Abstract
The Two Layer hypothesis is fast becoming the favoured narrative describing East Asian population history. Under this model, hunter-gatherer groups who initially peopled East Asia via a route south of the Himalayas were assimilated by agriculturalist migrants who arrived via a northern route across Eurasia. A lack of ancient samples from tropical East Asia limits the resolution of this model. We consider insight afforded by patterns of variation within the human pathogen Mycobacterium tuberculosis (Mtb) by analysing its phylogeographic signatures jointly with the human Y-chromosome. We demonstrate the Y-chromosome lineages enriched in the traditionally hunter-gatherer groups associated with East Asia’s first layer of peopling to display deep roots, low long-term effective population size, and diversity patterns consistent with a southern entry route. These characteristics mirror those of the evolutionarily ancient Mtb lineage 1. The remaining East Asian Y-chromosome lineage is almost entirely absent from traditionally hunter-gatherer groups and displays spatial and temporal characteristics which are incompatible with a southern entry route, and which link it to the development of agriculture in modern-day China. These characteristics mirror those of the evolutionarily modern Mtb lineage 2. This model paves the way for novel host-pathogen coevolutionary research hypotheses in East Asia.
Similar content being viewed by others
Introduction
With contributions from the fields of anthropometry and the study of ancient and contemporary DNA, a new paradigm has emerged in our understanding of East Asian population history1,2,3,4,5. This newly emerged consensus holds that East Asia was peopled in two main population movements or layers. The first layer consisted of hunter-gatherer groups with darkly pigmented skin, who arrived in East Asia via a dispersal route south of the Himalaya mountain range. These populations were later displaced and partially assimilated by agriculturalist groups with cold adapted physical features, who were likely to have reached East Asia via a route north of the Himalayas2.
Interestingly, the Two Layer model, and the closely related Dual Structure model of Japanese population history were originally rooted in the field of physical anthropology and received support from genetic studies only recently. The Two Layer model was initially proposed to explain the replacement of Hoabinhian hunter-gatherers exhibiting Australo-Melanesian features with agriculturalist groups exhibiting Northeast Asian features in the Southeast Asian archaeological record2,6,7,8,9,10. Similarly, the Dual Structure model describes the displacement and integration of hunter-gatherer Jomon groups by agriculturalist migrants across the Japanese archipelago. Under this theory, which draws from skeletal, odontal and craniometric data, Jomon-related ancestry was argued to persist mainly in the Indigenous Ainu and Ryukyuan groups from the northern and southern extremes of the island chain11,12,13,14,15,16,17,18,19,20. Genetic studies garnered support for both the Two Layer and Dual Structure models, by demonstrating the strong degree of shared genetic drift between ancient Jomon and modern Ainu genomes21, and between ancient Jomon and Hoabinhian genomes3.
Current studies now describe East Asian populations, at a basal level, to be modelled as mixtures of first and second layer ancestry, with the Indigenous Onge and the ancient Tianyuan genome from Neolithic northern China acting as surrogates for each respective layer1. Under these models, the Onge are argued to be amongst a number of modern-day human populations who represent largely unadmixed descendants of the first layer of peopling, including Papuans, the Ainu of Japan, and hunter-gatherers from the Philippines and Malaysia1,2,4.
While analyses of ancient DNA have afforded great insight into the Two Layer model, they are limited by the low availability of this form of data from the region. An underutilised means of deciphering aspects of human demography is by analysing patterns of genetic variation amongst human symbiotic species, including commensals and pathogens22,23,24.
In this study, we consider the insight afforded into the Two Layer model by analysing genetic variation within the genomes of the human obligate pathogen Mtb. We complement these inferences by investigating patterns of human genetic variation, inferred from Y-chromosome sequencing and genotype data. We pair these two marker systems, as they both possess ample phylogenetic signal, and are both inherited clonally, thus permitting the use of an identical suite of phylogenetic and phylogeographic tools in their analysis. The joint analysis of variation in these marker systems may reveal parallel phylodynamic and geographic signatures, which may be relevant to our understanding of the Two Layer model, and may furnish us with additional insights. We are also motivated to complete this investigation, as existing models describing the distribution of both Y-chromosome and Mtb genetic variation are largely incompatible with the Two Layer hypothesis.
Y-Chromosome data, while having played a seminal role in illuminating our African origins25,26, and describing male mediated demographic processes27,28, has not been robustly reconciled with the Two Layer model. Current models of East Asian Y-chromosome variation generally argue a single southern route origin of the four haplogroups which make up 90–95% of male lineages [C, D, and the NO clade (comprising haplogroups N and O)29,30,31,32], followed by subsequent northwards expansion30,31,33. These theories therefore don’t account for the proposed northern dispersal route associated with the second layer of peopling, or the implied genetic affinities shared by the present-day representatives of the first layer.
Similarly, genetic variation in the Mtb pathogen has also been well characterised in an East Asian context but has not been assessed in light of the Two Layer model. The East Asian Mtb landscape is dominated by isolates from the ‘evolutionarily ancient’ lineage 1, found at highest frequency in southern regions, and the ‘evolutionarily modern’ lineage 2, which display enhanced virulence and transmissibility34,35,36, and which predominates in northerly regions37. Under models proposing a Paleolithic38,39, as opposed to an argued Neolithic34,37,40,41 origin of Mtb, lineage 2 arrived in East Asia via a southern route, before expanding northwards, and radiating southwards again with the movement of the Han people39. This model doesn’t reconcile all aspects of the Two Layer hypothesis, including the proposed northern entry route of the second layer, nor the demographic processes and migrations which gave rise to the distribution of lineage 1.
Here we conduct a joint analysis of patterns of Mtb and Y-chromosome variation to propose a coevolutionary scenario compatible with the Two Layer model. In addition to illuminating a key chapter of human history, it will provide background context relevant to our understanding of the long-term interaction of humans and one of the deadliest diseases of all time.
Results
Y-chromosome variation and the Two Layer model
To explore East Asian patrilineal diversity we collated a large dataset of Y-chromosome haplogroup population frequencies. Studies were selected so as to unambiguously assign haplogroups to a fine degree of phylogenetic resolution42, and to cover a wide geographical extent (Supplementary Note 1; Supplementary Fig. 1).
Consistent with the results of previous studies31,33, we found haplogroups C, D and NO to account for around 95% of East Asian male lineages (Supplementary Note 1; Supplementary Figs. 2–7). When incorporating populations from Island Southeast Asia and Oceania, the K2b1 lineage, and three minor sister groups, K2c, K2d and P*, which are indistinguishable in most genotyping platforms, emerge as the fourth predominant lineage43.
We observed low frequencies of Q, R, F* and H at the western fringes of East Asia (Supplementary Note 1; Supplementary Figs. 8–11). We did not consider these lineages further, as prior studies have inferred a Central/North Eurasian origin for the former two31 (Supplementary Note 1; Supplementary Figs. 8, 9) and a South Asian origin for the latter pair44,45 (Supplementary Note 1; Supplementary Figs. 10, 11).
We also considered the frequencies of haplogroups summarised in previous studies involving traditionally hunter-gatherer groups associated with the first layer peopling of East Asia2,3,4,9 (Supplementary Note 1) and paired each with a geographically proximate non-Indigenous population. Doing so revealed a pattern of unity amongst the Indigenous populations, with the enrichment of lineages C, D and K2b1 when compared to each non-Indigenous group, which overwhelmingly carried lineages from the NO clade (Fig. 1).
The four Y-chromosome haplogroups which predominate in East Asian populations (C, D, K2b1/K* and NO) are designated with pink, blue, green and grey pie fillings respectively, with all other haplogroups designated black. *Note that in some datasets from which samples were drawn, the markers required to designate a K* haplotype as K2b1 were not typed, meaning that a small proportion of haplotypes in this category may belong to the rare haplogroups K2c, K2d and P* which have been shown to be entirely restricted to Island Southeast Asia43. Map was sourced from Google Maps using the using the ‘get_googlemap’ function of ggmap129.
Next, we explored the evolutionary dynamics of these Y-chromosome lineages, which can be inferred backwards across time using a phylogenetic toolkit. We inferred a phylogeny from Y-chromosome whole genome sequences from the Human Genome Diversity Panel (HGDP)46, which covers a wide and balanced selection of East Asian populations. To infer trajectories of effective population size for each of our four haplogroups we applied the Bayesian skyline technique, which does so from the distribution and spacing of coalescence events in that phylogeny47,48. Typically, this technique is applied to sequences sampled from a single homogenous population, however it is often applied to individual haplogroups when one has prior reason to associate a specific lineage with a particular population or movement of people49,50,51,52,53,54.
We see a strong contrast between the inferred historic populations sizes of the NO clade, and the three lineages which predominate in traditionally hunter-gatherer groups: C, D and K2b1 (Fig. 2). Lineages C, D and K2b1 are deeply rooted (Fig. 2a), with coalescence times exceeding 50Kya, and each being inferred to have sustained low population sizes until the present (Fig. 2b).
a Phylogeny of Y-chromosome haplogroups NO, K2b1, C and D. b Bayesian skyline plots estimating effective population size through time for each of these four haplogroups. The colour scheme used to designate each of these lineages is the same across all figure panels, and that used in Fig. 1. c Plot of lineages through time for each of the four Y-chromosome haplogroups, with shading indicating 95% highest posterior density (HPD) estimates across all iterations sampled by BEAST.
