Gasdermin D Deficiency Limits the Transition of Atherosclerotic Plaques to an Inflammatory Phenotype in ApoE Knock-Out Mice

Gasdermin D (GSDMD) is the key executor of pyroptotic cell death. Recent studies suggest that GSDMD-mediated pyroptosis is involved in atherosclerotic plaque destabilization. We report that cleaved GSDMD is expressed in macrophage- and smooth muscle cell-rich areas of human plaques. To determine the effects of GSDMD deficiency on atherogenesis, ApoE−/− Gsdmd−/− (n = 16) and ApoE−/− Gsdmd+/+ (n = 18) mice were fed a western-type diet for 16 weeks. Plaque initiation and formation of stable proximal aortic plaques were not altered. However, plaques in the brachiocephalic artery (representing more advanced lesions compared to aortic plaques) of ApoE−/− Gsdmd−/− mice were significantly smaller (115 ± 18 vs. 186 ± 16 × 103 µm2, p = 0.006) and showed features of increased stability, such as decreased necrotic core area (19 ± 4 vs. 37 ± 7 × 103 µm2, p = 0.03) and increased αSMA/MAC3 ratio (1.6 ± 0.3 vs. 0.7 ± 0.1, p = 0.01), which was also observed in proximal aortic plaques. Interestingly, a significant increase in TUNEL positive cells was observed in brachiocephalic artery plaques from ApoE−/− Gsdmd−/− mice (141 ± 25 vs. 62 ± 8 cells/mm2, p = 0.005), indicating a switch to apoptosis. This switch from pyroptosis to apoptosis was also observed in vitro in Gsdmd−/− macrophages. In conclusion, targeting GSDMD appears to be a promising approach for limiting the transition to an inflammatory, vulnerable plaque phenotype.


Introduction
Vulnerable atherosclerotic plaques are characterized by a large necrotic core formed by excessive necrotic cell death and inflammation [1]. Plaque cells can undergo different types of regulated necrosis although their significance in atherosclerosis is not always clear-cut. One of the best-defined forms of regulated necrosis is pyroptosis, a pro-inflammatory form of regulated cell death that is characterized by the formation of plasma membrane pores via members of the gasdermin (GSDM) protein family [2,3]. Six members of this family have been identified, including gasdermin D (GSDMD). GSDMD is N-terminally (NT) cleaved and is activated by caspase 1 and caspase 4/5 (homologous to caspase 11 in mouse) [3][4][5]. Subsequently, NT-GSDMDs oligomerize, translocate to the cell membrane and induce pore formation, which allows the release of cellular content and pro-inflammatory cytokines such as IL-1β and IL-18, and finally results in membrane disruption and cell lysis [6,7].
GSDMD is the common executor of both canonical pyroptosis (mediated by NLRP3 and other inflammasomes) and non-canonical pyroptosis. Moreover, GSDMD is required for IL-1β release, not only during pyroptosis but also in viable macrophages [7,31]. Interestingly, Gsdmd mRNA is upregulated in peripheral blood monocytes from patients with coronary artery disease and expression of GSDMD and NT-GSDMD is increased in ApoE −/− mice fed a high fat diet as compared to chow-fed controls [30]. These experiments indicate that GSDMD is actively involved in pyroptosis during atherogenesis in both humans and mice, and thus represents a promising target in plaques for inhibiting pyroptosis and inflammation. Genetic deletion of Gsdmd or pharmacological inhibition with necrosulfonamide reduces infarct size and heart failure in a mouse model of acute myocardial infarction [32], underlining the involvement of GSDMD in cardiovascular disease and the possibility for using it as a pharmacological target in atherosclerosis. Therefore, we aimed to evaluate the impact of Gsdmd gene deletion in atherogenesis. First, the effects of Gsdmd gene deletion were evaluated in macrophages and smooth muscle cells in vitro. We also analyzed the expression of cleaved GSDMD in human carotid lesions. Next, the effect of Gsdmd deletion on advanced atherogenesis was evaluated in ApoE −/− mice.

