You are here

Article Text

Article menu
Lupus nephritis
Original research
iTRAQ-based mass spectrometry screen to identify serum biomarkers in systemic lupus erythematosus
  1. Kamala Vanarsa1,
  2. Ting Zhang2,3,
  3. Jack Hutcheson2,
  4. Sneha Ravi Kumar1,
  5. Satyavani Nukala2,
  6. Haleigh Inthavong1,
  7. Bruce Stanley4,
  8. Tianfu Wu1,
  9. C C Mok5,
  10. Ramesh Saxena6 and
  11. Chandra Mohan1
  1. 1Department Biomedical Engineering, University of Houston, Houston, Texas, USA
  2. 2University of Houston, Houston, Texas, USA
  3. 3Rheumatology, The Second Affiliated Hospital of Zhejiang Chinese Medical University, Hangzhou, Zhejiang, China
  4. 4Penn State, University Park, Pennsylvania, USA
  5. 5Medicine, Tuen Mun Hospital, Hong Kong
  6. 6The University of Texas Southwestern Medical Center, Dallas, Texas, USA
  1. Correspondence to Professor Chandra Mohan; cmohan{at}Central.UH.edu

Abstract

Objective Systemic lupus erythematosus (SLE) is a complex systemic autoimmune disorder with no reliable serum biomarkers currently available other than autoantibodies.

Methods In the present study, isobaric tags for relative and absolute quantitation-based mass spectrometry was used to screen the sera of patients with SLE to uncover potential disease biomarkers.

Results 85 common proteins were identified, with 16 being elevated (≥1.3) and 23 being decreased (≤0.7) in SLE. Of the 16 elevated proteins, serum alpha-1-microglobulin/bikunin precursor (AMBP), zinc alpha-2 glycoprotein (AZGP) and retinol-binding protein 4 (RBP4) were validated in independent cross-sectional cohorts (Cohort I, N=52; Cohort II, N=117) using an orthogonal platform, ELISA. Serum AMBP, AZGP and RBP4 were validated to be significantly elevated in both patients with inactive SLE and patients with active SLE compared with healthy controls (HCs) (p<0.05, fold change >2.5) in Cohort I. All three proteins exhibited good discriminatory power for distinguishing active SLE and inactive SLE (area under the curve=0.82–0.96), from HCs. Serum AMBP exhibited the largest fold change in active SLE (5.96) compared with HCs and correlated with renal disease activity. The elevation in serum AMBP was validated in a second cohort of patients with SLE of different ethnic origins, correlating with serum creatinine (r=0.60, p<0.001).

Conclusion Since serum AMBP is validated to be elevated in SLE and correlated with renal disease, the clinical utility of this novel biomarker warrants further analysis in longitudinal cohorts of patients with lupus and lupus nephritis.

  • Autoantibodies
  • Autoimmune Diseases
  • Lupus Erythematosus, Systemic
  • Lupus Nephritis

Data availability statement

Data are available in a public, open access repository.

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

https://doi.org/10.1136/lupus-2022-000673

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    WHAT IS ALREADY KNOWN ON THIS TOPIC

    • Besides autoantibodies, few other serum proteins have disease-predictive potential in lupus.

    WHAT THIS STUDY ADDS

    • In a mass spectrometry-based screen, 16 proteins were noted to be elevated in lupus serum, highlighting the interaction between autoimmunity and metabolism.

    • Serum alpha-1-microglobulin/bikunin precursor (AMBP) was significantly correlated with renal systemic lupus erythematosus in two different cohorts.

    HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

    • The diagnostic potential of serum AMBP warrants further analysis in additional cross-sectional and longitudinal cohorts of patients with lupus.

    Introduction

    Systemic lupus erythematosus (SLE) is a complex autoimmune disorder affecting multiple organs with various manifestations. While some patients may only present with mild mucocutaneous symptoms, others may become critically ill with active life-threatening involvement. Thus, an accurate assessment of disease activities in patients with SLE is essential for timely treatment and optimal outcomes. Conventional biomarkers such as antinuclear antibodies, anti-dsDNA antibodies and complements help physicians diagnose patients with SLE. However, these conventional biomarkers do not always reflect disease activity or flare.1 Additional reliable markers are needed for tracking disease and optimising treatment in SLE.