The NO clade, which diversified more recently, experiences an initial population size increase around 15Kya, and begins a phase of exponential growth around 9Kya, expanding by a factor of almost 3 within the next 3000 years (Fig. 2b; from ~278,000 at 9Kya to ~776,000 at 6Kya). It is notable that the timeframe for this latter episode of growth coincides with the development of agriculture, which is inferred from archaeological data to have begun around 7000–6000BCE in Northern China55,56.
Importantly, we show that the contrasting temporal trends of the two sets of Y-chromosome lineages are consistent under different parameter choices for the skyline technique, when considering N and O lineages separately, and when considering the simpler ‘lineages through time’ trajectory for each haplogroup (Fig. 2c; Supplementary Note 2; Supplementary Figs. 14–16).
We next used the spatial frequency interpolation technique ‘kriging’ to visualise the distributions of the four predominant East Asian lineages, and assess these in light of the Two Layer model (Fig. 3a–d). The maps of population haplogroup frequencies were supplemented with measures of haplogroup diversity, generated from a dataset of over 5,000 STR profiles (Fig. 3; points). These techniques have a long history of use in human population genetics29,30,31,32,57 and allow researchers to infer likely origin points of haplogroups from regions of high diversity58,59,60,61.
a Haplogroup NO. b Haplogroup C. c Haplogroup D. d Haplogroup K2b1. Frequency and diversity colour scales differ between plots. Note that for the interpolation of haplogroup K2b1, only lineages which could be unambiguously assigned to this clade, by possessing derived alleles at markers for its sub-lineages M or S, were counted towards the frequency of this haplogroup. Several Indonesian populations possessed low frequencies of K-M526*. These lineages have been shown by Karafet et al. (2015)43 to mainly belong to K2b1, so the frequency of this lineage is likely to be slightly underestimated in these few Island Southeast Asian populations. Maps were obtained using the ggmap129 package of R.
From this analysis we observe that haplogroup C is found across a wide expanse of East Asia, Island Southeast Asia and Oceania, and at particularly high frequencies in inland and Northern Eurasia (Fig. 3b). We observe that estimates of haplogroup C diversity in inland and northern Eurasia are extremely low relative to modern-day China. Haplogroup D is found at highest frequencies across Tibet, Japan and the Andaman Islands, but also has a patchy distribution across mainland Southeast Asia (Fig. 3c). Haplogroup K2b1 is found exclusively in Papua New Guinea (PNG), Australia, the Philippines and Eastern Indonesia (Fig. 3d). Haplogroup NO is widely distributed, and is the only haplogroup of the four found in North-Western Eurasia (Fig. 3a). The diversity of this lineage appears highest in modern-day China, and Southeast Asia, and drops off when moving east towards the Philippines, and more dramatically when moving west towards India.
Based on the spatial and temporal characteristics of lineages C, D and K2b1, and their enrichment in traditionally hunter-gatherer first layer populations, we infer them to have arrived via the initial southern coastal route migration into East Asia. The NO lineage, which is underrepresented or absent from these populations, and which displays signatures of exponential population growth and spatial characteristics incompatible with a southern entry route, we infer to have arrived via the second layer of peopling.
Mtb genetic variation and the Two Layer model
In light of the above data, we now consider the distributions, frequencies, and phylogenetic characteristics of East Asian Mtb lineages. We visualised the spatial frequency of lineages 1 and 2 using data from a recent joint-analysis62 and supplemented these maps with estimates of lineage diversity inferred from spoligotypes from the SITVIT2 database63.
We found lineage 1 to be distributed around the rim of the Indian Ocean, and to display uniformly high diversity values in the Asian countries which border it (Fig. 4a). Consistent with phylogenetic studies, we calculate low diversity values for lineage 1 in East African countries, as well as the Philippines, where the majority of Mtb lineages belong to the endemic ‘Manila’ lineage, which has a relatively young root64.
(a) lineage 1. (b) lineage 2. Note that the diversity colour scales differ between (a) and (b). Maps were obtained using the ggmap129 package of R.
Lineage 2 is found at highest frequencies across Far East Asia, and extends west to Central Asia, as well as north to Russia (Fig. 4b). Lineage 2 isolates generally attain highest diversities in East Asian regions, including two sampling locations from China and one from Japan. The single highest L2 diversity value is reported for a sampling location in Bangkok, Thailand, however more southerly and westerly Southeast Asian locations consistently displayed lower diversity values. Much like Y-chromosome haplogroup NO, the frequency of lineage 2 drops when moving westwards across the Bay of Bengal into South Asia. Overall, levels of diversity, measured as the probability of selecting isolates with identical spoligotype patterns from a population, were far lower for lineage 2 than for lineage 1.
We propose these spatial frequency and diversity patterns suggest lineage 1 to show characteristics compatible with a southern route entry into East Asia, and lineage 2 to show opposing patterns, which are more consistent with a northern entry route, and more recent expansion. To support this association more formally, we assessed the correlation in frequency between putative first and second layer Mtb and Y-chromosome lineages across countries or major geographical regions in our two genotyping surveys. We found the frequencies of haplogroup NO and Mtb lineage 2 to show a strong positive correlation (Spearman’s rho = 0.65; p = 7.3 × 10−4), and the frequency of lineage 1 to be positively correlated with the summed frequencies of haplogroups C, D and K2b1 (Spearman’s rho = 0.44; p = 0.039; Supplementary Note 3; Supplementary Figs. 21 and 22).
Next, we inferred the phylogenetic characteristics of Mtb lineages 1 and 2 using a dataset of whole genome sequences, to observe features which may be shared with the Y-chromosome tree. We composed a dataset of isolates randomly sampled from available East Asian populations, which included Thailand, Myanmar, Cambodia, Vietnam, China and Japan. This sampling regime was designed to approximate the regime used for the HGDP, from which Y-chromosome dynamics were inferred. For the purposes of this analysis, we are examining qualitative similarities between the Mtb and Y-chromosome phylogenies, including relative divergence times and expansion trajectories of lineages putatively associated with each population movement. Due to inherent differences in the sampling designs of these studies, we did not expect exact congruence in the phylogenetic signatures observed.
As there is ongoing debate surrounding our ability to accurately estimate substitution rates in Mtb, and the stability of the instantaneous mutation rate of Mtb throughout time65,66,67, we chose not to calibrate our phylogeny according to a previously estimated rate. We chose instead to use the phylogenetic technique of ‘node calibration’, in which a tree is calibrated by assigning an age (or probability distribution over a range of ages) to a certain node, based off archaeological or biogeographic data, or signatures of co-divergence68,69,70,71. We chose this method in favour of the ‘tip calibration’ approach used by prior studies, which infer node ages and a substitution rate by incorporating heterochronous sequences into their phylogeny72,73. In addition to requiring adequate temporal structure74,75, tip dating is liable to the time dependent rate phenomenon76,77 (TDRP), which systematically underestimates the ages of deep nodes of a tree (see Discussion). Node calibration, while not subject to the TDRP, does, however, rely on a robust biogeographic or genetic model to justify the codivergence scenario proposed72,73,78 (see Discussion).
We used an approach analogous to Comas et al. (2013)38, who calibrate nodes of the Mtb phylogeny to match key divergence events in the human mitochondrial tree. As we infer Mtb lineage 1 to be the lineage carried by the first layer dispersal into East Asia, we assigned a prior distribution over its root age, such that it matches the average age of first layer Y-chromosome lineages C, D and K2b1 (56.5Ky, ±5Ky). Our inferred timepoint for the coalescence age of lineage 1 (56.5Kya) is therefore intermediate between the co-divergence scenarios proposed by Comas et al. (2013)38 (ranging from 6 to 100Kya), yet older than estimates derived from tip dating such as Menardo et al. (2021)79 (0.85Kya) and O’Neill et al. (2019)37 (2.38Kya; see Supplementary Note 5; Supplementary Table 1 for a comparison of models).
We find that calibrating the tree in this way reveals a correspondence between the root height of putative second layer Mtb lineage 2 (48.9Ky; HPD: 41.4–57.1Ky) and Y-chromosome haplogroup NO (49.8Ky; HPD: 47.8–51.8Ky). We also observed Mtb Lineage 1 to display similar qualitative characteristics to putative first layer Y-chromosome haplogroups C, D and K2b1. These lineages displayed early ‘lineage proliferation’ phases, in which the number of lineages initially expands, followed by a period of little subsequent increase (Fig. 5a).
a Phylogeny inferred from lineage 1 and 2 Mtb sequences from East Asian populations. b Lineage through time trajectories for L1 and L2 subtrees with 95% HPD intervals generated using BEAST. c Lineage through time trajectories for the four predominant East Asian Y-chromosome haplogroups shown in Fig. 2 with 95% HPD interval shading. The colour scheme used for (b and c) was devised so that orange corresponds to first layer Mtb and Y-chromosome lineages/ haplogroups, and grey corresponds to second layer lineages/ haplogroups.
