Neural cell adhesion molecule 1 is a novel autoantigen in membranous lupus nephritis.

Abstract

Membranous lupus nephritis is a frequent cause of nephrotic syndrome in patients with systemic lupus erythematosus. It has been shown in phospholipase A2 receptor positive membranous nephropathy that known antibodies can be detected within sera, determination of the target autoantigen can have diagnostic significance, inform prognosis, and enable non-invasive monitoring of disease activity. Here we utilized mass spectrometry for antigen discovery in laser captured microdissected glomeruli from formalin-fixed paraffin embedded tissue and tissue protein G immunoprecipitation studies to interrogate immune complexes from frozen kidney biopsy tissue. We identified neural cell adhesion molecule 1 (NCAM1) to be a target antigen in some cases of membranous lupus nephritis and within rare cases of primary membranous nephropathy. The prevalence of NCAM1 association was 6.6% of cases of membranous lupus nephritis and in 2.0% of primary membranous nephropathy cases. NCAM1 was found to colocalize with IgG within glomerular immune deposits by confocal microscopy. Additionally, serum from patients with NCAM1-associated membranous nephropathy showed reactivity to NCAM1 recombinant protein on Western blotting and by indirect immunofluorescence assay, demonstrating the presence of circulating antibodies. Thus, we propose that NCAM1 is a target autoantigen in a subset of patients with membranous lupus nephritis. Future studies are needed to determine whether anti-NCAM1 antibody levels correlate with disease activity or response to therapy.

Graphical Abstract

Introduction.

There are an estimated 121,000 to 402,000 patients living with Systemic Lupus Erythematosus (SLE) in the United States, which disproportionately affects females and African Americans ¹. SLE can involve multiple organ systems, and lupus nephritis is a serious complication that affects up to half of lupus patients ². Lupus nephritis causes significant morbidity in SLE patients, with 11 percent progressing to end-stage kidney disease and requirement for renal replacement therapies ³. Of lupus patients with biopsy-proven nephritis, 10-20 percent have membranous lupus nephritis, which can occur with or without a concurrent proliferative component ⁴. Membranous lupus nephritis is the leading cause of nephrotic syndrome in patients with SLE. It is characterized by the presence of subepithelial and intramembranous immune deposits, which are caused by circulating immune complexes and/or in situ immune complex formation within glomeruli. In the majority of cases of membranous lupus nephritis, the causative autoantigen is unknown. Recently, the exostosin (EXT) 1/2 protein complex was identified in the glomeruli of 34.6% of cases of SLE-associated membranous nephropathy (MN) ⁵. However, to date, circulating autoantibodies to EXT1 or EXT2 have not been identified.

Serum testing in idiopathic membranous nephropathy has proven useful to monitor disease and to inform treatment. Additionally, it has been shown that serologic remission can be detected months before resolution of proteinuria and evidence of clinical remission. For phospholipase A2 receptor (PLA2R) positive membranous nephropathy, the causative autoantigen in the majority of idiopathic MN cases ⁶, antibody testing can serve as a biomarker of disease activity and response to treatment, as measured by indirect immunofluorescence and/or enzyme-linked immunosorbent assay based testing ⁷. Thrombospondin-type 1 domain containing 7a, the second identified autoantigen in MN ^{8, 9}, also can be monitored by serum testing through indirect immunofluorescence-based assays ¹⁰. Identification of serum biomarkers in membranous lupus nephritis could enable disease monitoring similar to what is now possible with PLA2R-positive primary membranous nephropathy, and is the focus of these investigations.

Results.

Mass Spectrometry Analysis Identifies NCAM1 as a Membranous Antigen

Mass spectrometry of renal biopsies from patients with membranous lupus nephritis was performed for autoantigen discovery. Mass spectra of laser capture microdissected (LCMD) glomeruli from 13 cases of MN of unknown antigen type, including both lupus and non-lupus patients, were compared with twelve cases of known etiology (including 8 PLA2R and 4 EXT-associated MN) to identify proteins present uniquely in the unknown cases (supplemental figure 1. Three cases showed the presence of NCAM1 peptides uniquely in the unknown MN glomeruli with a total of eighteen unique peptides identified (supplemental figure 2). All three of these patients with NCAM1 detected had SLE.

The mass spectrometric profile of these LCMD glomeruli from three cases with abundant neural cell adhesion molecule 1 (NCAM1, also known as CD56) peptides, were compared with the twelve cases of MN of known type to identify the protein with the greatest fold change. The protein differential abundance for each type of MN was compared against the remaining two groups using normalized iBAQ values produced by MaxQuant. Statistical analysis was performed using Welch's t-test, and the results visualized on a volcano plot (Figure 1). To determine the ability of this approach to identify glomerular antigens, we first compared samples from patients with known membranous antigens, PLA2R and EXT1, to cases that did not stain for these antigens.PLA2R was the protein with the strongest fold increase in cases of PLA2R-associated MN (Figure 1A). EXT1 displayed the strongest fold changes in EXT-associated MN (Figure 1C). These results demonstrate that analysis of the laser capture microdissected glomeruli MS data in this way can identify proteins significant to the pathogenesis of disease in MN. When this same statistical analysis was used on the samples with NCAM1 identified, NCAM1 was identified as the protein with the highest fold increase (Figure 1E), and NCAM1 peptides were not present in the other samples. This confirms that NCAM1 is uniquely present in the glomeruli of a subset of patients with membranous nephropathy of unknown antigen type.

Figure 1.

