Cytoplasmic innate immune sensing by the caspase-4 non-canonical inflammasome promotes cellular senescence

Abstract

Cytoplasmic recognition of microbial lipopolysaccharides (LPS) in human cells is elicited by the caspase-4 and caspase-5 noncanonical inflammasomes, which induce a form of inflammatory cell death termed pyroptosis. Here we show that LPS-mediated activation of caspase-4 also induces a stress response promoting cellular senescence, which is dependent on the caspase-4 substrate gasdermin-D and the tumor suppressor p53. Furthermore, we found that the caspase-4 noncanonical inflammasome is induced and assembled in response to oncogenic RAS signaling during oncogene-induced senescence (OIS). Moreover, targeting caspase-4 expression in OIS showed its critical role in the senescence-associated secretory phenotype and the cell cycle arrest induced in cellular senescence. Finally, we observed that caspase-4 induction occurs in vivo in mouse models of tumor suppression and ageing. Altogether, we are showing that cellular senescence is induced by cytoplasmic LPS recognition by the noncanonical inflammasome and that this pathway is conserved in the cellular response to oncogenic stress.

Fig. 1. LPS-mediated caspase-4 activation induces a senescent phenotype in human primary fibroblasts.

A Schematic representation of LPS transfection experiments shown in B-D and Supplementary Fig. 1 E–G. IMR90 cells were infected with an empty pGIPZ vector (vector) or shRNA targeting either CASP1 (shC1) or CASP4 (shC4) before transfection with 0.1 μg LPS. The acquisition of senescent features after LPS transfection was assessed by immunofluorescence of senescence markers, mRNA expression analysis by RT-qPCR and SA-β-Galactosidase activity. B SA-β-Galactosidase activity was determined in IMR90 cells 4 days after LPS transfection. Representative images for SA-β-Gal activity are shown. Graph bars, error bars, and dots represent respectively the mean ± standard error of the mean (s.e.m.) and the individual values of 4 independent experiments. Statistical analysis was performed using one-way analysis of variance (ANOVA). C BrdU incorporation and p16^INK4a, p21^CIP1, and caspase-4 protein expression levels were measured by immunofluorescence in IMR90 cells 48 h after LPS transfection. Graph bars, error bars and dots represent respectively the mean ± s.e.m. and the individual values of 4 independent experiments. Statistical analysis was performed using two-tailed Student’s t test. D CDKN1A (p21^CIP1) and CDKN2A (p16^INK4a) mRNA relative expression was quantified by RT-qPCR in IMR90 cells 48 h after transfection. Graph bars, error bars and dots represent respectively the mean ± s.e.m. and the individual values of 3 independent experiments. Statistical analysis was performed using one-way analysis of variance (ANOVA). E CASP4 was overexpressed prior to LPS transfection in IMR90 cells. 5 ×105 control (vector) and CASP4 expressing IMR90 cells were transfected with increasing concentrations of LPS (0.1 or 1 μg), and cell proliferation was measured by BrdU incorporation 48 h after transfection. Graph bars, error bars, and dots represent respectively the mean ± s.e.m. and the individual values of 3 independent experiments. Statistical analysis was performed using one-way analysis of variance (ANOVA). F IMR90 cells were treated as in (E) and SA-β-Galactosidase activity was determined 4 days after LPS transfection. Representative images for SA-β-Galactosidase assay are shown. Graph bars, error bars, and dots represent respectively the mean ± s.e.m. and the individual values of 3 independent experiments. Statistical analysis was performed using one-way analysis of variance (ANOVA). G IMR90 cells were infected with an empty pGIPZ vector (vector) or shRNA targeting either CASP4 (shC4), GSDMD (shGSDMD) or TP53 (shP53) and protein expression analysis for Caspase-4, Gasdermin-D, p53, and β-Actin as loading control was performed by western blot. H–J IMR90 cells were transduced with an empty pGIPZ vector (vector) or shRNAs targeting either CASP4 (shC4), GSDMD (shGSDMD) or TP53 (shP53) prior to transfection with 0.1 μg LPS. BrdU incorporation (H) and the levels of p21^CIP1 (I) and p16^INK4a (J) were measured by immunofluorescence in IMR90 cells 48 h after LPS transfection. Graph bars, error bars and dots represent respectively the mean ± s.e.m. and the individual values of 3 independent experiments. Statistical analysis was performed using one-way analysis of variance (ANOVA). ****P < 0.0001, ***P < 0.001, **P < 0.01, and *P < 0.05. ns, not significant. Scale bar = 0.1 mm as indicated.

