T cell receptors employ diverse strategies to target a p53 cancer neoantigen

Adoptive cell therapy with tumor-specific T cells can mediate durable cancer regression. The prime target of tumor-specific T cells are neoantigens arising from mutations in self-proteins during malignant transformation. To understand T cell recognition of cancer neoantigens at the atomic level, we studied oligoclonal T cell receptors (TCRs) that recognize a neoepitope arising from a driver mutation in the p53 oncogene (p53R175H) presented by the major histocompatibility complex class I molecule HLA-A2. We previously reported the structures of three p53R175H-specific TCRs (38-10, 12-6, and 1a2) bound to p53R175H and HLA-A2. The structures showed that these TCRs discriminate between WT and mutant p53 by forming extensive interactions with the R175H mutation. Here, we report the structure of a fourth p53R175H-specific TCR (6-11) in complex with p53R175H and HLA-A2. In contrast to 38-10, 12-6, and 1a2, TCR 6-11 makes no direct contacts with the R175H mutation, yet is still able to distinguish mutant from WT p53. Structure-based in silico mutagenesis revealed that the 60-fold loss in 6-11 binding affinity for WT p53 compared to p53R175H is mainly due to the higher energetic cost of desolvating R175 in the WT p53 peptide during complex formation than H175 in the mutant. This indirect strategy for preferential neoantigen recognition by 6-11 is fundamentally different from the direct strategies employed by other TCRs and highlights the multiplicity of solutions to recognizing p53R175H with sufficient selectivity to mediate T cell killing of tumor but not normal cells.

Adoptive cell therapy with tumor-specific T cells can mediate durable cancer regression. The prime target of tumorspecific T cells are neoantigens arising from mutations in selfproteins during malignant transformation. To understand T cell recognition of cancer neoantigens at the atomic level, we studied oligoclonal T cell receptors (TCRs) that recognize a neoepitope arising from a driver mutation in the p53 oncogene (p53R175H) presented by the major histocompatibility complex class I molecule HLA-A2. We previously reported the structures of three p53R175H-specific TCRs (38-10, 12-6, and 1a2) bound to p53R175H and HLA-A2. The structures showed that these TCRs discriminate between WT and mutant p53 by forming extensive interactions with the R175H mutation. Here, we report the structure of a fourth p53R175H-specific TCR (6)(7)(8)(9)(10)(11) in complex with p53R175H and HLA-A2. In contrast to 38-10, 12-6, and 1a2, TCR 6-11 makes no direct contacts with the R175H mutation, yet is still able to distinguish mutant from WT p53. Structure-based in silico mutagenesis revealed that the 60-fold loss in 6-11 binding affinity for WT p53 compared to p53R175H is mainly due to the higher energetic cost of desolvating R175 in the WT p53 peptide during complex formation than H175 in the mutant. This indirect strategy for preferential neoantigen recognition by 6-11 is fundamentally different from the direct strategies employed by other TCRs and highlights the multiplicity of solutions to recognizing p53R175H with sufficient selectivity to mediate T cell killing of tumor but not normal cells.
Adoptive cell therapy (ACT) with tumor-specific T cells can promote durable regression of diverse cancers, including metastatic melanoma, colon, bile duct, cervix, and breast cancers (1)(2)(3)(4)(5). The therapeutic effect of these tumorinfiltrating lymphocytes (TILs) is mediated primarily by cytotoxic CD8 + T cells (6). The main target of tumor-specific T cells are neoantigens that result from DNA alterations during malignant transformation (7). Of special interest are neoantigens derived from oncogenes bearing driver mutations because these mutations are tumor-specific, important for tumor progression, and generally expressed by all tumor cells (8). In a pioneering study of ACT, a patient with metastatic colorectal cancer was treated successfully with four ex vivoexpanded CD8 + T cell clones specific for a neoepitope arising from the G12D driver mutation in the KRAS oncogene (2,9).
TP53 (tumor protein p53) was the first tumor suppressor gene identified and is inactivated in the large majority of human cancers (10,11). Mutations in TP53 effect most of the hallmarks of cancer cells, including proliferation, genomic instability, and metastasis (12,13). Hotspot positions include R175, G245, R248, R273, and R282, which cluster in the central DNA-binding domain of p53 and alter its DNA-binding properties (14). Mutations at these sites are attractive candidates for targeted immunotherapy because they confer a growth advantage to tumor cells and are associated with malignant progression.
