C4d in Native Glomerular Diseases

C4d is a degradation product of complement cascade and an evidence of complement activation. Complement system can be activated via 3 pathways – classical pathway (CP), lectin pathway (LP), and alternate pathway (AP). All 3 pathways converge in the formation of the C3 convertase which activates C3, producing active forms C3a and C3b, the latter combines with C3 convertase, forming C5 convertase which cleaves C5 into active forms C5a and C5b. C5b can further combine with C6, C7, C8, and C9 to form C5b-9, the membrane attack complex (MAC). MAC causes lysis in non-nucleated cells, but causes sub-lethal injury to nucleated cells including cell activation and tissue injury. CP is activated by the formation of antigen antibody complexes which ultimately results in the activation of C1s which cleaves C2 and C4 into active forms. LP is activated by the binding of mannose-binding lectins (MBL) or ficolins to carbohydrate moieties in bacterial organisms, which results in the activation of MBL-associated serine proteases (MASPs) which cleave C4 and C2. Both the CP and LP result in the formation of C4b2b which acts as the C3 convertase for CP/LP. In AP, there is spontaneous activation of C3 by hydrolysis of the thioester bond, as a result of which, factor B and factor D are recruited. Factor D cleaves factor B to form an active serine esterase (Bb) which then converts C3 to C3a and C3b. C3b associates with Bb to form the AP C3 convertase C3bBb (Figure 1).

The role of complement activation in various glomerular diseases, specifically the activation of AP, has been recently reviewed [1, 2]. C4d is a split product of C4 activation in CP and LP. Because of its thioester bond, it can covalently bind to cell surfaces and can serve as a marker of complement activation [3].

In renal transplantation, the presence of C4d in peritubular capillaries (PTC) along with the presence of allograft damage/renal injury and donor-specific antibody is diagnostic of acute antibody mediated rejection (AMR) and has been associated with worse prognosis [4-6]. On the other hand, the presence of PTC C4d staining in ABO incompatible allografts without histological damage is thought to represent graft accommodation [3].

C4d staining can be done by 2 methods - by indirect immunofluorescence (IF) on frozen tissue using monoclonal antibody to C4d or by immunohistochemistry (IHC) using polyvalent C4d antibody on formalin fixed paraffin embedded tissue. In renal transplantation, C4d staining by IHC and IF are found to be comparable though IF appears to be more sensitive [7, 8]. In native kidneys, C4d deposition in the mesangium and along the glomerular capillaries and in some renal arteriolar walls has been observed in frozen tissue and thought to represent transient glomerular deposition and clearance of immune complexes (IC) as part of natural antibody response [9]. Similar positive glomerular staining for C4d via IF on frozen tissue that is not noted on C4d staining via IHC in fixed tissue from native kidneys is noted in other studies [8, 10, 11].

This review focuses on the available data on glomerular C4d deposition in renal biopsies in various native glomerular kidney diseases (Table 1). Each section provides a brief background summary of complement activation in that glomerular disease followed by data on C4d in that glomerular disease.

IgA nephropathy has been noted to be associated with the activation of the AP and LP. [12, 13]. In a study of 60 patients with IgA nephropathy, while most patients (approximately 75%) were thought to have C3 deposition from alternate complement activation, about 25% (15 patients) had mesangial deposition of MBL - this was co-deposited with ficolin, MASP and C4d indicating activation of the LP of complement activation [14]. C1q was absent. Glomerular MBL deposition was associated with more histological damage (more mesangial proliferation, extracapillary proliferation, glomerular sclerosis, and interstitial fibrosis) and worse renal failure at the time of biopsy as well as more proteinuria compared with MBL negative cases.

In a retrospective study of IgA nephropathy patients, C4d (IHC) mesangial deposits were present in ∼68% of biopsies. Patients with C4d positivity were older, had greater percentage of HTN, lower percentage of macroscopic hematuria, more glomerulosclerosis and interstitial fibrosis. Patients with C4d positivity had a lower renal survival compared with C4d negative patients (44% compared with 91%) [15].

