Hyaluronan induces ERK activation with minimal transcriptomic changes in pancreatic α-cells

TL;DR

This study shows that hyaluronan activates ERK signaling in pancreatic α-cells but causes only minor changes in gene activity compared to T1D α-cells.

Contribution

The study reveals that hyaluronan alone does not fully mimic the gene changes seen in T1D α-cells, suggesting other factors are involved.

Findings

Hyaluronan activates ERK signaling in α-cells but causes minimal transcriptional changes.

Hyaluronan-stimulated and T1D α-cells share reduced PPAR signaling signatures.

Hyaluronan alone is insufficient to drive T1D-like α-cell transcriptomic remodeling.

Abstract

Type 1 diabetes (T1D) is driven by the immune-mediated destruction of pancreatic β-cells. While β-cells are selectively targeted and depleted in T1D, α-cells that secrete glucagon are relatively spared from this autoimmune attack. Hyaluronan (HA), a glycosaminoglycan component of the extracellular matrix, accumulates within pancreatic islets in patients with T1D and in animal models, and has been implicated in promoting chronic inflammation. Given that surviving α-cells reside within this HA-rich microenvironment, we investigated whether HA contributes to molecular alterations in α-cells associated with the T1D phenotype. A murine α-cell line, αTC1-6, was stimulated with low-molecular-weight HA, which induced sustained extracellular signal-regulated kinase (ERK) phosphorylation upon HA stimulation, indicating activation of intracellular signaling. To assess whether this signaling response was accompanied by transcriptional changes, RNA sequencing was performed, and the resulting profile was compared with publicly available RNA-seq datasets from α-cells isolated from T1D donors. Despite robust ERK activation, HA stimulation elicited only minimal transcriptional changes relative to the T1D α-cell transcriptome. Gene set enrichment analysis further revealed a modest shift toward reduced peroxisome proliferator-activated receptor (PPAR) signaling in both HA-stimulated and T1D α-cells, suggesting a shared pathway-level signature. These findings indicate that HA alone does not fully recapitulate the extensive transcriptional remodeling observed in T1D α-cells, and that additional inflammatory or microenvironmental cues are required.

•Low-molecular-weight hyaluronan activates ERK signaling in α cells.•Hyaluronan alone elicits modest transcriptional changes in αTC1-6 cells.•Low-molecular-weight hyaluronan induces minimal transcriptional changes in α-cells.•Hyaluronan-stimulated and T1D α-cells share reduced PPAR signaling signatures.•Hyaluronan alone is insufficient to drive T1D-like α-cell transcriptomic remodeling.

Low-molecular-weight hyaluronan activates ERK signaling in α cells.

Hyaluronan alone elicits modest transcriptional changes in αTC1-6 cells.

Low-molecular-weight hyaluronan induces minimal transcriptional changes in α-cells.

Hyaluronan-stimulated and T1D α-cells share reduced PPAR signaling signatures.

Hyaluronan alone is insufficient to drive T1D-like α-cell transcriptomic remodeling.

Figures

Low-molecular-weight HA signaling in αTC1-6 cells (A) Quantitative real-time PCR analysis of known HA receptors in αTC1-6 cells. Relative mRNA levels were normalized to 18S rRNA expression.(B and C) Western blot analysis of total and phosphorylated ERK in αTC1-6 cells stimulated with 10 mg/mL of LMW-HA for the indicated time points. Representative blots are shown in (B), and the corresponding phosphorylated/total protein ratios are presented in (C). Data are presented as the mean ± SD (A, n = 3, C, n = 4). Full-length membrane images and additional blots are shown in Supplementary Figure S1C.

