EVOLUTIONARY CO-OPTION OF AN ANCESTRAL CLOACAL REGULATORY LANDSCAPE DURING THE EMERGENCE OF DIGITS AND GENITALS.

The transition from fins to limbs has been a rich source of discussion for more than a century. One open and important issue is understanding how the mechanisms that pattern digits arose during vertebrate evolution. In this context, the analysis of Hox gene expression and functions to infer evolutionary scenarios has been a productive approach to explain the changes in organ formation, particularly in limbs. In tetrapods, the transcription of Hoxd genes in developing digits depends on a well-characterized set of enhancers forming a large regulatory landscape1,2. This control system has a syntenic counterpart in zebrafish, even though they lack bona fide digits, suggestive of deep homology3 between distal fin and limb developmental mechanisms. We tested the global function of this landscape to assess ancestry and source of limb and fin variation. In contrast to results in mice, we show here that the deletion of the homologous control region in zebrafish has a limited effect on the transcription of hoxd genes during fin development. However, it fully abrogates hoxd expression within the developing cloaca, an ancestral structure related to the mammalian urogenital sinus. We show that similar to the limb, Hoxd gene function in the urogenital sinus of the mouse also depends on enhancers located in this same genomic domain. Thus, we conclude that the current regulation underlying Hoxd gene expression in distal limbs was co-opted in tetrapods from a preexisting cloacal program. The orthologous chromatin domain in fishes may illustrate a rudimentary or partial step in this evolutionary co-option.

Sequential and directional insulation by conserved CTCF sites underlies the Hox timer in stembryos.

During development, Hox genes are temporally activated according to their relative positions on their clusters, contributing to the proper identities of structures along the rostrocaudal axis. To understand the mechanism underlying this Hox timer, we used mouse embryonic stem cell-derived stembryos. Following Wnt signaling, the process involves transcriptional initiation at the anterior part of the cluster and a concomitant loading of cohesin complexes enriched on the transcribed DNA segments, that is, with an asymmetric distribution favoring the anterior part of the cluster. Chromatin extrusion then occurs with successively more posterior CTCF sites acting as transient insulators, thus generating a progressive time delay in the activation of more posterior-located genes due to long-range contacts with a flanking topologically associating domain. Mutant stembryos support this model and reveal that the presence of evolutionary conserved and regularly spaced intergenic CTCF sites controls the precision and the pace of this temporal mechanism.

Context-dependent enhancer function revealed by targeted inter-TAD relocation.

The expression of some genes depends on large, adjacent regions of the genome that contain multiple enhancers. These regulatory landscapes frequently align with Topologically Associating Domains (TADs), where they integrate the function of multiple similar enhancers to produce a global, TAD-specific regulation. We asked if an individual enhancer could overcome the influence of one of these landscapes, to drive gene transcription. To test this, we transferred an enhancer from its native location, into a nearby TAD with a related yet different functional specificity. We used the biphasic regulation of Hoxd genes during limb development as a paradigm. These genes are first activated in proximal limb cells by enhancers located in one TAD, which is then silenced when the neighboring TAD activates its enhancers in distal limb cells. We transferred a distal limb enhancer into the proximal limb TAD and found that its new context suppresses its normal distal specificity, even though it is bound by HOX13 transcription factors, which are responsible for the distal activity. This activity can be rescued only when a large portion of the surrounding environment is removed. These results indicate that, at least in some cases, the functioning of enhancer elements is subordinated to the host chromatin context, which can exert a dominant control over its activity.

Sequential in cis mutagenesis in vivo reveals various functions for CTCF sites at the mouse HoxD cluster.

Mammalian Hox gene clusters contain a range of CTCF binding sites. In addition to their importance in organizing a TAD border, which isolates the most posterior genes from the rest of the cluster, the positions and orientations of these sites suggest that CTCF may be instrumental in the selection of various subsets of contiguous genes, which are targets of distinct remote enhancers located in the flanking regulatory landscapes. We examined this possibility by producing an allelic series of cumulative in cis mutations in these sites, up to the abrogation of CTCF binding in the five sites located on one side of the TAD border. In the most impactful alleles, the global chromatin architecture of the locus was modified, yet not drastically, illustrating that CTCF sites located on one side of a strong TAD border are sufficient to organize at least part of this insulation. Spatial colinearity in the expression of these genes along the major body axis was nevertheless maintained, despite abnormal expression boundaries. In contrast, strong effects were scored in the selection of target genes responding to particular enhancers, leading to the misregulation of Hoxd genes in specific structures. Altogether, while most enhancer-promoter interactions can occur in the absence of this series of CTCF sites, the binding of CTCF in the Hox cluster is required to properly transform a rather unprecise process into a highly discriminative mechanism of interactions, which is translated into various patterns of transcription accompanied by the distinctive chromatin topology found at this locus. Our allelic series also allowed us to reveal the distinct functional contributions for CTCF sites within this Hox cluster, some acting as insulator elements, others being necessary to anchor or stabilize enhancer-promoter interactions, and some doing both, whereas they all together contribute to the formation of a TAD border. This variety of tasks may explain the amazing evolutionary conservation in the distribution of these sites among paralogous Hox clusters or between various vertebrates.

