I wish to add a few words to the excellent comments from Alonzo Ross.
First, it is clear from numerous studies that in vivo, more ofter and
not, the neurotrophin is available to receptors on the nerve terminus,
and not to receptors on the cell body. Thus, some signal must get from
the nerve terminus to the cell body. For example, a system studied in my
lab (by Chris von Bartheld) is the chicken brain stem isthmo-optic
nuclei, neurons of which innervate the retina. One nucleus innervates
one retina, its twin innervates the other retina. BDNF injected into one
embryonic eye is transported to only one nucleus, and the developmental
cell death of only one eye is decreased. Injection of receptor
antibodies (either p75 or trkB) into one eye inhibits BDNF
transport and enhances cell death of only one nucleus. These data all
suggest that the important interaction is between the neurotrophin and
the nucleus at the nerve terminus.
Like the other work that Alonzo mentioned, these studies implicate both
p75 and trk receptors in neurotrophin retrograde transport.
Simple diffusion of a signal from the nerve terminus is unlikely - it is
not fast enough. Calculate the time required for a molecule (for example
calcium ion) to diffuse from the toe to the spinal cord motoneurons of a
giraffe. The answer exceeds the lifetime of the giraffe. For a number
of systems, it has been shown that the time required for the trophic
signal to get from the axon terminus to the cell body increases with the
length of the axon, and is precisely equal to the known rate of vesicular
transport.
As Alonzo suggested, the most likely model is that a
neurotrophin/receptor complex forms at the axon terminus, is internalized
as an endosome, and that endosome may then be transported back to the
cell body with the receptor signaling the whole time. For many other
growth factors, the signal terminates after endosomal internalization
because the low pH (5.0) of an endosome dissociates the growth factor
from the receptor. However, there is evidence that the endosomal
compartment of axons has a pH near pH 7.0. Furthermore,
unlike other growth factors (eg., EGF) neurotrophins retain substantial
affinity for their receptor at acidic pH. Thus, the transport of the
neurotrophin itself does not constitute the signal, but the transport of
the neurotrophin may be required to maintain the activity of the receptor
which which it likely remains associated during transpor. An open
question of course is whether the neurotrophin/receptor complex
transports "naked" or whether it is transported while associated with
various SH2-domain containing signaling molecules.
Finally, I second Alonzo's nomination of the Campenot chamber as the best
evidence that gene regulation is not the whole story, but that direct
signaling action of the neurotrophin at the nerve terminus, not requiring
gene regulation, is important. An interesting demonstration of the
phenomenon comes from recent studies by Erik Gunther in my lab. In
Campenot chambers, Erik has generated scenarios in which sensory neurons,
in a central chamber bifurcate their axons, sending one axon branch into
a chamber to the left containing NGF, and the other branch into a second
chamber on the right containing BDNF. One can show that the elongation
of the two branches of the axon are independently regulated by the NGF
and the BDNF in the two chambers. I would also note the interesting
paper published a few years ago by Bob Nichols and Eric Shooter, in which
physically enucleated PC12 cells grew neurites in response to NGF,
clearly demonstrating that gene action is not required.
--
Mark Bothwell
Dept. of Physiology & Biophysics
Box 357290
University of Washington
Seattle, WA 98195-7290
Tel: 206-543-7924
Fax: 206-543-0934