Glomerular C4d staining (noted via IHC in mesangium and capillaries) associated with albuminuria and tubular epithelium C4d staining correlated with higher WHO grade of IgA nephropathy (but not with albuminuria) in a study of 23 patients with IgA nephropathy [16].

In 283 patients with biopsy proven primary IgA nephropathy, Espinosa et al. [17] found 38.5% of biopsies to have glomerular C4d staining (mesangial) via IHC (also confirmed by IF in some patients). Glomerular C4d staining was associated with a higher proportion of hypertension, lower eGFR, higher proteinuria, greater proportion of patients with segmental sclerosis/interstitial fibrosis, and worse renal survival. Renal survival at 20 years was 28% in C4d-positive patients and 85% in C4d-negative patients. C4d deposition was an independent risk for the development of ESRD [17]. In 37 patients with IgA nephropathy, C4d (via IHC) was found in 16 patients in glomerulus (mesangial), a few patients also had C4d in the tubules and interlobular arteries [18]. Glomerular C4d was associated with endocapillary hyper cellularity, interstitial fibrosis/tubular atrophy and mesangial IgM deposition, with lower eGFR, worse proteinuria, and higher rate of hypertension at diagnosis. In 29 patients with IgA nephropathy, C4d staining (IHC) in mesangium and capillary wall was associated with global glomerular sclerosis/interstitial fibrosis as well as proteinuria and creatinine [19]. In 33 patients with IgA nephropathy, positive C4d mesangial staining (IHC) was associated with worse glomerulosclerosis, mesangial hyper cellularity, interstitial fibrosis as well as increased hypertension, proteinuria, and worse renal survival compared to C4d-negative patients [20].

In a study of 15 patients with IgA nephropathy examining the association of C4d deposition (via IHC) with histological features based on the Oxford classification, C4 deposition was noted mainly in the mesangium, as well as the glomerular capillary wall and in some cases in PTC, arterioles, and tubular epithelium. Segmental and global glomerular capillary deposition of C4d correlated best with endocapillary proliferation [21]. In a study of 74 patients with IgA nephropathy, several immunohistological markers were examined including C4d (via IHC) and CD3, a T cell marker. Glomerular C4d was positive in 34% patients, was associated with higher proteinuria, progressive kidney disease, with presence of tubulointerstitial fibrosis and immunoglobulin G (IgG). In addition, in multivariate analysis only baseline eGFR, C4d and CD3 were independently associated with a 50% decline in eGFR or progression to ESRD [22]. In a recent study of 96 patients with IgA nephropathy, the authors found mesangial deposition of C4d (via IHC) and MBL to be an independent predictor of urinary C4d and MBL levels. They found urinary levels of C4d and MBL to be sensitive and specific markers of mesangial C4d and MBL deposition [23]. Similar results in the pediatric population with mesangial C4d deposition (IHC) was noted as an independent predictor of decline in eGFR; renal survival was 8.6 years in C4d-positive patients versus 15.1 years in C4d-negative patients [24]. Most recently, Segarra et al. [25] studied 190 patients with IgA nephropathy, but with normal renal function (eGFR > 80 mL/min). Mesangial C4d deposition (via IHC) was noted in 20% patients and was associated with more proteinuria compared with C4d-negative patients at the time of diagnosis. On follow up, C4d-positive patients had a higher number of nephritic flares, received more immunosuppressive /immunomodulatory treatments, and had higher protein creatinine ratios compared with C4d-negative patients. In multivariate analysis, C4d positivity was identified as an independent predictor of eGFR slope and ESRD [25].

In summary, glomerular (mostly mesangial but in some studies also capillary wall and tubular epithelium) C4d deposition in IgA nephropathy is thought to reflect activation of the LP, and several studies have found that this is associated with worse histopathological features, proteinuria, and/or renal outcomes. Polymeric IgA is capable of binding to MBL and activating the LP [26] and potentially causing greater immune damage with complement activation but further studies on the mechanisms are needed.

