The basic helix-loop-helix protein BETA2 interacts with p300 to coordinate differentiation of secretin-expressing enteroendocrine cells

Abstract

The major epithelial cell types lining the intestine comprise a perpetually self-renewing population of cells that differentiate continuously from a stem cell in the intestinal crypts. Secretin-producing enteroendocrine cells represent a nondividing subpopulation of intestinal epithelial cells, suggesting that expression of the hormone is coordinated with cell cycle arrest during the differentiation of this cell lineage. Here we report that the basic helix-loop-helix protein BETA2 associates functionally with the coactivator, p300 to activate transcription of the secretin gene as well as the gene encoding the cyclin-dependent kinase inhibitor p21. Overexpression of BETA2 in cell lines induces both cell cycle arrest and apoptosis suggesting that BETA2 may regulate proliferation of secretin cells. Consistent with this role, we observed both reentry of normally quiescent cells into the cell cycle and disrupted cell number regulation in the small intestine of BETA2 null mice. Thus, BETA2 may function to coordinate transcriptional activation of the secretin gene, cell cycle arrest, and cell number regulation, providing one of the first examples of a transcription factor that controls terminal differentiation of cells in the intestinal epithelium.

Figure 1

Differentiation of secretin enteroendocrine cells. Schematic view of the small intestine in longitudinal section. Intestinal epithelial cells differentiate from an anchored stem cell population in the proliferating crypt compartment. Cells migrate vertically as a band up the crypt–villus axis to differentiate further. Secretin cells differentiate as nondividing, isolated mucosal endocrine cells in the upper villus. Cell loss is believed to occur near the villus tip near the end of the vertical migration pathway.

Figure 2

p300 coactivates BETA2-dependent transcription. HeLa cells were transfected transiently with a secretin–luciferase reporter gene and the indicated expression plasmids in the amounts (micrograms) indicated. (A) p300 potentiates transactivation by BETA2. (B) Immunoblot showing expression levels of transfected BETA2 in cells cotransfected with p300. (C) BETA2 immunoblot of proteins immunoprecipitated from βTC3 or HeLa cells with monoclonal antibodies against p300 or normal mouse serum. (D) E1A repression of BETA2-dependent transcription. BETA2 (2 μg) and E47 (2 μg) expression plasmids were included in the experiments shown in D. Results of transfections are expressed as relative luciferase activity compared to cells transfected with the reporter gene alone. Luciferase activity was measured in cell extracts 24 hr after transfection. Values shown represent the mean ± s.e.m. of at least five individual transfection experiments.

Figure 3

The bHLH domain of BETA2 associates with a domain of p300 that overlaps its E1A-binding site. (A) BETA2 binds to the carboxy-terminal third of p300 in vitro. In vitro-translated, [³⁵S]-labeled BETA2 was tested for its ability to bind to GST fusion proteins containing the amino-terminal (lane 3), middle (lane 4), and carboxy-terminal third (lane 5) of p300 or GST alone (lane 2). The autoradiograph was exposed for 2 days, using an intensifying screen. (B) Binding of E12 to p300. In vitro-translated ³⁵S-labeled E12 was tested for its ability to bind to a GST fusion protein containing either the carboxy-terminal third of p300 (lane 2) or BETA2 (lane 3). (C) In vitro-translated ³⁵S-labeled BETA2 proteins were examined for their ability to bind to amino acids 1572–2370 of p300 expressed as a GST fusion protein. Odd-numbered lanes show ∼10% of the input protein applied to the affinity matrix; even-numbered lanes show the radiolabeled proteins captured by the GST–p300 affinity matrix. (D) Structure of the BETA2 proteins used in the binding assay in C. (E,F) Different GST–p300 fusion proteins shown in G were tested for their ability to capture either ³⁵S-labeled full-length BETA2 (E) or BETA2 (1–158) (F). (H) HeLa cells were transfected transiently with a secretin–luciferase reporter gene, expression plasmids for BETA2 and E47, and either p300 or p300del30 . Values shown are the mean ± s.e.m. of at least three independent transfections and represent the relative reporter expression vs. the activity from cells that did not receive a p300 expression vector.

Figure 4

The bHLH domain of BETA2 associates with the carboxyl terminus of p300 in vivo. (A) A GST–p300 fusion protein encoding amino acids 1572–2116 of p300 or GST alone was expressed in HIT cells. Cell extracts were mixed with glutathione–agarose beads and bound proteins examined by immunoblotting for BETA2. (B) A mammalian two-hybrid assay was carried out in HeLa cells cotransfected transiently with expression vectors for one of four BETA2–E2 DNA-binding domain fusion proteins, a p300–VP16 fusion as indicated, and a luciferase reporter containing four copies of an E2-binding site. Luciferase activity was measured 24 hr after transfection and results expressed as activity relative to extracts transfected with the BETA2–E2 fusion plasmid alone. Values shown represent the mean ± s.e.m. of at least five independent experiments.

Figure 5

BETA2 induces expression of p21, cell cycle arrest, and apoptosis. (A) BETA2 and p300 increase p21 gene expression. HeLa cells were cotransfected with a p21–luciferase reporter plasmid and with expression plasmids for BETA2, p300, and E1A as indicated. Results are expressed as luciferase activity relative to the activity seen in extracts of cells transfected with the reporter gene alone. Luciferase activity was measured in cell extracts 24 hr after transfection. Values represent the mean  ± s.e.m. of at least five individual experiments. (B) CV-1 cells were transfected with influenza virus hemaglutinin epitope-tagged BETA2 and examined for BrdU incorporation (top row), p21 coexpression (middle row), and apoptosis by TUNEL assay (bottom row). (Top row left A single cell expressing HA–BETA2 localized with a cy3-conjugated secondary antibody; (middle) staining for BrdU localized with a FITC-conjugated primary antibody; (right) a multiple exposure photomicrograph confirming absence of BrdU incorporation in the cell expressing HA–BETA2. Induction of p21 is shown in the middle row. A single cell is seen coexpressing HA–BETA2 (left) and p21 (middle) localized with a FITC- and cy3-conjugated secondary antibodies. DNA staining of the same section with Hoechst dye (right). The arrow denotes cells expressing BETA2. (Bottom row) A single cell expressing HA-BETA2 (cy3, left) undergoing apoptosis by TUNEL assay (FITC, middle). DNA staining of the same section with Hoechst dye (right). The arrow denotes cell expressing BETA2.

Figure 6

Altered cell cycle arrest and cell number regulation in the small intestine of BETA2 null mice. (A) Immunostaining for PCNA and p21 in the small intestine of 2-day-old BETA2 −/− and +/− mice stained for β-galactosidase activity. Note brown DAB staining for PCNA colocalized with blue nuclear X-gal staining in −/− mice indicating cell cycle reentry not seen in X-gal-stained cells of heterozyous mice (top row). Note loss of peroxidase staining for p21 in X-gal-stained cells of −/− mice (bottom row). Insets show magnified view of cells denoted by arrows. (Bar) 10 μm; (insets) 6 μm. (B) The number of stained cells was counted in longitudinal sections of small intestine from three mice from each group. Results were expressed as the mean  ± s.e.m.per 100-μm intestinal length. Approximately threefold more β-gal-positive cells were present in the −/− mice (P < 0.02, two-sided Student’s t-test).