SOX9: dosage- and context-dependent functions in pancreatic development and β-cell biology

Abstract

SOX9 is a highly conserved transcription factor (TF) belonging to the SRY-related HMG-box (SOX) family and the high-mobility group (HMG) class of DNA-binding proteins. SOX9 integrates structural flexibility, DNA-dependent dimerization, and context-specific cofactor interactions to orchestrate organogenesis. In the pancreas, SOX9 acts as a dosage-sensitive gatekeeper: high expression in multipotent progenitors maintains proliferation and prevents premature endocrine differentiation, while its timely downregulation is a prerequisite for NEUROG3 (NGN3) induction and β-cell maturation. Genetic and clinical data from campomelic dysplasia and mouse models reveal that both haploinsufficiency and dominant-negative SOX9 variants disrupt pancreatic morphogenesis and endocrine formation, whereas inappropriate SOX9 reactivation in adult β-cells under metabolic or hypoxic stress drives dedifferentiation and diabetes-like phenotypes. Recent work further demonstrates that low-level SOX9 expression persists in mature β-cells, where it regulates alternative splicing and stress adaptation, underscoring its lifelong importance for β-cell function. Despite these insights, SOX9 has not emerged as a diabetes susceptibility gene in genome-wide association studies, suggesting that SOX9-related β-cell failure is primarily driven by rare, severe mutations and dysregulated expression rather than common variants. This review synthesizes current understanding of SOX9’s multifaceted roles across pancreatic development and adult β-cell biology, highlighting conserved mechanisms established through mouse genetics, species-specific considerations for translating findings to humans, and emerging opportunities for therapeutic intervention targeting SOX9-dependent pathways to preserve β-cell function and identity in diabetes.

Introduction

SOX9 is a highly conserved TF that integrates structural flexibility, DNA-dependent dimerization, and context-specific cofactor engagement to orchestrate multiple organogenic programs. Originally characterized for its essential roles in chondrogenesis and sex determination, SOX9 has since emerged as a dosage-sensitive master regulator of pancreatic development whose activity must be precisely modulated to maintain progenitor identity, enable endocrine differentiation, and preserve β-cell function throughout life [1, 2].

The central importance of SOX9 to human development is most dramatically evident in campomelic dysplasia, a severe multisystem disorder caused by heterozygous SOX9 mutations. In addition to skeletal and gonadal defects, these patients exhibit profound disruption of pancreatic architecture, reduced endocrine hormone expression, and impaired β-cell maturation, providing the first evidence that SOX9 insufficiency directly compromises human pancreatic organogenesis [3]. Mouse genetic studies have further demonstrated that SOX9 marks multipotent pancreatic progenitors, sustains their proliferation, and prevents premature endocrine differentiation; conditional SOX9 deletion results in progenitor depletion, pancreatic hypoplasia, and a near-complete block in β-cell formation [4–6].

A precision-regulated decline in SOX9 expression is required for NEUROG3 induction, which is essential for endocrine lineage commitment. Conversely, in mouse models, sustained SOX9 expression arrests pancreatic β-cell maturation, while early loss of SOX9 triggers premature differentiation and progenitor exhaustion [4]. These findings place SOX9 at the center of the transcriptional hierarchy that defines the pancreatic progenitor state, acting cooperatively with pancreatic and duodenal homeobox 1 (PDX1), forkhead box A2 (FOXA2), hepatocyte nuclear factor 6 (HNF6; encoded by ONECUT1), and hepatocyte nuclear factor 1 beta (HNF1B; encoded by TCF2) to establish lineage competence and control the timing of endocrine commitment [5, 6].

Although earlier studies showed that SOX9 is absent from adult endocrine cells [4, 7, 8], recent high-sensitivity transcriptomic studies have overturned this notion, revealing low but physiologically meaningful SOX9 expression in mature β-cells, where it regulates alternative splicing, stress adaptation, and insulin secretory competence [2]. In addition, diabetic stressors (hypoxia, inflammation, and hyperglycemia) have been shown to reactivate SOX9 in β cells in both mouse and human systems, promoting dedifferentiation and loss of mature β-cell identity [8–10]. Conversely, Sox9 haploinsufficiency produces maturity-onset diabetes of the young (MODY)-like glucose intolerance and reduced β-cell reserve in mice [5]. Together, these findings identify SOX9 as a bidirectional regulator of β-cell identity, where both insufficient and excessive expression destabilizes endocrine function.

Despite its profound functional impact, SOX9 has not emerged as a common diabetes risk gene in genome wide association studies (GWAS), highlighting a mechanistic model in which SOX9-mediated β-cell failure arises primarily from rare pathogenic variants, dosage imbalance, and stress-induced dysregulation rather than from common polymorphisms.

The aim of this review is to consolidate all existing studies that have examined SOX9 in pancreatic development and β-cell biology, and to organize this knowledge into a coherent developmental framework that tracks SOX9 expression stage-by-stage throughout human pancreatic development. By integrating findings from human fetal tissue analyses, mouse genetic models, stem-cell–derived systems, and β-cell stress paradigms, we provide a complete overview of the experimental strategies used to investigate SOX9 function. A major focus of this work is the direct comparison of human versus mouse SOX9 biology, highlighting conserved mechanisms as well as critical species-specific differences that limit interpretation of mouse data in a human context. Through this comparative analysis, we identify major gaps in the field, most notably the absence of true SOX9-null human models and outline the unresolved questions that must be addressed to fully understand how SOX9 governs pancreatic development and β-cell identity.

Structural and functional organization of the SOX9 protein

SOX9 is TF encoded on human chromosome 17q24.3, producing a 509 amino acid protein that functions as a master regulator of several developmental cell-fate programs, most prominently chondrogenesis, sex determination, neural crest specification, and endoderm-derived lineage commitment, including pancreatic, hepatic, intestinal, and pulmonary epithelial progenitors, where SOX9 plays a critical role in distal lung epithelial progenitor maintenance and branching morphogenesis [11–13]. Structurally, SOX9 contains three major functional regions: an N-terminal dimerization (DIM) domain, a central HMG box that mediates sequence-specific DNA binding, and two transactivation domains (TAD1 and TAD2) (Fig. 1A) [11, 14, 15]. One TAD is located in the middle (TAM) while the other is located at the extreme C-terminus (TAC). The PQA (proline, glutamine and alanine) domain is located between two transactivation domains.

Fig. 1.