Mesomelic dysplasias associated with the HOXD locus are caused by regulatory reallocations.

Human families with chromosomal rearrangements at 2q31, where the human HOXD locus maps, display mesomelic dysplasia, a severe shortening and bending of the limb. In mice, the dominant Ulnaless inversion of the HoxD cluster produces a similar phenotype suggesting the same origin for these malformations in humans and mice. Here we engineer 1 Mb inversion including the HoxD gene cluster, which positioned Hoxd13 close to proximal limb enhancers. Using this model, we show that these enhancers contact and activate Hoxd13 in proximal cells, inducing the formation of mesomelic dysplasia. We show that a secondary Hoxd13 null mutation in-cis with the inversion completely rescues the alterations, demonstrating that ectopic HOXD13 is directly responsible for this bone anomaly. Single-cell expression analysis and evaluation of HOXD13 binding sites suggests that the phenotype arises primarily by acting through genes normally controlled by HOXD13 in distal limb cells. Altogether, these results provide a conceptual and mechanistic framework to understand and unify the molecular origins of human mesomelic dysplasia associated with 2q31.

Dbx2 regulation in limbs suggests interTAD sharing of enhancers.

BACKGROUND: During tetrapod limb development, the HOXA13 and HOXD13 transcription factors are critical for the emergence and organization of the autopod, the most distal aspect where digits will develop. Since previous work had suggested that the Dbx2 gene is a target of these factors, we set up to analyze in detail this potential regulatory interaction. RESULTS: We show that HOX13 proteins bind to mammalian-specific sequences at the vicinity of the Dbx2 locus that have enhancer activity in developing digits. However, the functional inactivation of the DBX2 protein did not elicit any particular phenotype related to Hox genes inactivation in digits, suggesting either redundant or compensatory mechanisms. We report that the neighboring Nell2 and Ano6 genes are also expressed in distal limb buds and are in part controlled by the same Dbx2 enhancers despite being localized into two different topologically associating domains (TADs) flanking the Dbx2 locus. CONCLUSIONS: We conclude that Hoxa13 and Hoxd genes cooperatively activate Dbx2 expression in developing digits through binding to mammalian specific regulatory sequences in the Dbx2 neighborhood. Furthermore, these enhancers can overcome TAD boundaries in either direction to co-regulate a set of genes located in distinct chromatin domains.

Chromatin topology and the timing of enhancer function at the HoxD locus.

The HoxD gene cluster is critical for proper limb formation in tetrapods. In the emerging limb buds, different subgroups of Hoxd genes respond first to a proximal regulatory signal, then to a distal signal that organizes digits. These two regulations are exclusive from one another and emanate from two distinct topologically associating domains (TADs) flanking HoxD, both containing a range of appropriate enhancer sequences. The telomeric TAD (T-DOM) contains several enhancers active in presumptive forearm cells and is divided into two sub-TADs separated by a CTCF-rich boundary, which defines two regulatory submodules. To understand the importance of this particular regulatory topology to control Hoxd gene transcription in time and space, we either deleted or inverted this sub-TAD boundary, eliminated the CTCF binding sites, or inverted the entire T-DOM to exchange the respective positions of the two sub-TADs. The effects of such perturbations on the transcriptional regulation of Hoxd genes illustrate the requirement of this regulatory topology for the precise timing of gene activation. However, the spatial distribution of transcripts was eventually resumed, showing that the presence of enhancer sequences, rather than either their exact topology or a particular chromatin architecture, is the key factor. We also show that the affinity of enhancers to find their natural target genes can overcome the presence of both a strong TAD border and an unfavorable orientation of CTCF sites.

A complex regulatory landscape involved in the development of mammalian external genitals.