    Two proteomic discovery methods are commonly used to identify biomarkers. The first is targeted or focused proteomics using antibody or aptamer libraries against known antigens. We have previously documented multiple proteins, such as angiostatin, ferritin, activated leucocyte cell adhesion molecule and vascular cell adhesion molecule (VCAM)-1, among others, as potential SLE biomarkers using different focused proteomic platforms.2–6 However, this approach is limited by the quality and availability of antibodies or other ligands. An alternative method is to use unbiased proteomics. In this approach, labelled protein fragments are detected by mass spectrometry and reconstructed using computational methods. These methods theoretically allow investigation of the entire proteome. They include stable isotope labelling by amino acids in cell culture, isotope-coated affinity tag and isobaric tags for relative and absolute quantitation (iTRAQ), among others.

    iTRAQ accurately quantifies proteomic differences between small samples of patient sera. In brief, sera from up to eight subjects are digested, labelled individually with unique isobaric tags, pooled and subjected to mass spectrometry. The isobaric tags shift the mass-to-charge ratio (m/z) of each labelled peptide fragment to uniquely identify its source. Sample proteins are computationally reconstructed from identified fragments using known peptide libraries.7 Indeed, iTRAQ allows direct comparison between groups and has the potential to uncover post-translational modifications and protein isoforms.8

    In this work, we test the hypothesis that unbiased proteomics using iTRAQ has the potential to uncover novel biomarkers for SLE. The first aim of the study is to interrogate sera from patients with SLE and healthy controls (HCs) using iTRAQ, as diagrammed in online supplemental figure 1A. Second, proteins that discriminate patients with SLE from HCs are further validated in independent larger cohorts of patients with SLE, to assess if any of the identified biomarkers are predictive of disease activity.

    Supplemental material

    Materials and methods

    Patients and samples

    Patients were recruited from the renal clinic at the University of Texas Southwestern (UTSW) Medical Center, Dallas, Texas (Cohort I) (table 1) and Tuen Mun Hospital, Hong Kong, China (Cohort II) (table 2). All patients with SLE met the 2012 Systemic Lupus International Collaborating Clinics classification criteria for SLE. Clinical manifestations, SLE Disease Activity Index (SLEDAI), renal domain of SLEDAI (rSLEDAI), physician global assessment (PGA) and laboratory metrics including complement C3, C4, anti-dsDNA antibody, blood urea nitrogen (BUN) and creatinine were collected. Patients with SLE with SLEDAI scores ≤4 were categorised as having inactive SLE, and patients with SLEDAI scores ≥5 were categorised as having active SLE. Matched healthy individuals were enrolled as controls. Simple random sampling was used to retrieve samples from the archived Cohorts, with the operator being blinded to the clinical features or groupings of the subjects. Serum samples were obtained, aliquoted and stored at −80℃ until use.

    View this table:
    Table 1

    Characteristics of the UTSW validation cohort (Cohort I)

    View this table:
    Table 2

    Characteristics of the validation cohort from Hong Kong (Cohort II)

    High-abundance protein depletion and sample desalting

    A total of nine patients with SLE and five HCs from the UTSW cohort I were used for the iTRAQ screen (online supplemental table 1). Three sets of iTRAQ experiments were performed, with one HC sample (‘calibrator’) being included in all three runs, that was used to normalise all samples in all iTRAQ experiments. In the first set of 4-plex experiments, serum from the calibrator HC and three patients with SLE were included. In the second 4-plex experiment, serum from the calibrator HC, additional HC and two patients with SLE was included. In the final 8-plex experiment, serum from the calibrator HC, three additional HCs and four patients with SLE was included.

    14 µL of serum samples were added to 184 µL of buffer A (Agilent), loaded onto a 0.22 µm cartridge (Agilent 5185-5990) and centrifuged for 5 min at 14 000 rpm to remove debris before loading. The eluate was then spun on a high-capacity multiple affinity removal spin cartridge (MARS, Agilent #5188-5341) per the manufacturer’s instructions to deplete the six most abundant proteins: albumin, IgG, antitrypsin, IgA, transferrin and haptoglobin. Each cartridge was equilibrated with 4 mL of buffer A (Agilent) using a Luer Lock syringe. 200 µL of depleted eluate was centrifuged through the column, and flow-through from three washes was collected. The flow-through comprised of low-abundance proteins. The low-abundance flow-through was then added to spin concentrators (Agilent 5185-5991) and centrifuged for 20 min at 3800×g. The columns were washed three times with 50 mM NaHCO3, and the concentrated samples were recovered from the column. Serum samples before and after depletion of abundant proteins were collected and run through a sodium dodecyl sulphate (SDS)-polyacrylamide gel electrophoresis gel to verify depletion of the high-abundance proteins (online supplemental figure 1B).

    Protein digestion and iTRAQ labelling

    The protein concentration in each sample was measured using bicinchoninic acid assay per the manufacturer’s instructions (Fisher #23225). 100 µg of each sample were reconstituted to a concentration of 5 µg/µL in dissolution buffer (0.5 M triethyl ammonium bicarbonate, pH 8.5). One microlitre of denaturing agent (2% SDS) was added, and each sample was vortexed. One microlitre of reducing agent (100 mM tris(2-carboxyethyl)phosphine) was added, and each sample was vortexed. Tubes were incubated at 60°C for 1 hour, spun briefly, supplemented with 1 µL of 84 mM iodoacetamide (prepared fresh), vortexed, spun again and incubated for 30 min in the dark at room temperature. 10 µL of sequencing-grade trypsin (Promega #V5111, 1 mg/mL in 50 mM acetic acid) was added to digest each sample overnight at 48°C.

