Xia Wang1, Caidan Rezeng2, Yingfeng Wang1, Jian Li3, Lan Zhang1, Jianxin Chen3, Zhongfeng Li1. 1. Department of Chemistry, Capital Normal University, No. 105, Xisanhuanbeilu, Haidian District, Beijing 100048, PR China. 2. College of Pharmacy, Qinghai Nationalities University, No. 3 Bayizhong Road, Xining 810000, PR China. 3. Beijing University of Chinese Medicine, No. 11 Beisanhuandonglu, Chaoyang District, Beijing 100029, PR China.
Abstract
Renqingchangjue (RQCJ), a kind of Traditional Tibetan Medicine, has been widely utilized to treat various gastroenteritis diseases. However, the biosafety and toxicity of RQCJ was still indefinite because of toxic components in RQCJ, which included a variety of heavy metals. Thus, this study was aimed to evaluate the toxicity and expound the toxicological mechanism of RQCJ. In this study, rats were intragastrically administered with different doses of RQCJ for 15 days, and then, the restorative observation period lasted for 15 days. Liver and kidney tissues were collected for histopathological examination, and simultaneously serum and urine samples were collected for 1H nuclear magnetic resonance (1H NMR) spectroscopy analysis and biochemical analysis combined with inductively coupled plasma mass spectrometry (ICP-MS) measurement. The 1H NMR-based metabolomics analysis revealed that the administration of RQCJ significantly altered the concentrations of 14 serum metabolites and 14 urine metabolites, which implied disturbances in energy metabolism, amino acid metabolism, intestinal flora environment, and membrane damage. Besides, the biochemical analysis of serum samples was consistent with the histopathological results, which indicated slight hepatotoxicity and nephrotoxicity. The quantification of As and Hg in urine and serum samples by ICP-MS provided more evidence about the toxicity of RQCJ. This work provided an effective method to systematically and dynamically evaluate the toxicity of RQCJ and suggested that precautions should be taken in the clinic to monitor the potential toxicity of RQCJ.
Renqingchangjue (RQCJ), a kind of Traditional Tibetan Medicine, has been widely utilized to treat various gastroenteritis diseases. However, the biosafety and toxicity of RQCJ was still indefinite because of toxic components in RQCJ, which included a variety of heavy metals. Thus, this study was aimed to evaluate the toxicity and expound the toxicological mechanism of RQCJ. In this study, rats were intragastrically administered with different doses of RQCJ for 15 days, and then, the restorative observation period lasted for 15 days. Liver and kidney tissues were collected for histopathological examination, and simultaneously serum and urine samples were collected for 1H nuclear magnetic resonance (1HNMR) spectroscopy analysis and biochemical analysis combined with inductively coupled plasma mass spectrometry (ICP-MS) measurement. The 1HNMR-based metabolomics analysis revealed that the administration of RQCJ significantly altered the concentrations of 14 serum metabolites and 14 urine metabolites, which implied disturbances in energy metabolism, amino acid metabolism, intestinal flora environment, and membrane damage. Besides, the biochemical analysis of serum samples was consistent with the histopathological results, which indicated slight hepatotoxicity and nephrotoxicity. The quantification of As and Hg in urine and serum samples by ICP-MS provided more evidence about the toxicity of RQCJ. This work provided an effective method to systematically and dynamically evaluate the toxicity of RQCJ and suggested that precautions should be taken in the clinic to monitor the potential toxicity of RQCJ.
Renqingchangjue (RQCJ),
a kind of empirical Tibetan prescription,
is a pill composed by hundreds of precious components such as pearl,
cinnabar, sandalwood, agrwood, medlar, bezoar, artificial musk, saffron,
and so on.[1] Up to now, RQCJ has been utilized
to treat chronic gastroenteritis, atrophic gastritis, and gastric
ulcer.[2] However, similar to some other
ethnomedicines, Tibetan medicine also faces severe challenges, for
example, the wide use of heavy metals as pharmaceutical components
and toxic element.[3] As one of essential
components of RQCJ, Tsothel (Zuotai) is a notable metal-containing
Traditional Tibetan Medicine.[4] It is regarded
as the “king of essence” in Tibetan medicine because
it has been considered an important therapeutic ingredient for more
than 1000 years.[1] Tsothel contains 54%
mercuric sulphide (HgS) and is also rich in arsenic (As), plumbum
(Pb), copper (Cu), cadmium (Cd), and so forth.[5,6] It
has been reported that tsothel includes a high content of heavy metals
such asHg and As.[7] Therefore, it is of
great significance to investigate the safety of RQCJ and clarify its
potential mechanisms to assuage public concerns.Metabolomics,
based on the qualitative and quantitative analysis
of small molecule metabolites in biofluids or tissues, is an indispensable
approach to monitor the biochemical changes influenced by endogenous
and exogenous factors such as drug interventions and disease invasions.[8,9] As a kind of systemic approach, metabolomics has demonstrated great
potential in many fields such as toxicology research,[8,10,11] disease diagnosis,[12,13] and mechanism research.[14,15]1H nuclear
magnetic resonance (1HNMR) has proved to be one of the
widely analytical techniques in metabonomics research due to its characteristics
such as nondestruction, high throughput (5–8 min, possibly),[16] minimum sample preparation, and good reproducibility.[16−18] The NMR-based metabolomics approach has been widely applied to reveal
biochemical changes and disturbed pathways in body fluids such as
serum and urine.[19,20] As is reported, serum contains
various endogenous small-molecule metabolites which could reflect
the comprehensive metabolic changes in different levels of human bodies.[21] Simultaneously urine also plays an irreplaceable
role in toxicology research and biomarker screening, as it contains
less protein and lipid and can be accessible with noninvasive approach.[19] Thus, in this research, NMR-based serum and
urine metabolomics were applied to evaluate the toxicity of RQCJ.
