An autoregulatory pathway establishes the definitive chromatin conformation at the pit-1 locus

Abstract

The transcription factor Pit-1 (POU1-F1) plays a dominant role in cell lineage expansion and differentiation in the anterior pituitary. Prior studies of the mouse Pit-1 (mPit-1) gene revealed that this master regulatory locus is activated at embryonic day 13.5 (E13.5) by an early enhancer (EE), whereas its subsequent expression throughout adult life is maintained by a more distal definitive enhancer (DE). Here, we demonstrate that the sequential actions of these two enhancers are linked to corresponding shifts in their proximities to the Pit-1 promoter. We further demonstrate that the looping of the definitive enhancer to the mPit-1 promoter is critically dependent on a self-sustaining autoregulatory mechanism mediated by the Pit-1 protein. These Pit-1-dependent actions are accompanied by localized recruitment of CBP and enrichment for H3K27 acetylation within the Pit-1 locus. These data support a model in which the sequential actions of two developmentally activated enhancers are linked to a corresponding shift in higher-order chromatin structures. This shift establishes an autoregulatory circuit that maintains durable expression of Pit-1 throughout adult life.

FIG 1

The Pit-1 locus assembles development-specific and tissue-specific chromatin configurations. (A) Map of a 33-kb segment of the mouse genome encompassing the Pit-1 locus and surrounding regions. The Pit-1 gene is shown as a single unit (introns are not displayed). The EE is indicated by the open oval. The Pit-1 binding site arrays at the DE and at the promoter are indicated by gray ovals. The BglII restriction map is indicated below the genomic diagram by tick marks. The positions of the PCR primers used for the 3C analysis are indicated by labeled directional arrows. (B) 3C analysis of the Pit-1 locus using the Pit-1 promoter as the anchor fragment. 3C assays were performed on the GHFT1 cell line and on primary somatotrope-enriched pituitary and brain tissues of adult mice. The Pit-1p (Pit-1 promoter primer) site served as the anchor for these studies and was paired with each of the five additional 5′ primers in the PCR analysis of the 3C ligation products. The ligation frequency was normalized to the random ligation control (see Materials and Methods). Each thin vertical line represents the normalized relative ligation frequency (mean ± standard deviation [SD]) from four or more independent studies. The calculation of ligation frequencies is described in Materials and Methods. (C) 3C analysis of the Pit-1 locus in the adult using the Pit-1 promoter as the anchor fragment. Details of the diagram are as in panel B.

FIG 2

ChIP assays reveal the distinct distributions of H3K27 acetylation and dynamic recruitment of CBP, Pit-1, and Pol II at the Pit-1 locus in GHFT1 cells and in the adult pituitary. (A) Diagram of the Pit-1 locus (as in Fig. 1A). The positions of the amplicons used in the ChIP analyses are shown under the map. (B) Pattern of H3K27 acetylation at the Pit-1 locus in GHFT1 cells. The ChIP assays were performed using antibody against acetylated H3K27 and preimmune IgG, respectively. The y axis represents the relative enrichment (the ratio of bound fraction to that of the input; see Materials and Methods for details). The histogram values represent the relative enrichment and SD from three independent ChIP assays. (C) Pattern of H3K27 acetylation at the Pit-1 locus in the adult pituitary. The ChIP assays were performed as described for panel B, but on chromatin from adult somatotrope-enriched mouse pituitaries. (D) Distribution of CBP at the Pit-1 locus in GHFT1 cells. The ChIP assays were performed using antibody specific for CBP or normal IgG. The analysis of the relative enrichment was as described for panel A. (E) Distribution of CBP at the Pit-1 locus in the adult pituitary. The CBP ChIP assays were performed as described for panels A and C. (F) Pol II and Pit-1 occupancy at the Pit-1 locus in GHFT1 cells. The ChIP assays were performed as described for panel A using antibodies to Pol II (Ser5P Pol II) and Pit-1, along with a preimmune IgG control. The relative enrichment was calculated as described in Materials and Methods, and 1% of the input was used for normalization. (G) Pol II and Pit-1 occupancy at the Pit-1 locus in the adult pituitary. The ChIP assays were performed as described for panel A using antibody against Pit-1 and Ser5P Pol II. **, P < 0.01; *, P < 0.05.