    Digested samples were spun down and brought to room temperature. iTRAQ tag vials were reconstituted with 70 µL of ethanol. The 4-plex tags had m/z ratios of 114, 115, 116 and 117. The 8-plex tags comprised the 4-plex tags plus tags with m/z ratios of 113, 118, 119 and 121. Each separate tag was added to an individual sample, vortexed, spun and incubated at room temperature for one (4-plex) or 2 hours (8-plex). Samples were repeatedly pooled, vortexed, spun, freeze-dried and reconstituted with 100 µL of water to remove denaturing and reducing reagents that could potentially interfere with mass spectrometry. The sample digestion, iTRAQ labelling, nanoflow Liquid Chromatography (LC) peptide separation and spotting on Matrix-assisted laser desorption/ionisation (MALDI) plates and MS-MS/MS (Mass Spectrometry) analysis were performed at the Penn State College of Medicine Mass Spectrometry and Proteomics Core (https://scicrunch.org/resolver/RRID:SCR_017831)).

    ELISA validation

    Serum levels of alpha-1-microglobulin/bikunin precursor (AMBP) (Abcam, ab108884, 1:10 000 dilution), zinc alpha-2 glycoprotein (AZGP) (Raybio, EIA-ZAG, 1:2 dilution) and retinol-binding protein 4 (RBP4) (R&D, DRB400, 1:1000 dilution) were measured in healthy (n=11), inactive SLE (n=20) and active SLE (n=21) patients from the UTSW cohort (Cohort I) comprised of African-American and Hispanic individuals using an ELISA. Similarly, serum levels of AMBP and RBP4 were further measured in healthy (n=33), patients with inactive SLE (n=35) and active SLE (n=49) in the Chinese cohort (Cohort II). Briefly, appropriately diluted samples were added to capture antibody-coated plates, incubated and washed, after which biotinylated detection antibody, streptavidin-conjugated horseradish peroxidase, substrate reagent and stop solution were added stepwise. Optical densities were read at 450 nm using a microplate reader (ELX808 from BioTek Instruments, Winooski, Vermont, USA).

    Statistical analysis

    For mass spectrometry analysis, samples were analysed using standard mass spectrometry methods. In brief, labelled digested samples were ionised, transmitted through an electric field and detected to generate a histogram of mass-to--charge (m/z) ratios for each fragment. The resultant data were mapped to the respective parent proteins using Mascot advanced protein-MS ID software. For ELISA validation, data were plotted and analysed using GraphPad Prism V.5 (GraphPad, San Diego, California, USA) or R (R Foundation for Statistical Computing, Vienna, Austria).

    Group comparisons were analysed using the Mann-Whitney U test or χ2 test. Non-parametric Spearman correlation was performed for correlation analysis. Receiver operating characteristic (ROC) curves were used to compare the performance of biomarker candidates in discriminating both patients with SLE and healthy controls and active SLE and inactive SLE. Heatmaps and volcano plots were made using R. A two-tailed p value <0.05 was considered significant.

    Results

    Depletion of high-abundance serum proteins prior to iTRAQ analysis

    In mass spectrometry-based proteomic analyses, it is customary to remove the highest abundant proteins so that they do not overwhelm the peptide signal readouts, in order to discern lower-abundant proteins. This is important because most biomarkers are present at very low quantities in circulation. Accordingly, the following six high-abundance proteins were depleted from the serum samples using a multi-affinity spin cartridge (bearing antibodies to these six proteins): transferrin (≈80 kDa), albumin (≈66 kDa), antitrypsin (≈53 kDa), IgG (≈150 kDa), IgA (≈160 kDa) and haptoglobin (≈85 kDa). As shown in online supplemental figure 1B, the samples were effectively depleted of these high-abundance proteins.

    Several proteins are consistently elevated or reduced in the sera of patients with SLE versus HCs, as determined by iTRAQ proteomics

    Three independent iTRAQ screens were performed (online supplemental figure 2), and a common calibrator control was used to calibrate these screens. Each screen identified and quantified between 110 and 130 proteins in each iTRAQ run, as detailed in online supplemental figure 2. Of the 85 proteins commonly identified by iTRAQ across all three experiments (figure 1A), 16 were overexpressed (≥1.3 fold), including 6 that were significantly (p<0.05) elevated, and 23 were underexpressed (≤0.7 fold), including 4 that were significantly (p<0.05) downregulated, in SLE sera compared with HC sera (figure 1B). The fold changes comparing patients to HCs are presented as bar graphs in figure 2.

