Pancreatic α-cell sodium-glucose cotransporter 1 (SGLT1) does not appear to contribute to hyperglucagonemia and glucose intolerance in diabetic mice

Abstract

Pancreatic α-cells secrete glucagon, a hormone that elevates blood glucose levels. In type 2 diabetes, high plasma glucagon levels are associated with hyperglycemia. However, the underlying mechanisms of increasing glucagon secretion remain unclear. We focused on the intrinsic regulatory mechanisms of glucagon secretion in α-cells, in particular sodium–glucose cotransporter 1 (SGLT1), which is involved in the early steps of glucose sensing. We previously demonstrated that SGLT1 is expressed in α-cells and is significantly upregulated in diabetic mice compared with non-diabetic mice. In isolated islets from diabetic mice, SGLT1 knockdown attenuated glucagon hypersecretion, and in αTC1 cells, SGLT-specific substrates promoted glucagon secretion by raising intracellular calcium. On the basis of these findings, we hypothesized that SGLT1 upregulation in α-cells under diabetic conditions impairs the suppression of glucagon secretion, thereby contributing to hyperglycemia. However, a previous study showed that systemic SGLT1 knockout (KO) mice exhibit a higher proportion of α-cells in the islets and atypically high plasma glucagon levels. To clarify the roles of SGLT1 specifically in α-cells, we generated α-cell–specific SGLT1 KO mice using a tamoxifen-inducible Cre-loxP system and analyzed these mice fed a high-fat, high-sucrose diet. The results clearly showed that, inconsistent with the results from the systemic SGLT1 KO mice, SGLT1 deficiency specifically in α-cells did not affect glucagon secretion, glucose tolerance, or α-cell proportion in the islets under diabetic conditions. Thus, though SGLT1 is upregulated in diabetic α-cells, this does not appear to contribute to hyperglucagonemia and impaired glucose tolerance in diabetic mice.

Graphical Abstract

Introduction

Dysregulation of appropriate glucagon secretion from pancreatic α-cells in response to blood glucose levels is considered one of the contributing factors to hyperglycemia in type 2 diabetes [1]. Glucagon secretion is normally inhibited by rising blood glucose levels and/or paracrine signals derived from neighboring β- and δ-cells [2-4]. Accordingly, two major mechanisms have been proposed to underlie the dysregulation of glucagon suppression in type 2 diabetes. One is an “indirect” mechanism, in which reduced insulin or somatostatin secretion leads to insufficient paracrine inhibition in α-cells. The second is a “direct” mechanism, in which the intrinsic ability of α-cells to sense blood glucose is impaired and thereby suppression of glucagon secretion is hindered. However, the latter mechanism’s molecular details in particular have been largely unknown. Therefore, in order to elucidate the “direct” mechanism at the molecular level, we focused on the mechanism of glucose sensing in α-cells.

In pancreatic α-cells, it has been known that glucose transporter 1 (GLUT1) is expressed and regulates glucagon secretion autonomously by sensing variable blood glucose levels [5]. Indeed, glucagon secretion is suppressed by glucokinase activity in α-cells, which catalyzes the first phosphorylation step of glycolysis [6]. Importantly, we previously reported that in addition to GLUT1, sodium–glucose cotransporter 1 (SGLT1) is also expressed in pancreatic α-cells [7]. Our findings of SGLT1 expression in pancreatic α-cells have been subsequently confirmed in both mouse and human samples by several groups [8, 9].

While GLUT1 transports glucose according to the concentration gradient, SGLT1 mediates glucose uptake via the Na⁺ gradient across the plasma membrane, enabling glucose influx even under low extracellular glucose conditions. SGLT1 is abundantly expressed in epithelial cells of the small intestine and kidney, where it plays a major role in glucose reabsorption [10]. In intestinal L cells, glucose and Na⁺ uptake via SGLT1 triggers the secretion of glucagon-like peptide 1 (GLP-1) [11]. Importantly, we previously showed that SGLT-specific substrate α-methyl D-glucoside (αMG) raises the intracellular calcium concentration and stimulates glucagon secretion in αTC-1 cells [7]. In addition, SGLT-specific inhibitor and siRNA-induced knockdown of SGLT1 suppresses glucagon secretion in isolated islets. Moreover, SGLT1 expression levels in islet cells are higher in mice fed a high-fat, high-sucrose diet (HFHSD) and in db/db mice, in which glucagon secretion is higher than it is in mice fed normal chow (NC). Based on our previous findings, we hypothesized that SGLT1 upregulation in pancreatic α-cells under diabetic conditions impairs the suppression of glucagon secretion, thereby leading to hyperglycemia. However, previous reports have demonstrated that conventional (i.e., systemic) SGLT1 knockout mice exhibit a higher proportion of α-cells in pancreatic islets and atypically elevated glucagon secretion [12]. The role of SGLT1 in α-cells remains difficult to interpret because these mice have reduced blood glucose and plasma incretin levels due to ablation of intestinal SGLT1 and subsequent impaired glucose absorption.

