Takabuti, was a female who lived in ancient Egypt during the 25th Dynasty, c.660 BCE. Her mummified remains were brought to Belfast, Northern Ireland, in 1834 and are currently displayed in the Ulster Museum. To gain insight into Takabuti’s ancestry, we used deep sampling of vertebral bone, under X-ray control, to obtain non-contaminated bone tissue from which we extracted ancient DNA (aDNA) using established protocols. We targeted the maternally inherited mitochondrial DNA (mtDNA), known to be highly informative for human ancestry, and identified 38 single nucleotide variants using next generation sequencing. The specific combination of these SNVs suggests that Takabuti belonged to mitochondrial haplogroup H4a1. Neither H4 nor H4a1 have been reported in ancient Egyptian samples, prior to this study. The modern distribution of H4a1 is rare and sporadic and has been identified in areas including the Canary Islands, southern Iberia and the Lebanon. H4a1 has also been reported in ancient samples from Bell Beaker and Unetice contexts in Germany, as well as Bronze Age Bulgaria. We believe that this is an important finding because first, it adds to the depth of knowledge about the distribution of the H4a1 haplogroup in existing mtDNA, thus creating a baseline for future occurrences of this haplogroup in ancient Egyptian remains. Second, it is of great importance for archaeological sciences, since a predominantly European haplogroup has been identified in an Egyptian individual in Southern Egypt, prior to the Roman and Greek influx (332BCE).
Takabuti lived in Thebes, Egypt, during the 25th Dynasty (c. 660 BCE). Her mummified remains and the coffin containing them were brought to Belfast, Northern Ireland, in 1834 and are currently displayed in the Ulster Museum (https://www.nmni.com/our-museums/ulster-museum/Things-to-see/Takabuti-the-ancient-Egyptian-mummy.aspx ). Translation of the inscriptions on her coffin identify her as the daughter of Nespare, a priest of Amun at Thebes, and his wife Taseniret.
Takabuti’s initial morphological examination in 1834 revealed a female of 1.55 m (5 feet 1 inch) with exceptionally well-preserved dentition from which an age at death of 25–30 years was estimated. The decayed bandages, which had likely been dipped in resin, had covered the entire body including the head, arms, legs and feet (Konstantina Drosou, personal communication, December 15, 2018).
Four days after examination, the body was rewrapped in Irish linen, leaving the head, one arm and one foot exposed. More comprehensive studies of the mummy, carried out by X-ray imaging in 1987 and computerized tomography (CT) scanning in 2006, confirmed the age of 25–30 years at time of death and revealed a healthy individual with no evidence of serious childhood illness.
Although mitochondrial DNA (mtDNA) sequencing can be used to study population affinities of archaeological human remains1, DNA degradation in the Egyptian climate once limited the usefulness of this approach on mummified individuals2. Next generation sequencing, preceded by targeted hybridization capture can be used to enrich for particular components of the human genome3. Such techniques, have changed how DNA data can be obtained, providing rich analytical data from which sex4 , kinship5 and population affinity6 can be inferred. Using established hybridization capture methods and next generation sequencing we identified the mitochondrial haplotype of Takabuti and compare her haplogroup with the mitochondrial haplogroups previously identified in ancient Egyptian mummies.
The three bone biopsies were combined in one tube, and generated bone powder of 50 mg in total. Sample was then subjected to DNA extraction and library preparation followed by two rounds of in-solution hybridization capture directed at the mtDNA (NC_012920.1) and sequenced using Illumina technology. The library prepared after the first round of capture gave 18,903,330 sequence reads, of which 81,384 were initially mapped to the mtDNA (Table 1). After removal of duplicates, this mapped dataset was reduced to 991 reads with a mean coverage of 2.9×. After the second round of capture, 24,085,550 reads were obtained, of which 2,532,752 were initially mapped to the mtDNA and 3550, with a mean coverage of 9.8× were retained after duplicate removal. In order to increase the mapping stringency, the 3550 unique reads obtained from the second capture were individually tested against the mtDNA (NC_012920.1) by BLAST analysis. This left 3001 high-confidence reads, with a mean coverage of 9.8× and maximum coverage of 131×. Analysis of the read datasets with MapDamage revealed miscoding lesion distribution patterns typical of ancient DNA (Fig. 1).
