|
Post by Admin on Aug 18, 2023 8:26:29 GMT
High-coverage genome of the Tyrolean Iceman reveals unusually high Anatolian farmer ancestry Summary The Tyrolean Iceman is known as one of the oldest human glacier mummies, directly dated to 3350–3120 calibrated BCE. A previously published low-coverage genome provided novel insights into European prehistory, despite high present-day DNA contamination. Here, we generate a high-coverage genome with low contamination (15.3×) to gain further insights into the genetic history and phenotype of this individual. Contrary to previous studies, we found no detectable Steppe-related ancestry in the Iceman. Instead, he retained the highest Anatolian-farmer-related ancestry among contemporaneous European populations, indicating a rather isolated Alpine population with limited gene flow from hunter-gatherer-ancestry-related populations. Phenotypic analysis revealed that the Iceman likely had darker skin than present-day Europeans and carried risk alleles associated with male-pattern baldness, type 2 diabetes, and obesity-related metabolic syndrome. These results corroborate phenotypic observations of the preserved mummified body, such as high pigmentation of his skin and the absence of hair on his head. Graphical abstract Introduction The Tyrolean Iceman (hereafter referred to as the Iceman), also known as “Ötzi,” is the world's oldest glacier . Radiocarbon dating and stable isotope analysis have revealed that the Iceman lived during the Chalcolithic (Copper Age) in the Southern slopes of the eastern Italian Alps.1,2 His remains were found in the Italian part of the Ötztal Alps in 1991 and were directly dated to 3350–3120 calibrated BCE. In 2012, Keller et al. published the first whole-genome sequence of the Iceman.3 Comparative analyses based on autosomal data reported a close genetic affinity between the Iceman and present-day Sardinians. These findings were, however, published before genomes from a larger number of ancient western Eurasian individuals became available. Genomic data from European ancient individuals from 3000 to 4000 BCE, who we consider contemporaneous populations to the Iceman, showed that the genetic similarity between the Iceman and present-day Sardinians is due to common genetic components that were geographically widespread across Europe during the Neolithic period.4,5,6,7,8 The geographic region of the Alps, where the Iceman was discovered, remains, however, rather understudied. The first Iceman genome from 2012 was generated using the ABI SOLiD sequencer platform, which requires complex computational infrastructure at high economical cost.9 It had relatively low coverage (7.6×) compared with the high-coverage genome generated in this study and showed the presence of modern human DNA contamination. Therefore, thanks to the recent development of sequencing technologies (Illumina technology) with higher output and lower cost, which have become standard in the field of ancient DNA research,9 we produced a new high-coverage genome for the Iceman (15.3× coverage) from the same left iliac bone sample used for the 2012 study, with minimal modern human contamination (0.5% ± 0.06%). We found that the Iceman shows unusually high early Neolithic-farmer-related ancestry among the analyzed European individuals from the 4th millennium BCE. Moreover, we show that the two ancestry components from European hunter-gatherer-related ancestry and early Neolithic-farmer-related ancestry present in the Iceman admixed rather recently before the Iceman’s death (56 ± 21 generations ago, namely 4880 ± 635 BCE), suggesting the survival of hunter-gatherer groups south of the Alps as late as 5000 to 4000 BCE.
