Content uploaded by William Jeffrey Hurst
Author content
All content in this area was uploaded by William Jeffrey Hurst on Sep 23, 2014
Content may be subject to copyright.
T
he Maya archaeological site at Colha in
northern Belize, Central America, has
yielded several spouted ceramic vessels
that contain residues from the preparation
of food and beverages. Here we analyse dry
residue samples by using high-performance
liquid chromatography coupled to atmos-
pheric-pressure chemical-ionization mass
spectrometry, and show that chocolate
(Theobroma cacao) was consumed by the
Preclassic Maya as early as 600
BC, pushing
back the earliest chemical evidence of cacao
use by some 1,000 years. Our application of
this new and highly sensitive analytical
technique could be extended to the identifi-
cation of other ancient foods and beverages.
The site at Colha is known for its spe-
cialized production of lithic tools
1
and for
its collection of intact, spouted vessels
2
,
which were manufactured only during the
Preclassic period (900
BC to AD 250)
3
. These
have generally been recovered from burial
sites associated with elite individuals and
are relatively rare, typically being found
with other serving vessels such as bowls,
dishes and plates, and were probably used
to dispense liquid from the spout, in much
the same way as a teapot.
Based on epigraphic analysis of vessels
dating from the Classic (
AD 250–900) period
and documents written at the time of the
Spanish Conquest, liquid chocolate was
frothed to produce a foam — considered by
the Maya and the Aztecs to be the most
desirable part of the drink
4,5
— by pouring
the liquid from one vessel into another
4
. In
the earlier Preclassic spouted jars (Fig. 1),
frothing would also have been accomplished
by introducing air through the spout,
enabling the vessel to be used for preparing,
as well as pouring, liquid chocolate.
At the time of the Spanish Conquest,
chocolate was consumed with most meals
and was usually mixed with another ingre-
dient (for example, water, maize, chilli
and/or honey) and in different proportions
to produce a variety of drinks
5
. Our aim,
however, was to confirm the existence of
cacao residues in spouted vessels, rather
than to investigate the presence of other
components.
We used high-performance liquid chro-
matography (HPLC) coupled to atmos-
pheric-pressure chemical-ionization mass
spectrometry (APCI MS) to analyse each of
the samples collected from the 14 vessels
recovered from a series of burials at Colha
6,7
.
All of these vessels date to between 600
BC
and AD 250 (ref. 2). The procedure involved
withdrawing about 500 mg from each sam-
ple vial, and adding 3 ml distilled water at 80
7C to solubilize the materials. Before analy-
sis, we passed each sample through mem-
brane filters to eliminate particulate matter.
Cacao has a unique chemical composi-
tion of over 500 different compounds,
including members of the methylxanthine
class (primarily theobromine, with a lower
concentration of caffeine). As T. cacao is the
only Mesoamerican plant that contains
theobromine as the primary methylxan-
thine
8
, this compound can be used as a
marker for the presence of cacao. For exam-
ple, HPLC coupled to thermospray mass
spectroscopy has revealed cacao residues in
ceramic vessels found in an Early Classic
(
AD 460–480) tomb at the Maya site of Rio
Azul in northeastern Guatemala
8–10
.
For APCI MS, the probe was operated
in positive-ion mode to monitor for peaks
at m/z4181 (theobromine) and m/z4195
(caffeine), with the ultraviolet detector set
at 270 nm. The results of the HPLC MS
confirmed the existence of theobromine in
3 of the 14 samples analysed. Peaks from
the extract of the residue from vessel 13
(Fig. 1) were evident in the total-ion and
ultraviolet chromatograms and the selected-
ion-monitoring trace at m/z4181 (Fig. 2;
for peak calibration for standards, see sup-
plementary information) and was con-
firmed by the fact that the mass spectrum
(see supplementary information) and
ultraviolet chromatogram of this peak
show the same retention time as theo-
bromine (Fig. 2b).
To our knowledge, this is the first time
that this new technique has been used to
analyse dry residues from the interior sur-
faces of prehistoric pottery. The presence
of cacao in Maya spouted vessels at Colha
indicates that its usage pre-dates evidence
from Rio Azul by almost a millennium.
We now know that the Maya had a long,
continuous history of preparing and con-
suming liquid chocolate from the Preclassic
period through to the Spanish Conquest.
Cacao wood charcoal dating to the same
period has been found at several sites in the
region
11
, further supporting the idea that
cacao drinking has its roots in the Preclas-
sic, and indicating that this part of northern
Belize may have been one of the main pro-
duction areas for cacao during this period.
