Development and validation of a serum multi-omics based diagnostic model for systemic lupus erythematosus

doi:10.21203/rs.3.rs-1892917/v1

Download PDF

Research Article

Development and validation of a serum multi-omics based diagnostic model for systemic lupus erythematosus

https://doi.org/10.21203/rs.3.rs-1892917/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Objective.

The study aimed to identify potential biomarkers (metabonomics and proteomics) of systemic lupus erythematosus (SLE), and develop serum multi-omics based diagnostic models for SLE.

Methods.

Serum samples were collected from 111 SLE patients and 38 healthy controls (HCs). Gas chromatography-mass spectrometry (GC-MS) and Tandem Mass Tag (TMT) quantitative detection technology were respectively used to evaluate the subjects' serum metabolic spectrum and protein. Least absolute shrinkage and selection operator (LASSO) regression and stepwise logistic regression were used to select clinical characteristics, metabonomics and proteomics parameters associated with SLE and the severity of SLE. Three diagnosis models were developed and validated to diagnosis SLE and the severity of SLE. The receiver operating characteristic curve (ROC), sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were utilized to evaluated three diagnostic models. Three nomogram were plotted to visualize the three models.

Results.

12 variables of all variables (2 demographic characteristics, 9 laboratory indexes, 8 proteomic indexes and 339 metabolomic indexes) were selected, which can be used for the diagnosis of patients with different severity of SLE. Area under ROC (AUC) of three models were 0.932, 0.840 and 0.867, respectively. Meanwhile, the sensitivity and specificity of three models were high (Mode l: 0.956 and0.857, Model 2: 0.700 and 0.900, Model 3: 1.000 and 0.667).

Conclusion.

The changes of serum metabolites and proteins in patients with SLE were studied by multi-omics analysis, and three multi-omics based diagnosis models and nomograms for were developed and validated to diagnosis SLE and the severity of SLE.

Systemic Lupus Erythematosus

Multi-Omics

Diagnostic model

Nomogram

Systemic lupus erythematosus (SLE) is a persistent inflammatory autoimmune disease with high incidence and low quality of life [1]. SLE can occur at any age and sex, but the incidence is higher in young women [2]. SLE can affect multiple systems of the body in a variety of ways: from autoantibody production, fever and rash to pericarditis, polyarthritis and glomerulonephritis [3]. Because of the heterogeneity of the disease, no phenotype covers all the manifestations of SLE. The lack of consistency among patients with systemic lupus erythematosus complicates the study of the etiology of systemic lupus erythematosus [4]. Increased exposure to nuclear autoantigens and abnormal presentation of them to T cells has been observed in SLE patients [5, 6]. Therefore, T-cell-dependent stimulation of B cells causes the biosynthesis of antinuclear antibodies and the formation of immune complexes, which accumulate in various organs and tissues, resulting in dysfunction and clinically presenting as SLE [7, 8].

Tsokos et al believed that a variety of causes, such as genetic background and environmental factors, are related to the pathogenesis of SLE [9]. The clinical symptoms of SLE are relatively changeable, which makes the diagnosis of the disease challenging. Moreover, these symptoms overlap with symptoms of similar systemic autoimmune diseases, including primary Sjogren's syndrome and rheumatoid arthritis [10]. At present, the diagnostic markers widely used in SLE are still limited to antinuclear antibodies (ANA), anti-double-stranded DNA (anti-dsDNA) antibodies and complement activation [11]. In addition, anti-nucleosome antibodies, C-reactive protein (CRP), anti-C1q antibodies, B-cell subsets and erythrocyte-bound C4 have been reported to be associated with SLE [12]. However, the sensitivity and accuracy in the diagnosis of SLE are lacking.

With advances in technology and computing, the discovery of biomarkers has entered the "omics" era. Compared with genomics, proteomics and metabonomics may provide new insights into the principle of SLE progress by exploring the molecular domains downstream of the genome [13]. Proteomics can find some disease-specific protein molecules by comparing and analyzing the proteome between normal individuals and pathological individuals, and provide molecular markers for early diagnosis of diseases [14]. Metabolomics is an emerging technology that analyzes the concentration distribution of low-molecule metabolites in biological fluids and has been used in recent years to discover biomarkers for various diseases [15]. However, the values of multiple proteomic and small molecule metabolites in the clinical diagnosis of SLE remains unclear. Therefore, proteomics, metabolomics and laboratory indicators were used to find potential biomarkers, which is beneficial to understand the pathophysiological development of SLE. Here, we developed and validated serum multi-omics based diagnostic models for SLE and the severity of SLE. Our study provides a rich source of data to support further exploration of metabolic changes in SLE patients.

