Regulation of the unfolded protein response transducer IRE1α by SERPINH1 aggravates periodontitis with diabetes mellitus via prolonged ER stress

The hyperglycemic microenvironment induced by diabetes mellitus aggravates the inammatory response, in which the inositol-requiring enzyme-1α (IRE1α) signal transduction pathway of the unfolded protein response (UPR) participates. This study aimed to investigate the mechanism by which hyperglycemia regulates the IRE1α signaling pathway and affects endoplasmic reticulum (ER) homeostasis in human gingival epithelium in periodontitis with diabetes mellitus (DP). to mice Statistical to determine the relative mRNA level of SERPINH1; *p < the SEM. E. SERPINH1 expression in HGECs overexpressing SERPINH1 treated with a high concentration of glucose was detected by qRT-PCR. Statistical analysis of the qRT-PCR data was carried out to determine the relative mRNA level of SERPINH1; determined gene expression in was detected Statistical analysis of the data was carried out to determine the relative mRNA levels of IRE1α, XBP1-s, XBP1-u, GRP78 NLRP3 and IL-1β;


Abstract Background
The hyperglycemic microenvironment induced by diabetes mellitus aggravates the in ammatory response, in which the inositol-requiring enzyme-1α (IRE1α) signal transduction pathway of the unfolded protein response (UPR) participates. This study aimed to investigate the mechanism by which hyperglycemia regulates the IRE1α signaling pathway and affects endoplasmic reticulum (ER) homeostasis in human gingival epithelium in periodontitis with diabetes mellitus (DP).

Methods
Human gingival epithelium samples from healthy subjects, subjects with periodontitis and subjects with DP were collected, in vitro cultures of human gingival epithelial cells were challenged with a hyperglycemic microenvironment to observe the effects of diabetes on periodontal in ammation and to assess UPR-IRE1α signaling in human gingival epithelium in DP. Subsequently, RNA sequencing (RNAseq) data was analyzed to investigate the expression of ER-related genes in human gingival epithelium.
Furthermore, to explore the key role of serpin family H member 1 (SERPINH1) in the regulation of UPR-IRE1α signaling in a hyperglycemic microenvironment, experiments in SERPINH1-knockdown and SERPINH1-overexpression models were established in vitro.

Results
Diabetes causes a hyperin ammatory response in human gingival epithelium, which accelerates periodontal in ammation. A hyperglycemic microenvironment inhibited the inositol-requiring enzyme-1α / X-box binding protein 1 (IRE1α/XBP1) axis, decreased the expression of glucose regulated protein 78 (GRP78), and ultimately impaired the UPR, causing ER stress to be prolonged or more severe in human gingival epithelium. The RNA-seq and experiments revealed that the mechanism by which periodontitis is aggravated in individuals with diabetes mellitus may involve decreased SERPINH1 expression. SERPINH1 might act as an activator of IRE1α, maintaining human gingival epithelium homeostasis, suppressing nuclear factor-κB signaling pathway and reducing NOD-like receptor, pyrin domain containing protein 3 (NLRP3) and interleukin-1 beta (IL-1β) expression by preventing prolonged ER stress induced by highglucose conditions.

Conclusion
Regulation of the UPR transducer IRE1α by SERPINH1 alleviates DP by mitigating prolonged ER stress.

