Pets & Animals

Biochemical and physical correlates of DNA contamination in archaeological human bones and teeth excavated at Matera, Italy

Description
Biochemical and physical correlates of DNA contamination in archaeological human bones and teeth excavated at Matera, Italy
Categories
Published
of 9
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  Biochemical and physical correlates of DNAcontamination in archaeological human bones andteeth excavated at Matera, Italy M. Thomas P. Gilbert a, *, Lars Rudbeck b , Eske Willerslev a,c ,Anders J. Hansen c,d , Colin Smith e,1 , Kirsty E.H. Penkman e,2 ,Kurt Prangenberg e,f  , Christina M. Nielsen-Marsh e , Miranda E. Jans g ,Paul Arthur h , Niels Lynnerup i , Gordon Turner-Walker  j,3 , Martin Biddle k ,Birthe Kjølbye-Biddle k , Matthew J. Collins e,g,2 a Henry Wellcome Ancient Biomolecules Centre, Department of Zoology, University of Oxford,South Parks Road, Oxford OX1 3PS, UK  b Research Laboratory, Institute of Forensic Medicine, University of Copenhagen, Frederik V Vej 11,DK-2100 Copenhagen, Denmark c Department of Evolutionary Biology, Zoological Institute, University of Copenhagen, 5 Universitetsparken,DK-2100 Copenhagen Ø, Denmark d Department of Forensic Genetics, University of Copenhagen, Frederik V’s Vej 11, DK-2100 Copenhagen, Denmark e Fossil Fuels and Environmental Geochemistry, NRG, Drummond Building, University of Newcastle,Newcastle upon Tyne NE1 7RU, UK  f  Institut fu ¨ r Geowissenschaften, Sigwartstrasse 10, 72076 Tu ¨ bingen, Germany g Institute for Geo and Bioarchaeology, Vrije Universiteit Amsterdam, De Boelelaan 1085, 1081 HV Amsterdam, Holland  h Dipartimento di Beni Culturali, Via D. Birago, 64, University of Lecce, 73100 Lecce, Italy i Laboratory of Biological Anthropology, Institute of Forensic Medicine, University of Copenhagen,Frederik V Vej 11, DK-2100 Copenhagen, Denmark  j Institutt for arkeologi og kulturhistorie, NTNU, Vitenskapsmuseet, 7491 Trondheim, Norway k Hertford College, University of Oxford, Oxford OX1 3BW, UK  Received 28 October 2004; received in revised form 29 October 2004 Abstract The majority of ancient DNA studies on human specimens have utilised teeth and bone as a source of genetic material. In thisstudy the levels of endogenous contamination (i.e. present within the sample prior to sampling for the DNA analysis) are assessedwithin human bone and teeth specimens sampled from the cemetery of Santa Lucia alle Malve, Matera, Italy. This site is of exceptional interest, because the samples have been assayed for 18 measures of biochemical and physical preservation, and it is theonly one identified in a study of more than 107 animal and 154 human bones from 43 sites across Europe, where a significantnumber of human bones was well preserved. The findings demonstrate several important issues: (a) although teeth are more resilient * Corresponding author. Ecology and Evolutionary Biology, The University of Arizona, 1041 East Lowell Street, Tucson, AZ 85721, USA. Tel.: C 1 520 621 4881; fax:  C 1 520 621 9190. E-mail address:  mtpgilbert@spymac.com (M. Thomas P. Gilbert). 1 Present address: Department of Palaeobiology, Museo Nacional de Ciencias Naturales (CSIC), C/Jose ´ Gutie ´rrez Abascal, 228006 Madrid, Spain. 2 Present address: BioArch, The King’s Manor, University of York, York YO1 7EP, UK. 3 Present address: Institute for Cultural Heritage Conservation, National Yunlin University of Science and Technology, 123 University Road Sec.3, Touliu, Yunlin 640, Taiwan.0305-4403/$ - see front matter    2005 Elsevier Ltd. All rights reserved.doi:10.1016/j.jas.2004.12.008Journal of Archaeological Science 32 (2005) 785 e 793 http://www.elsevier.com/locate/jas  to contamination than bone, both are readily contaminated (presumably through handling or washing), and (b) once contaminatedin this way, both are difficult (if not impossible) to decontaminate. Furthermore, although assessed on  bone  samples, several of thespecific biochemical and physical characteristics that describe overall sample preservation, levels of microbial attack and relatedincreases in sample porosity directly correlate with the presence of observable contamination in both  bone  and  teeth  samples fromindividual samples. While we can only speculate on the cause of this relationship, we posit that they provide useful guides for theassessment of whether samples are likely to be contaminated or not.   2005 Elsevier Ltd. All rights reserved. Keywords:  Ancient DNA; Biopreservation; Bone; Contamination; Diagenesis; Human; Teeth 1. Introduction Bones and teeth are normally the longest lastingphysical evidence of human or animal presence at anarchaeological site, and