Animals with a backbone can look remarkably different from one another, like fish and birds, for example. Nevertheless, these animals – which are also known as vertebrates – have many genes in common that shape their bodies during development. These genes include a family called the Hox genes, which control how an animal’s body parts develop from its head to its tail and are needed to shape the animal’s limbs. Hox genes are found clustered in groups within a vertebrate’s DNA, and large regions of DNA on either side of a Hox cluster can, in some cases, physically interact with the Hox genes to regulate their expression.

So how do the same genes produce different body shapes? Different vertebrates regulate where and when their Hox genes are switched off and on in different ways. As such, it is likely that differences in gene regulation, rather than in the genes themselves, lead embryos to develop into the distinct shapes seen across the animal kingdom.

Snakes – for example – evolved from a lizard-like ancestor into elongated limbless animals as they have adapted to a burrowing lifestyle. However, it was not known if changes in how Hox genes are regulated have played a role in shaping the distinct body plan of snakes.

Guerreiro et al. have now compared how Hox genes are regulated in snakes, mice and other vertebrates, focusing on corn snakes and one particular cluster of Hox genes called the HoxD cluster. The comparison revealed that these Hox genes are regulated differently in developing snakes than in other vertebrate embryos. This is particularly the case for tissues that show the most differences when compared with other animals (such as the torso and genitals) or that are absent (such as the limbs). Although Hoxd genes are activated at different times and places in snakes than in other vertebrates, snake Hox genes appear to be regulated using the same general mechanisms as mouse Hox genes.

Guerreiro et al. suggest that changes to Hoxd gene regulation have contributed to the evolution of the snake’s shape and have most likely influenced the body shapes of other vertebrates as well. However, the findings also suggest that these gene regulatory changes have been constrained by an existing regulatory mechanism that has been maintained throughout evolution. It remains for future work to address whether these changes in Hox gene regulation are a cause or a consequence of the snake’s extreme body shape, or indeed a combination of the two.

Even though vertebrate species can display different morphologies, they all contain a strikingly similar repertoire of transcription factors and signalling molecules. In particular, genes with critical functions during embryonic development are often largely pleiotropic and highly conserved across species (for references, see Kirschner et al., 2005; Duboule and Wilkins, 1998). This universality of genetic and genomic principles has changed the evolutionary paradigm from the question of the nature of similarities to that of how distinct traits could evolve using such related developmental pathways (Carroll, 2008; De Robertis, 2008). Initially, Hox genes, as well as their structural and functional organization into genomic clusters were found well conserved across bilateria (Duboule and Dolle, 1989; Akam, 1989; Garcia-Fernandez and Holland, 1994; Graham et al., 1989; McGinnis et al., 1984). In addition, their mis-expression led to changes in the identity of both insect and vertebrate segments, illustrating their crucial role in the patterning of animal structures, even though the structures they specify are of very different nature in various taxa (e.g. (Maeda, 2006; Lewis, 1978; Krumlauf, 1994).

Tetrapods generally have four clusters of Hox genes (HoxA, HoxB, HoxC and HoxD), originating from genome duplications early in the vertebrate lineage (see e.g. Lemons and McGinnis, 2006) and located on different chromosomes, unlike fishes or some jawless vertebrates, which have more (Prince et al., 1998; Mehta et al., 2013; Amores et al., 1998). In addition, all vertebrate Hox clusters described thus far implement a particular type of regulatory process referred to as collinearity, whereby Hox genes are expressed sequentially in both time and space following their topological organization within each genomic cluster (Gaunt et al., 1988; Izpisua-Belmonte et al., 1991). This regulatory property is first observed during axial extension and, subsequently, in some structures such as the limbs (see (Deschamps, 2007; Deschamps and van Nes, 2005). In this latter case, and while the detailed underlying mechanism may be distinct from that at work in the major body axis (Kmita and Duboule, 2003), the general principle remains the same and was likely co-opted in the course of tetrapod evolution (Spitz et al., 2001), through the emergence of global enhancers located at remote positions on both sides of the cluster (Lonfat et al., 2014).

These complex regulations were extensively studied in the mouse, in particular at the HoxD locus, by using various targeted approaches in vivo. The HoxD cluster is surrounded by two gene deserts of approximately 1 Mb (megabase) in size, each one containing distinct sets of enhancers capable of activating specific sub-groups of target Hoxd genes depending on their location within the cluster. Each of these two gene deserts can be superimposed to a Topologically Associating Domain (TAD), i.e. a chromatin domain where DNA-DNA interactions in cis are privileged, for example between promoters and enhancers, and determined through chromosome conformation capture technologies (Dixon et al., 2012; Nora et al., 2012). The centromeric gene desert can activate the transcription of the Hoxd9 to Hoxd13 genes, whereas the telomeric gene desert, which is further subdivided into two sub-TADs (Andrey et al., 2013) controls the expression of Hoxd1 to Hoxd11 (see [Lonfat and Duboule, 2015]).

This bimodal regulation allows for the selected expression of Hoxd gene sub-groups in a series of secondary embryonic structures. During limb development, for instance, the telomeric TAD initially controls all genes from Hoxd3 to Hoxd11 in the proximal part of the limb bud, whereas more posterior genes such as Hoxd13 or Hoxd12 are controlled subsequently in the most distal aspect of the incipient limb by enhancers located within the centromeric TAD (Andrey et al., 2013). This latter regulatory landscape also controls transcription of the same posterior genes during the outgrowth of external genitalia (Lonfat et al., 2014). Since snakes are limbless animals and they display highly specialized and divergent external genitals (Tschopp et al., 2014), the existence of such a bimodal type of regulation at the snake HoxD locus was uncertain. Therefore, we set out to investigate Hox gene regulation in snakes. While these animals cannot yet be considered as genuine model systems (Guerreiro and Duboule, 2014; Milinkovitch and Tzika, 2007), recent advances in their genomic analyses make their study increasingly interesting in an Evo-Devo context (Castoe et al., 2013; Gilbert et al., 2014; Ullate-Agote et al., 2014; Vonk et al., 2013). These analyses revealed that snakes, regardless of their extreme morphologies, have a tetrapod-like complement of Hox genes with only a few exceptions (Vonk et al., 2013; Di-Poï et al., 2010). Consequently, the serpentiform body plan may have evolved either along with changes in time and space of Hox gene expression or with a different interpretation of Hox protein functions (see [Di-Poï et al., 2010; Woltering et al., 2009]).

The analysis of Hox gene expression in the developing corn snake (Pantherophis guttatus) revealed a surprisingly well conserved collinear mRNA distribution along the anterior-posterior axis. However, the rather strict correlation between morphological landmarks and the anterior borders of Hox transcript domains, usually seen in mammals and birds, was not always present in snakes (Burke et al., 1995; Woltering et al., 2009) (see also Head and Polly [2015]). It was thus concluded that some Hox proteins had likely changed (part of) their functionality. In addition, the fact that the most posterior Hox genes were poorly expressed in the extending tailbud was tentatively associated to the unusually large number of segments (Di-Poï et al., 2010), together with an increased pace in segmentation (Gomez et al., 2008).

In this work, we used a combination of experimental approaches to try and elucidate the nature of the differences in Hoxd gene regulation between snakes and mice at comparable stages of their early development. We find that, even though the structural organization of the corn snake HoxD cluster resembles that of tetrapods, the extreme body plan observed in snakes is associated with an extensive regulatory restructuring. In snakes, mesodermal enhancers are mostly located inside the cluster itself, whereas other vertebrates make use of long-range regulations located at remote positions. In addition, we show that despite the loss of limbs, the bimodal chromatin organisation at the Hoxd locus found in tetrapods is conserved in the snake lineage. However, we find that the regulation of snake Hoxd genes during the development of the external genitalia is different from that of other tetrapods, even though the general logic is conserved. In this latter case, the change in enhancer activity from a limb to an external genital specificity seems to have occurred. Altogether, we conclude that Hoxd gene regulation in the snake is in many ways distinct from the situation in mammals. We discuss the possible causative nature of these changes in the evolutionary transformation towards a serpentiform body plan.

