Apoptosis inhibitor 5 is an endogenous inhibitor of caspase-2

Abstract

Caspases are key enzymes responsible for mediating apoptotic cell death. Across species, caspase-2 is the most conserved caspase and stands out due to unique features. Apart from cell death, caspase-2 also regulates autophagy, genomic stability and ageing. Caspase-2 requires dimerization for its activation which is primarily accomplished by recruitment to high molecular weight protein complexes in cells. Here, we demonstrate that apoptosis inhibitor 5 (API5/AAC11) is an endogenous and direct inhibitor of caspase-2. API5 protein directly binds to the caspase recruitment domain (CARD) of caspase-2 and impedes dimerization and activation of caspase-2. Interestingly, recombinant API5 directly inhibits full length but not processed caspase-2. Depletion of endogenous API5 leads to an increase in caspase-2 dimerization and activation. Consistently, loss of API5 sensitizes cells to caspase-2-dependent apoptotic cell death. These results establish API5/AAC-11 as a direct inhibitor of caspase-2 and shed further light onto mechanisms driving the activation of this poorly understood caspase.

Keywords: apoptosis; apoptosis inhibitor 5; caspase‐2; cell death; pore‐forming toxins.

Figure 1. Apoptosis inhibitor 5 (API5) is identified in caspase‐2‐containing protein complexes

1. Mass spectrometry analysis of active caspase‐2 complexes. HeLa cells were pre‐incubated with biotin‐VAD‐fmk (B‐VAD, 50 μM) for 1 h and the cells were subsequently treated with 300 ng/ml of α‐toxin as mentioned in the Materials and Methods. The cells were then harvested and the active caspase‐2 complexes were precipitated by streptavidin agarose beads. The entire sample was subjected to trypsin digestion, and the proteins were identified by mass spectrometry. Shown is the MS/MS spectra of one of the API5 peptides identified.

2. The presence of API5 in caspase‐2 complexes was verified by immunoblot analysis.

3. Gel filtration analysis of control and PFT‐treated HeLa cell lysates. HeLa cells were treated with α‐toxin (150 ng/ml) for 4 h. Then, the cells were harvested, lysed and the crude protein extract was prepared for gel filtration analysis. The proteins were separated by size‐exclusion chromatography as detailed in the Materials and Methods section. The proteins from each collected fraction were precipitated, and the presence of proteins of interest was tested by Western blot analysis. The individual fractions are indicated.

Source data are available online for this figure.

Figure 2. Depletion of API5 sensitizes cells to PFT‐induced apoptosis

1. API5 is degraded during PFT‐mediated apoptosis. HeLa cells were treated with α‐toxin (300 ng/ml for 24 h) and the presence of API5, caspase‐2, and PARP was tested by immunoblots. Ponceau‐stained membrane is presented below to verify the loading (* indicates processed caspase‐2 and PARP).

2. API5 depletion sensitizes HeLa cells to PFT‐mediated apoptosis. ShControl and shAPI5 HeLa cells were treated with α‐toxin (600 ng/ml for 24 h), and the percentage of cell death was analysed by FACS. The Annexin V/PI staining pattern of control and toxin‐treated cells from a representative experiment is presented.

3. Quantification of experiments presented in (B) (n = 3, Mann–Whitney test, *P‐value = 0.0286), error bars represent ±SD of the mean.

4. HeLa cells were transfected with siRNAs (siControl or siAPI5#4) for 1 day prior to α‐toxin treatment. The cells were harvested for Western blot analysis at different time points (as indicated). FL: full length, *: processed PARP.

5. HeLa cells were treated for 24 h with α‐toxin (300 ng/ml), and the dead cells were washed away. The surviving cells were allowed to replicate for 48 h to check for clonogenic survival. Shown are data from a representative experiment. Scale bar, 250 μm.

6. For testing the viability of the cells, the crystal violet assay was performed on both control and shAPI5 treated with toxin and the surviving cells were quantified. The error bars represent the mean ± SD (n = 3, Mann–Whitney test, ***P < 0.0001).

7. The efficiency of the shAPI5 was verified by Western blot and vinculin was employed as a loading control.

Source data are available online for this figure.

