Forensic Science International 249 (2015) 2534

Contents lists available at ScienceDirect

Forensic Science International

journal homepage: www.elsevier.com/locate/forsciint

Recent advances in the characterization of hair of mummies from the

Chilean Andean coast
M. Fresnais a,b, P. Richardin b, A. Gimat b, M. Sepulveda c, E. Leize-Wagner a, A. Charrie a,*
a

Laboratoire de spectrometrie de masse des interactions et des syste`mes (LSMIS), 1 rue Blaise Pascal, 67008 Strasbourg, France
Centre de recherche et de restauration des musees de France (C2RMF), Palais du Louvre, Porte des Lions, 14 quai Francois Mitterrand, 75001 Paris, France
c
Laboratorio de Analisis e Investigacion arqueometrico, Departemento de Antropologa, Universidad de Tarapaca, 18 de sept. 2222, Casilla 6D, Arica, Chile
b

A R T I C L E I N F O

A B S T R A C T

Article history:
Available online 14 January 2015

Two pre-Hispanic mummies from the Andean coast, belonging to a corpus of 16 mummies from the San
Miguel de Azapa (Arica, Chile), were radiocarbon dated and analyzed in order to replace them in their
historical context and to study the conservation state of the hair bers and the heavy metal presence. The
radiocarbon dating placed both mummies in the Formative period (1700 years BC to 500 years AD).
Global and elemental analyses were performed using scanning electron microscopy coupled with energy
dispersive X-ray spectroscopy and using X-ray uorescence spectroscopy. These combined techniques
enabled to prove the good global conservation state of the mummies hair and to detect iron, lead,
bromide and also arsenic in some cases, in signicant amounts inside the hair bers. Fourier transformed
infra-red spectroscopy seemed to prove the good conservation state of the hair surface at a structural
level that is why the conservation of hair proteins at a molecular level will be investigated by a
proteomics approach in future work.
2015 Elsevier Ireland Ltd. All rights reserved.

Keywords:
Hair analysis
Andean mummies
SEM-EDS
XRF
FTIR
Qualitative proteomics

1. Introduction
Among human remains from museum collections, mummies
have intrigued researchers for a long time. They are subject to
much research in order to understand the mummication
processes, to improve conservation and restoration protocols,
and to infer fascinating clues about their civilizations. Mummication rituals in the pre-Hispanic Andean culture started
7000 years ago with the Chinchorro civilization in the Camarones
Valley in current Northern Chile, it spread to current Peru and to
the south of the Atacama Desert in Chile and have lasted up to
6000 years [1,2]. Many mummies found in excavation sites during
the last decades have not been dated yet and there is still a lot to
learn about them and about their civilizations.
Studying human remains, like bones, teeth or hair, is of great
importance in archaeology, since it enables a better understanding
of the customs and traditions of ancient civilizations. Bone or tooth
analysis is limited when it concerns the study of human remains

This paper is part of the special issue entitled Proceedings of the SoHT
Bordeaux 2014 meeting, June 11-13, 2014, Bordeaux, France. Guest edited by Dr
Pascal Kintz.
* Corresponding author. Tel.: +33 368851611.
E-mail address: acharrie@unistra.fr (A. Charrie).

http://dx.doi.org/10.1016/j.forsciint.2015.01.005
0379-0738/ 2015 Elsevier Ireland Ltd. All rights reserved.

from museum collections, whereas we can answer many questions

thanks to hair analysis, without compromising the integrity of the
remains. Hair is a very robust material that can be conserved
throughout centuries and it is also a powerful biological indicator
[35]. During its growth, hair integrates the different elements
present in the blood, nutriments as well as toxic elements coming
from the environment, the alimentation or some cosmetic treatments [6,7]. Along the hair bers, we can thus follow the evolution of
an individuals exposure to toxic compounds and update a lot of
information about the environment of the peoples or about some of
their customs, like their alimentation, their religious and cultural
practices or their technical development [3,4,812]. Furthermore,
the molecular ngerprint of the material have been modied by
anthropic or natural alterations, before and during the funeral ritual
or after the burial [13]. This complexity makes the analysis of
archaeological remains a true analytical challenge.
Previous research on hair from mummies of the Arica and
Camarones valley from Archaic to Late Intermediate periods (from
5000 years BC1 to 1400 years AD) have shown high levels of heavy
metals, such as arsenic and lead, using laser ablation inductively
coupled plasma mass spectrometry (LA-ICP-MS) [3,8]. This can be
explained by the high concentrations in heavy metals in the rivers

BC: before Christ; AD: Anno Domini.