By comparing the tree structures quantitatively using the ‘lineages through time’ metric, we also observe correspondence between the Y-chromosome and Mtb lineages thought to derive from first and second layer groups. Putative first layer lineage counts (Mtb L1, Fig. 5b; Y-chromosome lineages C, D and K2b1, Fig. 5c) gradually increase, while putative second layer lineage counts (Mtb L2, Fig. 5b; Y-chromosome lineage NO, Fig. 5c) expand dramatically closer to the tips of the trees. We also performed a more formal comparison by deriving rates of change from these trajectories38, and contrasting values obtained for putative first and second layer lineages. This approach revealed putative second layer lineages to be characterised by growth rates which were consistently faster than their first layer counterparts (Wilcoxon tests for trajectories L1 vs L2, NO vs C, NO vs D and NO vs K2b1; all p values < 0.038; Supplementary Note 4; Supplementary Fig. 23). This trend is particularly pronounced when considering the timeframe spanning the Neolithic period (10Kya) until the present (all p values < 2.62 × 10−6).
In line with the divergence ages we estimate, our inferred substitution rate (2.16 × 10−9s/s/y) is intermediate between the various estimates of Comas et al. (2013)38 (9.42 × 10−10 to 1.66 × 10−8), but slower than estimates derived from tip dating using contemporary samples80 (around 1×10−7). As detailed subsequently (see Discussion), this magnitude of substitution rate differences is typical of other pathogens when applying these two techniques over comparable timescales81,82. We also discuss the plausibility of the inferred substitution rate given available estimates from Mtb during latency83,84, and when considering patterns of rate heterogeneity within similar bacteria85,86. Although recent studies have implemented methods for modelling variation in substitution rates over different timescales70,76, the lack of additional high-confidence internal calibration points within the Mtb tree precluded us from applying these approaches with certainty.
In sum, these data draw a link between Mtb lineage 1 and Y-chromosome haplogroups C, D and K2b1 and the initial first layer hunter-gatherer presence in East Asia, and between Mtb lineage 2 and the second layer.
Novel deeply rooted Mtb L1 sublineages in Eastern Indonesia and PNG
We now describe a final line of suggestive evidence to link patterns of Mtb variation with the Two Layer model, by documenting a relevant aspect of lineage 1 phylogeographic diversity. Knowing that the Indigenous populations of Oceania and Eastern Indonesia represent largely unadmixed descendants of the first layer of peopling, we aimed to characterise the diversity of Mtb lineage 1 in this region. We reason that the identification of deeply rooted lineage 1 sublineages within these populations would support the association of lineage 1 and the first layer dispersal into the region.
We assembled a dataset of isolates from lineage 1.2.1, which is the predominant L1 sublineage in Island Southeast Asia, from three previous studies79,87,88. We chose to downsample the high number of isolates belonging to the well characterised ‘Manila’ and ‘Nonthaburi’ clades. To assess the relative divergence point of L1.2.1 with respect to other sublineages of L1, we incorporated isolates (where available) representing the single deepest split within the remaining L1 sublineages: L1.1.1, L1.1.2, L1.1.3 and L1.2.287. We inferred a phylogeny, and labelled L1.2.1 sublineages according to the nomenclature of three available schemes, but here refer to them as Clades 1 to 4 (Fig. 6).
Tip point colours and shapes correspond to geographical sampling locations, with red and blue circles designating Island Southeast Asia and PNG respectively, and grey shapes representing all other regions. L1.2.1 clades were labelled according to three alternate nomenclature schemes for additional context. Isolates representing the single deepest split in each of the remaining L1 sublineages are included. Each of these sublineage diversification events is marked with a purple circle, with purple tip labels indicating the sublineage of these isolates.
Figure 6 shows the L1.2.1 sublineage to be characterised by a series of deep splits, each of which results in descendant isolates sampled from either Island Southeast Asia or PNG. Clade 1, which includes the common Manila and Nonthaburi genotypes, occurs most commonly in Southeast Asia and shows evidence of recent expansion to other global regions. The most deeply rooted isolate from this clade was sampled from Borneo.
Clades 2, 3 and 4 form a monophyletic group, which diversified very soon after the origin of the L1.2.1 sublineage. Clade 3 was isolated only from patients in East Timor and PNG, and a single European patient79,87. The split between East Timorese and Papuan isolates within this clade was extremely deep. Clade 2 was found in patients from both Europe and Africa, but also possessed a deeply rooted isolate sampled from Borneo. Clade 4, a sublineage documented for the first time here, was isolated from 8 patients from Indonesia and one from Europe. The under-representation of Island Southeast Asia and Oceania in studies of Mtb genomics may mean that other deeply rooted L1.2.1 lineages have yet to be sampled.
We considered the relative divergence point of L1.2.1 sublineages and found them to match those of other sublineages of L1, implying that Mtb has been present in Island Southeast Asia for as long as it has been present in other regions around the rim of the Indian Ocean. We therefore propose that the strong regional structure and deep divergence of a sublineage of L1 in Oceania and Eastern Indonesia may couple this Mtb lineage with the human populations which arrived here as part of the first layer of peopling.
Synthesising the above data, we present in Fig. 7 our model of human-Mtb coexpansion as it pertains to the Two Layer hypothesis.
Discussion
We have conducted a joint analysis of the spatial and temporal dynamics of human Y-chromosome and Mtb sequences in East Asia, and synthesised a coevolutionary model compatible with the Two Layer hypothesis. Under this model, we propose Y-chromosome lineages C, D and K2b1 to have arrived in East Asia via an initial southern coastal route migration, by human populations which also carried Mtb lineage 1 (Fig. 7). We suggest the Y-chromosome NO clade to have arrived via a northern route expansion across Eurasia, carried by human populations carrying Mtb lineage 2 (Fig. 7).
Haplogroups C, D and K2b1 share the phylogenetic characteristics of deep roots and sparse internal branches, and are enriched in the traditionally hunter-gatherer human populations of East Asia. These haplogroups do not show evidence of extreme population growth associated with the development of agriculture and show sparse and scattered spatial frequency distributions. The diversity of the most prevalent of these haplogroups, haplogroup C, decreases dramatically when moving westwards into inland Eurasia.
We have inferred Mtb lineage 1 to also be characterised by a deep divergence time and sparse internal branches, and to display spatial frequency and diversity characteristics most compatible with a southern entry route. We propose that the presence of deeply rooted and regionally diverse lineage 1 sublineages in PNG and Eastern Indonesia also supports the association of this Mtb lineage with this layer of peopling. Prior archaeological studies have documented representatives of these populations to have settled soon after the initial dispersal of humans out of Africa89. This conclusion is reflected in the phylogenetic characteristics of both Y-chromosome and mitochondrial lineages amongst Indigenous Oceanians, which are deeply diverged from other worldwide lineages, and which undergo an early and rapid diversification from one another approximately ~45Kya90. The mirroring of these characteristics in the Mtb lineage 1 phylogenetic tree supports its association with the initial hunter-gatherer populations of the region.
Conversely, the temporal dynamics of the NO clade suggest it to have expanded at a timepoint concurrent with the development of agriculture, and to exhibit highest frequencies in Chinese populations. The connection of this lineage with this cultural development is further supported by its underrepresentation in traditionally hunter-gatherer populations in East Asia, and its strong enrichment in nearby agriculturalist groups. A dispersal trajectory of the NO lineage north of the Himalayas can be inferred from the extremely low diversities of this haplogroup in South Asian populations. We also note that the presence of a basal K2ba* haplogroup, which is a precursor to the NO lineage, retrieved from Ust'-Ishim in central Eurasia ~40Kya also supports a northern route origin of this lineage91.
We propose the phylogenetic characteristics of the NO clade to mirror those of Mtb lineage 2, as both NO and L2 display similar origin points relative to first layer lineages, and both undergo rapid phases of lineage expansion at similar phylogenetic depths. We infer the region of highest frequency and diversity of Mtb lineage 2 to be central East Asia, and find low frequencies and diversities of this lineage in regions along the proposed southern entry route.
There are some limitations in the scope of the presented model which, due largely to data availability, deals only with the populations of East Asia, and not with groups from more westerly regions. It is evident from both our analysis and prior surveys37,92, that Mtb lineage 1 has a strong presence in South Asian populations, particularly southern India, and that certain L1 sublineages are geographically restricted to this region79,87. Conducting a more detailed analysis of human and Mtb coevolution in South Asia, although not undertaken here, will therefore be necessary to fully account for the patterns of spatial frequency and diversity of lineage 1 depicted in Fig. 4a.
While we did not undertake this analysis, we do observe congruent features in the Mtb and Y-chromosome composition of South Asian populations. Prior studies have shown Indian populations to be modelled as mixtures of Ancestral Northern Indian (ANI) and Ancestral Southern Indian (ASI) ancestry, the latter component of which is linked to Indigenous first peoples of East Asia such as the Andamanese93. Populations enriched for ASI ancestry, which are predominantly located in southern India, possess high frequencies of haplogroup H44, which we note to possess the phylogenetic characteristics of deep divergence times and limited population growth46. Considering south Indian populations are also enriched for Mtb lineage 194, we tentatively propose Y-chromosome haplogroup H to be associated with the initial hunter-gatherer presence in South Asia, and to be associated with this Mtb lineage.
