|
Post by Admin on Jun 4, 2022 18:49:01 GMT
Comparison of the DNA preservation in different tissues We tested different tissues for DNA preservation and applied strict criteria for authenticity on the retrieved mitochondrial and nuclear DNA to establish authentic ancient Egyptian DNA. First, DNA extracts from several tissues (that is, bone, teeth, soft tissue and macerated teeth) from 151 individuals were screened for the presence of human mitochondrial DNA (mtDNA) resulting in a total of 2,157 to 982,165 quality filtered mitochondrial reads per sample, and 11- to 4,236-fold coverage. To estimate, identify and filter out potential contamination we applied the program schmutzi15 with strict criteria for contamination and kept only samples with less than 3% contamination for further analysis. For a comparison of different source material (soft tissue, bone and teeth) ten individuals (Supplementary Table 1) were sampled multiple times. Yields of preserved DNA were comparable in bone and teeth but up to ten times lower in soft tissues (Fig. 2a, Supplementary Table 1). Nucleotide misincorporation patterns characteristic for damaged ancient human DNA allowed us to assess the authenticity of the retrieved DNA13,14. The observed DNA damage patterns differed for the source materials with on average 19% damage in soft tissues and around 10% damage in bone tissue and teeth (Fig. 2b,c, Supplementary Table 1). Importantly, mtDNA haplotypes were identical for all samples from the same individuals. Our results thus suggest that DNA damage in Egyptian mummies correlates with tissue type. The protection of bone and teeth by the surrounding soft tissue or the embalmment of soft tissue may have contributed to the observed differences. Figure 2: DNA preservation and DNA damage of the samples used in this study. (a) coverage boxplots separated by tissue type (bone, mummified tissue, teeth), (b) boxplots showing damage of first base at the 3′ end separated by tissue type according to a, (c) damage on first base at the 5′ end of mapped reads separated by tissue type according to a and b. Generation of nuclear data In order to analyse the nuclear DNA we selected 40 samples with high mtDNA coverage and low mtDNA contamination. Using in solution enrichment for 1.2 million genome-wide SNPs26, we obtained between 3,632 and 508,360 target SNPs per sample (Supplementary Data 2). Overall, the nuclear DNA showed poor preservation compared to the mtDNA as depicted by a high mitochondrial/nuclear DNA ratio of on average around 18,000. In many samples, nuclear DNA damage was relatively low, indicating modern contamination. We sequenced two libraries per sample: one untreated library to assess DNA damage, and one library treated with enzymatic damage repair27, which was used for downstream analysis. We applied strict criteria for further analysis: we considered only male samples with at least 8% average cytosine deamination rates at the ends of the reads from the untreated library, and with at least 150 SNPs on the X chromosome covered at least twice, in order to estimate contamination levels reliably. Three out of 40 samples fulfilling these criteria had acceptable nuclear contamination rates: Two samples from the Pre-Ptolemaic Periods (New Kingdom to Late Period) had 5.3 and 0.5% nuclear contamination and yielded 132,084 and 508,360 SNPs, respectively, and one sample from the Ptolemaic Period had 7.3% contamination and yielded 201,967 SNPs. As shown below, to rule out any impact of potential contamination on our results, we analysed the three samples separately or replicated results using only the least contaminated sample.
