|
Post by Admin on Dec 17, 2021 23:57:18 GMT
Genetic Relatedness among Populations The genetic distance and MDS analyses based on MSY and mtDNA indicate that the MN and MA are highly diverged from the other populations for the MSY and mtDNA, respectively (supplementary fig. 5, Supplementary Material online). The MA and MN also show large differences from the other populations in the heat plots of Φst values (fig. 3B). However, in general, both MSY and mtDNA results show relatively larger genetic heterogeneity of the AA groups versus genetic homogeneity of the TK and ST groups (fig. 3B). The Mantel test of Φst values showed a significant correlation between the MSY and mtDNA Φst matrices (r = 0.4506, P < 0.01). After excluding these MA and MN as outliers, the MDS for the MSY showed that almost all AA-speaking groups are located along the edges of the plot, whereas most of the TK groups cluster in the center of the plot (fig. 4A), further supporting genetic heterogeneity of the AA and homogeneity of the TK populations. Interestingly, the SEA-specific O-M95* and O-M324* haplogroups (with several sublineages) differentiate the studied populations into at least two main paternal sources, and the frequencies of these two haplogroups correspond to the major differentiation in the MDS plot (fig. 4A). O-M95* is at high frequency in the populations on the left of the plot and gradually decreases to very low frequency in the populations on the right side in the first dimension, whereas the O-M324* frequency runs opposite to the O-M95* cline: O-M324* is at higher frequency in populations located on the right of the plot and decreases in frequency toward the left side (fig. 4A). The MDS plot and heat plot of MSY also indicates some Mon groups (MO1, MO3, MO5, and MO6) are close to the cluster of TK groups in the center of the plot (fig. 4A and C), indicating a close genetic relationship. In addition, non-SEA haplogroups lineages, for example, R*, H*, and J*, provide more support for genetic connections between Mon and Central Thais. FIG. 4. The two-dimensional MDS plot and five-dimensional MDS heat plot based on the Φst distance matrix for 57 populations (after removal of Maniq and Mlabri) of MSY (A and C) and mtDNA (B and D). For the MDS based on mtDNA (fig. 4B), the Mon generally showed genetic affinities with the TK groups in the center of the plot, with the exception of MO1, MO5, and MO6, which differ from the other Mon groups, as can be also seen in the MDS plot and heat plot (fig. 4B and D). Overall, we observe more genetic heterogeneity of the AA groups than the other linguistic groups and there are contrasting patterns of genetic relationships for the MSY versus mtDNA.
|
|
|
Post by Admin on Dec 18, 2021 21:41:59 GMT
Genetic Relatedness between Thai/Lao and Other Asian Populations The MDS based on the MSY Φst matrix of 73 populations from across Asia revealed that, in general, population clustering largely reflects linguistic affiliation (fig. 5), with some exceptions. In the first and second dimension, the AA populations are the most diversified, with the PL and MN appearing as outliers. There is one cluster of AA populations on the left, which also includes one TK group (BT2); the other AA populations are scattered along the main axis of the plot. Some Mon groups (MO2, MO4, and MO7) are relatively close to Indian and ISEA populations, indicating potential connections. Two central Thai groups (CT4 and CT7) are also relatively close to the Indian populations. The ST populations (Karen, Han Chinese, and Burmese) are rather close. The ISEA and Papuan populations are in closer proximity to South Asian populations (Indian, Bengali, and Punjabi). Generally, the haplogroup profile indicates genetic affinities between the Mon and South/Central Asian groups, which is consistent with the MDS plots (fig. 5) and results from previous mtDNA haplogroup analyses (Kutanan et al. 2017; Kutanan, Kampuansai, Brunelli, et al. 2018). FIG. 5. The two-dimensional MDS plot based on the MSY Φst distance matrix for 73 populations. Population details are listed in figure 1 and supplementary tables 1 and 7, Supplementary Material online.