We have also emitted some Y-chromosome and Mtb lineages from our model due to their low frequencies or lack of available data. One notable omission is Mtb lineage 4, which is present in low to moderate frequencies in some East Asian populations35,95,96. We omitted this lineage, as our focus was on Mtb and Y-chromosome lineages originating in East Asia, hence why Y-chromosome haplogroups Q, R, F* and H were not analysed extensively. As prior phylogeographic studies have shown the majority of L4 sublineages to display European origin points40, and a history of multiple introductions into Southeast Asia35, we chose to exclude this lineage from consideration. Interestingly, however, a small number of L4 sublineages are geographically restricted to East Asia, with highest frequencies in Chinese populations (L4.4 and 4.595,97). More detailed analysis may reveal these to be minor lineages associated with the second layer of peopling, or later demographic processes outside the scope of this investigation. For context, lineage 4 was only observed at a frequency >20% in the single Chinese cohort included in this analysis95. Spatial frequency interpolation also revealed low L4 frequencies in East Asia (Supplementary Note 3; Supplementary Fig. 20).
We wish to note that our model of historic human and Mtb co-expansion is one which infers a Paleolithic origin of contemporary Mtb variation. Recent molecular dating studies have inferred substitution rates which favour a Neolithic origin of Mtb, likely within the past 6000 years98,99,100, as opposed to the past 60,00038,39. We highlight three factors to note when considering Paleolithic vs Neolithic Mtb emergence theories.
Firstly, none of the recent studies inferring substitution rates on Mtb used genomes older than 350 years98,99, and the three most ancient Mycobacterium tuberculosis complex (MTBC) isolates analysed, incidentally all belonging to M. pinnipedii, were <1Ky old100. The phenomenon of ‘time dependency’ in substitution rate estimates is well documented in viruses and bacteria, with studies documenting systematic over-estimation of mutation rates when calibrating based on more recent samples67,73,101.
We note the differences in substitution rates inferred from our analysis (2.16 × 10−9) relative to tip dating80 (~1 × 10−7), are roughly within the ranges established for other pathogens81,82. Hepatitis B virus substitution rate estimates vary around two orders of magnitude when comparing human co-divergence calibration with tip dating81, and foamy virus substitution rates vary up to four when considering primate species co-divergence82. Available data for bacteria, although more limited, has suggested an order of magnitude difference in substitution rates inferred from contemporary samples vs samples approximately 1Ky old67. A recent review collating estimated mutation and substitution rates in Mtb specifically has also concluded the presence of a time-dependence effect102, although high quality DNA from more ancient Mtb genomes will be required to properly gauge its magnitude80.
Secondly, prior studies have proposed that the mutation rate of Mtb may not be static across time65. The mutation rate quantifies the instantaneous speed of mutation accumulation within Mtb bacteria, and therefore differs from the substitution rate, which describes a rate inferred using a temporal phylogeny103. Given the drastic transition in lifestyles from hunter-gatherers to agriculturalists in the Neolithic, it has been postulated that the Mtb pathogen also underwent a shift in life history strategy65,66. It is possible that in Paleolithic times, Mtb genomes spent more time in a latent state, and accumulated mutations at a slower rate than in the present day65.
Available data from other bacteria supports dormancy-related levels of rate variation which may be capable of explaining the substitution rate we infer for Mtb. Cui et al. (2012)85 find that branches of the Y. pestis phylogeny, a bacterial species with a similar genome size and general mutation rate to Mtb, can vary in substitution rates by 40-fold, from ~1.3 × 10−7 to 3.1 × 10−9. They attribute this variation to the bacteria’s ability to enter a dormant phase between epidemic periods. Similar substitution rates and levels of rate heterogeneity are also reported by Spyrou et al. (2019)86 for Y. pestis. It is plausible that the combination of latency and the relatively stronger effect of purifying selection in Mtb (dN/dS = 0.6104); relative to Y. pestis (dN/dS = 0.985); could explain the slow long-term substitution rate we infer.
In addition, data describing the mutation rate in human Mtb infection during latency supports a slowdown effect84, and mutation rates broadly consistent with Paleolithic emergence theories during long-term latency. When analysing samples with an apparent latency period of around 20 years, Colangeli et al. (2014)83, for instance, propose a mutation rate of 7.01 × 10−9. It should be noted, however, that these studies are sparse, have low sample sizes, and have produced conflicting results105. Collectively though, these data provide reasonable justification to credit the likelihood of a broad range of possible substitution rates for Mtb.
A final form of data which should be taken into account when considering Mtb emergence scenarios comes from the fields of paleomicrobiology and ancient DNA106,107,108. These studies, which amplify Mtb sequence motifs, have produced evidence consistent with the presence of the pathogen in samples from Neolithic Israel109,110, Syria111,112, Germany113, Hungary114,115,116, Egypt117 and Britain118, dating up to an estimated 10.8Kya. Multiple studies have also detected the TbD1 deletion110,113,117,118, implying that evolutionarily modern lineages were present at these timepoints. Further scrutiny will be required to verify these findings, and claims addressing the impact of mycobacterial contaminants, which have been previously brought into question119,120.
To facilitate a comparison of possible Mtb emergence theories, we present in Supplementary Table 1 the substitution rates and node ages inferred using our model and several alternate models based on either tip or node calibration. A key line of evidence in support of the Paleolithic emergence scenario we propose is that Mtb node ages appear to correspond to key demographic events in human history.
For instance, when calibrating the Mtb phylogeny so that the age of putative first layer lineage (L1) matches that of first layer Y-chromosome lineages, we also see a strong similarity between the inferred ages of putative second layer lineages and haplogroups (NO and L2; both ~49Kya). This date is also consistent with the date proposed by Matsumura et al. (2019)2 for the arrival of this second layer dispersal on the basis of archaeological data (45Kya).
Our model also predicts a coalescence point for the entire MTBC that is consistent with the deepest splits of human Y-chromosome lineages within the African continent. Mtb lineages 1 and 2 coalesce 112.4Kya (HPD: 94.1–131.9Kya), and prior phylogenetic analyses show the split of M. africanum lineages (L5 and L6) to be roughly 10% older than this (Comas et al. 2013)38. Aside from the rare West African A00 haplogroup discovered after extensive consumer genetic testing (divergence time: 250Kya), the divergence times of all major Y-chromosome haplogroups fall between roughly 60 and 190Kya27,121. The localisation of the most deeply rooted Mtb lineages to the African continent, with phylogenetic depths comparable with human uniparental lineages supports an out-of-Africa co-dispersal scenario for the pathogen22.
Paleolithic emergence scenarios also receive support from the geospatial dynamics of Mtb lineages, as well as their clinical characteristics and life history strategies. The ‘evolutionarily modern’ lineages 2, 3 and 4 show expansion centres corresponding to major centres of agriculture in Asia and Europe, and levels of transmissibility suggesting adaptation to high population sizes39,122,123. By contrast, the less virulent and transmissible lineage 1 has been hypothesised to be adapted to hunter-gatherer human populations66. Our model therefore provides an explicit demographic explanation for the distribution and dynamics of lineage 1, by linking it to the original hunter-gatherer populations of East Asia.
Available tip calibration scenarios, on the other hand, don’t provide node ages as immediately parsimonious with known human demographic events, or at least, no single model linking these dates and events has been proposed to date. This, however, does not necessarily mean that Mtb node ages and evolutionary dynamics dated using tip calibration are incompatible with Paleolithic emergence theories. As others have hypothesised67,80, substitution rates inferred using tip calibration may be appropriate for dating relatively shallow nodes within the tree, but not for the deep nodes we focus on in this study. The combination of ancient DNA, and research into the demographic factors underpinning the spread of Mtb lineages will clarify this matter.
The final limitation of this analysis we note is that inferences of admixture proportions in human populations can be confounded when analysing patterns of uniparental genetic variation. Biases in the relative contribution of male vs female lineages from a given ancestry group present themselves when considering most large-scale admixture events in modern times. These include, for instance, the events associated with the colonisation of the Americas124 and Greenland125, as well as the initial peopling of Polynesia126 and Madagascar127. We caution that Y-chromosome lineages from the first layer of peopling may have contributed in disproportionately low frequencies to the present-day genetic constitution of East Asian populations. This factor may explain the finding that putative first layer Mtb lineages are comparatively more numerous than first layer Y-chromosome lineages across many contemporary East Asian populations and show signs of more pronounced lineage expansion (Fig. 5). It follows that exploration of the dynamic of the Two Layer model from the perspective of maternal lineages may yield different inferred ancestry proportions and should be an avenue of future research.
Methods
Ethics declaration
All datasets included in this analysis were retrieved from prior peer reviewed publications. No datasets containing identifiable information were included. All studies selected document receiving informed consent and/ or ethical approval from relevant organisation(s).
Y-chromosome haplogroup frequency analysis
Y-chromosome haplogroup frequencies were collated from previous studies, described in greater detail in Supplementary Note 1. We ensured that individuals were genotyped to a high enough degree of resolution to preclude the possibility of spuriously classifying NO*, L, or T lineages as K-M526*42. The studies from which the frequencies of Y-chromosome haplogroups in traditionally hunter-gatherer populations were summarised are discussed in Supplementary Note 1, along with the rationale for the comparison of each non-Indigenous group.