Mice
Standard ApoE −/− mice (Jackson Laboratory, 002052, Bar Harbor, ME, USA) were crossbred with Gsdmd −/− mice (Genentech, South San Francisco, CA, USA), carrying a 1 bp insertion in the Gsdmd coding sequence (GAGTGATGTTGTCAGGCATGGGA becomes GAGTGATGTTtGTCAGGCATGGGA) created with CRISPR/Cas9. Litters were screened for the Gsdmd −/− genotype by PCR analysis using Gsdmd-specific primers (forward primer: GTTTCTTGTCGATGGGAACATTCAG, reverse primer: TGAGTCACACGCAGTATA) followed by Sanger sequencing using the reverse primer. Genotyping of the ApoE alleles was performed by PCR according to the instructions from the Jackson Laboratory. Thereafter, ApoE −/− Gsdmd −/− mice and ApoE −/− Gsdmd +/+ controls (all females, 6-8 weeks old) were fed a western-type diet (WD; TD.88137 supplemented with 21% fat and 0.2% cholesterol, Envigo, Indianapolis, IN, USA) to induce plaque formation. The animals were housed in a temperature-controlled room with a 12 h light/dark cycle and had free access to water and food. After 16 weeks WD, an overdose of sodium pentobarbital (250 mg/kg, i.p.) was administered and blood samples were collected via the retro-orbital plexus. Plasma levels of total cholesterol were measured using a commercially available kit (Randox laboratories, Crumlin, UK). Blood leukocyte subsets were analyzed on a BD accuri C6 flow cytometer as previously described [34]. All experiments were approved by the Ethical Committee of the University of Antwerp (Code 2019-24) and carried out in accordance with European Directive 2010/63/EEC.

Histological Analyses
The thoracic aorta was stained en face with Oil Red O to determine lipid burden. The heart, brachiocephalic artery, and proximal ascending aorta of ApoE −/− Gsdmd −/− and ApoE −/− Gsdmd +/+ mice were fixed in 4% formaldehyde (pH 7.4) for 24 h, dehydrated overnight in 60% isopropanol, and subsequently embedded in paraffin. The proximal ascending aorta was marked on the distal arch end and the brachiocephalic artery on the distal carotid end to ensure that they were always cut on the proximal side. Serial crosssections (5 µm) of the proximal parts of the brachiocephalic artery, proximal aorta and aortic root were prepared in random for histological analyses. Atherosclerotic plaque size and necrotic core area (defined as acellular areas with a threshold of 3000 µm 2 ) were analyzed on hematoxylin-eosin (H&E) stained sections. Total collagen content was measured on Sirius red stained sections. Apoptosis was analyzed using an ApopTag Plus Peroxidase In Situ Apoptosis kit (Millipore, S7101, Burlington, VT, USA). For immunohistochemistry, the following antibodies were used: anti-MAC3 (BD Pharmingen, 550292, San Diego, CA, USA) and anti-α-smooth muscle actin (αSMA, 12547, Sigma-Aldrich, St. Louis, MO, USA). Images were acquired with an Olympus BX43 microscope, which was calibrated for each magnification. Plaque size was measured based on pixels per µm, which was determined during the calibration of the microscope. Per mouse, one section was analyzed. Plaques were manually delineated in Image J software (National Institutes of Health) to establish the region of interest (ROI). Further analyses within the ROIs were performed using color thresholding or manual counting (apoptotic cells).

Flow Cytometry
Apoptosis was quantified with TdT-mediated dUTP-X nick end labeling (TUNEL). Briefly, BMDMs were detached with 0.25% trypsin-EDTA (Thermo Fischer Scientific, 25200072, Waltham, MA, USA). Detached BMDMs were fixed in 4% paraformaldehyde for 1 h at room temperature. After washing with PBS, BMDMs were permeabilized with 0.1% Triton X-100 in 0.1% sodium citrate solution for 2 min on ice. BMDMs were washed again and incubated with TUNEL reaction mixture using an in situ cell death detection kit (fluorescein, 11684795910, Roche, Switzerland) for 1 h at 37 • C. The samples were washed twice with FACS buffer (PBS with 0.1% bovine serum albumin and 0.05% sodium azide) and measured in the FL-1 channel on a BD Accuri C6 flow cytometer. At least 10,000 cells were measured. Debris was always gated out based on FSC/SSC scatter. Positive controls consisted of BMDMs treated with TNFα combined with cycloheximide. Negative controls consisted of untreated BMDMs. Unstained controls were included to exclude background signal.