|
|
|
Post by Admin on Dec 19, 2021 2:25:20 GMT
The Expansion of Male Lineages The Bayesian Skyline Plots (BSPs) of effective population size change (Ne) over time in each group reveal overall five different trends (fig. 6). The most common trend, found in Mon, Khmer, Htin, Central Thai, and Black Tai, showed Ne increasing gradually or remaining constant during 40–60 ka until a decline ∼5–7 ka, followed by rapid growth ∼5 ka and then a decrease ∼2.0–2.5 ka. The other trends differ from the first trend as follows: no population reduction ∼2.0–2.5 ka but population size either increases (Khon Mueang and Yuan) or remains stable (Lao Isan and Laotian); the Lue and Phuan show two increases in Ne, at about ∼5 ka and ∼10 ka; the Lawa show a stable population size since ∼30 ka and then a decline during the last 2 ka with a sudden increase ∼1 ka; and the Karen differ only slightly from the common trend with a population increase ∼1 ka. FIG. 6. The BSPs based on the MSY and mtDNA of 13 ethnicities from Thailand and Laos; Mon, Khmer, Htin, Central Thai, Black Tai, Khon Mueang, Yuan, Lao Isan, Laotian, Lue, Phuan, Lawa, and Karen. Solid lines are the median estimated effective population size (y axis) through time from the present in years (x axis). The 95% highest posterior density limits are indicated by dotted lines. By contrast, the BSP based on mtDNA sequences for each ethnicity show three common trends (fig. 6). The first trend is an increase in Ne during 40–50 ka, followed by stability and then decrease ∼2 ka, which was observed in Mon, Htin, Lawa, Khmer, Yuan, Phuan, and Lue. The second pattern, shown by the Khon Mueang, is an increase in Ne ∼ 40–50 ka, followed by stability and then increase again ∼10 ka, followed by a decline ∼2 ka. The Central Thai, Lao Isan, and Laotian show the third trend, in which population increases occur ∼40–50 and ∼10 ka. In general, the BSP by ethnicity indicated lower effective population sizes for the MSY than for mtDNA (fig. 6). We also plotted the BSP of several Asian populations from published MSY data (Karmin et al. 2015; Poznik et al. 2016) (fig. 7). Almost all of the MSEA and East Asian populations, that is, Kinh, Northern Han, Southern Han, and Japanese show a pronounced increase of the MSY Ne during ∼4–6 ka, except the Xishuangbanna Dai, in which there is an increase ∼2 ka. Around 5 ka, the Japanese show a decrease in Ne before a sudden increase, suggesting a bottleneck prior to demographic expansion. Interestingly, the ISEA population shows a large increase in Ne ∼ 35–40 ka and a smaller increase ∼2.5–3 ka. The South Asian populations, that is, Bengali, Punjabi, and Indians, also show two pulses of population increase at about the same times. The Punjabi also show an additional small increase in Ne change during ∼12 ka.
