Noncanonical Roles of Caspase-4 and Caspase-5 in Heme-Driven IL-1β Release and Cell Death

Abstract

Excessive release of heme from RBCs is a key pathophysiological feature of several disease states, including bacterial sepsis, malaria, and sickle cell disease. This hemolysis results in an increased level of free heme that has been implicated in the inflammatory activation of monocytes, macrophages, and the endothelium. In this study, we show that extracellular heme engages the human inflammatory caspases, caspase-1, caspase-4, and caspase-5, resulting in the release of IL-1β. Heme-induced IL-1β release was further increased in macrophages from patients with sickle cell disease. In human primary macrophages, heme activated caspase-1 in an inflammasome-dependent manner, but heme-induced activation of caspase-4 and caspase-5 was independent of canonical inflammasomes. Furthermore, we show that both caspase-4 and caspase-5 are essential for heme-induced IL-1β release, whereas caspase-4 is the primary contributor to heme-induced cell death. Together, we have identified that extracellular heme is a damage-associated molecular pattern that can engage canonical and noncanonical inflammasome activation as a key mediator of inflammation in macrophages.

Figure 1:. Heme induces IL-1β release that is increased in SCD patients.

(A) CD14+ monocytes were isolated from 5 healthy donors and differentiated into macrophages using GM-CSF for 7 days. When fully mature, cells were primed with or without LPS (100 ng/mL) for 3 h, washed and treated with or without heme (50 μM) in 0.1% FBS. Mature IL-1β levels were measured in cellular supernatants by ELISA at the indicated times. Error bars represent standard deviation of 5 independent experiments. ***p<0.001, ****p<0.0001 calculated by 1-way ANOVA. (B) GM-CSF-differentiated human macrophages from healthy donors were primed with or without LPS for 3 h (100 ng/mL), followed by treatment with or without heme (50 μM) in 0.1, 1, 5, or 10% FBS. After 20 h, IL-1β concentration was measured in cultured supernatants by ELISA. Error bars represent standard deviation of 4 independent biological replicates. (C) GM-CSF-differentiated human macrophages were isolated from healthy donors (control) or patients with sickle cell disease (SCD) and primed with or without LPS (100 ng/mL) for 3 h followed by treatment with or without heme (50 μM) in 0.1% FBS. IL-1β concentration was measured in cultured supernatants at 20 h by ELISA. Error bars represent standard deviation of 4 control and 4 SCD samples across 4 independent experiments. **p<0.01 calculated by 1-way ANOVA. (D) CD14+ monocytes were isolated from 3-5 healthy donors and differentiated in GM-CSF for 7 days. On day 7, macrophages were polarized either into the M2-like phenotype with IL-4 (50 ng/mL) or into the M1-like phenotype with GM-CSF for an additional 24 h. On day 8, both groups were washed and primed with LPS (100 ng/mL) for 3 h followed by treatment with or without heme (50 μM) in 0.1% FBS or ATP (5 mM) in 10% FBS. IL-1β concentration was measured in cultured supernatants at 20 h by ELISA. Error bars represent standard deviation of 3-5 independent experiments. *p<0.05 calculated by Student’s t test. (E) Representative phase-contrast images of M1 and M2 macrophages from (D) are shown. Scale bar represents 100 μm.

Figure 2:. Heme activates the inflammatory caspases.

GM-CSF-differentiated human macrophages isolated from healthy donors were transfected with the C1-Pro VC (300 ng) and C1-Pro VN (300 ng) (A); C4-Pro VC (500 ng) and of C4-Pro VN (500 ng) (B); or C5-Pro VC (1000 ng) and C5-Pro VN (1000 ng) (C); along with dsRedmito (50 ng) as a reporter for transfection. 24 h after transfection, cells were treated with or without LPS (100 ng/mL) for 3 h followed by treatment with or without heme (50 μM) in 0.1% FBS. After 1 h, FBS was reconstituted to 5% to inhibit extracellular heme. Cells were assessed for the percentage of dsRed-positive transfected cells that were Venus-positive at 20 h, determined from a minimum of 300 cells per well. Results are represented as percent Venus-positive cells over background (untreated cells). Error bars represent standard deviation of four independent experiments. *p< 0.05; **p< 0.01; calculated by Student’s t test. (D) Representative images show caspase BiFC in green and mitochondria in red. Scale bar represents 10 μm.