Page 3/27
Periodontitis is a global disease characterized by chronic in ammatory destruction of the periodontal supporting tissues, affecting the gingiva, periodontal ligament and alveolar bone, that eventually leads to tooth loss. Caused by dysbiosis of the oral microbiota, periodontitis is related to dysregulation of the immune-in ammatory response [1]. In recent years, epidemiological trends have shown that both the incidence and prevalence of type 2 diabetes mellitus have worsened worldwide. Diabetes mellitus causes a hyperin ammatory response to the periodontal microbiota and impairs the resolution of in ammation and repair, which accelerates periodontal in ammation and destruction [2]. In fact, periodontitis, the sixth most common complication of diabetes, is closely associated with hyperglycemia [3]. Therefore, the need to implement optimal treatment strategies for patients with periodontitis and diabetes mellitus is critical.
The endoplasmic reticulum (ER) is the key eukaryotic organelle for cellular protein synthesis and processing and maintains the stability of the intracellular environment [4]. Various extracellular and intracellular stresses can induce the continuous accumulation of misfolded or unfolded proteins in the ER and disturb the environment of the ER lumen, which results in ER stress that triggers cytoprotective signaling pathways, referred to as the unfolded protein response (UPR) [5]. Although activation of the UPR aims to restore cellular homeostasis, when the adaptive mechanisms put into motion by the UPR fail to compensate, a prolonged ER stress response can damage the target cells and trigger a proin ammatory response, ultimately inducing cell death [6,7]. Indeed, UPR activation has been acknowledged as the prominent feature of a number of pathological conditions, including diabetes, obesity, rheumatoid arthritis, and chronic in ammation [8][9][10][11]. Increasing evidence demonstrates that some UPR-related genes are upregulated in periodontitis [12][13][14]. In mammalian cells, the UPR is a signaling cascade initiated by three ER transmembrane sensors: inositol-requiring enzyme-1 (IRE1), double-stranded RNA-dependent protein kinase (PKR)-like ER kinase (PERK), and activating transcription factor 6 (ATF6). Compared to branches de ned by other stress sensors, the IRE1α branch is the most evolutionarily conserved signaling branch in the UPR. Activated IRE1α splices the mRNA encoding X-box binding protein 1 (XBP1) via its endoribonuclease (RNase) activity, thereby generating spliced XBP1 (XBP1-s), which activates the gene expression of UPR-associated regulators [15,16]. In the ER stress response, IRE1α activation and attenuation are related to cell fate and determine cellular survival or death [17]. Despite increasing progress toward an improved understanding of the functional importance of IRE1α signaling in in ammation, the molecular machinery that governs the dynamics of IRE1α activation and deactivation in periodontitis with diabetes mellitus remains largely elusive.
Serpin family H member 1 (SERPINH1), a major regulator of the UPR, is an ER chaperone that specializes in the maturation and tra cking of collagen and the most abundant ER client protein in mammals [18,19]. In certain models, SERPINH1 de ciency has been reported to result in ER stress-mediated apoptosis [20,21]. Recent studies also showed that SERPINH1 de ciency sensitized cells and animals to experimental ER stress, revealing the signi cance of SERPINH1 in maintaining cellular homeostasis. SERPINH1 can adjust IRE1α signaling to engage an adaptive UPR [22]. However, how SERPINH1 directly or indirectly regulates IRE1α signaling in the gingival epithelium in periodontitis with diabetes mellitus remains largely unknown. An understanding of the precise molecular mechanism by which SERPINH1 functions in the gingival epithelium may provide meaningful insight into the practical treatment of periodontitis with diabetes mellitus.
By conducting in vivo and in vitro experiments, we found the IRE1α/XBP1 pathway to be impaired and ER stress to be prolonged or more severe in the gingival epithelium in periodontitis with diabetes mellitus. In vitro SERPINH1-knockdown and SERPINH1-overexpression experiments revealed the molecular mechanism by which SERPINH1 regulates IRE1α signaling, suggesting a biological role for SERPINH1 in the gingival epithelium.

Materials And Methods
Cell culture and stimulation Human gingival epithelial cells (HGECs) obtained from the American Type Culture Collection (ATCC; Manassas, VA, USA) were cultured in DMEM-F12 medium containing 10% fetal bovine serum (FBS), 100 U/mL penicillin and 100 µg/mL streptomycin at 37°C in a humidi ed CO 2 incubator. To simulate the in ammatory microenvironment in periodontitis in vitro, 1 µg/mL P. gingivalis LPS (InvivoGen, San Diego, CA, USA) was added to the culture medium for 6 and 12 h. To simulate the microenvironment in type 2 diabetes in vitro, HGECs were stimulated with glucose at a high concentration (25 mM) in the culture medium for 48 h. HGECs cultured in medium with 5.5 mM glucose were used as negative controls. Later, the HGECs were harvested for analysis at the indicated time points.
RNA isolation and reverse transcription quantitative PCR (qRT-PCR) HGECs were suspended in 1 mL of TRIzol reagent (Invitrogen, Carlsbad, CA, USA) and stored at − 80°C.
Total RNA was puri ed using TRIzol reagent following the manufacturer's instructions and quanti ed with a Nanodrop 2000. cDNA was synthesized from 500 ng of total RNA using Prime Script RT Master Mix (Toyobo Co, Ltd, Osaka, Japan). Real-time PCR was performed in a 96-well optical reaction plate using SYBR PCR Master Mix (Roche, Indianapolis, IN, USA). mRNA expression was assayed on a Bio-Rad CFX96 detection system (Roche, Sweden). Details of the RT-qPCR primers used in this experiment are shown in Table 1. Relative SERPINH1, GRP78, IRE1, TRAF2, NLRP3, IL-1β, XBP1-s and XBP1-u expression was quanti ed with the 2-ΔΔCt method using GAPDH expression as an endogenous control.