From the 3001 sequence reads from the second round of capture, 38 SNVs were identified with search parameters set at minimum variant frequency of 100% and a minimum SNV coverage of 4× (Table 2). Five of these SNVs were also identified in the read dataset from the first capture. Of these 38 SNVs, five (NC_012920.1:m.1057G > A, m.2647G > A, m.3503C > T, m.12450C > T and m.14365C > T) were considered low confidence due to their location close the 5´ or 3´ termini of sequence reads, which means that they could potentially be miscoding lesions caused by ancient DNA degradation7. All of the SNVs were transitions with the exception of NC_012920.1:m.6088C > A and m.6597C > G. SNVs showed no matches to personnel’s mitochondrial profiles.
Examination of the SNVs indicated that Takabuti belonged to mtDNA haplogroup H4a1 (Fig. 2) with 90.57% confidence (Overall quality 0.906, HG Quality 1.000, Sample Quality 0.811). This assignment is based on the presence of eleven variants in total including three basal H4 variants (m.3992C > T, m.5004 T > C, m.9123G > A), along with m.4024A > G which characterizes H4a, and the low confidence SNV m.14365C > T which is associated with H4a18,9,10. An additional SNV, m.14582A > G, which is characteristic of H4a was present in the dataset but does not appear in Table 2 as the read coverage after second capture was only 2×. Of the remaining SNVs, m.3992C > T, m.4024A > G, m.4769A > G, m.8860A > G and m.9123G > A have been found before in mtDNA variants that have been assigned to haplogroup H4a111,12,13. The remaining of the SNVs were not included in data interpretation as they have not been described before.
We determined Takabuti’s mtDNA haplogroup from multiple overlapping sequence reads obtained after two rounds of hybridization capture directed at the mitochondrial genome. The sequences displayed damage patterns typical of ancient DNA, and the SNVs used to deduce the haplogroup were absent from the mtDNAs of all individuals involved in the DNA extraction and processing. Furthermore stringent steps were taken throughout all stages of this research to limit environmental and cross-contamination events (deep tissue sampling, PPE). This work was performed on bone tissue removed by minimally invasive techniques necessitating of only milligram quantities of bone and which minimised contamination. We therefore have confidence that the haplogroup we report for Takabuti is correct and not affected by modern contamination.
When interpreting our findings within the broader genetic record, the H super-haplogroup is the most common mtDNA lineage in Europe and is found also in parts of present-day Africa and western Asia9,14. The H4a1 variant possessed by Takabuti is relatively rare with a modern distribution including ~ 2% of a southern Iberian population15 , ~ 1% in a Lebanese population12 and ~ 1.5% of multiple Canary Island populations13.
Until our work neither H4 nor H4a1 has been reported in ancient Egyptian samples. However, in the archaeological record H4a1 has been reported in sixth–fourteenth century CE remains sourced from the Canary Islands, and three additional ancient DNA samples, two from Bell Beaker and Unetice contexts (2500–1575 BCE) at Quedlinburg and Eulau, both in Saxony-Anhalt, Germany10 , and one individual from early Bronze Age Bulgaria15 illustrating both the rare occurrence and sporadic distribution of this haplogroup.
The overall perspective from an examination of 97 samples from ancient Egypt with a mitochondrial haplogroup (Fig. 3) is of a complicated society with a rich mixture of established genetic backgrounds, indicative of a population shaped over time by migration into the region. This is perhaps not surprising since Egypt is situated at the only land gateway between Africa and the Middle East, a region known to have been populated through the centuries by nomadic tribes and rich trading routes. As the record currently stands these are represented by the U and M1a1 haplogroups throughout the first and second millennia BCE, expanded by J2a, R0, T1, T2, HV and I in the first millennium BCE4,5,6,16,17. Superimposed on this are individuals like Takabuti with rare haplogroups, which have not been previously identified in the background.