Membranous nephropathy (MN) is an immune-mediated disease with animal models (Heymann nephritis) pointing to the activation of the CP by binding of antibody to antigen present on the podocyte, resulting in the deposition of the C5b-9 complex, the MAC and damage to the podocyte and detection of C5b-9 in urine in mouse models [27]. Similarly, in humans urinary C5b-9 excretion was higher in MN compared with other glomerular diseases including focal sclerosis [28]. Other authors suggested that this increased urinary C5b-9 was not limited to MN, but was also present in focal segmental glomerulosclerosis (FSGS) and was related to proteinuria; they proposed that the elevation could be the result of complement activation via AP in renal tubular epithelial cells triggered by the proteinuria itself [29, 30]. Based on animal models, it is proposed that complement-mediated cytotoxicity to the podocyte results in sublethal injury that alters the function of the podocyte, results in dedifferentiation, change in cytoskeleton, foot process simplification, and proteinuria. However, while the CP is activated in mouse models, the exact complement activation in humans is unclear. The podocyte antigen in human primary MN has been recently discovered to be phospholipase A2 receptor (PLA2R) on the podocyte and is present in about 60–80% of patients, other podocyte antigens include thrombospondin. The IgG subtype in primary MN is IgG4, which is unable to fix complement via CP activation [31-33]. Hence, other ways of complement activation including AP or LP have been suggested. Deposition of MBL, a component of the LP, along the glomerular basement membrane has also been noted in 10 of 15 patients with MN by Lhotta et al. [34]. In a study of 12 patients with idiopathic MN, C4d deposition by IF in the glomeruli was found in 92% of the cases in close association with IgG deposits suggesting activation of the CP [35]. C4d deposition (via IHC) was present in a granular basement membrane distribution in 100% of biopsies in 31 cases of idiopathic MN and 5 cases of class 5 membranous lupus [36]. In a study comparing 21 membranous patients with 19 patients with minimal change disease (MCD), capillary C4d deposition (via IHC) was present in 100% of membranous patients and in none of the MCD patients [37]. Bally et al. [38] report a case of IMN in a patient with deficiency of MBL in whom complement activation was thought to occur via AP activation; they further evaluated 77 patients with idiopathic MN, and found similar MBL deficiency in 4 patients in whom there was weak C4d staining. However, all remaining patients with intact MBL had intense C4d staining (via IF) along glomerular capillary wall indicating activation via LP. In a recent study, lower levels of markers of LP and AP pathway activation were associated with greater chance of attaining complete remission [39]. In a Japanese cohort of 60 primary MN patients, 40 were positive for either PLA2R or thrombospondin (the intrinsic antigen related group) and 20 had neither PLA2R or thrombospondin (intrinsic antigen unrelated group – it is unclear how the diagnosis of primary MN was made in this group). C4d (via IF) was present along the capillary wall in all but 1 patient (59/60). MBL was present in 43% of patients suggesting LP activation, but C1q was also noted in about 46% of patients (in both intrinsic antigen related and unrelated), indicating CP activation. Patients with intrinsic antigen related group had more intense staining for MBL, which correlated with IgG4 staining and was associated with less chance of proteinuria remission and with renal dysfunction [40]. In a study of 69 patients with primary MN, C4d (via IHC) deposition with C1q was noted in 12 patients (indicating CP pathway activation), C4d without C1q deposition noted in 46 patients (indicating LP activation), and no deposition of C4d or C1q in 11 patients suggesting activation via AP pathway [41]. C4d deposition was associated with increased interstitial fibrosis in this study.

To summarize, C4d appears to be deposited in most patients with MN and this may help distinguish MN from other pathologies such as MCD. Several studies point to LP activation as the source of C4d deposition though some studies also suggest activation of the CP with resultant C4d deposition. Some data suggest worse outcome/renal pathology with MBL/C4d deposition respectively. There is evidence that MBL can bind to the Fc portion of IgG, which lacks galactose residues (such as in rheumatoid arthritis) and thereby activate the LP [42]. Recently polymorphisms in the MBL2 gene in idiopathic MN and lupus MN were suggested to be associated with LP activation in MN [43].

