GSDME in Endothelial Cells: Inducing Vascular Inflammation and Atherosclerosis via Mitochondrial Damage and STING Pathway Activation

The initiation of atherosclerotic plaque is characterized by endothelial cell inflammation. In light of gasdermin E’s (GSDME) role in pyroptosis and inflammation, this study elucidates its function in atherosclerosis onset. Employing Gsdme- and apolipoprotein E-deficient (Gsdme−/−/ApoE−/−) and ApoE−/− mice, an atherosclerosis model was created on a Western diet (WD). In vitro examinations with human umbilical vein endothelial cells (HUVECs) included oxidized low-density lipoprotein (ox-LDL) exposure. To explore the downstream mechanisms linked to GSDME, we utilized an agonist targeting the stimulator of the interferon genes (STING) pathway. The results showed significant GSDME activation in ApoE−/− mice arterial tissues, corresponding with atherogenesis. Gsdme−/−/ApoE−/− mice displayed fewer plaques and decreased vascular inflammation. Meanwhile, GSDME’s presence was confirmed in endothelial cells. GSDME inhibition reduced the endothelial inflammation induced by ox-LDL. GSDME was linked to mitochondrial damage in endothelial cells, leading to an increase in cytoplasmic double-stranded DNA (dsDNA). Notably, STING activation partially offset the effects of GSDME inhibition in both in vivo and in vitro settings. Our findings underscore the pivotal role of GSDME in endothelial cells during atherogenesis and vascular inflammation, highlighting its influence on mitochondrial damage and the STING pathway, suggesting a potential therapeutic target for vascular pathologies.


Introduction
Atherosclerosis is a prevalent disease that imposes a significant global burden due to its association with ischemia and infarction in vital organs, such as the heart and brain [1,2].The initial stages of atherosclerotic plaque formation involve the inflammation of endothelial cells, which release adhesion molecules and chemokines, such as intercellular adhesion molecule-1 (ICAM-1), vascular cell adhesion molecule-1 (VCAM-1), and monocyte chemotactic protein-1 (MCP-1).These molecules facilitate the recruitment of circulating monocytes to the intima, where they transform into foam cells within the expanding atherosclerotic plaque [3][4][5].
Gasdermins (GSDMs) are a family of proteins that play a crucial role in pyroptosis, a recently discovered form of programmed cell death characterized by cellular rupture and the release of pro-inflammatory cytokines.In humans, this protein family consists of six paralogous genes: GSDMA, GSDMB, GSDMC, GSDMD, GSDME (also known as DFNA5), and PJVK (also known as DFNB59) [6][7][8][9][10].Recent studies have revealed the widespread occurrence of pyroptosis in atherosclerosis [11,12], with GSDMD showing a propensity to promote the development of atherosclerosis [13].Similarly, GSDME, another member of the gasdermin family, induces pyroptosis via the caspase 3-GSDME pathway and is expressed in endothelial cells [14][15][16].Importantly, caspase 3, an upstream molecule of GSDME, has been identified as an independent risk factor for atherosclerosis in clinical investigations [17].
Aside from its direct involvement in inducing cell pyroptosis, the N-terminal fragment of gasdermin E (GSDME-NT), the activated fragment of GSDME, can aggregate on the mitochondrial membrane, leading to mitochondrial damage and the subsequent release of mitochondrial DNA (mtDNA) into the cytoplasm [18,19].Recent investigations have revealed that the cyclic GMP-AMP synthase-stimulator of the interferon gene (cGAS-STING) signaling pathway can be activated via excessive cytoplasmic endogenous DNA, including mtDNA or exogenous DNA.This activation results in an increased production of type I interferon [20,21].Additionally, STING regulates the downstream signaling pathway of nuclear factor kappa-B (NF-κB), which leads to the expression of genes encoding proinflammatory cytokines [22].Recent research suggests that cigarette smoke induces the release of mtDNA from mitochondria in endothelial cells, promoting atherosclerosis via activating the STING pathway.Free mtDNA has emerged as an independent predictor of atherosclerosis risk [17].
In this study, Gsdme-and apolipoprotein E-deficient (Gsdme −/− /ApoE −/− ) mice exhibited reduced mitochondrial damage, STING pathway activation, vascular inflammation, and atherosclerotic lesions compared to ApoE −/− mice after a WD.In in vitro models with human umbilical vein endothelial cells (HUVECs), the inhibition of GSDME was found to ameliorate oxidized low-density lipoprotein (ox-LDL)-induced STING pathway activation and endothelial cell inflammation.Intriguingly, when treated with SR-717, a specific STING agonist, there was a partial reversal of the inflammation and atherosclerosis suppression induced via GSDME inhibition, evident in both in vivo and in vitro settings.Collectively, our findings underscore the pivotal role of endothelial GSDME in atherogenesis, mediated through mitochondrial damage and the subsequent activation of the STING pathway.

