Sodium butyrate prevents cytokine‐induced β‐cell dysfunction through restoration of stromal interaction molecule 1 expression and activation of store‐operated calcium entry

Abstract

Sodium butyrate (NaB) improves β‐cell function in preclinical models of diabetes; however, the mechanisms underlying these beneficial effects have not been fully elucidated. In this study, we investigated the impact of NaB on β‐cell function and calcium (Ca²⁺) signaling using ex vivo and in vitro models of diabetes. Our results show that NaB significantly improved glucose‐stimulated insulin secretion in islets from human organ donors with type 2 diabetes and in cytokine‐treated INS‐1 β cells. Consistently, NaB improved glucose‐stimulated Ca²⁺ oscillations in mouse islets treated with proinflammatory cytokines. Because the oscillatory phenotype of Ca²⁺ in the β cell is governed by changes in endoplasmic reticulum (ER) Ca²⁺ levels, we explored the relationship between NaB and store‐operated calcium entry (SOCE), a rescue mechanism that acts to refill ER Ca²⁺ levels through STIM1‐mediated gating of plasmalemmal Orai channels. We found that NaB treatment preserved basal ER Ca²⁺ levels and restored SOCE in IL‐1β‐treated INS‐1 cells. Furthermore, we linked these changes with the restoration of STIM1 levels in cytokine‐treated INS‐1 cells and mouse islets, and we found that NaB treatment was sufficient to prevent β‐cell death in response to IL‐1β treatment. Mechanistic experiments revealed that NaB mediated these beneficial effects in the β‐cell through histone deacetylase (HDAC) inhibition, iNOS suppression, and modulation of AKT‐GSK‐3 signaling. Taken together, these data support a model whereby NaB treatment promotes β‐cell function and Ca²⁺ homeostasis under proinflammatory conditions through pleiotropic effects that are linked with maintenance of SOCE. These results also suggest a relationship between β‐cell SOCE and gut microbiome‐derived butyrate that may be relevant in the treatment and prevention of diabetes.

Sodium butyrate (NaB) augments insulin secretion in human islets from donors with type 2 diabetes and in INS‐1 β cells treated with pro‐inflammatory cytokines. NaB‐mediated improvements in insulin secretion were associated with restoration of Ca²⁺ oscillations, Stim1 expression, and store‐operated calcium entry (SOCE). Mechanistic experiments revealed that NaB rescued SOCE via inhibition of histone deacetylase activity and iNOS suppression. These results suggest a novel pathway through which NaB exerts protective metabolic effects in the pancreatic β cell via modulation of Ca²⁺ homeostasis.

Abbreviations

AKT

Ak strain transforming

Cyto

cytokine

Dz

diazoxide

ER

endoplasmic reticulum

ERK1/2

extracellular signal‐regulated kinase 1/2

Fura‐2‐AM

Fura‐2‐acetoxymethyl ester

GPCR

G‐protein‐coupled receptor

GPR41

G‐protein‐coupled receptor 41

GPR43

G‐protein‐coupled receptor 43

GSIS

glucose‐stimulated insulin secretion

GSK‐3

glycogen synthase kinase‐3

HDAC

histone deacetylase

HDI

histone deacetylase inhibitor

IBD

irritable bowel disease

IFN‐γ

interferon‐γ

IL‐1β

interleukin‐1β

IP3

inositol triphosphate

ITSA1

inhibitor of TSA‐1

KO

knockout

L‐NMMA

L‐NG‐monomethyl arginine acetate

LPS

lipopolysaccharide

NaB

sodium butyrate

NLRP3

nucleotide‐binding oligomerization domain‐like receptors family pyrin domain containing 3

ROI

region of interest

RRID

Research Resource Identifiers

SAB

secretion assay buffer

SAHA

Suberoylanilide hydroxamic acid

SCFA

short‐chain fatty acids

SERCA

sarco‐ER Ca²⁺ ATPase

SOCE

store‐operated calcium entry

SSA

singular spectrum analysis

STIM1

stromal interaction molecule 1

T1D

type 1 diabetes

T2D

type 2 diabetes

TG

thapsigargin

TNF‐α

tumor necrosis factor‐α

TSA

trichostatin A

WT

wild type

1. INTRODUCTION

Diabetes is a chronic condition that affects more than 34 million people in the United States and is caused by near complete loss of insulin secretion in the case of type 1 diabetes (T1D) or insulin resistance that is accompanied by inadequate insulin secretion in type 2 diabetes (T2D). ¹ Although inflammation‐induced β‐cell dysfunction is a prominent feature of both T1D and T2D, ² , ³ therapies that improve β‐cell health apart from those that increase insulin secretion are lacking in both conditions. ³ , ⁴ , ⁵

Recent findings show that alterations in the gut microbiome, coupled with changes in the repertoire of bacterially derived short‐chain fatty acids (SCFA), are linked to metabolic dysfunction and diabetes. ⁶ , ⁷ One prominent and largely consistent change observed in both T1D and T2D is reduced microbiome production of sodium butyrate (NaB), a four‐carbon SCFA produced from fermentation of dietary fiber. ⁸ , ⁹ Recently, Sanna and colleagues analyzed genome‐wide association data, gut metagenomic sequencing, and measurements of fecal SCFAs performed in over 900 non‐diabetic individuals from the LifeLines‐DEEP (LL‐DEEP) cohort. Using Mendelian randomization to assess causality, they showed that host‐genetic‐driven increases in gut production of NaB were associated with improved insulin secretion after an oral glucose tolerance test. ¹⁰ Similarly, strategies to increase NaB either by dietary manipulation or via direct augmentation have shown positive effects on metabolic measures including body weight and glycemia, although the effects are not consistent across different studies. ¹¹

Preclinic studies suggest that NaB treatment may be linked with improvements in glycemia, dyslipidemia, insulin resistance, and β‐cell health. ¹² , ¹³ Several mechanisms have been implicated in these beneficial effects, including: (1) inhibition of histone deacetylases (HDACs) and changes in cellular transcriptional programs, including those that regulate inflammation, ¹⁴ (2) activation of G‐protein‐coupled receptors (GPCR), (3) improvements in mitochondrial function, ¹² (4) inhibition of NF‐κB‐mediated production of inducible nitric oxide synthase (iNOS) and nitric oxide (NO), ¹⁵ and (5) reductions in inflammasome activation. ¹⁶ Of particular relevance to the β cell, a recent study showed that NaB protects mouse islets, INS‐1E β cells, and EndoC βH1 cells against cytokine‐induced reductions in insulin secretion and loss of insulin content, while a subset of gene expression changes induced by IL‐1β were reversed with NaB. ¹⁷ Subsequently, this same group showed that NaB acts as an HDAC inhibitor in cytokine‐treated β cells. ¹⁸