In FSGS, IgM and C3 deposition is frequently found in the sclerotic areas and sometimes in the mesangium in the non-sclerotic areas and have been thought to be nonspecifically trapped in areas of glomerular injury. However, animal model data suggest that complement activation is pathogenic in FSGS. In several proteinuric kidney diseases, intratubular complement activation and generation of C5b-9 is thought to mediate tubular cell injury and progressive tubulointerstitial damage [44]. In rat model of FSGS (Adriamycin induced), there was an increase in C5b-9 in proximal tubular cells and peritubular region along with increase in peritubular myofibroblast accumulation in mice with intact complement system, but renal C5b-9 deposition was absent and the peritubular myofibroblast accumulation decreased in C6 deficient mice [45]. This suggested that C5b-9 promotes peritubular myofibroblast accumulation in FSGS. In another animal model, development of FSGS was accompanied with glomerular C3 deposition, in addition complement activation was noted to occur via alternate and not classical complement pathway, mice deficient in either C3 or alternate complement factor (factor D) or MAC regulator (CD59) were protected from injury, and sustained less proteinuria and glomerulosclerosis/tubulointerstitial injury [46]. Similar role for AP activation was suggested in another rodent model of FSGS [47]. In another study, depletion of B cells by various methods resulted in reduced deposition of IgM in the glomerulus; this was accompanied with reduced C3 and C4 (including C4d) deposition in the glomerulus as well as reduced glomerulosclerosis and albuminuria [48]. The authors propose that IgM activates the complement system via CP in animal model of glomerulosclerosis, possibly by binding to neoepitopes exposed in the glomerulus after non-immune insults.

Studies in humans thus far show inverse correlation of plasma and urine levels (factor B fragment) with the eGFR at the time of diagnosis and end of study, as well as positive correlation of plasma and urine Ba with proteinuria in 19 patients in the FSGS clinical trial [49]; higher percentage of sclerotic lesions in patients with C3 deposition as well as IgM and C3 as an independent risk factor for refractoriness to treatment and renal dysfunction [50]; worse renal outcomes with deposition of IgM and C3 in non-sclerotic area in secondary FSGS (but not in primary) [51]; increased risk for ESRD with low serum C3 in 221 patients with FSGS, as well as correlation of moderate to severe tubulointerstitial injury with low serum C3, low AP activity, and high serum MAC [52].

In summary, the above data suggest that complement activation occurs in FSGS and it contributes to its pathogenicity. It is suggested that both CP and AP may be involved. However, there are no studies thus far specifically looking at C4d deposition or LP activation in FSGS.

MCD is generally not thought to involve the complement system, and renal biopsies with MCD are usually used as controls in studies evaluating complement pathways.

Membranoproliferative glomerulonephritis (MPGN) has historically been divided into type I, II, and III based on findings of deposits on EM – sub endothelial (type I), sub epithelial and sub endothelial (type III) and dense deposits in glomerular basement membrane (Type II or dense deposit disease DDD). This classification has been replaced by IF based classification into Immune Complex (IC) IC-mediated and complement-mediated MPGN. IC-mediated includes MPGN caused by infections, autoimmune diseases and dysproteinemias, and involves activation of CP by ICs. Complement-mediated MPGN, also called C3 glomerulopathy (C3G), includes C3 glomerulonephritis and DDD, both of which involve activation of AP and deposition of C3 with little or no deposition of immunoglobulin [53, 54].

In a study of patients with proliferative glomerulonephritis (GN), Sethi et al. [55] evaluated C4d staining (via IF) in biopsy samples. Of the 18 patients with IC GN including IgA/lupus/membranous and (IC mediated) MPGN, C4d staining was positive in all except the 2 specimens with IgA nephropathy and 1 patient with sclerosing MPGN. C4d staining was similar to the distribution of IgG and C3 in mesangium/capillary loop. C1q was positive in 5 biopsies with IC GN. In IC GN, C4d was a marker of activation of CP or LP activation. Of 30 biopsies with C3G, C4d staining was completely negative in 80% of patients with C3G and trace to 1+ in 20%. Of 13 biopsies with post infectious GN, C4d was negative in 6 (46%) suggesting possible role of AP in these patients. C1q was positive in 2 biopsies with post infectious GN. The sensitivity and specificity for negative or trace C4d for diagnosis of C3G (compared with IC-mediated GN) was 93 and 100 vs. 97% and 63% for negative C1q and 87 and 100% for negative IgG respectively [55]. The authors propose C4d as a valuable marker to distinguish IC GN from C3G.