    Figure 1
    Figure 1

    Identification of serum proteins in SLE using iTRAQ. (A) A total of 85 quantifiable proteins were identified in common across all three iTRAQ experiments and these are displayed as a heatmap. Each column represents a patient sample, while rows correspond to protein level as assayed by the iTRAQ study. Yellow represents upregulation and blue represents downregulation, comparing SLE sera to HC sera. Black indicates no difference between patients with SLE and HC. (B) Volcano plot displaying the upregulated and downregulated proteins in SLE sera compared with HC sera when comparing log2 fold change (FC) of protein expression versus the negative log10 p value, that is, biological significance versus statistical significance. Each dot represents a protein and its average value for that subset (SLE vs HC). HC, healthy control; iTRAQ, isobaric tags for relative and absolute quantitation; SLE, systemic lupus erythematosus.

    Figure 2
    Figure 2

    Upregulated and downregulated proteins in systemic lupus erythematosus (SLE) serum. A total of 85 overlapping proteins were identified and quantified in common across three independent isobaric tags for relative and absolute quantitation screens, comparing SLE sera to healthy control (HC) sera. Of these, 39 proteins were overexpressed (≥1.3 fold) or underexpressed(≤0.7 fold) in SLE sera versus HC sera. Data were log-transformed and filtered for fold changes ±30% or more, comparing SLE to healthy sera. Fold change was calculated from averages of HC and SLE samples. *p<0.05, SLE versus HCs.

    Validation of iTRAQ proteomic screen results using ELISA in independent SLE cohorts

    For the top-most eight proteins (as ranked in figure 2) that were elevated in SLE sera, and for which ELISA kits were commercially available, we were able to identify functional ELISA kits for three candidates—AMBP, AZGP and RBP4. Moreover, supporting renal literature suggested that these may be candidates of potential importance in SLE or glomerulonephritis. For instance, the alpha-1-microglobulin component of AMBP is markedly elevated in urine of patients with glomerulonephritis,9 while urinary RBP4 is significantly elevated in patients with SLE with active lupus nephritis (LN).10

    Measurement of AMBP, AZGP and RBP4 in Cohort I

    Serum levels of AMBP, AZGP and RBP4 were assayed by ELISA in Cohort I, comprised predominantly of African American and Hispanic patients and analysed using Student’s t-test to determine if they were significantly different in SLE. Serum AMBP, AZGP and RBP4 were all found to be significantly elevated in both patients with inactive SLE and patients with active SLE compared with HCs (p<0.05, fold change >2.5) (figure 3A). Of these proteins, AZGP showed the greatest discriminatory power by ROC analysis (AUC=0.93 for inactive SLE, 0.96 for active SLE), while AMBP exhibited the largest change in concentration in active SLE and inactive patients with SLE compared with HCs (fold change=5.96 for active SLE and 4.39 for inactive SLE), as shown in figure 3B. Correlation analysis was further performed relating the serum levels of these three proteins and established clinical parameters, including C3, SLEDAI, rSLEDAI (ie, just the four renal components of SLEDAI, summated) and serum creatinine. While AMBP showed a significant correlation with rSLEDAI, and RBP4 with BUN, the remaining correlations did not reach statistical significance (figure 4).

    Figure 3
    Figure 3

    ELISA validation of hits from iTRAQ screen in Cohort I. (A) ELISA validation of AMBP, AZGP and RBP4 levels in serum samples from healthy controls, patients with inactive SLE and patients with active SLE in Cohort I (from UTSW). (B) ROC curves of serum AMBP, AZGP and RBP4, displaying their ability to distinguish different groups of subjects, as indicated. *p<0.05; **p<0.01; ***p<0.001; ****p<0.0001. AMBP, alpha-1-microglobulin/bikunin precursor; AUC, area under the curve; AZGP, zinc alpha-2 glycoprotein; iTRAQ, isobaric tags for relative and absolute quantitation; RBP4, retinol-binding protein 4; ROC, receiver operating characteristic; SLE, systemic lupus erythematosus; UTSW, University of Texas Southwestern.

    Figure 4
    Figure 4

    Correlation of serum biomarkers with clinical parameters in Cohort I. Plotted are 2-D correlation plots relating ELISA-tested serum proteins to various clinical/laboratory parameters, including C3, SLEDAI, rSLEDAI (ie, the renal components of SLEDAI) and serum creatinine in Cohort I (from UTSW). Analyses were done using Spearman correlation. Corresponding r and p values are indicated in each plot. AMBP, alpha-1-microglobulin/bikunin precursor; AZGP, zinc alpha-2 glycoprotein; RBP4, retinol-binding protein 4; rSLEDAI, renal domain of SLEDAI; SLEDAI, Systemic Lupus Erythematosus Disease Activity Index; UTSW, University of Texas Southwestern.