In addition, the inductively coupled plasma mass spectrometry (ICP-MS),
which has been proved to be a kind of rapid, simple, accurate, and
reliable method,[22−24] was performed to identify and quantify the content
of heavy metals in serum and urine.In this study, the 1HNMR-based metabolomics approach
was utilized to identify potential biomarkers and reveal a series
of metabolic pathway perturbations in serum and urine from RQCJ-administered
rats. Finally, above results complemented with biochemical and histopathological
examinations and ICP-MS were used to expound the toxicological effects
of RQCJ. It may improve our cognition of potential toxicity of RQCJ
and be beneficial to its safety evaluation and rational application.
Results
Body Weight
Profiling
During the whole experiment,
there were no evident abnormalities in food intake of rats in neither
the normal control (NC) group nor the RQCJ dose groups. The body weights
of rats are recorded in Figure S1. There
was no significant difference in body weight among the four groups
at an early stage of RQCJ administration. However, on the 10th day,
compared with the NC group, there was a slower weight growth in RQCJ
groups, which was slowest on the 15th day. From the 18th to 30th day,
there was a steady weight growing in all groups, while the body weight
in RQCJ group rats decreased with the increase of administrated dose.
Clinical Biochemistry Parameters
Clinical biochemical
results of aspartate aminotransferase (AST), alanine aminotransferase
(ALT), blood ureanitrogen (BUN), and alkaline phosphatase (ALP) are
displayed in Figure and Table S1. Compared with the NC group,
levels of AST in MD and HD groups were increased, and ALT was slightly
elevated, which indicated a mild liver injury induced by RQCJ. Simultaneously,
decreased levels of BUN were found in the RQCJ administered groups,
suggesting the disorder of renal function. The ALP levels showed a
slight decrease between the RQCJ groups and NC group. All of the above
differences were gradually moderate after 15 days of recovery.
Figure 1
Distribution
of AST (A), ALT (B), BUN (C), and ALP (D) in serum
of rats. *p < 0.05 vs NC group. NC: normal control
group, LD: low-dose group, MD: middle-dose group, HD: high-dose group.
Distribution
of AST (A), ALT (B), BUN (C), and ALP (D) in serum
of rats. *p < 0.05 vs NC group. NC: normal control
group, LD: low-dose group, MD: middle-dose group, HD: high-dose group.
Histopathology
Histopathological
examination of liver
and kidney tissues is showed in Figure . There were no apparent histological changes in the
NC group on days 15 and day 30, whereas some morphological changes
such as inflammatory cell infiltration, slight necrosis of hepatocytes,
and mild renal tubular lesions were observed in RQCJ dosed groups
on day 15. The occurred morphological changes in liver and kidney
recovered on the 30th day.
Figure 2
Histopathological examination of liver and kidney
tissues of rats
among the four groups by H&E staining (10×). Renal tubular
lesions (*), inflammatory cell infiltration (↓), and necrosis
of hepatocytes (□). NC: normal control group, LD: low-dose
group, MD: middle-dose group, HD: high-dose group.
Histopathological examination of liver and kidney
tissues of rats
among the four groups by H&E staining (10×). Renal tubular
lesions (*), inflammatory cell infiltration (↓), and necrosis
of hepatocytes (□). NC: normal control group, LD: low-dose
group, MD: middle-dose group, HD: high-dose group.
1H NMR Spectroscopic Analysis of Serum and Urine
Samples
Representative 600 MHz 1HNMR spectra
of serum and urine samples from NC and HD groups on 15th day are displayed
in Figure . Metabolites
were identified based on the reported literature studies,[10,25−30] chenomx NMR Suite database (version 7.5, Chenomx, Canada) and the
Human Metabolome Database database (HMDB, http://www.hmdb.ca/). The identified
metabolites were further confirmed by two-dimensional spectra such
as1H–1H COSY (correlation spectroscopy), 1H–1H TOCSY (total correlation spectroscopy),
and 1H–13C HSQC (heteronuclear single
quantum correlation). Finally, a total of 27 metabolites in serum
and 30 metabolites in urine were identified.