NCAM1 is significantly enriched in glomeruli from a subset of membranous lupus nephritis patients. Normalized iBAQ values from MaxQuant were used for statistical analysis. Figures on the left (A, C, E) compare protein abundance in LCMD glomeruli and figures on the right (B, D, E) compare abundance of proteins obtained by immunoglobulin captured from kidney tissue. Dashed lines depict cutoff for p value=0.05. (A) Volcano plot comparing MS of laser capture microdissected glomeruli from eight PLA2R-positive MN cases compared with seven membranous nephropathy samples of other types identifies PLA2R as the protein with the highest fold change. (B) Volcano plot comparing MS of immunoglobulin capture from four PLA2R frozen kidney biopsy samples compared with seven membranous nephropathy samples of other types identifies identifies PLA2R as the protein with the highest fold change. (C) Volcano plot comparing MS of laser capture microdissected glomeruli from four EXT-positive membranous cases compared with eleven membranous nephropathy samples of other types identifies EXT1 as the protein with the highest fold change. (D) Volcano plot comparing MS of immunoglobulin capture from four frozen kidney biopsy samples that show EXT-positive MN compared with seven membranous nephropathy samples of other types identifies EXT1 and EXT2 as the proteins with the highest fold change. (E) Volcano plot comparing MS of laser capture microdissected glomeruli from three NCAM1-positive membranous cases with twelve membranous nephropathy samples of other types identifies NCAM1 as the protein with the highest fold change. (F) Volcano plot comparing MS of immunoglobulin capture from three frozen kidney biopsy samples that show NCAM1-positive MN with eight membranous nephropathy samples of other types identifies NCAM1 as the protein with the highest fold change. Each dot on the volcano blot represents a protein detected by mass spectrometry.

As a further independent discovery method, protein G co-immunoprecipitation was performed on fresh frozen tissue from a total of 11 kidney biopsies to identify proteins uniquely bound to immunoglobulin within the glomeruli of MN biopsies. This included 4 PLA2R MN, 4 EXT MN, and the 3 cases predicted as NCAM1 MN by LCMD-MS. All three patients predicted as NCAM1 MN had a history of systemic lupus erythematosus with membranous lupus nephritis diagnosed on kidney biopsy. A total of 1448 distinct proteins were identified in these samples. A similar analysis was performed as with the LCMD samples in that each group was evaluated for proteins with the greatest fold change when compared with MN of other types. The protein differential abundance for each type of MN was compared against the remaining two groups using normalized iBAQ values. Statistical analysis was performed using Welch's t-test, and the results visualized on a volcano plots (Figure 1). PLA2R was the protein with the strongest fold increase in cases of PLA2R-associated MN (Figure 1A-B). EXT1 and EXT2 displayed the strongest fold changes in EXT-associated MN (Figure 1 C-D). NCAM1 was identified as the protein with the highest fold increase in the NCAM1-associated cases (Figure 1E-F). NCAM1 was present in all three of the NCAM1-associated samples but none of the PLA2R or EXT-associated samples. These immunoprecipitations from frozen tissue confirms that IgG is bound to NCAM1 in the samples with increased glomerular NCAM1 detected by mass spectrometry.

NCAM1 co-localization with IgG:

A rabbit polyclonal antibody against NCAM1 showed strong positive staining along the glomerular capillary loops of cases with NCAM-1 positive MN with essentially complete co-localization with IgG by confocal microscopy. Co-localization of NCAM1 with IgG was 96.1 ± 2.3% in NCAM1-associated MN cases (Figure 2C) and 19.6 ± 5.8% of control PLA2R-positive MN cases (Figure 2F). This background staining in PLA2R-positive MN cases is artifactual, likely due to serum trapping within glomeruli in FFPE tissue. Co-localization with NCAM1 and IgG was significantly increased in NCAM1-associated MN cases, compared to PLA2R-positive MN controls (p<0.0001, one way-ANOVA). NCAM1 did not show significant co-localization with IgG in the glomeruli from control PLA2R-positive MN cases (additional images provided in supplemental figure 3). There are rare foci of NCAM1 positivity along tubular basement membranes (TBMs) with co-localization with IgG, however the majority of TBM deposits are NCAM1-negative (supplemental figure 4). Additionally, NCAM1 was not found to be positive in TBM deposits in every case of NCAM1 MN that contained IgG TBM deposits, so this finding is unlikely to be significant. No PLA2R-associated MN cases had tubular basement membrane deposits for evaluation.

Figure 2.

Co-localization of NCAM1 with IgG in cases of NCAM1-associated MN, but not PLA2R controls. A) IgG staining of a predicted NCAM1-positive case; B) NCAM1 staining of a predicted NCAM1-positive case; C) Overlay of IgG and NCAM1 of a predicted NCAM1 positive case; D) IgG staining of a PLA2R MN case; E) NCAM1 staining of a PLA2R MN case; F) Overlay of IgG and NCAM1 in a PLA2R MN case.

To further establish specificity of NCAM1 detection in NCAM1-associated MN, NCAM1 staining was also performed on 20 proliferative lupus nephritis biopsies (ISN/RPS class III or IV) and 10 EXT-associated MN biopsies. Zero of 20 proliferative lupus nephritis biopsies and 0 of 10 EXT-associated MN biopsies were NCAM1-positive (supplemental figure 5). Additionally, no proliferative lupus nephritis or EXT-associated MN cases showed NCAM1 staining within tubular basement membrane deposits (supplemental figure 5). To confirm antibody specificity, a second commercial antibody (Invitrogen anti-CD56, PA5-83479) demonstrated a similar pattern of staining within glomeruli in NCAM1-associated MN cases (data not shown).

