N,N,N′,N′-Tetrakis(2-pyridylmethyl)ethylenediamine induces endothelium-dependent hyperpolarization-mediated vasorelaxation via store-operated calcium entry mechanism in healthy and intestinal inflammatory mice

Abstract

Although the store-operated Ca²⁺ entry (SOCE) plays a critical role in maintaining Ca²⁺ homeostasis in vascular endothelial cells (VECs), its role in regulating endothelium-dependent hyperpolarization (EDH)-mediated vasorelaxation is largely unknown. Inflammatory bowel disease (IBD) and irritable bowel syndrome (IBS) are the most common gastrointestinal disorders with no effective cures. The present study applied N,N,N′,N′-tetrakis (2-pyridylmethyl)ethylenediamine (TPEN) as a Ca²⁺ chelator in the endoplasmic reticulum (ER) to study the SOCE/EDH-mediated vasorelaxation of micro-arteries and their involvements in the pathogenesis of IBD and IBS. Human submucosal arterioles and the second-order branch of 6–8 weeks male C57BL/6 mouse mesenteric arterioles were used, and TPEN-induced vasorelaxation was recorded by Danish DMT520A microvascular measuring system. The mice were fed water with 2.5 % dextran sulfate sodium for 7 days to induce mouse model of ulcerative colitis, and water avoidance stress was used to induce mouse model of IBS. The statistical significance of differences in the means of experimental groups was determined using a t-test for two groups or one-way ANOVA for more than two groups. TPEN concentration-dependently induced vasorelaxation of human colonic submucosal arterioles and the second-order branch of murine mesenteric arteries in endothelium-dependent manner. TPEN-induced vasorelaxation was much greater in the arteries pre-constricted by noradrenaline than those by high K⁺. While TPEN-induced vasorelaxation was unaffected by inhibitors of NO and PGI₂, it was significantly inhibited by the selective inhibitors of IK_Ca and SK_Ca channels but was potentiated by their activator. Moreover, TPEN-induced vasorelaxation was attenuated by selective inhibitors of NCX, NKA, SOCE, STIM translocation and Orai transportation. Finally, TPEN-induced vasorelaxation via SOCE/EDH was impaired in colitic mice but remained intact in IBS mice. Interestingly, TPEN could rescue vagus neurotransmitter ACh-induced vasorelaxation that was impaired in IBS mice. Therefore, since TPEN-induced SOCE/EDH-mediated vasorelaxation of mesenteric arteries is well-preserved to be able to rescue ACh-induced vasorelaxation impaired in IBS, TPEN has therapeutic potentials for IBS.

Abbreviations

TPEN

N,N,N′,N′-Tetrakis(2-pyridylmethyl)ethylenediamine

EDH

Endothelium-dependent hyperpolarization

SOCE

Store-operated calcium entry

cAMP

Cyclic adenosine monophosphate

H₂S

Hydrogen sulfide

NE

Norepinephrine

VSMC

Vascular smooth muscle cells

1. Introduction

The mesenteric circulation is vital in modulating hemoperfusion to maintain normal gastrointestinal (GI) functions, such as epithelial ion transports, motility, nutrient digestion, absorption and transportation [1]. Dysregulation in mesenteric circulation is involved in several GI diseases, such as epithelial injury, ulcers and inflammation [2]. The mesenteric arterial activity is fine controlled by multiple components, including vascular endothelium, smooth muscle and perivascular nerves [3]. The vascular endothelium could generate three acknowledged vasorelaxant factors, including nitric oxide (NO), prostacyclin (PGI₂) and endothelium-dependent hyperpolarization (EDH), to modulate mesenteric arterial activity to maintain normal mesenteric circulation [4,5].

While NO stimulates soluble guanylyl cyclase in vascular smooth muscle cells (VSMCs) to produce cyclic guanosine monophosphate (cGMP), leading to vasorelaxation of large conduit arteries [6], PGI₂ induces vasorelaxation via adenylyl cyclase/cyclic adenosine monophosphate (cAMP) signaling pathway [7]. Although EDH nature is not clear so far, Ca²⁺ influx into vascular endothelial cells (VECs) of small resistant arteries is indispensable for EDH mechanism. An elevation in the cytosolic concentration of Ca²⁺ ([Ca²⁺]cyt) triggers the activation of small conductance Ca²⁺-activated K⁺ channels (SK_Ca) as well as intermediate conductance Ca²⁺-activated K⁺ channels (IK_Ca) within vascular endothelial cells (VECs). This activation subsequently leads to the hyperpolarization of vascular smooth muscle cells, ultimately inducing vasorelaxation [8]. Hydrogen sulfide (H₂S), cytochrome P450-derived arachidonic acid metabolites and gap conjunctions may also contribute to EDH-mediated vasorelaxation [9,10]. Moreover, it is widely recognized that NO and PGI₂ occupy pivotal positions in modulating vasorelaxation in large conduit vessels, whereas EDH holds a crucial role in regulating the vasorelaxation of small resistant arteries, such as mesenteric arterioles [11]. Crucially, in scenarios where NO-mediated vasorelaxation is compromised in diseases like diabetes and hypertension, EDH serves as a compensatory mechanism, mitigating the reduced vasorelaxation effect of NO to a certain extent [12], implying its therapeutic potential for these diseases. However, little is currently known about the involvement of EDH-mediated vasorelaxation in intestinal inflammation.

Cell Ca²⁺ has an unreplaced role in modulating various human physiological processes [13]. The [Ca²⁺]_cyt in most cells is relative low in resting state, but it rapidly increases after cell stimulation, among which the store-operated Ca²⁺ entry (SOCE) mechanism plays a vital role [14]. The SOCE is comprised of STIM1 protein on the endoplasmic reticulum (ER) membrane and Orai1 protein family on plasma membrane, but recent studies found that TRP channels could also contribute to the SOCE [15]. The SOCE participates in multiple physiological activities, such as modulating vascular activity, neuronal and immune function, hormone secretion, and muscle development et al. [16,17] Although the SOCE participation in neuronal and immune function is extensively studied [18,19], little is known about its involvement in endothelium-dependent vasorelaxation except our previous report that cyclopiazonic acid (CPA), a selective inhibitor of the sarco (endo) reticulum calcium-ATPases (SERCA) to activate endothelial SOCE, could induce EDH-mediated vasorelaxation of healthy mouse mesentery [17]. However, so far it is totally unknown if endothelial SOCE/EDH is involved in intestinal inflammation.

