https://orcid.org/0000-0001-8820-0098
Abstract
Domestic chickens are less fearful, have a faster sexual development, grow bigger, and lay more eggs than their primary ancestor, the red junglefowl. Several candidate genetic variants selected during domestication have been identified, but only a few studies have directly linked them with distinct phenotypic traits. Notably, a variant of the thyroid stimulating hormone receptor (TSHR) gene has been under strong positive selection over the past millennium, but it’s function and mechanisms of action are still largely unresolved. We therefore assessed the abundance of the domestic TSHR variant and possible genomic selection signatures in an extensive data set comprising multiple commercial and village chicken populations as well as wild-living extant members of the genus Gallus. Furthermore, by mean of extensive backcrossing we introgressed the wild-type TSHR variant from red junglefowl into domestic White Leghorn chickens and investigated gene expression, hormone levels, cold adaptation, and behavior in chickens possessing either the wild-type or domestic TSHR variant. While the domestic TSHR was the most common variant in all studied domestic populations and in one of two red junglefowl population, it was not detected in the other Gallus species. Functionally, the individuals with the domestic TSHR variant had a lower expression of the TSHR in the hypothalamus and marginally higher in the thyroid gland than wild-type TSHR individuals. Expression of TSHB and DIO2, two regulators of sexual maturity and reproduction in birds, was higher in the pituitary gland of the domestic-variant chickens. Furthermore, the domestic variant was associated with higher activity in the open field test. Our findings confirm that the spread of the domestic TSHR variant is limited to domesticated chickens, and to a lesser extent, their wild counterpart, the red junglefowl. Furthermore, we showed that effects of genetic variability in TSHR mirror key differences in gene expression and behavior previously described between the red junglefowl and domestic chicken.
Introduction
Animal domestication is the process whereby populations change genetically and phenotypically in response to the selection pressure associated with a life under human supervision (Jensen and Wright 2014). As a result, domestic animals differ from their wild counterparts in a suite of phenotypic traits, including behavior, physiology, and morphology (Núñez‐León et al. 2019; Rosenfeld et al. 2020). The most numerous of all domestic species, the chicken, diverged from the ancestral red junglefowl around 8000 years ago in South Asia (Lawal et al. 2020), and had its genome further influenced through introgression from other Gallus species (Eriksson et al. 2008; Lawal et al. 2020). In comparison to red junglefowl, the domestic breeds are generally more docile, have a faster growth rate, lay more and larger eggs year-round, and reach sexual maturity at an earlier age (e.g., ∼20 weeks for White Leghorn vs ∼25 weeks for red junglefowl) (Schütz et al. 2001; Kerje et al. 2003; Jensen et al. 2005; Jensen 2006; Campler et al. 2009; Ericsson et al. 2014; Núñez‐León et al. 2019). Moreover, while the red junglefowl distribution is restricted to tropical and subtropical Asia (Delacour 1977; Tixier-Boichard et al. 2011), domestic chickens are spread worldwide over diverse regions, including temperate zones.
The extensive between-breed diversity of domestic chickens has been exploited to investigate links between genetic variants and phenotypic traits (Andersson 2012). However, while genome-wide studies have discovered a large number of domestication-related gene variants in the chicken (Rubin et al. 2010; Wragg et al. 2012; Fallahsharoudi et al. 2017; Lawal et al. 2018; Qanbari et al. 2019; Fogelholm et al. 2020), the biological functions and mechanisms of action of many of these variants remain unknown or speculative.
