PDX1 phosphorylation at S61 by mTORC1 links nutrient signaling to β cell function and metabolic disease

SUMMARY

PDX1 is a key transcription factor regulating insulin expression in response to glucose. Our previous work showed that PDX1 is also stimulated by amino acids (aa). Here, we demonstrate that PDX1 broadly mediates aa-regulated transcriptional programs in β cells, especially those controlling β cell proliferation and function. Mechanistically, mTORC1 phosphorylates PDX1 at serine 61 (S61), enhancing its protein stability and transcriptional activity. A certain monogenic diabetes mutation disrupts this phosphorylation and impairs PDX1 function. To investigate its physiological role, we generated mice carrying S61A and S61E mutations, mimicking unphosphorylated and phosphorylated states. S61 phosphorylation promoted insulin expression and β cell proliferation, leading to Western diet-induced hyperinsulinemia, obesity, and hepatic steatosis. These findings reveal the central role of aa-mTORC1-PDX1 signaling in coordinating β cell proliferation and function under both physiological and pathological conditions.

Graphical abstract

In brief

Fan et al. identify a mechanism where nutrient-mTORC1 signaling controls β cell function by phosphorylating PDX1 at S61, enhancing its expression and insulin transcription. S61 phosphorylation influences glucose homeostasis, promoting hyperinsulinemia, obesity, and steatosis under a Western diet. Mutations disrupting this site impair PDX1 activity, linking it to diabetes risk.

INTRODUCTION

Pancreatic β cells play a crucial role in glucose homeostasis in response to changes in circulating glucose levels.¹ They release insulin to regulate the uptake, utilization, and storage of glucose by various tissues.² In metabolic diseases, particularly diabetes mellitus, β cell function is often impaired or compromised.¹ Increased circulating insulin levels, referred to as hyperinsulinemia,³ promote excessive fat storage in adipose tissue by enhancing lipogenesis and inhibiting lipolysis.^4–7 Excessive adipose tissue, in turn, contributes to systemic insulin resistance and causes compensatory hyperinsulinemia, creating a vicious cycle. Chronic hyperinsulinemia also promotes hepatic steatosis that can progress to metabolic dysfunction-associated steatohepatitis (MASH).^8,9 Multiple environmental and inherited factors drive insulin hypersecretion, and reducing these burdens has been shown to prevent or ameliorate the disease.^10–12 Understanding this interplay is crucial for developing effective strategies to prevent and manage hyperinsulinemia-related metabolic disorders.

PDX1 is a key glucose-response transcription factor for insulin.¹³ It also plays a pivotal role in pancreatic β cell development, function, and phenotype maintenance.¹⁴ In mature β cells, PDX1 regulates the expression of key genes involved in glucose-stimulated insulin secretion, including insulin itself, GLUT2, and glucokinase.¹⁴ PDX1 dysfunction has been implicated in various forms of diabetes, as mutations in or reduced expression of PDX1 can reduce insulin production and β cell proliferation, leading to impaired cell survival.¹⁴ Several germline missense mutations, including C18R, P33T, Q59L, and D76N occurring within the N-terminal transactivation domain, have been associated with increased risks of developing type 2 diabetes (T2D) and multiple forms of monogenic diabetes, including maturity-onset diabetes of the young (MODY) and neonatal diabetes.^15–17 Understanding complex regulatory networks involving PDX1 is crucial for developing new therapeutic strategies to preserve β cell function and prevent or treat metabolic disorders.¹⁸

Amino acids (aa) represent major dietary components that are essential building blocks for protein and peptide synthesis in all organisms.¹⁹ They also serve as vital chemical signals for activating mTORC1 to regulate eukaryotic cell growth and metabolic homeostasis.^20–24 mTOR is an evolutionarily conserved protein serine-threonine kinase,²⁵ which constitutes the catalytic subunit of mTORC1 and mTORC2.^26,27 These multiprotein complexes have distinct functions, determined by different accessory subunits and sensitivity to rapamycin (Rap) inhibition. mTORC1 is a central controller of nutrient signaling in eukaryotes,^26,27 particularly aa signaling.²⁸ Under normal physiological conditions, mTORC1 plays a positive role in regulating β cell proliferation and survival, as evidenced by studies showing that Raptor or mTOR deficiency in β cells leads to reduced β cell mass and loss of β cell identity.^29–32 In contrast, hyperactivation of mTORC1 signaling due to chronic high-nutrient conditions has been observed in β cells and other metabolically active tissues in the setting of insulin resistance and T2D.^33–36 These opposing effects of the mTORC1 pathway activity are consistent with β cell adaptation and eventual failure during progression to T2D.³⁷ Although mTOR signaling is known to regulate β cell function in response to nutrients, the molecular mechanisms underpinning this regulation, particularly in metabolic disease states, remain poorly understood.

Recent advances yielded insights into the molecular mechanisms of aa sensing and signaling.³⁸ For example, RAB1A-mTORC1-mediated aa signaling operates independently of the RAG-mTORC1 axis.^39,40 At the organismal level, the aa-RAB1A-mTORC1 axis plays an important role in regulating insulin expression through PDX1 in pancreatic β cells.⁴¹ aa promote PDX1 transcriptional activity at the insulin gene promoter and β cell growth in an mTORC1- and PDX1-dependent manner. Postnatal induced homozygous Rab1A knockout results in reduced pancreatic β cell mass and hyperglycemia in mice.^41,42 However, the precise molecular mechanism and pathological significance by which mTORC1 regulates PDX1 function in pancreatic β cells remains unclear. In this study, we found that mTORC1 interacts with PDX1 and phosphorylates PDX1 at serine 61 (S61) to regulate PDX1 protein expression and transcriptional activity. We further showed that phosphorylation of PDX1(S61) promotes pancreatic β cell growth and function, both in vitro and in vivo, which is important for regulating glucose homeostasis. Persistent S61 phosphorylation promotes Western diet (WD)-induced hyperinsulinemia, obesity, and liver steatosis. These findings reveal the mechanism and roles of aa-mTORC1-PDX1 signaling in pancreatic β cell proliferation and function under physiological and pathological conditions.

RESULTS

aa-mTORC1 signaling broadly regulates genomic PDX1 binding in pancreatic β cells

We previously showed that aa regulate Pdx1 localization and transcriptional activation of the insulin promoter in pancreatic β cells in an mTORC1-dependent manner.^41,42 Moreover, aa and glucose independently and synergistically regulate PDX1.⁴¹ These observations suggest that aa play a major role in PDX1 regulation in pancreatic β cells. To gain further insight into the role of aa in PDX1-dependent transcriptional programs, we profiled genome-wide PDX1 binding sites using chromatin immunoprecipitation sequencing (ChIP-seq) in MIN6 cells in response to aa deprivation (−aa) and restimulation (re+aa) in the absence or in the presence of Rap (Figure S1A). aa markedly increased genomic binding peaks of PDX1 from ~12,000 under aa deprivation to >25,000 (Figures S1B and S1C). Rap blunted the aa-stimulated increase in PDX1 genomic binding events. Consistent with its role as a transcription factor, PDX1 binding was enriched in intergenic regulatory regions around transcription start sites (TSSs) (Figure S1D). De novo motif analysis was used to discover enriched motifs within peak regions. The most significantly enriched motif matches the known PDX1 consensus motif (Figure 1A),⁴³ demonstrating the efficiency and specificity of the PDX1 ChIP-seq approach. Together, these results document that aa broadly regulate the PDX1 DNA interactome in an mTORC1-dependent manner.

Figure 1. aa-mTORC1 signaling regulates global PDX1 DNA-binding activities in pancreatic β cells.

MIN6 cells were starved of aa for 24 h and restimulated with 1×aa for 3 h in the absence or presence of 100 nM rapamycin (Rap). Profiling of PDX1’s global DNA binding was carried out by anti-PDX1 ChIP-seq.

(A) The most enriched motif discovered by de novo motif analysis resembles the known PDX1 consensus binding motif.

(B) Venn diagram showing overlap in PDX1-bound regions between aa restimulation group (−aa versus re+aa) and aa stimulation + Rap group (re+aa versus re+aa+Rap). Select PDX1-bound target genes are annotated in the overlap groups.

(C) Functional categories of PDX1 target genes as revealed by pathway enrichment analysis. The top 20 enriched pathways in the overlapped PDX1 ChIP-seq targets are shown.

(D) ChIP-seq signal tracks of PDX1 on select target genes in MIN6 cells.

