Introduction
B cell regulation plays an important role in the pathogenesis of systemic lupus erythematosus (SLE) and lupus nephritis (LN) [
1,
2]. Many approaches have targeted blocking specific pathways of B cell activation for the treatment of the disease. It has also been shown that when B cells were depleted, this could ameliorate the progression of the disease [
3‐
7]. Anti-CD20 monoclonal antibody was able to normalize autoantibodies and serum complement levels even though its clinical efficacy for lupus nephritis was a failure in clinical trials [
8]. However, selection criteria for participants in the trials are vital, and this may partly be the explanation for the failure. A prognostic biomarker would therefore greatly help future clinical trials in selecting more appropriated participants.
B lymphocyte stimulator (BLyS) and a proliferation-inducing ligand (APRIL) are associated with mechanisms of disease. These proteins are found in plasma and membranes of monocytes, macrophages and dendritic cells. Both can activate B cells and plasma cells to proliferate, differentiate and engage in immunoglobulin G (IgG) class switching [
9‐
11]. Overexpression of BLyS has been shown to cause arthritis and nephritis in mice [
12] and by blocking its receptor, one could prevent the development of LN [
13]. High serum levels of both proteins are found in SLE patients [
14,
15]. Furthermore, aberrant expression of both cytokines in memory and plasma cells were tightly correlated with the disease activity score and autoantibody production [
16].
Both APRIL and BLyS are found in local tissue like lymph node, bone marrow, synovium and kidney [
17‐
20]. APRIL was also abundant in the cerebrospinal fluid of neuropsychiatric lupus [
21]. Since APRIL protein was predominantly found in active lupus nephritis and was locally produced by the glomerular mesangium [
22], therefore it may be able to determine the prognosis for lupus nephritis patients. In this study, we assessed the levels of serum and intrarenal APRIL levels in a prospective cohort of active lupus nephritis and its correlation with response to treatment.
Materials and methods
Research design
This is a longitudinal study that was conducted from March 2005 through June 2010 at the Lupus Research Unit, Department of Medicine, King Chulalongkorn Memorial Hospital (KCMH). All samples were stored at -80°C until analysis.
The study was approved by the Ethics Committee for Human Research of the Faculty of Medicine, Chulalongkorn University (IRB No. 201/52), and written informed consent was obtained from all patients.
Study population
All patients who had an SLE classification according to the 1997 criteria established by the American College of Rheumatology (ACR) were recruited from KCMH. Fifty-two patients with active LN were enrolled into the study. The criteria for active LN were urine proteinuria > 0.5 g/day and urine red blood cells (RBC)/white blood cells (WBC) > 5/high power field (HPF) or increased serum creatinine > 0.3 mg/dL [
23]. Active LN patients were also required to have a kidney biopsy with an adequate number of glomeruli (more than 10) so the severity of LN could be evaluated.
Twenty-seven patients with biopsy-proven ISN/RPS class III/IV LN were treated with mycophenolate mofetil (MMF) or intravenous cyclophosphamide plus prednisone. Blood samples from these patients were drawn at baseline and six months later. All patients were classified as complete response (CR) or non-CR (partial and non-response) according the ACR criteria [
24]. CR means urine protein less than 200 mg per day, urine red blood cells less than 5 per high power field and improved kidney function.
Data collection
Various data were collected: clinical assessment of SLE Disease Activity Index (SLEDAI), laboratory tests, histological parameters, blood levels of APRIL and BLyS, and measurements of APRIL and BLyS levels from the kidney biopsies. Blood levels of APRIL and BLyS were obtained on the day of the biopsy and on the next six month follow-up visit.
