GRAPHICAL ABSTRACT

Highlights

• Serpina3c is a key regulator of ER redox homeostasis in adipocytes.

• Serpina3c blocks HIF1α/Ero1α-PDI-mediated ER overoxidation and ER stress.

• Serpina3c loss in eWAT increases inflammation in eWAT, serum, and heart of obese mice.

INTRODUCTION

As unhealthy high-fat and high-calorie diets have become common in modern society, obesity has become a major public health epidemic worldwide, and this has greatly increased the risk of cardiovascular diseases [1,2]. The heart is an important target of obesity damage, which is characterized by increased cardiac inflammation, lipid deposition, interstitial fibrosis, and cardiomyocyte death. If not intervened early, it will ultimately progress to irreversible structural and systolic dysfunction, that is, obesity-related cardiomyopathy [3,4]. How to effectively antagonize metabolic heart damage during its early stage is an urgent scientific issue in the cardiovascular field.

The most primary feature of obesity is white adipose tissue (WAT) remodeling, which manifests as unhealthy WAT expansion, accompanied by dysfunctional effects such as inflammation and fibrosis [5]. WAT is classified anatomically as subcutaneous WAT (sWAT) and visceral WAT (vWAT). sWAT is chiefly responsible for lipid storage, and inguinal WAT (iWAT) is the typical representative of sWAT. vWAT prefers to manage circulatory inflammation levels via secreting pro-inflammatory and anti-inflammatory cytokines; consequently, it indirectly affects the inflammation levels of distant organs [6-8]. The most representative, abundant and widely studied type of vWAT is epididymal WAT (eWAT). The metabolic inflammation that occurs in obesity is a chronic low-grade inflammation, originating mainly from adipocyte-driven vWAT inflammation, and it progresses locally from vWAT to systemic inflammation. The critical link in metabolic inflammation is macrophage recruitment and M1 polarization to amplify vWAT inflammation, which is drove by adipocyte dysfunction [9,10]. vWAT inflammation is a prominent risk factor and initiator of obesity-induced circulatory inflammation and subsequent heart inflammation, which is the pathological basis for further metabolic heart damage [11-13]. Thus, effective relief of vWAT inflammation during obesity is a feasible strategy for further alleviating heart inflammation.

Serine (or cysteine) peptidase inhibitor, clade A, member 3C (Serpina3c) has been identified as a novel adipokine expressed primarily by mature adipocytes in WAT. Serpina3c expression in 3T3-L1 preadipocytes is negligible; however, the expression increases gradually upon differentiation, and this protein is highly expressed in mature adipocytes [14]. It has been reported that Serpina3c protein expression is down-regulated in eWAT from high-fat diet (HFD)-induced obese mice. Serpina3c global knockout (Serpina3c^–/–) mice exhibit weight gain and increased eWAT inflammation upon HFD feeding, partially due to the absence of inhibition of Serpina3c on cathepsin G, resulting in increased production of inflammatory cytokines in adipocytes [15]. However, the capacity of adipocytes to generate inflammatory cytokines is much lower than that of macrophages, so the contribution of Serpina3c to eWAT inflammation has not been well studied. Other actions and mechanisms of Serpina3c related to adipocytes remain largely unknown. In addition, it is unclear whether Serpina3c deficiency could affect heart damage via mediating eWAT inflammation.

Due to the massive WAT expansion and insufficient blood supply in states of nutritional overload, adipocytes are susceptible to local hypoxia pressure, and the transcription regulatory factor hypoxia-inducible factor 1α (HIF1α) is activated [16]. As a momentous organelle for sensing cellular stress, the endoplasmic reticulum (ER) triggers a series of ER stress (ERS) responses in adipocytes in response to hypoxia. Growing evidence suggests that sustained or excessive ERS can impair adipocyte function, such as cell survival and cytokine production [17,18]. The ER controls correct protein folding via oxidoreductases and chaperones on its membrane; this process is pivotal for maintaining ER homeostasis. ER oxidoreductase 1α (Ero1α) and the protein disulfide isomerase (PDI) family cooperate to regulate oxidized protein folding. Normally, Ero1α re-oxidizes and re-activates PDI, synergistically controlling the formation of protein disulfide bonds. However, this process consumes oxygen (O₂) and generates hydrogen peroxide (H₂O₂), which is the main source of reactive oxygen species (ROS) in the ER lumen [19]. Therefore, over-activation of the Ero1α-PDI signaling can cause ER overoxidation; this can disrupt the redox state of the ER and increase the frequency of abnormally folded proteins, thus facilitating ER oxidative stress and ERS [20]. Whether Serpina3c is involved in ER homeostasis in adipocytes remains unclear.

In the present study, we established a diet-induced obese mice model and a palmitic acid (PA)-stimulated lipotoxicity injury model in 3T3-L1 adipocytes to explore the function of Serpina3c. Our results suggest that targeting Serpina3c in eWAT may be a potential therapeutic strategy for alleviating obesity-related eWAT inflammation and heart damage.

METHODS

All methods are described in the Supplementary Methods section.

RESULTS

In vitro effect of Serpina3c on ER overoxidation, cytokine expression, and apoptosis in adipocytes

To explore the potential role of Serpina3c in adipocytes subject to lipotoxicity injury, we cross-analyzed RNA-sequencing data from Serpina3c knockdown (3cKD) 3T3-L1 adipocytes treated with PA for 24 hours compared to a control LV3 group, as well as from Serpina3c overexpression (3cOV) 3T3-L1 adipocytes compared to a control LV5 group. A total of 75 genes were upregulated in the 3cKD group and down-regulated in the 3cOV group, and 12 genes were down-regulated in the 3cKD group and up-regulated in the 3cOV group (Supplementary Fig. 1A). Next, we conducted Gene Ontology and Kyoto Encyclopedia of Genes and Genomes analyses of these 75 genes; the top three relevant enriched pathways were protein processing in the ER, the HIF-1 signaling pathway, and cytokine signaling pathway (Supplementary Fig. 1B). Gene set enrichment analysis consistently showed that 3cKD was correlated positively with the top three enriched pathways (Supplementary Fig. 1C). Based on these results, the following genes in the aforementioned three pathways attracted our attention: HIF1α, Ero1α, protein disulfide isomerase family A member 3 (PDIA3), protein disulfide isomerase family A member 4 (PDIA4), C-C motif chemokine ligand 2 (CCL2), C-C motif chemokine ligand 5 (CCL5), C-X-C motif chemokine ligand 1 (CXCL1), C-X-C motif chemokine ligand 10 (CXCL10), and interleukin-6 (IL-6); these genes showed up-regulated trends in the 3cKD group and down-regulated trends in the 3cOV group, while adiponectin yielded the opposite expression trend (Supplementary Fig. 1D and E).