Multivariate Statistical Analysis of Serum 1H NMR
Data
The serum 1HNMR data of the four groups
were subjected into principal component analysis (PCA) to investigate
the differences between all groups. LD group clusters was close to
the NC group while the MD and HD groups were significantly separated
from the controls in the score plots of 15th day (Figure A). After 15 days of recovery,
all RQCJ group clusters were close to the control group, and the separation
was less obvious (Figure B). Partial least squares discriminant analysis (PLS-DA) was
performed to obtain better group separation, and there were same trends
observed in PLS-DA score plots (Figure S2) on 15th day and 30th day, which indicate that the metabolic phenotype
of the rats was disturbed by RQCJ.
Figure 4
PCA score plot of serum (A,B) samples
from the four groups at day
15 and day 30 and PLS-DA score trajectory plots of urine (C) samples
from the four groups at days 0, 3, 6, 9, 12, 15, 18, 23, 26, and 30.
Day 0: one day before treatment, days 3–15: the time points
of administration, days 18–30: the time points of stop administration.
In PLS-DA score trajectory plots, each point represented the mean
position of a group (n = 12). NC: normal control
group, LD: low-dose group, MD: middle-dose group, HD: high-dose group.
PCA score plot of serum (A,B) samples
from the four groups at day
15 and day 30 and PLS-DA score trajectory plots of urine (C) samples
from the four groups at days 0, 3, 6, 9, 12, 15, 18, 23, 26, and 30.
Day 0: one day before treatment, days 3–15: the time points
of administration, days 18–30: the time points of stop administration.
In PLS-DA score trajectory plots, each point represented the mean
position of a group (n = 12). NC: normal control
group, LD: low-dose group, MD: middle-dose group, HD: high-dose group.In order to reveal the differences in metabolites
between two groups,
orthogonal partial least squares discriminant analysis (OPLS-DA) models
were further constructed. In the OPLS-DA score plots (Figures and S3), all RQCJ groups were well separated from the NC group. The OPLS-DA
color-coded coefficient plots (Figures and S3) displayed the metabolites
contributing to the group separation.
Figure 5
OPLS-DA score plots (A,C,E, and G) and
color-coded coefficient
plots (B,D,F, and H) of serum (A–D) and urine (E–H)
samples from the NC and HD groups on days 15 and day 30. Metabolite
variation could be visualized by the color-coded coefficient plots.
The alteration of metabolites between groups were identified according
to the first principal component [t(1)]. The upper
section of the coefficient plots indicated the increased metabolites
of the group in the positive direction of [t(1)],
and vice versa. NC: normal control group, HD: high-dose group.
OPLS-DA score plots (A,C,E, and G) and
color-coded coefficient
plots (B,D,F, and H) of serum (A–D) and urine (E–H)
samples from the NC and HD groups on days 15 and day 30. Metabolite
variation could be visualized by the color-coded coefficient plots.
The alteration of metabolites between groups were identified according
to the first principal component [t(1)]. The upper
section of the coefficient plots indicated the increased metabolites
of the group in the positive direction of [t(1)],
and vice versa. NC: normal control group, HD: high-dose group.Compared with the NC group on the 15th days, increased
levels (upper
section of the coefficient plots) of isoleucine, leucine, valine,
lactate, acetate, proline, methionine, acetoacetate, pyruvate, creatine,
choline, serine, threonine, and decreased levels (lower section of
the coefficient plots) of glucose were determined in serum samples
of the HD group. On the 30th day, the amounts of altered metabolites
in serum of HD group rats were reduced, including higher levels of
creatine and choline and lower levels of isoleucine, leucine, valine,
proline, glucose, and betaine. The results from statistical analysis
of NC versus each RQCJ groups based on serum 1HNMR data
are shown in Table .
Table 1
NMR Signals Assignments of Metabolites
with Significant Changes in Serum on Days 15 and 30, with Their Fold
Change Values, p Values, FDR p Values,
VIP Values and Correlation Coefficient Values (r)
Fold
change values were colored
according to log 2(fold) using color bar: red
and blue indicates increasing and decreasing concentration
of each group compared to NC.
The absolute of correlation coefficient
|r| obtained from PLS-DA model, the cutoff of |r| was set to 0.707.
The p values were
obtained from independent samples t-test. The chemical
shifts in boldface are used in calculating the relevant integrals
and p values (*p < 0.05, **p < 0.01).
Multiplicity: s-singlet; d-doublet;
t-triplet; q-quartet; dd-doublet of doublets; m-multiplet.
The FDR p values
were calculated from p values, the p values with no significant difference (P > 0.05)
were not calculated and expressed by “/”.
Fold
change values were colored
according to log 2(fold) using color bar: red
and blue indicates increasing and decreasing concentration
of each group compared to NC.The absolute of correlation coefficient
|r| obtained from PLS-DA model, the cutoff of |r| was set to 0.707.The p values were
obtained from independent samples t-test. The chemical
shifts in boldface are used in calculating the relevant integrals
and p values (*p < 0.05, **p < 0.01).Multiplicity: s-singlet; d-doublet;
t-triplet; q-quartet; dd-doublet of doublets; m-multiplet.The FDR p values
were calculated from p values, the p values with no significant difference (P > 0.05)
were not calculated and expressed by “/”.
