Biomolecular study of the human remains from tomb 5859 in theEtruscan necropolis of Monterozzi, Tarquinia (Viterbo, Italy)
Enrico Cappellini
a
*, Brunetto Chiarelli
a
, Luca Sineo
b
, Antonella Casoli
c
,Antonella Di Gioia
c
, Cristiano Vernesi
d
, Maria Cristina Biella
e
, David Caramelli
a
a
Dipartimento di Biologia Animale e Genetica “Leo Pardi”, Laboratori di Antropologia, Universita` di Firenze, Via del Proconsolo 12,Firenze 50122, Italy
b
Dipartimento di Biologia Animale, Universita` di Palermo, Italy
c
Dipartimento di Chimica Generale e Inorganica, Chimica Analitica, Chimica Fisica, Universita` di Parma, Italy
d
Dipartimento di Biologia, Universita` di Ferrara, Italy
e
Dipartimento di Scienze Storiche, Archeologiche e Antropologiche dell’Antichita`, Universita` di Roma “La Sapienza”, Italy
Received 17 December 2002; received in revised form 1 April 2003; accepted 22 October 2003
Abstract
Archaeological excavation in an Etruscan room tomb, from the Monterozzi necropolis in Tarquinia led to the recovery of fourindividuals. It was hypothesized that they could be members of a single family group. As both archaeological data and classicalanthropological analysis provided little information in this direction, ancient DNA (aDNA) was extracted from bone and toothfragments of the individuals. For each subject HVR-I of the mitochondrial DNA (mtDNA) was cloned and sequenced. To identifythe sex of the individuals, amelogenine and SRY genes were analysed. Short tandem repeat (STR) characterization was alsoperformed. DNA studies were preceded by the evaluation of amino acids racemization extent and thermogravimetric analysis(TGA), to evaluate, respectively, degradation and quantity of organic matter preserved in the samples.Results show that two subjects are males, whereas two are females. Furthermore, three of them share the same mtDNA sequence,and, as such, they could be related by maternal lineage. This evidence supports the hypothesis that the occupants of the tomb canbe considered members of a family group composing two parents and their son and daughter. Molecular study supplies newdata to better define the reconstruction previously proposed, based only on a morphological and archaeological approach.Multidisciplinary investigation also allows comparison of the di
ff 
erent methods and integration of their contributions.
2003 Elsevier Ltd. All rights reserved.
Keywords:
 Etruscans; Ancient DNA; Sex identification; Amino acid racemization; TGA; mtDNA; Amelogenine
1. Introduction
The Monterozzi necropolis of Tarquinia, in southernEtruria, nowadays in northern Latium, is one of themost important funerary complexes left by Etruscanculture. The necropolis is placed on a hill near theeminence where the ancient city stood [8]. It consists of more than 6000 burials of di
ff 
erent typology (roomtombs, pit tombs and hole tombs) spanning a chrono-logical period from 7th to 2nd century BC [8,9]. Bothinhumations and incinerations are present. Further-more, the fame of this necropolis is due to the presenceof the largest group of painted tombs, more than 200,left by the Etruscan culture [40]. The Monterozziarchaeological complex has not been completely investi-gated yet, despite the geophysical prospecting carriedout in this area by the Lerici foundation in the secondhalf of the last century. Through this approach a con-siderable number of tombs was located on the map [11].Knowledge of the complex since the Renaissance didnot favour preservation [30], consequently only aminority number of tombs was found unspoiled byarchaeologists. Between 1968 and 1970, during geo-physical prospecting in an area called “Calvario”, some
* Corresponding author. Tel.: +39-055-215-349;fax: +39-055-283-358
E-mail address:
 enrico.cappellini@unifi.it (E. Cappellini).Journal of Archaeological Science 31 (2004) 603–612
SCIENCE
 Journal of 
Archaeological
http://www.elsevier.com/locate/jas
SCIENCE
 Journal of 
Archaeological
http://www.elsevier.com/locate/jas0305-4403/04/$ - see front matter
2003 Elsevier Ltd. All rights reserved.doi:10.1016/j.jas.2003.10.012
 
complete and not profaned tombs were found andexcavated.In one of these tombs, namely n. 5859, the skeletalremains of four inhumated individuals, laying on bedscut in the tu
ff 
 (Fig. 1) were found, together with theburial artefacts. The study of the latter allowed thedating of the tomb between the end of the 4th andthe first third of the 3rd century BC [10]. Moreover, thehuman remains were morphologically and morpho-metrically analysed immediately after their discovery[25,26]. Combined anthropological and archaeologicalstudy led to the conclusion that the occupants of thetomb were: a child, 5859/14 (1) (numbering used here isthe one adopted by Mallegni [25]), 11 or 12 years old; aman, 5859/15 (2), about 35 years old; and two women,5859/16 and 17 (4 and 3), both about 40–45 years old.Due to the young age and poor preservation of theremains it was not possible to identify the sex of individual 5859/14 (1). Sex identification for subjects5859/16 and 17 (4 and 3) was tentative, as it appears inthe accurate morphological description by Mallegni [25].Cavagnaro Vanoni [10], reporting anthropologicalresults, suggested the identification of the subject5859/16 “
