Hox gene regulation and timing in embryogenesis.

G Model

ARTICLE IN PRESS

YSCDB-1614; No. of Pages 9

Seminars in Cell & Developmental Biology xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Seminars in Cell & Developmental Biology journal homepage: www.elsevier.com/locate/semcdb

Review

Hox gene regulation and timing in embryogenesis Thomas Montavon a,∗ , Natalia Soshnikova b,∗∗ a b

Department of Epigenetics, Max-Planck Institute of Immunobiology and Epigenetics, Stübeweg 51, 79108 Freiburg, Germany Institute of Molecular Biology gGmbH (IMB), Ackermannweg 4, 55128 Mainz, Germany

a r t i c l e

i n f o

Article history: Available online xxx Keywords: Gene clusters Collinearity Polycomb Enhancers Long-range interactions

a b s t r a c t Hox genes are critical regulators of embryonic development in bilaterian animals. They exhibit a unique mode of transcriptional regulation where the position of the genes along the chromosome corresponds to the time and place of their expression during development. The sequential temporal activation of these genes in the primitive streak helps determining their subsequent pattern of expression along the anterior–posterior axis of the embryo, yet the precise correspondence between these two collinear processes is not fully understood. In addition, vertebrate Hox genes evolved similar modes of regulation along secondary body axes, such as the developing limbs. We review the current understanding of the mechanisms operating during activation, maintenance and silencing of Hox gene expression in these various contexts, and discuss the evolutionary significance of their genomic organization. © 2014 Elsevier Ltd. All rights reserved.

Contents 1. 2. 3. 4. 5. 6. 7. 8. 9.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Genomic organization and collinear expression of Hox genes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cis and trans control of collinearity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coupling axial elongation and patterning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chromatin structure and Hox gene regulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Three-dimensional organization of Hox clusters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Long-range control in secondary axes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Morphological evolution and regulatory divergence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions and perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction Development of a multicellular organism from a single cell is an intricate process. It relies on myriads of transcription factors, which must be present in the right cell at the right time. Hox genes, which encode homeo-domain containing transcription factors, control the patterning of bilaterian embryos along their anterior–posterior axis, from the hindbrain down to the caudal end [1]. In both fruit flies and mice, deletion of a single Hox gene leads to altered axial

∗ Corresponding author. ∗∗ Corresponding author. Tel.: +49 6131 39 21530. E-mail addresses: [email protected] (T. Montavon), [email protected] (N. Soshnikova).

00 00 00 00 00 00 00 00 00 00 00

identities and transformation of specific embryonic structures into more anterior ones (e.g. Refs. [2–4]). Conversely, ectopic expression of a single Hox gene can also result in a posterior transformation or loss of the body structures [5–8]. In addition to this ancestral role, Hox genes also evolved novel function in the course of evolution. For instance, in tetrapods, Hox genes are essential for the outgrowth and patterning of limbs along both the anterior–posterior and proximal–distal axes [9]. Given the critical role of the Hox genes in animal development, the pattern and time of their expression must be tightly controlled. The regulation of Hox genes is achieved mostly at the transcriptional level, but translational control has also been documented [10]. The mechanisms of transcriptional regulation are dictated by the organization of the genes in clusters [11]. This type of genomic organization allows for the sharing of nuclear space, chromatin

http://dx.doi.org/10.1016/j.semcdb.2014.06.005 1084-9521/© 2014 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Montavon T, Soshnikova N. Hox gene regulation and timing in embryogenesis. Semin Cell Dev Biol (2014), http://dx.doi.org/10.1016/j.semcdb.2014.06.005

G Model

ARTICLE IN PRESS


T. Montavon, N. Soshnikova / Seminars in Cell & Developmental Biology xxx (2014) xxx–xxx

2

late posterior

A

early anterior

5’ 3’ 13 12 11 10 9 8 7 6 5 4 3 2 1 A B C D

K27me3

C activation

B

10

6

1

time

K27me3

K27me3

K27me3

K4me3

K27me3

K4me3

K4me3

K4me3

K4me3

K4me3

K4me3

K4me3

Ant. K27me3

K4me3

Post. K27me3

K4me3

Fig. 1. Hox gene clusters and collinearity. (A) In mammals, four Hox genes clusters (HoxA–D) share similar organization. Hox genes are classified in 13 paralogy groups. Genes located on the 3 side of each cluster are activated early in development, and in anterior tissues, while 5 located genes are expressed progressively later and in more posterior regions (temporal and spatial collinearity). (B) Chromatin dynamics and temporal collinearity. A transition in histone modifications over the cluster accompanies transcriptional activation during axis extension. The whole cluster is initially labeled with the repressive mark H3K27me3 (grey). This mark is progressively erased and replaced by H3K4me3 (green) in parallel with gene activation. The schemes on the left illustrate the progressive activation of Hox genes along with the extension of the anterior–posterior axis, with the expression of representative genes from different paralogy groups: Hox1 (blue), Hox6 (red) and Hox10 (green). (C) Higher-order chromatin organization and spatial collinearity. Hox clusters adopt compact structures in regions of the embryo where they are silent, such as the forebrain (top). Along the anterior–posterior axis, active and silent genes segregate in separate compartments (middle), labeled by H3K4me3 (active, green) or H3K27me3 (silent, grey). In more posterior areas, the majority of Hox genes are found in the active compartment. Panels B and C are inspired by Refs. [22] and [55], respectively.

structure, common regulatory elements, such as enhancers, and even promoters. As a result, the time and place of Hox genes expression are largely determined by the relative position of each gene within its cluster [11]. This review addresses our current knowledge of the mechanisms controlling Hox genes expression during embryogenesis, focusing largely on the mammalian Hox clusters and including possible insights that can be gained from other species. For a more comprehensive review of Hox regulation in Drosophila, we recommend several recent articles [12,13]. 2. Genomic organization and collinear expression of Hox genes In many animal species Hox genes are clustered, and the conservation of gene order between distant species indicate that this organization is ancestral [11]. Invertebrates and chordates possess a single Hox cluster containing 8–15 genes [14]. In contrast, vertebrates have four Hox clusters (HoxA, B, C and D) as a result of two rounds of whole genome duplications (Fig. 1A) [11,15]. The ancestor of teleost fishes experienced another duplication of genome, which produced additional Hox gene clusters [14]. Based on sequence homology and chromosomal location Hox genes are assigned to 13 paralogy groups, where Hox1 lies at the 3 and Hox13 is at the 5 end of the cluster. This exceptional genomic organization has important functional consequences, since the transcriptional activity of a given Hox gene depends on its position within the cluster, a phenomenon referred to as collinearity. Three distinct modes of collinearity have been

reported: (1) spatial collinearity is a correspondence between the position of each Hox gene within the cluster and its anterior boundary of expression (Fig. 1) [16,17]. Accordingly, the genes located at the 3 end of the cluster are transcribed in more anterior regions of the embryo compare to the genes situated at the 5 end, which are expressed in more posterior areas. Spatial collinearity was observed in all bilaterians; even in species where genomic clustering was completely lost, as in the larvacean Oikopleura, Hox genes are expressed in nested anterior–posterior territories reminiscent of the patterns observed for their orthologous counterparts in species that kept the clustered organization [18]. The presence of cis-regulatory elements in close proximity to the Hox genes (see below) might explain why disintegration of the clusters in two or more pieces would be compatible with correct spatial expression patterns of Hox genes during development [11]. (2) In vertebrates, Hox genes become activated in the posterior part of the primitive streak in a time order that reflects their location within the clusters. Genes at the 3 end of the clusters are activated first, whereas genes located at more 5 positions are activated subsequently, a process referred to as temporal collinearity (Fig. 1A) [19]. This process takes place in parallel with the progressive extension of the anterior–posterior body axis in vertebrates. A proper timing of Hox genes activation is necessary for the correct specification and growth along the anterior–posterior axis as precocious activation of the 5 located genes results in premature truncation of the embryos or loss of axial structures [7,20,21]. Genetic studies in mice demonstrated that Hox cluster integrity is essential for temporal collinearity [7,20,22], and accordingly, this