26

M. Fresnais et al. / Forensic Science International 249 (2015) 2534

and in the geological formations, due to intense volcanism of this

area. The concentrations in arsenic have probably led to chronic
poisoning and to a high infantile mortality in ancient pre-Hispanic
populations [8,14,15]. To our knowledge, no thorough analysis was
made in order to determine the conservation state of the hair
components at structural and molecular levels. This problematic
can however be essential in order to have a better view of the
preservation state of the bers and to see how the heavy metals
interact with the hair molecules.
In the present work, we used a combination of optical and
physical analysis techniques to have a global and structural
overview of the conservation state of hair from two mummies from
the Andean coast and belonging to a bigger corpus from Archaic to
Inca periods (6500 years BC to 1500 years AD). The coupling with
elemental analyses enabled to study the presence of heavy metals
and non-common elements in the bers. The optimization of the
protocol for proteomics analyses is currently in progress to
investigate the conservation of hair proteins at a molecular level
and the rst steps were applied to our archaeological samples.
2. Materials and methods
2.1. Materials
2.1.1. Chemicals and consumables
Sinapic acid (SA), ammonium bicarbonate (NH4HCO3) and
sodium dodecyl sulfate (SDS) were purchased from Fluka (Buchs,
Switzerland). Iodoacetamide (IAA), DL-dithiothreitol (DTT), sodium
hydroxide (NaOH), Trizma1 hydrochloride (TrisHCl, buffer solution
for proteins), thiourea, sodium deoxycholate (DCO), trichloroacetic
acid (TCA), triuoroacetic acid (TFA), formic acid (FA), dichloromethane, acetonitrile (ACN), acetone and silverwire for radiocarbon
dating were from Sigma Aldrich (St. Louis, MO, USA). Urea was
bought from Acros Organic (Geel, Belgium). Methanol (MeOH) and
propan-2-ol (iPrOH) were supplied by Hipersolv CHROMANORM1
(VWR Chemical Prolabo1, Radnor, Pa, USA). Ethanol was from
AnalaR Normapur1 (VWR Chemical Prolabo1, Radnor, PA, USA).
Sequencing porcine grade trypsin was from Promega (Madison, WI,
USA). Ultrapure water was produced with a water purication
Purelab1 UHQ system (ELGA LabWater VEOLIA Water, Anthony,
France). Cupper oxide CuO was from VWR International (Radnor, Pa,
USA), Quartz paper (glass lter prelters, APFA02500) was from
Millipore (Merck KGaA, Darmstadt, Germany) and dialysis cartridges

(Slide-A-Lyzer 3.5 kDa, 312 mL) were from Thermo Scientic Pierce
(Thermo Fisher Scientic Inc., Waltham, MA, USA).
2.1.2. Archaeological samples
We focused here on two hair samples belonging to a bigger
corpus studied in the frame of a project in collaboration between
the Laboratory of Archaeometric Analyses and Research from the
Department of Anthropology of the Tarapaca University (Arica,
Chile) and the Research and Restoration Center for French
Museums (C2RMF, Paris, France). The mummies of the corpus
were discovered in excavations led in the 1960s and 1970s in the
north of the Atacama Desert, on the coast of Arica and Parinacota
region in Chile. Mummy bundles were found sitting or with hyperbended legs, wrapped in textile mantles and they were deposited
in circular pits dug in the sand, which walls were covered with
plant mats. In general, different kind of offerings as ceramics,
artifacts of everyday life and some plants were disposed near the
mummies [16]. After the excavation, the mummies were brought
to the San Miguel de Azapa Museum (Arica, Chile). For their study,
mummy bundles were open and the heads were separated from
the rest of the body. More particularly, the two heads of mummies
which were studied in this work, PLM7_Cr3c and PLM7_T305
(Fig. 1), were found in the archaeological site Playa Miller 7 (Arica
valley) by Guillermo Focacci in 1974 [16]. According to previous
dates and observations from the same site, archaeologists
postulate that this site belonged to the Formative period
(1700 years BC500 years AD). The rst challenge of this study
was thus to replace the two mummies in their archaeological
context.
2.1.3. Modern samples
Modern hair samples were used as standards for different
analyses. They were taken from one European woman, who had
been exposed to chemical hair bleaching, for the physical and
chemical analyses, and from one European man, who had not been
exposed to any cosmetic or chemical treatment, in order to take the
inter-individual variability into account.
2.2. Hair treatments
2.2.1. Pre-wash cleaning
After a manual selection under the binocular microscope to
remove all solid contaminants, between 15 and 30 mg of hair were

Fig. 1. Photographs of heads of mummies PLM7_Cr3c (A) and PLM7_T305 (B) from the excavation site Playa Miller 7 in the Atacama Desert, in Arica and Parinacota region
(Chile).

M. Fresnais et al. / Forensic Science International 249 (2015) 2534

immersed in ultrapure water and placed in an ultrasonic bath until

all the remaining solids and mineral surface contaminations were
removed. Hair samples were then washed with a mixture of
CH2Cl2/MeOH 1:1 (v/v) to remove organic coating materials,
decomposition uids, and eventual varnishes and synthetic resins.
After this treatment, the samples were thoroughly rinsed once
with ultrapure water and acetone and three times with ultrapure
water.
2.2.2. Acid/alkali/acid wash treatment
The acid/alkali/acid method was applied to the pre-washed
hair. This protocol is mainly used for vegetal samples and can
remove all the contaminants or degraded amino acids [12]. This
consists of a series of extractions or washes at 60 8C for 1 h,
alternating with rinses with ultrapure water. The rst washing was
done with 5 mL of 0.5 M hydrochloric acid solution, followed by
5 mL of a fresh NaOH 0.0005 M aqueous solution and lastly, a new
acidic washing similar to the rst one.
2.2.3. Extraction and precipitation of hair proteins for radiocarbon
dating (Protocol 1)
Cortical proteins were extracted following the protocol
described by Richardin and coworkers (Fig. 2) [11,12]. This method
is able to destroy the disulde bonds in order to dissolve only the
cortical proteins. 5 mL of the extraction solution containing 0.13 M
DTT, 0.03 M SDS and 25 mM TrisHCl was used. SDS is an anionic
surfactant whose function is to denature the protein. It therefore
allows to partially or completely unfold the polypeptide chain.
Samples were immersed in the extraction solution without
agitation for 2 days at 50 8C. Under these conditions, the only
residue was the cuticle emptied of the cortex and medulla. After
extraction, the solution was ltered on quartz paper to eliminate
the solid waste, especially the cuticle. Then, the proteins were
puried by precipitation with 6 mL of a 2% DCO solution and
600 mL of a 100% TCA solution. The proteins then formed a
precipitate, also called ball. After centrifugation, the pellets were
puried with 10 mL of acetone at 0 8C. Finally, the precipitates
were abundantly rinsed with ultrapure water so that there was no
trace of acetone left, which could distort dates with the
contribution of dead carbon. The pad was then dried overnight
in an oven under partial vacuum.