Multivariate Statistical Analysis of Urine 1H NMR
Data
In order to dynamically observe the changes in urine
metabolism profile of rats, PCA was performed in urine 1HNMR data at days of 0, 3, 6, 9, 12, 15, 18, 23, 26, and 30. In
the PCA score plots (Figure S4), the separation
between RQCJ groups and the NC group were gradually obvious from day
1 to day 15. Then, from the 15th day, each dosed group cluster was
close to the NC group and could not be easily distinguished from the
NC group, especially on the 30th day. The dynamically metabolic changes
among 30 days are displayed by the PLS-DA score trajectory plots (Figure C), where each point
represented the mean position of a group. It could be seen that all
RQCJ groups were markedly separated from the control group, especially
the high dose group which was away from normal levels and started
to recover on the 18th day. The general recoveries of other two RQCJ
groups were found around the 15th day.Based on OPLS-DA score
plots, the altered metabolites in urine were further selected by using
the coefficient plots (Figures and S5). Compared with the NC
group, in the HD group there were the increased levels of leucine,
valine, lactate, taurine, glycine, betaine, hippurate, fumarate, and
benzoate (upper section of the coefficient plots) and decreased levels
of 2-oxoglutarate, phosphocholine, creatinine, 1-methylnicotinamide,
and formate (lower section of the coefficient plots) in urine samples
on the 15th day. After a 15 day recovery, in the HD group, there were
still an increased level of hippurate and benzoate and decreased levels
of leucine, succinate, citrate, malonate, and TMAO in urine. Based
on urine 1HNMR data, the results from statistical analysis
of NC versus RQCJ groups are shown in Table .
Table 2
NMR Signals Assignments
of Metabolites
with Significant Changes in Urine on days 15 and 30, with Their Fold
Change Values, p Values, FDR p Values,
VIP Values and Correlation Coefficient Values (r)a,b,c,d
Fold
change values were colored
according to log 2(fold) using color bar: red
and blue indicates increasing and decreasing concentration
of each group compared to NC.
The absolute of correlation coefficient
|r| obtained from PLS-DA model, the cutoff of |r| was set to 0.707.
The p values were
obtained from independent samples t-test. The chemical
shifts in boldface are used in calculating the relevant integrals
and p values (*p < 0.05, **p < 0.01).
Multiplicity: s-singlet; d-doublet;
t-triplet; q-quartet; dd-doublet of doublets; m-multiplet.
The FDR p values
were calculated from p values, the p values with no significant difference (P > 0.05)
were not calculated and expressed by “/”.
Fold
change values were colored
according to log 2(fold) using color bar: red
and blue indicates increasing and decreasing concentration
of each group compared to NC.The absolute of correlation coefficient
|r| obtained from PLS-DA model, the cutoff of |r| was set to 0.707.The p values were
obtained from independent samples t-test. The chemical
shifts in boldface are used in calculating the relevant integrals
and p values (*p < 0.05, **p < 0.01).Multiplicity: s-singlet; d-doublet;
t-triplet; q-quartet; dd-doublet of doublets; m-multiplet.The FDR p values
were calculated from p values, the p values with no significant difference (P > 0.05)
were not calculated and expressed by “/”.
Pathway Analysis
To investigate
the relevant metabolic
pathways disturbed by RQCJ, the selected potential biomarkers were
analyzed by MetaboAnalyst 4.0. The details of all matched pathways
are showed in Figure and Tables S2 and S3. Among them, based on impact value >0.1, pathways were
considered closely related to toxicity induced by RQCJ. Consequently,
nine pathways including valine, leucine, and isoleucine biosynthesis,
synthesis and degradation of ketone bodies, methane metabolism, glycine,
serine, and threonine metabolism, pyruvate metabolism, aminoacyl-tRNA
biosynthesis, cysteine and methionine metabolism, glycolysis or gluconeogenesis,
and butanoate metabolism in serum samples were selected, and five
pathways in urine samples including valine, leucine, and isoleucine
biosynthesis, taurine and hypotaurine metabolism, glycine, serine,
and threonine metabolism, nicotinate and nicotinamide metabolism,
and glyoxylate and dicarboxylate metabolism were selected. Metscape
is a plugin of Cytoscape for visualization and interpretation of metabolomics
data.[31] It allows users to track the connections
between metabolites and genes, visualize compound networks, and display
pathway information as well as reaction, enzyme, gene, and so forth.[32] To reveal the biochemical relationship associated
with these potential biomarkers, the compound networks analysis results
are shown in Figures S6 and S7. The network
also reflected complex biochemical relationships and provided evidence
for the involvement of altered metabolism in serum and urine influenced
by RQCJ.