Inflammatory bowel disease (IBD) refers to a chronic and recurring inflammatory condition affecting the gastrointestinal tract, encompassing both Crohn's disease (CD) and ulcerative colitis (UC). Present studies on the pathogenesis of IBD have been mainly focusing on intestinal mucosal barrier injury, immune and neural dysfunction, and intestinal microbiome [[20], [21], [22]]. However, the involvement of mesenteric circulation has been largely neglected although it was found that the vasorelaxation of mesenteric arteries was impaired in IBD patients, leading to hypoperfusion to the inflamed area of the intestinal mucosa to prompt IBD development [23,24]. Although the SOCE on immune cells and intestinal epithelial cells (IECs) participates in the pathogenesis of IBD [25,26], little is known about participation of endothelial SOCE in IBD. Furthermore, irritable bowel syndrome (IBS) is the most common functional GI disorder with no effective cures since its etiology remains unclear and current research mostly focuses on immune and gut microbiota [[27], [28], [29]]. Although recent studies found a low-grade intestinal inflammation in the pathogenesis of IBS [30], the involvements of mesenteric circulation and endothelial SOCE/EDH have not been explored so far. Furthermore, studies have shown many common symptoms in IBS and IBD, which share overlapping mechanisms of disease [31], and a recent study confirms that IBD is causally related to IBS [32].

The membrane-permeant heavy metal chelator N,N,N,N-tetrakis(2-pyridylmethyl)ethylenediamine (TPEN) exhibits a low affinity for Ca²⁺ (Kd ∼100 μM), rendering it uniquely suited for the rapid and reversible sequestration of Ca²⁺ within endoplasmic reticulum (ER) stores. Owing to its diminished binding affinity for Ca²⁺, TPEN does not significantly alter cytosolic Ca²⁺ concentrations ([Ca²⁺]cyt) [33,34]. These characteristics described above have led to the experimental use of TPEN as modulator of the ER Ca²⁺ concentrations for many years. However, it is largely unknown about the vascular pharmacology of TPEN and the mechanisms of action. Therefore, in the present study, we first investigated the actions of TPEN on resistance vessels and the underlying mechanisms, and then we applied TPEN to activate endothelial SOCE to further study its precise role in mesenteric vasorelaxation in health. Finally, we explored if the SOCE/EDH-mediated mesenteric vasorelaxation is involved in the pathogenesis of IBD and IBS due to their substantial overlaps [35,36].

2. Methods and materials

2.1. Animal study

All experiments were conducted on male C57BL/6 mice aged 6–8 weeks old and weighed 19–23 g. All mice were kept in cages in a temperature-controlled room with a 12 h light/dark cycle. The mice were provided with unlimited water and normal chow. Before the experiments, the mice were deprived of food and water for over 1 h. All experiments, including group identity and data collection, were conducted in a blinded manner. The procedures and policies adhered to in conducting experiments on animals were duly approved by the Medical College of Qingdao University (QDU-AEC-2023073).

2.2. Arteriole tissue preparations

The human colon samples from patients were acquired with their full informed consent for the study. The study of clinical samples was conducted under the approval of the Clinical Research Ethics Committee of the Army Medical University (AMU, Chongqing, China, AMUWEC2020368). The colonic tissue was collected from surgical patients and promptly placed in ice-cold Krebs–Henseleit solution. This tissue was then transported to the laboratory within 10 min for experimentation. Under a surgical microscope, the submucosal arterioles were isolated by carefully removing the surrounding fat and connective tissues.

The mice were euthanized via cervical dislocation, and the mesenteric loop was carefully excised. The excised mesenteric loop was immediately immersed in Krebs–Henseleit solution. Under surgical microscope observation, the second-order branch of mesenteric arteries (with an internal diameter of 150–250 μm and a length of approximately 2 mm) was isolated by removing adjacent fat and connective tissues. The Krebs–Henseleit solution was composed of (in millimolar concentrations): 118 NaCl, 11.1 D-glucose, 4.7KCl, 1.6 CaCl₂, 1.2 KH₂PO₄, 1.18 MgSO₄, and 25 NaHCO₃, maintaining a pH of approximately 7.4.

2.3. Vasorelaxation function tests

Using two tungsten wires, the isolated artery segment (approximately 2 mm in length) was securely mounted in a chamber bath containing 5 mL of Krebs–Henseleit solution. This chamber bath was continuously supplied with a gas mixture of 5 % CO₂ and 95 % O₂, maintaining a temperature of 37 °C to ensure optimal conditions for the experimental procedures. The fixed artery was stabilized at zero (under no external tension) for 20 min to adapt to the chamber situation and then continued normalization to optimal initial length. To establish a passive tension, the artery was stretched to achieve an internal circumference that corresponded to 90 % of its circumference under a transmural pressure of 100 mmHg. The vasorelaxation function was examined and recorded by Mulvany-style wire myograph (Model 520A, DMT, Aarhus, Denmark) and Powerlab analytical system (AD Instruments, Colorado Springs, CO, USA). At different concentrations, it took approximately 2∼3 min for TPEN/ACh-induced vasorelaxation immediately until the vascular tension is stable.

Cumulative concentration-response curve (CRC) to TPEN (10–75 μmol/L) and acetylcholine (ACh, 0.01–1000 μM) were performed in norepinephrine (NE, 5 μmol/L) -pre-constricted artery. In our previous study, NE-induced vasoconstriction was sustained for more than 40 min, and there were no statistical differences in the contractile tension. To delve into the underlying mechanisms, the stabilized arteries were subjected to treatment with various inhibitors and activators for a duration of 20–30 min. Isometric tension tests revealed that the vasorelaxation function of the artery remained unaffected by the respective solvent vehicles (less than 0.4 % DMSO or H₂O). In the control tissues, the presence of all inhibitors and activators during the construction of the CRC (concentration-response curve) did not significantly alter the pre-constricted tone, as indicated in Table 1.

To test the effects of TPEN on ACh-induced vasorelaxation in IBS, the fixed arteries were treated with 10 μM TPEN for approximately 20 min and continue the following experiments.

To eliminate the vascular endothelium, the luminal surface of the mesenteric arteries was gently rubbed several times using human hair. Successful endothelial denudation was confirmed by the absence (≤10 %) of a vasorelaxation response to 100 μM of carbachol (CCh). The subsequent experiments were conducted only after this successful removal of the vascular endothelium.