Limited transcriptional response of αTC1-6 cells to HA stimulation (A) MA plot showing the relationship between the mean value of logCPM and logFC for all genes in HA-treated vs. vehicle-treated αTC1-6 cells (n = 3 per group). (B) Volcano plot displaying logFC vs. -log10(*P*-value) for the same comparison. The dashed lined indicate the significance threshold (*P* < 0.05) and the fold-change cutoff. DEGs with logFC < −1 are shown in blue, and >1 in orange. Gray dots represent genes that are not significantly differentially expressed.Fig. 2

GSEA of the PPAR signaling pathwayGSEA of the KEGG pathway enrichment for the ranked gene lists derived from RNA-seq data of (A) vehicle-treated vs. HA-stimulated αTC1-6 cells and (B) α-cells isolated from non-diabetid control vs. T1D donors. Genes were ranked by logFC, and enrichment was assessed using KEGG pathway gene sets. Lower panels summarize enrichment statistics. The ranked gene lists for each comparison are shown in Supplementary Table S5 and S6.Fig. 3

Full text

Introduction

1

Type 1 diabetes (T1D) is characterized by the autoimmune destruction of pancreatic β-cells, resulting in a significant reduction or complete loss of insulin production [1]. As β-cells are the primary target of autoimmunity, they are almost completely lost as the disease progresses, whereas α-cells are relatively preserved and come to represent a greater fraction of the remaining islet cells [2]. It has been reported that in islets undergoing autoimmune destruction of β-cells, α-cell dysfunction is also observed [3,4]. Glucagon secretion, the primary role of α-cells, is regulated by multiple cues, including changes in circulating glucose levels and endocrine signals such as paracrine signaling by β-cell-derived insulin and δ-cell-derived somatostatin [5]. Loss of β-cells therefore removes a critical regulatory component of glucagon secretion, causing an imbalance in the orchestrated regulation of α-cell function and thereby disrupting their ability to appropriately sense and respond to changes in glucose levels [3]. As a result, characteristic abnormalities in glucagon regulation emerge in T1D. Shortly after diagnosis, patients often experience an inappropriate postprandial increase in glucagon secretion when it should normally be suppressed [4]. With disease progression, counter-regulatory glucagon responses to hypoglycemia become blunted due to the loss of the ability to sense changes in blood glucose levels, predisposing patients to hypoglycemia [3]. However, the islet microenvironmental cues that drive these changes remain poorly understood.

Recent studies have identified pathological accumulation of hyaluronan (HA), a glycosaminoglycan component of the extracellular matrix, within the pancreatic islets of patients with T1D [6]. Similar patterns of HA deposition have been observed in T1D mouse models, particularly in autoimmune-prone strains, whereas pharmacologically induced models such as streptozotocin-treated mice show significantly less accumulation [7]. Remarkably, HA accumulation occurs early in T1D progression, before immune cell infiltration and β-cell destruction, suggesting that it may play a key pathogenic role [8]. Furthermore, inhibition of HA synthesis prior to disease onset has been shown to prevent T1D development, supporting a causative role in disease pathogenesis [7,9].

HA is known to mediate pro-inflammatory signaling via its low-molecular-weight (LMW) fragments, which can induce cytokine and chemokine expression in various inflammatory diseases [10]. In the T1D-prone DO11.10xRIPmOVA (DORmO) model, HA accumulation in the islet microenvironment becomes more pronounced as the disease progresses, and increased infiltration of CD3^+^ T cells correlates with the loss of insulin-positive cells. These observations suggest an association between HA-rich islets, T cell infiltration, and β-cell loss in this model [7]. Furthermore, HA-mediated activation of CD44 signaling promotes phosphorylation of extracellular signal-regulated kinase (ERK) in human and murine CD4^+^ T cells and inhibits forkhead box p3-positive (FOXP3^+^) regulatory T cells (Tregs) induction [7]. Given the critical role of FOXP3^+^ Tregs in immune tolerance [11], HA-mediated suppression of Tregs is likely to accelerate autoimmune β-cell loss.