W. Jeffrey Hurst*, Stanley M. Tarka Jr*,
Terry G. Powis†, Fred Valdez Jr†,
Thomas R. Hester†
brief communications
NATURE
|
VOL 418
|
18 JULY 2002
|
www.nature.com/nature 289
Cacao usage by the earliest Maya civilization
Foaming chocolate prepared in spouted vessels made a delectable Preclassic drink.
Figure 2 Chromatographic analysis of residue from Maya cooking vessels. a, Mass-spectroscopy chromatogram of dry extract from
vessel 13. Red, total-ion chromatogram; green, selected-ion monitoring (SIM) at
m
/
z
4181 for theobromine; blue, SIM at
m
/
z
4195 for
caffeine. For the mass spectrum of the peak at 2.6 min, as well as details of methods and standards, see supplementary information.
b, Ultraviolet chromatogram of the same extract, showing a peak at 2.6 min, which is consistent with results in a. Dashed lines are
integration markers; arrows indicate the starts and ends of peaks. The peak at around 1.5 min is due to the chromatography solvent. The
peak for caffeine is not evident as its concentration in cocoa is about 10 times lower than that of theobromine.
Figure 1 Early Maya use of cacao (
Theobroma cacao
): spouted
vessel no. 13, which was found to contain cocoa residue. This
vessel is one of 14 excavated from Colha in northern Belize,
dating to between 600 BC and AD 250.
1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0
Time (min)
0
1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0
a
b
Relative
milliabsorbance
units
Relative peak
area counts
© 2002
Nature
Publishing
Group
Reprint, New York, 1941).
6. Potter, D. R. in The Colha Project, Second Season, 1980 Interim
Report (eds Hester, T. R., Eaton, J. D. & Shafer, H. J.) 173–184
(Center for Archaeological Research, San Antonio, Texas;
Centro Studi Ricerche Ligabue, Venice; 1980).
7. Potter, D. R. in Archaeology at Colha, Belize, 1981 Interim Report
(eds Hester, T. R., Shafer, H. J. & Eaton, J. D.) 98–122 (Center
for Archaeological Research, San Antonio, Texas; Centro Studi
Ricerche Ligabue, Venice; 1982).
8. Hurst, W. J., Martin, A. J. Jr, Tarka, S. M. Jr & Hall, G. D.
J. Chromatogr. 466, 279–289 (1989).
9. Hall, G. D., Tarka, S. M. Jr, Hurst, W. J., Stuart, D. &
Adams, R. E. W. Am. Antiquity 55, 138–143 (1990).
10.Stuart, D. Antiquity 62, 153–157 (1988).
11.Turner, B. L. & Miksichek, C. H. Econ. Bot. 38, 179–193 (1984).
Supplementary information accompanies this communication on
Nature’s website.
Competing financial interests: declared none.
Uday K. Tirlapur, Karsten König
Laser Microscopy Division, Institute of Anatomy II,
Friedrich Schiller University, Teichgraben 7,
07743 Jena, Germany
e-mail: utir@mti-n.uni-jena.de
1. Stephens, D. J. & Pepperkok, R. Proc. Natl Acad. Sci. USA 98,
4295–4298 (2001).
290 NATURE
|
VOL 418
|
18 JULY 2002
|
www.nature.com/nature
*Hershey Foods Technical Center, PO Box 805,
Hershey, Pennsylvania 17033, USA
e-mail: whurst@hersheys.com
†Department of Anthropology, University of Texas,
Austin, Texas 78712-1086, USA
1. Hester, T. R. & Shafer, H. J. in Archaeological Views from the
Countryside: Village Communities in Early Complex Societies
(eds Schwartz, G. M. & Falconer, S. E.) 48–63 (Smithsonian
Institution, Washington DC, 1994).
2. Valdez, F. Jr The Prehistoric Ceramics of Colha, Northern Belize.
Thesis, Harvard Univ. (1987).
3. Powis, T. G. & Hurst, W. J. Proc. 66th Annu. Meeting Soc. Am.
Archaeol. (New Orleans, 2001).
4. Coe, S. D. & Coe, M. D. The True History of Chocolate
(Thames & Hudson, London, 1996).
5. Tozzer, A. M. Landa’s Relación de Las Cosas de Yucatán (Kraus
brief communications
Cell biology
Targeted transfection by
femtosecond laser
T
he challenge for successful delivery of
foreign DNA into cells in vitro, a key
technique in cell and molecular biol-
ogy with important biomedical implications,
is to improve transfection efficiency while
leaving the cell’s architecture intact. Here
we show that a variety of mammalian
cells can be directly transfected with DNA
without perturbing their structure by first
creating a tiny, localized perforation in the
membrane using ultrashort (femtosecond),
high-intensity, near-infrared laser pulses.
