(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .



The activity of engrailed imaginal disc enhancers is modulated epigenetically by chromatin and autoregulation [1]

['Yuzhong Cheng', 'Division Of Developmental Biology', 'Eunice Kennedy Shriver National Institute Of Child Health', 'Human Development', 'National Institutes Of Health', 'Bethesda', 'Maryland', 'United States Of America', 'Fountane Chan', 'Judith A. Kassis']

Date: 2023-12

engrailed (en) encodes a homeodomain transcription factor crucial for the proper development of Drosophila embryos and adults. Like many developmental transcription factors, en expression is regulated by many enhancers, some of overlapping function, that drive expression in spatially and temporally restricted patterns. The en embryonic enhancers are located in discrete DNA fragments that can function correctly in small reporter transgenes. In contrast, the en imaginal disc enhancers (IDEs) do not function correctly in small reporter transgenes. En is expressed in the posterior compartment of wing imaginal discs; in contrast, small IDE-reporter transgenes are expressed mainly in the anterior compartment. We found that En binds to the IDEs and suggest that it may directly repress IDE function and modulate En expression levels. We identified two en IDEs, O and S. Deletion of either of these IDEs from a 79kb HA-en rescue transgene (HAen79) caused a loss-of-function en phenotype when the HAen79 transgene was the sole source of En. In contrast, flies with a deletion of the same IDEs from an endogenous en gene had no phenotype, suggesting a resiliency not seen in the HAen79 rescue transgene. Inserting a gypsy insulator in HAen79 between en regulatory DNA and flanking sequences strengthened the activity of HAen79, giving better function in both the ON and OFF transcriptional states. Altogether our data suggest that the en IDEs stimulate expression in the entire imaginal disc, and that the ON/OFF state is set by epigenetic memory set by the embryonic enhancers. This epigenetic regulation is similar to that of the Ultrabithorax IDEs and we suggest that the activity of late-acting enhancers in other genes may be similarly regulated.

Genes that control development are often used at different times and places in a developing embryo. Transcription of these important genes must be tightly regulated; therefore, these genes often have large arrays of regulatory DNA. In Drosophila, discrete fragments of DNA (enhancers) can be identified that turn genes on in patterns in the early embryo. In cells where the genes are transcriptionally ON, there are active modifications on chromatin, setting later enhancers in a transcription-permissive environment. In cells where the genes are OFF, repressive chromatin marks keep later enhancers inactive. In this paper we studied two late enhancers of the Drosophila en gene. We show that the correct activity of these enhancers is dependent on being next to other, earlier acting en enhancers. Our data also show that En can repress its own expression, likely directly by acting on these late enhancers. The chromatin-regulated activity of these en late enhancers is similar to what was described for a late enhancer of another Drosophila developmental gene, Ubx. We suggest that this mode of regulation is likely to be important for many late-acting developmental enhancers in many different organisms.

Funding: This work and all authors were funded by the Intramural Research Program of the Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

This is an open access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.

Here we study the activity of two en IDEs using three approaches 1) testing their activities in small transgenes 2) deleting them from HAen79, a 79kb transgene with HA-tagged En, that can rescue inv-en double mutants [ 27 ], and 3) deleting them from invΔ33, a chromosome that contains a 33kb deletion of inv DNA, creating a mimic at the endogenous en locus of the sequences present in HAen79 (called en80 in [ 27 ]). Our results suggest that the En protein directly represses its own expression through the imaginal disc enhancers and other sequences within the inv-en domain. Deletion of either imaginal disc enhancer from the HAen79 transgene causes a loss-of-function en phenotype, showing that these fragments are IDEs for en. In contrast, the same deletions do not cause phenotypes when deleted from the invΔ33 endogenous locus. Altogether our experiments show that the function of the imaginal disc enhancers is regulated by the chromatin environment of the endogenous inv-en domain.

en exists in a gene complex with invected (inv). inv encodes a closely related homeodomain protein that is largely co-regulated with en [ 17 , 18 ]. In the ‘OFF’ transcriptional state, H3K27me3, the repressive chromatin mark put on by the Polycomb protein complex PRC2, covers the entire inv-en domain, showing that inv-en is a target for Polycomb-mediated repression ([ 19 , 20 ]. Consistent with this, Polycomb group genes (PcG) are required to silence inv-en expression where they are not normally expressed in embryos and imaginal discs [ 21 – 24 ]. In our dissection of inv-en regulatory DNA we found two fragments of DNA that acted as enhancers of reporter genes in imaginal discs [ 12 ] but, unexpectedly, the reporter genes were expressed more strongly in the anterior compartment, the opposite of where En is expressed. Previous studies showed that overexpression of En via an inducible transgene can silence En expression in imaginal discs [ 25 , 26 ]. We hypothesized that when the en IDEs were outside of the inv-en domain they 1) were not silenced in the anterior compartment by PcG repressive marks and 2) were not covered by active chromatin marks in the posterior compartment and were susceptible to repression by En ( Fig 1 ).

