Proinflammatory cytokine-induced alpha cell impairment in human islet microtissues is partially restored by dual incretin receptor agonism

Abstract

Aims/hypothesis: In type 1 diabetes, the counterregulatory glucagon response to low plasma glucose is impaired. The resulting increased risk of hypoglycaemia necessitates novel strategies to ameliorate alpha cell impairment. Here, we aimed to establish an in vitro type 1 diabetes-like model of alpha cell impairment using standardised reaggregated human islet microtissues (MTs) exposed to proinflammatory cytokines. Additionally, we investigated the therapeutic potential of incretin receptor agonists in improving alpha cell responses to low glucose.

Methods: Human islet MTs were exposed to proinflammatory cytokines (IL-1β, IFN-γ and TNF-α) for 1 day (short-term) and 6 days (long-term). Alpha cell function was assessed by sequential glucose-dependent secretion assays at 2.8 and 16.7 mmol/l glucose, followed by glucagon measurements. Additional evaluations included ATP content, caspase-3/7 activity, chemokine secretion and content of islet transcription factors (aristaless-related homeobox [ARX] and NK6 homeobox 1 [NKX6.1]) and hormones. The effects of incretin receptor agonist treatment (glucose-dependent insulinotropic polypeptide [GIP] analogue [D-Ala²]-GIP with or without the glucagon-like peptide 1 [GLP-1] receptor agonist liraglutide) alongside or after cytokine exposure were also investigated, focusing on low-glucose-dependent glucagon secretion.

Results: Short-term cytokine exposure increased glucagon secretion at both 2.8 and 16.7 mmol/l glucose in islet MTs. In contrast, long-term cytokine exposure caused dose-dependent suppression of glucagon secretion at 2.8 mmol/l glucose, resembling a type 1 diabetes phenotype. Long-term cytokine exposure also diminished insulin and somatostatin secretion, reduced ATP content, increased caspase 3/7 activity and decreased islet content of ARX, NKX6.1, glucagon and insulin. Despite cytokine-induced impairment, alpha cells partially retained secretory capacity to L-arginine stimulation. Treatment with incretin receptor agonists during long-term cytokine exposure did not prevent alpha cell impairment. However, acute treatment with [D-Ala²]-GIP with or without liraglutide, or with the single-molecule dual agonist tirzepatide, after cytokine exposure partially restored glucagon secretion at low glucose.

Conclusions/interpretation: Long-term cytokine exposure of human islet MTs created a type 1 diabetes-like phenotype with impaired low-glucose-induced glucagon secretion. This cytokine-induced alpha cell impairment was partially restored by [D-Ala²]-GIP with or without liraglutide, or by tirzepatide.

Keywords: GIP; GLP-1; Glucagon secretion; Human model; Incretins; Liraglutide; Pancreatic alpha cells; Proinflammatory cytokines; Type 1 diabetes.

Fig. 1

Short-term cytokine exposure increases glucagon secretion. (a) Experimental schematic of the short-term (1 day) setup. (b, c) Glucagon secretion at 2.8 and 16.7 mmol/l glucose in control islet MTs (grey bars) and with increasing load of short-term cytokine exposure (blue bars) in donor 1 (b) and donor 2 (c). The dashed line denotes the baseline physiological response to low glucose of untreated control islet MTs. Data are presented as mean ± SD of a single donor in six technical replicates. *p<0.05, **p<0.01 and ***p<0.001 vs untreated control, by one-way ANOVA with Dunnett’s multiple comparisons test; ^†††p<0.001, for the two untreated controls at 2.8 vs 16.7 mmol/l glucose, by Student’s t test

Fig. 2

Long-term cytokine exposure impairs glucose-dependent glucagon secretion. (a) Experimental schematic of the long-term (6 day) setup. (b, d, f) Glucagon secretion at 2.8 and 16.7 mmol/l glucose in control islet MTs (grey bars) and with increasing load of long-term cytokine exposure (blue bars) in donor 3 (b), donor 4 (d) and donor 5 (f). The dashed line denotes the baseline physiological response to low glucose of untreated control islet MTs. (c, e) Accumulated glucagon secretion over the last 24 h of long-term cytokine exposure for donor 4 (c) and donor 5 (e). Data are presented as mean ± SD of a single donor in six technical replicates (five technical replicates for donor 5). *p<0.05, **p<0.01 and ***p<0.001 vs untreated control, by one-way ANOVA with Dunnett’s multiple comparisons test; ^††p<0.01 and ^†††p<0.001, for the two untreated controls at 2.8 vs 16.7 mmol/l glucose, by Student’s t test