Structural features of SOX9. A Domain architecture of SOX9. B Crystal structure of a HMG domain of SOX9 bound to DNA from human and mouse, represented in cartoon containing 3 helices (PDB ID: 4S2Q and 4EUW). C AlphaFold based predicted structure of SOX9 highlighting the prevalence of random coil beyond HMG domain

The central HMG domain (residues 101–180) is the critical functional core of SOX9, providing sequence-specific DNA recognition and imposing a characteristic 70–80° bend that enables chromatin remodeling and long-range enhancer activity [16–20]. The DIM domain (residues 61–83) enables SOX9 to form DNA-dependent homodimers, a property essential for cooperative enhancer binding and high-amplitude transcriptional activation [21–23]. At the opposite end of the protein, the C-terminal TAD1 (TAM) and TAD2 (TAC) domains function as flexible activation modules that mediate co-activator engagement (e.g., CBP/p300 and Mediator), supporting transcriptional activation downstream of SOX9 enhancer binding [14, 15]. The HMG domain remains the only region of SOX9 resolved at high structural resolution [16–19, 24], and Structural comparison of SOX9 HMG domain structures (complexed with cognate DNA) showed that both the mouse (4S2Q) and human (4EUW) SOX9 structures are highly similar (Fig. 1B).

By contrast, the N- and C-terminal regions have remained refractory to crystallographic or cryo-EM analysis owing to their intrinsic disorder, which confers substantial conformational flexibility [11, 14, 25]. Early biophysical studies using circular dichroism and limited proteolysis first demonstrated the absence of stable secondary structure in these regions, and computational predictors such as PONDR and IUPred consistently identify extensive disorder flanking the HMG box [26, 27]. AlphaFold modeling further supports this view, yielding high-confidence predictions only for the HMG domain and portions of the dimerization domain (DIM) region (Fig. 1C). This intrinsic structural plasticity is thought to facilitate SOX9’s interactions with a wide array of cofactors, including CBP/p300, SOX8/10, and β-catenin, and to enable context-dependent modulation of its regulatory activity [21]. Consistent with this flexibility, SOX9 displays the highest instability index among SOX family members, a property that likely reflects and supports its dynamic, interaction-rich functional behaviour [28]. Together, these features confer substantial conformational flexibility, enabling SOX9 to engage diverse cofactors and modulate its regulatory activity across distinct cellular contexts.

SOX9 binds both enhancers and promoters, yet the mechanisms governing these interactions are fundamentally different. At enhancers, SOX9 primarily engages DNA through the conserved SOX consensus motif (A/T A A/T C A A/T G), functioning either as a monomer or, when motifs are properly spaced and oriented, as a high-affinity cooperative dimer. This DNA-dependent homodimerization dramatically strengthens enhancer occupancy and is indispensable for robust transcriptional activation [21, 23, 26], and SOX9 is the best-characterized SOX family member whose high-output enhancer activity strictly depends on this dimeric mode. In chondrocytes, classical enhancer studies and genome-wide chromatin immunoprecipitation sequencing (ChIP-seq) have defined a motif-dependent Class II mode, in which SOX9 binds paired SOX motifs within distal, evolutionarily conserved enhancer clusters controlling COL2A1, COL9A2, COL11A2, and ACAN that often assemble into super-enhancer–like domains, and a promoter-associated, motif-independent Class I mode likely mediated by interactions with p300, RNA polymerase II, and the basal transcription machinery [29].

Outside the skeletal system, SOX9 directly regulates additional lineage-specific targets, including SOX10 in neural crest development [30] and PDX1 in endodermal and pancreatic progenitors [4], but the enhancer logic in these contexts remains far less defined. Although SOX9 binding to Sox10 regulatory elements has been demonstrated in other developmental settings [31] and SOX9-PDX1 co-occupancy at shared regulatory sites of pancreatic developmental genes has been reported in human embryonic stem cell (hESC)-derived pancreatic progenitors [6], comprehensive maps of SOX9-dependent enhancer architectures in developmental contexts outside chondrogenesis remain largely unresolved.

SOX9 expression dynamics during human pancreatic development

SOX9 displays a tightly regulated, stage-dependent expression pattern during human pancreatic development. It is initially expressed broadly across the foregut endoderm but becomes progressively restricted and enriched within the pancreatic epithelial progenitor/ductal compartment as pancreatic identity is established. SOX9 expression then diminishes as progenitors exit the multipotent state and commit to endocrine and exocrine lineages, mirroring the transition from endodermal competence to lineage restriction [3, 7].

Early pancreatic specification and initial SOX9 expression

Following gastrulation (~ post-conception days, PCD14–21), definitive endoderm is established (SOX17, FOXA2, CXCR4) and subsequently remodels during embryonic folding to form the primitive gut tube (~ PCD 25–27), which is regionalized into foregut, midgut, and hindgut domains, providing the anatomical framework for organ specification [32–38]. During the establishment of pancreatic identity (~ PCD 25–30), the posterior foregut endoderm is patterned through coordinated signals that repress Sonic Hedgehog (SHH) and permit activation of the pancreatic transcriptional program (Fig. 2Ai-ii). At ~ PCD 29–30, PDX1 expression first becomes detectable in the emerging dorsal and ventral pancreatic domains. At this developmental stage, SOX9 remains broadly distributed across the foregut endoderm, showing particularly prominent expression in the hepatic bud where it colocalizes with FOXA2. Within the presumptive pancreatic epithelium, SOX9 shows only limited overlap with PDX1, consistent with an early specification phase prior to its restriction to the pancreatic progenitor compartment (Fig. 2Bi) [4, 6, 34, 36].

SOX9 defines the multipotent pancreatic progenitor state

By ~ PCD 30–37, SOX9 expression becomes prominently associated with the dorsal and ventral pancreatic buds, marking an early progenitor state before lineage restriction. During this period, the pancreatic epithelium establishes a stabilized progenitor identity defined by co-expression of PDX1, SOX9, and NK6 homeobox 1 (NKX6.1), while markers of endocrine commitment such as NK2 homeobox 2 (NKX2.2) remain absent [3, 4, 34, 39]. This SOX9 + progenitor population maintains developmental plasticity, harboring the capacity to generate both ductal and endocrine lineages while remaining distinct from the emerging acinar compartment (Fig. 2Bii-iii). As branching morphogenesis initiates (~ PCD 37–40), SOX9 expression persists broadly across the developing epithelial network. The earliest evidence of spatial organization emerges with selective GATA binding protein 4 (GATA4) induction in peripheral tip cells, foreshadowing acinar specification, while SOX9 + trunk cells maintain progenitor characteristics (Fig. 2Ci) [34, 36, 40, 41].

