Plasma Metabolomics Signature of Gout and Asymptomatic Hyperuricemia

Background: Gout is a metabolic disease and is the most common form of inammatory arthritis affecting men. However, the pathogenesis of gout is still uncertain, and novel biomarkers are needed for early prediction and diagnosis of gout. To reveal the metabolic alterations in plasma of gout patients and hyperuricemia patients and discover novel molecular biomarkers for early diagnosis. Metabonomics was employed to screen and identify novel biomarkers of gout based on human plasma. High performance liquid chromatography-diode array detector (HPLC-DAD) and orthogonal signal correction partial least squares discriminate analysis (OPLS-DA) were also used for metabonomics study. Results: 80 and 62 features were selected as remarkable signicant variables in the two modes between gout and control group, 90 and 50 features between hyperuricemia (HUA) and control group, 63 and 60 features between gout and HUA group, respectively. 25 potential metabolic biomarkers which at least in two comparison groups were remained. Among 25 metabolites, 34% presented high area under the curve (AUC) values (AUC >0.75). Four metabolites including Lys-Ser, L-Pipecolic acid, glycine, arecoline were screened out. They were used to distinguish gout from hyperuricemia with AUC>0.75, which was greater than the AUC of uric acid. Conclusion: The differential metabolites of gout screened out were involved in amino acid metabolism, including glycine, serine and threonine metabolism, arginine biosynthesis, glutathione metabolism. Lys-Ser, L-pipecolic acid, glycine, arecoline were down-regulated in gout patients compared with hyperuricemia patients. The metabolomics signatures could serve as an ecient tool for early diagnosis and provide novel insights into the pathogenesis of gout. L-pipecolic acid, glycine, arecoline were down-regulated in gout patients compared with hyperuricemia patients. These results can provide novel insights into the development of gout. diagnosis of hyperuricemia acid ≥ 420µmo1/L ≥ The gout group also the diagnostic criteria of ACR-EULAR gout arthritis. Healthy control group: Normal blood uric acid levels. Exclusion criteria: malignant tumors/hematological tumors; pregnant/lactating women or women to become pregnant; chronic renal failure (CKD stage II and above, eGFR < 90 m1/min/1.73m²); abnormal liver function (ALT (cid:0) AST (cid:0) GGT (cid:0) T-Bil (cid:0) D-Bil > upper limit of normal value); hyperlipidemia (TG > 1.7 mmol/l, CHO > 5.17 mmol/l, LDL > 3.36 mmol/l); diabetes, hypertension, cardiovascular and cerebrovascular diseases; rheumatic diseases such rheumatoid arthritis, osteoarthritis, spondyloarthropathy, etc.; following immunosuppressants


Page 3/18
utilized a UPLC-MS to perform metabolomics analyses. We compared the metabolic products of age-and sex-matched gout group, hyperuricemia group and normal control group.

