Abstract
Wuwei Shexiang Pill (WSP) is a Tibetan traditional medicine, which has
been demonstrated to exhibit potent anti-inflammatory and anti-gout
effects. However, the specific pharmacological mechanism is not
elucidated clearly. In the present study, liquid chromatography-mass
spectrometry (LC-MS)-based metabolomics was applied to investigate the
alteration of serum metabolites induced by WSP treatment in MSU-induced
gouty rats. Subsequently, bioinformatics was utilized to analyze the
potential metabolic pathway of the anti-gout effect of WSP. The
pharmacodynamic data discovered that WSP could ameliorate ankle
swelling and inflammatory cell infiltration, as well as downregulate
the protein expression of IL-1β, p-NF-κB p65, and NLRP3 in the synovial
membrane and surrounding tissues of gouty ankles. LC-MS-based
metabolomics revealed that there were 30 differential metabolites in
the serum between sham-operated rats and gouty ones, which were mainly
involved in the metabolism of fructose and mannose, primary bile acid
biosynthesis, and cholesterol metabolism. However, compared to the
model group, WSP treatment upregulated 11 metabolic biomarkers and
downregulated 31 biomarkers in the serum. KEGG enrichment analysis
found that 27 metabolic pathways contributed to the therapeutic action
of WSP, including linoleic acid metabolism, phenylalanine metabolism,
and pantothenate and CoA biosynthesis. The comprehensive
analysis-combined network pharmacology and metabolomics further
revealed that the regulatory network of WSP against gout might be
attributed to 11 metabolites, 7 metabolic pathways, 39 targets, and 49
active ingredients of WSP. In conclusion, WSP could ameliorate the
inflammation of the ankle in MSU-induced gouty rats, and its anti-gout
mechanism might be relevant to the modulation of multiple metabolic
pathways, such as linoleic acid metabolism, phenylalanine metabolism,
and pantothenate and CoA biosynthesis. This study provided data support
for the secondary development of Chinese traditional patent medicine.
Keywords: Wuwei Shexiang pill (WSP), gout, serum metabolomics, linoleic
acid metabolism, phenylalanine metabolism, pantothenate and CoA
biosynthesis
1 Introduction
Gout is a common inflammatory arthritis that is caused by monosodium
urate (MSU) crystal deposition in the articular structure of the joint
of a person who suffers from hyperuricemia (HUA) HUA is a metabolic
disorder characterized by increased production and/or decreased
excretion of serum uric acid due to purine metabolism dysfunction
([48]FitzGerald et al., 2020). Then, long-term high levels of uric acid
in the blood may lead to the precipitation and deposition of MSU
crystals, triggering an innate immune response and intolerably painful
arthritis, which is known as gout flares. Clinically, gouty arthritis
(GA) is characterized by redness, swelling, and pain in the lower limb
joints, especially in the first metatarsophalangeal joint, which can
even affect physical function ([49]Dalbeth et al., 2021). Although most
patients with HUA will not develop gout in their lifetime, the
incidence rate of gout is rising year by year and showing a trend in a
younger age group ([50]Danve and Neogi, 2020; [51]Rao et al., 2022).
Unfortunately, the development of medical strategies for treating gout
is unsatisfactory. First-line anti-inflammatory and uric acid-lowering
drugs have shown various side effects and safety problems ([52]Jansen
et al., 2004; [53]MacDonald et al., 2014; [54]Wechalekar et al., 2014;
[55]Ragab et al., 2017). A promising small molecule compound MCC950,
which specifically targets NLRP3, was suspended in phase II clinical
trials because of its liver toxicity ([56]Mangan et al., 2018). Several
IL-1 antagonists (Anakinra, Canakinumab, and Rilonacept) have been
progressively applied in the treatment of refractory and recurrent
gouty arthritis, which was recommended by the European League Against
Rheumatism (EULAR) and the American College of Rheumatology (ACR)
([57]Arnold et al., 2022). However, these drugs are not yet available
in China. Hence, research for safer and more reliable anti-gout drugs
is imperative.
Traditional Chinese medicine (TCM) has received widespread attention
and is used clinically due to its preferable effectiveness and safety
([58]Wen et al., 2021; [59]Zhou et al., 2022). Chinese traditional
patent medicine Wuwei Shexiang pill (WSP), composed of Terminalia
chebula Retz., Aconitum pendulum Busch, Aucklandia lappa Decne., Acorus
calamus L., and artificial musk, has been used to treat various types
of arthritis in Tibetan-populated areas of China for centuries
([60]Chinese Pharmacopoeia Commission, 2020). Recently, WSP has also
been reported to significantly reduce serum uric acid levels and
inhibit ear swelling in mice ([61]Yamin and Lvyi, 2017). In our
previous study, we confirmed that WSP exerted an anti-gout effect by
inhibiting MAPK, NF-κB, and NLRP3 signaling pathways in MSU-stimulated
THP-1 macrophages ([62]Lang et al., 2022). Although our research has
unveiled the anti-inflammatory effect and mechanism in vitro, the
anti-gout pharmacodynamic activity and the metabolomic profiles of WSP
in vivo have not been reported.
Metabolomics is a systematic approach to analyzing the small molecules
metabolites in biological samples and unveiling the underlying
mechanisms of altered endogenous metabolites in physiological and
pathological states or stimulated by exogenous substances ([63]Fu et
al., 2023). Untargeted metabolomics is used to analyze changes in
biological endogenous metabolites and enrich for metabolic pathways via
high-throughput assays, which is widely used in the fields of
pharmacology, drug toxicology, modernization of Chinese medicine, and
so on ([64]Li et al., 2021; [65]Wang and Sun, 2022; [66]Wang et al.,
2023; [67]Huang et al., 2023). Recently, a large number of works of
literature have documented that metabolomics was a novel and
high-performance strategy to screen the predictive biomarkers of gout
flares and to unveil the molecular mechanism of anti-GA TCMs ([68]Huang
et al., 2019; [69]Lyu et al., 2019; [70]Wang and Sun, 2022; [71]Gu et
al., 2023; [72]Lei et al., 2023). Huang et al. proposed that the
metabolomics signatures of gout sufferings in the serum are mainly
comprised of several metabolic pathways, including purine metabolism,
branched-chain amino acids metabolism, and the tricarboxylic acid cycle
and bile secretion and arachidonic acid metabolism ([73]Huang et al.,
2020). Notably, in consideration of the sensitivity of LC-MS-based
metabolomics, the subtle changes in metabolites of biospecimen can be
detected, which is conducive to revealing the metabolic pathways of
TCMs.
