10-27. Relationship of morphological differentiation to restricted expression of Hox genes. From Averof and Akam, 1995. In the uniform thoracic region of branchiopod crustaceans, the three middle Hox genes (Antp, Ubx and abdA) have largely overlapping domains of expression (see lower left). In insects (at right) Antp is restricted to the thoracic segments that develop legs and wings, whereas Ubx and abdA are expressed in abdominal segments, where they repress distal-less and presumably regulate the nondevelopment of legs.
[Page 1167]
But this reasonable conjecture was then falsified because, just as in dipterans, Ubx does not turn on in the lepidopteran forewing imaginal disc, but achieves high levels of expression in the hindwing disc. Therefore, the growth of lepidopteran hindwings in the presence of Ubx must depend upon differences in the downstream T3 cascade of flies vs. butterflies. Warren et al. (1994, p. 461) conclude: “The most logical explanation is that the sets of downstream wing-patterning genes regulated by Ubx in these orders have diverged. In this view, Ubx operates in butterflies upon pattern regulating genes to differentiate hindwings from forewings, and in flies upon a different set of genes to distinguish halteres from wings.”
In further confirmation from detailed patterns at lower taxonomic levels, Weatherbee et al. (1999) then studied Hindsight, a homeotic mutation in butterflies that transforms parts of the hindwing into forewing identity. They found that these hindwing transformations in color and scale morphology occur in regions of the forewing where Ubx expression has been lost, thus sensibly explaining, under the general rule for Hox expression in butterfly wings, the apparent forewing identities of these altered regions. Reemphasizing the important principle previously illustrated for echinoderms vs. other triploblast phyla, but at this lower taxonomic level — that channels of internal homology also promote flexibility, not just limitation, through such mechanisms as cooptation and diversification of downstream cascades — Weather-bee et al. (1999, p. 113) write: “The diversity of insect hindwing patterns illustrates the broad range of possible morphologies that can evolve in homologous structures that are regulated by the same Hox gene.”
I turn, finally, to the two canonical and most anatomically extensive examples of evolution from homonomy to regional specialization and complexity in the evolution of insects and other arthropods — evolution from the plesiomorphic state of walking legs on all post-oral segments and, for pterygotes, from the ancestral condition (as revealed in the fossil record and preserved in modern mayfly larvae) of wings on all thoracic and abdominal segments. Data from evo-devo have effectively resolved the old debate about whether insect wings evolved as novel structures from hypothesized rigid extensions of the body wall in terrestrial forebears (the paranotal theory), or from dorsal branches of polyramous appendages of ancestral forms (the limb-exite theory).
Genetic data support the exite theory and provide a fascinating example of cooptation in evolution (the general subject of the subsequent Chapter 11). In the exite theory, insect wings and legs are, in some sense, serially homologous as specializations of different parts of an ancestral polyramous appendage — the wings from the dorsalmost branch (the exite) and the leg from the ventralmost-walking branch. In their major topological difference, wings develop as sheets, and legs as tubes. In Drosophila, the wing grows under the crucial influence of apterous, which is expressed only in dorsal cells and therefore maintains clear distinction between dorsal and ventral surfaces, thus abetting the growth of a sheet-like structure (Shubin et al., 1997). But the apterous gene does not function in the growth of tubular legs in Drosophila. However, [Page 1168] Averof and Cohen (1997) found apterous expression in the sheet-like dorsal branch of respiratory epipodites in a branchiopod crustacean, thus supporting the exite theory for the origin of wings. Shubin et al. (1997, p. 645) draw a reasonable phyletic conclusion consistent with this section's theme of evolution from homonomy to specialization: “This suggests that Recent wings evolved from the respiratory lobe of an ancestral polyramous limb, probably first appearing in the immature aquatic stages as gill-like structures, such as those found on all trunk segments of extinct Paleodictyoptera or extant mayfly larvae.”
