The goal of our research is to further our understanding of the molecular mechanisms that underlie axon guidance and synapse formation. Towards this end, we employ molecular, genetic and electrophysiological approaches in the model organism, Drosophila melanogaster.
We are pursuing two major specific aims:
We have identified Wnt5 as part of a novel pathway controlling axonal trajectories as they cross the ventral midline of the central nervous system. Furthermore, we have found that the Src Kinase Family member, src64B, is required for Wnt5-dependent signaling through the Derailed RYK receptor, thus identifying one of the first downstream effectors of this newly uncovered pathway. In collaborative studies, we have established that Wnt5 plays important roles in a) axon extension in the brain, b) sculpting the glomeruli, the region of the brain where olfactory neurons form topographically-mapped synapses with the interneurons which relay the sensory input to the central brain and c) in establishing the wildtype morphology and electrophysiology of the neuromuscular junction.
We have begun to dissect the roles of the isoforms encoded by the Drosophila dystrophin gene. Mutations in the human dystrophin gene underlie Duchennes Muscular Dystrophy (DMD), an unfortunately frequent and lethal genetic disease characterized by muscle wasting and mental retardation. We have shown that a postsynaptic large isoform of the Dystrophin protein, DLP2, plays an important role in the retrograde control of neurotransmitter release at the neuromuscular junction (NMJ). In an ongoing study, we have found that the central nervous system-specific Dystrophin isoform, Dp186, likely plays similar roles at interneuronal synapses in the brain. Finally, through use of transgenic RNA interference, we have provided evidence that a relatively uncharacterized short Dystrophin isoform, Dp117, is required to maintain the stability of the musculature.
The identification of dystrophin as the gene whose mutation underlies Duchennes Muscular Dystrophy (DMD) in humans almost two decades ago led to intense study of this large, complex gene, ultimately with the hope of developing a cure for the disease. DMD is characterized primarily by muscle wasting however, it has become clear that there are significant mental deficits in a subset of patients, suggesting that dystrophin also has important roles in the nervous system. In spite of dramatic progress in the field, e.g., the identification of Dystrophin’s interacting partners that form the Dystrophin associated Glycoprotein (DGC) complex, and the development of animal models of DMD, our understanding of Dystrophin’s likely many roles is still poor. Evidence has been presented that the DGC provides both the structural link between the cytoskeleton and extracellular matrix that underlies muscle contraction and acts to scaffold members of several signaling pathways.
Mammalian DMD models are complicated by genetic redundancy, e.g., the ability of utrophin to functionally substitute for dystrophin. We therefore are taking advantage of the significantly reduced number of DGC protein ortholog-encoding genes in the Drosophila. We have characterized the embryonic expression domains of all of the DGC members (Dekkers et al., 2004, Mech. Dev. GEP 4, 153), including the dystrophin isoforms (Figures 1 and 2). The fly DGC orthologs are predominantly expressed in the musculature and the central nervous system, similar to their mammalian counterparts. The overlap of a subset of their expression domains indicates that, at least some of them are likely to form DGC-like complexes in specific tissues. The large DLP2 isoform is found in the muscle in the postsynaptic density (Figure 2) and colocalized with actin at the sarcomeric I-bands. We have generated and characterized mutants lacking the large dystrophin isoforms and recently also generated mutants lacking the short CNS-specific dystrophin isoform, Dp186.
Figure 1: The Drosophila dystrophin gene, its promoters, protein isoforms and locations of P-elements and epitope regions used to generate antiserum. There are three transcripts, DLP1, DLP2 and DLP3 that encode the large protein dystrophin isoforms and a transcript, Dp186, encoding a short CNS-specific isoform. The EP(3)3397 and dysP-element were imprecisely excised to generate mutants specifically lacking DLP2 or Dp186, respectively. The conserved dystrophin domains are shown; yellow stars indicate the regions used to generate antibodies.
