Genetic locus responsible for diabetic phenotype in the insulin hyposecretion (ihs) mouse

Diabetic animal models have made significant contributions to understanding the etiology of diabetes and to the development of new medications. Our research group recently developed a novel diabetic mouse strain, the insulin hyposecretion (ihs)mouse. The strain involves neither obesity nor insulitis but exhibits notable pancreatic β-cell dysfunction, distinguishing it from other well-characterized animal models. In ihs mice, severe impairment of insulin secretion from pancreas has been elicited by glucose or potassium chloride stimulation. To clarify the genetic basis of impaired insulin secretion, beginning with identifying the causative gene, genetic linkage analysis was performed using [(C57BL/6 × ihs) F1 × ihs] backcross progeny. Genetic linkage analysis and quantitative trait loci analysis for blood glucose after oral glucose loading indicated that a recessively acting locus responsible for impaired glucose tolerance was mapped to a 14.9-Mb region of chromosome 18 between D18Mit233 and D18Mit235 (the ihs locus). To confirm the gene responsible for the ihs locus, a congenic strain harboring the ihs locus on the C57BL/6 genetic background was developed. Phenotypic analysis of B6.ihs-(D18Mit233-D18Mit235) mice showed significant glucose tolerance impairment and markedly lower plasma insulin levels during an oral glucose tolerance test. Whole-genome sequencing and Sanger sequencing analyses on the ihs genome detected two ihs-specific variants changing amino acids within the ihs locus; both variants in Slc25a46 and Tcerg1 were predicted to disrupt the protein function. Based on information regarding gene functions involving diabetes mellitus and insulin secretion, reverse-transcription quantitative polymerase chain reaction analysis revealed that the relative abundance of Reep2 and Sil1 transcripts from ihs islets was significantly decreased whereas that of Syt4 transcripts were significantly increased compared with those of control C57BL/6 mice. Thus, Slc25a46, Tcerg1, Syt4, Reep2 and Sil1 are potential candidate genes for the ihs locus. This will be the focus of future studies in both mice and humans.

Diabetic animal models have made significant contributions to understanding the etiology of diabetes and to the development of new medications. Our research group recently developed a novel diabetic mouse strain, the insulin hyposecretion (ihs)mouse. The strain involves neither obesity nor insulitis but exhibits notable pancreatic β-cell dysfunction, distinguishing it from other well-characterized animal models. In ihs mice, severe impairment of insulin secretion from pancreas has been elicited by glucose or potassium chloride stimulation. To clarify the genetic basis of impaired insulin secretion, beginning with identifying the causative gene, genetic linkage analysis was performed using [(C57BL/6 × ihs) F 1 × ihs] backcross progeny. Genetic linkage analysis and quantitative trait loci analysis for blood glucose after oral glucose loading indicated that a recessively acting locus responsible for impaired glucose tolerance was mapped to a 14.9-Mb region of chromosome 18 between D18Mit233 and D18Mit235 (the ihs locus). To confirm the gene responsible for the ihs locus, a congenic strain harboring the ihs locus on the C57BL/6 genetic background was developed. Phenotypic analysis of B6.ihs-(D18Mit233-D18Mit235) mice showed significant glucose tolerance impairment and markedly lower plasma insulin levels during an oral glucose tolerance test. Whole-genome sequencing and Sanger sequencing analyses on the ihs genome detected two ihs-specific variants changing amino acids within the ihs locus; both variants in Slc25a46 and Tcerg1 were predicted to disrupt the protein function. Based on information regarding gene functions involving diabetes mellitus and insulin secretion, reverse-transcription quantitative polymerase chain reaction analysis revealed that the relative abundance of Reep2 and Sil1 transcripts from ihs islets was significantly decreased whereas that of Syt4 transcripts were significantly increased compared with those of control C57BL/6 mice. Thus, Slc25a46, Tcerg1, Syt4, Reep2 and Sil1 are potential candidate genes for the ihs locus. This will be the focus of future studies in both mice and humans. a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 Introduction Type 2 diabetes (T2D) is a metabolic disorder characterized by abnormal glucose homeostasis due to a defect in the secretion and/or action of insulin. T2D affects more than 200 million individuals worldwide, constituting a major public health problem globally, and its prevalence is increasing in many countries [1]. T2D shows different pathologies depending on race. In general, T2D in East Asian countries is characterized by lower levels of insulin secretion from pancreatic islets and more frequently a non-obese type, differing from its pattern among Caucasians [2][3][4][5]. The etiology of T2D is complicated, involving multiple genetic and environmental factors [6]. Therefore, there has been increased interest in animal models of T2D which genetic and environmental factors that could influence the development of the disease and related complications can be precisely controlled in vivo. Animal models have made significant contributions to the study of diabetes mellitus, providing new information on its management and treatment in humans [7,8].
Recently, our research group established a novel, non-obese mouse strain with spontaneous diabetes, called the insulin hyposecretion (ihs) mouse [9]. During the oral glucose tolerance test (OGTT), ihs mice exhibit markedly impaired glucose tolerance caused by lower concentrations of plasma insulin, whereas their insulin sensitivity and insulin synthetic ability are found to be normal. The ihs mice show a severe defect in insulin secretion from pancreatic islets even when the islets are stimulated by potassium chloride (KCl) and sulfonylurea, which can directly cause β-cell plasma membrane depolarization followed by an influx of Ca 2+ and insulin granule exocytosis [10,11]. These results indicate that the insulin secretory disorder in the ihs mouse is caused by a defect in the Ca 2+ signaling pathway in β-cells [9]. Thus, these characteristics distinguish ihs mice from other well-characterized T2D animal models. Therefore, the ihs mouse offers the promise of identifying of a novel gene involved in insulin secretion through regulation of the Ca 2+ signaling pathway.
In this study, we performed genetic linkage analysis and quantitative trait loci (QTL) analysis of the glucose tolerance phenotype as a first step toward identifying the gene responsible for the impaired insulin secretion of ihs mice. Furthermore, we developed and analyzed a novel congenic mouse strain harboring the ihs locus on a C57BL/6 genetic background to confirm the ihs locus identified by genetic linkage analysis.