G Model

ARTICLE IN PRESS



mode of collinear regulation is not observed in species harboring a split Hox cluster, or dispersed Hox genes. How does the sequential activation of the Hox genes relate to their spatial distribution along the anterior–posterior axis? While the timing of activation usually impacts on the definitive expression territories, this is not always true, and the two collinear processes can be uncoupled [20]. Careful expression studies, analyses of transgenic and knock-out mice indicate that the early activation of Hox genes in the primitive streak and their subsequent expression in the axial structures is not one continuous event [20,23,24]. The establishment of definitive Hox gene expression territories involves an expansion of these early activation domains that does not simply reflect cell proliferation, and that is influenced by various signaling pathways [23,25] (see below). This is important to keep in mind, since most changes in Hox genes expression territories are documented at stages where their patterns are already well established, and how this relates to alterations in the timing of activation is often speculative. (3) Vertebrate Hox clusters were also recruited for the patterning of additional embryonic structures. These new domains of expression typically display some form of collinearity, yet this can be quite different from the situation in the main body axis. In tetrapods, coordinated expression of the 5 located Hoxd genes is essential for the development of digits [9]. Expression of the most 5 gene, Hoxd13, is the strongest in the distal limb mesenchyme, with a progressive decline in expression levels of Hoxd12 to Hoxd9. This phenomenon was termed quantitative collinearity [26], and reflects a preferential interaction of long-range enhancers with the 5 extremity of the HoxD cluster [27–29]. Such long-range regulatory mechanisms will be discussed in a separate section.

3. Cis and trans control of collinearity These highly coordinated expression profiles, and the general conservation of clustering among bilaterians, suggested some form of global transcriptional control for Hox clusters [30]. Surprisingly however, single Hox genes can often recapitulate their genuine anterior–posterior domain of expression when randomly integrated as transgenes, and as previously mentioned, clustering is not strictly required to maintain spatial collinearity. Indeed, multiple cis-acting elements directing spatially restricted expression patterns in the mesoderm or in neural tissues are located within the clusters themselves [1,31]. In contrast, neither transgenes nor a split HoxD cluster can recapitulate the proper sequence of activation [20,22,32]. Hox clusters, rather than individual genes, appear to be transcriptionally repressed at these early stages, yet the relevant regulatory sequences remain to be identified [20,33,34]. These distinct regulations could indicate that temporal, rather than spatial collinearity represents the ancestral constraint that kept Hox genes clustered, but that it was subsequently lost in some species. Alternatively, both processes might have originally relied on similar global controls, and more local enhancer elements could have evolved later, perhaps resulting in an increased robustness of the system. The search for regulators of Hox transcription in trans has revealed at least three candidates: retinoic acid (RA), fibroblast growth factors (FGF) and Wnts [1,31]. RA signaling is mediated by nuclear receptors, retinoic acid receptors (RARs) and retinoid X receptors (RXRs), which form heterodimers and associate with specific DNA sequences termed Retinoic Acid Response Elements (RARE) [35]. Multiple RAREs have been identified within cis-acting elements located in the vicinity of Hox1–Hox5 genes [36–39]. Accordingly, treatment with RA leads to the activation of the Hox1–Hox5 genes in motor neurons cultures [37] and to the ectopic, more anterior expression of these genes in mouse embryos [39].

3

The more “central” genes Hoxb6–Hoxb8 are also activated by RA, yet interestingly, these genes are sensitive to exogenous RA in progressively later time windows, when compared to anterior genes [40]. The expression of Hox6–Hox9 was also associated with Wnt or FGF signaling and Cdx transcription factors [31]. Compound Cdx1−/− ;Cdx2+/− mice display homeotic transformations in the axial skeleton accompanied by changes in Hox4–Hox9 expression domains [41]. In the chick neural tube, FGF treatment leads to anteriorization of Hoxb6–Hoxb9 expression domains, an effect mediated at least in part via an upregulation of Cdx genes [42]. Accordingly, it has recently been shown that the treatment of motor neurons cultures with Wnt and FGF leads to the activation of Cdx2 expression and to the recruitment of this transcription factor to the cis-regulatory elements proximal to the active Hox1–Hox9 genes [37]. Furthermore, both Cdx2 binding and FGF signals are required for the activation of Hox gene expression in motor neurons.

4. Coupling axial elongation and patterning Temporal collinearity was observed in vertebrates, but also in short-germ band insects, two phyla where the main body axis forms progressively by addition of new posterior tissues. It was therefore proposed that the sequential expression of Hox genes impose progressively more posterior identities to the newly appearing structures, along with axis extension [43]. In vertebrates, the initial activation of Hox genes takes place in cells that do not directly contribute to the main embryonic axis [23], but these early expression domains expand toward the node region, where new axial structures are being formed [44]. This results in the sequential addition, to the growing axis, of tissues expressing different combinations of Hox gene products, in a process involving both cell migration and cell-to-cell signaling [44,45]. The initiation and establishment of Hox gene expression patterns is thus closely linked to the extension of the anterior–posterior body axis. Accordingly, signaling pathways such as Wnt, Fgf and RA, as well as Cdx transcription factors, both control axial elongation or presomitic mesoderm (PSM) segmentation [41,46–48] and impact on Hox gene expression (see above). In addition, mutations in genes controlling the period of segmentation (or segmentation clock), such as transducers of the Notch pathway, lead to altered Hox expression in the PSM and to axial patterning defects [49,50]. Hence, the acquisition of Hox gene expression territories is coupled to the progressive formation of the main body axis, which might ensure a proper series of axial identities along the embryo. Reciprocally, Hox genes seem to influence the elongation of the axis itself. Central Hox genes indeed were shown to promote axis extension by their ability to rescue the caudal truncations observed in Cdx mutants. In contrast, ectopic expression of Hox13 genes leads to premature termination of axial growth [21]. The upstream regulators identified so far are able to activate rather large subsets of Hox genes, and are active in broad time windows along the anterior–posterior axis. How can these “generic” factors trigger the graded expression of Hox clusters, in both space and time? Exposing embryos to exogenous signals has shown that anterior or posterior Hox genes are responsive to these pathways at distinct stages of development [40,42]. These and other experiments indicated that Hox genes become progressively permissive for activation, in a time sequence reflecting their location within the clusters [44]. While additional trans-acting factors are likely involved in the acquisition of this permissiveness, the progressive “opening” of Hox clusters also relies on transitions in their chromatin structure.


