Fig 1.
Analysis of the Neurog3 misexpression efficiency in HNFN3OE mice following short-term tamoxifen induction.
(A-B) β-galactosidase activity assessment in the pancreata of HNFN3OE animals treated with vehicle (A) or Tam (B) for 2 weeks. A clear activity is noted solely in ductal cells. (C) Quantitative analysis of Neurog3 transcript levels by qPCR (n = 6 animals for each condition) outlining a 2.5-fold increase in the pancreata of Tam-treated HNFN3OE animals compared to controls. Statistics were performed using the Mann-Whitney test (D-E) By means of immunohistochemical analyses using antibodies raised against Neurog3, Neurog3-expressing cells are detected within the ductal epithelium of HNFN3OE animals treated with Tam for only 2 weeks (E), whereas Neurog3+ cells cannot be detected in their vehicle-treated counterparts (D). For clarity, when required, the ductal lumen is outlined with yellow lines and islets with red lines.
Fig 2.
The ectopic expression of Neurog3 triggered in adult pancreatic ductal cells results in hyperplastic islets with remodeled endocrine cell distribution but unchanged relative ratios.
(A-B) Mosaic photographs of pancreatic sections of control (A) and HNFN3OE animal treated with Tam for 12 months (B) and assayed for insulin expression using immunohistochemical detection. Note the increase in islet number and size in Tam-treated HNFN3OE animal (B) compared to vehicle-administered control (A). Control in A is age-matched with B. (C) HNFOE littermates were treated with Tamoxifen for the indicated duration at the specified starting ages. The islets size and number were then analyzed for each group and compared to age-matched controls. Note the progressive islet hypertrophy and increase in islet count, these depending on the duration of Tam administration. The ratio displayed in the table indicate the relative increase in islet surface and count ± SEM, n = 3 per group. (D-E) Immunostaining of islets of Langerhans from vehicle- (D) and Tam- (E) treated pancreata displaying the distribution of insulin-, glucagon- and somatostatin-expressing cells. The typical islet organization observed in controls, with a core of β-cells surrounded by non β-cells (D), appears altered in Tam-treated animals (E), with a preferential localization of α- and δ-cells at one pole of the islets, in proximity of ducts (E, S1 Movie). For clarity, the ductal lumen is outlined with yellow lines. (F) Relative distribution of the insulin-, glucagon-, and somatostatin-producing cell populations in Tam-treated HNFN3OE pancreata compared to controls (Controls n = 6, 1m n = 4, 3m n = 3, 12m n = 3). Interestingly, the increases in insulin-, glucagon-, and somatostatin-producing cell counts are found to be similar in Tam-treated HNFN3OE animals, their relative proportions remaining unchanged when compared to controls. Statistics were performed using the Mann-Whitney test or unpaired t-test with Welch’s correction.
Fig 3.
Observation of double and triple hormone-positive cells in adult Tam-treated HNFN3OE animals.
(A-D) The examination of adult Tam-treated HNFN3OE pancreata using confocal analyses outlines an abnormal localization of non β-cells, with a clustering of such cells at a pole of the islets close to ducts (A, Fig 2D and 2E, S1 Movie). The analysis of such cell clusters assessed using orthogonal projection of a Z-stack shows cells co-expressing two (white arrowheads) or even three (black arrowheads) endocrine hormones (C, D), as seen during pancreas morphogenesis. As expected, such multi-hormone-expressing cells were not detected in untreated controls (B).
Fig 4.
Lineage tracing experiments unravel the conversion of ductal cells into endocrine cells upon the sole Neurog3 misexpression.
(A) Taking advantage of the β-galactosidase tracer, we monitored the fate of the ductal cells ectopically expressing Neurog3. X-Gal staining reveals β-galactosidase-positive cells (previously ductal cells) within the islet of Langerhans (outlined with red lines) of Tam-treated HNFN3OE mice. (B) Quantitative RT-PCR analyses confirm the presence of β-galactosidase mRNA in the transcriptome of islets isolated from Tam-treated animals (n = 6 animals for each condition). Statistics were performed using the Mann-Whitney test (C-D) Immunohistochemical analyses combining β-galactosidase and insulin detection. While control pancreata are negative for β-galactosidase (C), their Tam-treated counterparts display cells positive for both insulin and β-galactosidase (D), indicating duct-to-endocrine cell conversion. Note that the apparent staining in the exocrine tissue is artefactual and caused by the antibody used. (E-F) Accordingly, several glucagon+ (E) or somatostatin+ (F) cells are also found labeled with the ductal cell tracer, β-galactosidase.
Fig 5.
Neurog3 misexpression is maintained in newly-formed endocrine cells in adult Tam-treated HNFN3OE pancreata.
The observation of islets in HNFN3OE animals treated with Tam for 10 months demonstrates that a clear majority of endocrine cells misexpress Neurog3 (A-B), this result being confirmed by qPCR analyses (C, n = 6 for each condition) Statistics were performed using the Mann-Whitney test.
Fig 6.
Phenotypical analyses of islet cells from Tam-treated HNFN3OE pancreata.
Representative pictures of immunohistochemical analyses performed on pancreas sections from untreated (A, C, E, G, I, K) and age-matched Tam-treated HNFN3OE mice (B, D, F, H, J, L) using the indicated antibody combinations. All insulin+ cells (preexisting and neogenerated) express the bona fide β-cell markers Nkx6.1 (A-B), NeuroD1 (C-D), Pdx1 (E-F), Rfx6 (G-H), Glut2 (I-J), and PC1/3 (K-L).
Fig 7.
HNFN3OE animals display a normal endocrine cell ultrastructure and exhibit an improved response upon glucose stimulation in vivo.
(A-D) Electron microscopy examinations (n = 3, 200 photographs analyzed per sample) reveal no difference in the ultrastructure of endocrine cells comparing control (A) and transgenic animals (B). Focusing on insulin-producing cells combining electron microscopy and immuno-gold labelling of insulin, all cells presenting a β-cell ultrastructure are found positive for insulin. Similarly, all cells labeled with insulin exhibit a β-cell ultrastructure (C-D and insets). (E) Upon being subjected to intraperitoneal glucose tolerance tests, HNFN3OE animals treated with Tam for one month display a lower peak in glycemia and a faster return to normoglycemia compared to their control counterparts. This improved tolerance is further enhanced following 4 months of Tam administration (Controls n = 5, 1m n = 5, 4m n = 4). (F) Assessment of basal insulin serum levels in HNFN3OE animals treated with Tam or not, indicating a normal basal insulin secretion despite an increased β-like cell mass (n = 3 for each condition). Statistics were performed using the Mann-Whitney test.
Fig 8.
Neurog3 misexpression in insulin+ cells does not alter β-cell function.
(A) A combination of immunohistochemical detection and Salmon-Gal staining was used to assess β-galactosidase activity in InsCre::Neurog3OE pancreata. 80% of insulin-producing cells appeared positive for β-galactosidase. For clarity, Salmon-Gal staining was converted to a green labeling in the photographs. (B) Intraperitoneal glucose tolerance tests of adult InsCre::Neurog3OE animals (5-month old, n = 5) do not show any difference in glucose response comparing transgenic animals and their age-matched transgene-negative littermates (n = 4), indicating that maintained Neurog3 misexpression in β-cells does not impair their ability to secrete insulin in response to glucose stimulation. Statistics were performed using the Mann-Whitney test.