Developmental genes are often controlled by large regulatory landscapes matching topologically associating domains (TADs). In various contexts, the associated chromatin backbone is modified by specific enhancer-enhancer and enhancer-promoter interactions. We used a TAD flanking the mouse HoxD cluster to study how these regulatory architectures are formed and deconstructed once their function achieved. We describe this TAD as a functional unit, with several regulatory sequences acting together to elicit a transcriptional response. With one exception, deletion of these sequences didn't modify the transcriptional outcome, a result at odds with a conventional view of enhancer function. The deletion and inversion of a CTCF site located near these regulatory sequences did not affect transcription of the target gene. Slight modifications were nevertheless observed, in agreement with the loop extrusion model. We discuss these unexpected results considering both conventional and alternative explanations relying on the accumulation of poorly specific factors within the TAD backbone.

The constrained architecture of mammalian Hox gene clusters.

In many animal species with a bilateral symmetry, Hox genes are clustered either at one or at several genomic loci. This organization has a functional relevance, as the transcriptional control applied to each gene depends upon its relative position within the gene cluster. It was previously noted that vertebrate Hox clusters display a much higher level of genomic organization than their invertebrate counterparts. The former are always more compact than the latter, they are generally devoid of repeats and of interspersed genes, and all genes are transcribed by the same DNA strand, suggesting that particular factors constrained these clusters toward a tighter structure during the evolution of the vertebrate lineage. Here, we investigate the importance of uniform transcriptional orientation by engineering several alleles within the HoxD cluster, such as to invert one or several transcription units, with or without a neighboring CTCF site. We observe that the association between the tight structure of mammalian Hox clusters and their regulation makes inversions likely detrimental to the proper implementation of this complex genetic system. We propose that the consolidation of Hox clusters in vertebrates, including transcriptional polarity, evolved in conjunction with the emergence of global gene regulation via the flanking regulatory landscapes, to optimize a coordinated response of selected subsets of target genes in cis.

Tail Bud Progenitor Activity Relies on a Network Comprising Gdf11, Lin28, and Hox13 Genes.

During the trunk-to-tail transition, axial progenitors relocate from the epiblast to the tail bud. Here, we show that this process entails a major regulatory switch, bringing tail bud progenitors under Gdf11 signaling control. Gdf11 mutant embryos have an increased number of such progenitors that favor neural differentiation routes, resulting in a dramatic expansion of the neural tube. Moreover, inhibition of Gdf11 signaling recovers the proliferation ability of these progenitors when cultured in vitro. Tail bud progenitor growth is independent of Oct4, relying instead on Lin28 activity. Gdf11 signaling eventually activates Hox genes of paralog group 13, which halt expansion of these progenitors, at least in part, by down-regulating Lin28 genes. Our results uncover a genetic network involving Gdf11, Lin28, and Hox13 genes controlling axial progenitor activity in the tail bud.

Noncoding copy-number variations are associated with congenital limb malformation.

PurposeCopy-number variants (CNVs) are generally interpreted by linking the effects of gene dosage with phenotypes. The clinical interpretation of noncoding CNVs remains challenging. We investigated the percentage of disease-associated CNVs in patients with congenital limb malformations that affect noncoding cis-regulatory sequences versus genes sensitive to gene dosage effects.MethodsWe applied high-resolution copy-number analysis to 340 unrelated individuals with isolated limb malformation. To investigate novel candidate CNVs, we re-engineered human CNVs in mice using clustered regularly interspaced short palindromic repeats (CRISPR)-based genome editing.ResultsOf the individuals studied, 10% harbored CNVs segregating with the phenotype in the affected families. We identified 31 CNVs previously associated with congenital limb malformations and four novel candidate CNVs. Most of the disease-associated CNVs (57%) affected the noncoding cis-regulatory genome, while only 43% included a known disease gene and were likely to result from gene dosage effects. In transgenic mice harboring four novel candidate CNVs, we observed altered gene expression in all cases, indicating that the CNVs had a regulatory effect either by changing the enhancer dosage or altering the topological associating domain architecture of the genome.ConclusionOur findings suggest that CNVs affecting noncoding regulatory elements are a major cause of congenital limb malformations.

Control of growth and gut maturation by HoxD genes and the associated lncRNA Haglr.