Spatial interpolation of haplogroup frequencies was carried out using the kriging procedure. This process includes fitting a variogram describing the relationship between geographic distance and haplogroup frequency differences for the chosen haplogroup, before interpolating the frequency of this haplogroup across a spatial surface. These steps were completed using various functions of the gstat package in R128 and visualised using ggmap129. We provide an illustration of the concordance between observed and interpolated haplogroup frequencies in Supplementary Note 1 (Supplementary Figs. 4–7), and also present kriging plots (Supplementary Note 1, Supplementary Figs. 8–11) for additional Y-chromosome haplogroups. Also included and separate kriging plots for haplogroups N and O (Supplementary Figs. 12 and 13).
Y-chromosome haplogroup diversity analysis
Y-chromosome STR profiles were obtained from previous studies29,31,59,60,130,131,132,133,134,135,136,137. Six short tandem repeats which were genotyped across all datasets were used for diversity calculations (DYS389I, DYS389II, DYS390, DYS391, DYS392 and DYS393). Minor discrepancies in the way the DYS389II repeat length was recorded between studies were corrected for, and haplotype diversity was measured using the approach of Nei (1987)138 using custom Unix scripts.
Y-chromosome phylogenetics
Genotypes called across the Y-chromosome for samples from the HGDP46 were restricted to biallelic SNPs and used to produce a fasta alignment of all variable sites. The regions of the Y-chromosome analysed were further restricted to the 10.4 Mb shown to be amenable to phylogenetic research71. BEAST139 was used to construct a Y-chromosome phylogeny describing the relationship between all lineages in this dataset falling into clades C, D, K2b1 and NO, which were identified from data presented in the supplementary material of Bergstrom et al. (2020)46. The BEAST parameters included a GTR model of nucleotide substitution, an uncorrelated lognormal relaxed clock to model rate heterogeneity between branches, a Coalescent Bayesian skyline tree prior, and a Y-chromosome specific mutation rate derived from the analysis of Fu et al. (2014)140. This rate was corrected to account for the fact that the alignment was restricted to variable sites only. BEAST was run in five independent replicates of 10,000,000 iterations, sampling every thousandth, before combining with LogCombiner. A Maximum Clade Credibility tree was produced using TreeAnnotator. Key splits in the phylogeny received high posterior support and were in fine agreement with those of Bergstrom et al. (2020)46 on the basis of both branching pattern and node age. Bayesian skyline estimates of effective population size were obtained after running BEAST on samples from each haplogroup individually for 30,000,000 iterations.
Mtb frequency and diversity analysis
Kriging of Mtb lineage frequencies was carried out using an identical approach to that described above for Y-chromosome haplogroups. We used the dataset of Wiens et al. (2018), which describes global lineage frequencies, to perform kriging and calculated spoligotype diversity using data from SITVIT263. We calculated diversity values for spoligotype families representative of lineages 1 (all ‘EAI’ clades) and 2 (‘Beijing’ clades) at each sampling location using the same diversity index as described for Y-chromosome haplogroups. We excluded isolates which were sampled from immigrants, and populations represented by a low number of isolates (n = 40) for each lineage. As per the rationale of prior studies37,79, we have not shown diversity values for countries in Western Europe or Australia, however frequency values for these countries are still shown. Plots showing the concordance of observed and interpolated Mtb lineage frequencies for lineages 1, 2, 3 and 4 are shown in Supplementary Figs. 17–20.
Mtb phylogenetics—lineage 1 and 2 analysis
We collated Mtb WGS data from prior studies which conducted random sampling within East Asian populations, and which weren’t enriched for drug resistant isolates35,95,96,141,142,143. To obtain an unbiased comparison of the dynamics of Mtb and Y-chromosome evolution, we down-sampled isolates from each of these studies so that the total number of Mtb lineage 1 and 2 isolates matched the number of Y-chromosome sequences from which the phylogeny was inferred. Fastq files for all isolates were downloaded from the Sequence Read Archive, and were quality trimmed using cutadapt144. Reads were then aligned to the H37Rv reference genome using BWA145, filtered for duplicates using Picard146, and subjected to variant calling using Pilon147. Poor quality and ambiguous variant calls were then set to missing, and a lineage was assigned to each sample using ‘fast-lineage-caller’148 and the nomenclature scheme of Coll et al. (2014)149. All per-sample vcf files were then merged, and both per-site and per-sample missingness filters were applied using bcftools150. Finally, variant calls were restricted to those in uniquely mappable regions of the Mtb genome before phylogenetic inference, using the coordinates of Brynildsrud et al. (2018)40.
As described in the Main Text, we used the phylogenetic technique of ‘node calibration’68,69,70,71 to scale our Mtb tree using signatures of human-Mtb co-divergence. We initially inferred an Mtb phylogeny from a sequence alignment incorporating all L1 and L2 isolates, assigning a prior distribution over the age of the L1 root. For this prior, we used a lognormal distribution with mean value of 56.5Ky (the average age of Y-chromosome haplogroups C, D and K2b1) ± 5Ky. Our phylogenetic parameters were otherwise the same as used for the Y-chromosome analysis, including the use of an uncorrelated lognormal relaxed molecular clock. We again ran BEAST in five independent replicates, before combining log files using LogCombiner. We obtained a substitution rate from this log file (2.16 × 10−9s/s/y; HPD: 1.92 × 10−9 to 2.40 × 10−9) after correcting for the fact that only polymorphic sites were included in the alignment. The resulting Maximum Clade Credibility tree is presented in Fig. 5a. To generate the accompanying ‘lineages through time’ trajectories for each lineage (Fig. 5b), we ran BEAST independently for 50,000,000 iterations on alignments which had been subset to include only L1 and L2 sequences respectively. We calibrated these trees by assigning the same prior distribution to the L1 root as described above (56.5Ky ± 5Ky), and setting a prior on the L2 root to match the age inferred in the tree in Fig. 5a (48.9Ky ± 5Ky).
Mtb phylogenetics—lineage 1.2.1 analysis
To infer a phylogeny representative of known lineage 1.2.1 diversity, isolates used in the investigations of Menardo et al. (2021)79 and Netikul et al. (2022)87 were collated. As detailed sublineage data was available for the isolates from Netikul et al. (2022)87, we were able to preferentially down-sample isolates from Clade 1 (Fig. 7; designated L1.2.2 by Netikul et al.) to 5 representative lineages. All isolates documented by Menardo et al. (2021)79 were include in our analysis, and we also included all L1.2.1 isolates from the study of Meumann et al. (2021)88. The protocol for downloading and calling variants on these isolates was identical to that used above for the Mtb ‘lineage 1 and 2’ dataset, with the only difference being an additional adapter inference and filtering step151, and appropriate pipeline adjustments for samples with single end reads. Representative isolates from lineages 1.1.1.1, 1.1.1, 1.1.2, 1.1.3, 1.2.2 and 5 were also downloaded and processed in the same manner as other samples. We used data presented by Netikul et al. (2022)87 to identify isolates representing the deepest split in each L1 sublineage. We inferred a phylogeny using RAxML152 using a GTR substitution model, an L5 outgroup, and annotated lineages according to various nomenclature schemes87,148,153.
Statistics and reproducibility
All studies from which data was drawn are referenced in the text. Only studies which used random sampling regimes (as opposed to preferentially selecting samples belonging to a particular Y-chromosome haplogroup, or with a particular drug resistant status) were used. Samples were only filtered on the basis of sequencing data quality. For the Mtb phylogenetic analyses [‘lineage 1 and 2 analysis’ (Fig. 5) and ‘lineage 1.2.1 analysis’ (Fig. 6)] that involved random sub-sampling, lists of the isolates that were retained are provided in Supplementary Data 1.
Multiple independent replicates of each BEAST phylogenetic analysis were run to check for consistency, and skyline trajectories were replicated after running BEAST with different parameter values (see Supplementary Note 2). A p value threshold of 0.05 was used for the statistical tests reported in Supplementary Notes 3 and 4.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
Data availability
All data analysed in this paper was drawn from previously published studies, as specified in the text and in Supplementary Data 1. Raw lineage through time values depicted in Figs. 2c, 5b and 5c are provided in Supplementary Data 1. The proportions depicted in the pie charts in Fig. 1 are listed in Supplementary Note 1, section iv.
References
Wang, C. C. et al. Genomic insights into the formation of human populations in East Asia. Nature 591, 413–419 (2021).
Matsumura, H. et al. Craniometrics Reveal “Two Layers” of Prehistoric Human Dispersal in Eastern Eurasia. Sci. Rep. 9, 1451 (2019).
McColl, H. et al. The prehistoric peopling of Southeast Asia. Science 361, 88–92 (2018).
Lipson, M. et al. Ancient genomes document multiple waves of migration in Southeast Asian prehistory. Science 361, 92–95 (2018).
Yang, M. A. et al. Ancient DNA indicates human population shifts and admixture in northern and southern China. Science 369, 282–288 (2020).