|
|
|
Post by Admin on Dec 19, 2021 3:36:38 GMT
The BSP by each major MSY haplogroup show four pulses of paternal Ne increases, at ∼9–11 ka, ∼5 ka, ∼2.0–2.5 ka, and ∼1.0 ka (fig. 8), in agreement with the plot by ethnicity. The early Holocene Ne increment is obviously noticed in O2a1c* and O2a2a*, whereas the Ne growth ∼5 ka is observed in O1b1a1a1b* and R*. Haplogroup O1a*, C* and D* show expansions in Ne ∼ 2.0–2.5 ka and haplogroup N* shows a recent expansion ∼1.0 ka. In addition, there are two expansion times for O1b1a1a1a* and O2a2b* (∼5 and ∼2 ka). FIG. 8. The BSPs for each major haplogroup. Solid lines are the median estimated paternal effective population size (y axis) through time from the present in years (x axis). The 95% highest posterior density limits are indicated by dotted lines. Demographic Models Previously, we used mtDNA genome sequences and demographic modeling to test different hypotheses about the origins of TK groups. Specifically, we tested whether different TK groups were primarily related to local AA groups (reflecting cultural diffusion, i.e., an AA group switching to a TK language), to a TK group from southern China (reflecting demic diffusion, i.e., spread of TK languages via migration from southern China), or were related to both (reflecting admixture between an incoming TK group from southern China and a local AA group). We found that the Khon Mueang (from northern Thailand), Lao Isan (from northeastern Thailand), and Laotian most likely originated via demic diffusion from southern China without substantial gene flow from AA groups (Kutanan et al. 2017). However, for the central Thai, the most likely scenario was demic diffusion with a very low level of gene flow between central Thai and Mon groups (Kutanan, Kampuansai, Brunelli, et al. 2018). Here, we use the same approach to test three demographic scenarios concerning the paternal origins of these major Thai groups (supplementary fig. 6, Supplementary Material online). For the Khon Mueang (KM) people (Test 1), the highest posterior probability (0.80) and rather highly selected classification trees (0.58) were found for the demic diffusion model (supplementary table 3, Supplementary Material online). By contrast, the cultural diffusion model is the most likely scenario for the Lao and central Thai groups. Both the combined Laotian (LA) and Lao Isan (IS) data sets (Test 2) and the separate LA data set (Test 3) weakly support the cultural diffusion model (for Test 2: posterior probability = 0.56 and selected classification tree = 0.37 and for Test 3: posterior probability = 0.56 and selected classification tree = 0.39). The IS data set (Test 4) supports cultural diffusion (with the present-day IS groups descended from local Khmer [KH] with the highest posterior probability [0.71] and classification trees selected slightly more often than for the other models [0.49]). For Test 5 (the central Thai [CT] data set), the cultural diffusion model had the highest posterior probability at 0.58 and was selected slightly more often among the classification trees (0.50) than the other models. However, a Principal Component Analysis plot shows that based on the first two PCs the observed data fall within the distributions simulated under the three models in only Test 4, whereas the other data sets fall within the simulated distributions for PCs 3 and 4, suggesting that there is low efficiency to reconstruct the variability of the observed data (supplementary fig. 7, Supplementary Material online). The parameter estimation for the best performing models in all five tests was able to obtain point estimates for each of the simulated effective population sizes (supplementary table 4, Supplementary Material online). However, the posterior distributions were generally flat (supplementary fig. 8, Supplementary Material online). We also calculated the MSY Φst and corrected pairwise difference among groups of populations used in ABC tests to estimate their genetic relationships (supplementary table 5, Supplementary Material online). The KM are closer to the Dai than the local AA group (Test 1), the ethnic Lao and Laotian showed similar genetic differences to both Dai and AA groups (Test 2 and Test 3), whereas the CT groups (Test 5) have closer genetic relationships to the local AA group than to Dai. In contrast, mtDNA Φst and corrected pairwise difference revealed that the KM and ethnic Lao are closer to the Dai than local AA, whereas the CT exhibited somewhat similar genetic distances to both Dai and AA. Overall, the simulations based on MSY sequences, compared with previous mtDNA simulation together with tests of genetic difference by Φst and corrected pairwise differences, suggest different demographic histories for males and females in the region.
|
|
|
Post by Admin on Dec 19, 2021 18:43:09 GMT
Discussion In order to gain more insights into MSEA genetic history, we here investigate the paternal genetic variation and structure by sequencing ∼2.3 mB of the MSY from representative ethnolinguistic groups from Thailand and Laos. In sum, most of the studied populations exhibit two major MSY haplogroups, O-M324* and O-M95* in different proportions, indicating two major paternal sources. O-M324* was widely spread in the TK groups, whereas O-M95* is predominant in the AA groups. However, some TK populations (BT2 and IS3) and some AA populations (PL, BO and MO4) exhibited the opposite pattern (fig. 1 and supplementary table 2, Supplementary Material online). We also compared patterns of MSY variation with mtDNA in the same set of populations and found some similar results, for example, overall lower genetic diversity and greater heterogeneity of AA groups than of TK and ST groups, large differences between the Mon and the other AA groups, and genetic connections between the Mon and central Thai (figs. 2–4). However, in many respects, the patterns of MSY and mtDNA variation are different, suggesting contrasting paternal and maternal genetic histories. Here, we focus on three groups of populations with different cultural practices and histories that also stand out in the genetic analyses: the hill tribes, the AA-speaking Mon, and the major TK-speaking groups.