Fig. 3

Long-term cytokine exposure induces cell death. (a–c) ATP content in control (grey bars) and with increasing load of long-term cytokine exposure (blue bars) in donor 3 (a), donor 4 (b) and donor 5 (c). Data are presented as mean ± SD of a single donor in six technical replicates. **p<0.01 and ***p<0.001 vs untreated control, by one-way ANOVA with Dunnett’s multiple comparisons test. (d) Caspase 3/7 activity after long-term cytokine exposure (dose 1, donor 8). Data are presented as mean ± SD of a single donor in five technical replicates. ***p<0.001, for untreated control vs cytokine exposure, by Student’s t test. RLU, relative light units

Fig. 4

Long-term cytokine exposure reduces transcription factor ARX and NKX6.1 expression. (a–d) Representative images for transcription factor staining in islet MTs from donor 9 with increasing load of long-term cytokine exposure (a) and associated quantification of DAPI-positive cell count (b), mean ARX intensity (c) and mean NKX6.1 intensity (d) expression. (e–h) Representative images for transcription factor staining in islet MTs from donor 10 (e) and associated quantification of DAPI-positive cell count (f), mean ARX intensity (g) and mean NKX6.1 intensity (h). Scale bar, 20 μm. Data are presented as mean ± SD of a single donor in 3–11 technical replicates (12–19 for control). *p<0.05 and ***p<0.001 vs untreated control, using one-way ANOVA with Dunnett’s multiple comparisons test

Fig. 5

Long-term cytokine exposure reduces islet hormone expression. (a–e) Representative images for islet hormone staining in islet MTs from donor 9 with increasing load of long-term cytokine exposure (a) and associated quantifications of DAPI-positive cell count (b), mean glucagon intensity (c), mean insulin intensity (d) and mean somatostatin intensity (e). (f–j) Representative images for islet hormone staining in islet MTs from donor 10 (f) and associated quantifications of DAPI-positive cell count (g), mean glucagon intensity (h), mean insulin intensity (i) and mean somatostatin intensity (j). (k–o) Representative images for islet hormone staining in islet MTs from donor 11 (k) and associated quantifications of DAPI-positive cell count (l), mean glucagon intensity (m), mean insulin intensity (n), and mean somatostatin intensity (o). Scale bar, 20 μm. Data are presented as mean ± SD of a single donor in 3–7 technical replicates (9–15 for control). ***p<0.001 vs untreated control, by one-way ANOVA with Dunnett’s multiple comparisons test

Fig. 6

Alpha cells partially retain secretory capacity in response to l-arginine after long-term cytokine exposure. Glucagon secretion at 2.8 and 16.7 mmol/l glucose in control (grey bars) and with increasing load of long-term cytokine exposure (blue bars) in islet MTs from donor 8 with and without 10 mmol/l l-arginine. The dashed line denotes the baseline physiological response to low glucose of untreated control islet MTs. Data are presented as mean ± SD of five technical replicates. **p<0.01 and ***p<0.001, for l-arginine stimulated MTs vs the respective unstimulated control and for l-arginine stimulated MTs vs the untreated control/baseline response, by two-way ANOVA with Šídák’s multiple comparisons test, L-arg, l-arginine