To elucidate the roles of SGLT1 specifically in pancreatic α-cells under diabetic conditions, we generated α-cell–specific SGLT1 knockout (αSGLT1KO) mice using a tamoxifen-inducible Cre/loxP system and analyzed these mice under HFHSD-fed conditions. If our hypothesis is correct, αSGLT1KO mice would exhibit attenuated glucagon hypersecretion and improved glucose tolerance under HFHSD-fed conditions in comparison with control mice. Here, we report metabolic phenotypes, including glucagon secretion and glucose tolerance, as well as pancreatic islet morphology, in αSGLT1KO mice under diabetic conditions.

Materials and Methods

Animal care

All experimental procedures were performed according to the Guide for the Care and Use of Laboratory Animals of the Science Council of Japan. The protocol for the animal experiments conducted in this study was formally approved by the Animal Care and Ethics Committee of Gunma University (approval number: 23-042). Mice were housed in a specific pathogen–free barrier facility, maintained under a 12-hour light/dark cycle, and were randomly assigned to consume normal chow (#CE-2, CLEA, Japan) or HFHSD (Oriental Yeast, Japan) containing 20% sucrose. The total energy in the HFHSD was derived from protein (17.2%), fat (54.5%), and carbohydrate (28.3%). HFHSD was provided from 8 weeks of age. Mice were fed HFHSD for more than 10 weeks, then given tamoxifen by injection. Water was provided ad libitum. None of the animals included in the data analyses displayed any health impairments.

Generation of Slc5a1 gene-floxed mice

Slc5a1^loxP/loxP mice were generated by the sequential electroporation method that we reported previously [13]. Sequences of target CRISPR RNAs (crRNAs) flanking exons 3 to 5 of the Slc5a1 gene, and corresponding donor single-stranded oligodeoxynucleotides (ssODNs) including loxP are shown in Table 1. Fertilized eggs were isolated from superovulated B6D2F1 female mice, 21 h post administration of human chorionic gonadotropin (hCG). The first electroporation to insert the left loxP into intron 2 was conducted at the 1-cell stage (24–26 h post hCG), and the second electroporation to insert the right loxP to intron 5 was conducted at the 2-cell stage (42–44 h post hCG). The sequentially electroporated embryos were then transferred to the oviducts of pseudopregnant ICR females. The floxed alleles of the obtained mice were confirmed by sequencing analysis. For PCR-based confirmation, the primers used for each loxP insertion site were as follows: for the left loxP insertion site, the forward primer was 5'-AGCTCAAGCAAATTCCTGCT-3', the reverse primer was 5'-TTTTCTATACCATTTCAAGTCTCTCC-3', and a 280-bp band is detected. For the right loxP insertion site, the forward primer was 5'-GTTCTGCCTCAGGCTCTGTT-3', the reverse primer was 5'-AAGACATCTGCCCAAAGCAG-3', and a 271-bp band is detected.

Table 1. Sequences of target crRNAs flanking exons 3 to 5 of the Slc5a1 gene and the corresponding donor single-stranded oligodeoxynucleotides (ssODNs) including loxP sites.