Factors Influencing Contrasting Genetic Variation in the Hill Tribes The hill tribes, who occupy the mountainous northern region of Thailand, are notable for their variation in patrilocal versus matrilocal residence pattern (Oota et al. 2001; Besaggio et al. 2007), as well as for their strong sense of group identity, which tends to isolate them from other groups (Schliesinger 2000; Nahhas 2007). If postmarital residence is influencing patterns of genetic variation, then the expectation is for larger between-group differences and smaller within-group diversity for patrilocal groups for the MSY, and the same trends for matrilocal groups for mtDNA. The first comparative study of mtDNA and MSY variation in patrilocal versus matrilocal groups was carried out in the northern Thai hill tribes and found a strong impact of postmarital residence on the mtDNA and MSY variation (Oota et al. 2001). However, previous studies compared genetic variation between partial mtDNA sequences and Y-STRs (Oota et al. 2001; Besaggio et al. 2007); here, we report the first comparison of mtDNA and MSY variation based on comparable sequence data.
Here, we analyzed the sequences of mtDNA genome and ∼2.3 mB of the MSY of the Khmu, Palaung, and Lawa groups, who practice patrilocality, whereas the Htin are matrilocal, similar to the ST-speaking Karen. The within-population genetic diversity values are in agreement with expectations, that is, greater diversity of matrilocal than patrilocal groups for MSY and the opposite trend in mtDNA (supplementary fig. 4, Supplementary Material online). Moreover, genetic differentiation between populations also goes in the direction predicted by postmarital residence pattern. However, in many cases, the differences between patrilocal and matrilocal groups are not significant, indicating that other factors are also having an effect. In particular, the Htin (TN1) and Lawa (LW3) exhibit very low within-population diversity for the MSY, whereas the Htin (TN1 and TN2) also show lower diversity for the mtDNA (fig. 2A–C).
One factor in particular that could influence the within-population genetic diversity and between-population differentiation is geographic isolation, which enhances genetic drift and inbreeding, thereby lowering within-population genetic diversity and increasing between-population differentiation. This could explain the very low internal diversity and high differentiation from other groups of some groups of Htin (TN1) and Lawa (fig. 4A and B and supplementary fig. 3, Supplementary Material online) that live in mountainous, isolated parts of northern Thailand. The Lawa furthermore favor intramarriage (Nahhas 2007) which would also reduce genetic variation in this group. The Htin (TN1) also show very low diversity and extreme divergence in genome-wide single nucleotide polymorphisms data (Xu et al. 2010) and both Htin (TN1–TN3) and Lawa (LW3) exhibit lower diversity and large differentiation in autosomal STRs (Kampuansai et al. 2017). Such drastic genetic drift effects could reduce the significance of the impact of postmarital residence on patterns of genetic diversity.
Moreover, these results are in keeping with previous observations that although the expected difference between patrilocal and matrilocal groups holds in some regions (Oota et al. 2001; Besaggio et al. 2007), in other regions patterns of mtDNA and MSY variation do not conform to expectations (Kumar et al. 2006; Arias et al. 2018). This is indeed to be expected given that many other factors, for example, other human cultures (e.g., linguistic exogamy), physical landscape, and subsistence strategies, influence patterns of genetic variation (Wilkins and Marlowe 2006; Chaix et al. 2007).