|
|
|
Post by Admin on Aug 19, 2023 20:35:41 GMT
Results A new high-coverage genome We obtained two samples of the left iliac bone and the surrounding tissue for a series of four extractions each in order to improve the amount of recovered ancient DNA and reduce modern human DNA contamination. In order to identify the extracts with the highest human DNA content for further processing, we compared the percentage of mapped human reads (i.e., endogenous DNA) after shotgun sequencing on an Illumina HiSeq platform, ranging from 1.58% to 51.02% endogenous human DNA (Table S1), for DNA libraries generated from these eight extracts (Table S1A). From the best two extracts (1412E2 and 1412E3), four double-stranded DNA libraries were generated with uracil DNA glycosylase (UDG), which reduces substitutions due to ancient DNA damage (i.e., deaminated cytosines) at the end of DNA fragments.10 We then performed paired-end shotgun sequencing on a total of 36 Illumina HiSeq sequencing lanes for all four libraries (Tables S1B and S1C). We processed the raw sequencing data with EAGER 1.92.211 and obtained a combined alignment with average genomic coverage of 15.3× after removing duplicates. In the end, we obtained 45.4% endogenous human DNA content (Table S1), with more than 90.6% of the genome covered by at least five reads (STAR Methods). We estimated the contamination level in the high-coverage genome using ANGSD based on the heterozygosity on the haploid X chromosome.12 The high-coverage Iceman genome has 0.5% ± 0.06% contamination, 10× less than the contamination level found in the previously published genome sequence3 (7.5% ± 0.25%) (Table S2). Genetic ancestry analysis It has been shown that early Neolithic European farmers derived most of their ancestry from early Anatolian farmers, suggesting that farming spread with people from the Near East through Anatolia and the Balkan peninsula starting around 7000 BCE.4,5,13 The arrival of farmers in Europe is followed by an increasing amount of admixture with local hunter-gatherers at variable levels during the initial expansion14,15,16 and subsequently into the 4th millennium BCE. By the end of the 4th millennium BCE, admixture between early Neolithic-farmer-related ancestry originating from Anatolia and European hunter-gatherer-related ancestry had become prevalent in most parts of Europe.5,15,16,17,18,19 Later, beginning from 2900 BCE, herders from the Pontic-Caspian steppe introduced substantial levels of so-called “Steppe-related ancestry” throughout Europe.17 After the 3rd millennium BCE until present day, all three ancestry components are found in almost all European populations.13 We examined the Iceman’s ancestry makeup in the context of those three ancestral components with corresponding representative proxies—western hunter-gatherers (“WHGs”), early Neolithic farmers from Anatolia (“Anatolia_N”) or Germany (“Germany_EN_LBK”), and herders from the Samara region (“Russia _Samara_EBA_Yamnaya”)—together with ancient populations from Germany, the northern Iberian Peninsula, Italy, and Sardinia also dated to the 4th millennium BCE (Figure 1; Table S3). Figure 1 Geographic location of the Iceman and analyzed published ancient western Eurasian groups See also Table S3. Projecting both the high-coverage and the previously published3 Iceman genome on modern western Eurasian genetic variation using principal-component analysis (PCA) (STAR Methods), the high-coverage genome is slightly shifted compared with the previously published one3 (Figure 2). The high-coverage genome of the Iceman clusters between the two groups in the PCA formed by (1) Middle-Neolithic and Chalcolithic Europeans dated to the 4th millennium BCE and by (2) early Neolithic European farmers, which in turn fall closely together with earlier farmers from Anatolia. In the genetic affinity test, the high-coverage Iceman genome shows a similar pattern, presenting the closest genetic affinity to contemporaneous Europeans from the 4th millennium BCE and early Neolithic European farmers (Figure S1).
|
|
|
Post by Admin on Aug 23, 2023 22:20:02 GMT
Figure 2 Principal-component analysis (PCA) We project the high-coverage Iceman genome, the previously published Iceman genome, and related published ancient western Eurasian individuals onto present-day western Eurasian populations. See also Table S3. In the PCA (Figure 2), the high-coverage Iceman genome locates closer to the cluster of early Neolithic Farmers from Europe (farmers associated with Linear Pottery culture, in short “LBK” hereafter) and Anatolia_N than other Middle-Late Neolithic to Chalcolithic Europeans (Spain_MLN, Italy_Sardinia_N, Italy_Sardinia_C, Italy_N.SG, Germany_MN_Baalberge, Germany_MN_Salzmuende, Germany_MN_Esperstedt, Italy_Broion_CA.SG),6,7,18,20 indicating that the Iceman may have more early Neolithic-farmer-related ancestry than other tested European individuals from the 4th millennium BCE. To calculate the exact proportions of ancestral components in the Iceman and other contemporaneous ancient European groups, we applied qpAdm modeling to test three proxies for the early Neolithic-farmer-related ancestry—Germany_EN_LBK, Anatolia_N, and Germany_EN_LBK_Stuttgart.DG (a 7,000-year-old high-coverage shotgun genome from the LBK culture in Germany) (Tables S3 and S4). We found that the Iceman derives 90% ± 2.5% ancestry from early Neolithic farmer populations when using Anatolia_N as the proxy for the early Neolithic-farmer-related ancestry and WHGs as the other ancestral component (Figure 3; Table S4). When testing with a 3-way admixture model including Steppe-related ancestry as the third source for the previously published3 and the high-coverage genome, we found that our high-coverage genome shows no Steppe-related ancestry (Table S5), in contrast to ancestry decomposition of the previously published Iceman genome.3,17 We conclude that the 7.5% Steppe-related ancestry previously estimated for the previously published Iceman genome3,17 is likely the result of modern human contamination.