Based on the RNA-sequencing analysis, we constructed LV3, 3cKD, LV5, and 3cOV 3T3-L1 adipocytes to investigate the role of Serpina3c (Fig. 1A). Compared to the control LV3 group, treating 3cKD group with 500 μM PA for 48 hours significantly reduced cell viability (Fig. 1B). Therefore, in subsequent experiments, the LV3, 3cKD, LV5, and 3cOV groups were stimulated with 500 μM PA for 48 hours. Consistent with the RNA-sequencing data, 3cKD markedly decreased the gene expression of the protective adipocytokine adiponectin and significantly promoted that of pro-inflammatory cytokines (IL-6, CCL2, CCL5, CXCL1, and CXCL10), HIF1α, Ero1α, PDIA3, and PDIA4 in PA-treated 3T3-L1 adipocytes. In contrast, 3cOV reversed the expression of these genes (Fig. 1C-E).

An interaction between Ero1α and PDI catalyzes a redox cycle that is essential for oxidative protein folding in the ER. When the Ero1α-PDI signaling is excessively up-regulated, H₂O₂ is overproduced; this leads to ER overoxidation, which is the main source of the ER oxidative stress mediator ROS, further inducing ERS [19]. We found that 3cKD significantly increased intracellular levels of H₂O₂ and ROS, as well as increased the protein levels of ERS markers, including glucose regulated protein 78 (GRP78), C/EBP homologous protein (CHOP), phosphorylated eukaryotic initiation factor 2α (p-eIF2α), cleaved activating transcription factor 6 (c-ATF6), and spliced X-box binding protein 1 (XBP1S), in adipocytes; 3cOV had the opposite effects (Fig. 1F-H). ERS is intimated tied to cell apoptosis and inflammation because it activates the mitogen-activated protein kinase (MAPK) signaling pathway; that is, it facilitates the phosphorylation of three major kinases: c-Jun N-terminal kinase (JNK), extracellular signal-regulated kinase (ERK) and P38 [21,22]. In contrast to 3cOV, 3cOV noticeably up-regulated the MAPK signaling pathway as evidenced by significantly increased protein levels of p-JNK, p-ERK, and p-P38; the marked increase in caspase-3 activity indicated increased adipocyte apoptosis (Fig. 1I and J). These results reveal that under PA-induced lipotoxicity, 3cKD in adipocytes enhances ER overoxidation and ERS to activative the MAPK signaling pathway, which promotes pro-inflammatory cytokine expression and cell apoptosis, and reduces adiponectin expression.

In obesity, macrophages play a critical role in eWAT inflammation, gathering around dead or dying adipocytes to form unique histological structures called the crown-like structures (CLSs). Pro-inflammatory cytokines secreted by adipocytes are pivotal causes of macrophage infiltration and M1 polarization in eWAT [23]. Therefore, we stimulated Raw264.7 cells for 24 hours with conditioned medium (CM) derived from PA-treated LV3, 3cKD, LV5, and 3cOV groups of 3T3-L1 adipocytes to determine the effect of CM on macrophage chemotaxis and M1 polarization. CM from 3cKD group obviously enhanced the migration ability and mRNA levels of M1 polarization makers in Raw264.7 cells, while CM from the 3cOV group blocked these effects (Supplementary Fig. 2A and B). Furthermore, we demonstrated that Serpina3c deficiency mainly indirectly regulates macrophage M1 polarization by affecting adipocyte function, rather than the direct effect of Serpina3c deficiency (Supplementary Fig. 2C and D). This evidence implies that 3cKD in adipocytes can promote the chemotaxis and M1 polarization of macrophages, the major contributors to eWAT inflammation in obesity.

Serpina3c deficiency aggravates ER overoxidation, inflammation, and apoptosis of eWAT from HFD-fed mice

Consistent with results from other studies [14,15], Serpina3c showed a highly restricted expression pattern in WAT of wild type (WT) mice (Supplementary Fig. 3A). Due to the central position of eWAT in metabolic inflammation, we focused our attention on eWAT. After HFD feeding for 12 weeks, lower Serpina3c protein level in eWAT of WT mice was detected (Supplementary Fig. 3B). A strategy based on the cyclization recombination enzyme (Cre)/locus of X-over P1 (LoxP) system was used to delete exon 3 of mouse Serpina3c by homologous recombination (Supplementary Fig. 4A). No Serpina3c expression was detected in eWAT from Serpina3c^–/– mice and the mice genotyping results were showed (Supplementary Fig. 4B and C). To determine the contribution of Serpina3c to metabolic inflammation, WT and Serpina3c^–/– mice were fed a chow diet (CD) or HFD for 12 weeks. After CD feeding, Serpina3c^–/– mice did not exhibit obvious abnormalities compared to WT mice (Fig. 2). However, Serpina3c^–/– mice were distinguishable from WT mice under feeding of HFD. Thus, for the following experiments, we describe only the results after HFD feeding.

After 12 weeks of HFD feeding, body weights were significantly higher in Serpina3c^–/– mice than WT mice (Fig. 2A and B). eWAT mass and eWAT weight to body weight ratios were also markedly higher in Serpina3c^–/– mice than WT mice (Fig. 2C-E). There were no significant differences in fasting blood glucose, food intake and serum lipid levels in Serpina3c^–/– and WT mice (Fig. 2F, G, and L). Hematoxylin-eosin (H&E), Masson’s trichrome and F4/80 immunohistochemistry (IHC) staining of eWAT respectively revealed larger adipocyte sizes, increased collagen deposition and greater CLS numbers in Serpina3c^–/– mice compared to WT mice (Fig. 2H-K). These observations indicate increased eWAT size, fibrosis and macrophage infiltration in HFD-fed Serpina3c^–/– mice relative to WT mice.