In a more recent study of patients with MPGN, the authors’ reclassified MPGN subtypes based on C3G consensus and subsequently reclassified based on C4d staining (via IHC). IC mediated MPGN had more intense C4d glomerular capillary wall staining compared with C3G. Absence of C1q but C4d positivity and positive C3 was an indicator of LP involvement and was found in both MPGN and C3GN [56].

Complement activation in lupus nephritis (LN) is thought to be via the CP with resultant deposition of C1q in addition to IgG and C3, C4. However, in a study looking at the deposition of MBL of the LP, MBL was deposited in most patients with LN in a pattern similar to the deposition of IgG [34]. Similarly another study found that in addition to CP activation, activation of AP and LP occurs in some patients with LN and is associated with more proteinuria [57]. In a study of 11 patients with LN, C1q along with IgG and C3 deposition was present in all biopsies, indicating CP activation; 9 (82%) of the biopsies also had a deposition of MBL suggestive of concomitant LP activation, these patients with MBL deposition had greater proteinuria compared with MBL-negative patients [58]. In a study of LN patients, all patients had C4d deposition (via IF) predominantly along the glomerular capillary loops. The intensity of C4d deposition significantly correlated with capillary deposition of IgG, IgA, C4, and C1q but not with the LN activity index. Hence, presence of C4d was noted to be a sensitive marker for the activation of CP, but not for disease activity [59]. Similar findings were noted in another study of 15 LN patients in which glomerular C4d (by IHC) along capillary wall and to a lesser extent in mesangium was found in 46% of biopsies and was associated with the presence of IgG, C3, and C1q but not with disease activity index [60]. However, in another study, glomerular C4d staining (via IHC) was found in 58% (14 of 24 patients) of patients with LN and correlated with higher activity index score and proteinuria [61]. This study did not provide data on C1q staining.

In a study investigating the presence of peritubular (PTC) capillary C4d staining in LN compared with patients with renal allograft rejection, C4d staining by IF was found in 6.8% (31 of 455) of patients with LN [62]. Peritubular C4 d staining in LN was granular, accompanied by deposition of IgG and C1q or C3, not associated with the presence of neutrophils or macrophages. This contrasted with kidney allografts with acute AMR in which C4d staining in PTC was more linear, not associated with IgG C3 or C1q deposition in PTC and associated with infiltration of neutrophils and macrophages in PTC. In addition, electron microscopy showed electron dense deposits along PTC basement membranes in LN but not in acute AMR. LN patients with positive PTC C4d staining had higher activity index score and lower chronicity score. LN patients with positive PTC C4d staining were younger, had more severe proteinuria and hematuria, higher percentage of positive serologies, lower C3, C4 levels and higher SLEDAI (Systemic Lupus Erythematosus Disease Activity Index) scores compared with C4d-negative LN patients. No association was noted between C4d positivity and thrombotic microangiopathy (TMA). The study did not comment on glomerular C4d staining. Controls included healthy kidneys and patients with IgA and MN and none were positive for C4d staining in PTC.

In a study of LN, Cohen et al. [63] found that C4d staining (via IHC) in the glomeruli was strongly associated with the presence of microthrombi and intense C4d staining was present in 7 of 8 biopsy samples with microthrombi. Neither C4d staining or presence of microthrombi was associated with the presence of antiphospholipid antibody status. C4d deposition was significantly correlated with the presence of capillary C3 and also correlated with capillary IgG. The authors suggest that activation of the classical complement pathway plays a pathogenic role in TMA in LN [63].