Results
Demographic characteristics of the study participants Totally, 30 male gout patients,30 age and body mass index (BMI)-matched hyperuricemia patients and 30 healthy controls (HCs) were included in this study. The average age of the participants was 33.78 ± 9.05 years and the mean BMI years was 25.50 ± 2.90 kg/m2. Gout patients and hyperuricemia patients had higher systolic blood pressure (SBP), uric acid (UA), glucose (GLU) and albumin (ALB) compared with healthy controls. Hyperuricemia patients had higher alanine transaminase (ALT) and triglyceride (TG) compared with healthy controls. Demographic characteristics of gout patients, hyperuricemia patients and healthy controls were shown in Table 1. Identi cation of plasma differential metabolites A total of 1055 features in positive ion mode and 1608 features in negative ion mode were detected using UPLC-MS. After peak alignment and removal of missing values, 1492 features in positive ion mode and 1589 features in negative ion mode were remained.
The remaining data were imported into SIMCA-P to conduct multivariate statistical analysis. The principal component analysis (PCA) analysis could outline the original distribution of metabolites. As shown in Fig. 1. a and h, there were no obvious outlier samples in modes. However, the scatter plot failed to show a clear separation in modes. Then, the OPLS-DA model was used to characterize the metabolic disturbances ( Fig. 2. a and f). Gout samples or HUA samples could be clearly distinguished from healthy control samples of positives and negative ion modes, but a small number of score plots crossed between gout group and HUA group. Permutation plots of two OPLS-DA models repeated 200 times veri ed the validation of the models (Fig. 3. a and f).
All the permuted R2 (cum) and Q2 (cum) values to the left were lower than the original point to the right and all the regression lines of the Q2 (cum) points have a negative intercept [10,11], which indicates that the models were valid with no over tting.
The standard value used in this project was that the P-value < 0.05, and Variable Importance in the Projection (VIP) value > 1.0,80 and 62 features were selected as remarkable signi cant variables in the two modes between gout and control group, 90and 50 features between HUA and control group, 63 and 60 features between gout and HUA group, respectively ( Fig. 4. a and f). For each group of comparison, the Euclidean Distance matrix was calculated for the quantitative values of the differential metabolites, and the differential metabolites were clustered by using the full linkage method, and the heatmap was used to demonstrate ( Fig. 5. a and f).
According to fold change > 1.2 or < 0.83, 25 potential metabolic biomarkers which at least in two comparison groups were remained. The Human Metabolome Database (HMDB) identi er, variable importance in projection values, P-values, fold change and AUC of each metabolite are shown in detail ( Table 2.) The differential metabolites can be categorized as amino acids, purine derivatives, fatty acyls, carboxylic acids and derivatives. hyperuricemia, compared to healthy control, the AUC was 0.991, no better diagnostic indicator than uric acid among the differential metabolites. But 4 metabolites AUC > 0.8. Using uric acid as a diagnostic indicator of gout, compared with healthy control, the AUC was 0.948. If using thymine as a diagnostic indicator of gout, the AUC was 0.949, and 5 metabolites AUC > 0.8. If uric acid levels were used as a diagnostic marker to identify patients with gout from hyperuricemia, the AUC = 0.565, and all metabolites screened out in this study were better than uric acid and 4 metabolites (Lys-Ser, L-Pipecolic acid, glycine, arecoline) AUC > 0.75 (Table 2.). These candidates could be used as novel diagnostic biomarkers.
Compared with hyperuricemia groups, the differential metabolites of gout screened out were involved in amino acid metabolism, including glycine, serine and threonine metabolism, arginine biosynthesis, glutathione metabolism (Fig. 6.)

Discussion
In this study, we enrolled the patients with hyperuricemia or gout and healthy control group, compared to the difference of plasma metabolites of three groups. Unlike previous studies, this study excluded the obesity, hyperlipidemia, hypertension, etc., which could be better re ect the differences of the above three groups metabolites. We still observed that blood glucose and systolic blood pressure were higher in the gout or hyperuricemia groups than in the healthy controls, which indicated that uric acid was related to glucose and salt metabolism [12][13][14][15].
In this study, we were able to discriminate gout patients or hyperuricemia patients from healthy controls in an OPLS-DA model based on plasma metabolites, and 25 signi cantly different metabolites were identi ed which were mainly involved in amino acid metabolism, purine metabolism, carbohydrate metabolism. However, many unknown components were still present. The identi cation of these unknown substances using HPLC-MS will be focused on in our future work. But a small number of score plots crossed between gout group and HUA group in an OPLS-DA model. Maybe we need more samples to separate the gout group from HUA group.
LC-MS-based metabolomics was a promising technique for biomarker discovery, which was put into this study, and reinforces the idea. In this study, we found that uric acid was still a biomarker for the diagnosis of hyperuricemia or gout, but it was limited to separate gout from hyperuricemia by uric acid. 4 metabolites including Lys-Ser, L-Pipecolic acid, glycine, arecoline were screen out. They were used to distinguish gout from hyperuricemia with AUC > 0.75, which was greater than the AUC of uric acid. Most of the metabolic differences found in this study were amino acids, which were different from previous studies. Many of the previous studies were lipid metabolites [7,9], which might be related to the inclusion criteria of our enrolled population. At the time of entry, we excluded patients with hyperlipidemia.
L-proline is one of the twenty amino acids used in living organisms as the building blocks of proteins. Proline is called an imino acid, although the International Union of Pure and Applied Chemistry (IUPAC) de nition of an imine requires a carbon-nitrogen double bond. Proline is a non-essential amino acid that is synthesized from glutamic acid. It is an essential component of collagen and is important for proper functioning of joints and tendons. Joint damage in patients with chronic gout may be associated with down-regulation of L-proline.
Glycine is a simple, nonessential amino acid, although experimental animals show reduced growth on low-glycine diets. The average adult ingests 3 to 5 grams of glycine daily. Glycine is involved in the body's production of DNA, phospholipids and collagen, and in release of energy. Glycine is an amino acid that is a component of endogenous antioxidant reductive glutathione. It is also sometimes referred to as a semi-essential amino acid and is often supplemented externally during severe stress. Recent studies have found that glycine also has anti-in ammatory and immunomodulatory effects, and has a protective effect on arthritis. Compared with patients with hyperuricemia, patients with gout had reduced glycine and reduced protective effect on joints.
Adenosine is a nucleoside that is composed of adenine and D-ribose. Adenosine or adenosine derivatives play many important biological roles in addition to being components of DNA and RNA. For instance, adenosine plays an important role in energy transfer as adenosine triphosphate (ATP) and adenosine diphosphate (ADP). Adenosine is an endogenous purine nucleoside.
Adenosine was down-regulated in gout patients compared with patients with hyperuricemia, which indicated that patients with gout have active purine metabolism.
In this study, we also found that the levels of thymine, sphingosine-1-phosphate and pyruvate were signi cantly lower and with high AUC (AUC > 0.75) in patients with gout or hyperuricemia than in healthy controls, and the levels of L-malic acid and cysteines-sulfate were signi cantly higher and with high AUC (AUC > 0.75) in patients with gout or hyperuricemia than in healthy controls.
These metabolites may be biomarkers for the diagnosis of gout and hyperuricemia.
This was a cross-sectional observational study, which found that some metabolites may be used in the diagnosis of gout or hyperuricemia, but the causal relationship cannot be determined. In order to better understand the physiological and pathological mechanisms of gout and hyperuricemia, the mechanism of the role of metabolites in the pathogenesis of gout needs further veri cation.

