XB-ART-39598Dev Biol April 15, 2009; 328 (2): 384-91.
Remodeling of insulin producing beta-cells during Xenopus laevis metamorphosis.
Insulin-producing beta-cells are present as single cells or in small clusters distributed throughout the pancreas of the Xenopus laevis tadpole. During metamorphic climax when the exocrine pancreas dedifferentiates to progenitor cells, the beta-cells undergo two changes. Insulin mRNA is down regulated at the beginning of metamorphic climax (NF62) and reexpressed again near the end of climax. Secondly, the beta-cells aggregate to form islets. During climax the increase in insulin cluster size is not caused by cell proliferation or by acinar-to-beta-cell transdifferentiation, but rather is due to the aggregation of pre-existing beta-cells. The total number of beta-cells does not change during the 8 days of climax. Thyroid hormone (TH) induction of premetamorphic tadpoles causes an increase in islet size while prolonged treatment of tadpoles with the goitrogen methimazole inhibits this increase. Expression of a dominant negative form of the thyroid hormone receptor (TRDN) driven by the elastase promoter not only protects the exocrine pancreas of a transgenic tadpole from TH-induced dedifferentiation but also prevents aggregation of beta-cells at climax. These transgenic tadpoles do however undergo normal loss and resynthesis of insulin mRNA at the same stage as controls. In contrast transgenic tadpoles with the same TRDN transgene driven by an insulin promoter do not undergo down regulation of insulin mRNA, but do aggregate beta-cells to form islets like controls. These results demonstrate that TH controls the remodeling of beta-cells through cell-cell interaction with dedifferentiating acinar cells and a cell autonomous program that temporarily shuts off the insulin gene.
PubMed ID: 19389350
PMC ID: PMC3863375
Article link: Dev Biol
Species referenced: Xenopus
Genes referenced: cela1.6 ins neurog1 neurog3 pcna trdn
Antibodies: Ins Ab2
Article Images: [+] show captions
|Fig. 2. Aggregation of pre existing β-cells give rise to bigger islets at the end of climax. (Aa–c) β-cell proliferation at different stages of metamorphosis. BrdU was injected intraperitoneally into staged animals every other day for 10 days and pancreases were collected on the 11th day from (a) NF55 tadpoles; (b) climax (NF59–NF66) and (c) growing froglets. BrdU (green) and insulin (red) were detected by antibodies. Cell cycle activity monitored by immunostaining with a PCNA antibody (green) and insulin (red). (d) NF55 tadpoles; (e) climax (NF59–NF66) and (f) growing froglets. (B) pElastase-GFP transgenic animal. (a–c) in situ hybridization of GFP mRNA; (d–f) immunohistochemistry for GFP (green) and insulin (red). (a, d) NF55; (b, e) NF62; (c, f) NF66. Scale bar 40 μm.|
|Fig. 3. Aggregation of β-cells is dependant on exocrine remodeling. Simultaneous immunohistochemistry of insulin (red) and GFP (green) of transgenic frogs at NF66. (A–C) pInsulin-TRDN-GFP; (D–F) pElastase-TRDN-GFP. The third panel of each (C, F) merge the first two panels. (G) Quantitative measurement of islet size and pie-distribution of islet cluster at NF66 of wild type (represented from Fig. 1 for comparison) and the two transgenic frogs. Scale bar 40 μm.|
|Fig. 4. Insulin mRNA in situ hybridization of control and transgenic pancreases at varying stages. These reactions were carried out together and developed for the same time. (A) Control premetamorphic pancreas has small islets that stain for insulin mRNA with similar intensity as (D) adult islet cells. (B) At NF62 most of the insulin mRNA is down regulated and reexpression begins at (C) the end of climax NF66. The pInsulin-TRDN transgenic pancreas does not down regulate insulin mRNA at (E) NF62 or (F) NF66, but its ability to aggregate by NF66 is not impaired. (G) pElastase-TRDN-GFP at NF66 showed similar insulin mRNA down regulation as that of wild-type but the β-cells do not aggregate. Scale bar equals 40 μm.|
|Fig. 5. Screening of progenitor cell markers revealed the existence of (A) ngn-3 positive cells at metamorphic climax. (a) Ngn-3 in situ at NF66; (b) insulin immunohistochemistry on the same section and (c) merger of a and b. Ngn3 cells never co localize with insulin staining cells. (B) Quantitative estimation of ngn3 positive cells at different developmental stages of control and the two different strains of transgenic animals. Scale bar equals 40 μm.|
|Fig. 1. Clustering of β-cells during spontaneous and T3-induced metamorphosis of X. laevis. (A) premetamorphic tadpoles (NF54); (B) froglet (NF66). A and B panels are stained with an insulin antibody (red) and DAPI (blue). (C) Average number of insulin cells per islet increases as the animal proceeds to metamorphosis. The pie-diagram above each bar represents the distribution of the different sized clusters marked with different colors as described in the insert. For example the blue color represents the fraction of total islets that consist of 1–5 β-cells. (a) The average cluster sizes at prometamorphosis and in the methimazole arrested animals is significantly higher (p < 0.05) than that of premetamorphic stages. Rearing tadpoles in methimazole (MMI) inhibits TH-induced changes including enlargement of islets. (b) The cluster size in froglets is larger than any of the tadpole islet sizes. (D) Exogenous triiodothyronine (T3) induces an increase in β-cell cluster size in premetamorphic tadpoles. Premetamorphic tadpoles were exposed to 10 nM T3 in rearing water and the insulin cluster sizes were calculated at 3 and 6-day of exposure and compared with the cluster size of control animals. There was a significant increase in cluster size on the 6th day of T3 treatment (p < 0.05). Scale bar (A and B) 20 μm.|
References [+] :
Accordi, Programmed cell death in the pancreas of Bufo bufo during metamorphosis. 2001, Pubmed