Animal Models and Procedures
Animal protocols were approved by the Institutional Animal Care and Use Committee of Shanghai Model Organisms Center, Inc., Shanghai, China (Protocol No. 2020-0045, 31 December 2020).Gsdme −/− mice and ApoE −/− mice on the C57BL/6J background were obtained from the Shanghai Model Organisms Center, Inc. (Shanghai, China) and crossed to generate Gsdme −/− /ApoE −/− mice.The genotyping of Gsdme −/− mice was performed by amplifying a 920-bp fragment for the wild-type (WT) allele and a 495-bp fragment for the conventional knockout allele using the forward primer TTGGGGCGGGAAAGGTC and the reverse primer AAGCAGGGCAGTTACAGGAG.Only male mice were included.They were fed a WD (21% fat, 0.15% cholesterol, SLAC, Shanghai, China) from 5 to 18 weeks of age and received intraperitoneal injections of SR-717 (0.3% dimethyl sulfoxide, 10 mg/kg per day; MCE, Monmouth Junction, NJ, USA) or phosphate buffer saline (PBS) in the last 4 weeks.ApoE −/− mice were sacrificed at 10, 14, or 18 weeks, and all Gsdme −/− /ApoE −/− mice were sacrificed at 18 weeks.

Hematoxylin and Eosin (HE) Staining
Aortic roots from anesthetized mice were fixed in 4% paraformaldehyde (Servicebio, Wuhan, China) for 4 h, washed thrice with PBS, and dehydrated in 30% sucrose overnight.Sections (10 µm thick) were cut, with six per slide, spaced by 80 µm, displaying a threevalve structure.Slides were stored at −80 • C. For HE staining, sections were ethanol dehydrated, cleared in xylene, and mounted with neutral balsam.Images were taken using a Leica microscope (Leica, Wetzlar, Germany).The average atherosclerotic lesion area was quantified from six sections per mouse using ImageJ software version 2 (National Institutes of Health, Bethesda, MD, USA).

Monocyte-Endothelial Cell Adhesion Assay
HUVECs were exposed to ox-LDL or vehicle for 24 h, with or without SR-717.THP-1 monocytes were labeled with calcein-AM (Beyotime Biotechnology, Beijing, China) and cocultured with HUVECs for 1 h.After washing to remove non-adherent THP-1 monocytes, cells were fixed, and ImageJ software was used to count cells in 10 random fields.

Statistical Analysis
Statistical analyses were performed using GraphPad Prism software version 8.0.Data are presented as mean ± SEM.Normality and equal variance were assessed using the Shapiro-Wilk test and F test, respectively.For two-group comparisons, an unpaired Student t-test was used, while one-way ANOVA with the Bonferroni post hoc test was employed for multiple group comparisons.Two-way ANOVA with the Bonferroni post-test was used for data with more than two categorical variables.A p-value less than 0.05 (p < 0.05) was considered statistically significant.

GSDME Promoted Vascular Inflammation and Atherosclerosis in ApoE −/− Mice
To elucidate the role of GSDME in atherogenesis, ApoE −/− mice were maintained on a WD regimen, and aortic samples were harvested from these mice aged between 5 and 18 weeks.The Western blot analysis revealed a progressive activation of GSDME in arterial tissues corresponding to the duration of WD (Figure 1A,B), indicating its involvement in atherogenesis.

