Monoclonal and polyclonal immunoglobulin G deposits on tubular basement membranes of native and pretransplant kidneys: A retrospective study

Monoclonal tubular basement membrane immune deposits (TBMID) are associated with progression of interstitial injury in renal allograft. However, the significance of monoclonal and polyclonal TBMID in the native kidney remains unclear. We retrospectively analyzed 1894 native kidney biopsies and 1724 zero‐hour biopsies performed between 2008 and 2018 in our institution. The rate of immunoglobulin G (IgG) TBMID was found to be 8.4% among native kidney biopsies and 0.4% among zero‐hour biopsies. Polyclonal TBMID is common in IgG4‐related tubulointerstitial nephritis (37.5%), diabetic nephropathy (31.3%) and lupus nephritis (25.5%). Monoclonal IgG TBMID was identified in seven cases, including three zero‐hour biopsies. The combination of IgG1κ was observed in two cases, IgG1λ in three, and IgG2κ in two. Electron microscopy revealed powdery electron‐dense deposits in all cases. Monoclonal gammopathy of undetermined significance was diagnosed in one case. Although one patient with focal segmental glomerulosclerosis developed renal failure, all others exhibited stable renal function. Monoclonal IgG TBMID in the native kidney is not associated with renal prognosis. However, this may be an interesting immunopathological finding that would help clarify the pathogenesis of TBM immune deposits. Further study for both monoclonal and polyclonal TBMID is required in the future.


INTRODUCTION
Immunofluorescence staining has been an important tool in renal biopsy to evaluate glomerular immune type deposition. Particularly, the distribution of immunoglobulin (Ig) G subclass and monotypic immunoglobulin deposition in glomeruli are important factors to characterize various glomerular diseases. [1][2][3] According to a recent consensus report, monoclonal IgG glomerulonephritis is characterized by monotypic IgG deposition in the glomeruli and/or along tubular basement membrane (TBM). 4 Monoclonal immunoglobulin glomerulonephritis is usually associated with an underlying monoclonal gammopathy, but not all patients. 5 On the other hand, immune complex glomerulonephritis is characterized by deposits of polyclonal immunoglobulin. 4 Moreover, it has been reported that the pathophysiology of monotypic immunoglobulin deposition is different in entity from conventional diseases. In one example, Nasr et al proposed anti-glomerular basement membrane (GBM) glomerulonephritis with monotypic immunoglobulin deposition as an atypical anti-GBM nephritis, distinguishing it from classic anti-GBM glomerulonephritis with polytypic linear GBM staining. 6 Furthermore, it has been reported that monotypic and polytypic IgA deposition are different in IgA nephropathy. 7 However, there are few studies on tubular basement membrane immune deposits (TBMID). TBMID is found in cases of lupus nephritis, membranoproliferative glomerulonephritis, membranous nephropathy, drug-induced tubulointerstitial nephritis, Sjögren's syndrome, IgG4-associated nephritis, and other autoimmune diseases. 8,9 Although previous studies have considered IgG subclass distribution in TBM for lupus nephritis and IgG4-related diseases, the significances still remain unclear. 10,11 Glomerulonephritis that is characterized by monotypic IgG deposition has been reported as proliferative glomerulonephritis with monoclonal immunoglobulin deposit (PGNMID). 12 PGNMID often lacks monoclonal gammopathy and clinical features and pathogenesis of PGNMID are different from other immune-related glomerulonephritis. 13 We have previously reported the association of monoclonal IgG TBMID with the formation of electron-dense deposition (EDD) in the TBM with subsequent progression of interstitial fibrosis and tubular atrophy (IFTA). 14 However, to the best of our knowledge, there are no reports on monoclonal TBMID in native kidneys or pretransplant kidneys.
In the present study, we analyzed biopsy specimens of polyclonal and monoclonal IgG TBMID in native kidneys and pretransplant kidneys. This retrospective clinicopathological study aimed to determine the clinicopathological features of polyclonal and monoclonal IgG TBMID.