Not only does this superior optical tech-
nique give high transfection efficiency and
cell survival, but it also allows simultaneous
evaluation of the integration and expression
of the introduced gene.
Previous techniques that have been devel-
oped for transfection of cells with DNA
1
include carrier-mediated transfer
2
andtrans-
fer by plasma-membrane permeabilization
3
,
as well as direct transfer
4
, but the efficiency of
targeted DNA delivery by these methods
may not be optimal. Moreover, none allows
contact-free, non-disruptive, stable trans-
fection of individual cells and concomitant
evaluation in situof transgene expression.
We directed a high-intensity (10
12
W
cm
12
), near-infrared, femtosecond-pulsed
laser beam (wavelength, 800 nm) from an
80-MHz titanium–sapphire laser, with a
mean power of 50–100 mW and tightly
focused using a high-numerical-aperture
objective, at a sub-femtolitre focal volume
at the cell membrane. This resulted in the
formation of a single, site-specific, transient
perforation in the cell membrane through
which DNA could enter. This mode of
targeted transfection differs from the less
precise nanosecond-pulsed, ultraviolet (355
nm) lasers used previously
5
and which were
found to disrupt cellular integrity
6
.
Using Chinese hamster ovarian (CHO)
and rat–kangaroo kidney epithelial (PtK2)
cells, we studied the process of transfection
mediated by intense near-infrared femto-
second laser pulses. Cells were suspended
inside a sterile miniaturized cell chamber in
0.5 ml culture medium containing 0.2 mg
plasmid DNA vector pEGFP-N1 (4.7 kilo-
bases) encoding enhanced green fluorescent
protein (EGFP)
7
. Transmission images of
cells were obtained at low power (*5 mW),
and the near-infrared laser beam was then
focused (under the same microscope) on
the edge of the membrane of a target cell,
which was exposed to an enhanced mean
laser power of 50–100 mW for 16 ms so that
transfection could occur. More than 200
cells of each type were targeted in each of 18
replicate experiments; it took 10–15 s to
prepare for the transfection of each cell.
We assessed the integration and expres-
sion efficiency of the EGFP gene in situ by
time-lapse two-photon fluorescence imag-
ing
8
at a mean laser power of *1 mW over a
period of 72 h, as well as by two-photon flu-
orescence-lifetime imaging (TPFLIM)
9
. Fig-
ure 1 shows that diffraction-limited focusing
of intense femtosecond near-infrared laser
pulses selectively facilitates transfection of
the target cells, but not of the adjacent cells.
Expression of EGFP in the transfected cell is
also demonstrated by TPFLIM, and the
measured fluorescence lifetime of about
2.4 ns is consistent with that reported for
mammalian cells expressing EGFP
10
.
Irrespective of cell type, the transfection
achieved by this technique was invariably
100%. This high level of selective and total
transfection, without any detrimental effects
on growth and division, and with virtually
no cell death or sign of apoptosis, together
with the ability to determine expression by
fluorescence-intensity imaging and TPFLIM
with the same microscope, demonstrate the
potential of this non-disruptive technique
in transfection and expression studies. The
ability to transfer foreign DNA safely and
efficiently into specific cell types (including
stem cells) — circumventing the need for
mechanical, electrical or chemical means —
will be an encouraging advance for a range
of ventures, including targeted gene therapy
and DNA vaccination.
Figure 1 Analysis of the targeted transfection of Chinese hamster
ovarian (CHO) cells with a plasmid encoding enhanced green fluor-
escent protein (EGFP) by
in situ
visualization, and measurement of
its expression by near-infrared, two-photon-excitation-evoked,
real EGFP fluorescence detection and fluorescence-lifetime imag-
ing. a, Real EGFP fluorescence image of several CHO cells trans-
fected with the pEGFP-N1 plasmid. b, Transmission image of a
single transfected CHO cell (arrow). c, Two-photon fluorescence-
lifetime image of the same cell expressing EGFP; colour scale
indicates the fluorescence lifetime (
t
) between 0 and 5 nano-
seconds. Inset, distribution of fluorescence lifetime (in picoseconds)
of EGFP throughout the entire transfected cell. Scale bar, 25 mm.
© 2002
Nature
Publishing
Group