Figure 6
Metabolic pathway analysis of potential biomarkers in serum (A)
and urine (B) serum: (1) valine, leucine and isoleucine biosynthesis;
(2) synthesis and degradation of ketone bodies; (3) methane metabolism;
(4) glycine, serine and threonine metabolism; (5) pyruvate metabolism;
(6) aminoacyl-tRNA biosynthesis; (7) cysteine and methionine metabolism;
(8) glycolysis or gluconeogenesis; (9) butanoate metabolism. Urine:
(1) valine, leucine and isoleucine biosynthesis; (2) taurine and hypotaurine
metabolism; (3) glycine, serine and threonine metabolism; (4) nicotinate
and nicotinamide metabolism; (5) glyoxylate and dicarboxylate metabolism.
Metabolic pathway analysis of potential biomarkers in serum (A)
and urine (B) serum: (1) valine, leucine and isoleucine biosynthesis;
(2) synthesis and degradation of ketone bodies; (3) methane metabolism;
(4) glycine, serine and threonine metabolism; (5) pyruvate metabolism;
(6) aminoacyl-tRNA biosynthesis; (7) cysteine and methionine metabolism;
(8) glycolysis or gluconeogenesis; (9) butanoate metabolism. Urine:
(1) valine, leucine and isoleucine biosynthesis; (2) taurine and hypotaurine
metabolism; (3) glycine, serine and threonine metabolism; (4) nicotinate
and nicotinamide metabolism; (5) glyoxylate and dicarboxylate metabolism.
Quantification of As and Hg in Urine and
Serum Samples by ICP-MS
The presence of heavy metal in RQCJ
has raised public concerns.
Even low metal concentrations may threaten the health of people.[33] Therefore, it was necessary to determinate the
heavy-metal contents in biofluid to evaluate the toxicological effects
of RQCJ. There were significant dose-dependent changes of As and Hg
in serum (Figure A,B)
and urine (Figure C,D) in RQCJ groups. On the 15th day, the concentrations of As and
Hg in serum (Figure and Table S4) samples of RQCJ groups
increased, especially in the HD group which was much higher than the
NC group. After a 15 day recovery, reduced levels of Hg and a slightly
decreasing of As were shown in serum samples of RQCJ groups. The dynamic
changes of As and Hg content in rats’ urine are recorded by Figure C,D and Table S5. From day 0 to 15, the contents of As
and Hg in urine of RQCJ groups raised gradually, which peaked the
highest level on the 15th day. Compared with the NC group, the RQCJ
groups showed reduced urinary levels of As and Hg from the 18th day.
Figure 7
Content
of As and Hg in rats serum (A,B) and urine samples (C,D)
from the four groups. *p < 0.05, **p < 0.01 vs control group. NC: normal control group, LD: low-dose
group, MD: middle-dose group, HD: high-dose group.
Content
of As and Hg in rats serum (A,B) and urine samples (C,D)
from the four groups. *p < 0.05, **p < 0.01 vs control group. NC: normal control group, LD: low-dose
group, MD: middle-dose group, HD: high-dose group.
Discussion
In this study, the toxic effects of RQCJ
were evaluated by 1HNMR-based metabolomics methods combined
with serum biochemistry,
histopathology, and ICP-MS. There were increased levels of AST and
ALT and decreased BUN in RQCJ groups from serum biochemistry results,
which was consistent with the results of histopathological examination:
slight necrosis, dilation of hepatocytes, and mild renal tubular lesions
in liver samples. Meanwhile, the ICP-MS results indicated dose-dependent
changes of As and Hg in urine and serum from RQCJ groups. The statistical
analysis of 1HNMR metabolomic profiles of urine and serum
samples revealed potential biomarkers and the relevant metabolic pathways
induced by RQCJ, such as disturbances in energy metabolism and amino
acid metabolism, imbalance of intestinal flora environment and membrane
damage, and injury of liver and kidney.Compared with the NC
group, there were lower levels of 2-oxoglutarate,
phosphocholine, creatinine, 1-methylnicotinamide, and formate and
higher levels of leucine, valine, lactate, taurine, glycine, betaine,
hippurate, fumarate, and benzoate in the urine samples of RQCJ groups.
In addition, compared with the NC group, levels of isoleucine, leucine,
valine, lactate, acetate, proline, methionine, acetoacetate, pyruvate,
creatine, choline, serine, and threonine increased significantly while
levels of glucose was decreased in serum samples. The disturbed metabolic
pathways are shown in Figure .
Figure 8
Schematic diagram of the disturbed metabolic pathways related to
toxicity induced by RQCJ, showing the interrelationship of the identified
metabolic pathways.
Schematic diagram of the disturbed metabolic pathways related to
toxicity induced by RQCJ, showing the interrelationship of the identified
metabolic pathways.Fumarate and 2-oxoglutarate,
regarded as important intermediates
of the tricarboxylic acid cycle (TCA cycle), were reported associated
with energy supply.[34] The level of 2-oxoglutarate
was decreased and fumarate was increased in urine samples from RQCJ
groups, suggesting the imbalance of the TCA cycle was caused by the
mitochondrial dysfunction and insufficient energy supply.[35] On this condition, glycolysis was used for energy
production. On account of the decreased level of glucose and increased
levels of pyruvate and lactate which could be metabolized from glucose,
energy provided from aerobic metabolism could not meet the body needs.
