Reaction of α-Acetoxy-N-nitrosopyrrolidine with Deoxyguanosine and DNA

Article abstract of DOI:10.1021/tx015521u

We investigated the reactions of α-acetoxy-N-nitrosopyrrolidine (α-acetoxyNPYR) with dGuo and DNA. α-AcetoxyNPYR is a stable precursor to the major proximate carcinogen of NPYR, α-hydroxyNPYR (3). Our goal was to develop appropriate conditions for the analysis of DNA adducts of NPYR formed in vivo. Products of the α-acetoxyNPYR-dGuo reactions were analyzed directly by HPLC or after treatment of the reaction mixtures with NaBH3CN. Products of the α-acetoxyNPYR-DNA reactions were released by enzymatic or neutral thermal hydrolysis of the DNA, then analyzed by HPLC. Alternatively, the DNA was treated with NaBH3CN prior to hydrolysis and HPLC analysis. The reactions of α-acetoxyNPYR with dGuo and DNA were complex. We have identified 13 products of the dGuo reaction - 6 of these were characterized in this reaction for the first time. They were four diastereomers of N2-(3-hydroxybutylidene)-dGuo (20,21), 7-(N-nitrosopyrrolidin-2-yl)Gua (2), and 2-(2-hydroxypyrrolidin-1-yl)deoxyinosine (12). Adducts 20 and 21 were identified by comparison to standards produced in the reaction of 3-hydroxybutanal with dGuo. Adduct 2 was identified by its spectral properties while adduct 12 was characterized by comparison to an independently synthesized standard. With the exception of adduct 2, all products of the dGuo reactions were also observed in the DNA reactions. The major product in both the dGuo and DNA reactions was N2-(tetrahydrofuran-2-yl)dGuo (10), consistent with previous studies. Several other previously identified adducts were also observed in this study. HPLC analysis of reaction mixtures treated with NaBH3CN provided improved conditions for adduct identification, which should be useful for in vivo studies of DNA adduct formation by NPYR.

Full text of DOI:10.1021/tx015521u

1428
Chem. Res. Toxicol. 2001, 14, 1428-1434
Meta bolic P r ofilin g of Ch r on ic Ca d m iu m Exp osu r e in
th e Ra t
J ulian L. Griffin,*,† Lee A. Walker,‡ Richard F. Shore,‡ and J eremy K. Nicholson†
Biological Chemistry, Biomedical Sciences Division, The Sir Alexander Fleming Building,
Exhibition Road, Imperial College of Science, Medicine and Technology, London, SW7 2AZ, U.K.,
and NERC Centre for Ecology and Hydrology, Monks Wood, Huntingdon,
Cambridgeshire, PE28 2LS, U.K.
Received May 2, 2001
A confounding problem with studying the effects of environmental exposure to contaminants
in wild populations is that analytical techniques are invasive, particularly where the
physiological effects of the toxin are assessed. In this study, a metabonomic approach to
investigate the biochemical effects of chronic oral exposure to environmentally realistic doses
1
of CdCl2 (low, 8 mg/kg; high, 40 mg/kg) is presented. H NMR spectra of urine from exposed
animals were analyzed using pattern recognition methods to identify biomarkers for a 94 day
exposure period. Creatinuria and both increased excretion and complexation of citrate was
detected after 19 days of exposure in both exposure groups. This was accompanied by a decrease
in plasma Ca2+/Mg2+ ratio in blood plasma after 94 days. Post mortem, magic angle spinning
(MAS) 1H NMR spectroscopy was used alongside conventional analytical techniques to
investigate intact tissue directly. According to atomic absorption sectroscopy, kidney tissue
accumulated 26.8 ( 2.5 µg of Cd2+/g dry wt (low) and 75.9 ( 4.3 µg of Cd2+/g dry wt (high).
1
Using high-resolution MAS H NMR spectroscopy altered lipid content was detected in kidneys
from animals exposed to Cd2+. However, unlike acute exposure, no testicular damage was
evident. This systemic approach to metabolism demonstrated the different physiological effects
of chronic subacute compared with an acute exposure to Cd2+
.
technique requires only the collection of a urine sample,
it also has the advantage of being noninvasive. Previ-
ously, analysis of urine using 1H NMR spectroscopy
identified various biomarkers of acute Cd2+ toxicity (11).
It was shown that urinary creatine concentration in-
creased as a result of testicular damage, while TCA cycle
intermediates were reduced due to inhibition of renal
carbonic anhydrase and subsequent alterations in TCA
cycle fluxes. However, these biomarkers were the result
of acute toxicity that followed intraperitoneal injection
of 6-24 mg of Cd2+/kg. Exposure of wild small mammals
in contaminated habitats is rarely this high and is also
primarily the result of ingestion of contaminated diet and
soil (7). Absorption of Cd2+ across the gut lining accounts
for only 3-5% of ingested Cd2+ (1, 9) and so the internal
dose of free living animals is much lower than that used
in the Nicholson and co-workers study (11) along with
other acute studies that involve intraperitoneal injection.
The dynamics of organ toxicity following environmental
Cd2+ exposure are also expected to vary, depending upon
whether exposure is acute or chronic.
In tr od u ction
Cadmium is a common contaminant of many natural
environments (1-3). While its impact on human and
animal health has been well studied in terms of acute
and subacute exposure (4-9), effects of low-level chronic
exposure, especially in the field, are poorly understood.
During acute exposure, Cd2+ induced necrosis and cel-
lular damage is known to occur in kidney, liver, and
testicular tissue; chronic exposure results in damage to
kidneys and bone (8-11).