We next performed a comparative analysis of PDX1 binding peaks under various conditions. aa restimulation enhanced PDX1 binding in the promoter region of 2,351 genes. Binding of PDX1 to the promoters of 1,901 genes was notably diminished in the Rap (re+aa+Rap) group when compared with aa stimulation (1,299 genes showed overlap between the two groups; Figure 1B). This observation suggests that a large subset of PDX1 target genes is regulated by mTORC1. KEGG pathway enrichment analysis of the overlapping 1,299 genes showed that aa-regulated, mTORC1-dependent PDX1 target genes are enriched in signaling pathways associated with pancreatic β cell proliferation and function (Figure 1C), including Ins2, Glp1r, Slc2a2, Klf11, and Ypel3, which are pivotal for the growth and function of pancreatic β cells (Figure 1D).

To validate the ChIP-seq results, we used small interfering RNAs (siRNAs) to knock down Pdx1 in MIN6 cells. All three siRNAs tested efficiently downregulated PDX1 expression at both the mRNA and protein levels (Figures S1E–S1H). aa significantly upregulated or downregulated PDX1 target genes (Figures S1I–S1N). PDX1 target genes that were upregulated by aa include Ins2, Glp1r, and Slc2a2 (Figures S1I–S1K). PDX1 target genes that were downregulated by aa include Csnk1e, Klf11, and Ypel3 (Figures S1L–S1N). Pdx1 knockdown blunted the response of both these upregulated and downregulated genes. We further characterized PDX1 binding to the promoters of Klf11 and Ypel3 under different conditions. Our results showed that PDX1 was bound near the TSS of these genes but not further upstream or within the coding regions (Figures S1Q–S1R). Moreover, binding of PDX1 near these TSSs was dependent on aa levels. These results indicate that they are bona fide PDX1 target genes.

mTORC1 interacts with PDX1 and phosphorylates PDX1 at S61

The findings described above underscore the pivotal role of aa-mTORC1-PDX1 signaling in governing the intricate regulation of pancreatic β cell proliferation and function. To better understand the relationship between mTORC1 and PDX1, we explored their potential interaction by co-immunoprecipitation (coIP) in MIN6 cells. PDX1 was found to be present in the anti-mTOR and anti-Raptor IPs (Figures 2A and 2B). Conversely, both mTOR and Raptor, but not Rictor, were co-immunoprecipitated by the anti-PDX1 antibody (Figure 2C), indicating that PDX1 interacts specifically with mTORC1, not mTORC2. Consistently, the interaction between mTOR and PDX1 was stimulated by aa, and the aa-stimulated interaction was attenuated by Rap (Figure 2D). The latter result was confirmed by the proximity ligation assay (PLA) in MIN6 cells (Figure 2E).

Figure 2. mTORC1 interacts with PDX1 and phosphorylates PDX1 at S61 in vitro.

(A–C) PDX1 interacts with mTORC1, not mTORC2. Endogenous mTOR (A), Raptor (B), or PDX1 (C) was immunoprecipitated from MIN6 cells and analyzed for the presence of PDX1, mTOR, and Raptor, respectively, by immunoblot.

(D) The interaction between mTOR and PDX1 is aa and mTORC1 dependent. MIN6 cells were starved of aa for 24 h and restimulated with 1×aa for 4 h in the absence or in the presence of 100 nM rapamycin. The interactions between mTOR and PDX1 protein were analyzed by co-immunoprecipitation using an anti-PDX1 antibody.

(E) mTOR interacts with PDX1 in the nucleus in an aa- and mTORC1-dependent manner. Endogenous mTOR and PDX1 proteins were assayed for their interaction in situ in MIN6 cells by the proximal ligation assay (PLA) (red). Representative confocal microscopic images are shown. The boxed areas were enlarged to show details. Nuclei were counterstained by DAPI (blue). Scale bar, 10 μm.

(F) S61 is located within a conserved mTORC1 consensus substrate motif. The mTORC1 consensus substrate motif and alignment of S61-adjacent aa sequences among humans, mice, and rats are shown.

(G–I) mTORC1 phosphorylates PDX1(S61) in vitro. An S61-containing PDX1 oligopeptide was incubated with mTOR isolated from 293T cells treated without or with 100 nM rapamycin for 4 h from HEK293 cell lysates. S61 phosphorylation was analyzed by ELISA using an anti-p-PDX1(S61) antibody. A phosphorylated S61 oligopeptide and an S61A oligopeptide were used as a positive and a negative control, respectively.

(J–L) mTORC1 phosphorylation of PDX1(S61) is dependent on mTOR kinase domain. An S61-containing PDX1 oligopeptide was incubated with mTOR isolated from 293T cells treated without or with 100 nM Torin for 4 h from HEK293 cell lysates. S61 phosphorylation was analyzed by ELISA using an anti-p-PDX1(S61) antibody. A phosphorylated S61 oligopeptide and an S61A oligopeptide were used as a positive and a negative control, respectively.

***p < 0.001 and n.s. p > 0.05. An unpaired Student’s t test for two groups was used. Data are represented as the mean ± SEM.

mTORC1 is a conserved serine/threonine protein kinase complex that phosphorylates various substrates,²⁴ predominantly favoring proline (P), hydrophobic (L and V), and aromatic (F, W, and Y) residues at the +1 position.^44,45 mTOR binding to PDX1 raises the possibility that PDX1 is a target substrate for mTORC1. Sequence alignment of vertebrate PDX1 proteins revealed that S61 is evolutionarily conserved and falls within a consensus mTORC1 substrate motif, suggesting it is a potential mTORC1 phosphorylation site (Figure 2F). Moreover, phosphoproteomic analyses revealed S61 as the principal PDX1 phosphorylation site in both MIN6 cells and the mouse pancreas.⁴⁶ We, therefore, asked if mTORC1 phosphorylates S61 by performing an in vitro kinase assay. mTOR was purified by IP from 293T cells treated without or with Rap (Figure 2G) and assayed for phosphorylation of an S61-containing PDX1 peptide with an ELISA using a phosphorylated (p)-PDX1(S61)-specific antibody. As a control, the p-PDX1(S61) antibody recognized the custom-made p-PDX1(S61) peptide but not the PDX1 peptide without phosphorylation, demonstrating assay specificity (Figures 2H and 2I). mTOR significantly phosphorylated PDX1(S61), which was inhibited by Rap (Figures 2H and 2I). Consistently, Torin, a selective small-molecule inhibitor of mTOR kinase, inhibited PDX1(S61) phosphorylation by mTOR (Figures 2J–2L), showing that the mTOR kinase domain is responsible for PDX1(S61) phosphorylation. These results establish that mTORC1 phosphorylates PDX1(S61) in vitro.

S61 phosphorylation regulates PDX1 protein level and transcriptional activity

To characterize S61 phosphorylation in pancreatic β cells, we used the custom p-PDX(S61) antibody. In immunoblots, it recognized wild-type FLAG-PDX1 but not FLAG-PDX1(S61A), a non-phosphorylatable mutant (Figure 3A), demonstrating antibody specificity. We found that aa starvation inhibited PDX1 S61 phosphorylation (Figures 3B, 3C, and S2). Conversely, aa restimulation increased p-PDX1(S61), which could be blocked by Rap (Figures 3C and S2B). Because aa starvation and Rap treatment affected the total PDX1 protein level, we immunoprecipitated FLAG-PDX1 and normalized the FLAG-PDX1 protein load for western blot, which confirmed the above results (Figure 3D). Additionally, Torin inhibited S61 phosphorylation, indicating that mTOR kinase is required (Figure 3D). PDX1 phosphorylation and target gene expression respond similarly to aa starvation and restimulation in an mTORC1-dependent manner in isolated mouse pancreatic islets, indicating that mTORC1 regulates PDX1 phosphorylation in a more physiologically relevant setting (Figures 3E and 3F).

Figure 3. aa and mTORC1 regulate PDX1(S61) phosphorylation in pancreatic β cells.

(A) Characterization of p-PDX1(S61)-specific antibody. FLAG-PDX1 and FLAG-PDX1(S61A) were expressed in MIN6 cells and analyzed for S61 phosphorylation by immunoblot using a custom anti-p-PDX1(S61) antibody.

(B) aa regulate PDX1 phosphorylation at S61. MIN6 cells were aa starved for different amounts of time. Phosphorylated PDX1 (p-PDX1) and total PDX1 protein were analyzed by immunoblot. β-actin was used as a loading control.

(C) aa regulate PDX1(S61) phosphorylation in an mTORC1-dependent manner. MIN6 cells were aa starved for 24 h and then restimulated with 1×aa for 4 h in the absence or in the presence of 100 nM rapamycin. p-PDX1(S61) and total PDX1 were analyzed by immunoblot. The S61 phosphorylation level under different conditions was normalized by the ratio of p-PDX1(S61) to PDX1.