27

2.2.4. Extraction of hair proteins for proteomics analyses (Protocol 2)

The extraction protocol followed the experimental procedure
given by Lee et al. and Barthelemy (Fig. 2) [17,18]. After washing
and degreasing, hair samples were immersed in 5 mL of the
extraction solution containing 7 M urea, 2 M thiourea, 50 mM DTT,
50 mM TrisHCl (pH 7.5) and 0.1% Triton X-100. Hair was
incubated without agitation for 18 h at 37 8C. After decantation,
the soluble and insoluble materials were separated and the protein
extract (soluble sample) was alkylated with 1 M iodoacetamide
(IAA) and 3 M TrisHCl for 10 min in the dark at room temperature.
The alkylated extract was then dialyzed in dialysis cartridges
(3.5 kDa, 312 mL) against water, with two changes of water after
2 h and then over 2 days. Finally, the extract was lyophilized
overnight and stored at 20 8C.
2.3. Analytical methods
2.3.1. Radiocarbon dating
Combustion The extracted proteins were combusted at high
temperature (5 h at 850 8C) under high vacuum (at 10 6 Torr) on a
semiautomatic combustion bench. 22.5 mg of the processed
samples was combusted in a quartz tube with 500 mg CuO (for
analysis) and a piece of silver wire (99.95%). The combustion gases
were separated, and the CO2 was collected in a sealed tube.
Graphitization and radiocarbon measurement The graphitization and all the measurements were achieved with the Artemis
AMS facility of Saclay, France [19]. The graphitization of the
obtained CO2 was realized by direct catalytic reduction with
hydrogen, using iron powder as catalyst at 600 8C and with an
excess of H2. During this process, the carbon was deposited on the
iron and the whole powder was mechanically pressed into a at
pellet. The radiocarbon activity was calculated by comparing the
measured intensities of the 14C, 13C and 12C beams from each
sample with those of CO2 standards prepared with HOx (I) oxalic
acid reference. It was then expressed as percent Modern carbon,
normalized with a d13C at 25 %. The radiocarbon ages were
calculated after correction of the isotope fractionation d13C
measured by AMS. The calendar ages were determined using
the software OxCal v4.2.3 [20,21] and the SHCal13 (Southern
Hemisphere radiocarbon calibration curve) calibration data set
[22]. The calculation was done online [23]. Calibrated radiocarbon

Fig. 2. Schematic description of the two protocols tested for hair protein extraction.
Protocol 1 was adapted from protocol for extraction and purication of hair keratin for AMS radiocarbon measurements Richardin et al. [11,12] and Protocol 2 was adapted
from Lee et al. [18] and Barthelemy [17].

28

M. Fresnais et al. / Forensic Science International 249 (2015) 2534

dates were presented using the 2-sigma values which account for
95.4% of the probability of the date falling within that particular
range, and dates were expressed in years cal BC or cal AD (before
Christ or Anno Domini).

different samples. Bruker Pepmix I and II were used for calibration

and spectra were processed with the Bruker FlexAnalysis software
(version 3.4).
3. Results and discussion

2.3.2. SEM-EDS analyses

Because they are not conductive materials, the selected hair
bers were metalized with carbon or with a Pt/Pd mixture and
placed on small adhesive carriers. Analyses were performed with
two microscopes (SEM Philips XL 30 CP for the EDS analyses and
FEG-SEM JEOL 7800F for the morphological observations). Elemental analyses of hair surface were realized with a detector for energy
dispersive X-ray spectroscopy (EDS, Oxford Instruments Silicon Drift
Detector X-MAX 50). Secondary electron (SE) mode was used for
morphological analyses (acceleration voltage 5 keV, work distance
4 mm), while the backscattered electron (BSE) mode was used in
complement of the EDS detector for the elemental analyses
(acceleration voltage 20 keV, beam size 5.3 mm, work distance
10 mm). In these operating conditions, the analyzed depth was
around 1 mm. All data were processed with the Aztec software.
2.3.3. XRF spectroscopy
X-ray uorescence analyses were carried out on hair strands
placed into paperboard carriers made for the study. The spectrometer was equipped with a Moxtek molybdenum anode X-ray tube
featured with a 50 mm-thick cobalt absorber, decreasing the
background in the area of interest (between 7.5 and 15 keV). The
detector was a Rontec ash detector (XFlash 2000A) with a 8 mm
beryllium window and the operating conditions of the tube were
40 kV and 800 mA. The distance X-ray tube-sample was about
1 cm, the beam impact angle 458 and the detection angle 908. The
depth of analysis was around 50100 mm, which is the order of the
diameter of hair. This technique allowed to achieve the global
elemental composition of hair bers and so not only of the hair
surface. Spectra were collected via a multichannel analyzer (Amtek
MCA8000A), using an acquisition time of 1200s and processed with
the dedicated PyMCA software (version 4.3.0). Using PyMCA and
comparison with standard targets, we were able to have a rst idea
of the amounts of the detected elements in ppm. Since the analysis
was performed in air, lights elements such as potassium,
aluminum, silicon, phosphorus and sulfur were detected only if
present in high amounts in the analyzed hair strand.

