Functional interaction between the Drosophila knirps short range transcriptional repressor and RPD3 histone deacetylase

Deciphering enhancer sequence using thermodynamics-based models and convolutional neural networks

Deciphering the sequence-function relationship encoded in enhancers holds the key to interpreting non-coding variants and understanding mechanisms of transcriptomic variation. Several quantitative models exist for predicting enhancer function and underlying mechanisms; however, there has been no systematic comparison of these models characterizing their relative strengths and shortcomings. Here, we interrogated a rich data set of neuroectodermal enhancers in Drosophila, representing cis- and trans- sources of expression variation, with a suite of biophysical and machine learning models. We performed rigorous comparisons of thermodynamics-based models implementing different mechanisms of activation, repression and cooperativity. Moreover, we developed a convolutional neural network (CNN) model, called CoNSEPT, that learns enhancer ‘grammar’ in an unbiased manner. CoNSEPT is the first general-purpose CNN tool for predicting enhancer function in varying conditions, such as different cell types and experimental conditions, and we show that such complex models can suggest interpretable mechanisms. We found model-based evidence for mechanisms previously established for the studied system, including cooperative activation and short-range repression. The data also favored one hypothesized activation mechanism over another and suggested an intriguing role for a direct, distance-independent repression mechanism. Our modeling shows that while fundamentally different models can yield similar fits to data, they vary in their utility for mechanistic inference. CoNSEPT is freely available at: https://github.com/PayamDiba/CoNSEPT.

A Looping-Based Model for Quenching Repression

We model the regulatory role of proteins bound to looped DNA using a simulation in which dsDNA is represented as a self-avoiding chain, and proteins as spherical protrusions. We simulate long self-avoiding chains using a sequential importance sampling Monte-Carlo algorithm, and compute the probabilities for chain looping with and without a protrusion. We find that a protrusion near one of the chain’s termini reduces the probability of looping, even for chains much longer than the protrusion–chain-terminus distance. This effect increases with protrusion size, and decreases with protrusion-terminus distance. The reduced probability of looping can be explained via an eclipse-like model, which provides a novel inhibitory mechanism. We test the eclipse model on two possible transcription-factor occupancy states of the D. melanogastereve 3/7 enhancer, and show that it provides a possible explanation for the experimentally-observed eve stripe 3 and 7 expression patterns.

Biological regulation-at-a-distance, whereby a transcription factor (TF) is able to generate susbstantial regulatory effects on gene expression even though it may be bound a large distance away from its target (500 bp–1 Mbp), is only partially understood. Using a biophysical model and a computer simulation that take dsDNA and TF volumes into account, we identify a downregulatory mechanism which functions at large distances, whereby a TF bound within ∼ 150 bp from an activator decreases the probability of looping-based interaction between the activator and the distant core promoter. This “eclipse” mechanism provides insight into the question of how enhancer architecture dictates gene expression.

Brinker possesses multiple mechanisms for repression because its primary co-repressor, Groucho, may be unavailable in some cell types

Transcriptional repressors function primarily by recruiting co-repressors, which are accessory proteins that antagonize transcription by modifying chromatin structure. Although a repressor could function by recruiting just a single co-repressor, many can recruit more than one, with Drosophila Brinker (Brk) recruiting the co-repressors CtBP and Groucho (Gro), in addition to possessing a third repression domain, 3R. Previous studies indicated that Gro is sufficient for Brk to repress targets in the wing, questioning why it should need to recruit CtBP, a short-range co-repressor, when Gro is known to be able to function over longer distances. To resolve this we have used genomic engineering to generate a series of brk mutants that are unable to recruit Gro, CtBP and/or have 3R deleted. These reveal that although the recruitment of Gro is necessary and can be sufficient for Brk to make an almost morphologically wild-type fly, it is insufficient during oogenesis, where Brk must utilize CtBP and 3R to pattern the egg shell appropriately. Gro insufficiency during oogenesis can be explained by its downregulation in Brk-expressing cells through phosphorylation downstream of EGFR signaling.

Context-dependent transcriptional interpretation of mitogen activated protein kinase signaling in the Drosophila embryo

Terminal regions of the Drosophila embryo are patterned by the localized activation of Mitogen Activated Protein Kinase (MAPK), which induces zygotic genes through relief of their repression by transcriptional repressor Capicua. The levels of MAPK activation at the anterior and posterior termini are close to each other, but the expression patterns of MAPK-target genes, such as zerknüllt (zen) and tailless (tll), display strong anterior-posterior (AP) asymmetry. This region-specific response to MAPK activation provides a clear example of context-dependent interpretation of inductive signaling, a common developmental effect that remains poorly understood. In the past, the AP asymmetry of zen expression was attributed to a mechanism that depends on MAPK substrate competition. We present data suggesting that the asymmetric expression of tll is generated by a different mechanism, based on feedforward control and multiple enhancers of the tll gene. A simple mathematical model of this mechanism correctly predicts how the wild-type expression pattern of tll changes in mutants affecting the anterior, dorsoventral, and terminal patterning systems and some of their direct targets.