In this study, a rat model of gouty arthritis induced by
intra-articular injection of MSU was adopted to evaluate the anti-gout
effect of WSP. Subsequently, the metabolomic profile on the effect of
WSP in the GA rats was monitored by LC-MS-based metabolomics, thus
revealing the anti-GA metabolomic pathways and the molecular mechanisms
of WSP.
2 Materials and methods
2.1 Regents and drugs
All the materials were obtained from the suppliers as follows: PBS
(Lot#8122153) (Gibco, NY, United States); uric acid (Lot#BCCC5658)
(Sigma-Aldrich, Darmstadt, GER); Etoricoxib tablets (Lot#U039558)
(J20180059) (Merck and Co., Inc., NJ, United States); WSP (Lot#2205001)
(Z51020967) (Jiuzhaigou Natural Pharmaceutical Group Co., Ltd., Aba,
CHN); 2-Chlorophenylalanine (Lot#20211126), Ammonium formate
(Lot#T2122203) (Aladdin, Shanghai, CHN); Acetonitrile (Lot#R142221)
(Dikma, CA, United States); formic acid (Lot#20221214) (TCI, Shanghai,
China); Anti-IL-1β antibody (Lot#BB09202951), Anti-p-NF-κB P65 antibody
(Lot# BB05301615), and Anti-NLRP3 antibody (Lot#BA12271673) (Bioss,
Beijing, CHN).
Preparation of MSU suspension: A total measurement of 168.1 mg of uric
acid and 818.6 mg of NaCl was dissolved in 100 mL of boiling water, and
1.25 mL of 1 mol/L NaOH was added, heated, and stirred until fully
dissolved. The solution was left at room temperature for 2–3 days. The
MSU crystal was obtained after filtering and drying and was sterilized
at 180°C for 2 h. Subsequently, it was weighed in a clean area and
prepared with sterile PBS as an MSU suspension (12 mg/mL).
2.2 Animals and treatments
A total of 66 male-specific pathogen-free Sprague-Dawley rats
(6∼8 weeks, 280∼300 g) were purchased from Hunan Slake Jingda
Experimental Animal Co., Ltd (License number: SCXK (Xiang) 2019-0004).
After 1 week of acclimatization, the rats were randomly divided into
six groups including the control group (CON), model group (MOD), WSP
low-dosage group (10 mg/kg, WSP-L), WSP medium-dosage group (20 mg/kg,
WSP-M), WSP high-dosage group (40 mg/kg, WSP-H), and Etoricoxib group
(32 mg/kg, ETO), which were orally administrated vehicle or drugs once
a day for 7 days, respectively. All the animals were housed under the
standard condition with regulated temperature and humidity with an
alternating 12-h light/12-h dark cycle, allowing them to take food and
water freely. After the experiment, the rats were sacrificed by
asphyxiation with CO[2] (KW-AL experimental animal euthanasia device,
Nanjing Calvin Biotech. Co., Ltd., Nanjing, CHN), which was delivered
into the cage for less than 5 psi per second. The death of the rats was
confirmed by a lack of respiration and consciousness. All the
experimental procedures were approved by the Experimental Animal Ethics
Committee of the Sichuan Academy of Chinese Traditional Medicine [Grant
number: SYLL (2021)-031].
2.3 Acute gouty arthritis model
After 1 h of administration on D 6, the animals were anesthetized with
1% sodium pentobarbital solution (40 mg/kg) i. p. When the absence of
the righting reflex and stable respiration was observed, acute gouty
arthritis in the rats was established according to Coderre’s approach
([74]Coderre and Wall, 1987). Briefly, a syringe was inserted from the
dorsum of the animal’s left ankle joint, and 0.1 mL of the prepared MSU
suspension (12 mg/mL) was injected into the articular cavity to create
the GA model, while the CON group was administrated with 0.1 mL sterile
PBS. After 1 h of administration on the last day, rats were
anesthetized with 1% pentobarbital sodium solution (40 mg/kg) i. p.,
and the blood was collected through the abdominal aorta, which was
allowed to stand at a room temperature for 30 min. The serum was
obtained by centrifugation at 4°C, 1500× g for 10 min, dispensed into
sterile tubes, and quickly transferred to a −80°C refrigerator after
touching the bottom of the tubes to change to liquid nitrogen for quick
freezing. After the animals were sacrificed using CO[2], they were
placed on an ice plate to quickly separate the joint ankle and
surrounding tissue samples and were put into 10% paraformaldehyde for
fixation.
2.4 Measurement of foot swelling
A line was drawn approximately 1 cm above the left ankle joint of the
rats before modeling, and the foot was measured and recorded along the
line using a foot measuring instrument (TECHMAN, Chengdu, CHN) at 0 h,
2 h, 4 h, 6 h, 8 h, and 24 h. Foot swelling was calculated as follows:
[MATH: Foot swelling mL=Foot volume after modeling mL−Foot volume at 0 h mL
:MATH]
2.5 Histopathology
The ankle joint and surrounding tissues were dehydrated, embedded in
paraffin, and cut into 3 μm thick slices. Slice samples were stained by
hematoxylin-eosin staining, and the pathological changes of the ankle
joint, synovial membrane, and surrounding tissues were observed under a
light microscope (Nikon, Tokyo, JPN). At least three areas were
randomly selected to be photographed. The pathologists who were blinded
to the experiment uniformly scored each photo, and the scoring criteria
which was set by our research team are shown in [75]Table 1,
([76]Zhu-jun et al., 2021).
TABLE 1.
Pathology scoring criteria.