Taken together, existing literature suggests a beneficial role for NaB in metabolic control and β‐cell function. However, the exact mechanisms by which NaB exerts potentially beneficial effects on the β cell are not fully understood, and the link between NaB and Ca²⁺ signaling in the β cell has not been explored. Herein, we highlight a novel pathway linking NaB treatment with upregulation of β‐cell expression of stromal interaction molecule 1 (STIM1), an endoplasmic reticulum (ER)‐resident protein that activates store‐operated calcium entry (SOCE) in response to ER Ca²⁺ depletion. ¹⁹ Our results reveal a unique ability of NaB to rescue IL‐1β‐mediated reductions in STIM1 and SOCE, with downstream impacts on glucose‐induced Ca²⁺ signaling, insulin secretion, and β‐cell viability. These results suggest a previously unrecognized relationship between SOCE and NaB that may have relevance in strategies to treat diabetes and/or improve metabolic health through manipulation of the gut microbiome.

2. MATERIALS AND METHODS

2.1. Materials

Sodium butyrate (NaB), diazoxide, verapamil, and recombinant human interleukin‐1β (IL‐β) were obtained from Millipore Sigma (St Louis, MO, USA). Thapsigargin (TG), suberoylanilide hydroxamic acid (SAHA), LMK235, MS275, EX527, Trichostatin A (TSA), inhibitor of TSA‐1 (ITSA1), L‐N^G‐monomethyl arginine acetate (L‐NMMA), and AR420626 (AR) were purchased from Cayman Chemicals (Ann Arbor, MI, USA). The mouse interferon‐γ (IFN‐γ) and human tumor necrosis factor‐α (TNF‐α) used in the incubation of β cells and isolated islets were purchased from Invitrogen (Carlsbad, CA, USA). Ca²⁺ imaging of mouse islets was performed with Fura‐2 AM (Fura‐2‐acetoxymethyl ester) acquired from Thermo Fisher Scientific (Waltham, MA, USA).

2.2. Animals, islets, and cell culture

C57BL/6J mice (Jackson Laboratories, Bar Harbor, ME, USA) were maintained under a protocol approved by the Indiana University Institutional Animal Care and Use Committee (IACUC). Mice were kept on a standard light–dark cycle with ad libitum access to food and water. Mouse pancreatic islets were isolated by collagenase digestion, hand‐picked, and allowed to recover overnight as described previously. ²⁰ Rat INS‐1832/13 (INS‐1) cells (RRID:CVCL_7226) were cultured as previously described. ²¹ Cadaveric human islets from non‐diabetic donors and donors with T2D were obtained from the Integrated Islet Distribution Program (IIDP) or the National Disease Research Interchange (NDRI). Characteristics of human islet donors are shown in Table S1. Upon receipt, human islets were hand‐picked and allowed to recover overnight in DMEM supplemented with 5.5 mM glucose, 10% fetal bovine serum, and 100 units/mL penicillin–streptomycin.

2.3. Glucose‐stimulated insulin secretion

Glucose‐stimulated insulin secretion (GSIS) assays were performed as previously described. ²² Briefly, cultured INS‐1 cells or isolated islets were incubated for 2 h in 1x secretion assay buffer (SAB) containing 2.5 mM glucose and then in 1x SAB containing 15 mM glucose for an additional 2 h. Insulin secretion was measured in the harvested SAB using either a human insulin ELISA kit (Cat#10‐1132‐01, Mercodia, Winston Salem, NC, USA) or a rat insulin ELISA kit (Cat#10‐1250‐01, Mercodia, Winston Salem, NC, USA) according to the manufacturer's instructions. The secreted insulin levels were normalized to total protein content or expressed as a stimulation index. The stimulation index in each sample was calculated using the formula below: insulin level at 15 mM glucose/insulin level at 2.5 mM glucose.

2.4. Ca²⁺ imaging

Intracellular Ca²⁺ was measured using the FLIPR Ca²⁺ 6 Assay Kit and a FlexStation 3 system (Molecular Devices, Sunnyvale, CA, USA). In brief, INS‐1 cells were plated in black wall/clear bottom 96‐multiwell plates (Costar, Tewksbury, MA, USA), cultured to ~90% confluency, and treated with drugs and/or stress conditions as indicated. INS‐1 cells were simultaneously treated with IL‐1β alone or in combination with NaB, other specified HDAC inhibitors, L‐NMMA, or AR420626. Following the indicated treatment, calcium 6 reagent was added to the medium, and cells were incubated for an additional 2 h at 37°C and 5% CO₂. Cells were then incubated for ~4 min in Ca²⁺‐free Hanks' balanced salt solution (HBSS) containing 5.5 mM glucose, 0.5 mM EGTA, 10 μM verapamil, and 200 μM diazoxide. The baseline fluorescence (F0) was measured for a minimum of 10 s in a Ca²⁺‐free HBSS medium. TG was then added at ~10 s to deplete ER Ca²⁺ stores, followed by supplementation with Ca²⁺ at ~140 s to reach 2 mM Ca²⁺ in the medium. SOCE was determined by the formula ΔF/F0, where ΔF is the increase in calcium 6 fluorescence intensity. Data were acquired at 1.52‐s intervals using an excitation wavelength of 485 nm and an emission wavelength of 525 nm.

Imaging of glucose‐stimulated Ca²⁺ oscillations in islets was performed using islets from mice expressing GCaMP6s, which is a genetically encoded fluorescent indicator of cytosolic Ca²⁺. The GCaMP6s expressing mice were obtained by crossing Ins1 ^Cre mice (The Jackson Laboratory, Strain #26801) with GCaMP6s Cre‐inducible mice (The Jackson Laboratory, Strain #028866). Isolated mouse islets were allowed to recover overnight and then treated for 24 h with NaB and/or a cytokine mixture containing 5 ng/mL IL‐1β, 100 ng/mL IFN‐γ, and 10 ng/mL TNF‐α. Baseline measurements were performed at 5.5 mM glucose and Ca²⁺ transients were measured in response to 11 mM glucose using a Zeiss Z1 microscope as described previously. ²² Analysis of Ca²⁺ imaging data and activity was performed according to previously described methods and quantitated as activity under basal 5.5 mM glucose conditions and during the first‐phase responses to glucose. ²³ , ²⁴ In brief, regions of interest (ROIs) were identified by automatic software algorithm. Significant changes in cytosolic Ca²⁺ were automatically annotated as events (with z score >3) within the ROIs. Events were characterized by their start time, peak time, and the width at half of their peak amplitude (events' halfwidth). Activity at each of the indicated phases was quantitated as the sum of AUC of individual detected events; sums were log₁₀ transformed for presentation.