Long- and short-range transcriptional repressors induce distinct chromatin states on repressed genes

Transcriptional repression is essential for establishing precise patterns of gene expression during development. Repressors governing early Drosophila segmentation can be classified as short- or long-range factors based on their ranges of action, acting either locally to quench adjacent activators or broadly to silence an entire locus. Paradoxically, these repressors recruit common corepressors, Groucho and CtBP, despite their different ranges of repression. To reveal the mechanisms underlying these two distinct modes of repression, we performed chromatin analysis using the prototypical long-range repressor Hairy and the short-range repressor Knirps. Chromatin immunoprecipitation and micrococcal nuclease mapping studies reveal that Knirps causes local changes of histone density and acetylation, and the inhibition of activator recruitment, without affecting the recruitment of basal transcriptional machinery. In contrast, Hairy induces widespread histone deacetylation and inhibits the recruitment of basal machinery without inducing chromatin compaction. Our study provides detailed mechanistic insight into short- and long-range repression on selected endogenous target genes and suggests that the transcriptional corepressors can be differentially deployed to mediate chromatin changes in a context-dependent manner.

Groucho corepressor functions as a cofactor for the Knirps short-range transcriptional repressor

Despite the pervasive roles for repressors in transcriptional control, the range of action of these proteins on cis regulatory elements remains poorly understood. Knirps has essential roles in patterning the Drosophila embryo by means of short-range repression, an activity that is essential for proper regulation of complex transcriptional control elements. Short-range repressors function in a local fashion to interfere with the activity of activators or basal promoters within approximately 100 bp. In contrast, long-range repressors such as Hairy act over distances >1 kb. The functional distinction between these two classes of repressors has been suggested to stem from the differential recruitment of the CtBP corepressor to short-range repressors and Groucho to long-range repressors. Contrary to this differential recruitment model, we report that Groucho is a functional part of the Knirps short-range repression complex. The corepressor interaction is mediated via an eh-1 like motif present in the N terminus and a conserved region present in the central portion of Knirps. We also show that this interaction is important for the CtBP-independent repression activity of Knirps and is required for regulation of even-skipped. Our study uncovers a previously uncharacterized interaction between proteins previously thought to function in distinct repression pathways, and indicates that the Groucho corepressor can be differentially harnessed to execute short- and long-range repression.

The H3K4 demethylase lid associates with and inhibits histone deacetylase Rpd3

JmjC domain-containing proteins have been shown to possess histone demethylase activity. One of these proteins is the Drosophila histone H3 lysine 4 demethylase Little imaginal discs (Lid), which has been genetically classified as a Trithorax group protein. However, contrary to the supposed function of Lid in gene activation, the biochemical activity of this protein entails the removal of a histone mark that is correlated with active transcription. To understand the molecular mechanism behind the function of Lid, we have purified a Lid-containing protein complex from Drosophila embryo nuclear extracts. In addition to Lid, the complex contains Rpd3, CG3815/Drosophila Pf1, CG13367, and Mrg15. Rpd3 is a histone deacetylase, and along with Polycomb group proteins, which antagonize the function of Trithorax group proteins, it negatively regulates transcription. By reconstituting the Lid complex, we demonstrated that the demethylase activity of Lid is not affected by its association with other proteins. However, the deacetylase activity of Rpd3 is greatly diminished upon incorporation into the Lid complex. Thus, our finding that Lid antagonizes Rpd3 function provides an explanation for the genetic classification of Lid as a positive transcription regulator.

Drosophila Ebi mediates Snail-dependent transcriptional repression through HDAC3-induced histone deacetylation

The Drosophila Snail protein is a transcriptional repressor that is necessary for mesoderm formation. Here, we identify the Ebi protein as an essential Snail co-repressor. In ebi mutant embryos, Snail target genes are derepressed in the presumptive mesoderm. Ebi and Snail interact both genetically and physically. We identify a Snail domain that is sufficient for Ebi binding, and which functions independently of another Snail co-repressor, Drosophila CtBP. This Ebi interaction domain is conserved among all insect Snail-related proteins, is a potent repression domain and is required for Snail function in transgenic embryos. In mammalian cells, the Ebi homologue TBL1 is part of the NCoR/SMRT-HDAC3 (histone deacetylase 3) co-repressor complex. We found that Ebi interacts with Drosophila HDAC3, and that HDAC3 knockdown or addition of a HDAC inhibitor impairs Snail-mediated repression in cells. In the early embryo, Ebi is recruited to a Snail target gene in a Snail-dependent manner, which coincides with histone hypoacetylation. Our results demonstrate that Snail requires the combined activities of Ebi and CtBP, and indicate that histone deacetylation is a repression mechanism in early Drosophila development.