(D) aa regulate exogenous PDX1(S61) phosphorylation in an mTORC1 kinase-dependent manner. MIN6 cells expressing FLAG-PDX1 were aa starved for 24 h and then restimulated with 1×aa for 4 h in the absence or in the presence of 100 nM rapamycin or Torin1. FLAG-PDX1 was immunoprecipitated using an anti-FLAG antibody, and p-PDX1(S61) was analyzed by an anti-PDX1(S61) immunoblot.

(E) aa regulate PDX1(S61) phosphorylation in mouse islets. Islets were isolated from Pdx1^+/+ mice, starved of aa for 4 h, and then restimulated with aa for 4 h in the absence or in the presence of 100 nM rapamycin. p-PDX1 and total PDX1 protein were analyzed by immunoblot.

(F) aa regulate gene expressions involved in β cell function in islets. Islets were isolated from Pdx1^+/+ mice, starved of aa for 4 h, and then restimulated with 1×aa for 4 h in the absence or in the presence of 100 nM rapamycin. mRNA expression of Pdx1, Ins2, Mafa, and Slc2a2 was analyzed by RT-qPCR.

(G) S61 phosphorylation mediates aa-mTORC1 signaling to regulate PDX1 protein expression. MIN6 cells expressing wild-type or mutant FLAG-PDX1 were aa starved for 24 h and then restimulated with 1×aa for 4 h in the absence or in the presence of 100 nM rapamycin. FLAG-PDX1 protein levels were analyzed by an anti-FLAG immunoblot.

(H) S61 phosphorylation regulates PDX1 transcription activity. MIN6 cells expressing wild-type or mutant PDX1 were assayed for the luciferase reporter activity driven by two different Ins2 promoter segments (n = 3).

***p < 0.001. An unpaired Student’s t test for two groups was used. Data are represented as the mean ± SEM.

We previously showed that PDX1 is regulated by both aa and glucose in an mTORC1-dependent manner. Indeed, both p-PDX1(S61) and p-S6K(T389) levels declined following glucose starvation (Figures S3A and S3B). Interestingly, Pdx1 mRNA levels increased moderately during glucose starvation, likely due to a feedback, compensatory mechanism (Figure S3C). The re-addition of glucose resulted in a time- and dose-dependent increase in p-PDX1(S61) (Figures S3D–S3F). Consistently, glucose re-addition stimulated the interaction between mTOR and PDX1 (Figures S3G and S3H). Additionally, p-PDX1 responds to variable aa and glucose concentrations (Figures S2E, S2F, and S3I–S3L). Collectively, these results show that PDX1 phosphorylation responds to changing aa and glucose levels.

To characterize the impact of S61 phosphorylation, we transiently expressed wild-type PDX1 or the S61 mutant in MIN6 cells treated under different conditions. As shown previously, aa privation markedly reduced the wild-type FLAG-PDX1 protein level (Figure 3G). In contrast, aa restimulation restored the FLAG-PDX1 level, which could be blocked by Rap (Figure 3G). The phosphomimetic mutant FLAG-Pdx1(S61E) showed an enhanced PDX1 protein level. Conversely, the unphosphorylated mimetic mutant FLAG-Pdx1(S61A) showed a reduced protein level (Figure 3G). Neither mutant responded to aa starvation/restimulation or Rap treatment, indicating that S61 phosphorylation is responsible for the regulation of PDX1 protein expression under these conditions. An Ins2 promoter-driven luciferase activity assay showed that FLAG-PDX1(S61E) exhibited enhanced PDX1 promoter activity and that FLAG-PDX1(S61A) exhibited reduced PDX1 promoter activity when compared with wild-type FLAG-PDX1 (Figure 3H). Moreover, the expression of FLAG-PDX1(S61E), not FLAG-PDX1(S61A), promoted MIN6 cell proliferation (as judged by Ki67 staining) and endogenous insulin mRNA expression (as judged by fluorescence in situ hybridization [FISH]) (Figure S4). Collectively, these results show that S61 phosphorylation increases PDX1-dependent β cell proliferation and transcription.

Monogenic diabetes mutation interferes with S61 phosphorylation and PDX1 transcriptional activity

Monogenic diabetes mellitus, caused by single-gene germline mutations, accounts for approximately 5% of all diabetes cases.⁴⁷ Mutations in PDX1 have been associated with T2D and multiple types of monogenic diabetes, including MODY and neonatal diabetes. Interestingly, several such mutations were mapped to the N-terminal activation domain of PDX1, in the vicinity of S61 (e.g., Q59L).¹⁶ Because of the proximity of these MODY mutations to S61, we investigated the potential effect on S61 phosphorylation. To this end, we generated the FLAG-PDX1(Q59L) mutant and transiently expressed it in MIN6 cells. Compared to the wild type, the PDX1(Q59L) mutant exhibited reduced S61 phosphorylation (Figures 4A and 4B) and decreased transcriptional activity toward the Ins2 promoter in a luciferase reporter assay (Figure 4C). Moreover, mTORC1 exhibited markedly reduced activity toward the PDX1 Q59L mutant substrate peptide (LGSPP) compared with the wild-type substrate peptide (QGSPP) in the in vitro kinase assay (Figures 4D and 4E). Collectively, these findings suggest that the MODY Q59L mutation interferes with PDX1 function by preventing mTORC1-dependent S61 phosphorylation.

Figure 4. The monogenic diabetes mutation Q59L interferes with PDX1(S61) phosphorylation.

(A and B) Pdx1Q59L has a decreased phosphorylation level at S61. MIN6 cells were cultured in normal medium and transfected with PDX1 wild-type or PDX1 mutant plasmids. After 48 h, phosphorylated PDX1 (p-PDX1) and total overexpressed PDX1 protein were analyzed by immunoblot. The p-PDX1 (S61) and PDX1 staining intensity ratios were calculated. β-actin was used as a loading control.

(C) PDX1Q59L regulates PDX1 transcription activity. MIN6 cells were cultured in normal medium and transfected with PDX1 or PDX1 mutant plasmids together with the Ins2 promoter-luciferase reporter plasmid. After 48 h, cell lysates were analyzed using the standard protocol. Triplicates were used (n = 3).

(D and E) mTORC1 phosphorylation assay in vitro. Four peptides, QG(pSer)PP, QGSPP, LGSPP, and QGAPP, were coated in 96 wells and phosphorylated by an immunoprecipitated mTOR complex from 293T cells. Phosphorylation was detected by the p-PDX1 (S61) antibody, followed by a standard ELISA. The OD450 nm measurement is shown in (E).

***p < 0.001. An unpaired Student’s t test for two groups was used. Data are represented as the mean ± SEM.

PDX1(S61) phosphorylation regulates pancreatic β cell proliferation in vivo

PDX1(S61) has been previously shown to be phosphorylated in developing chicken and mouse pancreas.⁴⁶ However, neither PDX1(S61A) nor the PDX1(S61E) mutant, when overexpressed, affected the pancreatic development of chicken embryos,⁴⁶ suggesting that phosphorylation of PDX1(S61) per se is not a key determinant of pancreas development. To assess the role of S61 phosphorylation in vivo, we engineered knockin mice (C57BL/6J) carrying S61A or S61E mutations in the Pdx1 genomic locus, which were verified by genotyping and targeted genomic DNA sequencing (Figures S5A–S5E). Pdx1^S61A/S61A and Pdx1^S61E/S61E mutant mice were born at expected Mendelian ratios and developed without discernible abnormalities. Moreover, the adult pancreas showed relatively normal histology, supporting previous observations that S61 phosphorylation status does not affect pancreas development.⁴⁶ Notably, immunohistochemical staining of insulin in the adult pancreas revealed that the Pdx1^S61A/S61A islet size was much smaller than that of wild-type animals, while the Pdx1^S61E/S61E islet size was larger (Figures 5A–5C). Our ChIP-seq analysis showed earlier that PDX1 regulates growth and proliferation genes in β cells and that PDX1 phosphorylation promotes the proliferation of MIN6 cells. Therefore, we stained the mouse pancreas tissue with PCNA, a cell proliferation marker. Ins⁺PCNA⁺ β cells increased in Pdx1^S61E/S61E islets compared to wild-type islets, while Ins⁺PCNA⁺ β cells decreased in Pdx1^S61A/S61A islets (Figures 5D and 5E), suggesting that S61 phosphorylation positively regulates β cell proliferation. Consistent with our cell-based results, PDX1 protein expression is reduced in β cells in Pdx1^S61A/S61A mice, compared with that in wild-type or Pdx1^S61E/S61E animals (Figures S5F–S5H). MAFA, a pancreatic β cell-specific transcription activator, whose expression remains relatively comparable among all three mouse groups, while GLUT2 protein expression is downregulated in Pdx1^S61A/S61A mice (Figure S6), suggesting that loss of S61 phosphorylation affects functions but not β cell maturity.

