, Introduction
Fabry disease (FD, OMIM number 301500) is an X-linked inherited lysosomal storage disease caused by mutations of the GLA gene. This biological phenomenon leads to the accumulation of globotriaosylceramide (Gb3) in podocytes, cardiac myocytes, endothelial cells, and neural cells because of defective or absent activity of the α-galactosidase A (α-Gal A) enzyme [1]. This may lead to life-threatening complications of hypertrophic cardiomyopathy, cerebrovascular events, and kidney injury [2]. Moreover, another sub-metabolite of globotriaosylsphingosine (lyso-Gb3), which is a deacylated form of Gb3, has been used as a biomarker for diagnosis and correlation with phenotype. Recently, many Gb3 isoforms and lyso-Gb3 analogs have been found in the plasma and urine of patients with FD [3–6].
Metabolomics is a global and multifaceted analysis of screening small molecules altered in biological samples [7]. Recent studies have approached the alteration of metabolic pathways in FD using a systematic and comprehensive metabolic analysis [8,9]. Previous metabolomics studies to identify new Gb3 or lyso-Gb3-related biomarkers in the plasma and urine of patients with FD have been conducted using a mass spectrometry metabolic approach, but these were focused on screening accumulation of glycosphingolipid analogs [4–6,10]. Although Ducatez et al. [11] identified 13 specific metabolites in the plasma of patients with FD from the French Fabry cohort through a network-based analysis, most of the targeted metabolites were glycerophospholipids, amino acids (AAs), and amines.
FD is associated with inflammation and oxidative stress. A recent study showed that α-Gal A–deficient podocyte dysregulated proteins involved in thermogenesis, lysosomal trafficking, metabolic activity, and cell-cell interaction and cycle [12]. In addition, the accumulation of Gb3 is related to an increase in reactive oxygen species (ROS) in endothelial cells, which may affect metabolic imbalance and mitochondrial dysfunction linked to the tricarboxylic acid cycle (TCA) cycle [13–15]. Thus, in this study, we focused on the analysis on systematic metabolomics including organic acids (OAs) along with TCA cycle metabolites, AAs, nucleosides (NSs), kynurenine pathway metabolites (KYNs), and fatty acids (FAs) as final products of lipid metabolism in the serum and urine of 20-week-old (20w) and 40-week-old (40w) FD mice compared to age-matched wild-type (WT) mice to reflect disease progression [16]. Our findings are expected to give a clue to the potential biomarker beyond the glycosphingolipids for predicting disease progression and provide therapeutic strategies in patients with FD.
Methods
Animals
Mice were maintained in a 12-hour light/dark cycle and had free access to standard chow and tap water. B6; 129-Glatm1Kul/J mice (α-Gal A knockout mice) were purchased from the Jackson Laboratory (JAX stock #003535). The male α-Gal A knockout (hemizygote) mice used in this study were selected by PCR-based genotyping using ear punch DNA samples. The primers used to amplify the GLA gene were as follows: oIMR5947 5’-AGG TCC ACA GCA AAG GAT TG-3’; oIMR5948 5’-GCA AGT TGC CCT CTG ACT TC-3’; and oIMR7415 5’-GCC AGA GGC CAC TTG TGT AG-3’. We used 20w and 40w α-Gal A knockout mice (n = 6 and 7 mice, respectively). Age-matched WT mice without a GLA mutation were used as healthy controls (n = 7 for each age group). Serum was collected from the retro-orbital sinus and centrifuged at 845 ×g for 10 minutes. Urine samples were collected in metabolic cages to examine the metabolites 2 days before sacrifice. The function data of the kidney and liver, and immunohistochemical staining in 20w and 40w FD mice and WT mice were shown in Supplementary Fig. 1A–E (available online).
All animal experiments were approved by the Institutional Animal Care and Use Committee of Chonnam National University Medical School and were conducted in accordance with the institution’s guidelines for experimental animal care and use (No. CNUHI-ACUC-23014).
