Axon guidance 3

Growth cones reprogram
1. Evidence
  1. Axons reprogram when intermediate targets (choice points) are encountered
    1. After transit of the midline, commissural axons lose their responsiveness to netrins
      1. Explains why, in the hindbrain, commisural axons are able to continue past the floor plate without turning
        But in the spinal cord, commissural axons turn after crossing the floor plate
    2. Commissural axons also become sensitive to repellants after crossing the floor plate
      1. Why don't these go straight on after crossing like those of the hindbrain?
        Inhibitory molecules in floor plate are semaphorins and slits
        Also present in ventral spinal cord
        Creating a channel through which commissural axons grow
2. How?
  1. Clues from drosophila
    1. As in vertebrate cells, insect midline glial cells express diffusible attractants (netrins) and cell surface repellants (slits)
    2. Also as in vertebrates, some axons cross forming commissures, and then turn to join longitudinal pathways formed by axons that have not crossed
    3. What controls whether axons cross or not, and if they cross, what prevents them from recrossing?
      1. Genetic screens identified mutants in which these processes went wrong
        Robo and comm mutants
Robo
1. Encodes a receptor for the inhibitory protein slit
2. Expressed at high levels on axons that dont cross the midline
  1. Commissural axons initially express low levels but high levels after they cross
3. Robo mutant
  1. No robo protein
  2. Slit is no longer detected
    1. All axons go back and forth across midline - forming roundabouts of axons
  3. Comm is expressed only in those neurons that normally cross the midline and is switched off after they cross
Comm
1. Comm mutant
  1. Comm protein missing
  2. Robo protein express at high levels in those cells that normally cross the midline
    1. Now extend their axons longitudinally
2. If expression forced in all neurons, robo is lost everywhere
  1. resulting in a phenotype just like robo mutant
    1. I.e. comm controls robo
      1. How?
        Robo is prevented from functioning before midline crossing in both flies and mice
        Comm encodes a traffiking protein that prevents robo from reaching the cell surface
        means that growth cone cannot receive slit inhibitory signals before crossing
        Also
        Vertebrate homologs of robo
        One of these, robo1, is expressed on commissural axons
        However, it is expressed both before and after crossing
        No comm homolog in vertebrates
        However, a second robo-like protein Rig1 (aka robo3) is expressed only in pre-crossing fibres
        Appears to block robo1 signalling until the midline is crossed
        Loss of Rig1 results in a failure of commissural axons to reach the mindline (floor plate)
        Conclusion
        Thus, although there may be differences in detail between vertebrates and invertebrates, the general mechanism is similar and axons are indeed reprogrammed along their pathway
Axon scaffold
1. Established by pioneer navigation
  1. Follower axons can follow
2. How to stay on the scaffold?
3. How to get off when target is reached
  1. FasII adhesion also controls defasiculation
    1. Overexpression of fasII leads to a 'by pass' phenotype
      1. The red and blue motor axons fail to defasiculate and so miss their targets
    2. FasII and other CAM adhesion can also be regulated by expression of other proteins
      1. E.g. BEAT
      2. Interfere with CAM-mediated adhesion
  2. How are targets chosen once in target area?
    1. Two main types
      1. Discrete targets
        "cellular"
        In Grasshopper and Drosophila, ablation of specific target muscles (by ablation of muscle precursor cells) leads to failure of relevant motor axons to leave main motor trunk at appropriate branch points
        Suggests axons are looking for specific labels on their targets
        Insect muscles carry a diverse set of molecules that together may constitute muscle 'address labels'
        Netrin
        Expressed in specific muscles
        Loss of netrin is exactly like ablating the muscles
        The axons wander and do not make synapses (even though the muscle is there)
        Ectopic netrin leads to axons innervating the wrong muscles
        Fasciclin 3
        Expressed in specific muscles and the motor axons that normally innervate them
        Ectopic expression of Fas3 leads Fas3-expressing axons (the red ones!) to innervate new targets
        Multiple cues seem to combine to make 'address labels'
      2. Topographic maps
        "multicellular"
        When neighbouring neurons send axons to neighbouring sites in their target to maintain the topology (order) in the target, e.g. retinotectal system
        could this be achieved?
        Sperry proposed two possibilities
        1) Each axon has a unique label complementary to a unique label in the target. (cf address labels in fly muscle)
        Number of different labels needed seems implausably high
        2) That a co-ordinate system, encoded by gradients of signalling molecules, stamps a "latitude and longitude" onto cells of the target. This would be read by complementary gradients of receptors expressed in the retinal ganglion cells.
        Evidence
        The “Stripe Assay” shows that cells from posterior tectum make a non-permissive factor that repels temporal retinal axons
        The inhibitory factor in the posterior tectum turns out to be two ephrins which are, as Sperry predicted, expressed in a gradient from posterior (hi) to anterior (lo)
        As also predicted by Sperry, an Eph receptor for ephrins A2 & A5 is expressed in the retina in a counter gradient from temporal (hi) to nasal (lo)
        In mice in which both Ephrin A2 and A5 are knocked out, temporal neurons project their axons into the posterior tectum and the topographic map is disordered
        Conclusion
        non-permissive, repellant factors can be used instructively - ie they can direct growth cones to specific places - to form topographic maps
4. Both involve controlling fasiculation
  1. In simplest form involves homophilic binding by cell adhesion molecules (CAM's)
    1. E.g. fasiculin II in insects
      1. Controls fasiculation in flies
        FasII mutants (no fasII) have many defasiculated axons
        Overexpression of fasII leads to novel fasiculations
    2. Homophilic 'like for like' interactions can bind two cell surfaces together
      1. In cells that do not normally adhere, FasII can cause aggregation
  2. Fasiculation = arrangement in bundles
Target derived factors
1. feedback from the target is key in determining the survival, phenotype and morphology of innervating neurons
2. target-derived factors are critical in formation of the monosynaptic stretch reflex
  1. Target-derived factors determine dendritic morphology and connectivity
    1. MNs innervating the triceps and pectoral muscles develop monosynaptic connections directly with PNs, whereas those innervating the cutaneous maximus (CM) and latissimus dorsi (LD) receive polysynaptic input from interneurons (IN).
      1. This is controlled by GDNF* secreted from the CM and LD muscles, which turns on the transcription factor (TF) Pea3 in the MNs
        Loss of Pea3 results in MNs innervating CM/LD that have the dendritic morphology of Tri and Pec-innervating MNs and aberrant proprioceptive connections
  2. Circuit completion relies on target feedback
    1. Similarly, muscle-expressed neurotrophin-3 (NT3) induces the expression of the TF Er81 by Ia proprioceptors (PNs)
      1. Knockout of Er81 leads to a failure of the Ia central projection to reach the ventral horn and form monosynaptic connections with the appropriate MNs
        Thus, feedback from the target determines the final patterns of both dendritic and axonal connections in the monosynaptic spinal reflex (and elsewhere)

Next up

Axon guidance 3

Description

Resource summary

Media attachments

Similar

	Created by Kristi Brogden over 10 years ago