| |
- The last decade has seen extraordinary progress in deciphering the
roles and mechanisms of action of Hedgehog proteins. From models in
the early 1990s of a short-range signal regulating pattern in the ectoderm
of the Drosophila embryo, Hh proteins are now recognized as acting both
locally and at long range to regulate a plethora of processes in vertebrate
as well as invertebrate development. At every step in the unraveling
of the signaling pathway there have been surprises, a fact that bears
ample testimony to the power of genetic analysis in uncovering novel
paradigms and principles. The first of these is the unusual autoprocessing
that generates the active Hh ligands, a process that simultaneously
couples them to cholesterol. As we have discussed above, this unique
lipid modification contributes to some key properties of the Hh signal,
mediating its controlled release and movement from its source. The process
of cell-tocell transport depends on two other components that appear
to be dedicated to Hh signaling: Dispatch mediated release of Hh-Np
from the sending cell and Toutvelu- dependent trafficking across the
target field. Cholesterol coupling may also ensure that Hh ligands are
concentrated in membranes, increasing the likelihood of ligand/receptor
interaction. Furthermore, the cholesterol anchor may target Hh ligands
to membrane subdomains that also localize Ptc or Hip1, thereby facilitating
either signaling or its termination by ligand sequestration. Robust
negative-feedback mechanisms are a hallmark of most signaling pathways,
and it is clear that in both the fly and the mouse, effective sequestration
of Hh by Ptc is dependent on its cholesterol linkage. A second lipid
modification, palmitoylation of the N terminus, also plays a key role,
but most likely not in membrane retention or movement. Rather, increasing
the overall hydrophobicity of this part of the protein some-how enhances
its ability to inactivate Ptc. Determining the structural basis of this
effect will be of key importance in understanding the dose-dependent
effects of Hh ligands. A further distinguishing feature of the Hedgehog
pathway is its mechanism of receptor-mediated activation. In most cases,
extracellular signals elicit their effects by binding to and activating
a membrane-anchored receptor that, in turn, activates intracellular
components of the pathway. Hh proteins, in contrast, act by repressing
their receptor, Ptc, which, in turn, controls the expression of Hh target
genes by repressing the activity of Smo. What is the logic of this unusual
mechanism? Presently the answer is unclear, but most likely it relates
to the peculiarities of Smo activation through some Ptc-dependent intracellular
trafficking process, a feature that Ptc might share with SCAP, another
SSD-containing regulatory factor. The recent discovery of the RABprotein
encoded by the opb gene apparently dedicated to this process, strengthens
this view. Future analysis of the subcellular behavior of Ptc and Smo
should yield some important new insights into this enigmatic process.
Finally, we have described the unusual way in which Hh signaling elicits
its effects at the transcriptional level by altering the sign of a bifunctional
transcriptional regulator. In Drosophila, the absence of Hh ligand allowsthe
cleavage of the Ci protein, converting it to a repressor form that can
bind target genes to block their transcription. Derepression of Smo,
in contrast, inhibits this cleavage and promotes the nuclear import
of activated full-length Ci, leading to the transcription of Hh target
genes. This highly economic process is further exemplified by the finding
that most, if not all, Hh signaling is mediated through Ci in Drosophila.
Why, then, do vertebrates use three distinct Gli proteins as transcriptional
effectors? One simple explanation could be that each operates similarly,
but that gene duplication and the acquisition of new regulatory motifs
have led to new tissues that incorporate Hedgehog signaling. Yet neither
the expression patterns nor the activities of the different Gli proteins
suggest this to be the case. Rather, there appears to have been a partial
separation of repressor and activator activities into individual Gli
proteins. At least one advantage of this elaboration would be to allow
more complex responses within a target field; thus, the response of
cells to Hh signaling would be dependent not only on the levels of ligand
to which they are exposed but also on the particular repertoire of Gli
genes that they express. Further analysis of the in vivo regulation
of Gli proteins and of their binding specificities for different Hedgehog
targets should help illuminate this aspect of the pathway. It is striking
that so much of what is known about this fascinating signaling mechanism
to date has been gleaned from genetic analysis, be it in flies or mice.
Yet whereas genetics has provided an elaborate framework for our understanding,
future progress will require a concerted effort to dissect the signaling
process at the biochemical and cell biological levels. The great advances
that have been made in identifying Hh-dependent processes and describing
the consequences of their activities must now be matched by elucidating
the ways in which Hh activities elicit these different cellular responses.
Characterization of the multimeric complex that regulates Ci activity
has provided a solid basis for this analysis, but many questions remain,
not least how Ptc and Smo interact and how Hh binding modulates their
interaction. The coming years promise to be at least as revealing as
the last. [Ingham, 2001]
- Posterior-compartment cells are often used as an assay system for
Hh signaling even though Hh signaling doesn't occur there. These cells
are ideally suited to addressing this question because they do not express
ci, hence the transgene-derived Gli protein can be studied in isolation;
they do not normally express ptc, hence activator function can be readily
observed (see above); they do express hh, simultaneously providing Hh
ligand and the possibility of using an hh reporter gene to assay repressor
activity; and finally, their ability to transduce the Hh signal can
be controlled by genetically manipulating the function of smoothened
(smo), which encodes an essential transducing component of the Hh pathway
(von Mering, 1999)
|
|