Consistent with the increased CLS numbers in eWAT from HFD-fed Serpina3c^–/– mice and our results of PA-treated 3cKD 3T3-L1 adipocytes, we found that adiponectin gene expression was markedly decreased, and the gene expression levels of proinflammatory cytokines, HIF1α, Ero1α, PDIA3, PDIA4, and ERS markers (GRP78, CHOP) were prominently increased in eWAT from Serpina3c^–/– mice compared with that from HFD-fed WT mice (Fig. 3A and B). Moreover, H₂O₂ levels in eWAT were higher, and the MAPK signaling pathway was evidently up-regulated in Serpina3c^–/– mice compared to HFD-fed WT mice (Fig. 3C and D). Consistently, caspase-3 activity and terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining revealed an apparent increase in eWAT apoptosis in Serpina3c^–/– mice compared to HFD-fed WT mice (Fig. 3E and F). In addition, transmission electron microscopy (TEM) showed more pronounced ER swelling in eWAT from Serpina3c^–/– mice compared with that from HFD-fed WT mice (Fig. 3G). In summary, these results imply that in the HFD-fed mice, Serpina3c deficiency exacerbates ER overoxidation and ERS in eWAT, thereby activating the MAPK signaling pathway and promoting eWAT inflammation and apoptosis.

Serpina3c deficiency promotes serum and heart inflammation in HFD-fed mice

During obesity, CLSs in eWAT release large amounts of proinflammatory cytokines into the bloodstream; these cytokines eventually reach distal organs, including the heart, causing persistent and chronic heart inflammation [23,24]. We found that Serpina3c deletion significantly increases the CLS number in eWAT after HFD feeding, but it is unclear whether its deletion further affects serum and heart inflammation in obesity. We found no significant differences in serum and heart inflammation between Serpina3c^–/– and WT mice fed a CD (Fig. 4). However, compared with WT mice fed the HFD, Serpina3c^–/– mice had significantly increased serum pro-inflammatory cytokine levels, including tumor necrosis factor-alpha (TNF-α), IL-6, and CCL2; IL-1β levels followed an upward trend that was not significantly different, and serum adiponectin levels were markedly decreased (Fig. 4A and B).

Although there were no significant differences in cardiac systolic and diastolic function, cardiomyocyte cross-sectional area, or triglyceride content in the hearts of Serpina3c^–/– and WT mice after 12 weeks of HFD feeding, Serpina3c^–/– mice exhibited higher serum levels of myocardial injury markers, including lactate dehydrogenase (LDH) and creatine kinase MB (CK-MB) isoenzyme, than WT mice (Fig. 4C, G, H, and I, Supplementary Fig. 5). H&E, Masson’s trichrome, and F4/80 IHC staining of the hearts showed that although cardiac muscle fibers in the WT and Serpina3c^–/– mice were arranged regularly, Serpina3c^–/– mice contained pronounced collagen deposition, interstitial fibrosis and macrophage infiltration compared to HFD-fed WT mice (Fig. 4D-F). In addition, the mRNA levels of inflammation and fibrosis genes, including TNF-α, transforming growth factor-β (TGF-β) and collagen type I, α 1 chain (COL1A1), in the heart were significantly increased in Serpina3c^–/– mice compared to HFD-fed WT mice (Fig. 4J).

To demonstrate the cross-talk between eWAT and the heart, primary cardiac fibroblasts (CFs) were extracted from the hearts of neonatal WT mice. We next cultured fresh eWAT from HFD-fed WT or Serpina3c^–/– mice in vitro for 12 hours to obtain CM; this CM was then added to CFs for 24 hours. Compared with CM from the eWAT of WT mice, CM from the eWAT of Serpina3c^–/– mice promoted the gene expression of fibrosis markers (TGF-β, COL1A1) in CFs (Fig. 4K). Taken together, these results suggest that in HFD-fed mice, Serpina3c deficiency induces more severe serum inflammation and heart damage.

Serpina3c inhibits ER overoxidation and its downstream effects in adipocytes partially via blocking HIF1α-mediated Ero1α-PDI signaling

Ero1α has been reported as a transcriptional target gene of HIF1α [25]. Whether HIF1α regulates the transcription of PDIA3 and PDIA4 remains to be assessed. Our dual-luciferase reporter assay results showed that HIF1α had a significant activating effect on the promoter activity of PDIA3 and PDIA4 in 293T cells, suggesting that PDIA3 and PDIA4 are target genes of HIF1α (Fig. 5A). To explore whether Serpina3c controls Ero1α-PDI signaling by suppressing HIF1α, 3cKD 3T3-L1 adipocytes were pretreated with 100 ng/mL recombinant mouse Serpina3c protein (rSerpina3c) or 5 μM of HIF1α inhibitor acriflavine (ACF) for 2 hours before stimulation with 500 μM PA for 48 hours. Compared with vehicle treatment, both rSerpina3c and ACF significantly up-regulated adiponectin expression, down-regulated the expression of pro-inflammatory cytokines, Ero1α, PDIA3, PDIA4, and ERS makers, reduced intracellular H₂O₂ accumulation and decreased MAPK pathway signaling and adipocyte apoptosis. There were no significant differences in these measurements between the rSerpina3c and ACF groups (Fig. 5B-H). Our results demonstrate that under lipotoxicity injury, Serpina3c inhibits HIF1α-mediated Ero1α-PDI signaling in adipocytes.

To confirm whether the worse ERS, pro-inflammatory cytokine expression and apoptosis caused by 3cKD in adipocytes was associated with Ero1α-PDI signaling, 3cKD 3T3-L1 adipocytes were pretreated with the Ero1α inhibitor EN460 (10 μM) or the PDI inhibitor propynoic acid carbamoyl methyl amide 31 (PACMA31; 0.4 μM) alone or in combination for 2 hours before stimulation with 500 μM PA for 48 hours. Compared with the vehicle treatment, EN460 and PACMA31 alone or in combination reversed the decreased adiponectin mRNA levels, the increased mRNA levels of pro-inflammatory cytokines and ERS markers, the increased intracellular H₂O₂ content and the enhanced MAPK pathway signaling and adipocyte apoptosis caused by 3cKD; the combination of EN460 and PACMA31 had the strongest rescue effects (Fig. 5I-L). Taken together, the above results indicate that under lipotoxicity injury, Serpina3c can abolish ER overoxidation and ERS, at least partially by inhibiting HIF1α-mediated Ero1α-PDI signaling to block the downstream MAPK signaling pathway, suppress pro-inflammatory cytokine expression and apoptosis, and promote adiponectin expression in adipocytes.

