Dynamic chromatin organization during foregut development mediated by the organ selector gene PHA-4/FoxA

Abstract

Central regulators of cell fate, or selector genes, establish the identity of cells by direct regulation of large cohorts of genes. In Caenorhabditis elegans, foregut (or pharynx) identity relies on the FoxA transcription factor PHA-4, which activates different sets of target genes at various times and in diverse cellular environments. An outstanding question is how PHA-4 distinguishes between target genes for appropriate transcriptional control. We have used the Nuclear Spot Assay and GFP reporters to examine PHA-4 interactions with target promoters in living embryos and with single cell resolution. While PHA-4 was found throughout the digestive tract, binding and activation of pharyngeally expressed promoters was restricted to a subset of pharyngeal cells and excluded from the intestine. An RNAi screen of candidate nuclear factors identified emerin (emr-1) as a negative regulator of PHA-4 binding within the pharynx, but emr-1 did not modulate PHA-4 binding in the intestine. Upon promoter association, PHA-4 induced large-scale chromatin de-compaction, which, we hypothesize, may facilitate promoter access and productive transcription. Our results reveal two tiers of PHA-4 regulation. PHA-4 binding is prohibited in intestinal cells, preventing target gene expression in that organ. PHA-4 binding within the pharynx is limited by the nuclear lamina component EMR-1/emerin. The data suggest that association of PHA-4 with its targets is a regulated step that contributes to promoter selectivity during organ formation. We speculate that global re-organization of chromatin architecture upon PHA-4 binding promotes competence of pharyngeal gene transcription and, by extension, foregut development.

Figure 1. Scanning mutagenesis of the pax-1 promoter.

(A) A cartoon depicting the pattern of PHA-4 expression during different stages of embryogenesis. Embryonic events that occur at specific developmental stages are annotated. PHA-4 expression from , myo-2 expression from and pax-1 expression from this study. (B) pax-1::GFP expression in 14 pharyngeal cells in a comma stage embryo (left) and GFP (green; anti-GFP Molecular Probes) co-stained with anti-intermediate filament antibody (magenta, right, [95]). Marginal cells and pm8 are visible, but e2 cells are faint in this image. GFP alone shown in the inset. (C) Linker scanning mutagenesis reveals two positive cis-regulatory sites: mutP (green) and mutA (magenta).% pharyngeal expression: number of independent lines with pharyngeal expression/total number of independent lines analyzed. Neg: negative. WT: wild type. (D) The architecture of the pax-1 promoter 70 bp upstream to the TSS revealing direct repeats (DR in grey) and inverted repeats (IR in blue).

Figure 2. PHA-4 associates with pharyngeal target promoters by the 8E (∼100 cell) stage.

CFP::LacI (depicted in green) and PHA-4::YFP (depicted in magenta) co-localization on pseudo-chromosomes bearing (A) myo-2 or (C) pax-1 promoters. Merge is white. Binding to pseudo-chromosomes is abolished by mutating the PHA-4 binding sites in myo-2 mutP or pax-1 mutP but is not affected when an unrelated activation site is mutated (pax-1 mutA). The cartoon illustrates the interpretation of the data. (B, D) Quantitation of embryos with co-localized CFP::LacI and PHA-4::YFP in transgenic lines bearing (B) a wild-type myo-2 promoter (solid and hatched black, WT) or one with mutated PHA-4 sites (solid and hatched green, mutP) and (D) a wild-type pax-1 promoter, a mutant promoter lacking PHA-4 binding sites (pax-1 mutP) or a mutant promoter inactivated for an unrelated activation site (pax-1 mutA) (white and dotted, mutA1). Numbers of embryos scored per stage shown in Figure S5C. Scale bar, 3 microns. Arrowheads indicate PHA-4 bound (co-localized) pseudo-chromosomes. Asterisks indicate arrays that lack associated PHA-4::YFP.

Figure 3. Decompaction of pseudo-chromosomes during pharyngeal differentiation.