Matsumura, H. et al. Morphometric affinity of the late Neolithic human remains from Man Bac, Ninh Binh Province, Vietnam: key skeletons with which to debate the “two layer” hypothesis. Anthropol. Sci. 116, 135–148 (2008).
Turner, C. G. Major features of Sundadonty and Sinodonty, including suggestions about East Asian microevolution, population history, and late Pleistocene relationships with Australian Aboriginals. Am. J. Phys. Anthropol. 82, 295–317 (1990).
Corny, J. et al. Dental phenotypic shape variation supports a multiple dispersal model for anatomically modern humans in Southeast Asia. J. Hum. Evol. 112, 41–56 (2017).
Bellwood, P. S. First Islanders: Prehistory and Human Migration in Island Southeast Asia (Wiley Blackwell, 2017).
Bellwood, P. S. (Ed.). The Global Prehistory of Human Migration (Wiley-Blackwell, 2015).
Hanihara, K. Dual structure model for the population history of Japanese. Jpn. Rev. 2, 1–33 (1991).
Hanihara, K. Origins and affinities of Japanese viewed from cranial measurements. Acta Anthropogenet 8, 149–158 (1984).
Suzuki, H. Microevolutionary changes in the Japanese population from the prehistoric age to the present day. J. Fac. Sci., Univ. Tokyo Sect. V. 3, 279–308 (1969).
Yamaguchi, B. The incidence of minor non-metric cranial variants in the protohistoric human remains from eastern Japan. Bull. Natl Sci. Mus., Ser. D. 11, 13–24 (1985).
Ossenberg, N. S., Dodo, Y., Maeda, T. & Kawakubo, Y. Ethnogenesis and craniofacial change in Japan from the perspective of nonmetric traits. Anthropol. Sci. 114, 99–115 (2006).
Fukase, H. et al. Facial characteristics of the prehistoric and early-modern inhabitants of the Okinawa islands in comparison to the contemporary people of Honshu. Anthropol. Sci. 120, 23–32 (2012).
Kawakubo, Y., Hanihara, T., Shigematsu, M. & Dodo, Y. Interpretation of craniometric variation in northeastern Japan, the Tohoku region. Anthropol. Sci. 117, 57–65 (2009).
Jinam, T. A., Kanzawa-Kiriyama, H. & Saitou, N. Human genetic diversity in the Japanese Archipelago: dual structure and beyond. Genes Genet. Syst. 90, 147–152 (2015).
Dodo, Y. & Ishida, H. Nonmetric analysis of doighama crania of the aenolithic yayoi period in western japan, in The Genesis of the Japanese population and Culture (Rokko-Shuppan, 1988).
Temple, D. H., Auerbach, B. M., Nakatsukasa, M., Sciulli, P. W. & Larsen, C. S. Variation in limb proportions between Jomon foragers and Yayoi agriculturalists from prehistoric Japan. Am. J. Phys. Anthropol. 137, 164–174 (2008).
Gakuhari, T. et al. Ancient Jomon genome sequence analysis sheds light on migration patterns of early East Asian populations. Commun. Biol. 3, 437 (2020).
Forni, D., Cagliani, R., Clerici, M. & Sironi, M. Disease-causing human viruses: novelty and legacy. Trends Microbiol. 30, 1232–1242 (2022).
Yuen, L. K. W. et al. Tracing Ancient Human Migrations into Sahul Using Hepatitis B Virus Genomes. Mol. Biol. Evol. 36, 942–954 (2019).
Moodley, Y. & Linz, B. Helicobacter pylori Sequences Reflect Past Human Migrations. Microb. Pathogenom. 6, 62–74 (2009).
Hammer, M. F. et al. Out of Africa and back again: nested cladistic analysis of human Y chromosome variation. Mol. Biol. Evol. 15, 427–441 (1998).
Underhill, P. A. et al. Y chromosome sequence variation and the history of human populations. Nat. Genet. 26, 358–361 (2000).
Poznik, G. D. et al. Punctuated bursts in human male demography inferred from 1,244 worldwide Y-chromosome sequences. Nat. Genet. 48, 593–599 (2016).
Jobling, M. A. & Tyler-Smith, C. Human Y-chromosome variation in the genome-sequencing era. Nat. Rev. Genet. 18, 485–497 (2017).
Rootsi, S. et al. A counter-clockwise northern route of the Y-chromosome haplogroup N from Southeast Asia towards Europe. Eur. J. Hum. Genet. 15, 204–211 (2007).
Shi, H. et al. Y chromosome evidence of earliest modern human settlement in East Asia and multiple origins of Tibetan and Japanese populations. BMC Biol. 6, 45 (2008).
Zhong, H. et al. Extended Y Chromosome Investigation Suggests Postglacial Migrations of Modern Humans into East Asia via the Northern Route. Mol. Biol. Evol. 28, 717–727 (2011).
Shi, H. et al. Genetic Evidence of an East Asian Origin and Paleolithic Northward Migration of Y-chromosome Haplogroup N. PLoS ONE 8, e66102 (2013).
Wang, C.-C. & Li, H. Inferring human history in East Asia from Y chromosomes. Investig. Genet. 4, 11 (2013).
Liu, Q. et al. China’s tuberculosis epidemic stems from historical expansion of four strains of Mycobacterium tuberculosis. Nat. Ecol. Evol. 2, 1982–1992 (2018).
Holt, K. E. et al. Frequent transmission of the Mycobacterium tuberculosis Beijing lineage and positive selection for the EsxW Beijing variant in Vietnam. Nat. Genet. 50, 849–856 (2018).
Coscolla, M. & Gagneux, S. Consequences of genomic diversity in Mycobacterium tuberculosis. Semin. Immunol. 26, 431–444 (2014).
O’Neill, M. B. et al. Lineage specific histories of Mycobacterium tuberculosis dispersal in Africa and Eurasia. Mol. Ecol. 28, 3241–3256 (2019).
Comas, I. et al. Out-of-Africa migration and Neolithic coexpansion of Mycobacterium tuberculosis with modern humans. Nat. Genet. 45, 1176–1182 (2013).
Luo, T. et al. Southern East Asian origin and coexpansion of Mycobacterium tuberculosis Beijing family with Han Chinese. Proc. Natl Acad. Sci. 112, 8136–8141 (2015).
Brynildsrud, O. B. et al. Global expansion of Mycobacterium tuberculosis lineage 4 shaped by colonial migration and local adaptation. Sci. Adv. 4, https://doi.org/10.1126/sciadv.aat5869 (2018).
Merker, M. et al. Evolutionary history and global spread of the Mycobacterium tuberculosis Beijing lineage. Nat. Genet. 47, 242–249 (2015).
Karafet, T. M. et al. New binary polymorphisms reshape and increase resolution of the human Y chromosomal haplogroup tree. Genome Res. 18, 830–838 (2008).
Karafet, T. M., Mendez, F. L., Sudoyo, H., Lansing, J. S. & Hammer, M. F. Improved phylogenetic resolution and rapid diversification of Y-chromosome haplogroup K-M526 in Southeast Asia. Eur. J. Hum. Genet. 23, 369–373 (2015).
Trivedi, R. et al. Genetic Imprints of Pleistocene Origin of Indian Populations: A Comprehensive Phylogeographic Sketch of Indian Y-Chromosomes. Int. J. Hum. Genet. 8, 97–118 (2008).
Cordaux, R. et al. Independent Origins of Indian Caste and Tribal Paternal Lineages. Curr. Biol. 14, 231–235 (2004).
Bergström, A. et al. Insights into human genetic variation and population history from 929 diverse genomes. Science 367, eaay5012 (2020).
Drummond, A. J., Rambaut, A., Shapiro, B. & Pybus, O. G. Bayesian Coalescent Inference of Past Population Dynamics from Molecular Sequences. Mol. Biol. Evol. 22, 1185–1192 (2005).
Ho, S. Y. & Shapiro, B. Skyline-plot methods for estimating demographic history from nucleotide sequences. Mol. Ecol. Resour. 11, 423–434 (2011).
Atkinson, Q. D., Gray, R. D. & Drummond, A. J. Bayesian coalescent inference of major human mitochondrial DNA haplogroup expansions in Africa. Proc. R. Soc. B: Biol. Sci. 276, 367–373 (2009).
Pennarun, E. et al. Divorcing the Late Upper Palaeolithic demographic histories of mtDNA haplogroups M1 and U6 in Africa. BMC Evolut. Biol. 12, 234 (2012).
Gandini, F. et al. Mapping human dispersals into the Horn of Africa from Arabian Ice Age refugia using mitogenomes. Sci. Rep. 6, 25472 (2016).
Soares, P. et al. The Expansion of mtDNA Haplogroup L3 within and out of Africa. Mol. Biol. Evol. 29, 915–927 (2012).
Duggan, A. T. et al. Maternal History of Oceania from Complete mtDNA Genomes: Contrasting Ancient Diversity with Recent Homogenization Due to the Austronesian Expansion. Am. J. Hum. Genet. 94, 721–733 (2014).
Watanabe, Y. et al. Analysis of whole Y-chromosome sequences reveals the Japanese population history in the Jomon period. Sci. Rep. 9, 8556 (2019).