    Validation of increase in serum AMBP and RBP4 in a second cohort: Cohort II

    Serum levels of AMBP and RBP4 were analysed in patients with inactive SLE, patients with active SLE and HCs from Cohort II, comprised of patients of Chinese origin. AZGP was not tested in this cohort due to insufficient samples. The levels of serum AMBP showed statistically significant differences among the three groups, as shown in figure 5A (p<0.01). ROC analysis indicated that the AUC values for discriminating patients with inactive SLE from HCs, patients with active SLE from HCs and patients with inactive SLE from patients with active SLE were 0.75, 0.87 and 0.70, respectively (figure 5B). Serum AMBP showed a significant correlation with C3, C4, PGA, anti-DNA levels and serum creatinine (p<0.05), as plotted in figure 6. Particularly striking was the strong correlation of serum AMBP with serum creatinine (r=0.60; p<0.001). In contrast to AMBP, serum RBP4 was not significantly higher in Chinese patients with SLE, compared with HCs. Possible reasons for the differences in the two cohorts (Cohort I vs Cohort II) are discussed below.

    Figure 5
    Figure 5

    (A) ELISA validation of AMBP and RBP4 in serum samples of healthy controls, patients with inactive SLE and patients with active SLE in Cohort II (from Hong Kong). (B) ROC curves of serum AMBP and RBP4, displaying their ability to distinguish different groups of subjects, as indicated. ns, non-significant; **p<0.01; ***p<0.001; ****p<0.0001. AMBP, alpha-1-microglobulin/bikunin precursor; AUC, area under the curve; RBP4, retinol-binding protein 4; ROC, receiver operating characteristic; SLE, systemic lupus erythematosus.

    Figure 6
    Figure 6

    Correlation of serum biomarkers with clinical parameters in Cohort II. Plotted are 2-D correlation plots relating ELISA-tested serum proteins to various clinical/laboratory parameters, including PGA, SLEDAI, anti-DNA, C3, C4 and serum creatinine in subjects from Cohort II (from Hong Kong). Analyses were done using Spearman correlation. Corresponding r and p values are indicated in each plot. AMBP, alpha-1-microglobulin/bikunin precursor; PGA, physician global assessment; SLEDAI, Systemic Lupus Erythematosus Disease Activity Index.

    Discussion

    iTRAQ is based on prelabelling peptides with mass tags, which allows parallel multiplex analysis and comparison of up to eight proteomes.11 Thus, iTRAQ has the potential to be a key tool in the area of quantitative proteomics by identifying and quantifying proteins.12 However, iTRAQ has not been extensively applied in biomarker discovery in SLE. Few studies have used iTRAQ to identify or validate proteins in tissues and peripheral blood mononuclear cells (PBMCs), and in one study in the urine of patients with SLE, several potential biomarkers have been revealed.12–16 In Murphy Roths Large/lymphoproliferative (MRL/lpr) mice, kidney tissue has been interrogated using iTRAQ, revealing increases in haptoglobin, immunoglobulin kappa v8-27 and integrin beta chain-2 in lupus.13 Proteins in the renal tissue of patients with LN, PBMCs of patients with SLE and LN urine have also been investigated using iTRAQ.12–16 To the best of our knowledge, iTRAQ has not been used to investigate the serum proteome in patients with SLE.

    Several potential serum biomarkers were uncovered by iTRAQ screening in this study. AZGP (a lipid mobilising factor) is a soluble adipokine produced in the skin, lung, pancreas, liver and kidney.17 While evolutionarily related to Major Histocompatibility Complex (MHC) class I, AZGP binds neither short peptides nor β2 microglobulin.18 19_ENREF_4 AZGP induces β3 adrenoreceptor-dependent lipolysis and mitochondrial uncoupling protein activity, which lead to cachexia.20 21 AZGP elevation has been associated with cachexia in chronic inflammatory conditions such as smoker’s lung and cystic disease of the breast and may also contribute to SLE.22–25 In this study, AZGP was significantly elevated in patients with active SLE and patients with inactive SLE compared with HCs._ENREF_11 AZGP also acts in the kidney to transport nephritogenic glycoprotein and may thus further contribute to nephritis in patients with SLE.26 27 Given that serum AZGP does not differentiate between inactive and active SLE and does not correlate with any clinical metric of disease in SLE, its use in clinical practice may be limited. _ENREF_7 Retinol binding protein 4 (RBP4) is a vitamin A carrier and member of the lipocalin family mainly expressed in liver. Interestingly, serum vitamin A is low in patients with SLE or other rheumatic diseases.28 RBP4 functions as an adipokine in serum, contributing to insulin resistance in adipose-specific Glut4 knockout mice.29 30 RBP4 also plays a role in inflammation by inducing the production of VCAM-1, Intercellular Adhesion Molecule 1 (ICAM-1), E-selectin, monocyte chemoattractant protein-1 (MCP-1) and interleukin-6 in endothelial cells, all of which have been implicated as biomarkers in SLE. In one study, urinary RBP4 was significantly elevated in patients with LN, especially in those with active disease.10