Figure 3:. Heme-induced caspase-5 activation and expression is reduced in M2 macrophages.

(A) GM-CSF-differentiated human macrophages isolated from healthy donors were polarized into M1 or M2 macrophages and transfected with the C1-Pro BiFC pair (300 ng of each), the C4-Pro BiFC pair (500 ng of each), or the C5-Pro BiFC pair (1000 ng of each) along with dsRedmito (50 ng) as a reporter for transfection. 24 h after transfection, cells were treated with or without heme (50 μM) in 0.1% FBS. After 1 h, FBS was reconstituted to 5% to inhibit extracellular heme. Cells were assessed for the percentage of dsRed-positive transfected cells that were Venus-positive at 20 h, determined from a minimum of 300 cells per well. Error bars represent standard deviation of three independent experiments. ***p=0.001; calculated by Student’s t test. (B) Cells were polarized to M1 or M2 macrophages as in (A) and treated with or without LPS (100 ng/mL) for 3 h followed by heme (50 μM) in 0.1% FBS. After 1 h, FBS was reconstituted to 5% to inhibit extracellular heme and 20 h later cell lysates were immunoblotted for caspase-1, caspase-4, caspase-5, or actin as a loading control. (C) Cells were polarized to M1 or M2 macrophages and transfected with the C5-Pro BiFC pair as in (A). Transfected hMDMs were treated with or without LPS (100 ng/mL) for 3 h followed by heme (50 μM) in 0.1% FBS. After 1 h, FBS was reconstituted to 5% to inhibit extracellular heme and 20 h later cells were assessed for the percentage of dsRed-positive transfected cells that were Venus-positive, determined from a minimum of 300 cells per well. Error bars represent standard deviation of three independent experiments. *p<0.05; calculated by Student’s t test.

Figure 4:. Heme activates caspase-4 and caspase-5 independently of canonical inflammasome interactions.

(A) GM-CSF-differentiated human macrophages isolated from healthy donors were transfected with the C1-Pro BiFC pair (300 ng of each), the D59R mutant C1-Pro BiFC pair (300 ng of each), the C4-Pro BiFC pair (500 ng of each), the D59R mutant C4-Pro BiFC pair (500 ng of each), the C5-Pro BiFC pair (1000 ng of each), or the D117R mutant C5-Pro BiFC pair (1000 ng of each) along with dsRedmito (50 ng) as a reporter for transfection. 24 h after transfection, cells were treated with or without heme (50 μM) in 0.1% FBS. After 1 h, FBS was reconstituted to 5% to inhibit extracellular heme. Cells were assessed for the percentage of dsRed-positive transfected cells that were Venus-positive at 20 h, determined from a minimum of 300 cells per well. Error bars represent standard deviation of four independent experiments. *p<0.05; calculated by Student’s t test. (B) GM-CSF-differentiated human macrophages isolated from healthy donors were transfected with the C1-Pro BiFC pair (300 ng of each), the C4-Pro BiFC pair (500 ng of each), or the C5-Pro BiFC pair (1000 ng of each), along with dsRedmito (50 ng) as a reporter for transfection with either siRNA against NLRP3 or a control siRNA (7.5 pmol). 24 h after transfection, cells were treated with or without heme (50 μM) in 0.1% FBS. After 1 h, FBS was reconstituted to 5% to inhibit extracellular heme. Cells were assessed for the percentage of dsRed-positive transfected cells that were Venus-positive at 20 h, determined from a minimum of 300 cells per well. Error bars represent standard deviation of three independent experiments. ***p<0.001; calculated by Student’s t test. (C) GM-CSF-differentiated human macrophages isolated from healthy donors were transfected with the C1-Pro BiFC pair (300 ng of each), the C4-Pro BiFC pair (500 ng of each), or the C5-Pro BiFC pair (1000 ng of each), along with dsRedmito (50 ng) as a reporter for transfection with either siRNA against ASC or a control siRNA (7.5 pmol). 24 h after transfection, cells were treated with or without heme (50 μM) in 0.1% FBS. After 1 h, FBS was reconstituted to 5% to inhibit extracellular heme. Cells were assessed for the percentage of dsRed-positive transfected cells that were Venus-positive at 20 h, determined from a minimum of 300 cells per well. Error bars represent standard deviation of three independent experiments. **p<0.01; calculated by Student’s t test.