|
|
|
Post by Admin on Jun 5, 2022 17:49:09 GMT
Analysis of mitochondrial genomes The 90 mitochondrial genomes fulfilling our criteria (>10-fold coverage and <3% contamination) were grouped into three temporal categories based on their radiocarbon dates (Supplementary Data 1), corresponding to Pre-Ptolemaic Periods (n=44), the Ptolemaic Period (n=27) and the Roman Period (n=19) (Supplementary Data 1). To test for genetic differentiation and homogeneity we compared haplogroup composition, calculated FST-statistics28 and applied a test for population continuity29 (Supplementary Table 2, Supplementary Data 3,4) on mitochondrial genome data from the three ancient and two modern-day populations from Egypt and Ethiopia, published by Pagani and colleagues17, including 100 modern Egyptian and 125 modern Ethiopian samples (Fig. 3a). We furthermore included data from the El-Hayez oasis published by Kujanová and colleagues30. We observe highly similar haplogroup profiles between the three ancient groups (Fig. 3a), supported by low FST values (<0.05) and P values >0.1 for the continuity test. Modern Egyptians share this profile but in addition show a marked increase of African mtDNA lineages L0–L4 up to 20% (consistent with nuclear estimates of 80% non-African ancestry reported in Pagani et al.17). Genetic continuity between ancient and modern Egyptians cannot be ruled out by our formal test despite this sub-Saharan African influx, while continuity with modern Ethiopians17, who carry >60% African L lineages, is not supported (Supplementary Data 5). To further test genetic affinities and shared ancestry with modern-day African and West Eurasian populations we performed a principal component analysis (PCA) based on haplogroup frequencies and Multidimensional Scaling of pairwise genetic distances. We find that all three ancient Egyptian groups cluster together (Fig. 3b), supporting genetic continuity across our 1,300-year transect. Both analyses reveal higher affinities with modern populations from the Near East and the Levant compared to modern Egyptians (Fig. 3b,c). The affinity to the Middle East finds further support by the Y-chromosome haplogroups of the three individuals for which genome-wide data was obtained, two of which could be assigned to the Middle-Eastern haplogroup J, and one to haplogroup E1b1b1 common in North Africa (Supplementary Table 3). However, comparative data from a contemporary population under Roman rule in Asia Minor, from the Roman city Ağlasun today in Turkey31, did not reveal a closer relationship to the ancient Egyptians from the Roman period (Fig. 3b,c). Figure 3: Analysis of 90 ancient Egyptian mitochondrial genomes. (a) Mitochondrial DNA haplogroup frequencies of three ancient and two modern-day populations, (b) Principal Component Analysis based on haplogroup frequencies: (sub-Saharan Africa (green), North Africa (light green), Near East (orange), Europe (yellow), ancient (blue), (c) MDS of HVR-I sequence data: colour scheme as above; note that ancient groups were pooled, (d) Skygrid plot depicting effective population size estimates over the last 5,000 years in Egypt. Vertical bars indicate the ages of the analysed 90 mitochondrial genomes (three samples with genome-wide data highlighted in red). Note that the values on y axis are given in female effective population size times generation time and were rescaled by 1:14.5 for the estimation of the studied population size (assuming 29-year generation time and equal male and female effective population sizes) (images by Kerttu Majander). Population size estimation using BEAST The finding of a continuous population through time allowed us to estimate the effective population size (Ne) from directly radiocarbon-dated mitochondrial genomes using BEAST32. Our results show similar values of effective population size in the different ancient time periods with an average value of between ca. 48,000 and 310,000 (average 95% CI) inhabitants in the region and period under investigation (Fig. 3d, Supplementary Fig. 2, Supplementary Table 4). This is important as it is the first time that such estimates can be contrasted with reported historic Egyptian census numbers from the neighbouring Fayum in the early Ptolemaic Period, which had a reported total population size of 85,000–95,000 inhabitants33.