Fig. 7

Treatment with [d-Ala²]-GIP with or without liraglutide alongside cytokine exposure does not prevent alpha cell impairment or cell death. (a) Experimental schematic for donor 8. (b, c) Glucagon secretion in islet MTs from donor 8 at 2.8 mmol/l glucose in control MTs (b), and long-term cytokine-exposed MTs (at dose ¼: 0.5, 2.5 and 2.5 ng/ml for IL-1β, IFN-γ and TNF-α, respectively) (c), without or with 1 µmol/l [d-Ala²]-GIP, liraglutide, or both. (d) Experimental schematic for donor 12 denoting addition of 24 h pre-treatment with [d-Ala²]-GIP and/or liraglutide before cytokine exposure and the addition during glucose-dependent hormone secretion assay. (e, f) Glucagon secretion in islet MTs from donor 12 at 2.8 mmol/l glucose in control MTs (no cytokines) (e), and long-term cytokine-exposed MTs (at dose ¼: 0.5, 2.5 and 2.5 ng/ml for IL-1β, IFN-γ and TNF-α, respectively) (f), without or with 1 µmol/l [d-Ala²]-GIP, liraglutide, or both. (g, h) Caspase 3/7 activity in islet MTs from donor 12 in control islet MTs (no cytokines) (g), and after long-term cytokine exposure (dose ¼: 0.5, 2.5 and 2.5 ng/ml for IL-1β, IFN-γ and TNF-α, respectively) (h) without or with 1 µmol/l [d-Ala²]-GIP, liraglutide, or both. Data are presented as mean ± SD of a single donor in five technical replicates. Results are not significant using one-way ANOVA with Tukey’s multiple comparisons test. RLU, relative light units

Fig. 8

Treatment with [d-Ala²]-GIP with or without liraglutide alters cytokine-induced secretion of CCL11 and CCL17. Accumulated CXCL8 (a), CXCL10 (b), CCL2 (c), CCL3 (d), CCL4 (e), CCL11 (f), CCL13 (g), CCL17 (h), CCL22 (i) and CCL26 secretion (j) in islet MTs from donor 12 during the last 24 h of the long-term cytokine exposure (dose ¼) without or with 1 µmol/l [d-Ala²]-GIP, liraglutide, or both. Data are presented as mean ± SD in 5 technical replicates (only three technical replicates were measured for untreated control and liraglutide bars due to pipetting error). *p<0.05, for all pairwise comparisons between treatment groups, by one-way ANOVA with Tukey’s multiple comparisons test

Fig. 9

Acute treatment with incretins following cytokine exposure boosts glucagon secretion of the type 1 diabetes phenotype. (a) Experimental schematic denoting acute incretin treatment during the hormone secretion assay. (b) Glucagon secretion in islet MTs from donor 8 at 2.8 mmol/l glucose in control islet MTs or in the islet MTs with increasing load of long-term cytokine exposure, with/without acute treatment with [d-Ala²]-GIP+liraglutide (1 µmol/l each) during the hormone secretion assay. (c, d) Glucagon secretion in islet MTs from donor 7 at 2.8 mmol/l glucose in control (c) and cytokine-exposed islet MTs (dose ¼: ¼: 0.5, 2.5 and 2.5 ng/ml for IL-1β, IFN-γ and TNF-α, respectively) (d) without or with acute [d-Ala²]-GIP+liraglutide (1 µmol/l each) or 1 µmol/l tirzepatide treatment. (e, f) As for (c, d) but in islet MTs from donor 13. (g, h) Glucagon secretion in islet MTs from donor 12 at 2.8 mmol/l glucose in control (g) and cytokine-exposed islet MTs (dose ¼: ¼: 0.5, 2.5 and 2.5 ng/ml for IL-1β, IFN-γ and TNF-α, respectively) (h) with/without acute treatment with 1 µmol/l [d-Ala²]-GIP, liraglutide, or both. (i, j) As for (g, h) but in islet MTs from donor 7. Data are presented as mean ± SD of a single donor in five technical replicates. *p<0.05, **p<0.01 and ***p<0.001, for [d-Ala²]-GIP+liraglutide treatment vs the respective untreated control, by two-way ANOVA with Šídák’s multiple comparisons test (b); *p<0.05, **p<0.01 and ***p<0.001, for [d-Ala²]-GIP+liraglutide or tirzepatide treatment vs the untreated control, by one-way ANOVA with Dunnett’s multiple comparisons test (c–f); *p<0.05, **p<0.01 and ***p<0.001, for all pairwise comparisons between treatment groups, by one-way ANOVA with Tukey’s multiple comparisons test (g–j). Ctrl, control