MATERIALS AND METHODS
We recruited all patients between April 2008 and March 2018 at Tokyo Women's Medical University with data of native kidney and zero-hour kidney biopsy. In a zero-hour biopsy, a cortical wedge is taken immediately after revascularization on completion of vascular anastomosis. All renal biopsies were examined with standard light microscopy (LM) and immunofluorescence (IF). For LM, all biopsies were routinely stained with hematoxylin and eosin, periodic acid-Schiff (PAS), Masson's trichrome and PAS methenamine silver. For IF, 2 μm cryostat sections were dried and stained with fluorescein isothiocyanate (FITC)-conjugated polyclonal antibodies against IgG; IgA; IgM; complement components 3, 1q and 4d (C3c, C1q and C4d); fibrinogen; κ light chain; and λ light chain (Dako Corp., Carpinteria, CA, USA). Measurement of IgG was performed on 2 μm cryostat sections stained with FITC-conjugated monoclonal antibodies against IgG1, IgG2, IgG3 and IgG4 (The Binding Site Inc., San Diego, CA, USA). Monoclonal IgG TBMID was identified by the presence of monotypic staining for a single light chain isotype and a single-heavy chain subclass in the TBM. Exclusion criteria of monoclonal TBMID was when multiple light chains were deposited in the TBM, even if the heavy chain subtype was isotypic. Moreover, patients with glomerular co-deposition were excluded. Therefore, we excluded two cases of LHCDD cases from monoclonal TBMID because of glomerular co-deposition. On the other hand, polyclonal TBMID was identified as the presence of polytypic staining for light chain isotype and heavy chain subclass in the TBM.
We performed electron microscopy (EM) for 57 of the 165 patients with IgG deposits in the TBM. EM was performed for 39 of the 91 patients in whom the IgG subclass was obtained, including 7 patients with monoclonal TBMID. Therefore, we compared the EM findings with those of 7 patients with monoclonal TBMID, 30 with polyclonal TBMID and 2 with LHCDD.
All biopsies were reviewed by three independent pathologists. Patient information was collected from medical records. The study was carried out in compliance with the Declaration of Helsinki and approved by the ethics committee at Tokyo Women's Medical University (No. 5415).

Distribution of IgG subclasses in the tubular basement membrane
We analyzed 1894 native renal and 1724 zero-hour biopsy specimens in the present study. Table 1 presents the IgG subclass distribution of TBMID in relation to various diseases. We identified IgG TBMID in 164 (8.4%) patients who underwent native renal biopsy and five (0.3%) who underwent zero-hour biopsy. The rate of TBMID is relatively higher in diabetic kidney disease (31.3%), lupus nephritis (25.5%), and IgG4-related disease (37.5%). On the other hand, TBMID was rare in IgA nephropathy (1.1%), tubulointerstitial nephritis (3.7%), and focal segmental glomerulosclerosis (7.7%). The rate of IgG1 deposition (92.3%) was significantly higher than the other subclasses (IgG2 (67.0%), IgG3 (58.2%), and IgG4 (53.8%)). In lupus nephritis, 17 of 38 cases that showed TBM deposits showed positive staining for all four subclasses. The rate of IgG3 and IgG4 deposition was lower than the other two IgG subclasses. In IgG4-related disease, all cases were positive for all four subclasses. In zero-hour biopsies, TBMID was exceedingly rare (0.3%), and all cases showed deposition of monotypic IgG subclass in the TBM.

Electron microscopic characteristics of monoclonal and polyclonal IgG TBMID
As the next step, we examined the EM findings of monoclonal and polyclonal TBMID. The EM findings are presented in Fig. 1. For cases where an IgG subclass was obtained, EM was performed for 39 of the 91 patients, including seven cases of monoclonal TBMID. In polyclonal TBMID, EDD was found in 10 of the 30 patients (33.3%); one patient each had powdery, nodular and fibrillary EDD, respectively. In the other seven cases, no specific morphology was observed, and they were classified as not otherwise specified (NOS). In LHCDD patients, 2 of 2 patients had powdery EDD. The EM findings of polytypic TBMID in each disease are shown in Table 2. EDD in the TBM was found in 38.4% of the patients with lupus nephritis and 50% of those with membranoproliferative glomerulonephritis. In all cases of IgG4-related disease and tubulointerstitial nephritis, EDD were detected. Alternatively, there was no EDD in any patient with diabetic kidney disease, membranous nephropathy, and FSGS. However, in monoclonal TBMID, the EM analysis revealed powdery EDD in all cases (Fig. 2). Only one patient showed slight EDD in the glomerulus (Case 1). In contrast, other patients did not show powdery EDD in the glomerulus.

