IgG Subclass Expression in Diabetic Nephropathy


 Background: This study aimed to analyze the distribution of IgG subclasses in diabetic nephropathy (DN) and its association with clinico-pathological features. Methods: Forty DN cases were analyzed to identify IgG subclasses, as well as collagen IV α5, CD34, and KIM-1.Results: Both IgG and its subclasses showed a linear expression and overlapped with collagen IV α5 on glomerular basement membrane (GBM) and some of tubular basement membrane (TBM), without complement deposition. Eleven cases of IgG subclass deposition along both GBM and TBM were associated with more proteinuria. Five cases of TBM-only IgG subclass deposition were accompanied with less KIM-1 positivity and more arteriosclerosis. The major IgG subclasses expressed on GBM were IgG1 and IgG2, while TBM expression was mainly IgG1 and IgG3. Glomerular IgG1-positive status was associated with less CD34 expression, while IgG2-positive status was associated with thicker GBM. Expression of multiple IgG subclasses along TBM showed less KIM-1 positivity and interstitial inflammation than those with isotype or no IgG subclass expression.Conclusions: IgG subclasses were selectively deposited along GBM and TBM in DN, which was determined by their profiles and severity of glomerular/tubular injury. IgG and its subclass deposition is not causal, but the consequence of renal injury and these positive statuses are associated with different DN injuries.


Introduction
Diabetic nephropathy (DN) has a unique histological pattern, including glomerular basement membrane (GBM) thickening, mesangial expansion, and glomerulosclerosis, among others [1]. As a nonimmunological-related renal disease, the detection of IgG deposition along GBM and/or tubular basement membrane (TBM) in some DN cases is surprising [2]. This linear pattern can not be detected by regular electron microscopy in DN, which is different from its deposition in immune complex-related glomerular diseases, such as membranous nephropathy (MN) [2]. Some hypotheses suggest that structural changes of the basement membrane lead to entrapment of serum proteins, including albumin and IgG [3,4]. One study demonstrated that up to 51.5% of the DN biopsies were IgG-positive, and the IgG intensity was associated with the progression of renal injury [2].
IgG has four subclasses (IgG1-G4), each subclass has a unique pro le regarding half-life, antigen binding, immune complex formation, complement activation, and triggering of effector cells. Oxelius VA found that serum IgG2 and IgG3 levels declined, while IgG1 and IgG4 were relatively normal in juvenile diabetes mellitus cases. Susanna M et al. found that IgG4 was selectively eliminated and urinary IgG4 could be a useful marker for preclinical stages of diabetic nephropathy [5,6]. Using the biopsy tissue, Hemminger J et al. conducted a large retrospective study of IgG subclasses in 1084 cases, which showed that IgG4dominant/codominant deposition with PLA2R-positive status was associated with primary MN, while IgG1 dominant/codominant with weak or absent IgG4 deposition was associated with autoimmune disease-related MN [7]. However, no IgG subclass in DN was included in this study. In 1984, Melvin T et al. studied IgG subclass in nine DN cases, and found that only IgG4 had the same glomerular linear deposition as IgG [8]. This is di cult to explain by using the IgG pro les, since IgG4 has an anionic charge and the lowest serum concentration compared with the other subclasses, and GBM is anionically charged [6,9]. Further, the effect of tubular IgG deposition remains to be studied. Therefore, we studied the distribution of IgG and its subclasses in 40 DN cases, and found more proteinuria in cases with both GBM and TBM IgG subclass deposition than negative ones. Less tubular injuries, but more arteriosclerosis were detected in TBM-only deposition group. There were more DN cases of GBM linear staining with the IgG1 or IgG2 isotype than those with the IgG4 isotype. Glomerular IgG1positive status was associated with least glomerular epithelial cell injury, while IgG2-positive status was associated with thicker GBM compared to other groups. Furthermore, DN with multiple IgG subclass expression alon TBM showed less tubular injury and interstitial in ammation than those with isotype IgG subclass expression.

Clinical and pathologic characteristics
The following clinical information was collected: age, gender, 24-hour proteinuria, serum albumin, serum creatinine (Scr), and the estimated glomerular ltration rate (eGFR).
All specimens were processed for light microscopy (LM), IF, immunohistochemistry (IHC), and electron microscopy (EM). Classi cation of DN and histological scoring were done according to the criteria reported by Tervaert et al [1]. Interstitial brosis and tubular atrophy (IFTA, 0-3 score), as well as arteria hyalinosis and sclerosis (0-3 score) were scored according to methods described by the previous study [1]. Diagnosis, classi cation, and the score of these pathological ndings were evaluated and con rmed by two renal pathologists. GBM and TBM thickness were measured under an electron microscope according to studies reported by Haas and Tyagi I [10,11]. Podocyte effacement was measured according to the degree of foot process fusion [12].