Several genomics studies have shown that a variant of the thyroid-stimulating hormone receptor (TSHR) gene, hereafter called the “domestic” variant, is fixed in the majority of domestic chicken breeds, presumably due to selection during domestication and breed generation (Rubin et al. 2010; Qanbari et al. 2015, 2019; Lawal et al. 2018). Furthermore, it has been shown that the domestic variant is fixed in commercial as well as in several different populations of indigenous village chickens (Rubin et al. 2010; Lawal 2018). Because the TSHR locus has been under extensive selection pressure during domestication, as indicated by the presence of a strong selective sweep, the TSHR has been identified as a prominent domestication-related locus in chickens (Rubin et al. 2010). Moreover, archaeogenetic data suggested that the increase in the allele frequency of the domestic TSHR variant started around 1100 years ago in Europe, coinciding with the intensification of chicken egg and meat production (Flink et al. 2014; Loog et al. 2017).
A nonconservative amino acid substitution in the TSHR, glycine to arginine at residue 558 (Gly558Arg), has been suggested as the potential causal mutation for the selective sweep in commercial domestic chicken breeds (Rubin et al 2010). A recent study, using molecular dynamics simulations, showed that this substitution may affect the TSHR signal transduction through whole-protein perturbation in the helix bundle dynamics (Grottesi et al. 2020). Therefore, the candidate substitution may have important functional consequences for the TSHR functionality and therefore affect dependent downstream processes.
The hypothalamus-pituitary-thyroid (HPT) axis plays a crucial role in the regulation of several physiological processes that are modified in domestic animals, including the onset of sexual maturity, reproduction, energy metabolism, behavior, and regulation of body temperature (Joseph-Bravo et al. 2015, 2017 Ortiga-Carvalho et al. 2016; Müller-Fielitz et al. 2017). Together with the hypothalamic thyrotropin-releasing hormone (TRH) and the thyroid-stimulating hormone (TSH) from the pituitary gland, the thyroid TSHR regulates the HPT axis. Moreover, photo-induced expression of the thyroid stimulating hormone β-subunit (TSHB), iodothyronine deiodinase 2 (DIO2), and Iodothyronine Deiodinase 3 (DIO3) in the pituitary gland affects hypothalamic TSHR and plays a key role in the regulation of onset of sexual maturity and reproduction in birds (Nakao et al. 2008; Yoshimura 2010). Consistent with a putative role of the domesticated TSHR variant in promoting an earlier onset of sexual maturation (Karlsson et al. 2016), juvenile domestic chickens are characterized by higher DIO2 and TSHB gene expression in the pituitary compared to red junglefowls (Fallahshahroudi et al. 2019).
In a previous study involving an intercross between domestic White Leghorn and red junglefowl, the domestic TSHR variant was associated with changes in hormone levels, faster sexual development in both male and female chickens and changes in tameness and aggressive behavior (Karlsson et al. 2015, 2016). Other mechanisms underlying phenotypic differences of TSHR variants, such as in gene expression, metabolism and adaptation to cold temperature, remain largely unexplored to date. At the same time, the low precision of intercrosses in targeting the TSHR locus can be improved with consecutive backcrossings (Luo et al. 2002).
In the present study, we first investigated the geographical distribution and prevalence of the TSHR domestic variant across hundreds of commercial and backyard domestic populations sampled across the world (Malomane et al. 2019), four wild Gallus species, and common pheasant (Lawal et al. 2020). This provided a much more comprehensive data set to test the hypothesis that the domestic TSHR variant is specific to Gallus gallus and is fixed in all domestic chicken populations across the world. We then introgressed the wild-type TSHR variant into domesticated White Leghorns and measured relevant gene expression in the hypothalamus, pituitary and thyroid glands, T3 and T4 levels in the blood, thermoregulation, and behavior. We predicted changes in gene expression, higher thyroid hormone levels, basal metabolic rate, and lower fearfulness in chickens possessing the domestic variant.
Materials and methods
Ethical statement
All experimental protocols were approved by the Linköping Ethical Board of Animal Experiments (permit no 50–13). The experiments were conducted in accordance with the approved guidelines.