Chemicals and reagents
All standards used and internal standards (ISs) for simultaneous profiling analyses of various metabolites, trimethylamine (TEA), and ammonium formate were purchased from Sigma-Aldrich and Tokyo Chemical Industry. Distilled water (DW), methanol, and acetonitrile (ACN) were purchased from J.T. Baker Inc. Formic acid and acetic acid were purchased from Wako Pure Chemical. Diethyl ether (DEE), ethyl acetate (EA), toluene, dichloromethane, and sodium chloride were purchased from Kanto Chemical. Tetrahydrofuran was purchased from Fisher Scientific Korea. Sulfuric acid and sodium hydroxide (NaOH) were purchased from Daejung Reagents Chemicals. O-Methoxyamine hydrochloride and N-methyl-N-(tert-butyldimethylsilyl) trifluoroacetamide (MTBSTFA) + 1% tert-butyldimethylchlorosilane were obtained from Thermo Scientific. All other chemicals were analytical grade.
Gas chromatography-tandem mass spectrometry (GC-MS/MS)
Metabolic profiling analyses of OAs and FAs were conducted by GCMS-TQ8040 (Shimadzu Corp.) equipped with an Ultra-2 (5% phenyl-95% methylpolysiloxane bonded phase; 25 × 0.20-mm inner diameter, 0.11 μm film thickness) cross-linked capillary column (Agilent Technologies) and interfaced with a triple quadrupole mass spectrometer. The profiling analyses were performed in multiple reaction monitoring (MRM) mode in electron impact ionization mode.
Liquid chromatography-tandem mass spectrometry (LC-MS/MS)
Metabolic profiling analyses of AAs, KYNs, and NSs were performed by a Shimadzu Nexera UPLC system (Shimadzu Corp.) coupled with an LCMS-8050 triple quadrupole mass spectrometer (Shimadzu Corp.). For chromatographic separation, an Intrada AA column (150 × 2 mm, 3 μm; Intakt) for AA profiling, a Zorbax Eclipse XDB-C18 (100 × 4.6 mm, 3.5 μm; Agilent Technologies) for KYN profiling, and a Hydro-RP C18 (150 × 4.6 mm, 4 μm; Phenomenex) for NS profiling were used. The MS system was operated in electrospray ionization mode and the profiling analyses were performed in MRM mode.
Sample preparation for organic acid and fatty acid profiling analyses in serum and urine by GC-MS/MS
Serum and urine samples from FD model mice were subjected to methoxime (MO)/tert-butyldimethylsilyl (TBDMS) derivatives for OA and FA profiling analysis by GC-MS/MS, as described in our previous studies [17–19]. Briefly, for deproteinization, ACN was added to serum (40 μL) and urine (20 μL) including ISs (13C2-succinic acid, 3,4-methoxybenzoic acid, lauric-d2-acid, and pentadecanoic acid), and they were mixed and centrifuged at 12,300 ×g for 3 minutes. Then, it was transferred to DW, the aqueous phase was adjusted to pH ≥12 with 5-M NaOH, and the MO reaction was performed with O-methoxyamine hydrochloride (1 mg) at 60 ℃ for 60 minutes. The aqueous phase was then acidified to pH ≤2 with 10% sulfuric acid and saturated with sodium chloride. Liquid-liquid extraction was conducted twice with 3 mL of DEE and 2 mL of EA. After 5 μL of TEA was added, the extracts were evaporated to dryness under a smooth stream of nitrogen at 40 ℃. To form the TBDMS derivative, the extracts were reacted with toluene (10 μL) and MTBSTFA (20 μL) for 60 minutes at 60 ℃. Then, 1 μL was injected for GC-MS/MS analysis.
Sample preparation for amino acid, kynurenine pathway metabolite, and nucleoside profiling analyses in serum and urine by LC-MS/MS
For deproteinization, ACN was added to serum (30 μL) and urine (10 μL) including ISs (13C1-phenylalanine and 13C1-leucine for AA profiling, 3,4-dimethoxybenzoic acid for KYN profiling, and 3-deazauridine for NS profiling), and they were mixed and centrifuged at 12,300 ×g for 3 minutes. After the supernatant was transferred to a Spin-X centrifuge filter tube and centrifuged at 12,300 ×g for 3 minutes, 1 μL was injected for LC-MS/MS analysis according to our previous method [20].
Star symbol pattern recognition analysis
The levels of OAs, FAs, AAs, KYNs, and NSs were calculated based on their standard calibration curves. The mean concentrations of each metabolite of the 20w and 40w FD groups were normalized to the corresponding mean values of the WT groups (20w and 40w, respectively). Each normalized value was plotted as a line emanating from a common central point. Star symbol patterns, which are used to readily observe metabolic alterations through a visualized format, were drawn with the normalized values using Microsoft Excel [21,22].