GSDME Mediated Ox-LDL-Induced Inflammation in HUVECs
We utilized immunofluorescence staining techniques to determine the cellular localization of GSDME within aortic tissues.Our analyses demonstrated that GSDME exhibits colocalization with the endothelial cell-specific marker CD31 in the aortas of 18-week-old ApoE −/− mice (Figure 2A).To delve deeper into the functional implications and underlying mechanisms of GSDME in endothelial cells during atherogenesis, we executed in vitro assays with HUVECs exposed to ox-LDL, thereby mimicking the pathological alterations characteristic of atherosclerosis.By adjusting the concentration of ox-LDL and modulating the exposure duration, we determined that GSDME becomes activated in HUVECs upon ox-LDL exposure.In particular, we noted a significant rise in GSDME-NT levels that directly correlated with both the duration of exposure and the concentrations of ox-LDL (Figure 2B-E).
To investigate the role of GSDME in HUVECs, we employed siRNA specifically targeting GSDME to inhibit its expression.Following 48 h of siRNA treatment, the effectiveness of this knockdown was validated via Western blot analysis, which revealed a marked reduction in GSDME expression with si-GSDME compared to the negative control siRNA (Figure 2F,G).After treatment with either si-GSDME or si-NC, HUVECs were subjected to 100 µg/mL ox-LDL or PBS for 24 h.The si-GSDME treatment significantly attenuated the ox-LDL-induced increase in ICAM-1 and VCAM-1 protein levels, as confirmed via Western blot analyses (Figure 2H-J).Consistent with these observations, monocyte-endothelial cell adhesion assays revealed a significant reduction in THP-1 cell adhesion after si-GSDME treatment in the presence of ox-LDL, relative to the si-NC-treated group (Figure 2K,L).

GSDME Damaged Mitochondria in Endothelial Cells and Activated STING Pathway
Based on the research findings suggesting that GSDME-NT preferentially localizes to the mitochondrial membrane rather than the cell membrane, leading to mitochondrial damage and subsequent translocation of mtDNA (belonging to double-stranded DNA) into the cytoplasm [10,11], we conducted immunofluorescence labeling for cytoplasmic double-stranded DNA (dsDNA) and CD31.This was performed on 5-week-old ApoE −/− mice (which were not subjected to a WD), 18-week-old ApoE −/− mice, and 18-week-old Gsdme −/− /ApoE −/− mice.Our findings revealed a marked elevation in dsDNA levels within endothelial cells following WD.However, this rise was notably mitigated in the absence of GSDME, as observed in the Gsdme −/− /ApoE −/− mice (Figure 3A).
Recent investigations have disclosed that cytoplasmic mtDNA, functioning as endogenous DNA, can activate the STING pathway, thereby precipitating an inflammatory response [20].Concurrently, additional studies have highlighted the activation of the STING pathway within endothelial cells throughout atherogenesis [23,24].In light of this, we evaluated the activation levels of the STING pathway in the aortas of both ApoE −/− and Gsdme −/− /ApoE −/− mice at 18 weeks of age following a WD regimen.Our data demonstrated that the phosphorylated activation of critical components, such as STING, TBK1, IRF3, and NF-κB p65, was markedly increased in the ApoE −/− mice, and this activation was noticeably diminished in the aortic tissues of Gsdme −/− /ApoE −/− mice compared to ApoE −/− (Figure 3B-F).In vitro findings were consistent with this, as after 24 h of ox-LDL incubation, HUVECs exhibited STING pathway activation, which was mitigated via the intervention of si-GSDME (Figure 3G-K).