3.1. Radiocarbon dating

The radiocarbon ages and calibrated calendar age ranges of hair
samples from PLM7_T305 and PLM7_Cr3c show a good reproducibility of the results (Fig. 3). Radiocarbon ages are 2590  30 and
2575  30 years BP (before present) for PLM7_T305, and 2645  30,
2650  30 and 2685  30 years BP for PLM7_Cr3c. Radiocarbon and
calibrated dates are shown in Fig. 3. These values seem to place both
mummies in the Formative period (from 2950 years BP to 1900 years
BP), which is consistent with the archaeologists assumption, which
were based on the rst dating results and observations from the Playa
Miller 7 excavation site.
3.2. Morphological analyses
First observations of the two untreated samples with the
binocular microscope showed the hair is brittle in both cases and
has a brown (PLM7_T305) to light brown (PLM7_Cr3c) color
(Fig. 4A and E). Many vegetal and mineral residues, probably
coming from the soil in which the mummies were buried, are also
clearly visible. Indeed, in the pits where the mummy bundles were
deposited, different kinds of plants, sand and textile bers were
found. In the case of PLM7_T305, it seems there is a red coloration
on the hair surface.
As macroscopic observations, SEM observations of the untreated hair show an important quantity of residues, deposits and
crystallized materials, which almost completely hide the hair
scales (Fig. 4B, C and F). On the images taken after the cleaning and
solvent extraction, we can clearly see the effectiveness of the
treatment since almost all observed residues and deposits are no
longer apparent (Fig. 4D and H). Globally, the cuticle of PLM7_Cr3c
seems to have a better conservation state. The hair scales are
visible and well-preserved on some areas, but they sometimes

2.3.4. FTIR spectroscopy

Infrared measurements were recorded using a Spectrum
2000 FTIR spectrometer from PerkinElmer, with a Golden GateTM
ATR accessory (Specac) featured with a single reection monolithic
diamond attenuated total reectance system, which allows the
direct analysis of hair without any sample preparation. Spectra
(4000400 cm 1, spectral resolution 4 cm 1, 150 scans) were
processed with OMNIC software (Perkin-Elmer).
2.3.5. MALDI-MS
The matrix assisted laser desorption-ionization mass spectrometry (MALDI-MS) analyses were conducted on a Bruker
Autoex II MALDI Time of Flight (TOF) mass spectrometer
(Bruker Daltonics, Bremen, Germany). The device was featured
with a pulsed nitrogen laser emitting at 337 nm and operated at an
extraction voltage of 20 kV. The samples were resuspended in the
extraction buffer (25 or 50 mM TrisHCl) and deposited on a
stainless steel target as dry droplets with a saturated solution of
sinapic acid (SA), in ACN/H20 0.1% TFA 1:1 (v/v). Gated suppression
was applied in order to prevent any saturation of the detector by
matrix ions. 1000 laser shots were averaged for each spectrum
and acquisitions were realized in reector and linear modes, with a
laser power optimized for each mode and kept constant for the

Fig. 3. Calibrated 14C ages for samples from PLM7_Cr3c (506) and from PLM7_T305
(511).

M. Fresnais et al. / Forensic Science International 249 (2015) 2534

29

Fig. 4. Macrophotographs of hair from PLM7_Cr3c (A) and PLM7_T305 (E). FEG-SEM images of untreated hair from PLM7_Cr3c (B and C) and SEM image of treated (UP water,
CH2Cl2/MeOH, acetone) hair from PLM7_Cr3c (D). SEM images of untreated and treated (UP water, CH2Cl2/MeOH, acetone) hair from PLM7_T305 (respectively F and H) and
EDS iron map of untreated hair from PLM7_T305 (G).

seem loosened or irregular and have even been removed on some

parts of the hair surface. On the other hand, no scales of the hair
sample from PLM7_T305 are visible and the hair surface is clearly
altered.
3.3. Elemental analyses
Elemental analyses with SEM-EDS establish both hair bers
from PLM7_T305 and PLM7_Cr3c have a similar elemental
composition on the surface (Fig. 5A and B). In particular, we nd
some of the major elements of hair (e.g. carbon, oxygen, nitrogen,
sulfur, and calcium) [10] for untreated and treated samples, and
also some natrium, magnesium, aluminum, silicon, phosphorus,
chloride and copper on the surface of the untreated bers only. EDS
mapping of these last elements proves the deposits and the
residues on untreated hair surface are salts and silicates probably
coming from the soil in which the mummies were buried. XRF
analyses (Fig. 6) lead to the detection of sulfur, calcium, copper and
zinc in untreated modern hair, as well as in untreated and treated
ancient hair from both PLM7_T305 and PLM7_Cr3c in relatively
similar proportions. The presence of these four elements in our
samples is then mostly due to the classic elemental composition of
human hair [24] with a small contribution of burial soil in the case
of copper, as shown with SEM-EDS analyses.
Some lead and strontium are also detected in ancient hair strands
before any treatment, but not after the cleaning steps, ascribing