Grade Description
0 The surface of the cartilage is flat and intact, with a clear
hierarchy of synovial tissue and no inflammatory cell infiltration in
the synovium and its surrounding tissue
1 The surface of the cartilage is discontinuous, the synovial tissue is
edematous, and the capillaries are dilated, congested, and infiltrated
with a few inflammatory cells
2 The surface of the cartilage is damaged and the synovial tissue is
edematous, congested, and infiltrated with a large number of
inflammatory cells
3 The cartilage is severely damaged and the membrane tissue structure
is disturbed, edematous, congested, and infiltrated with a large number
of inflammatory cells
[77]Open in a new tab
2.6 Immunohistochemistry
The ankle joint thick slices were dewaxed, antigen-retrieved, immersed
in 3% H[2]O[2], and incubated for 25 min at room temperature and
protected from light. Then, the tissues were blocked in 3% BSA for
25 min. Subsequently, primary antibodies (p-NF-κB p65, NLRP3, and
IL-1β, dilution: 1:100) were added to the tissues overnight at 4°C
Secondary antibody was added for incubation for 50 min at room
temperature. DAB solution was used for color reaction; the positive
protein expression was brown. Hematoxylin was used for the staining of
the nucleus, which showed as blue. The slices were observed under a
light microscope (Nikon, Tokyo, JPN) after dehydration and sealing. At
least three areas were randomly selected to be photographed.
Immunohistochemical images were processed by ImageJ software (NIH and
LOCI, United States) to calculate the positive areas.
2.7 Metabolomics analysis
2.7.1 Metabolic biomarkers collection
The serum samples were taken from the liquid nitrogen and slowly melted
at 4°C and vortexed. After centrifugation for 10 min (12,000 rpm, 4°C),
the supernatant was carefully dried, and then 150 µL of a
2-chlorophenylalanine solution prepared with 80% methanol (v:v = 4:1)
was added to resolubilize the samples. The supernatant was filtered
using a 0.22 μm microporous membrane and then added to the assay vial,
pending further analysis ([78]Demurtas et al., 2021).
2.7.2 QC samples
To correct for bias and systematic errors, 10 μL of each group sample
were mixed as QC samples, which were tested once for every 10 samples.
2.7.3 Liquid chromatography conditions
The LC analysis was performed on an ACQUITY UPLC System (Waters,
Milford, MA, United States). Chromatography was carried out with an
ACQUITY UPLC ^® HSS T3 (150 × 2.1 mm, 1.8 µm) (Waters, Milford, MA,
United States). The flow rate and injection volume were set at
0.25 mL/min and 2 μL, respectively, and the column temperature was set
at 40°C. Gradient elution procedures are shown in [79]Table 2
([80]Zelena et al., 2009).
TABLE 2.
Gradient elution procedures.
Time (min) Mobile phase
Positive mode Negative mode
0.1% formic acid in water (%) 0.1% formic acid in acetonitrile (%) 5 mM
ammonium formate (%) Acetonitrile (%)
0–1 98 2 98 2
1–9 98–50 2–50 98–50 2–50
9–12 50–2 50–98 50–2 50–98
12–13.5 2 98 2 98
13.5–14 2–98 98–2 2–98 98–2
14–20 98 2 98 2
[81]Open in a new tab
2.7.4 Mass spectrum conditions
Mass spectrometric detection of metabolites was performed on Q Exactive
(Thermo Fisher Scientific, United States) with an ESI ion source.
Acquisition parameters were set as follows: spray voltage was 3.50 kV
and −2.50 kV for positive mode and negative mode, sheath gas pressure
was 30 arb, and auxiliary gas pressure was 10 arb. The primary scan was
performed at 325°C in the range of m/z 100–1000 in the capillary tube.
Then, secondary cleavage was performed using HCD at a collision energy
of 30 eV and a secondary resolution of 17,500 ([82]Want et al., 2013).
2.7.5 Data processing
The raw data acquired from mass spectrometry was conversed into *mzXML
files by the Proteowizard software package (v3.0.8789) ([83]Smith et
al., 2006). The R XCMS software package was used for peak detection,
peak filtering, and peak alignment processing to obtain the substances
quantified list with parameters set to bw = 2, ppm = 15, peakwidth = c
(5, 30), mzwid = 0.015, mzdiff = 0.01, and method = “centWave”
([84]Navarro-Reig et al., 2015). Public databases such as HMDB
([85]Wishart et al., 2022), massbank ([86]Horai et al., 2010), and
LipidMaps ([87]Sud et al., 2007) were used to perform calibration tests
on metabolites. QC samples were corrected for data according to the
method reported previously ([88]Gagnebin et al., 2017), and QC samples
with RSD >30% will be discarded for quality control.
2.7.6 Pathway analysis
The MetaboAnalyst software package was used to perform functional
pathway enrichment and topological analysis for screening metabolic
biomarkers. The KEGG Mapper tool was used to visualize the pathways
([89]Xia and Wishart, 2011).
2.8 The comprehensive analysis combined with network pharmacology
Based on our previous network pharmacology analysis study, we combined
the metabolomics results with a network pharmacology approach for
further comprehensive analysis and a
metabolites-pathways-targets-ingredient network was constructed using
Cytoscape 3.7.2 software to determine the underlying mechanism of WSP.
2.9 Statistics analysis
All data were expressed as the mean ± standard deviation. One-way
analysis of variance (ANOVA) was used for multifactorial comparisons
using GraphPad Prism 9 (GraphPad Software, Inc., CA, United States).
Immunohistochemical images were processed by ImageJ software (NIH and
LOCI, United States) to calculate the positive areas. Pathology scoring
grades were performed with the Mann-Whitney U test using SPSS (IBM, NY,
United States). The p-value <0.05 was considered statistically
significant.
3 Results
3.1 WSP inhibits foot swelling in MSU-induced acute gouty arthritis in rats
As shown in [90]Figure 1A, B and [91]Table 3, all groups except the CON
group showed significant ankle swelling after receiving MSU crystal
injection. Compared with the CON group, the foot swelling of the model
rats gradually increased with time, demonstrating statistical
significance (p < 0.01), and the MSU-induced ankle swelling could be
maintained until 24 h. The inhibition of foot swelling was observed at
2 h in the ETO group, with significant inhibitions at every subsequent
time point. Compared with the MOD group, the WSP-H group primarily
showed a prominent therapeutic effect from 2 h to 6 h after MSU
injection, while the WSP-M group exhibited remarkable inhibition of
foot swelling from 6 h to 24 h. However, there was no conspicuous
effect observed in the ankle swelling of WSP-L rats.
FIGURE 1.
[92]FIGURE 1
[93]Open in a new tab
Foot swelling and pathological assay. (A,B) Foot swelling in each group
at each time point. (C) Histopathology staining with HE (left 20×,
right 200×). Black arrows indicate the infiltrated inflammatory cells
and green arrows indicate the structurally discontinuous synovial cell
layer.
TABLE 3.
The effect of each group on foot swelling in GA model rats (
[MATH: x¯±S
:MATH]
, n = 8–11).