Figure 5:. Heme-induced IL-1β release requires caspase-4 and caspase-5.

(A) THP-1 cells (WT), control sgRNA THP-1 cells (Ctrl) or THP-1 deficient in caspase-1 (ΔC1, generated by CRISPR/Cas9), were treated with PMA (10 ng/mL) for one day and allowed to recover for an additional day followed by incubation with IFNγ (20 ng/mL) and LPS (10 pg/mL) for 20 h to polarize them into an M1 phenotype. Cells were treated with or without LPS (100 pg/mL) for 3 h followed by heme (50 μM) in 0.1% FBS. After 1 h, FBS was reconstituted to 5% to inhibit extracellular heme. IL-1β concentration was measured in cultured supernatants by ELISA at 20 h. Error bars represent standard deviation of three independent experiments. *p<0.05; **p<0.01; calculated by Student’s t test. Significance of ΔC1 group is compared to Ctrl. (B) THP-1 cells (WT) and THP-1 cells deficient in caspase-4 (ΔC4, generated by CRISPR/Cas9), or caspase-5 (ΔC5, generated by CRISPR/Cas9) were treated and analyzed as in (A). Error bars represent standard deviation of 4 independent experiments. *p<0.05; calculated by Student’s t test. (C) Unstimulated or M1 polarized THP-1 cells (M1) of the indicated genotypes were treated with or without LPS (100 pg/mL) for 3 h followed by treatment with or without heme (50 μM) for 1 h in 0.1% FBS. 20 h later, cell lysates and culture supernatants were immunoblotted for the indicated proteins or actin as a loading control. FL, full length GSDMD; N, GSDMD-N-term; pro, pro-IL-1β; p17, cleaved IL-1β; light, short exposure; dark, long exposure; *, non-specific band. Lane numbers are indicated below.

Figure 6.. Caspase-4 contributes to heme-induced cell death.

(A) THP-1 cells were treated with or without LPS (100 ng/mL) for 3 h in the presence or absence of qVD-OPH (5 μM) followed by treatment with or without heme (50 μM) for 1 h in 0.1% FBS. After 1 h, FBS was reconstituted to 5% to inhibit extracellular heme. Cell death was assessed by flow cytometry for 7-AAD uptake 20 h later. Error bars represent standard deviation of 3 independent experiments. *p<0.05; **p<0.01; ***p<0.001 calculated by Student’s t test. (B-C) THP-1 cells or THP-1 cells deficient in the indicated caspases were treated with or without heme (50 μM) as in (A). Cell death was assessed by flow cytometry for 7-AAD uptake at 6 h (B) or 20 h (C). Error bars represent standard deviation of 3–4 independent experiments. *p<0.5, **p<0.01; calculated by Student’s t test. (D) PMA-primed THP-1 cells stably expressing the C1-Pro BiFC pair were treated with heme in the presence of qVD-OPH (5 μM) to prevent cells from lifting off due to apoptosis. Images were taken by confocal microscopy every 5 min for 24 h. Frames from the time-lapse show representative cells undergoing BiFC (green) prior to cell lysis as measured by the loss mCherry (red). Scale bars represent 5 μm. (E) Graph of the cells from (D) that became Venus-positive is shown. Each point on the mCherry graph (red) is scaled and aligned to each point on the caspase-1 BiFC graph (green) that represents the average intensity of mCherry or Venus in the cell at 5 min intervals where time=0 is the point of onset of mCherry loss, representing cell lysis. Arrow shows the point of onset of caspase-1 BiFC immediately prior to cell lysis. Error bars represent SEM of 9 individual cells.