Membranous nephropathy with masked polyclonal IgG deposits associated with primary Sjögren’s syndrome

Abstract

Tubulointerstitial nephritis and renal tubular acidosis are well-known renal involvements with primary Sjögren’s syndrome. However, several types of glomerulonephritis such as membranoproliferative glomerulonephritis and membranous nephropathy are also known to develop in patients with this syndrome. We here report a case of membranous nephropathy that developed 8 years after a diagnosis of primary Sjögren’s syndrome in a female patient. Interestingly, the deposition was not identified by routine immunofluorescence using snap frozen tissue, but was revealed by immunofluorescence on formalin-fixed paraffin-embedded sections treated with proteinase K. We further performed immunofluorescence analysis on the treated paraffin-embedded sections with the identified antigen but found that the deposited IgG was not monoclonal and that serum amyloid P, a sensitive marker for membranous-like glomerulopathy with masked IgG κ deposits, was not evident in the glomeruli. To the best of our knowledge, this report depicted the first case of masked polyclonal IgG deposits and further analysis is needed to clarify the underlying mechanisms of IgG masking and possible association with autoantibodies.

Introduction

Primary Sjögren’s syndrome (pSS) is a chronic, progressive autoimmune disorder characterized by lymphoplasmacytic infiltration of the exocrine glands, mainly the salivary and the lacrimal glands [1]. Systemic multi-organ involvement including the kidney can also occur in pSS and renal tubular acidosis is a common clinical presentation of the kidney involvement [2]. Histologically, tubulointerstitial nephritis is the major finding in pSS patients who have undergone a renal biopsy, but glomerulonephritis such as membranoproliferative glomerulonephritis (MPGN), membranous nephropathy (MN), and IgA nephropathy have also been described in these cases [2, 3]. Direct immunofluorescence (IF) analysis using frozen renal tissue can usually differentiate these conditions, but sometimes results in false-negative staining. As an alternative approach, IF staining of formalin-fixed paraffin-embedded tissue treated using an antigen retrieval step enables the antibodies to react with “masked” deposits in the glomeruli and thereby visualize them. This unmasking phenomenon is well known in paraprotein-related kidney diseases, particularly light-chain proximal tubulopathy [4, 5]. In 2014, Larsen et al. [6] reported 14 cases of peculiar glomerulopathy with subepithelial and frequent mesangial deposits in spite of negative staining by routine IF. They revealed the monoclonal deposition of IgG-κ by protease digestion, and thus coined the term ‘membranous-like glomerulopathy with masked IgG κ deposits’ (MGMID) [6]. Very recently, the same group identified serum amyloid P as a sensitive marker of MGMID [7].

We here report a patient with pSS who developed MN with masked IgG deposits. In contrast to MGMID, however, the deposits in our case did not show light-chain monoclonality and no serum amyloid P deposition was noted in the glomeruli. To our knowledge, this is the first case of membranous nephropathy with masked polyclonal IgG deposition.

Case presentation

An 85-year-old Japanese woman with pSS had developed 3 + proteinuria, revealed on a dipstick, with proteinuria of 3.13 g/gCr and decreased serum albumin (3.3 g/dL) was referred to the nephrology department of our hospital. Eight years prior to this episode, she was diagnosed with pSS due to sicca syndrome and showed positive results on a Schirmer test, and was serologically positive for SS-A and SS-B antibodies. Anti-nuclear antibodies also gave a positive result in a nucleolar pattern. She had been suffering only from sicca syndrome and had been attending our outpatient ward for follow-ups since her initial pSS diagnosis. She had no history of hypertension and has never developed Raynaud’s phenomenon. At the time of admission to our clinic, the patient was 147-cm tall and weighed 59.5 kg. Physical examination revealed pretibial edema, but no joint pain, neurological symptoms, or lupus-related skin lesions were evident. She had a blood pressure of 157/59 mmHg, and her laboratory test results were as follows: hemoglobin, 12.3 g/dL; total leucocyte count, 8600/mm³; lymphocyte, 950/mm³; platelets, 282,000/mm³. Her serum protein level was 7.9 g/dL, albumin 3.6 g/dL, total cholesterol 212 mg/dL, and triglyceride 106 mg/dL, urea nitrogen 16.5 mg/dL, serum creatinine 0.6 mg/dL eGFR 70.0 mL/min/1.73 m², sodium 137 mmol/L, potassium 3.8 mmol/L, chloride 103 mmol/L, calcium 9.1 mg/dL, phosphate 3.4 mg/dL. Urine pH was 7.0, and urinary protein excretion was 256 mg/day. The urinary sediment contained 1–4 erythrocytes and 0–1 leukocyte per high-power field, and urinary NAG level and β2-microglobuin was 16.1 U/L and 558 μg/L, respectively. Her serum IgG was slightly elevated at 2453 mg/dL and IgG4 at 192 mg/dL, but serum electrophoresis detected no monoclonal gammopathy and she was negative for Bence-Jones protein. Serological examinations revealed an anti-nuclear antibody titer of 1:640 in a speckled pattern and of 1:320 in a nucleolar pattern. Her levels of anti-ssDNA antibody were 9.2 U/mL (reference range < 6.9); anti-SS-A antibody, 7195 U/mL; and anti-SS-B antibody, 64.5 U/mL. She was negative for anti-dsDNA, anti-Sm, anti-phospholipid, anti-cardiolipin, anti-β2GP1 complexes, and anti-Scl70 antibodies. Serum cryoglobulin was absent, and serum complement was not decreased. She was subsequently hospitalized for a renal biopsy to differentiate the cause of her renal dysfunction.