One of the salient problems of monitoring chronic
exposure to environmental contaminants, such as Cd2+
,
is the lack of a suitable noninvasive technique to record
tissue Cd2+ accumulation and its biochemical effects. Cd2+
accumulation can be monitored using atomic absorption
spectroscopy of tissue extracts but this does not measure
the bioactivity of the metal. Although chemical assays
can record the increased expression of metallothionein
following Cd2+ exposure and histology can monitor neph-
rotoxicity, both are invasive and destructive techniques
involving considerable sample preparation (4, 5, 12, 13).
The use of coupled pattern recognition (PR) and 1H
NMR spectroscopy of biofluids [metabonomics (14)] has
proven to be a powerful technique in monitoring the
toxicological effects of many compounds in both animals
and man (11, 14-16). It provides information on the
degree of toxicity and its mode of action. Because the
In the present study, we have used a combination of
1
PR and H NMR spectroscopy to monitor the biochemical
changes induced in laboratory rats by a 94 day exposure
to environmentally realistic doses of Cd2+ incorporated
into their diet. At the end of the study, a range of organs
were analyzed using 1H NMR spectroscopy so that
biochemical changes could be correlated with tissue
damage. Conventional liquid-state NMR could be applied
to blood plasma, but, to determine whether renal or
testicular damage had occurred, high-resolution magic
* To whom correspondence should be addressed. Phone: +(44) 020-
7594-3230. Fax: +(44) 020-7594-3226. E-mail: j.griffin@ic.ac.uk.
† Biological Chemistry.
‡ NERC Centre for Ecology and Hydrology.
10.1021/tx015521u CCC: $20.00 © 2001 American Chemical Society
Published on Web 09/08/2001

Metabolic Profiling of Chronic Cadmium Exposure
Chem. Res. Toxicol., Vol. 14, No. 10, 2001 1429
angle spectroscopy (HRMAS) 1H NMR spectroscopy of the
intact tissue was used. NMR spectroscopy of intact tissue
is marred by broad resonances arising from dipole-dipole
intermolecular couplings, bulk magnetic field inhomog-
enieties, and short T2 relaxation. However, by spinning
intact tissue samples at the magic angle (θ ) 54.7°) and
at speeds in excess of the dipole-dipole couplings and
magnetic field inhomogenieties, high-resolution spectra,
comparable to those from tissue extracts, can be obtained
(17-21). Unlike extracts, all the metabolites present in
a tissue sample can be examined simultaneously using
this technique. This provided a holistic approach to
studying the perturbations induced in systemic metabo-
lism. These perturbations demonstrated how chronic
exposure to Cd2+ fundamentally differs in terms of target
organ toxicity and mechanism compared with an acute
insult.
Resonances from key metabolites were integrated using stan-
dard Bruker software (XWIN NMR version 2.5).
Blood plasma was also analyzed by 1H NMR spectroscopy
using both the conventional solvent presaturation pulse se-
quence described above (128 scans) and a Carr-Purcell-Meiboom-
Gill (CPMG) spin-echo experiment ([90 - (τ-180-τ)n - acqui-
sition]) as a T2 filter to attenuate broad signals from proteins
and lipoproteins (22). CPMG spectra were acquired with 256
scans with a total spin-spin (T2) relaxation delay of 80 ms, and
an interpulse delay of 2 s. Other acquisition parameters were
as described above. Each plasma sample consisted of 400 µL of
plasma diluted by 400 µL of D2O containing 1 mM TSP and 5
mM EDTA. EDTA was added to assess the cation content of
the blood plasma samples (23).
MAS 1H NMR Sp ectr oscop y An a lysis of Ren a l Tissu e.
HRMAS 1H NMR spectra were collected on 600 MHz Bruker
DRX spectrometers using a modified CPMG spin-echo pulse
sequence to remove the effects of lipids and large macromol-
ecules on spectral broadening (n ) 20; total spin-spin relaxation
delay (2nτ) ) 40 ms; relaxation delay ) 2 s between pulses).
Samples (∼20 mg) were spun at 5000 Hz at the magic angle
(54.7°) in the presence of D2O to provide a deuterium lock. For
both renal and testicular tissue, spectra were acquired with 256
scans into 32k data points with a spectral width of 4807.7 Hz
and a total pulse recycle delay of 9.82 s.
Cd 2+ Con cen tr a tion s in Ren a l Tissu e. Samples were dried
at 80 °C to constant weight, cold digested in 0.5 mL concentrated
nitric acid (Analar grade; BDH, Poole, U.K.), and then hot
digested at 120 °C until completely solubilized. Two 0.2 mL
aliquots of 30% w/v hydrogen peroxide (BDH, U.K.) were added
during digestion to aid sample oxidization. Digests were diluted
and analyzed by flame atomic absorption spectroscopy (Solar
969, ThermoUnicam, Cambridge, U.K.). The mean ((SE) %
recovery from a certified reference material (Reference material
186, Community Bureau of Reference) was 83.5 ( 2.29% (n )
3) and recoveries from two spiked blanks and four spiked tissue
samples were 104-110 and 109-121%, respectively. The limit
of detection (LoD) in the hepatic and renal samples (based on
mean sample weights) were 0.41 and 3.22 µg of Cd2+/g dry
weight (dw) respectively, while the LoD for the diet samples
was 2.77 µg of Cd2+/g wet weight. (ww).