Statistical analysis
Statistical analysis was performed with SPSS 17.0. Normally distributed data were expressed as mean ± SD. A comparison of clinical and pathological characteristics among groups was assessed by t-test or ANOVA for continuous variables, and by nonparametric tests for discontinuous variables. Categorical variables were expressed as percentages and comparisons among groups, which were assessed by chisquare test or the Fisher's exact test. A P-value of less than 0.05 was considered signi cant.

Clinical and morphological features
The average age at the time of renal biopsy of all 40 cases was 55.6 ± 15.7 years, and 28 cases comprised males. Patients showed proteinuria (urinary protein 4.8 ± 3.2 g/day), with normal to mild reduction of renal function (Scr 114.2 ± 61.0 umol/L, eGFR 78.4 ± 38.8 ml/min per 1.73 m 2 ) and serum albumin (30.4 ± 7.2 g/L). Renal biopsy showed seventy-ve percent of DN cases were in stage three. The average percentage of glomerulosclerosis was 17.5 ± 14.8%, and the average percentage of podocyte effacement was 76.6 ± 15.4%. IFTA score was 2.0 ± 0.8. All cases had minor to moderate linear IgG expression along the GBM, and IgG colocalized with collagen IV α5 staining (Fig. 1A). Thirty-one cases showed minor linear IgG along TBM (Fig. 1A). All cases showed negative or trace complements (C3, C4, C1q) along GBM or TBM (Fig. 1B).

Association between IgG subclass location and clinicopathological features
IgG subclasses were stained with the similar intensity of IgG, and the glomerular positive expression was also colocalized with collagen IV α5 expression (Fig. 1C). Positive IgG but negative or trace expression of IgG subclass (the None group) was observed in 11 cases. Thirteen cases showed IgG subclass expression along GBM only (GBM-only group), while ve cases were positive along TBM only (TBM-only group). The remaining 11 cases had positive expression on both GBM and TBM (the Both group) ( Table 1). A-sclerosis score 0.6 ± 0.5 0.5 ± 0. A, vs the "Both" group, P < 0.05; b, vs the "None" group, P < 0.05; c, vs the GBM-only group, P < 0.05; Among these four groups, higher proteinuria was detected in the Both group than the other groups, thereby indicating more glomerular injury. The TBM-only group showed less proteinuria and KIM-1 expression than other groups, thereby indicating less tubular injury. However, its arteriosclerosis score, a vascular injury marker, was higher than the None and the GBM-only groups (Table 1).

Association between GBM IgG subclass deposition and clinicopathological features
Twenty-four of the forty cases showed different IgG subclass expression on GBM, including thirteen cases from the GBM-only group and eleven cases from the Both group (Table 1). To determine whether glomerular IgG subclass was correlated with glomerular and tubulointerstitial injury, all forty cases were divided into six groups according to the positive status of IgG subclass expression on glomeruli: sixteen cases with trace to minor staining of IgG subclass (the None group); eight cases with IgG1-or IgG2positive status, respectively (IgG1-only or IgG2-only group); one case with IgG3-positive (IgG3-only group) status; three cases with IgG4-positive (IgG4-only group) status, and four cases with ≥ 2 IgG subclass expression (the Mixed group) ( Table 2 and Fig. 2A).

Association between TBM IgG subclass deposition and clinicopathological features
Sixteen of the forty cases showed different IgG subclass expression on TBM, including ve cases from the TBM-only group and eleven cases from the Both group (Table 1). To determine whether tubular IgG subclass was correlated with tubulointerstitial injury, all forty cases were divided into three groups according to the positive status of IgG subclass on tubules: twenty-four cases with trace to minor staining of IgG subclass (the None group); ten cases with one IgG subclass expression (the Isotype group); six cases with ≥ 2 IgG subclass expression (the Multiple types group) ( Table 3). IgG1 and IgG3 were major positive subclasses along TBM in the Isotype group (n = 4, respectively, Fig. 3A). Among these three groups, interstitial brosis, TBM thickness, and arterial injury score were similar. The multiple types group, in which three of the six cases were IgG subclass deposited along TBM only, showed less tubular KIM-1 density and interstitial in ammation than the None or Isotype groups (Fig. 3B).