them to an exogenous source. On the contrary, manganese is

detected in both untreated and treated hair from PLM7_Cr3c, and
manganese and nickel presence is also visible in treated bers, after
the elimination of a great part of iron on the hair from PLM7_T305.
The presence of these elements after the cleaning steps could thus
come from alimentation and drinking water.
XRF analyses highlight the presence of a relevant amount of
bromide in both untreated and treated hair samples (more than
1000 ppm against less than 500 ppm in typical modern hair) [10].
Pre-Hispanic Andean civilizations were settled on coastal areas.
Taking into account the great part of marine food in alimentation of
coastal civilizations since Archaic period [25], this observation is
not surprising. Indeed bromide, present in sea water, accumulates
in the organism of shes and sea food and, by extension, in that of
their predators (e.g. humans). Constant contact with sea water
might also have favored the accumulation of bromide in the
organism of these peoples.
With both SEM-EDS and XRF techniques, iron is detected in
the two untreated and treated ancient hair samples with a lower
amount of iron in treated hair, highlighting the contribution of an
exogenous iron source. For the mummy PLM7_Cr3c, the residual
iron amount in the treated bers (1100 ppm) is much higher than
that of modern hair (less than 50 ppm) [24], which is probably
due to the integration of iron in the hair bers through
alimentation or drinking water. The case of the mummy
PLM7_T305 is quite different. EDS analysis and XRF spectroscopy

Fig. 5. EDS spectra of modern hair and hair from PLM7_T305 and PLM7_Cr3c before (A) and after (B) the cleaning steps (UP water, CH2Cl2/MeOH, acetone). 10 nm-thin carbon
coating, acceleration voltage: 20 keV, work distance: 10 mm, spot size: 5.3 mm.

30

M. Fresnais et al. / Forensic Science International 249 (2015) 2534

Fig. 6. XRF spectra of modern hair and hair from PLM7_T305 and PLM7_Cr3c before and after the cleaning steps (UP water, CH2Cl2/MeOH, acetone). Red mark: Ar Ka ray Mo
anode X-ray tube, operating conditions: 40 kV and I = 800 mA, 50 mm-thick Co absorber (7.515 keV), acquisition time: 20 min.

highlight a surprisingly important amount of iron on the untreated

hair surface (almost 3.5%), while it is much lower in the treated bers
(around 100 ppm). Moreover, macrophotographs of sample, taken
previously to any treatment, showed the presence of a red coloration
and EDS map of iron revealed that this element is quite
homogeneously distributed on the hair surface (Fig. 4G). The
combination of these observations allows to assume that the red
coloration is due to the use of hematite (iron based red pigment) as a
colorant. The residual amount is still higher than the typical value,
which could be explained by the integration of iron either from the
red pigment through the cuticle or through alimentation or drinking
water. The use of hematite is also common in funerary practices in
Andean region. It can be used as painting directly on the body or on
hair, or to cover the mummy bundle and has been identied since
Archaic period [26].

Arsenic is also detected in the hair from PLM7_Cr3c, before

and after treatment. This could be expected since Chile has an
important volcanic activity and heavy metals are present in high
concentrations in river water, soil and geological formations (e.g.
mercury, lead and more specically arsenic). High arsenic
exposure via drinking water or plants using for food or textiles
was then very likely and previous studies have shown the
probability of a chronic arsenic poisoning of pre-Hispanic
Andean peoples [3,8,14,15]. The quantity of arsenic detected in
treated hair from PLM7_Cr3c is between 100 and 150 ppm, which
seems to conrm the hypothesis of arsenicism (arsenic concentrations in hair above 110 ppm are considered chronic
poisoning [8]). In the case of PLM7_T305, no arsenic was detected
by SEM-EDS or XRF analyses. This could show the absence of
arsenic, or at least, show that the amount of arsenic integrated

M. Fresnais et al. / Forensic Science International 249 (2015) 2534

into the ber is lower than the detection limit (about 10 ppm).
Further analyses of other mummies from the same area should be
performed in order to have a better vision of the arsenic exposure
of these peoples.
3.4. Structural analyses
Attenuated Total Reectance Fourier Transformed Infrared
(ATR-FTIR) spectra (Fig. 7.) taken from the two archaeological
samples are very similar and are typical of hair, with the presence
of the absorption bands of lipids, amides and cystine residues.
Spectra of untreated and treated hair exhibit no great difference,
although absorption regions of the CH2 and CH3 stretching and
deformation (30002800 cm 1 and 15001300 cm 1) have smaller intensity and area for both PLM7_T305 and PLM7_Cr3c,
highlighting the elimination of surface lipids after solvent
extraction [10]. One peak at 1316 cm 1 is also no longer visible
after the treatment, suggesting the elimination of a non-identied
compound. Spectra show the characteristic absorption bands of
keratin proteins: (i) CH stretching bands of amides A and B at
about 3300 and 3070 cm 1, (ii) n(C5
5O) and n(CN) stretching, and
d(NH) and d(O5
5C5
5N) deformation bands of amides I, II and III at
about 1630, 1520 and 1230 cm 1, (iii) absorption bands due to
cystine residues at 1121, 1070 and 1040 cm 1, according to the
oxidation state of the cystine [10,27,28].