Given the accumulation of HA in T1D islets and the persistence of α-cells within this HA-rich microenvironment, we investigated whether HA contributes to molecular alteration in α-cells associated with the T1D phenotype. In this study, we exposed a murine α-cell line, αTC1-6, to purified LMW-HA to assess HA-induced intracellular signaling responses and transcriptional changes using bulk RNA sequencing. Furthermore, we compared HA-responsive pathways with those observed in publicly available RNA-seq data from α-cells isolated from T1D donors to explore whether an HA-rich condition may contribute to specific aspects of α-cell transcriptional remodeling in T1D.

Materials and methods

2

Cell culture and reagents

2.1

The αTC1-6 cells (ATCC, Manassas, VA, USA) were cultured in low-glucose Dulbecco's Modified Eagle's Medium (DMEM) (Nacalai Tesque Inc., Kyoto, Japan) supplemented with 10% fetal bovine serum, 15 mM HEPES, 0.1 mM non-essential amino acids, 0.02% bovine serum albumin, penicillin (100 U/mL), and streptomycin (0.1 mg/mL) under 5% CO_2_ at 37 °C. LMW hyaluronan (GLR001, 32 kDa, R&D Systems, Minneapolis, MN, USA) was dissolved in phosphate-buffered saline and applied at indicated concentrations.

Western blotting

2.2

Cell lysates were prepared using radioimmunoprecipitation assay buffer (Fujifilm Wako, Osaka, Japan) supplemented with protease and phosphatase inhibitor cocktails (Nacalai Tesque Inc.). Equal amounts of protein were separated by polyacrylamide gel electrophoresis and subsequently transferred onto a polyvinylidene difluoride membrane (Merck-Millipore, Billerica, MA, USA). After blocking with Blocking-One or Blocking-One P solution (Nacalai Tesque Inc.), membranes were incubated with primary antibodies targeting total or phosphorylated ERK and AKT (Cell Signaling Technology, Danvers, MA), or β-actin (AC-15, Sigma Aldrich, St. Louis, MO, USA). After washing, membranes were incubated with horseradish peroxidase-conjugated secondary antibodies (Cell Signaling Technology). Blots were visualized using an enhanced chemiluminescence system (Immunostar LD; Wako), and digital images were captured using the ImageQuant LAS 4000 mini (GE Healthcare, Chicago, IL, USA). Band intensities were quantified using ImageJ software (National Institutes of Health, Bethesda, MD, USA).

Quantitative RT-PCR (qPCR)

2.3

Total RNA was extracted from αTC1-6 cells using a ReliaPrep RNA Miniprep Kit (Promega, Madison, WI, USA) and was reverse-transcribed into cDNA using a High-Capacity cDNA Reverse Transcription kit according to the manufacturer's instructions (Life Technologies, Gaithersburg, MD, USA). Quantitative real-time PCR analysis was subsequently performed using the THUNDERBIRD Next SYBR qPCR Mix (Toyobo, Osaka, Japan) on a StepOnePlus Real-Time PCR system (Applied Biosystems, Foster City, CA). Gene expression levels were normalized to 18S rRNA as an internal reference for each sample. Primer sequences are listed in Supplementary Table S1. Sequences with PrimerBank IDs were obtained from PrimerBank (https://pga.mgh.harvard.edu/primerbank/index.html); others were sourced from previously published studies [12,13].