are also the most widely usedsources of samples for ancient DNA (aDNA) studies.Post hoc explanations of their suitability as a sourceof ancient DNA have identified retarded rates of decomposition, arising from adsorption of DNA tohydroxyapatite [30], low water content [22], ‘mummifi- cation’ of individual cells [4,5] and physical exclusion of microbes and external contaminants [22]. Recently, anawareness of sample handling as a source of contami-nation has led researchers to investigate teeth as anaDNA source. One hypothetical benefit is protectionconferred by enamel [32]. Additionally, although histo-logical studies identify higher numbers of DNAcontaining cells per unit area of bone than teeth [12],several studies have reported better DNA yields in teeththan bone [29,32].Richards et al. [36] have argued convincingly thatcontamination, not DNA preservation, is the greatestproblem facing the field, although the two are evidentlylinked e sparse, damaged endogenous DNA is less likelyto be amplified than modern contamination. Although itis known that teeth and bone may become contaminatedprior to aDNA extraction [36], current techniques usedto decontaminate specimens e the application of bleach,exposure to UV light, and grinding or shot-blasting  e reflect a belief that contamination is concentrated in theouter surface of the material. Protocols designed to limitcontamination stress the prevention of contact betweensamples and previously amplified DNA (amplicons)[11,18]. Nevertheless, even when strict protocols arefollowed contaminants are frequently observed. Forexample, human DNA has been reported from cave bear[21], 500-year-old pig samples [36] and 109 out of 168 relatively recent fox teeth [43]. More seriously, severalstudies report significant numbers of human remainscontaminated with multiple human sequences [19,27].Obviously, decontamination methods are not 100%efficient, and contamination remains a serious threat tothe validity of ancient DNA studies, particularly onhuman templates.In this study we have investigated the presence andpersistence of contamination in teeth and femur samplescollected from human skeletons excavated at thecemetery of Santa Lucia alle Malve, Matera, Italy [8]. Bone  samples from the specimens have been assayed for18 measures of biochemical and physical preservation,allowing these parameters to be correlated with modernDNA contamination in both  bones  and  teeth . 2. Materials and methods Twenty-six teeth and eight femur samples were takenfrom 13 individuals excavated at the cemetery above thecave-church of Santa Lucia alle Malve, Matera, Italy [8](Table 1). The pH of the soil at the site ranges from 8.0to 8.3. Carbon-14 dating indicates that the samples dateto approximately the late 14th century (Clare Owen,Oxford RLAHA, pers. comm.). In a study thatinvestigated more than 107 animal and 154 humanbones from 41 sites across Europe (spanning fourclimatic regimes  e  Mediterranean, Continental, Mari-time (coastal) and Subarctic, and dating from 250 to5950 Y.B.P. [24]), this site was the only one in whicha significant number of human bones (7 of 14) were wellpreserved [23 e 25,38] (Table 1). Additionally, as part of  this study the extent of the combined asparagines/aspartic acid (Asx) racemization was determined in onetooth from each individual using the extraction methodof Poinar et al. [35] and the analytical method of Kaufman and Manley [26]. Furthermore, all have been stored together, and have undergone similar amounts of human handling post-excavation. 2.1. Ancient DNA extraction and amplification DNA was extracted from samples following strictancient DNA protocols in order to prevent samplecontamination with previously amplified DNA [15].Importantly, several different extractions (2 e 3) wereperformed per individual from distinct parts of the body(i.e. different teeth plus femur) to help identify both the‘endogenous’ DNA sequence and any contaminants.One extraction blank was used for every four samples to 786  M. Thomas P. Gilbert et al. / Journal of Archaeological Science 32 (2005) 785 e 793  monitor contaminants entering during the DNA extrac-tion. DNA extractions from bone used 0.2 g bonepowder, collected as in Barnes et al. [3]. Unlessotherwise stated, all DNA was extracted from teethfollowing Gilbert et al. [15]. All PCRs were performedon each sample at least twice, using the polymeraseenzyme Platinum Taq Hi-Fidelity (invitrogen), and in-corporating one PCR blank to every three samples.PCRs for human mitochondrial DNA (mtDNA) usedprimers L16209-H16356 [19] following Gilbert et al. [15]. We have previously demonstrated that these primers areexceedingly sensitive to low levels of