Fig. 2. LPS-mediated caspase-4 induced senescence is independent of inflammasome priming.

A CASP4, CASP1 or RAS^G12V were overexpressed in IMR90 cells, IL1A and IL1B mRNA relative expression levels were quantified by RT-qPCR. Graph bars, error bars and dots represent respectively the mean ± s.e.m. and the individual values of 3 independent experiments. Statistical analysis was performed using one-way analysis of variance (ANOVA). B Cells were treated as shown in Fig. 1A. CASP1 or CASP4 expression was targeted by shRNA prior to LPS transfection with 0.1 μg LPS. IL1B mRNA relative expression was quantified by RT-qPCR 48 h after LPS transfection. Graph bars, error bars and dots represent respectively the mean ± s.e.m. and the individual values of 3 independent experiments. Statistical analysis was performed using one-way analysis of variance (ANOVA). C–E IMR90 cells were infected with CASP4 or RAS^G12V expression vectors or empty vector (vector) control. After 3 h treatment with Pam2CSK4, cells were transfected with 0.1 μg LPS. IL1B mRNA relative expression (C) and BrdU incorporation (D) were measured by IF and RT-qPCR respectively 48 h after LPS transfection. E SA-β-Galactosidase activity was determined 4 days after LPS transfection. Representative images for SA-β-Galactosidase activity are shown. Graph bars, error bars and dots represent respectively the mean ± s.e.m. and the individual values of 3 independent experiments. Statistical analysis was performed using one-way analysis of variance (ANOVA). ****P < 0.0001, ***P < 0.001, **P < 0.01, and *P < 0.05. Scale bar = 0.1 mm as indicated.

Fig. 3. Caspase-4 mediated regulation of senescence is independent of its catalytical function.

A–C IMR90 cells were infected with wild-type (WT) CASP4, catalytically inactive (C258A) CASP4 or the empty vector (vector). Overexpression of RAS^G12V was used as a positive control for the induction of senescence. A Caspase-4 and IL-1β expression levels were investigated by immunoblotting. β-Actin immunoblot was performed for loading control. B Caspase-4 protein expression levels and BrdU incorporation measured by immunofluorescence, as well as SA-β-Galactosidase activity were assessed 4 days after equal number of cells were seeded. C Relative cell content (left) was quantified 15 days after equal number of cells were seeded; representative images (right) of crystal violet stained cells are shown. Graph bars, error bars and dots represent respectively the mean ± s.e.m. and the individual values of 3 independent experiments. Statistical analysis was performed using one-way analysis of variance (ANOVA). D–G (D) Schematic representation of the experiment shown in D-F and Supplementary Fig. 3G–H. IMR90 cells were infected with wild-type (WT) CASP4, catalytically inactive (C258A) CASP4 or the empty vector (vector) prior to transfection with 1 μg LPS. (E) Cell viability was measured 24 h after LPS transfection, (F) BrdU incorporation was measured by immunofluorescence 48 h after LPS transfection and (G) SA-β-Galactosidase activity was determined 4 days after LPS transfection. Graph bars, error bars and dots represent respectively the mean ± s.e.m. and the individual values of 3 independent experiments. Statistical analysis was performed using two-tailed Student’s t test. ****P < 0.0001, ***P < 0.001, **P < 0.01, and *P < 0.05. ns, not significant.

Fig. 4. The caspase-4 noncanonical inflammasome is activated in oncogene-induced senescence.