Study populations

A total of 111 SLE patients were admitted to the Affiliated Hospital of Nanjing University of Chinese Medicine, and 38 healthy volunteers from the Physical Examination Center of the Affiliated Hospital of Nanjing University of Chinese Medicine were selected as the control group. This study followed the principles of the Declaration of Helsinki and was approved by the Ethics Committee of the First Affiliated Hospital of Nanjing University of Chinese Medicine (2018NL165-02). All of the participants signed consent forms allowing their blood samples to be used for academic research.

Inclusion criteria include (1) 18–75 years old; (2) SLE disease activity index (SLEDAI) scores from 2 to 22; and (3) signed informed consent.

Exclusion criteria included: (1) patients with external infection (2) patients with tumor (3) patients with other connective tissue diseases. Serum samples from patients with SLE and healthy controls (HCs) were collected in procoagulant vessels and stored at − 80℃ until analysis.

Serum proteomic sample preparation

High-abundance proteins were removed using the ProteoMiner protein enrichment kit (Bio-Rad). The supernatant from each sample was digested with Trypsin Gold (Promega) at 1:50 enzyme-to-substrate ratio. After 16 h of digestion at 37°C, peptides were desalted with C18 cartridge to remove the high urea, and desalted peptides were dried by vacuum centrifugation. About 120 µg peptide mixture of each sample was labeled using Tandem Mass Tag (TMT)reagent according to the manufacturer’s instructions (Thermo Fisher Scientific). Finally, TMT-labeled digest samples were fractionated into 10 fractions by an increasing acetonitrile step-gradient elution according to instructions.

LC-MS/MS analysis

Shotgun proteomics analyses were performed using an EASY-nLC 1200 UHPLC system (Thermo Fisher) coupled with an Orbitrap Q Exactive HF-X mass spectrometer (Thermo Fisher)operated in the data-dependent acquisition (DDA) mode. A sample volume containing 2 µg of total peptides was injected onto a home-made (2 cm×100 µm, 5 µm). Peptides were separated on a home-made analytical column (15 cm×150 µm, 1.9 µm), using a linear gradient for TMT6-plex from 5 to 100% of eluent B (0.1% FA in 80% ACN) in eluent A (0.1% FA in H2O) in 90 min at a flow rate of 600 nL/min. Q Exactive HF-X mass spectrometer was operated in positive polarity mode with spray voltage of 2.3 kV and capillary temperature of 320°C. Full MS scans range from 350 to 1500 m/z were acquired at a resolution of 60000 (at 200 m/z) with an automatic gain control (AGC) target value of 3×106 and a maximum ion injection time of 20 ms. The 40 most abundant precursor ions from full MS scan were selected for fragmentation using higher energy collisional dissociation (HCD) at a resolution of 15000 (at 200 m/z) with an AGC target value of 1×105, a maximum ion injection time of 45 ms, a normalized collision energy of 32%, an intensity threshold of 8.3×103, and the dynamic exclusion parameter of 20 s.

Serum Metabolomics

Thawed the serum sample on ice, then mixed 40µL serum samples with 225µL cold methanol and vortexed for 10 seconds. Add 750 µL cold methyl tert butyl ether, vortexed for another 10 seconds, and shaked 10min at 4℃ in Thermo Mixer C (Eppendorf, Hamburg, Germany). Added 188 µL ultrapure water and vortexed for 20 seconds, then centrifuged for 2 min at 14,000 rcf at 4°C. Transfered the supernatant to a fresh tube, mixed with 750 µL methanol: isopropanol (1:1), and shaked in an orbital mixer for 10 minutes. Centrifuged for 2 min at 14,000 rcf at 4°C. The 475 µL of the supernatant was then dried in a SpeedVac sample concentrator for 2 hours at 45°C. The dried sample was mixed with 1 µL 1,2-¹³C2-myristic acid (5mg/mL) as the internal standard, and 30 µL methoxylamine hydrochloride was mixed in pyridine (10mg/mL). Shaking at 30°C for 90 min after vortexing for 1 min. Then shaken at 37°C for 30min after adding 30 µL 30 µL of N,O-bis (trimethylsilyl) trifluoroacetamide with 1% trimethylchlorosilane to the sample. Transfered the mixture to the sample bottle for GC-MS analysis. All serum samples were mixed into quality control (QC) samples, and QC samples were injected into every 10 samples after injection.

