Functional Evolution of
Embryonic Axis Determinants
We study the diversity of embryonic axis determinants in dipterans as a model for the origin and functional evolution of molecular innovations in early development. We found that global pattern organizer genes have changed frequently during the evolution of true flies and involve structurally distinct genes (Fig. 1). The reason for this evolutionary lability and the evolutionary paths that allow transitions between the functionally diverse mechanisms of axis specification remain unknown.
Fig. 1. Occurence of bicoid in Diptera. The phylogeny is based on Wiegmann et. al., 2011, PNAS 108: 5690-5695, for the occurence of bicoid see Lemke et al., 2010, Development 137: 1709-1719 and references therein. Indicated instances of missing bicoid orthologs (black) are based on genome sequences.
One specific research goal of this project is to understand the evolution and mechanism of action of a novel cysteine-clamp (C-clamp) gene that we named panish. Panish binds DNA and functions as a maternal anterior determinant in the mosquito model Chironomus riparius, similar to bicoid in Drosophila. The Drosophila gene bicoid is an important model for a variety of fundamental concepts in developmental and molecular biology, including long-range patterning, transcriptional and translational gene regulation, RNA transport, and biophysical aspects of protein gradients. Chironomus panish provides an independent experimental system to study bicoid-like morphogens and can inform the generality of the Bicoid model. Panish also provides an interesting new model for studying the interactions of cysteine-clamp (C-clamp) proteins with DNA because it is small (131 amino acids) and lacks (apart from the C-clamp) any other conserved domains. C-clamp proteins have been implicated in a variety of human diseases but their interaction with DNA remains poorly understood given that structural data are currently not available.
The bicoid paradigm of Drosophila:
The bicoid transcript is localized at the anterior pole of the Drosophila egg. Certain motifs of the bicoid mRNA 3’ untranslated region (UTR) are both necessary and sufficient for its proper localization. Following translation of the localized bicoid mRNA in the early embryo, the Bicoid protein diffuses to form a long-range concentration gradient. How the Bicoid gradient forms is the subject of many recent studies, as its precision directly affects the robustness of the early patterning process.
Bicoid is a homeodomain protein that functions as a transcriptional activator and in one instance (Caudal) as a translational repressor. Loss-of-function mutants of bicoid result in embryos with a second tail end in place of the head, thorax and parts of the abdomen. The bicoid mutant phenotype is asymmetric because of the gradient of a second protein, Hunchback, which forms in response to the posterior enriched translational regulator Nanos. Embryos lacking the maternal transcript of both bicoid and hunchback develop as symmetrical double abdomens. The same phenotype can be obtained by expressing nanos ectopically at the anterior pole of the egg to repress translation of bicoid and hunchback mRNA. However, ectopic expression of bicoid at the posterior pole overrides nanos activity and results in symmetrical double cephalons (embryos consisting of duplicated anterior head segments with opposite polarity). In the early embryo, bicoid is therefore both necessary and sufficient for anterior development.
Bicoid target genes are also regulated by the terminal system, a receptor tyrosine kinase (RTK) pathway that functions at both poles of the early embryo through the transmembrane receptor Torso. The Torso pathway counteracts the ubiquitous transcriptional repressor Capicua by phosphorylating it along with other repressors such as Runt and Krüppel, which are important for establishing the posterior expression boundaries of Bicoid target genes. Torso also phosphorylates Bicoid and Hunchback. Their presence in the anterior embryo causes a reduction of Capicua inactivation by phosphorylation (according to the model by substrate competition), thereby promoting the expression of Bicoid target genes that are repressed by Capicua. However, overexpression of bicoid can compensate for suppressed Torso pathway activity in the anterior embryo.
Identification of panish in Chironomus:
We identified panish by RNA sequencing of bisected Chironomus eggs. Expression profiling revealed a single panish transcript, significantly enriched in the anterior half (Fig. 4). The localization of panish transcript at the anterior pole of early embryos was confirmed by RNA in situ hybridization. In older embryos (after the beginning of blastoderm cellularization) panish transcript was not detected.
The panish transcript includes a predicted open reading frame (ORF) of 131 aa with a C-clamp DNA binding domain (29 aa). The C-clamp of Panish is most similar to the C-clamp of the Wnt signaling effector Pangolin/Tcf, Chironomus pangolin, but Panish does not possess the other conserved domains of Pangolin/Tcf (e.g. HMG-domain, ß-catenin-binding domain). Furthermore, the first two exons of panish (5’ UTR and 26 codons) upstream of the C-clamp domain are shared with a Chironomus homologue of the nucleoside kinase gene ZAP3, as revealed by transcriptome data and genomic mapping. Therefore, we refer to panish as a novel gene.
Developmental role of panish:
To test whether panish transcript is necessary for the anterior-posterior axis during embryonic development, we conducted a series of loss- and gain-of-function experiments. When early Chironomus embryos (before two-pole-cell stage) were injected with double-stranded RNA (dsRNA) against the panish ORF or 3’UTR to achieve knock down by RNA interference (RNAi), nearly all developed as mirror-image double-abdomens with similar survival rates between panish RNAi and controls. Cri-zap3 RNAi did not cause any obvious cuticle defects. RNAi against panish at the later blastoderm cellularization stage also resulted in embryos indistinguishable from wild-type controls, indicating that panish transcript is dispensable at later stages.
In order to confirm the requirement for panish mRNA in establishing the anterior domain, we performed rescue experiments by co-injecting either wild-type or out-of-frame mutated panish mRNA in combination with panish 3’UTR dsRNA. The formation of double-abdomens was suppressed in over 40 percent of the embryos with the injection of wild-type panish mRNA into the anterior third of the embryo compared to injection of mutated panish mRNA, injection buffer, bicoid mRNA and panish mRNA injection into the posterior third of the embryos. Injection of panish mRNA in the presence of endogenous panish activity also did not induce double-heads. These observations signify that panish controls head-to-tail polarity and that its activity is constrained to the anterior embryo.
We have built a new research program around the panish discovery. Our research addresses developmental and genetic questions, evolutionary questions, and biochemical questions. Some of the questions we started to address are:
(1) When and how did panish evolve?
(2) Which DNA-motifs does Panish bind?
(3) Does Panish function with (transcriptional) co-factors?
(4) Which target genes does Panish regulate at the transcriptional level and how does it interact with other components of the axis specification network?