Ethical statement
All animal experiments were approved by the President of NCGM and Kitasato University, following consideration by the Institutional Animal Care and Use Committee of NCGM (approval ID: no. 17056) and Kitasato University (approval ID: no. 17-099), and were carried out in accordance with institutional procedures, national guidelines, and the relevant national laws on the protection of animals.

Oral glucose tolerance test (OGTT)
Only male mice were used in this study, because the glucose intolerance is more prominent in males than in females [9]. The mice were fasted for 16 h, and tail blood glucose was measured at 0, 30, 60, 90, and 120 min after oral administration of glucose (2 g/kg body weight; Otsuka Pharmaceutical, Tokyo, Japan) by gavage. The blood glucose level was determined using Glutest Ace (Sanwa Kagaku Kenkyusho Co, Nagoya, Japan). For measurement of plasma insulin levels, plasma samples were collected from the retro-orbital venous plexus at 0, 15, and 30 min during OGTT. Plasma insulin levels were measured using a mouse insulin kit (Morinaga Institute of Biological Science, Inc., Kanagawa, Japan). Each area the under curve (AUC) of the blood glucose and insulin in the OGTT was calculated by the trapezoidal rule from glucose measurements at 0, 30, 60, 90, and 120 min and from plasma insulin measurements at 0, 15 and 30 min [9].

Genetic mapping
To enable genetic mapping for the ihs locus, F 1 progeny were generated by crossing male ihs mice with female B6 mice. Both male and female F 1 progeny were then backcrossed to ihs mice to produce N 2 progeny. Genomic DNA samples from male N 2 progeny were extracted from a tail clip by standard phenol/chloroform extraction [9] and were genotyped using polymorphic 59 simple sequence length polymorphism (SSLP) markers (S1 Table) as described previously [12]. The map positions of the SSLP and single nucleotide polymorphism (SNP) markers were established based on a mouse genome, Build GRCm38. In total, 96 male N 2 progeny were genotyped for the ihs locus according to their blood glucose levels measured at 60 min during OGTT. Mice with blood glucose levels less than 250 mg/dl at 60 min after the oral administration of glucose were taken to be normal glucose tolerance (NGT) type, and mice with blood glucose levels of 250 mg/dl or more at 60 min after the oral administration of glucose were considered the impaired glucose tolerance (IGT) type. The association between the ihs locus and each SSLP marker was individually evaluated via a chi-squared test using JMP7 statistics software (SAS Institute Japan, Tokyo, Japan) [12]. QTL analysis for blood glucose was performed using Map Manager QTXb20, a software program that uses a maximum likelihood algorithm with interval mapping and simultaneous search and that permits improved localization of loci and exclusion mapping [13]. Recombination frequencies (as percentages) were converted into genetic distance (centimorgans, cM) using the Kosambi map function. For each chromosome, a likelihood ratio statistic (LRS) score was calculated using 1000 random permutations of the trait values relative to the genotypes of the marker loci. Suggestive and significant threshold levels were calculated using 1000 random permutation tests, based on already-established guidelines [14]. Blood glucose levels at 60 min after the oral glucose administration were used as the quantitative trait. QTL data also were confirmed using QTL Cartographer 2.5 [15].