Clinical characteristics of patients with monoclonal IgG TBMID
Monoclonal IgG TBMID were identified in seven patients, including three who underwent zero-hour biopsy. Table 3 presents the clinical characteristics of patients withmonoclonal IgG TBMID including four patients performed of native kidney biopsy and three donor patients of zero-hour biopsy (four male and three female). The median age at transplantation was 53 (21-69) years and median serum creatinine at the time of biopsy was 0.81 (0.56-1.4) mg/dL. Median proteinuria at biopsy was 0.2 (negative-3.0) g/day. Hepatitis B or C infection or other infectious disease was not identified. Monoclonal gammopathy of undetermined significance was diagnosed in one patient (Case 6) based on IFE. Dysproteinemia was not detected in any other patients. Moreover, no patient was diagnosed with multiple myeloma. All four patients who underwent native kidney biopsy received treatment for glomerular disease. One patient received rituximab treatment (Case 1), while three received steroid treatment for IgA nephropathy and ANCAassociated vasculitis. All three cases of donor in zero-hour biopsy were observed without treatment. The median follow-up period from biopsy was 3.5 (0.5-6) years. Although one patient progressed to end-stage renal disease (ESRD) 6 years after biopsy despite rituximab treatment (Case 1), all other patients exhibited stable renal function. Median serum creatinine at the last follow-up was 0.8 (0.57-1.19) mg/dL.      vasculitis (n = 1), and focal segmental glomerulosclerosis (FSGS) (n = 1). The other two zero-hour biopsies showed only mild atherosclerosis. Neither tubulointerstitial nephritis nor cast in the tubules was identified in any patient. The median IFTA was 5% (range 0-30%). Glomerular IgG deposition was found to be negative in all cases by IF analysis. Monotypic IgG deposition was present in the proximal tubule in all cases. The combinations of IgG subclass and light chain in the TBM were IgG1κ (n = 2), IgG1λ (n = 3), and IgG2κ (n = 2). The IF staining pattern was linear, and C3c, C4 and C1q staining was negative in the TBM in all cases (Figs. 3,4; Table 4). In the biopsies of three renal allograft cases, deposits in the TBM weakened gradually (Fig. 4 shows Case 5 as a representative example).