template DNA,thus are unlikely to generate false negative data [16].DNA extracts that did not yield PCR products werescreened for the presence of PCR inhibitors (a commonphenomena in aDNA studies [33]) through ‘spiking’PCRs containing amplifiable DNA with an equalvolume (1 m l) of the potential inhibitor, and monitoringany reduction in PCR success [15]. All amplified humanPCR products were cloned. Up to 12 colonies weresequenced per cloned PCR (325 total, average 10.2clones per PCR), thus providing between 16 and 35cloned sequences across all independent extracts of eachindividual skeleton (average 25 per skeleton). Forfurther details refer to supplemental information.Endogenous and contaminant samples were identi-fied as described in the supplemental information. Inbrief, cloned PCR products were sequenced andcompared with the Cambridge Reference Sequence [2]to identify sample-specific motifs, both within the clonesfrom each extract, and between different extracts fromeach individual. Under an assumption that samplescontain authentic, uncontaminated DNA, it is expectedthat all cloned sequences will yield the same sequence(Fig. 1A). However, due to small amounts of post-mortem DNA damage (for example, hydrolytic de-amination of cytosine to uracil, providing isolated C / Tand G / A mutations, or hydrolytic deamination of adenine to hypoxanthine, providing isolated A / G andT / C mutations), aDNA sequences often exhibit smallamounts of variation around a consensus sequence(Fig. 1B). If however the sample is uncontaminated, allsequences will contain shared motifs that identify themas originating from one original source of DNA(Fig. 1A and B). However, should a sample becontaminated with non-endogenous sources of DNA,in most situations sequences that do not share motifswill be observed among the clones (Fig. 1C).This method has two potential weaknesses. Firstly, if a sample contains no endogenous DNA, but has beencontaminated with one source of exogenous DNA(which possibly due to the time-lapse since contamina-tion may even contain some evidence of damage-drivenmiscoding lesions), the results will appear authentic.However, as all samples investigated in this study havebeen handled by multiple individuals, there is noplausible reason why only one DNA source will berepresented among contaminant sequences. A moreproblematic issue arises when samples have beencontaminated by mtDNA sequences that are identical,or very similar (e.g. differing by 1 e 2 bp) to that of thesample. Based on the diversity observed among WesternEuropean mtDNA sequences (c.f. [37]), this is unlikelyto be an issue unless the sequence is exceedinglycommon, as in the case of basal haplogroup Hsequences (those that are identical to, or differ bya single step from, the Cambridge Reference Sequence(CRS) over the region of interest, and observed atfrequencies of up to 60% in Western Europe [41]). Insuch cases the lack of distinct motifs in the sequencemakes it exceedingly difficult to distinguish contaminantsequences from srcinal, damaged sequences (Fig. 1D).Therefore, such samples were left out of the analysis toprevent misidentification of contaminated samples. Forfurther details on this method and why we believe it toenable us to differentiate between authentic andcontaminant DNA sequences, we refer readers to thearguments presented elsewhere [16] and to the in-formation presented in the supplemental information. 2.2. Assessment of bone preservation correlateswith contamination Human bones suffer common patterns of alterationwith microscopic focal destruction resulting in areas of dense mineralization surrounding pores with diametersat 600 nm and 1.2  m m microbial alteration (‘‘ m ’’porosity, 600 nm e 1.5  m m [24,42]). We believe that thisdiagenetic feature is a characteristic of the rapidputrefaction of interred corpses by blood borne gutbacteria as suggested by Bell et al. [5]. In order tosimplify inter-sample variation, the mid-shaft of thefemur (as a large and commonly preserved element) wasassessed to determine the extent of diagenetic alteration.At the Matera site, only half of the sampled individualshad the putrefaction alteration which characterized 74%of all human skeletons in the European study. Theremaining femurs were remarkably well preserved,partly it would appear because of secondary mineralprecipitation. Matera was the only site of 50 studiedwith a significant number of well-preserved humanbones [25]. We still remain ignorant of why Matera is sounusual in this respect, the burial in limestone cuts, andextreme seasonal variation in temperature and moisture,may both be implicated. The detailed record of bonepreservation of the Matera samples, the