A IMR90 cells were infected with RAS^G12V expression vector to induce OIS. BrdU incorporation and SA-β-Galactosidase activity were measured 4 days after equal number of cells were seeded (left). Representative images (right) for SA-β-Galactosidase activity are shown. Graph bars, error bars and dots represent respectively the mean ± s.e.m. and the individual values of 3 independent experiments. Statistical analysis was performed using two-tailed Student’s t test. B RAS^G12V-OIS was induced as in (A) and CASP4 mRNA relative expression was quantified by RT-qPCR. Graph bars, error bars and dots represent respectively the mean ± s.e.m. and the individual values of 3 independent experiments. Statistical analysis was performed using two-tailed Student’s t test. C RAS^G12V-OIS was induced as in (A) and BrdU incorporation and caspase-4 protein levels were measured by immunofluorescence in RAS^G12V-OIS and control cells 4 days after equal number of cells were seeded. Graph bars, error bars and dots represent respectively the mean ± s.e.m. and the individual values of 3 independent experiments. Statistical analysis was performed using two-tailed Student’s t test. D IMR90 cells were infected with a control (ER:STOP) or an ER:RAS vector. Upon addition of 4OHT, ER:RAS cells undergo OIS. OIS can be detected by a reduction in BrdU incorporation (lower left) measured by immunofluorescence 5 days after 4OHT addition and an increase in SA-β-Galactosidase activity (lower right) one week after 4OHT addition. Graph bars, error bars and dots represent respectively the mean ± s.e.m. and the individual values of 3 independent experiments. Statistical analysis was performed using two-tailed Student’s t test. E Time-course experiment of CASP4 mRNA relative expression in OIS. IMR90 ER:STOP and ER:RAS cells were treated with 4OHT and CASP4 mRNA relative expression was quantified by RT-qPCR 0, 2, 4, 6 and 8 days after 4OHT addition. Graph lines and dots represent respectively the mean and the individual values of 3 independent experiments. Statistical analysis was performed using two-tailed Student’s t test. F Time-course experiment of IL1B mRNA relative expression in OIS. IMR90 ER:STOP and ER:RAS cells were treated with 4OHT and IL1B mRNA relative expression was quantified by RT-qPCR 0, 2, 4, 6, and 8 days after 4OHT addition. Graph lines and dots represent respectively the mean and the individual values of 3 independent experiments. Statistical analysis was performed using two-tailed Student’s t test. G IMR90 ER:STOP and ER:RAS were treated or not with 4OHT during eight days. Caspase-4, IL-1β and IL-8 protein levels were analyzed by immunoblotting. H To detect caspase-4 oligomers, IMR90 ER:STOP and ER:RAS cells were treated with 4OHT for five days, then cells were collected and subjected to disuccinimidyl suberate (DSS) crosslinking. After SDS-PAGE separation, DSS-cross-linked samples were probed for caspase-4 by immunoblotting. (I) Caspase-4 activity in OIS was measured using a fluorometric assay. IMR90 ER:STOP and ER:RAS cells were treated with 4OHT and LEVD-AFC cleavage was measured in low serum (0.5% FBS) cultured cells 0, 4, and 8 days after 4OHT addition. Graph bars, error bars and dots represent respectively the mean ± s.e.m. and the individual values of 3 independent experiments. Statistical analysis was performed using two-tailed Student’s t test. ***P < 0.001, **P < 0.01, and *P < 0.05 and ns, not significant. Scale bar = 250 μm as indicated.

Fig. 5. Caspase-4 activation controls the proinflammatory SASP.