The analysis of the Ca²⁺ oscillatory signal necessitates an understanding of the distinct frequencies present within the time series data. In brief, the oscillatory time series data underwent initial processing using singular spectrum analysis (SSA). This technique decomposes the time series data into a serial of singular functions with multiple eigen‐frequencies, each function possessing a corresponding amplitude coefficient that signifies its contribution to the original time series. ²⁵ This decomposition method facilitates the segregation of oscillation data into components with varying eigen‐frequencies, as well as minor fluctuations and overall trends. Components exhibiting lesser contributions were identified as noise and consequently excluded from subsequent analysis. Subsequently, the Fourier transform was employed to convert the components primarily responsible for the oscillatory signals from the time domain to the frequency domain. ²⁶ A peak detection algorithm was applied to the Fourier spectra to identify the significant frequencies within the oscillations. Finally, violin plots were generated to visualize the prevalent significant frequencies. Jenks optimization method was used to determine the different populations for all detected frequencies and the separation breaks between those populations. Two groups of oscillation frequencies were identified: the slow base oscillation and the fast oscillation. The threshold of separation between these two groups can be used to calculate the fast Ca²⁺ oscillation frequency percentage for each islet condition.

2.5. Cell viability assay

INS‐1 cells were cultured in black wall/clear bottom 96‐multiwell plates to reach ~70% confluence. Following treatment with NaB and/or IL‐1β for 24 h, 100 μL of CellTiter‐Glo Luminescent Cell Viability Assay reagent (Promega, Madison, WI, USA) was added to the medium. After a 10‐min incubation at room temperature, luminescence was measured with a plate reader (SpectraMax iD5 Multi‐Mode microplate reader, Molecular Devices). The luminescent reading of vehicle‐treated cells was set as 100%, and the percentage viability relative to vehicle groups was reported.

2.6. Immunoblot

Immunoblot experiments were performed as previously described ²⁷ , ²⁸ using antibodies against cleaved caspase‐3 (Cell Signaling Technology, Cat#9664, RRID:AB_2070042), procaspase‐3 (Cell Signaling Technology, Cat#9662, RRID:AB_331439), STIM1 (BD Biosciences, Cat#610954, RRID:AB_398267), inducible nitric oxide synthase (iNOS) (Millipore EMD, Cat#ABN26, RRID:AB_10805939), and β‐actin (Millipore Sigma, Cat#MAB1501, RRID:AB_2223041). All primary antibodies were used at a dilution of 1:1000, except STIM1 at 1:500. Species‐matched secondary antibodies (LI‐COR Bioscience, Cat# 926‐32212; RRID:AB_621847 and LI‐COR Bioscience, Cat# 926‐68 073; RRID:AB_10954442) were used. The images were analyzed using LI‐COR Biosciences Image Studio.

2.7. Quantitative RT‐PCR

Total RNA was extracted from INS‐1 cells or isolated islets using an RNeasy Mini Plus kit or RNeasy Micro Plus kit, respectively (Qiagen, Valencia, CA, USA). Isolated RNA was reverse‐transcribed into cDNA using M‐MLV Reverse Transcriptase (Invitrogen, Carlsbad, CA, USA) as described previously. ²⁹ The cDNA was subjected to quantitative PCR analysis using SsoAdvanced Universal SYBR Green Supermix (BioRad, Hercules, CA, USA). Relative gene expression levels were established against the housekeeping gene β‐actin, using the comparative delta Ct method, as described previously. ³⁰ Primer sequences are provided in Table 1.

TABLE 1.

List of primers used in the study.

Gene	Species	Forward primer sequence (5′ → 3′)	Reverse primer sequence (5′ → 3′)
STIM1	Mouse, rat	AGCCTCAGCCATAGTCACAG	TTCCACATCCACATCACCATTG
STIM2	Mouse	ACACTCCCCAGGATAGCAGT	GTTATGAGGTGGGCGTGTCA
STIM2	Rat	CAACGGCATCCTGGAGAAGT	CTGGAGGCTTCCTGAACTGG
Orai1	Mouse, rat	AGGTGATGAGCCTCAACGAG	CTGATCATGAGGGCAAACAG
Orai2	Mouse, rat	GGCCACAAGGGCATGGATTA	TGAGGGTACTGGTACTTGGTC
Orai3	Mouse, rat	GCCTGCACCACTGTGTTAGTA	TGTTGCTCACGGCTTCAATATG

2.8. HDAC activity assay

HDAC activity in INS‐1 cells treated with NaB or NaB/IL‐1β for 1 hour was determined using the HDAC‐Glo™ I/II Assay according to the procedure recommended by the manufacturer (Promega, Madison, WI, USA). This kit detects class I and class II HDAC activity.

2.9. Phosphorylation array assay

C57BL/6J mouse islets were isolated and allowed to recover overnight. Islets obtained from 14 mice were pooled and ~ 700 hand‐picked islets were distributed for each treatment group. The islets were treated with NaB and/or cytokines (IL‐1β, IFN‐γ, and TNF‐α) for 16 h. The treated islets were lysed using the lysis reagent provided in the phosphorylation array kit from RayBiotech Life (Peachtree Corners, GA, USA, Cat#AAH‐MAPK‐1‐8). The lysates were incubated with membranes pre‐coated with anti‐phosphorylated proteins overnight and incubated with anti‐rabbit IgG (detection antibody cocktail) overnight. The membranes were then incubated with HRP‐conjugated anti‐IgG. The membranes were incubated in a detection buffer and imaged with a chemiluminescent imaging system. Fiji software was used to analyze the resulting data.

2.10. Statistical analysis

Unless otherwise indicated, results are presented as the mean ± SEM. The statistical significance of differences between groups was analyzed using GraphPad Prism software (San Diego, CA, USA). Comparisons between the two groups were performed using an unpaired Student's t‐test, while comparisons among multiple groups used one‐way ANOVA with Tukey–Kramer post‐hoc analysis. A p‐value <.05 was considered statistically significant.

