Lu X

en oc


FIGURE 5.3 Domains of hox gene expression in developing arthropods and onychophorans. (A)-(C) Ubx and Abd-A proteins identified immunocytologically (red), together with Distal-less (Dll), (green), in embryos in which all trunk segments produce the same structures. (A) Centipede (Ethmostigmus rubripes). The Dll transcription factor is present in all appendages or outgrowths; anterior left. Ant, antenna; Md, mandible; Mx, maxillary segment; Tl, maxilliped, a specialized poison claw; T2, T3, and more posterior appendages, walking legs. Ubx/Abd-A is present in trunk limb-bearing


FIGURE 5.3 Domains of hox gene expression in developing arthropods and onychophorans. (A)-(C) Ubx and Abd-A proteins identified immunocytologically (red), together with Distal-less (Dll), (green), in embryos in which all trunk segments produce the same structures. (A) Centipede (Ethmostigmus rubripes). The Dll transcription factor is present in all appendages or outgrowths; anterior left. Ant, antenna; Md, mandible; Mx, maxillary segment; Tl, maxilliped, a specialized poison claw; T2, T3, and more posterior appendages, walking legs. Ubx/Abd-A is present in trunk limb-bearing


segments up to the TI/T2 boundary, and in all the limb buds (yellow). Inset: Ubx/Abd-A staining around base of limb bud. (B) Slightly older centipede embryo, anterior top left; Ubx/Abd-A staining in body wall up to TI/T2 boundary, now strongly in appendages T2-T22 (yellow), and parasegmentally, also in the nervous system (ns). (C) Onychophoran embryo (Acanthokara kaputensis; anterior, top left). Indicated body parts are Ant, antenna; J, jaws,; Sp, slime papillae; LI-LI 5, lobopods. All of these structures express the dll gene. Ubx/Abd-A are detected only in the last pair of lobopods, in LI5 (inset), partly overlapping with Dll (yellow). [(A-C) From Grenier et al. (1997) Curr. Biol. 7, 547-553.] (D) Summary of differences in developmental domains in which Ubx/Abd-A factors are detected in representative arthropod species; onychophorans are a sister group of the true arthropods. In chelicerates (here spiders) the body is divided into prosomal and opisthosomal domains, and the anterior boundary of ubx/abd-A expression (Damen et al., 1998; Abzhanov et al., 1999) lies between the first and second opisthosomal segments (A I and A2); i.e., following the segmental analysis of Damen et al., op. cit). Pdp, pedipalps; LI-L4, legs; AI-A4, opisthosomal segments. In myriapods (centipedes and millipedes) the expression boundary is further anterior [see (A, B)]; lighter color in trunk segment I (Trl) indicates an early expression domain in a centipede; Ic, intercalary segment; Mn, mandibles; Mxl and 2, maxillary segments. In crustaceans, as shown in more detail in (E), the expression boundary varies; in the most basal group, the branchiopods, it is at the junction between head and trunk, and in more derived groups more posterior (arrow). The anterior segments shown are An2, second antenna segment; and mandibular and maxillary segments. In insects (i.e., here Drosophila) ubx expression is dynamic and initially extends up to the anterior boundary of T3 (i.e., ps5; ps, parasegment), while Ubx/Abd-A protein is detected up to a boundary within segment AI (i.e, the ps6/7 boundary); but at the end of germ band extension ubx expression extends anteriorly into T2 and T3, symbolized by pink color. [(D) Slightly modified (by addition of onychophoran branch) from Abzhanov et al. (1999) Evol. Dev. 1, 77-89.] (E) Summary of changes in expression of ubx/abd-A genes, correlated with thoracic appendage morphology in the different crustacean clades. Representatives of several crustacean orders are included. Red appendages are locomotory thoracic limbs; gray appendages are specialized for feeding (maxillipeds). Three gnathal segments (appendages omitted) are indicated in white, followed by five thoracic segments. Red indicates domain of Ubx/AbdA protein as detected by immunocytology; orange, weak expression. [(E) From Averof and Patel (1997) Nature 388, 682-686, copyright Macmillan Magazines Ltd.] (F) Expression of anterior hox genes superimposed on scanning EMs of head surface structures in a crustacean, Porcellio scabar. Structures and segments are labeled as in (Fl). mxl, mx2, T2 as above; Tl/mxp, first thoracic segment/maxillipeds; pg, paragnaths; mn, mandibles; st, stomodaeum; lb, labrum; al, a2, first and second antennal segments. (Fl) labial (lab; hox I) expression domain; (F2) proboscepedia (pb; hoxl) expression domain; (F3) deformed (dfd; hox4) expression domain; and (F4) sex combs reduced (scr; hox5) expression domain. (G) Summary of lab, pb, dfd, and scr expression domains in head regions of crustaceans, insects, and chelicerates. The engrailed (en) domains serve as parasegmental (ps) markers, identifying the posterior

