Fig 2: Exogenous rHsp70 restored Hsp70 expressions in aged mice under sevoflurane anesthesia. Enzyme‐linked immunosorbent assay (ELISA) was used to measure the Hsp70 levels in serum (a) and hippocampus (b). Quantitative reverse transcription polymerase chain reaction (RT‐qPCR) was used to measure the mRNA expression of Hsp70 in the hippocampus (c). Hippocampal Hsp70 protein expression was determined by Western blot analysis (d). Twelve mice were used for each group. For the experiments of RT‐qPCR and Western blotting, the hippocampal homogenate from 12 mice in each group was mixed, and the experiments were repeated for four times. Data were represented by mean ± SD. *p < .05, **p < .01, ***p < .001

Fig 3: Blockade of Nav1.7 regulates chondrocyte biology through enhancing HSP70 and midkine secretion.a, Peptide-spectrum match (PSM)-based abundance of proteins identified from the 30–100 kDa fraction of conditioned medium that are unique to (red dots) or increased with (blue dots) PF-04856264 and ProTx II treatment. b, PSM-based abundance of proteins (red dot) identified from the 10–30 kDa fraction of conditioned medium that are unique to PF-04856264 and ProTx II treatment. c, COL2 and ACAN mRNA levels in human C28I2 chondrocytes stimulated with conditioned medium collected from cells treated with ProTx II (Pro-CM) or PF-04856264 (PF-CM) in the absence or presence of IgG or anti-HSP70 antibodies (n = 4 biological replicates). d, MMP13 and ADAMTS5 mRNA levels in C28I2 chondrocytes stimulated with IL-1β and conditioned medium collected from cells treated with ProTx II (Pro-CM) or PF-04856264 (PF-CM) in the absence or presence of control IgG or anti-midkine antibodies (n = 4 biological replicates). e, Safranin O and Fast Green-stained knee joint sections (n = 8). Ver, VER 155008. Scale bar, 50 µm. f, OARSI score from images represented in e. g, Two-minute travel distance and von Frey testing at the indicated time points with indicated treatments after DMM surgery (n = 8). Data are mean ± 95% confidence interval, P values by two-way ANOVA with Bonferroni post hoc test. *Vehicle versus PF, P < 0.05; **vehicle versus PF, P < 0.01; #PF versus PF + Ver + iMDK, P < 0.05; ##PF versus PF + Ver + iMDK, P < 0.01. c,d,f, Data are mean ± s.d., P values by one way ANOVA with Bonferroni post hoc test.Source Data

Fig 4: HSP70 enhances anabolism and midkine inhibits IL-1β induced catabolism in human chondrocytes.a, b, ELISA quantification of HSP70 in conditioned medium (a) and cell lysate (b) of human C28I2 cells treated with 25 nM ProTx II (Pro), 1 µM PF-04856264 (PF), or 10 µM CBZ for 48 h. c, d, ELISA quantification of Midkine in conditioned medium (c) and cell lysate (d) of human C28I2 cells treated with 25 nM Pro or 1 µM PF for 48 h. e, mRNA levels of COL2 and ACAN in human C28I2 cells treated with serial doses of HSP70 for 24 h. f, mRNA levels of MMP13 and ADAMTS5 in human C28I2 cells treated with serial doses of HSP70 and 10 ng/ml IL-1β for 24 h. g, mRNA levels of COL2 and ACAN in human C28I2 cells treated with serial doses of midkine for 24 h. h, mRNA levels of MMP13 and ADAMTS5 in human C28I2 cells treated with serial doses of midkine and 10 ng/ml IL-1β for 24 h. n = 4 biological replicates; Data are mean ± SD, P values are calculated by. i, j, ELISA quantification of HSP70 (i) and midkine (j) in conditioned medium of primary chondrocytes isolated from Nav1.7flox and Nav1.7chondrocye mice at 12 weeks post DMM surgery (n = 4 biological replicates). k, l, Levels of HSP70 (k) and midkine (l) in mouse sera collected from sham surgery control and from DMM surgery WT mice at 12 weeks after surgery (n = 24), assayed by ELISA. m, o, Levels of HSP70 (m) and midkine (o) in human sera collected from healthy individuals (n = 22) and from patients with OA (n = 165), assayed by ELISA. n, p, Correlation analysis (Pearson R and two-tailed P value) of HSP70 (n) and midkine (p) between matched sera and synovium fluids isolated from OA patients (n = 35). Data are means ± SD, P values are calculated by one way ANOVA with Bonferroni post-hoc test (a-h) and two-tailed unpaired Student’s t-test (i-l, m, o).

Fig 5: Ca2+ signalling in chondrocytes.a,b, F/F0 (a) and area under the curve (AUC) of intracellular Ca2+ (b) in human OA chondrocytes following ATP stimulation, measured by plate reader. ATP present from red arrow. c,d, HSP70 (c) and midkine (d) levels in conditioned medium of C28I2 cells pre-treated with BAPTA-AM, followed by ProTx II or PF-04856264. e, F/F0 of intracellular Ca2+ in KB-R7943 treated C28I2 chondrocytes following ATP stimulation, assayed by confocal fluorescence microscopy. f,g, HSP70 (f) and midkine (g) levels in conditioned medium of C28I2 cells treated with KB-R7943 in the presence or absence of PF-04856264. h, Expression of NCX isoforms in C28I2 cells. i, Knockdown efficiency of NCX1 in C28I2 cells. j,k, F/F0 (j) and AUC of intracellular Ca2+ (k) following ATP stimulation in C28I2 chondrocytes transfected with scramble or NCX1 siRNA measured by plate reader. l, HSP70 and midkine levels in conditioned medium of chondrocytes transfected with scramble or NCX1 siRNA and treated with ProTx II or PF-04856264. m, Model of mechanisms of chondrocyte- and DRG-expressed Nav1.7 in OA, and amelioration of OA and pain via Nav1.7 blockade. b–d,f,g,k,l, Data are mean ± s.d. b,f,g, P values calculated by one way ANOVA with Bonferroni post hoc test. k, Two-tailed unpaired Student’s t-test. c,d,l, Two-way ANOVA with Bonferroni post hoc test. b,f–i,k, n = 3 biological replicates. c,d,l, n = 4 biological replicates.Source Data