G Model YSCDB-1614; No. of Pages 9 4

ARTICLE IN PRESS T. Montavon, N. Soshnikova / Seminars in Cell & Developmental Biology xxx (2014) xxx–xxx

5. Chromatin structure and Hox gene regulation Chromatin plays an essential role in the transcriptional regulation of Hox genes. In early mouse embryos, prior to gastrulation, Hox genes are transcriptionally inactive [31]. Two large multiprotein complexes encoded by Polycomb group (PcG) genes are responsible for this silencing [51]. Polycomb Repressive Complex 1 and 2 (PRC1 and PRC2) contain a set of histone modifying enzymes, such as histone methyltransferases or histone demethylases, proteins binding methylated histone tails, as well as scaffold proteins essential for the stability of the complexes [52]. A component of PRC2, Ezh2 (a homologue of Drosophila Enhancer of zeste) catalyzes tri-methylation of histone H3 at lysine 27 (H3K27me3), which is essential for silencing and long-term repression of target genes [52]. Components of both PRC2 and PRC1 as well as H3K27me3 positive nucleosomes were detected over transcriptionally silent Hox clusters in embryonic stem (ES) cells [53,54]. Disruption of PRC2 activity leads to loss of H3K27 tri-methylation and de-repression of multiple Hox genes in ES cells [53]. In tissues where Hox genes remain silent, such as the forebrain, H3K27me3 is maintained over the Hox clusters for many cell divisions [34,55]. The same mechanism for maintenance of a stable repression of Hox genes operates in Drosophila as ectopic Hox gene expression was observed in PcG mutants [56]. Recruitment of PRC1 components to H3K27me3 marks further strengthens Hox gene silencing [57,58]. Firstly, E3 ubiquitin ligases Ring1A/1B mono-ubiquitylate histone H2A at lysine 119, which inhibits RNA Pol II at early steps of elongation [59–61]. Secondly, Ring1B promotes compaction of chromatin structure over Hox clusters [62]. Deletions of genes encoding for PRC1 components, including Bmi1, Cbx2, Phc1 and Phc2, cause homeotic transformations of several vertebrae accompanied by shifts in Hox gene expression in mouse embryos [13], suggesting that multiple components of PRC1 are either redundant or recruited only to a fraction of PcG target gene promoters. During progressive activation of Hox genes in the primitive streak chromatin states over the clusters change dramatically [22]. Hox genes located at the 3 end of the cluster are transcriptionally active at 8.5 days of embryonic development (E8.5) (Fig. 1B). Accordingly, H3K4me3 marks, generally associated with active transcription, are present at their promoters [63,64]. H3K4me3 is also enriched to a lower extent over the promoters of transcriptionally silent 5 located Hox genes suggesting that they are pre-labeled for future transcription. Conversely, repressive H3K27me3 mark localizes over the silent 5 located Hox genes. As development proceeds, 5 located Hox genes are sequentially activated, which is accompanied by the simultaneous appearance of activating H3K4me3 and loss of repressive H3K27me3 marks over their promoters. Consequently, the complete cluster is transcriptionally active and covered by H3K4me3 at E9.5 (Fig. 1B). The clustered organization of Hox genes seems essential for the sequential removal of the repressive H3K27me3 mark. Disruption of the HoxD cluster in two pieces leads to precocious loss of H3K27me3 and earlier appearance of H3K4me3 at Hox genes promoters, together with their ectopic activation [22]. The exact mechanism of the PRC2 recruitment to Hox clusters in vertebrates is not known. Systematic testing of specific DNA elements or potential consensus sites for PRC2 recruitment in vertebrates was not successful so far [34,65]. It has been proposed that CG rich promoters would be the docking pads for PRC2 binding [66–68]. However, most genes with CG rich promoters are not regulated by PRC2 complexes in mammalian cells [53,54]. There is a correlation between the distribution of PRC2 and several proteins binding DNA with low specificity, such as Jarid2 and AEBP2 [69–71]. Both Jarid2 and AEBP2 transiently interact with PRC2 components

[69–71], suggesting that a combination of such low affinity interactions might mediate PRC2 recruitment to specific loci. In addition, PRC2 components interact with various RNA molecules [72–75]. One class of these molecules, short promoter associated ncRNAs (PARS) transcribed from CpG-rich promoters, were proposed to recruit PRC2 and silence co-transcriptionally the target genes in cis [72]. Long non-coding RNAs (lncRNAs) also participate in PRC2 recruitment, and are believed to act both in cis and in trans [76]. The long ncRNA Hotair, transcribed from the 5 end of the HoxC cluster, binds both Suz12 and the Lsd1 H3K4me2/me1 demethylase and silence multiple genes, including 5 located Hoxd genes, in cultured cells [73,77]. However, recent reports suggest that RNA binding by PRC2 is rather promiscuous than sequencespecific [78]. Furthermore, attempts at confirming the function of Hotair in vivo yielded conflicting results [79,80].

6. Three-dimensional organization of Hox clusters Changes in higher-order chromatin organization also accompany Hox gene activation. This was first observed by visualizing the localization of specific Hox loci within the cell nucleus using Fluorescent in situ hybridization (FISH). The Hoxb1 and Hoxb9 gene loci, located 90 kb apart on the HoxB cluster, appear to colocalize in ES cells nuclei, where both genes are silent. The distance between these two loci increase upon RA-mediated differentiation, together with the transcriptional activation of the cluster, suggesting a decompaction of chromatin [81]. Similar differences were later observed in vivo: the cluster appears less compact in embryonic tissues expressing subsets of Hoxb genes than in areas where these genes are silent [82]. A relocation of Hoxb gene loci outside of their chromosome territory was also observed in parallel to their transcriptional activation [81,82]. The main limitation of FISH, however, is that its current resolution hardly allows discriminating loci separated by short genomic distances. This is particularly problematic for Hox clusters, which display an exceptionally high gene density for mammalian genomes (typically 10 genes in about 120 kb). The Chromosome Conformation Capture (3C) technique and its variants (such as 4C, 5C and Hi-C) helped overcome these limitations [83]. Briefly, 3C is based on the cross-linking and proximity-based ligation of DNA fragments that are in close physical proximity within the nucleus. The quantification of these new ligation products yields an estimate of the frequency of chromatin contacts within a cell population. These approaches revealed distinct spatial conformations for Hox clusters in cultured cells [84–86] or in various embryonic tissues in vivo [29,55,87,88]. In embryos, the three-dimensional organization of the clusters is tightly correlated to Hox genes expression status. In embryonic regions where Hox genes are not expressed, such as the forebrain, widespread contacts are observed between gene loci within each cluster, suggesting that each Hox cluster adopts a rather compact conformation in the silent state [29,55]. In contrast, in tissues expressing various subsets of Hox genes, the clusters appear to form two distinct compartments segregating active from silent genes [29,55,88]. This peculiar three-dimensional organization parallels spatial collinearity along the main body axis. In anterior regions of the embryo, transcribed Hox genes contact mostly other active genes within the same cluster, while silent genes group in a distinct compartment. In more posterior territories, where a larger subset of Hox genes is expressed, the majority of the genes participate in the active compartment (Fig. 1C) [55]. This suggests a model where Hox loci progressively move from an inactive to an active spatial compartment along with their transcriptional activation, in a stepwise remodeling of the cluster’s conformation. A temporal reorganization of Hox clusters conformation was indeed recently documented


G Model

ARTICLE IN PRESS



in the presomitic mesoderm and tail bud, at least for posterior genes [89]. This process might help ensuring a proper sequence of transcriptional activation, and prevent precocious expression of posterior genes. At later stages, it might also participate in keeping a memory of active versus inactive Hox loci. The mechanisms controlling Hox clusters conformation remain unknown. Importantly, active and silent compartments correspond to segments of the clusters decorated by H3K4me3 or H3K27me3, respectively [29,55,88]. This bimodal conformation might thus reflect a tendency for chromatin domains harboring similar histone modifications to aggregate together. Accordingly, PcG-dependent compaction of H3K27me3-decorated chromatin was observed in vitro [90], while ES cells deficient for either Ring1b (a PRC1 component) or Eed (PRC2) display less compact Hox clusters [62]. PcG-dependent interaction between Hox loci was also observed in Drosophila, where the 10 Mb-distant Ant-C and BX-C complexes segregate, together with other PcG targets, in specialized nuclear structures termed Polycomb bodies [91]. H3K4me3 might similarly facilitate interactions between active gene loci, although such a mechanism is so far not documented.

