Fig 2: Surface CD107a and phosphorylation of ERK of NK cells stimulated by immobilized NKG2D ligands A, BFlow cytometry analysis of the percentages of CD107a+ (A) cells pERK+ cells (B) under stimulation of different NKG2D ligands (MICA in red, MICB in orange, ULBLP3 in blue) compared with SA negative control (gray).C, DCorresponding quantification of percentages of CD107a+ cells and pERK+ NK cells in (A) and (B).E, FPlots and Pearson correlation analysis of NKG2D ligands stimulated percentages of CD107a+ (E) and pERK+ (F) NK cells with their reciprocals of EC50 to release IFN-?.GIFN-? release (one representative experiment of total three independent biological experiments) of periphery human NK cells under the stimulation of soluble NKG2D ligands at 100 nM. Data information: Every dot in (C) and (D) represents one independent biological experiment. Data are mean ± SEM for biological triplicate experiments. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (two-tailed unpaired t-test).

Fig 3: Mechanical force-induced allosteric regulation of NKG2D’s ligand dissociation and discrimination A, BSMD snapshots of NKG2D dissociation with MICA (A) and ULBP3 (B) in the presence of force (directions are indicated by black arrows).CThe force versus extension curves from the simulations shown in (A) and (B). Occurrence of the sudden extension changes are indicated in the shaded area and time points correspond with the snapshots in (A) and (B) are marked.D, EZoomed-in binding interfaces of NKG2D with MICA (D) or ULBP3 (E), corresponding with the configuration (shown as gray dashed box in (A) and (B), respectively).FDistance versus time curves for force-induced H-bond formation between indicated residues within NKG2D-MICA binding interfaces. The dashed red lines represent H-bonds whose distances are < 3.5 Å. Source data are available online for this figure.

Fig 4: Kinetic proofreading model analysis of Force-dependent ligand discrimination and amplification of NKG2D in mediating NK cell’s activation Schematic diagram of NKG2D-ligand combination and dissociation model. NKG2D binds and dissociates with multiple ligands at different initial on-rates (k on-initial) and off-rates (k off). Ubiquitous mechanical forces in vivo regulate the disassociation of NKG2D and ligands.Probabilities of productive signals of different ligands binding with NKG2D when at forces of 0 pN, 5 pN, 10 pN, and 15 pN, respectively.Probabilities of output signals at 300s varied with force-dependent off-rates of the three ligands, MICA, MICB, and ULBP3. Off-rates at different mechanical forces measured by experiments had been marked by specific symbols: pentagram, square, circle, and right-pointing triangle represented off-rates at 0 pN, 5 pN, 10 pN, and 15 pN, respectively.Contour plots showed probabilities of output signals at 300s produced by continuously variable initial on-rates and force-dependent off-rates.

Fig 5: Correlation analysis of NK cell functions with in situ force-dependent kinetics of NKG2D binding with ligands A–CPlots and Pearson correlation analysis of reciprocals of EC50 versus the force-dependent in situ k off at 5 pN (A), 10 pN (B) 15pN (C).DHeatmap of force-dependent in situ affinities A c K a of NKG2D binding with different ligands.E–GPlots and Pearson correlation analysis of reciprocals of EC50 versus the force-dependent effective in situ affinities A c K a at 5 pN (E), 10 pN (F), and 15 pN (G).HComparison of corresponding Pearson coefficients between all NKG2D ligands in situ binding kinetics and ligand-induced NK functions.IDetection range comparison in force-dependent affinity of NKG2D and indicated ligands under difference force. Bars in different colors are the ratios of the affinities of indicated ligands divided by that of ULBP1. Source data are available online for this figure.