To analyse the regulation of the snake HoxD cluster, we initially had to complement the available genomic information (Ullate-Agote et al., 2014) with high coverage sequencing of the gene cluster itself, along with the two flanking gene deserts. We screened a corn snake custom-made BAC library using as probes DNA sequences conserved from mammals to the anole lizard, present within this large DNA interval. A scaffold was built out of 13 overlapping BACs, which were selected for sequencing and from which a 1.3 Mb large DNA sequence of high quality was obtained (Figure 1—figure supplement 1A). The structural analysis of the corn snake HoxD cluster revealed that, as for other species of snakes whose genomes were recently released (Vonk et al., 2013; Castoe et al., 2013), all Hoxd genes but Hoxd12 are present and share the same transcriptional orientation within the cluster. When compared to the mouse, the corn snake cluster is about 1.5 fold larger (Figure 1A) (Vonk et al., 2013), likely due to a higher repeat content (Di-Poï et al., 2009) and consistent with the structures of the HoxD clusters of both the king cobra and the python (Castoe et al., 2013; Vonk et al., 2013).

Figure 1 with 2 supplements see all

Download asset Open asset

We plotted the sizes of the mammalian, reptile, bird and fish HoxD clusters against the size of their respective genome. When reptiles were excluded from the linear regression analysis, an R² value of 0.43 was scored, indicating significant correlation. However, when the corn snake, king cobra and Burmese python cluster sizes were added to the analysis, the R² value was reduced to 0.027 (Figure 1—figure supplement 1B). Even though snake Hox clusters show a size larger than what would be expected based on their genome size, the green anole lizard cluster is, from the vertebrate species analysed, the one with the lowest level of correlation with genome size (R²=0.0012). When we performed the same correlation analysis for the regulatory gene deserts that surround the HoxD cluster, high values of R² were scored both by excluding and including squamate values. However, the size of the squamate 3’ gene desert clearly showed a better correlation with genome size than the 5’ gene desert (Figure 1—figure supplement 1B).

Because the increased size of the cluster in Squamata correlated with the presence of a high number of transposable elements (Di-Poï et al., 2009, 2010), we investigated the number and type of repeats present in the HoxD locus. We found that the corn snake cluster contains more than twice as many repeats as the mouse counterpart. In addition, while the mouse HoxD cluster is mainly composed of SINEs (short interspersed elements), the corn snake cluster is composed of different types of transposable elements including LINEs (long interspersed elements) and DNA transposons (Figure 1—figure supplement 1C). The 5’ and 3’ gene deserts that surround the cluster contain a similar amount of repeats in the two species. Both gene deserts in mice include a wider range of repeat element types than in the cluster itself, yet SINEs remain the most represented transposable elements, whereas snake gene deserts mostly included LINEs (Figure 1—figure supplement 1C). Our deep DNA sequence of the entire corn snake HoxD locus, including both flanking gene deserts, allowed a global conservation analysis to be performed between non-coding regions amongst different vertebrate species. Surprisingly, we found that the pattern of conservation of the snake HoxD genomic landscape is almost identical to that of the chicken, when compared to the mouse sequence (Figure 1—figure supplement 2).

Regulatory potential of the HoxD cluster in vertebrates

In tetrapods, regulatory elements controlling Hox gene expression are found at various positions. The mouse HoxD cluster for instance contains regulatory elements, which are mainly involved in driving Hox gene expression along the anterior-posterior body axis during gastrulation, whereas remote enhancers located outside of the cluster itself regulate transcription in other organs or structures (Spitz et al., 2001; Lonfat and Duboule, 2015). Therefore, to try and identify snake-specific differences in the modes of regulations, we compared the regulatory potential of the snake HoxD cluster with that of other vertebrates by using a BAC transgenic approach in mice, whereby BACs containing HoxD clusters of either human, mouse, chicken, snake and zebrafish were randomly integrated in the mouse genome (Figure 2A). The expression of Hoxd4 was monitored by in situ hybridization with species-specific probes and, under these experimental conditions, all mammalian transgenic BACs showed transcript patterns restricted to the dorsal part of the main embryonic body axis (Figure 2B), resembling the pattern obtained when a single copy Hoxd4/LacZ transgene was used (Tschopp et al., 2012). Interestingly, however, this pattern represented only a subset of the full Hoxd4 expression pattern as seen either by WISH on control embryos (Figure 2B), or on previous reporter Hoxd4/lacZ transgenes likely integrated as tandem repeats (Zhang et al., 1997). Indeed, expression was scored mostly in the neural tube, yet not in the ventral mesodermal tissues of the upper trunk, i.e. above the level of hindlimbs. To better determine which mesodermal components had their Hox gene expression affected in the isolated human BAC line, we performed a Hoxd4 WISH in a sectioned embryo. We found that, at least at this stage of development, the human Hoxd4 gene was expressed only in the neural tube (Figure 2—figure supplement 1A).

Figure 2 with 3 supplements see all

Download asset Open asset

We then investigated the expression of Hoxd4 from either the chicken or the zebrafish BAC transgenic lines and found a similar expression pattern, again mostly limited to the neural tube as well as the dorsal-most part of the somites (Figure 2B). Altogether, these results suggested that enhancers controlling the robust expression of Hoxd genes in various mesodermal derivatives are, for the most part, located outside of the cluster itself. Alternatively, some mesodermal enhancers could be located inside the HoxD cluster, yet they may require additional sequences located at remote positions to properly impact upon the transcription of target genes in physiological conditions. Consequently, we searched for the location(s) of such enhancers outside the HoxD cluster by using a set of targeted deletions flanking the locus on either side of it.

Bimodal regulation in the snake HoxD regulatory locus

At the mouse HoxD locus, a bimodal regulatory strategy associated with particular chromatin conformations was reported to control Hox gene expression in a variety of organs and structures. Such global controls involve separate sets of target Hoxd genes, which are thus re-activated after the major body axis is laid down (Andrey et al., 2013; Spitz et al., 2001). Most of these structures, however, are either missing in snakes, such as the limbs or the intestinal cecum or whenever present, they are nevertheless substantially different from their mammalian counterparts. Because mouse Hoxd genes contact such remote enhancer sequences via long-range interactions included within two opposite TADs (Montavon et al., 2011), we set out to see whether such a bimodal type of regulatory topology would also exist in snakes, even in the absence of many of the related functionalities. We thus used whole mouse and snake embryos of similar size to characterize the interaction profile of Hoxd genes with their surrounding regulatory landscapes.

We used the 4C-seq version of chromosome conformation capture (Dekker et al., 2002; de Laat and Dekker, 2012) with four different Hoxd genes as viewpoints to assess their potential interaction tropism with either one of the flanking gene deserts. As observed in the mouse, significant interactions between the snake viewpoints and the centromeric TAD were observed, mostly when the Hoxd13 bait was used and, to a lower extent, with Hoxd11 (Figure 3A,B, compare tracks 1 and 2). In both cases however, substantial contacts were also observed with the opposite, telomeric TAD. These latter interactions increased when the snake Hoxd9 and Hoxd4 baits were used, whereas at the same time, interactions with the centromeric landscape almost disappeared (Figure 3A,B, tracks 3 and 4). Therefore, as previously reported in the case of mouse tissues and ES cells, genes located at different relative positions within the HoxD gene cluster show distinct interaction tropisms. 5’-located genes such as Hoxd13 interact primarily with the 5’-located gene desert, whereas genes located in a more 3’ part of the cluster, such as Hoxd4, interact mostly with the 3’-located gene desert (Figure 3C).