SOX9 downregulation accompanies lineage commitment

By ~ PCD 49–52, clear spatial and molecular segregation emerges between differentiating lineages. Acinar commitment in the peripheral tip domain is marked by GATA4 and carboxypeptidase A1 (CPA1) expression accompanied by reduced SOX9 and loss of NKX6.1. In contrast, trunk cells retain robust SOX9, NKX6.1, and HNF1β expression, maintaining ductal and endocrine potential [42–44]. This molecular bifurcation between GATA4 + CPA1+ tip cells and SOX9 + NKX6.1 + trunk cells represents a critical transition point where SOX9 expression becomes spatially restricted to cells retaining progenitor identity (Fig. 2Cii) [4, 34, 45].

Fig. 2.

Schematic overview of key developmental milestones during human pancreatic organogenesis. A Early endoderm formation and foregut patterning. (Ai) Formation of the three germ layers during gastrulation establishes the definitive endoderm (DE), marked by SOX17, FOXA2, CXCR4 expression. (Aii) Foregut formation and regional patterning convert the endodermal sheet into a primitive gut tube. Followed by the emergence of the dorsal and ventral pancreatic buds from the posterior foregut, marked by a FOXA2⁺SOX17⁺SHH⁻ early pre-pancreatic signature. B Pancreatic specification and progenitor establishment. (Bi) Onset of pancreatic specification, with both buds initiating PDX1 expression within distinct transcriptional contexts (NR2F1-enriched dorsal domain; TBX3-enriched ventral domain. SOX9 is broadly expressed across the foregut, including hepatic and biliary lineages, but only weakly overlaps with PDX1 at this stage. (Bii) Early pancreatic progenitor specification marked by induction of NKX6.1 (dorsal-biased) and strong co-expression of PDX1, SOX9, and FOXA2. (Biii) Gut rotation brings the ventral bud into contact with the dorsal bud, culminating in their fusion into a single pancreatic primordium composed of multipotent progenitor cells (PDX1⁺FOXA2⁺SOX9⁺NKX6.1⁺). C Branching morphogenesis and lineage segregation. (Ci) Initiation of branching morphogenesis generates a network of primitive epithelial tubes that are only one to two cell layers thick, marking the earliest architectural scaffold of the developing gland. (Cii) Emergence of acinar lineage commitment in distal tip domains, marked by GATA4 and CPA1 expression, with reduced SOX9 and absence of NKX6.1. Trunk cells retain SOX9, NKX6.1, and HNF1β expression, preserving ductal–endocrine potential. (Ciii) Emergence of the endocrine compartment, with small clusters of fetal β-cells defined by NKX2.2, NKX6.1, and FOXA2 expression. Progressive loss of SOX9 in endocrine cells, with minimal overlap in insulin⁺ or glucagon⁺ cells by mid-gestation. By ~ 14 post-conception weeks, distinct exocrine (GATA4⁺CPA1⁺), ductal (SOX9⁺), and endocrine (NKX6.1⁺NKX2.2⁺FOXA2⁺INS⁺) compartments are established.

SOX9 during the endocrine transition (8–15 Weeks)

Between ~ 8 and 15 weeks post-conception, SOX9 shifts from a broadly expressed pancreatic epithelial progenitor marker to a spatially restricted ductal/progenitor marker as endocrine and exocrine compartments emerge (Fig. 2C). Around 8–10 weeks, SOX9 is progressively downregulated across much of the epithelium coincident with NEUROG3 induction and the appearance of the first insulin-positive cells [1, 3, 7, 34]. Notably, a small transient population (~ 5% of SOX9 + cells at ~ 8–9 weeks) co-expressing SOX9 and NEUROG3 has been identified, capturing the molecular transition between progenitor competence and endocrine commitment [7, 34]. This brief intermediate state suggests that SOX9 downregulation is tightly coupled to, and may be required for, endocrine lineage entry—a hypothesis supported by functional studies demonstrating that sustained SOX9 expression blocks endocrine differentiation.

By ~ 10–12 weeks, newly formed endocrine cells aggregate into nascent islet-like clusters and lose SOX9 expression, while SOX9 becomes increasingly restricted to the CK19 + ductal–progenitor epithelium [7, 46]. By ~ 11–15 weeks, SOX9 is confined to ductal/centroacinar regions and distal tip-associated progenitors that co-express pancreas transcription factor 1 subunit alpha (PTF1A) and PDX1, likely representing acinar-biased progenitors with residual plasticity [47, 48]. By ~ 14–15 weeks, compartmentalization is largely established, with SOX9 definitively excluded from the endocrine lineage while maintained in ductal structures (Fig. 2Ciii) [7, 34, 49].

SOX9 in pancreatic progenitor specification and maintenance

Having established the temporal dynamics of SOX9 expression during human pancreatic development (Fig. 2), we next summarize its functional roles and the molecular mechanisms through which it regulates progenitor identity and lineage allocation, and differentiation, as well as emerging rols in adult β-cell function and stress responses (Fig. 3).Understanding these mechanisms is essential for interpreting SOX9’s role in adult β-cell biology and its potential involvement in diabetes pathogenesis.

Fig. 3.

Functional roles of SOX9 in pancreatic development and β-cell function. SOX9 regulates pancreatic lineage specification, progenitor maintenance, and endocrine differentiation during development, and supports ductal epithelial identity, β-cell function, and stress responses in adulthood. Under pathological conditions, SOX9 reactivation is associated with β-cell dedifferentiation and dysfunction. The inner green circle represents the sequential developmental stages during pancreatic organogenesis, from foregut endoderm formation to mature β-cells. The outer red circle highlights the diverse functional roles of SOX9 in pancreatic development, β-cell homeostasis, and stress responses

SOX9 and PDX1: cooperative control of pancreatic identity

While PDX1 has long served as a central marker of early pancreatic progenitors, its expression is not restricted to the pancreatic epithelium and extends into the adjacent posterior foregut endoderm during early patterning [50–54]. In this context, SOX9 emerged as a more specific marker of multipotent pancreatic progenitors and a core component of the pancreatic transcriptional network, with transcriptomic profiling confirming that combined PDX1 and SOX9 expression distinguishes pancreatic progenitors from other endodermal lineages [4–6].