Perhaps the most intriguing aspect of our findings, which is of great archaeological interest and importance, is the observation of a predominantly European haplogroup in an Egyptian individual located in Southern Egypt. What is fascinating is that the individual pre-dates the Roman and Greek influx (332BCE). At face value, the current genetic evidence suggests a high degree of isolation from migration into Southern Egypt. However, this finding challenges that assertion, suggesting that further investigative work could be carried out to gain a better understanding of the genetic makeup of ancient Southern Egypt.
The simplified representation of mitochondrial haplogroups in ancient Egypt in Fig. 3 demonstrates the importance of studying individuals, in order to strengthen the archaeological maternal genetic record of ancient Egypt. This extends beyond our key finding of an individual who is clearly not characteristic of the background maternal lineages based on the currently known haplotyped population. For example, in one of our previous papers we identified the M1a1 haplogroup in two mummies5, pushing back the earliest observation of this haplogroup in Egyptian mummies by 500 years. Similarly, a study by Loreille et al., 20184 pushed back the recorded chronology of the U haplogroup in ancient Egypt by almost 1000 years. Therefore, single-case studies add to existing knowledge in this field, challenging and updating our current understanding.
Our results add to the growing body of reports demonstrating the utility of hybridization capture as a means of obtaining authentic ancient DNA sequences from Egyptian mummies4,5,6 and the importance of those findings to better understand and interpret ancient Egyptian populations.
Deep tissue sampling
Sampling was performed at the Ulster Museum in the gallery where the mummy is on display, using strict anti-contamination controls. Following isolation of the display area, the mummy was placed on a medium-density fibreboard supported by trestles that were constructed in situ for the purpose of sampling (Fig. 4). Personnel were equipped with lead jackets, forensic suits (Tyvek), boot covers, hair nets, face masks, goggles and gloves. Deep bone tissue samples (Fig. 4) were obtained with the assistance of a portable C-arm X-ray imaging intensifier, (Philips BV Endura), positioned next to the mummy to accurately locate the vertebrae, as the mummy remained wrapped to preserve her morphological integrity (Fig. 4). Three biopsies were performed in total using one biopsy needle (Murphy M2 Diamond Tip, 11 g×15 cm, UK Medical), from the lumbar vertebral body 3 (L3). The first layer of superficial bandage was retracted using a sterile surgical retractor enabling the biopsy to be performed through the deeper bandaging layer opposite the level of the L3 vertebra in the mid-line (Fig. 5). Prior to sampling, each entry point was treated with DNA-Away (Molecular BioProducts), in order to reduce contamination from dust particles. Bone powder from the biopsies was immediately transferred into three sterile 50 ml falcon tubes (one tube for each biopsy to avoid reopening) which were then wrapped in two layers of UV irradiated aluminium foil.
Ancient DNA extraction and sequencing
Samples were transferred to the Manchester Institute of Biotechnology where DNA extraction and sequencing library preparation were performed in a set of physically isolated, restricted access laboratories, each equipped with an ultrafiltered air supply system maintaining positive displacement pressure. The laboratories were periodically sterilized by UV irradiation when not in use. All surfaces were cleaned with 5% sodium hypochlorite solution and 70% ethanol, and all utensils and equipment such as pipettes were treated with DNA-Away before and after use. Consumables such as tubes were UV irradiated (254 nm, 120,000 μJ cm−2 for 2 × 5 min, with 180° rotation between exposure) before use. Personal protective equipment included Tyvek forensic suits, face masks, hair nets, goggles, boot covers and two pairs of sterile gloves. DNA extraction was carried out in a Class II biological safety cabinet, and sequencing libraries and polymerase chain reaction (PCR) mixes were prepared in a laminar flow cabinet. DNA extraction was accompanied by a DNA-free negative control (normal extraction but without sample) followed by a DNA-free PCR negative control and a library preparation negative control, the latter controls set up with water rather than DNA extract5. To test for potential contamination during sampling and DNA processing, mouth swabs were taken from all individuals present during the tissue sampling and from individuals working in the Manchester ancient DNA labs, and these samples were anonymized and the mtDNA for each sample was typed. Samples were obtained by informed consent and all steps in this process were performed in accordance with the University of Manchester ethics regulations. Institutional governance check confirms that the work was conducted in accordance with the University of Manchester policy.