Clinical and prevalence data for NCAM1-associated membranous nephropathy:

A series of consecutive membranous lupus nephritis biopsies were identified over a 6-month period that could be tested for the prevalence of NCAM1-associated MN. A total of 216 consecutive cases of membranous lupus nephritis with or without proliferative changes by light microscopy were diagnosed in our laboratory during the study time period. Among these, there were 212 cases with sufficient tissue remaining to stain for the type of membranous present. All cases were stained for NCAM1 by immunofluorescence. A total of 14 of 212 of membranous lupus nephritis cases were NCAM1-positive (6.6%). EXT-staining was also performed on this cohort (of which 209 cases had residual tissue for staining), and 33 of 209 cases were EXT2-positive (15.8% of MLN cases). NCAM1-associated MN was rarely positive in a separate analysis of 101 cases of idiopathic MN stained for NCAM1 (2 of 101 cases, 2.0%) Twenty-four patients in this cohort were EXT2-positive, however our biorepository frequency is over-representative of the true frequency, as the data are not from consecutive patients and there was increased recruitment of EXT-positive patients into our biorepository. Nineteen of 20 cases were mutually exclusive, with one membranous lupus nephritis case identified with dual EXT and NCAM1 positivity.

Clinical details for patients with NCAM1-associated MN are shown in Table 1. The patients were on average 34 years of age (± 12.1 years) and were predominantly female (14 of 20 patients, 70%). The mean serum creatinine was 1.35 ± 0.88 mg/dL. Quantitative proteinuria was on average 6.7 ± 9.5 grams/day, determined by either urine protein-to-creatinine ratio or 24 hour urinary protein levels. The mean serum albumin level was depressed (2.2 ± 0.78 g/dL).

Table 1.

Clinical and serologic features of cases NCAM1-associated MN. Abbreviations: Cr = creatinine; SLE = systemic lupus erythematosus; Dx = diagnosis.

Case	Diagnosis	Age (years)	Sex	Cr (mg/dL)	Albumin (g/dL)	Proteinuria (grams/day)	Serologies (positive)	SLE Dx	Neuropsychiatric Symptoms	Treatment
1	Lupus V	42	F		3.2	1.4	ANA+, dsDNA+, Smith+	Yes	Seizures	Hydroxychloroquine
2	Lupus IV + V	36	M	2.7	3.1			Yes		Hydroxychloroquine Mycophenolate Prednisone
3	Membranous	40	M	0.8	0.9			No
4	Lupus V	44	M	1.5	2.9	4	ANA+, dsDNA+	Yes	Stroke	Prednisone, Azathioprine
5	Lupus V	23	F	0.5		3.8		Yes	Seizures, Lupus cerebritis	Monthly IV cyclophosphamide, Prednisone, Hydroxychloroquine
6	Lupus V	38	F	2.1	2	40	ANA+, dsDNA+	Yes	Psychiatric disorder	Hydroxychloroquine Methylprednisolone
7	Lupus V	28	F	1.5	1.5	6.8	ANA+	Yes	Seizures	Prednisone, Hydroxychloroquine, Mycophenolate
8	Membranous	54	M	1		5		No		Hydroxychloroquine Mycophenolate, ACEI
9	Lupus V	24	F	0.7	2.3	0	ANA+, Smith+	Yes		ACTHar, ACEI
10	Lupus V	33	F	0.9		1.8		Yes
11	Lupus III + V	60	M	0.7		2		Yes		Cyclophosphamide Prednisone
12	Lupus III + V	27	F	0.6	1.9	8.2	ANA+, dsDNA+	Yes		Toclizumab, ARB, Prednisone, Mycophenolate
13	Lupus V	23	F	0.8	2.2	3		Yes		Mycophenolate Prednisone
14	Lupus III + V	22	F	2.3	2.7	2.1	ANA+	Yes	Lupus cerebritis	Hydroxychloroquine Mycophenolate Prednisone
15	Lupus III + V	26	F	2	2.5	4.6	ANA+	Yes		Belimumab Prednisone
16	Lupus V	47	F	3.7	1.8	11	ANA+	Yes		Methylprednisolone
17	Lupus V	12	F	0.3	2.1	3+ UA, no quantitation	ANA+, ENA+	Yes		Hydroxychloroquine Mycophenolate
18	Lupus V	44	F	1.6	3.2	2.5	ANA+, SS-A+, SS-B+	Yes		Hydroxychloroquine Mycophenolate Prednisone
19	Lupus V	24	M	0.8	2.16	1.3	ANA+, Smith+, RNP+	Yes	Central serous retinopathy	Prednisone, Mycophenolate
20	Lupus V	33	F	1.2	16	0.5	ANA+, SS-A+	Yes	Inflammatory cerebritis, positive AchR antibodies	Prednisone, Hydroxychloroquine

Serologic studies showed that 13 of 13 patients with serologic data available at the time of biopsy had a positive ANA (100%), 4 of 13 had a positive dsDNA (30%), and 3 of 13 had anti-Smith antibodies (23%). Two patients had SS-A antibodies, and one patient had anti-RNP antibodies. Serum complements (C3 and C4 levels) were normal in 6 of 12 patients with available data (50%) and were low in 6 of 12 NCAM1-positive patients (50%). Three patients had a positive family history of SLE. Eight patients had evidence of neuropsychiatric disease in addition to nephritis (40%). Neuropsychiatric manifestations in SLE patients with NCAM1-associated MN included seizures (n=3), cerebritis (n=3, one with acetylcholine receptor antibodies), serous central retinopathy (n=1), psychosis (n=1), and stroke (n=1) in a young patient suspected to be related to anti-phospholipid antibody syndrome. One patient had seizures along with cerebritis. A summary of the clinicopathologic data of NCAM1-associated MN patients is provided in Table 1.