3. RESULTS

3.1. NaB increases insulin secretion in human islets from donors with T2D and INS‐1 cells under IL‐1β stress

To test whether NaB has beneficial effects on insulin secretion in humans, islets isolated from human cadaveric donors with T2D were incubated in the absence or presence of NaB for 24 h prior to a GSIS assay. NaB treatment significantly increased insulin secretion in islets from donors with T2D in comparison to vehicle‐treated T2D islets (Figure 1A). To explore whether NaB had similar effects under pro‐inflammatory conditions that typify T2D, ³¹ INS‐1 cells and islets from healthy human donors were treated with proinflammatory cytokines. As expected, cytokine treatment of INS‐1 cells (IL‐1β, Figure 1B) and human islets (IL‐1β, IFN‐γ, and TNF‐α, Figure 1C) significantly reduced GSIS. NaB was able to reverse this secretory defect in INS‐1 cells (Figure 1B), and in cytokine‐treated human islets from non‐diabetic donors, insulin secretion was modestly increased by NaB co‐treatment, but not to a significant extent (Figure 1C).

FIGURE 1.

NaB increases insulin secretion in human islets from donors with T2D and INS‐1 cells under IL‐1β stress. (A) Human islets from two donors with type 2 diabetes (T2D) were incubated with or without 0.5 mM NaB for 24 h, and GSIS was performed. Insulin levels were determined using a human insulin ELISA kit and normalized to total protein content in the lysates. The open circles indicate ELISA replicates from two donors. (B) INS‐1 cells were treated with 5 ng/mL IL‐1β, and GSIS was performed. Insulin levels were measured using a rat insulin ELISA kit and data were normalized to total protein content. The values obtained in 3–4 independent experiments are represented by open circles. (C) Human islets from four healthy donors were incubated with 5 ng/mL IL‐1β, 100 ng/mL IFN‐γ, and 10 ng/mL TNF‐α (cytokine) in the absence or presence of 0.5 mM NaB for 24 h, and GSIS was performed. Insulin levels were determined using a human insulin ELISA kit. The stimulation index was obtained by calculating the fold change between the insulin level at 15 mM glucose and the insulin level at 2.5 mM in the same group. Results are shown as the mean ± SEM; *p < .05, **p<0.01, ***p < .001, ****p < .0001.

3.2. NaB prevents cytokine‐induced dysfunction in Ca²⁺ oscillations

Impaired insulin secretion is associated with defective Ca²⁺ homeostasis and impaired stimulus–secretion coupling. ² We have previously linked changes in Ca²⁺ oscillations with alterations in glucose‐stimulated Ca²⁺ responses in islets. ²² Therefore, to test whether NaB impacts Ca²⁺ homeostasis, islets were isolated from Ins‐1 cre‐dependent GCaMP6s mice that express a fluorescent indicator specifically in β cells. ³² The isolated islets were treated for 24 h with a combination of IL‐1β, IFN‐γ, and TNF‐α in the presence or absence of NaB, and glucose‐stimulated Ca²⁺ oscillations were measured. In the absence of cytokine treatment, vehicle‐ and NaB‐treated islets showed apparent first‐phase responses followed by second‐phase Ca²⁺ oscillations in response to glucose stimulation, as previously reported ²² (Figure 2A,B). However, upon cytokine treatment, the first‐phase Ca²⁺ response was significantly reduced, while co‐treatment with NaB restored this response (Figure 2C–E). Quantitative analysis of the fraction of active β cells during glucose stimulation also showed that cytokine treatment resulted in loss of the first‐phase glucose response, which was restored by NaB treatment (Figure 2F). To determine the effect of NaB on second‐phase responses, frequencies from the Ca²⁺ oscillations from each group were extracted and represented in violin plots (Figure 2G). All violin plots exhibited a comparable distribution in the lower frequency range, suggesting a fundamental slow oscillatory signal common to all conditions. However, a significant number of frequencies detected from cytokine‐treated islets (32.09%) were in the fast oscillation (high frequency) range, as compared to the low percentage in this range seen in vehicle‐treated and NaB‐treated islets (5.54% and 1.21%, respectively) (Figure 2G). We also observed that the accelerated frequencies in the second‐phase Ca²⁺ oscillations due to cytokine stress were drastically diminished by co‐treatment with NaB (1.56% of detected frequencies were fast oscillations). These data suggest that NaB treatment suppresses cytokine‐dependent effects on both first‐phase and second‐phase Ca²⁺ responses in islets.

FIGURE 2.

NaB prevents cytokine‐induced dysfunction in Ca²⁺ oscillations. Islets obtained from mice carrying the Ins‐Cre inducible GCaMP6s Ca²⁺‐binding fluorescent indicator were treated with 5 ng/mL IL‐1β, 100 ng/mL IFN‐γ, and 10 ng/mL TNF‐α in the absence or presence of 0.5 mM NaB for 24 h. The number of islets measured in each group was as follows: vehicle (n = 14), NaB (n = 11), cytokine (n = 27), and cytokine + NaB (n = 14). Individual cells were determined as regions of interest (ROIs) obtained by an automated segmentation algorithm. We discarded ROIs with number of events below 5. (A–D) Top panels: [Ca²⁺] events' halfwidth duration through time. Color indicates the statistical significance in terms of z‐score. Middle panels: A raster plot of cytoplasmic [Ca²⁺] showing events in individual ROIs. Color indicates the duration of events, with long events indicated in blue and short events in red. Lower panels: Representative cytosolic Ca²⁺ traces after treatment with 11 mM glucose from a number of ROIs that represent the average activity of the islets. (E) Activity for the baseline and phase 1 and phase 2 responses is displayed as the log10 value. Results are displayed as means ± SEM; **p < .01, ****p < .0001. (F) Pooled data showing the fraction of active cells during glucose‐dependent activation. (G) Violin plots of the significant frequencies detected from the Fourier transform of singular spectrum analysis (SSA) components. Dashed line indicates the cluster separation threshold determined by Jenks optimization method.