ELISA measurements for APRIL and BLyS
Enzyme-linked immunosorbent assay (ELISA) was used to quantify levels of APRIL. ELISA plates (R&D Systems, Inc., Minneapolis, MN, USA) were coated with mouse anti-human APRIL polyclonal antibody. Standards and samples were pipetted into the wells. Any APRIL present in the patients' samples were bound by the antibody coated on the plate. After washing the plate for any unbound substances, polyclonal antibodies against APRIL conjugated to horseradish peroxidase were added to the wells. After that, the plate was washed again to remove any unbound antibody-enzyme reagent. Next, a substrate (tetramethylbenzidine) was added to the wells. The plate was incubated at room temperature in the dark for a couple of minutes. When the color changed from blue to yellow, sulfuric acid was added to stop the color development. The amount of APRIL correlated with the intensity of the color. The plate was read at a wavelength of 540 nm or 570 nm.
BlyS was quantitated by the BlyS Platinum ELISA kit (eBioscience, San Diego, CA, USA). The procedure of the ELISA is the same as that used to quantitate the levels of APRIL except for the antibodies used, which would be more specific in detecting BlyS.
For serum APRIL or BLyS levels, measurement was performed in duplication for each samples. The coefficient variation (CV) was calculated and any measurement with CV above 20% was excluded from the analysis. Then, five out of fifty-two samples were excluded due to CV > 20%.
RNA extraction and cDNA synthesis
RNA was extracted from the renal biopsy specimens of the LN patients. Total RNA was purified by RNeasy purification kit (Qiagen, Valencia, CA, USA). The first strand cDNA was synthesized from 250 ng of total RNA by using MultiScribe™ Reverse Transcriptase enzyme (Applied Biosystems, Carlsbad, CA, USA). The reaction was performed at 45°C for 45 min.
Quantification of intrarenal mRNA levels of APRIL and BLyS
Intrarenal mRNA levels of APRIL and BLyS were measured by real-time quantitative polymerase chain reaction (RT-PCR). The sequences of the primers are as follows: APRIL sense 5'-AGAGTCTCCCGGAGCAGAGTT-3', APRIL antisense 5'-CTGGTTGCCACATCACCTCTGT-3', BLyS sense 5'-TCACCGCGGGACTGAAAATCTT-3' and BLyS antisense 5'-AGCTGAGAAGCCATGGAACAAATG-3'. Each PCR was carried out in a 20 μL reaction volume composed of 2 μL cDNA template, 10 μL of SYBR Green PCR Master Mix (Applied Biosystems, Carlsbad, CA, USA), 0.5 mM forward primer, and 0.5 mM reverse primer. The amplification reaction was performed by using an Applied Biosystems 7500 Real-Time PCR System (Applied Biosystems, Carlsbad, CA, USA). The comparative Delta Ct method was used to determine the mRNAs levels of APRIL and BLyS. 18s ribosomal RNA was used as the endogenous reference. Dissociation curve analysis confirmed amplification of one specific product per primer pair.
Renal histology
Paraffin sections were stained with hematoxylin and eosin, periodic acid-Schiff, trichrome, and silver for light microscopy. Activity and severity of the specimens were scored by using the Austin criteria [
25]. The maximum scores of the activity and chronicity indices were 24 and 12, respectively. The activity index was calculated by summing the semi-quantitative scores of the following parameters: endocapillary proliferation, fibrinoid necrosis, cellular crescents, leukocyte infiltration, hyaline thrombi and interstitial infiltration. The chronicity index was calculated by adding the semi-quantitative scores of the following parameters: glomerular sclerosis, fibrous crescents, interstitial fibrosis and tubular atrophy. The percentage of each parameter was calculated by the following equation:
All biopsies were examined by one pathologist (WK) who was not aware of the molecular results. The histological samples of LN were classified according to the ISN/RPS classification [
26].
Renal SLEDAI
Renal SLE disease activity index of the SLE patients at the time of visit was measured by using MEX-SLEDAI scoring system [
27]. From the MEX-SLEDAI scoring system, the renal disease activity was scored on a 0 to 16 point scale according to the MEX-SLEDAI index. A higher score indicated a greater disease activity. Renal SLEDAI was calculated by summing the scores of the following parameters: urinary casts (heme-granular or red blood cell casts), hematuria (> 5 RBC/HPF by excluding stones, infections, or other causes), proteinuria (> 0.5 gram/24 hours) and pyuria (> 5 WBC/HPF by excluding infections). The maximum score of renal SLEDAI was 16.