There were two NCAM1-positive patients without a known history of SLE and showing no autoimmune manifestations or a clinical suspicion for lupus nephritis. One patient was a 54 year old Caucasian male with a medical history of hypertension, chronic obstructive pulmonary disease, obesity, asthma, obstructive sleep apnea, and bronchiectasis related to repeated episodes of pneumonia due to common variable immunodeficiency (CVID). Membranous nephropathy is a common renal manifestation of CVID ¹¹. The second patient was a 40 year old male with a history of hypertension, obesity, and anxiety. On renal biopsy, both cases lacked tissue ANA staining, tubular basement membrane staining, or a "full-house" pattern of immunofluorescence.

Histopathology of NCAM1-associated membranous nephropathy:

Histopathology was reviewed from all cases of NCAM1-associated MN (n=20; including 2 cases of primary MN and 18 cases of membranous lupus nephritis. NCAM1-positive cases occurred in cases with and without a concurrent proliferative component (n=5, including ISN/RPS class III or IV) and was not restricted to pure membranous lupus nephritis cases (ISN/RPS class V). Four of twenty cases had focal lupus nephritis in addition to membranous lupus nephritis (ISN/RPS class III + V) and one case had diffuse lupus nephritis in addition to membranous lupus nephritis (ISN/RPS class IV + V). Therefore, 25% of cases had a proliferative component overall, manifested by the presence of endocapillary hypercellularity or crescents. In NCAM1-positive cases, all of which shown NCAM1 positivity in a similar distribution as IgG 18 of 20 (90%) had diffuse and global IgG staining, while 2/20 (10%) had an incomplete global to segmental IgG staining pattern. Eighteen of 20 cases (90%) demonstrated staining for other immune reactants in addition to IgG, kappa, and lambda light chains, with IgA in 13 of 20 cases (65%), IgM in 13 of 19 cases (one case lacked tissue available on sections, 68%), C3 in 17 of 20 cases (85%), and C1q in 11 of 20 cases (55%). Eight of 20 (40%) of the NCAM1-positive cases showed "full house" immunofluorescence with staining for all immunoglobulin heavy chains tested (including IgA, IgG, and IgM), as well as both complement components (C1q and C3). There were no cases of monoclonal membranous nephropathy, with all cases showing approximately equal kappa and lambda light chain staining with a distribution similar to IgG. IgG subclasses were performed on 13 cases and did not show a consistent pattern of staining (Supplemental Table 1). However, only 2 of 13 cases were IgG4 dominant or co-dominant. This is dissimilar to primary membranous nephropathy, where IgG4 is restricted or co-dominant within glomerular capillary loop deposits. Extraglomerular staining was present in some NCAM1-associated MN cases, with 11 of 20 cases showing tubular basement membrane immune deposits (55%), 3 of 20 cases showing vascular immune deposits (15%), and 5 of 20 cases having a tissue ANA pattern (25%). Therefore, NCAM1-associated MN can have varying histopathologic features (Figure 3 and Table 2).

Figure 3.

Histopathology of NCAM1-associated MN. (A) NCAM1 positivity in a granular capillary loop distribution within glomeruli (NCAM1 labeled with rhodamine red X; original magnification × 200, scale bar = 20 μm) (B) Glomerulus with endocapillary hypercellularity (Periodic acid Schiff; original magnification ×400, scale bar = 20 μm) (c), Spikes and holes seen along the glomerular capillary loops by Jones methenamine silver staining (original magnification ×400, scale bar = 20 μm). (D and E) IgG immunofluorescence shows (D) global granular glomerular capillary loops staining (original magnification ×400, scale bar = 20 μm) and (E) granular tubular basement membrane staining (original magnification ×200, scale bar = 50 μm). (F), Subepithelial electron dense deposits by electron microscopy (original magnification ×5000).

Table 2.

Histopathologic features of cases of NCAM1-associated MN. Abbreviations: GS = glomerulosclerosis; IF/TA = interstitial fibrosis and tubular atrophy; ANA = antinuclear antibodies.

