Immunoglobulin isotype switching of antibodies to vimentin is associated with development of transplant glomerulopathy following human renal transplantation

Background: Immune responses to tissue-restricted self-antigens are thought to play a role in chronic rejection after solid organ transplantation. De novo development of antibodies (Abs) to vimentin have been reported to be associated with interstitial fibrosis/tubular atrophy after kidney transplant, and it has been suggested that immunoglobulin isotype switching of Abs to vimentin may occur during this process. We aimed to determine the correlation between immunoglobulin isotype switching of Abs to vimentin and development of transplant glomerulopathy (TG) after kidney transplant, and to determine whether citrullinated modification of vimentin is required for de novo anti-vimentin development. Methods: Sera were collected from 24 patients with TG (diagnosed on biopsy), 24 matched stable kidney transplant recipients (KTxRs) and 22 normal healthy subjects who did not undergo transplant. Serum vimentin Abs concentrations were measured using enzyme-linked immunosorbent assay (ELISA). Immunoglobulin isotypes of anti-vimentin were determined using isotype-specific Abs conjugated with horseradish peroxidase. Samples were considered positive to vimentin Abs if the values were above mean + 2 x standard deviations of the levels in the healthy control subjects. Specificities of anti-vimentin for mutated citrullinated vimentin and anti-mutated citrullinated vimentin were measured by ELISA. Results: In this retrospective analysis of 24 KTxRs with TG, 16/24 (67%) patients with biopsy -proven TG developed Abs to vimentin (645 +/- 427 ng/ml). In contrast, only 4/24 (17%) stable KTxRs had detectable Abs to vimentin (275 +/- 293 ng/ml; p = 0.001). Of the patients with TG, 15/24 (63%) developed Abs to vimentin of IgG isotype (572 +/- 276 ng/ml), whereas only 6/24 (25%) stable KTxRs (310 +/- 288 ng/ml) had anti-vimentin of IgG isotype (p = 0.002). However, no significant difference was noted in the concentration of IgM isotype anti-vimentin between KTxRs with TG (9/24 [38%], 407 +/- 401 ng/ml) and stable KTxRs (5/24 [21%], 348 439 ng/ml; p = 0.631). The serum concentration of Abs specific for the mutated form of citrullinated vimentin was not significantly different between KTxRs with TG and stable KTxRs. Conclusions: Patients with biopsy -proven TG demonstrated significantly increased levels of anti-vimentin Abs of the IgG isotype compared with stable 'Calls. Anti-vimentin in stable KTxRs was primarily of IgM isotype. Therefore, the observed isotype switching of anti-vimentin from IgM to IgG isotype strongly suggests ongoing immune responses to vimentin in KTxRs diagnosed with TG. Furthermore, as opposed to patients with rheumatoid arthritis (who develop immune responses primarily to citrullinated vimentin), KTxRs diagnosed with TG developed immune responses to non-citrullinated vimentin, suggesting that modification of vimentin protein via citrullination is not required for the de novo anti-vimentin response seen in patients with TG.