A Trace 1310 Gas Chromatograph equipped with an AS 1310 autosampler connected to a TSQ 8000 triple quadrupole mass spectrometer (Thermo Fisher Scientific, Waltham, MA, USA)was used for chromatographic separation. The standard n-alkane mixture (C8–C40) was injected into the GC/MS to detect and eliminate retention time shifts for the analysis of each sample.

Data Analysis

According to the diagnosis and severity of SLE, the study subjects were divided into health group, mild group, moderate group and severe group. The mild, moderate and severe patients were combined as patients group, and combined with health group to form dataset 1, which was used to develop the diagnostic model of SLE. Moderate and severe patients were combined as moderate & severe group, and combined with mild patients to form data set 2, which was used to develop the diagnostic model of moderate & severe patients. Moderate and severe patients were combined to form dataset 3 to develop a diagnostic model for severe patients (Fig. 1).

After the quantitative variables were standardized, three datasets were randomly divided into the development sets and the validation sets by the ratio of 8:2. First, in the development set, the student t test or Wilcoxon rank sum test or χ² test were performed to compare the differences of laboratory indicators, proteomics and metabolomics indicators between the health group and patients group, mild group and moderate & severe group, moderate group and severe group. Then, variables with P < 0.05 from the above analysis were included in the LASSO regressions, which was repeated for 20 times. Next, variables that were selected more than 10 times in 20 LASSO regressions were considered as candidate variables to establish diagnostic models, and final diagnostic models were established by stepwise logistic regressions. Meanwhile, the forest plot and nomogram were plotted to visualize diagnostic models. Finally, the diagnostic models were validated by validation sets and evaluated by receiver operating characteristic curve (ROC), sensitivity, specificity, positive predictive value and negative predictive value.

All Statistical analyses and visualizations were performed by R Version 4.0.3 (R Foundation for Statistical Computing, Vienna, Austria). The two-tailed statistical test level α = 0.05 was adopted.

Baseline characteristic

A total of 149 subjects were admitted in this study, including 128 females (85.91%) and 21 males (14.09%). The mean age was 39.7 ± 12.7 years old, with the minimum age of 18 years old and the maximum age of 75 years old. There were 111 SLE patients, including 43 mild patients (38.74%), 30 moderate patients (27.03%), 23 severe patients (20.72%) and 15 patients (13.51%) with unknown severity. The clinical information was summarized in Table 1.

Table 1

Clinical Characteristics of the Subjects
Characteristic	SLE	HC
Characteristic	N = 111	N = 38
Sex, n (%)
Male	5 (4.50)	16 (42.11)
Female	106 (95.5)	22 (57.89)
Age (years), Mean ± SD	41.90 ± 13.34	33.21 ± 7.71
ALT (U/L), Median (P₂₅ ~ P₇₅)	21.00 (15.00 ~ 29.00)	16.66 (13.25 ~ 21.64)
AST (U/L), Median (P₂₅ ~ P₇₅)	21.00 (18.00 ~ 27.00)	19.00 (17.00 ~ 21.02)
WBC (×10^9cells/L), Median (P₂₅ ~ P₇₅)	4.56 (3.26 ~ 5.84)	5.88 (5.23 ~ 6.30)
RBC (×10^9cells/L), Median (P₂₅ ~ P₇₅)	4.12 (3.58 ~ 4.46)	4.59 (4.35 ~ 4.75)
Hb (g/L), Median (P₂₅ ~ P₇₅)	117.00 (94.50 ~ 131.50)	139.08 (130.84 ~ 142.19)
PLT (×10^9/L), Median (P₂₅ ~ P₇₅)	152.0 (113.5 ~ 194.5)	217.2 (198.4 ~ 226.9)
NEUT%, Median (P₂₅ ~ P₇₅)	66.70 (57.70 ~ 75.30)	58.55 (56.27 ~ 61.70)
LYM%, Median (P₂₅ ~ P₇₅)	20.90 (15.30 ~ 28.90)	31.07 (29.26 ~ 31.96)
NEUT# (×10^9cells/L), Median (P₂₅ ~ P₇₅)	4.10 (2.26 ~ 7.00)	32.01 (28.34 ~ 33.53)
ALT: alanine aminotransferase; AST: aspartate aminotransferase; WBC: white blood cell; RBC: red blood cell; Hb: hemoglobin; PLT: platelet; NEUT: neutrophils; LYM: lymphocyte;