Western blot analysis
Total protein samples from cultured HGECs were extracted using radioimmunoprecipitation assay (RIPA) buffer (Millipore, MA, USA) for 30 min on ice. The lysates were further homogenized with an ultrasonic homogenizer and centrifuged at 12,000 rpm at 4°C for 20 min. Then, the total protein concentration in the RIPA-extracted lysates was quanti ed using a bicinchoninic acid (BCA) protein assay kit (CWBioTech, Beijing, China). Equal amounts of proteins were separated using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gels and subsequently transferred onto polyvinylidene uoride (PVDF) membranes (Millipore, MA, USA). Buffer containing 5% nonfat milk was used to block the PVDF membranes for 60 min at room temperature. To detect the proteins, PVDF membranes containing the proteins were sequentially incubated with primary antibody overnight at 4°C overnight, and anti-tubulin and anti-GAPDH antibodies were used as loading controls. The membranes were washed three times in TBST for 10 min and then incubated with horseradish peroxidase-conjugated secondary antibody for 1 h at room temperature. Image acquisition was performed using a chemiluminescence kit (Millipore, MA, USA), and target band densitometric analysis was performed using ImageJ software (National Institutes of Health, Bethesda, Maryland, USA). The antibodies and their concentrations are listed in Table 2.

IHC staining
The human gingival epithelium was collected, xed with 4% paraformaldehyde and then embedded in para n. Four-micrometer-thick sections were cut and stained for IHC analysis. Antigen retrieval was performed by placing the sections in 0.01 mol/L sodium citrate buffer (pH 6.0) at 95-100°C in a microwave oven for 20 min. Then, the sections were incubated overnight at 4°C with the corresponding primary antibody. After removal of the primary antibody, the sections were incubated with a goat antimouse/rabbit secondary antibody (GeneTech, Shanghai) at room temperature for 1 h. After removal of the secondary antibody, the sections were developed using a diaminobenzidine (DAB) solution (GeneTech) according to the manufacturer's protocols, counterstained with hematoxylin for 3 min, dehydrated, mounted with neutral gum and imaged. Staining results were measured with Image-Pro Plus 6.0 by the average optical density (AOD). The antibodies and their concentrations are listed in Table 2.

Intracellular calcium concentration detection
The intracellular calcium concentration was measured with Fluo-4AM (Shanghai Beyotime Bio-Tech Co., Ltd.) according to manufacturer's instructions. In detail, cells were washed twice with phosphate-buffered saline (PBS) and digested with trypsin for 5 min at 37°C. An equal quantity of DMEM supplemented with 10% FBS was added to terminate digestion. The cells were centrifuged at 1,000 rpm for 5 min, the supernatant was removed, and the cells were washed once with PBS and centrifuged again to obtain a cell pellet. Subsequently, the cells were stained with 200 µL of Fluo-4 AM ( nal concentration of 5 µM), incubated for 30 min in PBS at 37°C, washed three times with PBS and incubated in PBS for an additional 15 min in the absence of Fluo-4AM to allow complete de-esteri cation of the dye. The uorescence intensity was obtained by ow cytometry with 488-nm laser excitation and a 512-520-nm emission lter.
Bioinformatic analysis RNA-seq data of the gingival epithelium of mice with periodontitis with diabetes mellitus and mice with periodontitis were obtained as we reported before [23]. Bioinformatic analysis were processed and executed by RStudio software and limma [24] package. Differentially expressed genes (DEGs) analysis was performed based on an adjusted p value of < 0.05 and an absolute log2 (fold change) of > 2. Volcano plots and heat maps have been constructed to present the ndings of the experiment.

Transient siRNA transfection and interference assay
HGECs at an appropriate density were seeded into plates. After 24 h of culture in antibiotic-free normal growth medium, the HGECs were transfected with SERPINH1 siRNA (50 nM, RiboBio Co., Ltd., Guangzhou, China) and negative control siRNA (si-NC, 50 nM, RiboBio Co., Ltd., Guangzhou, China) by employing riboFect™ CP (RiboBio Co. Ltd., Guangzhou, China) transfection reagent according to the manufacturer's protocol. Untreated HGECs and HGECs treated with nontargeting scrambled siRNA were used as controls. The HGECs were harvested 48 h after transfection for qRT-PCR and western blotting.

Establishment of stable SERPINH1-overexpressing HGEC lines
To generate HGECs stably expressing SERPINH1, HBLV-h-SERPINH1-3x ag-ZsGreen-PURO (Han Biotech, Shanghai, China) was transfected into HGECs. The titers of lentivirus employed in this experiment were 1.5 × 10^8 TU/mL, and the multiplicity of infection (MOI) was 100. Brie y, HGECs were seeded in 24-well plates at a density of 1.5 × 10 5 cells per well. When the cells reached 30-50% con uence, HBLV-h-SERPINH1-3x ag-ZsGreen-PURO was added to each well following the manufacturer's guidelines. For viral transfection, 4 µg/mL polybrene (Han Biotech, Shanghai, China) was added to the lentiviral supernatant for 24 h to improve the infection e ciency. After 72 h, stable transfectants were selected by the addition of puromycin (2 µg/mL, Han Biotech, Shanghai, China) to the transfected cells for 14 days. Stable SERPINH1-overexpressing HGEC lines were established and con rmed by uorescence microscopy, western blot analysis and real-time PCR. Then, the successfully transfected cells were subjected to high-glucose conditions for 48 h. Null-HGECs were infected with lentivirus expressing the anti-puromycin gene and ZsGreen and used as a negative control.

