Principal Component Analysis (PCA) for Highland Bolivian Creole
(HBC), Lowland Bolivian Creole (LBC), Holstein (Ho) and Zebu breeds (ZEB).

Page 22/25
Figure 3
Manhattan plots for the pairwise FST between: a. Highland Bolivian Creole (HBC) and Lowland Bolivian
Creole (LBC); b. Highland Bolivian Creole (HBC) and Holstein (Ho); c. Lowland Bolivian Creole (LBC) and
Holstein (Ho). A threshold of 1% was set to determine signicant FST values (red line).

Page 23/25
Figure 4
Manhattan plots for Rsb between: a. Highland Bolivian Creole (HBC) and Lowland Bolivian Creole (LBC);
b. Highland Bolivian Creole (HBC) and Holstein (Ho); c. Lowland Bolivian Creole (LBC) and Holstein (Ho;
c). A threshold of 1% of the values was set to determine signicant SNPs (black dotted line).

Page 24/25
Figure 5
Manhattan plots for XP-EHH between: a. Highland Bolivian Creole (HBC) and Lowland Bolivian Creole
(LBC); b. Highland Bolivian Creole (HBC) and Holstein (Ho); c. Lowland Bolivian Creole (LBC) and
Holstein (Ho). A threshold of 1% of the values was set to determine signicant SNPs (black dotted line).
Figure 6
Manhattan plots for ROH in Highland Bolivian Creole (HBC; a), Lowland Bolivian Creole (LBC; b) and
Holstein (Ho; c).

Page 25/25
Figure 7
Haploview plots for Highland Bolivian Creole (HBC; a), Lowland Bolivian Creole (LBC; b) and Holstein (Ho;
c).
Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.
FigS1.PCAwithoutzebu.pdf
FigS2.Chromosomes.pdf
FigS3.VennPlot.png
TableS1.ROHparameters.docx
TableS2.FST.xlsx
TableS3.RsbXP.xlsx
TableS4.PathwaysandGOterms.xlsx
SupplementaryMaterial.docx