7. Long-range control in secondary axes In addition to their ancestral function in patterning the main body axis, vertebrate Hox clusters have acquired novel expression territories during evolution, for instance along the limb proximal–distal axis, or in the developing external genitalia and gastro-intestinal tract [26,30,92]. These expression patterns are typically not shared by all clusters, suggesting that they evolved after the two rounds of genome duplication in the vertebrate lineage. In contrast to the main body axis (see above), these novel regulatory specificities cannot be recapitulated by local sequences, but rely on global, long-range regulations. These regulations typically involve multiple, partially redundant enhancer elements located at a distance, within gene-poor regions flanking the clusters, and sharing their activity over contiguous sets of Hox genes (Fig. 2A) [29,87,88,93]. The distant location of these elements relative to their targets might be related to the high density of genes within Hox clusters, and to the presence of numerous regulatory sequences involved in their regulation along the main body axis: evolving new sets of control elements within the clusters could potentially interfere with their critical ancestral function. Once established, these global regulations could have participated in the consolidation of the structure of Hox clusters, to maintain or reinforce the coordinated activation of their targets [11]. The control of Hoxd gene expression in growing limbs is the best understood of these global regulations. Transcriptional activation of Hoxd genes in the limbs follows two phases, corresponding to the patterning of distinct limb segments [94,95]. In the first phase, necessary for the specification of the arm and forearm (or leg and lower leg), a group of enhancers spread along a gene desert flanking the cluster on its telomeric side (Hoxd1 side) activates Hoxd1–Hoxd11 along with the growth of the limb buds. This phase displays both temporal and spatial collinearities, such that the genes located closest to the enhancers are transcribed earlier and in a broader territory within the bud [87]. Subsequently, a distinct set of enhancers, spanning the gene desert located on the (opposite) centromeric side of the cluster, activates posterior genes (Hoxd13–Hoxd10) in the distal extremity of the limb bud, which will give rise to the digits [29]. Strikingly, both the proximal and the distal regulatory landscapes match adjacent topological domains (or TADs, for topologically associating domains) partially overlapping the HoxD cluster (Fig. 2A). TADs consist of megabase-scale regions of preferential chromatin interactions, and seem relatively stable between

5

cell types, leading to the hypothesis that they represent an inherent structural property of genomes [96,97]. Indeed, Hoxd genes located on the centromeric or telomeric extremities of the cluster mostly form long-range interactions within their respective TAD (that is, the centromeric or telomeric gene deserts, respectively), regardless of their expression status. Transcriptional activation during the early or late phase involves changes in histone modifications (such as acetylation) within the gene deserts, together with a limited reconfiguration of the contacts, rather than the establishment of drastically different spatial conformations [29,87]. In contrast, genes occupying a more central position within the cluster, such as Hoxd9–Hoxd11 are located near the boundary between the centromeric (C-DOM) and telomeric (T-DOM) TADs. While initially implementing the telomeric/proximal regulation (Fig. 2B), these genes switch to a centromeric control in the subset of distal cells that initiate the late phase of expression. This switch is accompanied by a redirection of their long-range interactions from the T-DOM to the C-DOM (Fig. 2C) [87]. Subsequent extension of the developing limb leads to a physical separation of cells responding to the telomeric or centromeric regulations, leading to the non-overlapping expression domains that will give rise to the proximal and distal limb compartments (Fig. 2D). Therefore, a temporal transition in long-range interactions is translated into a spatial pattern of Hoxd gene expression along the limb axis, with the two limbs compartments expressing distinct sets of Hoxd genes. In other words, the apparent spatial collinearity in the limbs is, in fact, the product of two independent regulatory controls that need to be activated in a timely fashion. This model raises several questions: which trans-acting factors control both phases of activation, and the reallocation of central genes from one TAD to the other? What mechanism leads to the physical separation of the two domains? It has been hypothesized that the growth of the limb increases the distance between the two cell pools and signaling centers secreting signals necessary for the relevant enhancer’s activity, thus leading to the progressive extinction of the early regulation in distal cells, and conversely [87]. FGFs from the apical ectodermal ridge (AER) and Shh from the zone of polarizing activity (ZPA) indeed both impact on the late phase of expression [98,99]. Interestingly, multiple enhancers were also shown to regulate Hoxa genes in distal limbs, and some of these depend on Shh for their activation [93].

8. Morphological evolution and regulatory divergence The critical function of Hox genes in organizing the body plan in bilateria, together with their peculiar mode of regulation, make them a paradigmatic gene family in the field of evolutionary developmental biology [100]. Structural changes in the organization of Hox clusters and/or alterations in their transcriptional regulation have often been proposed as causative events underlying key morphological adaptations. For instance, shifts in the expression domains of Hox genes along the trunk of chicken or goose, when compared to mice, are correlated to different numbers of cervical vertebrae in these species [101,102]. Snakes represent an extreme body plan amongst vertebrates, with an elongated and de-regionalized anterior–posterior axis. While relying on similar mechanisms as those described for other amniotes, the segmentation of the PSM of snake embryos occurs at a much higher relative pace [103], which suggests potential heterochrony in Hox gene activation. Comparative expression studies suggested alterations in Hox genes expression patterns, as well as in the function of Hox proteins themselves [104,105]. In particular, Hoxa13 and Hoxd13 transcripts were not detected in the snake’s somitic mesoderm [104]. As group 13 proteins promote the termination of axial extension in the mouse embryo [21], this could be


G Model

ARTICLE IN PRESS



6

A

C-DOM

digits

T-DOM

proximal limb

HoxD

cen

tel

digestive tract

external genitalia?

100 kb

distal limb

B

Hoxd12-13

C

Hoxd1-11

T-DOM active

Hoxd12-13

time

C-DOM silent

proximal limb

C-DOM silent

Hoxd1-11

T-DOM active

Hoxd1-4 Hoxd12-13

Hoxd9-13

D

time

C-DOM active

T-DOM silent

C-DOM silent

Hoxd1-4 Hoxd12-13

Hoxd9-13

C-DOM active

T-DOM silent

C-DOM silent

Hoxd1-11

T-DOM active

Hoxd1-11

T-DOM active

Fig. 2. Long-range regulation of Hoxd cluster genes and dynamic topology in developing limbs. (A) Map of the Hoxd cluster and its genomic neighborhood. Two adjacent topological domains encompass parts of the gene cluster as well as the flanking gene deserts on each side (C-DOM and T-DOM, red pattern above the map). Multiple enhancers controlling Hoxd gene expression in developing digits are scattered within the centromeric (cen) desert (blue ovals). The telomeric (tel) domain contains enhancers activating Hoxd genes in proximal limbs (arms and forearms, green) and digestive tract (purple). Expression of Hoxd genes in external genitalia requires sequences centromeric from the cluster, yet the relevant enhancers are not yet mapped (orange arrow). (B–D) Switch in long-range interactions during limb development. Two phases of Hoxd genes transcription occur in limbs. (B) During early limb budding, Hoxd genes segregate into distinct spatial compartments (C-DOM and T-DOM). Transcriptional activation of Hoxd1–11 requires interaction with active enhancers (green ovals with arrows) within the T-DOM. At this stage, the C-DOM, including distal enhancers (grey ovals) as well as the Hoxd12–13 genes, is silent (grey). (C) Later on, most cells maintain this pattern of contact and gene expression (right). However, a group of distal cells (left) activate C-DOM enhancers (blue ovals) and start expressing Hoxd12–13. Hoxd9–11 relocate from the T-DOM to the C-DOM and are therefore also transcribed in this domain. In contrast, the T-DOM, including Hoxd1–8, is silenced. (D) These distinct topologies are maintained at later stages, and limb growth leads to physical separation of the two cell populations, giving rise to non-overlapping expression domains that correspond to the patterning of distal (digits) or proximal (arms and forearm) limb segments. Figure inspired by Ref. [87].