Statistical analyses
Comparisons between groups were evaluated by unpaired two-tailed Student's t test or one-way analysis of variance (ANOVA) with Tukey's multiple comparison test. Correlation analyses were performed by Pearson's correlation analysis. Statistical analyses were performed with Prism software (GraphPad).
Differences for which P < 0.05 were considered statistically signi cant.

A high-glucose microenvironment aggravates periodontal in ammation
To determine whether a high-glucose microenvironment aggravates in ammation, in vitro experiments were performed to investigate the in ammatory response in human gingival epithelial cells (HGECs) in a high-glucose microenvironment. As shown in Fig. 1A-1F, compared with those in HGECs without highglucose treatment, the expression levels of p-p65, NOD-like receptor, pyrin domain containing protein 3 (NLRP3) and interleukin-1 beta (IL-1β) in HGECs treated with high glucose were signi cantly increased.
The signi cant differences in p65, NLRP3, and IL-1β expression in the human gingival epithelium (n = 14) between the periodontitis (P) group and periodontitis with diabetes mellitus (DP) group were con rmed by immunohistochemical staining. Compared with that in the P group, p65, NLRP3, and IL-1β immunoreactivity in the human gingival epithelium was signi cantly upregulated in the DP group ( Fig. 1G-1L). These results showed that the hyperglycemic microenvironment induced by diabetes mellitus can lead to p65 phosphorylation; p65 nuclear accumulation; nuclear factor-κB (NF-κB) activation; and IL-1β production.
A hyperglycemic microenvironment induced prolonged ER stress responses in the human gingival epithelium Gingival epithelium ER homeostasis play a key role in defending against exogenous infection. An increasing number of studies has indicated that in cells are subjected to prolonged ER stress, the intracellular calcium concentration increases and disrupts ER homeostasis [25]. To clarify how a highglucose microenvironment impairs HGEC function, ow cytometry analysis was performed to examine cytosolic Ca 2+ stained with Fluo-4AM. Intriguingly, a high-glucose microenvironment induced a signi cantly increase in the intracellular calcium ion concentration in HGECs ( Fig. 2A, 2B), suggesting that a high-glucose microenvironment induces prolonged ER stress responses and perturbs ER homeostasis in HGECs.
To investigate the mechanisms responsible for prolonged ER stress responses in periodontitis under diabetic conditions, we sought to understand the role of IRE1α signaling within HGECs under hyperglycemia. IRE1α is a bifunctional enzyme that processes an RNase domain and a kinase domain. We detected IRE1α activity in HGECs; the treatment of HGECs with glucose at a high concentration (25 mM) for 48 h diminished the expression of IRE1α and increased the levels of phosphorylated IRE1α (p-IRE1α) compared to those in HGECs treated with 5.5 mM glucose (Fig. 2C-2E), indicating that the expression and activity of IRE1α are defective in HGECs cultured in high-glucose conditions. IRE1α plays an important role in maintaining ER homeostasis by initiating the unconventional splicing of XBP1 mRNA to create a translational frame shift in XBP1 mRNA. This produces a potent transcription factor, XBP1-s, which regulates the expression of genes with functions in ER protein folding and tra cking and ERassociated degradation to preserve ER homeostasis [26]. We used real-time polymerase chain reaction (PCR) to measure the mRNA expression levels of ER stress markers in HGECs treated with 5.5 mM or 25 mM glucose for 48 h. The level of un-spliced XBP1 (XBP1-u) mRNA was increased in HGECs treated with 25 mM glucose compared with HGECs treated with 5.5 mM glucose, but a corresponding increase in XBP1-s was absent, indicating a defect in the processing of XBP1-u to XBP1-s by IRE1α RNase activity despite the phosphorylation of IRE1α (Fig. 2E). To determine whether a progressive decline in XBP1-s in hyperglycemia also affects the regulation of UPR target gene expression, we studied the effect of glucose on glucose regulated protein 78 (GRP78) activation by treating HGECs with glucose at one of two concentrations, 5.5 mM and 25 mM. The treatment of HGECs with glucose at a high concentration (25 mM) for 48 h decreased the level of GRP78 compared to that upon treatment with 5.5 mM glucose ( Fig. 2C-2E), indicating the impaired resolution of ER stress under hyperglycemia. In summary, these results demonstrated that the IRE1α signaling pathway is inhibited in HGECs under high-glucose conditions, indicated that failure of the ER stress response and UPR activation completely overwhelm cytoprotective mechanisms in response to hyperglycemia, resulting in prolonged or severe ER stress.
Subsequently, to mimic the in ammatory environment, HGECs were stimulated with Porphyromonas gingivalis lipopolysaccharide (LPS). Intriguingly, the results of RT-qPCR and western blotting also indicated that IRE1α, XBP1-s and GRP78 levels remained high, whereas XBP1-u levels remained low after HGECs were stimulated with 1 µg/mL P. gingivalis LPS (Fig. 2C-2E). Through its RNase activity, activated IRE1α directly splices XBP1 mRNA to produce XBP1-s, which then upregulates the expression of genes downstream of IRE1α such as GRP78, thereby reducing misfolded or unfolded proteins in the ER. Our results suggest that P. gingivalis LPS-stimulated HGECs may be responsible for the initiation of adaptive ER stress to restore homeostasis, but in the presence of hyperglycemia, cellular stress exceeds the capacity of the UPR, and the UPR machinery is damaged. This notion also re ects the poorer control of periodontal in ammation by conventional periodontal treatments in patients with periodontitis with diabetes mellitus compared with periodontitis patients.
Next, to analyze the expression of IRE1α and GRP78 in the human periodontium, we performed immunohistochemical staining of the human gingival epithelium for IRE1α and GRP78 (n = 14). IRE1α and GRP78 immunoreactivity in the human gingival epithelium was signi cantly downregulated in the DP group and signi cantly upregulated in the P group compared with the healthy group (Fig. 2F-2I). Pearson's correlation analysis was conducted and revealed a positive relationship between IRE1α and GRP78 expression within tissues of the human gingival epithelium (Fig. 2J-2L). These results demonstrate impairment of the IRE1α signaling pathway and prolonged ER stress in the gingival epithelium in periodontitis with diabetes mellitus and suggest that prolonged and exaggerated ER stress is related to the presence of hyperglycemia.
Decreased SERPINH1 expression in the gingival epithelium in periodontitis with diabetes mellitus To better understand the mechanism by which severe in ammation in the periodontium occurs under diabetic conditions, we subjected the gingival epithelium of mice with periodontitis with diabetes mellitus and mice with periodontitis to RNA sequencing (RNA-seq) data analysis. We analyzed ER-related genes and identi ed differentially expressed genes based on biological process. Among the ER-related genes, SERPINH1, ATP2A1, CASQ1, CYP2E1, STBD1, RBL21 and TMEM38A showed signi cantly different expression between the DP group and P group (Fig. 3B, 3C). RT-qPCR analysis further con rmed that SERPINH1 mRNA expression levels were signi cantly decreased in HGECs under high-glucose (25 mM) conditions compared with those without high-glucose treatment (Fig. 3A). The decrease in the SERPINH1 level in HGECs under high-glucose conditions was further con rmed by western blot analysis (Fig. 3D,   3H). We further evaluated the expression level of SERPINH1 in the human gingival epithelium (n = 14) using immunohistochemistry. A sharp contrast in SERPINH1 tissue staining was observed between the DP and P groups. According to the results of immunohistochemical staining for SERPINH1, the expression of SERPINH1 in the human gingival epithelium was lower in the DP group and higher in the P group than in the healthy group (Fig. 3I, 3J). Taken together, these data suggest that decreased SERPINH1 protein expression is associated with diabetes. Furthermore, Pearson's correlation analysis was conducted, and a positive relationship between SERPINH1 and IRE1α expression within human gingival epithelial tissues was found (Fig. 3E-3G), which indicates that signi cant downregulation of SERPINH1 in a hyperglycemic microenvironment may lead to prolonged or severe ER stress.