(A) Pseudo-chromosomes bearing wild-type myo-2 promoters within the pharynx (P region) or outside, at the indicated stages. Decompacted (arrow) and compacted (asterisk) pseudo-chromosomes are noted. PHA-4::YFP was used to identify pharyngeal cells (not shown). Scale bar, 3 microns. (B) Cumulative distributions of areas for pseudo-chromosomes bearing wild-type myo-2 or pax-1 promoters at the indicated developmental stages. The horizontal axis represents the area of individual pseudo-chromosomes multiplied by 10. The vertical axis represents the cumulative proportion of pseudo-chromosomes with an equal or smaller area. Curves shifted to the right, indicate a greater proportion of pseudo-chromosomes with large areas, for pharyngeal cells (magenta) relative to cells outside of the pharynx (green). Areas of pseudo-chromosomes increased as embryos developed (p = 0.00003 for myo-2, p = 0.0002 for pax-1). For myo-2, n = 2 lines, 10 embryos per stage per line. For pax-1, n = 1 line, 5 embryos per stage.

Figure 4. PHA-4 is required for chromatin decompaction.

(A) Pseudo-chromosomes bearing mutated PHA-4 binding sites within myo-2 either within the pharynx (P region) or outside, at the indicated stages. PHA-4::YFP was used to identify pharyngeal cells (not shown). Scale bar, 3 microns. Cumulative distributions of pseudo-chromosome areas for (A) mutant myo-2 or (B) mutant pax-1 pseudo-chromosomes. Lines analyzed were mutated for PHA-4 binding sites within myo-2 (MutP), the PHA-4 binding site within pax-1 (MutP) or an alternative activation site within pax-1 (mutA). The horizontal axis represents the area of individual pseudo-chromosomes multiplied by 10. The vertical axis represents the cumulative proportion of pseudo-chromosomes with an equal or smaller area. Note the overlap of pseudo-chromosome areas for PHA-4-binding mutations within the pharynx (magenta) and outside of the pharynx (green), indicating no induced decompaction. For myo-2, n = 2 lines, 10 embryos per stage, per line. For pax-1, n = 1 line each mutant, 5 embryos per stage. (C) Pseudo-chromosomes bearing 3X PHA-4 binding site repeats within the pharynx (P region) or outside, at the bean stage (upper) and 2-Fold stage (lower). PHA-4::YFP was used to identify pharyngeal cells (not shown). Scale bar, 3 mm. Note the decompaction with pharyngeal cells (arrowheads) relative to non-pharyngeal cells (asterisk). (D) Quantitation of embryos with co-localized CFP::LacI and PHA-4::YFP in transgenic lines bearing 3X repeats.

Figure 5. PHA-4 binding and activity is limited in the intestine.

(A) PHA-4::YFP (depicted in magenta) does not associate with pseudo-chromosomes (marked with CFP::LacI, green) bearing the pax-1 mutA promoter in the intestine. The cartoon illustrates the interpretation of the data. (B) Over-expression of PHA-4 under a heat-shock promoter leads to widespread expression of PAX-1::GFP (green) in multiple tissues but not in the intestine (magenta; J126 (lower panel). Control embryos that did not receive heat-shock express PAX-1::GFP only in marginal cells (upper panel). (C) PHA-4 is expressed in all tissues after heat shock, including the intestine (magenta; J126). Scale bar, 10 microns.

Figure 6. Emerin inhibits PHA-4 binding in the pharynx.

(A) CFP::LacI (depicted in green) and PHA-4::YFP (depicted in magenta) co-localization on pseudo-chromosomes bearing pax-1MutA after No RNAi, emr-1(RNAi) or met-2(RNAi). (B) Percentage of embryos with at least one co-localized dot (No RNAi 9/19, emr-1(RNAi) 13/23, set-1(RNAi) 14/29, and met-2(RNAi) 16/22) (C) Percentage of pharyngeal nuclei with bound PHA-4::YFP among embryos with co-localization. After emr-1 reduction (Ω, n = 13), embryos had a ∼3 fold increase in PHA-4::YFP binding compared to No RNAi controls (n = 9). For met-2 and set-1 n = 16 and 14, respectively D) The proportion of embryos bearing de-condensed arrays for each RNAi treatment is graphed. See Table S2 for the number of embryos assayed. (E). Example of pseudo-chromosomes after emr-1(RNAi) or met-2(RNAi). Decompaction within the pharynx (arrowhead) presumably reflects PHA-4 association. Scale bar, 3 microns.