    In our study, serum RBP4 is significantly elevated in patients with SLE compared with HCs in Cohort I, which is comprised predominantly of African-American and Hispanic subjects, but not in Cohort II, composed of Chinese subjects. A couple of differences between these cohorts may have contributed to these differences. First and foremost, the two cohorts have strikingly different ethnicities. Whether genetic polymorphisms that dictate the expression levels of RBP4, directly or indirectly, are at play is currently unknown. Second, clinical manifestations in the two cohorts also vary. For example, a higher percentage of the patients in Cohort I had renal disease, while higher percentages of patients in Cohort II had mucocutaneous and/or musculoskeletal diseases, as detailed in tables 1 and 2. Indeed, previous reports have described symptom-specific proteins in SLE serum.31 Hence, it is tempting to speculate that the increased serum RBP4 levels in Cohort I may have been driven by the increased numbers of African-American and Hispanic subjects with renal involvement, and this needs to be examined in more detail using larger cohorts with subset analyses. However, since serum RBP4 does not differentiate between inactive and active SLE and does not correlate with any clinical metric of disease in SLE, its use in clinical practice may be limited.

    The AMBP gene encodes two plasma glycoproteins: alpha-1-microglobulin, an immunosuppressive lipocalin, and bikunin, a member of the plasma serine proteinase inhibitor family. In the liver, the precursor protein is cleaved by a furin-like enzyme to release the constituent proteins.32 As a lipocalin, AMBP removes free radicals and oxidising agents in the serum, including haem, and exerts immunosuppressive effects.33 34 In our study, serum AMBP is significantly upregulated in patients with inactive SLE and patients with active SLE compared with HCs in both cohorts. Importantly, serum AMBP could discriminate patients with inactive SLE from patients with active SLE in the Chinese cohort. It is tempting to speculate that serum AMBP did not discriminate patients with inactive SLE from patients with active SLE in Cohort I since 35% of the patients with inactive SLE had a history of LN, raising the interesting possibility that serum AMBP may be a marker of past and active LN. The significant correlations between serum AMBP and several clinical parameters in both cohorts, including PGA, renal components of SLEDAI, C3, C4, anti-DNA antibodies and serum creatinine, highlight the potential utility of AMBP in monitoring disease status and renal involvement in SLE.

    Several aspects of this study could have been improved. It would be important to extend these studies to larger cohorts of cross-sectional and longitudinal samples in order to confirm if serum AMBP is indeed predictive of past or active disease activity or renal disease in SLE and whether it can be used to track disease activity serially during patient follow-up. Larger cohorts will also allow adjustment for confounding factors including ethnicity, disease manifestations as well as medication history. Given that there were also differences in medication use between these two cohorts, these need to be further investigated as potential confounding factors. Efforts should also be undertaken to examine treatment naïve SLE samples, before commencing induction therapy. Finally, mechanistic studies are warranted to explore how AMBP might contribute to disease pathogenesis in SLE, given its potential association with disease activity and renal status.

    Data availability statement

    Data are available in a public, open access repository.

    Ethics statements

    Patient consent for publication

    Ethics approval

    This study was approved by the institutional review boards of UTSW, Tuen Mun Hospital and the University of Houston. Informed consent was obtained from each patient. Participants gave informed consent to participate in the study before taking part.

    Acknowledgments

    We acknowledge the assistance of Dr Tianfu Wu, Haleigh Inthavong and Sonja Vodehnal.