Geographical spread and signature of selection in TSHR
We used previously published genotypes in the SYNBREED project (Malomane et al. 2019) to analyze the geographic distribution of variants and further evaluate the evidence for selective sweeps in the TSHR. The SYNBREED chicken diversity panel comprises high density single nucleotide polymorphism (SNP) data from 174 chicken populations from Asia, Europe, South America, and Africa. First, we used SHAPEIT2 v2.r904 (Delaneau et al. 2013) to phase variants in the region containing the TSHR and 2 Mbp of flanking sequence. We used the UCSC Lift Over tool to map the SNP positions to the latest version of the chicken reference genome (GCRg6a). After excluding samples and SNPs with more than 10% missing values, we phased 3047 samples on 1176 SNPs. To investigate the frequency of the domestic TSHR variant within each population, we labeled all variants based on the core region of the TSHR, spanning 14 markers. Assuming that White Leghorn chickens are fixed for the domestic TSHR variant, we labeled all other SYNBREED haplotypes (a proxy for the domestic TSHR variant) within this core region based on whether they exactly matched to any haplotype in White Leghorns. Then, we re-evaluated evidence for a TSHR selective sweep (Rubin et al. 2010) in the SYNBREED sample by plotting observed heterozygosity, and the integrated site-specific extended haplotype homozygosity score iES (Tang et al. 2007), estimated with the rehh R package (version 3.0.1; Gautier and Vitalis 2012). We excluded the White Leghorn population from the iES analysis because of its low-haplotype diversity, which makes haplotypes extend to the border of the region. In order to analyze TSHR diversity in wild junglefowl (details provided in Supplemental File S1), we used previously published sequence data (Wang et al. 2015; Ulfah et al. 2016; Lawal et al. 2020) from red junglefowl (Gallus gallus; n = 11), green junglefowl (Gallus varius, n = 12), gray junglefowl (Gallus sonneratii; n = 3), Ceylon junglefowl (Gallus lafayettii: n = 8), and pheasants (Phasianus colchicus; n = 2).
Study animals
By selective backcrossing, we generated a population of almost pure domestic White Leghorn chickens homozygous for either the wild-type or the domesticated TSHR allele. The details regarding the breeding of this backcross line can be found in the extended material and methods (Supplementary File S1). We generated two backcross batches. The first batch was used for gene expression analysis (n = 24, 12 of each sex) and the second batch was used for investigating behavior, hormonal levels, and metabolic rate (n = 70). Birds were hatched in a small incubator with 37.5°C, 55% relative humidity and regular rotation of the eggs, and were then kept until the end of the experiment, at 6 weeks of age, in a 2 × 3 m pen with 12 h light/day, ad libitum food and water, a littered floor, heat lamp, and perches.
Gene expression
At the age of 6 weeks, we sacrificed 24 individuals (6 for each sex and gene variant combinations) by decapitation. The midbrain, containing thalamus and hypothalamus, as well as the pituitary and the thyroid glands were dissected and immediately snap frozen in liquid nitrogen (Løtvedt et al. 2017). The tissues were, afterwards, stored in −80°C until further processing. In each tissue, we measured expression of TSHR, TSHB, DIO2, and DIO3. Details regarding RNA extraction, RT-qPCR procedures are provided in Supplemental File S1.
Hormones and metabolic rate
Thyroid hormone levels were measured in the blood of 40 birds (10 of each gene variant and sex combinations). All blood samples (around 1 mL, drawn from the brachial (wing) vein into heparinized tubes) were collected at the age of 6 weeks and between 10 and 11:30 in the morning. The samples were stored in ice, and subsequently centrifuged (2000 G for 10 min in a refrigerated centrifuge) to separate plasma from the blood cells. The plasma samples were stored in a −80°C freezer and within a month were used for analysis of TSH (using ELISA), T3 and T4 levels (using UPLC–MS/MS).
When animals were 5––6 weeks, oxygen consumption, a proxy of metabolic rate, was measured overnight while progressively decreasing ambient temperature, from thermoneutral (30°C), to 5°C following previously implemented procedures (Lindholm et al. 2017). Further details of hormonal analyses and oxygen consumption are provided in Supplementary File S1.