Group n Dose (mg/kg×d) Foot swelling (mL)
0 h 2 h 4 h 6 h 8 h 24 h
CON 9 - 0.00 ± 0.26 0.23 ± 0.21 0.21 ± 0.26 0.11 ± 0.26 0.15 ± 0.22
0.05 ± 0.21
MOD 11 - 0.00 ± 0.19 0.58 ± 0.29^## 0.72 ± 0.22^## 1.03 ± 0.27^## 1.29
± 0.36^## 1.43 ± 0.42^##
WSP-L 8 10 × 7 0.00 ± 0.20 0.36 ± 0.33 0.51 ± 0.25 0.88 ± 0.44 1.08 ±
0.71 1.37 ± 0.67
WSP-M 9 20 × 7 0.00 ± 0.12 0.54 ± 0.20 0.62 ± 0.25 0.68 ± 0.35[94] ^b
0.81 ± 0.34[95] ^c 0.94 ± 0.26[96] ^c
WSP-H 9 40 × 7 0.00 ± 0.21 0.33 ± 0.17[97] ^b 0.42 ± 0.37[98] ^b 0.56 ±
0.42[99] ^c 0.96 ± 0.59 1.24 ± 0.44
ETO 9 32 × 7 0.00 ± 0.10 0.30 ± 0.22[100] ^b 0.30 ± 0.15[101] ^c 0.55 ±
0.30[102] ^c 0.72 ± 0.29[103] ^c 0.87 ± 0.27[104] ^c
[105]Open in a new tab
^^a
[MATH:
p < 0.05
:MATH]
,^##
[MATH:
p < 0.01
:MATH]
vs CON.
^^b
[MATH:
p < 0.05
:MATH]
.
^^c
[MATH:
p < 0.01
:MATH]
vs MOD.
3.2 WSP attenuated ankle joint changes in gouty rats
As shown in [106]Figure 1C, HE pathological section analysis indicated
the pathomorphological features of the synovial membrane and its
surrounding tissues of the joint. The cartilage surface of the CON
group appeared smooth and intact; the structure of synoviocytes was
clear, and there was no apparent infiltration of inflammatory cells
found in the synovial and surrounding tissues. In contrast, in the
joint samples from the MOD group, the synovial cell layer was seen to
have structural discontinuity (green arrows), and the synovial tissue
structure was damaged and infiltrated by a large number of inflammatory
cells (black arrows). Furthermore, the Mann-Whitney U test analysis
showed that the average score of the ankle in gouty rats was much
higher than that in the pseudosurgical rats, as illustrated in
[107]Table 4. On the contrary, WSP administration could alleviate the
pathological lesions of the rats to significant degrees, which is
displayed in [108]Figure 1C and [109]Table 4. In the WSP-H group, the
articular cartilage surface was smooth and flat, and only a small
amount of inflammatory cell infiltration was seen in the synovial and
surrounding tissues. The above results suggested that MSU could cause
serious damage to the joint synovial tissue. However, WSP treatment
exhibited a protective effect on the gouty ankle induced by MSU, which
is characterized by the improvement of the pathological changes and the
reduction of the infiltrated inflammatory cells.
TABLE 4.
Effect of WSP on the histopathological scoring of the ankle in gouty
rats (
[MATH: x¯±S
:MATH]
, n = 8–11).
Group Dose (mg/kg ×d) n Grades (samples) Weighted average score
0 1 2 3
CON - 9 6 3 0 0 0.3
MOD - 11 0 0 8 3 2.3[110] ^a
WSP-L 10 × 7 9 0 3 6 0 1.7[111] ^b
WSP-M 20 × 7 8 0 4 4 0 1.5[112] ^c
WSP-H 40 × 7 9 0 2 7 0 1.8[113] ^b
ETO 32 × 7 9 0 5 4 0 1.4[114] ^c
[115]Open in a new tab
^^a
[MATH:
p < 0.01
:MATH]
vs CON.
^^b
[MATH:
p < 0.05
:MATH]
.
^^c
[MATH:
p < 0.01
:MATH]
vs MOD (Mann-Whitney U test).
3.3 WSP downregulated the expression of GA-related proteins in the synovium
and surrounding tissues of the rats after MSU injection
In gout flares, the initiation of NLRP3 inflammasome was considered a
key step ([116]Dalbeth et al., 2021). NLRP3 inflammasome is closely
related to NF-κB signaling, which regulates the gene expression of all
components of inflammasome assembly and activation ([117]Joosten et
al., 2010). Activated NLRP3 inflammasome signaling ultimately leads to
the production and release of IL-1β, triggering a local inflammatory
response. As shown in [118]Figure 2, the protein expression of IL-1β,
p-NF-κB P65, and NLRP3 in the synovium and surrounding tissues of the
MOD group after MSU injection was significantly increased compared with
the CON group, while the expression of all three proteins was
downregulated by WSP treatment.
FIGURE 2.
[119]FIGURE 2
[120]Open in a new tab
Immunohistochemistry. (A) IHC staining of IL-1β (200×). (B) IHC
staining of p-NF-κB P65 (200×). (C) IHC staining of NLRP3 (200×). (D)
Statistics of the ratio of positive area.
3.4 Effects of WSP on the metabolic pathways of MSU-induced acute gouty
arthritis model rats
3.4.1 BPC, TIC, and QC
The base peak chromatogram (BPC) depicts the continuous spectrum formed
by the maximum ion intensity at different time points during the
continuous scan of the mass spectrometry. In both positive and negative
modes, all samples showed good peak shapes, intensity, and obvious peak
separation, and the general trend of the base peaks in each group was
similar, indicating good reproducibility and reliable results
([121]Supplementary File S1). Meanwhile, there were visible differences
in the number, intensity, and type of peaks in each group, suggesting
differences in the types and amounts of metabolites in the samples.
Total ion chromatogram (TIC) records the total ion signal generated by
the metabolites, which reflects the overall information of the sample
as well as BPC (shown in [122]Supplementary File S1), and the
distribution of QC samples (red dots) almost completely overlaps,
indicating that the systematic error of this experiment is small, the
experiment is reproducible, and the results are reliable
([123]Supplementary File S1).
