Origin and evolutionary history of domestic chickens inferred from a large population study of Thai red junglefowl and indigenous chickens

Determination of the mtDNA D-loop haplotypes of indigenous chicken breeds and red junglefowl in Thailand

We determined the nucleotide sequences of the 780 bp fragments of the mtDNA D-loop region, including the hypervariable segment I, in 125 individuals from 10 indigenous chicken breeds (a list of breeds is shown in Table 1), and 279 red junglefowls from two subspecies (G. g. gallus and G. g. spadiceus) within 12 populations in Thailand. A total of 44 haplotypes with 62 variable sites, consisting of 26 singletons and 36 parsimony informative sites were identified (Supplementary Tables S1, S2; accession No. LC542982 to LC543385). Table 2 summarizes the details of the haplotypes found in the 10 indigenous chicken breeds and 12 red junglefowl populations, and Fig. 1 shows the composition of each haplogroup of indigenous chickens and two subspecies of red junglefowl. Forty-four haplotypes were temporally classified into eight common haplogroups; A, B, C, D, E, F, H, and J (Supplementary Figs. S1, S2), according to Liu et al.⁴ and Miao et al.⁵. In the present study, we treated haplogroups C and D as one unit (CD) as they were not clearly separated. In addition, haplogroup J was closely related to haplogroup CD. Haplogroups A, B, and E were predominant in the Thai indigenous chickens (Table 2), and their frequencies were almost the same (Fig. 1). Haplogroup CD was predominant in the G. g. spadiceus population, but rare in indigenous chickens. In the G. g. gallus population, the haplogroups B, CD, and E were detected at almost the same frequency; however, haplogroup A was not detected. The frequency of haplogroup J, which was mainly found in the Si Sa Ket population, was much higher in G. g. gallus compared with indigenous chicken and G. g. spadiceus populations (Fig. 1). BT, NK-W, NK-B, LHK, CH, PHD, Decoy, fighting chicken, and seven red junglefowl populations (Huai Sai [Ggg], Huai Sai [Ggs], Sa Kaeo, Chanthaburi, Khao Kho, Chaiyaphum, and Khok Mai Rua) each exhibited breed- or population-specific haplotypes (Supplementary Table S2): Hap_04 (BT); Hap_07 and 08 (NK-W); Hap_09 (NK-B); Hap_10 and Hap_11 (LHK); Hap_14 (CH); Hap_17 (PHD); Hap_18 to 20 (Decoy); Hap_29 to 33 (fighting chicken); Hap_21 (Huai Sai [Ggg]); Hap_22 (Huai Sai [Ggs]); Hap_24 and 25 (Sa Kaeo); Hap_27 and 28 (Chanthaburi); Hap_36 (Khao Kho); Hap_38 and 39 (Chaiyaphum); and Hap_43 and 44 (Khok Mai Rua). Twenty-nine out of 44 D-loop haplotypes which had not been previously deposited in GenBank, were newly identified in the present study.

The topologies of the Bayesian tree and the maximum-likelihood (ML) tree based on the HKY + G + I model of evolution, which were selected as the best-fit substitution model, were fundamentally similar. Although the Bayesian posterior probability of the internal nodes and the ML bootstrap values were relatively low due to the short internal branches (multifurcations) of the phylogenetic trees, the haplogroups A, B, F–I, K, Y, and Z were supported by a Bayesian posterior probability of greater than 0.97 (Supplementary Figs. S1, S2). Both The trees revealed that the D-loop sequences obtained in this study could be classified into six haplogroups: A, B, CD, E, F, and J, and a complex group of rare haplogroups (H, I, K, W, and X), except for an unclassified haplotype, Hap_38.

