![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Vol. 12, No. 12, pp. 1751-1762, June 15, 1998
Department of Biochemistry and Molecular Biophysics, Howard Hughes Medical Institute, Columbia University College of Physicians and Surgeons, New York, New York 10032 USA
LIN-12/Notch proteins function as
receptors for intercellular signals during development. Many aspects of
LIN-12/Notch-mediated signaling have been elucidated
through studies of cell-cell interactions that occur during
Caenorhabditis elegans and Drosophila melanogaster development. The basic principles that operate in these lower organisms
have also been shown to apply to vertebrates (for review, see Gridley
1997 Genetic studies have identified many conserved components that are
important for LIN-12/Notch signaling. Table
1 summarizes the C. elegans and
Drosophila genes involved in LIN-12/Notch
signaling mentioned here, and their vertebrate counterparts. Of
particular note is that transmembrane protein ligands of the conserved
Delta/Serrate/LAG-2 (DSL) family activate
LIN-12/Notch signal transduction (for review, see
Weinmaster 1997
Introduction
Top
Introduction
References
). Molecular features defined in lower organisms have also been
shown to be conserved in vertebrates, including components of the
signaling and signal transduction systems (for review, see Weinmaster
1997). The focus of this paper is on what has been learned about
LIN-12/Notch signaling from invertebrates. First, a
description of roles for LIN-12/Notch proteins in
development is given, using different model cell fate decisions to
illustrate various features. A discussion of the mechanism of
LIN-12/Notch signal transduction follows, including new
in vivo evidence that favors the direct participation of the
intracellular domain of LIN-12/Notch proteins in
regulating target gene expression. Finally, other influences on
LIN-12/Notch activity are discussed, particularly protein
turnover and protein processing.
The LIN-12/Notch "pathway"
) and that many LIN-12/Notch outputs are
mediated by a transcription factor named Suppressor of Hairless
[Su(H)] in Drosophila and LAG-1 in C. elegans (see
Schweisguth and Posakony 1994; Bailey and Posakony 1995; LeCourtois and
Schweisguth 1995; Christensen et al. 1996). One "pathway"
involving Su(H) is considered below and is diagrammed in Figure 2
(below). However, there is evidence that certain responses to activated
Notch do not depend on Su(H) (Lecourtois and Schweisguth 1995; Shawber
et al. 1996; Wang et al. 1997) and, as described below, there are other
potential roles for LIN-12/Notch proteins that may use
other signal transduction components.
Table 1.
C. elegans and Drosophila genes
featured
| |
LIN-12/Notch proteins mediate cell-cell interactions |
|---|
The fundamental role inferred from genetic studies of
LIN-12/Notch proteins is that they mediate inductive or
lateral cell-cell interactions that specify cell fate. Inductive
interactions involve signaling between nonequivalent cells. There are
numerous examples of inductive interactions mediated by
LIN-12/Notch proteins, and in many of these cases, the
expression of ligand in the inducing cell or tissue appears to be the
critical regulatory step. For example, two inductive interactions
mediated by the C. elegans LIN-12/Notch protein
GLP-1 have been studied in some detail. GLP-1 is expressed during
embryogenesis by two equivalent blastomeres, ABa and ABp (Crittenden et
al. 1994); the ligand APX-1 is expressed by a nonequivalent blastomere,
P2, and induces ABp to follow a distinct fate from ABa (Mango
et al. 1994; Mello et al. 1994; Mickey et al. 1996). GLP-1 is also
expressed in the germ line; the ligand LAG-2 produced by a somatic
gonadal cell activates GLP-1 in the germ line to promote mitosis
(Crittenden et al. 1994; Henderson et al. 1994).
Lateral interactions occur within a population of equivalent cells and
result in the generation of cells of different types. This process has
been termed "lateral inhibition," because it has been envisaged
as a competition between cells for one of the fates, with one cell
deciding to adopt a default fate (the fate it would adopt if it were
isolated from the other cells) and then inhibiting others from adopting
that fate (Wigglesworth 1940; Lawrence 1992). However, results from
genetic studies of lin-12 in C. elegans and
Notch in Drosophila suggest that cells communicate with each other prior to commitment to either fate, so that the cells
essentially specify each other; hence, I prefer the term "lateral
specification" for this process. In some cases, other factors influence the
outcomes of signaling between cells that are essentially equivalent.
| |
Feedback mechanisms during lateral specification |
|---|
In the absence of other influences, lateral specification is stochastic, in that any cell in the population could adopt the default fate. Studies of lin-12/Notch activity in C. elegans and Drosophila have suggested that during some cell fate decisions, differences in lin-12/Notch activity are amplified by feedback mechanisms that affect both ligand and receptor expression. These principles are exemplified by a simple decision involving only two equivalent cells in C. elegans gonadogenesis and a decision involving a small group of equivalent cells in Drosophila sense organ development.