Although SOX9 and PDX1 are initially expressed independently, their relationship becomes functionally interdependent as development proceeds. SOX9 is required to sustain PDX1 expression, and loss of SOX9 causes progressive reduction in PDX1 levels, depletion of the SOX9⁺PDX1⁺ progenitor pool, and decreased PDX1⁺ cells in both epithelial progenitors and emerging endocrine lineages [4–6].This dependency is particularly evident beyond mid-gestation during the secondary transition and major wave of β-cell differentiation, underscoring SOX9–PDX1 coupling as essential for preserving progenitor identity and endocrine competence.

Genome-wide occupancy studies reveal complementary DNA-binding strategies: PDX1 enriches at distal enhancers while SOX9 shows marked preference for promoter-proximal regions, indicating distinct modes of transcriptional regulation [6]. Despite these differences, SOX9 and PDX1 co-occupy regulatory regions of key pancreatic developmental genes including PTF1A, paired box 6 (PAX6), and NEUROG3. Notably, some co-bound targets remain transcriptionally inactive in progenitors, suggesting that SOX9 and PDX1 may participate in repression of alternative endodermal programs, thereby reinforcing pancreatic identity. In human fetal pancreas, SOX9 and PDX1 are broadly co-expressed in early epithelial progenitors (8–11 weeks) but progressively diverge, with SOX9 becoming restricted to ductal/centroacinar compartments while PDX1 persists in differentiating endocrine and exocrine cells [7].

SOX9-mediated progenitor maintenance and proliferation

In mouse models, SOX9 supports progenitor proliferation, survival, and maintenance of a multipotent epithelial state. Loss of SOX9 results in premature activation of NEUROG3, transient expansion of endocrine-committed cells, and subsequent depletion of the progenitor pool, ultimately impairing pancreatic growth.

[4]. In mouse genetic models, pancreas-specific Sox9 loss results in progressive progenitor depletion, impaired organ growth, and severe hypoplasia, with the dorsal pancreatic domain showing particularly strong SOX9 dependence [4–6]. Early Sox9 deletion produces severe hypoplasia, while deletion after endocrine specification has minimal impact, defining a critical developmental window for SOX9 function [5].

Mechanistically, SOX9 operates through both Notch-dependent and Notch-independent programs. In mouse studies, SOX9 directly sustains Hes family bHLH transcription factor 1 (HES1) expression in PDX1⁺ progenitors positioned adjacent to NGN3⁺ endocrine precursors, consistent with Notch-mediated lateral inhibition [4, 53]. However, Sox9 deficiency produces substantially more severe phenotypes than Hes1 loss despite only partial HES1 reduction, indicating essential Notch-independent functions for progenitor survival [4, 55, 56]. In mouse embryos, SOX9 prevents premature endocrine commitment by reinforcing Notch–HES1–mediated restraint of NGN3, with Sox9-deficient embryos showing early emergence of NGN3⁺ and glucagon⁺ cells that rapidly depletes the progenitor reservoir [4, 57].

SOX9⁺ progenitors as the source of endocrine lineages

Mouse lineage-tracing studies and human fetal analyses together establish SOX9⁺ progenitors as the cellular source of pancreatic endocrine lineages. Although SOX9 does not directly specify individual endocrine subtypes, genetic and human studies demonstrate that it governs endocrine lineage formation by maintaining the progenitor pool and regulating NGN3-dependent endocrine commitment [4, 7, 57]. Lineage tracing analysis further confirms that SOX9⁺ embryonic epithelium give rise to both endocrine and exocrine compartments, with endocrine cells arising via continued neogenesis from SOX9⁺ progenitors [4]. Consistent with this hierarchical organization, endocrine-committed NGN3⁺ cells are frequently intercalated within SOX9⁺ epithelium, reflecting their emergence from this progenitor niche. During early human pancreatic development, SOX9 co-expresses with key endocrine regulators, including NGN3, NKX2.2, NKX6.1, and PAX6 but is excluded from terminally differentiated hormone-producing endocrine cells, positioning SOX9 as an upstream regulator of endocrine specification [4, 7, 58].

Mouse loss-of-function studies underscore the essential role of SOX9 in endocrine lineage formation, showing that SOX9 depletion markedly reduces pancreatic endocrine hormone–producing cells, including insulin-, glucagon-, somatostatin-, and pancreatic polypeptide–expressing populations [4]. Notably, another study further suggests a lineage imbalance under SOX9-deficient conditions, characterized by impaired β-cell differentiation with reduced insulin expression alongside a relative increase in α-cell specification, indicated by elevated glucagon expression [57].

Functionally, Sox9 haploinsufficiency reduces islet cell mass by ~ 50% and leads to aberrant endocrine cell composition. Conversely, SOX9 overexpression in human fetal pancreatic epithelium enhances endocrine transcriptional programs (NGN3, PAX6, NKX6.1, INS) without broadly activating ductal identity programs [4, 7]. Collectively, these findings establish SOX9 as a central regulator of endocrine lineage formation, while highlighting that its precise, cell type–specific roles across individual pancreatic endocrine populations remain incompletely understood and warrant further investigation.

SOX9 in adult pancreas: from developmental regulator to disease marker

SOX9 expression in adult ductal epithelium and context-dependent plasticity

In adult pancreas, SOX9 remains expressed in ductal and centroacinar epithelium, motivating the hypothesis that it marks an epithelial population with latent regenerative potential [1, 4]. Mouse lineage-tracing studies demonstrate injury- and context-dependent reactivation: under moderate hyperglycemia combined with long-term gastrin and epidermal growth factor (EGF) exposure, SOX9⁺ ductal cells can undergo stepwise conversion toward insulin-producing cells and improve glycemic control, whereas this response is absent under normoglycemia or extreme hyperglycemia [59, 60]. These findings support the concept of a facultative progenitor pool in adult SOX9⁺ epithelium whose potential can be recruited under specific metabolic and signaling conditions. However, in humans, direct evidence remains sparse and concentrated in early gestation, leaving unresolved whether human pancreas maintains a sustained reservoir of multipotent SOX9⁺ progenitors beyond initial endocrine commitment [47].

