Fig. 1 |. Constructing thermal-specific gene switches.

a, Schematic of a panel of six thermal gene-switch constructs consisting of two to seven HSEs upstream of the HSPB1 core promoter (labelled 2H-B1 to 7H-B1). Uppercase base pairs within HSE were conserved while base pairs indicated as n were randomized. b,c Gluc reporter expression by Jurkat T cells following heating as a function of HSE number (b) or temperature (c). NS, not significant; *P = 0.0248 between 2H-B1 and 4H-B1 at 42 °C; **P = 0.0037 between 5H-B1 and 7H-B1 at 42 °C; two-way analysis of variance (ANOVA) and Tukey post-test and correction; mean ± s.e.m. is depicted; n = 3 biologically independent wells. Three independent experiments were performed with similar results. d, Kinetics of Gluc reporter expression by primary human T cells following heat treatments at indicated temperatures. ****P < 0.0001; two-way ANOVA and Tukey post-test and correction; mean ± s.e.m. is depicted; n = 3 biologically independent wells. Two independent experiments were performed with similar results. e, Activity of thermal gene switches containing different core-promoter constructs following heat treatments in primary human T cells. *P = 0.0140 between HSPA6 and untransduced (UTD) groups at 37 °C; one-way ANOVA and Tukey post-test and correction; mean ± s.e.m. is depicted; n = 3 biologically independent wells. Two independent experiments were performed with similar results. f, Gluc expression in primary human T cells that were incubated 24 h at the displayed temperatures. Two-way ANOVA and Tukey post-test and correction; mean ± s.e.m. is depicted; n = 3 biologically independent wells. g, Activity of 7H-YB in primary human T cells from multiple human donors following 30 min heat treatment at the indicated temperatures. ****P < 0.0001; two-tailed t-test; mean ± s.e.m. is depicted; n = 3 biologically independent wells. h, Activity of 7H-YB compared with endogenous HSP70 and HSPA6 promoters in primary human T cells following exposure to CoCl₂ to mimic hypoxia or i, to CdCl₂ to model heavy-metal toxicity. *P = 0.011; ****P < 0.0001; two-way ANOVA and Tukey post-test and correction; mean ± s.e.m. is depicted; n = 3 biologically independent wells.