Biology Reference
In-Depth Information
demonstrated phenotypic effects of perturbing signaling.
This was initially done first by ectopic transplantation of
signaling tissues, and later by numerous biochemical and
molecular biological means. But, because of the structure
of the developmental regulatory system, trying to under-
stand development by focusing on signaling alone leads
inexorably to a phenomenological cul de sac. It is like
studying the output effects of flipping various switches on
the panel of an unfamiliar electronic device. Signals per se
have effects on development only indirectly, and only by
causing alterations of regulatory state in the cells receiving
the signal. The nature of the effects and the causal relations
between the signal and the developmental consequence
depend in detail on the structure of the GRN circuitry
called into play by the change in regulatory state that has
been effected by the signal. That is to say, the exact
consequences of signals for development depend in any
given case on the genomic regulatory wiring that mandates
GRN circuitry. This must be true, because the same few
signaling systems are used over and over in development.
Developmentally important signals work in general as
follows. Most signals utilized for developmental purposes
are protein ligands encoded by genes activated as part of
the function of regionally active GRNs. The signal ligands
are externalized, and the receiving cells express trans-
membrane protein receptors (as part of the transcriptional
function of their GRNs) that specifically recognize and
bind the signaling ligands. The consequence of extracel-
lular ligand
There are three other aspects of developmental
signaling that when viewed from the vantage point of
regulatory state effects are general enough to require
mention here. One is the 'Janus-like' behavior of many
signal transduction systems
[24,25]
. If the transcription
factor that serves as the early response factor for a given
signal is in a cell receiving the signal, it produces a newly
active regulatory factor as just discussed, but if not, the
same factor will act as an obligate repressor. Therefore,
were such a factor present in all cells of an embryo, it
would act as a global control agent, binding the sites which
it recognizes in its target genes in all cells and repressing
these genes everywhere except where the signal is being
received, while in these last cells it would contribute to
their activation. A second often encountered aspect is that
diffusible signals sometime function over distances as
vectorial patterning devices, that is, different regulatory
states are installed in response to the signal closer to its
source where it is stronger, and others are installed farther
away where the ligand concentration is lower. In some
well-analyzed examples specific types of GRN circuitry
can be shown to be responsible for discontinuous, hyster-
etic responses to different levels of signal input
[24,
26
e
31]
, and this may be the general explanation for the
often observed Boolean regulatory states set up in response
to graded signaling ligand distributions. Thirdly, inductive
signals are always transient. A predictable feature of
developmental GRNs downstream of the primary signal
response genes is circuitry that transfers dependence of the
new regulatory state from the external signal to internal
cross-regulatory functions encoded in the GRN architec-
ture. For example, it is commonly found that the gene
activated by a signal might activate other genes, which
engage in a positive feedback loop (e.g.,
[14]
). Once such
a loop is set up it not only obviates further dependence on
signal reception but also resets the levels of transcriptional
output downstream
[32]
.
Propagation of regulatory states within territorial
domains: There is also another kind of signaling, in which
cells of a territory all both send and receive the signal.
Termed community effect signaling
[33]
, this ensures that
the regulatory states of these cells are homogeneous, and
may be a fundamental mechanism by which the cells of
multicellular tissues remain in lockstep at the level
of developmental regulatory state. In several cases of
community effect signaling a common circuitry feature is
observed: the gene encoding the ligand responds to the very
same early response transcription factor that is activated on
reception of the signal. Its consequence is that all cells
receiving the signal also send the signal, so that within the
multicellular territory all the cells are locked into the
mutual embrace of an intercellular feedback system. Any
genes downstream which depend on the same activated
transcription factor will now be expressed in these cells at
receptor binding is to alter molecular struc-
ture on the intracellular end of the receptor protein
complex, which in turn affects intracellular biochemistry
in such a way as to alter the regulatory activity of a dedi-
cated target transcription factor present in the cytoplasm.
Some signals cause the target transcription factor to
acquire a cofactor it needs for activity, others cause it to be
released from a cytoplasmic docking protein, and still
others cause it to be chemically modified so as to activate it
or
e
stimulate its
transit
into the nucleus. Different
ligand
receptor pairs affect intracellular biochemistry in
different ways and utilize different target transcription
factors, but the underlying logic is always the same:
reception of the signal results in a specific, qualitative,
gain of function change to the regulatory state. Thus
signals indirectly result in new specification functions in
those cells positioned to receive the signal. The key
consequence is to cause new regulatory states to be set up
in given spatial domains. This is what used to be called
inductive signaling, defined phenomenologically by the
signal-dependent appearance of a new state of specifica-
tion in the receiving cells different from that of the sending
cells. Since progressive installation of spatial regulatory
states is causal for the whole developmental process,
perturbation of signaling often produces dramatic pheno-
typic results.
e