Genetic Variation and Origin of the Mon The Mon groups showed genetic differences from other AA populations but closer relatedness to the TK populations, especially the central Thai, in both MSY and mtDNA (figs. 2A, B, 3A, and B). Our previous simulation results, based on mtDNA, also supported admixture among the Mon and central Thai groups (Kutanan, Kampuansai, Brunelli, et al. 2018). In addition, some Mon groups (MSY: MO3, MO5, MO6 and mtDNA: MO2, MO3 and MO4) exhibit genetic affinities with the Karen (fig. 3B), reflecting genetic heterogeneity and contrasting genetic patterns between MSY and mtDNA. Admixture might be an important factor influencing the genetic structure of the lowland AA-speaking Mon. Archaeological evidence indicates that the Dvaravati civilization of the Mon was centered in present-day central Thailand and southern Myanmar and had expanded to a large part of MSEA during the sixth to seventh century AD (Diffloth 1984; Guillou 1999; Saraya 1999). After the intensification of Thai and Burmese kingdoms, the Mon in Myanmar were conquered by the Burmese during the 18th century AD; the ethnic Mon in Myanmar are currently concentrated in the Mon and Karen States (Pon Nya 2001). In Thailand, the present-day Mon are distributed in central Thailand and surrounding areas, with some groups living in the North and the Northeast. However, they are not considered to be the descendants of the ancient Mon Dvaravati civilization in Thailand, but rather political refugees that fled from Myanmar to Thailand during the 16th to 19th centuries AD (Ocharoen 1998). However, based on linguistic evidence, the remnants of the Dvaravati Mon population are now considered a distinct ethnic group known as the Nyahkur (BO) whose communities are restrict found in hilly areas along the border between central and northeastern Thailand (Diffloth 1984). In contrary to linguistic evidence, the Nyahkur has no shared haplotype or related to any specific Mon groups, indicating their genetic differences. However, Nyahkur show genetic sharing in both MSY and mtDNA with the Khmer groups (fig. 3A) which reflects their previous connection. In addition, the high frequency of MSY haplogroup O2a* and C* (fig. 1), close genetic relationship to many TK- and ST-speaking groups (fig. 3B) and highest MPD value for MSY (fig. 2C) indicated later extensive gene flow, promoting the paternal difference of Nyahkur from the Mon and also other AA groups.
Previous genetic studies of G6PD mutations reported a high prevalence of the Mahidol type G6PD deficiency in the Mon, Burmese, and Karen, different from Thai, Laotian, and Khmer groups exhibiting the Vientiane-type G6PD mutation (Iwai et al. 2001; Matsuoka et al. 2005; Nuchprayoon et al. 2008). Thus, both our results and previous studies indicate a close genetic relationship among Mon, Burmese, and Karen in Myanmar, suggesting a common origin or extensive gene flow. Our previous mtDNA study also revealed genetic relations between some Mon groups (MO1 and MO5) and Burmese, with both of them close to some Indian populations, whereas other Mon groups are closer to the Karen groups (MO2, MO3, and MO4) (see details in Kutanan, Kampuansai, Brunelli, et al. [2018]). In general, genetic mixing among Mon, Karen, and Myanmar might have happened before the arrival of the Mon to Thailand, whereas mixing among the Mon and central Thai would have occurred after the arrival of the Mon. However, MSY data for the Burmese are limited, and further MSY studies of populations from Myanmar are needed to confirm this scenario.
A connection between Indian groups and the Mon is suggested by South/Central Asian MSY lineages in the Mon, for example, R*, H*, J*, L*, and Q* (fig. 1), consistent with some mtDNA lineages, for example, W3a1b, M6a1a, M30, M40a1, M45a, and I1b (Kutanan et al. 2017; Kutanan, Kampuansai, Brunelli, et al. 2018). Thus, both mtDNA and the MSY indicated contact between the ancestors of the Mon and Indian. Archaeological evidence also suggests Indian influences, for example, the symbolism on the Dvaravati coin which indicates the importance of royalty, and includes several motifs associated with Indian precedents of the first to fourth century AD (Higham and Thosarat 2012).
|
|