FIG 3

Pattern of p300 distribution at the Pit-1 locus. (A) Map of the Pit-1 locus. (B) Distribution of p300 at the Pit-1 locus in GHFT1 cells. The ChIP assays were performed as described for Fig. 2 using antibody specific for p300. (C) Distribution of p300 at the Pit-1 locus in the adult pituitary. The ChIP assays were performed as described for Fig. 2 using antibody specific for p300. **, P < 0.05. The error bars represent SD.

FIG 4

Pit-1 is associated with the interacting distal enhancer and Pit-1 promoter regions in the adult pituitary and in Pit-1/Triple cells. (A) Map of the Pit-1 locus as described for Fig. 1A. (B) ChIP-3C analysis. ChIP-3C was performed on chromatin isolated from somatotrope-enriched adult pituitary and Pit-1/Triple cells (see Materials and Methods). The Pit-1 promoter primer (Pit-1p) was used as the anchor fragment in the assay. The PCR product between the DE and Pit-1p from the Pit-1 ChIP of adult pituitary and Pit-1/Triple cross-linked chromatin were redigested by BglII to confirm their identities as ligation products (not shown). (C) Pit-1 is recruited to the distal enhancer region and the promoter at the Pit-1 locus in Pit-1/Triple cells. The ChIP assays were performed as described for Fig. 2F using an antibody specific to Pit-1. **, P < 0.01; *, P < 0.05.The error bars represent SD.

FIG 5

The definitive (maintenance) chromatin configuration at the Pit-1 locus in Pit-1/Triple cells is Pit-1 dependent. (A) Pit-1 protein depletion in Pit-1/Triple cells. Cells transfected with the indicated shRNA-bearing plasmids were incubated for 5 days and then assayed for levels of Pit-1 by Western blotting. Quantification of the Western blot signals, normalized to a lamin B loading control (Cont), is shown in the histogram. The level of Pit-1 protein in the control sample was defined as 1.0. The histogram values are the averages from two different shRNA clones. The error bars represent SD. (B) Impact of Pit-1 protein depletion on the chromatin conformation of the Pit-1 locus in Pit-1/Triple cells. The 3C assays were performed on chromatin from the Pit-1/Triple cells (see Materials and Methods). The positions of the PCR primers are labeled below the map of the Pit-1 locus. The Pit-1p primer was used as the anchor primer for the assays. The relative ligation frequency was calculated after normalization to that of the random control and the ubiquitously expressed Ercc3 locus. The primers for the Ercc3 locus were located in two adjacent BglII fragments in the mouse genome (see Materials and Methods for details).

FIG 6

The recruitment of CBP and Pol II to the definitive (maintenance) Pit-1 locus in Pit-1/Triple cells is dependent on the actions of Pit-1. (A) Pit-1 depletion by shRNA decreases recruitment of CBP and Pol II to the Pit-1 promoter. ChIP assays were performed on chromatin from Pit-1/Triple cells transfected with plasmids carrying control shRNA (Cont.) or Pit-1 shRNA. The relative enrichment was calculated as the ratio of bound fraction to the 1% input control. The histogram values represent the averages and SD from three independent experiments. (B) Pit-1 depletion decreases recruitment of CBP to the DE of the Pit-1 locus. Study details are as for panel A.

FIG 7

The chromatin configuration at the Pit-1 locus in GHFT1 cells is unaffected by Pit-1 depletion. (A) Pit-1 protein depletion in GHFT1 cells. Pit-1 shRNA transfection was as described for Fig. 5A. A Western blot analysis of Pit-1 and lamin B (loading control) is shown at the top, and a histogram of the band intensities (n = 2) is shown below the blot. The error bars represent SD. (B) 3C analysis of looping between the EE and the promoter at the Pit-1 locus in GHFT1 cells with (dashed line) and without (solid line) Pit-1 depletion. The 3C assays were performed as described for Fig. 5B.