almost for sure
” (translation from Italian byE.C.) as a male. She also stressed how this reconstruc-tion could define, in archaeological terms, an unusualcontext, because next to the skull of this subject a coupleof bronze rings, identified as earrings and commonlyconsidered a female ornament, were fonnd.Sex identification based solely on archaeological data,in presence of insu
cient morphological evidence, couldlead to imprecise conclusions. Particularly, in absenceof epigraphical data, conclusions are based on burialartefacts analysis only, assuming that specific objects canbe attributed exclusively to one of the two sexes. From areview of archaeological contexts currently in progress,it emerges that the attribution of a burial to a male or afemale based only on some particular elements amongthe burial artefacts can be not particularly precise. Atypical example is represented by the mirror and thestrigil [13]. It is a rooted custom to claim the mirror, anobject strictly related with beauty and body care, as a
Fig. 1. Map of tomb 5859, with position and numbers attributed to each subject (from [10], modified).
E. Cappellini et al. / Journal of Archaeological Science 31 (2004) 603–612
604
 
typical female attribute, while the strigil, whose functionis linked to physical activity, is commonly associatedwith a male. In the classical epoch, also in the Etruscanculture, some female burials with strigils are reported[7,27,41], as well as male burials in which a mirror is partof the burial artefacts. For this reasons it is clear thatsex identification achieved in this manner can only beconsidered preliminary and not conclusive.It is interesting to note that in tomb 5859, at the footof subject 5859/17 (3), both a mirror (inventory number19) and a strigil (inv. numb. 11) were found. Further-more, among the burial artefacts of subject 5859/16 (4),along with a couple of bronze rings (inv. numb. 43 e 44),identified as unadorned earrings, a strigil was alsodiscovered (inv. numb. 33).A seemingly similar condition was reported, in tomb5957, from the same necropolis, where one of the twoburials, was unmistakably attributed by [12] to a male,according with [26], despite the presence of a couple of  earrings (inv. numb. 5).
The opportunity to analyse ancient DNA (aDNA)preserved in human remains can give a useful contri-bution in this discussion. Among the first applications of this alternative approach to ancient samples from theItalian area, some studies to identify sex were performedon human remains of an extensive group of subjectscoming from some Etruscan necropolises [42,43]. In par-ticular DNA analysis performed on human remains fromtomb 5957 corroborated the identification of the male sexfor both the subjects found in that tomb [6], showing it tobe a useful instrument strictly within this context.
The investigation of DNA preserved in ancient bonetissues has enabled the clarification of many archaeo-logical and anthropological contexts [23,24,33]. Inparticular it is now possible to identify sex even in casesof very poor or fragmented remains, as well as in casesof children and infants [19,28], thus surmounting the limits of classical anthropological analysis, despite theattempts to improve morphological sex identification ininfant and juvenile skeletons [39]. Even more, through the analysis of mitochondrial DNA (mtDNA) it hasbeen possible to perform population genetics studies[5,44] and, in a low number of well documented cases, ithas also been possible to identify possible kinship inmaternal lineage [34].STR profiles on modern human DNA samples areconsidered the best tool for personal identification. Anine STR loci profile o
ff 
ers a MP (matching probability)value of 10
11
between two modern unrelated DNAsamples [22].The possiblity of obtaining personal identi-fication and kinship reconstruction for the samples fromtomb 5859 was tried, even though STR studies appliedto aDNA could lead to inaccurate conclusions, due tothe degradation of DNA. Nevertheless, the use of STRon ancient samples has been applied already in somecases [17,18,38]In this framework, the study of aDNA from theremains of the occupants of tomb 5859 appeared to be auseful approach to:
 identify their sex
 evaluate the hypothesis regarding their kinship
 appreciate coherence within di
ff 
erent molecular dataand with morphological and archaeological ones.