Serpina3c overexpression or HIF1α knockdown in eWAT alleviates eWAT, serum, and heart inflammation in Serpina3c^–/– mice

To investigate whether adipocyte-specific overexpression of Serpina3c or knockdown of HIF1α in eWAT could mitigate metabolic inflammation in Serpina3c^–/– mice, 8-week-old Serpina3c^–/– mice were fed HFD for 2 weeks, and then their bilateral eWAT was injected with adiponectin-promoter-driven adeno-associated virus serotype 8 (Adipoq-AAV8) containing the Serpina3c coding sequence or an HIF1α interference sequence or with a negative control vector (Fig. 6A and B). After 12 weeks of HFD feeding, body weights and eWAT volumes were obviously lower in AAV-Serpina3c mice and AAV-shHIF1α mice than AAV-Vector mice (Fig. 6C-F). Histopathological eWAT staining showed decreases in adipocyte size, collagen deposition and CLS numbers in AAV-Serpina3c mice and AAV-shHIF1α mice compared to HFD-fed AAV-Vector mice (Fig. 6G). Furthermore, TEM results showed that the degree of ER swelling in the eWAT of AAV-Serpina3c mice and AAV-shHIF1α mice was less than that of HFD-fed AAV-Vector mice (Fig. 6H). Compared with the HFD-fed AAV-Vector mice, AAV-Serpina3c mice and AAV-shHIF1α mice had upregulated adiponectin expression, down-regulated mRNA levels of pro-inflammatory cytokines, Ero1α, PDIA3, PDIA4, and ERS markers, and reduced H₂O₂ content and apoptosis in eWAT; these data were not significantly different between the AAV-Serpina3c and AAV-shHIF1α mice (Fig. 6I-L).

Consistently, the serum levels of pro-inflammatory cytokines (TNF-α, IL-6, and CCL2) and myocardial injury makers (LDH, CK-MB) were significantly increased, while serum adiponectin levels were markedly decreased in AAV-Serpina3c mice and AAV-shHIF1α mice compared to the HFD-fed AAV-Vector mice (Supplementary Fig. 6A-C). Our histopathological data showed that compared with HFD-fed AAV-Vector mice, both AAV-Serpina3c mice and AAV-shHIF1α mice exhibited reduced collagen deposition, interstitial fibrosis and macrophage infiltration in their hearts (Supplementary Fig. 6D). TNF-α, TGF-β, and COL1A1 mRNA levels in the heart were significantly lower in AAV-Serpina3c mice and AAV-shHIF1α mice than HFD-fed AAV-Vector mice (Supplementary Fig. 6E).

In addition, CFs were treated with CM from in vitro-cultured fresh eWAT of HFD-fed AAV-Vector, AAV-Serpina3c, or AAV-shHIF1α mice. Compared with CM from the eWAT of HFD-fed AAV-Vector mice, that from the eWAT of AAV-Serpina3c and AAV-shHIF1α mice inhibited the gene expression of fibrosis markers in CFs (Supplementary Fig. 6F). There were no significant differences in using CM from the eWAT of AAV-Serpina3c and AAV-shHIF1α mice. Together, these results demonstrate that in HFD-fed Serpina3c^–/– mice, adipocyte-specific 3cOV or HIF1α knockdown in eWAT can improve ER overoxidation and ERS in eWAT to decrease eWAT inflammation and apoptosis, thereby alleviating serum inflammation and heart damage.

DISCUSSION

Under physiological conditions, eWAT can secrete protective adipokines such as adiponectin and omentin. However, in states of overnutrition, the secretion of anti-inflammatory adipokines decreases, and large amounts of pro-inflammatory cytokines are secreted by eWAT [12,26]. Therefore, obese individuals with preferential vWAT expansion (i.e., apple-shaped or abdominal obesity) are more prone to metabolic inflammation than those who have preferential sWAT enlargement (i.e., pear-shaped or gluteofemoral obesity) [27]. The eWAT inflammation is closely related to the occurrence and development of heart damage during obesity [28]. Given the centrality of eWAT in obesity and metabolic inflammation, as well as the technical difficulties of direct heart intervention targets, exploring eWAT as a target could result in feasible anti-heart damage prevention and treatment methods. In the present study, we found that Serpina3c deficiency exacerbated inflammation in the eWAT, serum and heart of HFD-fed mice; importantly, these effects could be alleviated by adipocyte-specific 3cOV or HIF1α knockdown in eWAT. Furthermore, consistent with the changes observed in Serpina3c^–/– mice, obese WT mice with eWAT localized knockdown of Serpina3c also exhibited more severe eWAT remodeling (expansion, inflammation, and fibrosis), serum inflammation and heart damage compared to control mice (Supplementary Fig. 7).

The main source of eWAT inflammation in obesity is the CLSs, which can continuously produce inflammatory cytokines such as TNF-α and IL-6. These cytokines enter the bloodstream to cause serum inflammation, and they reach the distal heart to induce chronic and sustained heart inflammation, which ultimately progresses to obesity-related heart damage [23,24]. Our data show that Serpina3c deficiency significantly increased the CLS numbers in the eWAT of HFD-fed mice. During obesity, a vicious circle forms between hypertrophic adipocytes and infiltrating macrophages in eWAT. Dysfunctional adipocytes can secrete chemokines, especially CCL2, which is coupled with adipocyte death, to recruit monocytes from the blood to eWAT, forming CLS with the resident macrophages and promoting macrophage M1 polarization [29]. In this study, we discovered that Serpina3c deficiency promoted adipocyte apoptosis, increased the expression of pro-inflammatory cytokines, including IL-6, CCL2, CCL5, CXCL1, and CXCL10, in vivo and in vitro, and decreased the expression of adiponectin. Consistent with these results, CM from PA-treated 3cKD adipocytes facilitated the migration ability and M1 polarization of Raw264.7 cells compared to that from the control LV3 group. Our data support the vital role of Serpina3c ablation in CLS formation in the eWAT of HFD-fed mice by promoting adipocyte apoptosis and chemokine production.