3.3. NaB improves SOCE in IL‐β‐treated INS‐1 cells

Our previous studies showed that changes in Ca²⁺ oscillatory activity were associated with impaired SOCE. ²² Therefore, to determine if NaB impacts cellular Ca²⁺ homeostasis and SOCE under cytokine stress, INS‐1 cells were treated for 24 h with IL‐1β in the absence or presence of NaB. The depletion of ER Ca²⁺, as shown by the increase of Ca²⁺ fluorescence at the first 140 s, and subsequent SOCE‐mediated replenishment of Ca²⁺ stores, shown by the increase of Ca²⁺ fluorescent after 140 s, were characterized as detailed in Figure 3A. First, the uptake of Ca²⁺ from the extracellular milieu was eliminated by blockade of plasma membrane Ca²⁺ channels with diazoxide and verapamil and incubation of the cells in Ca²⁺ free buffer with EGTA chelation. The first measured peak of cytosolic Ca²⁺ reflects the release of stored Ca²⁺ from the ER via dosing with TG, which inhibits SERCA‐mediated Ca²⁺ uptake into the ER. Next, extracellular Ca²⁺ was supplemented, allowing Orai‐ and STIM1‐mediated SOCE, which was measured in the second peak of cytosolic Ca²⁺. As shown previously, ²² treatment with IL‐1β significantly reduced ER Ca²⁺ stores (a negligible first peak in response to TG) and SOCE (reduced second peak following Ca²⁺ supplementation) (Figure 3B–D). In contrast, simultaneous incubation of NaB and IL‐1β partially maintained ER Ca²⁺ stores (Figure 3B,C) and SOCE responses (Figure 3B,D) in INS‐1 cells.

FIGURE 3.

NaB improves ER Ca²⁺ store and SOCE in IL‐β‐treated INS‐1 cells. (A–E) Levels of stored ER Ca²⁺ and store‐operated Ca²⁺ entry (SOCE) were measured with Calcium 6 fluorescent dye in INS‐1 cells after 24‐h incubation performed under vehicle, IL‐1β (5 ng/mL), NaB (0.5 mM), and IL‐1β + NaB (5 ng/mL and 0.5 mM, respectively) treatment conditions. (A) The assay paradigm is illustrated. (B) Representative experimental baseline‐normalized fluorescence tracings. (C) Quantification of baseline‐normalized Ca²⁺ release from the ER. (D) Quantification of baseline‐normalized SOCE. The values obtained in eight independent experiments are represented by open circles, and the results are shown as the mean ± SEM; ***p < .001, **** p < .0001.

3.4. NaB improves cell viability

Impairments in SOCE have been linked with reduced ER Ca²⁺ and increased β‐cell death. ²² To test whether NaB treatment improves β‐cell viability, INS‐1 cells were exposed for 24 h to IL‐1β in the absence or presence of NaB, and cell viability was determined. As expected, IL‐1β treatment significantly reduced cell viability to 42.3% compared with that in cells without any treatment; however, co‐treatment with NaB significantly increased viability to 62.4% (Figure 4A), which is a 20.1% increase in the β‐cell viability. Consistent with this observation, NaB reduced the production of cleaved caspase‐3 in IL‐1β‐treated INS‐1 cells by 36.9% (Figure 4B,C).

FIGURE 4.

NaB improves cell viability in IL‐1β stressed β cells. (A–C) Cells were incubated in the absence or presence of IL‐1β (5 ng/mL) and NaB (0.5 mM) for 24 h prior to assays of cell viability and caspase‐3 cleavage. (A) CellTiter‐Glo luminescent cell viability assay. Data are from two independent experiments with four repeats in each experiment. (B) Immunoblot of cleaved caspase‐3 and procaspase‐3; Data are from four independent experiments. (C) Quantification of relative signals within (B) for the ratio of cleaved caspase‐3 normalized to procaspase‐3. The results are shown as the mean ± SEM; **p < .01, ***p < .001, ****p < .0001.

3.5. NaB mediates rescued expression of SOCE‐related genes

NaB exerts genomic effects through increasing histone acetylation and has been linked with increased acetylation in non‐histone proteins, such as SP1. ³³ Furthermore, NaB functions as a transcription modulator and an anti‐inflammatory agent to regulate metabolic pathways. ¹⁴ , ³⁴ Therefore, we performed temporal studies to characterize the mechanisms underlying NaB‐mediated improvements in SOCE response in INS‐1 cells. NaB treatment improved SOCE in INS‐1 cells exposed to IL‐1β under chronic (24 h) but not acute (4 or 8 h) conditions (Figure 5A). This finding raised the possibility that the beneficial effects of NaB involved changes in gene expression, which usually require time to manifest. Therefore, quantitative RT‐PCR (qRT‐PCR) was performed on INS‐1 cells with or without NaB/cytokine treatment for 24 h to measure the expression level of genes known to regulate SOCE. The expression of Orai2 was unchanged by the treatment conditions, while Orai3 expression was increased in IL‐1β‐treated INS‐1 cells and decreased upon co‐treatment with IL‐1β and NaB (Figure 5B). Interestingly, NaB treatment led to upregulated expression of Stim1, Stim2, and Orai1 mRNA in IL‐1β‐treated cells (Figure 5B). However, only the expression of Stim1 was reduced by IL‐1β (Figure 5B), suggesting that NaB‐mediated upregulation of Stim1 expression may improve SOCE and overall β‐cell function.

FIGURE 5.

NaB mediates rescued expression of SOCE‐related genes. (A and B) INS‐1 cells were incubated in the absence or presence of IL‐1β (5 ng/mL) and NaB (0.5 mM) over a time course of 4–24 h followed by Ca²⁺ imaging and analysis as described in Figure 3. Samples were also collected at 24 h for quantitative RT‐PCR analysis of gene expression levels. (A) Baseline‐normalized quantitation of the SOCE response (ΔF/F0) from imaging tracings performed at 4, 8, and 24 h of co‐incubation. (B) Expression levels of SOCE‐related genes (Stim1, Stim2, Orai1, Orai2, and Orai3). The expression level of β‐actin served as an internal control for normalization of SOCE gene expression; n = 3. (C) Mouse islets were incubated in the presence of 5 ng/mL IL‐1β, 100 ng/mL IFN‐γ, and 10 ng/mL TNF‐α in the absence or presence of 0.5 mM NaB for 24 h and were subjected to quantitative RT‐PCR analysis of genes related to SOCE, including Stim1, Stim2, Orai1, Orai2, and Orai3. The expression level of β‐actin served as an internal control for normalization of SOCE gene expression and data are expressed relative to vehicle incubation; n = 5. Cyto, cytokine. (D and E) INS‐1 cells were incubated in the absence or presence of IL‐1β (5 ng/mL) and NaB (0.5 mM). Samples were collected at 24 h for immunoblotting. (D) A representative immunoblot of STIM1 protein levels. The level of β‐actin protein served as an internal control. (E) Quantitation of the ratio of STIM1 to β‐actin protein levels for (A). Data are expressed relative to the vehicle group; n = 4. The results are shown as the mean ± SEM; *p < .05, **p < .01, ***p < .001, ****p < .0001.