Case	% Global GS	IF/TA	Proliferative component	IgA	IgG	IgM	C3	C1q	Full house	Mesangial deposits	Subendo deposits	TBM deposits	Tissue ANA	Vascular deposits	EM Stage
1	0	Absent	No	2	3	0	2	0.5	No	Yes	No	Absent	Absent	Absent	2
2	0	Mild	Yes	1	3	1	1	3	Yes	Yes	Yes	Present	Absent	Absent	3
3	2.5	Absent	No	0	3	0	1	0	No	No	No	Absent	Absent	Absent	2
4	35	Severe	No	0	3	1	0	0	No	Yes	No	Absent	Absent	Absent	3
5	1	Absent	No	0.5	3	1	2	1	Yes	Yes	No	Present	Present	Absent	2
6	16	Mild	No	0.5	3		2	2	Yes	Yes	No	Present	Absent	Absent	1
7	33	Moderate	No	0	3	0	0.5	0	No	Yes	No	Absent	Absent	Absent	3
8	2	Mild	No	2	3	0	3	0	No	Yes	No	Absent	Absent	Absent	3
9	3	Absent	No	0	3	0	2	0	No	Yes	No	Absent	Present	Absent	2
10	18	Moderate	No	1	2	1	0.5	0	No	Yes	Yes	Absent	Absent	Absent	3
11	18	Moderate	Yes	0	3	2	0.5	0	No	Yes	No	Present	Absent	Present	3
12	3	Moderate	Yes	1.5	3	1	2.5	2.5	Yes	Yes	No	Present	Present	Absent	2
13	7	Mild	No	0	3	0.5	0.5	0.5	No	Yes	No	Absent	Absent	Present	1
14	42	Moderate	Yes	1	3	0	0	0	No	Yes	N/A	Absent	Present	Absent	N/A
15	11	Severe	Yes	0	2	2	0.5	1	No	Yes	Yes	Present	Absent	Absent	2
16	23	Severe	No	2	3	2	2	3	Yes	Yes	Yes	Present	Absent	Present	3
17	0	Absent	No	2	2	2	1.5	1.5	Yes	Yes	No	Present	Present	Absent	3
18	33	Moderate	No	1	1	0.5	0	0	No	Yes	No	Present	Absent	Absent	3
19	0	Absent	No	2	3	2	2	0.5	Yes	Yes	No	Present	Absent	Absent	3
20	0	Moderate	No	3	2.5	3	2	0.5	Yes	Yes	No	Present	Absent	Absent	3

Electron microscopy was available for 19 of the 20 NCAM1-positive cases. Mesangial electron-dense deposits were present in glomeruli in 18 of 19 cases (95%) and subendothelial electron-dense deposits in 4 of 19 cases (21%). Subepithelial electron-dense deposits were present in all cases and had variable stages of maturation present by the Ehrenreich and Churg classification¹² (stage 1 to stage 3).

Serum reactivity against NCAM1 in patients and controls:

Serum from NCAM1 patients reacted with recombinant NCAM1 120 kDa protein in 2 of 2 patients with NCAM1-associated MN with sera on non-reducing western blots (Figure 4) and not under reducing conditions (supplemental figure 6), but serum from patients with PLA2R MN (0 of 14), negative controls and 0 of 4 EXT controls (Figure 4) and supplemental figure 7. These serum reactivity studies confirmed that circulating anti-NCAM1 antibodies exist in a subset of lupus patients. Loss of seroreactivity under reducing conditions suggests that the epitope may be dependent on internal disulfide bonds).

Figure 4.

Serum reactivity against NCAM1 identified in patients with NCAM1-associated MN. Western blotting shows immunoreactivity against NCAM1 recombinant protein in NCAM1-associated MN but not PLA2R-associated MN controls. A band of 120 kDa was identified, corresponding to that obtained after incubation with an anti-NCAM1 primary antibody.

Additionally, indirect immunofluorescence assays shown seroreactivity against NCAM1-expressing HEK-293 cells, but not within vector controls within NCAM1-associated MN. PLA2R-positive MN controls lacked seroreactivity (Supplemental Figure 8).

Discussion.

This newly identified target antigen in membranous lupus nephritis, NCAM1, is a member of the immunoglobulin superfamily of proteins. In adults, NCAM1 expression is seen at high levels within the central nervous system (including the cerebral cortex, cerebellum, and hippocampus), peripheral nerves, thyroid gland, adrenal gland, heart, stomach (within gastric chief cells), and cells within the immune system ¹³. Immune cells expressing NCAM1 include natural killer (NK) cells, γΔ T cells, CD8⁺ T cells that have an activated phenotype, and dendritic cells. NCAM1 is expressed in the kidney during development within the metanephrogenic mesenchyme, and with maturity, NCAM1 expression decreases. In adults, NCAM1 expression is mostly restricted to interstitial cells that are in proximity to the corticomedullary junction¹⁴, but also shows expression within podocytes ¹⁵ . Increased NCAM1 expression within interstitial cells has been associated with interstitial fibrosis in glomerulonephritis, including lupus nephritis¹⁶ . However, increased interstitial fibrosis was not observed in the NCAM1-associated membranous lupus nephritis cases that we evaluated. NCAM1 is also present in urinary exosomes ¹⁷, of which the significance is unclear Further work will be required to determine the pathophysiology of NCAM1-associated MN.

Despite NCAM1 expression occurring within podocytes, we do not observe a podocyte staining pattern within normal kidney biopsies or non-NCAM1 membranous cases. While NCAM1-associated MN cases were primarily within lupus patients, we cannot exclude the possibility that it could also serve as a primary podocyte antigen.

There are no studies to date that explore NCAM1 in membranous lupus nephritis. However, increased NCAM1 within urine has been shown in a subset of lupus patients with proliferative lupus nephritis with correlation to disease activity ¹⁸. When lupus sera were reacted against human renal glomerular endothelial cells, NCAM1 expression was significantly reduced. This is thought to be due to soluble NCAM1 competing with the membrane-bound form ¹⁹. It is possible that in some lupus nephritis patients, membrane-bound NCAM1 could be an autoantigen that drives disease.

Given that some patients had neuropsychiatric manifestations concurrent with nephritis, and NCAM1 is expressed at high levels within the central and peripheral nervous system, it is possible that NCAM1 autoreactivity may indicate a subset of patients at risk for both neuropsychiatric manifestations and nephritis. Of note, NCAM1 has been implicated as a biomarker for epilepsy within cerebrospinal fluid for patients with seizures ²⁰, and shows increased expression within reactive astrocytes²¹ . Although our frequency of neuropsychiatric manifestations was higher than reported in SLE patients overall, it is unknown whether it is increased within lupus patients with MLN, as at this time, there is no data in the literature on the frequency of MLN and neuropsychiatric disease. Additionally, given that our cohort is relatively small, it may be premature to determine clinical associations of lupus manifestations and NCAM1-associated MN.

