Disruption of heme homeostasis by nuclear receptor Nur77 induces pyroptosis through granzyme B-dependent GSDMC cleavage

Abstract

Pyroptosis plays a crucial role in physiological and pathological processes. As melanoma cells are resistant to apoptosis but express gasdermin proteins, it is appealing to counter melanoma with the induction of gasdermin-executed pyroptosis. GSDMC, initially cloned from metastatic melanoma cells, has been demonstrated as a potential executioner of pyroptosis. However, no lead compounds that trigger GSDMC-mediated pyroptosis have been reported, which limits the in-depth investigation of GSDMC functions. Here, we discovered a chemical compound, dodecyl 1H-benzo[d]imidazole-5-carboxylate (DdBIC), that targeted the nuclear receptor Nur77 to induce pyroptosis through cleaving GSDMC by granzyme B in melanoma cells. Upon DdBIC binding, Nur77 was translocated to the mitochondria to activate the hemoprotein SDHA to overconsume succinyl-CoA, subsequently disrupting the homeostasis of heme in the SDH complex and resulting in electron leakage to induce mito-ROS production. This mito-ROS signal was sensed by the mitochondrial protease OMA1 via oxidation, which led to downstream OPA1 cleavage and subsequent released into the cytoplasm. Cytosolic OPA1 activated PERK to induce the integrated stress response (ISR), which further activated granzyme B to cleave GSDMC, culminating in the induction of pyroptosis. Together, this study elucidates a signal cascade from Nur77-impaired homeostasis of heme metabolism to PERK-mediated ISR activation, and reveals a novel paradigm, by which granzyme B, rather than caspases, cleaves GSDMC for pyroptotic induction and provides a new strategy for the therapeutic treatment of melanoma by lead compound DdBIC.

Fig. 1

DdBIC induces GSDMC-mediated pyroptosis by regulating heme levels. Melanoma A375 cells were treated with DdBIC (20 μM) for 1 h to measure the level of succinyl-CoA, for 4 h to measure the level of heme, and for 8 h to assess pyroptotic features, unless otherwise specified. a Top, GSDMC-knockdown cells were treated with different compounds (n = 63) for 8 h, and LDH levels were detected. Bottom, chemical structures of DdBIC, the DdBIC probe (DdBIC-P) and its corresponding negative probe DdBIC-NP. b Cells were treated with DdBIC at the indicated doses to detect pyroptotic morphology and LDH levels. c In GSDMC-knockdown cells, pyroptotic morphology and LDH levels were evaluated. d DdBIC induced GSDMC cleavage in a dose-dependent manner. e Cells were cotreated with hemin (5 μM) and DdBIC to detect pyroptosis. f Cells were incubated with DdBIC-P or DdBIC-NP (100 μM) for 8 h, and pyrotopsis was performed as indicated. g DdBIC-P modified by a click reaction with rhodamine-N3 (DdBIC-P-R) was incubated with cells for 4 h. Mitochondria and nuclei were stained with an anti-Hsp60 antibody and DAPI. h, i Cells were treated with DdBIC, and the levels of cytosolic and mitochondrial heme (h) or succinyl-CoA (i) were shown. The cytosolic and mitochondrial fractions were removed to exclude contamination (h, right). j, k Cells were cotreated with DM-αKG and DdBIC to measure the levels of heme in mitochondria (j) and pyroptosis (k). l, m SUCLG1 was knocked down first in cells, and pyroptosis (l) and the levels of mitochondrial heme and succinyl-CoA (m) were detected. Statistics: two-way ANOVA with Tukey’s test to c, e and j–m; one-way ANOVA with Tukey’s test to b, f and i; unpaired two-tailed Student’s t test to h. P values are shown

Fig. 2

Mito-ROS are elevated by DdBIC through the regulation of the SDH complex in the ETC. Melanoma A375 cells were treated with DdBIC (20 μM) for 4 h to measure the levels of heme and mito-ROS and to observe the morphology of the mitochondria, for 8 h to assess pyroptotic features, unless otherwise specified. a Cells were cotreated with hemin, α-VE (25 μM) or Mito-Q (0.5 μM) and DdBIC as indicated to assay the levels of mito-ROS. b, c Cells were cotreated with α-VE or Mito-Q and DdBIC to observe mitochondrial morphology (b) and detect pyroptosis (c). Mitochondria and nuclei were shown after TOM20 and DAPI staining (b). d Cells were cotreated with DBM (1 mM) or TTFA (100 μM) and DdBIC to detect pyroptotic features and mito-ROS levels. e, f SDHA and SDHC were separately knocked down to detect pyroptosis (e) and the level of mito-ROS; mitochondrial morphology was shown, and mitochondrial swelling was quantified (f). g Left, model of the relationships among SDH, heme and mito-ROS. Right, cells were treated with DdBIC, and the heme level in the SDH complex was measured. h Cells were treated with DdBIC for the indicated times, and the levels of fumarate and CoQH₂ were measured. i, j Cells were cotreated with DMM (7 mM) and DdBIC to assay the levels of mitochondrial heme and mito-ROS (i) and pyroptosis (j). Statistics: two-way ANOVA with Tukey’s test to a–f, i and j; one-way ANOVA with Tukey’s test to h; unpaired two-tailed Student’s t test to g. P values are shown