Development of congenic strain
To introgress the ihs locus derived from ihs mice into B6 mice, the F 1 (ihs × B6) progeny were backcrossed to B6 mice for 12 generations. The homozygous B6.ihs-(D18Mit233-D18Mit235) (hereafter, B6-ihs) were subsequently produced and maintained by sib-mating. The primer sets used to create B6-ihs are listed in S2 Table. Whole-genome sequencing and candidate gene search DNA was isolated from tail clips of female ihs mice using NucleoSpin Gel and PCR Clean-up kit (Takara Bio Inc., Shiga, Japan). A DNA library for the whole-genome sequence (WGS) was then prepared using NEBNext UltraTM DNA Library Prep kit (Illumina, San Diego, CA, USA), following the manufacturer's instructions. Prepared libraries were sequenced in pairedend 150-nucleotide reads using an Illumina NovaSeq 6000 (Illumina). Sequencing and library construction were performed by GENEWIZ (South Plainfield, NJ, USA). Data were aligned with the Burrows-Wheeler Alignment tool [16]; the SNV/InDel was selected using the Genome Analysis Toolkit pipeline (Broad Institute, Cambridge, MA, USA) [17], and was annotated by ANNOVAR [18]. The WGS data were registered with the DNA Data Bank of Japan (accession no. DRA010083). The variants were confirmed as not already present in the mouse genetic variation database (the Mouse Genome Project, a reference database of genetic variations in laboratory mice). An amino acid substitution that affects protein function was predicted by the online software tools SIFT [19] and PROVEAN [20].
For Sanger sequencing, genomic DNAs from ihs, B6 and ICGN mice were amplified with the gene-specific primer sets listed in S3 Table. The PCR products were purified using ExoSA-P-IT Express (Thermo Fisher Scientific, Waltham, MA, USA) and used as templates in the sequence reaction. Sequencing was performed using BigDye terminator v3.1 cycle sequencing kit (Thermo Fisher Scientific) and ABI 3130xl Genetic Analyzer (Thermo Fisher Scientific).

Reverse-transcription quantitative polymerase chain reaction (RT-qPCR)
Pancreatic islets were isolated from the pancreas by collagenase digestion as previously described [21]. Briefly, male mice at age 10 weeks were anesthetized with Sevoflurane (Mylan, Inc., Southpoint, PA, USA), followed by injection of collagenase into the common bile duct. The pancreas was removed by dissection and was incubated at 37˚C for 12 min. The islets were then collected under a stereoscopic microscope. Total RNAs from islets were extracted with the Isogen RNA extraction kit (Nippon Gene, Tokyo, Japan), and cDNA was prepared with ReverTra Ace (Toyobo, Osaka, Japan). For quantification, gene-specific primers (S4 Table) were used with Thunderbird SYBR qPCR Mix (Toyobo, Osaka, Japan) in a 7900HT Fast Real-Time PCR System (Applied Biosystems, Foster City, CA, USA). The expression level of each target gene was normalized to the level of Actb mRNA. Each experiment was performed in three biological replicates.

Statistics
The data are expressed as means ± standard error. Student's t-test was used for comparisons of two independent groups. The results for blood glucose levels and plasma insulin levels during OGTT were analyzed using repeated-measures analysis of variance (ANOVA). A P value <0.05 was considered statistically significant.