Membranous nephropathy can be the first presentation of SLE in a subset of lupus patients, where other clinical and serologic sequelae may not be seen but can arise at a later time. In the patients with “idiopathic” membranous nephropathy who are NCAM1 positive, following these patients to determine if they later develop lupus would be important to determine if NCAM1 positivity or anti-NCAM1 antibodies could serve as a harbinger of SLE.

There are several limitations of this study. The study is of a small cohort of NCAM1-associated MN patients, with a relatively small sample size (n= 20 patients). Currently, we are unable to determine if there are histopathologic features that could distinguish NCAM1-associated MN cases from other cases of membranous lupus nephritis, as the histopathologic and immunologic patterns on the kidney biopsies in this small cohort had variable appearances. Determining a distinguishing factor could help the nephropathologist decide when to stain membranous biopsies with NCAM1 ⁵. It is possible that anti-NCAM1 antibodies could be used as a biomarker to predict the likelihood of remission and monitor response to therapy, but we have insufficient data at this time for that conclusion. Whether anti-NCAM1 antibodies in serum correlate with proteinuria is currently unknown, as we only had two serum samples from NCAM1-positive patients available for evaluation, without serial sampling. At this time, it is also unknown if NCAM1-associated MN cases have differential response to common therapies used in the treatment of MN, such as rituximab, cyclophosphamide, or calcineurin inhibitors.

In conclusion, NCAM1-associated membranous nephropathy is enriched in patients with systemic lupus erythematosus, has a female predominance, and presents in younger patients than seen with PLA2R or THSD7A-associated membranous nephropathy. NCAM1-associated MN makes up 7.1% of all membranous lupus nephritis cases, which have variable histopathologic patterns. There may also be an association with neuropsychiatric manifestations of systemic lupus erythematosus, but more work is required for confirmation.

Materials and Methods.

Renal biopsy processing techniques

Standard renal biopsy processing techniques were used including light, immunofluorescence, and electron microscopy as previously described ^22-24. All light microscopy samples were stained with hematoxylin and eosin, Jones methenamine silver, Masson trichrome, and periodic acid-Schiff reagent. All direct immunofluorescence sections were cut at 4 μm and reacted with fluorescein-tagged polyclonal rabbit anti-human antibodies to IgG, IgA, IgM, C3, C1q, fibrinogen, and κ-, and λ-light chains (Dako, Carpenteria, CA, USA) for 1 h. For electron microscopy, thin sections were examined in a Jeol JEM-1011 electron microscope (Jeol, Tokyo, Japan).

Preparation of Laser Capture Microdissected Glomeruli for Mass Spectrometry

Renal biopsy tissue from formalin fixed paraffin embedded tissue was cut at a thickness of 10 μm onto Leica PET-membrane frame slides. Four tissue sections were taken, with approximately 40 glomeruli dissected per case. These slides were then stained with Meyer’s hematoxylin. The glomeruli were microdissected into microcentrifuge tubes using a Leica DM6000B microscope into PBS buffer. The microdissected glomeruli were lysed in 2% SDS and 0.1M DTT at 99°C for 1h and processed by filter assisted sample preparation (FASP). The clarified lysate was transferred onto Vivacon 500 concentrators (MWCO of 30kDa; Sartorius, Gottingen, Germany). SDS was removed by repeat washes with 8M Urea in 0.1M Tris/Cl, pH 8.5, and the samples alkylated with 0.05M iodoacetamide. Iodoacetamide was removed by 3 washes with 8 M urea /0.1 M Tris/Cl, pH 8.5, followed by 3 washes with 0.05 M ammonium bicarbonate. Proteins were digested with trypsin (sequencing grade, Promega, Madison WI) at a 40:1 w/w ratio at 37°C for 16 hrs. Peptides were collected by centrifugation and desalted on C18-Stage tips (Thermo-Fisher Scientific, Waltham, MA).

Preparation of Tissue IgG Co-Immunoprecipitation for Mass Spectrometry

Protein G co-immunoprecipitation was performed from OCT frozen tissue of 4 PLA2R positive, 4 EXT positive, and 3 NCAM1 positive renal biopsies for independent confirmation of the mass spectrometry results from laser capture microdissection as shown previously²⁵. To do this, the residual OCT frozen tissue from archived renal biopsies was thawed, washed four times in phosphate buffered saline (PBS), and lysed by mechanical disruption of tissue cores in Pierce IP lysis buffer by bead beating. Protein extracts (approximately 100 μl each) were incubated with 50 μl of protein G magnetic Dynabeads (Dynabeads Protein G, product #10004D, Invitrogen) at room temperature for 1 h with shaking. Samples were then washed four times with PBS to reduce non-specific binding interactions. Proteins were digested from the beads prior to mass spectrometric analysis.

For preparation of peptides for mass spectrometry, the beads were resuspended and reduced in 0.1 M DTT/ 0.1 M Tris/Cl pH 7.6 at 60°C for 20 min. Samples were transferred onto Vivacon 500 concentrators (MWCO of 30kDa; Sartorius: Goettingen, Germany), centrifuged and washed repeatedly with 8 M urea /0.1M Tris/Cl, pH 8.5, then alkylated with 0.05M iodoacetamide. Iodoacetamide was removed by 3 washes with 8 M urea /0.1 M Tris/Cl, pH 8.5, followed by 3 washes with 0.05 M ammonium bicarbonate. Proteins were digested with trypsin (sequencing grade, Promega: Madison, WI) at a 40:1 w/w ratio at 37°C for 16 hrs. Peptides were collected by centrifugation and protein concentration determined by BCA (Pierce: Rockford, IL). Aliquots containing equal protein amounts were desalted on SepPak C18 cartridges (Waters: Taunton, MA) and dried in a speedvac before being submitted to NanoLC/MS/MS.