Western Blotting
Tissues were homogenized in RIPA buffer containing protease and phosphatase inhibitors. Protein concentrations were determined using the BCA method. Samples were then 1:1 diluted in Laemmli sample buffer (Bio-Rad, Hercules, CA, USA) containing 5% β-mercaptoethanol (Sigma-Aldrich) and heat-denatured for 5 min at 100 • C. Samples were loaded on Bolt 4-12% Bis-Tris gels (Invitrogen) and after electrophoresis transferred to Immobilon-FL PVDF membranes (Millipore) according to standard procedures. Subsequently, membranes were blocked for one hour in Odyssey Li-COR blocking buffer. After blocking, membranes were probed with primary antibodies diluted in Odyssey Li-COR blocking buffer followed by 1 h incubation with IRDye-labeled secondary antibodies at room temperature. Membranes were visualized with an Odyssey SA infrared imaging system (Li-COR Biosciences, Lincoln, NE, USA).

Statistical Analyses
Statistical analyses were performed using Graphpad Prism 9. All data were expressed as mean ± SEM, except in boxplots where medians are shown. Dots represent the number of samples from independent experiments or individual mice. Statistical tests are specified in the text and figure legends. Differences were considered significant when p < 0.05.
Interestingly, the expression of caspase 1 p10 was significantly higher in Gsdmd −/− BMDMs as compared to Gsdmd +/+ controls ( Figure 1E). Because active caspase 1 can also act in a pro-apoptotic fashion [39,40], a TUNEL assay was performed on LPS-primed BMDMs after treatment with nigericin for 1 h ( Figure 1F). A significant increase in TUNEL positivity was observed after nigericin treatment in Gsdmd −/− BMDMs, while PI positivity was not increased, indicating that the plasma membranes of Gsdmd −/− cells were intact but that DNA fragmentation typical of apoptosis occurred. In contrast, TUNEL positivity did not change in Gsdmd +/+ BMDMs after nigericin treatment while PI positivity did increase significantly ( Figure 1F). Similar findings were observed after treatment with ATP, another classical caspase 1-dependent pyroptosis inducer ( Figure 1G).

Cleaved GSDMD Is Expressed in Human Carotid Lesions
To confirm that GSDMD is not only active in mice, cell-specific expression of cleaved GSDMD was evaluated in human carotid lesions ( Figure 3). In general, cleaved GSDMD was expressed in the plaque, especially in the shoulder regions, and media ( Figure 3B-D). Double immunohistochemical staining of cleaved GSDMD and CD68 showed that GSDMD is cleaved in plaque areas rich in macrophages, albeit not all CD68-positive cells contained cleaved GSDMD ( Figure 3B). On sections stained separately for cleaved GSDMD and αSMA ( Figure 3C), the positivity pattern appeared very similar. Indeed, double immunohistochemical staining of cleaved GSDMD and αSMA showed a clear overlapping of red (cl-GSDMD) and purple (αSMA) signal, confirming colocalization of cleaved GSDMD and smooth muscle cells. The CD31-positive intima delineated the vessel lumen ( Figure 3D). However, no cleaved GSDMD positivity was observed in CD31-positive endothelial cells. analyzed via flow cytometry and cell death was measured using PI labelling (two-way ANOVA followed by Sidak's multiple comparison, n = 3-5 independent experiments). * p < 0.05, ** p < 0.01, *** p < 0.001.