Based on our RNA-sequencing analysis, 3cKD appears to be positively correlated with the following pathways: protein processing in ER pathway, HIF-1 signaling pathway and cytokine-cytokine receptor interaction pathway. In this research, we attempted to clarify the relationships among these pathways. The oxidation and isomerization of intramolecular disulfide bonds are essential for proper protein folding and maturation in the ER. Ero1α/PDI signaling constitutes a pivotal pathway for oxidative protein folding in the ER. Ero1α catalyzes the de novo formation of disulfide bonds by coupling PDI oxidation to O₂ reduction, which results in H₂O₂ production, and oxidized PDI can introduce disulfide bonds into the reduced substrates. The Ero1α/PDI oxidative folding pathway releases H₂O₂ as the major source of ROS in the ER. Over-activation of Ero1α/PDI signaling implies H₂O₂ overproduction in the ER, leading to abnormal increases in ER oxidative stress and misfolded protein accumulation, ultimately causing persistent or excessive ERS [30,31].

There are three typical signals of unfolded protein responses in response to ERS: inositol-requiring enzyme 1 and XBP1; protein kinase R-like ER kinase and eIF2α; and ATF6. It is well known that ERS is closely related to cell apoptosis and inflammatory responses, in which the MAPK signaling pathway plays a pivotal role, especially JNK and P38 MAPK [32,33]. ERS has also been reported to down-regulate adiponectin expression in adipocytes [34]. In this study, we identified PDIA3 and PDIA4 as HIF1α target genes for the first time. Yan et al. [25] previously reported that Ero1α is the target gene of HIF1α, and hyperhomocysteinemia promotes H₂O₂ accumulation by up-regulating the HIF1α-Ero1α pathway, thus activating ER overoxidation and ERS in adipocytes; this pathway can be suppressed by the HIF1α inhibitor ACF. It has also been reported that hyperhomocysteinemia can up-regulate Ero1α expression to activate ERS-induced macrophage apoptosis, which can be attenuated by the Ero1α inhibitor EN460 [35]. Su et al. [36] found that the PDI inhibitor PACAM31 reversed palmitate-induced increases in ERS and the expression of pro-inflammatory cytokine, as well as decreases in adiponectin expression in adipocytes. Our data show that compared with the PA-treated control group, 3cKD in adipocytes up-regulated HIF1α/Ero 1α-PDI signaling, thus activating ER overoxidation, ERS and the MAPK signaling pathway, and these effects could be blocked by rSerpina3c, ACF, EN460, or PACAM31. In addition, ERS inhibitor tauroursodeoxycholate could reverse the pro-inflammatory response in PA-treated 3cKD 3T3-L1 adipocytes (Supplementary Fig. 8). There were no significant differences in basal levels of H₂O₂ and ERS among LV3, 3cKD, LV5, and 3cOV 3T3-L1 adipocytes (Supplementary Fig. 9). One limitation of this study is that we did not elucidate the mechanism through which Serpina3c regulates HIF1α; we plan to investigate this in future studies.

In conclusion, we have revealed a new mechanistic insight into the role of Serpina3c in alleviating metabolic inflammation in obese mice (Supplementary Fig. 10). We identified Serpina3c as an important regulator of adipocyte ER redox homeostasis in vitro and in vivo, and this protein acts via inhibiting HIF1α/Ero1α-PDI-mediated ER overoxidation and ERS. Ultimately, the Serpina3c/HIF1α/Ero1α-PDI signaling in eWAT may be a potential therapeutic target for obesity-related eWAT dysfunction and heart damage.

SUPPLEMENTARY MATERIALS

NOTES

CONFLICTS OF INTEREST

No potential conflict of interest relevant to this article was reported.

AUTHOR CONTRIBUTIONS

Conception or design: Y.J., Y.Y.Y.

Acquisition, analysis, or interpretation of data: all authors.

Drafting the work or revising: Y.J., G.S.M., Y.Y.Y.

Final approval of the manuscript: all authors.

FUNDING

This research was supported by National Natural Science Foundation of China (grant number NSFC 82370443).

ACKNOWLEDGMENTS

None

Fig. 1.

Serine (or cysteine) peptidase inhibitor, clade A, member 3C (Serpina3c) knockdown (3cKD) in adipocytes enhanced endoplasmic reticulum overoxidation and endoplasmic reticulum stress (ERS), and promoted the expression of pro-inflammatory cytokines and adipocyte apoptosis under lipotoxicity injury. (A) Western blotting of Serpina3c protein level in 3cKD 3T3-L1 adipocytes and its control LV3 group, as well as Serpina3c overexpression (3cOV) 3T3-L1 adipocytes and its control LV5 group. (B) Cell counting kit-8 (CCK8) assay measured the cell viability (%) of LV3 group and 3cKD group after 500 μM palmitic acid (PA)-treated for 24 or 48 hours. (C-J) In these experiments, LV3, 3cKD, LV5, and 3cOV groups were treated by 500 μM PA for 48 hours. (C, D) The relative mRNA levels of indicated genes in 3T3-L1 adipocytes. (E) The protein levels of indicated genes in 3T3-L1 adipocytes and quantification of the relative protein band density (n=3 for each group). (F) Hydrogen peroxide (H₂O₂) level in 3T3-L1 adipocytes was determined. (G) The protein levels of ERS makers in 3T3-L1 adipocytes and quantification of the relative protein band density (n=3 for each group). (H) Reactive oxygen species (ROS) level in 3T3-L1 adipocytes was detected. (I) The protein levels of mitogen-activated protein kinase signaling pathway in 3T3-L1 adipocytes and quantification of the relative protein band density (n=3 for each group). (J) Caspase-3 activity in 3T3-L1 adipocytes was determined. Data were presented as mean±standard error of the mean (n=5 for each group unless otherwise mentioned). NS, no significance; IL-6, interleukin-6; CCL2 or CCL5, C-C motif chemokine ligand 2 or 5; CXLC1 or CXCL10, C-X-C motif chemokine ligand 1 or 10; HIF1α, hypoxia-inducible factor 1α; Ero1α, endoplasmic reticulum oxidoreductase 1α; PDIA3 or PDIA4, protein disulfide isomerase family A member 3 or 4; GRP78, glucose regulated protein 78; CHOP, C/EBP homologous protein; p-eIF2α, phosphorylated eukaryotic initiation factor 2α; c-ATF6, cleaved activating transcription factor 6; XBP1S, spliced X-box binding protein 1; p-JNK, phosphorylated c-Jun N-terminal kinase; p-ERK, phosphorylated extracellular signal-regulated kinase. ^aP<0.05, ^bP<0.01, ^cP<0.001, ^dP<0.0001.

Fig. 2.