SOX9 in adult β-cells: low-level expression and functional roles

SOX9 is progressively excluded from the endocrine lineage during development and is generally undetectable in mature islets by conventional assays [1, 7, 8]. Sporadic reports of SOX9 in islet samples have been attributed to methodological artifacts including genomic DNA contamination, substrate diffusion in in situ hybridization, and antibody non-specificity [1, 4]. However, recent work in mouse models using β-cell-specific Sox9 deletion combined with high-sensitivity single-molecule RNA detection detected low-level Sox9 transcripts in adult β-cells from mice older than 10 months [2]. This low-level expression appears functionally significant. Consistent with dosage-sensitive regulation observed for other MODY TFs (HNF1A, HNF4A, PDX1, NEUROD1), SOX9 exhibits dose-dependent control over metabolic phenotypes. Pancreas-specific Sox9 heterozygosity leads to progressive, age-dependent glucose intolerance exacerbated by metabolic stress despite largely preserved islet architecture [5]. Similarly, β-cell-restricted Sox9 deletion after endocrine specification causes worsening glucose intolerance with age, elevated fasting glycemia, and increased body weight, with heterozygotes showing intermediate phenotypes consistent with gene-dosage effects [2]. These mice exhibit elevated basal insulin secretion with blunted glucose-stimulated responses without overt peripheral insulin resistance, indicating primary β-cell-intrinsic defects. Mechanistically, SOX9 maintains β-cell secretory competence through control of transcript isoform programs and stress-response pathways. β-cell-restricted Sox9 deletion reduces the splicing regulator serine/arginine-rich splicing factor 5 (SRSF5), accompanied by widespread isoform changes enriched in ER stress/unfolded protein response pathways and impaired insulin secretion [2]. Consistent with this, SOX9 overexpression in β-cells under cytokine-induced inflammatory stress attenuates apoptosis and partially restores insulin secretion [61]. These findings suggest that low-level SOX9 helps maintain β-cell function by supporting splicing fidelity and stress adaptation.

SOX9 reactivation as a marker of β-cell dedifferentiation

Beyond quantitative β-cell loss, diabetes-associated failure can involve erosion of mature β-cell identity (dedifferentiation), characterized by loss of insulin, key TF, and glucose-sensing machinery [62–64]. Evidence of β-cell identity loss has been reported in islets from individuals with type 2 diabetes [65, 66].

In a β-cell-specific Vhlh knockout model that enforces chronic hypoxia-inducible factor (HIF)-driven pseudohypoxic stress, Sox9 expression increased progressively in islets from ~ 8 weeks while mice remained normoglycemic [8]. SOX9 reactivation was detected in endocrine cells that had downregulated insulin and exhibited reduced PDX1, coinciding with induction of the progenitor factor Hnf6 and re-engagement of developmental programs including Notch (Hes1, hairy/enhancer-of-split related with YRPW motif protein 1 (Hey1)), Hedgehog (Ptc, GLI family zinc finger 1 (Gli1)), and Wnt/β-catenin (Axin2) components [8]. These data position SOX9 upregulation as a molecular hallmark of β-cell identity erosion preceding overt metabolic decompensation.

Chronic hyperglycemia, a major diabetogenic stressor, induces SOX9 at both mRNA and protein levels in β-cells [67]. SOX9 induction was accompanied by reduced proliferation, increased apoptosis, elevated oxidative and inflammatory markers (TNF-α, IL-2), and impaired insulin secretion, whereas SOX9 knockdown partially mitigated these effects. β-cell-directed SOX9 overexpression is sufficient to drive diabetes and dedifferentiation-like outcomes in vivo [8]. Together, these findings support the hypothesis that stress-induced SOX9 upregulation represents a maladaptive response contributing to β-cell dysfunction and identity destabilization during diabetes progression.

Metabolic cofactors and genetic associations

SOX9 activity can be tuned at the protein level by metabolic cofactors. SOX9 dimerization and cooperative DNA binding are enhanced by Flavin adenine dinucleotide (FAD), a vitamin B2–derived cofactor, and Pyridoxal 5′-phosphate (PLP), the active form of vitamin B6, which bind near the dimer interface and modulate SOX9–DNA interactions [68]. In pancreatic progenitor systems, supplementation with these cofactors increased SOX9 and NGN3 expression. Notably, deficiencies in these same micronutrients have been associated with impaired glycemic control and β-cell dysfunction: PLP insufficiency correlates with hyperglycemia, and riboflavin/FAD deficiency has been linked to altered β-cell function and diabetes progression [69–71]. Together, these observations support a model in which micronutrient status may indirectly shape β-cell development and resilience by tuning SOX9 activity, with potential implications for maintaining β-cell identity in diabetes. A related concept has been proposed for vitamin D signalling, where active vitamin D signaling has been reported to upregulate SOX9 expression in equine chondrocytes, although whether this regulatory relationship is conserved in pancreatic β-cells remains unclear [72]. More broadly, these findings raise the possibility that metabolic cofactors and micronutrient pathways represent underexplored levers for modulating SOX9-dependent gene programmes relevant to β-cell maintenance and diabetes. Although SOX9 appears in the genome-wide association studies (GWAS) Catalog, no significant associations have been identified between SOX9 variants and diabetes-related traits. This suggests that while SOX9 plays critical functional roles in β-cell development and maintenance, its influence on diabetes risk is not driven by common genetic variants. Rare deleterious mutations in SOX9, such as those underlying Campomelic Dysplasia, highlight that SOX9-related pathologies arise primarily from severe mutations rather than common polymorphisms. Nevertheless, dysregulation of SOX9 expression through epigenetic or inflammatory mechanisms may still contribute to β-cell dysfunction, warranting future studies integrating functional genomics and pharmacological approaches to determine whether modulating SOX9 activity could enhance β-cell survival and regeneration in diabetes.