3.4.2 Multivariate statistical analysis
Metabolomics data are characterized by high latitude and multivariate,
and in order to investigate the potential multidimensional data, we
need to further process the obtained information, namely, principal
component analysis (PCA), partial least squares discriminant analysis
(PLS-DA), and orthogonal-partial least squares discriminant analysis
(OPLS-DA) ([124]Thevenot et al., 2015). PCA visually reflects the
overall distribution characteristics of all samples and trends. As
shown in [125]Figure 3, each group had a good separation status and
less overlap compared with the MOD group, especially in the negative
mode. The R^2X parameter of PCA was the main interpretability parameter
of the model, and it was better when R^2X > 0.5, as shown in
[126]Supplementary File S2, and R^2X > 0.5 for each group in both
positive and negative modes, suggesting that the model was accurate.
The analysis results and parameters of PLS-DA and OPLS-DA are shown in
[127]Supplementary File S2 and [128]Supplementary File S3.
FIGURE 3.
[129]FIGURE 3
[130]Open in a new tab
The statistical results of PCA analysis.
3.4.3 Identification and analysis of metabolic biomarkers
Compound structures could be inferred by primary MS, and further
secondary MS is required to obtain precise information to improve the
accuracy of the results. We performed metabolic biomarkers confirmation
in HMDB, massbank, LipidMaps, mzcloud, and other databases based on
MS/MS fragmentation patterns. Metabolic biomarkers were screened using
VIP >1 as a condition and then were obtained using a t-test, p < 0.05
([131]Kieffer et al., 2016).
After MS/MS screening, a total of 30 metabolic biomarkers were
identified after modeling by MSU compared with the CON group, including
dodecanoic acid, taurohyocholate, mesaconate, 2-hydroxycinnamic acid,
trans-ferulic acid, nicotinic acid, etc. A total of 15 biomarkers were
upregulated and 15 were downregulated in the MSU-induced gouty rats
(Shown in [132]Table 5).
TABLE 5.
Metabolic biomarkers between the MOD group and CON group.
Metabolic biomarkers m/z Retention time s) Formula -log10 P) VIP Trend
in MOD
Dodecanoic acid 200.97 71 C[12]H[24]O[2] 3.45 2.39 ↑
Taurohyocholate 516.30 527.1 C[26]H[45]NO[7]S 2.69 2.20 ↓
Mesaconate 129.02 81 C[5]H[6]O[4] 3.10 2.18 ↑
2-Hydroxycinnamic acid 165.05 209.1 C[9]H[8]O[3] 2.22 2.15 ↑
trans-Ferulic acid 195.14 625.1 C[10]H[10]O[4] 2.36 2.10 ↑
Nicotinic acid 124.04 104.5 C[6]H[5]NO[2] 2.44 2.08 ↑
L-Methionine S-oxide 166.05 149 C[5]H[11]NO[3]S 2.36 2.06 ↑
4-Guanidinobutanal 130.09 813.1 C[5]H[11]N[3]O 2.26 2.03 ↑
9,10-Epoxyoctadecenoic acid 296.23 778.3 C[18]H[32]O[3] 2.12 2.02 ↑
D-Xylose 149.01 821.9 C[5]H[10]O[5] 2.63 2.02 ↓
Aflatoxin B[1] 312.36 770.7 C[17]H[12]O[6] 1.88 2.02 ↑
26-Hydroxyecdysone 480.28 527 C[27]H[44]O[7] 2.08 2.01 ↓
Mannitol 182.98 33 C[6]H[14]O[6] 2.09 1.99 ↑
3-Dehydroecdysone 462.27 589.9 C[27]H[42]O[6] 1.94 1.96 ↓
3-Methyloxindole 148.08 498.9 C[9]H[9]NO 1.94 1.94 ↓
D-Fructose 179.06 927.6 C[6]H[12]O[6] 2.37 1.93 ↓
Ursodeoxycholic acid 373.27 838.6 C[24]H[40]O[4] 1.96 1.86 ↑
25-Hydroxycholesterol 401.09 471.8 C[27]H[46]O[2] 2.00 1.85 ↑
4-Chlorobenzoate 138.99 980.2 C[7]H[5]ClO[2] 1.55 1.84 ↓
L-Olivosyl-oleandolide 516.30 761.3 C[26]H[44]O[10] 1.71 1.83 ↓
Caryophyllene alpha-oxide 203.18 816.1 C[15]H[24]O 1.68 1.80 ↓
Catechol 111.02 754.2 C[6]H[6]O[2] 1.71 1.79 ↓
3-Epiecdysone 464.28 669 C[27]H[44]O[6] 1.53 1.76 ↓
Triacetate lactone 127.04 34.6 C[6]H[6]O[3] 1.54 1.75 ↓
Eucalyptol 153.96 69.7 C[10]H[18]O 1.56 1.74 ↑
Bilirubin 585.27 794.2 C[33]H[36]N[4]O[6] 1.45 1.70 ↑
Nonadecanoic acid 297.24 802.5 C[19]H[38]O[2] 1.49 1.62 ↓
FAPy-adenine 154.07 246.4 C[5]H[7]N[5]O 1.36 1.61 ↓
Glycocholic acid 464.30 534.1 C[26]H[43]NO[6] 1.36 1.48 ↑
β-D-Fructose 6-phosphate 259.02 82.2 C[6]H[13]O[9]P 1.32 1.45 ↓
[133]Open in a new tab
Furthermore, a total of 42 metabolic biomarkers were screened in GA
rats treated with WSP, including 2-hydroxycinnamic acid,
hydroxykynurenine, nicotinic acid, 2-methoxyestradiol, and
N6-Acetyl-L-lysine. As displayed in [134]Table 6, a total of 11
metabolites in the serum were upregulated and 31 were downregulated by
the administration of WSP.
TABLE 6.
Metabolic biomarkers between the WSP group and MOD group.
