T cell receptor variable β 20-1 harbors a nucleotide binding pocket in the CDR2 β loop

Abstract

Novel aspects of T cells containing TCRVβ20-1 are numerous, ranging from pathogen specific reactivity to specific tissue homing, or possible T cell subsets. Recently, it was demonstrated that TCR itself could become reactive by binding to small molecules free of the pHLA interface. Our work here was to identify a natural ligand binding to an identified pocket on the CDR2β loop of these TCR. Using docking of suspected ligands, we were able to show Guanine and Adenine di- and tri-nucleotides readily bind to the identified site. Comparing these with small molecule sites found on other TCR types, we show this interaction is novel. With further molecular dynamic simulations, these sites are shown to be plausible by conducting simple computational based solubility tests as cross validation. Combined with simple proliferative responses, the identified nucleotides are also shown to have functional consequences by inducing T cell proliferation for CD4/Vβ20-1 + T cells, while failing to induce proliferation in other T cell isolates. Merging computational and simple cell assays, this work establishes a role of nucleotides in T cells found to contain this TCR sub-type.

Keywords: Adverse Drug Reactions; Autoimmune Diseases; T Cell Receptor; T Cell Receptor Variable Domain.

Figure 1. Sequence alignments.

(A) Comparison of the CDR2 loop of human TCRVβ20-1 with representative organisms. Green letters represent >60% conservation and Red, >85% across species. The loop is highly conserved across mammalian species, with high similarity even across distant species such as X. leavis. X. leavis retains two close matches, while other organisms contain one. (B) Comparison of human related TCRV with significant similarity. Only two TCRVα, 7-1 and 20-1, show any homology. Red letters, amino acid hydrogen bond with ligands, Green, backbone hydrogen bonds with ligand. Both show mutations rendering the loop incapable of binding nucleotides, indicated with blue bar underneath. (C) Alignments of the two human TCRVα showing overlapped K- > L/F at position 55, and T- > M/L at position 57. Mutated residues occupy most of the loop pocket.

Figure 2. Docked ligands on TCR models.

(A) Control TCRVβ9/Vα12-2 with bound SMX in CDR3α loop (B) Control TCRVβ5-1/Vα9-2 with bound SMX to two sites on the TCRVβ domain. Yellow, Vβ Blue Vα. (C) TCRVβ20-1/Vα17-1 with bound SMX in the CDR2β loop. (D) Same as (C), with GDP bound in same pocket. Residues shown in (C), (D) make up part of the ligand binding pocket. In (A), (B) residues shown are for orientation of the TCR to the viewer. (A) and (C) both represent data shown in ref [2].

Figure 3. Sulfanilamide Proliferation Responses.

(A) TCRVβ9/Vα12-2 sulfanilamide cross reactivity, (B) TCRVβ5-1/Vα9-2 cross reactivity, (C) TCRVβ20-1/Vα17-1 cross reactivity, and (D) TCRVβ20-1/CD4+ pull down pool cross reactivity. Solid second x-axis line across graphs represents background cut off. In all EBV-BCLC were used as APC, which were non-autologous for the pull down pool. Shown are means of triplicates experiments with three wells each experiment, grey bars, with STDV as black error bars. (A), (C) were both shown as figure in prior work [2].

Figure 4

TCR bound ligand molecular dynamic simulation in solvent. Shown are 10 nanosecond simulations of indicated ligand starting from docked conformations in the CDR2 loop of TCRVβ20-1. Colored line coordinates to mean distance from Center of mass (COM) of small molecule from TYR58 as central residue in the pocket. Light fluctuating lines are actual position at 5 picosecond intervals. Both SMX, blue and GDP, red, remain bound in the pocket, and SMX adopts a slightly tighter bound conformation within the pocket at 5.5 nanoseconds. SDM, green, leaves the pocket almost immediately.

Figure 5. Nucleotide Induced Proliferation.

(A) TCRVβ9/V α12-2 nucleotide proliferation (B) TCRVβ5-1/Vα9-2 nucleotide proliferation and (C) TCRVβ20-1/CD4 + pull down nucleotide proliferation. In all, SMX is used also as a positive control, along with PHA. In (B), a slight response above background was noted, but not significant. Solid secondary x-axis line in all across graph indicates background. Non-continuous line in (B), (C) shows background column used for 1 way ANOVA significance tests for each individual nucleotide, or SMX in (C). Each is triplicate experiments with three wells each experiment, means, grey bars, STDV, black error bars. Autologous APC were used for each.