Ma ter ia ls a n d Meth od s
Collection of Biologica l Sa m p les. Sprague Dawley male
rats (Strain, Crl, CD(SD) IGS BR, Charles River, U.K.) were
kept throughout the study in standard plastic cages in a
constant environment (15.5 light; 8.5 dark, 20 °C). They were
initially fed ad libitum on standard laboratory feed (RM1;
Special Diet Services, Witham, U.K.). The rats were subdivided
into three groups of six, placed in separate metabolic cages and
allowed to acclimatize for 24 h. Urine was then collected over
ice packs for 24 h; urine collectors contained a small quantity
(0.01%) of sodium azide to prevent bacterial contamination. Any
food particles or faeces were removed from the urine sample by
centrifugation, and the samples were stored at -20 °C prior to
NMR analysis. These urine collections were used to provide
predose measurements for all three rat groups. Subsequently,
animals were maintained for 94 days on either the normal
(control) diet, or the same diet with nominal concentrations of
8 (low) or 40 (high) mg/kg CdCl2. Urine samples were subse-
quently collected (as described above) at 6 h, 24 h, 2 days, 3
days, 9 days, 19 days, 26 days, 44 days, 58 days, 75 days, and
94 days after animals were placed on the Cd2+ enriched diet.
Meta lloth ion ien Con ten t in Hep a tic Tissu e. Hepatic
metallothionein residues were determined by cadmium-satura-
tion (24). Liver samples were homogenized in 0.25 M sucrose
in Tris-HCl buffer (pH 7.5) and centrifuged at 4 °C. The
supernatant was removed and the metallothionein present
saturated with CdCl2. Excess Cd2+ was then removed by binding
with haemoglobin (2% w/v bovine haemoglobin; Sigma, Poole
U.K.) solution, which was then denatured by heat treatment
and removed from the sample by centrifugation. The samples
were then acidified with nitric acid and the Cd2+ content of the
acidified sample determined by flame-atomic absorption spec-
troscopy. The concentration of metallothionein in the liver was
calculated, assuming that 6 Cd2+ bind per metallothionein
protein which has a molecular weight of 6050 (25). The LoD for
metallothionein in the liver samples was 3.91 µg of protein/g
tissue ww. Metallothionein concentrations in blanks included
with the unknowns were below the LoD.
At the end of the experiment rats were sacrificed by cervical
dislocation and a terminal trunk bleed was immediately col-
lected in heparinised tubes. Blood was centrifuged at 10 000 rpm
for 10 min, and the plasma separated and stored at -20 °C.
At post mortem any gross anatomical changes were noted
visually. The left kidney was then removed within 20 s and ∼20
mg sections were taken from the inner renal cortex and papilla
for HRMAS spectroscopy. The right kidney was removed for
Cd2+ analysis by atomic absorption spectroscopy (see below).
About 20 mg of testicular tissue (general tubular region) was
removed and analyzed by HRMAS spectroscopy. Samples of liver
were also removed to determine metallothionein concentration
(see below). All tissues were snap-frozen in liquid nitrogen
immediately after dissection and stored at -70 °C prior to
analysis.
NMR Sp ectr oscop y of Ur in e a n d Blood P la sm a . 1H NMR
spectra were obtained on a Bruker Avance DRX-600 Spectrom-
eter operating at 600.13 MHz 1H frequency at 298 K. Urine
samples were examined using a broad band inverse-detected
flow injection probe operated using the BEST automation
package (Bruker, Germany). Each urine sample consisted of 600
µL of urine, 300 µL of sodium phosphate buffer (pH 7.1), and
100 µL of D2O (internal lock) with TSP (final concentration 1
mM) as an internal reference standard (δ ) 0 ppm). 1H NMR
spectra were collected using a selective solvent presaturation
sequence with 32 scans, an acquisition time of 2.34 s and a
further relaxation delay of 5 s (22). The spectra were acquired
using 7 kHz spectral width and 32k time domain data points.
P a tter n Recogn ition In vestiga tion of NMR Sp ectr a . For
the purposes of pattern recognition analysis, 1H NMR spectra
were data reduced (AMIX v.2 program, Bruker, Germany) by
being sub-divided into 0.04 ppm designated regions from δ 0.4-
10.0 for urine and δ 0.4-4.5 for blood plasma and MAS spectra
(15, 26, 27). The residual water peak area was excluded from
analysis and then the area of the resonances within each
remaining region integrated. Each spectral region was normal-
ized to the summation of spectral regions for the entire
spectrum, which accounted for dilution or bulk mass differences
between samples. Correlation Principal Components Analysis
(PCA) was performed using Pirouette v. 2.6 (InfoMetrix, Wood-
inville, WA). PCA is an unsupervised method of classification,
requiring no training set of data, that mathematically correlates
Spectra were Fourier transformed following multiplication of
an exponential term equivalent to 0.3 Hz line broadening.

1430 Chem. Res. Toxicol., Vol. 14, No. 10, 2001
Griffin et al.
Ta ble 1. Mea n Con cen tr a tion s of Ca d m iu m in th e Diets
(µg of Cd /g of w w ), Ren a l, a n d Hep a tic Tissu es (both µg
of Cd /g of d w ) a n d Hep a tic Meta lloth ion ein Resid u es (µg
of MT/g of w w ) for Ra ts F ed Con tr ol, Low (8 µg of Cd 2+/g)
a n d High (40 µg of Cd 2+/g) Ca d m iu m -Con ta m in a ted
Dietsa
treatment
Cd in diet
renal Cd
hepatic Cd hepatic MT
control
low Cd
high Cd
<2.77
<3.22
<0.4
3.0 ( 0.3
8.9 ( 1.1
<3.9
13.2 ( 1.7
31.0 ( 2.9
8.8 ( 0.3
26.8 ( 2.5
35.7 ( 1.65 75.9 ( 4.3
a
N ) 6 for dietary, renal and hepatic Cd concentration, and
N) 5 for hepatic MT residues.
the variation detected in a data set to a spatial representation
of the data. The first PC generated by this method represents
the maximum variation that can be correlated together for the
data. Subsequent PCs represent less variation and are chosen
to be orthogonal to the first PC.