|
|
|
Post by Admin on Aug 25, 2023 3:14:45 GMT
Figure 3 Global and local ancestral proportions and admixture time of ancient groups in the 4th millennium BCE (A) qpAdm estimates on the proportion of western hunter-gatherer (WHG) and Anatolia_N ancestry in Iceman and contemporaneous ancient European groups. (B) Dating the admixture time using DATES for target populations shown in panel (A). Horizontal bars in (A) and (B) represent ±1 standard error (SE) estimated by qpAdm and DATES correspondingly. (C) Local ancestry assignments across 22 autosomes of the Iceman genome. The majority of the Iceman genome is inferred to be of farmer origin (LBK, orange), and a small fraction is inferred to be of WHG origin (blue), consistent with the global ancestry proportion estimates from qpAdm. Gray area represents short arms on the chromosomes, which do not contain any unique genetic material. Compared with the Iceman, the analyzed contemporaneous European populations from Spain and Sardinia (Italy_Sardinia_C, Italy_Sardinia_N, Spain_MLN) show less early Neolithic-farmer-related ancestry, ranging from 27.2% to 86.9% (Figure 3A; Table S4). Even ancient Sardinian populations,7 who are located further south than the Iceman and are geographically separate from mainland Europe, derive no more than 85% ancestry from Anatolia_N (Figure 3; Table S4). The higher levels of hunter-gatherer ancestry in individuals from the 4th millennium BCE have been explained by an ongoing admixture between early farmers and hunter-gatherers in the Middle and Late Neolithic in various parts of Europe, including western Europe (Germany and France), central Europe, Iberia, and the Balkans.5,15,16,17,18,19 Only individuals from Italy_Broion_CA.SG found to the south of the Alps present similarly low hunter-gatherer ancestry as seen in the Iceman.21 We conclude that the Iceman and Italy_Broion_CA.SG might both be representatives of specific Chalcolithic groups carrying higher levels of early Neolithic-farmer-related ancestry than any other contemporaneous European group. This might indicate less gene flow from groups that are more admixed with hunter-gatherers or a smaller population size of hunter-gatherers in that region during the 5th and 4th millennium BCE. Recent admixture between early farmers and hunter-gatherers in southern Europe Given the high proportion of early Neolithic-farmer-related ancestry in the Iceman genome, we also tested if using the early Neolithic-farmer-related ancestry as a single source is sufficient. We found that qpWave results suggest neither Anatolia_N nor Germany_EN_LBK as an appropriate single source, confirming that the European hunter-gatherer-related ancestry is low but significantly present in the Iceman’s genome (p < 0.05; Table S6). We estimated the admixture date between the early Neolithic-farmer-related (using Anatolia_N as proxy) and WHG-related ancestry sources using DATES22 to be 56 ± 21 generations before the Iceman’s death, which corresponds to 4880 ± 635 calibrated BCE assuming 29 years per generation23 (Figure 3B; Table S7) and considering the mean C14 date of this individual. Alternatively, using Germany_EN_LBK as the proxy for early Neolithic-farmer-related ancestry, we estimated the admixture date to be 40 ± 15 generations before his death (Table S7), or 4400 ± 432 calibrated BCE, overlapping with estimates from nearby Italy_Broion_CA.SG, who locate to the south of the Alps7,18,20 (Figure 3B). While compared with the admixture time between early Neolithic farmers and hunter-gatherers in other parts of southern Europe, for instance in Spain and southern Italy, we found that, particularly, the admixture with hunter-gatherers as seen in the Iceman and Italy_Broion_CA.SG is more recent (Figure 3B; Table S3), suggesting a potential longer survival of hunter-gatherer-related ancestry in this geographical region.