The greater group to which the arthropods belong is termed the "panarthro-pods." These include, in addition to the arthropods, the tardigrades (or water bears), plus onychophorans (Nielsen, 1995). Spatial differences in box gene expression within the homologous segmental domains of various panarthropod groups are illustrated in Fig. 5.3. A perfectly situated outgroup for the arthropods within the panarthropods is represented by the onychophorans (i.e., velvet worms). These animals provide a revealing side light on the process of panarthropod diversification. The onychophorans are living descendants of a lobopo-dian fauna that had diverged from the ancestors of the arthropods by the later Early Cambrian (Budd, 1996; Wheeler et al., 1993). Lobopods differ from the arthropods in that their appendages are not jointed as are arthropod appendages, but like the arthropods their bodies are segmented, and they share many anatomical features with the arthropods (Nielsen, 1995). As in myriapods, the locomo-tory appendages on onychophoran trunk segments are all alike. In Fig. 5.3A-C the distribution of Ubx and Abd-A protein (assayed together with a single antibody) in a developing centipede is compared to that in a developing onychophoran (Grenier et al., 1997). In both species the distal-less (dll) gene is expressed in all the appendages, as displayed by an antibody against the Distal-less protein (green). But the distribution of the Ubx/Abd-A protein (red) in these embryos is entirely different. In the onychophoran the ubx/abd-A genes apparently have little to do with the morphological plan of most of the appendage-bearing segments, since expression is confined to the very last of the 15 lobopod pairs (Fig. 5-3C). In contrast, in the centipede (Fig. 5.3A, B) Ubx/Abd-A is detected in every trunk segment (i.e., T2-T22), excepting only the first, which generates a special appendage. Expression of Ubx/Abd-A in the body segments also turns out to differ greatly within the arthropods. A summary is shown in Fig. 5.3D (Abzhanov et al., 1999; see legend for details and further references).

Two conclusions strike one in staring at Fig. 5.3D: first that the same box genes evidently intervene in many different patterning processes, since these animals generate diverse structures on their trunk segments (there are no wing blades on shrimp!); and second, that the expression domains of the ubx/abd-A genes with respect to these segments has been anything but immutable in evolution. This point is extended in Fig. 5.3E, where it can be seen that within crustaceans Ubx/ Abd-A distribution is correlated with diversity in the organization of the appendages on the thoracic segments (Averof and Patel, 1997). The appendage assignments with respect to segment are diverse in these crustaceans, and the rule is that part of each segment or anterior part of each parasegment. Segments are, at top, for crustaceans, oc, ocular; and the remainder as in (F). At bottom, for chelicerates, ch, chelicerate segment; pp, pedipalps, and the remainder as in (D). Columns indicate homologous segments. Horizontal bars represent domains of expression, and light colors denote weak, late, or transient expression. [(F-G) From Abzhanov and Kaufman (1999) Proc. Natl. Acad. Sci. USA 96, 10224-10229, copyright National Academy of Sciences, USA.]

locomotory appendages form in domains where Ubx/Abd-A is present, while the specialized feeding appendages (maxillipeds) arise on segments lacking these factors. The conclusion from these and additional observations (Abzhanov and Kaufman, 2000) is that the various panarthropods have coopted the ubx and the abd-A genes for segmental pattern formation processes particular to each clade, and the same is true of the antennapedia gene. It follows that the insects and many of the crustaceans have independently reorganized the regulatory systems controlling segmental appendage development; in the common ancestor from which both derived the same appendages were probably present on all the trunk segments (Averof and Akam, 1995; Abzhanov and Kaufman, 2000).