RNA-sequencing (RNA-seq) analysis

2.4

Total RNA was extracted using the ReliaPrep RNA Miniprep Kit (Promega). RNA concentration was quantified using the ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA), and RNA integrity was assessed using the TapeStation (Agilent Technologies, Santa Clara, CA, USA). The sequencing libraries were prepared from 200 ng of total RNA using the MGIEasy rRNA Depletion Kit and MGIEasy RNA Directional Library Prep Set (MGI Tech Co., Ltd., Shenzhen, China), according to the manufacturer's instructions. The libraries were sequenced on the DNBSEQ-G400 FAST Sequencer (MGI Tech Co., Ltd.) using a paired-end (2 × 150 bp) strategy. All sequencing reads were trimmed of low-quality bases and adapters using Trimmomatic (v.0.38) [14]. Gene-level reads for mouse samples were estimated using RSEM version 1.3.0 in conjunction with Bowtie2 against the mouse reference genome GRCm39 [15,16]. Publicly available human α-cell RNA-seq datasets were obtained from Gene Expression Omnibus (GEO) under accession number [GSE106148](GSE106148), and gene-level count matrices were downloaded [17]. Differentially expressed genes (DEGs) were identified using the edgeR program [18]. Normalized counts per million (CPM) values, log fold-changes (logFC), and P-values were obtained from the gene-level raw counts. P-values were adjusted for multiple testing using the Benjamini-Hochberg method to obtain false discovery rates (FDR). Genes with |logFC| > 1 and P-value <0.05 were considered differentially expressed.

Gene set enrichment analysis

2.5

Gene set enrichment analysis (GSEA) was performed using the GSEA algorithm described by Subramanian et al. [19]. For each comparison, genes were ranked using the standard Signal2Noise metric. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway gene sets were used as the predefined gene sets. Enrichment scores were calculated using 1000 permutations, and gene sets with sizes between 15 and 500 genes were included. Pathways with a nominal P-value <0.05 and an FDR q-value <0.25 were considered significantly enriched. The top 20 positively and negatively enriched pathways for each comparison are presented in the Supplementary Table S4–S8.

Statistical analysis

2.6

Quantitative data are presented as the mean ± standard deviation (SD). Statistical analyses were conducted using JMP Pro 16.0.0 (SAS Institute Inc., Cary, NC, USA). A *P-*value less than 0.05 was considered significant.

Results

3

Hyaluronan receptors in α-cells

3.1

To identify potential HA receptors in α-cells, we performed qPCR analysis using αTC1-6 cells, a mouse pancreatic α-cell line. We examined the expression of several cell-surface HA receptors, including CD44 [20], receptor for hyaluronic acid-mediated motility (RHAMM) [20], lymphatic vessel endothelial hyaluronan receptor 1 (LYVE-1) [21], and G protein-coupled receptor class C group 5 member C (GPRC5C) [22]. Among these, RHAMM exhibited the highest mRNA expression, approximately 60-fold higher than CD44, the most extensively studied HA receptor, suggesting a predominant RHAMM-dependent signaling (Fig. 1A).Fig. 1. Low-molecular-weight HA signaling in αTC1-6 cells (A) Quantitative real-time PCR analysis of known HA receptors in αTC1-6 cells. Relative mRNA levels were normalized to 18S rRNA expression.(B and C) Western blot analysis of total and phosphorylated ERK in αTC1-6 cells stimulated with 10 mg/mL of LMW-HA for the indicated time points. Representative blots are shown in (B), and the corresponding phosphorylated/total protein ratios are presented in (C). Data are presented as the mean ± SD (A, n = 3, C, n = 4). Full-length membrane images and additional blots are shown in Supplementary Figure S1C. *P < 0.05, for indicated comparisons, by the Wilcoxon rank test.Fig. 1