31

ATR-FTIR is surface specic and the depth of analysis is

known to be less than 1.5 mm at 1600 cm 1 for keratinized bers.
This technique allows to have an overview of the structural
conservation state of hair cuticle by studying the amide I and II
bands and the cystine residue bands. For our samples, the
position of amide I band (1633 cm 1 for PLM7_T305 and
1631 cm 1 for PLM7_Cr3c) is consistent with a well-preserved
cuticle (absorption band around 1630 cm 1 with a shift by
about 10 cm 1 to higher values for decuticulated hair) [27,28].
The intensity ratio of the amide I to amide II band is also an
indicator of the good preservation state of the cuticle: a ratio
around 1.3 would be consistent with well-preserved hair
surface while a much higher ratio of about 12.5 would
highlight the absence of the cuticle [27,28]. In our study, the ratio
for the two treated hair samples is around 1.1, which seems to
conrm the good conservation state of the cuticle. The sulfonate
vibrational region reveals the presence of partially oxidized
cystine derivatives (e.g. cystine monoxide at 1075 cm 1) before
any treatment and also the presence of cystic acid before and
after the cleaning (sulfonate group vibration at 1040 cm 1) [27].
This oxidation of cystine is probably due to the aging of the
mummies and the inuence of environmental conditions when
they were buried.
The good global conservation state of the ancient hair surface is
also proven by comparing spectra of modern hair with spectra of

Fig. 7. ATR-FTIR spectra of modern hair and of hair from PLM7_Cr3c before and after the cleaning steps (UP water, CH2Cl2/MeOH, acetone).

M. Fresnais et al. / Forensic Science International 249 (2015) 2534

32

Fig. 8. Mass spectra of modern hair proteins extracted with Protocol 1 (C-D) and Protocol 2 (A-B). Dry droplet deposit; Matrix: SA saturated in ACN/H20 1:1 (v/v); Positive
ionization mode: Detection: linear mode (A and C) and linear mode with COVALX (detection system for high-mass compounds) (B and D).

hair from PLM7_Cr3c and PLM7_T305 since the obtained proles

are clearly similar.
3.5. Proteomics approach
Proteomics analysis of modern hair is well described in
literature [17,18,29] and has led to the molecular characterization

of cortical proteins (mainly keratins and proteins associated to

keratins, also called KAP). Our proteomics approach for the
characterization of hair in archaeological context will, of course,
be based on these studies but will also use specic protocols,
adapted to be robust, reproducible and applicable on small
amounts of samples. The aim is not to specically characterize
keratins and KAP but to see if these proteins are still present and

Fig. 9. Photographs of blank test (A), modern hair (B) and archaeological hair (C) after protein extraction with Protocol 2.

M. Fresnais et al. / Forensic Science International 249 (2015) 2534

33

Fig. 10. Mass spectra of modern hair proteins (A and B) and ancient hair proteins (C and D) extracted with Protocol 2. Matrix: SA 10 mg/mL in ACN/H20 1:1 (v/v); Positive
ionization mode, Detection: linear mode (A and C) and linear mode with COVALX (detection system for high-mass compounds) (B and D).

also to study the alteration products, the post-translational

modications (PTMs) and the possible interactions with exogenous compounds integrated in the ber, like heavy metals for
instance. These aspects are difcult and challenging since there are
poor agreement with existing databases for modern hair and no
database for hair proteins in archaeological context.
Two protocols were selected. Protocol 1 was adapted from the
protocol for cleaning, extraction and purication of hair proteins
for AMS measurements [11,12], and Protocol 2 was adapted from
Lee et al. and Barthelemy [17,18] (Fig. 2). To preserve the
archaeological samples, the optimization of extraction protocol
and of MALDI-TOF parameters was done on modern hair. Both
protocols require the same time for the whole process, can be
performed on the same amounts of hair and are well reproducible
for modern hair. Protocol 2 is however easier to execute because it
requires no precipitation step, step which can be difcult for low
sample quantities. Protocol 2 also gives higher extraction rates
(ratio of nal mass of extracted proteins and initial mass of raw
sample) than Protocol 1. MALDI spectra performed before the
freeze-drying of intact proteins reveal a clear keratin prole
(keratins and KAPs have masses in the range 600060000 Da
[17,30]) for Protocol 2 but not for Protocol 1 (Fig. 8), which is
probably due to the high concentrations in salts used for the
extraction and precipitation, and the lack of a dialysis step. After
proceeding the extraction step on our two archaeological samples,
rst observations showed that hair was almost completely
dissolved in the extraction solution, while there was a substantial
insoluble material in the case of modern hair (Fig. 9). Preliminary
MALDI-MS results show that the two archaeological samples have
a similar signature that substantially differs from the one of
modern hair. Modern hair displays a clear protein prole. On the
contrary, ancient hair has no dened prole and no intact protein
was identied (Fig. 10). This seems to show that the molecular
interactions which ensure protein cohesion inside the ber have
been altered making the ancient keratin much less robust. These
observations seem to question the conclusions of FTIR analyses and