Figure 3 with 1 supplement see all

Download asset Open asset

Interestingly, while the general tendency in the bimodal distribution of interactions was thus comparable between mouse and snake full embryos, the pattern of contacts presented important differences between the two species. In the mouse developing proximal limb for instance, Hoxd11 and Hoxd9 preferentially interact with the telomeric sub-TAD referred to as region ‘b’ rather than the sub-TAD ‘a’ (Figure 3—figure supplement 1A and C) (Andrey et al., 2013). Mouse region ‘b’ thus likely contains important proximal limb regulatory sequences. In whole embryo tissue, we found that the contacts were rather equally distributed between the two regions ‘a’ and ‘b’ (Figure 3—figure supplement 1A and C). In contrast, Hoxd4 preferentially interacted with the sub-TAD ’a’ (Figure 3A; brackets), similar to Hoxd1 in proximal limbs (Andrey et al., 2013). In snake embryonic cells, however, Hoxd9 and to a much lesser extent Hoxd11, displayed an interaction pattern related to that of Hoxd4 with contacts enriched within the sub-TAD ’a’, suggesting the absence of strong regulatory controls located in region ‘b’ of the snake gene desert (Figure 3—figure supplement 1B and D).

Divergent evolution of genital bud-specific regulatory sequences

In the mouse, strong contacts between Hoxd13 and its flanking regulatory landscape were associated with its function during the development of both digits and external genitals (Lonfat et al., 2014; Montavon et al., 2011). As snakes lack digits, we investigated whether the conservation of this particular chromatin domain was related to the existence of external genital organs. Male snakes display hemipenes (HP), resulting from symmetrical genital buds during development. As it was proposed that the genitals of mammals have a different embryonic origin than those of other amniotes such as squamates (Tschopp et al., 2014), the existence of the same global regulation in snakes was unclear.

In situ hybridization revealed that 3’-located genes such as Hoxd4 are expressed in the snake HP (Figure 4A, top), in contrast to the mouse where neither Hoxd3 nor Hoxd4 are transcribed in this structure (Lonfat et al., 2014). This difference was also scored when Hox gene expression was analysed in the various BAC transgenic lines. While the human, mouse, chicken and zebrafish Hoxd genes were expressed mostly along the main body axis and transcribed neither in the limbs, nor in the external genitals, in agreement with previous results (see above and (Lonfat and Duboule, 2015) (Figure 4A, bottom and Figure 4—figure supplement 1), the snake BAC expressed the Hoxd11 to Hoxd4 genes in the developing limbs and genital bud (Figure 4A and Figure 4—figure supplement 1). This likely reflects a lack of repression of these genes into such structures, rather than the presence of limb and genital enhancers located inside the cluster. We investigated this issue by RT-qPCR in the incipient genitals of both mouse and snake, using as a control a region of the trunk located at the exact same anterior-posterior level. In such conditions, while the mouse Hoxd9 to Hoxd3 genes were expressed at much higher levels in the trunk when compared to the genitals, the steady state levels of snake mRNAs were nearly the same in both tissues (Figure 4B). Therefore, these results suggested that the snake HoxD cluster lacks the sequences necessary to prevent Hox gene expression from the trunk lateral mesoderm, a critical factor to properly develop limbs and genitals in mouse (see discussion).

Figure 4 with 1 supplement see all

Download asset Open asset

Enhancer evolution

To further explore this difference in Hox gene regulation between murine and snake external genitalia, we assessed whether Hoxd gene regulation in the developing genitals also relied upon enhancer sequences located in the regulatory landscape upstream of Hoxd13. We performed 4C-seq analyses using embryonic mouse GT and snake HP tissues at comparable stages, as well as control trunk tissues. As expected, both the mouse Hoxd11 and Hoxd13 genes showed more interactions with the centromeric gene desert in the GT material than in control trunk material (Figure 4C). In contrast, the snake Hoxd13 and Hoxd11 interaction profiles were not significantly different, when either HP or the control samples were used. We searched the snake gene desert for the presence of the GT1 and GT2 sequences, two DNA segments described in the mouse counterpart to specifically interact with Hoxd13 in the developing GT (Lonfat et al., 2014) (Figure 4C, bottom left) and could identify them (Figure 4C, bottom right). However, even though these sequences are well conserved in the snake, they did not significantly increase their interactions with Hoxd13 during the development of the snake genitals (Figure 4C, right). In fact, the comparative analysis of the snake 4C-seq data revealed that, if anything, only the Prox sequence, a known mouse GT and limb enhancer (Gonzalez et al., 2007) appeared to gain interactions in the snake HP sample, when compared to control tissue.

Therefore, in addition to the fact that snakes are limbless and that their external genitalia may derive from a different embryonic origin (Tschopp et al., 2014), our results pointed to distinct modalities in the implementation of global gene regulation at the HoxD locus. Consequently, we searched for the presence within the snake centromeric TAD of the digit- and genital bud-specific enhancers previously identified in mammals (Lonfat et al., 2014; Montavon et al., 2011). Surprisingly, all these murine enhancers conserved within mammals and up to birds showed some level of conservation in snakes (Figure 5A). To try and assess the functional potential of these sequences, we isolated the snake sequences Prox, GT2 and Island I (Figure 5A,B) and used them separately in a mouse transgenic enhancer assay. The mouse counterpart of the Prox sequence displayed activity in both the developing digits and GT. The mouse GT2 sequence is a genital only-specific enhancer and the mouse Island I sequence displays limb-only enhancer specificity when placed upstream of a lacZ reporter in a transgenic assay (Montavon et al., 2011; Gonzalez et al., 2007; Lonfat et al., 2014) (Figure 5C, top).

Figure 5

Download asset Open asset

The snake Prox element elicited a robust lacZ staining in the developing mouse GT. Interestingly however, and in contrast to the mouse sequence, staining was not detected in the growing limb buds, indicating that the snake Prox sequence had lost its potential to drive transcription in digits (Figure 5C, bottom). Likewise, the snake GT2 sequence was able to drive reporter gene expression in the mouse GT. However, the level of lacZ staining obtained was consistently weaker than with the mouse GT2 sequence (Figure 5C, bottom), perhaps related to the weak (if any) contact observed by 4C between the snake Hoxd13 gene and this sequence (Figure 4C). Therefore, in the two cases where a mouse sequence displayed an enhancer potential for the developing GT, the cognate snake sequence appeared to share part or all of this potential (Figure 5C, bottom).

We then turned to the snake Island I, a mouse limb-specific enhancer sequence and, noteworthy, the snake version was able to drive reporter gene expression in the mouse GT while unable to elicit any staining in the developing limb buds (Figure 5C, top). Therefore, in this case, the same regulatory sequence conserved between the mouse and the snake was interpreted either as a limb- or as a genital-specific enhancer by the mouse, when introduced as transgenes. To further evaluate this striking change in enhancer potential, we cloned and investigated the regulatory capacity of both the chicken and the green anole lizard Island I sequences. The chicken construct elicited staining in limbs but not in the GT, i.e. in a pattern reminiscent of the mouse rather than the snake Island I (Figure 5C, top). Noteworthy, while the lizard Island I also drove reporter gene expression in limbs, weak staining was scored in the GT, somehow displaying an intermediate enhancer specificity between the snakes and the other amniotes assayed (Figure 5C, top).

Proper sequential Hox gene activation in time and space during the elongation of the main body axis is critical for the correct patterning of the axial skeleton (Deschamps and van Nes, 2005). While the underlying regulatory mechanisms remain to be fully understood (see e.g. (Gaunt, 2015), they likely involve control sequences located both inside and outside of the Hox gene clusters (Tschopp et al., 2009; Tschopp and Duboule, 2011) as well as concurrent epigenetic modifications (Soshnikova and Duboule, 2009).

Transgenic approaches have identified several cis-regulatory elements, which could reproduce, for the most part, a Hox-like expression pattern in the main body axis and which mapped close to their target gene(s) (for examples, see (Bel-Vialar et al., 2002; Brend et al., 2003; Charité et al., 1995; Kwan et al., 2001; Oosterveen et al., 2003; Sharpe et al., 1998). However, our various transgenic mouse lines containing HoxD clusters from different vertebrate species showed that, in most vertebrates, regulatory sequences located in the gene deserts flanking the HoxD cluster are necessary for proper expression in the embryonic trunk mesoderm. Indeed, when using the expression of Hoxd4 as a read-out of the regulatory potential contained either in the mouse, the human, the chicken or the zebrafish clusters, ventral mesodermal expression was not detected. This result is at odds with previous reports describing the presence of mesodermal enhancers on a short mouse transgene derived from this locus (Zhang et al., 1997; Morrison et al., 1997). Similarly, we found Hoxd3 expression to be absent from the ventral mesoderm in mouse embryos that lack the telomeric gene desert while previous work has reported that a small single-copy construct containing the Hoxd3 locus was able to drive reporter gene expression in this tissue (Tschopp et al., 2012).