Similar to what we observed in INS‐1 cells (Figure 5B), cytokines reduced Stim1 expression in mouse islets, while co‐treatment with NaB rescued Stim1 expression (Figure 5C). The expression of Stim2, Orai1, and Orai3 did not vary among islet treatment groups, while Orai2 expression levels were augmented by NaB treatment (Figure 5C). Consistent with the mRNA expression results reported above, increased STIM1 protein levels were detected in lysates prepared from INS‐1 cells co‐incubated with NaB and IL‐1β (Figure 5D,E), suggesting an association between NaB, rescued SOCE responses, and the expression levels of STIM1.

3.6. NaB improves ER Ca²⁺ stores and SOCE through HDAC inhibition and iNOS suppression

NaB is known to inhibit class I and class IIa HDACs, including HDAC 1, 2, 3, 4, 5, 7, 8 and 9. ³³ To test whether NaB functions as an HDAC inhibitor in our model, we measured HDAC activity in INS‐1 cells treated with IL‐1β or IL‐1β + NaB for 1 h. Co‐treatment with IL‐1β and NaB significantly reduced HDAC activity to 65.4% of the activity present in IL‐1β‐treated cells (Figure 6A), verifying that NaB is acting as an HDAC inhibitor in IL‐1β‐treated INS‐1 cells. Next, to determine if a pan‐HDAC inhibitor showed similar effects on β‐cell Ca²⁺ regulation, we used trichostatin (TSA, an inhibitor of class I and II histone deacetylases) and ITSA1 (an HDAC activator via suppressing TSA). We found that TSA was able to rescue ER Ca²⁺ stores and SOCE responses in IL‐1β‐treated INS‐1 cells, while co‐treatment with TSA and ITSA1 blocked the rescue effects of TSA on Ca²⁺ released from the ER (ER Ca²⁺ stores) and SOCE (Figure 6B,C). These findings show that HDAC inhibition via treatment with either NaB or TSA alters β‐cell Ca²⁺ stores and regulation.

FIGURE 6.

NaB improves ER Ca²⁺ stores and SOCE through HDAC inhibition and iNOS suppression. (A) HDAC activity was measured using a luminescence‐based assay in INS‐1 cells treated with IL‐1β (5 ng/mL) or IL‐1β + NaB (0.5 mM) for 1 h. Data are expressed relative to the IL‐β group. n = 7. (B and C) ER Ca²⁺ stores (B) and SOCE response (C) were measured in IL‐1β‐stressed INS‐1 cells treated with a pan‐HDAC inhibitor, trichostatin (TSA), in the absence or presence of ITSA, an HDAC activator via suppression of TSA, for 24 h. n = 8. (D and E) HDAC inhibitors with defined specificity, including SAHA, LMK235, MS275, and EX527, were employed in IL‐1β‐treated INS‐1 cells, and (D) ER Ca²⁺ stores and (E) SOCE response were measured at 24 h. n = 8. (F) Representative immunoblot image of protein level of inducible nitric synthase (iNOS) in IL‐1β‐treated INS‐1 cells at 24 h. (G) Quantification of normalized iNOS expression. n = 4. (H) ER Ca²⁺ stores and (I) SOCE response in INS‐1 cells co‐treated with IL‐1β and L‐N^G‐monomethyl arginine acetate (L‐NMMA), a nitric oxide synthase inhibitor, for 24 h to measure. n = 4. (J) ER Ca²⁺ stores and (K) SOCE responses in IL‐1β‐treated INS‐1 cells incubated with either NaB or AR420626 (AR) for 24 h. n = 8. The results are shown as the mean ± SEM; *p < .05, **p < .01, ***p < .001, ****p < .0001.

To test other HDAC inhibitors and provide insight into which HDACs might play a role in modulating SOCE, INS‐1 cells were treated with IL‐1β in the presence or absence of several HDAC inhibitors with defined specificity, including SAHA (pan‐HDAC inhibitor), LMK235 (inhibits HDAC 1, 2, 4, 5 and 6), MS275 (inhibits HDAC 1, 2, and 3), and EX527 (inhibits SIRT1). We found that SAHA, LMK235, and MS2755 rescued ER Ca²⁺ stores in IL‐1β‐treated INS‐1 cells, while EX527 did not (Figure 6D). Furthermore, we found that SAHA and LMK235 rescued the SOCE response in IL‐1β‐treated INS‐1 cells, similar to the results observed using NaB (Figure 6E). However, EX527 and MS275 did not modify SOCE (Figure 6E), suggesting that pan‐HDAC inhibition exerts a strong effect on rescuing SOCE through HDAC 4 and HDAC 5 under IL‐1β stress.

NO produced by iNOS plays a major role in diabetes‐associated β‐cell dysfunction, ³⁵ and NaB has been shown in other cell types to modulate iNOS. ³⁶ , ³⁷ Therefore, we determined the protein expression of iNOS in INS‐1 cells co‐treated with IL‐1β and NaB. As expected, IL‐1β significantly increased iNOS expression relative to the vehicle group, while NaB treatment decreased iNOS expression (Figure 6F,G). Consistent with this finding, co‐treatment with IL‐1β and L‐NMMA, an iNOS inhibitor, significantly increased ER Ca²⁺ stores and SOCE compared to IL‐1β treatment alone (Figure 6H,I), suggesting an important role of iNOS in NaB‐mediated benefits in IL‐1β‐treated INS‐1 cells.

NaB has been shown to bind free fatty acid receptors 2 and 3 (FFAR2/3, also known as G‐protein‐coupled receptor 43/41 (GPR43/41)) to modulate cAMP levels and protein kinase C activation. ³⁸ Next, we utilized the GPR41 agonist, AR420626, to mimic the activation of this GPCR in INS‐1 cells. However, we found that AR420626 was unable to rescue ER Ca²⁺ levels or SOCE responses in IL‐1 β‐treated INS‐1 cells (Figure 6J,K), indicating that GPCR signaling may play a minor role in modulation of Ca²⁺ signaling under inflammatory stress in β cells.