The mode of inheritance of insulin hyposecretion in ihs mice
In a previous study, our research group demonstrated that ihs mice showed marked impaired glucose tolerance due to impaired insulin secretion [9]. Therefore, the blood glucose levels after glucose loading were adopted as an indicator of impaired insulin secretion. To determine the mode of inheritance of insulin hyposecretion in ihs mice, F 1 and N 2 progeny were generated, and OGTTs were performed in these mice. As demonstrated in a previous study, ihs mice showed a remarkable impairment of glucose tolerance. However, the glucose tolerance of F 1 mice was comparable to that of B6 mice (Fig 1A). The glucose AUC also showed no statistically significant difference between F 1 and B6 mice (F 1 : 20,499.0 ± 2,331.5, vs. B6: 23,197.5 ± 1,358.5, P = 0.35, Fig 1B). Although the blood glucose levels at 60 min after glucose administration in the N 2 progeny showed consecutive distribution (Fig 1C), there seemed to be a distinct threshold to be able to be divided into two groups. Based on the result of OGTTs in B6, ihs and F 1 mice (Fig 1A), we divided them into two groups (NGT and IGT type) at 250 mg/dl (Fig 1C). In contrast, it was difficult to divide those of 120 min after glucose administration into groups due to a continuous trait (S1 Fig). Those N 2 progeny with blood glucose levels less than 250 mg/dl at 60 min after the oral administration of glucose were labeled the NGT type, and those N 2 progeny with blood glucose levels of 250 mg/dl or more at 60 min after the oral administration of glucose were called the IGT type. Of the 96 male offspring, 53 (55.2%) and 43 (44.8%) individuals at 10-15 weeks of age were classified as NGT and IGT type, respectively, indicating that hyposecretion of insulin in ihs mice is inherited in a recessive fashion ( Fig 1C).

Genetic linkage analysis of ihs locus
To identify the locus responsible for impaired insulin secretion in ihs mice (ihs locus), DNA samples were genotyped from 96 male N 2 progeny using 59 SSLP markers, in order to cover the whole mouse genome (S1 Table). The association could then be examined between the genotype of each SSLP marker and the phenotype, which had already been defined as either the B6 or the ihs type, in terms of glucose tolerance for each individual. Using a chi-squared test, a significant association was detected between the ihs locus and SSLP markers on chromosomes 4, 6, 18, 19, and X, with an especially highly significant association at D18Mit233 (P = 1.78E-18) ( Table 1). Since it was possible that the ihs locus was located on chromosome 18, haplotype analysis was performed using additional three SSLP markers on chromosome 18. This haplotype analysis revealed that the ihs locus was located around an interval of 14.9-Megabase (Mb) between D18Mit233 and D18 Mit235 (Fig 2A and 2B).
Since blood glucose is considered a measurable phenotype that depends on the cumulative actions of many genes and the environment, QTL analysis was performed for blood glucose levels at 60 min after the oral glucose administration. One significant QTL peak was identified on chromosome 18 between D18Mit233 and D18Mit235 with a likelihood ratio statistic (LRS) score of 124.5 (Table 2 and Fig 2C), and there were two suggestive QTL peaks on chromosome 4 at D4Mit13 and chromosome 11 at D11Mit38 with LRS scores of 7.2 and 8.3, respectively ( Table 2). There were no QTL peaks beyond the significant threshold level (LRS = 12.1) on any other chromosome except for chromosome 18 (Tables 2 and S5). In addition, QTL analysis for blood glucose levels at 120 min after glucose administration also showed a highly significant QTL peak between D18Mit132 and D18Mit53 on chromosome 18 (S6 Table) and a significant QTL peak at D11Mit38 on chromosome 11.

Glucose tolerance and insulin secretion of B6.ihs-(D18Mit233-D18Mit235) during the oral glucose tolerance test
To confirm that the gene responsible for the ihs locus (the ihs gene) was present in the region around between D18Mit233 and D18Mit235, a congenic mouse (B6-ihs), was produced by introgression of the region between D18Mit233 and D18Mit235 from the ihs mouse on the B6 genetic background. Upon OGTT, the B6-ihs mice showed marked impaired glucose tolerance compared to B6 mice (Fig 3A). The glucose AUC values after glucose loading in the B6-ihs mice also were higher than those of the B6 mice (B6-ihs: 36,366 ± 1,737.7 vs. B6: 24,594 ± 980.3, P < 0.01, Fig 3B). Furthermore, markedly lower plasma insulin levels during OGTT was also observed in B6-ihs mice (Fig 3C). The AUC for plasma insulin levels of ihs mice was significantly lower than those of B6 mice (B6-ihs: 7.49 ± 1.68 vs. B6: 25.37 ± 4.17, P < 0.01, Fig 3D). Female B6-ihs mice also showed impaired glucose tolerance and lower plasma insulin concentration than B6 mice during OGTT (Fig 3E, 3F, 3G and 3H). These results demonstrate that the ihs gene was definitely present in the region from D18Mit233 to D18Mit235, a region that contained 199 genes.