    References

    1. ↵
      1. Arriens C,
      2. Wren JD,
      3. Munroe ME, et al
      . Systemic lupus erythematosus biomarkers: the challenging quest. Rheumatology (Oxford) 2017;56:i32–45. doi:10.1093/rheumatology/kew407
      OpenUrl
    2. ↵
      1. Wu T,
      2. Xie C,
      3. Wang HW, et al
      . Elevated urinary VCAM-1, P-selectin, soluble TNF receptor-1, and CXC chemokine ligand 16 in multiple murine lupus strains and human lupus nephritis. J Immunol 2007;179:7166–75. doi:10.4049/jimmunol.179.10.7166
      OpenUrlAbstract/FREE Full Text
    3. ↵
      1. Celiac Disease Center
      . Celiac disease facts and figures. Chicago, USA: University of Chicago, 2011.
    4. ↵
      1. Spits H,
      2. Artis D,
      3. Colonna M, et al
      . Innate Lymphoid cells--a proposal for uniform nomenclature. Nat Rev Immunol 2013;13:145–9. doi:10.1038/nri3365
      OpenUrlCrossRefPubMed
    5. ↵
      1. Stanley S,
      2. Vanarsa K,
      3. Soliman S, et al
      . Comprehensive aptamer-based screening identifies a spectrum of urinary biomarkers of lupus nephritis across ethnicities. Nat Commun 2020;11:2197. doi:10.1038/s41467-020-15986-3
    6. ↵
      1. Vanarsa K,
      2. Soomro S,
      3. Zhang T, et al
      . Quantitative planar array screen of 1000 proteins uncovers novel urinary protein biomarkers of lupus nephritis. Ann Rheum Dis 2020;79:1349–61. doi:10.1136/annrheumdis-2019-216312
      OpenUrlAbstract/FREE Full Text
    7. ↵
      1. Ross PL,
      2. Huang YN,
      3. Marchese JN, et al
      . Multiplexed protein quantitation in Saccharomyces cerevisiae using amine-reactive Isobaric tagging reagents. Mol Cell Proteomics 2004;3:1154–69. doi:10.1074/mcp.M400129-MCP200
      OpenUrlAbstract/FREE Full Text
    8. ↵
      1. Zieske LR
      . A perspective on the use of iTRAQ™ reagent technology for protein complex and profiling studies. J Exp Bot 2006;57:1501–8. doi:10.1093/jxb/erj168
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. Scherberich JE,
      2. Mösbauer A,
      3. Meissner A, et al
      . Kidney- and serum derived proteins in urine of patients suffering from renal diseases or arterial hypertension. Klin Wochenschr 1989;67 Suppl 17:44–7.
      OpenUrl
    10. ↵
      1. Go DJ,
      2. Lee JY,
      3. Kang MJ, et al
      . Urinary vitamin D-binding protein, a novel biomarker for lupus nephritis, predicts the development of proteinuric flare. Lupus 2018;27:1600–15. doi:10.1177/0961203318778774
      OpenUrl
    11. ↵
      1. Núñez EV,
      2. Domont GB,
      3. Nogueira FCS
      . iTRAQ-based shotgun proteomics approach for relative protein quantification. Methods Mol Biol 2017;1546:267–74. doi:10.1007/978-1-4939-6730-8_23
      OpenUrl
    12. ↵
      1. Wang L,
      2. Dai Y,
      3. Qi S, et al
      . Comparative proteome analysis of peripheral blood mononuclear cells in systemic lupus erythematosus with iTRAQ quantitative proteomics. Rheumatol Int 2012;32:585–93. doi:10.1007/s00296-010-1625-9
      OpenUrlPubMed
    13. ↵
      1. Liao D-J,
      2. Cheng X-P,
      3. Li N, et al
      . A comparative study on the incidence, aggravation, and remission of lupus nephritis based on iTRAQ technology. Comb Chem High Throughput Screen 2020;23:649–57. doi:10.2174/1386207323666200416151836
      OpenUrl
    14. ↵
      1. Sui W,
      2. Tang D,
      3. Zou G, et al
      . Differential proteomic analysis of renal tissue in lupus nephritis using iTRAQ reagent technology. Rheumatol Int 2012;32:3537–43. doi:10.1007/s00296-011-2207-1
      OpenUrlPubMed
    15. ↵
      1. Wang L,
      2. Dai Y,
      3. Peng W, et al
      . Differential expression of serine-threonine kinase receptor-associated protein in patients with systemic lupus erythematosus. Lupus 2011;20:921–7. doi:10.1177/0961203311399304
      OpenUrlCrossRefPubMed
    16. ↵
      1. Turnier JL,
      2. Brunner HI,
      3. Bennett M, et al
      . Discovery of SERPINA3 as a candidate urinary biomarker of lupus nephritis activity. Rheumatology (Oxford) 2019;58:321–30. doi:10.1093/rheumatology/key301
      OpenUrl
    17. ↵
      1. Tada T,
      2. Ohkubo I,
      3. Niwa M, et al
      . Immunohistochemical localization of Zn-alpha 2-glycoprotein in normal human tissues. J Histochem Cytochem 1991;39:1221–6. doi:10.1177/39.9.1918940
      OpenUrlCrossRefPubMedWeb of Science
    18. ↵
      1. Freije JP,
      2. Fueyo A,
      3. Uría JA, et al
      . Human Zn-Α2-glycoprotein: complete genomic sequence, identification of a related pseudogene and relationship to class I major histocompatibility complex genes. Genomics 1993;18:575–87. doi:10.1016/s0888-7543(05)80359-3
      OpenUrlCrossRefPubMedWeb of Science
    19. ↵
      1. Sánchez LM,
      2. López-Otín C,
      3. Bjorkman PJ