DNA was extracted from 50 mg of bone powder (resulting from a total of three biopsies) following the procedure of Dabney18, modified by the addition of 10% (w/v) N-lauroylsarcosine to the extraction buffer and 5 M NaOAc and 5 M NaCl to the PB buffer19,20.
Preliminary analyses involved two overlapping PCRs21 directed at the mtDNA hypervariable region I (HVRI) loci (NC_012920.1:m16028-16,195 and NC_012920.1:m16210-16,340) to assess DNA preservation and endogenous DNA content, as well as a multiplex PCR using previously established primers22 targeting the amelogenin locus (NC_000023.11:g.11314994_11315100 (chrX) and NC_000024.10:g.6738028_6738138 (chrY)) to confirm the sex of the mummy (Table 3). PCRs were performed with the Multiplex PCR kit (Qiagen) in a final volume of 25 μl consisting of 3 μl DNA extract, 10 μM each primer and 1×Qiagen master mix. Thermocycling conditions were: 95 °C for 15 min; 44 cycles of 94 °C for 0.5 min, annealing temperature for 1.5 min, 72 °C for 1.0 min; 72 °C for 15 min. PCR products were examined in 1.5% and 4.0% agarose gels, purified (Qiagen MinElute Purification kit) and sequenced from both ends by the Sanger method (GATC Biotech, Cologne). Sequences were mapped to the rCRS using Geneious v8.1.823. The amelogenin test failed probably due to poor preservation, whereas preliminary PCRs showed no variations.
Three dual-indexed NGS libraries were prepared24,25 including one library and one extraction negative control. Only the sample library was enriched twice by in-solution hybridization capture (Arbor Biosciences) according to the manufacturer's instructions for degraded samples, using a baitset covering the entire mitochondrial genome. A total of 12 and 15 cycles of post-capture PCR was performed with the enriched products using the IS5 (5′-AATGATACGGCGACCACCGA-3′) and IS6 (5′-CAAGCAGAAGACGGCATACGA-3′) primers. Libraries (one enriched and two non-enriched) were quantified by qPCR (Roche LightCycler 480) and fluorimetry (Qubit 2.0), and their length distribution assessed using a bioanalyzer (Agilent). The concentration of the extraction and library controls were undetectable and therefore were not enriched or sequenced. The remaining library was then purified using the MinElute Purification kit (Qiagen) and both enrichment rounds sequenced using the MiSeq Reagent kit (v3) 2×75. Sequence data are curated at the European Nucleotide Archive under project accession number PRJEB38492.