Figure 5. PDX1(S61) phosphorylation promotes islet β cell proliferation and insulin expression in vivo.

(A and B) PDX1(S61) phosphorylation status impacts islet size and β cell area. IHC staining of insulin protein in pancreatic tissues of 8-month-old wild-type and Pdx1^S61A/S61A and Pdx1^S61E/S61E knockin mice under a normal chow diet was performed. Ins⁺ areas were quantified using ImageJ software. Pdx1^+/+ (n = 6), Pdx1^S61A/S61A (n = 6), and Pdx1^S61E/S61E (n = 5). At least 5 islets were surveyed in each mouse. Scale bar, 100 μm. ND, normal diet.

(C) The whole slides of the above tissues were scanned to get the image of the whole pancreas. The entire Ins+ area and the whole pancreas area were calculated using ImageJ. 3 slides were used for each mouse, with 4 mice for each genotype group.

(D and E) PDX1(S61) phosphorylation promotes pancreatic β cell proliferation. Immunofluorescence (IF) analysis of insulin (red) and PCNA (green) in 8-week-old wild-type and Pdx1^S61A/S61A and Pdx1^S61E/S61E knockin mice was performed. The boxed areas were enlarged to show details. The white arrows show cells with positive PCNA staining in the nucleus (D). Scale bar, 100 μm. Proliferation was calculated by the percentage of Ins⁺PCNA⁺ cells/Ins⁺ cells per mouse in each group (E). At least 400 Ins⁺ cells per mouse were included

(F and G) PDX1(S61) phosphorylation negatively regulates Klf11 expression. Klf11 expression was analyzed by RT-qPCR in Pdx1^+/+, Pdx1^S61A/S61A, and Pdx1^S61E/S61E mouse pancreas at 7 months old (n = 3 for each group).

(H) PDX1(S61) phosphorylation promotes insulin expression. Ins2 expression was analyzed by RT-qPCR in Pdx1^+/+, Pdx1^S61A/S61A, and Pdx1^S61E/S61E mouse pancreas at 7 months old (n = 3 for each group).

(I and J) PDX1(S61) phosphorylation status regulates circulating insulin level. Serum insulin was measured by ELISA in Pdx1^+/+ and Pdx1^S61A/S61A mice (37 weeks old) (H) or in Pdx1^+/+ and Pdx1^S61E/S61E mice (41 weeks old) (I).

All results are presented as the mean ± SEM. Representative results are shown. ***p < 0.001, **p < 0.01, and *p < 0.05. An unpaired Student’s t test for two groups was used.

Klf11, a PDX1 target gene identified in our ChIP-seq analysis, was previously shown to be a negative growth regulator.⁴⁸ Indeed, Klf11 mRNA expression was upregulated in Pdx1^S61A/S61A islets and downregulated in Pdx1^S61E/S61E islets, compared to wild-type islets (Figures 5F and 5G). These results suggest that phosphorylation of PDX1(S61) positively contributes to the regulation of pancreatic β cell proliferation through its downstream target genes. Consistent with our observations in MIN6 cells, Ins2 mRNA expression was reduced in Pdx1^S61A/S61A islets but increased in Pdx1^S61E/S61E islets, compared to wild-type islets (Figure 5H). Consistently, circulating insulin levels were reduced in Pdx1^S61A/S61A mice compared to wild-type animals (Figure 5I). Interestingly, there was no significant difference in circulating insulin levels between Pdx1^S61E/S61E and wild-type animals (Figure 5J). This suggests that insulin protein stockpiles in β cells are maximal in wild-type animals and are not significantly altered by the presence of additional insulin-encoding mRNAs under normal dietary conditions.

PDX1 phosphorylation at S61 drives hyperinsulinemia and the development of obesity and hepatic steatosis under WD conditions

The WD, characterized by high intake of processed foods, refined carbohydrates, and saturated fats, has been strongly linked to the rising prevalence of obesity and T2D worldwide.⁴⁹ This dietary pattern often leads to excessive calorie consumption and nutritional imbalances, which can promote weight gain and insulin resistance.⁴⁹ To gain insight into the potential role of PDX1(S61) phosphorylation in WD-induced metabolic pathogenesis, we monitored metabolic parameters of the knockin mice under a 45 kcal WD for an extended period. Interestingly, Pdx1^S61E/S61E mice exhibited more accelerated body weight gain under the WD than their wild-type and Pdx1^S61A/S61A counterparts (Figures 6A and 6B), despite similar food intake among all three animal groups (Figure 6C). The white adipose tissues (WATs) and liver tissues, two pivotal players in WD-induced metabolic syndromes, such as obesity, liver diseases, and diabetes, were further examined. The results showed that the WD-induced WAT gain was much greater in Pdx1^S61E/S61E mice than in the other two animal groups (Figure 6D), which was due to increased adipocyte size (Figures 6E and 6F). In contrast, the WD-induced WAT gain was smaller in Pdx1^S61A/S61A mice (Figures 6E and 6F). The WD-induced liver weight gain was also much greater for Pdx1^S61E/S61E mice than for their wild-type counterparts (Figures 6G and 6H). The difference is particularly pronounced when considering the liver/body weight ratio (Figure 6H). Consistently, the WD-fed Pdx1^S61E/S61E mice exhibited more severe hepatic steatosis than the other two animal groups (Figure 6I). In contrast, the Pdx1^S61A/S61A mice exhibited resistance to WD-induced WAT gain and liver steatosis (Figures 6D–6I). These findings suggest that Pdx1(S61) phosphorylation contributes to WD-induced obesity and liver steatosis.

Figure 6. PDX1(S61) phosphorylation promotes WD-induced obesity and hepatic steatosis in mice.

(A–C) PDX1(S61) phosphorylation contributes to WD-induced weight gain. Body weight (A), body weight gain (B), and food intake (C) were monitored in Pdx1^+/+, Pdx1^S61A/S61A, and Pdx1^S61E/S61E male mice fed a 45 kcal WD from 8 to 32 weeks old. Weight changes and food intake were measured weekly.

(D–F) PDX1(S61) phosphorylation promotes WD-induced obesity. Representative images are shown for inguinal white adipose tissue (iWAT) and epididymal white adipose tissue (eWAT) (D) and H&E staining in iWAT (E) in 45-kcal-WD-fed mouse Pdx1^+/+, Pdx1^S61A/S61A, and Pdx1^S61E/S61E pancreas at 36 weeks. Scale bar, 100 μm. Adipose cell sizes were quantified using ImageJ software (F). Pdx1^+/+ (n = 4), Pdx1^S61A/S61A (n = 4), and Pdx1^S61E/S61E (n = 4). ≥6 adipose cells were analyzed in each mouse.

(G–I) PDX1(S61) phosphorylation promotes WD-induced liver steatosis. Representative images are shown for the whole liver (G) and H&E staining (I) of 45-kcal-WD-fed Pdx1^+/+, Pdx1^S61A/S61A, and Pdx1^S61E/S61E mice at 36 weeks. Scale bar, 100 μm. The liver/body weight ratio was calculated in WD-fed mouse Pdx1^+/+ (n = 4), Pdx1^S61A/S61A (n = 4), and Pdx1^S61E/S61E (n = 4) pancreas at 36 weeks (H).

All results are presented as the mean ± SEM. Representative results are shown. ***p < 0.001 and *p < 0.05. An unpaired Student’s t test for two groups was used.

Hyperinsulinemia plays a crucial role in the development of WD-induced obesity and liver steatosis by promoting fat storage and inhibiting fat breakdown.^4–7 Because PDX1(S61) phosphorylation promotes β cell proliferation and insulin expression in knockin animals on a normal diet (ND), we examined their circulating insulin levels under a WD. The WD caused elevated blood insulin levels in wild-type mice versus the ND (Figures 7A and 7B). This elevated circulating insulin level was further manifested in the WD-fed Pdx1^S61E/S61E animals but was lower in the WD-fed Pdx1^S61A/S61A mice (Figures 7A and 7B). Similar to the ND-fed animals, islet size was larger in WD-fed Pdx1^S61E/S61E mice than in WD-fed wild-type mice, while that of islets from WD-fed Pdx1^S61A/S61A animals was smaller (Figures 7C–7E). Immunohistochemistry (IHC) staining of PCNA showed that differences in pancreatic islet sizes among different animal groups were correlated with differential β cell proliferation as judged by PNCA staining (Figures 7F and 7G). Notably, although Pdx1^S61E/S61E mice on the WD displayed moderately higher fasting blood glucose levels, they exhibited slightly better glucose tolerance than the other two animal groups during the study period (Figures 7H and 7I). No statistically significant difference was observed in insulin sensitivity among all three study groups (Figure 7J). While Pdx1^S61A/S61A animals exhibited reduced PDX1 and GLUT2 protein expression, the maturity of their β cells appeared to be maintained as judged by MAFA staining (Figure S7). Collectively, these observations suggest that Pdx1(S61) phosphorylation promotes hyperinsulinemia, aggravating metabolic dysfunctions caused by WD.

