RSPO1, a potent inducer of pancreatic β cell neogenesis

Abstract

Inducing the neogenesis of pancreatic insulin-producing β cells holds great promise for diabetes research. However, non-toxic compounds with such activities remain to be discovered. Herein, we report the identification of RSPO1, a key agonist of the Wnt/β-catenin pathway, as an inducer of β cell replication. Specifically, we provide evidence that RSPO1 promotes a significant increase in β cell neogenesis in vitro, ex vivo, and in vivo. Importantly, RSPO1 administration is sufficient to activate Wnt/β-catenin signaling in β cells and counter chemically induced or autoimmune-mediated diabetes. Similarly, an optimized analog of RSPO1, allowing for weekly administration, also prevents diabetes in vivo. Lastly, the treatment of transplanted human islets with RSPO1 induces a significant 2.78-fold increase in human β cell numbers in only 60 days, these cells being functional. Such activities of RSPO1 to promote β cell neogenesis could therefore represent an unprecedented hope in the continued search for diabetes alternative therapies.

Keywords: Rspo1; Wnt/β-catenin signaling; diabetes; endocrine pancreas; islets of Langerhans; β cell replication.

Graphical abstract

Figure 1

Rspo1 and Lgr4 expression and localization in the mouse pancreas (A) Determination of Rspo1 transcript levels by RT-qPCR in pancreatic cell lines. (B) Assessment of Rspo1 transcripts in whole pancreas versus isolated murine islet extracts. (C and D) Rspo1 localization within the pancreas using RNAscope. 21-day-old WT pancreata labeled with a probe specifically recognizing the Rspo1 transcript (C). Pancreatic section from the same animal labeled using a negative control probe (D). (E) RT-qPCR analyses of four different pancreatic cell lines to characterize Lgr4 expression within the mouse pancreas. (F) Determination of Lgr4 expression in whole pancreas and isolated mouse islets. All data shown represent mean ± SEM of n = 5. Results were considered significant if p < 0.0001 (∗∗∗∗), p < 0.001 (∗∗∗), p < 0.01 (∗∗), and p < 0.05 (∗) using a one-way ANOVA (A, E, and F) or a Student’s t test (B). See also Figure S1.

Figure 2

RSPO1 induces β cell replication in MIN6 cells, in isolated islets, and in adult WT mice (A) Assessment of MIN6 cell number upon 24-h incubation with increasing doses of native RSPO1 or saline. (B) Representative photographs of pancreatic islets isolated from WT adult mice incubated with either RSPO1 (at 0.4, 1, or 2 μM) or saline. Aiming to label replicating cells, isolated islets were co-incubated with BrdU during the last 24 h and then stained for insulin (red) and BrdU (green). (C) Quantification of the percentage of BrdU⁺ β cells in control islets and islets incubated for 72 h with increasing doses of RSPO1. (D) Pancreatic sections obtained from adult WT mice administered intraperitoneally for 5 consecutive days with different doses of RSPO1 and stained for insulin (red) and Ki67 (green). (E) Quantitative assessment of Ki67⁺ cells per islet after 5 consecutive RSPO1 administrations. All data shown represent mean ± SEM of n = 5. Results were considered significant if p < 0.0001 (∗∗∗∗), p < 0.001 (∗∗∗), p < 0.01 (∗∗), and p < 0.05 (∗) using one-way ANOVA. See also Figure S2.

Figure 3

RSPO1 stimulates β cell replication via the activation of Wnt signaling The putative activation of the canonical Wnt signaling pathway upon RSPO1 incubation was assessed in vitro. (A) MIN6 cell number quantification upon incubation for 24 h with increasing doses of two RSPO1 mutants Lgr4^M and Znrf3^M. (B) MIN6 cell number assessment upon a 24-h treatment with saline, 0.1% DMSO, 100 nM MSAB, or 500 nM cardamonin and co-incubation of MIN6 cells with either MSAB or cardamonin and native RSPO1 at 400 nM. (C) Quantification of BrdU⁺ cells in ex vivo murine islets incubated for 72 h with saline, 100 nM MSAB, 1 μM RSPO1, or a combination of the 2. (D) β-catenin protein levels upon RSPO1 treatment assessed by ELISA following a 3-h incubation with saline, native RSPO1, RSPO1 Lgr4^M, or RSPO1 Znrf3^M at 400 nM. Data shown represent mean ± SEM of n = 5. Results were considered significant if p < 0.0001 (∗∗∗∗), p < 0.001 (∗∗∗), p < 0.01 (∗∗), and p < 0.05 (∗) using one-way ANOVA. See also Figure S3.