Burial conditions i.e. the tomb typology (a hypogeousroom excavated in the rock) as well as the kind of depo-sition (on beds obtained from the tufaceous bank)favoured the preservation of human remains and aDNA.Only few stone fragments, fallen from the roof of thetomb, came in contact with human remains. Furthermore,environmental parameters remained near constant duringthe year. In particular, temperature slightly fluctuatesbetween 13
 (
C and 18
 (
C, while relative humidity isalmost constant at 80% [3]. These conditions limited theaction of the various diagenetic agents, biological or other.
MtDNA from the four individuals was extracted,cloned and sequenced and STR profile was assayed.Molecular sex identification was performed analysingamelogenine and SRY genes [36].DNA studies were preceded by chemical analyses toestimate the amount and the degree of degradation of organic matter preserved in the samples. Gas chroma-tography coupled with mass spectrometry (GC-MS), hasbeen used to evaluate the amino acid racemizationextent. Thermogravimetric analysis (TGA) was alsoperformed to determine the quantity of organic matterpreserved. Despite disagreement among researchers [14],aspartic acid racemization extent is considered a usefulindicator of the possibility to amplify aDNA in thesample [32]. On the other side, TGA evaluates the decrease in the mass of a sample during temperatureincrease in dynamic air atmosphere, making the esti-mation of the quantity of organic matter preserved inthe sample possible. TGA has some good qualities: itrequires few milligrams of sample taken from the samefraction for racemization analysis, it is fast, no elaboratetreatment of the sample is necessary and it is alsoinexpensive and safe because it does not need hazardousreagents. To define the number of copies of mtDNAtemplate, competitive PCR has also been performed.Overall, this work is an attempt to evaluate thepotential of the biomolecular approach in the study of human skeletal remains from archaeological con-texts dated at within the 1st millennium BC from theMediterranean area in view of a broader applicationof these techniques in archaeological investigations asa whole.
2. Materials and methods
The skeletal remains of the occupants of tomb 5859were housed at the University of Pisa (Italy). For each
E. Cappellini et al. / Journal of Archaeological Science 31 (2004) 603–612
 605
 
subject two distinct bone fragments, one (about 5 cm inlength) taken from a rib and one taken from limb bone,were sampled by E.C. and D.C. between 1998 and 2001.The bones had not been washed or treated in any way bythe archaeologists, anthropologists or conservers beforesampling. For subject 5859/16 (4) the root of an upperpremolar was also sampled. During manipulation of sample collection disposable gloves and masks wereused.
 2.1. Precautions against contamination
Due to the poor preservation of the DNA of the boneremains, all the procedures adopted to analyse geneticmaterial from human skeletal remains are prone tonumerous contaminations. Thus, specific precautionswere adopted to handle specimens during DNA extrac-tion and amplification. In order to limit the riskof contamination by exogenous DNA, the criteriaproposed by [16] were followed as closely as possible.DNA extraction, PCR amplification and analysisof the PCR products were performed in separatelaboratory rooms. Specimens were handled (usingdisposable mask, gloves and laboratory coats) in an areawhere no modern DNA studies had previously beenperformed. The DNA extraction and the setting upof PCR reactions of ancient DNA templates werecarried out under two di
ff 
erent hoods (each in adi
ff 
erent room), daily irradiated with UV rays (254 nm).Only disposable sterile tubes, filtered tips, sterilereagents and solutions, exclusively dedicated forancient DNA studies, were used. Di
ff 
erent setsof pipettes were used for DNA extraction, PCRamplification and analysis of the PCR products. Nega-tive controls were included in each set of extraction andamplification.
 2.2. Amino acids racemization analysis
Approximately 5 mg of each sample were dissolved in2 ml of 6N HCl, transferred into a screw-capped tubeunder a nitrogen atmosphere and hydrolysed for 5 h inan oil bath at 100
 (
C. During hydrolysis, proteins weretransformed into amino acids. Hydrolysed samples werespiked with 50 µg of D-L norleucine (10 µl of a 5 g/lsolution) and 5 µg of D-L norvaline (10 µl of a 0.5 g/lsolution) per 1 mg of weighted sample and were com-pletely evaporated. The residue was dissolved in 3 ml of 2N HCl in 2-propanol (Acros, New Jersey, USA), andkept in a screw-capped tube for 1 h at 90
 (
C. After theevaporation of the solvent, the residue was dissolvedin 2 ml of dichloromethane (Baker, HPLC grade,Deventer, Holland) and treated with 0.2 ml of trifluoro-acetic anhydride (Aldrich, Steinheim, Germany) for 1 hat 60
 (
C. After cooling, the tube was opened, and thesolvent evaporated. The derivatization led to the
-trifluoroacetyl-
O
-2-propyl esters. The residue of sample was recovered in 0.2 ml of dichloromethane, andused for GC/MS injection.A standard mixture of amino acids (Sigma, St Louis,MO, USA) in 6N HCl at a concentration of 100 mg/mleach, was prepared and derivatized with the same pro-cedure as above stated and was further used as areference for gas chromatographic retention times andmass spectra of the amino acid derivatives.The analysis of amino acids was carried out by meansof a HP-6890 Series gas chromatograph coupled to anHP-5973 Mass Selective Detector (Hewlett-Packard,Palo Alto, CA, USA). Chromatographic separationswere accomplished using a Chirasil-L-Val (25 m
0.25 mm i.d.