These discrepancies may derive from a qualitative aspect whereby a short transgene may reveal the potentialities absent from a more complex BAC environment, such as in the case of Hoxd3. They may also reflect quantitative differences, for example when comparing (close to) single copy BAC integrations with a large number of head to tail integrations of a shorter transgene, as for Hoxd4. In any case, a different result was obtained when the snake HoxD BAC was used since, unlike the other vertebrate clusters tested, the snake transgene appeared to contain the regulatory elements necessary for expression in the trunk mesoderm. The H3K27 acetylation profiles in mouse and corn snake trunks supported this result for only few acetylation peaks were scored outside of the snake cluster compared to the mouse. In an evolutionary context, it is possible that in snakes, long-range mesodermal enhancers were progressively complemented by local enhancers to regulate Hoxd gene expression in the developing body axis, perhaps to provide an increased fine-tuned control in the expression balance between single Hoxd genes during embryonic development.

In the absence of any genetic approach to study snake development, it is difficult to evaluate the functional relevance of this difference in regulation. It is however worth noting that the snake HoxD cluster BAC was the only transgenic configuration where expression was strong in the limbs and external genitals, whereas BACs derived from animals with limbs did not elicit an expression of Hoxd genes into the transgenic limbs, even though the endogenous genes were strongly expressed there. Also, isolated mouse mesodermal enhancers were often described to activate reporter gene expression in secondary structures such as limbs (Charité et al., 1995; Kwan et al., 2001; Sharpe et al., 1998; Renucci et al., 1992). In tetrapods, this apparent paradox may reflect the necessity for a highly specific type of regulation to control both Hoxa and Hoxd genes in specific domains of the growing limb buds (Andrey et al., 2013; Berlivet et al., 2013; Woltering et al., 2014). Implementing such global limb regulations may require previous regulatory inputs to be terminated. Our results using the corn snake BAC suggest that, in the absence of limbs, this negative control may have been lost in the course of evolution, leading to the maintenance of transcription in all mesoderm derivatives. Whether or not this increased ‘mesoderm potential’ present in the snake HoxD cluster may somehow relate to the presence of ventral mesodermal enhancer remains to be clarified.

Hox clusters of jawed vertebrates show high levels of chromatin compaction (Noordermeer et al., 2011; Fabre et al., 2015), a feature associated with the unusual level of gene packaging and organization, which occurred at the roots of the vertebrate lineage (Duboule, 2007) together with the emergence and generalisation of long-range regulations at these loci (Darbellay and Duboule, 2016). Squamate Hox clusters however seem to slightly deviate from this rule by having accumulated a large number of transposable elements (Di-Poï et al., 2009), a situation rarely found around genetic loci of developmental importance (Simons et al., 2007) due to the potential effects of such sequences to elicit genetic and morphological variations (Kidwell and Lisch, 1997; Friedli and Trono, 2015). While the presence of such repeated elements within and around the snake HoxD cluster may have been associated with differences in the location of enhancer sequences, the distribution of chromatin modifications did not point to any drastic regulatory re-organization of this gene cluster in snakes. Indeed, the H3K9me3 histone mark, which in some cases is associated with TEs such as LTRs, LINEs and SINEs (Friedli and Trono, 2015; Mikkelsen et al., 2007) was not found within the HoxD cluster itself. In addition, the analysis of other chromatin marks present in trunk tissue during the sequential activation of the gene cluster displayed distributions similar to those found in the mouse cluster.

These global similarities between mouse and snake in the structure of the HoxD cluster were also noticed when interaction profiles were considered. There again, the snake embryonic material displayed the bimodal distribution of contacts on both sides of the gene cluster, as expected either from several studies using specific mouse samples (Lonfat et al., 2014; Andrey et al., 2013), or from the full embryonic material used in this work. In all cases, most Hoxd genes tend to naturally interact within the ‘telomeric’ (3’-located) TAD, whereas Hoxd13 was strongly associated with the ‘centromeric’ (5’-located) TAD. However, a significant difference was scored in the interaction profiles of Hoxd9, which displayed more contacts further away of region CNS39 in the mouse than in the snake sample. As this regulatory region is particularly active in proximal limbs, the lower frequency of contacts observed in snake embryos was not unexpected. This observation indicates that, while the general bimodal TAD regulatory structure is conserved at the HoxD locus, differences between vertebrate species may exist either in the extent, or in the internal organization of interactions within the TADs.

This latter point was of particular interest as the mouse centromeric TAD was initially defined due the presence of many limb and genital-specific enhancers (Montavon et al., 2011; Lonfat et al., 2014), which were not necessarily expected to be conserved in the snake orthologous landscape due to both the absence of limbs and the presumably distinct origin of snake external genitalia following the relative shift of the cloaca over the course of evolution (Tschopp et al., 2014). However, tetrapod limb-specific enhancers are often conserved in snakes and a significant overlap between limb and genital cis-regulatory mechanisms was recently reported (Infante et al., 2015), in agreement with the similarities between the molecular mechanisms employed to generate the two structures (Kondo et al., 1997; Cohn, 2011; Yamada et al., 2006). Indeed when a mouse limb- and genital-specific Tbx4 enhancer sequence was isolated from snake, it could only recapitulate genital expression, thus having lost the limb-specific regulatory potential (Infante et al., 2015).

Consistent with this observation, we find that the snake counterpart of the mouse limb and genital enhancer Prox (Gonzalez et al., 2007) has lost its limb regulatory potential, while keeping a strong capacity to drive expression in the developing genitalia. This suggests that the mouse Prox consists of two regulatory modules, while the snake Prox has kept the genital specificity only. More strikingly however, Island I, which in the mouse is a limb-only specific enhancer (Montavon et al., 2011), switched its regulatory capacity in the snake to become a genital-only specific enhancer, accompanied by elevated enhancer-promoter interactions as assayed by 4C (Figure 4C – asterisk). On the other hand, the chicken Island I revealed enhancer specificities almost identical to those of the mouse sequence despite the fact that birds are more closely related to squamates than to mammals, suggesting that Island I had a limb-only activity at the time of divergence between mammals and reptiles/birds. Interestingly, the lizard sequence displayed a weak activity in the genital bud, in addition to the limb, indicating that the co-option to a genital function likely occurred in the squamate clade and thus preceded limb loss in snakes.

External genitalia of squamates, unlike that of mammals, were proposed to have the same embryonic origin as limbs (Tschopp et al., 2014). Although, this could bias our mouse-based transgenic analysis, the snake Prox sequence could drive reporter gene expression specifically in the mouse GT and not in the limb, making it unlikely that differences in embryonic origin could interfere with this particular analysis. While the generation of the same transgenics in snakes would be necessary to fully rule out this bias, such experiments are not feasible for the moment.

Altogether our results show that, while the general bimodal regulatory strategy is conserved, some profound differences in the regulation are scored at the HoxD locus between two species displaying strikingly distinct morphologies. It is as yet unclear if such changes were causative of the extensive morphological changes that snakes experienced over the course of evolution, or whether they are merely consequential. It nevertheless indicates that vertebrates with extreme variations in those systems known to be under the control of Hox genes (vertebral number and identities, limbs, genitals) rely upon the same general regulatory architecture and principles at Hox loci. In this view, vertebrate morphological evolution was accompanied by changes in Hox gene regulation, yet such variations were constrained within the general regulatory framework found at these loci. This may reflect selective pressures that impose essential basic properties to vertebrate body plans, while other more subtle morphological specificities, less likely to result in adverse effects, may arise in the course of evolution.