Raw data were demultiplexed using BBmap v38.07 (sourceforge.net/projects/bbmap/) and adapter sequences removed using AdapterRemoval v2.126 specifying the following parameters: trimns, minlength 25, trimqualities, minquality 30, collapse. Reads with overlaps of at least 11 bp were collapsed into single sequences, whereas reads that did not overlap were processed separately using BEDTools v2.26.027 and only confirmed forward and reversed pairs retained. Mapping was performed with BWA v.0.7.5a-r40528 using default aln, samse and sampe commands with the exception of the l1024 parameter which is suggested for ancient DNA data. The dataset was mapped against the rCRS reference genome29. Subsequently, the sam files were cleaned and sorted by coordinate system and converted to bam format using default CleanSam and Sortsam commands from Picard Tools v2.18.27 (https://broadinstitute.github.io/picard). Mapped reads were extracted with Samtools v1.230 using the parameters b, q30, and F4. Removal of duplicate sequences was performed with the script aweSAM collapser (https://gist.github.com/jakeenk/). Single nucleotide variations (SNVs) were called using HaplotypeCaller from GATK v.431 and validated using Geneious v.R823. The depth of SNV coverage was set at DP > 4 and only those SNVs with a 100% read coverage were retained. Mitochondrial DNA haplogroup assignment was performed using Haplogrep32 sand data authenticity was assessed with MapDamage v2.07,33. BLAST34 searches were carried out online (https://blast.ncbi.nlm.nih.gov/Blast.cgi).
Brown, T. & Brown, K. Biomolecular Archaeology: An Introduction (Wiley, Chichester, 2011).
Marchant, J. Ancient DNA: Curse of the Pharaoh’s DNA. Nature 472, 404–406 (2011).
Lippold, S. et al. Human paternal and maternal demographic histories: Insights from high-resolution Y chromosome and mtDNA sequences. Investig. Genet. 5, 1–17 (2014).
Loreille, O. et al. Biological sexing of a 4000-year-old egyptian mummy head to assess the potential of nuclear DNA recovery from the most damaged and limited forensic specimens. Genes (Basel).9, 135 (2018).
Drosou, K., Price, C. & Brown, T. A. The kinship of two 12th Dynasty mummies revealed by ancient DNA sequencing. J. Archaeol. Sci. Rep. 17, 793–797 (2018).
Schuenemann, V. J. et al. Ancient Egyptian mummy genomes suggest an increase of Sub-Saharan African ancestry in post-Roman periods. Nat. Commun. 8, 15694 (2017).
Jónsson, H., Ginolhac, A., Schubert, M., Johnson, P. L. F. & Orlando, L. MapDamage2.0: Fast approximate Bayesian estimates of ancient DNA damage parameters. Bioinformatics29, 1682–1684 (2013).
Herrnstadt, C. et al. Reduced-median-network analysis of complete mitochondrial DNA coding-region sequences for the major African, Asian, and European haplogroups. Am. J. Hum. Genet. 70, 1152–1171 (2002).
Roostalu, U. et al. Origin and expansion of haplogroup H, the dominant human mitochondrial DNA lineage in west Eurasia: The Near Eastern and Caucasian perspective. Mol. Biol. Evol. 24, 436–448 (2007).
Brotherton, P. et al. Neolithic mitochondrial haplogroup H genomes and the genetic origins of Europeans. Nat. Commun. 4, 1764 (2013).
Olivieri, A. et al. Mitogenome Diversity in Sardinians: A Genetic Window onto an Island’s Past. Mol. Biol. Evol. 34, 1230–1239 (2017).
Matisoo-Smith, E. et al. Ancient mitogenomes of Phoenicians from Sardinia and Lebanon: A story of settlement, integration, and female mobility. PLoS ONE 13, 1–19 (2018).
Fregel, R. et al. Mitogenomes illuminate the origin and migration patterns of the indigenous people of the Canary Islands. PLoS ONE 14, 1–24 (2019).
Bekada, A. et al. Introducing the Algerian Mitochondrial DNA and Y-Chromosome Profiles into the North African Landscape. PLoS ONE 8, e56775 (2013).
Hernández, C. L. et al. The distribution of mitochondrial DNA haplogroup H in southern Iberia indicates ancient human genetic exchanges along the western edge of the Mediterranean. BMC Genet. 18, 1–14 (2017).
Khairat, R. et al. First insights into the metagenome of Egyptian mummies using next-generation sequencing. J. Appl. Genet. 54, 309–325 (2013).
Oras, E. et al. Multidisciplinary investigation of two Egyptian child mummies curated at the University of Tartu Art Museum, Estonia (Late/Graeco-Roman Periods). PLoS ONE 15, 1–27 (2020).