of particular relevance to the evolution of elongated morphologies in snakes. Such shifts in expression boundaries might be the consequence of an altered activity of signaling pathways during axis extension. Alternatively, it could also reflect changes in the structure of Hox clusters themselves. While mammalian Hox clusters are particularly compact and devoid of long repetitive elements, several reptile and amphibian species display enlarged clusters that harbor large numbers of transposable elements [104,106,107]. These elements are typically decorated by different sets of chromatin modifications, such as high levels of tri-methylated H3K9 and DNA methylation [108,109]. How such chromatin features impact on Hox gene regulation is not known, but it could interfere with the progressive removal of H3K27me3 during axial extension [22]. Of note, the mammalian HoxB cluster harbors a large stretch of

repeated elements between Hoxb9 and Hoxb13, and this segment does not participate in either the active or the silent interaction domains in the mouse embryo [55], further suggesting that invasion of transposable elements could affect transcriptional regulation within Hox clusters. Morphological transitions are also associated with the acquisition of novel long-range regulations. Early comparisons of gene expression patterns in mouse or zebrafish suggested that the distal phase of Hox gene expression in the limb had no counterpart in fishes, leading to the proposal that digits are a tetrapod-specific structure, and that the acquisition of a distal Hox gene expression domain accompanied this transition [110]. However, this conclusion became controversial when more fish species were analyzed [111,112]. 4C analyses indeed indicated that the bimodal pattern of interactions that characterize Hox gene activation in mouse limbs


G Model

ARTICLE IN PRESS



is also present in zebrafish fins [113]. Yet surprisingly, putative fish enhancers, or even entire regulatory landscapes, typically elicit proximal, rather than distal limb patterns when tested as transgenes in mice [113–115]. Therefore, while the general regulatory logics underlying Hox genes activation in limbs seems conserved amongst vertebrate, regulatory elements responsive to distal signals are only found around tetrapod clusters. 9. Conclusions and perspectives In the past few years, important progress was made toward a better understanding of Hox gene regulation. The influence of chromatin structure and higher order organization in this process opens new perspectives for future research. Of the various implementations of collinearity, the early events leading to the progressive activation of Hox clusters remain the most elusive. Further efforts will be required to characterize both cis- and transregulators, at this early stage. The role of the various signaling pathways impacting on Hox gene expression along the axis was indeed mostly documented during the expansion phase, and their influence on the initial activation of Hox gene transcription is not firmly documented. While genetics studies suggest both activating and repressive influences from the genomic areas surrounding the clusters, the relevant regulatory elements are not identified so far [116]. In this view, the long-range regulatory mechanisms that begin to be deciphered in the context of secondary axes might give new clues to help understanding the ancestral collinearity along the main body axis [87]. The involvement of a global chromatin architecture, rather than discrete cis-regulatory elements, and binding sites for a given transcription factor, in controlling the temporal activation of Hox genes in this early phase would be an attractive mechanism that could help explaining the evolutionary conservation of their clustered organization. The connection between the temporal sequential activation of Hox genes and their subsequent spatial distribution remains unclear. The chromatin profiles over the HoxD cluster from early mouse embryos [22] suggest that the cell pool becoming “open for transcription” (that is, free from repressive histone modifications) is considerably larger that the actual expression domain of the corresponding Hox genes, at this stage. This could help define a cell population where a given set of genes are primed for subsequent activation during the anterior expansion of Hox expression territories, along with axis elongation. Resolving these complex issues is not a trivial task, owing in part to the difficulty of capturing the dynamics of Hox gene expression at early stages. Live imaging of ex vivo cultured embryos expressing fluorescence-labeled Hox proteins might help in this respect [117]. Such an approach would also allow interfering chemically with candidate signaling pathways that might influence the time of Hox gene activation. Despite the advances in mapping the regulatory circuits controlling Hox genes activation, the actual process of collinearity, that is, the sequential, and directional response of the various genes within the cluster, remains largely elusive. Such a differential behavior is observed in every context of Hox gene expression, strongly suggesting that it relies on an intrinsic property of the clusters, irrespective of the actual signal triggering their activation. While early hypotheses of a progressive transition in chromatin structure were largely confirmed [22,30], we still do not know what controls the directionality and the kinetics or pace of this opening. In others words, how these specific cell populations, within the embryo, can sense time and translate it into a developmental pattern. Acknowledgements We thank Guillaume Andrey for discussions and sharing artwork, and an anonymous reviewer for valuable suggestions. We

7

apologize to colleagues whose work could not be cited here due to space constraints.

References [1] Alexander T, Nolte C, Krumlauf R. Hox genes and segmentation of the hindbrain and axial skeleton. Annu Rev Cell Dev Biol 2009;25:431–56. [2] Gendron-Maguire M, Mallo M, Zhang M, Gridley T. Hoxa-2 mutant mice exhibit homeotic transformation of skeletal elements derived from cranial neural crest. Cell 1993;75:1317–31. [3] Kaufman TC. Cytogenetic analysis of chromosome 3 in DROSOPHILA MELANOGASTER: isolation and characterization of four new alleles of the proboscipedia (pb) locus. Genetics 1978;90:579–96. [4] Rijli FM, Mark M, Lakkaraju S, Dierich A, Dolle P, Chambon P. A homeotic transformation is generated in the rostral branchial region of the head by disruption of Hoxa-2, which acts as a selector gene. Cell 1993;75: 1333–49. [5] Denell RE, Hummels KR, Wakimoto BT, Kaufman TC. Developmental studies of lethality associated with the antennapedia gene complex in Drosophila melanogaster. Dev Biol 1981;81:43–50. [6] Grammatopoulos GA, Bell E, Toole L, Lumsden A, Tucker AS. Homeotic transformation of branchial arch identity after Hoxa2 overexpression. Development 2000;127:5355–65. [7] Tschopp P, Duboule D. A regulatory ‘landscape effect’ over the HoxD cluster. Dev Biol 2011;351:288–96. [8] van de Ven C, Bialecka M, Neijts R, Young T, Rowland JE, Stringer EJ, et al. Concerted involvement of Cdx/Hox genes and Wnt signaling in morphogenesis of the caudal neural tube and cloacal derivatives from the posterior growth zone. Development 2011;138:3451–62. [9] Zakany J, Duboule D. The role of Hox genes during vertebrate limb development. Curr Opin Genet Dev 2007;17:359–66. [10] Kondrashov N, Pusic A, Stumpf CR, Shimizu K, Hsieh AC, Xue S, et al. Ribosome-mediated specificity in Hox mRNA translation and vertebrate tissue patterning. Cell 2011;145:383–97. [11] Duboule D. The rise and fall of Hox gene clusters. Development 2007;134:2549–60. [12] Maeda RK, Karch F. The bithorax complex of Drosophila an exceptional Hox cluster. Curr Top Dev Biol 2009;88:1–33. [13] Mallo M, Alonso CR. The regulation of Hox gene expression during animal development. Development 2013;140:3951–63. [14] Pascual-Anaya J, D’Aniello S, Kuratani S, Garcia-Fernandez J. Evolution of Hox gene clusters in deuterostomes. BMC Dev Biol 2013;13:26. [15] Ohno S. Gene duplication and the uniqueness of vertebrate genomes circa 1970–1999. Semin Cell Dev Biol 1999;10:517–22. [16] Gaunt SJ, Sharpe PT, Duboule D. Spatially restricted domains of homeo-gene transcripts in mouse embryos: relation to a segmented body plan. Development (Suppl) 1988;104:169–70. [17] Graham A, Papalopulu N, Krumlauf R. The murine and Drosophila homeobox gene complexes have common features of organization and expression. Cell 1989;57:367–78. [18] Seo HC, Edvardsen RB, Maeland AD, Bjordal M, Jensen MF, Hansen A, et al. Hox cluster disintegration with persistent anteroposterior order of expression in Oikopleura dioica. Nature 2004;431:67–71. [19] Izpisua Belmonte JC, Falkenstein H, Dollé P, Renucci A, Duboule D. Murine genes related to the Drosophila AbdB homeotic genes are sequentially expressed during development of the posterior part of the body. EMBO J 1991;10:2279–89. [20] Tschopp P, Tarchini B, Spitz F, Zakany J, Duboule D. Uncoupling time and space in the collinear regulation of Hox genes. PLoS Genet 2009;5:e1000398. [21] Young T, Rowland JE, van de Ven C, Bialecka M, Novoa A, Carapuco M, et al. Cdx and Hox genes differentially regulate posterior axial growth in mammalian embryos. Dev Cell 2009;17:516–26. [22] Soshnikova N, Duboule D. Epigenetic temporal control of mouse Hox genes in vivo. Science 2009;324:1320–3. [23] Deschamps J, Wijgerde M. Two phases in the establishment of HOX expression domains. Dev Biol 1993;156:473–80. [24] Gaunt SJ, Strachan L. Forward spreading in the establishment of a vertebrate Hox expression boundary: the expression domain separates into anterior and posterior zones, and the spread occurs across implanted glass barriers. Dev Dyn 1994;199:229–40. [25] Deschamps J, van Nes J. Developmental regulation of the Hox genes during axial morphogenesis in the mouse. Development 2005;132:2931–42. [26] Dolle P, Izpisua-Belmonte JC, Brown JM, Tickle C, Duboule D. HOX-4 genes and the morphogenesis of mammalian genitalia. Genes Dev 1991;5: 1767–77. [27] Kmita M, Fraudeau N, Herault Y, Duboule D. Serial deletions and duplications suggest a mechanism for the collinearity of Hoxd genes in limbs. Nature 2002;420:145–50. [28] Montavon T, Le Garrec JF, Kerszberg M, Duboule D. Modeling Hox gene regulation in digits: reverse collinearity and the molecular origin of thumbness. Genes Dev 2008;22:346–59. [29] Montavon T, Soshnikova N, Mascrez B, Joye E, Thevenet L, Splinter E, et al. A regulatory archipelago controls Hox genes transcription in digits. Cell 2011;147:1132–45.