Six haplotypes from seven indigenous and two red junglefowl populations (LHK, CH, PHD, KP, BT, Decoy, fighting chicken, Huai Sai [Ggs], and Petchaburi) belonged to haplogroup A (Fig. 2a). Seven haplotypes from seven indigenous chicken breeds and five red junglefowl populations (LHK, CH, PHD, KP, Decoy, fighting chicken, DT, Sa Kaeo, Huai Sai [Ggg], Huai Sai [Ggs], Khao Kho, and Petchaburi) were classified into haplogroup B (Fig. 2b). Haplogroup CD contained 12 haplotypes, which were identified in two Thai indigenous chicken breeds (PHD and fighting chicken) and seven red junglefowl populations (Chanthaburi, Khok Mai Rua, Chaing Rai, Huai Sai [Ggs], Khao Kho, Chaiyaphum, and Huai Yang Pan) (Fig. 2c). Eight haplotypes in four indigenous chicken breeds (CH, BT, NK-W, and NK-B) and four red junglefowl populations (Roi Et, Khok Mai Rua, Chaiyaphum, and Huai Yang Pan) belonged to haplogroup E (Fig. 2e). Haplogroup F contained one haplotype, which was only found in two indigenous chicken breeds (LHK and PHD) (Fig. 2f). Eight haplotypes of haplogroup J (Hap_10, Hap_11, Hap_24 to 26, Hap_34, Hap_40, and Hap_42) were found in two indigenous chicken breeds (LHK and fighting chicken) and seven red junglefowl populations (Sa Kaeo, Chabthaburi, Si Saket, Roi Et, Khok Mai Rua, Chaing Rai, and Huai Yang Pan) (Fig. 2c). Only one haplotype of haplogroup H (Hap_05) was detected in two indigenous chicken breeds (BT and fighting chicken) (Fig. 2d). Hap_38, which was found in three individuals of the Chaiyaphum population, did not belong to any known haplogroups; however, the haplotype was more closely related to haplogroup CD than the other haplogroups (Fig. 2c; Supplementary Figs. S1, S2).

Divergence times for each haplotype were determined using BEAST analysis (Supplementary Table S2) and were 0.24–0.45 kilo years ago (KYA) for haplogroup A, 0.15–0.39 KYA for haplogroup B, and 0.14–0.37 KYA for haplogroup CD; the haplotypes of haplogroup E exhibited a wide range of divergent times, ranging from 0.12 to 0.70 KYA (0.41 KYA on average). One haplotype in haplogroups F and H had possibly diverged at 0.33 and 0.34 KYA, respectively. The divergence times of haplotypes in haplogroup J ranged from 0.10 to 0.60 KYA. Hap_38 exhibited a markedly earlier divergence time, which was estimated to be approximately 12,000 years ago (Supplementary Table S2; Supplementary Fig. S1).

Genetic diversity of mtDNA D-loop sequences

The number of D-loop haplotypes in each population (H) ranged from 1 (G. g. gallus population at Huai Sai and Si Sa Ket) to 10 (fighting chicken) (Table 3). Among the Thai indigenous chicken breeds, LHK, CH, PHD, KP, BT, Decoy and fighting chicken exhibited relatively higher genetic diversity (pi, 0.005 for KP and Decoy to 0.009 for LHK, PHD, and fighting chicken; Theta-w, 2.86 for KP to 8.30 for fighting chickens) than in NK-W, NK-B, and DT (pi, 0.001 for NK-W, NK-B, and DT; Theta-w, 0.35 for NK-B to 0.71 for NK-W) (Table 3). With regard to the red junglefowl populations, all populations excluding Si Sa Ket and Petchaburi exhibited similar levels of genetic diversity. The Petchaburi population had two haplotypes, and the genetic diversity was relatively low. The low genetic diversity in the Si Sa Ket population was attributed to the fact that all 30 examined individuals shared only one haplotype. Seven out of the 10 indigenous chicken breeds (LHK, CH, PHD, Decoy, fighting chicken, NK-W, and DT) exhibited negative Tajima’s D values, suggesting that the chickens were bred under purifying selection within each population; however, even though the Tajima’s D values of all populations were not statistically significant (p > 0.05).

