The Signal from the Target Travels Home

A sympathetic neuron’s cell body can sit centimetres away from the axon terminals where it collects NGF. The terminals receive the signal. The nucleus decides whether the cell lives.

You met the neurotrophic hypothesis in that lesson: neurons are overproduced, targets make a limited amount of a survival factor, and neurons that do not get enough die. Stated that way, the hypothesis skips a step. The target releases NGF near the axon terminals. The machinery that decides between life and death, the Bcl-2 family proteins, the transcription of survival genes, sits in and around the cell body. Something has to cross the distance between them. This lesson asks what crosses, how we know, and why it matters for which target factor keeps a neuron alive.

How does NGF transduce its signal?

The short answer comes from the Cell Signaling course. NGF binds its receptor tyrosine kinase, TrkA, at the axon terminal. The ligand and the active receptor are pulled into the cell together in a small membrane vesicle, the signaling endosome, which keeps signaling while it is carried (see that lesson). The motor protein dynein drags it along microtubules toward the cell body (see that lesson). Once there, kinase pathways reach transcription factors such as CREB (see that lesson).

That is the mechanism. The developmental question is different. Why must the survival signal travel at all, and what happens to a neuron whose signal cannot make the trip?

The evidence that something travels

The first evidence was anatomical. In 1974, Hendry, Stöckel, Thoenen and Iversen injected radioactive, iodine-labeled NGF into the eye of mice and found the label building up in the superior cervical ganglion on the same side, the sympathetic ganglion whose axons innervate the eye. In 1979, Judy Brunso-Bechtold and Viktor Hamburger did the embryonic version. They implanted a small gel pellet soaked in radioactive NGF under the leg skin of chick embryos at about ten days of incubation. Eight hours later, the lumbar dorsal root ganglia on the implanted side were labeled. Nor were the lateral motor columns, ventral roots or dorsal roots, though motor axons ran into the same leg.

That selectivity is the point. If the label had leaked into the blood, both sides would be labeled. If any nerve took it up, the motor column would be labeled. Instead it went to the sensory neurons, which respond to NGF. But note what the experiment could not see. Autoradiography locates radioactivity. It cannot tell you whether NGF arrived intact, whether it was active, or whether its arrival had anything to do with survival.

Robert Campenot’s dish answered the survival question. In 1977 he described a culture in which a divider, sealed to the dish, separated a central chamber from side chambers. Rat sympathetic neurons were plated in the centre, and their axons grew under the seal into the side chambers. The fluids did not mix. Two results followed. When NGF was removed from a side chamber, the axons there stopped growing and later degenerated, even though their cell bodies sat in NGF. And when cell bodies had no NGF, many still survived, provided their axons reached a chamber that had it. In 2002, MacInnis and Campenot found that NGF coupled to beads and supplied to the distal axons of rat sympathetic neurons kept over 80% of the neurons alive for 30 hours.

So a signal from the distal axon can keep the cell body alive. Does it reach the genome? Antonella Riccio and colleagues in David Ginty’s laboratory gave NGF only to the distal axons of cultured rat sympathetic neurons and found CREB phosphorylated in the nuclei. Stimulating the far end of the axon was enough to change the state of a transcription factor at the other end.

A Problem

Chris Deppmann poses the problem as three questions. How far can a signal travel? How close must it get to the nucleus? What qualities should a long-distance signal have?

Start with distance. For a molecule moving by diffusion, the typical time to cover a distance grows with the square of that distance.

time ≈ (distance)² / (2 × D)

A soluble protein in cytoplasm has a diffusion coefficient D on the order of 10 square micrometres per second. For 1 cm, which is 10,000 micrometres, that gives 10⁸ / 20, or about 5 million seconds: roughly two months. A phosphorylated kinase would be dephosphorylated long before then. Dynein-driven endosomes move at roughly a micrometre per second, so the same centimetre takes a few hours.

Distance tells you the signal needs a vehicle. The other questions tell you what kind. A good long-distance signal should stay active for hours, should carry the identity of the factor that started it, and should not be confused with signals the neuron picks up along the way. A membrane vesicle holding the ligand bound to its active receptor fits all three. That is why the signaling endosome became the favoured model.

Favoured is not proven. Heerssen, Pazyra and Segal reported in 2004 that dynein transports activated Trk receptors and that this transport is needed for the survival of target-dependent neurons. That is evidence for a carried signal. But imagine the alternative: each stretch of axon passes the signal to the next, and each local step needs dynein for its own reasons. Blocking dynein would stop that relay just as completely. The relay is not hypothetical. MacInnis and Campenot reported in 2002 that retrograde signals could support survival without retrograde transport of NGF itself. A dynein block, by itself, cannot tell a courier from a relay. You need to label the receptor at the tip and watch it arrive, or compare the speed of the signal with the speed of the motor.

NGF versus NT-3: why the second factor must travel

Development supplies the sharpest test. Sympathetic axons in the mouse grow out along blood vessels before reaching the organs they innervate. The vessels, an intermediate target, make NT-3. The final targets make NGF. Both factors can activate TrkA.

Kuruvilla, Ginty and colleagues asked what each does when it reaches only the distal axons. NGF supported local axon growth and kept the cell bodies alive. NT-3 supported local growth through TrkA, but not survival. The difference was in trafficking. NGF-TrkA complexes were internalized and carried back to the cell body. NT-3-TrkA complexes were not. In mice lacking NT-3, sympathetic axon growth and target innervation were impaired. The paper also proposed a handoff: retrograde NGF signaling raises the neuron’s level of p75, which reduces its response to NT-3, so the neuron stops listening to the intermediate target once it reaches the final one.

The logic matters for the neurotrophic hypothesis. The factor that decides survival must be the one whose signal can reach the nucleus. A factor that acts only locally can steer an axon, but it cannot select a neuron. Competition among neurons for the final target, the subject of that lesson, therefore depends on the trafficking of the receptor, not only on binding.

Design problem

Sympathetic axons pass an intermediate target rich in NT-3 before reaching a final target rich in NGF. Both factors activate TrkA on axons. You grow mouse sympathetic neurons in compartmented cultures with no factor at the cell bodies, and give NT-3 or NGF to distal axons only. Predict what each supports, and explain the difference.

Does a single flavor of endosome mediate all of these functions?

NGF arriving at the cell body does more than prevent death. It regulates dendrite growth, transmitter phenotype, cell size and, through the dendrites, synapse formation. One carrier, or several?

Work from Deppmann’s laboratory and its collaborators addresses this. Sharma, Deppmann, Harrington and colleagues (Neuron, 2010) found that in sympathetic neurons, NGF-TrkA signaling endosomes travel from distal axons to cell bodies and dendrites and promote assembly of the postsynaptic specialization, a long-distance relationship between target NGF and the synapses a neuron receives. Suo and colleagues (Nature Neuroscience, 2014) identified Coronin-1, a protein whose expression NGF induces, as an endosomal effector that keeps NGF-TrkA endosomes from fusing with lysosomes, and found that it is required for developmental competition for survival. Together these results suggest that biasing an endosome’s maturation changes which function it serves, so that distinct endosome pools may govern distinct developmental processes. The review by Barford, Deppmann and Winckler (Developmental Neurobiology, 2017) sets out this view. Whether every message rides in its own kind of endosome is still open.

In disease models, dynein and dynactin mutations that impair retrograde transport are linked to motor neuron degeneration, and a mouse model of Down syndrome shows failed NGF transport with degeneration of cholinergic neurons. What happens to an axon left without NGF is the subject of that lesson.