Figure 5.3F, G deal with anterior box gene expression in the head regions of arthropods from spiders to Drosopbila. Each of the four box genes included in Fig. 5.3G is expressed in a particular domain of the head of a crustacean in which certain structures arise. No doubt these box gene expressions are required for some aspect of morphogenesis in each region. But when the domains of expression of these same four genes are compared in a spider, in a crustacean, and in Drosopbila (Fig. 5.3G), we again see that, except for the labial QhoxT) gene, they are all expressed in a unique fashion in each animal (Abzhanov and Kaufman, 1999). As Fig. 5-3G indicates, different combinations of box genes are expressed in the homologous anterior segments of these three clades.

In considering Fig. 5.3 we must keep in mind that downstream of each of the box gene expression domains will lie regulatory patterning networks that include multiple targets for their products, just as illustrated in Fig. 5.2 for Ubx in the Drosopbila haltere. What is described as the general A/P patterning function of box cluster genes consists actually of an amazing variety of different networks, different linkages, and different regulatory functions. These have evidently been rebuilt over and over, even within clades that construct their body plans in more or less similar ways.

Off the A/P Axis

Hox genes are expressed all over the bilaterian body, though not anterior of the antennal segments in arthropods nor anterior of the hindbrain in vertebrates (McGinnis and Krumlauf, 1992; van der Hoeven et al., 1996; Akam, 2000). Colinear A/P patterns of expression occur in the endoderm as well as the trunk mesoderm and neuroectoderm (Grapin-Botton and Melton, 2000; Beck et al., 2000; Fig. 4.4A). But colinear expression of hox cluster genes also occurs off the A/P axis, in the limb buds of tetrapods, and as we see below, in much more remote contexts as well.

The posterior hox cluster genes are expressed in an interesting and dynamic way in developing amniote limb buds. A summary of hoxa and hoxd cluster expression patterns in chick wing and leg buds is shown in Fig. 5.4A (Nelson et al., 1996; some genes of the other clusters are utilized in limb buds as well). There are three phases of expression. In phase 1, the hoxd9 and hoxdlO genes are initially expressed throughout the bud, and in its initial distal outgrowth (Fig. 5.4A1). Following this, in phase 2, a nested pattern of expression is set up such that hoxd9, 10, 11, and 12 are all expressed at the posterior margin of the wing bud but only the more 3' of this group, i.e., hoxd9 and hoxdlO, are expressed toward the anterior side of the bud (Fig. 5.4A2). But in phase 3 a different pattern is installed in which the orientation of the phase 2 expression is completely reversed, so that the most 5' gene (hoxdl3) is now expressed toward the distal anterior margin of the bud (Fig. 5.4A3). Numerous comparative, mutational, and expression perturbation studies show that box gene expression is essential for limb development. Phase 2 expression is a component of the patterning system that sets up the developmental domains from which derive major proximal and distal structural elements of the limb (e.g., our upper arms and forearms), while phase 3 expression is associated with patterning of the terminal portion of the limb (the autopod, e.g., our hand and its digits); (reviewed by Johnson and Tabin, 1997; Nelson et al., 1996; Shubin et al., 1997; Zakany et al., 1997; Peichel et al., 1997). The argument could be made that expression of the 5' box genes at the posterior margin of the limb buds (Fig. 5.4A2) is just another case of colinear A/P box gene use. However, activation of these genes in this order in the limb bud depends on signaling from a site at the posterior margin of the limb bud. This signaling system is set in place through an independent mechanism operating completely within the growing limb bud (reviewed by Johnson and Tabin, 1997). The reversal of orientation in the phase 3 expression pattern clinches the matter in any case: here the vectoral box gene expression system is used along a transient that is not only independent of the A/P axis of the trunk, but is nearly opposite to it. The anteriorly curved secondary axis which is marked by phase 3 box gene expression serves as a mount for the secondary skeletal growths that produce the digits (Fig. 5.4B3); (Shubin et al., 1997).

The expression of boxdll in mouse and zebrafish forelimb limb buds is compared in Fig. 5.4B (Sordino et al., 1995). Early on, the same posterior pattern of expression is seen in the two species (Fig. 5.4B1, B4). But then something different occurs in the mouse limb buds: the axis of expression curves anteriorly, while the axis of the developing fish appendage (i.e., its pectoral fin) remains straight (Fig. 5.4B2-6).

