Osteoprotegerin and Denosumab Stimulate Human Beta Cell Proliferation through Inhibition of the Receptor Activator of NF-κB Ligand Pathway

Abstract

Diabetes results from a reduction of pancreatic β-cells. Stimulating replication could normalize β-cell mass. However, adult human β-cells are recalcitrant to proliferation. We identified osteoprotegerin, a bone-related decoy receptor, as a β-cell mitogen. Osteoprotegerin was induced by and required for lactogen-mediated rodent β-cell replication. Osteoprotegerin enhanced β-cell proliferation in young, aged, and diabetic mice. This resulted in increased β-cell mass in young mice and significantly delayed hyperglycemia in diabetic mice. Osteoprotegerin stimulated replication of adult human β-cells, without causing dedifferentiation. Mechanistically, osteoprotegerin induced human and rodent β-cell replication by modulating CREB and GSK3 pathways, through binding Receptor Activator of NF-κB (RANK) Ligand (RANKL), a brake in β-cell proliferation. Denosumab, an FDA-approved osteoporosis drug, and RANKL-specific antibody induced human β-cell proliferation in vitro, and in vivo, in humanized mice. Thus, osteoprotegerin and Denosumab prevent RANKL/RANK interaction to stimulate β-cell replication, highlighting the potential for repurposing an osteoporosis drug to treat diabetes.

Figure 1. OPG is required for PRL-induced rodent β-cell proliferation, and enhances rodent β-cell replication in young, aged, and STZ-treated mice

(A) Percent BrdU-positive β-cells in male OPG KO and WT mice infused with Veh or PRL for 7 days. β-cell proliferation in WT-Veh mice is depicted as 100% (n=4-5 mice/group); *p<0.05 vs other three groups; (See also Figure S1). (B) Representative images of pancreatic sections stained for insulin (green) and BrdU (red) from mice injected daily for 7 days with vehicle (Veh) and 1.0μg mOPG-Fc/g body weight. Arrows indicate BrdU- and insulin-positive cells. (C) Percentage of BrdU-positive β-cells in 10-week old mice injected daily for 7 days with Veh or different doses of mOPG-Fc; (n=5-10 mice/group, >1000 β-cells counted/mouse; See also Figure S2). (D) Percentage of pHH3-positive β-cells in mice treated as in (B). (E) Percentage of BrdU-positive β-cells in mice injected every alternate day with Veh or 1.0μg/g mOPG-Fc/g for 30 days; (n=5 mice/group; See also Figure S3). (F) β-cell mass in mice treated as in (E); *p=0.05 vs Veh. (G) Percent BrdU-positive β-cells in one-year old mice injected daily with Veh or 1.0μg/g mOPG-Fc for 7 days (n=6-7 mice/group). (H) Percent BrdU-positive β-cells in MLDS mice injected daily with Veh or 1.0μg/g mOPG-Fc for 16 days (n=7-8 mice/group). (I) Percent TUNEL-positive β-cells in mice treated as in (H). (J) Diabetes incidence, defined as blood glucose >250mg/dl, in mice treated as in (H); #p<0.05 vs Veh by chi-square test. All values are presented as mean ± SEM. *p<0.05 vs Veh, except where specified otherwise.

Figure 2. OPG stimulates β-cell proliferation, and increases phosphorylation of CREB and GSK3 in rodent and human islets in vitro

(A) Percent BrdU-positive β-cells in mouse islet cell cultures treated with Veh or 0.1μg/ml of mOPG-Fc for 24h in the presence of BrdU (n=3). (B) Representative confocal images of human islet cell cultures treated with Veh or 0.1μg/ml of hOPG-Fc and BrdU for 24h, and stained for insulin (green), BrdU (red), and DAPI (blue). (C) Percent BrdU-positive human β-cells treated with Veh or different doses of hOPG-Fc for 24 or 72h; (n=4-6 human islet preps; See also Figure S4). (D) Percent Ki67-positive human β-cells as treated in (C). (E) Representative western blot analysis and quantitation of phCREB/tubulin ratio in mouse islets treated with Veh or mOPG-Fc for varying times. (F) Analysis as in (E) of phGSK3/tubulin ratio. (G) Representative western blot analysis and quantitation of phCREB/tubulin ratio in human islets treated with Veh or hOPG-Fc for varying times. (H) Analysis as in (G) of phGSK3/tubulin ratio. (n=3-6 islet preps). All values are presented as mean ± SEM. *p<0.05 vs Veh.

Figure 3. OPG induces human β-cell proliferation by interfering with RANKL/RANK, which acts as a brake in β-cell replication

(A) Competition assay in which human islet cell cultures treated with either Veh (−), 0.1μg/ml of hOPG-Fc, or hRANKL, or the combination of hRANKL/hOPG-Fc at a 1:1, 1:5, or 5:1 ratio, and BrdU for 24h, were assessed for percent BrdU-positive β-cells (Veh as 100%; n=3-8 human islet preps); *p<0.05 vs the four remaining groups. (B) Representative images of islet cell cultures from RANK-lox mice transduced with Ad-LacZ or Ad-Cre for 72h, and stained for insulin (red), Ki67 (green), and DAPI (blue). (C) Percent Ki67-positive β-cells in Ad-LacZ (as 100%) and Ad-Cre transduced RANK-lox islet cells (n=3); *p<0.05 vs Ad-LacZ. All values are presented as mean ± SEM.

Figure 4. DMB increases human β-cell proliferation in vitro and in vivo

(A) Percent BrdU-positive β-cells in human islet cell cultures treated with Veh (−) or different concentrations of DMB, and BrdU for 24h. (B) Human islet cell cultures treated with Veh or 0.1μg/ml of DMB for 24h, and stained for insulin (red), Ki67 (green), and DAPI (blue). (C) Percent Ki67-positive human β-cells treated with Veh (−) (as 100%) or 0.1μg/ml of DMB for 24 or 72h; (n=3-5 human islet preps). (D) Experimental design of human islets transplanted under the kidney capsule of male euglycemic NOD/SCID mice. (E) Representative images of human islet grafts in the kidney capsule of Veh or DMB (1μg/g) treated mice stained for insulin (green) and BrdU (red). Arrow represents BrdU-positive β-cell. (F) Percent BrdU-positive β-cells in human islet grafts of Veh and DMB-treated mice (n=5 human islet preps; 1427±121 β-cells counted/prep). All values are presented as mean ± SEM. *p<0.05 vs Veh.