show the need of a future work on molecular analysis of mummies

hair.
4. Conclusion
This work enables to better situate the two mummy heads from
the Andean coast in their historical and environmental context.
Radiocarbon results were consistent with the rst hypotheses of
the archaeologists and placed the two individuals in the Formative
period (1700 years BC500 years AD).
Arsenic was detected in only one of the two samples at a level of
about 100 ppm. This seems to suggest that some individuals
suffered from chronic arsenic intoxication, which is consistent
with the environmental conditions and the hypothesis of high
arsenic exposure. Other heavy metals, like iron, manganese or lead,
were also detected in both samples. In particular, the detection of a
very high amount of iron and the iron-map using SEM-EDS on
untreated hair from PLM7_T305 help to hypothesize the use of an
iron-based red pigment, probably the hematite, on this mummy
head. Furthermore, important amounts of bromide were detected
for the two hair samples. These peoples lived on the Pacic coast,
the marine inuence was then of great inuence in their
alimentation and everyday life.
FTIR spectroscopy seemed to show a good global conservation
state of hair cuticle at a structural level for both samples, but this
has to be further tested with the study of the conservation state of
hair cuticle and hair cortex at a molecular level. This will be
performed thanks to a proteomics approach that is currently
in progress and could also help to visualize the interactions
between heavy metals and keratin molecules. In order to do
this, it was crucial to develop, as a rst step, the extraction
protocols of hair protein. Two approaches were tested but one,
dedicated to proteomics analysis of hair, seem to provide
better results according to the rst MALDI-MS analyses of
intact proteins extracted from modern hair. This protocol was
performed on a rst archaeological sample and rst observations

34

M. Fresnais et al. / Forensic Science International 249 (2015) 2534

suggest that molecular interactions between hair proteins

are altered.
This rst part of the study brought new relevant information
about the two mummies and enabled to nd new evidences
consistent with the chronological and environmental hypotheses.
The rst step of the proteomics approach gave very promising
results and it now needs to be continued with the optimization of
nanoLC and ESI-MS parameters and nally, with the analysis of the
whole corpus.
Acknowledgments
The authors would like to thank the team of the Artemis AMS
facility in Saclay for the radiocarbon measurements, Thomas
Calligaro and Eric Laval for the help with the elemental analyses
and the laboratory of bio-electrochemistry and spectroscopy for
lending us their freeze-drying device. Project was nanced by the
French MESR (Ministe`re de lEnseignement Superieur et de la
Recherche) for the Ph.D. thesis scholarship, and by Proyecto UTA
o UTA2014. 374114. We also thank Convenio de Desempen
MINEDUC. We acknowledge the SFTA (Societe Francaise de
Toxicologie Analytique) for the grant they offered us to attend
the Analytical, Clinical and Forensic Toxicology Meeting in
Bordeaux in June 2014. Finally, the authors thank the editor and
the three reviewers for their constructive comments and suggestions which have greatly contributed to improve this paper.
References

[10]

[11]

[12]

[13]

[14]
[15]

[16]
[17]

[18]

[19]
[20]

[21]
[22]

[1] B. Arriaza, V. Standen, V. Cassman, C.M. Santoro, Chinchorro culture pioneers of

the coast of the Atacama Desert, in: H. Silverman, W.H. Isbell (Eds.), Handb. South
Am. Archaeol., Springer, New York, 2008, pp. 4558.
[2] P.A. Marquet, C.M. Santoro, C. Latorre, V.G. Standen, S.R. Abades, M.M. Rivadeneira, et al., Emergence of social complexity among coastal hunter-gatherers in
the Atacama Desert of northern Chile, Proc. Natl. Acad. Sci. 109 (2012) 14754
14760.
[3] S. Byrne, D. Amarasiriwardena, B. Bandak, L. Bartkus, J. Kane, J. Jones, et al., Were
Chinchorros exposed to arsenic? Arsenic determination in Chinchorro mummies
hair by laser ablation inductively coupled plasma-mass spectrometry (LA-ICPMS), Microchem. J. 94 (2010) 2835.
[4] I. Kakoulli, S.V. Prikhodko, C. Fischer, M. Cilluffo, M. Uribe, H.A. Bechtel, et al.,
Distribution and chemical speciation of arsenic in ancient human hair using
synchrotron radiation, Anal. Chem. 86 (2014) 521526.
ez, A rapid approach for assessment of
[5] S. Steely, D. Amarasiriwardena, J. Jones, J. Yan
arsenic exposure by elemental analysis of single strand of hair using laser ablationinductively coupled plasma-mass spectrometry, Microchem. J. 86 (2007) 235240.
[6] I. Nicolis, E. Curis, P. Deschamps, S. Benazeth, Arsenite medicinal use, metabolism, pharmacokinetics and monitoring in human hair, Biochimie 91 (2009)
12601267.
[7] H.C. Hopps, The biologic bases for using hair and nail for analyses of trace
elements, Sci. Total Environ. 7 (1977) 7189.
[8] B. Arriaza, D. Amarasiriwardena, L. Cornejo, V. Standen, S. Byrne, L. Bartkus, et al.,
Exploring chronic arsenic poisoning in pre-Columbian Chilean mummies, J.
Archaeol. Sci. 37 (2010) 12741278.
[9] L. Bartkus, D. Amarasiriwardena, B. Arriaza, D. Bellis, J. Yanez, Exploring lead
exposure in ancient Chilean mummies using a single strand of hair by laser

[23]
[24]
[25]

[26]

[27]
[28]

[29]

[30]

ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS), Microchem. J. 98 (2011) 267274.