Role of NAD binding and catalytic residues in the C-terminal binding protein corepressor

CtBP corepressor proteins potentiate the activity of many metazoan transcriptional repressors. These proteins are homologous to prokaryotic D-2-hydroxyacid dehydrogenases, possessing an NAD/NADH binding fold and conserved active site residues. When expressed in Drosophila, a catalytic site mutant retains biological activity, however, we find that an NAD binding mutant lacks biological activity. The NAD mutant, similar to a dimerization mutant, is expressed at low levels, indicating that binding of NAD/NADH may affect CtBP stability. These data support the idea that the ancestral dehydrogenase activity is not required for CtBP function, and NAD binding may play a regulatory, rather than catalytic, role.

Enhancer responses to similarly distributed antagonistic gradients in development

Formation of spatial gene expression patterns in development depends on transcriptional responses mediated by gene control regions, enhancers. Here, we explore possible responses of enhancers to overlapping gradients of antagonistic transcriptional regulators in the Drosophila embryo. Using quantitative models based on enhancer structure, we demonstrate how a pair of antagonistic transcription factor gradients with similar or even identical spatial distributions can lead to the formation of distinct gene expression domains along the embryo axes. The described mechanisms are sufficient to explain the formation of the anterior and the posterior knirps expression, the posterior hunchback expression domain, and the lateral stripes of rhomboid expression and of other ventral neurogenic ectodermal genes. The considered principles of interaction between antagonistic gradients at the enhancer level can also be applied to diverse developmental processes, such as domain specification in imaginal discs, or even eyespot pattern formation in the butterfly wing.

The early development of the fruit fly embryo depends on an intricate but well-studied gene regulatory network. In fly eggs, maternally deposited gene products—morphogenes—form spatial concentration gradients. The graded distribution of the maternal morphogenes initiates a cascade of gene interactions leading to embryo development. Gradients of activators and repressors regulating common target genes may produce different outcomes depending on molecular mechanisms, mediating their function. Here, we describe quantitative mathematical models for the interplay between gradients of positive and negative transcriptional regulators—proteins, activating or repressing their target genes through binding the gene's regulatory DNA sequences. We predict possible spatial outcomes of the transcriptional antagonistic interactions in fly development and consider examples where the predicted cases may take place.

Transcriptional regulation by C-terminal binding proteins

C-terminal binding protein family members function predominantly as transcriptional corepressors in association with sequence specific DNA-binding transcriptional repressors. The vertebrates have two CtBP genes while the invertebrates contain a single gene. Genetic studies indicate that the CtBP genes play pivotal roles in animal development. The vertebrate C-terminal binding proteins (CtBP1 and CtBP2) are highly related and are functionally redundant for certain developmental processes and non-redundant for others. The animal C-terminal binding proteins exhibit structural and functional similarity to d-isomer-specific 2-hydroxy acid dehydrogenases (D2-HDH). They function as dimers, recruiting transcriptional regulators through two protein-binding interfaces in each monomer. The corepressor complex of CtBP1 contains enzymatic constituents that mediate coordinated histone modification by deacetylation and methylation of histone H3-Lysine 9 and demethylation of histone H3-Lysine 4. CtBP also recruits the small ubiquitin-related modifier (SUMO) conjugating E2 enzyme UBC9 and a SUMO E3 ligase (HPC2), suggesting that CtBP-mediated transcriptional regulation may also involve SUMOylation of transcription factors. In addition to gene-specific transcriptional repression, CtBP1 appears to antagonize the activity of the global transcriptional coactivators, p300/CBP. Genetic evidence also suggests that the fly CtBP (dCtBP) and the vertebrate CtBP2 might activate transcription in a context-dependent manner. The transcriptional regulatory activity of CtBP is modulated by the nuclear NADH/NAD+ ratio and hence appears to be influenced by the metabolic status of the cell. The nuclear dinucleotide ratio may differentially influence the repression activities of factors that recruit CtBP through PLDLS-like motifs and those through non-PLDLS-motifs.

Developmental expression and phylogenetic conservation of alternatively spliced forms of the C-terminal binding protein corepressor

The C-terminal binding protein (CtBP) is an evolutionarily conserved transcriptional corepressor found in multicellular eukaryotes. Multiple forms of the protein are typically found in animal cells, produced from separate genes and by alternative splicing. CtBP isoforms have also been implicated in cytoplasmic functions, including Golgi fission and vesicular trafficking. All forms of CtBP contain a conserved core domain that is homologous to alpha-hydroxyacid dehydrogenases, and a subset of isoforms (CtBP(L)) contain extensions at the C terminus. Despite distinct developmental profiles and knockout phenotypes in the mouse, the properties of different isoforms of the protein are found to be similar in many transcriptional assays. We have investigated the expression and conservation of distinct isoforms of the CtBP protein in insects and found that the expression of multiple, developmentally regulated isoforms is widely conserved. In a variety of Drosophila species, the relative abundance of CtBP(L) to CtBP(S) drops sharply after embryogenesis, revealing a conserved developmental shift. Despite the overall lower levels of this isoform, bioinformatic analysis reveals that exons encoding the C-terminal extension in CtBP(L) are conserved from Diptera to Coleoptera, suggesting that the CtBP(L) isoform contributes an important, evolutionarily conserved function.