In collaboration with Dr. Jaap Plomp in the Departments of Neurology and Neurophysiology at the LUMC, we have established our ability to perform electrophysiological recordings at a well-studied Drosophila larval NMJ, a technique used in only a handful of labs throughout the world. Through immunohistochemical staining of the pre- and postsynaptic structures using monoclonal antibodies and electrophysiological recording in the mutants and isogenic controls, we have begun to elucidate the roles of dystrophin at the NMJ. Our results show a novel requirement for the large DLP2 dystrophin isoform in retrograde signaling from the muscle to the motoneuron to maintain synaptic homeostasis at the Drosophila NMJ (van der Plas et al., 2006, J. Neurosci. 26, 333). Lack of the DLP2 isoform, which is expressed postsynaptically in the musculature, leads to increased neurotransmitter release and muscle hyperdepolarization (Figure 3). This phenotype can be rescued by transgenic postsynaptic expression of DLP2, clearly indicating that postsynaptic DLP2 plays a role in regulating neurotransmitter release from the presynaptic apparatus.
Figure 2: The large dystrophin Isoforms are localized at the postsynaptic side of the NMJ. Third larval instar body walls were stained with anti-DysCO2H, which recognizes all Dystrophin isoforms (A, C, E), anti-Dyslarge, which recognizes only the large isoforms (D) or double labeled with anti-Synaptotagmin and anti-DysCO2H (B). (A) Dystrophin protein is expressed at the wildtype 3rd instar larval NMJ. (B) Double labeling of a wildtype larval body wall with anti-Synaptotagmin, staining presynaptic structures, and anti-DysCO2H reveals that the Dystrophin large isoforms are predominantly postsynaptically localized at the NMJ. (C) The DLP2 isoform protein accumulates highly at the NMJ after overexpression in the muscle. (D) Overexpressed DLP2 is also visualized at the NMJ using the large isoform specific anti-Dyslarge antisera. (E) Dystrophin protein is severely reduced in the dysGE20705 mutant. (F) RNA in situ hybridization of wildtype larval VNC (solid arrow), brain (arrowhead) and associated eye-antennal discs (open arrow) with a large isoform specific antisense probe reveals no apparent expression of large dystrophin isoform mRNAs in the VNC or brain. (G) Dp186 specific RNA in situ hybridization of the wildtype larval VNC (solid arrow), brains (arrowhead) and associated eye and antennal discs reveals that Dp186 mRNA is most highly abundant in the VNC, brain and the eye and antennal discs (open arrow). (H) Western blot analysis of embryo extracts prepared from wildtype, dysGE20705, elav-Gal4/GS12472 (overproducing DLP2) and elav-Gal4/UAS-Dp186 (overproducing Dp186) using the indicated antibodies and showing their specificity. Figure 3: Quantal content, the amount of neurotransmitter released after stimulation, is increased in the dystrophin mutant. Besides using the direct method (corrected EJP/mEJP), the quantal content was calculated using the failure analysis (Ln (Stimuli/Failures)) and the variance method (mean EJP/standard deviation) 2. Both methods confirmed the increase of the quantal content in the DLP2 mutant dysE6 compared to the wild type larvae at 0.25mM Ca.
Interestingly, the mutants that lack only the CNS specific isoform, Dp186, also show increased neurotransmitter release at inter-neuronal synapses in the Drosophila brain (Figure 4; Fradkin et al., submitted). Disruption of homeostasis at dystrophin-deficient inter-neuronal brain synapses may contribute to the poorly understood mental impairments observed in a number of DMD patients. When the expression of all dystrophin isoforms is reduced via transgenic RNA interference, we observe strong progressive muscle degeneration during the larval and pupal stages, further indicating the validity of the development of Drosophila as a model for DMD (van der Plas et al., in revision at Mech. Develop.). As lack of either the DLP2 or CNS-specific isoforms fails to result in alteration of muscle structure, we analyzed the effects of reducing the expression levels of the recently identified Dp117 Dystrophin isoform by use of transgenic RNA interference. Our results indicate that Dp117 is, at least in part, necessary for maintaining the musculature during development. Ongoing studies include localizing Dp117 protein expression domains and determining with which other genes it interacts.
Figure 4: Dp186 mutants, but not the DLP2 E6 mutant, show increased evoked responses at interneuronal synapses reflecting increased neurotransmitter release. Central interneuronal recordings were performed on first instar larvae of the indicated genotypes. The Dp186 mutants show significantly higher evoked responses than either the wildtype control or the DLP2 mutants. The spontaneous responses (miniatures) of all genotypes were identical (not shown), indicating that quantal content is increased at Dp186 mutant central interneuronal synapses.