Comparison of characteristics between health group and patient groups, mild group and moderate & severe group, moderate and severe group showed that 358 variables (2 demographic characteristics, 9 laboratory indicators, 8 proteomic indicators, and 339 metabonomic indicators) were statistically significant between SLE patients and health group (P < 0.05), 168 variables (1 laboratory indicator, 167 metabolomics indicators) were significantly different between mild and moderate & severe group (P < 0.05), 104 variables (104 metabolomics indicators) were significantly different between moderate and moderate group (P < 0.05).

LASSO regressions

In the three development sets, variables with statistically significant differences between the groups were respectively included in the LASSO regression, and 20 times were repeated. The selected variables and their frequencies were shown in Supplementary Fig. 1-Figure3 and Table S1-Table S3.

Diagnostic model

In 3 development sets, variables selected more than 10 times in 20 LASSO regressions were considered as candidate variables for stepwise logistic regressions. The results showed that female, ITIH3 and TAG 57:0 were statistically associated with SLE; WBC, LPC 15:0(SN2), PC (11:0/22:2), TAG 48:5, LPC 22:5 and PC 36:2 were statistically associated with moderate & severe SLE; palmitic acid, PC (9:0/24:2) and PC 44:8e were statistically associated with severe SLE, as shown in Table 2–4. In addition, three forest plots were conducted using the variable coefficients, 95% confidence intervals and corresponding P values of each model, as shown in Fig. 2.

Table 2

Variables associated with SLE in stepwise logistic model
Variable	β	SE	OR (95%CI)	P value
Sex female	3.078	1.171	21.715(2.187-215.654)	0.009
ITIH3	-1.195	0.422	0.303(0.132–0.692)	0.005
TAG 57:0	-4.214	0.869	0.015(0.003–0.081)	< 0.001

Table 3

Variables associated with moderate and severe SLE in stepwise logistic model
Variable	β	SE	OR(95%CI)	P value
WBC	-1.206	0.414	0.299(0.133–0.674)	0.004
LPC 15:0(SN2)	0.833	0.404	2.300(1.041–5.081)	0.039
PC (11:0/22:2)	1.453	0.504	4.276(1.594–11.478)	0.004
TAG 48:5	1.906	0.732	6.726(1.601–28.264)	0.009
LPC 22:5	-1.055	0.479	0.348(0.136–0.891)	0.028
PC 36:2	-1.69	0.581	0.185(0.059–0.576)	0.004

Table 4

Variables associated with severe SLE in stepwise logistic model
Variable	β	SE	OR(95%CI)	P value
palmitic acid	-1.675	0.69	0.187(0.048–0.725)	0.015
PC (9:0/24:2)	-1.719	0.674	0.179(0.048–0.672)	0.011
PC 44:8e	1.164	0.548	3.203(1.095–9.375)	0.034

Evaluation and visualization of three models

According to the ROC (Fig. 3), the AUC (95%CI) of the three diagnostic models were 0.932 (0.833-1.000), 0.840 (0.6599-1.000) and 0.867 (0.637-1.000), respectively. However, the AUC of clinical characteristics to diagnosis SLE and moderate & severe patients was 0.621 (0.435–0.807) and 0.715 (0.452–0.978), as shown in Fig. 3A and Fig. 3B. The sensitivity, specificity, positive predictive value and negative predictive value corresponding to the three models are shown in Table 5. In short, according to the evaluation results, the diagnostic capabilities of the three models were high. In particular, the sensitivity and specificity of Model 1(diagnostic model of SLE) were 0.956 and 0.857, which means the model could identify 95.6% true SLE patients and 85.7% false patients.