The first evidence that LIN-12/Notch-mediated lateral
specification involves the amplification of a stochastic small
difference between equivalent cells came from a study of
LIN-12-mediated signaling between two cells of the hermaphrodite gonad
of C. elegans (see Fig. 1). These gonadal
cells, named Z1.ppp and Z4.aaa, are initially equivalent in their
developmental potential in that each has an equal chance of becoming
the anchor cell (AC), a terminally differentiated cell type, or a
ventral uterine precursor cell (VU), which contributes descendants to
the ventral uterus (Kimble and Hirsh 1979). However, in any given
hermaphrodite, only one of these cells will become the AC, whereas the
other becomes a VU, depending on interactions between them (Kimble
1981; Seydoux and Greenwald 1989).
|
The interactions between Z1.ppp and Z4.aaa are mediated by
lin-12 and lag-2, a gene encoding a DSL ligand
(Greenwald et al. 1983; Lambie and Kimble 1991; Henderson et al. 1994;
Tax et al. 1994). Mutations that constitutively activate
lin-12 have no AC, and mutations that eliminate
lin-12 activity have two ACs (Greenwald et al. 1983). Thus,
activation of lin-12 results in the VU fate; failure to
activate lin-12 results in the AC fate.
Genetic mosaics in which lin-12(+) and lin-12(0)
cells are juxtaposed suggested the existence of a feedback mechanism
(Seydoux and Greenwald 1989). In genetic mosaics in which either Z1.ppp or Z4.aaa lacked lin-12 activity [lin-12(0)] while
the other cell was lin-12(+), the lin-12(0) cell
always became an AC, presumably because it could no longer receive the
signal. In addition, the lin-12(+) cell always became a VU
and therefore behaved differently in the genetic mosaic than in wild
type, where it has an equal chance of becoming an AC. These
observations suggested that a stochastic variation in ligand
and/or receptor activity between Z1.ppp and Z4.aaa is
amplified by a feedback mechanism in both cells.
Analysis of the patterns of lin-12 and lag-2
expression during the AC/VU decision suggested that
transcriptional control is a component of the feedback mechanism
(Wilkinson et al. 1994). Initially, lin-12 and lag-2
are expressed in both Z1.ppp and Z4.aaa. However, the expression
patterns change in a reciprocal manner, so that lin-12
expression becomes restricted to the presumptive VU and lag-2
expression becomes restricted to the presumptive AC. These changes in
gene expression occur prior to commitment. Furthermore, they are
influenced by lin-12 activity, which appears to promote
expression of lin-12 and to repress expression of
lag-2. If the later phase of lin-12 expression in the
presumptive VU is reduced by deleting a 5' regulatory element, a VU
is not made, indicating that positive autoregulation of lin-12
transcription in the presumptive VU is necessary to specify the VU fate.
Studies in Drosophila have suggested that such feedback
mechanisms operate in other cell fate decisions (Heitzler and Simpson 1991; see also Huppert et al. 1997), as exemplified by specification of
sensory organ precursors (SOPs) that give rise to bristles on the
notum. During development of the peripheral nervous system, equivalent
proneural cells within a small cluster interact via Notch so
that one becomes the SOP, giving rise to a neuron and accessory cells,
whereas the others become epidermis (Stern 1954; Shellenbarger and
Mohler 1978; Hartenstein and Posakony 1990; Simpson 1990).
A feedback mechanism was again inferred by the observation of a bias in
cell fate choice in genetic mosaics: When Notch+ cells are
juxtaposed to Notch
cells, the Notch cells
always become SOPs and the Notch+ cells always become
epidermis (Heitzler and Simpson 1991). Furthermore, the opposite bias
in cell fate choice was observed in genetic mosaics involving the
ligand Delta: When Delta+ cells are juxtaposed to
Delta cells, the Delta cells become epidermis
and the Delta+ cells become SOPs (Heitzler and Simpson
1991). Support for the involvement of transcriptional control in this
feedback mechanism during SOP specification has been suggested by
analysis of genetic mosaics involving genes of the
achaete-scute (ac-sc) complex (Heitzler et al.
1996), known transcriptional activators of Delta (Kunisch et
al. 1994), and the analysis of the pattern of Delta expression in the
SOP lineage (Parks et al. 1997).
In Drosophila, the downstream effects of activating Notch have
been intensively studied. Although the circuitry underlying Notch
positive autoregulation is not known, the circuitry underlying the
feedback loop that represses ligand expression upon receptor activation
during SOP specification has been well described (Fig. 2; Heitzler et
al. 1996). Su(H) is the key effector upon Notch activation in SOP specification (Schweisguth and Posakony 1994; Bailey
and Posakony 1995; Lecourtois and Schweisguth 1995; Schweisguth 1995).