	Target	ssODN
Left loxP	AGACTGCTGAGTAGGCCCAT	AGCACAGCTCTACTGTGGTCATAGACCCTCTCCCGGGTCCTCTCTACTGCCAGCTCCCATATAACTTCGTATAGCATACATTATACGAAGTTATATCGATGGGCCTACTCAGCAGTCTTCCAGTAAGCTCCATGCCTGGTGACAAATCCCAGTGAGCTCG
Right loxP	CTGTATCCATGGAACTCTTA	ACCCAAGAGCATCATTCTTATCCCTCTGAGTGGTGAGCTCACCTGTATCCATGGAACTCTATAACTTCGTATAGCATACATTATACGAAGTTATGGTACCTAGGGAGCAATTAAACATGGACCCTGATGGGAAACCTCTTAGAGCAACAGAGCCTGAGGC

Generation of α-cell–specific SGLT1 knockout mice and study design

We crossed Slc5a1^loxP/loxP mice with Glucagon-Cre^ERT2 mice [14] to generate Slc5a1^loxP/loxP; Glucagon-Cre^ERT2 mice. To induce a diabetic phenotype, mice were fed a HFHSD beginning at 8 weeks of age. Tamoxifen was administered intraperitoneally at a dose of 100 μg/g body weight per day for three consecutive days between 18 and 20 weeks of age to induce Cre-mediated recombination. These mice are referred to as HFHSD-fed αSGLT1KO mice. Phenotypic analyses were conducted at 23 to 30 weeks of age. In all experiments, HFHSD-fed, tamoxifen-injected, Cre-negative Slc5a1^loxP/loxP littermates were used as controls. Only male mice were used in this study.

Confirmation of exon 3–5 deletion by genomic PCR

Genomic DNA was extracted from tails, ileum, and pancreatic islets of HFHSD-fed αSGLT1KO mice using phenol–chloroform extraction followed by isopropanol precipitation. PCR was performed using these DNA samples as templates with the following primers: P1: 5'-ATGGAAAATGTGATTGTAGCAAGCAGT-3', P2: 5'-GTTCTGCCTCAGGCTCTGTT-3', and P3: 5'-GCCAAGAGCTGACAGTCCTTCTATGTA-3'. Product sizes were 1,143 bp (deletion band) and 589 bp (floxed band), respectively. PCR products were separated on 2% agarose gels, stained with ethidium bromide, and visualized using the Printgraph Classic imaging system (#WSE-5400, ATTO Corporation, Japan).

Mouse genotyping

Mouse tail DNA was used for detecting the presence of the Gcg-Cre^ERT2 and Slc5a1^loxP/loxP alleles. PCR primers and reactions used for genotyping studies are listed in Table 2.

Table 2. PCR primers used for genotyping Gcg-Cre^ERT2 and Slc5a1^loxP/loxP alleles.

Target	Forward Primer	Reverse Primer	Product size (bp)
Gcg-CreERT2	1) AGAGGGTTTCCCTGCCACA 2) TGCCAGTCACTTGGGATG	GGTTACCAGGTGGTCATGTCT	Wt: 302 KI: 100
Slc5a1loxP/loxP Left loxP	AGCTCAAGCAAATTCCTGCT	TTTTCTATACCATTTCAAGTCTCTCC	Wt: 240 KI: 280
Slc5a1loxP/loxP Right loxP	GTTCTGCCTCAGGCTCTGTT	AAGACATCTGCCCAAAGCAG	Wt: 231 KI: 271

Immunohistochemistry

Immunohistochemistry was performed as previously described [15], with minor modifications. Briefly, mice were deeply anesthetized and perfused transcardially with ice-cold 0.05 M phosphate-buffered saline (PBS, pH 7.4), followed by fixation with 4% paraformaldehyde (PFA). The pancreas was excised, post-fixed in 4% PFA, and embedded in OCT compound, after which 6 μm thick frozen sections were cut using a cryostat. Sections were permeabilized and incubated in Tris-buffered saline (TBS) containing 5% normal donkey serum (NDS) for 1 h at room temperature to block nonspecific binding. Rat monoclonal anti-glucagon antibody (1:1,000, #52A1A, Immuno-Biological Laboratories, Japan) and guinea pig polyclonal anti-insulin antibody (1:400, #A0564, Dako, USA) were diluted in TBS/5% NDS and applied overnight at 4°C. After washing with TBS, sections were incubated with appropriate fluorescent-conjugated secondary antibodies (1:200, Jackson ImmunoResearch, USA) for 1 h at room temperature. Fluorescence images were captured using a confocal laser-scanning microscope (#FV3000, Olympus, Japan).