Ancestral chordates did not have paired appendages such as derive from the limb buds of tetrapods. These uses of the ¿ox cluster patterning systems off the A/P body axis are obvious evolutionary cooptions, and the cladistic evidence for this is summarized in Fig. 5.4C (Shubin et al., 1997). There are no appendages on modern invertebrate chordates, e.g., larvacean urochordates, or cephalochordates such as amphioxus; nor on basal jawless vertebrates, such as the hagfish; nor on the earliest fossil vertebrate or invertebrate chordates Qanvier, 1996; Shu et al., 1999; Chen et al., 1999). The first appendages to appear in the vertebrate lineages were unpaired fore and aft median fins as at (a) in Fig. 5.4C, followed by paired pectoral fins as at (b). Only the jawed vertebrates (gnathostomes) have two sets of paired appendages (Janvier, 1996). As Fig. 5.4C indicates, a cooption of box

FIGURE 5.4 Evolutionary cooption of posterior hox genes for development of tetrapod limbs. (A) Phases of expression of posterior genes of the hoxa and hoxd clusters in chick wing and leg buds, based on in situ hybridization data. Distal is to right, and five successive developmental stages are shown in each row. In phase I hoxd9 and hoxd 10 are expressed throughout the distal region of both wing and leg buds. In phase 2 a nested set of expressions is established, oriented from posterior to anterior in a colinear manner with respect to the 5' to 3' chromosomal sequence of the genes, such that the hoxd 11 and hoxd 12 genes are expressed most posteriorly and the hoxd9 gene most anteriorly to the bud. In phase 3 a reverse order of expression appears: hoxa 13 and hoxd 13 are now expressed more anteriorly than are hoxd 10 and hoxd 11. Phase 3 is similar in wing and leg, while phase 2 is not. [(A) From Nelson et al. (1996) Development 122, 1449-1466 and The Company of Biologists, Ltd.] (B) Expression of hoxd 11 in tetrapod (mouse) forelimb bud (BI—3) and in teleost (zebrafish) pectoral fin bud (B4-6); anterior left; distal up. (BI) hoxd 11 is expressed posterior to the proximodistal axis of the mouse limb bud (dashed line between arrowheads). (B2) At day 12 expression extends across the anterior region of the bud and with reference to the proximodistal axis, the axis bounding the expression domain is now curved (dashed line). (B3) The terminal domain of expression shown in (B2) is where the autopod (hand), including the digits, will form as indicated by cartilage condensations (green stain). (B4) Expression of hoxdll in a posterior region of zebrafish fin bud. (B5) Elongating fin bud, in which hoxdll expression has decreased, but remains posterior. The expression domain continues to be bounded by the straight proximodistal axis (dashed line). (B6) The straight axis is reflected in the organization of cartilage condensations in the fin. [(B) From Sordino et al. (1995) Nature 375, 678-681, copyright Macmillan Magazines Ltd.] (C) Summary of evolutionary development of paired appendages in vertebrates; inferred states of hox cluster expression during limb development and structural organization of the autopod indicated in blue. As shown at (a), jawless fish such as the Silurian Jamoytius have median elongate fins; or, as at (b) the armored, jawless osteostracans display paired pectoral fins (janvier, 1996). Two pairs of appendages emerging from the trunk body wall appear only in the jawed vertebrates (gnathostomes); in modern gnathostomes (c) the posterior hox genes are expressed in the buds from which anterior as well as posterior appendages develop [cf. (A)]. The teleosts, represented in (B4-6) by the zebrafish, are ray finned fishes (actinopterygians), and as shown in (B6), their fin axes are straight. This character is retained by the sister group of the tetrapods, the extinct osteolepids, represented here by Eusthenopteron (d), while even primitive tetrapods such as the Devonian Icthyos-tega (e) display the curved axis seen in the mouse in (B5,6). Development of the autopod is inferred to depend on the curved phase 3 pattern of hox cluster expression (A3); the zeugopod, adjoining the autopod, denotes the major structural component of the lower limb (i.e., forearm or calf of leg), and its development depends on phase 2 expression (cf. A2). [(C) Modified slightly from Shubin et al. (1997) Nature 388, 639-648, copyright Macmillan Magazines Ltd.] (D) Expression of lacz construct driven by a zebrafish hoxdll enhancer in a transgenic mouse embryo. The arrow indicates expression in the forelimb bud. Expression occurs also in the caudal trunk region, the genital eminence, and the hindlimb bud. [(D) From Beckers et al. (1996) Dev. Biol. 180, 543-553.]

Was this article helpful?

0 0

Post a comment