The renal biopsy revealed that 25 glomeruli were contained, nine of which showed global sclerotic changes, and interstitial inflammation was minimal (Fig. 1a). Stippling of the glomerular basement membrane was obvious on silver methenamine-stained sections (Fig. 1b). Although no positive staining was found in the glomeruli on IF staining using frozen tissue, electron microscopy revealed subepithelial dense deposits along the glomerular capillary wall with occasional intramembranous lucent deposits (Fig. 1c, d). This strongly indicated stage 3 membranous nephropathy. No mesangial or subendothelial deposits were noted.

Fig. 1.

Light, immunofluorescence, and electron microscopy findings for the renal biopsy specimen. a No tubulointerstitial inflammation was evident on light microscopy (Periodic acid–Schiff staining; original magnification, × 100). b The glomerular basement membrane showed a bubble-like appearance on periodic acid–methenamine-silver staining (original magnification, × 400). c No IgG deposits within the glomeruli were identified on routine immunofluorescence. d Electron microscopy revealed subepithelial deposits. Intramembranous electron lucent deposits were also noted (arrows) (bar 1 µm)

To confirm the immunoglobulin deposition in the glomeruli, immunofluorescence using formalin-fixed paraffin-embedded (FFPE) sections treated with proteinase K was performed as previously described [8]. This immunofluorescence analysis revealed granular positive IgG staining along the glomerular capillary wall with faintly positive staining for C3, suggesting possible immune complex deposition, and IgG4 was the dominant deposit among the IgG subclasses (Fig. 2). We further tested the possibility that the patient had MGMID, but no monoclonal light-chain deposits were evident (Fig. 3a, b). Larsen et al. [7] recently identified serum amyloid P as a sensitive and specific marker of MGMID, but we could not detect it in the glomeruli in our current case (Fig. 3c–e). Immunofluorescence for phospholipase A2 receptor (PLA2R) was also negative (data not shown). Lastly, we tested the possibility of false-negative results of our routine immunofluorescence in which we use polyclonal rabbit FITC-labeled anti-human IgG antibody (F0202) purchased from Agilent technologies (CA, USA). This time, we performed immunofluorescence using Dylight 488-conjugated polyclonal goat anti-human IgG antibody (NBP1-74980) purchased from Novus Biologicals (CO, USA), but again failed to reveal the deposition (data not shown). Taken all together, we concluded that our patient had developed stage 3 MN with masked polyclonal IgG deposition.

Fig. 2.

Immunofluorescence findings of formalin-fixed paraffin-embedded sections treated with proteinase K digestion. Positive IgG staining was identified within the glomeruli and C3 was also faintly positive. IgG4 deposits were dominant among the IgG subclasses (a IgG, b C3, c IgG1, d IgG2, e IgG3, f IgG4)

Fig. 3.

Additional staining of paraffin-embedded tissue treated with proteinase K digestion. The deposited IgG was polyclonal and amyloid P was not found in the glomeruli (a κ, b λ, c IgG, d serum amyloid P, e merged image of IgG and serum amyloid P staining)

After evaluation of the biopsy, we administered a diuretic agent to our patient and her proteinuria was decreased. Her most recent serum creatinine and urine protein–creatinine ratio (measured four months after the kidney biopsy) were 0.59 mg/dL and 1.53 g/gCr, respectively.

