Is Citrullination Required for the Presence of Restricted Clonotypes Reacting With Type II Collagen? Comment on the Article by Chemin et al

We read with great interest the recently accepted article by Chemin et al (1), in which they identified and functionally characterized citrullinated (Cit) collagen type II (CII)-specific T cell epitopes relevant to RA. Chemin and coll. enrolled a cohort of 15 HLA-DRB1*10:01 positive RA patients and healthy donors, stimulated with candidate CII peptides, demonstrating that the citrullination process is able to generate T cells specific for Cit-epitopes from CII. Particularly interesting is their report of the influence of citrulline residues on presentation and recognition of peptides by T cells. Furthermore the authors show also that amino acids not directly contacting the binding groove affect T cell recognition and somehow the peptide interaction with the HLA pocket. We studied in depth this interaction using a computational modelling where we explored the three-dimensional model for the ternary complex TCR-Vbeta, not citrullinated CII 261-273 and HLA-DR4 (1). The generation of the three-dimensional structure of TCR/DR4-collagen complex defined a roadmap that led to the identifications of small-molecule inhibitors able to block proliferation of CII261-273/HLA-DR4 specific T cells (2). These molecules can prove to be a valid starting point for the discovery of new and more potent derivatives able to disrupt protein-protein interaction between TCR and collagen/HLA-DR4 complex, blocking T cells proliferation through a mechanism of action that differs to those of therapeutic agents used currently against RA. In this context it is important to define the role of HLADRB1*10 in the pathogenesis of RA, in particular because it represents a very small portion in general population (less then 10%) and in particular RA population. Peculiar interest resides in the interaction of different HLA-DRB1 alleles, as we suggested in a recently published work (3); we demonstrated that shared, autoreactive, CII261-273 specific TRBV25-TRBJ2.2+ T cells could be selected on the background of HLADRB1*04 or HLA-DRB1*01 or HLA-DRB1*11 as second alleles, in a cohort of 27 (out 90) HLADRB1*04 ERA patients leading to the identification of a subgroup of patients with a more severe disease and a marked resistance to DMARDs (4). Moreover as a proof of concept we showed that TRBV25-TRBJ2.2+ T cells secrete IL-17, thus contributing directly to the aggressiveness of the disease. The possible mechanism underling this association could be in our opinion the influence carried by both DR alleles and their capability to modulate the affinity to the binding of the peptide, that were not citrullinated in that study. On the basis of the results of Chemin et al. we re-examined our cohort (90 ERA patients) where the HLA-DRB1*10 allele frequency were only 6,7% (6 out 90); interestingly all the patients presented a “double gene-dose” of the so-called shared epitope (R(Q)K(R)RAA): 4 out of the 6 HLA-DRB1*10 patients carried also a HLA-DRB1*04 allele and the other 2 carried HLADRB1*01 allele. The mean DAS of those patients (n=6) at the diagnosis were 3.7, and moreoverthey were all positive for ACPA and Rheumatoid Factor. Although none of them displayed the expansion of our study-target TCR rearrangement (TRBV25-TRBJ2.2), we performed immunoscope analysis also of other TCRs specific for CII261-273, previously identified in a group of RA patients (5). In this re-examination we found that other CII261-273 specific TCRs were expanded in this group of DR4+/DR10+ subjects, namely TRBV6.4-TRBJ2.2 (1 patient), TRBV14-TRBJ2.5 (2 patients), TRBV7_4-TRBJ2.6 (2 patients) and TRBV19-TRBJ2.5 (1 patient). We then compared these data with the results obtained by Chemin et al. (ref. Figure 5B Chemin et al.), and we found similarities with our six ERA patients, although they were stimulated with another not-citrullinated CII peptide; the possible explanation of those similarities, although in a very small cohort of patients in both studies, could reside in the influence of HLA allele combination more than on the citrullination of CII peptide, also on the basis of the poor differences between percentages of cit-CII(311-325)/unstimulated HEC impact on the total repertoire of PBMC of 4 out 5 HLA-DRB1*10 RA patients (ref. Figure 4 Chemin et al.). It could be interesting to analyse the influence of the other HLA-DRB1 allele associated with DR10. Finally it is important to stress the association of TCR repertoire with the clinical and the outcome of the patients. In fact, this subpopulation of patients that represent less than 10% of total RA population, are all good responders to a treat to target strategy in our clinical records, although carrying a double gene-dose of shared epitope. In conclusion we believe that restricted clonotypes can be found even with uncitrullinated collagen type II stimuli, that the DRB1* genotypes involved are several but that the rearranged TCR is skewed towards few selected TRVB chains among which the TRBV25-TRBJ2.2 appears to be the more represented. In addition these TCR rearrangements appears to be of special importance in defining the more refractory subsets of RA.