Mass Spectrometry Analysis of Laser Capture Microdissected Glomeruli and Tissue IgG Co-Immunoprecipitation

Digested peptides were analyzed by NanoLC/MS/MS using a Thermo Orbitrap Fusion Lumos mass spectrometer. The peptides were loaded onto a reverse phase trap column (Integra-frit, New Objective, MA) containing Waters XSelect C18 CSH 2.5 μm resin coupled to a 150 mm X 0.075 mm analytical column containing the same reverse phase resin as used in the trap. A nanoAcquity UPLC system (Waters Corp, Milford, MA) was then used to generate a 60 min gradient from 98:2 to 60:40 buffer A:B ratio (Buffer A=0.1% formic acid, 0.5% acetonitrile; buffer B=0.1% formic acid, 99.9% acetonitrile). Peptides were eluted from the column with an integrated spray tip (picofrit, New Objective) and ionized by electrospray (2.0 kV) followed by MS/MS analysis using higher energy collision induced dissociation (HCD). Survey scans of peptide precursors were performed at 240K resolution (at 400 m/z) with a 5 × 10⁵ ion count target. Tandem MS was performed by isolation at 1.6 Th with the quadrupole, HCD fragmentation with normalized collision energy of 30, and rapid scan MS analysis in the ion trap. The obtained MS/MS data was searched against the most recent Uniprot human database containing both the Swiss Prot and the TREMBL entries using MaxQuant. Visualization of data was done using Scaffold v4.6. The false discovery rate was set at 1% for the peptide-to-spectrum matches. Normalized iBAQ values from MaxQuant were used for quantitation. iBAQ distributions for each sample were median-adjusted to control for differences in loading. iBAQ values equal to zero were removed from the data set. For statistical hypothesis testing, a two-sample Welch's t-test was performed for each protein using normalized iBAQ values for the two groups. If a protein was only detected in one group, a one-sample Welch's t-test was performed, using the smallest detected iBAQ value as the null hypothesis.

NCAM1 staining

Formalin-fixed paraffin-embedded sections, cut at 3 μm, were deparaffinized and antigen retrieval was performed by incubation at 99°C. The sections were reacted with rabbit polyclonal anti-NCAM1 antibody (Sigma, catalog #HPA039835, St. Louis MO), followed by a Rhodamine red X-conjugated goat anti-rabbit secondary which was solid-phase adsorbed to ensure minimal cross-reaction with human IgG (1:100; Jackson Immunoresearch, West Grove, PA). A second commercial antibody was also tested for specificity (Invitrogen anti-CD56 rabbit polyclonal antibody, catalog #PA5-83479). Each case was run with positive and negative controls. The stain was evaluated by standard immunofluorescence microscopy. The stain was judged to be positive if there was granular capillary loop staining in the glomeruli and negative if there was no capillary loop staining in glomeruli. Co-localization of IgG and NCAM1 along the glomerular basement membranes was examined by confocal microscopy using a Leica SBA DMI8 confocal laser scanning microscope. For this analysis, polyclonal (FITC-conjugated) rabbit anti-human IgG (1:40; Agilent, Santa Clara, CA) was reacted with FFPE tissue following antigen retrieval in heat and high pH citrate buffer and staining for NCAM1 as described above. Negative controls were performed to ensure antibody specificity by omitting primary antibodies. For identification of control MN cases, immunohistochemistry on MN cases was performed for PLA2R (PLA2R rabbit polyclonal antibody, Sigma, cat #HPA012657), and immunofluorescence for thrombospondin type 1 containing 7A (THSD7A, THSD7A rabbit polyclonal antibody, Atlas antibodies, catalog # AMAB91234), and exostosin-1 (EXT1, EXT1 rabbit polyclonal antibody, Invitrogen, catalog # PA5-60699).

For co-localization studies of IgG with NCAM1, 5 NCAM1 glomeruli from biopsies with NCAM1-positive immunofluorescent staining and 5 glomeruli from biopsies with PLA2R positive membranous nephropathy were examined. Confocal images were captured using sequential scanning of Z-stack images that were overlayed for maximal projection. Co-localization was quantified using the co-localization Leica image analysis software, where the percentage of co-localization of Rhodamine Red X (NCAM1) and FITC (IgG) was determined within glomeruli. Pearson correlation coefficients were determined for each glomerulus. A one-way analysis of variance was used to determine differences in percent co-localization between groups, using GraphPad Prism v5.0 software.