The immunogenicity of p53 mutations in cancer patients has been demonstrated by the detection of T cell responses against several p53 neoantigens, most notably R175H in which arginine at position 175 is replaced by histidine (15,16). This driver mutation is the most frequently observed mutation in TP53 as well as the most common mutation in any tumor suppressor gene (17). A number of T cell receptors (TCRs) have been isolated from TILs of epithelial cancer patients that target a neoepitope corresponding to residues 168 to 176 of p53R175H (HMTEVVRHC; mutant amino acid in bold) (15,16). The TCRs are restricted by HLA-A*02:01, which is the most frequent major histocompatibility complex (MHC) class I allele in the U.S. population (18). These TCRs may prove effective in eliminating tumors expressing HLA-A2*02:01 and the p53R175H mutation when transduced into a patient's peripheral blood lymphocytes for ACT (15,16).
With the aim of understanding TCR recognition of cancer neoantigens at the atomic level, we previously determined crystal structures of three p53R175H-specific TCRs (12-6, 38-10, and 1a2) in complex with HLA-A*02:01 and the neoepitope p53R175H (19). The structures revealed that these TCRs discriminate between WT and mutated p53 by focusing on the R175H mutation, with which they make extensive interactions. Here, we report the structure of a fourth p53R175H-specific TCR (6)(7)(8)(9)(10)(11) bound to the p53R175H peptide and HLA-A*02:01. In sharp contrast to 12-6, 38-10, and 1a2, TCR 6-11 makes no contacts with the R175H mutation, yet is nevertheless able to distinguish mutant from WT p53. Collectively, these structures demonstrate that there are multiple distinct solutions to recognizing the p53R175H neoepitope with sufficient on-target affinity and specificity to mediate the killing of tumor cells expressing mutant p53 without affecting normal cells expressing WT p53, a critical consideration for avoiding adverse clinical events in ACT due to off-target TCR recognition (20).
Overview of the 6-11-p53R175H-HLA-A2 complex To understand how TCR 6-11 discriminates between WT and mutant p53 epitopes, and to compare discrimination by 6-11 with that by 12-6, 38-10, and 1a2, we determined the structure of the 6-11-p53R175H-HLA-A2 complex to 3.33 Å resolution (Table S1) ( Fig. 2A). The interface between TCR and pMHC was in unambiguous electron density for each of the four complex molecules in the asymmetric unit of the crystal (Fig. 2B). The rmsd in α-carbon positions for the TCR VαVβ and MHC α1α2 modules, including the p53R175H peptide, ranged from 0.18 Å to 0.35 Å for the four 6-11-p53R175H-HLA-A2 complexes, indicating close similarity. Therefore, the following description of TCR-pMHC interactions applies to all molecules in the asymmetric unit of the crystal.
T cell receptor 6-11 docks over p53R175H-HLA-A2 in a canonical diagonal orientation, with variable α (Vα) over the α2 helix of HLA-A2 and variable β (Vβ) over the α1 helix. The crossing angle of TCR to pMHC (21) is 35 , which is similar to the crossing angles of 38-10 (34 ) and 1a2 (30 ) but more acute than that of 12-6 (51 ) (Fig. 3, A-D). The incident angle (22), which corresponds to the degree of tilt of TCR over pMHC is 19 for 6-11, compared to 20 for 12-6, 27 for 38-10, and 1 for 1a2. Thus, the 6-11 complex is most like the 38-10 complex with respect to crossing angle and most like the 12-6 complex with respect to incident angle.
T cell receptor 6-11, like TCRs 12-6, 38-10, and 1a2 (19), is shifted toward the C-terminus of the p53R175H peptide, which is the site of the driver mutation at P8. To quantitate the shifts, we projected the positions of the TCR centers onto the pMHC plane, where the x-axis is aligned with the peptide and a more positive x value indicates a C-terminal shift (Table S2). Of note, 6-11 exhibits the seventh-highest C-terminal shift among 137 reported TCR-pMHC structures, which is nearly as much as 38-10 (third-highest) but more than 12-6 (23rdhighest) or 1a2 (27th-highest). The C-terminal shift of these TCRs is the key to their ability to discriminate between WT and mutant p53 peptides (see below).