Crawford, G. W. “East Asian Plant Domestication” in Archaeology of Asia, (Ed. Stark, M.), 77–95 (Blackwell Publishing, 2006).
Zhang, C. & Hung, H. C. “East Asia: archaeology” in The Global Prehistory of Human Migration, (Ed. Bellwood, P. S.) 534–553, (Blackwell Publishing, 2014).
Zhong, H. et al. Global distribution of Y-chromosome haplogroup C reveals the prehistoric migration routes of African exodus and early settlement in East Asia. J. Hum. Genet. 55, 428–435 (2010).
Barbujani, G. Geographic patterns: how to identify them and why. Hum. Biol. 72, 133–153 (2000).
Sengupta, S. et al. Polarity and Temporality of High-Resolution Y-Chromosome Distributions in India Identify Both Indigenous and Exogenous Expansions and Reveal Minor Genetic Influence of Central Asian Pastoralists. Am. J. Hum. Genet. 78, 202–221 (2006).
Trejaut, J. A. et al. Taiwan Y-chromosomal DNA variation and its relationship with Island Southeast Asia. BMC Genet. 15, 77 (2014).
Zhang, X. et al. Y-chromosome diversity suggests southern origin and Paleolithic backwave migration of Austro-Asiatic speakers from eastern Asia to the Indian subcontinent. Sci. Rep. 5, 15486 (2015).
Wiens, K. E. et al. Global variation in bacterial strains that cause tuberculosis disease: a systematic review and meta-analysis. BMC Med. 16, 196 (2018).
Couvin, D., David, A., Zozio, T. & Rastogi, N. Macro-geographical specificities of the prevailing tuberculosis epidemic as seen through SITVIT2, an updated version of the Mycobacterium tuberculosis genotyping database. Infect., Genet. Evol. 72, 31–43 (2019).
Phelan, J. E. et al. Mycobacterium tuberculosis whole genome sequencing provides insights into the Manila strain and drug-resistance mutations in the Philippines. Sci. Rep. 9, 9305 (2019).
Barbier, M. & Wirth, T. The Evolutionary History, Demography, and Spread of the Mycobacterium tuberculosis Complex. Microbiol. Spect. 4 https://doi.org/10.1128/microbiolspec.TBTB2-0008-2016 (2016).
Gagneux, S. Ecology and evolution of Mycobacterium tuberculosis. Nat. Rev. Microbiol. 16, 202–213 (2018).
Duchêne, S. et al. Genome-scale rates of evolutionary change in bacteria. Microb. Genom. 2, e000094 (2016).
Balanovsky, O. Toward a consensus on SNP and STR mutation rates on the human Y-chromosome. Hum. Genet. 136, 575–590 (2017).
Duchêne, S., Lanfear, R. & Ho, S. Y. W. The impact of calibration and clock-model choice on molecular estimates of divergence times. Mol. Phylogenet. Evol. 78, 277–289 (2014).
Forni, D., Cagliani, R., Clerici, M., Pozzoli, U. & Sironi, M. You Will Never Walk Alone: Codispersal of JC Polyomavirus with Human Populations. Mol. Biol. Evol. 37, 442–454 (2019).
Poznik, G. D. et al. Sequencing Y Chromosomes Resolves Discrepancy in Time to Common Ancestor of Males Versus Females. Science 341, 562–565 (2013).
Rieux, A. & Balloux, F. Inferences from tip-calibrated phylogenies: a review and a practical guide. Mol. Ecol. 25, 1911–1924 (2016).
Ho, S. Y. W. et al. Time-dependent rates of molecular evolution. Mol. Ecol. 20, 3087–3101 (2011).
Duchêne, S. et al. Bayesian Evaluation of Temporal Signal in Measurably Evolving Populations. Mol. Biol. Evol. 37, 3363–3379 (2020).
Eaton, K. et al. Plagued by a cryptic clock: insight and issues from the global phylogeny of Yersinia pestis. Commun. Biol. 6, 13 (2023).
Membrebe, J. M., Suchard, M. A., Rambaut, A., Baele, G. & Lemey, P. Bayesian Inference of Evolutionary Histories under Time-Dependent Substitution Rates. Mol. Biol. Evol. 36, 1793–1803 (2019).
Aiewsakun, P. & Katzourakis, A. Time-Dependent Rate Phenomenon in Viruses. J. Virol. 90, 7184–7195 (2016).
Ho, S. Y. W. et al. Biogeographic calibrations for the molecular clock. Biol. Lett. 11, 20150194 (2015).
Menardo, F. et al. Local adaptation in populations of Mycobacterium tuberculosis endemic to the Indian Ocean Rim. F1000Research 10, 60 (2021).
Menardo, F., Duchêne, S., Brites, D. & Gagneux, S. The molecular clock of Mycobacterium tuberculosis. PLOS Pathog. 15, e1008067 (2019).
Zehender, G. et al. Enigmatic origin of hepatitis B virus: An ancient travelling companion or a recent encounter? World J. Gastroenterol. 20, 7622 (2014).
Aiewsakun, P. & Katzourakis, A. Time dependency of foamy virus evolutionary rate estimates. BMC Evolut. Biol. 15, 119 (2015).
Colangeli, R. et al. Whole Genome Sequencing of Mycobacterium tuberculosis Reveals Slow Growth and Low Mutation Rates during Latent Infections in Humans. PLoS ONE 9, e91024 (2014).
Colangeli, R. et al. Mycobacterium tuberculosis progresses through two phases of latent infection in humans. Nat. Commun. 11 https://doi.org/10.1038/s41467-020-18699-9 (2020).
Cui, Y. et al. Historical variations in mutation rate in an epidemic pathogen, Yersinia pestis. Proc. Natl Acad. Sci. 110, 577–582 (2012).
Spyrou, M. A. et al. Phylogeography of the second plague pandemic revealed through analysis of historical Yersinia pestis genomes. Nat. Commun. 10, 4470 (2019).
Netikul, T. et al. Whole-genome single nucleotide variant phylogenetic analysis of Mycobacterium tuberculosis Lineage 1 in endemic regions of Asia and Africa. Sci. Rep. 12, 1565 (2022).
Meumann, E. M. et al. Tuberculosis in Australia’s tropical north: a population-based genomic epidemiological study. Lancet Reg. Health - West. Pac. 15, 100229 (2021).
O’Connell, J. et al. When did Homo sapiens first reach Southeast Asia and Sahul? Proc. Natl Acad. Sci. 115, 8482–8490 (2018).
Karmin, M. et al. Episodes of Diversification and Isolation in Island Southeast Asian and Near Oceanian Male Lineages. Mol. Biol. Evol. 39, msac045 (2022).
Kivisild, T. The study of human Y chromosome variation through ancient DNA. Hum. Genet. 136, 529–546 (2017).
Netikul, T., Palittapongarnpim, P., Thawornwattana, Y. & Plitphonganphim, S. Estimation of the global burden of Mycobacterium tuberculosis lineage 1. Infect., Genet. Evol. 91, 104802 (2021).
Reich, D., Thangaraj, K., Patterson, N., Price, A. L. & Singh, L. Reconstructing Indian population history. Nature 461, 489–494 (2009).
Gutierrez, M. C. et al. Predominance of Ancestral Lineages of Mycobacterium tuberculosis in India. Emerg. Infect. Dis. 12, 1367–1374 (2006).
Lin, D. et al. The geno-spatio analysis of Mycobacterium tuberculosis complex in hot and cold spots of Guangxi, China. BMC Infect. Dis. 20, 462 (2020).
Iwamoto, T. et al. Overcoming the pitfalls of automatic interpretation of whole genome sequencing data by online tools for the prediction of pyrazinamide resistance in Mycobacterium tuberculosis. PLoS ONE 14, e0212798 (2019).
Stucki, D. et al. Mycobacterium tuberculosis lineage 4 comprises globally distributed and geographically restricted sublineages. Nat. Genet. 48, 1535–1543 (2016).
Sabin, S. et al. A seventeenth-century Mycobacterium tuberculosis genome supports a Neolithic emergence of the Mycobacterium tuberculosis complex. Genome Biol. 21, 201 (2020).
Kay, G. L. et al. Eighteenth-century genomes show that mixed infections were common at time of peak tuberculosis in Europe. Nat. Commun. 6, 6717 (2015).
Bos, K. I. et al. Pre-Columbian mycobacterial genomes reveal seals as a source of New World human tuberculosis. Nature 514, 494–497 (2014).
Biek, R., Pybus, O. G., Lloyd-Smith, J. O. & Didelot, X. Measurably evolving pathogens in the genomic era. Trends Ecol. Evol. 30, 306–313 (2015).
Stritt, C. & Gagneux, S. How do monomorphic bacteria evolve? The Mycobacterium tuberculosis complex and the awkward population genetics of extreme clonality. BioRxiv https://doi.org/10.32942/x2gw2p (2022).
Drummond, A. J., Pybus, O. G. & Rambaut, A. Inference of Viral Evolutionary Rates from Molecular Sequences. Adv. Parasitol. 54, 331–358 (2003).
Chiner-Oms, A., Lopez, M. A., Moreno-Molina, M., Furió, V. & Comas, I. Gene evolutionary trajectories in Mycobacterium tuberculosis reveal temporal signs of selection. Proc. Natl Acad. Sci. 119, e2113600119 (2022).