2.4. HUVEC culture and patch clamp study

HUVEC (Human umbilical vein endothelial cells, Sciencell, USA, Catalog #8000) were cultivated in endothelial cell medium (ECM, Sciencell, USA, Cat. #1001), supplemented with 10 % fetal bovine serum (Gibco, USA) and 1 % penicillin-streptomycin (Beyotime Biotechnology, China). These cells were maintained at 37 °C in an environment with 5 % CO₂ and saturated humidity. Cell passages between 10 and 20 were selected for experimentation. Prior to the experiments, HUVEC were plated onto glass coverslips for 24 h.

To record macroscopic currents in HUVEC, a patch clamp technique was employed, utilizing a bath solution containing various compounds. The patch clamp system was equipped with a HEKA EPC10 amplifier, operated by the PatchMaster software. During the recordings, the membrane potential was maintained at 0 mV, and the currents were elicited through a specific protocol: a 400-ms step at −100 mV, followed by a 500-ms voltage ramp ranging from −100 to +100 mV, and then a 400-ms step at 100 mV, all executed in 2-s intervals.

The extracellular buffer comprised (in millimolar concentrations): 140 NaCl, 5 KCl, 2 CaCl₂, 2 MgCl₂, and 10 HEPES, adjusted to a pH of 7.3. The pipette solution contained 140 mM CsCl, 5 mM EGTA, 3 mM Mg-ATP, and 10 mM HEPES, also with a pH of 7.3. The osmolality of all solutions was approximately 300 mOsmol/kg.

2.5. DSS-induced colitis of mouse model

Male C57BL/6 mice were randomly assigned to either a control group or an experimental group. The control group received plain drinking water, while the experimental group was administered water containing 2.5 % DSS for a duration of 7 days. During the experimental period, all mice were monitored daily for changes in body weight, rectal bleeding, and water consumption [8]. Upon completion of the experimental protocol, the mice were euthanized via cervical dislocation to measure the length of their colons. The statistical indicators of DSS-induced colitis mice referred to previous experiments [37].

2.6. Water avoidance stress (WAS)-induced IBS model mice

Given that stress is recognized for its disruptive impact on the gut microbiome and its ability to exacerbate symptoms related to irritable bowel syndrome (IBS), the current study employed a mouse model of IBS induced by water avoidance stress (WAS). This approach was adopted in line with previous experimental designs [38,39]. The specialized testing apparatus comprised a transparent plastic tank measuring 45 cm in length, 32 cm in width, and 26 cm in height. At the bottom center, a cuboid acrylic block with dimensions of 3 cm length, 3 cm width, and 9 cm height was securely fastened. The experimental animals were divided into two groups: sham and water avoidance stress (WAS), with at least five mice in each group. For the IBS group, the mice were placed on an identical platform inside the tank, which contained fresh room temperature water (25 °C) reaching up to 1 cm below the acrylic bar. This procedure was carried out for 1 h daily. In contrast, the mice in the sham group were also placed on an identical platform but without any water in the tank, for 1 h per day. At the end of each 1-h session for both the WAS and sham groups, the number of fecal pellets in the container was counted. All mice were monitored by measuring body weight every day.

2.7. Statistics

All results are presented as mean values ± standard error of the mean (SEM), where ‘n' denotes the number of animals in each experimental group. No data points were omitted from the analysis. The cumulative concentration-response curve (CRC), maximum relaxation (Rmax), area under the curve (AUC), and the concentration required for 50 % of the maximum effect (EC₅₀) were calculated using GraphPad Software 8.0 (San Diego, CA). Statistical significance between experimental group means was assessed using a t-test for two groups or a one-way ANOVA for more than two groups (GraphPad Prism 8.0, GraphPad Software, Inc., RRID:SCR_002798). It was determined that there was no significant variance inhomogeneity. A P-value less than 0.05 was considered statistically significant.

2.8. Materials

All salts were procured from Sangon Biotech in Shanghai, China. Acetylcholine, carbachol (CCH), Nω-nitro-L-arginine, and indomethacin were obtained from Sigma-Aldrich, located in St. Louis, USA. NE was purchased from GRANDPHARMA co. LTD in China. TPEN, TRAM-34, apamin, SKA-31, SN-6, ouabain, GSK-7975 A, SKF96365, ML-9, and Brefeldin A (BFA) were supplied by MedChemExpress (MCE), headquartered in New Jersey, USA. Dextran sulfate sodium (DSS) was provided by meilunbio. TPEN, TRAM-34, apamin, SKA-31, SN-6, ouabain, GSK-7975 A, SKF96365, ML-9, and BFA were dissolved in DMSO at a final concentration not exceeding 0.1 %, which did not introduce any significant alterations to vascular activities during the experiments.

3. Results

3.1. TPEN induced vasorelaxation of human colonic submucosal arterioles likely via EDH mechanism

Given their status as well-recognized resistance vessels, human colonic submucosal arterioles play a crucial role in preserving mucosal barrier function by regulating blood perfusion in healthy individuals [40]. Therefore, we first examined vasorelaxation effects of TPEN on human colonic submucosal arterioles. As shown in the left panel of Fig. 1A, the application of TPEN induced a concentration-dependent vasorelaxation effect on arterioles that had been pre-constricted with norepinephrine (NE), but it scarcely elicited vasorelaxation in arterioles pre-constricted with high potassium (K⁺). Notably, there were significant differences in the cumulative concentration-response curve (CRC), maximum relaxation (Rmax), area under the curve (AUC), and concentration for 50 % maximal effect (EC₅₀) between the arterioles pre-constricted with NE and those pre-constricted with high K⁺ (the right three panels of Fig. 1A). The loss of vasorelaxation in arterioles pre-constricted with high K⁺ following TPEN application indicates a possible involvement of K⁺ channels in the TPEN-induced vasorelaxation effect. These findings strengthen the conclusion that TPEN is capable of inducing vasorelaxation in human colonic submucosal arterioles.

Fig. 1.

TPEN induced EDH-mediated vasorelaxation of human colonic submucosal arterioles.

(A) Summarized data shows TPEN-induced CRC, R_max, AUC, and EC₅₀ of human colonic submucosal arterioles pre-constricted with NE (5 μM, n = 6) or KCl (80 mM, n = 6). (B) Summarized data shows TPEN-induced CRC, R_max, AUC, and EC₅₀ of human colonic submucosal arterioles pre-constricted with NE in the absence (control, Ctrl, n = 6) or the presence of either 100 μM L-NNA plus 10 μM INDO (n = 6) or L-NNA + INDO (&) + 3 μM apamin (Apa) + 30 μM TRAM-34 (TRAM, n = 6). Data were expressed as a percentage of NE- or KCl-induced vasoconstriction and shown as means ± SEM. **P < 0.01, ***P < 0.001, ****P < 0.0001 and ns: no significance.