G Model YSCDB-1614; No. of Pages 9 8


[30] Dollé P, Izpisua Belmonte JC, Falkenstein H, Renucci A, Duboule D. Coordinate expression of the murine Hox-5 complex homoeobox-containing genes during limb pattern formation. Nature 1989;342:767–72. [31] Young T, Deschamps J. Hox, Cdx, and anteroposterior patterning in the mouse embryo. Curr Top Dev Biol 2009;88:235–55. [32] Deschamps J, van den Akker E, Forlani S, De Graaff W, Oosterveen T, Roelen B, et al. Initiation, establishment and maintenance of Hox gene expression patterns in the mouse. Int J Dev Biol 1999;43:635–50. [33] van der Hoeven F, Zakany J, Duboule D. Gene transpositions in the HoxD complex reveal a hierarchy of regulatory controls. Cell 1996;85:1025–35. [34] Schorderet P, Lonfat N, Darbellay F, Tschopp P, Gitto S, Soshnikova N, et al. A genetic approach to the recruitment of PRC2 at the HoxD locus. PLoS Genet 2013;9:e1003951. [35] Rhinn M, Dolle P. Retinoic acid signalling during development. Development 2012;139:843–58. [36] Gavalas A, Studer M, Lumsden A, Rijli FM, Krumlauf R, Chambon P. Hoxa1 and Hoxb1 synergize in patterning the hindbrain, cranial nerves and second pharyngeal arch. Development 1998;125:1123–36. [37] Mazzoni EO, Mahony S, Peljto M, Patel T, Thornton SR, McCuine S, et al. Saltatory remodeling of Hox chromatin in response to rostrocaudal patterning signals. Nat Neurosci 2013;16:1191–8. [38] Nolte C, Jinks T, Wang X, Martinez Pastor MT, Krumlauf R. Shadow enhancers flanking the HoxB cluster direct dynamic Hox expression in early heart and endoderm development. Dev Biol 2013;383:158–73. [39] Packer AI, Crotty DA, Elwell VA, Wolgemuth DJ. Expression of the murine Hoxa4 gene requires both autoregulation and a conserved retinoic acid response element. Development 1998;125:1991–8. [40] Oosterveen T, Niederreither K, Dolle P, Chambon P, Meijlink F, Deschamps J. Retinoids regulate the anterior expression boundaries of 5 Hoxb genes in posterior hindbrain. EMBO J 2003;22:262–9. [41] van den Akker E, Forlani S, Chawengsaksophak K, de Graaff W, Beck F, Meyer BI, et al. Cdx1 and Cdx2 have overlapping functions in anteroposterior patterning and posterior axis elongation. Development 2002;129: 2181–93. [42] Bel-Vialar S, Itasaki N, Krumlauf R. Initiating Hox gene expression: in the early chick neural tube differential sensitivity to FGF and RA signaling subdivides the HoxB genes in two distinct groups. Development 2002;129: 5103–15. [43] Duboule D. Temporal colinearity and the phylotypic progression: a basis for the stability of a vertebrate Bauplan and the evolution of morphologies through heterochrony. Dev Suppl 1994:135–42. [44] Forlani S, Lawson KA, Deschamps J. Acquisition of Hox codes during gastrulation and axial elongation in the mouse embryo. Development 2003;130:3807–19. [45] Iimura T, Pourquie O. Collinear activation of Hoxb genes during gastrulation is linked to mesoderm cell ingression. Nature 2006;442:568–71. [46] Aulehla A, Wehrle C, Brand-Saberi B, Kemler R, Gossler A, Kanzler B, et al. Wnt3a plays a major role in the segmentation clock controlling somitogenesis. Dev Cell 2003;4:395–406. [47] Dubrulle J, McGrew MJ, Pourquie O. FGF signaling controls somite boundary position and regulates segmentation clock control of spatiotemporal Hox gene activation. Cell 2001;106:219–32. [48] Moreno TA, Kintner C. Regulation of segmental patterning by retinoic acid signaling during Xenopus somitogenesis. Dev Cell 2004;6:205–18. [49] Zakany J, Kmita M, Alarcon P, de la Pompa JL, Duboule D. Localized and transient transcription of Hox genes suggests a link between patterning and the segmentation clock. Cell 2001;106:207–17. [50] Cordes R, Schuster-Gossler K, Serth K, Gossler A. Specification of vertebral identity is coupled to Notch signalling and the segmentation clock. Development 2004;131:1221–33. [51] Schuettengruber B, Cavalli G. Recruitment of polycomb group complexes and their role in the dynamic regulation of cell fate choice. Development 2009;136:3531–42. [52] Aloia L, Di Stefano B, Di Croce L. Polycomb complexes in stem cells and embryonic development. Development 2013;140:2525–34. [53] Boyer LA, Plath K, Zeitlinger J, Brambrink T, Medeiros LA, Lee TI, et al. Polycomb complexes repress developmental regulators in murine embryonic stem cells. Nature 2006;441:349–53. [54] Lee TI, Jenner RG, Boyer LA, Guenther MG, Levine SS, Kumar RM, et al. Control of developmental regulators by Polycomb in human embryonic stem cells. Cell 2006;125:301–13. [55] Noordermeer D, Leleu M, Splinter E, Rougemont J, De Laat W, Duboule D. The dynamic architecture of Hox gene clusters. Science 2011;334:222–5. [56] Simon J, Chiang A, Bender W. Ten different Polycomb group genes are required for spatial control of the abdA and AbdB homeotic products. Development 1992;114:493–505. [57] Cao R, Wang L, Wang H, Xia L, Erdjument-Bromage H, Tempst P, et al. Role of histone H3 lysine 27 methylation in Polycomb-group silencing. Science 2002;298:1039–43. [58] Endoh M, Endo TA, Endoh T, Isono K, Sharif J, Ohara O, et al. Histone H2A mono-ubiquitination is a crucial step to mediate PRC1-dependent repression of developmental genes to maintain ES cell identity. PLoS Genet 2012;8:e1002774. [59] de Napoles M, Mermoud JE, Wakao R, Tang YA, Endoh M, Appanah R, et al. Polycomb group proteins Ring1A/B link ubiquitylation of histone H2A to heritable gene silencing and X inactivation. Dev Cell 2004;7:663–76.

