IntroductionT-cell receptors (TCRs) recognize cognate peptides presented by major histocompatibility complexes (pMHCs), playing a central role in adaptive immunity by defending against pathogens and tumor cells.1,2,3 This recognition, aided by coreceptors such as CD8, must be highly specific to ensure effective immune surveillance and prevent autoimmunity.4,5,6,7,8 However, pathogens and tumor cells can often evade the immune surveillance through mutations in the antigens that weaken TCR–pMHC binding.9,10,11 To counteract this evasion, strategies have been developed to enhance TCR binding affinity. Yet, these approaches frequently result in off-target interactions with self-antigens, leading to tissue damage and limiting the success of TCR-based T-cell therapies.12,13,14,15,16,17,18 Understanding the molecular mechanisms underlying the specificity of naturally evolved TCRs is essential to address these challenges. It is also crucial to investigate how CD8 influences off-target cross-reactivity in engineered high-affinity TCRs and to elucidate the mechano-chemical basis governing TCR–pMHC interactions that dictate specificity and prevent off-target effects for improving the safety and efficacy of TCR-based immunotherapies.Traditionally, TCR specificity and function were thought to depend on three-dimensional (3D) binding affinity, measured through methods like surface plasmon resonance (SPR). Higher TCR–pMHC affinity was assumed to improve specificity and activation.19,20 However, studies have shown that 3D affinity does not always correlate with TCR specificity and functional effectiveness.21,22,23 For instance, a high-affinity mutant of 2C-TCR (m33-TCR) was found to be less effective than its parental counterpart in slowing tumor growth in mice.24 These findings suggest that 3D affinity measurements may not accurately reflect the dynamics of TCR–pMHC interactions that occur at the two-dimensional (2D) interface where T cells engage antigen-presenting cells (APCs), often with assistance from coreceptors such as CD8. Previous studies have often relied on pMHC-tetramer staining to measure binding kinetics of pMHCs with TCRs expressed on T cells, but this method is prone to yield high false-positive rates for T cell activation, raising concerns about its reliability.23 Consequently, 2D binding assays have emerged as more physiologically relevant approaches for characterizing TCR–pMHC interactions. These assays better capture the conditions under which T cells interact with APCs, providing insights into the true specificity and functional effectiveness of TCRs in a biologically meaningful context.While measurements of 2D binding affinity better correlate with TCR potency, they remain inadequate in distinguishing highly specific TCRs from those prone to off-target cross-reactivity.22,25,26,27 This limitation may stem from the exclusion of CD8 in many 2D assays, despite evidence indicating that CD8 influences TCR binding by reducing dissociation rates or prolonging bond lifetimes.28,29 Prolonged TCR–pMHC bond lifetimes are generally associated with enhanced T cell activation and TCR specificity.6 However, the “focusing hypothesis” posits that an optimal, rather than maximal, dwell-time is crucial for regulating T cell activation and TCR specificity.5,30,31 Extended bond lifetimes may, in some cases, inhibit T cell activation, as proposed by the “serial triggering model”.19,32,33 The strength of CD8–MHC binding also modulates TCR–pMHC dissociation and ligand recognition, further underscoring CD8’s selective role in shaping TCR specificity.34,35 Furthermore, the precise mechanism by which TCR–pMHC interactions coordinate with CD8 to regulate TCR specificity remains unclear. Key questions include whether CD8 selectively assists TCR binding to certain antigens, how TCR–pMHC interactions influence CD8 cooperation, and whether CD8 contributes to the cross-reactivity of engineered high-affinity TCRs.Recent studies have underscored the critical role of mechanical forces in TCR antigen recognition.1,2,36,37,38,39,40,41,42,43 In the process of antigen scanning, T cells generate dynamic traction and tensile forces on TCR–pMHC complexes,44,45 which can induce conformational changes to promote the formation of catch bonds that stabilize TCRs' interactions with agonistic pMHCs.23,27 Unlike slip bonds, which weaken and shorten bond lifetimes under force, catch bonds strengthen TCR–pMHC interactions under force, making them uniquely suited to ensure TCR specificity.23,27 However, both naturally evolved TCRs with high specificity and engineered high-affinity TCRs with off-target