We previously showed that the large majority (80%) of contacts between TCRs 38-10, 12-6, and 1a2 and the p53R175H peptide involves C-terminal residues P7 Arg and P8 His, and that these contacts are about evenly distributed between these two residues (19) (Fig. 5, A-C). These TCRs achieve highly specific recognition of mutant p53 peptide relative to WT by minimizing interactions with the central and N-terminal portions of p53R175H, which are structurally identical in the WT peptide, and instead focusing on the mutation at P8. In sharp contrast to 38-10, 12-6, and 1a2, 6-11 makes no interactions with P8 His (Fig. 5B) (Table S4), despite the ability of this TCR to discriminate between mutant and WT p53 (Fig. 1, A and B). The 6-11-p53R175H-HLA-A2 complex crystallized at pH 8.5. The imidazole group of P8 His should be uncharged at this pH. Instead of P8 His, the principal focus of 6-11 is on the P7 Arg side chain, with which it forms four hydrogen bonds: 6-11 Asp93α Oδ1-Nη2 P7 Arg, 6-11 Asp93α Oδ2-Nη1 P7 Arg, 6-11 Asp93α Oδ2-Nη2 P7 Arg, and 6-11 Pro96α O-Nη2 P7 Arg (Fig. 5C). Computational alanine scanning in Rosetta (23) with the 6-11-p53R175H-HLA-A2 complex as input (Table 3) supports the dominance of P7 Arg in 6-11 TCR recognition. In addition, 6-11 Tyr95α makes hydrophobic contacts with P5 Val and P6 Val. However, since P5 Val, P6 Val, and P7 Arg are conserved and highly superimposable in crystal structures of the unbound WT p53-HLA-A2 and mutant p53R175H complexes (19), the mechanism whereby TCR 6-11 distinguishes WT from mutant p53 is not obvious.
To resolve this conundrum, we evaluated the effect of replacing P8 His by Arg, which corresponds to reversion to the WT p53 peptide, we carried out in silico mutagenesis using Rosetta (23). The peptide substitution was modeled in the X-ray structure of the 6-11-p53R175H-HLA-A2 complex, followed by side-chain minimization and energetics-based scoring to calculate ΔΔG. The predicted ΔΔG value was 1.6 Rosetta energy units (REU; analogous to kcal/mol) (Table 3), consistent with the substantial (60-fold) loss in 6-11 binding affinity for WT p53 peptide that we measured by SPR (Fig. 1B). To investigate the mechanistic basis for this affinity loss, the individual Rosetta scoring function terms comprising the predicted ΔΔG were obtained (Table 3). This revealed that the energetic cost of desolvating P8 Arg during complex formation with TCR 6-11 dominated the reduction in binding affinity, contributing 1.5 out of 1.6 REU of the predicted affinity change. Structurally, the limited space around P8 for an Arg residue at that position leads to likely packing interactions of the Arg side chain with the 6-11 TCR CDR loops and its unfavorable desolvation (Fig. S1). The unfavorable effect of desolvating P8 Arg versus His by TCR 6-11 is in accordance with the n-octanol to water amino acid transfer energies of Fauchere and Pliska (24), which had an approximately 1.5 kcal/ mol hydrophobic energy difference between His and Arg side chains, as well as more recent computed amino acid hydrophobic energies (25) that showed an approximately 2 kcal/mol difference for Arg versus His residue desolvation. As with the Rosetta-computed ΔΔG values, these Arg versus His amino acid desolvation energy differences are comparable to, albeit slightly less than, the 60-fold binding affinity loss (corresponding to ΔΔG of approximately 2.4 kcal/mol) observed for TCR 6-11 due to the P8 His to Arg substitution. By contrast, a similar previous analysis for TCRs 38-10, 12-6, and 1a2 showed that disruption of hydrogen bonds involving P8 His was mainly responsible for affinity losses of 38-10 and 1a2 for WT p53, while loss of van der Waals interactions accounted for the affinity reduction of 12-6 (19).