Vascular endothelial cells are well-recognized for their ability to produce three key relaxation factors: endothelium-dependent hyperpolarization (EDH), nitric oxide (NO), and prostacyclin (PGI₂), which collectively contribute to the mediation of vasorelaxation [5]. Therefore, we further examine the contribution of three relaxation factors to TPEN-induced vasorelaxation of human colonic submucosal arterioles. As shown in Fig. 1B, upon application of Nω-nitro-L-arginine (L-NNA, 100 μM) and indomethacin (INDO, 10 μM) to inhibit the production of nitric oxide (NO) and prostacyclin (PGI₂), TPEN-induced vasorelaxation remained largely unaffected, indicating that NO and PGI₂ do not play significant roles in TPEN's vasorelaxant effect. However, in the presence of L-NNA and INDO, the vasorelaxation induced by TPEN was significantly reduced by the selective inhibitors of IK_Ca and SK_Ca, TRAM-34 (30 μM) and apamin (3 μM, Fig. 1B). While the precise nature of endothelium-dependent hyperpolarization (EDH) remains elusive, the crucial roles played by SK_Ca and IK_Ca are well-established [8]. Therefore, these findings suggest that TPEN likely mediates vasorelaxation in human colonic submucosal arterioles via an EDH-dependent mechanism. In essence, TPEN appears to modulate EDH-mediated vasorelaxation in human colonic submucosal arterioles under physiological conditions.

3.2. TPEN induced endothelium-dependent vasorelaxation of the second-order branches of mouse mesenteric arteries

Since we previously demonstrated that functional characterization of human colonic mesenteric arterioles was identical to that of the second-order branch of mouse mesenteric arteries as resistance vessels in terms of endothelium-dependent vasorelaxation [8,41]. To determine if both vessels exhibit similar responses to TPEN-induced endothelium-dependent vasorelaxation, we initially investigated the effect of TPEN on the second-order branch of mouse mesenteric arteries that had been pre-constricted with norepinephrine (NE). Our findings revealed that, analogous to its effect on human colonic mesenteric arterioles, TPEN induced a concentration-dependent vasorelaxation in the mouse mesenteric arteries as well (Fig. 2A).

Fig. 2.

TPEN induced endothelium-dependent vasorelaxation of the second-order branch of mice mesenteric arteries in a concentration-dependent manner.

(A) Summarized data shows TPEN-induced CRC, R_max, AUC and EC₅₀ of the second-order branch of mice mesenteric arteries with intact endothelium (EC+, n = 6) or denuded endothelium (EC-, n = 6). (B) Summarized data shows TPEN-induced CRC, R_max, AUC and EC₅₀ of the second-order branch of mice mesenteric arteries without (ctrl, n = 6), or with either 100 μM L-NNA (n = 6), or 10 μM INDO (n = 6), or 100 μM L-NNA plus 10 μM INDO (n = 6). Data were expressed as a percentage of NE-induced vasoconstriction and shown as means ± SEM. **P < 0.01, ****P < 0.0001 and ns: no significance.

Second, to test if TPEN relies on the endothelium, we compared its vasorelaxation of endothelium-intact vs endothelium-denuded arteries. As shown in the left panel of Fig. 2A, TPEN-induced vasorelaxation disappeared in the endothelium-denuded arteries. TPEN-induced R_max and AUC of the endothelium-intact arteries was obviously greater than the endothelium-denuded arteries, while TPEN-induced EC₅₀ was obviously smaller than the endothelium-denuded arteries (Fig. 2A). In summary, TPEN predominantly triggers endothelium-dependent vasorelaxation with similar functional characteristics in both human colonic mesenteric arterioles and the second-order branch of mouse mesenteric arteries. Given that TPEN modulates EDH-mediated vasorelaxation identically in both types of vessels, and considering the challenges associated with obtaining human colonic submucosal arterioles, we chose to utilize the second-order branch of mouse mesenteric arteries in subsequent experiments.

3.3. TPEN induced mesenteric arterial vasorelaxation via EDH mechanism

We further verified the contribution of three relaxation factors in TPEN-induced vasorelaxation. As depicted in the left panel of Fig. 2B, the vasorelaxation induced by TPEN remained unaffected by the individual or combined use of L-NNA and INDO, thus excluding the involvement of NO and PGI₂. Interestingly, when arteries were pre-constricted with high K⁺, the endothelium-dependent vasorelaxation triggered by TPEN was abolished, pointing towards the participation of EDH (Fig. 3A). To delve deeper into the role of SK_Ca and IK_Ca in TPEN-induced vasorelaxation, we administered apamin plus TRAM-34, which are selective inhibitors of SK_Ca and IK_Ca, as well as SKA-31, a selective activator of both SK_Ca and IK_Ca at a concentration of 0.3 μM. Notably, when L-NNA and INDO eliminated the production of NO and PGI₂, the residual vasorelaxation induced by TPEN was reduced by the selective inhibitors of SK_Ca and IK_Ca (apamin plus TRAM-34) but was enhanced by the selective activator of SK_Ca and IK_Ca (SKA-31, Fig. 3B). These findings further strengthen the crucial role of EDH in TPEN-induced vasorelaxation.

Fig. 3.

TPEN induced vasorelaxation of the second-order branch of mice mesenteric arteries via EDH mechanisms predominantly.

(A) Summarized data shows TPEN-induced CRC, R_max, AUC and EC₅₀ of the second-order branch of mice mesenteric arteries pre-constricted with NE (5 μM, n = 6) or KCl (80 mM, n = 6). (B) Summarized data shows TPEN-induced CRC, R_max, AUC and EC₅₀ of the second-order branch of mice mesenteric arteries with 100 μM L-NNA plus 10 μM INDO (n = 6), L-NNA + INDO (&) + 0.3 μM SKA-31(n = 6), or L-NNA + INDO (&) + 3 μM apamin (Apa) + 30 μM TRAM-34 (TRAM, n = 6). (C) Summarized data shows TPEN-induced CRC, R_max, AUC and EC₅₀ of the second-order branch of mice mesenteric arteries with L-NNA + INDO (n = 6), or L-NNA + INDO (&) + 10 μM SN-6 (n = 6), or L-NNA + INDO (&) + 100 μM ouabain (n = 6). Data were expressed as a percentage of NE-induced vasoconstriction and shown as means ± SEM. **P < 0.01, ***P < 0.001, ****P < 0.0001 and ns: no significance.