Dabney, J. et al. Complete mitochondrial genome sequence of a Middle Pleistocene cave bear reconstructed from ultrashort DNA fragments. Proc. Natl. Acad. Sci. USA 110, 15758–15763 (2013).
Damgaard, P. B. et al. Improving access to endogenous DNA in ancient bones and teeth. Sci. Rep. 5, 1–12 (2015).
Allentoft, M. E. et al. Population genomics of Bronze Age Eurasia. Nature 522, 167–172 (2015).
Bouwman, A. S., Brown, K. A., Prag, A. J. N. W. & Brown, T. A. Kinship between burials from Grave Circle B at Mycenae revealed by ancient DNA typing. J. Archaeol. Sci. 35, 2580–2584 (2008).
Butler, E. & Li, R. K. Genetic markers for sex identification in Forensic DNA analysis. J. Forensic Investig. 2, 10 (2014).
Kearse, M. et al. Geneious Basic: An integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics 28, 1647–1649 (2012).
Kircher, M., Sawyer, S. & Meyer, M. Double indexing overcomes inaccuracies in multiplex sequencing on the Illumina platform. Nucleic Acids Res. 40, 1–8 (2012).
Meyer, M. & Kircher, M. Illumina sequencing library preparation for highly multiplexed target capture and sequencing. Cold Spring Harb. Protoc. 5, 5448 (2010).
Schubert, M., Lindgreen, S. & Orlando, L. AdapterRemoval v2: Rapid adapter trimming, identification, and read merging. BMC Res. Notes 9, 1–7 (2016).
Quinlan, A. R. & Hall, I. M. BEDTools: A flexible suite of utilities for comparing genomic features. Bioinformatics 26, 841–842 (2010).
Li, H. & Durbin, R. Fast and accurate short read alignment with Burrows–Wheeler transform. Bioinformatics 25, 1754–1760 (2009).
Andrews, R. M. et al. Reanalysis and revision of the Cambridge reference sequence for human mitochondrial DNA. Nat. Genet. 23, 147 (1999).
Li, H. et al. The sequence alignment/Map format and SAMtools. Bioinformatics 25, 2078–2079 (2009).
McKenna, A. The Genome Analysis Toolkit: A MapReduce framework for analyzing next-generation DNA sequencing data. Genome Res. 20, 1297–1303. https://doi.org/10.1101/gr.107524.110 (2010).
Kloss-Brandstätter, A. et al. HaploGrep: A fast and reliable algorithm for automatic classification of mitochondrial DNA haplogroups. Hum. Mutat. 32, 25–32 (2011).
Ginolhac, A., Rasmussen, M., Gilbert, M. T. P., Willerslev, E. & Orlando, L. mapDamage: Testing for damage patterns in ancient DNA sequences. Bioinformatics 27, 2153–2155 (2011).
Altschul, S. F., Gish, W., Miller, W., Myers, E. W. & Lipman, D. J. Basic local alignment search tool. J. Mol. Biol. 215, 403–410 (1990).
We thank the Ulster Museum for permission to take samples from Takabuti, and the Friends of the Ulster Museum who funded the DNA analysis. We are grateful to Mr Mark Regan and the Kingsbridge Private Hospital for providing assistance with the portable CT scanner, and the staff of National Museums NI who provided stalwart support. We also thank the Genomics Sequencing Facility and the Computational Shared Facility (CSF3) at the University of Manchester. Special thanks go to Dr David Tosh for coordinating the project, Emeritus Professor Terence A Brown, Emerita Professor Rosalie David and Dr Odile Loreille for their helpful discussions, suggestions and comments.
The authors declare no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Drosou, K., Collin, T.C., Freeman, P.J. et al. The first reported case of the rare mitochondrial haplotype H4a1 in ancient Egypt. Sci Rep 10, 17037 (2020). https://doi.org/10.1038/s41598-020-74114-9