Data analysis
Statistical analysis was performed by using SPSS software (version 16, SPSS Inc, Chicago, IL, USA). Continuous variables were summarized as mean ± SE or median (interquartile range: IQR) where appropriate and categorical variables were used as frequencies and percentages. Significant differences were calculated by using Mann-Whitney tests for continuous variables. Correlation coefficients were calculated by Spearman's test. Significant factors obtained from univariate analysis were entered into a multiple logistic regression model for control of potential confounding effects. Box plot graphs were done by using GraphPad Prism version 4.03 for Windows (GraphPad Software, San Diego, CA, USA). The receiver operating characteristic curve (ROC) curve of APRIL levels was plotted by using SPSS software. Differences with a P value of < 0.05 were considered statistically significant.
For serum APRIL or BLyS levels, measurement was performed in duplication for each sample. The coefficient variation (CV) was calculated and any measurement with CV > 20% was excluded from the analysis.
Discussion
This study demonstrated that B cell activation markers such as APRIL and BLyS are tightly correlated with SLE and LN disease activity. Serum levels of BLyS were influenced by immunosuppressant whereas serum levels of APRIL could accurately predict the status of the kidney and severity of the disease (Table
2). Patients with high serum levels of APRIL have severe proliferative nephritis such as endocapillary proliferation, fibrinoid necrosis and polymorph infiltration. In addition, serum levels of APRIL more than 4 ng/mL may predict that the patients will experience treatment failure (Table
3 and Figure
2). Moreover, in non-responder patients, intrarenal mRNA levels of APRIL and BLyS were higher compared to the responders (Figure
4). This supported the notion that blocking BLyS and APRIL may be useful in treating lupus nephritis.
BLyS and APRIL have been shown to be involved in B cell-mediated pathology of SLE. In an animal model, it has been shown that BLyS transgenic mice had an elevated number of B cells with prolonged survival times [
28]. These transgenic mice exhibited lupus-like phenotypes such as having high titer levels of anti-dsDNA antibody, proteinuria, and glomerulonephritis [
28,
29]. It has been shown that by blocking the receptors of BLyS and APRIL, TACI-Ig can decrease levels of proteinuria in NZB/W F1 lupus-prone mice [
29]. On the other hand, by blocking only BLyS, BAFF-R-Ig has been shown to be able to prolong the lives of NZB/W F1 mice and prevent renal inflammation [
30].
These data obtained from the animal studies were consistent with findings in lupus patients. Chu
et al. [
16] found expressions of BLyS and APRIL on CD19+ B cells that were closely correlated with the activities of the disease and autoantibody levels in SLE patients. We also observed an elevated level of circulating BLyS [
29], which corresponded to the serologic indices of SLE patients (Table
2). Interestingly, dosages of immunosuppressive drugs affected serum levels of BLyS and APRIL differently (Figure
3A and
3B). Serum levels of both cytokines in this study may release from different B-cell populations [
31]. BLyS represents early preimmune B-cell stages and could be easily inhibited by steroids or MMF [
32]. In contrast, APRIL may lately respond to the immunosuppressants. We observed high APRIL levels associated with lupus patients without renal involvement and only used low-dose prednisone (data not shown). This was previously demonstrated by Morel
et al. [
33].
Regarding circulating BLyS levels, we observed less serum BlyS levels (1.32 ± 0.15 ng/mL) as compared to other studies (range 1.5 to 19.1 ng/mL) [
34]. In contrast to previous published works [
34,
35], all participants, in this study, were diagnosed lupus glomerulonephritis and had marked proteinuria. As Petri M and Stohl W
et al. pointed out that lupus nephritis may have a urinary loss of BLyS protein. Therefore, a loss of correlation between circulating BLyS and disease activity may occur in lupus nephritis [
34,
35]. Unlike other published works, we did not find a correlation between BLyS and anti-dsDNA. This could be explained by the urine loss of BLyS protein. Moreover, potent immunosuppressants (steroids and MMF) may decrease BlyS production while such treatment had less influence on the anti-dsDNA titer [
32]. Therefore, different results among published works could be explained by different stages of disease and studied populations.