Discussion

MGMID have been well described by Larsen et al. and most commonly arise in women under 40; these cases show positive autoimmune serologic study results such as antinuclear antibodies but few fulfill the criteria for well-defined autoimmune diseases such as SLE or RA [9]. These authors also found that IgG deposits were not observed by routine immunofluorescence methods and pronase treatment of paraffin-embedded sections was required to reveal them. They further noted monotypic deposits of κ light chain and dominant IgG1 deposits among a few cases tested for IgG subclasses, though no case presented with hematological malignancies and more than 95% of the patients were negative on serum electrophoretic studies [9]. MGMID is a very rare form of glomerulonephritis identified only in 0.13% of native kidney biopsies [6]. Until this description of MGMID, these cases were likely to be misdiagnosed as C3 glomerulopathy, or as an atypical form of infection-associated glomerulonephritis or membranous nephropathy [6]. Recently, the same group additionally reported that serum amyloid P was colocalized with IgG deposits along the glomerular basement membrane, and thus concluded that amyloid P is a sensitive and specific marker of MGMID [7]. In our current MN case, proteinase K digestion revealed masked IgG deposits along the GBM. However, this deposited IgG was not monotypic and amyloid P was not detected in the glomeruli. These findings made MGMID less likely in our patient. Class V lupus nephritis might be another possible disease in our case, but serum titers of anti-DNA, anti-dsDNA, and anti-ssDNA autoantibodies were too low to suggest SLE, and the patient did not develop other clinicopathological findings to be diagnosed as SLE.

Several prior case series have described glomerular diseases that have developed in pSS patients, with the rate of MN among pSS patients reported to vary from 4.0 to 35.9% [2, 10–14]. Interestingly, the rate of MN in Western and Japanese cohorts of pSS cases is less than 6% whilst rates of 15.6% and 35.9% are reported for MN development in Chinese cohorts. The specific high prevalence of MN in pSS patients in China could be partly due to higher air pollution, with particulate matter of 2.5 mm (PM2.5) reported by Xu et al. [15]. These authors reported that the frequency of MN in China had increased from 12.2% in 2004 to 24.9% in 2014, with the adjusted odds for MN having increased by 13% annually over that 11-year period but that for other major glomerulopathies remaining stable, and long-term exposure to high levels of PM2.5 showing an association with an increased risk of MN [15].

Pathological findings have been reported that can differentiate secondary MN from primary MN. In general, IgG4 predominance among the IgG subclasses, and PLA2R or thrombospondin type-1 domain-containing 7A (THSD7A) positivity are the best known IF results that indicate idiopathic membranous nephropathy [16–18]. In addition, neural epidermal growth factor-like 1 protein (NELL-1) has recently been added as a novel target protein in primary MN [19]. In the case of IgG4 and PLA2R staining patterns in the glomeruli of pSS cases, few reports have described the results of IF analysis for these markers. Yabuuchi et al. have reported negative staining for PLA2R whilst positive IgG4 signals were detectable along the glomerular basement membrane in a Japanese pSS patient [20]. To date, IgG4 deposits have tended to be positive in primary MN but not exclusively. Huang et al. reported that IgG4 was not dominant in 24% of their primary MN cases, and they stated that IgG1 was the dominant IgG subclass deposit in 64% of early stage (stage 1) MN cases. This indicated that IgG subclass staining was a less sensitive test for primary or secondary MN. In addition, PLA2R has been found not to be specific for primary MN. Its positivity is much higher among primary MN cases but the prevalence of PLA2R deposits in the glomeruli of cancer patients is reported to be 25% [21]. Immunofluorescence results, therefore, should be interpreted very carefully to determine a minor population of MN as primary or not.

The mechanisms of IgG masking have not been elucidated. In cases with MGMID, Larsen et al. [7] speculated that lack of Ca²⁺ in the media and buffers used for IF might disrupt amyloid P and IgG interactions, thus resulting in negative staining using snap frozen sections whilst formalin fixation would preserve such interactions. In our case, some unknown molecules other than SAP could interact with the deposited IgG and prevent its detection by routine immunofluorescence. In addition, autoantibodies might play some role in the masking phenomenon. Our case develops several kinds of autoantibodies such as SS-A, SS-B and anti-nuclear antibodies. MGMID cases were also known to develop autoimmune serologic results, with 55% of patients showing positivity of anti-nuclear antibodies and/or anti-dsDNA antibody [9]. Because masked IgG deposits with Sjögren’s syndrome have not been reported so far, our case indicated that autoantibodies showing nucleolar pattern such as anti-U3-RNP, anti-7-2 RNP, anti-RNA polymerase, anti-NOR90 antibodies might be attributable to false negativity in routine IF although we did not confirm the existence of such antibodies in our case due to lack of symptom suggesting progressive systemic sclerosis or mixed connective tissue disease. Interestingly, C3 seemed to be also masked in our case, although for unknown reasons, and the similar phenomenon was reported in MPGN with masked monotypic IgG deposits; two out of 16 cases of such unique MPGN showed false-negative staining of C3 by routine IF [22]. Additional cases with pSS or autoantibodies showing nucleolar pattern should be recruited for further analysis of renal biopsy specimens to elucidate the clinical and pathological characteristics in more detail. Mass spectrometry analysis of similar cases would be also helpful to dissect the molecules involved in the deposits and mechanisms of masking antigen, and to determine MN cases as primary or secondary.