We also investigated the energetic contribution of P8 His of p53R175H to binding TCR 6-11 by mutating P8 His to Ala. As noted in Table 3, this substitution was predicted to have a substantial effect on 6-11 TCR binding by Rosetta (23) (ΔΔG: 1.2 REU), which is unexpected, since the 6-11-p53R175H-HLA-A2 structure revealed no interactions between TCR and P8 His using standard cut-off distances of 4.0 Å for van der Waals contacts and 3.5 Å for hydrogen bonds (Table S4). As measured by SPR (Fig. 1C), 6-11 bound p53R175A-HLA-A2 with K D = 98.4 ± 4.2 μM, which is 25-fold lower affinity than for p53R175H-HLA-A2 (K D = 3.5 μM). This destabilization is in accordance with the in silico modeling, which likewise predicted less 6-11 binding disruption for P8 His to Ala versus His to Arg. In the case of the P8 Ala substitution, the attractive van der Waals term dominated the predicted binding energy loss, indicating that while relatively shortrange TCR contacts of the P8 His (<4.0 Å) are not present in the structure, other proximal TCR contacts of the P8 His side chain, for example 6-11 Pro96α, which is <5 Å from P8 His, are favorable interactions that are lost upon Ala substitution. To assess the TCR 6-11 binding impact of additional substitutions at P8, we performed computational mutagenesis to model the effects of all 19 non-His amino acids at that position (Table S5). While certain hydrophobic amino acid residues may allow 6-11 binding, based on this analysis, several charged and polar residues at P8 (e.g., Asp, Gln, and Glu) are predicted to cause major disruptions in 6-11 binding (ΔΔG > 1.0 REU), in addition to Arg and Ala. While computational mutagenesis in Rosetta has been relatively accurate in the context of other TCR-pMHC interfaces (26,27), due to possible limitations of the Rosetta conformational sampling or scoring function, future experimental binding measurements can confirm these structure-based predictions of hotspots or affinity changes.

Discussion
During protein-protein complex formation, water molecules are largely excluded from the interface between the interacting partners. The removal of waters exacts a large desolvation penalty that must be offset by attractive hydrophobic and electrostatic contributions in order to form a stable complex. Our computational analysis of the 6-11-p53R175H-HLA-A2 structure revealed that the lower affinity of TCR 6-11 for WT p53-HLA-A2 (K D = 236 μM) than for mutant p53R175H-HLA-A2 (3.8 μM) is primarily due to the higher energetic cost of desolvating P8 Arg in the WT p53 peptide than P8 His in the mutant. Importantly, this unusual strategy for distinguishing WT from mutant p53 does not rely on direct contacts between TCR 6-11 and P8 His, in marked contrast to the more typical strategies employed by TCRs 38-10, 12-6, and 1a2, which depend on direct contacts.
Although we do not observe direct contacts between TCR 6-11 and P8 His, we cannot rule out indirect interactions mediated by bound water molecules. The limited resolution of the 6-11-p53R175H-HLA-A2 structure (3.33 Å) does not permit the identification of ordered waters with confidence, and none were included in the final model. However, in several high resolution TCR-pMHC structures (≤2.5 Å), interfacial waters have been found to form bridging hydrogen bonds that enhance polar interactions and neutralize unpaired hydrogenbonding groups (28).
In addition to TCRs, antibodies are also under active investigation for immunotherapeutic targeting of cancer neoantigens. A monoclonal antibody (H2) specific for p53R175H-HLA-A2, the exact same pMHC targeted by TCR 6-11, was recently reported (29). In the crystal structure of Fab H2 bound to p53R175H-HLA-A2, the V L CDR3 and V H CDR1-3 loops form a tight cage around P7 Arg and P8 His in which the imidazole side chain of P8 His is part of a hydrogen bonding network with V L CDR3 Tyr94 and V H CDR2 Asp54. Thus, antibody H2, like TCRs 38-10, 12-6, and 1a2, distinguish WT Figure 4. Interactions of TCRs with HLA-A2. A, interactions between 6-11 and the HLA-A2 α1 helix. The side chains of contacting residues are drawn in stick representation with carbon atoms in pink (TCR α chain), blue (TCR β chain) or light gray (HLA-A2), nitrogen atoms in dark blue, and oxygen atoms in red. Hydrogen bonds are indicated by red dashed lines. B, interactions between 6-11 and the HLA-A2 α2 helix. C, interactions between 38-10 and the HLA-A2 α1 helix. D, interactions between 38-10 and the HLA-A2 α2 helix. E, interactions between 12-6 and the HLA-A2 α1 helix. F, interactions between 12-6 and the HLA-A2 α2 helix. G, interactions between 1a2 and the HLA-A2 α1 helix. H, interactions between 1a2 and the HLA-A2 α2 helix. TCR, T cell receptor.