The Drosophila engrailed (en) gene encodes a homeodomain transcription factor whose best-known functions are in embryonic segmentation and specification of the posterior compartment in larval imaginal discs, precursors of the external structures of the adult [ 8 – 10 ]. En is expressed in the embryo in a series of stripes in the ectoderm, and subsets of cells in the central and peripheral nervous systems, hindgut, fat body, posterior spiracles, and head [ 11 ]. Using a reporter gene in transgenic flies, we identified 20 embryonic enhancers spread over a 66kb region including DNA upstream, within, and downstream of the 4kb en transcription unit [ 12 ]. However, we were unable to identify a fragment of DNA that drove expression of a reporter gene in the posterior compartment of imaginal discs in an en-like pattern. We speculated that, like the imaginal disc enhancers of the Ultrabithorax (Ubx) gene [ 13 – 16 ], the ‘ON-OFF’ state of the en imaginal disc enhancers is set by the embryonic expression pattern and remembered throughout development through epigenetic memory; without this epigenetic memory, the en imaginal disc enhancers could not regulate a reporter gene in the appropriate pattern ( Fig 1 ).

Developmentally important transcription factors are expressed in spatially and temporally restricted patterns in the precursors of many different cell types. These complex gene expression patterns are generated by a large number of enhancers, traditionally defined by their abilities to stimulate patterned gene expression in transgenes (reviewed in [ 1 ]). Many developmental genes have so-called “shadow enhancers”; that is, more than one enhancer that can drive transcription in a similar pattern. Enhancers with overlapping functions are thought to impart robustness to transcription of these important genes [ 2 – 5 ]. In addition to pattern setting enhancers (which contain binding sites for both transcriptional activators and repressors [ 1 ]), developmental genes are regulated by the Polycomb (PcG) and Trithorax group genes (TrxG). Studies in Drosophila show that PcG and TrxG genes can impart a memory of the early pattern by setting the chromatin in an ON or OFF transcriptional state (reviewed in [ 6 , 7 ]). We are interested in how chromatin environment influences the enhancer activity of developmental genes.

Results

The inv and en genes are contained within a 113kb domain flanked by the genes E(Pc) and tou (Fig 2A). en is required for both embryonic and adult development. In contrast, the inv gene is not required for viability or fertility in the laboratory [18]. In many experiments in this paper, we use either a large transgene (HAen79) or a mutated inv-en domain (invΔ33) that encode no Inv protein to study the function of the imaginal disc enhancer (IDE) (Fig 2A). Table 1 contains a list of the transgenes and inv-en mutants used in our experiments. Inv and En are co-expressed in embryos and imaginal discs (Fig 2B) [12,18]. In some experiments with transgenes, we examined Inv expression from the wildtype inv-en domain in order to compare expression of the endogenous locus with the HA-en transgene (see below).

PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 2. Map of invΔ33, transgenes, PREs and imaginal disc enhancers. (A) Diagram of the inv-en region of the genome with flanking genes. The black boxes labeled O and S are the locations of the IDEs studied in this paper. Vertical lines show the locations of the constitutive inv and en PREs. The arrows denote the direction and extent of the transcription units. The DNA deleted in invΔ33, enΔ110, and enE is shown by dotted lines. Bottom, the extent of the DNA present in the two large transgenes used in this study is shown by black lines. In these transgenes, En is labeled on the N-terminus with a single HA-tag [12]. (B) Expression pattern of En and Inv in a wild-type wing disc. A fate map of a wing imaginal disc is shown on the right. A-anterior, P-posterior, D-dorsal, V-ventral. Diagram is from [46]. https://doi.org/10.1371/journal.pgen.1010826.g002

A strong correlation between En protein and repression of the S enhancer We used the variegated expression of HA-en from HAen79ΔS as a tool to address the correlation between En expression and repression of the S enhancer. We constructed a genotype HAen79ΔS enΔ110/ HAen79ΔS enE; S-Gal4@attP2/+ and examined En and Gal4 distribution in wing discs. enE is a 41kb deletion that removes En and produces a truncated Inv protein that lacks the homeodomain (Fig 2A). In this background, the only source of En is from the HAen79ΔS transgene. Strikingly, in the posterior compartment of the wing pouch, S-Gal4 is repressed in the cells where En is expressed (Fig 4B). These data, along with ChIP data that show En binding to S, support the hypothesis that En can directly repress the S-enhancer in the wing pouch.

