It was previously assumed that the complement system was not involved in the pathogenesis of AAV, because little immunoglobulin or complement deposition is found in AAV, and AAV is generally not associated with hypocomplementemia [19
]. However, increasing evidence suggests that activation of the complement system, via the alternative pathway, is crucial for the development of AAV [2
]. The complement system comprises more than 30 plasma and membrane-bound proteins. Among these proteins, C5a, together with its receptor, seems to play a central role in the pathogenesis of AAV, demonstrated by both animal and in vitro
]. The current study further investigated C5a as well as its receptor, CD88 and C5L2, in human AAV.
In the current study, only patients with positive MPO-ANCA were included, for two reasons. First, previous studies that investigated the role of the complement system in the pathogenesis of AAV focused mainly on MPO-ANCA-positive vasculitis [2
]. Second, our previous studies suggested that MPO is the most important ANCA target antigen for Chinese patients with AAV; MPO-ANCA-positive patients constituted about 80% to 90% of Chinese patients with AAV [20
In the current study, we investigated the plasma and urinary levels of C5a. The potent anaphylatoxin, C5a, is released when C5 is cleaved by C5 convertase [27
]. The measurement of cleavage products such as C5a in plasma, therefore, can provide a sensitive index of complement activation. It was found that the plasma level of C5a was about 10 times higher in patients with AAV in the active phase than that in normal controls. In the remission phase of AAV, the plasma level of C5a was comparable to that of normal controls. It might indicate an especially important role of C5a in AAV, compared with some other autoimmune diseases, such as primary antiphospholipid syndrome, in which plasma level of C5a is normal [22
]. The urinary level of C5a, after adjustment for urinary creatinine, was even higher (about 560 times) in patients with AAV than that in normal controls. Our results confirmed and further extended a recent observation [28
]. Increased urinary C5a might come from two sources. First, C5a may derive from the circulation because of glomerular basement membrane damage. Second, the complement may be activated locally in kidneys, and C5a is therefore released to urine [29
]. Overall, it may indicate the pathogenic role of excessive or uncontrolled production of C5a in AAV, in particular, in the glomerulonephritis of AAV.
The two types of C5a receptors, CD88 and C5L2, both have high affinity for C5a [6
]. Although in vitro
studies have found that mesangial cells could express CD88 [31
], the current study demonstrated that CD88 is expressed mainly in proximal tubules, distal tubules, and collecting ducts, but scarcely at all expressed in glomeruli and vasculature in patients with AAV. Our results are in line with those of previous studies [32
The expression of CD88 in the patients with AAV was significantly lower than that in normal controls, and, more important, was closely correlated with renal function and the extent of interstitial infiltration. Moreover, double-labeling immunofluorescence demonstrated that CD88 is scanty on neutrophils, monocytes, and macrophages. In a different opinion, CD88 is expressed in myeloid cells and is unable to be detected in tubular epithelial cells [35
]. However, in the current study, CD88 staining was confirmed by two other anti-CD88 monoclonal antibodies, and similar results were obtained (see Additional files 3
, and Figures S3, S4, S5, S6). Moreover, in positive controls, neutrophils could be stained by anti-CD88 monoclonal antibodies. Therefore, our results on neutrophil CD88 staining are reliable.
To confirm that the small amount of glomerular CD88 in AAV patients was not because of the lack of neutrophils in the glomeruli, neutrophils in the glomeruli of AAV patients were confirmed by both HE staining and immunohistochemical staining of CD66b (see Additional file 1
, Figure S1; Additional file 2
, Figure S2). Obviously, more neutrophil infiltration was found in the renal section of AAV patients than in normal controls. Immunohistochemical examination showed prominent expression of CD66b in glomeruli of AAV patients, whereas expression of CD66b was scanty in glomeruli of normal controls. Moreover, in HE and CD88 counterstaining, little or no CD88 staining was noted in renal infiltrated neutrophils (see Additional file 7
, Figure S7). Therefore, it was clear that the small amount of CD88 in glomeruli of AAV patients was because neutrophils in the glomeruli "lost" CD88, not because of the lack of neutrophils in the glomeruli.
Schreiber et al.
] suggested that CD88 in cells of myeloid origin plays a crucial role in the development of AAV. It is well known that CD88 on neutrophils interacts with C5a to produce a series of functional responses such as chemotaxis, enzyme release, degranulation, and respiratory burst [6
]. All of these functions induce inflammatory responses and are involved in ANCA-mediated tissue damage [4
]. However, the precise role of CD88 in renal tubular epithelial cells in AAV patients remains less defined. In the renal ischemia-reperfusion injury model, the interaction between C5a and CD88 on tubular epithelial cells induces a local inflammatory response resulting in cellular dysfunction and leads to renal function loss [36
]. CD88 is known to be rapidly internalized after treatment with C5a [37
]. The low expression of CD88 in AAV patients might be attributed to C5a-mediated internalization. The downregulation of CD88 expression in AAV patients might contribute to the self-protection mechanism, so as to alleviate the C5a-mediated inflammation. We speculate that binding of C5a to the CD88 in renal tubular epithelial cells may also be involved in tubulointerstitial injury in AAV patients. In AAV, the role of CD88 expression in the tubulointerstitium and the mediators responsible for downregulation of CD88 expression in the tubulointerstitium await further studies.
In contrast, the role of C5L2 remains controversial. C5L2 has been shown to be a nonsignaling decoy receptor and to compete with CD88 for binding to C5a. Thereby, C5L2 attenuates the proinflammatory C5a response. However, a study demonstrated that C5L2 is involved in the pathogenesis of asthma-like airway hyperresponsiveness and inflammation [7
]. Thus, C5L2 plays complex and dual roles in the pathogenesis of inflammation. To the best of our knowledge, the current study was the first one to investigate C5L2 in AAV. We found that the expression of C5L2 in glomeruli in AAV patients is significantly higher than that in normal controls. In ANCA-associated glomerulonephritis, glomerular intrinsic cells proliferation was not common. To some extent, the increase of cells in the glomerular tuft can reflect the glomerular inflammatory infiltrate. In our study, the expression of C5L2 in glomeruli positively correlated with the numbers of cells in the glomeruli. Moreover, C5L2 also expressed on neutrophils (or monocytes) and macrophages, besides on renal intrinsic cells. The significantly increased expression of C5L2 in renal specimens suggested that expression of C5L2 is regulated differently from that of CD88 in AAV. Unlike CD88, C5L2 did not show internalization on C5a binding [38
]. Moreover, inflammatory cells (that is, neutrophils, monocytes, and macrophages) infiltrated in kidneys could enhance the C5L2 expression. Whether C5L2 participated in proinflammatory or antiinflammatory responses in AAV requires further study.