During embryonic development, Hox genes participate in the building of a functional digestive system in metazoans, and genetic conditions involving these genes lead to important, sometimes lethal, growth retardation. Recently, this phenotype was obtained after deletion of Haglr, the Hoxd antisense growth-associated long noncoding RNA (lncRNA) located between Hoxd1 and Hoxd3 In this study, we have analyzed the function of Hoxd genes in delayed growth trajectories by looking at several nested targeted deficiencies of the mouse HoxD cluster. Mutant pups were severely stunted during the suckling period, but many recovered after weaning. After comparing seven distinct HoxD alleles, including CRISPR/Cas9 deletions involving Haglr, we identified Hoxd3 as the critical component for the gut to maintain milk-digestive competence. This essential function could be abrogated by the dominant-negative effect of HOXD10 as shown by a genetic rescue approach, thus further illustrating the importance of posterior prevalence in Hox gene function. A role for the lncRNA Haglr in the control of postnatal growth could not be corroborated.

Hotair Is Dispensible for Mouse Development.

Despite the crucial importance of Hox genes functions during animal development, the mechanisms that control their transcription in time and space are not yet fully understood. In this context, it was proposed that Hotair, a lncRNA transcribed from within the HoxC cluster regulates Hoxd gene expression in trans, through the targeting of Polycomb and consecutive transcriptional repression. This activity was recently supported by the skeletal phenotype of mice lacking Hotair function. However, other loss of function alleles at this locus did not elicit the same effects. Here, we re-analyze the molecular and phenotypic consequences of deleting the Hotair locus in vivo. In contrast with previous findings, we show that deleting Hotair has no detectable effect on Hoxd genes expression in vivo. In addition, we were unable to observe any significant morphological alteration in mice lacking the Hotair transcript. However, we find a subtle impact of deleting the Hotair locus upon the expression of the neighboring Hoxc11 and Hoxc12 genes in cis. Our results do not support any substantial role for Hotair during mammalian development in vivo. Instead, they argue in favor of a DNA-dependent effect of the Hotair deletion upon the transcriptional landscape in cis.

A role for HOX13 proteins in the regulatory switch between TADs at the HoxD locus.

During vertebrate limb development, Hoxd genes are regulated following a bimodal strategy involving two topologically associating domains (TADs) located on either side of the gene cluster. These regulatory landscapes alternatively control different subsets of Hoxd targets, first into the arm and subsequently into the digits. We studied the transition between these two global regulations, a switch that correlates with the positioning of the wrist, which articulates these two main limb segments. We show that the HOX13 proteins themselves help switch off the telomeric TAD, likely through a global repressive mechanism. At the same time, they directly interact with distal enhancers to sustain the activity of the centromeric TAD, thus explaining both the sequential and exclusive operating processes of these two regulatory domains. We propose a model in which the activation of Hox13 gene expression in distal limb cells both interrupts the proximal Hox gene regulation and re-enforces the distal regulation. In the absence of HOX13 proteins, a proximal limb structure grows without any sign of wrist articulation, likely related to an ancestral fish-like condition.

Combined function of HoxA and HoxB clusters in neural crest cells.

The evolution of chordates was accompanied by critical anatomical innovations in craniofacial development, along with the emergence of neural crest cells. The potential of these cells to implement a craniofacial program in part depends upon the (non-)expression of Hox genes. For instance, the development of jaws requires the inhibition of Hox genes function in the first pharyngeal arch. In contrast, Hox gene products induce craniofacial structures in more caudal territories. To further investigate which Hox gene clusters are involved in this latter role, we generated HoxA;HoxB cluster double mutant animals in cranial neural crest cells. We observed the appearance of a supernumerary dentary-like bone with an endochondral ossification around a neo-Meckel's cartilage matrix and an attachment of neo-muscle demonstrating that HoxB genes enhance the phenotype induced by the deletion of the HoxA cluster alone. In addition, a cervical and hypertrophic thymus was associated with the supernumerary dentary-like bone, which may reflect its ancestral position near the filtrating system. Altogether these results show that the HoxA and HoxB clusters cooperated during evolution to lead to present craniofacial diversity.

A switch between topological domains underlies HoxD genes collinearity in mouse limbs.

Hox genes are major determinants of the animal body plan, where they organize structures along both the trunk and appendicular axes. During mouse limb development, Hoxd genes are transcribed in two waves: early on, when the arm and forearm are specified, and later, when digits form. The transition between early and late regulations involves a functional switch between two opposite topological domains. This switch is reflected by a subset of Hoxd genes mapping centrally into the cluster, which initially interact with the telomeric domain and subsequently swing toward the centromeric domain, where they establish new contacts. This transition between independent regulatory landscapes illustrates both the modularity of the limbs and the distinct evolutionary histories of its various pieces. It also allows the formation of an intermediate area of low HOX proteins content, which develops into the wrist, the transition between our arms and our hands. This regulatory strategy accounts for collinear Hox gene regulation in land vertebrate appendages.