Cleaved GSDMD Is Expressed in Human Carotid Lesions
To confirm that GSDMD is not only active in mice, cell-specific expression of cleaved GSDMD was evaluated in human carotid lesions (Figure 3). In general, cleaved GSDMD was expressed in the plaque, especially in the shoulder regions, and media ( Figure 3B-D). Double immunohistochemical staining of cleaved GSDMD and CD68 showed that GSDMD is cleaved in plaque areas rich in macrophages, albeit not all CD68-positive cells contained cleaved GSDMD ( Figure 3B). On sections stained separately for cleaved GSDMD and αSMA ( Figure 3C), the positivity pattern appeared very similar. Indeed, double immunohistochemical staining of cleaved GSDMD and αSMA showed a clear overlapping of red (cl-GSDMD) and purple (αSMA) signal, confirming colocalization of cleaved GSDMD and smooth muscle cells. The CD31-positive intima delineated the vessel lumen ( Figure 3D). However, no cleaved GSDMD positivity was observed in CD31-positive endothelial cells.
In plaques of the proximal aorta ( Figure 6) and aortic root (Supplementary Figure S3), no differences in plaque size, necrotic core area, cell infiltration, and total collagen content were observed between ApoE −/− Gsdmd −/− and ApoE −/− Gsdmd +/+ mice. Nevertheless, MAC3 immunoreactivity was significantly decreased while the αSMA to MAC3 immunoreactivity ratio was significantly increased in plaques of the proximal aorta from ApoE −/− Gsdmd −/− mice, again indicating a shift in plaque composition and inflammatory state (Figure 6C). In contrast to what was observed in vitro and in the brachiocephalic artery from ApoE −/− Gsdmd −/− mice, TUNEL positivity was not increased in plaques in the proximal aorta ( Figure 6D). to count apoptotic cells (dotted boxes are magnified, scale bar = 20 µm). * p < 0.05, ** p < 0.01 (independent samples t-test, boxplot: Mann-Whitney test, n = 10-18 mice per group). Scale bar = 100 µm. Representative images are shown.
Previous studies have demonstrated that GSDMD, NT-GSDMD, and Gsdmd mRNA are upregulated in aortas of hyperlipidemic mice and in LPS/oxLDL-treated mouse peritoneal macrophages [30]. Moreover, the expression of NT-GSDMD is increased in plaques of LDLr −/− mice [45]. Gsdmd mRNA is also upregulated in peripheral blood monocytes of atherosclerotic patients [30]. However, to the best of our knowledge, the expression of cleaved NT-GSDMD has not yet been evaluated in human atherosclerotic plaques. Therefore, we performed immunostaining on human carotid lesions. We demonstrated that NT-GSDMD was expressed in human plaques, both in macrophage-and VSMC-rich regions. However, it should be noted that this finding was based on analysis of a limited number of plaques and comparison with plaque-free arteries was not made. Other groups have shown that components of the canonical pyroptosis pathway colocalize with plaque macrophages and contribute to macrophage pyroptosis and atherogenesis [12,15,24,27]. Interestingly, in the present study, colocalization of NT-GSDMD with αSMA-positive smooth muscle cells was even more pronounced than with CD68-positive macrophages. This is in line with a recent study reporting that caspase 1-dependent pyroptosis occurs in VSMCs and contributes to the progression of atherosclerosis [22]. Importantly, pyroptosis has also been described in plaque endothelial cells [46][47][48]. However, we did not observe any NT-GSDMD immunoreactivity in the luminal endothelial layer of human carotid lesions. Similarly, Rajamäki and colleagues also did not observe any NLRP3 immunoreactivity in the endothelium of human coronary plaques [15].
Disulfiram (used to treat alcohol use disorder) has recently been identified as a potent GSDMD inhibitor [49]. Furthermore, dimethyl fumarate (used as an immunomodulator in multiple sclerosis) also covalently binds GSDMD thereby inhibiting pyroptosis in vitro and in vivo in a mouse model of LPS-induced shock [50]. The availability of safe, EMAand FDA-approved drugs that inhibit GSDMD makes it an interesting therapeutic target to address pyroptosis in atherosclerosis. Therefore, we crossbred atherosclerotic ApoE −/− mice with Gsdmd −/− mice to evaluate the effect of GSDMD deficiency on atherogenesis. After 16 weeks WD, Oil red O staining was performed on the thoracic aorta but no change in overall lipid burden was observed. In line with this finding, plaque size was not altered in the proximal aorta. However, the plaque size in the brachiocephalic artery of ApoE −/− Gsdmd −/− mice was significantly decreased as compared to ApoE −/− Gsdmd +/+ mice. Although both the ascending proximal aorta and the brachiocephalic artery of ApoE −/− mice are atherogenesis-prone sites [51], plaques in the brachiocephalic artery enter an advanced, human-like stage more rapidly [52,53]. Indeed, plaques in the brachiocephalic artery of ApoE −/− mice easily reveal a vulnerable plaque phenotype and thinning of the fibrous cap may even lead to plaque