As a core constituent of ketone bodies, acetoacetate is formed according
to incomplete oxidation of fatty acid.[36] When the lipolysis was enhanced, the long chain fatty acyl-CoA increased
and accumulated. In mitochondria, acetyl-CoA, especially long-chain
acyl-CoA, increased and inhibited citric acid synthase by isomerization,
which made it difficult for acetyl-CoA to enter TCA cyclic oxidation
and condenses acetyl-CoA accumulated in liver to form ketones.[37] Increased levels of acetate and acetoacetate
in serum samples of RQCJ-treated groups suggested energy metabolism
disorders which might be related to the increased levels of ketone
bodies.Branched chain amino acids (BCAAs), including valine,
leucine,
and isoleucine, are mainly produced by daily intake of protein foods.[38] The levels of BCAAs were upregulated to a certain
extent in serum samples of RQCJ treated rats. On the one hand, it
might indicate an inhibition of the synthesis of proteins, branched
fatty acids, and neurotransmitters, which induced the accumulation
of valine, leucine, and isoleucine. On the other hand, the feedback
against Hgtoxicity could promote leucine and isoleucine to synthesize
oxidoreductases or DNA-repair enzymes for repairing oxidative damage
caused by Hg,[39] which might contribute
to the upregulation of BCAA levels in RQCJ treated rats. As is known,
BCAAs could be converted to asparagine and further produce aspartate.
Aspartate was a synthetic precursor of lysine, threonine, methionine,
and so forth. Accompanied by the increased levels of BCAAs, asparagine
was also upregulated which might affect the metabolic processes of
threonine and methionine. It has been reported that the increased
level of methionine might result from the intake of As.[40] Hence, the upregulation of methionine supported
that the concentrations of As in RQCJ treated groups were much higher
than the NC group.As a nonessential amino acid, serine is a
precursor of synthetic
purine, thymine, and choline.[41] In addition,
serine is used for promoting the metabolism of fats and fatty acids
and maintaining the normal function of immune system.[42] If necessary, serine will be synthesized from glycine.
However, excessive accumulation of serine might cause immune suppression
and psychological symptoms, either.[43] In
this study, the increased serum concentration of serine and urine
concentration of glycine provided evidence for adverse effects of
RQCJ on immune system.Creatinine is produced by the catabolism
of creatine and is excreted
by the kidney. Both creatinine and creatine are associated with the
processes of ATP delivery and consumption.[44] In this study, the levels of creatinine decreased in the urine after
RQCJ intervention. At the same time, a decreased level of 1-methylnicotinamide
was detected, which was considered as a sign of tubular dysfunction.[45] These results indicated the dysfunction of energy
metabolism and renal function in RQCJ groups.The level of choline
was increased in serum, and phosphocholine
was decreased in urine of RQCJ groups. As the major component of the
cell membrane, choline can be produced through decomposition of phosphatidylcholine
and plays an important role in maintaining cell integrity.[46] An increasing serum level of choline in RQCJ
groups suggested the impairment of membrane in fluidity, which had
been verified in the previous studies on the toxicityassessment of
Arisaematis Rhizoma in rats.[41] Meanwhile,
it has been reported that As could cause membrane damage by destroying
the membrane structure and increasing membrane permeability.[47] Thus, high levels of As in serum of RQCJ groups
also supported membrane damage. Besides, as a metabolite of choline,
the increased level of betaine might also indicate disorders in choline
metabolism.There are several important physiological functions
of Taurine,
including promoting glycolipid metabolism, antioxidation accompanied
with regulating osmotic pressure, and so forth.[27] Taurine is considered as a specific marker of hepatotoxicity[48] for it is produced by the liver in response
to a toxic injury and subsequently released from damaged cells. Thus,
the increased level of taurine in urine indicated the liver dysfunction
induced by RQCJ, consistent with the histological examination and
clinical chemistry results.Hippurate is synthesized from benzoate
and glycine in the liver.
Moreover, benzoate is a metabolite of intestinal microbiota.[49] Therefore, an increase of the hippurate level
in urine implied a disruption or imbalance of intestinal microbiota.
According to the previous study, urinary hippurate synthesis was considered
to be influenced by metabolites of As.[50] In this study, the elevated hippurate level in urine might support
the high concentration of As.After 15 days of recovery, there
was an obvious trend of resumption
for most of the altered metabolites in both serum and urine. The restored
metabolites mainly included acetoacetate, 3-HB, valine, lysine, threonine,
methionine, serine, and taurine. However, there were still remained
some significantly altered metabolites, which were involved in the
disturbances of energy metabolism, perturbation of intestinal microenvironment,
membrane damage, and so forth.