Metabolic biomarkers m/z Retention time s) Formula -log10 P) VIP Trend
in WSP
2-Hydroxycinnamic acid 165.05 209.1 C[9]H[8]O[3] 2.82 2.27 ↓
Hydroxykynurenine 225.03 613.6 C[10]H[12]N[2]O[4] 2.03 2.20 ↓
Nicotinic acid 124.04 104.5 C[6]H[5]NO[2] 2.44 2.15 ↓
2-Methoxyestradiol 283.17 825.9 C[19]H[26]O[3] 2.58 2.13 ↓
N6-Acetyl-L-lysine 189.12 146.6 C[8]H[16]N[2]O[3] 2.17 2.10 ↓
Caryophyllene alpha-oxide 203.18 816.1 C[15]H[24]O 1.73 2.06 ↑
9,10-Epoxyoctadecenoic acid 296.23 778.3 C[18]H[32]O[3] 1.96 2.03 ↓
L-2-Hydroxyglutaric acid 146.96 667.7 C[5]H[8]O[5] 2.35 2.01 ↓
12-Hydroxydodecanoic acid 215.01 928.9 C[12]H[24]O[3] 2.2 2.00 ↓
Ribose 1,5-bisphosphate 309.17 736.9 C[5]H[12]O[11]P[2] 2.09 1.97 ↓
Palmitoleic acid 237.22 890 C[16]H[30]O[2] 1.98 1.96 ↓
Ursodeoxycholic acid 373.27 838.6 C[24]H[40]O[4] 1.77 1.95 ↓
L-Cysteine 120.98 52.6 C[3]H[7]NO[2]S 1.88 1.95 ↓
D-Erythritol 4-phosphate 202.03 468.7 C[4]H[11]O[7]P 2.1 1.94 ↑
L-2,4-diaminobutyric acid 118.07 281.9 C[4]H[10]N[2]O[2] 1.9 1.93 ↓
(S)-2-Methylmalate 148.03 317.5 C[5]H[8]O[5] 1.8 1.93 ↓
Benzaldehyde 107.05 283 C[7]H[6]O 1.82 1.92 ↓
Bilirubin 585.27 794.2 C[33]H[36]N[4]O[6] 1.77 1.91 ↓
Phenylacetaldehyde 120.06 410.9 C[8]H[8]O 1.74 1.89 ↓
4-Quinolinecarboxylic acid 171.91 35.4 C[10]H[7]NO[2] 1.46 1.81 ↓
α-dimorphecolic acid 279.23 794.2 C[18]H[32]O[3] 1.55 1.81 ↓
9,10-DHOME 313.24 703.5 C[18]H[3]4O[4] 1.7 1.79 ↓
D-Phenylalanine 165.02 921.4 C[9]H[11]NO[2] 1.75 1.79 ↑
Neocnidilide 177.13 530.2 C[12]H[18]O[2] 1.36 1.79 ↓
γ-Glutamylcysteine 248.96 93.7 C[8]H[14]N[2]O[5]S 1.77 1.78 ↓
Sodium deoxycholate 414.32 848.2 C[24]H[39]O[4]Na 1.49 1.77 ↑
L-Fucose 164.07 928.7 C[6]H[12]O[5] 1.63 1.76 ↑
Pterin 163.04 931 C[6]H[5]N[5]O 1.5 1.75 ↑
5-Methyl-2′-deoxycytidine 240.10 254.5 C[10]H[15]N[3]O[4] 1.57 1.75 ↓
5-Hydroxyindoleacetic acid 173.98 117 C[10]H[9]NO[3] 1.49 1.71 ↑
β-Tyrosine 164.07 953.9 C[9]H[11]NO[3] 1.4 1.71 ↓
6-Hydroxyhexan-6-olide 130.07 315.9 C[6]H[10]O[3] 1.39 1.70 ↑
Mannose 6-phosphate 260.02 146.3 C[6]H[13]O[9]P 1.57 1.70 ↓
6-Keto-prostaglandin F1a 371.26 660.4 C[20]H[34]O[6] 1.43 1.70 ↓
4-Guanidinobutanal 130.09 813.1 C[5]H[11]N[3]O 1.39 1.69 ↓
L-Valine 118.09 126.1 C[5]H[11]NO[2] 1.39 1.68 ↑
(2Z,4S,5R)-2-Amino-4,5,6-trihydroxyhex-2-enoate 178.09 295.7
C[6]H[11]NO[5] 1.43 1.67 ↓
6-Hydroxynicotinic acid 138.02 536.2 C[6]H[5]NO[3] 1.53 1.66 ↓
γ-Linolenic acid 277.22 864.5 C[18]H[30]O[2] 1.49 1.63 ↓
Dodecanoic acid 200.97 71 C[12]H[24]O[2] 1.36 1.63 ↓
15-Deoxy-d-12,14-PGJ[2] 315.20 757.2 C[20]H[28]O[3] 1.31 1.59 ↑
D-Xylose 149.01 821.9 C[5]H[10]O[5] 1.33 1.55 ↑
[135]Open in a new tab
Subsequently, the above results were collected for the agglomerative
hierarchical clustering analysis, and the dramatically altered
biomarkers were plotted in the form of heat maps. The metabolic
biomarkers which might share the same metabolic function or metabolic
pathways were clustered to visualize the changes and classification.
The results are shown in [136]Figures 4A, B.
FIGURE 4.
[137]FIGURE 4
[138]Open in a new tab
Metabolic biomarkers clustering heatmap (A) MOD vs CON. (B) WSP vs MOD.
The results of metabolic pathways enrichment analysis (C) MOD vs CON.
(D) WSP vs MOD.
3.4.4 Metabolic pathway enrichment analysis
We further enriched and investigated the information by using the MetPA
database to elucidate the metabolic pathways in the model rats that
might be affected by the administration of WSP through KEGG metabolic
pathway enrichment analysis. As shown in [139]Figures 4C, D, 27
metabolic pathways were enriched in the MOD group compared with the CON
group, mainly involving fructose and mannose metabolism, primary bile
acid biosynthesis, cholesterol metabolism, and other pathways. As shown
in [140]Figures 4C, D a total of 39 metabolic pathways were enriched in
the GA rats with WSP administration mainly in linoleic acid metabolism,
phenylalanine metabolism, lysosomal, ferroptosis, pantothenate and CoA
biosynthesis, PPAR signaling pathway, etc. [the top six pathways ranked
by -log10P)]. Based on the results of metabolic biomarkers screening
and the enrichment of metabolic pathways, we mapped a metabolic network
in which WSP might affect GA ([141]Figure 5).
FIGURE 5.
[142]FIGURE 5
[143]Open in a new tab
Schematic diagram of the metabolic networks regulated by WSP. Green
arrow indicates decreased metabolites in the serum and red arrow
indicates metabolites in the serum.