Fig. 3

OMA1 senses DdBIC-induced mitochondrial ROS signals to cleave OPA1. Melanoma A375 cells were treated with DdBIC (20 μM) for 4 h to detect OMA1 oxidation and OPA1 cleavage, for 8 h to assess pyroptotic features, unless otherwise specified. a Cells were cotreated with α-VE, Mito-Q, hemin, DBM or TTFA and DdBIC as indicated to detect the oxidation of OMA1. b OMA1 was knocked down in cells to detect pyroptosis in response to DdBIC stimulation. c, d The 12 cysteines in OMA1 were all mutated to generate a mutant OMA1^12CS. OMA1 oxidation was detected (c); after reintroducing either OMA1 or OMA1^12CS into OMA1-knockdown cells, pyroptosis was detected (d). e After reintroducing either OMA1 or OMA1^E328Q (the active site of OMA1) into OMA1-knockdown cells, pyroptosis was detected. f DdBIC elevates mito-ROS-driven OMA1 activity. OMA1 activity was measured via a fluorogenic substrate assay in A375 cells treated with DdBIC alone or cotreated with Mito-Q and DdBIC. g Cells were cotreated with α-VE or Mito-Q and DdBIC (left) or with OMA1 knockdown (right) to detect OMA1 activity and OPA1 cleavage. h OPA1 was knocked down in cells to detect pyroptosis. i S-OPA1 was overexpressed in cells to detect pyroptosis in the presence of DdBIC. j, k Cells were treated with DdBIC for the indicated times (j) with or without cotreatment with α-VE, Mito-Q, hemin, DBM or TTFA as indicated (k), and the cytosolic fractions were prepared to measure the levels of S-OPA1. Statistics: two-way ANOVA with Tukey’s test to b, f, h and i; one-way ANOVA with Tukey’s test to d–f. P values are shown

Fig. 4

DdBIC activates the PERK-mediated ISR pathway to induce pyroptosis. Melanoma A375 cells were treated with DdBIC (20 μM) for 5 h to detect the dimerization of PERK and phosphorylation of PERK and eIF2α for 8 h to assess pyroptotic features, unless otherwise specified. a PERK was knocked down in cells, after which pyroptosis was detected. b PERK was knocked down to detect the phosphorylation of eIF2α. c, d eIF2α and eIF2α^S51A were reintroduced into eIF2α-knockdown cells to detect the phosphorylation of eIF2α (c) and pyroptosis (d). e PERK or OPA1 was knocked down first in cells, the dimerization of PERK was determined. f A cytosolic fraction was prepared to detect DdBIC-induced OPA1 dimers through nonreducing SDS-PAGE and BN-PAGE. g In OPA1-overexpressing cells, DdBIC-induced interactions between OPA1 and endogenous PERK were detected in the presence of hemin, α-VE or TTFA, as indicated. h In OPA1-knockdown cells, PERK phosphorylation (upshift band) was detected (top), which was abolished by incubation with CIAP (bottom). i Cells expressing PERK^WT-HA or a phosphorylation-deficient PERK^T982A-HA mutant were treated with DdBIC. PERK was immunoprecipitated (IP) via an anti-HA antibody, and phosphorylation levels were assessed via a Phos-tag assay and immunoblotting with an anti-panphospho-Ser/Thr antibody. j Cells were cotreated with hemin, DBM, TTFA, α-VE or Mito-Q and DdBIC as indicated to determine the phosphorylation levels of PERK and eIF2α. k In OMA1- or OPA1-knockdown cells, the phosphorylation levels of PERK and eIF2α were determined. Statistics: two-way ANOVA with Tukey’s test to a; one-way ANOVA with Tukey’s test to d. P values are shown

Fig. 5

Granzyme B cleaves GSDMC in response to DdBIC stimulation. Melanoma A375 cells were treated with DdBIC (20 μM) for 6 h to determine the level and activity of granzyme B (GZMB), for 7 h to determine the interaction between GZMB and GSDMC, or for 8 h to assess pyroptotic features, unless otherwise specified. a Cells were cotreated with 3,4-DCI (10 μM) for 8 h to detect pyroptosis. b GZMB was transfected into cells, and the GZMB level in the cytosolic fraction was detected in response to DdBIC stimulation. c GZMB and GSDMC were overexpressed for the TurboID-based assessment of the interaction between GZMB and GSDMC. d GZMB was knocked down to assay pyroptosis. e GZMB and GZMB^S183A were transfected into GZMB-knockdown cells to detect pyroptosis. f The cleavage of GSDMC by recombinant GZMB was detected in vitro. g, h Cells were cotreated with ISRIB and DdBIC or with PERK knockdown to determine the mRNA (g) or protein (h) levels of GZMB and GZMB activity. i GSDMC and GSDMC^D276A were reintroduced into GSDMC-knockdown cells to detect pyroptosis. Statistics: two-way ANOVA with Tukey’s test to a, d, g and h; one-way ANOVA with Tukey’s test to e and i. P values are shown