Whole-genome sequencing and candidate gene search
To identify the ihs mutation, WGS analysis was applied to the ihs genome to detect the differences between the ihs genome and the reference sequence (GRCm38) of the B6J strain. Nonsynonymous single nucleotide variants (nsSNVs) and/or insertion deletion (InDels) were sought that fulfilled the following four criteria: 1) nsSNVs and/or InDels in the coding regions of genes that were located within the ihs locus, 2) those that were homozygous genotype, because ihs mice had an autosomal recessive disorder, and 3) those that were ihs-specific variants, which were confirmed as not being in the mouse genetic variation databases, because hypoinsulin secretion from islets is a striking characteristic of the ihs strain, and 4) those that were confirmed by Sanger sequencing. The variants that fulfilled these criteria were evaluated as candidates for ihs mutation(s). WGS analysis of ihs genome revealed that 28,970 nsSVNs existed within the ihs locus (S7 Table), and the two ihs-specific nsSVNs in Slc25a46 and Tcerg1 genes satisfied the four criteria just described (Table 3). These nsSNVs were not present in either the dbSNP database or the 36 common inbred strains genome database in the Sanger Center (https://www.sanger.ac.uk/science/data/mouse-genomes-project). In addition, there were 10,442 InDels within the ihs locus (S8 Table); however, no InDel satisfied the criteria. Although two nsSNVs (c.A537C and c.T539C) and two InDels (c.527_529delACT and c.541_542insTAC) in Pcdhgb7 were identified by WGS analysis, these variants changing amino acids could not be confirmed by Sanger sequencing, indicating that the nsSNVs and InDels in Pcdhgb7 were detected by a WGS sequence error and/or errors due to the parameters for the SNP and InDel labeling used for data analysis.
Since the ihs mice were derived from the ICGN mice, we compared the variants changing amino acids found in the ihs mice with those in ICGN mice. The c.A683G in Slc25a46 was detected only in the ihs genome, whereas the c.C1586T in Tcerg1 was confirmed in both the ihs and ICGN genome (Table 3).
In terms of in silico prediction on whether an amino acid substitution has an impact on the biological function of a protein, PROVEAN [20] predicted a "deleterious" effect of p.N228S in Slc25a46 and of p.T529I in Tcerg1, and these effects are believed to disrupt the function of the The glucose tolerance and insulin secretion of B6-ihs mice during OGTT. a, e. Plot of the blood glucose (mg/ dl) of B6-ihs (male; n = 6, female; n = 9) and B6 (male; n = 6, female; n = 9) mice at 9-12 weeks of age. The blood glucose levels were measured at 0, 30, 60, 90, and 120 min after the oral glucose administration. b, f. The area under the curve (AUC) of the 0-120 min glycemic responses. c, g. The plasma insulin levels of B6-ihs (male; n = 5, female; n = 8) and B6 (male; n = 5, female; n = 7) mice at 10-17 weeks of age during OGTT. The plasma insulin levels were measured at 0, 15, and 30 min after glucose loading. d, h. AUC for insulin during 0-30 min. AUC was calculated using the trapezoidal rule. � P < 0.05, �� P < 0.01. https://doi.org/10.1371/journal.pone.0234132.g003 protein, whereas SIFT [19] suggested that both p.N228S in Slc25a46 and p.T529I in Tcerg1 have little effect on these protein function (Table 3).
In addition, given the information regarding the candidate genes for the ihs locus based on PubMed and MGI database, synaptotagmin 4 (Syt4), polycystic kidney disease 2 like 2 (Pkd2l2), receptor accessory protein 2 (Reep2), and endoplasmic reticulum chaperone SIL1 homolog (Sil1), which have been found involved in diabetes mellitus and/or insulin secretion, were the candidate genes for the ihs locus. However, WGS analysis of the ihs genome indicated that there were no nsSNVs and InDels in the protein coding regions of Syt4, Pkd2l2, Reep2, and Sil1genes (S7 and S8 Tables). Therefore, the mRNA expression of candidate genes in the pancreatic islets from B6 and ihs mice was next examined via RT-qPCR analysis. The relative abundance of Reep2 and Sil1 transcripts from the ihs islets were 1.4-fold and 1.7-fold lower than those from B6 islets, respectively, whereas the abundance of Syt4 transcripts was 1.3-fold higher than for those from B6 (Fig 4B). The Pkd2l2 transcripts could not be detected in either ihs or B6 islets (Fig 4A). In summary, Slc25a46, Tcerg1, Syt4, Reep2, and Sil1 can be tentatively proposed as potential candidate genes for the ihs locus.