3.7. NaB rescues SOCE responses in cytokine‐stressed β cells via the AKT‐GSK‐3 pathway

Because beneficial effects of NaB have been linked with modulation of a variety of signaling pathways in other tissues, ³⁹ we assayed changes in phosphorylation that may contribute to the NaB‐mediated effects in the β cell. Isolated mouse islets were treated with cytokines (IL‐1β, IFN‐γ, and TNF‐α) in the presence or absence of NaB for 16 h, and a phosphorylation array was used to screen for regulation of select proteins in the MAPK signaling pathway. Exposure of islets to cytokines significantly decreased phosphorylation of AKT, CREB, ERK, GSK‐3α, GSK‐3β, JNK, MEK, and MKK6 when compared to that in the islets without any treatment (Figure 7). However, co‐treatment of NaB and cytokines upregulated the phosphorylation levels in ERK, GSK‐3α, GSK‐3β, and MEK when compared to that in the group receiving cytokines treatment (Figure 7). NaB treatment also showed a trend (p = 0.0788) to increase AKT phosphorylation in cytokines‐stressed mouse islets (Figure 7).

FIGURE 7.

NaB rescues SOCE in cytokine‐stressed β cells via the AKT‐GSK‐3 pathway. Signal transduction via the MAPK pathway was evaluated in WT C57BL/6J mouse islets after treatment with cytokines (5 ng/mL IL‐1β, 100 ng/mL IFN‐γ, and 10 ng/mL TNF‐α) in the presence or absence of NaB (0.5 mM) for 16 h. The islets were lysed and subjected to the phosphorylation array. The signals for each protein target were quantified and normalized to the positive controls. The levels of each protein target observed in the experimental group samples were expressed relative to the vehicle group. The values obtained in three independent experiments are represented by open circles, and the results are shown as the mean ± SEM; *p < .05, **p < .01, ***p < .001.

4. DISCUSSION

NaB is a simple four‐carbon SCFA ⁴⁰ that has been shown to have beneficial effects on pancreatic β‐cell function in mouse islets, INS‐1E cells, and human EndoC‐βH1 cells. ¹⁷ At present, the mechanisms underlying these effects have not been fully elucidated. Here, we found that NaB augments insulin secretion in human islets from donors with T2D and in INS‐1 β cells treated with pro‐inflammatory cytokines. Furthermore, co‐treatment of INS‐1 cells or mouse islets with cytokines and NaB partially rescued Ca²⁺ oscillations. We have shown previously that changes in Ca²⁺ oscillatory activity are associated with impaired SOCE, ²² leading us to hypothesize that NaB may exert beneficial effects in the β cell through SOCE modulation. Experiments to test this hypothesis revealed that NaB treatment led to the maintenance of Stim1 expression in cytokine‐treated β cells and showed that STIM1 restoration was associated with improved SOCE responses and cell viability. Mechanistic experiments linked NaB‐mediated restoration of SOCE with inhibition of HDAC activity and iNOS suppression, while a phosphorylation array highlighted the pleiotropic nature of NaB by showing that the compound also modulated AKT‐GSK‐3 signaling. Taken together, these results suggest a novel pathway through which NaB exerts protective metabolic effects in the diabetic β cell via modulation of Ca²⁺ homeostasis and SOCE.

The primary dietary source of NaB is host microbial fermentation of undigested dietary fiber, such as that found in legumes. ⁴⁰ Dysbiosis or alterations in the microbiome lead to reduced gut NaB production, which has been associated with adverse health outcomes. ⁴¹ Conversely, a large Mendelian randomization study showed that host‐genetic‐driven increases in gut production of NaB were associated with improved insulin secretion. ¹⁰ The anti‐inflammatory effects of NaB have been studied in multiple cell types and disease states, ¹⁴ and NaB has been shown to suppress NF‐κB signaling in intestinal biopsies and isolated lamina propria cells from individuals with Crohn's disease, ⁴² reduce pro‐inflammatory cytokine production in response to lipopolysaccharide (LPS) in endothelial cells, ⁴³ and decrease activation of the nucleotide‐binding oligomerization domain‐like receptors family pyrin domain containing 3 (NLRP3) inflammasome in adipocytes. ¹⁴ In addition, NaB may bind FFAR2/3, resulting in modulation of cAMP levels, altered Ca²⁺ concentrations, and protein kinase C activation. ³⁸ However, the NaB‐mediated effects through FFAR2/3 are still controversial, as opposite observations have also been reported in the literature. ⁴⁴

Of direct relevance to our study, NaB was shown to improve β‐cell function and glucose homeostasis in hyperglycemic rats treated with streptozotocin, where it was found to modulate p38/ERK MAPK signaling and apoptosis. ⁴⁵ Furthermore, NaB improved glycemia in streptozotocin and streptozotocin/high‐fat diet‐treated mice, and mechanistically these effects were linked with reduced serum IL‐1β levels, decreased NF‐kB signaling in the total pancreas, decreased β‐cell apoptosis, and reduced islet ER stress. ⁴⁶ More recently, NaB was shown to reduce the expression of IL‐1β‐induced inflammatory genes in the β cell ¹⁷ and protect β cells from chronic IL‐1β‐induced dysfunction via HDAC inhibition. ¹⁸

At the present time, NaB is under clinical investigation for co‐treatment along with probiotics for irritable bowel disease (IBD) (ClinicalTrials.gov, Identifier: NCT05013060) and pediatric obesity ⁴⁷ , ⁴⁸ and as an add‐on therapy to maintain remission in patients with ulcerative colitis. ⁴⁹ A clinical test of butyrate supplementation was carried out to determine whether butyrate was effective to improve inflammation, kidney parameters, serum metabolites, and gastrointestinal symptoms in individuals with T1D, albuminuria, and intestinal inflammation. ⁵⁰ Along these lines, strategies to increase NaB production in the gut include supplementation with the butyrylated form of high amylose maize starch (HAMS), which has been reported to lower body weight, reduce serum insulin, and decrease blood glucose levels in a rat model of T2D. ⁵¹ Butyrylated and acetylated forms of HAMS are currently being tested in a small pilot study to determine whether this intervention is able to modulate the gut microbiome and improve β‐cell function in youth with newly diagnosed T1D. ⁵² Similar studies are underway in individuals with T2D. ⁵³ , ⁵⁴

Given this background, we aimed to explore a link between NaB, Ca²⁺ signaling, and SOCE in the β cell. We and others have shown that impaired SOCE in the β cell results in a number of detrimental effects including reduced insulin secretion in response to glucose and fasiglifam, impaired glucose‐stimulated Ca²⁺ oscillations, decreased ER Ca²⁺ levels, increased ER stress, reduced cell viability, and decreased β‐cell identity. ²² , ²⁴ , ⁵⁵ Conversely, overexpression of STIM1 in human islets from donors with T2D improves insulin secretion, suggesting that restoration of STIM1 and SOCE may improve β‐cell health and function in the context of diabetes. ²² Interestingly, NaB has been reported to trigger apoptosis in colon cancer cells and nasopharyngeal cancer cells in a SOCE‐dependent manner. ³⁴ Here, we found that in addition to improving insulin secretion and glucose‐stimulated Ca²⁺ responses, NaB restored Stim1 expression in cytokine‐treated islets and β cells, whilst STIM1 restoration was linked with improved SOCE responses and cell viability.