Figure 4

RSPO1 administration protects mice from developing hyperglycemia in different T1D models (A) Glycemia follow-up of two-month-old WT mice administered intraperitoneally for 80 consecutive days with either saline or 0.8 mg/kg of RSPO1 and co-treated with 50 mg/kg of streptozotocin (STZ) from day 16 to day 18 (n = 10). (B) IPGTT performed following 74 days of RSPO1 (or saline) treatment (n = 10). (C) Glycemia follow-up of six-week-old WT mice first administered for 3 consecutive days (1–3) with 50 mg/kg of STZ and subsequently injected daily with either saline or RSPO1 at 0.8 mg/kg once their glycemia reached 250 mg/dL (approximately at day 7) (n = 10). (D) IPGTT performed 63 days after first STZ administration (n = 10). (E) Weekly monitoring of random glycemia in 10-week-old NOD females injected intraperitoneally daily for 18 weeks with either saline or 0.4 or 0.8 mg/kg of RSPO1 (n = 15 in control group, n = 10 to 13 in treated groups). (F) Quantification of the whole β cell mass in NOD mice administered daily with either saline or 0.4 or 0.8 mg/kg of native RSPO1. The β cell mass of 10-week-old NOD females was used to evaluate the insulin⁺ area at the beginning of the study (n = 15 in control group, n = 10 to 13 in treated groups). All data shown represent mean ± SEM. Results were considered significant if p < 0.0001 (∗∗∗∗), p < 0.001 (∗∗∗), p < 0.01 (∗∗), and p < 0.05 (∗) following a two-way ANOVA (A–D); an unpaired Student’s t test or Mann-Whitney test (B), a mixed-effect model (restricted maximum likelihood, REML) with a Dunnett’s comparison test (C); and a one-way ANOVA, a Mann-Whitney test, or a Kruskal-Wallis test (E). See also Figures S4 and S5.

Figure 5

An FC-coupled RSPO1 induces pancreatic β cell neogenesis in vitro, ex vivo, and in vivo (A) Assessment of MIN6 cell number upon 24-h incubation with saline or increasing doses of FC-coupled RSPO1 protein (n = 5). (B) Quantification of the percentage of BrdU+ cells in saline-treated islets and islets incubated for 72 h with 200 nM or 1 or 3 μM FC-RSPO1 (n = 5). (C) Weekly monitoring of random glycemia in 10-week-old NOD females injected intraperitoneally weekly for 18 weeks with either saline or 2.4 mg/kg of FC-RSPO1 (n = 12 in control group, n = 10 in treated group). (D) Quantification of the whole β cell mass in NOD mice weekly administered with either saline or 2.4 mg/kg of FC-coupled RSPO1. The β cell mass of 10-week-old NOD females was used to evaluate the insulin⁺ area at the beginning of the study (n = 12 in control group, n = 10 in treated group). (E–H) Pancreatic sections from FC-RSPO1-treated NOD mice stained for the β cell markers Pdx1 (E), PC1/3 (F), Nkx6.1 (G), and Glut2 (H). All data shown represent mean ± SEM. Results were considered significant if p < 0.0001 (∗∗∗∗), p < 0.001 (∗∗∗), p < 0.01 (∗∗), and p < 0.05 (∗) following a one-way ANOVA (A and B); a one-way ANOVA, a Mann-Whitney test, or a Kruskal-Wallis test (C); or an unpaired Student’s t test (D).

Figure 6

The administration of RSPO1 to immunodeficient mice transplanted with human islets induces functional human β cell neogenesis (A) IPGTT on transplanted RAG12N2 mice performed 28 days following daily saline or RSPO1 treatment initiation (0.4 and 0.8 mg/kg). (B) Assessment of human c-peptide circulating levels in control and RSPO1-treated mice at basal level and 15 min after glucose administration (n = 4). (C) Representative image of an immunohistochemical assay performed on grafted human islets following 60 days of RSPO1 administration (0.4 mg/kg) and labeling insulin in green and BrdU in red. (D) Quantitative analyses showing the percentage of human BrdU⁺/Ins⁺ double-positive cells in control grafts and in grafts obtained from mice treated with either 0.4 or 0.8 mg/kg of RSPO1. (E) Immunohistochemical quantification of the total insulin volume (mm³) within the transplants of control RAG12N2 mice and mice treated daily with either 0.4 or 0.8 mg/kg of RSPO1. (F) Correlation between the mean of the human c-peptide secretion (y axis) and the β cell mass (volume of insulin, x axis) calculated per animal. All data shown in (A), (B), (D), and (E) represent mean ± SEM of n = 4. Results were considered significant if p < 0.0001 (∗∗∗∗), p < 0.001 (∗∗∗), p < 0.01 (∗∗), and p < 0.05 (∗) following a two-way ANOVA (A and B) or a one-way ANOVA (D and E). See Figure S6.