3.5 Comprehensive analysis
In our previous study, we obtained information on the effective pathway
of WSP in treating gout via KEGG enrichment (details shown in
[144]Supplementary File S4). To further elaborate the regulatory
network of WSP, the comprehensive analysis combined metabolomics and
network pharmacology was applied to construct a
metabolites-pathways-targets-ingredient network (Shown in [145]Figure
6), which involved 11 metabolites, 7 pathways, 39 targets, and 49
ingredients of WSP. The results showed that WSP mainly intervened in
arachidonic acid metabolism, ABC transporters, and PPAR signaling
pathway.
FIGURE 6.
[146]FIGURE 6
[147]Open in a new tab
Metabolites-pathways-targets-ingredient network. Markers of the
ingredients checklist are shown in [148]Supplementary File S5, and the
bigger nodes represented greater degrees. HZ: Terminalia chebula Retz.;
TBC: Aconitum pendulum Busch; MX: Aucklandia lappa Decne.; ZCP: Acorus
calamus L.; SX: Artificial musk.
4 Discussion
WSP has been used in China, especially among Tibetan people, for the
treatment of inflammation and inflammation-related diseases, such as
tonsillitis, pharyngitis, and rheumatoid arthritis. Our preliminary
study found that WSP might exert an anti-gout effect through multiple
targets and pathways ([149]Lang et al., 2022). However, the specific
mechanism is unclear. In this study, we explored the effect of WSP on
MSU-induced gouty arthritis in rats and used LC-MS untargeted
metabolomics approach to reveal its possible mechanism for the first
time.
Early anti-inflammatory treatment to control pain and the progression
of arthritis is the current consensus treatment in acute gout flares
([150]FitzGerald et al., 2020). An acute gouty inflammatory response is
triggered by the precipitation of MSU crystals due to elevated uric
acid levels in the blood, which is mainly a self-limiting inflammatory
response mediated by autoimmune cells. Clinically, gout is
characterized by joint swelling and heat pain, with a short time from
flare to peak (usually shorter than 12 h), and severe pain can even
affect activities and walking ([151]Taylor et al., 2015). Thus, early
anti-inflammatory treatment can result in greater benefits for gout
patients. In view of pathogenesis, the cellular mechanism of gout is
complex and it involves Toll-like receptor signaling initiation, NF-κB
signaling activation, oxidative stress, and NLRP3 signaling pathway
([152]Liu-Bryan et al., 2005; [153]Bauernfeind et al., 2009;
[154]Dominic et al., 2022). Stimulation of pathogen-associated
molecular patterns or damage associated molecular patterns triggers the
migration of monocytes-macrophages and exerts the body’s autoimmune
defense. MSU crystals can stimulate innate immune pathways and are
closely associated with the initiation of NLRP3 inflammasome
([155]Martinon et al., 2006). Activation of NLRP3 inflammasome is a
necessary signal of gouty inflammation. Upon activation of upstream
signaling, cytoplasmic pattern recognition receptors (PRRs) such as
NLRP3 perform a series of reactions including recruitment, aggregation,
and assembly. Pro-caspase-1 protein is recruited by ASC, forming NLRP3
inflammasome ([156]Schroder et al., 2010). Then, it will mediate the
maturation and release process of IL-1β, causing a severe inflammatory
response. Hence, we used an in vivo model to validate the anti-gout
effect of WSP. In this model, the foot swelling, pathological, and
immunohistochemical performance after MSU injection were assayed to
observe the effect of WSP on gout. Furthermore, metabolomics was used
to detect metabolic biomarkers expression in model animals and after
administration of WSP, to provide further evidence of metabolic
regulation of the anti-gout effect.
In MSU-stimulated acute arthritis of the ankle joint, the foot swelling
of the MOD group increased significantly at all time points from 2 h
post-modeling compared to the CON group, and the foot swelling
basically peaked at 8 h and lasted for 24 h. It is suggested that MSU
triggered a severe inflammatory response from 2 h and was in the
ascending phase of inflammation for 24 h. This result was consistent
with previous reports ([157]Goo et al., 2021; [158]Zhou et al., 2022).
Acute gouty arthritis caused severe joint pain and swelling and might
lead to changes in levels of serum metabolic biomarkers. However, the
study of specific mechanisms of TCM and its preparations has become
fairly complicated due to its complex components and multi-target
therapeutic characteristics. Recently, network pharmacology has been
used in the discovery of active compounds and the interpretation of the
overall mechanisms in the TCM or TCM preparations, by collecting and
analyzing data from bioinformation databases associated with herbs,
disease targets, and pathways. Nevertheless, network pharmacology has
several limitations if it is used alone, such as prediction-based
false-positive results, varying relative abundance of compounds in TCM,
and debatable ADME-based screening ([159]Jiashuo et al., 2022). Thus,
an integrated strategy combining network pharmacology with other
approaches is increasingly being used. For instance, network
pharmacology methodology is combined with multi-omics studies such as
proteomics ([160]Cheng et al., 2022; [161]Sun et al., 2022),
metabolomics ([162]Pan et al., 2020; [163]Zhou et al., 2020b; [164]Li
et al., 2021), transcriptomics ([165]Xiao et al., 2021; [166]Zhou et
al., 2022a), and lipidomics ([167]Chen et al., 2022). Furthermore, gut
microbiomics ([168]Goo et al., 2021; [169]Yao et al., 2022) and
meta-analysis ([170]Yi et al., 2020) were adopted for overall analysis
combined with network pharmacology methodology. Taking into account the
multi-faceted aspects and complexities of TCM or TCM preparations at
hand, the integration strategy combining multi-omics with network
pharmacology is beneficial for the accuracy of prediction results. In
this study, LC-MS-based untargeted metabolomics assays identified 30
potential biomarkers in the pathology of acute gouty arthritis, 15 of
them were upregulated and 15 were downregulated. A total of 42 serum
metabolic biomarkers were identified in model rats after treatment with
WSP, 11 of them were upregulated and 31 were downregulated. The above
results were enriched in metabolic pathways, mainly involving linoleic
acid metabolism, phenylalanine metabolism, lysosomal, ferroptosis,
pantothenate and CoA biosynthesis, PPAR signaling pathway, etc.
Linoleic acid is an essential nutrient for the human body, but
disorders of linoleic acid metabolism can lead to disease. Linoleic
acid is a synthetic precursor of arachidonic acid, which is metabolized
to produce γ-linolenic acid. γ-linolenic acid can participate in the
arachidonic acid (AA) metabolic pathway to generate AA, which exhibits
pro-inflammatory and pro-thrombotic potential. Eicosanoids are locally
acting bioactive signaling lipids, which are considered
pro-inflammatory mediators for inflammation, immunity, and allergies,
and eicosanoids are derived from arachidonic, γ-linolenic acid, and
polyunsaturated fatty acids ([171]Dennis and Norris, 2015).