Given that the EDH mechanism is mediated by Na⁺-K⁺ ATPase (NKA) and Na⁺/Ca²⁺ exchanger (NCX) in vascular smooth muscle cells (VSMC) [17,42], we further explored their potential involvement in TPEN-induced vasorelaxation. As illustrated in Fig. 3C, the vasorelaxation induced by TPEN was significantly reduced by the application of 100 μM ouabain, a selective NKA inhibitor, or 10 μM SN-6, a selective NCX inhibitor. This further underscores the pivotal role of EDH in the vasorelaxation response triggered by TPEN.

3.4. TPEN induced EDH-mediated vasorelaxation likely via STIM1/Orai1-constructed SOCE in mice and humans

Since TPEN may activate the store-operated Ca²⁺ entry (SOCE) by chelating the ER/Ca²⁺ [15], we first examined the involvement of the SOCE in TPEN-induced vasorelaxation. As depicted in Fig. 4A, the vasorelaxation effect of TPEN on mouse mesenteric arteries was notably diminished by two selective SOCE blockers, SKF96365 (30 μM) and GSK-7975A (30 μM), when administered in the presence of L-NNA and INDO.

Fig. 4.

TPEN induced EDH-mediated vasorelaxation via STIM1/Orai1-dependent SOCE mechanisms.

(A) Summarized data shows TPEN-induced CRC, R_max, AUC and EC₅₀ of the second-order branch of mice mesenteric arteries in the presence of either 100 μM L-NNA plus 10 μM INDO (n = 6), L-NNA + INDO (&) + 30 μM GSK-7975A (n = 6) or L-NNA + INDO (&) + 30 μM SFK96365 (n = 6). (B) Summarized data shows TPEN-induced CRC, R_max, AUC and EC₅₀ of the second-order branch of mice mesenteric arteries with L-NNA + INDO (n = 6), L-NNA + INDO (&) + 10 μM ML-9 (n = 6), or L-NNA + INDO (&) + 5 μM brefeldin A (BFA, n = 6). (C) Summarized data shows TPEN-induced CRC, R_max, AUC and EC₅₀ of human colonic submucosal arterioles with L-NNA + INDO (n = 6), L-NNA + INDO (&) + 30 μM GSK-7975A (n = 6), or L-NNA + INDO (&) + 10 μM ML-9 (n = 6). Data were expressed as a percentage of NE-induced vasoconstriction and shown as means ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Second, it is widely recognized that SOCE is molecularly composed of the STIM1 protein located on the endoplasmic reticulum membrane and the Orai1 protein situated on the plasma membrane [42]. To validate this mechanism, we employed ML-9, an inhibitor of STIM1 translocation at 10 μM, and brefeldin A (BFA), an inhibitor of Orai1 transportation at 5 μM [43,44]. As seen in Fig. 4B, in the presence of L-NNA and INDO, TPEN-induced CRC, R_max, and AUC of mouse mesenteric arteries were notably suppressed by either ML-9 or BFA. This supports the involvement of STIM1/Orai1 in TPEN-induced vasorelaxation in mice. Lastly, TPEN-induced vasorelaxation in human colonic submucosal arterioles, in the presence of L-NNA and INDO, was also significantly reduced by GSK-7975A (Fig. 4C), indicating the participation of SOCE. Furthermore, TPEN-induced EDH-mediated vasorelaxation in human colonic submucosal arterioles was also inhibited by ML-9 (Fig. 4C), thus confirming the crucial role of STIM1/Orai1 in TPEN-induced vasorelaxation in humans. In conclusion, TPEN induces EDH-mediated vasorelaxation in both human and mouse mesenteric arteries through the STIM1/Orai1-mediated SOCE under physiological conditions.

3.5. Confirmation of TPEN-induced SOCE in HUVECs

To confirm whether TPEN triggers endothelial SOCE in human umbilical vein endothelial cells (HUVECs), we conducted a patch clamp study. Initially, we recorded the temporal evolution of membrane current changes at +100 mV. Fig. 5A depicts representative traces of membrane current alterations, demonstrating that 10 μM TPEN significantly induced membrane currents, which were notably diminished by 10 μM GSK-7975A. Fig. 5B summarizes the inhibitory effect of GSK-7975A on TPEN-induced currents in HUVECs. Fig. 5C shows representative I–V curves in the absence or presence of TPEN alone or TPEN combined with GSK-7975A. Notably, the reversal potentials of these I–V curves were close to 0 in all three conditions, indicating the electrophysiological characteristics of non-selective cation currents, including SOCE. These findings further corroborate the hypothesis that TPEN induces endothelial SOCE.

Fig. 5.

TPEN promoted transmembrane non-selective cation currents in HUVEC.

(A) Representative transmembrane non-selective cation currents in the absence or the presence of 10 μM TPEN alone or a combination of 10 μM TPEN plus 10 μM GSK-7975A. (B) Summary data of the currents measured at 100 mV under same experimental conditions in A. Data were shown as means ± SEM. ****P < 0.0001, n = 5 cells. (C) Representative current-voltage curves in response to voltage steps from −100 to +100 mV in the absence or the presence of 10 μM TPEN alone or a combination of 10 μM TPEN plus 10 μM GSK-7975A.

3.6. TPEN-induced SOCE/EDH-mediated vasorelaxation in ulcerative colitis

Given that the mesenteric vasorelaxation induced by the vagus neurotransmitter ACh is compromised in IBD patients, leading to decreased blood flow in inflamed regions and contributing to IBD progression [23,24], we further investigated whether TPEN-induced EDH-mediated vasorelaxation, mediated through SOCE, plays a role in the pathogenesis of ulcerative colitis. Therefore, we first created DSS-induced ulcerative colitis mouse model. As shown in Fig. 6A and B, the body weight and colon length of colitis mice those ingested water containing 2.5 % DSS were significantly reduced (P value of the seventh day of weight between two groups <0.0001). On day 4, the DSS-induced colitis mice started to have bloody stools, and their stool scores and disease activity index (DAI) were obviously higher compared to normal mice (P value of the seventh day of stool score and disease activity index between two groups <0.0001, Fig. 6C and D). Finally, Fig. 6E clearly demonstrates that the ACh-induced vasorelaxation was notably compromised in mice with DSS-induced colitis, in contrast to healthy control mice. This observation suggests that the functionality of the mesenteric arterial endothelium is adversely affected in ulcerative colitis.

