«Joshua Frederick Coulcher UCL Submitted for the Degree of Doctor of Philosophy September 2011 Declaration I, Joshua Frederick Coulcher, confirm that ...»
Accepting all these similarities of gene expression as evidence of serial homology, the leg coxa would be serially homologous to the maxillary stipes, the mandibular coxa and to the labial prementum. The leg subcoxa would be serially homologous to the maxillary cardo, the mandibular subcoxa and the labial postmentum (fig.7.1 and fig.7.2).
However, there are some subtle differences between the proximal domains of Tc dac of the leg appendages compared to the gnathal appendages. These differences may reflect fundamentally different patterning mechanisms of these coxal segments, or may result from the different morphology of the limbs based on the presence of endites in the gnathal appendages.
There are in fact three domains of expression of Tc dac in the developing leg (Prpic et al., 2001), a distal domain in the femur, a larger proximal domain which overlaps with the coxa domain of Tc ser, and another smaller spot of expression which is co-expressed with the subcoxa domain of Tc ser (data not shown). The proximal domain of Tc dac expression in the leg is co-expressed with the coxa domain of ser expression. This is in contrast to Tc dac expression in the gnathal appendages where the proximal expression domain is expressed in between the coxa and subcoxa domains of Tc ser expression domains in the developing coxa.
Does the Tc ser subcoxal ring domain of the maxilla correlate to the cardo segment boundary?
The evidence for the homology between the mandible subcoxa and coxa to the maxilla subcoxa and coxa has been presented above. One question that remains to be answered is whether the developing cardo-stipes segment boundary is the subcoxal-1 domain of Tc ser expression.
As I had trouble to locate unambiguously the precise position of the developing cardo-stipes segment boundary, the subcoxal-1 domain of the maxilla may not relate at all to the position of the cardo. Instead the proximal-most domain-0 of Tc ser may represent the future cardo segment (see fig.3.5L and star in fig.3.7A,C). This domain is significantly different to the subcoxal-1 domain in terms of timing of expression and relative expression of the PD domain genes and Tc prd as an endite marker.
The subcoxal-1 domain may be involved in patterning some aspect of the proximal gnathal appendages, for example, by providing positional information to pattern the endites. The gnathal appendage endites are all present in the coxa immediately distal to the subcoxal-1 domain of Tc ser (see fig.3.4D).
It was not possible to perform in situ hybridizations with Tc ser, or to obtain SEM images of the developing maxilla on very late stage embryos, later than those shown in the figures above. These might have confirmed the position of the division of the maxilla protopodite into cardo and stipes segments. As a result, there is uncertainty as to the position of the cardo-stipes boundary on the developing maxilla.
It is plausible that the subcoxa ring of Tc ser expression relates to the future cardostipes boundary, and certainly marks the position of some sort of boundary in the developing maxilla (see fig.3.7E,F).
It seems just as reasonable to hypothesize that the proximal stripe of Tc ser (proximal-0 Tc ser domain) that appears later in maxilla development than the subcoxa ring domain could relate to the position of the future cardo and stipes division. In which case, the subcoxal-1 Tc ser domain in the maxilla would relate to some other aspect of development than the development of the cardo/stipes segment boundary.
The subcoxal-1 Tc ser domain could reflect the evolutionary history of the maxillary appendage as a vestigial segmental expression domain, or it could be involved in some other aspect of appendage development, such as the development of the endites.
In order to determine whether the subcoxa and coxa are serially homologous between the mandible and maxilla, and also to the labial and leg appendages, it is necessary to compare the expression of the PD domain genes relative to the Notch signalling pathway in other arthropod taxa. Evidence from one taxon, Tribolium, of similar expression patterns in different appendages is not significant enough to provide strong support for particular serial homology hypotheses. By relating the expression of ser to particular leg segments and to PD domain gene expression it is possible to define the precise segmental expression the PD domain genes. With this knowledge it will be much easier to compare segments between different appendages and between diverse taxa without requiring functional genetics (which is a major constraint in evolutionary developmental studies). Study of the expression of PD domain genes and the Notch signalling pathway in biramous limbs could also be informative, by relating the molecular development of the PD axis to the morphological definitions of the protopodite (defined as where the two branches of the biramous limb, the telopodite and exopodite, attach). In particular, the coxal-2 domain of Tc ser expression, considering its early activation in all appendages with telopodites and its proximity to the likely boundary between the protopodite and telopodite may relate to the protopodite/telopodite boundary. If this is the case, then this ser domain may be present in the distal segment of the protopodite in other appendage types like biramous limbs.
If relationships between the Notch signalling pathway and the PD domain gene expression reveal further similarities between the subcoxa and coxa of the mandible to the maxilla, it would support serial homology of the subcoxa of the mandible and maxilla, and maybe to the leg appendages. The plausibility of a plesiomorphic three segmented protopodite could then be seriously evaluated.
4.1 Introduction In Drosophila, cap’n’collar (cnc) has been shown to pattern the mandibular segment (Mohler et al., 1995). cnc differentiates the mandibular segment from the maxillary segment by repressing the Hox gene Dfd. In this chapter, the function of the Tribolium homologue of cnc was explored by knocking it down by parental RNAi to determine whether it has a conserved role in patterning the mandibular segment of Tribolium.