Both mean-centered and autoscaling routines were applied
to the spectral data prior to independent PCA. Mean-center
analysis involved subtraction of the mean of the integral for a
designated spectral region and measures variance about this
mean. Thus, PCs in mean centered data are influenced by
regions with the maximum variance about the mean. Autoscaled
analysis sets each integral region to unit variance, ensuring the
sum of a single PC across all observables will be zero. This
weights spectra so that features with lower spectral intensity
and the high concentration metabolites contribute equally to
the PC analysis. Spectral data were reduced from 240 spectral
descriptors to a number of PCs (less than 5) as determined by
the maximization of cumulative total variation in the data that
is predicted by those components. Loading plots from the PC
analysis were used to identify resonances to predict metabolite
changes within renal tissue exposed to Cd2+. These predictions
were then cross validated with spectra both visually and by
manual integration of the regions.
F igu r e 1. Urinary profiles from control (A) and high dose (B)
CdCl2 (40 ppm) exposed rats after 19 days of exposure. Key: 1,
acetate; 2, acetylated glycoproteins; 3, glutamate; 4, R-ketoglu-
tarate; 5, oxaloacetate; 6, citrate; 7, creatine/creatinine; 8,
phosphocholine and taurine (underlying triplet); 9, trimethyl-
amine-N-oxide; 10, creatine; 11, creatinine.
Sta tistica l An a lysis. Data are represented as means (
standard error of the mean. Students t-test and Tukey tests
were performed using Instat (Graphpad Inc, CA).
Resu lts
Cd 2+ Accu m u la tion a n d Meta lloth ion ein Exp r es-
sion . Rats fed either diet contaminated with CdCl2
accumulated Cd2+ in both renal and hepatic tissue (Table
1). The ∼4-fold increase in the mean Cd2+ concentrations
from the low to high cadmium contaminated diets cor-
responded to a difference in both renal and hepatic Cd2+
concentrations of ∼3-fold. Increased metallothionein
expression was also detected in hepatic tissues from rats
fed the contaminated diets, with a ∼2-fold increase in
expression between the high and low dose exposure
groups (Table 1).
F igu r e 2. Mean centered and autoscaled preprocessing PC
analysis of urinary profiles at 19 days of exposure to CdCl2 in
feed. Key: (0) control, ([) low dose (8 ppm), (b) high dose.
complete separation apparent for all the high, medium,
and control dosed animals, at subsequent time points the
majority of spectra (>3) from both the high and medium
dosed groups could be shown to occupy separate PC space
compared with the remainder of the spectra. Further-
more, all the animals exposed to high and low doses of
Cd2+ did not cluster with the control group for at least
one time point. Mean PC loadings for each time point
showed that the animals exposed to high and low doses
of Cd2+ occupied a different PC space than the control
cluster (Table 2). Inspection of the metabolites that
contributed to the PC loadings indicated that both citrate
and creatine contributed to this separation. Between 19
days and 58 days, separation occurred in either PCs 1,
2, or 3, but, after this time period, higher order PCs,
which represented less of the total variation described
in the spectra were the only ones that separated the
groups.
NMR Sp ectr oscop y of Ur in e Dem on str a tes Cr e-
1
a tin u r ia a n d Ca tion Ch ela tion by Citr a te. The H
NMR urinary profiles (Figure 1) were analyzed by PR to
determine the metabolic consequence of Cd2+ insult
across the time course. Separation of some urine profiles
from the cluster of control spectra was apparent from 19
days for animals in the high dose group and three
animals from the low dose group for both mean centered
and autoscaled PC analysis (Figure 2). The separation
was more evident following mean centered analysis
indicating the metabolites correlated most to cadmium
toxicity were caused by high concentration metabolites.
The separation was caused by four effects: an increase
in sugars, an increase in creatine, a decease in hippurate
and a low-frequency shift in the chemical shift of the
citrate resonance (Figure 1). While at no time point was

Metabolic Profiling of Chronic Cadmium Exposure
Chem. Res. Toxicol., Vol. 14, No. 10, 2001 1431
Ta ble 2. P r in cip le Com p on en ts th a t Ca u se Clu ster in g of Cd 2+ Dosed Ra ts in Sep a r a te P C Sp a ce to th a t of Con tr olsa
PC(s)
causing
separation
percentage
variance
of PC(s)
time/days
19
mean loading of PC
(A) Mean Centered
metabolites contributing
PC1
35
high: 10.7 ( 6.30**
citrate (+); creatine (+); taurine (-)
low: 8.7 ( 4.8**
control: -20.6 ( 1.38
high: -0.56 ( 1.34
low: -2.80 ( 3.26
control: 4.57 ( 1.93
26
PC4
7.5
creatine (+); citrate (chemical shift and +); hippurate (-)
44
58
PC1/PC4
PC3
50.3
13.5
citrate (chemical shift); creatine (+)
creatine (+); citrate (chemical shift and +); taurine (-)
high: -4.70 ( 2.00**
low: -1.42 ( 1.15*
control: 5.35 ( 1.62
75
96
no separation
PC4/PC5
7.2
citrate (chemical shift); creatine (+); hippurate (+)
(B) Autoscale
high: 0.21 ( 1.88
19
PC2
21.1
creatine (+); lactate (+); hippurate (-); taurine (-);
glutamate (+); lecuine (+), isoleucine (+); valine (+)
low: 3.64 ( 2.03
control: -2.50 ( 0.98
26
44
no separation
PC2/PC4
26.3
32.5
sugars/glucose (+); lactate (+); glutamate (+);
hippurate (-); taurine (-)
creatine (+); citrate (chemical shift and +). sugars (+);
low concentrated aromatic amino acids (+)
58
PC1
high: 2.28 ( 2.56
low: 4.55 ( 2.74*
control: -5.64 ( 2.50
75
96
PC2/PC4
PC5
29.3
3.8
creatine (+), sugars (+); glutamate (+);
citrate (chemical shift and +)
creatine (+); glutamate (+); sugars/glucose (+);
phenylalanine (-); tyrosine (-)
high: 1.83 ( 1.13*
low: 0.51 ( 1.10
control: -1.88 ( 0.62
a
PCs are listed for the different time points where clustering was evident along with the amount of spectral variation represented (%)
by the PC and the main constituents of the PC for both autoscale and mean centered analysis. Mean ((SE) PCs for an exposure group
have been calculated where separation was predominantly across one PC. (*) p < 0.05, (**) p < 0.005 for significant difference from
control group.