from mutant p53 via direct contacts with P8 His, which is fundamentally different from the indirect strategy utilized by TCR 6-11.

Crystallization and data collection
For crystallization of the TCR 6-11-p53R175H-HLA-A2 complex, TCR 6-11 was mixed with p53R175H-HLA-A2 in a 1:1 M ratio at a concentration of 14 mg/ml. Crystals were obtained at room temperature by vapor diffusion in hanging drops. The 6-11-p53R175H-HLA-A2 complex crystallized in 20% (w/v) PEG 3350, 0.1 M Bis-Tris propane (pH 8.5), and 0.2 M potassium thiocyanate. For data collection, the crystals were cryoprotected with 20% (w/v) glycerol and flash-cooled. X-ray diffraction data were collected at beamline 23-ID-D of the Advanced Photon Source, Argonne National Laboratory. The diffraction data were indexed, integrated, and scaled using the program HKL2000 (30). Data collection statistics are shown in Table S1.

Structure determination and refinement
Data reduction was performed using the CCP4 software suite (31). The TCR 6-11-p53R175H-HLA-A2 structure was solved by molecular replacement with the program Phaser (32) and refined by Phenix with NCS constraints (33). The model was further refined by manual model building with Coot (34) based on 2F o -F c and F o -F c maps. The α chain of a CD1bspecific TCR (PDB accession code 6OVN) (35), the β chain of dengue virus-specific TCR D30 (5WKF) (36), and p53R175H-HLA-A2 (6VR5) (19) with the CDRs and peptide removed were used as search models to determine the orientation and position of the 6-11-p53R175H-HLA-A2 complex. Refinement statistics are summarized in Tables S1. Contact residues were identified with the CONTACT program (31) and were defined as residues containing an atom 4.0 Å or less from a residue of the binding partner. Figures were prepared using PyMOL (https://pymol.org/).

Surface plasmon resonance analysis
The interaction of TCR 6-11 with p53-HLA-A2 and p53R175H-HLA-A2 was assessed by SPR using a BIAcore T100 biosensor. Biotinylated p53-HLA-A2 or p53R175H-HLA-A2 was immobilized on a streptavidin-coated BIAcore SA chip (GE Healthcare) at 3000 resonance units (RU). An additional flow cell was injected with free biotin alone to serve as a blank control. For the analysis of TCR binding, solutions containing different concentrations of 6-11 were flowed sequentially over chips immobilized with p53-HLA-A2, p53R175H-HLA-A2, or the blank. The time point before the ending injections was used as the equilibrium level. Dissociation constants were calculated by fitting equilibrium and kinetic data to a 1:1 binding model using BIA evaluation 3.1 software.

Calculation of TCR centers
Calculations of TCR center positions were performed as described previously (19). The 6-11-p53R175H-HLA-A2 complex was translated and rotated into a reference frame used in our previous study (19), with the MHC helix plane aligned to the x-y plane. This reoriented complex was then used to calculate the TCR variable domain center and its projection onto the x-y plane, giving its x position and y position over the centered and oriented MHC in Ångstrom units. These positions were compared with values calculated for other structurally characterized TCRpeptide-MHC class I complexes in that same reference frame, obtained from the TCR3d database (38); all positions except the 6-11 TCR position were reported in our recent study (19).

Data availability
Atomic coordinates and structure factors for the TCR 6-11-p53R175H-HLA-A2 complex have been deposited in the Protein Data Bank under accession code 7RM4.
Supporting information-This article contains supporting information (31)