Figure 7. PDX1(S61) phosphorylation promotes Western diet-induced hyperinsulinemia in mice.

(A) Pdx1^S61E/S61E knockin mice fed a Western diet (WD) exhibit elevated serum insulin levels. Serum insulin was measured by ELISA in Pdx1^+/+, Pdx1^S61A/S61A, and Pdx1^S61E/S61E fed a WD for 28 weeks.

(B) Pdx1^S61E/S61E knockin mice fed a WD exhibit elevated serum C-peptide levels. Serum C-peptide was measured by ELISA in Pdx1^+/+, Pdx1^S61A/S61A, and Pdx1^S61E/S61E fed a WD for 33 weeks.

(C and D) PDX1(S61) phosphorylation leads to increased islet β cell area under WD conditions. IHC staining of insulin protein in pancreatic tissues from Pdx1^+/+, Pdx1^S61A/S61A, and Pdx1^S61E/S61E knockin mice fed a WD is shown. Ins⁺ areas were quantified using ImageJ software. Pdx1^+/+ (n = 4), Pdx1^S61A/S61A (n = 4), and Pdx1^S61E/S61E (n = 4). At least 5 islets were included in each mouse. Scale bar, 100 μm.

(E) The whole slides were scanned to get the image of the whole pancreas. The entire Ins+ area and the whole pancreas area were calculated using ImageJ. 3 slides were used for each mouse, with 4 mice for each genotype group.

(F and G) PDX1(S61) phosphorylation promotes pancreatic β cell proliferation under WD conditions. IF analysis of insulin (red) and PCNA (green) in the pancreas of Pdx1^+/+, Pdx1^S61A/S61A, and Pdx1^S61E/S61E fed a WD for 36 weeks was performed. The boxed areas were enlarged to show details. The white arrows label the cells with positive PCNA staining in the nucleus (F). Scale bar, 100 μm. Proliferation was calculated by the percentage of Ins⁺PCNA⁺ cells/Ins⁺ cells per mouse in each group. ≥900 Ins⁺ cells per mouse were counted.

(H) Fasting glucose analysis of knockin mice on a WD. The 12 h fasting blood glucose level was monitored every 4 weeks in Pdx1^+/+, Pdx1^S61A/S61A, and Pdx1^S61E/S61E mice fed a WD.

(I) Glucose tolerance analysis of knockin mice under WD conditions. Intraperitoneal glucose tolerance tests were performed in Pdx1^+/+ (n = 14), Pdx1^S61A/S61A (n = 15), and Pdx1^S61E/S61E (n = 15) male mice fed a WD at 30 weeks of age.

(J) Insulin sensitivity analysis of knockin mice under WD conditions. Intraperitoneal insulin tolerance tests were performed in Pdx1^+/+ (n = 14), Pdx1^S61A/S61A (n = 15), and Pdx1^S61E/S61E (n = 15) male mice fed a WD at 32 weeks of age.

All results are presented as the mean ± SEM. Representative results are shown. ***p < 0.001, **p < 0.01, and *p < 0.05. An unpaired Student’s t test for two groups was used.

DISCUSSION

PDX1 is a critical transcription factor that regulates insulin expression in pancreatic β cells in response to circulating glucose levels. Glucose-mediated regulation of PDX1 is complex, which can be explained in part by the phosphorylation of PDX1 by GSK3.^50,51 When glucose levels are low, GSK3 phosphorylates PDX1.^49–51 These phosphorylation events trigger degradation and turnover of the PDX1 protein. In contrast, when glucose levels are high, GSK3-mediated phosphorylation of PDX1 is reduced, allowing accumulation of PDX1 protein. PDX1 has also been reported to be a phosphorylation target for ERK, which is important for glucose response.⁵² Thus, PDX1 is regulated by multiple signaling inputs in response to glucose levels. We previously found that both aa and glucose regulate PDX1 protein expression and activity in an mTORC1-dependent manner.^41,42 In this study, we showed that mechanistically, mTORC1 interacts with PDX1 and phosphorylates PDX1 at S61 within its N-terminal transactivation domain. This post-translational modification is dependent on the levels of aa and glucose, which are sensitive to the mTORC1 inhibitor Rap. Importantly, S61 phosphorylation enhances PDX1 protein levels and its transcriptional activity. Our findings elucidated a molecular mechanism by which mTORC1 integrates aa and glucose signaling to regulate the key pancreatic β cell transcription factor PDX1. This phenomenon adds an important layer of regulation to nutrient-dependent control of insulin gene expression and β cell growth and function.

S61 is located within the N-terminal transactivation domain that synergistically activates enhancer-mediated insulin transcription in conjunction with other transcriptional coactivators.⁵³ Interestingly, several missense germline mutations, including C18R, P33T, Q59L, and D76N, occur within this subdomain. These single aa changes have been associated with an increased risk of developing T2D and multiple types of monogenic diabetes, including MODY and neonatal diabetes.^15–17 For example, in a family carrying an autosomal-dominant Q59L mutation, circulating insulin levels progressively decrease with age, giving rise to a late-onset form of T2D. We found that the Q59L mutation interferes with mTORC1 phosphorylation of S61 in vitro and markedly reduces S61 phosphorylation in MIN6 cells, with concomitantly diminished transactivation activity toward the Ins2 promoter. Consistently, Pdx1^S61A/S61A mutant mice displayed reduced β cell mass and hypoinsulinemia. These findings suggest a potential mechanism by which this monogenic mutation affects insulin production and suggest that conditions interfering with S61 phosphorylation can diminish circulating insulin levels and increase the risk of developing T2D.

In the setting of obesity, peripheral insulin resistance develops, impairing glucose uptake in target tissues. To compensate, pancreatic β cells increase insulin secretion, leading to hyperinsulinemia.³ Excess insulin in turn promotes fat storage in adipose and liver tissues by enhancing lipogenesis and inhibiting lipolysis, further contributing to obesity and hepatic steatosis.³ The mTOR pathway acts as a nutrient sensor, responding to nutrient abundance, particularly the levels of aa and glucose. Overnutrition or excess diet intake, a major risk factor for obesity and metabolic syndromes, leads to β cell mTORC1 activation. Obese patients are known to have elevated circulating aa (particularly the branched-chain aa V, L, and I)⁵⁴ that can synergize with high glucose to exacerbate mTORC1 activity in β cells. Sustained mTORC1 hyperactivation plays a crucial role in the development of hyperinsulinemia, a hallmark of obesity. It enhances insulin expression and biosynthesis. The resulting hyperinsulinemia further exacerbates obesity by shifting the metabolic balance toward tissue fat accumulation, contributing to the progression of obesity and liver steatosis.

The Pdx1^S61E/S61E mice showed worsened fasting blood glucose compared to the wild-type and Pdx1^S61A/S61A animals. However, they appear to have normal or slightly better glucose tolerance and insulin sensitivity. One possible explanation is compensatory hyperinsulinemia or masked insulin resistance, which is heightened in β cell response to early insulin resistance.⁵⁵ In this scenario, slightly better glucose tolerance in the glucose tolerance test (GTT)/insulin tolerance test (ITT) assays might be maintained by elevated insulin secretion, even if subtle insulin resistance exists. Another possible explanation is altered hepatic glucose production.⁵⁶ For example, in mice lacking insulin receptors only in the liver (LIRKO mice), fasting glucose is high due to unchecked gluconeogenesis, but peripheral tissues may still be insulin sensitive. It is also plausible that both mechanisms could be involved.