Conclusions
In the present study,
the 1HNMR-based metabonomics
approach combined with histopathology, clinical biochemistry, and
ICP-MS were applied to the systemic evaluation of RQCJ toxicity on
rats. These findings demonstrated that the administration of RQCJ
might induce obvious metabolic changes in serum and urine samples,
which is related to perturbation in energy metabolism, amino acid
metabolism, gut microbiota environment, membrane damage, and so forth.
In addition, long-term administration of RQCJ might induce slight
and reversible hepatotoxicity and nephrotoxicity. In addition, As
and Hg were detected at a high level both in serum and urine after
RQCJ administration, which might be caused by toxic constituents of
RQCJ. Therefore, precautions should be taken in the clinic to monitor
the potential toxicity of RQCJ. Our present study explored the toxicity
of RQCJ from a systematic and holistic perspective and provided data
support for the application of RQCJ in clinic.
Experimental Procedures
Chemicals
RQCJ was obtained from Tso-Ngon Tibetan Medicine
Hospital (QingHai, China). Analytical pure K2HPO4 and NaH2PO4 were purchased from Sigma-Aldrich
(St. Louis, MO, U.S.A.). Deuterium oxide (D2O, 99.9% D)
and 3-(trimethylsilyl)-propionic-2,2,3,3-d4 acid sodium salt (TSP, 98.0% D) were purchased from Cambridge Isotope
Laboratories, Inc. (MA, U.S.A.). Standard stock solutions of mercury
and arsenic were obtained from the Central Iron & Steel Research
Institute (Beijing, China).
Animals
Forty-eight SPF Sprague-Dawley
male rats (200–220
g, 7 weeks old, Rodent license no. SCXK 2011-0004) were purchased
from Sibeifu Laboratory Animal Technology Co., Ltd. (Beijing, China).
The experimental animals were housed at 22 ± 2 °C and the
relative humidity was 55 ± 5%, with a 12/12 h light–dark
cycles. All rats were acclimatized for a week before administration.
The animal experiments were approved by the Animal Ethics Committee
of Beijing University of Chinese Medicine and were in strict accordance
with the guidelines for Care and Use of Laboratory Animals.
Drug Administration
and Sample Collection
Rats were
randomly divided into four groups (n = 12): NC group,
low-dose (LD) group, middle-dose (MD) group, and high-dose (HD) group.
The NC group was intragastrically administered with normal saline
(1 mL/100 g). LD, MD, and HD groups were intragastrically administered
with equal volume RQCJ at a dose of 250.0, 666.7, and 1666.7 mg·kg–1, respectively. The experimental doses were equivalent
to 15, 40, and 100 times of the human (60 kg adults) clinical dose
(1 g per human per day, Chinese Pharmacopoeia, version 2015). Rats
were intragastrically administered for 15 consecutive days and then
recovered for 15 days. The total experiment lasted for 30 days.On the 15th day, half of rats in each group were sacrificed, while
the rest half were sacrificed on the 30th day. The blood samples were
collected and separated by centrifugation (13,000 rpm, 4 °C,
15 min). Urine samples were collected in ice-cooled tubes containing
50 μL sodium azide (1%, w/v) at day 0 (one day before treatment),
days 3, 6, 9, 12, and 15 (the time points of administration) and days
18, 23, 26, and 30 (the time points of stop administration). All of
the samples were stored at −80 °C for further 1HNMR spectroscopy analysis, serum biochemical and ICP-MS determination.
The liver and kidney tissues were immediately removed from rats, rinsed
with cold phosphate buffer saline, and immersed in 10% neutral-buffered
formalin solution for histopathological examination.
Clinical Biochemistry
and Histopathology
The clinical
biochemical assays of serum samples were performed by an Olympus 2700
biochemistry analyzer (Olympus Co., Japan). Biochemical parameters
included serum ALT, AST, ALP, and serum and ureanitrogen (BUN).The formalin-fixed liver and kidney slices were dehydrated, embedded
in paraffin, and then cut into 5 μm sections. The sections were
stained with hematoxylin and eosin (H&E) for microscopic observation.
Sample Preparation for NMR Measurement
The serum and
urine sample were thawed at room temperature. For serum, each 200
μL of sample was mixed with 400 μL of deuterated phosphate
buffer (NaH2PO4/K2HPO4, 0.045 mol/L, pH = 7.47). For urine, 55 μL of deuterated phosphate
buffer (NaH2PO4/K2HPO4, 1.5 mol/L, containing 0.01% TSP, pH = 7.47) was added to 550 μL
of sample. All of the mixtures were allowed to stand at room temperature
for 5 min and then centrifuged for 15 min to remove precipitates (13,000
rpm, 4 °C). The supernatant (550 μL) was transferred into
a 5 mm NMR tube and maintained at 4 °C until NMR measurement.