Behavior
Behavioral tests were conducted starting at 2 weeks of age, in standard, fully enclosed test arenas (80 × 80 × 100 cm). Video records of the behavior were later analyzed using EthoVision® XT (Noldus, the Netherlands) to measure activity and ranging behavior inside the test arena. Chicks were first assayed in an open field (OF) test to quantify general activity and fearfulness-related behavior (Forkman et al. 2007). One week following the OF test, chicks were then assayed in a social reinstatement (SR) test, to quantify their sociability. Chickens with higher social dependency move toward social companions earlier and spend more time close to them (Suarez and Gallup 1983; Johnsson et al. 2016). Details of the behavioral tests are reported in Supplemental File S1.
Statistical analysis
All statistical analyses were conducted in R version 3.6.1 (R Core Team 2019). Plots were generated using the ggplot2 R package (Wickham 2016). To study the effects of TSHR variant and sex on gene expression, hormones, and behavior, we fitted linear models by robust regression (due to the presence of possible outliers), using the “rlm” function from the MASS package (Venables and Ripley 2002). All models included sex and genotype as predictors; model assumptions were checked and p-values were obtained using the f.robftest () function from the sfsmisc package (Maechler 2017) which compute a robust F-test, i.e., a Wald test for multiple coefficients of the rlm object. For the analysis of metabolic rate, due to the presence of multiple individual measures, we fitted linear mixed-effects model with the lme function from the nlme R package (Pinheiro et al. 2018). The model included metabolic rate as response variable, temperature, variant, sex, and the interaction between variant and temperature as predictors, and random individual intercepts and slopes for temperature.
Data availability
Supplementary File S1 provides the extended material and methods including details of protocols for measuring gene expression, hormone levels, metabolic rate, behaviors, and generation of the intercross and backcross lines. It also contains information related to the population genetics study, used primers for genotyping and qPCR, and accession numbers for the sequence data. Supplementary File S2 includes all data used in the analyses and producing the figures, including genotypes, gene expression, metabolic rate, hormone levels, and all measured behaviors. Supplementary Files S1 and S2 Dataset are stored in the Figshare repository and can be accessed through the following link: https://doi.org/10.6084/m9.figshare.13387913. The SNP data set used in the populations genetics analysis is stored in the Figshare repository and can be accessed through the following link: https://doi.org/10.6084/m9.figshare.8003909.
Results
Haplotype sharing and heterozygosity
We used the available high-density SNP chip genotyping data from the SYNBREED panel to study the prevalence of the domestic TSHR variant in diverse chicken populations (including two red junglefowl populations). Our analysis showed that the domestic TSHR variant is prevalent across all domestic populations, including both commercial and local breeds. Furthermore, the domestic variant was detected at a substantial frequency in two different red junglefowl populations, occurring in different ecological regions in Thailand and representative of two subspecies (G.g. gallus and G.g. spadiceus) (Figure 1A). To confirm the presence of the previously detected TSHR sweep (Rubin et al. 2010), we plotted observed heterozygosity and the variant-based sweep statistic iES in the mega base region around TSHR from the SYNBREED panel. All studied populations showed reduced heterozygosity and a peak in iES at TSHR (chr5: 40973965-41020273), indicative of a selective sweep (Figure 1B). Number of haplotypes (local breeds grouped by broad geographic regions, main commercial breeds and two red junglefowl populations) are provided in Supplementary File S1. Based on available sequencing data from Ceylon, green and gray junglefowl, as well as pheasants, there was no evidence of reduced heterozygosity in the TSHR region in any of these species (Figure 1C). Similarly, the domestic TSHR Gly558Arg missense variant was absent in all the pheasants and junglefowl species, except the red junglefowl, in which it reached a frequency of 27%.
