Incretin-responsive human pancreatic adipose tissue organoids: A functional model for fatty pancreas research

Abstract

Objective: Infiltration of adipocytes into the pancreatic parenchyma has been linked to impaired insulin secretion in individuals with increased genetic risk of T2D and prediabetic conditions. However, the study of this ectopic fat depot has been limited by the lack of suitable in vitro models.

Methods: Here, we developed a novel 3D model of functionally mature human pancreatic adipose tissue organoids by aggregating human pancreatic adipose tissue-derived stromal vascular fraction (SVF) cells into organoids and differentiating them over 19 days.

Results: These organoids carry biological properties of the in situ pancreatic fat, presenting levels of adipogenic markers comparable to native pancreatic adipocytes and improved lipolytic and anti-lipolytic response compared to conventional 2D cultures. The organoids harbour a small population of immune cells, mimicking in vivo adipose environment. Furthermore, they express GIPR, allowing investigation of incretin effects in pancreatic fat. In accordance, GIP and the dual GLP1R/GIPR agonist tirzepatide stimulate lipolysis but had distinct effects on the expression of proinflammatory cytokines.

Conclusions: This novel adipose organoid model is a valuable tool to study the metabolic impact of incretin signalling in pancreatic adipose tissue, revealing potential therapeutic targets of incretins beyond islets. The donor-specific metabolic memory of these organoids enables examination of the pancreatic fat-islet crosstalk in a donor-related metabolic context.

Keywords: Adipogenesis; Incretins; Inflammation; Organoids; Pancreatic adipose tissue.

Figure 1

Pancreatic adipose tissue organoids display improved adipogenesis. (A) Schematic representation of in vitro generation of functionally mature pancreatic adipocytes: human pancreatic fat tissue is digested with collagenase to isolate primary mature adipocytes and SVF cells. SVF cells are expanded, seeded onto conventional adherent plates (2D; monolayer cell culture) or ultra-low-attachment plates (3D; spheroid cell culture) and subjected to adipogenic induction (D0, day 0) for 7 days, followed by a 12-day differentiation period. Representative brightfield images of 2D and 3D cells at D0, D7, D14 and D19. Scale bar 300 μm. (B) Percentage or proliferative (Ki67⁺cells) during 2D and 3D differentiation. Results are expressed as mean ± SEM for n = 4–5. No proliferation was detected in organoids. (C) Representative microscopy images of in vitro differentiated mature adipocytes (monolayer and organoids; confocal fluorescent image) and of in situ mature adipocytes (human pancreatic tissue section; brightfield chromogenic image) immunostained for the lipid droplet membrane protein perilipin-1 (PLIN1). Nuclei are stained in red. Scale bar 200 μm. (D–J) Relative mRNA levels (RT-PCR) of selected genes in the course of (left panels) cell differentiation as monolayer (grey lines) or organoids (blue lines) and (right panels; orange bars) in the primary adipocytes and SVF cells. Following genes were analyzed: (D)PPARG, (E) ADIPOQ (adiponectin), (F)ADRB1 (beta 1 adrenergic receptor), (G)ADRB2 (beta 2 adrenergic receptor), (H)ADRB3 (beta 3 adrenergic receptor), (I)INSR (insulin receptor) and (J)LEP (leptin). RPS13 was used as housekeeping. Results are presented as mean ± SEM for n = 8–10 donors. Statistical analysis was done using two-way ANOVA; &p < 0.05 vs D0 monolayer; ∗p < 0.05 vs D0 organoid; #p < 0.05 monolayer vs organoids. (K) Secreted adipokines by organoid cell cultures at D0 and D19 of differentiation. Results are expressed as mean ± SEM for n = 4 donors. (L) Pearson correlation of ADIPOQ mRNA levels at D19 with the donor’s BMI for monolayer and organoids of n = 9 independent donors.