What is the retrograde signaling pathway(s) through which DLP2 and Dp186 acts at the NMJ and interneuronal synapse, respectively? We find that Wishful thinking (wit), a presynaptic BMP receptor previously implicated in neuromuscular homeostasis, is required for the increased neurotransmitter release observed in the dystrophin mutants. However, the expression and activation of the well-characterized SMad effectors downstream of wit is unchanged in the dystrophin mutant or when dystrophin is overexpressed. Another possible retrograde signal candidate is nitric oxide (NO), a small membrane permeable signaling molecule known to be expressed at the NMJ. NO has previously been shown to stimulate Ca2+-independent vesicle release at the Drosophila NMJ. Furthermore, lack of the dystrophin protein in mammals is associated with loss of NO signaling. Ongoing work is directed at furthering our understanding of these retrograde signaling pathways and determining how reduced expression levels of Dystrophin isoforms impact them.
Wnt genes are highly conserved, secreted glycoproteins that function in a variety of developmental processes. Recently, Wnt genes have also been shown to be required for various processes in nervous system development, such as, cell fate determination, synapse formation, axon guidance and neurite outgrowth. We have generated Drosophila mutants that lack the Wnt5 gene whose protein is expressed on subsets of axon tracts in the central nervous system. We found that Wnt5 plays an important role in guidance of commissural axons across the ventral midline in the Drosophila embryo (Figure 1). Recently, members of the RYK/Drl receptor tyrosine kinase family have been shown by others to act as receptors for Wnt proteins (Yoshikawa et al. (2003), Nature 422, 583). Other members of the Wnt/RYK/Drl signal transduction pathway remain unknown. In order to better our understanding of Wnt5 signaling, we developed a dominant gain of function wnt5 over-expression assay useful for screening for interacting genes (Figure 1) and have performed a yeast-two-hybrid screen with the intracellular domain of the Drl receptor to identify its downstream effectors. We have found that Wnt5-mediated signaling requires the non-receptor tyrosine kinase, Src64B (Wouda et al., submitted). Lack of src64B leads to axon defects similar to those observed in wnt5 and drl mutants and src64B and drl are required for a dominant gain of function phenotype of wnt5. Src64B and Drl are expressed in an overlapping set of neurons and both proteins are transported out along the axon. Therefore, src64B function is essential for wnt5 signaling during nervous system development.
Figure 1: Abnormal commissural projections in the Wnt5 mutants and when wnt5 is overexpressed at the ventral midline. (A) Wildtype embryo stained with anti-BP102 that labels all neurons in the central nervous system showing the anterior (AC) and posterior (PC) commissures present in each hemisegment. (B) Wnt5 null mutant showing the loss of the separation between AC and PC (arrow). (C) Ectopic Wnt5 expression in the midline glia results in complete loss of the AC (arrow).
We are collaborating with three laboratories, each focusing on different aspects of nervous system development, to further our analyses of Wnt signaling. With Dr. H. Hing (University of Illinois, U. S. A.), we have found that Wnt5 can act as a powerful organizer of the olfactory map in the Drosophila adult brain (Nature Neuroscience, in revision). Loss of Wnt5 results in severe derangement of the glomerular pattern, while over-expression of wnt5 results in the formation of ectopic glomeruli (Figure 2). Interestingly, the Drl receptor seems to acts as a negative regulator of Wnt5 signaling in this tissue. With the Hassan lab (Leuven, Belgium) we have uncovered a role for wnt5 in an integrated system of Wnt, FGF and JNK signaling that regulates axon extension in the Drosophila brain (Figure 3, Srahna et al., 2006, PLOS Biol. 4, 2076). Lastly, we have collaborated with the Fornerod lab (N.K.I., Amsterdam) to show a novel role for RanBP3, a factor involved in the nuclear export of b-catenin, in Wnt signaling (Hendrickson et al., 2005, J. Cell Biol. 171, 785).
Figure 2: Wnt5 is required for olfactory map development. Wildtype (A and C) and wnt5 mutant (B and D) antennal lobes are shown during early (A and B) and late (C and D) pupal stages. The antennal lobe neuropile is labeled in red and the Olfactory Receptor Neurons are labeled in green. In the wnt5 mutant, the antennal lobe is poorly developed and the commissure (white arrow) is not formed.
Figure 3: Wnt5 is required for correct extension of axons into the adult Drosophila optic tectum. Adult brains labeled with a marker for a subset of DCN neurons that project into the optic system and the commissure connecting the two brain hemispheres are shown. In the absence of wnt5, only approximately half the numbers of axons extend into the optic lobe.