In a WD-induced obesity model, Pdx1^S61E/S61E mice showed accelerated body weight gain despite similar food intake, increased inguinal WAT (iWAT) mass and adipocyte size, greater liver weight gain, and more severe hepatic steatosis versus control wild-type mice. These mice also exhibited enhanced β cell proliferation, pancreatic islet mass, and circulating insulin levels. In contrast, Pdx1^S61A/S61A mice exhibited partial resistance to WD-induced iWAT gain and blunted increases in β cell proliferation, pancreatic islet mass, and blood insulin levels versus WD-fed wild-type mice. Our observations suggest that PDX1(S61) phosphorylation is a crucial effector of β cell mTOR signaling, contributing to hyperinsulinemia and diet-induced metabolic dysfunctions. Our findings enhance the understanding of mTOR signaling in pancreatic β cells, which may inform the development of novel therapeutic strategies to normalize circulating insulin levels, disrupt the vicious cycle of hyperinsulinemia and insulin resistance characteristic of obesity, and potentially mitigate the development of liver steatosis.

Limitations of the study

How PDX1 phosphorylation changes dynamically during development, in response to dietary changes, fasting/feeding cycles, and under pathological conditions, remains an important and unanswered question. In this study, we were unable to address it due to time constraints and the inability of the current phospho-PDX1 antibody to work in mouse tissues. Future efforts will focus on developing better reagents that are compatible with these materials. We did not note a significant difference between male and female animals, suggesting sex does not play a major role in this context. However, only one mouse strain was used. The findings reported herein will also require further confirmation in other model systems, especially those more closely reflecting human physiology. While the current models offer valuable mechanistic understanding, they may not fully capture the complexity of human biology. Therefore, future studies will aim to replicate and extend these results in more physiologically relevant systems (e.g., human tissues or organoid models) to better assess their translational potential and clinical relevance.

RESOURCE AVAILABILITY

Lead contact

Requests for further information and resources should be directed to and will be fulfilled by the lead contact, Prof. X.F. Steven Zheng (zhengst@cinj.rutgers.edu).

Materials availability

All unique reagents generated from this study are available from the lead contact with a completed materials transfer agreement.

STAR★METHODS

EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS

Animal

All animal care and treatment were carried out in compliance with Rutgers University Institutional Animal Care and Use Committee (IA-CUC) guidelines. Pdx1 S61A and S61E knock-in mice were made using CRISPR-Cas9. C57BL/6J embryos were microinjected with a mixture containing High Fidelity Cas9 protein (IDT), an sgRNA (MilliporeSigma) and ssODN (IDT) which contained homology arms and either the S61A or S61E mutation, separately. The sgRNA used was ACTGGAGCAGGGAAGTCCTCCGG. The S61A donor oligo sequence was GCCGGGTCGTCGGAGGCGAGCGGGGGCACTTCGTATGGGGAGATGTCGGGAGGAgcTCCCTGCTCCAGTGATCCCAGCGAGCTTGTAAACTGGGGTGGCGGCGGA (S61A sequence change in lower case bold), and the S61E donor oligo sequence was GCCGGGTCGTCGGAGGCGAGCGGGGGCACTTCGTATGGGGAGATGTCGGGAGGctcTCCCTGCTCCAGTGATCCCAGCGAGCTTGTAAACTGGGGTGGCGGCGGA (S61E sequence change in lower case bold). Founders were screened by PCR and digested with restriction enzyme SacI (NEB) for S61A and MwoI (NEB) for S61E then NGS sequencing (Azenta Life Sciences) or Sanger sequencing to confirm the changes. Confirmed founders were bred and progeny were screened using PCR primers ACCATGAACAGTGAGGAGCAGTAC and CACTGGCCTTTCCACGCGTGAG and presence of restriction site or by Sanger sequencing of PCR products. Wild type and homozygous alleles mice were used in the next experiment. Both male and female mice were used in this study.

Cell lines

Low passage mouse insulinoma cell line MIN6 cells were obtained from Dr. Sally Radovick (Rutgers Robert Wood Johnson Medical School). MIN6 cells were routinely maintained in Dulbecco’s modified medium (DMEM, Gibco, 10566-016) with 25mM glucose, supplemented with 15% FBS, and 50 μmol/L 2-mercaptoethanol and at passages less than 30 in a humidified atmosphere with 5% CO₂ at 37°C. Cell lines are authenticated by short tandem repeats (STR) and tested for mycoplasma using Thermo Fisher Mycoplasma Detection Kit.

METHOD DETAILS

RNA extraction, reverse transcription and quantitative PCR

Total RNA from MIN6 cells or tissues were extracted using TRIzol reagent (Thermo Scientific) and RNeasy Mini Kit (Qiagen) according to manufacturer’s instruction. cDNA synthesis was conducted using TaqMan Reverse Transcription Reagents (Applied Biosystems) per manufacturer’s instructions. Quantitative PCR was performed with SYBR Green PCR master mix (Applied Biosystems) using ViiA 7 Real-Time PCR system (Applied Biosystems). Rn18S or β-actin was used as the internal control. Data were normalized by using the average Ct value of the internal control and further analyzed by the 2−ΔΔCt method.

Plasmid DNA and siRNA transfection

Myc-DDK tagged mouse Pdx1 plasmid was purchased from Origene (MR227421). Mouse Pdx1 S61A and S61E mutants were generated using Phusion Site-Directed Mutagenesis Kit (Thermo Scientific, F541). siRNAs targeting mouse Pdx1 were either purchased or synthesized from Sigma. Transient transfections were performed using Lipofectamine 3000 or RNAi Max (Invitrogen) according to the manufacturer’s instructions. To transfect MIN6 cells using Lipofectamine 3000, seed cells one day prior to transfection to reach approximately 70–80% confluency. Prepare two solutions: one with plasmid DNA and P3000 reagent diluted in Opti-MEM, and another with Lipofectamine 3000 also diluted in Opti-MEM. Mix the two solutions, incubate for 10 min at room temperature to allow complex formation, and then add the mixture dropwise to the cells in antibiotic-free medium. After 6-8h, replace the transfection medium with fresh complete growth medium. Gene expression can typically be assessed 24-48h post-transfection. Grow MIN6 cells so that they are healthy and in the exponential growth phase, and seed them such that they reach ~60–80% confluency at the time of transfection. Dilute your siRNA (or miRNA mimic/inhibitor) and in a separate tube dilute RNAiMAX in the same medium. Combine the diluted RNA and the diluted reagent gently and incubate ~5–20 min at room temperature to allow complexes to form, and then add the complexes to the cells. After transfection, incubate the cells for 24–72 h before assessing knockdown of target mRNA or protein.

Cell culture and treatments

MIN6 were maintained in DMEM (GIBCO, 10566-016) with 25mM glucose, supplemented with 15% FBS, 100 U/mL penicillin, 100 μg/mL streptomycin and 50 μmol/L 2-mercaptoethanol. Cells in different assays typically grown to 80% confluence unless otherwise indicated. AA starvation and restimulation were conducted as described previously.⁵⁹ Briefly, cells were rinsed twice with PBS and incubated in RPMI-1640 without AA supplemented with high glucose (GIBCO, A2494001) and 15% dialyzed FBS (GIBCO, A3382001) for indicated time. 1×AA mixture in the above medium was added for indicated times. The AA mixtures are diluted from MEM Amino Acids Solution (50×)(#11130-051), MEM non-Essential Amino Acids Solution (100×)(#11140-050) and GlutaMAX Supplement (100×)(#35050-061). The final AA mixture composition and concentration in the culture medium are L-Arginine hydrochloride (0.599mM), L-Cystine (0.1mM), L-Histidine hydrochloride-H2O (0.2mM), L-Isoleucine (0.4mM), L-Leucine (0.4mM), L-Lysine hydrochloride (0.396mM), L-Methionine (0.101mM), L-Phenylalanine (0.2mM), L-Threonine (0.4mM), L-Tryptophan (0.05mM), L-Tyrosine (0.199mM), L-Valine (0.4mM), Glycine (0.1mM), L-Alanine (0.1mM), L-Asparagine (0.1mM), L-Aspartic acid (0.1mM), L-Glutamic Acid (0.1mM), L-Proline (0.1mM), L-Serine (0.1mM), L-alanyl-L-glutamine (4mM). Cell samples were analyzed by quantitative PCR, IF or Western Blot. 100% glucose is 4.5g/L in RPMI-1640.

Duolink proximal ligation assay (PLA)

Duolink assay was carried out based on the manufacturer’s instruction. MIN6 cells with different treatments were cultured on Chamber-slides (154526, Thermo Fisher Scientific), and fixed using 4% paraformaldehyde after cell culture. Cells were blocked and incubated with primary antibodies at 4°C overnight. Antibodies against mTOR and Pdx1 from different species were utilized. A single antibody was used as a negative control. All Chamber-slides were mounted using Mounting Media with DAPI (0100-20, SouthernBiotech). Images were visualized and captured using Nikon A1R laser scanning confocal microscope.