Table 5

The evaluation index of three models
Indicator	Model 1	Model 2	Model 3
Sensitivity	0.956	0.700	1.000
Specificity	0.857	0.900	0.667
Pos Pred Value	0.956	0.875	0.714
Neg Pred Value	0.857	0.750	1.000

In addition, as shown in Fig. 4–6, three models were visualized as nomograms to facilitate clinical application of them, users of the nomogram could roughly calculate an individual’s risk of SLE based on certain patient’s characteristics.

SLE is an autoimmune disease caused by pathogenic autoantibodies, which may cause multiple organ damages. [16]. It covers a worldwide scale, mostly among Africans and Asians [17]. Because of its heterogeneity, there is still no effective therapy for SLE. At present, the early diagnosis of SLE need more research and clinical experience. In practice, the main diagnostic markers of SLE are antinuclear antibodies, several autoantibodies and complement activation [18]. Patients still have excessive damage accumulation, morbidity and death, which indicates that there are a lot of medical needs in SLE. Previous studies have observed abnormal changes in many proteins and metabolites in patients with SLE. For example, patients with SLE have lower levels of most amino acids, glycolysis and TCA circulating metabolites, and higher levels of fatty acids and markers of oxidative stress [19]. In a study by Rashmi Madda et al, 14 significantly expressed proteins were identified (9 up-regulated and 5 down-regulated). Most proteins, including alpha 2 macroglobulin, complement C4, complement factor H, fibrinogen beta chain, and alpha-1-antitrypsin, play a key role in the pathogenesis of SLE [20]. These studies are helpful for us to understand the pathogenesis of SLE. However, considering the complexity of diseases and the pathological differences between active and inactive diseases, looking for a single biomarker cannot fully understand the pathogenesis of SLE. Most of the previous studies are limited by the analysis platform or sample size, which cannot identify a sufficient number of biomarkers. We used gas chromatography-mass spectrometry and TMT quantitative detection technology for metabonomic and preliminary proteomic analysis. Comprehensive multi-omics analysis will provide new insights into the molecular mechanism of SLE.

We have established three models of SLE preliminary, moderate & severe diagnosis and severe diagnosis. 12 indicators related to SLE diagnosis were screened out from 358 variables by lasso regression analysis, including sex (female), ITIH3, TAG, WBC, LPC, PC and palmitic acid. Previous studies have confirmed that SLE mostly occurs in women of childbearing age, with a male-to-female ratio of about 1:9 [21, 22]. This is consistent with the results of our preliminary diagnostic model analysis. Inter-alpha-trypsin inhibitor (ITI) family is a serine protease inhibitor composed of two precursor proteins, including a light chain (bikunin) and five homologous heavy chains [23]. Different homologous heavy chains constitute a total of five members of ITIH1-5. Different homologous heavy chains constitute five members of ITIH1-5, among which ITIH3 is located in the closely linked region of chromosome 3p21. During the assembly of mature ITI proteins in the liver, ITIH3 undergoes extensive post-translational modification [24], mainly involving C-terminal pruning [25]. Previous studies have shown that ITI is closely related to inflammation, such as colitis [26], pancreatitis [27], Sjögren’s syndrome [28], polyarthritis [29] or sepsis [30]. Our results confirm that ITIH3 is down-regulated in patients with SLE. Leukopenia is one of the important hematological features of SLE. According to the latest diagnostic classification criteria of SLE jointly issued by the European League Against Rheumatism (EULAR) and the American College of Rheumatology (ACR), leukopenia < 4000/mm³scores 3 points in clinical diagnosis. Our statistical results show that the number of WBCs is negatively correlated with SLE, and it can be used to diagnose and distinguish mild patients from moderate and severe patients. Maintaining the dynamic balance of lipid metabolism is a basic feature of normal organisms. The increased level of oxidative stress in patients with SLE leads to the disorder of lipid metabolism in patients with SLE, which is closely related to its pathogenesis [31]. It can be seen in our study that palmitic acid, LPC 15:0(SN2), LPC 22:5, PC (11:0/22:2), PC (9:0/24:2), PC 36:2, PC 44:8e, TAG 57:0 and TAG 48:5 can be used as lipid markers in different degrees of SLE. Palmitic acid, LPC 15:0(SN2), LPC 22:5, PC (11:0/22:2), PC (9:0/24:2), PC 36:2, PC 44:8e, and TAG 48:5 can distinguish mild patients from moderate and severe patients, while palmitic acid, PC (9:0/24:2) and PC 44:8e can further distinguish moderate patients from severe patients. Abnormal lipid metabolism caused by SLE may induce cardiovascular disease, which increases the mortality of patients. Finally, after screening the diagnostic markers, we verify the diagnostic effectiveness of the three models. Complex diseases are generally considered to be caused by the interaction of a variety of environmental exposure factors (macro) and inherent genetic characteristics (micro). Clinical diagnosis alone is increasingly unable to meet the needs of disease diagnosis. Different levels of genetic information in organisms have a relatively large amount of data, which constitute different combinatorial data. Through the comparison of ROC in Fig. 3, it can be seen that clinical and omics co-diagnosis is more effective than simple clinical diagnosis, especially in the diagnosis of severe SLE.