F igu r e 3. Plot of PC5 versus time for three animals exposed to either a diet containing no ([), low (0), or high Cd2+ (2). Time
points plotted at predose (-1 day), 1, 2, 19, 26, 44, 58, and 94 days. Other time points omitted due to changes in solvent suppression
quality across the spectra.
To investigate further the temporal variance in urinary
profiles, the mean spectral profile (a summation of the
spectra for each individual in a treatment group) were
analyzed using mean centered PCA across the majority
of time points (time points where data had bad water
suppression were excluded from this analysis). This
demonstrated that the high and low Cd2+ dosed animals
were most affected between 26 and 58 days after initial
exposure (Figure 3). This was largely due to a change in
the chemical shift and relative intensity of the citrate
resonances at δ 2.58 and 2.70.
To confirm the occurrence of creatinuria in Cd-dosed
animals after 19 days, the concentration of creatine was
quantified by comparing the singlet resonance at δ 3.02
with two easily quantifiable metabolite resonances: ac-
etate (singlet δ 1.91; invariant in urine between the three
groups) and taurine (triplet δ 3.41; decreased in urine of
animals exposed to Cd2+). The ratio of creatine/acetate
was not significantly different (high Cd2+ exposed ) 4.54
( 0.20; low Cd2+ ) 4.01 ( 0.20; control ) 3.84 ( 0.30)
but the creatine/taurine ratio was increased in Cd-dosed
animals (high Cd2+ exposed ) 3.26 ( 0.48*; low Cd2+
exposed ) 3.40 ( 0.62*; control ) 1.63 ( 0.22; *p < 0.05
for difference from control group).
The chemical shift change recorded for citrate in
animals exposed to Cd2+ may have been produced either
by an alteration in pH or as a result of chelation of cations
by the citrate. However, all urine samples were buffered
to a pH of ∼7.1 and this was confirmed for the urine
samples in question. To determine whether chelation was
responsible for the chemical shift alteration, 5 mM EDTA
was added to the 19 day urine samples. EDTA behaves

1432 Chem. Res. Toxicol., Vol. 14, No. 10, 2001
Griffin et al.
F igu r e 6. Mean centered analysis of blood plasma profiles
following 94 days of exposure to CdCl2. Key: (0) control, ([)
low dose (8 ppm), (b) high dose.
resonances at δ 3.11. Cadmium exposed rats had in-
:
creased complexed Ca2+ Mg2+ ratio compared with
control rats (singlets at δ 2.61 and 2.75, respectively,
control ) 1.12 ( 0.03; low ) 1.16 ( 0.05; high ) 1.51 (
0.06; p < 0.001 for difference between control and high
dose and low and high dose groups). Exposure of rats to
Cd2+ also affected the spectra obtained from both the
inner cortex and the papilla of the kidney. Renal tissue
from Cd2+ exposed animals had decreased resonance
intensities for CH3 lipid triglycerides and phosphocholine
but increased resonance intensities from sugars, choline,
and phosphatidylcholine/inositol, as determined using PR
and visual inspection of the integral regions (data not
shown). Exposure to Cd2+ had no biochemical effect in
HRMAS spectra of testicular tissue. To confirm that there
was no change in the creatine content of testicular tissue
in the high dose rats, the resonance of creatine (δ 3.04)
was integrated along with that of choline (δ 3.20; as-
sumed to remain constant). The ratio of testicular cre-
atine/choline was the same for Cd2+ exposed and control
animals (high dose Cd2+ exposed ) 2.7 ( 0.5; control )
2.2 ( 0.2). To examine the redox potential of the tissue,
the ratio of alanine to lactate was estimated by compar-
ing the ratio of the resonances at δ 1.44 and 1.33,
respectively; again exposure to Cd2+ had no detectable
effect (high dose Cd2+ exposed ) 3.1 ( 0.6; control ) 3.4
( 0.4).
F igu r e 4. CPMG 1H NMR spectra of blood plasma from rats
fed control diet (A) or diet contaminated with 40 ppm. CdCl2
(B). Key: 1, VLDL and LDL CH3-lipid groups; 2, leucine, valine,
and isoleucine resonances; 3, VLDL and LDL CH2CH2CH2 lipid
groups; 4, lactate; 5, alanine; 6, acetate; 7, CH2CdC lipid
resonances and glycoproteins; 8, methionine; 9, [Ca-EDTA]2-
(ethylenic protons); 10, [Mg-EDTA]2- (ethylenic protons); 11,
[Ca-EDTA]2- (acetate protons); 12, free EDTA4- and [Mg-
EDTA]2- (acetate protons); 13, free EDTA4-; 14, glucose reso-
nances; 15, lactate.