Immunoblot and co-immunoprecipitation assay

Cells were collected and resuspended in lysis buffer (CelLytic, Sigma-Aldrich) supplemented with protease inhibitor cocktail (cOmplete, Roche) and phosphatase inhibitor (PhosSTOP, Sigma-Aldrich) for 30 min on ice, followed by centrifugation at 16,000 × g for 10 min to pellet cell debris. The supernatant was collected as a whole cell lysate (WCL) and protein concentration was determined with BCA assay. For co-immunoprecipitation assay, cell lysates were incubated with primary antibody overnight at 4°C, and then with protein A/G plus agarose beads for 1 h at 4°C. Immunoprecipitants were collected by centrifugation and washed four times with lysis buffer, and then beads were re-suspended in electrophoresis sample buffer. For Western Immunoblot, protein samples were mixed with electrophoresis sample buffer and heated at 95°C for 10 min. The samples were separated by SDS-polyacrylamide gel electrophoresis (PAGE) and were transferred onto PVDF membranes (Bio-Rad). Membranes were blocked with 5% skim milk for 1 h at room temperature and incubated with primary antibodies overnight at 4°C. Antibodies were diluted according to the manufacturer’s instructions. Membranes were further incubated with HRP-conjugated secondary antibodies, followed with enhanced chemiluminescence detection reagents (NEL105001EA, PerkinElmer). Protein levels were analyzed using the ChemiDoc XRS+ system (Bio-Rad) and ImageJ v.6.0.1 software.

Immunofluorescence assay

MIN6 cells or frozen tissue slides were fixed with 4% paraformaldehyde and blocked in blocking buffer (PBS, 0.25% Triton X-100, 5% BSA, 10% goat serum). Following blocking, the specimens were incubated with the specified primary antibodies overnight at 4°C. Subsequently, cells or tissue slides were exposed to Alexa Fluor 488 or 561 secondary antibodies for 1 h at room temperature. Dilutions for both primary and secondary antibodies adhered to the manufacturer’s instructions. Cells or tissue slides were subsequently incubated with DAPI (4083S, Cell Signaling Technology) for 15 min and mounted with fluorescence mounting medium (0100-20, SouthernBiotech). Immunofluorescence patterns were visualized and captured using Nikon A1R laser scanning confocal microscope.

Histology and immunohistochemistry

Mouse tissues were fixed in 10% neutral buffered formalin solution (A5472, Sigma-Aldrich) for 24 h. Fixed tissues were transferred to 70% ethanol and then embedded in paraffin. H&E staining was performed as described previously.⁶⁰ For IHC, detection was performed with the Elite ABC Kit (PK-6100, Vector Laboratories) and DAB Substrate (8059, Cell Signaling Technology). Epredia Cytoseal Mounting medium (831016, Thermo Fisher Scientific) was used after dehydration. Pancreas, liver, and adipose tissues were dissected and processed uniformly. 8 mm thick sections were cut using a cryostat. Islets areas were determined by ImageJ software as described.⁶¹

Islet isolation

Islets were isolated from 6 months old male mice as previously described.^62,63 Briefly, three critical steps are included. First, clamping of the common bile duct near the liver and injecting collagenase XI at the ampulla of Vater. Second, enzymatic digestion and mechanical separation of the exocrine pancreas. Third, Histopaque-1077 gradient purification step. After density gradient centrifuge and HBSS buffer wash, islets were resuspended in RPMI-1640 culture media. The islets were hand-picked under microscope and transferred to RPMI1640 without AAs supplemented with high glucose (GIBCO, A2494001), 15% dialyzed FBS (GIBCO, A3382001) and 1% Penicillin-Streptomycin. After 4h AA starvation, the islets were readded into full AAs in the absence or presence of rapamycin for 4h. The islets were collected for protein extraction or RNA extraction.

Mouse metabolic characterization

Metabolic studies were performed according to the recommendations of the Mouse Metabolic Phenotyping Center.⁶⁴ Body weight and blood glucose of Pdx1 mutant and control mice were monitored every 4 weeks. For glucose tolerance test (GTT), mice were fasted for 12 h (overnight) and blood glucose was evaluated with a glucometer. Mice were then weighed and injected with glucose intraperitoneally at 2 mg/g body weight. Blood glucose level was measured every 30 min. For the insulin tolerance test (ITT), mice were fasted for 4 h (morning fast) and then injected intraperitoneally with insulin at 0.75 U/kg body weight. Blood glucose levels were measured at indicated times. For in vivo insulin and C-peptide measurements, mice were fasted for 2h in the morning, blood was collected from the tail vein and centrifuged, serum was collected and measured for insulin or C-peptide concentration using the Insulin ELISA kit (10-1247-01, Mercodia) or C-peptide ELISA kit (90050, Crystal Chem). For circulating glucagon measurement, serum was collected from the tail vein from mice fasted overnight. Glucagon was measured using Rodent glucagon ELISA kit (10-1281-01, Mercodia). Fasting chemistries were measured using reagents from Thermofisher (triglycerides, TR22421; cholesterol, TR13421).

(WD)-induced T2D mouse model

A type 2 diabetes (T2D) murine model induced by a Western diet (WD) was established following the described protocol.⁶⁵ Eight-week-old male mice were provided unrestricted access to either a standard chow (TD.2018S, ENVIGO) with 6.2% fat content or a Western diet (TD.08811, ENVIGO) with 45% kcal from fat (21% MF, 2% SBO) for a duration of 24 weeks. Glucose and insulin tolerance tests were performed to confirm the development of hyperglycemia induced by the Western diet. At the conclusion of the 34-week period, mice were euthanized, and tissues were collected for further analysis.

RNA-seq and analysis of differentially expressed genes

RNA was harvested using TRIzol reagent (Thermo Scientific) and RNeasy Mini Kit (Qiagen) according to manufacturer’s instruction. RNA library construction, RNA quality assessment, and sequencing (Illumina, 150bp paired end) were performed by Azenta. Raw transcriptome-sequencing reads were aligned to Mus musculus reference genome (GRCm38.p6) using subread-aligner (subread 2.0.1). Differentially expressed genes were determined using limma 3.46.0 with a combination of fold change ≥2 and false discovery rate (FDR, by Benjamini-Hochberg algorithm) adjusted P-value ≤0.05 (n = 3). Data visualization was plotted by SRplot.⁶⁶

Chromatin immunoprecipitation (ChIP) assay and ChIP-seq

ChIP was performed according to manufacturer’s instruction (9003, Cell Signaling Technology). Briefly, MIN6 cells were suspended in 3.7% formaldehyde at room temperature for 10 min and then blocked by glycine to a final concentration of 0.25M for 5 min on ice. After centrifugation at 2000 × g for 5 min, cell pellets were resuspended in ice-cold lysis Buffer A for 10 min. After centrifugation, nuclei pellets were resuspended in ice-cold Buffer B. Then micrococcal nuclease was added to digest DNA to length of approximately 150–900 bp at 37°C for 20 min. Sonication with Diagenode’s Bioruptor (B01020001, Diagenode) was used to break nuclear membrane to release chromatin-DNA fragments, which were collected in the supernatant after centrifugation at 9400 × g for 10 min at 4°C. The supernatant was incubated with anti-Pdx1 (5679, Cell Signaling Technology) at 4°C overnight and then with Protein G Magnetic Beads (9006, Cell Signaling Technology) for another 1 h at 4°C. After 3 times high salt/low salt wash, chromatin was eluted from Magnetic Beads using elution buffer at 65°C for 30min with gentle vortexing. The eluted supernatant was reversed cross-linked by adding 6 μL 5M NaCl and 2 μL Proteinase K and incubated 2 h at 65°C. Samples immediately proceeded to DNA purification using Spin Columns. The purified DNA fragment was used for sequencing or RT-PCR at the next step.