In conclusion, we made a statistical analysis on the demographic characteristics, laboratory indicators, proteomic and metabonomic indicators of patients with SLE, diagnostic markers were screened to provide a reference for the diagnosis of SLE. We have observed changes in many important metabolites in patients with SLE, and multi-omics techniques can identify and quantify various biomolecules in blood and different types of tissues. It accelerates the development of life science research and contributes to a better understanding of the mechanism, diagnosis and treatment of the disease.

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Author Contributions

S-jL, Y-zW, X-xW, W-W and F-G conceived of and designed the project. Y-zW, W-W and X-xW performed the experiments and analyzed the data. Q-Y, Q-zH and P-W contributed to the sample collection and storage. F-G, Q-Y and Q-zH were responsible for searching and confirming the clinical data. Y-zW and W-W wrote the manuscript. All the authors approved the final version of the manuscript to be submitted.

Funding

This work is ﬁnancially supported by the National Natural Science Foundation of China (No. 82074241 and 81774096), the Open Projects of the Discipline of Chinese Medicine of Nanjing University of Chinese Medicine Supported by the Subject of Academic priority discipline of Jiangsu Higher Education Institutions (NO. ZYX03KF031), the Jiangsu Provincial Medical Youth Talent (No. QNRC2016642), the Jiangsu Provincial Medical Youth Talent (No. QNRC2016642), the Six Talent Project of Jiangsu Province, China (No. WSW-014), the general project of medical scientific research of Jiangsu Provincial Health Commission (No. M2021001).

Acknowledgments

We thank all the participants who contributed their efforts to this study, as well as the staff of the Affiliated Hospital of Nanjing University of Traditional Chinese Medicine who helped us collect samples.

Data Availability Statement

The data presented in this study are available on request from the corresponding author. The data are not publicly available due to privacy restrictions.