A regulatory archipelago controls Hox genes transcription in digits.

The evolution of digits was an essential step in the success of tetrapods. Among the key players, Hoxd genes are coordinately regulated in developing digits, where they help organize growth and patterns. We identified the distal regulatory sites associated with these genes by probing the three-dimensional architecture of this regulatory unit in developing limbs. This approach, combined with in vivo deletions of distinct regulatory regions, revealed that the active part of the gene cluster contacts several enhancer-like sequences. These elements are dispersed throughout the nearby gene desert, and each contributes either quantitatively or qualitatively to Hox gene transcription in presumptive digits. We propose that this genetic system, which we call a "regulatory archipelago," provides an inherent flexibility that may partly underlie the diversity in number and morphology of digits across tetrapods, as well as their resilience to drastic variations.

A transcriptionally silent RXRalpha supports early embryonic morphogenesis and heart development.

Retinoic acid (RA) receptors (RARs) alpha, beta, and gamma heterodimerized with rexinoid receptors (RXRs) alpha, beta, and gamma mediate the RA signal. To analyze the contribution of the transcriptional activity of RXRalpha, the main RXR during embryogenesis, we have engineered a mouse line harboring a transcriptionally silent RXRalpha mutant that lacks the activation functions AF1 and AF2. All homozygous mutants (Rxra(afo)) display the ocular defects previously observed in compound Rar-null and Rxra/Rar-null mutants, thus demonstrating that a transcriptionally active RXRalpha is required during eye development. In contrast, the vast majority of Rxra(afo) fetuses do not display the Rxra-null mutant hypoplasia of the myocardium, thus demonstrating that RXRalpha can act as a transcriptionally silent heterodimerization partner. Similarly, a transcriptionally silent RXRalpha mutant can support early embryogenesis, as Rxra(afo)/Rxrb-null embryos display a normal morphology, contrasting with the severe malformations exhibited by compound Rxra/Rxrb-null embryos. Along the same line, we show that a silent RXRalpha mutant is sufficient to allow the initial formation of the placental labyrinth, whereas later steps of trophoblast cell differentiation critically requires the AF2, but not the AF1, function of RXRalpha.

Genetic and pharmacological evidence that a retinoic acid cannot be the RXR-activating ligand in mouse epidermis keratinocytes.

Using genetic and pharmacological approaches, we demonstrate that both RARgamma/RXRalpha heterodimers involved in repression events, as well as PPARbeta(delta)/RXRalpha heterodimers involved in activation events, are cell-autonomously required in suprabasal keratinocytes for the generation of lamellar granules (LG), the organelles instrumental to the formation of the skin permeability barrier. In activating PPARbeta(delta)/RXRalpha heterodimers, RXRalpha is transcriptionally active as its AF-2 activation function is required and can be inhibited by an RXR-selective antagonist. Within repressing RARgamma/RXRalpha heterodimers, induction of the transcriptional activity of RXRalpha is subordinated to the addition of an agonistic ligand for RARgamma. Thus, the ligand that possibly binds and activates RXRalpha heterodimerized with PPARbeta(delta) cannot be a retinoic acid, as it would also bind RARgamma and relieve the RARgamma-mediated repression, thereby yielding abnormal LGs. Our data also demonstrate for the first time that subordination of RXR transcriptional activity to that of its RAR partner plays a crucial role in vivo, because it allows RXRs to act concomitantly, within the same cell, as heterodimerization partners for repression, as well as for activation events in which they are transcriptionally active.

Ligand-dependent contribution of RXRbeta to cholesterol homeostasis in Sertoli cells.

We show that mice expressing retinoid X receptor beta (RXRbeta) impaired in its transcriptional activation function AF-2 (Rxrb(af20) mutation) do not display the spermatid release defects observed in RXRbeta-null mutants, indicating that the role of RXRbeta in spermatid release is ligand-independent. In contrast, like RXRbeta-null mutants, Rxrb(af20) mice accumulate cholesteryl esters in Sertoli cells (SCs) due to reduced ABCA1 transporter-mediated cholesterol efflux. We provide genetic and molecular evidence that cholesterol homeostasis in SCs does not require PPARalpha and beta, but depends upon the TIF2 coactivator and RXRbeta/LXRbeta heterodimers, in which RXRbeta AF-2 is transcriptionally active. Our results also indicate that RXRbeta may be activated by a ligand distinct from 9-cis retinoic acid.