Serine (or cysteine) peptidase inhibitor, clade A, member 3C (Serpina3c) deficiency increased epididymal white adipose tissue (eWAT) size, fibrosis and macrophage infiltration in high-fat diet (HFD)-fed mice. Male 8-week-old wild type (WT) and Serpina3c global knockout (Serpina3c^–/–) mice were fed a chow diet (CD) or HFD for 12 weeks. (A) Body weight change of mice. (B) Representative photographs of mice. (C) Representative pictures of eWAT. (D) eWAT weight. (E) eWAT weight to body weight ratio. (F) Fasting blood glucose of mice. (G) Food intake of mice. (H) Representative images of hematoxylin-eosin (H&E) staining, Masson’s trichrome staining and macrophage marker F4/80 immunohistochemistry (IHC) staining of eWAT sections. Scale bar: 100, 50, 100 µm, respectively. (I, J, K) Quantitative analysis of average adipocyte area, fibrosis area of eWAT (%), and crown-like structure (CLS) number of per field in (H). (L) Serum levels of triglyceride (TG), total cholesterol (TC), non-esterified fatty acid (NEFA) in mice. Data were presented as mean±standard error of the mean (n=5 for each group). NS, no significance. ^aP<0.05, ^bP<0.01, ^cP<0.001.

Fig. 3.

Serine (or cysteine) peptidase inhibitor, clade A, member 3C (Serpina3c) ablation exacerbated endoplasmic reticulum (ER) overoxidation, inflammation and apoptosis of epididymal white adipose tissue (eWAT) in the high-fat diet (HFD)-fed mice. These mice were the same batch of 12 week-HFD-fed wild type (WT) and Serpina3c global knockout (Serpina3c^–/–) mice as shown in Fig. 2. (A) The relative mRNA levels of indicated genes in eWAT. (B) The protein levels of indicated genes in eWAT and quantification of the relative protein band density (n=3 for each group). (C) Hydrogen peroxide (H₂O₂) level in eWAT was measured. (D) The protein levels of mitogen-activated protein kinase signaling pathway in eWAT and quantification of the relative protein band density (n=3 for each group). (E) Caspase-3 activity of eWAT was determined. (F) Representative images of terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) stained eWAT sections. Scale bar: 100 µm. (G) Representative photographs of ER in ultrathin eWAT sections observed by transmission electron microscopy (TEM) at 50,000 magnification (scale bar: 200 nm; red arrow: ER). Data were presented as mean±standard error of the mean (n=5 for each group). TNF-α, tumor necrosis factor-alpha; IL-6, interleukin-6; CCL2 or CCL5, C-C motif chemokine ligand 2 or 5; CXLC1 or CXCL10, C-X-C motif chemokine ligand 1 or 10; HIF1α, hypoxia-inducible factor 1α; Ero1α, endoplasmic reticulum oxidoreductase 1α; PDIA3 or PDIA4, protein disulfide isomerase family A member 3 or 4; GRP78, glucose regulated protein 78; CHOP, C/EBP homologous protein; p-JNK, phosphorylated c-Jun N-terminal kinase; p-ERK, phosphorylated extracellular signal-regulated kinase; DAPI, 4’,6-diamidino-2-phenylindole. ^aP<0.05, ^bP<0.01, ^cP<0.001.

Fig. 4.

Serine (or cysteine) peptidase inhibitor, clade A, member 3C (Serpina3c) ablation exacerbated serum and heart inflammation in high-fat diet (HFD)-fed mice. These mice were the same batch of 12 week-HFD-fed wild type (WT) and Serpina3c global knockout (Serpina3c^–/–) mice as shown in Fig. 2. (A) Detection of serum pro-inflammatory cytokines in mice. (B) Detection of serum adiponectin level in mice. (C) Determination of serum myocardial injury indexes like lactate dehydrogenase (LDH) and creatine kinase MB isoenzyme (CK-MB). (D) Representative images of hematoxylin-eosin (H&E) staining, Masson’s trichrome staining and F4/80 immunohistochemistry (IHC) staining of heart sections (scale bar: 50 µm). (E, F) Quantitative analysis of interstitial fibrosis of heart (%) and F4/80 positive area of heart (%) in (D). (G) Representative images of wheat germ agglutinin (WGA) stained heart sections and quantitative analysis of cardiomyocyte cross-sectional area (scale bar: 20 µm). (H) Representative images of Oil Red O stained heart sections. (I) Determination of triglyceride (TG) content in the heart. (J) The relative mRNA levels of tumor necrosis factor-alpha (TNF-α), transforming growth factor-β (TGF-β) and collagen type I, α 1 chain (COL1A1) in the heart. (K) Cardiac fibroblasts were treated for 24 hours with conditioned medium (CM) obtained from in vitro 12-hour-cultured fresh epididymal white adipose tissue (eWAT) in HFD-fed WT or Serpina3c^–/– mice, and then the relative mRNA levels of f ibrosis markers were detected. Data were presented as mean±standard error of the mean (n=5 for each group). IL-6, interleukin-6; IL-1β, interleukin-1β; CCL2, C-C motif chemokine ligand 2; NS, no significance; CD, chow diet. ^aP<0.05, ^bP<0.01.

Fig. 5.