Fig. 6

Nur77 is the target of DdBIC. Melanoma A375 cells were treated with DdBIC (20 μM) for 0.5 h to display Nur77 mitochondrial localization and its interaction with other proteins, for 4 h to measure the levels of heme and mito-ROS, for 8 h to assess pyroptotic features, unless otherwise specified. a Nur77 was knocked down to detect pyroptosis. b DdBIC bound to a pocket formed by helices 5, 8, and 10 of the LBD. The LBD is shown as a light blue cartoon, amino acids near DdBIC are shown as sticks, and hydrogen bonds are shown as yellow dashed lines. c Measuring the binding affinity of DdBIC with either the LBD or LBD^4mt via ITC. The titration data were analyzed via the program MicroCal PEAQ-ITC Analysis Software and fitted with a one-site binding model. The levels of LBD or LBD^4mt were indicated (right). d Cells were treated with DdBIC for the indicated times, and the expression levels of Nur77 were detected. e Cells were treated with DdBIC to observe Nur77 localization. Nur77 and mitochondria were indicated with their corresponding antibodies, and nuclei were shown by DAPI staining. f Cells were transfected with HK2 and Nur77, and their interaction was determined (top). Nur77 was transfected into HK2-knockdown cells, and the mitochondrial localization of Nur77 was determined (bottom). g HK2 was knocked down to detect pyroptosis. h Binding surface of the LBD for HK2. The binding of DdBIC to the LBD induces conformational changes in residues Q528, R563 and E445 (white sticks) from those in the apo structure (purple sticks). i Nur77 and SDHA were overexpressed, and the mitochondrial fraction was prepared to determine the interaction between Nur77 and SDHA. j The levels of mitochondrial heme and mito-ROS were determined in Nur77-knockdown cells. k Nur77 and HK2 were separately knocked down in cells, and OMA1 activity, OPA1 cleavage, phosphorylation of PERK and eIF2α, and GZMB expression levels were analyzed. Statistics: two-way ANOVA with Tukey’s test to a, g and j. P values are shown

Fig. 7

DdBIC inhibits tumor growth by inducing Nur77-mediated and GSDMC-induced pyroptosis in mouse models. For the xenograft model, melanoma A375 cells were injected subcutaneously into the posterior flanks of nude mice (n = 8). When the tumors reached a diameter of approximately 3–4 mm, the mice were intraperitoneally administered DdBIC (20 mg/kg) every other day for 2 weeks and then sacrificed. a Comparison of the sizes, volumes and weights of xenograft tumors treated with different concentrations of DdBIC. b Detection of the cleavage of OPA1 and GSDMC and ISR activation (represented by eIF2α phosphorylation) in xenograft tumor samples from the same tumors as in (a). c Nur77 or GSDMC was separately knocked down in A375 cells. The mice were treated with DdBIC. Images of xenograft tumors and the tumor volumes and weights were shown. d, e Phosphorylation of eIF2α and heme levels were detected in the same tumors as in (c). f Nur77 was knocked down first in A375 cells, and Nur77^WT and Nur77^4mt were reintroduced into the cells. The mice were treated with DdBIC. Images of xenograft tumors and the tumor volumes and weights were shown. g The cleavage of OPA1 and GSDMC, the phosphorylation of eIF2α and the heme level were measured in the same tumors as in (f). h, i GSDMC was knocked down first in A375 cells, and GSDMC^WT and GSDMC^D276A were reintroduced into the cells. The mice were treated with DdBIC. Images of xenograft tumors and the tumor volumes and weights were shown (h). The cleavage of GSDMC in the same tumors as in (h) was also detected (i). j C57BL/6J mice bearing murine melanoma B16 cell-derived allografts were treated intraperitoneally with DdBIC for 2 weeks. Tumors and tumor weights are indicated (n = 8). The heme level, cleavage of OPA1 and GSDMC, and ISR activation were indicated. k B16 cell-derived allograft tumors were collected after DdBIC administration for 24 h, and the proportion and activation status of immune cells within the TME (n = 5) were analyzed. l C57BL/6J mice bearing subcutaneous B16 melanoma cells were treated with i.p. injections of DdBIC and anti-PD-1 antibody every other day. Representative images and weights of tumors at the endpoint were shown (n = 8 mice per group). Statistics: two-way ANOVA with Tukey’s test to c, e and l; one-way ANOVA with Tukey’s test to a; unpaired two-tailed Student’s t test to f, g, h, j and k. P values are shown

Fig. 8

Working model for the DdBIC function. DdBIC directly targets the nuclear receptor Nur77. With the assistance of HK2, Nur77 is translocated into the mitochondria to negatively regulate the heme level by increasing the activity of SDHA, which causes leakage of electrons, leading to increased mito-ROS. The mitochondrial protease OMA1 senses this ROS signal via oxidation and then cleaves its downstream OPA1, which converts OPA1 into S-OPA1 that subsequently releases into the cytoplasm. Cytosolic S-OPA1 activates PERK to cause an integrated stress response (ISR) and then activates granzyme B activity, which further cleaves GSDMC and induces pyroptosis. Figure 8, created using BioRender.com