On the related issue of evolutionary suppression of wings on most segments and their restriction to two pairs in most insects, and to one in dipterans (see Fig. 10-28), Carroll et al. (1995, p. 58) demonstrate that “wing formation is not promoted by any homeotic gene, but is repressed in different segments by different homeotic genes.” Against the older view (consistent with Lewis's original additive model of phenotypic complexification) that Antp positively regulates the formation of wings and halteres on T2 and T3 of Drosophila, Carroll et al. (1995) present evidence that other Hox genes repress wing primordia on the remaining body segments. For example, Scr is expressed in both labial and Tl segments; in mutant embryos lacking Scr expression, “flight appendage primordia arise in the Tl segment” (p. 58). As for suppression of more posterior wings, the oldest information about homeotic mutations in Drosophila documented the development of complete wings on T3, where the vestigial halteres usually form. We now know that this bithorax phenotype (which gave its name to the previous designation of the posterior Drosophila Hox series as the “bithorax complex” or BX-C) results from a mutation that represses the Ubx gene in T3. Another mutation of Ubx leads to the growth of wing primordia on Al as well.
Carroll et al. (1995) propose that when wings existed on all post-oral body segments of a homonomous ancestor, “there was no homeotic gene input into their number or design” (p. 59). Carroll et al. then hypothesize that elimination of wings from most segments occurred as the Hox genes became
10-28. Differentiation of flight in the evolution of insects as a consequence of repression of wings on posterior segments by various Hox genes. The fossil nymph in B possessed wings on all segments. These are reduced but still present on all segments in the fossil mayfly nymph at C. From Carroll et al., 1995.
[Page 1169]
regionalized and individualized, leading to suppression by different elements in various parts of the body: “The evolution of Scr-responsive elements led to the modification or elimination of prothoracic wings and the evolution of abdA and Ubx-responsive elements led to the elimination of abdominal wings and, in the Diptera, to the reduction of metathoracic wings.”
For the most general question of specialization of appendages for a wide variety of forms and functions from their uniform state on all segments posterior to the head of a homonomous ancestor, several lines of evidence identify walking appendages (either uniramous, or biramous with an upper gill branch) as the ancestral state for a homonomous ancestor, and as a continuing “ground state” for modern more differentiated forms as well. First, the most homonomous modern groups — the Myriapoda among the arthropods, and the Onychophora as a sister group to the entire arthropod phylum — bear leg-like structures on each segment. Second, numerous Cambrian arthropods that cannot be placed into modern groups share the common property of nearly identical biramous appendages on all postoral segments, and only a pair or two of antennae on any preoral segments — as in Marrella, the most common fossil in the Burgess Shale (see Fig. 10-29 and Gould, 1989c). Third, as discussed previously (pp. 1132–1134), the extensive suite of thoracic segments that bear identical leg-like appendages in many modern Crustacea also show extensive and complete overlap of expression for several Hox genes.
Proceeding down the AP axis of complexified arthropods, we first note that antennae develop in the most anterior segments where no Hox expression occurs. This situation probably marks retention of the ancestral condition. Even the most homonomous forms, including myriapods and onychophores, exhibit some specialization in the head segments at the anterior end, and only grow identical appendages on subsequent postoral segments. Thus, the original Hox complex probably never regulated development at the extreme anterior end around the mouth, and antennae probably represent the plesiomorphic condition for se
gments with no Hox action. Interestingly, and in confirmation, the suppression of all Hox activity in Tribolium yields the lethal
10-29. Drawing of Marrella from Gould, 1989c, to show homonomy of nearly identical biramous appendages on all postoral segments. Drawing by Marianne Collins.
[Page 1170]
consequence of a dead larva with antennae on all segments (Stuart et al., 1991; Shubin et al., 1997, p. 664; see also Cassares and Mann, 1998, on antennal-determining genes repressed by Hox action in Drosophila).