He J, Chan T, Hong X, Zheng F, Zhu C, Yin L, Dai W, Tang D, Liu D, Dai Y. Microbiome and Metabolome Analyses Reveal the Disruption of Lipid Metabolism in Systemic Lupus Erythematosus. Front Immunol. 2020 Jul 31;11:1703. doi: 10.3389/fimmu.2020.01703. PMID: 32849599; PMCID: PMC7411142.
Formiga F, Moga I, Pac M, Mitjavila F, Rivera A, Pujol R. Mild presentation of systemic lupus erythematosus in elderly patients assessed by SLEDAI. SLE Disease Activity Index. Lupus. 1999;8(6):462-5. doi: 10.1177/096120339900800609. PMID: 10483015.
Costenbader KH, Karlson EW. Cigarette smoking and systemic lupus erythematosus: a smoking gun? Autoimmunity. 2005 Nov;38(7):541-7. doi: 10.1080/08916930500285758. PMID: 16373259.
Speyer CB, Costenbader KH. Cigarette smoking and the pathogenesis of systemic lupus erythematosus. Expert Rev Clin Immunol. 2018 Jun;14(6):481–487. doi: 10.1080/1744666X.2018.1473035. Epub 2018 May 11. PMID: 29724134; PMCID: PMC7069669.
Nakajima M, Nakajima A, Kayagaki N, Honda M, Yagita H, Okumura K. Expression of Fas ligand and its receptor in cutaneous lupus: implication in tissue injury. Clin Immunol Immunopathol. 1997 Jun;83(3):223-9. doi: 10.1006/clin.1997.4352. PMID: 9175910.
Rosen A, Casciola-Rosen L, Ahearn J. Novel packages of viral and self-antigens are generated during apoptosis. J Exp Med. 1995 Apr 1;181(4):1557-61. doi: 10.1084/jem.181.4.1557. PMID: 7699336; PMCID: PMC2191952.
Sekigawa I, Naito T, Hira K, Mitsuishi K, Ogasawara H, Hashimoto H, Ogawa H. Possible mechanisms of gender bias in SLE: a new hypothesis involving a comparison of SLE with atopy. Lupus. 2004;13(4):217 – 22. doi: 10.1191/0961203304lu1012ed. PMID: 15176655.
Piotrowski P, Gasik R, Lianeri M, Cieślak D, Wudarski M, Hrycaj P, Łacki JK, Jagodziński PP. Asp327Asn polymorphism of sex hormone-binding globulin gene is associated with systemic lupus erythematosus incidence. Mol Biol Rep. 2010 Jan;37(1):235–9. doi: 10.1007/s11033-009-9639-7. Epub 2009 Aug 1. PMID: 19649728.
Tsokos GC, Lo MS, Costa Reis P, Sullivan KE. New insights into the immunopathogenesis of systemic lupus erythematosus. Nat Rev Rheumatol. 2016 Nov 22;12(12):716–730. doi: 10.1038/nrrheum.2016.186. PMID: 27872476.
Bengtsson AA, Trygg J, Wuttge DM, Sturfelt G, Theander E, et el. Metabolic Profiling of Systemic Lupus Erythematosus and Comparison with Primary Sjögren's Syndrome and Systemic Sclerosis. PLoS One. 2016 Jul 21;11(7):e0159384. doi: 10.1371/journal.pone.0159384. PMID: 27441838; PMCID: PMC4956266.
Shin TH, Kim HA, Jung JY, Baek WY, Lee HS, Park HJ, Min J, Paik MJ, Lee G, Suh CH. Analysis of the free fatty acid metabolome in the plasma of patients with systemic lupus erythematosus and fever. Metabolomics. 2017 Dec 14;14(1):14. doi: 10.1007/s11306-017-1308-6. PMID: 30830319.
Yan B, Huang J, Zhang C, Hu X, Gao M, Shi A, et el. Serum metabolomic profiling in patients with systemic lupus erythematosus by GC/MS. Mod Rheumatol. 2016 Nov;26(6):914–922. doi: 10.3109/14397595.2016.1158895. Epub 2016 May 11. PMID: 26915395.
Eddy S, Mariani LH, Kretzler M. Integrated multi-omics approaches to improve classification of chronic kidney disease. Nat Rev Nephrol 2020; 16(11): 657–68.
Dubin RF, Rhee EP. Proteomics and Metabolomics in Kidney Disease, including Insights into Etiology, Treatment, and Prevention. Clin J Am Soc Nephrol 2020; 15(3): 404–11.
Clish CB. Metabolomics: an emerging but powerful tool for precision medicine. Cold Spring Harb Mol Case Stud 2015; 1(1): a000588.
Borchers AT, Naguwa SM, Shoenfeld Y, et al. The geoepidemiology of systemic lupus erythematosus. Autoimmun Rev 2010;9(5):A277–A287. doi:10.1016/j.autrev.2009.12.008.
Kim HA, Lee HS, Shin TH, Jung JY, Baek WY, Park HJ, Lee G, Paik MJ, Suh CH. Polyamine patterns in plasma of patients with systemic lupus erythematosus and fever. Lupus. 2018 May;27(6):930–938. doi: 10.1177/0961203317751860. Epub 2018 Jan 7. PMID: 29308729.