Silencing of SERPINH1 signi cantly prolonged ER stress responses and initiated an abnormal in ammatory response in HGECs
To ascertain the effect of SERPINH1 inhibition of ER stress responses and the in ammatory response, we transfected SERPINH1 siRNA into HGECs. RT-qPCR and western blot analysis con rmed the good transfection e ciency of SERPINH1 siRNA (Fig. 4A-4C). Next, we evaluated the effect of SERPINH1 interference on IRE1α, GRP78, XBP1-s, and XBP1-u expression in the HGECs. The expression levels of IRE1α, GRP78 and XBP1-s were signi cantly decreased, and those of XBP1-u and p-IRE1α were increased by SERPINH1 siRNA transfection in the HGECs, suggesting that UPR failure resulted in prolonged or severe ER stress (Fig. 4D, 4G, and 4I).
In addition, we explored the effects of SERPINH1 siRNA transfection on in ammation in HGECs. Activation of the NF-kappa B pathway plays an important role in the progression of in ammation. ER stress aggravates the in ammatory response through the IRE1α-associated NF-κB signaling pathway [27]. Therefore, we detected p-p65, NLRP3 and IL-1β expression in HGECs using RT-qPCR and western blotting. The expression levels of p-p65, NLRP3 and IL-1β were signi cantly increased by SERPINH1 siRNA transfection (Fig. 4D-4I). Taken together, these results indicate that silencing SERPINH1 signi cantly prolonged ER stress responses and initiated or aggravated an abnormal in ammatory response in HGECs.