Discu ssion
Acute Cd2+ toxicity has been well documented both in
laboratory animals and man (1-3), but subacute chronic
exposure to the metal is still poorly understood. This is
despite environmental exposure to Cd2+ most commonly
involving long term chronic exposure, rather than an
acute incident. In the present study, we used urinary
profiles to metabolically profile the changes in systemic
biochemistry during chronic exposure to Cd2+. These
changes were cross-correlated with evidence for toxicity
within organs targeted by Cd2+. Chronic exposure to Cd2+
both induced different biochemical processes within
tissues and targeted different organs compared with
acute toxicity.
F igu r e 5. High resolution MAS 1H spectra of renal tissue
(inner cortex) (A) and testicular tissue (B) from a rat exposed
to a diet containing 40 ppm CdCl2. 1, CH3 lipid groups; 2,
leucine, isoleucine, and valine; 3, CH2 lipid groups; 4, lactate;
5, alanine; 6, CH2CH2CO lipid; 7, acetate; 8, CH2CdC lipid; 9,
glutamate; 10, CH2CdC lipid + glutamate; 11, CH2CdC lipid
+ glutamate + glutamine; 13, CH2CO; 14, creatine; 15, choline
resonances (choline, phosphocholine, and phosphatidylcholine);
16, glycine; 17, glucose resonances.
as a competitive chelation agent and caused the citrate
resonances to move to low frequencies, confirming that
the shift in citrate was caused by metal complexation.
Alter ed Bioch em istr y in Blood P la sm a a n d Kid -
n ey Tissu e, Bu t Not in th e Testes. Blood plasma
(spiked with EDTA to investigate cation content; Figure
4) and samples of renal (inner cortex and papillar; Figure
5a) and testicular tissue (Figure 5b) were examined by
Creatinuria was one of the major impacts of Cd2+ insult
during high and low dose chronic exposure. This was
detected by pattern recognition and confirmed in terms
of crude metabolite ratios (creatine/taurine ratio). El-
evated urinary creatine is also the dominant feature of
acute Cd2+ toxicity in male rats (22, 28, 29) and is caused
by testicular atrophy from vascular constriction (22),
although Cd2+ toxicity in the testes may also affect
testicular tissue directly (30). However, in the present
study, there was no reduction in creatine in the testes of
1
MAS H NMR spectroscopy to investigate organ toxicity.
CPMG spectra of blood plasma were examined by PC
analysis. Both high and low dose groups readily sepa-
rated from the control group along PC2 (Figure 6), caused
by decreases in the resonance intensity of (CH2)n lipo-
proteins (δ 1.33), acetate (δ 1.94) and alanine (δ 1.44).
There was also an increase in intensity of the EDTA

Metabolic Profiling of Chronic Cadmium Exposure
Chem. Res. Toxicol., Vol. 14, No. 10, 2001 1433
rats exposed chronically to either high or low Cd2+ doses.
Furthermore, the redox potential of testicular tissue was
similar between Cd2+ exposed tissue and control tissue,
demonstrating that testicular toxicity had not occurred.
There was also no evidence of reduced creatine content
in renal tissue. Thus, another source of creatine must
have been responsible for the creatinuria detected after
19 days of exposure.
a decrease in CH3 lipids was detected in blood plasma in
this study, indicating that some of the lipid profile
changes were caused by alterations in lipid and fatty acid
transport. While the kidney is the primary site of Cd2+
storage, the liver also accumulates significant amounts
of the metal (Table 1; refs 1, 2, 8, 9, 34). Hepatic tissue
stores and modifies lipids, and reductions of CH3 lipid
moieties in blood plasma and the kidneys may indicate
hepatic damage.
Two exposure levels were used during this study, with
the higher dose being four times larger than the lower.
However, accumulation of Cd2+ within renal and hepatic
tissue only increased 3-fold. Furthermore, metallothio-
nein expression only increased 2-fold. This suggests that
at higher exposure levels, Cd2+ is either stored peripher-
ally to the kidney and liver, or excretory mechanisms are
increased. This emphasizes again the difference in pa-
thology induced by acute and chronic cadmium toxicity.
While acute exposure to cadmium causes the metal to
be rapidly accumulated within liver and kidney tissue,
the burden of chronic exposure is reduced in these two
tissues.
In this study, we have recorded the physiological effects
of a subacute, chronic exposure to Cd2+ toxicity. Unlike
acute exposure to the metal, the vascular system of the
testes was not targeted. While renal metabolism is
perturbed in both acute and chronic exposure, adaptive
mechanisms counteract renal tubular acidosis during
chronic exposure. This is not the case following acute
exposure. However, following chronic exposure the kidney
is also affected by an alteration in lipid content, possibly
caused by mitochondrial proliferation. Analysis of both
blood plasma and urine indicated that there was a
perturbation in systemic metabolism caused by chronic
exposure.
A principal feature of acute Cd2+ toxicity is renal
tubular acidosis. This is caused by direct inhibition of
carbonic anhydrase (11). Without carbonic anhydrase to
modulate renal pH, the kidney consumes glutamine to
+
generate NH4 ions and so modulates pH (31-33). This
is accompanied by a reduction in renal glutamate content,
as previously observed in the renal tissue of bank voles
(Clethrionomys glareolus) (9). While no change in gluta-
mate content was detected in rat renal tissue in the
present study, a requirement by the kidney for glutamine
would explain the urine creatinuria detected. Muscle is
rich in both glutamine and creatine and the tissue
exports glutamine to stimulate other organs, particularly
the kidney, during pH regulation (31, 32). This process
impairs muscular metabolism, and may have caused the
transient creatinuria detected in this study. Creatinuria
was less marked later in the study, and no excess of
creatine or glutamine was detected in blood plasma.