Discussion
Spontaneous diabetic rodent models are invaluable tools for both understanding the pathogenesis of diabetes and identifying a novel diabetic gene. Our research group recently developed a novel insulin hyposecretion mouse that is suitable for identifying a novel gene involved in insulin secretion through regulating the Ca 2+ signaling pathway [9]. This study showed that a recessively acting locus responsible for the hyposecretion of insulin in the ihs mouse was mapped to a 14.9-Mb region on chromosome 18 between D18Mit233 and D18Mit235 ( Fig  2B), a region that contained 199 genes. This region overlapped the Nidd2, a NON-derived diabetogenic locus that controls blood glucose [22,23].
We also found that D4Mit13 showed a significant association between genotype and glucose tolerance (Table 1) and an LRS peak beyond suggestive threshold level in QTL analysis ( Tables 2 and S6). This region overlapped the Nidds locus, which is associated with hyperglycemia and decreased β-cell mass [24,25]. Furthermore, we identified a significant QTL peak (LRS = 13.1) of the blood glucose levels at 120 min after OGTT at D11Mit38 on chromosome 11 (S6 Table). This region overlapped the Nidd1n locus, which is associated with glucose tolerance and fasting and non-fasting blood glucose levels in NSY mice [26]. Given the consecutive distribution of blood glucose levels in N 2 progeny (Figs 1C and S1) and a considerable difference in the blood glucose levels at 120 min after OGTT between ihs ( Fig  1A) and B6-ihs (Fig 3A) mice, the Nidds locus on chromosome 4 and/or the Nidd1n locus on chromosome 11 might have an influence on the clearance of blood glucose after glucose loading in ihs mice. With respect to WGS analysis on the ihs genome, two ihs-specific nsSNVs in the ihs locus were identified. PROVEAN predicted that both p.N228S in Slc25a46 and p.T529I in Tcerg1 would be harmful to protein function (Table 3). In particular, p.N228S in Slc25a46 was detected in only the ihs mouse but not in the ICGN mouse from which the ihs mouse was derived (Table 3). Slc25a46 is a member of the mitochondrial solute carrier 25 protein family and is proposed to play a role in mitochondrial dynamics and maintenance of cristae by interacting with mitochondrial proteins involved in mitochondrial fusion and fission such as optic atrophy1 (Opa1), mitofusin 1 (Mfn1), and mitofusin 2 (Mfn2), and by interacting with the mitochondrial contact site (MICOS) complex involved in the maintenance of cristae [27]. Although the substrates of Slc25a46 have not been identified [28], dysfunction of this gene results in neuropathy including Charcot-Marie-Tooth type 2 [29], Leigh syndrome [30], and optic atrophy [31]. Mitochondria play an important role in insulin secretion [32]. Mitochondria produce ATP by glucose metabolism, resulting in closing the ATP-sensitive potassium channels causing β-cell plasma membrane depolarization followed by an influx of Ca 2+ and insulin granule exocytosis. Therefore, Slc25a46 might be involved in the regulation of insulin secretion [32].
Tcerg1 is a nuclear protein that regulates transcription and pre-mRNA splicing by interacting with RNA polymerase II and pre-mRNA splicing factor SF1 [33,34]. It has been suggested that this gene is directly and/or indirectly involved in regulating the expression of 900 or more genes [35]. Although the involvement of this gene in diabetes and insulin secretion has not been reported, Tcerg1 might indirectly affect insulin secretion through the regulation of insulin secretion related gene expression.