Haplotype sharing and sweep mapping suggest prevalence of selective sweep and domestic variant in all studied populations. (A) Haplotype sharing in the TSHR in the SYNBREED data broken down into African, Asian, European, and South American breeds, White Leghorns, and red junglefowl (captured in Thailand). The bars show counts of haplotypes that are identical to any of the White Leghorn haplotypes in the TSHR, as a proxy for the domestic TSHR variant. (B) Heterozygosity (black lines) and haplotype-based sweep statistic iES (gray lines) in the region around TSHR the SYNBREED data. The black bar shows the location of the TSHR, and the red dot the location of the candidate Gly558Arg missense variant (rs13587540). The iES scores have been standardized by dividing by the largest iES value within the region. The iES test statistic cannot be estimated in White Leghorn due to the lack of haplotype diversity. Wild represents the combination of the two red junglefowl populations in SYNBREED data. (C) Heterozygosity (black lines) in 40 kbp sliding windows and haplotype-based sweep statistic iES (gray lines) in the region around TSHR in sequence data from junglefowl. The black bar and red dot show the location of TSHR and the candidate missense variant. The iES scores have been standardized by dividing by the largest iES value within the region. Note the different vertical axes for heterozygosity and iES.
Gene expression
Expression of TSHR in the hypothalamus was significantly lower in chickens carrying the domestic variant than in those with the wild-type variant (Estimate wild-type vs domestic variant: −0.44 ± 0.1 SE, P < 0.001; Figure 2A). Sex did not affect hypothalamus TSHR expression (Estimate for female vs male: −0.02 ± 0.1, P = 0.85). Neither variant (0.02 ± 0.09, P = 0.83), nor sex (−0.15 ± 0.09, P = 0.11) were related to hypothalamic DIO2 expression (Figure 2A). The expression of DIO3 was significantly lower in the hypothalamus of the domestic-variant chickens compared to wild-type (−0.28 ± 0.13, P = 0.04). Furthermore, hypothalamic DIO3 expression was higher in males than females (0.30 ± 0.13, P = 0.03). Expression of TSHB in the hypothalamus was below the designated detection cutoff and therefore was excluded from the analysis (Ct-value > 32).
Relative gene expression of TSHR, TSHB, DIO2, and DIO3 in HPT axis is affected by the variant of TSHR. “Wild-type” and “Domestic” represent the red junglefowl and White Leghorn variant, respectively. Colored circles represent observed values; black dots and associated bars show estimated mean and 95% confidence intervals. For variant effects: *P < 0.05, **P < 0.01, and NS P > 0.05.
In the pituitary gland, TSHB and DIO2, but not DIO3, were expressed at higher levels in chickens possessing the domestic variant (TSHB: 0.80 ± 0.13, P < 0.001; DIO2: 0.83 ± 0.20, P < 0.001; DIO3: −0.15 ± 0.17, P = 0.39; Figure 2B). Furthermore, expression of pituitary DIO2 was lower in males than females (−0.61 ± 0.20, P = 0.01). However, there was no evidence for sex differences in the expression of TSHB (−0.21 ± 0.13, P = 0.13) or DIO3 (−0.03 ± 0.17, P = 0.85). Expression of pituitary TSHR was lower than cut-off value.
Finally, in the thyroid gland, TSHR was expressed at a marginally higher level in chickens possessing the domestic variant (0.30 ± 0.17, P = 0.1; Figure 2C). However, sex (−0.6 ± 0.17, P = 0.73) did not affect thyroid expression of TSHR. Expression of DIO2 and TSHB were below the detection limit, whereas DIO3 expression was not affected by neither variant (−0.14 ± 0.24, P = 0.57) nor sex (−0.14 ± 0.14, P = 0.56).