However, a reduction in blood plasma alanine was
detected in rats exposed to the high dose of Cd2+, and a
combination of alanine amino transferase and glutamate
dehydrogenase could supply renal tissue with NH4+ ions
to reduce the effects of Cd2+ induced long-term renal
acidosis. This demonstrates that, while acute toxicity
causes transient renal acidosis, systemic metabolism is
altered during chronic exposure to maintain normal renal
function.
A transient increase in citrate excretion, and a change
in the chemical shift of four resonances of the molecule
were detected in high and low dose exposed rats. Nichol-
son and co-workers (11) previously found that acute
exposure to Cd2+ caused a decrease in citrate concentra-
tion, as well as other TCA cycle intermediates. This was
caused by a decrease in TCA cycle flux following renal
tubular acidosis. The changes seen in the present study
following chronic exposure to Cd2+ are fundamentally
different. The chemical shift change of the citrate reso-
nances was not induced by urine acidosis, but by chela-
tion of ions by citrate. Cd2+ displaces Ca2+ within the body
(2) and an increase in extracellular free Ca2+ is detri-
mental to renal tissue as excess calcium precipitates as
calcium phosphate. The decrease in taurine urinary
excretion that was detected also indicates that there was
a change in osmoregulatory homeostasis, as taurine is a
known osmoregulatory compound. Increased Ca2+ con-
tent, compared with Mg2+ plasma concentration, was also
detected in blood plasma samples, which is consistent
In conclusion, our study clearly demonstrates the very
different effects that a contaminant may have depending
upon whether exposure is either acute or chronic. The
consequences are that different organs are targeted, and
different biochemical pathways within the affected tissue
are induced. The results have clear implications about
the potential pitfalls in extrapolating the results of acute
exposure studies to environmental situations, and dem-
onstrate the ease of a “metabonomics” approach to
following environmental toxicology.
Ack n ow led gm en t. This work was supported by an
Environmental Diagnostic award made by the Natural
Environment Research Council, UK. J .L.G. is a Royal
Society University Research Fellow.
Refer en ces
(1) Bremner, I. (1979) The Chemistry, Biochemistry and Biology of
Cadmium (Webb, M., Ed.) pp 175-193, Elsevier, Amsterdam.
(2) Goyer, R. A. (1995) in Casarett and Doull’s Toxicology (Klassen,
C. D., Ed.) 5th ed., pp 699-702, McGraw-Hill Professional
Publishing, Singapore.
(3) Peterson, A. P., and Alloway, B. J . (1979) The Chemistry,
Biochemistry and Biology of Cadmium (Webb, W., Ed.) pp 45-
92, Elsevier, New York.
(4) Nicholson, J . K., Kendall, M. D., and Osborn, D. (1983) Cadmium
and mercury nephrotoxicity. Nature 304, 633-635.
(5) Nicholson, J . K., and Osborn, D. (1983) Kidney lesions in pelagic
seabirds with high tissue levels of cadmium and mercury. J . Zool.
London 200, 99-118.
(6) Roels, H., Bernard, A., Buchet, J . P., Goret, A., Lauwerys, R.,
Chettle, D. R., Harvey, T. C., and Haddad, I. A. (1979) Critical
concentration of cadmium in renal cortex and urine. Lancet 1,
221.
with Cd2+-induced displacement of tissue Ca2+
.
In rats exposed to the high dose of Cd2+, renal tissue
had decreased CH3 lipid but increased choline/phospho-
choline content. We have previously demonstrated an
alteration in the lipid profile of bank voles chronically
exposed to the high Cd2+ diet for 14 days (26). This was
thought to have been caused either by a necrosis-induced
change in cell populations, infiltration of immune cells
to repair damaged renal tissue, or mitochondrial prolif-
eration, which is known to accompany chronic Cd2+
exposure in both birds and laboratory mice (5). However,

1434 Chem. Res. Toxicol., Vol. 14, No. 10, 2001
Griffin et al.
(7) Shore, R. F., and Douben, P. E. T. (1994) The ecotoxicological
significance of cadmium intake and residues in terrestrial small
mammals. Ecotoxicol. Environ. Saf. 29, 101-112.
(8) World Health Organisation (1980) Recommended Health-Based
Limits in Occupational Exposure to Heavy Metals, WHO, Geneva.
(9) World Health Organisation (1992) Environmental Health Criteria
134: Cadmium, WHO, Geneva.
(10) Dunnick, J ., and Fowler, B. A. (1988) in Handbook on Toxicity of
Inorganic Compounds (Seller, H. G., and Siegel, H., Eds.) pp 115-
174, Marcel Dekkar, New York.
(21) Millis, M., Weybright, P., Campbell, N., Fletcher, J . A., Fletcher,
C. D., Cory, D. G., and Singer, S. (1999) Classification of Human
Liposarcoma and Lipoma using Ex vivo Proton NMR spectroscopy.
Magn. Res. Med. 41, 257-267.
(22) Nicholson, J . K., Foxall, P. J ., Spraul, M., Farrant, R. D., and
Lindon, J . C. (1995) 750 MHz 1H and 1H-13C NMR spectroscopy
of human blood plasma. Anal. Chem. 67 (5), 793-811.
