A pathologic association between FH autoantibodies and DEAP-HUS has been reported in several cohorts but the prevalence of these antibodies has not been assessed in other diseases. Due to involvement of complement in SLE, RA and LA+/APS and a frequent presence of various autoantibodies in these diseases, we have performed a pilot study evaluating frequency of FH autoantibodies in these patient groups. Furthermore, we assessed prevalence of autoantibodies directed against two other soluble complement inhibitors C4BP and FI. We found no association of C4BP or FI autoantibodies and RA, SLE or LA+. To the contrary, significant increase in frequency of FH-autoantibodies compared to matched controls was found not only as previously reported in aHUS but also in RA and LA+.
The analytical specificity for these three autoantibody assays was evaluated by inhibition experiment using excess antigen. Addition of excess purified FH, FI or C4BP to several samples with high autoantibody titers diminished the signal (not shown) indicating specificity of the autoantibodies and demonstrating that the binding of these autoantibodies is not restricted to plastic absorbed protein. Furthermore, RA patient antibodies eluted from a PVDF membrane fragment containing FH transferred from SDS/PAGE gel, recognized purified FH but not pre-albumin separated by SDS/PAGE and transferred to a PVDF membrane (not shown). Incubation of HEK293 cell lysate spiked with FH with such specifically eluted patient FH-autoantibodies revealed mainly the signal corresponding to FH (150 kDa) and only very minor signals for bands around 270 kDa and 50 to 60 kDa. Thus, we concluded that FH-autoantibodies appear to be specific for FH and do not react with other human proteins.
FH is the main soluble inhibitor of the alternative complement pathway due to several mechanisms [5
]. FH acts as cofactor to a serine protease FI in a degradation of C3b but it also inhibits the formation and accelerates the decay of the alternative pathway C3-convertase. The main region responsible for these activities is N-terminal fragment composed of CCP1-4. Furthermore, C-terminal CCPs 19 to 20 are crucial for the attachment of FH to cellular surfaces in order to provide protection. The mechanism by which FH autoantibodies contribute to aHUS is still under investigation but several interesting observations have been made. In most cases reported so far, FH autoantibodies were directed against CCP19 to 20 of FH and did not inhibit complement inhibitory activity of FH in the fluid phase but rather blocked binding of FH to C3bBb convertase [16
] and to cell surfaces [17
]. The C-terminus of FH is also the region in which most of the aHUS-associated mutations are located. It has been suggested that under conditions of enhanced complement activation, a lower local concentration of FH at the cell surface may lead to cell damage. In case of aHUS, the main targeted tissue may be endothelium. In rheumatoid arthritis, joint inflammation causes exudation of plasma proteins including complement factors and inhibitors into synovial fluid. There is strong evidence of ongoing activation of complement in synovial fluid [9
], stimulated among others by molecules released during inflammation from the cartilage, such as fibromodulin [35
] and cartilage-oligomeric matrix protein [36
]. FH has been shown to attenuate complement activation initiated by these molecules [37
] and autoantibodies against FH, which we now observed in RA patients, also bind to and functionally impair the N-terminus of the protein, which may result in enhanced complement activation, ensuing inflammation and tissue damage in joints.
We found that RA patients from two independent cohorts have significantly increased FH autoantibody frequency compared to controls. We did not find any positive correlation between FH-autoantibody positivity and type of treatment and other analytical data in the Swedish RA cohort. Analysis of the Dutch RA patients did not reveal any significant differences between the FH-autoantibodies positive- and negative patients, regarding sex, age, autoantibody status, inflammation or smoking. However, the low number of FH-autoantibody-positive individuals highly limits the power to reliably find such differences. Furthermore, the presence of FH-autoantibodies in several chosen RA patients varied during disease course but it was not associated with infections. A larger study is required to find out if FH-autoantibodies are associated with disease flares or severity of RA.
When analyzing samples from SLE patients, FH-autoantibodies could only be detected at the time point with higher disease activity but no significant correlation between FH-autoantibody titers and disease activity (SLEDAI), or with any particular SLE ACR criteria were found. This may be due to the low power to detect differences with only a few FH-autoantibody positive patients and a low overall number of patients in this exploratory study. Interestingly, in DEAP-HUS FH autoantibodies were clearly lower at remission than at disease onset [32
]. Potential association of FH autoantibodies with nephritis would be of particular interest to study further due to recently reported genetic association with FH polymorphisms and SLE [38
], as well as association with non-synonymous mutations in FH and CD46 with faster onset of nephritis in SLE patients [39
So far little information on prevalence of mutations or polymorphisms in complement inhibitors, such as FH, has been published for RA or APS. A recent study found no association between common FH polymorphisms predisposing to age-related macular degeneration and RA [40
]. However, our current data together with previously published observations suggest that generation of FH-autoantibodies is specifically associated with deficiency of CFHR1 in many different disease groups. So far there is no clear explanation for the association between these two phenotypes. It has been suggested that autoantibodies generated in the context of a CFHR1 deficiency are targeted to a region of the FH molecule that is critical for the development of aHUS, that is, CCP19 to 20, and somehow is related to structural and functional similarity between CFHR1 and FH [41
]. However, this is not the case for the RA patients in the current study who carry FH-autoantibodies against many different regions of FH. To what extent this is the effect of intra-molecular epitope spreading or de-novo
recognition of epitopes outside CCP19 to 20 remains unknown. However, our results suggest that FH-autoantibodies found in RA, SLE and LA+ patients are polyclonal in nature since they interact with several different regions of FH. Another conclusion from these data could be that the combination of a deletion of CFHR1 and autoantibodies against FH is a risk factor for APS, or at least positivity for the lupus anticoagulants' test. The number of patients analyzed in this study was limited to support such strong conclusions and future replication studies in well-defined thrombosis and APS patients will have to be performed to support this observation.