dipolar patternof preservation and the similar treatment of the bonessince excavation presents an ideal opportunity toinvestigate correlations between contamination andpreservation indices.Unpublished data shows that bone is more proneto secondary contamination than dentine (M.T.P.G, 788  M. Thomas P. Gilbert et al. / Journal of Archaeological Science 32 (2005) 785 e 793  unpublished data), as the enamel partially protects thedentine from allochthnaous DNA. We, therefore, choseto attempt to extract DNA from the dentine, in additionto extractions from the mid-shaft femur samples,reasoning that although we were unable to conduct thesame suite of analyses on the dentine, the femur datashould give an insight into the deterioration of the wholeskeleton. 3. Results All extraction and PCR blanks were consistentlynegative throughout the study, indicating that theresults are unlikely to derive from contaminants in theextraction or PCR processes. Furthermore, the extent of racemization of aspartic acid and alanine in the sampleswas below the threshold over which DNA is unlikely tosurvive (following [35]). 3.1. Preservation and contamination of Matera teeth and bones Nine Matera teeth samples contained only one DNAsequence, and appear to be uncontaminated, while 12teeth extracts contained multiple sequences and wereidentified as contaminated (Table 1). The majority of  1.A 201 210 220 230 240 250 260caagcaagtacagcaatcaaccctcaactatcacacatcaactgcaactccaaagccacccctc A......................t................................t........B......................t................................t........C......................t................................t........D......................t................................t........E......................t................................t........F......................t................................t........G......................t................................t........H......................t................................t........ 1.B 201 210 220 230 240 250 260caagcaagtacagcaatcaaccctcaactatcacacatcaactgcaactccaaagccacccctc A......................t...g............................t........B......................t................................t........C......................t................................t........D..........t....g......t................................t........E....a.................t................................t........F.......................................................t........G......................t................................t........H......................t................................t........ 1.C 201 210 220 230 240 250 260caagcaagtacagcaatcaaccctcaactatcacacatcaactgcaactccaaagccacccctc A......................t...........g....................t........B......................t................................t........C........c.......................................................D......................t................................t......c.E........c.......................................................F......................t................................t........G........c.......................................................H........c....................................................... 1.D 201 210 220 230 240 250 260caagcaagtacagcaatcaaccctcaactatcacacatcaactgcaactccaaagccacccctc A.......................................................t........B......................t.........................................C.......................................................t........D................................................................E................................................................F.......................................................t........G................................................................H.....................................g.......................... Fig. 1. The identification of contaminated sequences from a cloned PCR product. (A) Cloned PCR products from uncontaminated, undamagedDNA extracts show no sequence variation. (B) Uncontaminated, but damaged PCR products show some sequence variation, but also conservedsequence motifs among the clones. (C) Cloned PCR products containing 1 old, damaged DNA source, plus a modern, contaminant source of DNA.(D) In some situations it is not possible to differentiate whether variation among cloned sequences is due to contamination or post mortem damage.For full details on figure refer to main text. For each decision applied to samples in this study refer to supplemental information.789 M. Thomas P. Gilbert et al. / Journal of Archaeological Science 32 (2005) 785 e 793

Grama

Apr 16, 2018
Search
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks