Innate TCRβ-chain engagement drives human T cells toward distinct memory-like effector phenotypes with immunotherapeutic potentials

Clonotypic αβ T cell responses to cargoes presented by major histocompatibility complex (MHC), MR1, or CD1 proteins underpin adaptive immunity. Those responses are mostly mediated by complementarity-determining region 3 motifs created by quasi-random T cell receptor (TCR) gene rearrangements, with diversity being highest for TCRγδ. Nonetheless, TCRγδ also displays nonclonotypic innate responsiveness following engagement of germline-encoded Vγ-specific residues by butyrophilin (BTN) or BTN-like (BTNL) proteins that uniquely mediate γδ T cell subset selection. We now report that nonclonotypic TCR engagement likewise induces distinct phenotypes in TCRαβ+ cells. Specifically, antibodies to germline-encoded human TCRVβ motifs consistently activated naïve or memory T cells toward core states distinct from those induced by anti-CD3 or superantigens and from others commonly reported. Those states combined selective proliferation and effector function with activation-induced inhibitory receptors and memory differentiation. Thus, nonclonotypic TCRVβ targeting broadens our perspectives on human T cell response modes and might offer ways to induce clinically beneficial phenotypes in defined T cell subsets.

The TRBV6-5 antibody was optimized for affinity, cross-reactivity to non-human primates and stability using molecular evolution strategies.Three rounds of molecular evolution using five different phage/yeast libraries were employed.Each round constituted creating a DNA library of scFv mutants, expressing the library on the surface of M13 filamentous phage and/or yeast (Saccharomyces cerevisiae), enriching the phage or yeast population for improved clones by incubating the library at different conditions (low antigen concentration, competition for scarce antigen in the presence of excess competitor, high-temperature stress) nonpermissive for the parental clone followed by high-throughput individual clone screening by ELISA and/or flow cytometry.For the thermostability branch, prior to incubation with 10 nM antigen, library phage was incubated at temperatures at or above the Tm.For affinity screens, bacterial periplasmic extracts containing soluble scFv mutants were tested by ELISA against coated huTRBV6-5 and cynomolgus TRBV antigen in the absence (direct ELISA) and presence of excess parental antibody (competitive ELISA).The final clones derived from the screens were cloned into an expression vector that allowed generation of monovalent scFv-Fc fusion.Following purification, antibodies were tested for binding to human and cynomolgus antigen by Surface plasmon resonance and thermostability by Differential scanning fluorimetry.The variant that showed the best human/cynomolgus cross reactivity and thermal stability (defined as affinity matured, AM) was chosen as lead for further studies.

Supplementary Figure 1 .
a-CD3e AF647 a-hIgG + a-TRBV6-5 PAR AF647 a-hIgG + a-TRBV6-5 AM AF647 a-mIgG + a-TRBV6-5 H131 Analysis of the binding of a-TRBV antibodies.A. Flow cytometry analysis of the interference between a-TRBV and a-CD3e antibodies.Dot plots (top row) show the gating strategy.Overlayed histograms (bottom row) show the consistently lower a-CD3e staining of cells co-stained by each a-TRBV antibodies (blue) compared to other subsets (black).Numbers indicate a-CD3e MFI.Representative of 15 donors.B. Full TCR complex structure showing that the accessible region of the Vb domain (grey) is located on the same side as and on top of the CD3e subunits (dark blue), suggesting that the binding of a-TRBV antibodies interferes with a-CD3e staining.Derived from PDB 6JXR (69).CDR1, CDR2, CDR3 and HV4 regions are highlighted in green, yellow, cyan and magenta, respectively.C. J76 cells expressing the indicated TCR were stained for CD3 expression without (Control) or with prior incubation with increasing doses of the indicated antibodies.Primary flow cytometry histograms for the maximum dose (100 µg/ml) and summary data are shown on the top and bottom rows, respectively.Representative of two independent experiments.D. Representative histograms showing the staining of the indicated TRBV6-5 variants and TRBV6-6 negative control by the indicated antibodies.Related to Fig 1C.E. Summary flow cytometry analysis of the staining of the indicated TRBV5-1, TRBV12-3 and TRBV20-1 mutants and their respective negative controls (TRBV5-4, TRBV12-5 and TRBV29-1).MFI were normalized to that of a-CD3e and expressed as % of the wild-type (wt) control stainings.Representative of 3 independent transductions.F. Mapping of the aminoacids mutated in D. The Vb domain structures were derived from PDB 5BRZ (TRBV5-1) (70), 6UK4 (TRBV12-3) (71), and 4PJ8 (TRBV20-1) (72).The CDR1, CDR2, CDR3 and HV4 regions are highlighted in green, yellow, cyan and magenta, respectively.

4 a
a -C D 3 e h S P 3 4 a -T R B V 6 -5 A M U n s t i m a -C D 3 e h S P 3 4 a -T R B V 6 -5 A M U n s t i m a -C D 3 e h S P 3 4 a -T R B V 6 -5 A M U n s t i m a -C D 3 e h S P 3 -T R B V 6 -5 A M %T-BET + %T-BET + T-BET MFI (×10 3 ) .5 a-CD8a .5 a-CD8a FITC a-CD4

IRF4a
-5 AM (C7) a-CD3e hSP34 (C4) a-TRBV6-5 AM (C5) a-CD3e hSP34 (C0) -T R B V 6 -5 A M ( C l u s t e r 7 ) a -C D 3 e h S P 3 4 ( C l u s t e r 4 ) a -T R B V 6 -5 A M ( C l u s t e r 5 ) a -C D 3 e h S P 3 4 ( C l u s t e r 0 ) t i m a -C D 3 e h S P 3 4 U n s t i m a -T R B V 6 -5 A M a -T R B V 6 -5 A M a -C D 3 e h S P

Supplementary Figure 7 .
Identification of a-TRBV-regulated transcription factors.A. Waterfall plot showing the Log2FC of transcription factors differentially expressed in a-TRBV6-5 AM -targeted versus -non-targeted cells in CD4 (top) and CD8 (bottom) cells.Related to Fig 7A.B. Waterfall plot showing the Log2FC of transcription factors differentially expressed in a-TRBV6-5 AM -stimulated cells, comparing CD25 + versus CD25 NEG CD8 cells.Related to Fig 7A.C. Violin plots showing the expression of IRF4 (imputed expression).Log2FC, Log2 fold-change; ****, p<0.0001 (Wilcoxon rank sum test with Bonferroni correction, from the DEG analysis).D,E.Purified T cells were cultured for 7 days without (Unstim) or with the indicated plate-bound antibodies then intracellularly stained for analysis of IRF4 expression by flow cytometry, after gating on CD4 or CD8 cells (right and left columns respectively).Plots representative of n=6 donors are shown in D. Summary data in E are the mean of the percentage of IRF4 + (top row) cells, and their MFI (bottom row).ns, not significant; *, p<0.05; **, p<0.01; ***, p<0.001 (paired ANOVA with Dunnett's method, comparing a-TRBV6-5 AM to every other condition individually).