Activation of the STING Pathway Altered the Protective Effects of GSDME Deficiency on Atherogenesis and Endothelial Inflammation
In our investigation of the interaction between the STING pathway and GSDME, we utilized SR-717, a STING agonist.The structural and functional analyses confirmed that SR-717 functions as a direct cyclic guanosine monophosphate-adenosine monophosphate (cGAMP) mimetic, inducing the characteristic 'closed' conformation of STING [25].Mice models, both ApoE −/− and GSDME −/− /ApoE −/− , on a WD regimen, were administered daily intraperitoneal doses of 10 mg/kg SR-717 or a vehicle control from 14 to 18 weeks of age.Utilizing in situ imaging and HE staining, we observed that the activation of the STING pathway led to an enlargement of the plaque area in both GSDME −/− /ApoE −/− and ApoE −/− mice, thereby counteracting the suppressive influence of GSDME deficiency on atherosclerosis progression (Figure 4A,C).Subsequent analysis employing immunofluorescence methodologies revealed an amplified macrophage infiltration in both mouse models post-STING pathway activation (Figure 4B,D).This intensified inflammatory response served to neutralize the inhibitory effects of GSDME deficiency on vascular inflammation in ApoE −/− mice.This finding was further corroborated by the altered expression levels of pivotal inflammatory markers, including ICAM-1, VCAM-1, and MCP-1 (Figure 4E-H).Moreover, in the backdrop of GSDME deficiency, a partial restoration in the phosphorylation and activation of the STING pathway was noted upon its activation (Figure 4I-M).
To investigate the influence of the STING pathway on endothelial inflammation mediated via GSDME, HUVECs were initially exposed to siRNA.Following this, they were pre-treated with SR-717 or a vehicle control for a duration of 1 h and subsequently exposed to either ox-LDL or PBS for 24 h.Our results demonstrated that in HUVECs treated with ox-LDL, SR-717 elevated the expression of ICAM-1 and VCAM-1 protein, observed in both the si-NC and si-GSDME groups.Notably, this upregulation counteracted the suppressive influence of si-GSDME on protein expression in the presence of ox-LDL (Figure 5A-C).Simultaneously, our observations revealed that SR-717 enhanced the monocyte-endothelial cell adhesion in response to ox-LDL, counteracting the inhibitory effects of si-GSDME (Figure 5D-E).Additionally, the results from Western blot assays revealed that SR-717 mitigated the reduction in STING pathway phosphorylation, a consequence linked to si-GSDME, in ox-LDL-treated HUVECs (Figure 5F-J).