1H NMR Spectroscopy Analysis of Urine and Serum Samples
NMR spectra of the serum and urine samples were acquired on a VARIAN
VNMRS 600 MHz NMR spectrometer (Varian Inc, Palo Alto, Calif) equipped
with a 5 mm inverse-proton (HX) triple resonance probe and z-axis gradient coil, operated at 1H frequency
of 599.901 MHz and a temperature of 298 K. The urine spectra were
measured using the following parameters: nuclear overhauser enhancement
spectroscopy (NOESY) pulse sequence [relaxation delay (RD)–90°–t1–90°–tm–90°–acquisition] with water suppression
[RD = 2.0 s, mixing time (tm) = 100 ms,
short delay (t1) = 4 μs]. The 90°
pulse length was adjusted to approximately 10 μs, 64 K data
points were obtained form 128 transients using a spectral width of
20 ppm. Prior to fourier transformation, the free induction decays
(FIDs) were weighted by an exponential function with a line-broadening
factor of 0.5 Hz. The serum spectra were measured using the following
parameters: the water-suppressed standard Carr–Purcell–Meibom–Gill
pulse sequence (RD −90°–(τ–180°−τ)–acquisition) with a fixed total spin–spin
RD 2nτ of 320 ms. Weakening the broad NMR signals
from macromolecules (such as proteins) by the sequence, the signals
of micromolecules were obviously observed. The FIDs were collected
into 64 K data points over a spectral width of 12,000 Hz with 128
scans. The FIDs were zero-filled to double size and multiplied by
an exponential line-broadening factor of 0.5 Hz before fourier transformation.
ICP-MS Measurement of Urine and Serum Samples
Urine
and serum samples were thawed at room temperature and then digested
with HNO3 + H2O2 (5:1) dissolution
system using a modified acid digestion method on microwave digestion
instrument (CEM Co., MARS). The digested solution was then detected
by ICP-MS (Agilent Technologies, 7500 Ce) for determination of the
tracer elements of 75As and 202Hg. The Hg and
As standard solutions (calibration curve, 0.1, 1, 10, and 100 ng/mL)
were obtained by dilution 100 mg/L standard stock solutions with ultrapure
deionized water (containing 5% (w/w) HNO3). Quantification
was based on 72germanium (72Ge) as internal
standards for As and 209bismuth (209Bi) for
Hg.
Data Processing
All of the 1HNMR spectra
of serum and urine were processed using MestReNova7.1.2 software (Mestrelab
Research SL, Spain). Each spectrum manually phased adjustment and
baseline corrected. Shift correction is based on lactate at δ
1.33 (serum samples) or TSP at δ 0.00 (urine samples). Then,
the region of δ 9.0–0.5 in serum spectra and δ
9.5–0.5 in urine spectra were binned into 0.002 ppm integrated
spectral buckets. In order to eliminate the influence of the residual
water signals, the δ 5.2–4.7 region both in urine spectra
and serum spectra were removed. For the urine spectra, the region
of δ 6.0–5.45 was also excluded to avoid the effects
of urea signals. The remaining integrated regions were normalized
to total area of the spectra.The normalized NMR data were imported
into SIMCA-P+12.0 (Umetrics Inc, Sweden) for multivariate statistical
analysis. PCA was performed using a mean-centered method, which can
reflect the inherent clustering trends and determine the outliers.
The analysis results were visualized using scores plots. The normalized
data were subjected to unit variance-scaled approach for PLS-DA and
OPLS-DA. PLS-DA was utilized to analyze the relationship between NMR
data (X variable) and grouping variable (Y variable) and obtain variables that contributed most to
the separation (variable projection importance (VIP) > 1). The
quality
and validity of PLS-DA model were assessed using a sevenfold cross-validation
method and 200 permutation tests, while R2 and Q2 were utilized to indicate the
interpretability of variables and predictability of the models. OPLS-DA
was utilized to maximize the difference between the groups, and the
coefficient plot corresponding to OPLS-DA model was generated using
MATLAB R2012a software (Version 7.1, Mathworks Inc, USA) combined
with self-programming. The coefficient plots were color-coded according
to the absolute value of coefficients (r), and the
metabolites with hot color (red) contributed more to the group separation
than the ones with cold color (blue). Then, the independent sample t-test was followed by multivariate statistical analysis
to further compare the significant difference in selected metabolites
via SPSS 17.0 (SPSS Inc, USA), and the significance threshold was
set to P < 0.05. In this research, metabolites
were identified as potential biomarkers on the basis of correlation
coefficients (|r| > 0.707, n =
12),
variable importance (VIP > 1) and statistically significant (P < 0.05). The identified potential biomarkers were analyzed
for metabolic pathway using the MetaboAnalyst 4.0 (http://www.metaboanalyst.ca). In order to visualize the metabolic network, we used MetScape
(http://metscape.ncibi.org./) to integrate the compound network.
Authors: Hector C Keun; Timothy M D Ebbels; Henrik Antti; Mary E Bollard; Olaf Beckonert; Götz Schlotterbeck; Hans Senn; Urs Niederhauser; Elaine Holmes; John C Lindon; Jeremy K Nicholson Journal: Chem Res Toxicol Date: 2002-11 Impact factor: 3.739
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8