All other specializations down the AP axis are apparently derived and dependent upon differentiation and regionalization, or elimination in some cases, of expression in various Hox genes. For example, in gnathal segments just posterior to segments bearing antennae in many groups, the most homonomous modern forms develop mouthparts of essentially leg-like form (as in myriapods). These leg-like appendages express Distal-less at their distal tips, the typical situation for ordinary arthropod legs. But Distal-less is not expressed at the distal ends of more specialized (and non leg-like) feeding appendages of insects and crustaceans. “These data,” Shubin et al. write (1997, p. 644), “agree with fossil evidence suggesting that crustacean and insect mandibles were reduced from the primitive whole-limb mandible by truncation of the mandibular proximodistal axis.”
I have already discussed, in previous parts of this section, the role of Hox restrictions and repressions in the evolution of all other outstanding phenotypic specializations in more posterior regions of arthropod bodies, including the differentiation of maxillipeds from legs on the previously homonomous crustacean thorax (pp. 1132–1134), the restriction of wings to just one or two thoracic segments in insects (p. 1165), and the complete suppression of legs on the insect abdomen, with localized Hox repression to permit the growth of prolegs on the abdominal segments of lepidopteran larvae (p. 1165).
When we turn to the history of vertebrates, we first encounter an apparent exception to the generality that phenotypic specialization correlates with reduction in number of Hox genes and regionalization of their action. Amphioxus, the modern cephalochordate surrogate for an ancestral form, has only one Hox cluster, while gnathostome vertebrates have four — so duplication, occurring at least twice, clearly marks a major feature of vertebrate evolution, with obvious implications for correlating the complexity of our phylum with this marked increase in the total number of Hox genes, and in apparent contradiction to the opposite relation of phenotypic elaboration with genetic restriction, as discussed throughout this section.
But the single cluster of amphioxus contains homologs of the first 10-paralogy groups of vertebrate Hox genes, arranged in the usual colinear array. Moreover, the amphioxus genome includes at least two AbdB-like genes, indicating that tandem duplication of these posterior Hox elements was already underway in the cephalochordates, even though true vertebrates have carried the process further (Carroll, 1995; Coates and Cohn, 1998). Therefore, essentially the full Hox complement had already been established when the genome of an immediate vertebrate ancestor included only one set of Hox genes. Moreover, the full fourfold amplification had already been completed by the origin of jaws in early fishes because all modern gnathostomes — that is, all living species of vertebrates except for the two small lineages of agnathan fishes, the lampreys, with three Hox sets, and the hagfishes — have four sets. [Page 1171]
Thus, the common ancestor of all 40,000 or so modern gnathostome species already had four Hox sets, and only the handful of agnathan species has fewer sets among modern vertebrates. Thus, our large clade of 40,000 species evolved under the general rule featured throughout this section: phenotypic specialization correlated with Hox deletions” and restrictions of expression. Or, to put the matter somewhat facetiously, you start with all you will ever get, and work “down” from there — an optimal formula for the evolutionary importance of historical constraint.
As Coates and Cohn (1998, p. 375) write (see also Fig. 10-30): “During the period since these gene duplication episodes, jawed vertebrate Hox cluster evolution seems to have been characterized by gene deletions.” Moreover, as Figure 10-30 also shows, teleost fishes, which did not originate until Mesozoic times, evolve different patterns of deletion from those found in mammals, a group with a Paleozoic ancestry from a very different vertebrate lineage — thus “indicating quite separate patterns of gene loss in tetrapod and teleost lineages” (Coates and Cohn, 1998, p. 375).
The relatively homonomous architecture of the postcranial skeleton of many early fishes (and many early tetrapods as well) has evolved in the conventionally “higher” tetrapods, primarily in mammals, into a more complex, specialized and regionalized axial skeleton with clear and often quite sharp distinctions, in both form and function, from cervical to thoracic to lumbar, sacral and caudal regions of the vertebral column. Burke et al. (1995) have demonstrated an interesting basis for much of this phenotypic complexity and regionalization in the establishment of definite boundaries of action for particular paralogy groups of the Hox clusters, thus repeating the general arthropod correlation of Hox regionalization with phenotypic specialization along the AP axis.