The arthropod mandible is an appendage generally adapted for biting and chewing. All post-antennal arthropod appendages have evolved from a biramous limb (Boxshall, 2004; Waloszek et al., 2007; Chen, 2009). The biting/chewing structure that forms the gnathal edge of the mandible is present on the base (the protopodite) of this ancestral biramous limb.
The mandible has most likely evolved from a biramous maxilla-like precursor by modification of the proximal endite to form the gnathal edge. Such a maxilla like precursor is present in numerous stem lineage arthropod fossils such as Martinssonia elongata (Muller and Waloszek, 1986) and Phospatocopida (Siveter et al., 2001;
Edgecombe, 2010; Rota-Stabelli et al., 2011). Numerous variations to the structure of the ancestral mandible have occurred, such as loss of the mandibular palps. However the defining characteristic of the mandibular appendage shared by mandibulate arthropods is the presence of this gnathal edge, consisting of a molar process and incisor process, on the protopodite. This gnathal edge is widely considered to be a synapomorphy of the clade Mandibulata (Kraus, 2001; Edgecombe et al., 2003;
Arthropod phylogeny and the evolution of the mandible
The mandible is present in two extant arthopod lineages, Pancrustacea (insects and crustaceans) and Myriapoda (millipedes and centipedes). Insects and myriapods were previously grouped together in the Atelocerata (also known as Tracheata). As a result of robust molecular phylogenetic evidence, together with a suite of morphological characters, insects are now grouped with crustaceans to form the Pancrustacea.
The position of the Myriapoda is less clear. Two competing hypotheses are favoured by different molecular phylogenies: placing the myriapods as sister group to the Pancrustacea to form Mandibulata, or placing the myriapods as a sister group to the Chelicerata to form Myriochelata. Morphological evidence strongly favours Mandibulata over Myriochelata. More recent molecular phylogenies that have larger datasets, include evidence from rare genomic changes and take more care over outgroup choice favour Mandibulata (see chapter one for literature reviews).
These two competing phylogenetic hypotheses have different implications regarding the evolution of the mandible. The most commonly suggested path of mandible evolution is of a single origin in the ancestor to the myriapods, crustaceans and insects (Snodgrass, 1938; Edgecombe, 2010). This hypothesis is compatible with the clade Mandibulata, not surprisingly as the name Mandibulata suggests. There is also the formal possibility, considered unlikely, that the mandible has evolved independently more than once within Mandibulata.
If myriapods were grouped with the chelicerates in Myriochelata it would suggest that the mandible had evolved independently in the Myriapoda and Pancrustacea (Mayer and Whitington, 2009). The homologous segment of the chelicerates to the mandibular segment is the first leg segment (L1). The ancestral leg appendage of the Chelicerata on the L1 segment was almost certainly a biramous limb.
The ancestor to Myriochelata would have therefore possessed a leg appendage on the mandibular/L1 segment. A single evolutionary origin of the mandible appears extremely unlikely under the Myriochelata hypothesis. As sister group to the mandibulate myriapods, the chelicerate first leg appendage would have had to evolve from a mandible.
If the mandible is a homologous appendage across Mandibulata it would suggest that there may be significant similarities in the embryonic development of the mandible and the mandibular segment between diverse lineages of this clade. By studying the genes involved in patterning the mandible of Tribolium, a mandibulate insect, and by comparing these genes to other mandibulate taxa, it may be possible to demonstrate the homology of the mandible across mandibulates and inform the phylogenetic dispute regarding the placement of the Myriapoda. In addition, the manner in which the mandible is patterned may show evidence for the evolution of the mandible from a maxilla-like precursor.
Mandible segment patterning genes in Drosophila
As a preliminary step, candidate mandibular segment patterning genes were chosen from Drosophila to be considered for research in Tribolium. Research on Drosophila melanogaster has revealed several segment patterning genes, cap’n’collar (cnc), Deformed (Dfd) and apontic (apt), to be required for patterning the mandibular segment. There is however, no known homologue of apt in Tribolium and therefore effort was concentrated on studying the function of cnc and Dfd.
Hox genes are master regulatory genes that give segments their individual identity. Hox genes activate downstream targets that define and pattern segments.
One Hox gene, Dfd, is expressed in the mandibular segment of all mandibulate arthropods. In Drosophila Dfd is expressed in the mandibular and maxillary segments during embryogenesis. Drosophila that are mutant for Dfd are missing maxillary and mandibular derived structures (Merrill et al., 1987; Regulski et al., 1987; McGinnis et al., 1998; Brown et al., 1999; Veraksa et al., 2000).
cnc function in Drosophila
Dfd is responsible for patterning both the mandibular and maxillary segments, but does not differentiate the mandibular segment from the maxillary segment. For this function another gene, cnc, is involved. cnc, a basic Leucine zipper family gene (bZIP), is necessary for the development of labral and mandibular derived structures and achieves this in part by repressing the maxilla patterning function of Dfd. cnc is expressed in an anterior ‘cap’ domain in the labrum and a posterior ‘collar’ domain in the mandibular segment. cnc null mutants lose both labral and mandibular segment derived structures and have a duplication of maxilla derived structures (Mohler et al., 1995; McGinnis et al., 1998; Veraksa et al., 2000).