cross-reactivity can form catch bonds with the same antigens,26 complicating efforts to distinguish between them.In this study, we demonstrate that the strength of TCR catch bonds significantly influences CD8’s role in determining TCR specificity. Using single-molecule biophysical and functional analyses, we reveal that naturally evolved TCRs form optimal catch bonds with cognate pMHCs. These optimal bonds facilitate CD8 engagement to enhance TCR specificity by enabling sequential, force-induced conformational changes in TCR, pMHC, and CD8 molecules. In contrast, engineered high-affinity TCRs form overly tight binding interfaces with cognate pMHCs, disrupting the necessary force-induced conformational changes and impairing proper CD8 interaction with pMHCs. Paradoxically, these high-affinity TCRs can also establish moderate-strength catch bonds with non-stimulatory or self pMHCs, leading to off-target cross-reactivity, diminished TCR specificity, and reduced therapeutic potential. Our findings highlight the importance of balancing TCR catch-bond strength to optimize CD8-mediated TCR specificity while minimizing off-target cross-reactivity, offering critical insights for designing effective TCR-based cell therapies that can achieve enhanced potency and safety.ResultsAffinity enhancement of 2C-TCR attenuates T cell activation, antigen sensitivity and specificityStudies have shown that the high-affinity variant (m33-TCR) was less potent in delaying tumor growth than its parental natural 2C-TCR in mice,24 prompting us to investigate how the 3D binding affinity \(({K}_{{{{\rm{d}}}}})\) of the TCR–pMHC bi-molecular interaction can influence TCR specificity and off-target cross-reactivity. 2C-TCR specifically recognizes the SIYRYYGL peptide (R4) presented by H-2Kb (R4-MHC), where the critical residue at position 4 (arginine) plays a vital role in initiating and determining 2C-TCR antigen recognition via its interaction with 2C-TCR-CDR3β27,46 to stabilize the 2C-TCR–R4-MHC complex (Fig. 1a). Mutation of this residue to leucine alters the peptide sequence to SIYLYYGL (L4), significantly reducing the TCR–pMHC 3D binding affinity and rendering the peptide non-stimulatory for the 2C-TCR (Fig. 1b). Additionally, the binding strength is influenced by interactions between the TCR’s complementarity-determining regions (CDRs), particularly CDR3α motif, and the α1α2 domains of pMHC (Fig. 1b).20 Therefore, we engineered two TCR variants by mutating the essential GFASA motif within 2C-TCR-CDR3α to LHRPA or LERPY, generating the m33- or m67-TCRs, respectively.20 These engineered TCRs were then individually expressed, with or without CD8, on 58α-β- hybridoma T cells (Supplementary information, Fig. S1a, b) to investigate the dynamic molecular mechanism by which CD8 and TCR–pMHC binding cooperatively regulate TCR specificity and off-target cross-reactivity.Fig. 1: Strengthening TCR–pMHC 3D binding affinity by TCR CDR3α mutagenesis impairs the TCR specificity in the presence of CD8.a Structural overview of 2C-TCR (cornflower blue), m33-TCR (medium purple), and m67-TCR (cyan) in complex with R4-MHC are depicted. The R4 (SIYRYYGL) peptide, H-2Kb and β2-microglobulin within R4–MHC complex are highlighted in gold, purple, and yellow, respectively. A zoomed-in view of the interactions between the R4 peptide (SIYRYYGL, upper panel) or the L4 peptide (SIYLYYGL, bottom panel) and CDR3s of 2C-TCR is shown in dashed boxes on the left; the peptide is represented in yellow ribbon with the hotspot residue shown in stick, and CDR3s are represented in surface colored according to electrostatic properties (positive charge in red and negative charge in blue) on the right. b 3D binding affinity \(({K}_{{{{\rm{d}}}}})\) of 2C-, m33-, and m67-TCRs binding to R4- or L4-MHCs, respectively. c–f IL-2 production of 2C, m33 and m67 hybridoma T cells expressing CD8 or not, when stimulated by RMA-S cells pulsed with different concentrations of R4 (c, d) or L4 (e, f) peptide. The data for both the presence and absence of CD8 were obtained from the same experimental batch. The TCR sensitivity is determined by the lowest antigen concentration that started to elicit the IL-2 production from hybridoma T cells (5% potency, or P5). The presented data represent one of three independent experiments. The analyses were conducted using the Mann–Whitney U-test. g, h The comparison of TCR specificity of 2C-, m33-, or m67-TCRs in the absence (g) or presence (h) of CD8. Error bars are ± SEMs. The analyses were conducted using unpaired student's t-tests. Statistical significance was indicated as follows: ns not significant, *P