[From Bill Powers (2005.07.31.1951 MDT)]
In the 14 July 2005 Nature, starting on p.193, is a review article by
Carmeliet and Tessier-Lavigne titled “Common mechanisms of nerve and
blood vessel wiring.” It is about the interesting fact that blood
vessels grow right along with nerve axons during development, often
following almost identical paths, under the influence of the same
chemical “cues.” But the most interesting aspect of the article
from our point of view is that the people doing this research, while
astonishinly good at what they do, desperately need control theory to
make sense of what they’re finding.
I’ve snipped two excerpts, reproduced below. In the first excerpt, we see
the confusion that results from not understanding negative feedback
control. Somehow, “cues” can act to “attract” or
“repel” growing axons, and cause them to pass on to the next
stage of their growth, all as if the power to cause these things lay in
the chemical cues rather than in control systems in the growing axon. The
first passage:
···
============================================================================
Angiogenic and axon terminal sprouting both serve the same function:
to provide coverage of a target tissue. In each case, target
cells direct sprouting through regulation of growth factor release.
For vessels, a hypoxic tissue secretes VEGF, thus beckoning its
vascularization; VEGF expression is downregulated when target
cells receive appropriate oxygen supply5. For axon terminals, target
cells devoid of synaptic input secrete growth factors such as nerve
growth factor (NGF) that similarly beckon their innervation and are
downregulated when appropriate electrical stimulation of the target
is successful4.
=======================================================================
That, of course, is a description of two negative feedback control
systems, one with a high reference level for oxygen and the other with a
high reference level for synaptic input (or something that synaptic input
produces). VEGF and NGF are the output signals.
Of course two other control systems are also implied, one in the growing
blood vessel and the other in the growing axon. “Nerve growth
factor” or NGF is clearly a controlled input for a growing axon; the
axon seeks a high reference level of NGF, and steers toward it, up the
gradient. similarly for the growing blood vessel which puts out (literal)
“feelers” at the growing tip.
The authors continue in a perfectly approriate way, even though they
never mention control systems:
==========================================================================
Thus, unlike the more stereotyped patterning of larger
vessels and nerves (see below), formation of capillary networks and
axon terminal arborizations is non-stereotyped, but governed by the
metabolic or electrical needs of the target tissue. Furthermore, in
both systems sprouting requires not merely the presence of growth
factors, but also of appropriate gradients of these factors�we
review
this first for VEGF-mediated stimulation of angiogenesis. (p. 194)
=========================================================================
In the next excerpt, some feeble attempts are made to understand what’s
going on. You’ll find this passage:
"What molecules guide axons?
The task of orchestrating the complex wiring of neural circuits�and
possibly also of the vascular networks�is largely carried out by a
limited number of guidance
cues." Somehow, those cues do it all. This is the old-fashioned
chemist’s way of seeing the body: the proximal cause is all that is
considered. The closed loop is simply ignored. Yet it doesn’t take much
imagination to see that we’re talking about interacting control systems
of the kind we’re all familiar with. Those “second messengers”
inside the cells, if traced out functionally and seen as components of a
system, would show control systems all over the place. I do wish we had
some neuroscientists and biochemists around who can recognize a control
system when they see one.
Anyway, read on:
============================================================================
We now discuss the second parallel
in the formation of neural and
vascular networks�in this case, involving common molecular
cues�that is, how larger vessels and axons follow highly stereotyped
anatomical patterns (Fig. 1). How can axons navigate to far away
targets over distances exceeding the neuronal cell body diameter by
more than a thousand fold? They simplify this task by breaking
up their long trajectory into smaller segments bounded by
intermediate
targets1
. Thus, the complex task of
projecting long distances
is reduced to the simpler task of navigating a series of short
segments�each perhaps a few hundred micrometres long�from
one intermediate target to the next. Although endothelial cells do
not
migrate over distances anywhere near approaching those of the
longest axons, they often migrate over trajectories that are similar
in length to the shorter segments navigated by axons.
The guidance of axons and endothelial cells is directed by specific
cues in the extracellular environment. Over the past decade,
considerable
progress has been made in understanding axon guidance
mechanisms1�3
. Guidance cues come in four
varieties: attractants and
repellents, which may act either at short range (being cell- or
matrixassociated)
or at longer range (being diffusible). Intermediate targets
are often the source of long-range attractive signals that lure
axons,
and of short- or long-range repellent signals that expel axons
that have entered the target, or prevent their entry altogether. In
between intermediate targets, axons and vessels are often guided
through tissue corridors by attractive cues made by cells along the
corridors, and by repulsive signals that prevent them from entering
surrounding tissues.
What molecules guide axons? The task of orchestrating the
complex wiring of neural circuits�and possibly also of the vascular
networks�is largely carried out by a limited number of guidance
cues. In the 1990s, four families of axon guidance cues were
discovered: the netrins, semaphorins, ephrins and Slits, and their
receptors (Fig. 4; and see below). In addition, growth factors (like
neurotrophins and hepatocyte growth factor (HGF)) function as
axonal attractants and morphogens (Hedgehog, Wnt and BMP
proteins) have recently been implicated as both attractive and
repulsive cues (reviewed in ref. 26). The responses of axons to
these guidance cues show remarkable versatility and plasticity. A
single cue may be attractive or repulsive, or regulate axonal
branching,
depending on the complement of receptors expressed by
the responsive neuron or the activity of second messengers in the
neuron, and an individual axon may change its response to cues as it
develops1,2,27
. This plasticity allows an axon to
be initially attracted,
and then to switch its response such that it becomes repelled by
cues
in the intermediate target, causing it to move on to the next leg of
its
trajectory. Examples of this versatility, and the mechanisms that
make it possible, are discussed in detail below.
=========================================================================
Sad, isn’t it?
Best,
Bill P.