Human SOX9 mutations: insights from campomelic dysplasia

While experimental models have established SOX9’s essential roles in pancreatic development, human genetic evidence directly linking SOX9 dysfunction to pancreatic phenotypes remains limited. Germline mutations in SOX9 cause campomelic dysplasia (CD), a rare congenital disorder primarily characterized by skeletal malformations, respiratory insufficiency, and, in approximately 70% of XY individuals, disorders of sex development [73–77]. The disorder follows an autosomal dominant inheritance pattern and arises most frequently from de novo mutations, although familial recurrence due to parental germline mosaicism has been reported [78]. The complete loss of both SOX9 alleles is embryonically lethal and has not been observed in humans, underscoring SOX9’s essential dosage-dependent developmental role [73, 77].At the molecular level, disease severity reflects both the position and functional impact of SOX9 mutations. Variants affecting the HMG DNA-binding domain impair DNA binding, nuclear localization, and transcriptional activity, with partial loss-of-function mutations often associated with ACD, whereas mutations that severely disrupt DNA interaction typically result in classic, lethal CD [26, 73–77, 79]. Consistent with the requirement for cooperative, DNA-dependent SOX9 dimerization, mutations in the SOX9 DIM identified in CD patients (e.g., A76E and Δ66–75) abolish dimer formation on DNA while preserving monomeric binding, leading to a marked reduction in transcriptional activation of SOX9 target genes [22]. In addition, CD arises from two major classes of genetic alterations: coding mutations that directly impair SOX9 protein function and structural variants that disrupt long-range regulatory elements required for proper SOX9 expression [80–82]. While many truncating mutations result in haploinsufficiency through nonsense-mediated decay, others escape degradation and generate truncated proteins that retain DNA-binding capacity but interfere with wild-type SOX9 activity, producing dominant-negative effects and more severe phenotypes [27, 73, 83]. Notably, no strict genotype–phenotype correlation exists, as identical SOX9 mutations can give rise to widely variable clinical outcomes, reflecting differences in residual transcriptional activity, dominant-negative interference, modifier genes, and developmental context [73, 77, 79, 84]. Beyond the skeletal and gonadal phenotypes, CD provides rare human genetic evidence implicating SOX9 in pancreatic development: histological analysis of three CD pancreatic tissue revealed abnormal epithelial architecture, with less densely packed epithelium and poorly demarcated islets exhibiting variable expression of endocrine hormones and β-cell markers [3]. Collectively, SOX9-associated CD provides a powerful human genetic framework illustrating how subtle perturbations in SOX9 dosage, DNA binding, dimerization, and transcriptional regulation can profoundly alter developmental trajectories, with direct relevance to SOX9 function across multiple embryonic lineages, including the pancreas [27].

Modeling SOX9 function in pancreatic development

Understanding SOX9’s role in pancreatic islet development and function requires integrating evidence across multiple experimental systems. While SOX9 is highly conserved across vertebrates, with strong sequence conservation in the HMG DNA-binding domain and key regulatory regions, translating findings from animal models to human biology requires careful consideration of species-specific differences in developmental timing, tissue architecture, and compensatory mechanisms. The limited availability of stage-resolved human fetal pancreatic tissue means that much of our mechanistic understanding of SOX9 function has historically relied on mouse genetics, which underpins much of the current SOX9-dependent regulatory framework (Table 1). Core transcriptional relationships including the SOX9–PDX1–HES1–NGN3 axis appear to be broadly conserved between mouse and human, making murine studies highly informative for establishing causal gene function and regulatory hierarchy. In mouse models, Notch–HES1 signaling represses endocrine differentiation and maintains progenitor identity, with loss of HES1 leading to premature endocrine commitment [55, 56]. Consistent with this, SOX9 sustains HES1 expression and prevents early activation of NEUROG3, as Sox9 deletion results in increased NGN3⁺ cells and depletion of the progenitor pool [4]. Human fetal studies support a similar regulatory framework, where SOX9 is associated with progenitor maintenance and regulation of endocrine differentiation, although the precise SOX9–HES1–NEUROG3 interactions remain less mechanistically defined in human systems [7]. Importantly, mouse, human, and stem cell–derived systems provide complementary insights into SOX9 function. Mouse models define causal mechanisms, human fetal studies provide physiological context, and directed differentiation systems allow controlled experimental interrogation, although they may not fully recapitulate human developmental dynamics. Nonetheless, translating mouse findings to human development requires caution, as species-.

Table 1.