Fig. 6.

DSS-induced colitis of mouse model successfully.

(A) Summary data shows the time courses of body weight in control mice (n = 6) or colitis mice treated with 2.5 % DSS (po) for 7 days (n = 6). (B) Representative photos of colon length in control and colitis mice. Moreover, summary data shows colon length in control mice (n = 6) and colitis mice (2.5%DSS, n = 6). (C) Summary data shows the time courses of stool scores in control mice (n = 6) or colitis mice (2.5%DSS, n = 6). (D) Summary data shows the time courses of disease activity index in control mice (n = 6) or colitis mice (2.5%DSS, n = 6). (E) Summary data showing the ACh-induced (10 nM - 1 mM) CRC, R_max and AUC of the second-order branch of control (Ctrl) mice and colitis mice (2.5%DSS, n = 4). Data were shown as means ± SEM. ****P < 0.0001.

We further analyzed the TPEN-induced vasorelaxation in both normal and colitis mice. As depicted in Fig. 7A, the TPEN-induced CRC of vasorelaxation was noticeably diminished in colitis mice. Specifically, the R_max and AUC values in colitis mice (R_max: 53.63 % ± 4.597 %, AUC: 2100 ± 261.8) were significantly lower compared to those in normal mice (Rmax: 96.13 % ± 4.743 %, P < 0.0001; AUC: 4078 ± 226.6, P < 0.001).

Fig. 7.

Impairments of the TPEN/SOCE/EDH-mediated vasorelaxation in colitis.

(A) Summary data shows the CRC, R_max, AUC and EC₅₀ of TPEN-induced vasorelaxation of the second-order branch of control (Ctrl, n = 6) mice and colitis mice treated with 2.5 % DSS (po) for 7 days (2.5%DSS, n = 6). (B) Summary data shows the CRC, R_max, AUC and EC₅₀ of TPEN-induced vasorelaxation of the second-order branch of control (Ctrl, n = 6) mice and colitis mice treated with 2.5 % DSS (po) for 7 days (2.5%DSS, n = 6) with 100 μM L-NNA plus 10 μM INDO (n = 6). Data were expressed as a percentage of NE (5 μM)-induced vasoconstriction and shown as means ± SEM. **P < 0.01, ***P < 0.001, ****P < 0.0001 and ns: no significance.

To further elucidate the role of EDH, we compared the TPEN-induced vasorelaxation in control and colitis mice under the influence of L-NNA and INDO. As illustrated in Fig. 7B, the R_max (47.97 % ± 9.404 %) and AUC (1908 ± 500.8) values in colitis mice were significantly reduced compared to normal mice (R_max: 97.78 % ± 5.134 %, P < 0.001; AUC: 4558 ± 184.9, P < 0.01). Moreover, the EC₅₀ in colitis mice (38.5 ± 1.663 μM) was significantly higher than that in normal mice (EC₅₀: 28.68 ± 1.361 μM). Collectively, these findings indicate that TPEN-induced SOCE/EDH-mediated vasorelaxation is substantially impaired in colitis, aligning with previous research [17].

3.7. TPEN-induced vasorelaxation of WAS-induced acute IBS mice

Some research revealed that a substantial overlaps between irritable bowel syndrome (IBS) and IBD, suggesting a pathogenic role of inflammation in IBS [36]. However, it is not yet known about the involvement of TPEN-induced vasorelaxation in IBS.

We first made WAS-induced acute IBS murine model for 1 day. As shown in Fig. 8A, fecal pellets of WAS-induced IBS mice were significantly increased compared to control mice, indicating successful setup of WAS-induced acute IBS murine model for 1 day [38]. Next, we assessed whether the vagus neurotransmitter ACh-induced mesenteric vasorelaxation was altered in mice with acute IBS. As depicted in Fig. 8B, ACh-induced vasorelaxation was notably impaired in IBS mice compared to healthy controls. To delve deeper into the contribution of three endothelium-derived relaxing factors in ACh-induced vasorelaxation in IBS, we utilized L-NNA and INDO to test this. As displayed in Fig. 8C, the application of L-NNA and INDO had minimal effect on ACh-induced vasorelaxation, suggesting that ACh/EDH-induced vasorelaxation was compromised in IBS. Furthermore, in the presence of L-NNA and INDO, ACh/EDH-induced vasorelaxation was further hindered by the selective STIM1 inhibitor BFA, strengthening the involvement of STIM1 (Fig. 8C).

Fig. 8.

The beneficial role of TPEN in acute WAS-induced IBS.

(A) Summary data of WAS mice's stool pellets versus sham mice in one day. (B) Summarized data shows acetylcholine (ACh, 10 nM - 1 mM)-induced CRC, R_max and AUC of the second-order branch from the sham mice (n = 6), WAS-induced IBS mice (n = 6) in the absence or presence of 10 μM TPEN. (C) Summarized data shows acetylcholine (ACh, 10 nM - 1 mM)-induced CRC, R_max and AUC of the second-order branch from WAS-induced IBS mice (n = 6) in the absence or presence of 100 μM L-NNA plus 10 μM INDO, or L-NNA + INDO (&) + 5 μM brefeldin A (BFA, n = 6). (D) Summarized data shows TPEN-induced CRC, R_max, AUC and EC₅₀ of healthy mice (Ctrl, n = 6), sham mice (n = 6) or WAS-induced IBS mice (n = 6). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 and ns: no significance.

Subsequently, we investigated whether TPEN-induced mesenteric vasorelaxation was affected in acute IBS mice. As shown in Fig. 8D, TPEN-induced vasorelaxation remained unaffected in IBS mice. Interestingly, 10 μM TPEN significantly alleviated the impairment of ACh-induced vasorelaxation in IBS mice (Fig. 8B). This observation suggests that TPEN may enhance EDH-mediated vasorelaxation, thereby improving the compromised ACh-induced vasorelaxation in IBS mice. Consequently, TPEN-induced vasorelaxation remains intact in acute IBS, whereas ACh-induced vasorelaxation is impaired, yet it can be rescued by TPEN.