Hormone levels and metabolic rate
There was no evidence for any effect of variant (−0.06 ± 0.04, P = 0.17: Figure 3A) or sex (0.04 ± 0.04, P = 0.33) on plasma TSH levels. Thyroid hormone, T3 was marginally lower in domestic variant chickens (−1.2 ± 0.68, P = 0.08), whereas sex did not have any clear effect on T3 (−0.9 ± 0.69, P = 0.19). Plasma levels of T4 were neither affected by variant (−0.92 ± 0.62, P = 0.14) nor by sex (−0.55 ± 0.62, P = 0.39).
Effects of TSHR variant on hormone levels and metabolic rate were not statistically significant. (A) Plasma levels of TSH, T3, and T4. Colored circles represent observed values; black dots and associated bars show estimated mean and 95% confidence intervals. (B) Metabolic rate, presented as mass-specific oxygen consumption (mL O2·g-1·h-1), according to ambient temperature and TSHR variant, in female (left) and male (right) chickens. Colored dots represent observed metabolic rates. Fitted model regression lines and confidence interval ribbons are presented. “Wild-type” and “Domestic” represent the red junglefowl and White Leghorn variant respectively. For variant effects: NS P > 0.05.
Overall, decreasing ambient temperature was associated with a significant increase in metabolic rate (−0.047 ± 0.002, P < 0.001; Figure 3B). However, there was no effect of variant, either alone (0.02 ± 0.08, P = 0.77) or in interaction with temperature (−0.001 ± 0.003, P = 0.88), nor any clear effect of sex (0.06 ± 0.04, P = 0.15).
Behavior
In the OF test, latency to move was not affected by TSHR variant (−1.3 ± 7.4 seconds, P = 0.86; Figure 4A), while males started to move in the arena marginally later than females (12.58 ± 7.24 sec, P = 0.09). However, total distance moved differed between variants, with domestic-variant chicks showing higher total movement than wild-type ones (7.2 ± 2.99 m, P = 0.02; Figure 4B), whereas sex-differences were unsupported (−2.15 ± 2.88 m, P = 0.45). On the other hand, the ratio between time spent in periphery versus center, a proxy for fearfulness, was neither affected by genotype (−7.62 ± 10.9, P = 0.5), nor by sex (−2.63 ± 10.64, P = 0.8).
Domestic TSHR affect activity in the open field (OF) test but does not affect social behaviors in social reinstatement (SR) test. (A) latency of the first step (s), and (B) total distance moved (m) in the OF test. (C) latency of reaching the social zone (s), and (D) total time spent in the designated social zone (s), for the social reinstatement test. “Wild-type” and “Domestic” represent the red junglefowl and White Leghorn variant, respectively. Colored circles represent observed values; black dots and associated bars show estimated mean and 95% confidence intervals. For variant effects: *P < 0.05 and NS P > 0.05.
In the sociability test, TSHR variant was neither a significant predictor of latency to reach the social zone (10.66 ± 11.98 s, P = 0.38; Figure 4C) nor of total time spent within it (2.42 ± 27.31 s, P = 0.93; Figure 4D). However, males reached the social zone later than females (27.18 ± 11.62 s, P = 0.02), whilst there was no evidence that total time spent in the social zone differed between the sexes (−5.1 ± 26.48 s, P = 0.85).
Discussion
We have shown that the domestic variant of the TSHR is widespread in production and regional chicken breeds as well as in red junglefowl populations. Importantly, we demonstrated significant effects of the domestic variant on downstream gene expression throughout the HPT axis, likely to affect development and reproduction. These effects of TSHR-variant further translated into behavioral differences, and resonate with the intense positive selection that has occurred on this locus.