Figure EV1. Depletion of API5 sensitizes cells to PFT‐mediated apoptosis

1. A, B

   HeLa cells were transfected with either control or API5 siRNAs for 1 day and then treated with 300 ng/ml of PFT for 24 h. The cells were harvested and the dead cells were measured by FACS analysis after Annexin V/PI staining as detailed in the Materials and Methods. Shown in (B) are data from three (n = 3; left panel) or four (n = 4, middle panel) independent experiments. The dead cells include Annexin V‐positive early apoptotic as well as Annexin V/PI double‐positive cells, indicating the late apoptotic/secondary necrotic populations as analysed by flow cytometry. The efficiency of the knockdown with siRNA#1 was monitored by immunoblots (B, right panel).

2. C

   Microscopy analysis of API‐5 depleted cells upon α‐toxin treatment. HeLa cells were transfected with siRNA and treated with PFT as mentioned before. The cells were treated with in situ caspase‐3/7 substrate (green) for 30 min as mentioned in the Materials and Methods. The images were acquired after 6 h post‐toxin treatment. Scale bar, 300 μm.

Source data are available online for this figure.

Figure 3. Depletion of API5 sensitizes HeLa cells to PFT, but not to other inducers of apoptosis

1. A, B

   ShControl and shAPI5 cells were treated for 24 h with α‐toxin (150 ng/ml), staurosporine (STS, 125 nM), TNF‐α (20 ng/ml)/CHX, camptothecin (CPT, 4 μM), etoposide (ETO, 50 μM), cisplatin (Cis, 40 μM) or brefeldin A (BrefA, 5 μM). Cells were harvested and labelled with propidium iodide for cell death assay (B) or lysed for Western blot analysis (A) (* indicates processed caspase‐2 and PARP). The Ponceau staining of the entire membrane is shown below. For (B) n = 3, two‐way ANOVA with a Bonferroni test, ***P‐value < 0.001.

Source data are available online for this figure.

Figure EV2. Depletion of API5 sensitizes NCI‐H1650 cells to PFT‐mediated apoptosis

1. A–C

   NCI‐H1650 cells (lung adenocarcinoma) cells were transfected with control or API5 siRNAs employing Saint‐Red reagent. After 24 h, the cells were treated with α‐toxin (150 ng/ml), TNF‐α (20 ng/ml)/CHX, camptothecin (4 μM) or etoposide (50 μM) for 24 h. Cells were harvested and labelled with propidium iodide for analysing cell death by (A, B) FACS analysis (n = 3 in A, n = 1 in B) or for (C) Western blot analysis.

Source data are available online for this figure.

Figure 4. Depletion of API5 enhances caspase‐2 activation and caspase‐2‐dependent cell death

1. Depletion of API5 enhances caspase‐2 dimerization and activation. ShControl and shAPI5 HeLa cells were incubated with biotin‐VAD‐fmk (B‐VAD, 50 μM) 1 h prior to α‐toxin (300 ng/ml) treatment and incubated for 18 h. The active caspase‐2 complexes were precipitated with streptavidin–agarose beads as mentioned in the Materials and Methods section. The samples were then tested for the presence of active caspase‐2 and API5 by immunoblot analysis. The relative intensity of caspase‐2 bands was measured by ImageJ software and is indicated below the bands.

2. HeLa cells were transfected with siRNAs, and 24 h later, they were challenged with α‐toxin at different concentrations and the processing of caspase‐2 was monitored by immunoblots. FL: full‐length caspase‐2.

3. Microscopy analysis of API5‐ and caspase‐2‐depleted cells upon α‐toxin treatment. ShControl and shCaspase‐2 HeLa cells were transfected with siRNAs and 24 h later, the cells were pre‐incubated with fluorescent Magic Red Caspase substrate and green fluorescent YOYO‐1 for 30 min following the manufacturers' instructions. The cells were treated with α‐toxin (300 ng/ml) for 24 h. The images were acquired at 12 h post‐treatment in the IncuCyte imaging system. Scale bar, 300 μm.