Conclusions
In summary, the present study identi ed the plasma metabolomics signatures of gout through LC-MS. The differential metabolites of gout screened out were involved in amino acid metabolism, including glycine, serine and threonine metabolism, arginine biosynthesis, glutathione metabolism. Lys-Ser, L-pipecolic acid, glycine, arecoline were down-regulated in gout patients compared with hyperuricemia patients. These results can provide novel insights into the development of gout.

Study Design and Population
This study was an across-sectional observational study. From October 2018 to May 2019, 60 participants and 30 volunteers were enrolled from the outpatient department or hospitalized at the Center for Endocrine Metabolism and Immune Disease Center at Beijing Lu He Hospital, Capital Medical University. Inclusion criteria for gout and hyperuricemia group: 18-75 years old; de nite diagnosis of hyperuricemia (serum uric acid ≥ 420µmo1/L (7 mg/dl), ≥ 2 times in the last month; willing to participate in this study. The gout group also met the diagnostic criteria of ACR-EULAR gout arthritis. Healthy control group: Normal blood uric acid levels. Exclusion criteria: malignant tumors/hematological tumors; pregnant/lactating women or women planning to become pregnant; chronic renal failure (CKD stage II and above, eGFR < 90 m1/min/1.73m²); abnormal liver function (ALT AST GGT T-Bil D-Bil > upper limit of normal value); hyperlipidemia (TG > 1.7 mmol/l, CHO > 5.17 mmol/l, LDL > 3.36 mmol/l); diabetes, hypertension, cardiovascular and cerebrovascular diseases; rheumatic diseases such as rheumatoid arthritis, osteoarthritis, spondyloarthropathy, etc.; have taken the following drugs in the last 6 months: glucocorticoids, immunosuppressants (mercaptopurine, 6-mercaptopurine, cyclosporine, cyclophosphamide), chemotherapy drugs; those who are allergic to drugs such as non-steroidal anti-in ammatory drugs, colchicine, febuxostat, allopurinol, or hypersensitivity; those with severe infections, active tuberculosis; have accepted urate-lowering therapy in the past three months; persons with mental illness who cannot cooperate; combined with other diseases affecting the e cacy or poor compliance. Data pre-processing and statistical analysis MS raw data (.wiff) les were converted to the mzXML format by ProteoWizard, and processed by R package XCMS (version 3.2).
The process includes peak deconvolution, alignment and integration. Minfrac and cut off are set as 0.5 and 0.6 respectively. Inhouse MS2 database was applied for metabolites identi cation.
In this study, the original data were preprocessed rstly. Then, the missing values were lled up by half of the minimum value.
Also, internal standard normalization method was employed in this data analysis. The nal dataset containing the information of peak number, sample name and normalized peak area was imported to SIMCA15.0.2 software package (Sartorius Stedim Data Analytics AB, Umea, Sweden) for multivariate analysis. Data was scaled and logarithmically transformed to minimize the impact of both noise and high variance of the variables. After these transformations, PCA, an unsupervised analysis that reduces the dimension of the data, was carried out to visualize the distribution and the grouping of the samples. 95% con dence interval in the PCA score plot was used as the threshold to identify potential outliers in the dataset. In order to visualize group separation and nd signi cantly changed metabolites, supervised orthogonal projections to OPLS-DA was applied. Then, a 7-fold cross validation was performed to calculate the value of R2 and Q2. R2 indicates how well the variation of a variable is explained and