Mouse islet isolation and culture

Islets were isolated from mice by collagenase digestion as described previously [16]. Mice were anesthetized and sacrificed by cervical dislocation. Collagenase (#C7657, Sigma) solution was injected into the pancreatic duct, and the pancreas was removed and digested with collagenase at 37°C for 17 min. The digest was then vigorously shaken and transferred to ice-cold M-199 medium (#M0393, Sigma) supplemented with NaHCO₃. Islets were hand-picked under a microscope and cultured overnight in RPMI-1640 medium (#189-02025, Fujifilm Wako Chemicals, Japan) containing 10% heat-inactivated fetal bovine serum (FBS) and cultures at 37°C, 5% CO₂.

For the islet secretion assay, 15 size-matched islets were picked and pre-incubated in Krebs-Ringer bicarbonate buffer (KRB) containing 7 mM glucose for 30 min at 37 °C. The islets were then incubated for 120 minutes in KRB supplemented with 1 mM glucose, 20 mM glucose, or 60 mM KCl. After incubation, islets were collected and lysed in acid ethanol (75% EtOH; 0.18 N HCl) to determine total hormone content. Glucagon secretion levels were normalized to total intracellular glucagon content.

Measurements of plasma concentration of hormones and metabolic analyses

Blood samples were collected using heparinized syringes and placed into tubes with a mixture of aprotinin and EDTA, and plasma glucagon and insulin levels were measured with a glucagon sandwich ELISA (#10-1281-01, Mercodia, Sweden), a GLP-1(9-36/37) ELISA (#27788, Immuno-Biological Laboratories, Japan) and an insulin ELISA (#27705, Immuno-Biological Laboratories, Japan), respectively.

Insulin tolerance tests were conducted after a 6-hour fast, and glucose tolerance tests were performed following overnight fasting. Insulin (0.75 U/kg body weight) was injected intraperitoneally. Glucose (2 g/kg body weight) was administered orally. Blood glucose was measured using Glutest Ai (#GT-1840, Sanwa Kagaku Kenkyusho, Japan).

Statistical Analysis

Data are presented as mean ± SEM (standard error of the mean). The significance of differences was assessed with Welch’s t-test. Data were analyzed using commercial software (Prism 8, GraphPad; Excel, Microsoft). A p-value <0.05 was considered statistically significant. The statistical tests and sample sizes used for the individual experiments are described in the figure or figure legends.

Results

Sodium-binding sites predicted to be essential for transporter activity exist in exon 3 of Slc5a1 gene

Putative sodium-binding sites, referred to as the Na⁺ sites, have been proposed in SGLT1 (Fig. 1). Among these, the Na2 site is conserved across many members of the amino acid-polyamine-organocation (APC) superfamily [17] and is also present in another representative sodium–glucose co-transporter, SGLT2 [18]. It is proposed that extracellular sodium first binds to the Na2 site, triggering the opening of the external gate and allowing external glucose to flow into the transporter [19]. In addition, a distinct sodium-binding site known as the Na3 site has been identified in SiaT, a bacterial sodium/sialic acid symporter [20]. This site has been shown to be conserved within SGLT1 on the basis of sequence alignment and has been suggested to regulate transporter activity (Fig. 1).

Fig. 1. Effective target regions for disrupting SGLT1 transporter activity.

Amino acid sequences of mouse and human SGLT1 and SGLT2 were aligned using Clustal Omega [27] and further processed manually. Predicted transmembrane helices (TMs) are indicated by gray bars. Amino acid residues involved in the Na2 site are highlighted with blue boxes, and the residue corresponding to the Na3 site is shown in a red box. Residues mentioned in the text (e.g., H83, S77) are indicated in bold. Regions encoded by exon 3, 4, and 5 of mouse SGLT1 gene are highlighted in purple, yellow, and green, respectively. Amino acids shown in red indicate residues that are conserved between mouse SGLT1 and human SGLT1 or mouse SGLT2.

Among the transmembrane helices of SGLT1, transmembrane helix 1 (TM1) contributes to the formation of both the Na2 and Na3 sites (Fig. 1). This suggests that TM1 is a critical region for transporter function. Indeed, substitution of the histidine 83 (H83) or serine 77 (S77), which is thought to be part of the Na3 site in TM1, abolishes substrate uptake by SGLT1 [17, 21]. Therefore, to ensure efficient loss of SGLT1 transport activity in α-cells, we targeted exon 3 of the Slc5a1 gene, which encodes TM1, for conditional deletion using the Cre/loxP system.

