Origin of the Arthropod Mandible

Origin of the arthropod mandible

SIR - Arthropods, vast in number and for investigating the structure of the with enormous variation in body forms, mandibles (whole limb versus limb base are a fascinating group. We have found only). In the millipede Oxidus gracilis (a in that myriapods (millipedes, centipedes the figure), Dll is expressed in the distal and allies) have different mandibular ori part of the mandibles, as predicted in ref. gins from insects and crustaceans, which is 4, indicating their whole-limb structure. of consequence for resolving phylogenetic In light of the recent discovery of the relationships among major groups of Cambrian fossil whose head and trunk arthropods. appendages were long and Ieg-like8, the For more than a century, the phylo whole-limb mandibles of today's myriapods genetic relationships among the main probably represent an ancestral arthropod arthropod lineages have been a topic of state. Thus, we have a testable hypothesis: lively discussion. Almost every imaginable if myriapods and insects are indeed sister combination has been proposed, but at taxa, then Dll should also be expressed in present only two hypotheses are seriously insect mandibles. But Panganiban et al. 9 considered: the 'TCC' view, which sepa have shown that Dll is not expressed in rates trilobites, crustaceans and cheliccr mandibles of modem insects. To examine ates from the rest of the arthropods1, and whether the absence of Dll is characteristic the 'mandibulate' theory, which groups of the whole insect lineage, we included in together crustaceans, insects and myri our analysis the primitively wingless insect apods2. One feature is common to both: Thennobia domestica, and found that Dll is the close relationship between myriapods not expressed in the mandibles of this and insects. These two arthropod groups species (b in the figure); this further sug were traditionally united into Atelo gests that insect mandibles are formed cerata3, because they share five adult from the limb base and may be similar to characteristics: a tracheal system, mal the mandibles of adult crustaceans. pighian tubules, absence of appendages Both 0. gracilis and T. domestica under corresponding to the second antennae of go direct development, where immature crustaceans, unbranched legs, and a stages differ from the adults mainly in the mandible (jaw) composed of a whole limb. development of the gonads and genitalia. The existence of the Atelocerata has This allows us to correlate directly embry recently been questioned by two indepen onic changes in Dll expression with struc dent studies reporting molecular data to tural changes in adult mandibles. In con infer arthropod phylogeny4•5• Friedrich and trast, only crustaceans that undergo larval Tautz4 suggested that crustaceans, and not development have been studied so far6• Expression pattern of DI/ in the mandibular seg• myriapods, are the sister group of insects, Consequently, the finding that Dll is ments of the mi llipede Oxidus gracilis (a), the arguing that the first three characteristics expressed throughout the mandibles of primitively wingless insect Thermobia domestica common to both myriapods and insects are crustacean nauplius larvae (c in the figure) (b). the crustacean nauplius larva Artemia francis• convergent adaptations to terrestrial life is not informative because the larval cells cana (c) and the terrestrial isopod Armadillidium vu/gare (d) . An, antenna! segment; Mn, man• and thus do not reflect a common ancestry. expressing Dll do not contribute to the dibular segment; Mx, maxillary segment. Amm• Panganiban et al. 6 have shown also that adult structures (as noted in ref. 6). To infer heads indicate the mandibular appendages. differences between branched and the origins of the mandibles in adult crus unbranched legs can be caused by a simple taceans, it is necessary to study species that it directly supports the hypothesis that developmental switch. In the light of these undergo direct development. We therefore crustaceans, not myriapods, are the arguments, it is essential to evaluate criti included the terrestrial isopod Annadillidi sister group of insects4. The overall simi cally the fifth feature shared by myriapods um vulgare, a direct developer, in our larity of nervous and visual systems in and insects, namely, a similarity in the analysis. We found that crustacean both insects and crustaceans provides structure of their mandibles. mandibles are indeed composed of a limb independent support for this hypothesis 10. Traditionally, crustacean mandibles are base only, as is evident by the lack of Dll These findings lessen the case for uniting regarded as being formed from a limb expression in ectoderm (din the figure). insects and myriapods into Atelocerata. base, and ateloceratean mandibles as In summary, our data are consistent Aleksandar Popadlc, Douglas Rusch being composed of a whole limb 7, but this with earlier predictions2·4·6 that the arthro Michael Peterson, Bryan T. Rogers view has been challenged recently2. One pod mandible was originally composed of Thomas C. Kaufman way to resolve this dispute is to investigate a whole limb and was similar to the pre Howard Hughes Medical Institute, the structure of arthropod mandibles at sent-day mandibles of the myriapods. Department of Biology, the molecular level, as suggested by Further, our data suggest that during Indiana University, Friedrich and Tautz4 • arthropod evolution, the mandible struc Bloomington, Indiana 4 7405, USA We studied the pattern of expression of ture changed from a whole limb (a in the 1. Cisne, J. L. Science 186, 13-18 (1974). the homeobox gene Distal-less (Dll) in the figure) to a limb base only, the latter type 2. Kukalova-Peck, J. Can. J. Z oo/. 70, 236-255 (1992). 3. Te~ord, M. J. & Thomas, R. H. Nature 376, 123-124 (1995). mandibles of the millipede, and compared being a shared feature between insects 4. Friedrich, M. & Tautz. D. Nature 376, 165-167 ( 1995). it with results from insects and crus and crustaceans (b , d). This finding has 5. Boore, J. L. et al. Nature 376, 163-165 (1995). 6. Panganiban, G. et al. Science 270, 1363- 1366 (1995). taceans (Dll antibody was kindly provided two important implications. First, it 7. Manton, S. M. TheArthropoda: Habits, Functional by G. Panganiban). This gene specifies the argues against the traditional view that Morphology and Evolution (Clarendon, Oxford, 1977). 8. Robi son, R. A. Nature 343, 163- 164 (1990). distal part of appendages, and therefore insect and crustacean mandibles are fun 9. Panganiban, G. et al. Curr. Biol. 4, 671- 675 (1994). can be used as a molecular marker damentally different7; and second, 10. Osorio, D. et al. Trends Ecol. Evol. 10, 449- 454 (1995). NATURE · VOL 380 · 4 APRIL 1996 395