Discussion
Our findings provide evidence of GSDME activation during the progression of atherosclerosis and confirm that the deficiency of GSDME exhibited significant effects on reducing atherogenesis and diminishing vascular inflammation in ApoE −/− mice.Furthermore, we confirmed the localization of GSDME within endothelial cells of arterial tissues.This finding was complemented by in vitro studies, where GSDME deficiency attenuated the inflammatory response in endothelial cells exposed to ox-LDL.Probing further into the mechanistic underpinnings within ApoE −/− mice, we found that GSDME deficiency correlated with decreased cytoplasmic dsDNA accumulation in endothelial cells, indicating a diminished activation of the STING pathway.Thus, our findings suggest the STING pathway as a potential downstream effector in the cascade of GSDME-induced endothelial inflammation and atherosclerosis.To further elucidate this connection, we employed a STING-specific agonist to activate the STING pathway in GSDME-deficient ApoE −/− mice and HUVECs.Interestingly, this activation counteracted some of the protective effects of GSDME deficiency, suggesting that the STING pathway operates downstream of GSDME in mediating endothelial inflammation and atherosclerosis.
GSDME, an important member of the gasdermin family responsible for mediating pyroptosis, is involved in multiple inflammatory diseases [26][27][28].Numerous studies have demonstrated that pyroptosis contributes to inflammation in the progression of atherosclerosis [29][30][31].However, the role and mechanism of GSDME in atherogenesis remain unclear.This study presents compelling evidence for GSDME's role in endothelial cells, contributing to vascular inflammation and the onset of atherosclerosis, as demonstrated both in animal models and at the cellular level.In a recent in vitro study using macrophages, it was found that the ablation of GSDME suppressed inflammation and macrophage pyroptosis induced by ox-LDL [32].However, our findings bring to light a novel aspect, revealing that GSDME mediates endothelial inflammation, thereby promoting vascular inflammation and the initiation of atherosclerosis.This underscores the importance of endothelial cells when considering GSDME's influence on atherosclerosis.
While prior research has posited the involvement of endothelial cell pyroptosis in atherosclerotic progression, it is important to note that the occurrence rate of this pyroptosis remains relatively low [33].Given this context, our study was geared toward identifying alternative mechanisms by which GSDME exacerbates vascular inflammation and atherosclerosis beyond the scope of cell pyroptosis.Our study observed elevated cytoplasmic dsDNA levels in ApoE −/− mice, a trend that significantly receded in the absence of GSDME.This is consistent with earlier studies suggesting that GSDME-NT tends to aggregate on mitochondrial membranes, thereby triggering mitochondrial dysfunction and resulting in the release of materials, such as dsDNA, into the cytoplasm [19].Importantly, our results showed that inhibiting GSDME reduced the activation of the STING pathway in both vascular tissues and endothelial cells.Consequently, we explored the potential role of SR-717, a known STING agonist [25].Experiments conducted on mice and HUVECs demonstrated that SR-717 administration could partially counteract the inhibitory effects of GSDME deficiency on atherosclerosis, vascular inflammation, and ox-LDL-induced endothelial cell inflammation.This implies that the STING pathway may function downstream of GSDME-mediated effects.To our knowledge, this is the inaugural study to propose the mtDNA-STING axis as a potential downstream mechanism influenced by GSDME during atherogenesis and endothelial inflammation.
Moreover, our findings indicate that STING agonists exacerbate atherosclerosis, vascular inflammation, and ox-LDL-induced endothelial cell inflammation in ApoE −/− mice.This observation is particularly intriguing given the recent surge of interest in STING agonists.Numerous studies have highlighted the potential therapeutic benefits of STING agonists across various clinical settings, with a particular focus on their promising role in cancer therapy [34][35][36].However, our results underscore the need for caution.While the therapeutic potential of STING agonists is undeniable, it is imperative to consider their potential side effects, especially in the context of vascular health.Future studies should delve deeper into the mechanisms underlying these observations to ensure the safe and effective application of STING agonists in clinical settings.
In this investigation, we employed ApoE −/− and GSDME −/− /ApoE −/− mouse models.For subsequent studies, it may be beneficial to use endothelial-specific Gsdme-deficient mice to delve deeper into GSDME's role in atherogenesis.Our findings suggest that GSDME activity modulates the mtDNA-STING pathway.However, it is crucial to consider that other signaling cascades may also be influenced downstream of GSDME during atherogenesis.The field has seen recent advancements, with the identification of alternative innate immune sensors for cytosolic mtDNA.In a noteworthy study, ZBP1 was identified to potentiate IFN-I responses to cytosolic mtDNA and, in turn, to exacerbate inflammation [37].Given these insights, future research should focus on uncovering additional potential pathways, aiming to deepen our understanding of the intricate mechanisms through which GSDME affects endothelial inflammation and atherosclerosis.As we continue to unravel these mechanisms, the urgency to develop and validate GSDME protein inhibitors in mouse atherosclerosis models becomes evident.Such advancements are pivotal for enhancing our mechanistic insights and further pave the way for potential clinical drug development.

Conclusions
This research underscores the pivotal role of GSDME, an often overlooked member of the gasdermin family, in the pathogenesis of atherosclerosis, specifically in endothelial cells.Through both in vivo and in vitro analyses, we have unearthed GSDME's contribution to vascular inflammation, a key facet of atherogenesis.Our findings underscore the significant effects of GSDME deficiency on reducing atherogenesis and pinpoint the STING pathway as a downstream mediator of these processes.The compelling evidence presented here broadens our understanding of endothelial inflammation within the context of atherosclerosis, highlighting the nuanced interplay between GSDME, mitochondrial dysfunction, and the STING pathway.Importantly, this work opens new avenues for therapeutic strategies targeting GSDME-mediated mechanisms, with potential implications in the prevention and treatment of atherosclerosis.