1. 1. Isabel Guerreiro
   2. Denis Duboule

   (2016) Reorganization of HoxD regulatory landscapes during the evolution of a snake-like body plan

   Publicly available at NCBI Gene Expression Omnibus (accession no: GSE79048).

   http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE79048
2. 1. Isabel Guerreiro
   2. Denis Duboule

   (2016) CornSnake_HoxD_scaffold

   Publicly available at NCBI GenBank (accession no: KU866087).

   http://www.ncbi.nlm.nih.gov/nuccore/KU866087

1. 1. Akam M

   (1989) Hox and HOM: homologous gene clusters in insects and vertebrates

   Cell 57:347–349.

   https://doi.org/10.1016/0092-8674(89)90909-4

   • Google Scholar
2. 1. Amores A
   2. Force A
   3. Yan YL
   4. Joly L
   5. Amemiya C
   6. Fritz A
   7. Ho RK
   8. Langeland J
   9. Prince V
   10. Wang YL
   11. Westerfield M
   12. Ekker M
   13. Postlethwait JH

   (1998) Zebrafish hox clusters and vertebrate genome evolution

   Science 282:1711–1714.

   https://doi.org/10.1126/science.282.5394.1711

   • Google Scholar
3. 1. Andrey G
   2. Montavon T
   3. Mascrez B
   4. Gonzalez F
   5. Noordermeer D
   6. Leleu M
   7. Trono D
   8. Spitz F
   9. Duboule D

   (2013) A switch between topological domains underlies HoxD genes collinearity in mouse limbs

   Science 340:1234167.

   https://doi.org/10.1126/science.1234167

   • Google Scholar
4. 1. Bel-Vialar S
   2. Itasaki N
   3. Krumlauf R

   (2002)

   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 129:5103–5115.

   • Google Scholar
5. 1. Berlivet S
   2. Paquette D
   3. Dumouchel A
   4. Langlais D
   5. Dostie J
   6. Kmita M

   (2013) Clustering of tissue-specific sub-TADs accompanies the regulation of HoxA genes in developing limbs

   PLoS Genetics 9:e1004018.

   https://doi.org/10.1371/journal.pgen.1004018

   • Google Scholar
6. 1. Brend T
   2. Gilthorpe J
   3. Summerbell D
   4. Rigby PW

   (2003) Multiple levels of transcriptional and post-transcriptional regulation are required to define the domain of Hoxb4 expression

   Development 130:2717–2728.

   https://doi.org/10.1242/dev.00471

   • Google Scholar
7. 1. Burge C
   2. Karlin S

   (1997) Prediction of complete gene structures in human genomic DNA

   Journal of Molecular Biology 268:78–94.

   https://doi.org/10.1006/jmbi.1997.0951

   • Google Scholar
8. 1. Burke AC
   2. Nelson CE
   3. Morgan BA
   4. Tabin C

   (1995)

   Hox genes and the evolution of vertebrate axial morphology

   Development 121:333–346.

   • Google Scholar
9. 1. Carroll SB

   (2008) Evo-devo and an expanding evolutionary synthesis: a genetic theory of morphological evolution

   Cell 134:25–36.

   https://doi.org/10.1016/j.cell.2008.06.030

   • Google Scholar
10. 1. Castoe TA
    2. de Koning APJ
    3. Hall KT
    4. Card DC
    5. Schield DR
    6. Fujita MK
    7. Ruggiero RP
    8. Degner JF
    9. Daza JM
    10. Gu W
    11. Reyes-Velasco J
    12. Shaney KJ
    13. Castoe JM
    14. Fox SE
    15. Poole AW
    16. Polanco D
    17. Dobry J
    18. Vandewege MW
    19. Li Q
    20. Schott RK
    21. Kapusta A
    22. Minx P
    23. Feschotte C
    24. Uetz P
    25. Ray DA
    26. Hoffmann FG
    27. Bogden R
    28. Smith EN
    29. Chang BSW
    30. Vonk FJ
    31. Casewell NR
    32. Henkel CV
    33. Richardson MK
    34. Mackessy SP
    35. Bronikowski AM
    36. Yandell M
    37. Warren WC
    38. Secor SM
    39. Pollock DD
    40. Bronikowsi AM

    (2013) The Burmese python genome reveals the molecular basis for extreme adaptation in snakes

    PNAS 110:20645–20650.

    https://doi.org/10.1073/pnas.1314475110

    • Google Scholar
11. 1. Charité J
    2. de Graaff W
    3. Vogels R
    4. Meijlink F
    5. Deschamps J

    (1995) Regulation of the Hoxb-8 gene: synergism between multimerized cis-acting elements increases responsiveness to positional information

    Developmental Biology 171:294–305.

    https://doi.org/10.1006/dbio.1995.1282

    • Google Scholar
12. 1. Cohn MJ

    (2011) Development of the external genitalia: conserved and divergent mechanisms of appendage patterning

    Developmental Dynamics 240:1108–1115.

    https://doi.org/10.1002/dvdy.22631

    • Google Scholar
13. 1. Darbellay F
    2. Duboule D

    (2016) Topological Domains, Metagenes, and the Emergence of Pleiotropic Regulations at Hox Loci

    Current Topics in Developmental Biology 116:299–314.

    https://doi.org/10.1016/bs.ctdb.2015.11.022

    • Google Scholar
14. 1. David FP
    2. Delafontaine J
    3. Carat S
    4. Ross FJ
    5. Lefebvre G
    6. Jarosz Y
    7. Sinclair L
    8. Noordermeer D
    9. Rougemont J
    10. Leleu M

    (2014) HTSstation: a web application and open-access libraries for high-throughput sequencing data analysis

    PLoS One 9:e85879.

    https://doi.org/10.1371/journal.pone.0085879

    • Google Scholar
15. 1. de Laat W
    2. Dekker J

    (2012) 3C-based technologies to study the shape of the genome

    Methods 58:189–191.

    https://doi.org/10.1016/j.ymeth.2012.11.005

    • Google Scholar
16. 1. De Robertis EM

    (2008) Evo-devo: variations on ancestral themes

    Cell 132:185–195.

    https://doi.org/10.1016/j.cell.2008.01.003

    • Google Scholar
17. 1. Dekker J
    2. Rippe K
    3. Dekker M
    4. Kleckner N

    (2002) Capturing chromosome conformation

    Science 295:1306–1311.

    https://doi.org/10.1126/science.1067799

    • Google Scholar
18. 1. Delpretti S
    2. Montavon T
    3. Leleu M
    4. Joye E
    5. Tzika A
    6. Milinkovitch M
    7. Duboule D

    (2013) Multiple enhancers regulate Hoxd genes and the Hotdog LncRNA during cecum budding

    Cell Reports 5:137–150.

    https://doi.org/10.1016/j.celrep.2013.09.002

    • Google Scholar
19. 1. Delpretti S
    2. Zakany J
    3. Duboule D

    (2012) A function for all posterior Hoxd genes during digit development?

    Developmental Dynamics 241:792–802.

    https://doi.org/10.1002/dvdy.23756

    • Google Scholar
20. 1. Deschamps J
    2. van Nes J

    (2005) Developmental regulation of the Hox genes during axial morphogenesis in the mouse

    Development 132:2931–2942.

    https://doi.org/10.1242/dev.01897

    • Google Scholar
21. 1. Deschamps J

    (2007) Ancestral and recently recruited global control of the Hox genes in development

    Current Opinion in Genetics & Development 17:422–427.

    https://doi.org/10.1016/j.gde.2007.07.008

    • Google Scholar
22. 1. Di-Poi N
    2. Montoya-Burgos JI
    3. Duboule D

    (2009) Atypical relaxation of structural constraints in Hox gene clusters of the green anole lizard

    Genome Research 19:602–610.

    https://doi.org/10.1101/gr.087932.108

    • Google Scholar
23. 1. Di-Poï N
    2. Montoya-Burgos JI
    3. Miller H
    4. Pourquié O
    5. Milinkovitch MC
    6. Duboule D

    (2010) Changes in Hox genes' structure and function during the evolution of the squamate body plan

    Nature 464:99–103.

    https://doi.org/10.1038/nature08789

    • Google Scholar
24. 1. Dixon JR
    2. Selvaraj S
    3. Yue F
    4. Kim A
    5. Li Y
    6. Shen Y
    7. Hu M
    8. Liu JS
    9. Ren B

    (2012) Topological domains in mammalian genomes identified by analysis of chromatin interactions

    Nature 485:376–380.

    https://doi.org/10.1038/nature11082

    • Google Scholar
25. 1. Duboule D
    2. Dollé P

    (1989)

    The structural and functional organization of the murine HOX gene family resembles that of Drosophila homeotic genes

    The EMBO Journal 8:1497–1505.