Immunoprecipitated and purified DNA was prepared as a standard Illumina library using NEBNext Ultra II DNA Library Prep Kit according to the manufacturer’s protocol by Novogene. Samples were quantified with Qubit 2.0, and the insertion size of the library is detected with NGS3K. Sequencing was performed according to the concentration and the demand of data amount on Illumina NovaSeq platform. Raw reads from PDX1 ChIP-seq were trimmed to remove low quality bases and potential adapter sequences. Next, the clean reads after trimming were mapped to the Mus musculus reference genome (GRCm38.p6) using BWA (0.7.12)⁶⁷ considering the small fragment size of ChIP-seq, and duplicates were removed using SAMBLASTER.⁶⁸ MACS2 (2.1.0)⁶⁹ was used for PDX1 peak calling and then peaks were annotated by HOMER for their locations relative to nearby genes. Peak visualization was performed using IGV software.⁷⁰ MEME (4.10.2)⁷¹ software was used to detect significant motif sequence in the peak. GO and KEGG enrichment analysis of overlapping genes were performed using Goseq/topGO Bioconductor (2.13). Data visualization was plotted by SRplot.⁶⁶

QUANTIFICATION AND STATISTICAL ANALYSIS

Statistical comparisons between groups were conducted using the unpaired two-tailed Student’s t test via GraphPad Prism v.10 software or Excel (Microsoft Office), unless otherwise specified. The presentation of data utilized means ± standard error of the mean (S.E.M), with statistical significance set at p < 0.05. Sample sizes were predetermined based on established norms for the experiment and are delineated for each instance. Rigorous randomization procedures were adhered to in the execution of experiments. The investigators were not blinded during the experiments and outcome assessment, unless explicitly stated otherwise. Each experiment consisted of biological replicates and was repeated 3 times at least. The statistical parameters for analyses are detailed in the figure legends.

Supplementary Material

Supplemental information can be found online at https://doi.org/10.1016/j.celrep.2025.116811.

KEY RESOURCES TABLE

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Rabbit monoclonal anti-Flag (D6W5B)	Cell signaling Technology	#14793; RRID: AB_2572291
Rabbit monoclonal anti-Raptor (24C12)	Cell signaling Technology	#2280; RRID: AB_561245
Rabbit monoclonal anti-Rictor (D16H9)	Cell signaling Technology	#9476; RRID: AB_10612959
Rabbit monoclonal anti-mTOR (7C10)	Cell signaling Technology	#2983; RRID: AB_2105622
Rabbit polyclonal anti-mTOR	This paper	N/A
Rabbit monoclonal anti-p70 S6K	Cell signaling Technology	#2708; RRID: AB_390722
Rabbit monoclonal anti-phospho-p70 S6K (Thr389)	Cell signaling Technology	#9234; RRID: AB_2269803
Rabbit monoclonal anti-phospho-Akt (Ser473)	Cell signaling Technology	#4060; RRID: AB_2315049
Mouse monoclonal anti-Akt (Pan)	Cell signaling Technology	#2920; RRID: AB_1147620
Rabbit monoclonal anti-β-Actin	Cell signaling Technology	#8457; RRID: AB_10950489
Rabbit monoclonal anti-PCNA	Cell signaling Technology	#13110; RRID: AB_2636979
Rabbit monoclonal anti-Ki67	Cell signaling Technology	#9129; RRID: AB_2687446
Mouse monoclonal anti-Ki67	Cell signaling Technology	#9449; RRID: AB_2797703
Rabbit monoclonal anti-Insulin	Abcam	#ab181547; RRID: AB_2716761
Mouse monoclonal anti-Insulin Clone No.:4B6A7	Proteintech	#66198-1-Ig
Mouse monoclonal anti-Insulin	Cell signaling Technology	#8138; RRID: AB_10949314
Rabbit monoclonal anti-Glucagon (EP3070)	Abcam	#ab92517; RRID: AB_10561971
Mouse monoclonal anti-Glucagon (K79bB10)	Abcam	#ab10988; RRID: AB_297642
Rabbit monoclonal anti-Pdx1	Cell signaling Technology	#5679; RRID: AB_10706174
Rabbit polyclonal anti-Pdx1	This paper	N/A
Rabbit polyclonal anti-phospho-Pdx1 (Ser61)	This paper	N/A
Mouse monoclonal anti-KLF11	Sigma-Aldrich	#SAB1404583; RRID: AB_10740381
Rabbit monoclonal anti-MAFA	Cell signaling Technology	#79737; RRID: AB_2799938
Rabbit polyclonal anti-GLUT2	Abcam	#ab54460; RRID: AB_880234
Biological Samples
Mouse whole blood	This paper	N/A
Mouse serum	This paper	N/A
Chemicals, Peptides, and Recombinant Proteins
Rapamycin	LC Laboratories	#53123-88-9
cOmplete Protease Inhibitor Cocktail	Roche	#11836145001
PhosSTOP™	Sigma-Aldrich	#P9620
Paraformaldehyde	Sigma-Aldrich	#P6148
D- (+)-Glucose	Sigma Aldrich	#G7021
Human Insulin solution	Novo Nordisk	#0169-1833-11
Critical Commercial Assays
Mouse Insulin ELISA Kit	Mercodia	#10-1247-01; RRID: AB_2783837
Mouse C-peptide ELISA Kit	Crystal Chem	#90050
Rodent Glucagon ELISA Kit	Mercodia	#10-1281-01; RRID: AB_2783839
Triglycerides Reagent	Thermo Scientific	#TR22421;
Total Cholesterol Reagent	Thermo Scientific	#TR13421;
Clarity BG1000 Blood Glucose Meter and Strips	Clarity Diagnostics	#75840-798-BX
Stellaris FISH Probes, Mouse Ins 1/2 with Quasar 570 Dye	LGC Biosearch tech	#VSMF-3611-5
SimpleChIP® Enzymatic Chromatin IP Kit	Cell signaling Technology	#9003
RNeasy Mini Kit	Qiagen	#74106
Lipofectamine™ 3000 Transfection Reagent	Life Technologies	#L3000015
Lipofectamine™ RNAiMAX Transfection Reagent	Life Technologies	#13778075
TaqMan™ Reverse Transcription Reagent	Life Technologies	#N8080234
PowerUp™ SYBR™ Green Master Mix for qPCR	Applied Biosystem	#A25742
Dual-Glo Luciferase Assay System	Promega	#E2940
Deposited Data
Raw and processed sequencing data for ChIP-seq	This paper	GSE278789
Raw and processed sequencing data for RNA-seq	This paper	GSE278790
Experimental Models: Cell Lines
Mouse: MIN6	Dr. Radovick lab	RRID: CVCL_0431
Experimental Models: Organisms/Strains
Mouse Pdx1S61A/S61A, C57BL6	This paper	N/A
Mouse Pdx1S61E/S61E, C57BL6	This paper	N/A
Oligonucleotides
Mouse Pdx1 siRNA #1, GGGAACUUAACCUAGGCGUTT	Babu et al.57	N/A
Mouse Pdx1 siRNA #2, GGUUAAGAAUAACGAAAGGUU	Babu et al.57	N/A
Mouse Pdx1 siRNA #3 Sigma esiRNA	Sigma Aldrich	#EMU046321
Recombinant DNA
pCMV6-mPdx1-Myc-DDK	OriGene	#MR227421;
pCMV6-mPdx1S61A-Myc-DDK	This paper	N/A
pCMV6-mPdx1S61E-Myc-DDK	This paper	N/A
pCMV6-mPdx1-tGFP	OriGene	#MG227421;
pGL4.84	Promega	#E752A
pGL4.84-Ins2 promoter	Zhang et al.41	N/A
Software and Algorithms
ImageJ	NIH	https://imagej.nih.gov/ij/
Prism	GraphPad	https://www.graphpad.com/scientific-software/prism/
Photoshop	Adobe	https://www.adobe.com/products/photoshop.html
Adobe Illustrator	Adobe	https://www.adobe.com/products/illustrator.html
Integrative Genomics Viewer (IGV)	Robinson et al.58	https://software.broadinstitute.org/software/igv/
Other
MEM Amino Acids Solution (50×)	Thermo Fisher Scientific	#11130-051
MEM Non-Essential Amino Acids Solution (100×)	Thermo Fisher Scientific	#11140-050
GlutaMAX Supplement	Thermo Fisher Scientific	#35050-061
Glucose Solution	Thermo Fisher Scientific	#A24940-01
Dialyzed FBS	Thermo Fisher Scientific	#26400-044
FBS	Biowest	#S162H
RPMI 1640 Medium Modified w/o L-Glutamine, w/o Amino acids, Glucose	USBiological	#R9010-01
DMEM, high glucose, GlutaMAX Supplement	Thermo Fisher Scientific	#10566-016
RPMI 1640 Medium, GlutaMAX Supplement	Thermo Fisher Scientific	#61870-036
TRIzol LS Reagent	Life Technologies	#10296028
Antifade mountant with DAPI	Life Technologies	#P36931

Highlights.

• Amino acids regulate PDX1 functions in pancreatic β cells

• mTORC1 phosphorylates PDX1 at S61, promoting β cell proliferation and insulin expression

• Monogenic diabetes mutation disrupts S61 phosphorylation

• S61 phosphorylation drives hyperinsulinemia, obesity, and liver steatosis under a Western diet

Data and code availability

• PDX1 ChIP-seq data have been deposited at the GEO repository and are available under accession number GEO: GSE278789.

• This paper does not report original code.

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