Ding H, Mohan C. Connective tissue diseases: promises and challenges of metabolomics in SLE. Nat Rev Rheumatol. 2016;12(11):627–628.
Madda R, Lin SC, Sun WH, Huang SL. Differential expressions of plasma proteins in systemic lupus erythematosus patients identified by proteomic analysis. J Microbiol Immunol Infect. 2019 Oct;52(5):816–826.
Pons-Estel GJ, Alarcon GS, Scofield L, Reinlib L, Cooper GS. Understanding the epidemiology and progression of systemic lupus erythematosus. Semin Arthritis Rheum. 2010;39(4):257–68.
Sekigawa I, Naito T, Hira K et al (2004) Possible mechanisms of gender bias in SLE: a new hypothesis involving a comparison of SLE with atopy. Lupus 13:217–222.
Jiang X, Bai XY, Li B, Li Y, Xia K, Wang M, Li S, Wu H. Plasma Inter-Alpha-Trypsin Inhibitor Heavy Chains H3 and H4 Serve as Novel Diagnostic Biomarkers in Human Colorectal Cancer. Dis Markers. 2019 Jul 31;2019:5069614. doi: 10.1155/2019/5069614. PMID: 31481982.
Heron A, Bourguignon J, Calle A, Borghi H, Sesboue R, Diarra-Mehrpour M, Martin JP: Post-translational processing of the inter-alpha-trypsin inhibitor in the human hepatoma HepG2 cellline. Biochem J 1994, 302 (Pt 2):573–580.
Salier JP, Rouet P, Raguenez G, Daveau M: The inter-alpha-inhibitor family: from structure to regulation. Biochem J 1996, 315 (Pt 1):1–9.
de la Motte CA, Hascall VC, Drazba J, Bandyopadhyay SK, Strong SA: Mononuclear leukocytes bind to specific hyaluronan structures on colon mucosal smooth muscle cells treated with polyinosinic acid:polycytidylic acid: inter-alpha-trypsin inhibitor is crucial to structure and function. Am J Pathol 2003,163(1):121–133.
Ohnishi H, Kosuzume H, Ashida Y, Kato K, Honjo I: Effects of urinary trypsin inhibitor on pancreatic enzymes and experimental acute pancreatitis. Dig Dis Sci 1984, 29(1):26–32.
Liao CC, Chou PL, Cheng CW, et el. Comparative analysis of novel autoantibody isotypes against citrullinated-inter-alpha-trypsin inhibitor heavy chain 3 (ITIH3)(542–556) peptide in serum from Taiwanese females with rheumatoid arthritis, primary Sjögren's syndrome and secondary Sjögren's syndrome in rheumatoid arthritis. J Proteomics. 2016 Jun 1;141:1–11. doi: 10.1016/j.jprot.2016.03.031. Epub 2016 Apr 9. Erratum in: J Proteomics. 2019 Mar 1;194:227. PMID: 27072115.
Kikuchi H, Tanaka S, Matsuo O: Plasminogen activator in synovial fluid from patients with rheumatoid arthritis. J Rheumatol 1987, 14(3):439–445.
Opal SM, Lim YP, Siryaporn E, Moldawer LL, Pribble JP, Palardy JE, Souza S: Longitudinal studies of inter-alpha inhibitor proteins in severely septic patients: a potential clinical marker and mediator of severe sepsis. Crit Care Med 2007, 35(2):387–392.
Lu L, Hu C, Zhao Y, He L, Zhou J, Li H, et al. Shotgun Lipidomics Revealed Altered Profiles of Serum Lipids in Systemic Lupus Erythematosus Closely Associated with Disease Activity. Biomolecules. 2018 Oct 3;8(4):105. doi: 10.3390/biom8040105.
Yukovich M, Vostretsova K, Chen W, Avina-Zubieta JA. Overall and cause-specific mortality in pateints with systemic lupus erythematosus: a meta-analysis of observational studies. Arthritis Care Res (Hoboken). 2014;66:608–16

No competing interests reported.

Supplementarymaterials.pdf

Download PDF

Version 1

posted

You are reading this latest preprint version

Development and validation of a serum multi-omics based diagnostic model for systemic lupus erythematosus

Status:

Version 1

Abstract

Objective.

Methods.

Results.

Conclusion.

Figures

Introduction

Materials And Methods

Study populations

Serum proteomic sample preparation

LC-MS/MS analysis

Serum Metabolomics

Data Analysis

Results

Baseline characteristic

LASSO regressions

Diagnostic model

Evaluation and visualization of three models

Discussion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1