Generation of pancreatic α-cell–specific SGLT1 knockout (αSGLT1KO) mice

As exon 3 is likely to encode an essential region for the transporter activity of SGLT1, we targeted exon 3 of SLC5A1 gene to ensure effective ablation of SGLT1 activity in pancreatic α-cells. Using CRISPR-Cas9 genome editing, we inserted loxP sites flanking exons 3 to 5 of the SLC5A1 gene (Fig. 2A). We then crossed Slc5a1^loxP/loxP mice with Glucagon-Cre^ERT2 mice [14] to generate Slc5a1^loxP/loxP; Glucagon-Cre^ERT2 mice. After treatment with tamoxifen, Cre recombinase is activated only in pancreatic α-cells and intestinal L-cells, where the proglucagon promoter is active, leading to excision of the loxP-flanked region of SLC5A1 gene in these cells. Because Cre recombinase activity in intestinal L-cells is transient due to their rapid turnover and subsequent replacement [14], SGLT1 should be specifically deleted in pancreatic α-cells in Slc5a1^lox/lox; Glucagon-Cre^ERT2 mice 4 weeks after tamoxifen treatment. We then confirmed the specific deletion of the region spanning exon 3 and 5 of SLC5A1 gene by deletion PCR using the primers indicated by P1 to P3 in Fig. 2A. The results clearly demonstrate that the deletion band using P1 and P3 was detected only in the islets, but not in the tails and ileums, indicating that the region spanning exon 3 to 5 of the Slc5a1 gene is specifically deleted in α-cells of these mice (Fig. 2B). Therefore, these mice are hereafter referred to as αSGLT1KO mice.

Fig. 2. Generation of αSGLT1KO mice.

(A) Schematic diagram illustrating the strategy used to generate αSGLT1KO mice and control mice. LoxP sites were inserted into the Slc5a1 locus, and floxed mice were crossed with Gcg-Cre^ERT2 mice. Subsequent tamoxifen administration induced Cre recombinase activity in the proglucagon promoter-activated cells, resulting in the deletion of exons 3 to 5 in pancreatic α-cells, which we termed αSGLT1KO mice, whereas these exons remained intact in control mice. αSGLT1KO mice and control mice were littermates obtained from the same breeding pairs. P1, P2, and P3 indicate the positions of primer 1, primer 2, and primer 3, respectively. (B) Genomic PCR was performed using DNA extracted from islets, tail, and ileum. PCR products amplified with P2 and P3 (floxed band) were detected in all tissues, whereas the deletion band amplified with P1 and P3 was detected only in the islets of αSGLT1KO mice. C: control mice; KO: αSGLT1KO mice; N: negative control (no DNA template). (C) Pancreatic sections from Glucagon-Cre^ERT2; Rosa26^tdTomato mice treated with tamoxifen were subjected to immunohistochemistry using a rat monoclonal anti-glucagon antibody. Native tdTomato fluorescence is shown in red, glucagon staining in green, and nuclei (DAPI) in blue. Merged images (yellow) indicate colocalization of tdTomato and glucagon. Most glucagon-positive cells were also tdTomato-positive in tamoxifen-treated Glucagon-Cre^ERT2; Rosa26^tdTomato mice. Scale bars = 100 μm. All analyses were performed in a blinded manner.

We next verified the deletion efficiency in αSGLT1KO mice by crossing Glucagon-Cre^ERT2 mice with Rosa26tdTomato mice [22]. In the tamoxifen-treated Glucagon-Cre^ERT2; Rosa26^tdTomato mice, tdTomato is visualized only in the Cre recombinase–expressing cells. Therefore, we conducted immunostaining with anti-glucagon antibodies followed by FITC-labeled secondary antibodies of the pancreatic sections of tamoxifen-treated Glucagon-Cre^ERT2; Rosa26^tdTomato mice and then counted the numbers of glucagon-positive cells and tdTomato-positive cells. As shown in Fig. 2C, tdTomato signals were detected in 86.6% (1,266/1,462 cells) of glucagon-positive α-cells, indicating that SGLT1 should be deleted in nearly 90% of α-cells in αSGLT1KO mice.