(23) Nicholson, J . K., Buckingham, M. J ., and Sadler, P. J . (1983) High
resolution 1H nmr studies of vertebrate blood and plasma.
Biochem. J . 211, 605-615.
(11) Nicholson, J . K., Hingham, D. P., Timbrell, J . A., and Sadler, P.
J . (1989) Quantitative high resolution 1H NMR Urinalysis studies
on the biochemical effects of cadmium in the rat. Mol. Pharmacol.
36, 398-404.
(12) Larison, J . R., Likens, G. E., Fitzpatrick, J . W., and Crock, J . G.
(2000) Cadmium toxicity among wildlide in the Colorado Rocky
Mountains. Nature 411, 181-183.
(13) Richardson, M. E., Fox, M. R., and Fry, B. E. (1973) Pathological
changes produced in J apanese Quail by ingestion of cadmium. J .
Nutr. 104, 323-338.
(14) Nicholson, J . K., Lindon, J . C., and Holmes, E. (1999) ‘Metabo-
nomics’: understanding the metabolic responses of living systems
to pathophysiological stimuli via multivariate statistical analysis
of biological NMR data. Xenobiotica 11, 1181-1189.
(15) Anthony, M. L., Sweatman, B. C., Beddell, C. R., Lindon, J . C.,
and Nicholson, J . K. (1994) Pattern recognition classification of
the site of nephrotoxicity based on metabolic data derived from
proton nuclear magnetic resonance spectra of urine. Mol. Phar-
macol. 46, 199-211.
(16) Holmes, E., Nicholls, A. W., Lindon, J . C., Ramos, S., Spraul, M.,
Neidig, P., Connor, S. C., Connelly, J ., Damment, S. J ., Haselden,
J ., and Nicholson, J . K. (1998) Development of a model for
classification of toxin-induced lesions using 1H NMR spectroscopy
of urine combined with pattern recognition. NMR Biomed. 11,
235-244.
(24) Onsaka, S., and Cherian, M. G. (1982) Comparison of metallothio-
nein Determination by polarographic and cadmium-saturation
methods. Toxicol. Appl. Pharmacol. 63, 270-274.
(25) Kagi, J . H. R., and Vallee, B. L. (1961) Metallothionein:
a
cadmium and zinc containing protein from equine renal cortex.
II. Physicochemical properties. J . Biol. Chem. 236, 2435-2442.
(26) Griffin, J . L., Walker, L. A., Troke, J ., Osborn, D., Shore, R. F.,
and Nicholson, J . K. (2000) The initial pathogenesis of cadmium
induced renal toxicity. FEBS Lett. 478, 147-150.
(27) Holmes, E., Nicholls, A. W., Lindon, J . C., Ramos, S., Spraul, M.,
Neidig, P., Connor, S. C., Connelly, J ., Damment, S. J ., Haselden,
J ., and Nicholson, J . K. (1998) Development of a model for
classification of toxin-induced lesions using 1H NMR spectroscopy
of urine combined with pattern recognition. NMR Biomed. 11,
235-244.
(28) Gunn, S., Gould, T. C., and Anderson, W. (1982) The selective
injurious response of testicular and epididymal blood vessels to
cadmium and its prevention by zinc. Am. J . Pathol. 42, 685-696.
(29) Samarickmansinghe, G. P. (1979) Biological effects of cadmium
in mammals. In The Chemistry, Biochemistry and Biology of
Cadmium (Webb, M., Ed.) Elsevier/North-Holland Biomedical
Press, Amsterdam.
(30) J anecki, A., J akubowiak, A., and Steinberger, A. (1992) Effect of
Cadmium Chloride on Transepithelial Electrical Resistance of
Sertoli Cell Monloayers in Two-Compartment Cultures. Toxicol.
Appl. Pharmacol. 112, 51-57.
(31) Hems, D. A. (1972) Metabolism of glutamine and glutamic acid
by isolated perfused kidneys of normal and acidotic rats. Biochem.
J . 130, 671-680.
(32) Newsholme, E. A., and Leech, A. R. (1983) Biochemistry for the
Medical Sciences, p 512, Wiley, London.
(33) Goldstein, L. (1980) Adaptation of renal ammonia production to
metabolic acidosis. The Physiologist 23, 19-25.
(34) Shore, R. F. Myhill, D. G., Routledge, E. J ., and Wilby, A. (1995)
Impact of an environmentally realistic intake of cadmium on
calcium, magnesium and phosphate metabolism in bank voles,
Clethrionomys glareolus. Arch. Environ. Contam. Toxicol. 29,
180-186.
(17) Andrew, E. R., and Eades, R. G. (1959) Removal of dipolar
broadening of NMR spectra of solids by specimen rotation. Nature
183, 1802.
(18) Cheng, L. L., Ma, M. J ., Becerra, L., Hale, T., Tracey, I., Lackner,
A., and Gonzalez, R. G. (1997) Quantitative neuropathology by
high-resolution magic angle spinning proton magnetic resonance
spectrscopy. Proc. Natl. Acad. Sci. U.S.A. 94, 6408-6413.
(19) Garrod, S., Humpfer, E., Spraul, M., Connor, S. C., Polley, S.,
Connelly, J ., Lindon, J . C., Nicholson, J . K., and Holmes, E. (1999)
High-resolution MAS-1H NMR spectroscopic studies on intact rat
renal cortex and medulla. Magn. Reson. Med. 41, 1108-1118.
(20) Millis, K., Maas, E., Cory, D. G., and Singer, S. (1997) Gradient
high-resolution magic angle spinning nuclear magnetic resonance
spectroscopy of human adipocyte tissue. Magn Reson. Med. 38,
399-403.
TX015521U