|
|
|
Post by Admin on Jun 6, 2022 17:27:51 GMT
Population genetic analysis of nuclear DNA On the nuclear level we merged the SNP data of our three ancient individuals with 2,367 modern individuals34,35 and 294 ancient genomes36 and performed PCA on the joined data set. We found the ancient Egyptian samples falling distinct from modern Egyptians, and closer towards Near Eastern and European samples (Fig. 4a, Supplementary Fig. 3, Supplementary Table 5). In contrast, modern Egyptians are shifted towards sub-Saharan African populations. Model-based clustering using ADMIXTURE37 (Fig. 4b, Supplementary Fig. 4) further supports these results and reveals that the three ancient Egyptians differ from modern Egyptians by a relatively larger Near Eastern genetic component, in particular a component found in Neolithic Levantine ancient individuals36 (Fig. 4b). In contrast, a substantially larger sub-Saharan African component, found primarily in West-African Yoruba, is seen in modern Egyptians compared to the ancient samples. In both PCA and ADMIXTURE analyses, we did not find significant differences between the three ancient samples, despite two of them having nuclear contamination estimates over 5%, which indicates no larger impact of modern DNA contamination. We used outgroup f3-statistics38 (Fig. 5a,b) for the ancient and modern Egyptians to measure shared genetic drift with other ancient and modern populations, using Mbuti as outgroup. We find that ancient Egyptians are most closely related to Neolithic and Bronze Age samples in the Levant, as well as to Neolithic Anatolian and European populations (Fig. 5a,b). When comparing this pattern with modern Egyptians, we find that the ancient Egyptians are more closely related to all modern and ancient European populations that we tested (Fig. 5b), likely due to the additional African component in the modern population observed above. By computing f3-statistics38, we determined whether modern Egyptians could be modelled as a mixture of ancient Egyptian and other populations. Our results point towards sub-Saharan African populations as the missing component (Fig. 5c), confirming the results of the ADMIXTURE analysis. We replicated the results based on f3-statistics using only the least contaminated sample (with <1% contamination estimate) and find very similar results (Supplementary Fig. 5), confirming that the moderate levels of modern DNA contamination in two of our samples did not affect our analyses. Finally, we used two methods to estimate the fractions of sub-Saharan African ancestry in ancient and modern Egyptians. Both qpAdm35 and the f4-ratio test39 reveal that modern Egyptians inherit 8% more ancestry from African ancestors than the three ancient Egyptians do, which is also consistent with the ADMIXTURE results discussed above. Absolute estimates of African ancestry using these two methods in the three ancient individuals range from 6 to 15%, and in the modern samples from 14 to 21% depending on method and choice of reference populations (see Supplementary Note 1, Supplementary Fig. 6, Supplementary Tables 5–8). We then used ALDER40 to estimate the time of a putative pulse-like admixture event, which was estimated to have occurred 24 generations ago (700 years ago), consistent with previous results from Henn and colleagues16. While this result by itself does not exclude the possibility of much older and continuous gene flow from African sources, the substantially lower African component in our ∼2,000-year-old ancient samples suggests that African gene flow in modern Egyptians occurred indeed predominantly within the last 2,000 years. Figure 4: Principal component analysis and genetic clustering of genome-wide DNA from three ancient Egyptians. (a) Principal Component Analysis-based genome-wide SNP data of three ancient Egyptians, 2,367 modern individuals and 294 previously published ancient genomes, (b) subset of the full ADMIXTURE analysis (Supplementary Fig. 4). Figure 5: Shared drift and mixture analysis of three ancient Egyptians with other modern and ancient populations. (a) Outgroup f3-statistics measuring shared drift of the three ancient Egyptian samples and other modern and ancient populations, (b) The data shown in a, compared with the same estimates for modern Egyptians, ordered by shared drift with modern Egyptians, (c) Admixture f3-statistics, testing whether modern Egyptians are mixed from ancient Egyptians and some other source. The most negative Z-scores indicate the most likely source populations.
|
|
|
Post by Admin on Jun 6, 2022 21:45:34 GMT
Supplementary Table 3: Y-Chromosomal haplotype results. SampleID YHaplogroup Comment
JK2134 J This individual was assigned to haplogroup J based on mutations: CTS8938/PF4577: 18567169T→G, F2817/PF4579: 18695159C→T, F4299/PF4589: 21144431T→A, S22619/Z7820: 21144432C→A, F4300: 21144433T→A, YSC0000228: 22172960G→T, L778/PF4616/YSC0000236: 23088142T→C.
JK2888 E1b1b1a1b2 This individual was assigned to haplogroup E1b1b1a1b2 based on mutation: V22: 6859957T→C, and to upstream E1b1b1a: CTS2661: 14410669C→T and E1b1b1: PF1619: 13848122T→C, CTS2620: 14393170A→C, M5360: 23618826C→T.