SERPINH1 overexpression regulated the ER stress response and alleviated in ammation in HGECs cultured under high-glucose conditions
The results of in vitro SERPINH1 silencing and the decreased expression levels of SERPINH1 in the human gingival epithelium under periodontitis with diabetes mellitus prompted us to assess the therapeutic potential of SERPINH1 overexpression in an in vitro model. We hypothesized that SERPINH1 overexpression would ameliorate prolonged ER stress and the in ammatory response. To further investigate gingival epithelial barrier function and in ammatory responses, HBLV-h-SERPINH1transfected HGECs were successfully constructed, and the mRNA and protein expression levels of SERPINH1 were signi cantly upregulated in the HBLV-h-SERPINH1 HGECs (Fig. 5A-5D). Next, we detected the expression levels of related genes and proteins, including SERPINH1, IRE1α, p-IRE1α, GRP78, XBP1-s, XBP1-u, p-p65, NLRP3, and IL-1β, by RT-qPCR and western blot analysis to evaluate the ability of SERPINH1 to alleviate the prolonged ER stress response and in ammation. In the high-glucose group compared to the group without high-glucose treatment, the expression levels of XBP1-u, p-IRE1α, p-p65, NLRP3, and IL-1β were higher, and the expression levels of SERPINH1, IRE1α, GRP78, XBP1-s were lower.
However, following treatment with HBLV-h-SERPINH1 under high glucose, the expression levels of XBP1-u, p-IRE1α, p-p65, NLRP3, and IL-1β were signi cantly decreased, and the expression levels of SERPINH1, IRE1α, GRP78, and XBP1-s were signi cantly increased compared with those in the high-glucose group. (Fig. 5E-5K). Our results proved that SERPINH1 overexpression mitigated the prolonged ER stress response and in ammation in HGECs under high-glucose conditions.
Collectively, in vitro SERPINH1-knockdown and SERPINH1-overexpression experiments indicated the broad applicability of SERPINH1 overexpression as it diminished the in ammatory response, which may play a role in the development of periodontitis with diabetes mellitus, in HGECs under high-glucose conditions and that the IRE1α signaling pathway may be involved in regulating in ammation in HGECs under high-glucose conditions.