HA stimulation activates the ERK signaling in α-cells

3.2

We next assessed the downstream signaling pathways associated with RHAMM. The phosphoinositide 3-kinases (PI3K)-mediated AKT and ERK signaling pathways are the most implicated in HA-RHAMM signaling [[23], [24], [25]]. Stimulation of αTC1-6 cells with LMW-HA induced dose-dependent ERK phosphorylation, with 10 μg/mL sufficient to trigger activation (Supplementary Fig. S1A). A time-course analysis revealed ERK phosphorylation within 5 min of stimulation, with sustained activation for at least 60 min (Supplementary Fig. S1B). As phosphorylation levels continued to rise, extended stimulation up to 24 h demonstrated persistent ERK phosphorylation (Fig. 2B and C). In contrast, AKT phosphorylation was not affected by HA stimulation at any time point tested (Supplementary Fig. S2). These results suggest that prolonged HA stimulation leads to sustained activation of the ERK signaling pathway in α-cells, indicating potential downstream transcriptional regulation.Fig. 2. Limited transcriptional response of αTC1-6 cells to HA stimulation (A) MA plot showing the relationship between the mean value of logCPM and logFC for all genes in HA-treated vs. vehicle-treated αTC1-6 cells (n = 3 per group). (B) Volcano plot displaying logFC vs. -log10(P-value) for the same comparison. The dashed lined indicate the significance threshold (P < 0.05) and the fold-change cutoff. DEGs with logFC < −1 are shown in blue, and >1 in orange. Gray dots represent genes that are not significantly differentially expressed.Fig. 2

Minimal transcriptional changes observed in HA-stimulated αTC1-6 cells

3.3

To investigate whether HA-induced prolonged ERK activation leads to transcriptional changes, we performed RNA-seq analysis on αTC1-6 cells stimulated with LMW-HA for 24 h. MA and volcano plots showed that LMW-HA induced only a modest transcriptional response, with most genes exhibiting a log fold change (logFC) close to zero (Fig. 2A and B). Full differential expression results for all detected genes are provided in Supplementary Table S2. Differential expression analysis identified relatively few genes, 31 upregulated and 37 downregulated, exceeding our criteria of |logFC| > 1 and P < 0.05 (Fig. 2B–Supplementary Table S3). These findings indicate that LMW-HA alone has minimal effects on gene expression in αTC1-6 cells.

Peroxisome proliferator-activated receptor (PPAR) signaling pathway is downregulated in HA-stimulated αTC1-6 cells and T1D α-cells

3.4

To determine whether HA induces coordinated, pathway-related changes, we performed gene set enrichment analysis (GSEA). In HA-stimulated cells, several pathways potentially related to ERK signaling, including the calcium signaling pathway and glycerol phospholipid metabolism, showed a trend toward enrichment (nominal P-values <0.05), although they did not pass the predefined FDR threshold (q-value <0.25) (Supplementary Table S4). In contrast, pathways associated with inflammatory responses, such as cytokine-cytokine receptor interaction and systemic lupus erythematosus, tended to be enriched in vehicle-treated cells (nominal P-value <0.05, FDR q-value >0.25; Supplementary Table S4).

We next conducted GSEA on publicly available RNA-seq data from human α-cells isolated from T1D and non-diabetic control donors (GSE106148). Although several HA-responsive pathways showed a similar directional shift in T1D α-cells, the PPAR signaling pathway was the only KEGG pathway that met our significance criteria (nominal P-value <0.05, FDR q-value <0.25) in both datasets and consistently exhibited enrichment in vehicle/control cells (Fig. 3A and B; Supplementary Table S4–S8). PPAR signaling has been reported to be negatively regulated downstream of the ERK activation [[26], [27], [28], [29]], suggesting a potential mechanistic link between HA-induced ERK activation and reduced PPAR activity. Taken together, these results indicate that while HA stimulation alone induces only minimal gene expression changes in αTC1-6 cells, an HA-rich condition may contribute to a shift toward reduced PPAR signaling, a feature also observed in human T1D α-cells.Fig. 3GSEA of the PPAR signaling pathwayGSEA of the KEGG pathway enrichment for the ranked gene lists derived from RNA-seq data of (A) vehicle-treated vs. HA-stimulated αTC1-6 cells and (B) α-cells isolated from non-diabetid control vs. T1D donors. Genes were ranked by logFC, and enrichment was assessed using KEGG pathway gene sets. Lower panels summarize enrichment statistics. The ranked gene lists for each comparison are shown in Supplementary Table S5 and S6.Fig. 3