In a larger prospective study of 155 patients with LN, intensity of glomerular C4d staining (via IHC) was significantly related to glomerular microthrombi as well as the presence of IgG anti-beta2 glycoprotein (anti-B2GP1) antibodies. In addition, intensity of glomerular C4d staining was associated with lower serum C3, higher LN activity index, and greater proteinuria [64]. However, in another study using CD61+ platelets as marker for microthrombi and CD68 as marker for macrophage infiltration, it was found that microthrombi associate with macrophage infiltration but not with C4d (via IHC) [65].

Kim et al. [66] reviewed complement including C4d staining in 20 patients with LN. Glomerular C4d was observed in 13 patients mostly along capillary wall, some mesangium. C4d deposition with IgG, C3 and C1q suggested CP activation (seen in 7 of 20 patients) and C4d and C3 deposition in the absence of C1q indicated LP activation. Class 5/membranous LN was associated with a strong granular staining along capillary wall, this could be used to identify patients with membranous LN if tissue is not available for IF. C4d deposition did not correlate with either the SLE disease activity or the histological features [66].

Hence, in LN C4d deposition is found in the glomerulus in varying intensities but especially strong in class 5 LN, and some patients can have C4d deposition in the PTC. C4d along with C1q indicates activation of CP and C4 deposition in the absence of C1q suggests activation of LP. C4d staining is associated with worse histological and clinical findings in some but not all studies. In addition, some studies have found C4d to be associated with microthrombi in LN; however, this has not been noted in other studies.

Multiple studies in animal models have shown that AP is activated in anti-neutrophilic cytoplasmic antibody (ANCA) mediated GN and results in the formation of C5a which primes neutrophils for the activation by ANCA and also is a potent chemoattractant for neutrophils [67]. Xiao et al. [68] have shown that transfer of anti-Myeloperoxidase (MPO) IgG into wild-type mice or anti-MPO splenocytes into immunodeficient mice caused crescentic GN that could be blocked by complement depletion. In addition, C5 and factor B knockout mice were protected from anti-MPO IgG mediated GN, but C4 knockout mice developed necrotizing GN similar to wild-type mice indicating that complement is activated via the AP and not via the CP or LP [68]. In a small study of 7 patients with MPO-ANCA positive GN, there was positive glomerular and small blood vessel staining for C3d, factor B, P (both factors B and properdin indicating AP activation) and MAC but there was no staining for MBL or C4d [69]. The same group reviewed renal biopsies from 12 patients with ANCA negative pauci-immune GN and noted that 8 of 12 were C4d positive and 6 of these 8 were MBL positive indicating activation of LP with resultant deposition of C4d. Intensity of C4d and MBL staining was much higher in those on dialysis versus in those not on dialysis [70]. In a study of 85 patients with pauci-immune GN, glomerular (capillary wall and mesangial) C3d and C4d staining (via IHC) was observed in ∼49 and 45% of the biopsies, respectively. While positive glomerular C3d staining was noted to be an independent predictor for renal survival, there was no association with C4d positivity with renal outcome [71]. C1q staining was positive in only 1 patient in this study. In a recent study in patients with ANCA-associated GN, glomerular C3d deposition was found in 58% of biopsies (43 of 74 biopsies), glomerular C4d in mesangiocapillary pattern (via IF) was positive in ∼71%, properdin in ∼39%, and MBL in 30% of biopsies. C1q was positive only in 5.3% of biopsies. Renal biopsies positive for C3d or properdin showed more cellular crescents (versus C3d or properdin negative biopsies) and C3d-positive biopsies had more interstitial fibrosis/tubular atrophy, C3d-positive patients had more proteinuria and worse renal function at baseline and on follow up compared with C3d-negative patients. No such association with C4d positivity was seen [72].

Hence, in ANCA-associated vasculitis, while there is predominantly activation of the AP, there is some evidence of activation of LP or possibly CP resulting in C4d deposition.