      . Biochemical characterization and crystalization of human Zn-Α 2 -glycoprotein, a soluble class I major histocompatibility complex homolog. Proc Natl Acad Sci U S A 1997;94:4626–30. doi:10.1073/pnas.94.9.4626
      OpenUrlAbstract/FREE Full Text
    20. ↵
      1. Hassan MI,
      2. Waheed A,
      3. Yadav S, et al
      . Zinc alpha 2-glycoprotein: a multidisciplinary protein. Mol Cancer Res 2008;6:892–906. doi:10.1158/1541-7786.MCR-07-2195
      OpenUrlAbstract/FREE Full Text
    21. ↵
      1. Russell ST,
      2. Zimmerman TP,
      3. Domin BA, et al
      . Induction of lipolysis in vitro and loss of body fat in vivo by zinc-Alpha2-glycoprotein. Biochim Biophys Acta 2004;1636:59–68. doi:10.1016/j.bbalip.2003.12.004
      OpenUrlCrossRefPubMed
    22. ↵
      1. Garcia-Gonzalez A,
      2. Gonzalez-Lopez L,
      3. Valera-Gonzalez IC, et al
      . Serum leptin levels in women with systemic lupus erythematosus. Rheumatol Int 2002;22:138–41. doi:10.1007/s00296-002-0216-9
      OpenUrlCrossRefPubMedWeb of Science
    23. ↵
      1. Vanni H,
      2. Kazeros A,
      3. Wang R, et al
      . Cigarette smoking induces overexpression of a fat-depleting gene Azgp1 in the human. Chest 2009;135:1197–208. doi:10.1378/chest.08-1024
      OpenUrlCrossRefPubMed
    24. ↵
      1. Freije JP,
      2. Fueyo A,
      3. Uría J, et al
      . Human Zn-Α2-glycoprotein cDNA cloning and expression analysis in benign and malignant breast tissues. FEBS Lett 1991;290:247–9. doi:10.1016/0014-5793(91)81271-9
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      1. Díez-Itza I,
      2. Sánchez LM,
      3. Allende MT, et al
      . Zn-alpha 2-glycoprotein levels in breast cancer cytosols and correlation with clinical, histological and biochemical parameters. Eur J Cancer 1993;29A:1256–60. doi:10.1016/0959-8049(93)90068-q
      OpenUrlCrossRef
    26. ↵
      1. Shibata S,
      2. Miura K
      . Plasma Zn-Alpha2-glycoprotein as a second source of nephritogenic glycoprotein in urine. Nephron 1982;31:170–6. doi:10.1159/000182638
      OpenUrlPubMedWeb of Science
    27. ↵
      1. Gordon JN,
      2. Green SR,
      3. Goggin PM
      . Cancer cachexia. QJM 2005;98:779–88. doi:10.1093/qjmed/hci127
      OpenUrlCrossRefPubMedWeb of Science
    28. ↵
      1. Comstock GW,
      2. Burke AE,
      3. Hoffman SC, et al
      . Serum concentrations of Α tocopherol, Β carotene, and retinol preceding the diagnosis of rheumatoid arthritis and systemic lupus erythematosus. Ann Rheum Dis 1997;56:323–5. doi:10.1136/ard.56.5.323
      OpenUrlAbstract/FREE Full Text
    29. ↵
      1. Yang Q,
      2. Graham TE,
      3. Mody N, et al
      . Serum retinol binding protein 4 contributes to insulin resistance in obesity and type 2 diabetes. Nature 2005;436:356–62. doi:10.1038/nature03711
      OpenUrlCrossRefPubMedWeb of Science
    30. ↵
      1. Kanai M,
      2. Raz A,
      3. Goodman DWS
      . Retinol-binding protein: the transport protein for vitamin A in human plasma. J Clin Invest 1968;47:2025–44. doi:10.1172/JCI105889
      OpenUrlCrossRefPubMedWeb of Science
    31. ↵
      1. Lefranc D,
      2. Launay D,
      3. Dubucquoi S, et al
      . Characterization of discriminant human brain antigenic targets in neuropsychiatric systemic lupus erythematosus using an immunoproteomic approach. Arthritis Rheum 2007;56:3420–32. doi:10.1002/art.22863
      OpenUrlCrossRefPubMedWeb of Science
    32. ↵
      1. Bratt T,
      2. Olsson H,
      3. Sjöberg EM, et al
      . Cleavage of the alpha 1-microglobulin-bikunin precursor is localized to the Golgi apparatus of rat liver cells. Biochim Biophys Acta 1993;1157:147–54. doi:10.1016/0304-4165(93)90058-g
      OpenUrlPubMed
    33. ↵
      1. Kumar S,
      2. Bandyopadhyay U
      . Free Heme toxicity and its detoxification systems in human. Toxicol Lett 2005;157:175–88. doi:10.1016/j.toxlet.2005.03.004
      OpenUrlCrossRefPubMedWeb of Science
    34. ↵
      1. Akerström B,
      2. Lögdberg L,
      3. Berggård T, et al
      . Alpha(1)-Microglobulin: a yellow-brown lipocalin. Biochim Biophys Acta 2000;1482:172–84. doi:10.1016/s0167-4838(00)00157-6
      OpenUrlCrossRefPubMed

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Footnotes

    • KV and TZ contributed equally.

    • Correction notice This article has been corrected since it was published. Author name ‘C Mok’ has been corrected to ‘C C Mok’.

    • Contributors CMohan conceived and planned the study. KV, SRK, SN and BS undertook the experiments. KV, JH, TZ and CMohan analysed the data and wrote the manuscript. CMok and RS provided the clinical samples. CMohan serves as the guarantor. All authors read the manuscript and approved the contents.

    • Funding This work was supported by NIH, Grant numbers R01 AR074096.

    • Competing interests None declared.

    • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.