Unchanged islet morphology, α-cell mass and glucagon secretion from isolated islets in HFHSD-fed αSGLT1KO mice

We previously reported that the expression levels of SGLT1 are elevated in αTC1 cells cultured in high glucose and in the islets isolated from HFHSD-fed mice compared with those in low glucose and islets isolated from normal chow-fed mice, respectively [7]. Therefore, in order to investigate the pathophysiological roles of SGLT1 under diabetic conditions, Slc5a1^lox/lox; Glucagon-Cre^ERT2 mice were fed with HFHSD beginning at 8 weeks of age. After 10–12 weeks feeding with HFHSD, we administered tamoxifen to delete SGLT1 specifically in α-cells to generate HFHSD-fed αSGLT1KO mice. After 8–10 weeks, mice were sacrificed and their pancreatic sections were immunostained with anti-glucagon antibodies and anti-insulin antibodies. In contrast to the previous report showing an increased proportion of α-cells and decreased proportion of β-cells in the systemic SGLT1 knockout mice [12], islet morphology was unaltered in HFHSD-fed αSGLT1KO mice compared with that in the HFHSD-fed control mice (Fig. 3A). In addition, neither α-cell mass, assessed by measuring the area of glucagon-positive cells, nor β-cell mass, assessed by measuring the area of insulin-positive cells, differed between HFHSD-fed αSGLT1KO mice and HFHSD-fed control mice (Fig. 3B, C). The α/β ratio was also comparable between these two groups (Fig. 3D). Because we previously showed that an SGLT-specific substrate stimulates glucagon secretion and that the expression level of SGLT1 is elevated in the islets of diabetic mice [7], we hypothesized that elevated SGLT1 expression in α-cells accounts for increased glucagon secretion in diabetic mice. If so, glucagon secretion should be decreased in the islets isolated from HFHSD-fed αSGLT1KO mice compared with secretion in islets from the HFHSD-fed control mice. However, glucagon secretion from the islets in response to low (1 mM) or high (20 mM) glucose, or to KCl (60 mM) did not differ between HFHSD-fed αSGLT1KO mice and HFHSD-fed control mice (Fig. 3E).

Fig. 3. Islet morphology and glucagon secretion from isolated islets in HFHSD-fed αSGLT1KO mice.

(A) Immunohistochemical staining of pancreatic sections from HFHSD-fed αSGLT1KO mice and control mice at 28 to 30 weeks of age. Glucagon, insulin, and nuclei (DAPI) are shown in red, green, and blue, respectively. Scale bars = 100 μm. (B–D) Quantitative analysis of islet morphology. α cell area (B), β cell area (C), and the ratio of α cell area to β cell area (D). Analyses were performed in a blinded manner; n = 3 per group. (E) Glucagon secretion from isolated islets of αSGLT1KO mice and control mice treated with 1 mM glucose (G1), 20 mM glucose (G20), or 60 mM KCl (KCL) for 120 min. Sample sizes: (G1) n = 3–4 per group, (G20) n = 5–6 per group, (KCl) n = 3–4 per group. Data are presented as means ± SEM. ns indicates not significant, as determined by Welch’s t-test.

Unchanged glucagon secretion, glucose tolerance, and insulin sensitivity in HFHSD-fed αSGLT1KO mice

We monitored the body weights and random-fed blood glucose levels in HFHSD-fed αSGLT1KO mice before and after tamoxifen administration. As shown in Fig. 4A, B, no significant differences were observed in the body weights and blood glucose levels between HFHSD-fed αSGLT1KO mice and HFHSD-fed control mice even after tamoxifen administration. We previously reported that plasma glucagon levels at fasting and 30 min after glucose load were significantly higher in type 2 diabetes patients than in healthy individuals [23]. If those observations could be accounted for by the increased expression of SGLT1 in pancreatic α-cells under diabetic conditions, plasma glucagon levels should be lower and thereby glucose tolerance should be better in HFHSD-fed αSGLT1KO mice than in HFHSD-fed control mice. To test this hypothesis, we performed an oral glucose tolerance test (OGTT). However, comparable glucose tolerances were observed between HFHSD-fed αSGLT1KO mice and HFHSD-fed control mice (Fig. 4C). Furthermore, inconsistent with our hypothesis, there were no significant differences in plasma glucagon levels between HFHSD-fed αSGLT1KO mice and HFHSD-fed control mice before and 10 min after glucose loading (Fig. 4D). In addition, neither plasma insulin levels nor GLP-1 levels differed between these mice (Fig. 4E, F). We also performed an insulin tolerance test (ITT) in these mice and found that insulin sensitivity did not differ between HFHSD-fed αSGLT1KO mice and HFHSD-fed control mice (Fig. 4G). Taken together, unexpectedly, the results indicate that the ability to secrete glucagon was not affected by SGLT1 ablation in pancreatic α-cells under HFHSD-fed conditions, and consequently blood glucose homeostasis was not altered in αSGLT1KO mice.