0.12 µm d.f.) fused-silica column(Chrompack Italia, Milan) according to the followingtemperature programme: initial 50
 (
C held for 1 min,then increased to 145
 (
C at 4
 (
C/min, held at 145
 (
C for1 min, then increased to 170
 (
C at 10
 (
C/min and heldat 170
 (
C for 1 min. The injector was kept at 280
 (
C,while helium gas flow was approximately 0.66 ml/min.The injection (1 µl) was split (split ratio of 20:1).MS conditions were as follows: interface temperature190
 (
C, ion source temperature 240
 (
C and electronimpact 70 eV. The mass spectrometer operated in SIM(Selected Ion Monitoring) mode in order to tentativelymeasure the corresponding D/L ratios. The followingtarget ions were selected: m/z 140 for alanine (Ala), m/z168 for norvaline, and norleucine, m/z 184 for asparticacid (Asp), m/z 198 for glutamic acid (Glu).The contribution of the conditions for hydrolysis(HCl 6N, 100
 (
C, 5 h) of the amino acid racemizationwas determined applying the previously described pro-cedure on rat tail collagen and fresh bovine bone,making measurements in triplicate.The D/L Asp values detected were: 0.0062
0.0009 inrat tail collagen and 0.0047
0.0004 in fresh bovinebone. These results suggested that this type of hydrolysisdid not sensibly a
ff 
ect the extent of racemizationdetected in a fossil bone.The low extent of racemization of aspartic acid showshow the condition chosen for sample preparation, inparticular the hydrolytic step, did not artificially increasethe D/L ratio measurements.
 2.3. Thermogravimetric analysis
5 mg of bone powder were subjected to thermogravi-metric analysis using a Perkin-Elmer TGA7. Tempera-ture grew costantly from 50
 (
C to 950
 (
C with a rate of 10
 (
C/min.
 2.4. DNA extraction
All samples were manually brushed, then UV (
=254nm) perpendicularly irradiated for 45 min on each
E. Cappellini et al. / Journal of Archaeological Science 31 (2004) 603–612
606
 
surface in order to eliminate as much contaminant DNAas possible [15]. Soon after, all fragments were manually powdered in a mortar, abundantly washed with bleachand ethanol before the preparation of each sample.0.5–1 g bone powder was mixed for 10 min with 2volumes SSC 1
bu
ff 
er (NaCl 0.15 M and sodiumcitrate 0.015 M, pH 7.2) and 0.05% (w:v) SDS. Themixture was then subjected to a standard phenol-chloroform extraction [35]. The DNA was concentrated with Centrex UF-2 (Schleicher & Schuell) and purifiedby a silica-based [21] procedure using MERmaid
SPIN
e
 Kit (Bio 101 Inc.). At the end of the procedureabout 50 µl of DNA extract were recovered.
 2.5. DNA quantification and “long” amplificatedetection
DNA quantification was performed via a competitivePCR, as described by [20]. A competitor containing a 95 bp deletion (from position 16131 to 16225), in themitochondrial D-Loop region, was used. PCR reagentswere those described for HVR-I, while the primers werethe same used for the second fragment amplification.The amplification thermal profile was: 95
 (
C for 10 minfollowed by 45 cycles (94
 (
C/50 s, 53
 (
C /50 s and72
 (
C/50 s), with a final extension step at 72
 (
C for5 min.Appropriate molecular behaviour was also tested byamplification of longer mtDNA fragments (443 bp and724 bp), which have been reported as very unusual forancient DNA. PCR conditions were those described formtDNA analysis below, primers used for 443 bp frag-ment were L 15995 and H 16401 (see below for completeprimers sequences), while for 724 bp fragment primersused were L 16247 and H 00360 (5
#
-TGG TGT TAGGGT TCT TTG TT-3
#
).