Statistical analysis
Data from animal experiments are presented as the mean ± standard deviation. After converting concentration data to log10-transformed data, a statistical comparison for the univariate analysis was performed using the Wilcoxon rank-sum test to distinguish discriminatory features between the two groups. The significant comparison results are indicated by a p-value of <0.05. Multivariate statistical analyses of metabolomics data, including partial least squares discriminant analysis (PLS-DA), hierarchical clustering heatmap analysis, univariate receiver operating characteristic (ROC) curve analysis, and pathway analysis, were performed using MetaboAnalyst 5.0 (https://www.metaboanalyst.ca) [23–25]. The cross validation of PLS-DA was confirmed with parameters, a correlation coefficient (R2), a cross-validation correlation coefficient (Q2), and an accuracy.
Results
Altered metabolic profiles and univariate analysis
Serum
Among 207 metabolites in our database according to the standard solutions, 91 metabolites including 17 OAs, 20 FAs, 36 AAs, seven KYNs, and 11 NSs were determined in the 20w and 40w serum of FD model (Supplementary Table 1, available online). The 20w and 40w FD groups were normalized to the corresponding 20w and 40w WT groups, respectively. Star symbol patterns were drawn to visually monitor altered metabolism and discrimination of WT and FD states from the metabolic profiles (Fig. 1). As a result of the Wilcoxon rank-sum test, eight metabolites, including glutathione disulfide (GSSG), isoleucine, and N4-acetylcytidine, showed significant changes (p < 0.05) between the 20w WT group and the 20w FD group, and 24 metabolites, including 5’-deoxy-5’-methylthioadenosine (MTA), 5,6-dihydrouridine, and N4-acetylcytidine, showed significant changes between the 40w WT group and the 40w FD group.
Urine
A total of 99 metabolites including 52 OAs, 24 AAs, seven KYNs, and 16 NSs were determined in the 20w and 40w urine of FD model mice (Supplementary Table 2, available online). The 20w and 40w FD groups were normalized to the corresponding 20w and 40w WT groups, respectively. Star symbol patterns were drawn to visually monitor altered metabolism and discrimination of WT and FD states from the metabolic profiles (Fig. 2). As a result of the Wilcoxon rank-sum test, nine metabolites, including pyruvic acid, 2-hydroxybutyric acid, and glycolic acid, showed significant changes (p < 0.05) between the 20w WT group and the 20w FD group, and 19 metabolites, including 3-methyladipic acid, ethylmalonic acid, and malonic acid, showed significant changes between the 40w WT group and the 40w FD group.
Multivariate analyses
Serum
To select metabolites that could discriminate between the WT group and the FD group, 27 metabolites (glutamic acid, 1-methylguanosine, MTA, N4-acetylcytidine, arachidonic acid, docosahexaenoic acid, docosatetraenoic acid, tryptophan, pipecolic acid, α-aminoadipic acid, creatine, citrulline, lysine, picolinic acid, 5-hydroxyindoleacetic acid, serotonin, 5,6-dihydrouridine, pseudouridine, N2-methylguanosine, 5-methylcytidine, N2,N2-dimethylguanosine, isoleucine, aspartic acid, GSSG, and cytidine) were considered key metabolites from serum of the 20w and 40w model mice; these were chosen based on p-values (<0.05) and variable importance in projection (VIP) scores (>1.0) from PLS-DA analysis (Table 1). Statistical analyses were performed with the 27 metabolites selected. The WT and FD groups of 20w and 40w model mice were separated in PLS-DA (Fig. 3A, D), but were not classified in heatmap analysis (Supplementary Fig. 2A, B; available online). In the serum of 20w model mice, the R2, Q2, and accuracy of PLS-DA were 0.88, 0.63, and 0.93, respectively, while in the serum of 40w model mice, they were 0.70, 0.58, and 0.87, respectively.