Fig. 4. Body weight, random-fed blood glucose, glucose tolerance, plasma glucagon, insulin, GLP-1 levels, and insulin sensitivity in HFHSD-fed αSGLT1KO mice.

(A and B) Body weights (A) and random-fed blood glucose levels (B) of αSGLT1KO mice and control mice during the experimental period. Arrows indicate the timing of tamoxifen injection. (C–F) HFHSD-fed αSGLT1KO mice and HFHSD-fed control mice were subjected to oral glucose tolerance tests (OGTTs). Blood glucose levels during OGTT (C) and plasma glucagon (D), insulin (E), and GLP-1 (F) levels at 0 min and 10 min after glucose loading are shown. (G) Blood glucose levels during insulin tolerance testing (ITT). Sample sizes: (A and B) n = 6–8 per group, (C and G) n = 8–11 per group, (D) n = 14–19 per group, (E) n = 6–9 per group, (F) n = 7–15 per group. Data are presented as means ± SEM. ns indicates not significant, as determined by Welch’s t-test.

Discussion

In this study, we hypothesized that α-cell SGLT1, whose expression is elevated in diabetic conditions, hinders the suppression of glucagon secretion, leading to an increase in blood glucose levels. To test this hypothesis, we generated tamoxifen-inducible αSGLT1KO mice, which were made diabetic and obese by feeding with HFHSD and then treated with tamoxifen to specifically delete SGLT1 in pancreatic α-cells. Through phenotypic validation of these mice, several insights regarding SGLT1 in pancreatic α-cells were obtained. First, knockout of SGLT1 specifically in α-cells under HFHSD-fed conditions did not alter islet morphology or glucagon secretion from isolated islets. Second, α-cell–specific SGLT1 knockout mice showed no changes in body weight, random-fed blood glucose levels, glucose tolerance, or insulin sensitivity. Third, no significant changes in plasma glucagon, insulin, or GLP-1 levels were observed in α-cell-specific SGLT1 knockout mice before and 10 minutes after glucose loading. On the basis of these findings, upregulation of SGLT1 in α-cells under diabetic conditions does not appear to contribute to the impaired suppression of glucagon secretion and subsequent hyperglycemia commonly observed in diabetes. We previously reported that dapagliflozin (a specific inhibitor of SGLT2) increases glucagon secretion, whereas canagliflozin (a less specific inhibitor of SGLT2) does not increase glucagon secretion via SGLT1 inhibition [7]. Nonetheless, clinical reports so far have not demonstrated dramatic clinical differences between these two SGLT2 inhibitors. This study, showing that SGLT1 in pancreatic α-cells does not appear to contribute to hyperglucagonemia and glucose intolerance in diabetic mice, may explain why there are no significant differences in clinical characteristics between SGLT2 inhibitors despite their different specificities for SGLT2.

In our previous study, we demonstrated that the addition of SGLT inhibitor or SGLT1 knockdown reduced glucagon secretion in the islets isolated from HFHSD-fed wild type mice [7]. However, the present study did not reproduce this finding. In our earlier experiments, SGLT1-expressing β and δ-cells [8, 24] were also affected by SGLT1 inhibitor. This may have altered insulin and somatostatin secretion, thereby reducing glucagon secretion through a paracrine mechanism. By contrast, the αSGLT1KO in the present study had no influence on β-cells and δ-cells, allowing us to evaluate the primary effect of SGLT1 deficiency in α-cells on glucagon secretion. Another possible explanation for the unaltered glucagon secretion in αSGLT1KO mice is the incomplete efficiency of Cre-mediated recombination. Consistent with the previous report [14], approximately 90% of the α-cells in the Glucagon-Cre^ERT2 mice expressed Cre recombinase (Fig. 2C). Therefore, the remaining approximately 10% of α-cells maintained SGLT1 expression, which may have been sufficient to maintain plasma glucagon levels [25]. This represents an inherent limitation of the Cre/loxP system used in this study. A third possible explanation is that glucagon secretion may have been influenced by reduced expression of GLUT1, another glucose transporter expressed in α-cells [7, 10]. It is known that GLUT1-mediated glucose uptake in α-cells activates glucokinase, leading to a decrease in glucagon secretion [6, 26]. In αSGLT1KO mice, the effect of SGLT1 deletion in α-cells may have been compensated for by the decreased GLUT1 expression in α-cells, and plasma glucagon levels may have been maintained. On the other hand, whether SGLT2 is expressed in α-cells has been debated, but accumulating evidence generally indicates that SGLT2 is not expressed in α-cells [7, 8, 24]. Therefore, it is unlikely that SGLT1 deletion caused a compensatory change in SGLT2 in α-cells, leading to the maintenance of glucagon secretion.