Serine (or cysteine) peptidase inhibitor, clade A, member 3C (Serpina3c) inhibited endoplasmic reticulum overoxidation and its downstream in adipocytes at least partially by suppressing hypoxia-inducible factor 1α (HIF1α)/endoplasmic reticulum oxidoreductase 1α (Ero1α)-protein disulfide isomerase (PDI) signaling. (A) The effect of transcription regulatory factor HIF1α on the promoter activity of protein disulfide isomerase family A member 3 (PDIA3) and PDIA4 in 293T cells was detected by Dual-luciferase reporter assay. (B-H) Serpina3c knockdown (3cKD) 3T3-L1 adipocytes were pretreated with 100 ng/mL recombinant mouse Serpina3c protein (rSerpina3c) or 5 μM HIF1α inhibitor acriflavine (ACF) for 2 hours before being stimulated with 500 μM palmitic acid (PA) for 48 hours. (B) The mRNA levels of indicated genes in 3cKD 3T3-L1 adipocytes. (C, D) The protein levels of indicated genes in 3cKD 3T3-L1 adipocytes and quantification of the relative protein band density (n=3 for each group). (E) Hydrogen peroxide (H₂O₂) level in 3cKD 3T3-L1 adipocytes was determined. (F, G) The protein levels of mitogen-activated protein kinase (MAPK) signaling pathway in 3cKD 3T3-L1 adipocytes and quantification of the relative protein band density (n=3 for each group). (H) Caspase-3 activity in 3cKD 3T3-L1 adipocytes was determined. (I-L) 3cKD 3T3-L1 adipocytes were pretreated with 10 μM Ero1α inhibitor EN460 or 0.4 μM PDI inhibitor propynoic acid carbamoyl methyl amide 31 (PACMA31) alone or in combination for 2 hours before being stimulated with 500 μM PA for 48 hours. (I) The mRNA levels of indicated genes in 3T3-L1 adipocytes. (J) H₂O₂ level in 3cKD 3T3-L1 adipocytes was measured. (K) The protein levels of MAPK signaling pathway in 3cKD 3T3-L1 adipocytes and quantification of the relative protein band density (n=3 for each group). (L) Caspase-3 activity in 3cKD 3T3-L1 adipocytes was determined. Data were presented as mean±standard error of the mean. MCS, multiple cloning site; 6xHis, hexahistidine; IL-6, interleukin-6; CCL2 or CCL5, C-C motif chemokine ligand 2 or 5; CXLC1 or CXCL10, C-X-C motif chemokine ligand 1 or 10; GRP78, glucose regulated protein 78; CHOP, C/EBP homologous protein; p-JNK, phosphorylated c-Jun N-terminal kinase; p-ERK, phosphorylated extracellular signal-regulated kinase; NS, no significance. ^aP<0.05, ^bP<0.01, ^cP<0.001, ^dP<0.0001.

Fig. 6.

Adeno-associated virus (AAV) mediated adipocyte-specific overexpression of serine (or cysteine) peptidase inhibitor, clade A, member 3C (Serpina3c) or knockdown of hypoxia-inducible factor 1α (HIF1α) in epididymal white adipose tissue (eWAT) alleviated endoplasmic reticulum (ER) overoxidation, fibrosis and macrophage infiltration in eWAT of high-fat diet (HFD)-fed Serpina3c global knockout (Serpina3c^–/–) mice. (A) Scheme of Serpina3c overexpression or HIF1α knockdown in the eWAT of Serpina3c^–/– mice. (B) Western blotting of Serpina3c protein expression in eWAT of mice. (C) Body weight change of mice. (D) Body weight of mice at 12-week-HFD feeding. (E) Representative photographs of mice. (F) Representative pictures of eWAT. (G) Representative images of hematoxylin-eosin (H&E) staining, Masson’s trichrome staining and F4/80 immunohistochemistry (IHC) staining of eWAT sections (scale bar: 100, 100, 50 µm, respectively). (H) Representative photographs of ER in ultrathin eWAT sections observed by transmission electron microscopy (TEM) at 50,000 magnification (scale bar: 200 nm; red arrow: ER). (I, J) The relative mRNA levels of indicated genes in eWAT. (K) Hydrogen peroxide (H₂O₂) level in eWAT was measured. (L) Caspase-3 activity in eWAT was determined. Data were presented as mean±standard error of the mean (n=5 for each group). Adipoq-AAV8, adiponectin-promoter-driven adeno-associated virus serotype 8; NS, no significance; TNF-α, tumor necrosis factor-alpha; IL-6, interleukin-6; CCL2 or CCL5, C-C motif chemokine ligand 2 or 5; CXLC1 or CXCL10, C-X-C motif chemokine ligand 1 or 10; Ero1α, endoplasmic reticulum oxidoreductase 1α; PDIA3 or PDIA4, protein disulfide isomerase family A member 3 or 4; GRP78, glucose regulated protein 78; CHOP, C/EBP homologous protein. ^aP<0.05, ^bP<0.01, ^cP<0.001.