Urine
To select metabolites that could discriminate between the WT and FD groups, the 23 metabolites (2-hydroxybutyric acid, 3-hydroxypropionic acid, 1-methyladenosine, N6-methyladenosine, kynurenic acid, phenylacetic acid, malonic acid, ethylmalonic acid, isovaleryglycine, methylsuccinic acid, hexanoylglycine, adipic acid, 3-methyladipic acid, vanillic acid, 3-indolecarboxylic acid, vanillylmandelic acid, picolinic acid, 5-hydroxyindoleacetic acid, xanthosine, pyruvic acid, glycolic acid, α-ketoglutaric acid, and 4-hydroxyphenylpyruvic acid) were considered as key metabolites from urine of 20w and 40w model mice; these were chosen based on p-values (<0.05) and VIP scores (>1.0) from the PLS-DA analysis (Table 2). Statistical analyses were performed with the 23 metabolites selected. The WT group and the FD group of 20w and 40w model were separated in PLS-DA (Fig. 4A, D), and were classified in heatmap analysis, except for one case (Supplementary Fig. 2C, D; available online). In the urine of 20w model mice, the R2, Q2, and accuracy of PLS-DA were 0.99, 0.54, and 0.70, respectively, while in the urine of 40w model mice, they were 0.99, 0.81, and 0.90, respectively.
Identification of differential metabolites
Serum
Univariate ROC analysis, a practical tool for evaluating the performance of diagnostic tests, was performed on the 27 metabolites. 11 Metabolites in the 20w serum and 23 metabolites in the 40w serum showed an area under the curve (AUC) values (>0.8) (Table 1). Particularly, four metabolites including isoleucine (AUC = 0.95, log2 fold change [FC] = 0.50), aspartic acid (AUC = 0.93, log2FC = –0.60), GSSG (AUC = 1.00, log2FC = –1.49), and cytidine (AUC = 0.86, log2FC = 0.45) in the 20w serum were significantly altered (p < 0.05, VIP score > 1.0, AUC > 0.8) and thus may represent altered metabolites of an early diagnosis of FD. In addition, 19 metabolites including cis-aconitic acid (AUC = 0.88, log2FC = –0.35), arachidonic acid (AUC = 0.90, log2FC = 0.76), docosahexaenoic acid (AUC = 0.90, log2FC = 0.42), docosatetraenoic acid (AUC = 0.84, log2FC = 0.39), tryptophan (AUC = 0.84, log2FC = –0.57), pipecolic acid (AUC = 0.90, log2FC = –0.93), α-aminoadipic acid (AUC = 0.88, log2FC = –1.33), threonine (AUC = 0.84, log2FC = –0.39), creatine (AUC = 0.90, log2FC = –0.47), citrulline (AUC = 0.94, log2FC = –0.67), lysine (AUC = 0.86, log2FC = –0.54), picolinic acid (AUC = 0.86, log2FC = –0.31), 5-hydroxyindoleacetic acid (AUC = 0.86, log2FC = –0.71), serotonin (AUC = 0.92, log2FC = 0.44), 5,6-dihydrouridine (AUC = 0.98, log2FC = 1.25), pseudouridine (AUC = 0.90, log2FC = 0.64), N2-methylguanosine (AUC = 0.90, log2FC = 0.96), 5-methylcytidine (AUC = 0.90, log2FC = 0.56), and N2,N2-dimethylguanosine (AUC = 0.90, log2FC = 0.64) in the 40w serum were significantly altered (p < 0.05, VIP score > 1.0, AUC > 0.8) and thus may represent altered metabolites of progressive FD. Importantly, four metabolites (glutamic acid, 1-methylguanosine, MTA, and N4-acetylcytidine) in both 20w and 40w serum showed changes in equal directions (p < 0.05, VIP score > 1.0, AUC > 0.8). Glutamic acid increased more from 20w to 40w, and 1-methylguanosine, MTA, and N4-acetylcytidine decreased more from 20w to 40w. These metabolites could be interpreted as serum altered metabolites of progressive FD. Furthermore, GSSG was significantly increased in the 20w serum, but this change was not present in the 40w serum.