SOX9-related pancreatic models summarized from key studies

Animal Model	Cre Driver Promoter	Targeted Tissue	Purpose	Outcome	Findings	Reference
Pdx1-Cre; Sox9flox/flox	Pdx1	Pancreatic progenitors	Conditional deletion of Sox9 in pancreatic epithelium	Severe pancreatic hypoplasia by E11.5 due to loss of progenitor maintenance.	SOX9 maintains pancreatic progenitor cells by regulating Notch signaling via co-expression of HES1; links SOX9 loss to premature endocrine differentiation.	(Seymour, Freude et al. [4])
Sox9-eGFP mice	None	SOX9+epithelium	Short-term in vivo lineage tracing of SOX9+ cells	Compartment of multipotent progenitors persists after the secondary transition of pancreas development.	SOX9 + epithelium simultaneously gives rise to NGN3+ endocrine progenitors.	(Seymour, Freude et al. [57])
Pdx1-Cre; Sox9flox/+	Pdx1	Pancreatic progenitors	Test whether Sox9 dosage affects pancreatic differentiation	~ 50% reduction in islet cell mass; exocrine compartment largely unaffected.	Dosage-sensitive requirement for SOX9 in initiating endocrine fate.	(Seymour, Freude et al. [57])
Zp3-Cre; Sox9flox/flox Prm1-Cre; Sox9flox/flox	Zp3; Prm1	Germline	Determine whether Sox9 is required for pancreatic specification	Reduced Pdx1 expression in dorsal and ventral pancreatic buds; early embryonic lethality prevented analysis beyond E11.5.	Sox9 is dispensable for pancreatic fate assignment and early pancreatic bud outgrowth.	(Shih, Seymour et al. [6])
Foxa3-Cre; Sox9flox/flox	Foxa3	Foregut endoderm	Early endoderm deletion and gene-dosage interaction with Pdx1 (Pdx1+/−; Sox9+/+, Pdx1+/−; Sox9+/Δgut, Pdx1+/−; Sox9Δgut/Δgut)	Deletion captured by ~E9.5. Sox9+/Δgut: normal bud size. Pdx1+/−; Sox9Δgut/Δgut: severely hypoplastic dorsal pancreas and absent ventral pancreas. Pdx1+/−; Sox9+/Δgut: reduced ventral bud size.	Sox9 heterozygosity alone does not impair pancreas formation. Sox9 is required for ventral pancreas initiation/outgrowth and for maintenance/expansion of dorsal progenitors after initial specification. Sox9 and Pdx1 are dispensable for initiation, but required for Ptf1a and Nkx6-1, respectively; SOX9 and PDX1 bind largely distinct regulatory regions and restrict non-pancreatic lineages.	(Shih, Seymour et al. [6])
Rosa26mCherry−tetO−Sox9 (Sox9GOF); Pdx1tTA	None (tTA system)	Pdx1-expressing foregut progenitors	Gain-of-function: forced Sox9 overexpression in Pdx1 + progenitors	Increased Pdx1 levels in duodenal region; repression of intestinal markers (Cdx2, Onecut2).	SOX9 and PDX1 together suppress intestinal lineage programs, but are not sufficient alone to initiate the pancreatic program.	(Shih, Seymour et al. [6])
Pdx1-Cre; Sox9flox/flox	Pdx1	Pancreatic progenitors	Define SOX9 role in regulating key transcription factors during early pancreas development	At E10.5, PDX1 normal; by E12.5, PDX1 markedly reduced in Sox9-deleted cells; HNF6, FOXA2, and TCF2 unchanged.	SOX9 is required to maintain PDX1 expression in pancreatic progenitors.	(Dubois, Shih et al. [5])
R26Cre™; Sox9flox/flox	Rosa26	All tissues after tamoxifen induction	Assess Sox9 function during differentiation (secondary transition) rather than initiation; timed deletion after E12.5	Induction at E12.5: ~90% loss of SOX9 in luminal epithelium by E15.5; pancreas size/morphology preserved; PDX1+ cells reduced; fewer PDX1+/INS+ cells; HNF6/FOXA2/TCF2 unchanged. Induction at E14.5: noticeable reduction in PDX1+ cells.	SOX9 maintains PDX1 during the secondary transition and supports endocrine commitment and beta-cell development; required for endocrine cell formation even late in development.	(Dubois, Shih et al. [5])
Pdx1-Cre; Sox9flox/+ (Sox9+/Δpan)	Pdx1	Pancreatic progenitors	Test whether reduced Sox9 dosage in pancreas causes diabetes-like phenotypes	At E18.5, beta-cell mass ~ 50% lower; islet organization preserved; remaining beta-cells functionally competent; postnatal proliferation partially compensates; limited adaptation to metabolic stress.	Reduced SOX9 dosage yields a MODY-like transient glucose intolerance that improves as beta-cells expand postnatally and restore functional mass.	(Dubois, Shih et al. [5])
Ins2-Cre; Sox9flox/flox	Ins2	Beta cells	Beta-cell-specific Sox9 deletion to test role in postnatal beta-cell integrity and regeneration	Impaired regeneration after STZ injury; poor beta-cell recovery, reduced insulin, increased apoptosis.	SOX9 is essential for maintaining beta-cell integrity and promoting regeneration following injury	(Wang, Chen et al. [61])
Ins-Cre; VhlhloxP/loxP	Ins2	Beta cells	Delete Vhlh in beta-cells to model chronic metabolic/pseudohypoxic stress (T2D-like)	Age-dependent Sox9 induction in islets from ~ 8 weeks (pre-diabetes); loss of insulin and reduced PDX1.	Chronic metabolic stress triggers ectopic SOX9 re-expression in adult beta-cells, marking loss of identity and a progenitor-like, NGN3-independent state.	(Puri, Akiyama et al. [8])
Ins-Cre; CAG-Sox9/HA	Ins2	Beta cells	Ectopic SOX9 overexpression in beta-cells	Hyperglycemia by 6–10 months; reduced insulin and PDX1; downregulation of beta-cell genes; disrupted GLUT2 membrane localization.	Ectopic SOX9 drives loss of beta-cell identity and function, producing a diabetes-like phenotype.	(Puri, Akiyama et al. [8])
Ins-Cre; Sox9flox/flox	Ins2	Beta cells	Delete Sox9 specifically in β cells to test whether Sox9 is required for maintenance/function of mature β cells; assess residual Sox9 expression by RNAscope	RNAscope detects low but measurable Sox9 transcripts in nuclei of insulin+ β cells (in controls); β-cell Sox9 loss causes functional impairment.	SOX9 is present at very low levels in adult β cells, but is still required to maintain β-cell functional competence; deletion causes progressive dysfunction without overt loss of β-cell identity.	(Puri, Maachi et al. [2])

specific differences in developmental tempo, signaling dynamics, and transcriptional regulation can alter when, and in what cellular context, conserved factors operate. This is particularly relevant for human pluripotent stem cell (hPSC) differentiation platforms, which aim to mimic embryogenesis by sequentially modulating key pathways to generate endoderm, foregut, pancreatic progenitors, and endocrine lineages [85–88]. Despite these limitations, stem cell–based systems provide important mechanistic insight by enabling controlled modulation of signaling pathways and developmental progression [85, 86]. In this context, SOX9 is likely to influence the efficiency and timing of pancreatic progenitor formation and endocrine differentiation, consistent with its established dosage-sensitive role in pancreatic development. Because much of the pathway logic informing these protocols originates from mouse embryology, it remains uncertain how faithfully in vitro–derived pancreatic cells recapitulate human fetal developmental trajectories and stage-specific regulatory dependencies [34].

Beyond molecular timing, the human and mouse pancreas also differ in gross organization and tissue architecture. The human pancreas is a compact, encapsulated organ with anatomically defined head, body, and tail regions, whereas the mouse pancreas is a diffuse, dendritic tissue distributed through the intestinal mesentery. Humans exhibit clearly defined lobes and lobules with substantial mesenchymal contribution, while mice possess three loosely connected lobes (duodenal, splenic, and gastric) separated by adipose/connective/lymphatic tissue, complicating direct anatomical comparisons and tissue sampling [89]. These structural differences reflect broader ontogenetic divergence and shape how developmental processes manifest at the tissue level.

A major cross-species distinction is developmental pace, with mouse gestation lasting ~ 3 weeks while human pancreatic development unfolds over months. Consistent with this compressed timeline, murine endocrine differentiation is frequently described as occurring in two broad transitions (“primary” followed by “secondary”), during which NEUROG3⁺ endocrine progenitors arise within the trunk epithelium and endocrine lineages expand as the organ rapidly patterns [1, 5]. In contrast, human endocrine specification does not resolve into two discrete waves. Instead, endocrine differentiation proceeds as a single, prolonged program that begins gradually once the SOX9⁺/PDX1⁺ progenitor epithelium has expanded and segregated into tip and trunk domains (approximately 4–10 weeks post-conception, PCW). During this period, NEUROG3 induction increases progressively rather than in temporal bursts, with endocrine progenitors continuously generating hormone-expressing cells over an extended window; maturation continues through mid-gestation and endocrine functional competence is consolidated postnatally [3, 34, 39].

Early inductive logic is also conserved but deployed differently. In mouse and chick, notochord–foregut contact contributes to dorsal pancreatic specification through repression of SHH in adjacent endoderm. In humans, a comparable interaction has been observed around 25–27 days post-conception (Carnegie stage 12–13), but appears shorter in duration and less morphologically distinct than in mice, suggesting that conserved mechanisms operate under species-specific spatial–temporal constraints [34, 39]. Collectively, these differences imply that conserved regulators may be embedded in divergent buffering and compensatory networks, strengthening the case for integrating human fetal datasets and stem-cell–derived models to capture human-specific dynamics [34, 90, 91].