JK2911 J This individual was assigned to haplogroup J based on mutations: CTS687/PF4503:6953311A→T, CTS1250/PF4510/YSC0001255:7296343G→T, PF4513/NA:7759610C→T, PF4519/NA:8669451C→G, PF4524/NA:10009851G→A, PF4530/NA:13597365C→T, CTS2769/PF4538:14476551T→A, F1973/PF4546/YSC0001304:15581303G→A, F2114/PF4551:16262942G→A, CTS5628/PF4555:16401405C→G, CTS5678/PF4556:16427564A→T, F2502/PF4564:17495914G→A, CTS7738/PF4568:17637446T→C, CTS7832/PF4569:17693210A→G, F2769/PF4576:18552360G→C, F2973/PF4585/YSC0001312:19194316C→T, F4299/PF4589:21144431T→A, S22619/Z7820:21144432C→A, F4300/NA:21144433T→A, F3176/PF4592/YSC0001314:21329083T→C, PF4595/NA:21858778C→A, YSC0000228/NA:22172960G→T, M304/Page16/PF4609:22749853A→C, L778/PF4616/YSC0000236:23088142T→C, CTS11571/PF4617:23163701C→A, CTS11750/PF4618/YSC0001250:23250894C→T, CTS12047/YSC0001253:23443976A→G
Estimating phenotypes Finally, we analysed several functionally relevant SNPs in sample JK2911, which had low contamination and relatively high coverage. This individual had a derived allele at the SLC24A5 locus, which contributes to lighter skin pigmentation and was shown to be at high frequency in Neolithic Anatolia41, consistent with the ancestral affinity shown above. Other relevant SNPs carry the ancestral allele, including HERC2 and LCT, which suggest dark-coloured eyes and lactose intolerance (Supplementary Table 9).
|
|
|
Post by Admin on Jun 6, 2022 22:36:50 GMT
Discussion This study demonstrates that the challenges of ancient DNA work on Egyptian mummies can be overcome with enrichment strategies followed by high-throughput DNA sequencing. The use of ancient DNA can greatly contribute towards a more accurate and refined understanding of Egypt’s population history. More specifically, it can supplement and serve as a corrective to archaeological and literary data that are often unevenly distributed across time, space and important constituents of social difference (such as gender and class) as well as modern genetic data from contemporary populations that may not be fully representative of past populations.
The archaeological site Abusir el-Meleq was inhabited from at least 3250BCE until about 700CE and was of great religious significance because of its active cult to Osiris, the god of the dead, which made it an attractive burial site for centuries2. Written sources indicate that by the third century BCE Abusir el-Meleq was at the centre of a wider region that comprised the northern part of the Herakleopolites province, and had close ties with the Fayum and the Memphite provinces, involving the transport of wheat, cattle-breeding, bee-keeping and quarrying42. In the early Roman Period, the site appears to have been the main centre in its own district42. Abusir el-Meleq’s proximity to, and close ties with, the Fayum are significant in the context of this study as the Fayum in particular saw a substantial growth in its population during the first hundred years of Ptolemaic rule, presumably as a result of Greek immigration33,43. Later, in the Roman Period, many veterans of the Roman army—who, initially at least, were not Egyptian but people from disparate cultural backgrounds—settled in the Fayum area after the completion of their service, and formed social relations and intermarried with local populations44. Importantly, there is evidence for foreign influence at Abusir el-Meleq. Individuals with Greek, Latin and Hebrew names are known to have lived at the site and several coffins found at the cemetery used Greek portrait image and adapted Greek statue types to suit ‘Egyptian’ burial practices2,45. The site’s first excavator, Otto Rubensohn, also found a Greek grave inscription in stone as well as a writing board inscribed in Greek46. Taken together with the multitude of Greek papyri that were written at the site, this evidence strongly suggests that at least some inhabitants of Abusir el-Meleq were literate in, and able to speak, Greek45. However, a general issue concerning the site is that several details of the context of the individuals analysed in this study were lost over time. All of the material was excavated by Rubensohn in the early twentieth century, whose main interest was to obtain literary papyri from cartonnage rather than to excavate human remains47. As is customary for the time, Rubensohn’s archaeological records are highly incomplete and many of the finds made by him were removed undocumented from their contexts. Furthermore, many of his excavation diaries and notes were destroyed during the Second World War19. This lack of context greatly diminishes the possibility of ‘thick description’ of the analysed individuals, at least in terms of their names, titles and materially expressed identity. However, the finds nevertheless hold much promise for a long-term study of population dynamics in ancient Egypt. Abusir el-Meleq is arguably one of the few sites in Egypt, for which such a vast number of individuals with such an extensive chronological spread are available for ancient DNA analysis. Although we only analysed mummified remains, there is little reason to believe that the burials Rubensohn excavated belonged exclusively to a group of prosperous inhabitants on the basis of the far published references to excavation diaries and Rubensohn’s preliminary reports that permit a basic reconstruction. Rather it seems arguable that the complete spectrum of society is represented, ranging from Late Period priests’ burials that stand out by virtue of their size and contents to simple inhumations that are buried with little to no grave goods2. The widespread mummification treatments in the Ptolemaic and Roman Periods in particular, leading to a decline in standards and costs48 and the generally modest appearance of many burials further supports this assessment.