Urine
Univariate ROC analysis was performed on the 23 metabolites. Nine metabolites in the 20w urine and 19 metabolites in the 40w urine showed AUC values (>0.8) (Table 2). In particular, four metabolites including pyruvic acid (AUC = 1.00, log2FC =–0.89), glycolic acid (AUC = 0.97, log2FC = –0.36), α-ketoglutaric acid (AUC = 0.86, log2FC = –0.95), and 4-hydroxyphenylpyruvic acid (AUC = 0.86, log2FC = –1.02) in the 20w urine were significantly altered (p < 0.05, VIP score > 1.0, AUC > 0.8) and thus may represent altered metabolites of an early diagnosis of FD. In addition, 14 metabolites including phenylacetic acid (AUC = 0.86, log2FC = –0.62), malonic acid (AUC = 0.94, log2FC = –0.77), ethylmalonic acid (AUC = 0.96, log2FC = –1.03), isovaleryglycine (AUC = 0.88, log2FC = –0.58), methylsuccinic acid (AUC = 0.94, log2FC = –0.38), hexanoylglycine (AUC = 0.84, log2FC = –1.17), adipic acid (AUC = 0.88, log2FC = –0.92), 3-methyladipic acid (AUC = 1.00, log2FC = –1.07), vanillic acid (AUC = 0.86, log2FC = 0.87), 3-indolecarboxylic acid (AUC = 0.94, log2FC = –0.31), vanillylmandelic acid (AUC = 0.88, log2FC = –0.43), picolinic acid (AUC = 0.86, log2FC = –2.34), 5-hydroxyindoleacetic acid (AUC = 0.90, log2FC = –0.34), and xanthosine (AUC = 0.94, log2FC = –0.81) in the 40w urine were significantly altered (p < 0.05, VIP score > 1.0, AUC > 0.8) and thus may represent altered metabolites of progressive FD. Importantly, four metabolites (2-hydroxybutyric acid, 3-hydroxypropionic acid, 1-methyladenosine, and N6-methyladenosine) in both 20w and 40w urine showed changes in equal directions. More specifically, 2-hydroxybutyric acid and 3-hydroxypropionic acid increased in both the 20w and 40w urine, while 1-methyladenosine and N6-methyladenosine decreased in both the 20w and 40w urine. Thus, these metabolites could be interpreted as altered metabolites of progressive FD in urine. Interestingly, kynurenic acid was significantly decreased in the 20w urine, while it was significantly increased in the 40w urine, showing a change in the opposite direction.
Metabolic pathway analysis
Serum
To confirm which metabolic pathways the 27 metabolites were associated with, pathway analysis was performed. Alanine, aspartate, and glutamate metabolism, arginine and proline metabolism, arginine biosynthesis, and D-glutamine and D-glutamate metabolism were evaluated as key pathways in the 20w serum (p < 0.05, impact score > 0.1) (Fig. 3C). In addition, arginine and proline metabolism, arginine biosynthesis, D-glutamine and D-glutamate metabolism, tryptophan metabolism, arachidonic acid metabolism, and lysine degradation were evaluated as key pathways in the 40w serum (p < 0.05, impact score > 0.1) (Fig. 3F).
Urine
To confirm which metabolic pathways the 23 metabolites were associated with, pathway analysis was performed. Pyruvate metabolism, glycolysis/gluconeogenesis, and the TCA cycle were evaluated as a key pathway in the 20w urine (p < 0.05, impact score > 0.1) (Fig. 4C). Tyrosine metabolism (p = 0.221, impact score = 0.08) and the TCA cycle (p = 0.61, impact score = 0.10) were found to be related pathways in the 40w urine (Fig. 4F).
Discussion
In the present study, we conducted metabolic profiling analysis to identify 91 metabolites in the serum and 99 metabolites in the urine of mice with FD. Multivariate analysis was used to identify differential metabolites. We selected 27 key metabolites associated with FD that were different from those found in the serum of WT mice (p < 0.05, VIP score > 1.0). Specifically, eight metabolites were altered at 20w, and 23 metabolites were altered at 40w (p < 0.05, VIP score > 1.0, AUC > 0.8). In addition, 23 metabolites were found to be different in the urine of mice with FD compared with WT mice (p < 0.05, VIP score > 1.0). Especially, nine metabolites were altered at 20w and 19 metabolites were altered at 40w (p < 0.05, VIP score, > 1.0, AUC > 0.8). Pathway analysis revealed that tryptophan, arachidonic acid metabolism, arginine biosynthesis, and glutamine and glutamate metabolism were affected in the serum of mice with FD. Our findings could be differentiated from previous human studies in that we performed systemic metabolomics analyses including OAs, kynurenine, or NSs in serum and urine at early and late age points to find the potential biomarker beyond Gb3 and lyso-Gb3 analogs.