In vitro experiments have shown that acetyl histone species accumulate in cultured mammalian cells treated with NaB. Subsequently, NaB was shown to suppress HDAC activity in vivo and in vitro to promote histone acetylation. ¹⁴ , ⁵⁶ The HDAC inhibition activity of NaB was confirmed in HEK293 cells, HMEC‐1 cells (endothelial cells), L6 myotubes (skeletal muscle cells), MAC‐Ts cells (bovine mammary epithelial cells), ³³ and more recently in β cells. ¹⁸ Notably, the suppression of HDACs by HDAC inhibitors has been linked to a number of beneficial metabolic effects, including protection against diet‐induced obesity in mice, potentiation of GSIS through enhanced GLP‐1R agonism, and protection of β cells from cytokine‐induced apoptosis. ⁵⁷ , ⁵⁸ Histone hyperacetylation induced by NaB occurs through class I and II HDACs, ³³ and NaB modifies gene expression through genetic mechanisms including regulation of gene chromatin accessibility, expression and/or modulation of signaling mechanisms that control transcription, or changes in levels of transcription factors and their complexes. ¹⁴

To identify potential mechanisms through which NaB modulates β‐cell Ca²⁺ signaling, we tested whether NaB mediates its effects via HDAC inhibition in our model system. Consistent with results observed by Pedersen et al., ¹⁸ we found that HDAC activity was reduced by NaB treatment in cytokine‐treated INS‐1 cells. We also treated INS‐1 cells with other pan‐HDAC inhibitors, such as TSA and SAHA, and found that they were able to similarly improve ER Ca²⁺ and SOCE, suggesting that both NaB and pan‐HDAC inhibitors rescue ER Ca²⁺ and SOCE under proinflammatory cytokine stress. In addition, the TSA‐specific HDAC activator ITSA1 reversed these beneficial effects, further confirming that HDAC inhibition plays a significant role in improved ER Ca²⁺ stores and SOCE responses in β cells.

Cytokines such as IL‐1β induce the expression of iNOS and the production of NO, which contributes to β‐cell dysfunction and damage. ⁵⁹ , ⁶⁰ In agreement with the previously reported anti‐inflammatory effects of NaB in other cell types, we found that NaB co‐treatment reduced iNOS expression in response to IL‐1β treatment in INS‐1 cells. Furthermore, treatment with the iNOS inhibitor, L‐NMMA, recapitulated improvements in ER Ca²⁺ and SOCE in IL‐1β‐stressed INS‐1 cells, indicating that iNOS/NO signaling also plays a role in the observed NaB‐mediated beneficial effects. ⁶⁰ , ⁶¹ Finally, NaB has been shown to increase the expression of GPR41, PI3K, and phosphorylated AKT in rat neuronal cells. ⁶² In liver cells, NaB is known to improve hepatic steatosis induced by a high‐fat and fiber‐deficient diet through activation of GPR43/GPR41‐mediated GPCR signaling. ⁶³ However, in our model system, SOCE responses were not rescued by co‐treatment with the GPR41 agonist AR420626 in IL‐1β‐treated INS‐1 cells, indicating that activation of GPR41 may play a minor role in NaB‐mediated SOCE improvements in β cells. Notably, our results are consistent with Pedersen et al who also failed to uncover the effects of NaB via FFAR2/3 agonism in mouse islets exposed to low‐level IL‐1β for 10 days. ¹⁸

Given the abundance of literature demonstrating various pleiotropic effects of NaB, we performed an exploratory phosphorylation array to identify signaling changes that may be modulated by NaB treatment in cytokine‐treated mouse islets. Our dataset showed that NaB exhibited a trend (p = 0.0788) to rescue the phosphorylation level of AKT in cytokine‐stressed islets, and the activation of AKT was reported previously in streptozotocin‐treated T1D rats. ³⁹ Therefore, this assay identified changes in AKT‐GSK3 signaling in response to NaB treatment in β cells. These findings were notable, as AKT signaling regulates diverse cellular functions, including apoptosis, metabolism, and cell proliferation. ⁶⁴ Our findings extend these observations and implicate that changes in GSK‐3 signaling may also contribute to NaB effects.

Taken together, our data indicate that NaB promotes β‐cell health and function under proinflammatory conditions through HDAC inhibition, iNOS suppression, STIM1‐mediated control of SOCE, and AKT‐mediated inhibition of GSK‐3. Further studies will determine whether NaB treatment could be a viable therapy for the treatment of diabetes, with implications for strategies that employ either direct augmentation or efforts to manipulate the gut microbiome.

AUTHOR CONTRIBUTIONS

C.L. contributed to the conception and design of the study, data analysis and interpretation, collection and assembly of data, and manuscript writing. T.K. directed the study conception and design, contributed to the acquisition, analysis, and interpretation of data, and provided critical revision of the manuscript. F.S. contributed to the conception and design of the study and revision of the manuscript. S.A.W. contributed to the data collection and revision of the manuscript. P.S. contributed to data collection and revision of the manuscript. W.W. contributed to data collection and analysis. G.C. contributed to data analysis. J.L. contributed to the conception of data analysis. M.S.R. contributed to data analysis. C.E.‐M. directed funding acquisition, study conception and design, participated in the collection and assembly of data, contributed to data analysis, directed manuscript writing, and gave final approval of the manuscript. C.E.‐M. is the guarantor of this work and, as such, has full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. All authors gave final approval of the manuscript.

DISCLOSURES

The authors declare no conflicts of interest.

Supporting information

Table S1.

Lee C‐C, Kono T, Syed F, et al. Sodium butyrate prevents cytokine‐induced β‐cell dysfunction through restoration of stromal interaction molecule 1 expression and activation of store‐operated calcium entry. The FASEB Journal. 2024;38:e23853. doi: 10.1096/fj.202302501RR

DATA AVAILABILITY STATEMENT

The data supporting the findings in this study are available in the Methods and/or Supplementary Material of this article. Further information and requests for resources and reagents should be directed to and will be fulfilled by the corresponding author. The study did not generate new unique reagents.