Figure 2

Pancreatic adipose tissue organoids exhibit functional enhancements characteristic of mature adipocytes. (A–B) Lipolytic performance of pancreatic adipocytes differentiated in monolayer (white bars) or as organoids (blue bars). The cells were incubated with insulin, isoproterenol and forskolin for 3 h, supernatant was collected and fatty acids and glycerol release was measured as described in the methods. Lipolysis was measured as release of (A) free fatty acids (FFA) and of (B) glycerol and presented as fold change over respective control. Results are presented as mean ± SEM for (A) n = 5 and (B) n = 3 independent donors. (C–G) Representative western blot (C) and relative quantifications (D–G) of P^Ser660HSL, HSL, P^Ser473PKB, PKB and GAPDH, in pancreatic adipocytes differentiated in monolayer (white bars) and organoid (blue bars) cultures. Results are expressed as FC (fold change) over respective control (Con) and presented as mean ± SEM of n = 3 donors. Statistical analysis was done using two-way ANOVA ∗p < 0.05, or student’s t-test #p < 0.05.

Figure 3

Pancreatic adipose tissue organoids secrete proinflammatory chemokines. (A–C) Relative mRNA levels assessed by RT-PCR of proinflammatory cytokines (A)IL-6, (B)MCP-1 and (C)IL1B during in vitro cell differentiation (left panels) as monolayer (grey lines) or as organoids (blue lines), and in the native adipocytes and SVF cells (orange bars, right panels). RPS13 was used as housekeeping. Results are presented as mean ± SEM for n = 4–9 donors. Statistical analysis was done using two-way ANOVA &p < 0.05 vs D0 monolayer; ∗p < 0.05 vs D0 organoids; p < 0.05 # monolayer vs organoid. (D–F) Relative mRNA levels (RT-PCR) of proinflammatory cytokines (D)IL-6, (E)MCP-1 and (F)IL1B in pancreatic adipocytes differentiated (D19) as monolayer (white bars) or organoids (blue bars). The cells were preincubated in the absence or presence of the TLR4 inhibitor, CLi095 (5 μM; 1 h) before LPS treatment (100 ng/ml; 24 h). Results are presented as mean ± SEM for n = 3 donors. (G–I) Secretion of (G) IL-6, (H) MCP-1 and (I) IL1B in organoids at D0 and D19. Results are presented as mean ± SEM for n = 3 donors. ∗p < 0.05 vs D0 organoids. (J) % of CD45⁺ leukocytes of viable organoid non-adipocyte cells and (K) % of macrophages in undifferentiated (D0) and differentiated (D19) organoids. Results are expressed as mean ± SEM for n = 3 donors. Student’s t-test; ∗∗p ≤ 0.01. (L) Representative fluorescent confocal image of a pancreatic adipocyte organoid section immunostained for ATGL (white), CD68 (green). Nuclei are stained in blue. Scale bar 200 μm.

Figure 4

GIPR agonism modulates lipolysis and inflammation in pancreatic adipose tissue organoids. (A) Relative mRNA levels of GIPR (RT-PCR) during in vitro cell differentiation (left panel) as monolayer (gray line) or as organoid (blue line) culture and in the native adipocytes and SVF cells (right panel). Results are expressed as mean ± SEM for n = 4 independent experiments. Statistical analysis was done using two-way ANOVA ##p < 0.01 monolayer vs organoid. (B) Lipolytic performance of pancreatic adipocyte organoids quantified as glycerol release as described in the methods. (C, D) Representative immunoblots of (C) P^Ser660HSL and HSL and (D) relative quantification P^Ser660HSL in the organoids used for the lipolytic assays presented in (B). Results are expressed as mean ± SEM of n = 4 independent experiments. (E–I) Secretome analysis in organoids at D19 after 24 h treatment with test substances as indicated. Supernatant was collected and protein levels of (E) adiponectin, (F) adipsin, (G) IL1B, (H) IL-6 and (I) MCP-1 were measured as described in the methods. (J) Lipolytic performance of pancreatic adipocyte organoids quantified as released free fatty acids (FFA) as described in the methods. Organoids were treated during differentiation (from D7 to D19) with GIP (20 nM) or tirzepatide (10 nM). At D19, cells were starved (3 h) and exposed to GIP (100 nM), tirzepatide (10 nM) or isoproterenol (1 μM) for another 3 h. Supernatant was collected and released FFA were measured and normalized to respective protein amount. Results are presented as mean ± SEM for n = 3 donors. Statistical analysis was done using one-way ANOVA ∗p < 0.05.