DISCUSSION
In this study, we examined the distribution of IgG subclass in the TBM in various stages of renal disease. Additionally, we found seven cases of monoclonal IgG TBMID and examined their clinicopathological characteristics.
In humans, there are four subclasses of IgG (IgG1-4). Each IgG subclass is associated with immunological and functional roles. In glomerular disease such as membranous nephropathy, the specific deposition of IgG subclasses in the glomerulus has diagnostic significance. 15 The roles of IgG1 and IgG3 are to activate the complement system, whereas IgG2 and IgG4 have little involvement with this activation. 16 The present study found TBMID to be common in the context of diabetic nephropathy, membranous glomerulonephritis, and lupus nephritis, with the rate of IgG1 deposition being much higher than other IgG subclasses. However, no differences were observed between the other three IgG subclasses. About half of the cases of lupus nephritis showed deposition of all subclasses, with IgG1, IgG2 and IgG3 being predominant. These findings are consistent with prior studies on IgG subclass deposition in the glomeruli. 17 On the other hand, TBMID is exceedingly rare in IgA nephropathy and living transplant donors. All these cases showed monotypic IgG deposition in the TBM. This suggests that examination of IgG subclass in the TBM should be carried out for such cases.
In the seven cases of monoclonal TBMID that were found in the present study, IgG1 was the most common subclass, and the ultrastructure of the EDD in the TBM exhibited a powdery structure. In contrast, approximately only 30% of the patients with polytypic TBMID showed EDD in the EM analysis, and the characteristics of EDD differed as per the underlying disease. Moreover, PGNMID is characterized by non-organized EDD and the most common subclass is IgG3. 13 We have previously reported the presence of  IgG1λ Linear Zerohour biopsy   monoclonal IgG TBMID in renal allograft and found monoclonal IgG1κ TBMID to exhibit non-organized EDD formation in the TBM and to be associated with the progression of IFTA. 14 However, the present study did not identify remarkable IFTA in any case. Taken together, our findings suggest that monoclonal IgG TBMID in native kidney occurs via a different mechanism than in the case of polyclonal TBMID, PGNMID, and monoclonal IgG TBMID in renal allograft. This may be because the pathogenesis of the TBM immune complex formation is different from that in the glomeruli and renal allograft. 10,18 The pathological findings of the present study reflect monoclonal immunoglobulin deposition disease (MIDD). The most common glomerular manifestation of light chain deposition disease (LCDD) is nodular glomerulosclerosis, while many cases display normal glomeruli in LM. Moreover, some cases of light chain deposition disease (LCDD) exhibit deposits localized to the renal tubules and powdery EDD in EM analysis. 19 On the other hand, there are some differences between the cases presented here; while light chain and heavy chain deposition disease (LHCDD) and heavy chain deposition disease (HCDD) are often characterized by deposition of C3c and C1q, 20 we did not observe such deposition in the TBM. Moreover, while LHCDD is well known for its poor prognosis, almost all our cases had good renal and life prognoses. The pathological significance of these differences is unclear. Previous reports have demonstrated that the various subtypes of Bence-Jones protein have different roles in the kidney, 21 which suggests that pathological and clinical changes depend on the type of free light and heavy chain. Moreover, in patients with a renal allograft, deposits in the TBM had weakened at the time of follow-up biopsies. Previous reports have shown that transmitted IgA nephropathy and membranous glomerulonephritis resolved after renal transplantation. 22,23 We also reported the full-house immunofluorescence staining in renal allograft that resolved gradually on follow-up biopsies. 24 These findings indicate that monoclonal TBMID can be interpreted as a donor-derived non-pathognomonic deposition. In this study, one patient progressed to ESRD, which was diagnosed as FSGS and showed nephrotic-range proteinuria and relatively high serum creatinine at the time of renal biopsy. Some reports have stated that renal insufficiency is an independent risk factor for MIDD. 25,26 In this case, tubular lesions as well as glomerular lesions may contribute to the progression of ESRD. These results suggest that early detection of renal dysfunction and comprehensive histologic evaluation, including IF and EM, is important in monoclonal IgG TBMID.
There are several limitations of this study which should be acknowledged. First, due to the retrospective nature, some laboratory data were not obtainable. In this study, dysproteinemia was evaluated using SPEP/UPEP. However, only one patient (Case 6) was evaluated using both SPEP/UPEP and IFE. In this case, dysproteinemia was not detected using SPEP/UPEP but detected by IFE. SPEP/UPEP is less sensitive than IFE. Therefore, we cannot exclude the low burden disease from the negative results. All patients with monotypic TBMID should be evaluated with IEE in the future.
Second, monotypic IgG and light chain deposition was only evaluated by IF. The affinity of IgG to the TBM varies depending on the type of immunoglobulin and light chain. Although each IgG subclass except IgG4 and both κ and λ light chain were detected, we cannot deny that the affinity of immunoglobulin and light chain for the TBM may have affected our results. Future study requires molecular studies such as proteomics analysis. Third, IgG subclass was not examined in all cases. In our laboratory, IgG subclass staining is not routinely performed; for example, IgG subclass staining is carried out to distinguish between idiopathic or secondary membranous nephropathy. However, it is known that there are many nonspecific IgG deposits in the TBM in diabetic nephropathy. Thus, many cases did not undergo IgG subclass staining and there may be more cases of monoclonal IgG TBMID than our results suggest. Moreover, renal diseases that are complicated with monoclonal IgG TBMID may be more diverse. Fourth, we did not perform EM for all patients. Moreover, although we obtained several EM images, including the tubules, the EM images mainly focused on the glomerulus in many cases. Thus, EDD may be undetectable if the deposits are focal or in small amounts. In this article, approximately 60% of the patients with lupus nephritis did not have EDD in the TBM; however, this could be attributable to insufficient examination.
In conclusion, IgG TBMID in the native kidneys is associated with formation of EDD on the TBM but does not influence renal prognosis or IFTA progression. However, IgG TBMID might be a potentially useful immunopathologic parameter for cases of MIDD lacking monotypic glomerular IgG deposition or where there is asymptomatic underlying hematologic disorder. Careful ultrastructural and immunohistochemical analysis of the type of deposit on the TBM as well as hematological investigations could be beneficial in routine clinical practice.

AUTHOR CONTRIBUTIONS
AS designed the research and wrote manuscript. ST, KH and JK supported in writing this report and revised it. SH, AS and YN analyzed pathological information. MO, TT, HI, KT MH and KN analyzed clinical information. All authors read and approved the final manuscript.

HUMAN AND ANIMAL RIGHTS
The study was in compliance with the Declaration of Helsinki. Informed consent was obtained by optout as it was a retrospective study. This study and all its protocols were approved by the ethics committee at Tokyo Women's Medical University (No. 5415).