We evaluated the prevalence of the domestic TSHR variant in a large dataset comprising 174 breeds and 3235 individuals, sampled across Europe, Asia, Africa, and South America (Malomane et al. 2019). In agreement with previous studies (Rubin et al. 2010; Lawal et al. 2018), but over a much more comprehensive sample, we found that the domestic variant and selective sweep are present at high frequency in all investigated domestic breeds across the world. Our results further suggest that the domestic variant also exists at intermediate to high frequencies in two wild-caught red junglefowl populations (G. g. spadiceus 28% and G. g. gallus 75%; Figure 1A). This is in agreement with a previous study (Lawal et al. 2018) showing fixation of domestic TSHR locus in four out of six studied red junglefowl populations. Next, to investigate whether the wild-type variant is indeed ancestral, we analyzed additional sequence data from all living species of the Genus Gallus and the common pheasant as an outgroup. We demonstrated that the domestic variant is not present in other junglefowl species, suggesting that it was not widespread in the common ancestor of the Gallus species.
Two scenarios may possibly explain the prevalence of the domestic variant in red junglefowls. As previously suggested (Rubin et al. 2010), the domestic variant may have introgressed into wild populations as a result of hybridization with domestic chickens. Supporting gene flow from domestic chicken into red junglefowl, admixture with domestics has been detected across several red junglefowl populations over a wide geographical range (Mukesh et al. 2013; Gering et al. 2015). Furthermore, it has been shown that red junglefowl populations carry other domestic gene variants such as Beta-Carotene Oxygenase 2 (BCO2), further supporting admixture with domestic chickens (Lawal et al. 2018, 2020). The alternative scenario is that the domestic variant had been present amongst red junglefowl populations ancestral to both domestic chicken breeds and present-day red junglefowls. In line with this scenario, Wang et al. (2020) sequenced red junglefowl from multiple subspecies and found that the G. g. spadiceus is the closest one to the domestic chickens and has the highest frequency (94%) of the domestic TSHR variant, with G. g. gallus at intermediate frequency (50%), and other red junglefowls a low frequency (5.4%), consistent with our results.
Further confirming previous studies (Rubin et al. 2010; Qanbari et al. 2015, 2019), we also found low heterozygosity and a signature of selective sweep covering the TSHR locus (Figure 1B). The region of low heterozygosity (chr5:40973965-41020273) covered the TSHR and overlapped with the previously reported domestication related selective sweep in the TSHR locus (Rubin et al. 2010). However, the observed low heterozygosity in TSHR in the wild populations was unexpected. Therefore, we explored the possibility that rather than positive selection, low heterozygosity in TSHR may be caused by unidentified ancient structural variants such as inversions, which may reduce heterozygosity through recombination suppression (Coughlan and Willis 2019). Hence, we investigated heterozygosity in the TSHR region using sequence data from red and other junglefowl species. However, we did not find any evidence of reduced heterozygosity in TSHR in any of the above-mentioned species (Figure 1C), leaving positive selection as the most likely causal factor for the sweep. Hence, it is likely that presence of TSHR sweep in the SYNBREED red junglefowl populations is an outcome of their potential sustained hybridization with domestic populations.
Patterns of gene expression mirrored previously reported differences between domestic and wild-type chickens (Fallahshahroudi et al. 2019), and strongly suggest that the domestic variant causes earlier sexual maturity and potentially enhanced HPT activity. However, we did not find any significant difference between domestic- and wild-type variant chickens in circulating pituitary hormone TSH and thyroid hormone T3 and T4 levels. Multiple factors, such as amount of utilized food (McNabb 2007) and stress (Helmreich and Tylee 2011), can affect circulating thyroid hormones at a given timepoint. Thus, the apparent lack of association between gene expression and hormone levels may be attributed to biological fluctuations in the levels of circulating thyroid hormones and the small number of studied animals, limiting statistical power. Accordingly, using a bigger sample size (n = 114) we have previously shown that juvenile (10 weeks old) intercross male chicks possessing the domestic TSHR variant have higher T3 and lower T4 than those possessing the wild-type variant (Karlsson et al. 2015). Yet another possibility is that the TSHR variants exert their effects locally in the brain, independently of hormone levels in the blood. Supporting this, pituitary TSHB was found to promote sexual maturity in quails by increasing local concentration of thyroid hormones within the hypothalamus, without modifying the HPT axis activity (Nishiwaki-Ohkawa and Yoshimura 2016). It should also be remembered that the birds studied here were still young, while hormonal differences associated with the variant may only become apparent at the time of sexual maturity. In agreement with this, we previously demonstrated that the domestic variant is associated with earlier onset of sexual maturity in intercross birds (Karlsson et al. 2016). Thus, studying juvenile chicks, enabled us to investigate the developmental mechanisms underlying the phenotypic effects of the domestic TSHR variant on hormone levels and onset of sexual maturity.
In agreement with previous research (Heldmaier and Ruf 1992), metabolic rate, measured by oxygen consumption, increased with experimentally decreasing ambient temperature. However, we found no effect of variant, either alone or in interaction with temperature. Hence, this study does not support an adaptive role for the domestic TSHR variant in regulating acute responses to cold temperature, though we cannot exclude that our measure of metabolic rate was not sensitive enough to detect a small size effect, or that stress effects on metabolism confounded and obscured TSHR-linked variation.
Mirroring formerly differences between White Leghorns and red junglefowls (Campler et al. 2009), domestic-variant chickens were significantly more active during the OF test compared to their wild-type counterparts. This suggests that the TSHR variant may causally underlie some of the observed behavioral differences between the aforementioned breeds. While higher activity levels can be interpreted as a sign of reduced fear (Forkman et al. 2007) other factors such as social dependence and explorative tendency of the birds may also contribute to the behaviors observed in the OF test, and hence, caution must be taken in the interpretation of this test (Jones and Carmichael 1997; Forkman et al. 2007). Furthermore, TSHR genotype did not affect other fear-related behaviors such as latency of the first step and the ratio of time spent in periphery to center of the arena. It is worth noting that studies in rats showed a positive association between high activity in the OF test and circulating thyroid hormone levels (Hara et al. 2009; Helmreich and Tylee 2011).
On the other hand, we found no evidence for any effect of TSHR variant in the SR test. The domesticated variant has previously been associated with reduced aggression in adult chickens (Karlsson et al. 2015), which led us to hypothesize that social behavior might be affected as well. We found however a sex-difference in the latency to reach the social zone, with females approaching their peers faster than males, suggesting higher sociability for females.
We decided to introgress the wild-type TSHR variant into White Leghorn rather than the opposite because of two main reasons. First, selection of the domestic TSHR variant coincided with intensification of chicken production in Europe. Therefore, it is likely that other domestication related gene variants with potential epistatic interaction with TSHR were also selected during this period. Hence, we believe understanding the current functionality of this variant necessitate investigating it in the genetic background of the domestic breed. Second, we assumed that the more homogenous genetic background of the White Leghorn compared to that of red junglefowl may enable detecting smaller effects. However, it is possible that the domestic TSHR variant exerts more pronounced effects in the genetic background of red junglefowl. Given the prevalence of the domestic variant in the present red junglefowl populations, future works should compare red junglefowls differing in the TSHR variant.
In conclusion, we found that the domestic TSHR variant is common in a wide range of chickens throughout the world, including modern production breeds and regional breeds as well as in different red junglefowl subspecies. However, this domestic variant was absent in other species of the genus Gallus. Future studies should address the biological function and fitness consequences of these haplotypes in red junglefowl populations. Furthermore, TSHR variants had significant effects on downstream gene expression throughout the HPT axis, supporting an important role for this variant in many facets of chicken domestication, such as reproduction, physiology and behavior.
Acknowledgments
We thank Julia Buskas for her technical assistance.
Funding
The project was supported by grants from Swedish Research Council (VR): 2015–05444, 2015-4870, and 2017–06218, and the European Research Council (ERC; Advanced Grant 322206 GENEWELL).
Conflicts of interest
None declared.
Literature cited
R Core Team.
Author notes
Amir Fallahshahroudi and Martin Johnsson authors contributed equally to this work.