Phylogenetic relationships among mtDNA D-loop haplotypes in red junglefowl in Asia

Haplogroups A, B, CD, E, and J were frequently identified in red junglefowl in Thailand (Fig. 1; Supplementary Table S2). Haplotypes from Thailand, Vietnam, Laos, and Myanmar were located in internal nodes of the haplogroup A, B, CD, and E, and haplotypes from China were derived from the haplotypes in Southeast Asia (Fig. 3). Haplogroup J exclusively consisted of haplotypes from Thailand, Vietnam, and Cambodia. In haplogroup F, haplotypes from Cambodia exhibited the ancestral haplotypes of Chinese red junglefowl. Three haplotypes of red junglefowl from Indonesia were observed in haplogroup K. A small number of haplotypes from the other rare haplogroups G and W to Z were only observed in Chinese red junglefowl.

Genetic characteristics of indigenous chicken breeds and red junglefowl estimated by 28 microsatellite DNA markers

Two-hundred and ninety-eight red junglefowls in two subspecies (G. g. gallus and G. g. spadiceus) from 12 populations and 138 chickens from 10 indigenous chicken breeds, were used for the genetic diversity analyses using 28 microsatellite markers (Table 1; Supplementary Table S3). The allelic richness (AR) values ranged from 1.40 for MCW103 to 1.93 for MCW0014 (1.77 on average) (Supplementary Table S3). Na ranged from 2.14 for MCW0103 to 7.59 for LEI0192 (2.86 on average). F_IS varied from – 0.08 for LEI0166 to 0.54 for MCW0014 (0.05 on average). The F_ST and F_IT values fell within the 0.07 (MCW0098) to 0.31 (MCW0247) range and 0.09 (MCW0123) to 0.63 (MCW0014) range, respectively (F_ST = 0.17, F_IT = 0.22 on average). Two markers, MCW0222 and MCW0014, showed the null allele frequency across all populations (NAF) higher than 0.2 (Supplementary Table S3). Looking at each population, null allele frequencies higher than 0.2 were detected in seven, three, and three populations for MCW0014, MCW0222, and LEI0192, respectively, and in less than one or two populations for the other 11 markers (Supplementary Table S4).

Significant departures from Hardy–Weinberg equilibrium were observed for LEI0234, MCW0014, and MCW0123 in more than 10 populations (p < 0.05) (Supplementary Table S5). The Khok Mai Rua and Huai Sai (Ggs) populations exhibited significant departures at 13 and 11 loci, respectively, while the other red junglefowl and indigenous chicken populations exhibited significant departures at less than 10 loci (p < 0.05).

Out of the 22 populations examined, the BT population exhibited the least genetic diversity (AR = 2.26; mean number of alleles [MNA]) 3.32; Ne = 2.01; Ho = 0.45) (Table 3). In the other nine indigenous chicken breeds, AR ranged from 2.76 (NK-B) to 3.25 (LHK); MNA ranged from 3.43 (NK-B and DT) to 6.07 (fighting chicken); Ne ranged from 2.50 (NK-W and NK-B) to 3.25 (fighting chicken); and Ho ranged from 0.51 (PHD) to 0.64 (NK-B). In the red junglefowl populations, AR ranged from 2.78 (Roi Et) to 3.59 (Chanthaburi); MNA ranged from 2.82 (Huai Sai [Ggg]) to 7.07 (Khok Mai Rua); Ne ranged from 2.37 (Huai Sai [Ggg]) to 3.94 (Chanthaburi); and Ho ranged from 0.45 (Huai Sai [Ggg]) to 0.68 (Chanthaburi). Chanthaburi (AR = 3.59; MNA = 6.29; Ne = 3.94; Ho = 0.68) and Khok Mai Rua (AR = 3.48; MNA = 7.07; Ne = 3.49; Ho = 0.60) exhibited comparatively higher genetic heterogeneity among the populations examined.