Regarding the statistical analysis conducted on the serum of mice with FD, a total of eight metabolites including four AAs and four NSs were significantly altered in mice with FD at 20w (p < 0.05, VIP score > 1.0, AUC > 0.8). The kidneys are involved in the synthesis and exchange of several AAs between different organs [26]. Glutathione metabolism is considered the most important antioxidant activity and various physiologic functions, including detoxification, modulation of redox-regulated signal transduction, storage and transport of cysteine, regulation of cell proliferation and immune responses, and metabolism of leukotrienes and prostaglandins [27]. Under oxidative stress, glutathione is synthesized from L-glutamate, L-cysteine, and glycine through glutathione synthetic enzymes and is oxidized to GSSG. A recent study showed that at the time of diagnosis, patients with FD exhibit altered glutathione metabolism and elevated glutathione levels [28]. Similar to the results of this previous study, we found that the levels of glutathione and GSSG were higher in mice with FD than in WT mice at 20w, indicating that an increase in glutathione metabolism could be an adaptive response to oxidative stress. However, the GSSG level in mice with FD was found to be similar to that of WT mice at 40w. These results may be indicative of decreased activity of glutathione peroxidase and GSSG reductase as well as an increase in ROS production, which could lead to a decrease in the GSSG level with the progression of FD [29]. Moreover, increased concentrations of ROS reduce the bioavailability of nitric oxide (NO) due to its rapid oxidative degradation and dysfunction of endothelial NO synthase, the enzyme responsible for producing NO in endothelial cells [30]. In particular, the accumulation of Gb3 in the vascular endothelium of FD is linked to an increase in the production of ROS [15,31]. Citrulline is a non-essential AA that is produced as an intermediate product of the urea cycle. It is also a byproduct of NO synthesis by endothelial NO synthase in the presence of tetrahydrobiopterin as a cofactor [30]. In this light, the increase in citrulline observed in mice with FD at 40w may be due to an increase in the byproducts of NO production, which could serve as a defense mechanism against oxidative stress.
Furthermore, the kynurenine pathway of tryptophan catabolism is known to be dysregulated in conditions associated with inflammation, such as cardiovascular disease, atherosclerosis, and chronic kidney disease [32,33]. A recent deep plasma targeted proteomic profiling and network analysis study also showed that several inflammation-related proteins are dysregulated in patients with FD [34]. Notably, our findings demonstrated that the levels of metabolites such as tryptophan, picolinic acid, and serotonin, which are involved in the kynurenine pathway, are altered in the serum of mice with FD at 40w. In addition, our final pathway analyses of the serum suggest that various pathways including those involved in tryptophan metabolism, arachidonic acid metabolism, arginine biosynthesis, proline metabolism, and glutamine and glutamate metabolism can be affected in mice with FD with high oxidative stress, inflammation, and immune intolerance.
Urinary NSs and deoxynucleosides are metabolites that result from oxidative DNA damage or RNA degradation [35]. Degraded normal NSs can be salvaged as nucleotide components, while the modified NSs are excreted into the urine. Therefore, identifying metabolic changes of metabolites in NSs in the urine in FD is important. Metabolites associated with NSs metabolism showed higher concentrations in the urine of mice with FD than those with WT mice at 40w. This finding suggests that urinary excretion of modified NSs were increased in the later stages of FD in mice. Interestingly, picolinic acid, which is another tryptophan metabolite of the downstream kynurenine pathway, was significantly increased in the urine at 40w in mice with FD. Picolinic acid has been implicated in neuroprotection and immune modulation [36,37]. Furthermore, the levels of α-ketoglutaric acid, which is a dicarboxylate OA and a key intermediate metabolite in the TCA cycle involved in maintaining mitochondrial metabolism homeostasis [38,39], were found to be increased in the urine of mice with FD collected at 20w compared with the levels in the urine of WT mice. This increased level of α-ketoglutaric acid in the urine may indicate mitochondrial disorders. Although substantial differences in AAs were observed at 40w between mice with FD and WT mice, there were no statistically significant differences in the urine collected at 20w and 40w.
In conclusion, this systematic metabolomics study revealed that 27 metabolites in the serum and 23 metabolites in the urine were associated with FD in mice at 20w and 40w. These altered metabolites were associated with oxidative stress, inflammation, NO biosynthesis, and immune regulation. Through a better understanding of the differential metabolic pathways involved in FD mice, we may be able to identify potential biomarkers to predict FD progression and provide insight into disease prognosis by assessing metabolic responses to treatment in early and late stages. Furthermore, these findings may provide novel therapeutic strategies to protect and reverse end-organ damage. Future studies are needed to validate the alteration of metabolic profiles related to oxidative stress, inflammation, and NO biosynthesis in the FD patients cohort.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Supplement table 1
Print
