Engineering T cell receptors for improved therapeutic T regulatory cell (Treg) function

MS is a debilitating autoimmune disorder in which autoreactive T cells attack and cause the destruction of the myelin sheath that protects neurons in the central nervous system. It is understood that this autoimmune attack is caused by autoreactive T cells with TCRs that bind their cognate peptide-MHC (pMHC) with atypical geometry and low affinity in the absence of functional Tregs. We aim to develop a therapeutic population of Tregs by transfecting naïve T cells with a TCR engineered to have higher affinity and normal binding geometry. We have developed and validated a single plasmid display system that expresses a myelin basic protein (MBP)-specific TCR, Ob.1A12 (Ob)—derived from a patient with MS—in its native, two-chain format on the surface of yeast. We have used this system to generate a library of MBP-specific Ob variants, from which we have selected clones with higher affinity. We have also developed a transposase-mediated system to stably express two chain TCRs on 58α-/β- T cells, which we can use to test various TCR’s ability to activate the cells upon binding. By cotransfecting FoxP3 with our TCR, we are using this system to demonstrate that TCRs engineered to have higher affinity have a greater ability and to suppress effector T cells. The full variable and constant domains for both α and β chains of Ob were cloned into the yeast expression vector pCT302. (Fig1) To create our library, we used NNS degenerate codon primers to randomize the residues of CDR1β. EBY100 S. cerevisiae cells expressing TCR were stained with anti-human TCR Vβ2, and cognate pMHC tetramer, MBP·DRB1*1501 (MBP·DRB) or a negative control tetramer, clip·DRB1*1501 (clip·DRB). Library variants were compared to wild type Ob, and ones with higher affinity for with MBP·DRB tetramer were selected with fluorescence assisted cell sorting (FACS). Consistent with Ob’s weak affinity (\u3e100μM), in wild type Ob we see only 11% of the TCRβ positive cells binding MBP·DRB at a detectable level, and 18% binding clip·DRB. Six variants from the fourth sort were isolated evaluated, and one variant, NS4B3, showed significantly improved binding—51% of TCRβ positive—to MBP·DRB with a modest decrease in non-specific binding to clip·DRB. (Fig2) We have cloned this variant into our PiggyBac vector, and will determine whether this variant confers more sensitive activation in vitro by tranfecting into 58α-/β- T cells and measuring the IL-2 release—an indicator of T cell activation—upon stimulation with cognate pMHC. We have demonstrated that we can generate a library of TCR variants using a two-chain yeast display system, while previous TCR libraries have depended upon single-chain formats. We believe this this system will generate TCRs with specificity and affinity that translates more readily to in vitro and in vivo T cell activity. Ultimately, we will transfuse the engineered Tregs with the best performance into humanized MS model mice to evaluate their potential to suppress further autoimmune attack. This work has been supported by the National Science Foundation Graduate Research Fellowship Program (DGE-1110007) and Welch Foundation Grant F-1767. Please click Additional Files below to see the full abstract