The utility of biomarkers in clinical nephrology has been successfully demonstrated by leaders in the field [
36,
37]. Since serum APRIL levels were associated with the histology of nephritis, and high serum levels (> 4 ng/mL) of APRIL may predict treatment failure in patients with class III/IV LN on standard course of therapy. Thus, levels of APRIL may be used as a biomarker to prognose or guide the treatment for LN. This would indicate that patients with class III/IV LN and high serum levels of APRIL will need additional therapy in conjunction with the standard immunosuppressive treatment in order to successfully treat the disease.
The use of monoclonal antibodies such as belimumab together with standard therapy has been shown to work in treating patients with mild forms of SLE without renal or neurological organ involvement [
38‐
40]. Although belimumab is the only monoclonal antibody that has already been approved to treat SLE, monoclonal antibodies to APRIL/BLyS are ongoing in clinical trials. Current treatment for severely active LN are only limited to immunosuppressants such as steroids, cyclophosphamide or mycophenolate. For non-responsive LN patients, they will inevitably progress to end-stage kidney disease.
Currently, there are clinical trials assessing the safety and efficacy of monoclonal antibodies to BLyS/APRIL such as atacicept. The recombinant protein, atacicept, was studied in phase II/III trial together with MMF and steroid in newly active LN. Unfortunately, patients who received atacicept had profound hypogammaglobulinemia and infection [
41]. Combination use of atacicept with MMF is therefore prohibited. In order to sufficiently evaluate the safety and efficacy of these antibodies, selection of participants with active LN is extremely crucial. Based on the results obtained from our study, we postulate that LN patients with high serum levels of APRIL may benefit from using atacicept or other targeted APRIL therapy.
Equally important, our study has shed some light on the pathogenesis of LN. We postulate that local factors may control these gene expressions. BLyS/APRIL mRNAs and proteins were previously observed in the glomerulus and inflammatory infiltrating cells [
22]. This would indeed explain why our patients with high intrarenal mRNAs were resistant to treatment. Excess levels of APRIL and BLyS may promote survival of autoreactive B cells in the kidney, causing an increased production of pathogenic autoantibodies, and finally resulting in immune-complex-induced glomerulonephritis.
There were some limitations encountered in this study. Aside from the rather small sample size in this study, the studied samples were collected from patients who received different immunosuppressive therapies. Therefore, a further validation study on samples collected before initiation of immunosuppressant should be done in a prospective clinical trial. Nonetheless, this study clearly showed that the B cell activation biomarkers were present in the circulation and kidney tissues of active lupus nephritis patients.
Acknowledgements
The authors thank the Chula Clinical Research Center (N. Buathong) for statistical consultancy. These studies were supported by the Faculty of Medicine, Chulalongkorn University's Ratchadaphiseksomphot Fund (WT), the Cluster and Program Management Office (CPMO) National Science and Technology Development Agency (P-10-10221), the Higher Education Research Promotion and National Research University Project in Thailand (HR 1163A). PT received post-doctoral support from the Ratchadaphiseksomphot Fund, Chulalongkorn University.
Part of this study was presented as an oral presentation at the American Society of Nephrology Meeting, Denver, CO, USA. April 16, 2010 (Treamtrakanpon W, Avihingsanon Y, Eiam-ong S: B lymphocyte related signatures are associated with active lupus nephritis. J Am Soc Nephrol 2010; 21:25A).
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
WT and PT performed experiments, analyzed data, and wrote the manuscript. TB and PS collected the samples and performed the ELISA. WK reviewed and interpreted the kidney histological results. SE and AL reviewed the manuscript. NH and YA led the project, reviewed the clinical data, provided patient care, and finalized the manuscript. All authors have read and approved the manuscript for publication.