Ford, C. B. et al. Use of whole genome sequencing to estimate the mutation rate of Mycobacterium tuberculosis during latent infection. Nat. Genet. 43, 482–486 (2011).
Donoghue, H. D. Tuberculosis and leprosy associated with historical human population movements in Europe and beyond – an overview based on mycobacterial ancient DNA. Ann. Hum. Biol. 46, 120–128 (2019).
Donoghue, H. D. Paleomicrobiology of Human Tuberculosis. Microbiol. Spect. 4 https://doi.org/10.1128/microbiolspec.PoH-0003-2014 (2016).
Buzic, I. & Guiffra, V. The paleopathological evidence on the origins of human tuberculosis: a review. J. Prevent. Med. Hyg. 61, E3–E8 (2020).
Hershkovitz, I. et al. Detection and Molecular Characterization of 9000-Year-Old Mycobacterium tuberculosis from a Neolithic Settlement in the Eastern Mediterranean. PLoS ONE 3, e3426 (2008).
Hershkovitz, I. et al. Tuberculosis origin: The Neolithic scenario. Tuberculosis 95, S122–S126 (2015).
Baker, O. et al. Human tuberculosis predates domestication in ancient Syria. Tuberculosis 95, S4–S12 (2015).
Baker, O. et al. Prehistory of human tuberculosis: Earliest evidence from the onset of animal husbandry in the Near East. Paléorient 43, 35–51 (2017).
Nicklisch, N. et al. Rib lesions in skeletons from early neolithic sites in Central Germany: On the trail of tuberculosis at the onset of agriculture. Am. J. Phys. Anthropol. 149, 391–404 (2012).
Masson, M. et al. Osteological and Biomolecular Evidence of a 7000-Year-Old Case of Hypertrophic Pulmonary Osteopathy Secondary to Tuberculosis from Neolithic Hungary. PLoS ONE 8, e78252 (2013).
Masson, M. et al. 7000 year-old tuberculosis cases from Hungary – Osteological and biomolecular evidence. Tuberculosis 95, S13–S17 (2015).
Pósa, A. et al. Tuberculosis in Late Neolithic-Early Copper Age human skeletal remains from Hungary. Tuberculosis 95, S18–S22 (2015).
Zink, A. R. et al. Molecular history of tuberculosis from ancient mummies and skeletons. Int. J. Osteoarchaeol. 17, 380–391 (2007).
Taylor, G. M., Young, D. B. & Mays, S. A. Genotypic Analysis of the Earliest Known Prehistoric Case of Tuberculosis in Britain. J. Clin. Microbiol. 43, 2236–2240 (2005).
Wilbur, A. K. et al. Deficiencies and challenges in the study of ancient tuberculosis DNA. J. Archaeolog. Sci. 36, 1990–1997 (2009).
Nelson, E. A., Buikstra, J. E., Herbig, A., Tung, T. A. & Bos, K. I. Advances in the molecular detection of tuberculosis in pre-contact Andean South America. Int. J. Paleopathol. 29, 128–140 (2020).
Karmin, M. et al. A recent bottleneck of Y chromosome diversity coincides with a global change in culture. Genome Res. 25, 459–466 (2015).
Hershberg, R. et al. High Functional Diversity in Mycobacterium tuberculosis Driven by Genetic Drift and Human Demography. PLoS Biol. 6, e311 (2008).
Comas, I. & Gagneux, S. A role for systems epidemiology in tuberculosis research. Trends Microbiol. 19, 492–500 (2011).
Conley, A. B. et al. A Comparative Analysis of Genetic Ancestry and Admixture in the Colombian Populations of Chocó and Medellín. G3 Genes|Genomes|Genet. 7, 3435–3447 (2017).
Bosch, E. et al. High level of male-biased Scandinavian admixture in Greenlandic Inuit shown by Y-chromosomal analysis. Hum. Genet. 112, 353–363 (2003).
Kayser, M. et al. Melanesian and Asian Origins of Polynesians: mtDNA and Y Chromosome Gradients Across the Pacific. Mol. Biol. Evol. 23, 2234–2244 (2006).
Pierron, D. et al. Genomic landscape of human diversity across Madagascar. Proc. Natl Acad. Sci. 114, E6498–E6506 (2017).
Gräler, B., Pebesma, E. & Heuvelink, G. Spatio-Temporal Interpolation using gstat. R. J. 8, 204 (2016).
Kahle, D. & Wickham, H. ggmap: Spatial Visualization with ggplot2. R. J. 5, 144 (2013).
Balaresque, P. et al. Y-chromosome descent clusters and male differential reproductive success: young lineage expansions dominate Asian pastoral nomadic populations. Eur. J. Hum. Genet. 23, 1413–1422 (2015).
Cai, X. et al. Human Migration through Bottlenecks from Southeast Asia into East Asia during Last Glacial Maximum Revealed by Y Chromosomes. PLoS ONE 6, e24282 (2011).
Delfin, F. et al. The Y-chromosome landscape of the Philippines: extensive heterogeneity and varying genetic affinities of Negrito and non-Negrito groups. Eur. J. Hum. Genet. 19, 224–230 (2011).
He, J. D. et al. Patrilineal Perspective on the Austronesian Diffusion in Mainland Southeast Asia. PLoS ONE 7, e36437 (2012).
Gayden, T. et al. Y-STR diversity in the Himalayas. Int. J. Leg. Med. 125, 367–375 (2011).
Lappalainen, T. et al. Migration Waves to the Baltic Sea Region. Ann. Hum. Genet. 72, 337–348 (2008).
Malyarchuk, B. et al. Y-chromosome diversity in the Kalmyks at the ethnical and tribal levels. J. Hum. Genet. 58, 804–811 (2013).
Xue, Y. et al. Male Demography in East Asia: A North–South Contrast in Human Population Expansion Times. Genetics 172, 2431–2439 (2006).
Nei, M. Molecular evolutionary genetics. (Columbia University Press, 1987).
Bouckaert, R. et al. BEAST 2: A Software Platform for Bayesian Evolutionary Analysis. PLoS Comput. Biol. 10, e1003537 (2014).
Fu, Q. et al. Genome sequence of a 45,000-year-old modern human from western Siberia. Nature 514, 445–449 (2014).
Edokimov, K. et al. Whole-genome sequencing of Mycobacterium tuberculosis from Cambodia. Sci. Rep. 12, 7693 (2022).
Maung, H. M. W. et al. Geno-Spatial Distribution of Mycobacterium Tuberculosis and Drug Resistance Profiles in Myanmar–Thai Border Area. Trop. Med. Infect. Dis. 5, 153 (2020).
Ajawatanawong, P. et al. A novel Ancestral Beijing sublineage of Mycobacterium tuberculosis suggests the transition site to Modern Beijing sublineages. Sci. Rep. 9, 13718 (2019).
Martin, M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet. J. 17, 10 (2011).
Li, H. & Durbin, R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 25, 1754–1760 (2009).
McKenna, A. et al. The Genome Analysis Toolkit: A MapReduce framework for analyzing next-generation DNA sequencing data. Genome Res. 20, 1297–1303 (2010).
Walker, B. J. et al. Pilon: An Integrated Tool for Comprehensive Microbial Variant Detection and Genome Assembly Improvement. PLoS ONE 9, e112963 (2014).
Freschi, L. et al. Population structure, biogeography and transmissibility of Mycobacterium tuberculosis. Nat. Commun. 12, 6099 (2021).
Coll, F. et al. A robust SNP barcode for typing Mycobacterium tuberculosis complex strains. Nat. Commun. 5, 4812 (2014).
Li, H. A statistical framework for SNP calling, mutation discovery, association mapping and population genetical parameter estimation from sequencing data. Bioinformatics 27, 2987–2993 (2011).
Andrews, S. FastQC: A Quality Control Tool for High Throughput Sequence Data. http://www.bioinformatics.babraham.ac.uk/projects/fastqc/ (2010).
Stamatakis, A. RAxML version 8: a tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics 30, 1312–1313 (2014).
Napier, G. et al. Robust barcoding and identification of Mycobacterium tuberculosis lineages for epidemiological and clinical studies. Genome Med. 12, 114 (2020).
Acknowledgements
The authors acknowledge support from the National Health and Medical Research Council, Australia APP1172853 (S.J.D) and the US National Institutes of Health U19AI162583 (S.J.D).
Author information
Authors and Affiliations
Contributions
M.S. designed the study and performed the analysis. M.S. and S.J.D. wrote the paper.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
This paper has been previously reviewed at another Nature Portfolio journal. This document only contains reviewer comments and rebuttal letters for versions considered at Communications Biology. Communications Biology thanks the anonymous reviewers for their contribution to the peer review of this work. Primary Handling Editor: George Inglis. A peer review file is available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Silcocks, M., Dunstan, S.J. Parallel signatures of Mycobacterium tuberculosis and human Y-chromosome phylogeography support the Two Layer model of East Asian population history. Commun Biol 6, 1037 (2023). https://doi.org/10.1038/s42003-023-05388-8
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s42003-023-05388-8
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.