Discussion
Homeostasis of the innate host immune defense function of gingival epithelial tissue is the rst barrier of the periodontal tissue against invading pathogens and plays an important role in the occurrence and development of periodontitis [28]. In this study, we provide evidence that a hyperglycemic microenvironment aggravates the in ammatory response in the gingival epithelium. It now appears that in periodontitis with diabetes mellitus, periodontal tissues undergo long-term, low-level in ammation induced by a hyperglycemic state, and innate immunity of the gingival epithelial tissue is abnormal, which is the main way in which diabetes exacerbates periodontal tissue in ammation [29]. Thus, exploring the molecular mechanism by which diabetes mellitus aggravates in ammation of the human gingival epithelium and identifying potential therapeutic targets are important research areas.
The signaling events often associated with innate immunity and host de ance are triggered under ER stress. UPR signaling has been proven to modulate in ammation in multiple ways and contributes substantially to disease progression [30]. The initial intent of the UPR is to reestablish homeostasis and normal ER function by activating transcriptional programs to induce the expression of genes that can enhance the protein-folding capacity of the ER and ER-assisted degradation (ERAD) genes to enhance the clearance of misfolded or unfolded proteins. These adaptive features of the UPR probably play essential roles in reestablishing cellular homeostasis and in sustaining the normal cellular physiology. Transition of the adaptive UPR to the maladaptive UPR is induced when proteostatic imbalance occurs in the ER, which in turn contributes to the development and progression of disease [7].
In this study, we have de ned a mechanism through which periodontitis with diabetes mellitus impedes the most conserved branch of the UPR and impairs ER function. The attenuation of IRE1α signaling during prolonged or severe ER stress is a key step in cell fate determination after induction of the UPR.
Although the role of IRE1α in disease has been extensively studied, the molecular mechanisms that mediate its regulation are poorly understood. ER stress has been linked to multiple pathological conditions, ranging from neurodegeneration to metabolic disorders and in ammation [31]. Studies have demonstrated that under prolonged ER stress, IRE1α is turned off [32], which promotes the occurrence and development of disease. Small molecules targeting the activity of IRE1α have shown protective effects in various disease models [33,34]. IRE1α is the only protein reported to have kinase activity coupled to RNase activity [35]. Notably, our ndings demonstrated that both activities of IRE1α and XBP1s mRNA are defective under hyperglycemic conditions despite the phosphorylation and sustained expression of IRE1α. This observation might be explained by uncoupling of the kinase and RNase activities of the IRE1α protein in hyperglycemia, which thus inhibits IRE1α RNase activity and XBP1 mRNA splicing. This is consistent with studies in the literature reporting that in metabolic disease-related chronic in ammation, the kinase and RNase activities of the IRE1α protein are uncoupled, which suppresses IRE1α RNase activity and XBP1 mRNA splicing [36]. In accordance with the literature, IRE1α phosphorylation levels are modulated by phosphatases and kinases [37,38], whereas IRE1α RNase activity is modulated by IRE1α oligomerization status [39], and IRE1α oligomerization is assisted by nonmuscle myosin II (NMII) and the actin cytoskeleton [40]. The selectivity of the RNase activity of IRE1α for the XBP1 mRNA may depend on the oligomerization status of IRE1α [41,42]. Therefore, regulation of the dual functionalities of IRE1α is complex and multitiered. As con rmed in our study, IRE1α signaling triggered by hyperglycemic conditions differs from that activated under the adaptive UPR. Additionally, IRE1α has been implicated in multiple signaling pathways that lead to immune activation and in ammation. Evidence of the involvement of IRE1α in modulating in ammatory cytokine production is compelling. Kinases responsible for activation of the transcription factor NF-κB may be induced by the UPR-induced alarm signal [43]. Chen J. and colleagues reported that ER stress impairs human umbilical vein endothelial cells by increasing in ammation through the IRE1-associated NF-κB signaling pathway [27]. Our results also indicated the role of IRE1α signaling in regulating innate immunity and in ammation in the gingival epithelium. Thus, IRE1α signaling represents a regulated process involving multiple factors that may play an important role in the development of periodontitis with diabetes mellitus.
SERPINH1 is an ER stress target gene [44,45], and complete SERPINH1 de ciency has been reported to result in unresolved ER stress and rapid apoptosis in speci c cellular models [20,21]. In this study, we demonstrated that low SERPINH1 expression may be responsible for the aggravation of periodontitis under diabetic conditions, con rming reduced IRE1α signaling upon SERPINH1 deletion, which results in prolonged ER stress and aggravated in ammatory response in HGECs. Previous studies have demonstrated that SERPINH1 is an ER-localized chaperone that enhances IRE1α activation. SERPINH1 contains an RDEL motif at its C-terminus and is localized in the lumen of the ER [46,47]. In fact, Sepulveda and colleagues recently showed in mouse embryonic broblasts that SERPINH1 physically interacts with the luminal domain of IRE1α, which promotes the oligomerization of IRE1α and plays a role in regulating the UPR [22]. Recent studies by Yoneda A. et al. con rmed that SERPINH1 acts as a regulator of IRE1α activity through formation of a complex consisting of SERPINH1 with IRE1α [48]. In summary, SERPINH1 may bind the luminal region of IRE1α in the ER and regulate IRE1α activity in HGECs. Intriguingly, this study showed that silencing SERPINH1 in HGECs resulted in the inhibition of IRE1α but not its kinase activity and prevented the generation of su cient levels of XBP1-s while enhancing the NF-κB signaling pathway. In hyperglycemia, SERPINH1 overexpression enhanced IRE1α activation, potentiated its downstream signaling, and inhibited the kinase domain of IRE1α, resulting in the mitigation of ER stress and in ammation. These results indicate that regulation of the dual functionalities of IRE1α by SERPINH1 is complex and multitiered and that SERPINH1 might activate IRE1α activity to maintain gingival epithelium homeostasis and reduce proin ammatory cytokine expression by preventing prolonged ER stress induced by high-glucose environments. Consequently, our observations provide a molecular mechanism underlying defective IRE1α signaling in the gingival epithelium under hyperglycemic conditions, which may be a critical determinant of ER function in the development of periodontitis with diabetes mellitus. In conclusion, the current study shows that SERPINH1 overexpression activated the IRE1α/XBP1 axis, increased the expression of GRP78, and mitigated prolonged ER stress, thus alleviating the in ammatory response in HGECs under high-glucose conditions (graphic summary, Fig. 6).