    • Google Scholar
26. 1. Duboule D
    2. Wilkins AS

    (1998) The evolution of 'bricolage'

    Trends in Genetics 14:54–59.

    https://doi.org/10.1016/S0168-9525(97)01358-9

    • Google Scholar
27. 1. Duboule D

    (2007) The rise and fall of Hox gene clusters

    Development 134:2549–2560.

    https://doi.org/10.1242/dev.001065

    • Google Scholar
28. 1. Fabre PJ
    2. Benke A
    3. Manley S
    4. Duboule D
    5. Behke A

    (2015) Visualizing the HoxD Gene Cluster at the Nanoscale Level

    Cold Spring Harbor Symposia on Quantitative Biology 80:9–16.

    https://doi.org/10.1101/sqb.2015.80.027177

    • Google Scholar
29. 1. Featherstone MS
    2. Baron A
    3. Gaunt SJ
    4. Mattei MG
    5. Duboule D

    (1988) Hox-5.1 defines a homeobox-containing gene locus on mouse chromosome 2

    PNAS 85:4760–4764.

    https://doi.org/10.1073/pnas.85.13.4760

    • Google Scholar
30. 1. Friedli M
    2. Trono D

    (2015) The developmental control of transposable elements and the evolution of higher species

    Annual Review of Cell and Developmental Biology 3131:429–451.

    https://doi.org/10.1146/annurev-cellbio-100814-125514

    • Google Scholar
31. 1. Garcia-Fernández J
    2. Holland PW

    (1994) Archetypal organization of the amphioxus Hox gene cluster

    Nature 370:563–566.

    https://doi.org/10.1038/370563a0

    • Google Scholar
32. 1. Gaunt SJ
    2. Sharpe PT
    3. Duboule D

    (1988)

    Spatially restricted domains of homeo-gene transcripts in mouse embryos: relation to a segmented body plan

    Development 104:169–179.

    • Google Scholar
33. 1. Gaunt SJ

    (2015) The significance of Hox gene collinearity

    The International Journal of Developmental Biology 59:159–170.

    https://doi.org/10.1387/ijdb.150223sg

    • Google Scholar
34. 1. Gérard M
    2. Chen JY
    3. Gronemeyer H
    4. Chambon P
    5. Duboule D
    6. Zákány J

    (1996) In vivo targeted mutagenesis of a regulatory element required for positioning the Hoxd-11 and Hoxd-10 expression boundaries

    Genes & Development 10:2326–2334.

    https://doi.org/10.1101/gad.10.18.2326

    • Google Scholar
35. 1. Gilbert C
    2. Meik JM
    3. Dashevsky D
    4. Card DC
    5. Castoe TA
    6. Schaack S

    (2014) Endogenous hepadnaviruses, bornaviruses and circoviruses in snakes

    Proceedings of the Royal Society B: Biological Sciences 281:20141122.

    https://doi.org/10.1098/rspb.2014.1122

    • Google Scholar
36. 1. Gomez C
    2. Ozbudak EM
    3. Wunderlich J
    4. Baumann D
    5. Lewis J
    6. Pourquié O

    (2008) Control of segment number in vertebrate embryos

    Nature 454:335–339.

    https://doi.org/10.1038/nature07020

    • Google Scholar
37. 1. Gonzalez F
    2. Duboule D
    3. Spitz F

    (2007) Transgenic analysis of Hoxd gene regulation during digit development

    Developmental Biology 306:847–859.

    https://doi.org/10.1016/j.ydbio.2007.03.020

    • Google Scholar
38. 1. Graham A
    2. Papalopulu N
    3. Krumlauf R

    (1989) The murine and Drosophila homeobox gene complexes have common features of organization and expression

    Cell 57:367–378.

    https://doi.org/10.1016/0092-8674(89)90912-4

    • Google Scholar
39. 1. Guerreiro I
    2. Duboule D

    (2014) Snakes: hatching of a model system for Evo-Devo?

    The International Journal of Developmental Biology 58:727–732.

    https://doi.org/10.1387/ijdb.150026dd

    • Google Scholar
40. 1. Head JJ
    2. Polly PD

    (2015) Evolution of the snake body form reveals homoplasy in amniote Hox gene function

    Nature 520:86–89.

    https://doi.org/10.1038/nature14042

    • Google Scholar
41. 1. Hérault Y
    2. Rassoulzadegan M
    3. Cuzin F
    4. Duboule D

    (1998) Engineering chromosomes in mice through targeted meiotic recombination (TAMERE)

    Nature Genetics 20:381–384.

    https://doi.org/10.1038/3861

    • Google Scholar
42. 1. Infante CR
    2. Mihala AG
    3. Park S
    4. Wang JS
    5. Johnson KK
    6. Lauderdale JD
    7. Menke DB

    (2015) Shared Enhancer Activity in the Limbs and Phallus and Functional Divergence of a Limb-Genital cis-Regulatory Element in Snakes

    Developmental Cell 35:107–119.

    https://doi.org/10.1016/j.devcel.2015.09.003

    • Google Scholar
43. 1. Izpisúa-Belmonte JC
    2. Falkenstein H
    3. Dollé P
    4. Renucci A
    5. Duboule D

    (1991)

    Murine genes related to the Drosophila AbdB homeotic genes are sequentially expressed during development of the posterior part of the body

    The EMBO Journal 10:2279–2289.

    • Google Scholar
44. 1. Kidwell MG
    2. Lisch D

    (1997) Transposable elements as sources of variation in animals and plants

    PNAS 94:7704–7711.

    https://doi.org/10.1073/pnas.94.15.7704

    • Google Scholar
45. Book

    1. Kirschner M 
    2. Gerhart J 
    3. Ebrary INC

    (2005)

    The Plausibility of Life Resolving Darwin's Dilemma

    New Haven: Yale University Press.

    • Google Scholar
46. 1. Kmita M
    2. Duboule D

    (2003) Organizing axes in time and space; 25 years of colinear tinkering

    Science 301:331–333.

    https://doi.org/10.1126/science.1085753

    • Google Scholar
47. 1. Kondo T
    2. Zákány J
    3. Innis JW
    4. Duboule D

    (1997) Of fingers, toes and penises

    Nature 390:29.

    https://doi.org/10.1038/36234

    • Google Scholar
48. 1. Krumlauf R

    (1994) Hox genes in vertebrate development

    Cell 78:191–201.

    https://doi.org/10.1016/0092-8674(94)90290-9

    • Google Scholar
49. 1. Kwan CT
    2. Tsang SL
    3. Krumlauf R
    4. Sham MH

    (2001) Regulatory analysis of the mouse Hoxb3 gene: multiple elements work in concert to direct temporal and spatial patterns of expression

    Developmental Biology 232:176–190.

    https://doi.org/10.1006/dbio.2001.0157

    • Google Scholar
50. 1. Lee EC
    2. Yu D
    3. Martinez de Velasco J
    4. Tessarollo L
    5. Swing DA
    6. Court DL
    7. Jenkins NA
    8. Copeland NG

    (2001) A highly efficient Escherichia coli-based chromosome engineering system adapted for recombinogenic targeting and subcloning of BAC DNA

    Genomics 73:56–65.

    https://doi.org/10.1006/geno.2000.6451

    • Google Scholar
51. 1. Lee TI
    2. Johnstone SE
    3. Young RA

    (2006) Chromatin immunoprecipitation and microarray-based analysis of protein location

    Nature Protocols 1:729–748.