Western blotting:

NCAM1 recombinant protein (120 kDa isoform, R & D Systems, Minneapolis, MN) was electrophoresed at 0.4 μg/lane on Bolt 10% Bis-Tris Plus gels (Thermo-Fisher Scientific, Waltham, MA), transferred onto 0.4 μm nitrocellulose membranes, and blocked with 5% bovine serum albumin solution in phosphate buffered saline containing 0.1% Tween (PBST). The membranes were reacted with sera at 1:50 dilution in 2% bovine serum albumin solution in PBST for 2 h at room temperature, washed three times in PBST, and incubated with Affinipure rabbit anti-human IgG-HRP at 1:2000 dilution (catalog # 309-035-082, Jackson Immunoresearch, West Grove, PA) for 1 h. The immunoblots were washed with PBST and developed using 3, 3'-diaminobenzidine tetrahydrochloride (DAB, Agilent Technologies, Santa Clara, CA) for 3 minutes. Two serum samples from patients with NCAM1 positivity on kidney biopsy and serum samples from 14 patients with PLA2R-positive membranous nephropathy and 4 patients with EXT-associated MN were tested, after incubation at 56°C for 1 hour for complement inactivation. Reactivity of the anti-rabbit NCAM1 antibody (Sigma, catalog # HPA039835, St. Louis, MO) at 1:500 dilution was used as a positive control, with Affinipure goat anti-rabbit IgG-HRP used as a secondary antibody (catalog # 111-035-144, Jackson Immunoresearch, West Grove, PA).

Indirect immunofluorescence assay:

Human embryonic kidney (HEK)-293 cells were transfected with 0.5 μg NCAM1-expressing pCMV6 plasmid (Origene, CD56 NCAM1 NM_181351 human tagged ORF clone, cat #RC207890) per well of a 6-well plate for 4 hours in a 37°C incubator in minimal essential medium without FBS. pCMV6 plasmid transfections (pCMV6-entry mammalian vector with C-terminal Myc-DDK tag, catalog # PS100001) were used as negative controls. Ten percent fetal bovine serum was then added and the cells were incubated overnight. The cells were harvested using 0.25% trypsin solution, resuspended, and seeded onto slides coated with 0.1% poly-L-lysine solution, which were incubated overnight at 37°C. The slides were washed in PBS, fixed in 4% paraformaldehyde for 15 minutes at room temperature and permeabilized with 0.2% Triton-X100 solution in PBS for 5 minutes at room temperature. The slides were washed with PBS, then incubated with NCAM1 or PLA2R serum at a 1:10 dilution for 1 hour at room temperature. The slides were then washed in PBS, followed by incubation with anti-human IgG-FITC direct conjugated antibody (Kent Laboratories, catalog #1912MJ6) for 1 hour at room temperature. Slides were washed in PBS, then visualized under immunofluorescence. Anti-NCAM1 rabbit polyclonal antibody was used as a positive control.

Supplementary Material

1

Supplemental Figure 1. Cohort diagram showing identification of NCAM1-positive membranous nephropathy (MN) cases. A discovery cohort, through laser capture microdissection and protein G immunoprecipitation identified NCAM1 to be present in PLA2R-negative, THSD7A-negative, EXT-negative MN (3/13 cases, with 12 additional cases of known antigen type utilized as controls). Two validation cohorts were utilized, with immunofluorescence screening of cases for NCAM1-positive MN. The first series was a series of 218 membranous lupus nephritis (MLN) cases, of which 14 were NCAM1 positive (4.9%). Additional cases were identified through screening of biorepository samples (containing MN and MLN cases) in which identified 3/101 NCAM1-positive cases (3.0%), two of which had matched sera for analysis.

Supplemental Figure 2. Peptide based heat map shows a total of 18 NCAM1 peptides detected in cases of NCAM1-associated membranous nephropathy, which were not present in other membranous nephropathy cases assayed, which included PLA2R-positive MN controls, EXT-associated MN controls, and 10 additional cases of unknown type. Red coloring indicates the presence of peptides, while blue coloring indicates the absence of peptides.

Supplemental Figure 3. NCAM1 co-localizes with IgG in NCAM1-associated MN, but not PLA2R-positive MN controls. Three representative glomeruli from each category are shown, with IgG in green (FITC conjugate) to the left, NCAM1 in red (with rhodamine red X) in the middle of each panel, and the panels to the right indicate overlay images from the green and red channels.

Supplemental Figure 4. NCAM1 shows rare positive staining (arrow) in tubular basement membranes that co-localizes with IgG immune deposits in a biopsy with NCAM1-associated MN cases that contained extraglomerular staining. No PLA2R-positive MN control cases contained TBM deposits for evaluation.

Supplemental Figure 5. NCAM1 is expressed within glomeruli of NCAM1-associated MN cases, but not EXT-associated MN or proliferative lupus nephritis controls. The images are representative, with a total of 20 NCAM1-associated MN cases, 20 proliferative lupus nephritis cases, and 10 EXT-associated MN cases stained in total.

Supplemental Figure 6. Sera from NCAM1 patients did not react with NCAM1 recombinant protein under reducing conditions. Both NCAM1 cases and representative PLA2R western blot strips are shown. An anti-rabbit NCAM1 antibody reacting with the NCAM1 recombinant protein serves as a positive control.

Supplemental Figure 7. EXT-associated membranous lupus nephritis serum did not show reactivity against NCAM1-recombinant protein under non-reducing conditions, demonstrating specificity to NCAM1-associated membranous nephropathy. Reactivity of a rabbit polyclonal anti-NCAM1 antibody is shown as a positive control.

Supplemental Figure 8. NCAM1 sera reacted with NCAM1 over-expressing HEK-293 cells in an indirect immunofluorescence assay, but not pCMV6 vector control or HEK-293 cells alone (top panels). PLA2R sera did not show seroreactivity. Reactivity with rabbit polyclonal anti-NCAM1 antibody was used as a positive control.

Supplemental Table 1. IgG subclass staining in NCAM1-associated MN. Thirteen of 20 NCAM1-associated MN cases had tissue available for IgG subclass staining.