One copy of the HAen79 transgene is haploinsufficient Flies survive well with one copy of the inv-en domain and have no known phenotypes. That is not the case with the HAen79 transgenes. While homozygous HAen79 enΔ110 flies survive with only minor wing defects, flies with only one copy of HAen79 in a homozygous enΔ110 background have wing defects and survive poorly (S4 Fig and Table 2). Some HAen79 enΔ110/enΔ110 flies hatch and then stick to the sides of the vial or fall in the food and die, suggesting a defect in nervous system development. In contrast, HAen79ΔO or HAen79ΔS enΔ110/three different inv-en deletions (enΔ110, enE and enX31) die as pharate adults with severe leg defects, and wings that are usually not expanded (S4C Fig). A rare HAenΔO enΔ110/enE fly with an expanded wing showed a lack of wing veins throughout most of the wing and deformed legs (S4C Fig). HAen79ΔSS2 enΔ110/inv-en deletion flies survive with wing defects similar to those seen in HAen79ΔSS2 enΔ110 homozygotes and have no leg defects. These data suggest that ΔS takes out more regulatory sequences than does ΔSS2. Thus, although the HAen79 transgene can rescue inv-en double mutants, it is a not equivalent to a wildtype en locus. Deleting either the O or S fragment from the HAen79 transgene leads to leg defects when these transgenes are the only source of En (S4 Fig). This provides further evidence that these fragments are also enhancers in leg discs. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Table 2. Viability and phenotype of flies with a single copy of the en gene. https://doi.org/10.1371/journal.pgen.1010826.t002

En expression from invΔ33 is not sensitive to the loss of the O or SS2 enhancers invΔ33 was created as a mimic of the HAen79 transgene at the endogenous locus (called en80 in [27], Fig 2A). At invΔ33, in addition to the 79kb present in HAen79, there is 1kb of DNA just downstream of the E(Pc) transcription stop site. E(Pc) and tou transcription form the boundaries of the inv-en domain [31]. We left 1kb downstream of E(Pc) because we did not want to risk interfering with E(Pc) transcription termination. We used CRISPR/Cas9 to delete either fragment O or SS2 from invΔ33 and saw no difference in the En expression pattern in wing discs (Fig 5). We next tested whether invΔ33, invΔ33ΔO or invΔ33ΔSS2 were sufficient as single copies by crossing them to three inv-en deletion mutants (Table 2). All three lines survive well over all the deletion mutants; none have wing vein or leg defects. Some flies hold their wings out, a phenotype associated with loss of inv [20]. We wondered whether adding a HA-tag to En would impair its function. We tagged En with HA on both a wildtype chromosome and on invΔ33ΔSS2, making HAinvΔ33ΔSS2. We found no evidence that the HA-tag compromised En function, as HAinvΔ33ΔSS2 flies survive well as heterozygotes and do not have wing vein defects (Table 2). In summary, these data show that the endogenous invΔ33 domain is resilient to the loss of a single imaginal disc enhancer. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 5. Deletions of disc enhancers from invΔ33 reveals the stability of the endogenous locus. En in invΔ33, invΔ33ΔO and invΔ33ΔSS2 wing imaginal discs looks like WT. En in invΔ33ΔOΔSS2 homozygous and invΔ33ΔOΔSS2/enΔ110 wing discs is variegated. At least 10 discs were examined for each genotype and a representative disc is shown. https://doi.org/10.1371/journal.pgen.1010826.g005 We next deleted both fragments O and SS2 and found that En expression is variegated both in invΔ33ΔOΔSS2 homozygotes and invΔ33ΔOΔSS2/enΔ110 wing discs (Fig 5). This variegated expression is consistent with the hypothesis that these enhancers are regulated by chromatin modifications. invΔ33ΔOΔSS2 survive well as heterozygotes (Table 2). Consistent with the variegated expression patterns, some wings have vein defects, whereas others do not. We hypothesized that the endogenous locus was more stable to enhancer deletions than the HA-en transgene because it has boundaries that stabilize the chromatin state of the locus. For example, H3K27me3, the Polycomb chromatin mark, stops at the 3’ ends of the E(Pc) and tou genes at the endogenous locus [31]. On the other hand, the transgene does not have boundaries, and the H3K27me3 spreads from the en DNA in both directions many kilobases, stopping at actively transcribed genes [27]. We hypothesize that this destabilizes the transgene in both the ON and OFF transcriptional states, making it less stable and more sensitive to the loss of enhancers.

[END]
---
[1] Url: https://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1010826

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.