Complement activation contributes to glomerular injury in animal models of anti-GBM induced renal disease [73]. In an animal model of anti-GBM disease, Otten et al. [74] suggested the activation of CP and AP but not the LP. In a recent study of 10 patients with anti-GBM crescentic GN, both classical and alternate complement pathways were thought to be activated [75]. In addition to linear IgG deposition, there was deposition of C1q, C3d, C5b-9, factor B, P and C4d along the glomerular capillary wall. While there was deposition of MBL, it was noted to be diffuse; in addition to glomerular capillary wall, it was deposited in the mesangium, Bowman’s capsule, and within the crescents. In addition, MBL did not co-localize with C5b-9 and only partially co-localized with C4d. The authors concluded that LP was not likely involved in anti-GBM GN and that MBL deposition was due to MBL binding to carbohydrate ligands exposed during glomerular damage.

Complement activation plays a key role in atypical hemolytic uremic syndrome (aHUS) with 60% of patients having mutations in complement proteins, many of which involve the AP [76, 77].

Recent data suggest that complement activation may be involved in other TMA settings such as shiga toxin-producing Escherichia coli (STEC-HUS), post hematopoietic stem cell transplant (HSCT), pregnancy-related HUS or drug-related HUS [78, 79].

As stated above [63], glomerular C4d deposition was found to be associated with TMA in patients with LN and was thought to be deposited as a result of CP activation.

Mii et al. [80] describe TMA in post HSCT, likely thought to be associated with diffuse endothelial injury from chronic graft versus host disease (GVHD) which the authors term as “chronic humoral GVHD”. C4d (via IF) was diffusely deposited in the glomerular capillaries and had patchy distribution in PTC and tubular BM consistent with complement activation in this setting [80]. In a subsequent small series of 7 patients with TMA post HSCT, the same authors describe findings of acute GVHD in 7 patients, 4 of who also had chronic GVHD. There was diffuse endothelial injury in glomeruli, PTC, and small arteries. C4d staining (via IF in biopsy patients and via IHC in autopsy cases) diffuse in glomeruli and patchy in PTCs and tubular BM was noted in 4 of 7 of these patients, suggesting antibody-mediated renal endothelial injury [81]. C1q staining was negative.

In a study of tissue samples in children with HSCT, C4d deposition in patients who had transplantation associated TMA (TA-TMA) was compared with patients with HSCT who either had normal kidneys or had another non-TMA related kidney pathology [82]. Deposition of arteriolar C4d (via IHC) was significantly more common in patients with TA-TMA compared with controls (75 vs. 8%). Deposition of C4d in PTC or tubular basement membrane was rare but found only in TA-TMA not in controls. Glomerular C4d deposition was common and similar in both TA-TMA and in control patients and was thought to be nonspecific. C1q staining data was not available. Arteriolar C4d deposition was thought to be a marker of TA-TMA.

In a retrospective study of 42 renal pathology specimens from 36 patients (28 patients with native kidney disease and 8 patients with renal transplants) with histologically confirmed TMA, C4d deposition (via IHC) was present in ∼88% of samples [83]. Terminal complement complex C5–9 was present in 78.6% samples, C1q was present in ∼92% samples, and MBL in 28.6% samples. Most of the C4d/C5–9 depositions were in the glomeruli and the arterioles and small/none, respectively, in PTC. Etiologies of TMA included SLE-related TMA, hematopoietic SCT-related TMA, and atypical HUS among others. CP activation was defined as the presence of C1q and absence of MBL or presence of IgM and C4d. LP activation was defined as the presence of MBL and absence of C1q. CP activation was noted either alone (61.9%) or in combination with LP (28.6%) and was therefore activated in 90.5% of TMA cases. The authors conclude that C4d deposition represented CP activation driven by antibody-mediated endothelial cell injury, for example, in setting of SLE or after HSCT. But in other situations, such as aHUS, complement activation with resultant C4d deposition may result from or in addition to endothelial damage (and thereby trapping of IgM and subsequently classical complement pathway activation) or by ischemic endothelial injury as does occur in ischemia reperfusion injury.