3.8. TPEN-induced vasorelaxation of WAS-induced chronic IBS mice

We further examined if TPEN-induced vasorelaxation is also involved in chronic IBS after making WAS-induced IBS murine model for 10 days. As shown in Fig. 9A, fecal pellets of WAS-induced IBS mice were significantly increased compared to control mice although their body weights were similar, indicating successful setup of WAS-induced chronic IBS murine model for 10 days [38]. Second, vagus neurotransmitter ACh-induced vasorelaxation was obviously impaired in chronic IBS mice, which could be rescued by TPEN (Fig. 9B). Thirdly, the ability of TPEN to induce vasorelaxation remained unaffected in chronic IBS mice, as evidenced in Fig. 9C. Taken together, TPEN-induced vasorelaxation is well preserved in IBS while ACh-induced vasorelaxation is impaired, which could be rescued by TPEN.

Fig. 9.

The beneficial role of TPEN in chronic WAS-induced IBS.

(A) Summary data shows the time courses of body weight and stool pellets between the sham mice and WAS-induced IBS mice in ten days. (B) Summarized data shows acetylcholine (ACh, 10 nM - 1 mM)-induced CRC, R_max and AUC of the sham mice (n = 6), chronic WAS-induced IBS mice (n = 6) in the absence or presence of 10 μM TPEN. (C) Summarized data shows TPEN-induced CRC, R_max, AUC and EC₅₀ of healthy mice (Ctrl, n = 6), sham mice (n = 6) or chronic WAS-induced IBS mice (n = 6). *P < 0.05, ***P < 0.001, ****P < 0.0001 and ns: no significance.

4. Discussion

In the present study, we revealed that TPEN induces endothelium-dependent vasorelaxation of human submucosal arterioles and murine mesenteric arteries. Mechanistically, TPEN, as a chelator for the ER/Ca²⁺, activates endothelial SOCE to induce EDH-mediated vasorelaxation. TPEN-initiated SOCE/EDH-mediated vasorelaxation is impaired in UC but not in IBS. Moreover, TPEN can partially rescue vagus neurotransmitter ACh-induced vasorelaxation impaired in IBS. The followings are experimental evidence obtained from our study: 1) TPEN concentration-dependently induced vasorelaxation of human submucosal arterioles and murine mesenteric arteries in an endothelium-dependent manner; 2) TPEN induced EDH-mediated vasorelaxation via the SOCE likely composed of STIM/Orai; 3) TPEN-induced vasorelaxation via SOCE/EDH mechanism remained intact in IBS but was significantly impaired in UC; 4) ACh-induced vasorelaxation was impaired in IBS, which could be rescued by TPEN.

It is well established that vascular endothelial cells produce three relaxation factors: nitric oxide, PGI₂, and EDH to induce vasorelaxation. While NO and PGI₂ have been extensively researched, the exact nature of EDH remains largely unexplored [42]. A crucial aspect of the EDH mechanism is an increase in endothelial Ca²⁺ signaling [43,44]. EDH plays a pivotal role in regulating the activity of small resistant arteries in healthy conditions [11]. Furthermore, when NO-induced vasorelaxation is compromised in certain vascular diseases, EDH acts as a compensatory mechanism [12]. Therefore, investigating the role of EDH in both health and disease is of utmost importance, and this is a key research focus in our laboratory.

Previously, we have shown that EDH-mediated vasorelaxation could be a potential therapeutic target for IBD (inflammatory bowel disease) and sepsis, as it improves tissue blood flow [8,45]. Moreover, we have demonstrated the significant role of EDH in regulating vasorelaxation in small resistant arteries, and elucidated its regulatory mechanisms involving several Ca²⁺-permeable TRP channels [8,41]. However, the precise role of endothelial SOCE in regulating EDH-mediated vasorelaxation remains largely unclear.

The SOCE plays a critical role in modulating Ca²⁺ hemostasis in VECs to regulate vascular activity [17,42,43]. Despite its clarified role in NO-mediated vasorelaxation, the involvement of SOCE in EDH-mediated vasorelaxation remains enigmatic. In our current study, we employed TPEN, an endoplasmic reticulum Ca²⁺ chelator that stimulates SOCE, to delve deeper into the SOCE/EDH-mediated vasorelaxation of human submucosal arterioles and murine mesenteric arteries. Our findings indicate that: 1) TPEN triggers endothelium-dependent vasorelaxation, which is more pronounced in arteries pre-constricted by norepinephrine (NE) compared to those constricted by high K⁺. 2) The vasorelaxation induced by TPEN is unaffected by selective inhibitors of NO and PGI₂. 3) Inhibition of IK_Ca and SK_Ca channels significantly diminishes this vasorelaxation, while their activation enhances it. 4) Vasorelaxation induced by TPEN is attenuated by inhibitors of NCX (sodium-calcium exchanger) and NKA (sodium-potassium ATPase). 5) TPEN-induced vasorelaxation is reduced not only by SOCE blockers but also by inhibitors that affect STIM translocation and Orai transportation. 6) In human umbilical vein endothelial cells (HUVECs), TPEN-induced non-selective cation currents are profoundly inhibited by a SOCE blocker. Collectively, our results suggest that TPEN likely induces EDH-mediated vasorelaxation in human submucosal arterioles and murine mesenteric arteries via STIM1/Orai1-mediated SOCE. However, further molecular biological investigations are imperative to validate this hypothesis.

Since TPEN as the ER/Ca²⁺ chelator that does not influence [Ca²⁺]_cyt [33,34], it is likely more specific than CPA to decrease the ER/Ca²⁺ contents to activate the SOCE. Although TPEN is also a cytoplasmic Zn²⁺ ([Zn²⁺]_cyt) chelator, it was reported that [Zn²⁺]_cyt induced a marked endothelium-dependent vasorelaxation [3]. Therefore, TPEN would induce vasoconstriction if it acts as a [Zn²⁺]_cyt chelator in blood vessels, excluding the possibility of TPEN as a [Zn²⁺]_cyt chelator in the present study. In addition, the SOCE is also a major source for refilling the ER/Ca²⁺ stores in VSMC; however, it would prompt vasoconstriction by increasing [Ca²⁺]_cyt in VSMC [44]. Therefore, TPEN at least at the concentration applied in the present study selectively activate the SOCE on VECs to induce EDH-mediated vasorelaxation, which has been finally verified by our patch clamp study.