By comparing ancient individuals from Abusir el-Meleq with modern Egyptian reference populations, we found an influx of sub-Saharan African ancestry after the Roman Period, which corroborates the findings by Henn and colleagues16. Further investigation would be needed to link this influx to particular historic processes. Possible causal factors include increased mobility down the Nile and increased long-distance commerce between sub-Saharan Africa and Egypt49. Trans-Saharan slave trade may have been particularly important as it moved between 6 and 7 million sub-Saharan slaves to Northern Africa over a span of some 1,250 years, reaching its high point in the nineteenth century50. However, we note that all our genetic data were obtained from a single site in Middle Egypt and may not be representative for all of ancient Egypt. It is possible that populations in the south of Egypt were more closely related to those of Nubia and had a higher sub-Saharan genetic component, in which case the argument for an influx of sub-Saharan ancestries after the Roman Period might only be partially valid and have to be nuanced. Throughout Pharaonic history there was intense interaction between Egypt and Nubia, ranging from trade to conquest and colonialism, and there is compelling evidence for ethnic complexity within households with Egyptian men marrying Nubian women and vice versa51,52,53. Clearly, more genetic studies on ancient human remains from southern Egypt and Sudan are needed before apodictic statements can be made.
The ancient DNA data revealed a high level of affinity between the ancient inhabitants of Abusir el-Meleq and modern populations from the Near East and the Levant. This finding is pertinent in the light of the hypotheses advanced by Pagani and colleagues, who estimated that the average proportion of non-African ancestry in Egyptians was 80% and dated the midpoint of this admixture event to around 750 years ago17. Our data seem to indicate close admixture and affinity at a much earlier date, which is unsurprising given the long and complex connections between Egypt and the Middle East. These connections date back to Prehistory and occurred at a variety of scales, including overland and maritime commerce, diplomacy, immigration, invasion and deportation54. Especially from the second millennium BCE onwards, there were intense, historically- and archaeologically documented contacts, including the large-scale immigration of Canaanite populations, known as the Hyksos, into Lower Egypt, whose origins lie in the Middle Bronze Age Levant54.
Our genetic time transect suggests genetic continuity between the Pre-Ptolemaic, Ptolemaic and Roman populations of Abusir el-Meleq, indicating that foreign rule impacted the town’s population only to a very limited degree at the genetic level. It is possible that the genetic impact of Greek and Roman immigration was more pronounced in the north-western Delta and the Fayum, where most Greek and Roman settlement concentrated43,55, or among the higher classes of Egyptian society55. Under Ptolemaic and Roman rule, ethnic descent was crucial to belonging to an elite group and afforded a privileged position in society55. Especially in the Roman Period there may have been significant legal and social incentives to marry within one’s ethnic group, as individuals with Roman citizenship had to marry other Roman citizens to pass on their citizenship. Such policies are likely to have affected the intermarriage of Romans and non-Romans to a degree55. Additional genetic studies on ancient human remains from Egypt are needed with extensive geographical, social and chronological spread in order to expand our current picture in variety, accuracy and detail.
However, our results revise previous scepticism towards the DNA preservation in ancient Egyptian mummies due to climate conditions or mummification procedures8. The methodology presented here opens up promising avenues for future genetic research and can greatly contribute towards a more accurate and refined understanding of Egypt’s population history.
|
|