In addition to WGS analysis, a candidate gene approach was performed to identify the gene mutation responsible for hyposecretion of insulin in the ihs mice. Given the gene functions of the candidate genes at this locus, Syt4, Pkd2l2, Reep2, and Sil1 genes were the candidates for the ihs gene. Syt4, a non-Ca 2+ binding paralogue of the major β-cell Ca 2+ sensor Syt7, interacts with Syt7, and is involved in controlling the Ca 2+ sensitivity of insulin vesicle secretion during β-cell maturation. Syn4-deficient (Syt4-/-) mice showed a modest defect in glucose clearance and impaired GSIS. A significant higher basal GSISS was observed in pre-weaned Syt4-/-islets, implying that Syt4 increases the threshold of Ca 2+ needed to trigger insulin secretion [36]. In addition, islet-specific overexpression of Syt4 induced impairment of glucose tolerance during OGTT and impaired insulin secretion from isolated islets elicited by glucose [36]. RT-qPCR analysis of isolated islets in ihs mice showed a relative abundance of Syt4 transcripts that was 1.3-fold higher than those in B6 (Fig 4B), suggesting Syt4 gene is a potential candidate for the ihs mutation. Sil1 is a nucleotide exchange factor for the binding immunoglobulin protein (BiP), which is an endoplasmic reticulum chaperone [37,38]. In the β-cells, Sil1 is involved in insulin biosynthesis and insulin secretion processes [39]. Furthermore, Sil1-deficient mice showed a phenotype similar to that of ihs mice, such as having severe impairment of glucose tolerance and GSIS, and normal insulin sensitivity [9,39]. Reep2 belongs to the receptor expression enhancing protein superfamily and enhances the functions of taste receptor type1 member 3 (T1R3), which is a sweet taste receptor, by interaction of these proteins [40]. In the β-cells, T1R3 has been reported to promote insulin secretion by facilitating the metabolic pathway of mitochondria and subsequently increasing ATP production, suggesting that Reep2 might affect insulin secretion via T1R3 [41]. A decrease of 30% and 42% in the expression levels of the Sil1 and Reep2 genes, respectively, were seen in ihs islets compared with B6 islets (Fig  4B), suggesting that Sil1 and Reep2 gene are also the potential candidate for ihs mutation. Pkd2l2 is a member of transient receptor potential (TRP) superfamily and functions as a Ca 2+ permeable cation channel [42,43]. Although the physiological roles of this gene are unclear, Pkd2l2 might participate in insulin secretion by regulating intracellular Ca 2+ influx in β-cells.
However, Pkd2l2 transcripts could not be detected in either ihs or B6 islets (Fig 4A), and the ihs-specific variants changing amino acids could not be found on the Pkd2l2 gene (S7 and S8 Tables), suggesting that the Pkd2l2 gene is not responsible for the ihs locus.
Although the result of the WGS analysis showed no ihs-specific variants changing amino acids on exons of these genes (S7 and S8 Tables), the RT-qPCR analysis revealed that the ihs mice showed significantly higher expression levels of Syt4 gene and significantly lower expression levels of Sil1 and Reep2 than B6 mice (Fig 4B), suggesting that the ihs mutation is located in the promoter or intron region of these candidate genes. In general, the frequency of variations in exons of the genes is much lower than that in promoters and introns. As the ihs mice are spontaneous mutants derived from ICGN mice, which provide an animal model for congenital nephrosis [9], there is no suitable control strain. Therefore, we focused on the variants on the exons of the genes in the ihs locus as a first step toward identifying the ihs mutation responsible for impaired insulin secretion. Although further study, such as comprehensive gene expression analysis in ihs islets using RNA sequencing, is required for a better understanding of the molecular pathology of the ihs mice, Slc25a46, Tcerg1, Syt4, Reep2 and Sil1 are tentatively proposed as the potential candidates for the ihs gene.

Conclusions
The ihs locus was identified in the region on chromosome 18 between D18Mit233 and D18Mit235 (14.9-Mb). Furthermore, Slc25a46, Tcerg1, Syt4, Reep2 and Sil1 in this region are tentatively proposed as potential candidate genes for the ihs gene. This will be a focus of future studies in both mice and humans. The identification of the gene responsible for the impaired insulin secretion of the ihs mouse is expected to lead to further understanding of the regulation of the Ca 2+ signaling pathway and insulin granule exocytotic machinery in pancreatic β-cells.