In summary, many of the conditions that cause TMA are associated with the deposition of C4d likely from CP activation in most cases and from LP activation in some cases.

While post infectious (post streptococcal) GN and infection associated GN are IC-mediated diseases with activation of CP, in post infectious GN activation of AP is thought to play a role as well [84, 85].

In post infectious GN (post streptococcal GN), activation of LP has been noted as well with staining of MBL and C4d (IHC) in the glomerulus (mesangium) in the absence of C1q and immunoglobulins [86]. In a study of 18 patients with post streptococcal GN, all patients had evidence of AP activation but 7 patients also had activation of LP (with positive staining for C4, C4 binding protein, MBL, and MASP) [87]. In the study by Sethi et al. [55], about 7 of 13 patients listed as post infectious GN had C4d deposition but also mild staining for IgG, suggesting activation of CP or LP since C1q was positive in only 2 biopsies. However, some patients included in the post infectious group likely had infection-related GN and not post infectious GN.

In infection-related GN, activation of LP has also been noted. In a study of 16 patients, mixed cryoglobulinemia from hepatitis C infection, MBL, and MASP could be isolated in the cryoprecipitate along with C1q, IgG, IgM, and C4d. In 3 patients who had histological studies done, there was deposition of MBL, MASP, and C4d in the glomerulus while the deposition of C1q was variable [88]. In the study by Sethi et al. [55], in the 1 patient with IC GN who had MPGN from HCV, there was deposition of IgG with C4d but without C1q deposition.

In fibrillary GN, IgG is usually present along with C3, suggesting CP activation, in addition C1q is present in lesser number of patients [89, 90]. In a study by Sethi et al. [55], of the 2 patients with fibrillary GN, both had C4d deposition along with the presence of IgG and C3 in the biopsy.

C4d staining has diagnostic utility in unmasking hidden immune deposits, as shown by Sethi et al. [91]. They describe a patient with membranous-like nephropathy and another with MPGN in whom routine IF was negative for Immunoglobulins and showed some C3, but had bright C4d staining (via IF) along capillary walls. Repeat IF following pronase digestion of formalin fixed paraffin embedded tissue revealed IgG kappa.

Recently a new entity C4G has been described in a small series of patients in whom there is bright C4d deposition with minimal or no deposition of C3 and no deposition of C1q or immunoglobulin [92]. These patients had a mesangial or membranoproliferative pattern of injury, with electron microscopy showing mostly mesangial or subendothelial deposits that corresponded to C4d. Laser microdissection and mass spectrometry confirmed deposits to be C4. Depending on the EM findings, patients were described to have either C4 DDD or C4G. Given bright C4d deposits and absence of C1q, immunoglobulin and C3, this deposition likely represented the activation of LP. Additional case reports have been published including case with C4 DDD occurring in the setting of TMA and monoclonal gammopathy [93] and in setting of preeclampsia at 18 weeks along with renal biopsy findings of TMA and C4G [94].

C4d deposition in the glomerulus can result from either activation of CP or LP of complement. Deposition of C4d along with C1q and immunoglobulin would suggest CP activation. Deposition of C4d with or without immunoglobulin and the absence of C1q would indicate activation of LP. Immunoglobulins such as IgA as well as galactose-deficient IgG are thought to activate the LP as noted above. In addition, some data points to IgM capable of LP activation in an ischemia reperfusion injury model in the heart [95]. C4d deposition in different glomerular diseases appears to be associated with worse renal prognosis in some but not all studies. Use of C4d staining along with other components of complement deposition will help delineate different complement activation pathways in glomerular diseases and help understand disease pathogenesis. In addition, in some diseases, C4d staining may help identify a subset of patients with worse renal prognosis which may help guide treatment decisions. As newer drugs affecting complement pathways become available, identifying key complementary pathways in different glomerular diseases will be crucial. C4d staining by IHC is a simple stain that can be done in most labs and is valuable when tissue is not available for IF. Further studies in animal models to understand the pathogenesis and prospective studies in humans, specifically studying the deposition and activation of different complement pathways and their role in different glomerular diseases are needed.