[60] Stock JK, Giadrossi S, Casanova M, Brookes E, Vidal M, Koseki H, et al. Ring1-mediated ubiquitination of H2A restrains poised RNA polymerase II at bivalent genes in mouse ES cells. Nat Cell Biol 2007;9:1428–35. [61] Wang H, Wang L, Erdjument-Bromage H, Vidal M, Tempst P, Jones RS, et al. Role of histone H2A ubiquitination in Polycomb silencing. Nature 2004;431:873–8. [62] Eskeland R, Leeb M, Grimes GR, Kress C, Boyle S, Sproul D, et al. Ring1B compacts chromatin structure and represses gene expression independent of histone ubiquitination. Mol Cell 2010;38:452–64. [63] Bernstein BE, Humphrey EL, Erlich RL, Schneider R, Bouman P, Liu JS, et al. Methylation of histone H3 Lys 4 in coding regions of active genes. Proc Natl Acad Sci U S A 2002;99:8695–700. [64] Santos-Rosa H, Schneider R, Bannister AJ, Sherriff J, Bernstein BE, Emre NC, et al. Active genes are tri-methylated at K4 of histone H3. Nature 2002;419:407–11. [65] Woo CJ, Kharchenko PV, Daheron L, Park PJ, Kingston RE. A region of the human HOXD cluster that confers polycomb-group responsiveness. Cell 2010;140:99–110. [66] Ku M, Koche RP, Rheinbay E, Mendenhall EM, Endoh M, Mikkelsen TS, et al. Genomewide analysis of PRC1 and PRC2 occupancy identifies two classes of bivalent domains. PLoS Genet 2008;4:e1000242. [67] Lynch MD, Smith AJ, De Gobbi M, Flenley M, Hughes JR, Vernimmen D, et al. An interspecies analysis reveals a key role for unmethylated CpG dinucleotides in vertebrate Polycomb complex recruitment. EMBO J 2012;31: 317–29. [68] Mendenhall EM, Koche RP, Truong T, Zhou VW, Issac B, Chi AS, et al. GCrich sequence elements recruit PRC2 in mammalian ES cells. PLoS Genet 2010;6:e1001244. [69] Kim H, Kang K, Kim J. AEBP2 as a potential targeting protein for Polycomb Repression Complex PRC2. Nucleic Acids Res 2009;37:2940–50. [70] Peng JC, Valouev A, Swigut T, Zhang J, Zhao Y, Sidow A, et al. Jarid2/Jumonji coordinates control of PRC2 enzymatic activity and target gene occupancy in pluripotent cells. Cell 2009;139:1290–302. [71] Shen X, Kim W, Fujiwara Y, Simon MD, Liu Y, Mysliwiec MR, et al. Jumonji modulates polycomb activity and self-renewal versus differentiation of stem cells. Cell 2009;139:1303–14. [72] Kanhere A, Viiri K, Araujo CC, Rasaiyaah J, Bouwman RD, Whyte WA, et al. Short RNAs are transcribed from repressed polycomb target genes and interact with polycomb repressive complex-2. Mol Cell 2010;38:675–88. [73] Rinn JL, Kertesz M, Wang JK, Squazzo SL, Xu X, Brugmann SA, et al. Functional demarcation of active and silent chromatin domains in human HOX loci by noncoding RNAs. Cell 2007;129:1311–23. [74] Zhao J, Sun BK, Erwin JA, Song JJ, Lee JT. Polycomb proteins targeted by a short repeat RNA to the mouse X chromosome. Science 2008;322:750–6. [75] Kaneko S, Bonasio R, Saldana-Meyer R, Yoshida T, Son J, Nishino K, et al. Interactions between JARID2 and noncoding RNAs regulate PRC2 recruitment to chromatin. Mol Cell 2014;53:290–300. [76] Rinn JL, Chang HY. Genome regulation by long noncoding RNAs. Annu Rev Biochem 2012;81:145–66. [77] Tsai MC, Manor O, Wan Y, Mosammaparast N, Wang JK, Lan F, et al. Long noncoding RNA as modular scaffold of histone modification complexes. Science 2010;329:689–93. [78] Davidovich C, Zheng L, Goodrich KJ, Cech TR. Promiscuous RNA binding by Polycomb repressive complex 2. Nat Struct Mol Biol 2013;20:1250–7. [79] Schorderet P, Duboule D. Structural and functional differences in the long non-coding RNA hotair in mouse and human. PLoS Genet 2011;7: e1002071. [80] Li L, Liu B, Wapinski OL, Tsai MC, Qu K, Zhang J, et al. Targeted disruption of Hotair leads to homeotic transformation and gene derepression. Cell Rep 2013;5:3–12. [81] Chambeyron S, Bickmore WA. Chromatin decondensation and nuclear reorganization of the HoxB locus upon induction of transcription. Genes Dev 2004;18:1119–30. [82] Chambeyron S, Da Silva NR, Lawson KA, Bickmore WA. Nuclear reorganisation of the Hoxb complex during mouse embryonic development. Development 2005;132:2215–23. [83] de Laat W, Dekker J. 3C-based technologies to study the shape of the genome. Methods 2012;58:189–91. [84] Ferraiuolo MA, Rousseau M, Miyamoto C, Shenker S, Wang XQ, Nadler M, et al. The three-dimensional architecture of Hox cluster silencing. Nucleic Acids Res 2010;38:7472–84. [85] Wang KC, Yang YW, Liu B, Sanyal A, Corces-Zimmerman R, Chen Y, et al. A long noncoding RNA maintains active chromatin to coordinate homeotic gene expression. Nature 2011;472:120–4. [86] Fraser J, Rousseau M, Shenker S, Ferraiuolo MA, Hayashizaki Y, Blanchette M, et al. Chromatin conformation signatures of cellular differentiation. Genome Biol 2009;10:R37. [87] Andrey G, Montavon T, Mascrez B, Gonzalez F, Noordermeer D, Leleu M, et al. A switch between topological domains underlies HoxD genes collinearity in mouse limbs. Science 2013;340:1234167. [88] Delpretti S, Montavon T, Leleu M, Joye E, Tzika A, Milinkovitch M, et al. Multiple enhancers regulate Hoxd genes and the Hotdog LncRNA during cecum budding. Cell Rep 2013;5:137–50. [89] Noordermeer D, Leleu M, Schorderet P, Joye E, Chabaud F, Duboule D. Temporal dynamics and developmental memory of 3D chromatin architecture at Hox gene loci. eLife 2014, http://dx.doi.org/10.7554/eLife.02557.