Our results in αSGLT1KO mice differ significantly from those in systemic SGLT1 KO mice [12]. There are two important points to note. First, the change in islet morphology observed in systemic SGLT1 KO mice—the increased proportion of α-cells—was not observed in αSGLT1KO mice. In contrast to systemic SGLT1 KO mice, in which SGLT1 is ablated from the fetal stage, αSGLT1KO mice lack SGLT1 in α-cells only in adulthood because we injected tamoxifen into adult mice. Therefore, our model could not evaluate the roles of SGLT1 on islet morphology during pancreas development. In other words, our model allowed us to evaluate the effects of SGLT1 in mature α-cells under diabetic conditions without considering changes in islet morphology during pancreas development. Second, the atypical increase in plasma glucagon levels during OGTT observed in systemic SGLT1 KO mice was not reproduced in αSGLT1KO mice. Because systemic SGLT1 KO mice lack SGLT1 in various cell types throughout the body, including proximal renal tubules, intestinal L-cells, and pancreatic islet cells other than α-cells (β- and δ-cells), the roles of SGLT1 specifically in α-cells could not be evaluated. In addition, the systemic SGLT1 KO mice used in the previous report were fed a glucose-deficient, fat-enriched diet because they lack SGLT1 in intestinal cells. Therefore, the influence of the special diet on the proportion of α-cells and glucagon secretion could not be ruled out. In the present study, we used αSGLT1KO mice to analyze the effects of selective SGLT1 deletion in pancreatic α-cells, enabling us to specifically evaluate the role of SGLT1 in α-cells in the diabetic conditions. It should be noted, however, that the efficiency of Cre/loxP-mediated recombination in our αSGLT1KO mice was approximately 90%. Therefore, we cannot completely exclude the possible influence of the remaining SGLT1-expressing α-cells (~10%). Future studies using a more efficient knockout system will be necessary to fully elucidate the role of SGLT1 in pancreatic α-cells.

In conclusion, α-cell–specific SGLT1 deletion in HFHSD fed mice did not affect glucagon secretion, glucose tolerance, or islet morphology. These findings suggest that increased SGLT1 expression in α-cells in the diabetic state does not appear to contribute to the impaired suppression of glucagon secretion and impaired glucose tolerance commonly observed in diabetes (see the Graphical Abstract).

Graphical Abstract.

Author Contributions

Y. Ikeuchi: Conceptualization, Methodology, Formal analysis, Investigation, Data curation, Writing – original draft, Visualization, O. Kikuchi and M. Kobayashi: Methodology, Formal analysis, Investigation, Writing – review & editing, Y. Tabei and H. Hashimoto: Writing – review & editing, R. Kobayashi, T. Horii and I. Hatada: Methodology, Formal analysis, Investigation, Resources, Writing – review & editing, T. Miyatsuka: Resources, Writing – review & editing, T. Kitamura: Conceptualization, Project administration, Writing – review & editing, Supervision.

Data Availability

Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Funding

This work was supported by Japan Society for the Promotion of Science (JSPS) KAKENHI (Grant Number 22H03125) and Platform Project for Supporting Drug Discovery and Life Science Research (Basis for Supporting Innovative Drug Discovery and Life Science Research (BINDS)) from AMED under Grant Number JP21am0101120 to I.H..

Disclosure

None of the authors have any conflicts of interest associated with this study. Takeshi Miyatsuka and Tadahiro Kitamura are members of Endocrine Journal’s Editorial Board.