Project 1

Sensory axon pathfinding in the PNS

Funded by ARC Discovery Grant 2005-07

The sensory nervous system of the Drosophila embryo provides us with a simple system for studying the mechanisms by which axons navigate through the early part of their trajectory. Axons from each of the 42 sensory neurons in abdominal hemisegments of the embryo grow towards the CNS following a stereotypic route (see image and diagram below).

The full set of sensory neurons in a single abdominal hemisegment of a Drosophila embryo stained with monoclonal antibody 22c10. The drawing below shows the course followed by the axon of a single neuron from each of the dorsal, lateral, ventral' and ventral clusters as they head towards the CNS. Image by Kerri-Lee Harris.

We have previously described in detail the pathways that these sensory axons follow during normal development (Harris & Whitington 2001). This has revealed the series of choice points faced by the sensory axons along their trajectory. We are now using this knowledge to:

• Identify molecules that play key roles in guiding these sensory axons at these choice points.
• Understand how these molecules work together to cause sensory growth cones to grow in a particular direction at each choice point.

Schematic view of pathways followed by sensory axons. Choice points, where axons leave one growth substrate and associate with a new one, are indicated by the small circles:

lateral view

cross-section view

We are using two main approaches to find sensory axon guidance molecules.

• Embryos carrying loss-of-function mutations in candidate axon guidance genes are collected and the pattern of axon growth from the sensory neurons is examined using specific staining techniques.
• We over-express candidate genes either in the neurons or in their growth substrates using the GAL4-UAS system (Brand and Perrimon, Development 118, 401-415 (1993)) and examine the effects on sensory axon growth.

A defective pattern of axon growth in a mutant or transgenic embryo suggests that the protein encoded by that gene plays a role in axon guidance. The normal pattern of tissue expression of that molecule is examined to provide further evidence for such a role and to shed light on its mechanism of action.

For example, in robo2 mutants, the dorsal cluster sensory axons often fail to grow ventrally along the trachea - their normal route. Instead, they may project dorsally, anteriorly or posteriorly, sometimes crossing into the next segment - see image below (Parsons et al, 2003).

Sensory axon defects in a robo2 mutant embryo. In the top hemisegment, the dorsal cluster axons project posteriorly and dorsally, and join the dorsal sensory neurons in the adjacent segment (blue arrow). In contrast, the dorsal sensory axons in the lower segment project in a normal, ventral direction (red arrow). Image by Veronica Martin.

Robo2 protein is expressed on the tracheal branch normally followed by the dorsal sensory axons but not on the sensory axons themselves (see image below). These data suggest that Robo2 acts as an attractant for the dorsal sensory axons. When Robo2 protein is missing, as in a robo2 mutant embryo, these axons don't reliably recognise their normal growth substrate, the trachea, and head off in an alternative direction.

An antibody against Robo2 protein stains the trachea followed by the sensory axons (blue staining indicated by red arrows) but does not stain the sensory neurons themselves (here stained brown with the 22c10 antibody, indicated by blue arrow). Image by Veronica Martin.

 

Robo2 is not the only gene involved in guidance of the dorsal sensory axons. We have seen defective axon growth from these neurons in other mutants, e.g. Notch , flamingo , sema2a . To work out how these and other genes interact to guide the sensory axons, we make embryos that carry mutations in two or more of these genes. A change in the frequency of axon defects in these embryos compared to what is seen in the single mutants provides evidence of an interaction between these genes.

An important tool in understanding how this genetic interaction leads to changes in the behaviour of the sensory growth cones is time-lapse microscopy.

Project 2

How do sensory axons find their synaptic targets in the CNS?

Funded by NH&MRC Projects Grant 2005-07

Once a sensory axon reaches the ventral nerve cord, its journey is far from over. It must then navigate through the complex terrain of the CNS to find a specific target neuron, with which it will make synaptic connections. It is likely that the recognition by a sensory axon of its target cell depends upon an interaction between molecules expressed on the axon and its target. However little is known about the identity of such "targeting" molecules or how they work.

The goal of this project, carried out in collaboration with Dr David Merritt, University of Queensland, is to identify sensory axon targeting molecules, using a genetic approach. We reason that a mutation in an axon targeting molecule will result in defects in the morphology of the sensory axon terminals. We are therefore examining sensory axon morphology in the CNS of embryos carrying mutations in candidate axon targeting genes. To reveal sensory axon morphology in the CNS of these mutants, we use two methods:

• We inject individual sensory neurons with DiI (see image and drawing below).

Part of the ventral nerve cord of a late stage embryo visualised with DIC optics. The cell bodies of central neurons and the neuropile can be clearly seen. lch5-5 neurons in two adjacent hemisegments have been injected with DiI to reveal their axon branching patterns within the neuropile. The lower axon has attained its mature morphology, while the upper axon has not yet turned anteriorly. Image by Veronica Martin.

• We express a membrane targeted form of Green Fluorescent Protein in specific groups of sensory neurons.

We have identified a role for the Epidermal Growth Factor receptor (EGFr) in triggering sensory axons to arborise within the CNS. When EGFr levels are knocked down in sensory neurons by driving a dominant negative form of the gene, the lch5-5 axon fails to turn anteriorly. Instead it stalls at the stage shown by the upper axon in the image above. We are currently trying to determine whether EGF ligands instruct the lch5-5 axon to grow along the specific longitudinal path within the connectives taken by the lower cell in this image.

Project 3

Mapping neurons in the CNS of the Drosophila embryo.

Drosophila along with many other invertebrates, is a highly suitable model organism for a wide range of problems in neurobiology and development because it possesses identified neurons. Being able to analyse neural differentiation at the level of single, identified cells has been decisive in advancing our understanding of the genetic basis for axon guidance and determination of neural fate.

All of the neurons in the PNS of the Drosophila embryo, but only a relatively small subset of the neurons in the CNS, have been individually identified. Studies of neural development in the CNS have been hampered by this lack of data on axon morphology of interneurons at the single neuron level.

In conjunction with Professor Gerd Technau, Christof Rickert and Thomas Kunz at the University of Mainz, we have embarked upon a project to characterise the axon morphology of all of the ~ 300 interneurons in each abdominal hemi-segment of the late stage wild-type Drosophila embryo.

We are systematically dye-filling each central neuron with DiI, digitally imaging the stained neurons (see below) and recording aspects of their axon morphology in a visual database. We have now filled virtually all of the interneurons and are currently writing up a description of their morphologies for publication in late 2006/early 2007.

Image by Kerri-Lee Harris.