The translational consequences are evident in monogenic disease genetics. While loss of NEUROG3 or RFX6 causes complete endocrine loss in mice, comparable human mutations produce variable phenotypes with residual endocrine differentiation [55, 90, 91]. Similarly, HNF1B/TCF2 haploinsufficiency causes MODY5 with prominent pancreatic defects in humans [92], yet Tcf2⁺/⁻ mice do not develop diabetes and show minimal pancreatic pathology [93], consistent with greater compensatory capacity in murine β-cells. This notion aligns with observations that reduced Sox9 dosage can be tolerated in mice through adaptive β-cell expansion and preserved glucose homeostasis [5], whereas the inherently low proliferative capacity of human β-cells predicts greater vulnerability to reduced gene dosage and impaired long-term maintenance [94].

In humans, therefore, the most informative route to infer developmental gene function has been the study of naturally occurring monogenic mutations, complemented by induced PSC (iPSC)-derived pancreatic organoids that enable mechanistic dissection of human-specific regulatory dependencies [95–99]. Notably, despite the central role of SOX9 established in animal models, only one report has directly linked SOX9 mutation to pancreatic hypoplasia in humans [3]. Moreover, the only stem-cell–based study interrogating SOX9 loss in a human pancreatic context used an inducible strategy that deletes SOX9 after endocrine differentiation, thereby failing to address SOX9’s earliest roles in human pancreatic specification and organogenesis [2]. These gaps underscore why mouse genetics remains essential for mechanistic depth, while human genetics and human stem-cell systems are critical for determining which aspects of SOX9 biology and which compensatory routes are truly conserved in human pancreatic development.

Finally, the diversity of available murine Sox9 alleles and Cre drivers has enabled stage- and lineage-specific interrogation of Sox9 function across pancreas development. Conditional Sox9 loss is most commonly achieved using the Sox9^flox allele [100] combined with pancreas drivers such as Pdx1-Cre (deleting after pancreatic specification) [4], or broader early endoderm drivers such as Foxa3-Cre (capturing earlier pre-pancreatic endoderm stages) [6]. Because early Sox9 deletion produces severe hypoplasia that obscures later roles, inducible systems (e.g., R26Cre^TM/ER) have been used to delete Sox9 after E12.5 to interrogate functions during the endocrine differentiation window [5]. Post-developmental requirements have also been tested using β-cell–targeted strategies (e.g., Ins-Cre; Sox9^flox/flox) and complementary β-cell lines such as MIN6 and RIN-m5F, which are useful for mechanistic studies but require cautious interpretation given immortalization-associated alterations [2, 8, 61, 101].

Conclusions and future directions

SOX9 emerges as a central orchestrator of pancreatic development whose functions extend from embryonic progenitor specification through adult β-cell maintenance and stress adaptation. During development, SOX9 defines multipotent pancreatic progenitors, cooperates with PDX1 to establish and reinforce pancreatic identity, and controls the timing of endocrine differentiation through both Notch-dependent and independent mechanisms. Its tightly regulated spatial and temporal expression, shifting from broad foregut distribution to progressive restriction within ductal and progenitor compartments, mirrors the fundamental developmental transition from pluripotency to lineage commitment. SOX9⁺ progenitors serve as the cellular source of endocrine lineages, with SOX9 downregulation representing a critical checkpoint that permits NGN3 induction and β-cell differentiation.

In adult β-cells, SOX9 exhibits context-dependent functions that challenge simple models of its role. While normally expressed at low levels where it supports secretory competence through splicing regulation and stress adaptation, aberrant SOX9 reactivation under diabetogenic stress marks β-cell dedifferentiation and identity erosion. This dual nature, protective at physiological levels, pathological when dysregulated, positions SOX9 as both a potential therapeutic target and a biomarker of β-cell dysfunction. The dosage sensitivity observed across mouse genetic models and human campomelic dysplasia reinforces that subtle perturbations in SOX9 activity can profoundly affect pancreatic outcomes, with implications extending from developmental hypoplasia to age-dependent glucose intolerance.

Critical gaps remain in translating these insights to human islet biology and diabetes therapeutics. First, the scarcity of stage-resolved human fetal pancreatic tissue and limited genetic evidence directly linking SOX9 to human pancreatic phenotypes necessitate greater investment in human stem cell models and detailed metabolic phenotyping of campomelic dysplasia survivors. Second, the mechanisms governing SOX9 expression in adult β-cells, including epigenetic regulation, metabolic inputs through vitamin-derived cofactors, and stress-induced transcriptional reactivation, remain incompletely understood and represent attractive targets for pharmacological intervention. Third, whether SOX9 reactivation in failing β-cells represents a maladaptive dedifferentiation program or a partially protective stress response remains unresolved and has direct implications for therapeutic strategy.

Several promising research directions emerge from this framework. Single-cell multi-omic approaches applied to human fetal and adult pancreatic tissues could define the precise transcriptional networks downstream of SOX9 across developmental stages and disease states, identifying stage-specific vulnerabilities and potential intervention points. Genome editing in human pluripotent stem cells could systematically interrogate SOX9 dosage effects, cooperative interactions with other MODY TFs, and the functional consequences of dimerization-disrupting mutations observed in campomelic dysplasia. Metabolic cofactor supplementation trials, guided by the discovery that FAD and PLP enhance SOX9 dimerization, could test whether micronutrient optimization improves β-cell function in at-risk populations. Finally, lineage-tracing studies in human pancreatic organoids and improved animal models could definitively establish whether adult SOX9⁺ ductal epithelium retains latent regenerative capacity that can be harnessed for β-cell replacement.

Ultimately, SOX9 exemplifies how developmental regulators can be repurposed in adult tissues for homeostatic maintenance, with their dysregulation contributing to disease pathogenesis. Understanding the molecular switches that govern SOX9’s transition from developmental driver to maintenance factor, and from protective to pathological in the diabetic context, will be essential for designing rational therapies that preserve β-cell identity, enhance regenerative potential, and prevent dedifferentiation in diabetes.

Authors’ contributions

NS contributed to the design, data collection, and writing the manuscript. ME, ZI, and WH contributed to data collection. EMA and PRK contributed to the design, review, and editing of the manuscript. All authors read and approved the final version of the manuscript.

Funding

P.R.K. were funded by Qatar Biomedical Research Institute (QBRI) IGP6, and E.M.A. were funded by a budget from Sidra Medicine (Project No. SDR400215; SDR400217).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.