The vasorelaxation of mesenteric arteries was previously reported to impair in IBD patients, resulting in a reduction of blood perfusion in the inflammatory area to prompt IBD progression [23,24]. Therefore, improving mesenteric microcirculation is a potential strategy for IBD therapy. Certainly, our research has uncovered the involvement of EDH-mediated mesenteric artery in the underlying mechanisms of ulcerative colitis (UC) [8,17]; nonetheless, the precise role it plays in intestinal inflammation remains elusive. Similarly, although the role of SOCE in immune cells and intestinal epithelial cells has been implicated in the pathogenesis of inflammatory bowel disease (IBD), its specific mechanisms are still under investigation [26], little is known about participation of endothelial SOCE in IBD. Our current research has uncovered a significant finding: the TPEN-induced EDH-mediated vasorelaxation is compromised in colitis mice. This finding, coupled with our previous report regarding the impairment of CPA-induced vasorelaxation in colitis mice, strongly points to the crucial role played by endothelial SOCE/EDH in the pathogenesis of ulcerative colitis (UC). This suggests that targeting SOCE/EDH could be a promising therapeutic strategy for UC treatment.

IBS is the most common functional GI disorder, affecting up to 10%–20 % of the population without effective treatments [45]. Since current study has been focusing on immune system and gut microbiota [29], the involvement of mesenteric circulation and endothelial SOCE have not been explored so far. Recent study revealed a low-grade intestinal inflammation in IBS and substantial overlaps between IBS and IBD [30], suggesting a pathogenic role of inflammation in IBS [36]. Our findings have compelled us to delve deeper into the role of endothelial SOCE/EDH-mediated vasorelaxation in the development of inflammatory bowel disease (IBD). In our current study, we make a groundbreaking revelation: ACh-induced EDH-mediated vasorelaxation is significantly compromised in mice with water avoidance stress (WAS)-induced irritable bowel syndrome (IBS), indicating the involvement of mesenteric microcirculation in IBS. Intriguingly, TPEN-induced endothelial SOCE/EDH-mediated vasorelaxation remains unaffected in IBS mice, suggesting a potential compensatory mechanism that could rescue the impaired ACh-induced vasorelaxation. Therefore, TPEN-induced SOCE/EDH-mediated vasorelaxation is a potential strategy for IBS therapy.

Taken together, we revealed that TPEN/EDH-induced vasorelaxation of mesenteric arteries is preserved in IBS but impaired in UC. This may be due to the protective effect of TPEN in intestinal low-grade inflammation (IBS) but not in high-grade inflammation (IBD). Furthermore, our results demonstrate that TPEN is capable of partially reversing the impairment in ACh-induced vasorelaxation observed in IBS, as illustrated in Fig. 10. In our current study, we hypothesized that TPEN may restore the compromised ACh-induced vasorelaxation through its effects on EDH mechanisms. EDH mechanisms were involved in IK_Ca, SK_Ca on vascular endothelial cell, and NCX, NKA of vascular smooth muscle cell et al. Therefore, it needed further experiments to test the specific target and mechanisms of TPEN-improved the impaired vasorelaxation. Indeed, due to the deleterious effect of TPEN on cells, such as induction of apoptosis, DNA damage, and intracellular production of reactive oxygen species, it needed the further experiments to test the safety and effectiveness on the whole animals. It was acknowledged that the mesenteric circulation plays an irreplaced role in maintaining the GI functions and prompting the recovery after injury via blood perfusion.

Fig. 10.

The underlying mechanisms of theTPEN/SOCE/EDH-mediated vasorelaxation of mesenteric arterioles in health and its role in IBS and IBD.

The SOCE/EDH-mediated vasorelaxation in health (left panel), its role in IBS (middle panel) and colitis (right panel). (A) TPEN that can chelate Ca²⁺ in the ER to activate endothelial STIM1/Orai1-dependent SOCE and increase [Ca²⁺]_cyt signaling to activate IK_Ca and SK_Ca on VEC, leading to K⁺ efflux. An increase in extracellular K⁺ between VEC and VSMC activates NKA to eventually cause vasorelaxation through hyperpolarization. Moreover, NKA activation reduces [Na⁺]_i in VSMC, which stimulates NCX activity to decreases [Ca²⁺]_i, resulting in further vasorelaxation. (B) In the acute or chronic WAS-induced IBS, ACh-induced vasorelaxation was impaired, but TPEN-induced vasorelaxation remains intact, which can rescue the impaired ACh-induced vasorelaxation. (C) However, TPEN/SOCE/EDH-induced vasorelaxation was significantly impaired in DSS-induced colitis. VEC: vascular endothelial cells; VSMC: vascular smooth muscle cells; ER: endoplasmic reticulum; IK_Ca and SK_Ca: intermediate and small conductance of Ca²⁺-activated K⁺ channels; NKA: Na⁺/K⁺-ATPase; NCX: Na⁺/Ca²⁺-exchanger.

In summary, TPEN triggers EDH-mediated vasorelaxation in mesenteric arteries, likely through a mechanism involving the STIM/Orai complex of SOCE, which is well-preserved in IBS but impaired in UC. Therefore, our current study reveals that the SOCE/EDH may rescue endothelial dysfunction, highlighting it as a potential target for IBS therapy.

Key points

• 1.TPEN prompts EDH-mediated vasorelaxation in mesenteric arteries, presumably mediated by the STIM/Orai complex of store-operated calcium entry (SOCE).
• 2.TPEN/EDH-induced vasorelaxation is preserved in IBS but impaired in UC.
• 3.TPEN exhibits the ability to partially mitigate the compromised ACh-induced vasorelaxation observed in IBS.

Funding statement

This work was supported by research grants from the National Natural Science Foundation of China (82273115 to HD).

Ethics stataments

The experiments conducted on animals adhered strictly to the procedures and policies approved by the Medical College of Qingdao University (Approval No. QDU-AEC-2023073). Additionally, the study involving clinical samples was conducted with the approval of the Clinical Research Ethics Committee of the Army Medical University, located in Chongqing, China (Approval No. AMUWEC2020368).

Data availability statement

All data generated or analyzed throughout this study are comprehensively presented in this article. If any further data is required, we are willing to provide it upon request.

CRediT authorship contribution statement

Luyun Zhang: Writing – original draft, Methodology, Investigation. Shaoya Rong: Formal analysis. Jianxin Wang: Formal analysis. Hanxing Wan: Data curation. Feng Xu: Validation. Hui Dong: Writing – review & editing.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests Hui Dong reports financial support was provided by The National Natural Science Foundation of China. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article.