A pituitary-specific enhancer of the POMC gene with preferential activity in corticotrope cells

Abstract

Cell-specific expression of the pituitary proopiomelanocortin (POMC) gene depends on the combination of tissue- and cell-restricted transcription factors such as Pitx1 and Tpit. These factors act on the proximal POMC promoter together with transcription factors that integrate inputs from signaling pathways. We now report the identification of an upstream enhancer in the POMC locus that is targeted by the same subset of transcription factors, except Pitx1. This enhancer located at -7 kb in the mouse POMC gene is highly dependent on Tpit for activity. Whereas Tpit requires Pitx1 for action on the promoter, it acts on the -7-kb enhancer as homodimers binding to a palindromic Tpit response element (TpitRE). Both half-sites of the TpitRE palindrome and Tpit homodimerization are required for activity. In vivo, the enhancer exhibits preferential activity in corticotrope cells of the anterior lobe whereas the promoter exhibits preference for intermediate lobe melanotropes. The enhancer is conserved among different species with the TpitRE palindrome localized at the center of conserved sequences. However, the mouse and human -7-kb enhancers do not exhibit conservation of hormone responsiveness and may differ in their relative importance for POMC expression. In summary, pituitary expression of the POMC gene relies on an upstream enhancer that complements the activity of the proximal promoter with Tpit as the major regulator of both regulatory regions.

Fig. 1.

Identification of a novel POMC regulatory region. A, Whole-genome Stat3 ChIP-chip analyses in AtT-20 (18) cells reveal recruitment of Stat3 at the POMC promoter and in a region located about 7 kb upstream (green boxes, top diagram). Blue boxes and lines represent the POMC gene exons and introns, respectively. The “mammalian conservation” diagram represents sequence conservation around the POMC locus derived from 19 available mammalian species using mm9 assembly on the UCSC genome browser (32, 33). The position of previously identified hypothalamic regulatory elements nPE1 and nPE2 (16) as well as that of the promoter and −7-kb enhancer are indicated by black boxes. B, Schematic representation of the −7-kb enhancer indicating putative regulatory elements and transcription factor binding sites as identified by in silico analyses with MatInspector software (Genomatix) C, ChIP analyses of the −7-kb POMC enhancer for recruitment of the indicated transcription factors. ChIP experiments performed in AtT-20 cells used hormone-stimulated cells for Nurr1 (+CRH), Stat3 (+LIF), and GR (+Dex). ChIP enrichments are normalized relative to a control IgG ChIP and an unbound control region of the MyoD gene. Data are representative of at least three independent experiments. ^#Pitx1 ChIP is the only one to exhibit no significant recruitment. D, Schematic representation of the POMC promoter indicating the position of known regulatory elements. E, ChIP analyses of transcription factor recruitment to the POMC promoter in the same experiment as presented for the enhancer in C. All recruitments are significant compared with IgG control.

Fig. 2.

The −7-kb POMC conserved region has enhancer properties. A, Activities of luciferase reporter constructs containing either 6.9- or 7.7-kb upstream 5′ flanking sequences upon transfection in AtT-20 cells. B, Activity (relative to minimal promoter construct) of various luciferase reporters upon transfection of AtT-20 cells. The light gray box represents the −480/−34-bp proximal promoter while the darker gray box represents the enhancer located at −7 kb in the POMC locus. C, Hormone response of POMC luciferase reporters containing either minimal (−34/+63-bp) promoter, −7-kb enhancer, −480-bp proximal promoter or −7.7-kb 5′ flanking sequences. Activation is relative to nontreated cells (Ctl). All data represent the means (± sem) of independent experiments [n = 3 (A), 7 (B), and 4 (C)] each performed in duplicates.

Fig. 3.

Enhancer activity is dependent on palindromic Tpit binding site (TpitREpal). A, DNA sequence of the mouse and human TpitREpal elements of the −7-kb enhancer compared with a consensus TbxRE. The nucleotide changes used to create mutations in each or both half sites of the TpitREpal are indicated. B, Activity of enhancer reporters upon transfection into AtT-20 cells. The X over the TpitRE boxes indicates mutagenesis of either or both half sites. C, Activity of various reporters upon transfection in AtT-20 cells. The X in the reporter diagram indicates the element(s) that have been mutated in either enhancer and/or promoter. D, Activity of oligonucleotide luciferase reporters upon transfection into AtT-20 cells. The reporters contain either consensus TbxRE or TpitREpal oligonucleotides or their mutants as indicated. All data represent the means (± sem) of independent experiments [n = 5 (B), 3 (C), and 3 (D)] each performed in duplicates.