Cyclooxygenase (COX) is one of the metabolic mediators of the
eicosanoids synthesis pathways, which controls a wide range of
inflammatory processes ([172]Smith et al., 2000). For instance, AA can
be converted into prostaglandins enzyme-catalyzed by COX. During gouty
inflammation, the first-line treatment strategy is COX inhibitors, such
as unselective COX inhibitors (ibuprofen and aspirin) and
COX-2-targeted agents (Etoricoxib) ([173]Wu et al., 2022; [174]Burkett
et al., 2023). The above inflammatory metabolites and mediators are
worthy of attention in the management of gout. Our metabolomic results
showed that γ-linolenic acid, α-dimorphecolic acid, and
9,10-Epoxyoctadecenoic acid and 9,10-DHOME, as well as all the
downstream of the linoleic acid metabolic pathway, were downregulated
by WSP treatment, suggesting that WSP might inhibit the linoleic acid
metabolic pathway and further affect AA metabolism to exert the
anti-inflammatory effect. Interestingly, this conjecture was again
confirmed in the integrative analysis combined with network
pharmacology, which revealed a nonnegligible role of AA metabolism in
WSP modulatory networks. Thus, key indicators of the linoleic acid
metabolic pathway, such as lipoxygenase and cytochrome P4501A2, could
be subsequently validated by qRT-PCR, Western blot, and ELISA
technology ([175]Wang et al., 2022). In consideration of the inhibition
of WSP on the MAPK/NF-κB signaling pathway ([176]Lang et al., 2022), we
hypothesized that WSP inhibited the combination of NF-κB and target
DNA, thereby suppressing COX expression and further reducing the
production of related metabolites catalyzed by COX1/2, such as PGs. In
the present study, we elucidated the role of WSP in the NF-κB signaling
pathway from a metabolomics perspective for the first time.
Phenylalanine metabolism is one of the important metabolic pathways of
amino acids, which has also been reported to be closely associated with
hyperuricemia and gout ([177]Jiang et al., 2017; [178]Zhou et al.,
2022). Phenylpyruvate and 2-hydroxycinnamic acid are products of the
phenylalanine metabolic pathway, and phenylpyruvate can further
generate phenylacetaldehyde, which generates phenylacetal-CoA, which is
involved in the biosynthesis of pantothenic acid and CoA via Acetyl-CoA
([179]Wu et al., 2018). In our study, D-phenylalanine expression was
upregulated after administration of WSP, while the expression of the
downstream phenylacetaldehyde and 2-hydroxycinnamic acid was
downregulated. We speculated that WSP might be able to correct the
abnormal phenylalanine metabolic pathway which led to the reduction of
downstream phenylacetaldehyde and 2-hydroxycinnamic acid. An abnormal
serum level of phenylalanine in patients with gout has been reported in
several previous studies ([180]Zhang et al., 2018; [181]Huang et al.,
2020; [182]Shen et al., 2021), suggesting that it might be one of the
metabolic biomarkers of gout, which was consistent with our findings.
Furthermore, the levels of several amino acids such as valine,
phenylalanine, tyrosine, and cysteine were significantly altered in
this study. Notably, a previous study on serum metabolic biomarkers in
gout patients also showed that the disorder of the amino acid metabolic
pathways could be involved in the pathogenic mechanism of gout
([183]Shen et al., 2021). Moreover, the phenylalanine metabolic process
would continue to affect the downstream biosynthetic pathways of
pantothenic acid and CoA, which was consistent with the KEGG enrichment
results. Pantothenic acid is the precursor of CoA, which is the
cofactor of various metabolic reactions and its synthesis is regulated
by acetoacetyl-CoA synthetase (AACS) ([184]Shan et al., 2021). CoA
plays an essential role in energy metabolism and participates in the
metabolism of glucose, protein, and lipids via the TCA cycle
([185]Hrubsa et al., 2022; [186]Wang et al., 2023). The TCA cycle is
the center of energy metabolism, which is closely related to
mitochondrial function. As we know, MSU could trigger mitochondrial
dysfunction and cause an increase in mitochondrial ROS, thus inducing
caspase-1-independent IL-1β secretion ([187]Abderrazak et al., 2015).
In this condition, an increment of mitochondrial fatty acid oxidation
could induce the increased generation of acetyl CoA, followed by the
augmented NADH and FADH2 levels in the TCA cycle, which amplified more
ROS formation, reinforcing a vicious cycle for the activation of
inflammasome ([188]de Mello et al., 2018; [189]Cojocaru et al., 2023).
In general, pantothenic acid and CoA play an essential role in cellular
metabolism. Therefore, we hypothesized that WSP indirectly affected
mitochondrial function and the TCA cycle by regulating upstream
phenylalanine metabolism and biosynthetic pathways of pantothenic acid
and CoA, thus modulating the alteration of metabolic pathways. Indeed,
some limitations of this study should be noted. Firstly, the effects of
WSP in rodent models of gout did not reflect the changes it may cause
in patients with gout. Therefore, it is essential to conduct clinical
trials to reveal the alteration of serum metabolite composition after
WSP intervention in gouty patients. Besides, integrated with network
pharmacology analysis, we predicted the active anti-gout ingredients in
WSP that might affect metabolic pathways associated with gout flares.
Additional in vitro experiments should be performed to further validate
the anti-gout potential and specific mechanisms of the predicted active
compounds, thus clarifying the material basis of WSP.
5 Conclusion
In the acute gouty arthritis model of rats, WSP treatment inhibited
significant foot swelling, attenuated pathological lesions, and
downregulated the related protein expression of NLRP3 inflammasome
signaling in synovial tissue of the ankle. Further metabolomics
combined with network pharmacology analysis suggested that the
therapeutic effect of WSP involved 11 biomarkers and 7 metabolic
pathways. We speculated that WSP might regulate linoleic acid
metabolism, phenylalanine metabolism, and pantothenate and CoA
biosynthesis, then further affect the arachidonic acid metabolic
pathway and mitochondrial function, thus inhibiting gouty inflammation.
The above results suggested that WSP could be a prospective candidate
as a novel anti-gout agent for the secondary development of Chinese
traditional patent medicine.
Acknowledgments