rupture, similarly to human plaques [53][54][55]. In contrast, aortic plaques of ApoE −/− mice resemble stable lesions in humans, based on fibrous cap stress analysis [53]. Altogether, these observations suggest that GSDMD deficiency does not affect plaque initiation (as lipid burden in the thoracic aorta and proximal aortic plaque size were not decreased in ApoE −/− Gsdmd −/− mice) but rather the transition to and growth of a vulnerable plaque. This is supported by the increased αSMA/MAC3-immunoreactivity ratio observed in plaques of both the brachiocephalic artery and the proximal aorta of ApoE −/− Gsdmd −/− mice as compared to controls. This finding suggests a lower degree of plaque inflammation and vulnerability, and was also observed in plaques of the proximal aorta despite the unchanged plaque size, demonstrating a general decrease in plaque inflammation. Moreover, macrophage content was significantly decreased in plaques in the proximal aorta of ApoE −/− Gsdmd −/− mice as compared to controls. This is in line with a recent study in which GSDMD expression was suppressed in ApoE −/− mice using an adeno-associated virus-5 (AAV-5) delivery system [30]. The authors reported a decrease in F4/80-positive macrophage content in the aorta and aortic valve of AAV-5-GSDMD-treated ApoE −/− mice as compared to AAV-5-control-treated ApoE −/− mice. Of note, the authors also reported a decrease in lipid burden and plaque size in the aorta and aortic valve when GSDMD expression was suppressed, however, only a limited number of mice was included and preliminary data were reported [30]. Another study reported a decreased lesion area in the aortic root of LDLr antisense oligonucleotide-treated Gsdmd −/− mice compared to Gsdmd +/+ controls, which was attributed to decreased IL-1β release resulting in decreased foam cell formation and ATP release in plasma [45]. Importantly, experimental differences such as a longer period of feeding WD in the present study, evaluation of different vascular sites, and the use of different genetic models make comparison with these studies difficult.
In plaques of the brachiocephalic artery, the number of TUNEL positive cells per mm 2 was significantly increased in ApoE −/− Gsdmd −/− mice compared to controls. Increased TUNEL positivity is not specific and DNA fragmentation is reported to occur during apoptotic death as well as during different forms of necrotic death [56,57]. However, as necrosis is decreased in plaques of ApoE −/− Gsdmd −/− mice, it is plausible to conclude that a switch to apoptosis occurs when pyroptosis is defective. Accordingly, in vitro TUNEL positivity and levels of pro-apoptotic caspase 1 p10 were increased in Gsdmd −/− BMDMs while PI positivity did not increase. Similarly, Opoku and colleagues recently reported increased apoptosis, characterized by phosphatidyl serine exposure on the cell surface, when pyroptosis was defective in Gsdmd −/− macrophages [45]. Importantly, a switch to apoptosis, which is non-lytic and non-inflammatory (in contrast to pyroptosis), will limit plaque progression, inflammation, and destabilization and thus, is regarded to be atheroprotective at this stage of atherosclerosis in ApoE −/− mice. However, as previously reported by our lab, accumulation of apoptotic bodies can induce secondary necrosis, eventually resulting in expansion of the necrotic core and plaque size during later stages of atherosclerosis [58]. Accordingly, large plaques with an inflammatory phenotype were observed in the aortic root from both ApoE −/− Gsdmd −/− and ApoE −/− Gsdmd +/+ mice, possibly because after 16 weeks WD plaques in the aortic root are in a more advanced stage compared to plaques in the brachiocephalic artery. Thus, the effects in the longer term of targeting GSDMD and the concomitant switch in cell death modality remain to be elucidated.

Conclusions
We report that cleaved GSDMD is present in human atherosclerotic plaques and is required for inflammatory pyroptosis in murine macrophages and smooth muscle cells in vitro. GSDMD deficiency in ApoE −/− mice does not inhibit plaque initiation and the formation of stable aortic plaques, but plays a role in the formation of inflammatory plaques in the brachiocephalic artery. Indeed, a shift toward a less inflammatory plaque composition and delayed plaque progression were observed in plaques of the brachiocephalic artery from ApoE −/− Gsdmd −/− mice as compared to plaques with a more vulnerable phenotype observed in ApoE −/− Gsdmd +/+ controls. This is accompanied by less plaque necrosis and a switch to apoptosis when GSDMD is deficient, which is also observed in vitro in BMDMs. Therefore, targeting GSDMD appears to be a promising approach for limiting the transition to an inflammatory and vulnerable plaque phenotype, and subsequently plaque destabilization. However, the pharmacological translation and therapeutic value in atherosclerosis patients should be further investigated.