Fig. 4.

Activation of luciferase reporter constructs by Tpit in Tpit-deficient mouse αT3 cells. A, Tpit dose-response curves for activation of POMC luciferase reporters represented schematically on left. B, Effect of wild-type or dimerization-deficient M86R Tpit (50 or 250 ng of expression vectors) on activity of indicated reporters. C, Dose-responses to Tpit and Tpit M86R of −7.7-kb at −6.9-kb POMC promoter reporters. D, Responsiveness of oligonucleotide reporters to either Tpit or Tpit M86R. All data are presented as means (± sem) of at least three independent experiments, each performed in duplicates.

Fig. 5.

Tpit dimers bind the TpitREpal in vitro. A, Gel retardation was used to assess binding of either wild-type (WT) Tpit or mutant M86R to the consensus TbxRE palindrome and to TpitREpal probes, or to probes mutated for each half-site of the TpitREpal, as indicated. Equivalent amounts of ³²P-labeled probes were used. The first panel is an overnight film exposure, whereas the last three panels are 3 d exposures of the same gel. Tpit binding to the consensus TbxRE probe is stronger than the TpitREpal (compare lanes 2 and 7), but Tpit:DNA complexes are supershifted by Tpit antibody in both cases (lanes 3 and 8). The M86R Tpit dimerization-mutant exhibit monomeric binding on the consensus TbxRE, contrasting with an absence of TpitREpal probe retardation (lanes 4 and 9). Mutant probes for either half-site of the TpitREpal do not exhibit any binding (lanes 11–20).

Fig. 6.

Preferential activity of −7-kb enhancer transgene in corticotropes compared with promoter bias toward melanotropes. A, The different LacZ transgene reporters used to assess activities of POMC promoter, enhancer, or both together, are represented schematically at the top together with the number of transgenic e17.5 pups that express the transgene over the total number of transgenic pups analyzed. Coimmunofluorescence for Tpit (green) and βGal transgene (red) is shown for three different pituitaries harboring each of the three transgenes analyzed. Insets show magnifications of colocalization in anterior lobe. B, Quantitation of βGal transgene reporter expression in AL corticotropes relative to IL melanotropes. The expression of βGal transgenes was scored over the total number of Tpit-positive cells in AL and IL separately, and their ratio is presented for each transgene. For each expressing transgenic embryo, at least four pituitary sections were scored. Pairwise P values were calculated using t test for the enhancer constructs relative to the −480-bp promoter construct. Whereas the −480-bp promoter exhibits a preference for IL melanotrope expression, both reporters containing the enhancer exhibit preference for AL corticotropes.

Fig. 7.

Properties of human −7-kb POMC enhancer. A, Sequence comparison of conserved Tpit regulatory elements of the promoter and −7-kb enhancer. The TpitREpal of the −7-kb enhancer is more conserved across species than the Tpit/PitxRE of the promoter. TpitRE sequences are shown in blue and PitxRE in red. B, Activity of luciferase reporters containing either human (Hs) POMC promoter, −7-kb enhancer or both upon lipofection into AtT-20 cells. Results are the means ± sem of four independent experiments each performed in duplicate. C, Activation of indicated luciferase reporters by increasing amounts of Tpit expression vector upon lipofection into αT3 cells. A representative of two independent experiments is shown. D, Activation of rat, mouse, and human Tpit/PitxRE reporters by indicated amounts of Tpit expression plasmid after lipofection in αT3 cells. Conserved nucleotides within the Tpit binding site are indicated by capital letters. E, Schematic representation of the human −7-kb enhancer indicating the position of predicted regulatory elements/transcription factor binding sites as identified by in silico analyses with MatInspector software (Genomatix). F, Hormone response of the indicated luciferase reporters upon lipofection into AtT-20 cells. Representative data of four independent experiments, each performed in duplicate (± sem).