4. ShControl and shCaspase‐2 HeLa cells were transfected with siRNAs, and 1 day later, the cells were challenged with α‐toxin (300 ng/ml). The samples were subjected for Annexin V/PI measurements by flow cytometer as mentioned in the Materials and Methods. The error bars represent the mean ± SD (n = 5), **P‐value = 0.00897 (Student's t‐test).

Source data are available online for this figure.

Figure EV3. Depletion of API5 enhances caspase‐2 activation and processing

1. Measuring caspase‐2 activity. ShControl and shCaspase‐2 HeLa cells were transfected with siRNAs, and 24 h later, they were challenged with α‐toxin (300 ng/ml); 24 h post‐treatment, the samples were subjected to in vitro caspase‐2 activity measurement as indicated in the Materials and Methods and following manufacturer's instructions.

2. The cells were treated as above and 24 h later were subjected to Western blot analysis. FL: full length, *: processed form.

Source data are available online for this figure.

Figure EV4. Caspase‐2 is activated in response to potassium ion depletion in a caspase‐3/7‐independent manner

Wild‐type or caspase‐3/7 double KO MEFs were treated with valinomycin (30 μM) for 24 h. Cells were then collected and treated with FAM‐VDVAD‐FMK‐FLICA reagent. Caspase‐2 activity was measured by FACS analysis following the manufacturer's protocol (ImmunoChemistry). Shown are data from a single representative experiment.

Figure 5. API5 directly interacts with the CARD domain and prevents caspase‐2 dimerization

1. API5 directly interacts with caspase‐2. In vitro‐translated Flag‐tagged caspase‐2 was incubated with recombinant API5 and its truncated versions for 2 h and caspase‐2 was pulled down with Flag‐M2 beads. The samples were eluted and subjected to Western blot analysis to check for binding between caspase‐2 and API5. 11B4 refers to the antibody clone raised against caspase‐2. The various constructs of API5 and their respective amino acid compositions are indicated above. FL: full length, aa: amino acid.

2. The interaction between recombinant API5 and the CARD domain of caspase‐2 was tested by GST pull‐down experiments as indicated in the Materials and Methods section. The samples were eluted and separated via SDS–PAGE and stained with Coomassie blue. The red arrow indicates the anticipated size of API5. The green arrow shows the anticipated size of caspase‐2–GST–CARD. PD: pull‐down.

3. Human recombinant caspase‐2 was incubated with different concentrations of human recombinant API5 (1× = 100 ng) for 1 h at 37°C and subjected to caspase activity measurement using a fluorescent plate reader as mentioned in the Materials and Methods.

4. In vitro‐translated caspase‐2‐Flag and caspase‐2‐His were co‐incubated with increasing concentrations of in vitro‐translated API5 protein for 30 min and caspase‐2‐Flag was immunoprecipitated. IP Flag: co‐immunoprecipitation, input: total sample before IP, FL: full length.

5. API5 overexpression inhibits caspase‐2 activity; 293T were transected with caspase‐2 with increasing amounts of API5 for 48 h. Cells were then collected and treated with FAM‐VDVAD‐FMK‐FLICA for measuring caspase‐2 activity following manufacturer's instructions (ImmunoChemistry) (n = 2).

6. Schematic view of API5‐mediated inhibition of caspase‐2 dimerization and activation. N: N‐terminus, C: C‐terminus of the protein.

Source data are available online for this figure.

Figure EV5. API5 did not interact with caspase‐9 and caspase‐1

1. HeLa cells were incubated in serum‐free EBSS (SI, starvation‐induced) media or treated with α‐toxin, and the samples were lysed and subjected to caspase‐9 immunoprecipitation (IP) (see Materials and Methods) 24 h post‐treatment. The total lysates and the immunoprecipitated‐eluted samples were analysed by Western blot.

2. 293T cells were transfected with pCMV2‐Flag‐Caspase‐1 or pCMV3‐Flag‐Caspase‐9 for 48 h. The Flag‐tagged caspases were precipitated by anti‐Flag M2 Magnetic Beads. The anti‐Flag M2 Magnetic Beads were washed 5 times and then incubated overnight at 4°C with recombinant API5. After washing (3×), the beads were re‐suspended in SDS buffer and subjected to Western blot analysis.

Source data are available online for this figure.