REFERENCES

• 1. Welsh A, Hammad M, Pina IL, Kulinski J. Obesity and cardiovascular health. Eur J Prev Cardiol 2024;31:1026-35.ArticlePMCPDF
• 2. Joynt Maddox KE, Elkind MS, Aparicio HJ, Commodore-Mensah Y, de Ferranti SD, Dowd WN, et al. Forecasting the burden of cardiovascular disease and stroke in the United States through 2050-prevalence of risk factors and disease: a presidential advisory from the American Heart Association. Circulation 2024;150:e65-88.
• 3. Samson WK, Yosten GL, Remme CA. A primer on obesity-related cardiomyopathy. Physiol Rev 2022;102:1-6.Article
• 4. Swiatkiewicz I, Patel NT, Villarreal-Gonzalez M, Taub PR. Prevalence of diabetic cardiomyopathy in patients with type 2 diabetes in a large academic medical center. BMC Med 2024;22:195.
• 5. Rabadan-Chavez G, Diaz de la Garza RI, Jacobo-Velazquez DA. White adipose tissue: distribution, molecular insights of impaired expandability, and its implication in fatty liver disease. Biochim Biophys Acta Mol Basis Dis 2023;1869:166853.Article
• 6. Weijie Z, Meng Z, Chunxiao W, Lingjie M, Anguo Z, Yan Z, et al. Obesity-induced chronic low-grade inflammation in adipose tissue: a pathway to Alzheimer’s disease. Ageing Res Rev 2024;99:102402.
• 7. Gilani A, Stoll L, Homan EA, Lo JC. Adipose signals regulating distal organ health and disease. Diabetes 2024;73:169-77.PMC
• 8. Zhang X, Ha S, Lau HC, Yu J. Excess body weight: novel insights into its roles in obesity comorbidities. Semin Cancer Biol 2023;92:16-27.Article
• 9. Dahdah N, Tercero-Alcazar C, Malagon MM, Garcia-Roves PM, Guzman-Ruiz R. Interrelation of adipose tissue macrophages and fibrosis in obesity. Biochem Pharmacol 2024;225:116324.Article
• 10. Bradley D, Deng T, Shantaram D, Hsueh WA. Orchestration of the adipose tissue immune landscape by adipocytes. Annu Rev Physiol 2024;86:199-223.Article
• 11. Vasamsetti SB, Natarajan N, Sadaf S, Florentin J, Dutta P. Regulation of cardiovascular health and disease by visceral adipose tissue-derived metabolic hormones. J Physiol 2023;601:2099-120.ArticlePDF
• 12. Rana MN, Neeland IJ. Adipose tissue inflammation and cardiovascular disease: an update. Curr Diab Rep 2022;22:27-37.ArticlePDF
• 13. Li C, Qin D, Hu J, Yang Y, Hu D, Yu B. Inflamed adipose tissue: a culprit underlying obesity and heart failure with preserved ejection fraction. Front Immunol 2022;13:947147.ArticlePubMedPMC
• 14. Choi Y, Choi H, Yoon BK, Lee H, Seok JW, Kim HJ, et al. Serpina3c regulates adipogenesis by modulating insulin growth factor 1 and integrin signaling. iScience 2020;23:100961.ArticlePMC
• 15. Li BY, Guo YY, Xiao G, Guo L, Tang QQ. SERPINA3C ameliorates adipose tissue inflammation through the Cathepsin G/Integrin/AKT pathway. Mol Metab 2022;61:101500.ArticlePubMedPMC
• 16. AlZaim I, de Rooij LP, Sheikh BN, Borgeson E, Kalucka J. The evolving functions of the vasculature in regulating adipose tissue biology in health and obesity. Nat Rev Endocrinol 2023;19:691-707.ArticlePDF
• 17. Ajoolabady A, Lebeaupin C, Wu NN, Kaufman RJ, Ren J. ER stress and inflammation crosstalk in obesity. Med Res Rev 2023;43:5-30.ArticlePubMedPDF
• 18. Ma K, Zhang Y, Zhao J, Zhou L, Li M. Endoplasmic reticulum stress: bridging inflammation and obesity-associated adipose tissue. Front Immunol 2024;15:1381227.ArticlePubMedPMC
• 19. Jha V, Kumari T, Manickam V, Assar Z, Olson KL, Min JK, et al. ERO1-PDI redox signaling in health and disease. Antioxid Redox Signal 2021;35:1093-115.ArticlePubMedPMC
• 20. Victor P, Sarada D, Ramkumar KM. Crosstalk between endoplasmic reticulum stress and oxidative stress: focus on protein disulfide isomerase and endoplasmic reticulum oxidase 1. Eur J Pharmacol 2021;892:173749.ArticlePubMed
• 21. Hotamisligil GS, Davis RJ. Cell signaling and stress responses. Cold Spring Harb Perspect Biol 2016;8:a006072.ArticlePubMedPMC
• 22. Pan B, Sun J, Liu Z, Wang L, Huo H, Zhao Y, et al. Longxuetongluo Capsule protects against cerebral ischemia/reperfusion injury through endoplasmic reticulum stress and MAPK-mediated mechanisms. J Adv Res 2021;33:215-25.ArticlePubMedPMC
• 23. Kolb H. Obese visceral fat tissue inflammation: from protective to detrimental? BMC Med 2022;20:494.ArticlePubMedPMCPDF
• 24. Piche ME, Tchernof A, Despres JP. Obesity phenotypes, diabetes, and cardiovascular diseases. Circ Res 2020;126:1477-500.ArticlePubMed
• 25. Yan Y, Wu X, Wang P, Zhang S, Sun L, Zhao Y, et al. Homocysteine promotes hepatic steatosis by activating the adipocyte lipolysis in a HIF1α-ERO1α-dependent oxidative stress manner. Redox Biol 2020;37:101742.ArticlePubMedPMC
• 26. Pestel J, Blangero F, Watson J, Pirola L, Eljaafari A. Adipokines in obesity and metabolic-related-diseases. Biochimie 2023;212:48-59.PubMed
• 27. Fu J, Hofker M, Wijmenga C. Apple or pear: size and shape matter. Cell Metab 2015;21:507-8.ArticlePubMed
• 28. Oikonomou EK, Antoniades C. The role of adipose tissue in cardiovascular health and disease. Nat Rev Cardiol 2019;16:83-99.ArticlePDF
• 29. Chavakis T, Alexaki VI, Ferrante AW Jr. Macrophage function in adipose tissue homeostasis and metabolic inflammation. Nat Immunol 2023;24:757-66.ArticlePubMedPDF
• 30. Wang L, Wang CC. Oxidative protein folding fidelity and redoxtasis in the endoplasmic reticulum. Trends Biochem Sci 2023;48:40-52.ArticlePubMed
• 31. Shergalis AG, Hu S, Bankhead A 3rd, Neamati N. Role of the ERO1-PDI interaction in oxidative protein folding and disease. Pharmacol Ther 2020;210:107525.ArticlePubMedPMC
• 32. Celik C, Lee SY, Yap WS, Thibault G. Endoplasmic reticulum stress and lipids in health and diseases. Prog Lipid Res 2023;89:101198.ArticlePubMed
• 33. Almanza A, Carlesso A, Chintha C, Creedican S, Doultsinos D, Leuzzi B, et al. Endoplasmic reticulum stress signalling: from basic mechanisms to clinical applications. FEBS J 2019;286:241-78.ArticlePubMedPDF
• 34. Hosogai N, Fukuhara A, Oshima K, Miyata Y, Tanaka S, Segawa K, et al. Adipose tissue hypoxia in obesity and its impact on adipocytokine dysregulation. Diabetes 2007;56:901-11.ArticlePubMedPDF
• 35. Zhang N, Zhu L, Wu X, Yan R, Yang S, Jiang X, et al. The regulation of Ero1-alpha in homocysteine-induced macrophage apoptosis and vulnerable plaque formation in atherosclerosis. Atherosclerosis 2021;334:39-47.ArticlePubMed
• 36. Su SC, Chien CY, Chen YC, Chiang CF, Lin FH, Kuo FC, et al. PDIA4, a novel ER stress chaperone, modulates adiponectin expression and inflammation in adipose tissue. Biofactors 2022;48:1060-75.ArticlePubMedPDF

Figure & Data

References

Citations

Citations to this article as recorded by  

About this article

Jiang Y, Guo JQ, Wu Y, Zheng P, Wang SF, Yang MC, Ma GS, Yao YY. Serpina3c Mitigates Adipose Tissue Inflammation by Inhibiting the HIF1α-Mediated Endoplasmic Reticulum Overoxidation in Adipocytes. Diabetes Metab J. 2025 May 22. doi: 10.4093/dmj.2024.0441. Epub ahead of print.

Received: Jul 31, 2024; Accepted: Feb 25, 2025

DOI: https://doi.org/10.4093/dmj.2024.0441.

Download citation ↓