M. Cotte, Developpement de la microscopie infrarouge avec rayonnement synchrotron pour letude de pratiques pharmaceutiques anciennes Exemples des
emplatres de plomb et des tissus humains, (The`se de doctorat), Universite Paris
6, 2004.
P. Richardin, M. Coudert, N. Gandolfo, J. Vincent, Radiocarbon dating of mummied human remains, Application to a series of Coptic mummies from the Louvre
Museum, Radiocarbon 55 (2013) 345352.
P. Richardin, N. Gandolfo, P. Carminati, P. Walter, A new protocol for radiocarbon
dating of hair and keratin type samples application to an Andean mummy from
the National Museum of Natural History in Paris, Archaeol. Anthropol. Sci. 3
(2011) 379384.
A. Charrie-Duhaut, J. Connan, N. Rouquette, P. Adam, C. Barbotin, M.-F. de
Rozie`res, et al., The canopic jars of Rameses II: real use revealed by molecular
study of organic residues, J. Archaeol. Sci. 34 (2007) 957967.
C. Boston, B. Arriaza, Arseniasis and teratogenic anomalies in the Atacama Desert
coast of ancient Chile, Interciencia 34 (2009) 338343.
B. Arriaza, Arseniasis as an environmental hypothetical explanation for the origin
of the oldest articial mummication practice in the world, Chungara Rev.
Antropol. Chil. 37 (2005) 255260.
G. Focacci, Excavaciones en el cementerio Playa Miller 7, Arica (Chile), Chungara
Rev. Antropol. Chil. 3 (1974) 2374.
N. Barthelemy, Proteomique qualitative et quantitative, une passerelle pour relier
lexpression genomique a` la construction des edices biologiques Application a`
la comprehension de la structure moleculaire du cheveu humain, (The`se de
doctorat), Universite de Strasbourg, 2011.
Y.J. Lee, R.H. Rice, Y.M. Lee, Proteome Analysis of Human Hair Shaft From Protein
Identication to Posttranslational Modication, Mol. Cell. Proteomics 5 (2006)
789800.
E. Cottereau, M. Arnold, C. Moreau, D. Baque, D. Bavay, I. Caffy, et al., Artemis, the
new 14C AMS at LMC14 in Saclay, France, Radiocarbon 49 (2007) 291299.
C.B. Ramsey, M. Scott, H. van der Plicht, Calibration for archaeological and
environmental terrestrial samples in the time range 2650 ka cal BP, Radiocarbon
55 (2013) 20292034.
C.B. Ramsey, S. Lee, Recent and planned developments of the program OxCal,
Radiocarbon 55 (2013) 720730.
A.G. Hogg, Q. Hua, P.G. Blackwell, M. Niu, C.E. Buck, T.P. Guilderson, et al., SHCal13
southern hemisphere calibration, 050,000 years cal BP, Radiocarbon 55 (2013)
18891903.
Radiocarbon Calibration, 2013 http://c14.arch.ox.ac.uk/oxcal/OxCal.html.
L. Bertrand, Approche structurale et bioinorganique de la conservation de bres
keratinisees archeologiques, (The`se de doctorat), Universite Paris 6, 2002.
A. Roberts, F. Donald Pate, B. Petruzzelli, C. Carter, M.C. Westaway, C.M. Santoro,
et al., Retention of huntergatherer economies among maritime foragers from
Caleta Vitor, northern Chile, during the late Holocene: evidence from stable
carbon and nitrogen isotopic analysis of skeletal remains, J. Archaeol. Sci. 40
(2013) 23602372.
M. Sepulveda, H. Rousselie`re, E. Van Elslande, B. Arriaza, V. Standen, C.M. Santoro,
et al., Study of color pigments associated to Archaic Chinchorro mummies and
grave goods in Northern Chile (70003500 BP), J. Herit. Sci. 2 (2014), http://www.
biomedcentral.com/content/pdf/2050-7445-2-7.pdf (accessed 27.05.14).
H. Zhang, K.R. Millington, X. Wang, A morphology-related study on photodegradation of protein bres, J. Photochem. Photobiol. B 92 (2008) 135143.
E. Wojciechowska, M. Rom, A. Wochowicz, M. Wysocki, A. Weseucha-Birczynska, The use of Fourier transform-infrared (FTIR) and Raman spectroscopy
(FTR) for the investigation of structural changes in wool bre keratin after
enzymatic treatment, J. Mol. Struct. 704 (2004) 315321.
N.R. Barthelemy, A. Bednarczyk, C. Schaeffer-Reiss, D. Jullien, A. Van Dorsselaer, N.
Cavusoglu, Proteomic tools for the investigation of human hair structural proteins
and evidence of weakness sites on hair keratin coil segments, Anal. Biochem. 421
(2012) 4355.
I. Szeverenyi, A.J. Cassidy, C.W. Chung, B.T.K. Lee, J.E.A. Common, S.C. Ogg, et al.,
The human intermediate lament database: comprehensive information on a
gene family involved in many human diseases, Hum. Mutat. 29 (2008) 351360.