Discussion

4

HA, a major component of the extracellular matrix, has been reported to accumulate in pancreatic islets of individuals with T1D [[6], [7], [8], [9]]. We tested whether HA can directly stimulate α-cells to reprogram gene expression. Although LMW-HA induced sustained ERK phosphorylation in αTC1-6 cells, it elicited only a minimal transcriptional response. HA alone was therefore insufficient to drive the transcriptional reprogramming characteristic of α-cells in T1D. However, GSEA revealed that there was a shift toward reduced PPAR signaling pathway in both HA-stimulated αTC1-6 cells and human T1D α-cells, raising the possibility that the HA-rich condition may contribute to selective pathway-level changes observed in T1D α-cells.

Several factors may underlie the limited transcriptional change after HA stimulation. One possibility is the inflammatory milieu, which had not been considered in the present experimental setting. In vivo, α-cells are exposed not only to HA but also to inflammatory cytokines, altered extracellular matrix environments, and paracrine inputs from immune cells as well as other islet cells, all of which may act synergistically with HA to amplify gene expression changes. Another contributing factor may be the lack of heterogeneity in HA in our model. We only used LMW-HA in our experimental setting based on a prior report showing that the predominant size of HA found in autoimmune T1D model mice is LMW-HA fragments [30]. However, higher molecular weight species or dynamic HA degradation might induce additional responses.

Among the PPARs, PPARγ and PPARα are expressed in pancreatic α-cells, where they contribute to the regulation of glucagon production [31,32]. Activation of the ERK signaling pathway has been shown in multiple cell types to reduce PPAR γ activity, thereby attenuating its transcriptional function [[26], [27], [28]]. Similar ERK-dependent suppression of PPARα signaling has also been reported in hepatocytes under hypoxic conditions [29]. In support of a functional role for PPAR signaling in α-cells, stimulation of αTC1-6 cells with the PPARγ agonist troglitazone has been shown to reverse hyperglycemia-induced glucagon dysregulation [33]. Our observation that HA-induced ERK activation is accompanied by a pathway-level shift toward reduced PPAR signaling is consistent with prior reports describing ERK-mediated repression of PPAR pathways and may represent a mechanism contributing to altered regulation of glucagon secretion observed in T1D.

A key limitation of this study is the substantial inherent difference in the gene expression profiles between the immortalized αTC1-6 cell line and isolated human α-cells. Although murine αTC1-6 cells share a considerable portion of their transcriptome with human α-cells [34], species- and model-specific regulatory differences likely influence their response. Another limitation is that the degree of HA accumulation in the T1D donor islets used for the comparison is unknown. Therefore, the pathway-level similarity observed in PPAR signaling should be interpreted cautiously. Future studies using human α-cell models, such as iPSC-derived α-like cells or intact human islets, will be necessary to better understand the consequences of HA accumulation in T1D islets.

In conclusion, our data indicate that HA alone is insufficient to replicate the transcriptional phenotypes of T1D α-cells. Instead, HA-induced ERK activation is associated with a shift toward reduced PPAR signaling, a feature also observed in T1D α-cells. These findings suggest that HA accumulation in T1D islets may act as one of several microenvironmental cues that collectively modulate α-cell behavior.

Declaration of generative AI and AI-assisted technologies in the manuscript preparation process

The authors used ChatGPT (OpenAI, GPT-5.2) to assist with language refinement and improve the readability of the manuscript. The scientific content, data analysis, and conclusions were solely determined by the authors. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the article.

Funding sources

This study was supported by grants from JST 10.13039/501100025019SPRING JPMJSP2136 to 10.13039/501100009051SS.

CRediT authorship contribution statement

Suguru Sonoyama: Funding acquisition, Investigation, Visualization. Akiko Mizokami: Conceptualization, Investigation, Methodology, Visualization, Writing – original draft. Tomomi Sano: Validation, Writing – review & editing. Eijiro Jimi: Writing – review & editing. Masafumi Moriyama: Writing – review & editing. Takashi Kanematsu: Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.