Discussion
In our study, 67.2% of the DN cases showed linear IgG staining by IF, similar to that observed in previous studies [2]. Unlike other immune complex-mediated glomerular nephropathy, such as membranous nephropathy, DN with positive IgG had negative or trace complements along with GBM or TBM, and electron microscopy showed negative dense deposits along GBM/TBM. Eluates from DN kidneys do not contain anti-GBM antibody [15]. Furthermore, not only IgG but also albumin showed linear expression along GBM in DN cases, suggesting this kind of IgG deposition might occur more frequently as a consequence of renal injury [3,4,16].
Not all DN cases had positive expression of IgG subclass, and they showed different positive statuses among DN cases. IgG subclass is a very useful diagnostic tool for several renal diseases, including MN, heavy and light chain deposition disease, proliferative glomerulonephritis with polyclonal IgG deposition, and so on [7]. In our study, we found that IgG1 and IgG2 were major subclasses of GBM IgG expression in DN, which was consistent with another recent paper; however, researchers found no relation between the IgG subclass distribution along the GBM and clinicopathological data [17]. We also found that IgG1 was expressed along both GBM and TBM, while IgG2 was more restrictive along GBM, and IgG3 was more restrictive along TBM. The different locations of IgG subclass deposition are determined by both immunoglobulin pro les and the local environment. While the molecular size is similar among four subclasses of IgG, the cationic charge and serum concentration decreased in the order of IgG1, IgG2, IgG3, and IgG4 [6,9,18]. Owing to the anionic charge feature of GBM, the a nity of IgG subclass to GBM should be higher for IgG1 and IgG2 than IgG4 [19,20]. Podocyte injury during the early stage of DN recognized the GBM matrix and induced GBM thickness [21]. The glomerular clearance of IgG4 increased, while IgG1 clearance decreased reciprocally in diabetes compared to that in the control [22]. Whether these changes of foamy GBM matrix contribute to all the different IgG subclass attractions remains unclear. For the TBM deposit of IgG, two possible mechanisms may be proposed. First, the glomeruliltered IgG is reabsorbed by tubular epithelial cells (TECs), and nally deposited along the TBM. TEC induces endocytosis of IgG via megalin/cubulin complex, and induces transcytosis via neonatal Fc receptor (FcRn) for IgG [23]. In DN, megalin/cubulin expression is increased at early stage and decreased at late stage [24,25]. One study showed that tubular reabsorption of IgG1 was moderately reduced and that of IgG4 remained at the same level in diabetes compared to that in controls [22]. TBM thickness is also detected in DN. Whether this kind of change contributes to the IgG deposit along TBM is unknown. We found that IgG1 and IgG3 were major subclasses for TBM deposition, potentially arising due to tubular selective reabsorption. Second, the interstitial IgG that leaks from peritubular capillary is trapped by TBM. We found that the group with IgG subclass deposition along TBM-only had a higher level of arteriosclerosis, suggesting this possibility. However, in contrast, no megalin/cubulin expression on the TBM side suggests a lower possibility of this mechanism.
Although IgG deposit is considered as a consequence rather than the cause of DN injury, several studies have indicated that IgG positivity is associated with poorer renal prognosis in DN [2,17]. In our study, cases with both GBM and TBM IgG subclass deposition showed more proteinuria than those with TBMonly deposition, suggesting severe glomerular injury. However, the TBM-only group showed the lowest serum albumin levels, but not the highest proteinuria levels, suggesting well preserved tubular reabsorption, which was veri ed by less tubular KIM-1 expression. IgG subclasses are also critical for diagnosis and may correlate with progression, such as in MN [7]. In our study, IgG1 in glomeruli was associated with CD34 loss, thus indicating increased glomerular epithelial cell injury, while IgG2 in glomeruli was associated with GBM thickness, which might be due to IgG2 being the most stable antibody in terms of structure and long-term retention [26]. However, GBM thickening in DN is thought to be the result of podocyte injury, which disrupts the balance of GBM matrix synthesis and degradation [21]. Thickened GBM accompanied by glomerular endothelial cell injury prefers to trap IgG1 and IgG2, which is supported by complete overlap staining of IgG and its subclasses with collagen IV α5 [27]. Furthermore, cases with multiple IgG subclass deposition along TBM had less tubular injury, interstitial in ammation, as well as similar TBM thickness and IFTA score than those with isotype IgG subclass deposition, potentially indicating well preserved and less selective reabsorption by tubular epithelial cells in early-stage DN [22].
The limitations of this study should be noted. First, this is a descriptive study with an interesting nding, but mechanistic studies need to be performed in the future. Second, the sample size of the IgG3 positive group was limited, that resulted in di culty of analysis of the clinicopathological features of this group, and which negatively impacted the comparison with other groups. Lastly, serum level and urine secretion of IgG subclass could not be analyzed due to the retrospective nature of the present study, and further prospective research is warranted.
In summary, IgG1 and IgG2 were major IgG subclasses deposited along GBM, while IgG1 and IgG3 were major IgG subclasses deposited along TBM in DN, which were determined by their pro les and severity of glomerular/tubular injury. IgG and its subclass deposition are not causal, but the consequence of renal injury and the positive statuses are associated with different DN injuries.