A Schematic diagram of the experimental approach. ER:STOP and ER:RAS IMR90 cells were targeted with either control (non-targeting pool, siNTP) or CASP4-targeting pool siRNA (siCASP4). All cells were treated with 4OHT from day 0. RNA was extracted 5 and 8 days after the addition of 4OHT and three independent biological replicates were subjected to transcriptomic analysis. Differentially expressed gene (DEG) analysis was performed and the number of significant upregulated and downregulated genes 5 and 8 days after the addition of 4OHT upon CASP4-targeting in ER:RAS cells is shown. B Normalized Enriched Scores (NES) of a set of 50 curated hallmark gene signatures were calculated based on the DEG analysis performed between control and CASP4-knockdown ER:RAS samples after 5 and 8 days of 4OHT treatment. Gene sets with a false discovery rate (FDR) q-value of ≤ 0.25 at least in one of the timepoints are shown. P-values for each gene set are indicated next to the corresponding bar. C Heatmap of the log2FC values of all 175 genes included in the “INFLAMMATORY RESPONSE” GSEA gene set of control ER:STOP and CASP4-knockdown ER:RAS compared to control ER:RAS after 5 days of 4OHT treatment. The top 25 differentially expressed signature genes in RAS^G12V-OIS are zoomed in. D IL1A and (E) IL1B mRNA relative expression levels were quantified by RT-qPCR after 5 days of 4OHT treatment in ER:STOP and ER:RAS cells transfected with the indicated siRNAs. Graph bars, error bars and dots represent respectively the mean ± s.e.m. and the individual values of 3 independent experiments. Statistical analysis was performed using one-way analysis of variance (ANOVA). F SAA1 and (G) SAA2 mRNA relative expression levels were quantified by RT-qPCR after 8 days of 4OHT treatment in ER:STOP and ER:RAS cells transfected with the indicated siRNAs. Graph bars, error bars and dots represent respectively the mean ± s.e.m. and the individual values of 3 independent experiments. Statistical analysis was performed using one-way analysis of variance (ANOVA). H IMR90 ER:STOP and ER:RAS cells were transfected with control (NTP), CASP1 or CASP4-targeting siRNA and treated with 4OHT or not during 8 days as indicated. Lysates were subjected to immunoblotting analyses with the indicated antibodies. I and J IMR90 ER:STOP and ER:RAS cells transfected with control (NTP), CASP1 or CASP4-targeting siRNA and treated with 4OHT or during 8 days. Full-length IL-1β (I) and mature IL-1β (J) levels were analyzed by immunoblotting and band intensities were quantitated. Graph bars, error bars and dots represent respectively the mean ± s.e.m. and the individual values of 3 independent experiments. Statistical analysis was performed using two-tailed Student’s t test. K Secreted IL-1β was quantified by ELISA in IMR90 ER:STOP and ER:RAS cells were treated or not 8 days with 4OHT as indicated. Graph bars, error bars and dots represent respectively the mean ± s.e.m. and the individual values of 3 independent experiments. Statistical analysis was performed using one-way analysis of variance (ANOVA). L IMR90 ER:STOP and ER:RAS cells were transduced with an empty retroviral vector (vector) or two different shRNAs targeting CASP4 (shCASP4) and treated with 4OHT or not as indicated during 8 days. CASP4, IL6, IL8, and IL1B mRNA relative expression levels were quantified by RT-qPCR. ****P < 0.0001, ***P < 0.001, **P < 0.01, and *P < 0.05.

Fig. 6. Caspase-4 contributes to the arrest in cell proliferation in OIS.