    https://doi.org/10.1038/nprot.2006.98

    • Google Scholar
52. 1. Lemons D
    2. McGinnis W

    (2006) Genomic evolution of Hox gene clusters

    Science 313:1918–1922.

    https://doi.org/10.1126/science.1132040

    • Google Scholar
53. 1. Lewis EB

    (1978) A gene complex controlling segmentation in Drosophila

    Nature 276:565–570.

    https://doi.org/10.1038/276565a0

    • Google Scholar
54. 1. Lonfat N
    2. Duboule D

    (2015) Structure, function and evolution of topologically associating domains (TADs) at HOX loci

    FEBS Letters 589:2869–2876.

    https://doi.org/10.1016/j.febslet.2015.04.024

    • Google Scholar
55. 1. Lonfat N
    2. Montavon T
    3. Darbellay F
    4. Gitto S
    5. Duboule D

    (2014) Convergent evolution of complex regulatory landscapes and pleiotropy at Hox loci

    Science 346:1004–1006.

    https://doi.org/10.1126/science.1257493

    • Google Scholar
56. 1. Maeda RK
    2. Karch F

    (2006) The ABC of the BX-C: the bithorax complex explained

    Development 133:1413–1422.

    https://doi.org/10.1242/dev.02323

    • Google Scholar
57. 1. Margueron R
    2. Reinberg D

    (2011) The Polycomb complex PRC2 and its mark in life

    Nature 469:343–349.

    https://doi.org/10.1038/nature09784

    • Google Scholar
58. 1. Martens JH
    2. O'Sullivan RJ
    3. Braunschweig U
    4. Opravil S
    5. Radolf M
    6. Steinlein P
    7. Jenuwein T

    (2005) The profile of repeat-associated histone lysine methylation states in the mouse epigenome

    The EMBO Journal 24:800–812.

    https://doi.org/10.1038/sj.emboj.7600545

    • Google Scholar
59. 1. McGinnis W
    2. Garber RL
    3. Wirz J
    4. Kuroiwa A
    5. Gehring WJ

    (1984) A homologous protein-coding sequence in Drosophila homeotic genes and its conservation in other metazoans

    Cell 37:403–408.

    https://doi.org/10.1016/0092-8674(84)90370-2

    • Google Scholar
60. 1. Mehta TK
    2. Ravi V
    3. Yamasaki S
    4. Lee AP
    5. Lian MM
    6. Tay BH
    7. Tohari S
    8. Yanai S
    9. Tay A
    10. Brenner S
    11. Venkatesh B

    (2013) Evidence for at least six Hox clusters in the Japanese lamprey (Lethenteron japonicum)

    PNAS 110:16044–16049.

    https://doi.org/10.1073/pnas.1315760110

    • Google Scholar
61. 1. Mikkelsen TS
    2. Ku M
    3. Jaffe DB
    4. Issac B
    5. Lieberman E
    6. Giannoukos G
    7. Alvarez P
    8. Brockman W
    9. Kim TK
    10. Koche RP
    11. Lee W
    12. Mendenhall E
    13. O'Donovan A
    14. Presser A
    15. Russ C
    16. Xie X
    17. Meissner A
    18. Wernig M
    19. Jaenisch R
    20. Nusbaum C
    21. Lander ES
    22. Bernstein BE

    (2007) Genome-wide maps of chromatin state in pluripotent and lineage-committed cells

    Nature 448:553–560.

    https://doi.org/10.1038/nature06008

    • Google Scholar
62. 1. Milinkovitch MC
    2. Tzika A

    (2007) Escaping the mouse trap: the selection of new Evo-Devo model species

    Journal of Experimental Zoology Part B: Molecular and Developmental Evolution 308:337–346.

    https://doi.org/10.1002/jez.b.21180

    • Google Scholar
63. 1. Montavon T
    2. Soshnikova N
    3. Mascrez B
    4. Joye E
    5. Thevenet L
    6. Splinter E
    7. de Laat W
    8. Spitz F
    9. Duboule D

    (2011) A regulatory archipelago controls Hox genes transcription in digits

    Cell 147:1132–1145.

    https://doi.org/10.1016/j.cell.2011.10.023

    • Google Scholar
64. 1. Morrison A
    2. Ariza-McNaughton L
    3. Gould A
    4. Featherstone M
    5. Krumlauf R

    (1997)

    HOXD4 and regulation of the group 4 paralog genes

    Development 124:3135–3146.

    • Google Scholar
65. 1. Noordermeer D
    2. Leleu M
    3. Splinter E
    4. Rougemont J
    5. De Laat W
    6. Duboule D

    (2011) The dynamic architecture of Hox gene clusters

    Science 334:222–225.

    https://doi.org/10.1126/science.1207194

    • Google Scholar
66. 1. Nora EP
    2. Lajoie BR
    3. Schulz EG
    4. Giorgetti L
    5. Okamoto I
    6. Servant N
    7. Piolot T
    8. van Berkum NL
    9. Meisig J
    10. Sedat J
    11. Gribnau J
    12. Barillot E
    13. Blüthgen N
    14. Dekker J
    15. Heard E

    (2012) Spatial partitioning of the regulatory landscape of the X-inactivation centre

    Nature 485:381–385.

    https://doi.org/10.1038/nature11049

    • Google Scholar
67. 1. Oosterveen T
    2. Niederreither K
    3. Dollé P
    4. Chambon P
    5. Meijlink F
    6. Deschamps J

    (2003) Retinoids regulate the anterior expression boundaries of 5' Hoxb genes in posterior hindbrain

    The EMBO Journal 22:262–269.

    https://doi.org/10.1093/emboj/cdg029

    • Google Scholar
68. 1. Prince VE
    2. Joly L
    3. Ekker M
    4. Ho RK

    (1998)

    Zebrafish hox genes: genomic organization and modified colinear expression patterns in the trunk

    Development 125:407–420.

    • Google Scholar
69. 1. Ramírez F
    2. Dündar F
    3. Diehl S
    4. Grüning BA
    5. Manke T

    (2014) deepTools: a flexible platform for exploring deep-sequencing data

    Nucleic Acids Research 42:W187–191.

    https://doi.org/10.1093/nar/gku365

    • Google Scholar
70. 1. Renucci A
    2. Zappavigna V
    3. Zàkàny J
    4. Izpisúa-Belmonte JC
    5. Bürki K
    6. Duboule D

    (1992)

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

    The EMBO Journal 11:1459–1468.

    • Google Scholar
71. 1. Sharpe J
    2. Nonchev S
    3. Gould A
    4. Whiting J
    5. Krumlauf R

    (1998) Selectivity, sharing and competitive interactions in the regulation of Hoxb genes

    The EMBO Journal 17:1788–1798.

    https://doi.org/10.1093/emboj/17.6.1788

    • Google Scholar
72. 1. Simons C
    2. Makunin IV
    3. Pheasant M
    4. Mattick JS

    (2007) Maintenance of transposon-free regions throughout vertebrate evolution

    BMC Genomics 8:470.

    https://doi.org/10.1186/1471-2164-8-470

    • Google Scholar
73. 1. Soshnikova N
    2. Duboule D

    (2009) Epigenetic temporal control of mouse Hox genes in vivo

    Science 324:1320–1323.

    https://doi.org/10.1126/science.1171468

    • Google Scholar
74. 1. Spitz F
    2. Gonzalez F
    3. Peichel C
    4. Vogt TF
    5. Duboule D
    6. Zákány J

    (2001) Large scale transgenic and cluster deletion analysis of the HoxD complex separate an ancestral regulatory module from evolutionary innovations

    Genes & Development 15:2209–2214.

    https://doi.org/10.1101/gad.205701

    • Google Scholar
75. 1. Spitz F
    2. Herkenne C
    3. Morris MA
    4. Duboule D

    (2005) Inversion-induced disruption of the Hoxd cluster leads to the partition of regulatory landscapes

    Nature Genetics 37:889–893.

    https://doi.org/10.1038/ng1597

    • Google Scholar
76. 1. Tschopp P
    2. Christen AJ
    3. Duboule D

    (2012) Bimodal control of Hoxd gene transcription in the spinal cord defines two regulatory subclusters