 2.6. Sex identification
For each subject two aDNA samples wereextracted from bone or tooth fragments. On each aDNAsample, three amplifications were performed. As startingmaterial for the subjects 5859/14 (1), 5859/15 (2) and5859/17 (3) two fragments each from a di
ff 
erent bonewere used, while for subject 5859/ 16 (4) a fragment frombone and one from tooth were used.Starting from 2 µl of extracted DNA, two loci wereamplified by PCR using a multiplex approach. The firstintron of the amelogenin homologous gene and a smallportion, 93 bp, of the SRY, sex determining region Ygene [2] were co-amplified with the following thermalprofile: 94
 (
C for 10 min (activation of AmpliTaq Gold,Perkin Elmer), 65 cycles (94
 (
C/1 min, 60
 (
C/1 min and72
 (
C/1 min) and a final extension step at 72
 (
C for5 min. The fastest transitions available between eachtemperature step were applied by a MJ Research PTC100 thermal cycler. The 25 µl reaction mix contained 2 Uof Taq polymerase, 200 µM of each dNTP, 2.5 mMMgCl
2
, 9 pmoles of each primer (XY1: 5
#
-CCC TGGGCT CTG TAA AGA ATA GTG-3
#
; XY2: 5
#
-ATCAGA GCT TAA ACT GGG AAG CTG-3
#
; SRY15
#
-GCA CTT CGC TGC AGA GTA CCG A-3
#
; SRY25
#
-ATA AGT ATC GAC CTC GTC GGA A-3
#
), and2.5 µl of the relative 10
reaction bu
ff 
er. 25 µl of eachPCR product were controlled by electrophoresis, on a5% HR agarose gel with a 2 h run at 50 V, thenvisualized with ethidium bromide.
 2.7. HVR-I mtDNA sequences
For each subject HVR-I mtDNA sequence was deter-minated from two di
ff 
erent aDNA samples extractedfrom two distinct bones. Due to degraded aDNA tem-plate the 360 bp long HVR-I, from position 16024 to16383 according to [1], was subdivided into three distinct overlapping fragments, each less than 200 bp long. Foreach DNA fragment, two amplifications were per-formed. Each amplification product was cloned and twoclones were sequenced. Briefly, the consensus sequenceof each fragment was reconstructed from 8 amplicons,four from the first and four from the second DNAextract. Therefore for each individual the entire HVR-Isequence has been determined by the alignment of 24clones.Starting from 2 µl of extracted DNA HVR-I wasamplified with the following thermal profile: 94
 (
C for10 min, 50 cycles (94
 (
C/45 s, 53
 (
C/1 min and 72
 (
C/1 min) and finally 72
 (
C for 10 min. The 50 µl reactionmix contained 2 U of AmpliTaq Gold (Applied Biosys-tems), 200 µM of each dNTP, 2.5 mM MgCl
2
 and 1 µMof each primer. The following primer pairs were used: L15995 (5
#
-CCA CCA TTA GCA CCC AAA G-3
#
)/H16132 (5
#
-CTA CAG GTG GTC AAG TAT TTA TGGT-3
#
) for the first fragment, L 16107 (5
#
-CGC TAT GTATTT CGT ACA TTA CTG C-3
#
)/H 16261 (5
#
-TGGTAT CCT AGT GGG TGA GG-3
#
), for the secondfragment and L-16247 (5
#
-CAA CTA TCA CAC ATCAAC TGC AA-3
#
)/H 16401 (5
#
-GAT TTC ACG GAGGAT GGT G-3
#
) for the third fragment—all positionnumbers are according to [1]. MtDNA PCR products were cloned using TOPO TA Cloning Kit (Invitrogen)according to the manufacturer’s instructions. Screeningof white recombinant colonies was accomplished byPCR, transferring the colonies into a 40 µl reaction mix(67 mM Tris HCl [pH 8.8], 2 mM MgCl
2
, 1 µM of eachprimer, 0.125 mM of each dNTP, 0.75 units of TaqPolymerase) containing M13 forward and reverse uni-versal primers. After 5 min at 92
 (
C, 30 cycles of PCR(90
 (
C/30 s, 50
 (
C/1 min, 72
 (
C 1 min) were carriedout and clones with insert of the expected size wereidentified by agarose gel electrophoresis. After puri-fication of these PCR products with Microcon PCR
E. Cappellini et al. / Journal of Archaeological Science 31 (2004) 603–612
 607
of 10