Genetic relationships among indigenous chicken breeds and red junglefowl in Thailand

Two cladograms based on Da and Dxy genetic distances constructed with mtDNA D-loop sequences exhibited similar topologies (Fig. 4a,b); six breeds of Thai indigenous chickens (LHK, CH, PHD, KP, Decoy, and fighting chicken) and the DT chicken breed showed close genetic relationships in both trees, whereas three other Thai indigenous chicken breeds (BT, NK-W and NK-B) were phylogenetically located far from the six breeds. Two red junglefowl subspecies were not phylogenetically separated from each other. However, the phylogenetic tree based on the F_ST genetic distance constructed using microsatellite markers indicated that all red junglefowl populations excluding Huai Sai (Ggg) and Huai Sai (Ggs) had a close genetic relationship (Fig. 4c). Five indigenous chicken breeds (Decoy, CH, LHK, PHD, and KP) fell in a cluster distinct from the red junglefowl populations; however, the fighting chickens were genetically closer to red junglefowl. Four indigenous chicken breeds (BT, NK-W, NK-B and DT) were obviously genetically different from the other indigenous chicken breeds and the red junglefowl populations. The phylogenetic tree topology based on R_ST genetic distances was largely different from those based on Da, Dps, and F_ST genetic distances (Fig. 4d). In the unrooted phylogenetic trees among individuals based on Da and Dps genetic distances, most of the 22 populations were classified into distinct clusters (only the tree based on the Da genetic distance is presented in Supplementary Fig. S3). While BT, NK-W, NK-B, and DT each formed a distinct cluster, the LHK, CH, PHD, and KP individuals formed a mixed cluster, where a few Decoy and DT individuals were included. The clusters of all the red junglefowl populations were basically distinct, although no obvious genetic differentiation was observed between the two subspecies, G. g. gallus and G. g. spadiceus (illustrated by open and filled circles, respectively).

We performed STRUCTURE analysis using 26 MS markers, excluding two markers, MCW0222 and MCW0014, which showed NAF higher than 0.2; however, we realized that the resulting STRUCTURE plots did not differ from those obtained using all 28 markers. Thus, we used all of the 28 markers for STRUCTURE analysis in this study. Analysis using STRUCTURE HARVESTER revealed that K = 2 was optimal for the 22 populations (delta K = 6.59) and K = 4 was the second highest (delta K = 2.05) (Fig. 5a). The STRUCTURE plot at K = 2 subdivided four red junglefowl populations (Sa Kaeo, Chanthaburi, Chaing Rai, and Chaiyaphum) from the other red junglefowl populations and indigenous chicken breeds. The 22 populations were classified into four clusters at K = 4 as follows: (1) indigenous chickens, such as ornamental chicken breeds (LHK, CH, PHD, and KP), Decoy, fighting chicken, and two red junglefowl populations (Si Sa Ket and Roi Et); (2) indigenous meat chicken breeds (BT and two NK breeds) and DT; (3) two red junglefowl populations (Sa Kaeo and Chanthaburi); and (4) the other populations of red junglefowl, consisting of Khok Mai Rua, Chaing Rai, Huai Sai (Ggg), Huai Sai (Ggs), Khao Kho, Chaiyaphum, Petchaburi, and Huai Yan Pan (Fig. 5b). However, Huai Yang Pan and Khok Mai Rua were mixed with other populations, such as Si Sa Ket and Roi Et, as well as Sa Kaeo and Chanthaburi. At K = 6, two populations of red junglefowl (Si Sa Ket and Roi Et) formed a cluster independent from the other six Thai indigenous chicken breeds (LHK, CH, PHD, KP, Decoy, and fighting chicken).

Analysis of molecular variance (AMOVA) revealed that 83.85% and 16.00% of the total genetic variance in 12 red junglefowl populations was attributed to genetic variance within populations and among groups of each subspecies, respectively, and only 0.16% of the total genetic variance was attributed to variance between two subspecies. This result indicates that the genetic variation of red junglefowl is mostly explained by the genetic divergence among populations, and that the genetic difference between subspecies is very low.

Source: Ecology - nature.com