    Development 139:929–939.

    https://doi.org/10.1242/dev.076794

    • Google Scholar
77. 1. Tschopp P
    2. Duboule D

    (2011) A regulatory 'landscape effect' over the HoxD cluster

    Developmental Biology 351:288–296.

    https://doi.org/10.1016/j.ydbio.2010.12.034

    • Google Scholar
78. 1. Tschopp P
    2. Sherratt E
    3. Sanger TJ
    4. Groner AC
    5. Aspiras AC
    6. Hu JK
    7. Pourquié O
    8. Gros J
    9. Tabin CJ

    (2014) A relative shift in cloacal location repositions external genitalia in amniote evolution

    Nature 516:391–394.

    https://doi.org/10.1038/nature13819

    • Google Scholar
79. 1. Tschopp P
    2. Tarchini B
    3. Spitz F
    4. Zakany J
    5. Duboule D

    (2009) Uncoupling time and space in the collinear regulation of Hox genes

    PLoS Genetics 5:e1000398.

    https://doi.org/10.1371/journal.pgen.1000398

    • Google Scholar
80. 1. Ullate-Agote A
    2. Milinkovitch MC
    3. Tzika AC

    (2014) The genome sequence of the corn snake (Pantherophis guttatus), a valuable resource for EvoDevo studies in squamates

    The International Journal of Developmental Biology 58:881–888.

    https://doi.org/10.1387/ijdb.150060at

    • Google Scholar
81. 1. Vonk FJ
    2. Casewell NR
    3. Henkel CV
    4. Heimberg AM
    5. Jansen HJ
    6. McCleary RJR
    7. Kerkkamp HME
    8. Vos RA
    9. Guerreiro I
    10. Calvete JJ
    11. Wuster W
    12. Woods AE
    13. Logan JM
    14. Harrison RA
    15. Castoe TA
    16. de Koning APJ
    17. Pollock DD
    18. Yandell M
    19. Calderon D
    20. Renjifo C
    21. Currier RB
    22. Salgado D
    23. Pla D
    24. Sanz L
    25. Hyder AS
    26. Ribeiro JMC
    27. Arntzen JW
    28. van den Thillart GEEJM
    29. Boetzer M
    30. Pirovano W
    31. Dirks RP
    32. Spaink HP
    33. Duboule D
    34. McGlinn E
    35. Kini RM
    36. Richardson MK

    (2013) The king cobra genome reveals dynamic gene evolution and adaptation in the snake venom system

    PNAS 110:20651–20656.

    https://doi.org/10.1073/pnas.1314702110

    • Google Scholar
82. 1. Woltering JM
    2. Noordermeer D
    3. Leleu M
    4. Duboule D

    (2014) Conservation and divergence of regulatory strategies at Hox Loci and the origin of tetrapod digits

    PLoS Biology 12:e1001773.

    https://doi.org/10.1371/journal.pbio.1001773

    • Google Scholar
83. 1. Woltering JM
    2. Vonk FJ
    3. Müller H
    4. Bardine N
    5. Tuduce IL
    6. de Bakker MA
    7. Knöchel W
    8. Sirbu IO
    9. Durston AJ
    10. Richardson MK

    (2009) Axial patterning in snakes and caecilians: evidence for an alternative interpretation of the Hox code

    Developmental Biology 332:82–89.

    https://doi.org/10.1016/j.ydbio.2009.04.031

    • Google Scholar
84. 1. Yamada G
    2. Suzuki K
    3. Haraguchi R
    4. Miyagawa S
    5. Satoh Y
    6. Kamimura M
    7. Nakagata N
    8. Kataoka H
    9. Kuroiwa A
    10. Chen Y

    (2006) Molecular genetic cascades for external genitalia formation: an emerging organogenesis program

    Developmental Dynamics 235:1738–1752.

    https://doi.org/10.1002/dvdy.20807

    • Google Scholar
85. 1. Zákány J
    2. Kmita M
    3. Alarcon P
    4. de la Pompa JL
    5. Duboule D

    (2001) Localized and transient transcription of Hox genes suggests a link between patterning and the segmentation clock

    Cell 106:207–217.

    https://doi.org/10.1016/S0092-8674(01)00436-6

    • Google Scholar
86. 1. Zappavigna V
    2. Renucci A
    3. Izpisúa-Belmonte JC
    4. Urier G
    5. Peschle C
    6. Duboule D

    (1991)

    HOX4 genes encode transcription factors with potential auto- and cross-regulatory capacities

    The EMBO Journal 10:4177–4187.

    • Google Scholar
87. 1. Zhang F
    2. Pöpperl H
    3. Morrison A
    4. Kovàcs EN
    5. Prideaux V
    6. Schwarz L
    7. Krumlauf R
    8. Rossant J
    9. Featherstone MS

    (1997) Elements both 5' and 3' to the murine Hoxd4 gene establish anterior borders of expression in mesoderm and neurectoderm

    Mechanisms of Development 67:49–58.

    https://doi.org/10.1016/S0925-4773(97)00104-4

    • Google Scholar
88. 1. Zhang Y
    2. Liu T
    3. Meyer CA
    4. Eeckhoute J
    5. Johnson DS
    6. Bernstein BE
    7. Nusbaum C
    8. Myers RM
    9. Brown M
    10. Li W
    11. Liu XS

    (2008) Model-based analysis of ChIP-Seq (MACS)

    Genome Biology 9:R137.

    https://doi.org/10.1186/gb-2008-9-9-r137

    • Google Scholar

Author details

1. Isabel Guerreiro

   Department of Genetics and Evolution, University of Geneva, Geneva, Switzerland

   Contribution

   IG, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   Competing interests

   The authors declare that no competing interests exist.

2. Sandra Gitto

   Department of Genetics and Evolution, University of Geneva, Geneva, Switzerland

   Contribution

   SG, Acquisition of data, Analysis and interpretation of data

   Competing interests

   The authors declare that no competing interests exist.

3. Ana Novoa

   Instituto Gulbenkian de Ciência, Lisbon, Portugal

   Contribution

   AN, Did all the transgenic animals for the first submission, Acquisition of data

   Competing interests

   The authors declare that no competing interests exist.

4. Julien Codourey

   Department of Genetics and Evolution, University of Geneva, Geneva, Switzerland

   Contribution

   JC, Produced transgenic animals for the revised version, Acquisition of data

   Competing interests

   The authors declare that no competing interests exist.

5. Thi Hanh Nguyen Huynh

   Department of Genetics and Evolution, University of Geneva, Geneva, Switzerland

   Contribution

   THNH, Acquisition of data, Analysis and interpretation of data

   Competing interests

   The authors declare that no competing interests exist.

6. Federico Gonzalez

   Department of Genetics and Evolution, University of Geneva, Geneva, Switzerland

   Present address

   Institute for Bioengineering of Catalonia, Barcelona, Spain

   Contribution

   FG, Acquisition of data, Contributed unpublished essential data or reagents

   Competing interests

   The authors declare that no competing interests exist.

7. Michel C Milinkovitch

   Department of Genetics and Evolution, University of Geneva, Geneva, Switzerland

   Contribution

   MCM, Drafting or revising the article, Contributed unpublished essential data or reagents

   Competing interests

   The authors declare that no competing interests exist.

8. Moises Mallo

   Instituto Gulbenkian de Ciência, Lisbon, Portugal

   Contribution

   MM, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   Competing interests

   The authors declare that no competing interests exist.

   "This ORCID iD identifies the author of this article:" 0000-0002-9744-0912

9. Denis Duboule

   1. Department of Genetics and Evolution, University of Geneva, Geneva, Switzerland
   2. School of Life Sciences, Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland

   Contribution

   DD, Conception and design, Analysis and interpretation of data, Drafting or revising the article

   For correspondence

   denis.duboule@epfl.ch

   Competing interests

   The authors declare that no competing interests exist.

   "This ORCID iD identifies the author of this article:" 0000-0001-9961-2960

• 3,737

  views

• 703

  downloads

• 28

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.