|
|
|
Post by Admin on Aug 25, 2023 23:42:55 GMT
Recent admixture between early farmers and hunter-gatherers in southern Europe Given the high proportion of early Neolithic-farmer-related ancestry in the Iceman genome, we also tested if using the early Neolithic-farmer-related ancestry as a single source is sufficient. We found that qpWave results suggest neither Anatolia_N nor Germany_EN_LBK as an appropriate single source, confirming that the European hunter-gatherer-related ancestry is low but significantly present in the Iceman’s genome (p < 0.05; Table S6). We estimated the admixture date between the early Neolithic-farmer-related (using Anatolia_N as proxy) and WHG-related ancestry sources using DATES22 to be 56 ± 21 generations before the Iceman’s death, which corresponds to 4880 ± 635 calibrated BCE assuming 29 years per generation23 (Figure 3B; Table S7) and considering the mean C14 date of this individual. Alternatively, using Germany_EN_LBK as the proxy for early Neolithic-farmer-related ancestry, we estimated the admixture date to be 40 ± 15 generations before his death (Table S7), or 4400 ± 432 calibrated BCE, overlapping with estimates from nearby Italy_Broion_CA.SG, who locate to the south of the Alps7,18,20 (Figure 3B). While compared with the admixture time between early Neolithic farmers and hunter-gatherers in other parts of southern Europe, for instance in Spain and southern Italy, we found that, particularly, the admixture with hunter-gatherers as seen in the Iceman and Italy_Broion_CA.SG is more recent (Figure 3B; Table S3), suggesting a potential longer survival of hunter-gatherer-related ancestry in this geographical region.
Effective population size and heterozygosity The high-coverage genome allows for additional analyses, such as estimating effective population sizes through time and genome-wide heterozygosity, that are not possible with lower-coverage or SNP-captured genomes. Specifically, we estimated the population-size history of the population represented by the Iceman and the two source populations represented by an early Neolithic farmer from Stuttgart in Germany (“Germany_EN_LBK_Stuttgart.DG”) and by a Mesolithic hunter-gatherer from Loschbour in Luxembourg (“Luxembourg_Loschbour.DG”)13 using MSMC2.24 The demographic histories estimated using the aforementioned three ancient high-coverage genomes share the same population bottleneck between 25,000 and 200,000 years ago, similar to that obtained from a present-day Sardinian individual, and they show a slight population size increase in a recent time epoch from 20,000–25,000 years ago (Figure S2). We observe a higher population size in recent times for the Iceman and the Germany_EN_LBK_Stuttgart.DG individual (both with high early Neolithic-farmer-related ancestry) compared with the Luxembourg_Loschbour.DG hunter-gatherer, which is possibly linked to the larger population size of early farming populations versus the hunter-gatherer populations in recent times.25 We estimated the rate of heterozygosity for the Iceman, Germany_EN_LBK_Stuttgart.DG, Luxembourg_Loschbour.DG, and a present-day Sardinian individual and plotted the per-chromosome estimate together with the standard error calculated from a weighted jackknife procedure in Figure S3 (Table S8). Both Germany_EN_LBK_Stuttgart.DG and the Iceman show higher heterozygosity levels than Luxembourg_Loschbour.DG, but the Iceman shows a relatively lower level than Germany_EN_LBK_Stuttgart.DG. This is consistent with the supposed relative isolation of the Iceman and the low WHG-derived ancestry seen in his genome.
|
|