Conclusion
This study has demonstrated that regulation of the UPR transducer IRE1α by SERPINH1 alleviates periodontitis with diabetes mellitus by mitigating prolonged ER stress. Therefore, SERPINH1 overexpression in HGECs might become a powerful therapeutic modality against periodontitis with diabetes mellitus. This nding provides insight for the further study of periodontitis with diabetes mellitus.  E. NLRP3 and IL-1β gene expression was detected by qRT-PCR. Statistical analysis of the qRT-PCR data was used to determine the relative mRNA levels of NLRP3 and IL-1β; *p < 0.05. Error bars represent the SEM. F. Statistical analysis of western blot data was carried out to determine the relative intensity of p-p65; *p < 0.05. Error bars represent the SEM. G. Immunohistochemical staining of the human gingival epithelium for NF-κB p65 was carried out in the healthy group, periodontitis group and periodontitis with diabetes mellitus group (n = 14).   Silencing of SERPINH1 signi cantly prolonged ER stress responses and initiated an abnormal in ammatory response in HGECs. A. SERPINH1 expression in HGECs transfected with SERPINH1 siRNA was determined by western blot analysis. B. Statistical analysis of the western blot data was carried out to determine the relative intensity of SERPINH1; *p < 0.05. Error bars represent the SEM. C. SERPINH1 gene expression was detected by qRT-PCR. Statistical analysis of the qRT-PCR data was carried out to determine the relative mRNA level of SERPINH1; *p < 0.05. Error bars represent the SEM. D. The expression of IRE1α, p-IRE1α, IL-1β and GRP78 in HGECs transfected with SERPINH1 siRNA was determined by western blot analysis. E. p-p65 expression in HGECs transfected with SERPINH1 siRNA was determined by western blot analysis. F. NLRP3 expression in HGECs transfected with SERPINH1 siRNA was determined by western blot analysis. G. Statistical analysis of the western blot data was carried out to determine the relative intensity of IRE1α, p-IRE1α, GRP78, NLRP3 and IL-1β; *p < 0.05. Error bars represent the SEM. H. Statistical analysis of the western blot data was carried out to determine the relative intensity of p-p65; *p < 0.05. Error bars represent the SEM. I. IRE1α, XBP1-s, XBP1-u, GRP78 NLRP3 and IL-1β gene expression was detected by qRT-PCR. Statistical analysis of the qRT-PCR data was carried out to determine the relative mRNA levels of IRE1α, XBP1-s, XBP1-u, GRP78, NLRP3 and IL-1β; *p < 0.05. Error bars represent the SEM. SERPINH1 overexpression regulated the ER stress response and alleviated in ammation in HGECs cultured under high-glucose conditions. A. Representative images of stable SERPINH1-overexpressing human gingival epithelial cell lines at 100× magni cation is shown. B. SERPINH1 expression in HGECs overexpressing SERPINH1 was determined by western blot analysis. C. Statistical analysis of the western blot data was carried out to determine the relative intensity of SERPINH1; *p < 0.05. Error bars represent the SEM. D. SERPINH1 gene expression was detected by qRT-PCR. Statistical analysis of the qRT-PCR data was carried out to determine the relative mRNA level of SERPINH1; *p < 0.05. Error bars represent the SEM. E. SERPINH1 expression in HGECs overexpressing SERPINH1 treated with a high concentration of glucose was detected by qRT-PCR. Statistical analysis of the qRT-PCR data was carried out to determine the relative mRNA level of SERPINH1; *p < 0.05. Error bars represent the SEM. F. The expression of IRE1α, p-IRE1α, GRP78 and IL-1β in HGECs overexpressing SERPINH1 treated with a high concentration of glucose was determined by western blot analysis. G. The expression of NLRP3 and SERPINH1 in HGECs overexpressing SERPINH1 treated with a high concentration of glucose was determined by western blot analysis. H. p-p65 expression in HGECs overexpressing SERPINH1 treated with a high concentration of glucose was determined by western blot analysis. I. Statistical analysis of the western blot data was carried out to determine the relative intensity of p-p65; *p < 0.05. Error bars represent the SEM. J. Statistical analysis of the western blot data was carried out to determine the relative intensity of SERPINH1, IRE1α, p-IRE1α, GRP78, NLRP3 and IL-1β in HGECs overexpressing SERPINH1 treated with a high concentration of glucose; *p < 0.05. Error bars represent the SEM. K. IRE1α, XBP1-s, XBP1-u, GRP78

Abbreviations
NLRP3 and IL-1β gene expression in HGECs overexpressing SERPINH1 treated with a high concentration of glucose was detected by qRT-PCR. Statistical analysis of the qRT-PCR data was carried out to determine the relative mRNA levels of IRE1α, XBP1-s, XBP1-u, GRP78 NLRP3 and IL-1β; *p < 0.05. Error bars represent the SEM. Schematic model depicting the mechanism by which SERPINH1 inhibits the in ammatory response under the high-glucose microenvironment through the IRE1α signaling pathway.