For example, different groups of vertebrates vary greatly in the number of vertebrae per region, but the boundary between regions may still remain
10-30. Following the evolution of four Hox clusters in vertebrates, the major pattern of change has not resulted in further addition, but rather in elimination — in different patterns in various groups. From Coates and Cohn, 1998, p. 375.
[Page 1172]
sharp (Burke et al. refer to the shifts of these boundaries to emplace more or fewer vertebrae within any given region, as “transpositions”). In their most intriguing conclusion, Burke et al. (1995) found that some of these phenotypic transitions correlate precisely with anterior expression boundaries of particular Hox genes. For example, Hoxc-6 marks the transition between cervical and thoracic vertebrae, despite the highly variable number of cervical vertebrae, ranging from 3 or 4 in frogs, to 7 in mice (as in virtually all mammals, including giraffes), to 17 in geese. The thoracic-lumbar transition generally correlates with the expression of Hoxa-9, Hoxb-9, and Hoxc-9, whereas the Hoxd-9 boundary tends to be shifted backwards to the lumbosacral transition. Carroll (1995, p. 483) comments on these differences in the ninth paralogy group: “This may be significant because the thoracic-lumbar distinction is not general among tetrapods. It may be that shifts within the Hox-9 group were important in the evolution of this transition from a more uniform trunk, perhaps even in the evolution of the tetrapods from fish.”
These regularities of Hox regionalization may help us to understand both the limitations and flexibilities of vertebrate anatomy in terms of historical constraint. In an early article, for example, Tabin (1992) suggested that tetra-pod limbs may now be constrained to five digits per limb (despite the presence of up to 8 digits in the earliest tetrapods of the Late Devonian Period — see Coates and Clack, 1990; and Gould, 1993e) because the Hoxd series that plays such a major role in patterning limbs may now only generate five “addresses” for the development of distinct digits. Many polydactylous mutants (and experimental manipulations) exist in vertebrates, but the supernumeraries are always phenotypic replicates of one of the five distinct digits, so the general hypothesis holds (see also Shubin et al., 1997, pp. 642-643).
A related classical question asks why tetrapods, honoring their name, never grow more than four limbs, whereas the other major terrestrial group of arthropods usually evolves phenotypes with more appendages (even though we might imagine, on functional grounds, that an increased number of supports would be even more valuable in large vertebrates with much lower area to volume ratios — for supporting strength of bone scales as cross-sectional area). Coates and Cohn (1998, p. 379) note that “the nearest approach to a third pair of lateral appendages may be the lateral caudal keels of certain fishes, such as tuna and various sharks.” But a true anatomical third pair has
never evolved in any tetrapod or extant fish. (The extinct acanthodian fishes evolved the only vertebrate departure from the principle of two primary limb pairs.)
But as Hox rules constrain, they can also be tweaked to win interesting flexibility. Cohn and Tickle (1999), for example, studied Hox expression in the axial skeleton of pythons, which can grow more than 300 essentially identical vertebrae, and which retain hindlimb rudiments but express no forelimb development at all. Except for the atlas, every vertebra anterior to the rudimentary hindlimb develops ribs (a thoracic feature) as well as ventral hypopophyses (a cervical feature), suggesting to Cohn and Tickle (1999, p. 474) that “information encoding thoracic identity may have extended into [Page 1173] the cervical region and partially transformed these segments. Thus the entire trunk resembles an elongated thorax.”
Cohn and Tickle studied the expression of Hoxb-5, Hoxc-6, and Hoxc-8 in the ontogeny of pythons. In both teleosts and tetrapods, the anterior expression boundaries of all three genes in lateral plate mesoderm occurs “at the forelimb/pectoral fin level, where they are involved in specifying forelimb position and shoulder development” (p. 475). But pythons develop no phenotypic expression of the forelimb at all, thus suggesting that a suppression of this positional boundary, and a forward expansion of expression in these genes, might be causally related to the vast increase in number, and identity in thoracic form, of snake vertebrae — and potentially helping to explain one of the most striking functional novelties ever evolved in vertebrates.
The Structure of Evolutionary Theory Page 186