G Model YSCDB-1614; No. of Pages 9


[90] Grau DJ, Chapman BA, Garlick JD, Borowsky M, Francis NJ, Kingston RE. Compaction of chromatin by diverse Polycomb group proteins requires localized regions of high charge. Genes Dev 2011;25:2210–21. [91] Bantignies F, Roure V, Comet I, Leblanc B, Schuettengruber B, Bonnet J, et al. Polycomb-dependent regulatory contacts between distant Hox loci in Drosophila. Cell 2011;144:214–26. [92] Yokouchi Y, Sakiyama J, Kuroiwa A. Coordinated expression of Abd-B subfamily genes of the HoxA cluster in the developing digestive tract of chick embryo. Dev Biol 1995;169:76–89. [93] Berlivet S, Paquette D, Dumouchel A, Langlais D, Dostie J, Kmita M. Clustering of tissue-specific sub-TADs accompanies the regulation of HoxA genes in developing limbs. PLoS Genet 2013;9:e1004018. [94] Tarchini B, Duboule D. Control of Hoxd genes’ collinearity during early limb development. Dev Cell 2006;10:93–103. [95] Nelson CE, Morgan BA, Burke AC, Laufer E, DiMambro E, Murtaugh LC, et al. Analysis of Hox gene expression in the chick limb bud. Development 1996;122:1449–66. [96] Nora EP, Lajoie BR, Schulz EG, Giorgetti L, Okamoto I, Servant N, et al. Spatial partitioning of the regulatory landscape of the X-inactivation centre. Nature 2012;485:381–5. [97] Dixon JR, Selvaraj S, Yue F, Kim A, Li Y, Shen Y, et al. Topological domains in mammalian genomes identified by analysis of chromatin interactions. Nature 2012;485:376–80. [98] Haramis AG, Brown JM, Zeller R. The limb deformity mutation disrupts the SHH/FGF-4 feedback loop and regulation of 5 HoxD genes during limb pattern formation. Development 1995;121:4237–45. [99] Laufer E, Nelson CE, Johnson RL, Morgan BA, Tabin C. Sonic hedgehog and Fgf4 act through a signaling cascade and feedback loop to integrate growth and patterning of the developing limb bud. Cell 1994;79:993–1003. [100] Holland PW. Evolution of homeobox genes. Wiley Interdiscip Rev Dev Biol 2013;2:31–45. [101] Burke AC, Nelson CE, Morgan BA, Tabin C. Hox genes and the evolution of vertebrate axial morphology. Development 1995;121:333–46. [102] Gaunt SJ. Conservation in the Hox code during morphological evolution. Int J Dev Biol 1994;38:549–52. [103] Gomez C, Ozbudak EM, Wunderlich J, Baumann D, Lewis J, Pourquie O. Control of segment number in vertebrate embryos. Nature 2008;454:335–9. [104] Di-Poi N, Montoya-Burgos JI, Miller H, Pourquie O, Milinkovitch MC, Duboule D. Changes in Hox genes’ structure and function during the evolution of the squamate body plan. Nature 2010;464:99–103.

9

[105] Woltering JM, Vonk FJ, Muller H, Bardine N, Tuduce IL, de Bakker MA, et al. Axial patterning in snakes and caecilians: evidence for an alternative interpretation of the Hox code. Dev Biol 2009;332:82–9. [106] Voss SR, Putta S, Walker JA, Smith JJ, Maki N, Tsonis PA. Salamander Hox clusters contain repetitive DNA and expanded non-coding regions: a typical Hox structure for non-mammalian tetrapod vertebrates? Hum Genomics 2013;7:9. [107] Di-Poi N, Montoya-Burgos JI, Duboule D. Atypical relaxation of structural constraints in Hox gene clusters of the green anole lizard. Genome Res 2009;19:602–10. [108] Karimi MM, Goyal P, Maksakova IA, Bilenky M, Leung D, Tang JX, et al. DNA methylation and SETDB1/H3K9me3 regulate predominantly distinct sets of genes, retroelements, and chimeric transcripts in mESCs. Cell Stem Cell 2011;8:676–87. [109] Rebollo R, Karimi MM, Bilenky M, Gagnier L, Miceli-Royer K, Zhang Y, et al. Retrotransposon-induced heterochromatin spreading in the mouse revealed by insertional polymorphisms. PLoS Genet 2011;7:e1002301. [110] Sordino P, van der Hoeven F, Duboule D. Hox gene expression in teleost fins and the origin of vertebrate digits. Nature 1995;375: 678–81. [111] Woltering JM, Duboule D. The origin of digits: expression patterns versus regulatory mechanisms. Dev Cell 2010;18:526–32. [112] Schneider I, Shubin NH. The origin of the tetrapod limb: from expeditions to enhancers. Trends Genet 2013;29:419–26. [113] Woltering JM, Noordermeer D, Leleu M, Duboule D. Conservation and divergence of regulatory strategies at Hox Loci and the origin of tetrapod digits. PLoS Biol 2014;12:e1001773. [114] Schneider I, Aneas I, Gehrke AR, Dahn RD, Nobrega MA, Shubin NH. Appendage expression driven by the Hoxd Global Control Region is an ancient gnathostome feature. Proc Natl Acad Sci U S A 2011;108:12782–6. [115] Amemiya CT, Alfoldi J, Lee AP, Fan S, Philippe H, Maccallum I, et al. The African coelacanth genome provides insights into tetrapod evolution. Nature 2013;496:311–6. [116] Zakany J, Kmita M, Duboule D. A dual role for Hox genes in limb anterior–posterior asymmetry. Science 2004;304:1669–72. [117] Ahn Y, Mullan HE, Krumlauf R. Long-range regulation by shared retinoic acid response elements modulates dynamic expression of posterior Hoxb genes in CNS development. Dev Biol 2014;388:134–44.


SnapShot: Hox gene regulation.

GalR and Hox Proteins.

Chromatin boundary elements organize genomic architecture and developmental gene regulation in Drosophila Hox clusters.

Evidence for positive and negative regulation of the Hox-3.1 gene.

The Hox gene family in transgenic mice.

Hox gene dysregulation in acute myeloid leukemia.

Zygotic genome activation and imprinting: parent-of-origin gene regulation in plant embryogenesis.

The regulation of the murine Hox-2.5 gene expression during cell differentiation.

Onychophoran Hox genes and the evolution of arthropod Hox gene expression.

HOX gene activation by retinoic acid.

Gene Regulation by the AGL15 Transcription Factor Reveals Hormone Interactions in Somatic Embryogenesis.

Coordinate regulation of HOX genes in human hematopoietic cells.

Multiple sites of Hox-7 expression during mouse embryogenesis: comparison with retinoic acid receptor mRNA localization.

HOX gene complement and expression in the planarian Schmidtea mediterranea.

HOX gene expression in normal and neoplastic human kidney.

Comparison of mouse and human HOX-4 complexes defines conserved sequences involved in the regulation of Hox-4.4.

Post-transcriptional regulation of early embryogenesis.

Embryonic timing, axial stem cells, chromatin dynamics, and the Hox clock.

Hox-1.11 and Hox-4.9 homeobox genes.

Hox gene duplications correlate with posterior heteronomy in scorpions.

Timing is everything: GTPase regulation in phototransduction.

Genetic Regulation of Puberty Timing in Humans.

Murine Hox-1.11 homeobox gene structure and expression.

Characterisation of the murine Hox-3.3 gene and its promoter.