A IMR90 ER:STOP and ER:RAS cells were transfected with control (NTP), two individual CASP4-targeting siRNAs (CASP4-1 and CASP4-2) or a pool of 4 different siRNA sequences targeting CASP4 (CASP4-p), and treated with 4OHT or not as indicated. BrdU incorporation was measured by immunofluorescence 5 days after 4OHT addition. Graph bars, error bars and dots represent respectively the mean ± s.e.m. and the individual values of 3 independent experiments. Statistical analysis was performed using two-tailed Student’s t test. B, C IMR90 cells were transduced with an empty retroviral vector (vector) or shRNAs targeting either CASP4 (shCASP4) or TP53 (shP53). B Lysates were subjected to immunoblotting analyses with the indicated antibodies eight days after 4OHT addition. C On day 0, equal number of cells were subjected to 4OHT treatment. Fifteen days after 4OHT addition, plates were fixed and stained with crystal violet. Crystal violet was extracted and used to quantify cell content. Graph bars, error bars and dots represent respectively the mean ± s.e.m. and the individual values of 4 independent experiments. Statistical analysis was performed using two-tailed Student’s t test. D Related to Fig. 5A. DEG analysis was performed between control ER:STOP and CASP4-knockdown ER:RAS compared to control ER:RAS after 5 days of 4OHT treatment. Heatmap of the log2FC values from the indicated genes. E IMR90 ER:STOP and ER:RAS cells were transfected with control (NTP), CASP1 or CASP4- targeting siRNAs and treated with 4OHT during 4 days. Cell lysates were subjected to immunoblotting analyses with the indicated antibodies. F IMR90 ER:RAS cells were transfected with control (NTP), CASP1 or CASP4-targeting siRNA and mRNA relative expression of the indicated genes was quantified by RT-qPCR after 5 days of 4OHT treatment. Graph bars, error bars and dots represent respectively the mean ± s.e.m. and the individual values of 3 independent experiments. Statistical analysis was performed using one-way analysis of variance (ANOVA). ***P < 0.001, **P < 0.01, *P < 0.05 and ns, not significant.

Fig. 7. Caspase-11 expression is induced in senescence in vivo.

A Immunohistochemistry showing Ki-67 and caspase-11 staining in sections from Pdx-cre WT and Pdx-cre Kras^G12D pancreas (left panels). Black arrows indicate acinar pancreatic cells, white arrows indicate PanIN cells. B Close up images showing PanINs with high and low expression of caspase-11. Quantification of Ki-67 positive cells of total PanIN cells from a total of 11 mice of 7 to 15 weeks of age. PanINs were classified according to the expression of caspase-11 as indicated. The percentage of Ki-67 positive cells was calculated scoring all cells of PanINs classified as high or low caspase-11 expression per mouse as indicated. Scatter plots were generated from total cells from high and low caspase-11 expressing PanINs with individual points representing the mean Ki-67 percentage positivity for each mouse, with horizontal lines representing group mean and s.e.m. Statistics: Mann–Whitney U test. ***p < 0.001. C Analysis of caspase-11 expression was conducted by immunohistochemistry in lung sections from wild type (WT) or nfkb1 knock out mice (nfkb1^-/-) at 9.5 months of age. 10–15 random images were captured per mouse and average percentage positivity calculated for airway epithelial compartments. Scatter plots represent mean percentage positivity for each animal with horizontal line representing group median. Broad-band autofluorescence (an indicator of lipofuscin accumulation) was acquired from paraffin-embedded sections excited at 458 nm with fluorescence emission captured above 475 nm using a fluorescence microscope (Leica DM550B). Fluorescence intensity was analyzed using ImageJ. At least 10 small airways were analyzed per mouse and an average intensity calculated per animal. Scatter plots represent average value per animal with the horizontal line representing group median. Statistics: Mann–Whitney U test. *p < 0.05, **p < 0.01. Representative images of caspase-11 staining by immunohistochemistry in airway epithelial cells from wt and nfkb1-/- mice, captured using x40 objective. D Linear regression graph between representing caspase-11 expression from samples in (C), and p21 expression from the same samples, which were previously published in Hari et al. [14]. E Representative image of Immuno-FISH for γH2A.X (green) and telomeres (red) in lung alveoli cells from 6.5-month and 24-month-old mice captured using X100 oil objective. Arrows point to γH2A.X foci co-localizing with telomeres (TAF). Statistics: Mann–Whitney U test. ***p < 0.001. F Analysis of caspase-11 expression by immunohistochemistry in lung sections of wt mice at 6.5 months of age (Young) and 24 months of age week (Old). Scatter plots were generated from 10–15 random images captured per animal with individual points representing mean percentage positivity for each mouse with horizontal line representing group median. Statistics: Mann–Whitney U test. **p < 0.01. Representative images of caspase-11 staining by immunohistochemistry (positive, brown; negative, blue) in alveolar cells from wt mice 6.5 and 24 months of age, captured using x40 objective. Scale bar = 100 μm and 25 μm as indicated.