Su(H) promotes expression of genes of the Enhancer of split [E(spl)] complex encoding basic helix-loop-helix (bHLH)
proteins (Bailey and Posakony 1995; Lecourtois and Schweisguth 1995),
which combine with a protein called Groucho (which maps to the
E(spl) complex but is not a bHLH protein) to create a
repressor (Paroush et al. 1994; Fisher et al. 1996). The E(spl)
bHLH-Groucho protein complex represses genes of the ac-sc
complex, which encodes multiple proteins of a different bHLH class
(Oellers et al. 1994; Heitzler et al. 1996). As the Ac-Sc proteins
activate expression of Delta (Kunisch et al. 1994), this
feedback loop operates to repress Delta expression in the cell
in which Notch has been activated.
|
There are two additional points pertaining to SOP specification that
should be noted here. First, the ac-sc genes also act upstream of Notch, in that expression of ac-sc
complex genes establish the proneural clusters. Global positional cues
govern the pattern of ac-sc expression to establish the
proneural clusters that undergo Notch-mediated cell-cell
interactions (for review, see Skeath and Carroll 1994). The
Notch-mediated interactions lead to increased expression of
ac-sc in the presumptive SOP and repression of
ac-sc in other cells of the cluster. Second, for certain
macrochaetes (large bristles) the position of the SOP is not always in
the center of the cluster; however, the SOP is generated at a
reproducible position within a cluster (Cubas et al. 1991), suggesting
that other factors may influence the specification of the SOP.
| |
Biased LIN-12/Notch-mediated interactions |
|---|
LIN-12/Notch proteins are also involved in specifying the fates of cells in invariant lineages. In these cases, although each cell of an equivalence group has the potential to adopt either the default or the alternative fate, intrinsic or extrinsic factors lead to a predictable outcome.
Asymmetric segregation of modulating factors: Drosophila Numb
A notable example of a lineage-based mechanism that biases a
Notch-mediated interaction occurs during Drosophila SOP
development, where the asymmetric segregation of Numb modulates Notch
activity in one of two daughter cells (Fig. 3). The
SOP normally undergoes a defined lineage, producing two daughters: One
daughter produces a hair cell and a socket cell; the other daughter
produces a neuron and a sheath cell (Hartenstein and Posakony 1989).
Temperature-shift experiments have established that Notch and
Delta function at each step of the lineage (Hartenstein and
Posakony 1990; Parks and Muskavitch 1993): If Notch activity
is reduced around the time of the first SOP division, the two daughters
each produce a neuron and a sheath cell; if Notch activity is
reduced at both the first and second divisions in the SOP lineage, four
neurons are generated. The involvement of Notch and
Delta at these steps of the SOP lineage imply that cell-cell
interactions between the SOP daughters and between SOP granddaughters
are important for generating differences within the SOP lineage.
|
numb was identified by its effects on Notch-mediated cell fate
decisions (Uemura et al. 1989). numb appears to have a
cell-intrinsic influence on Notch-mediated signaling in several
lineages. For example, at each successive step in the SOP lineage, the
Numb protein is believed to be newly synthesized and then
preferentially segregated into one of the SOP daughters (Rhyu et al.
1994; Knoblich et al. 1995) (see Fig. 3). Jan and Jan (1995) have
suggested that numb influences Notch-mediated signaling at
successive steps in the SOP lineage by reducing Notch activity
in the cell that contains Numb, thereby biasing the interaction between
sister cells that signal each other. They envisage the
numb-imposed bias as a mechanism to increase the reliability
of Notch-mediated signaling in situations requiring rapid decision
making, when there may be insufficient time to activate
transcription-based feedback mechanisms (Jan and Jan 1995).
The asymmetric localization of Numb also has been shown to influence a
Notch-mediated inductive interaction in a central nervous system (CNS)
lineage. Two sister cells, dMP2 and vMP2, adopt different fates during
CNS development as a result of an inductive signal emanating from
outside the MP2 lineage (Spana and Doe 1996). The preferential
segregation of Numb into the dMP2 cell appears to account for the
different response of the sister cells to the inductive signal (Spana
and Doe 1995, 1996; Spana et al. 1995).
Numb is a novel membrane-associated protein (Uemura et al. 1989). The
amino-terminal portion of Numb appears to interact physically with the
amino-terminal portion of the Notch intracellular domain (Guo et al.
1996). The mechanism by which Numb binding to Notch inhibits Notch
activity is not known; one possibility is that Numb interferes with the
interaction of Notch with Su(H) or another transcription factor.
Differential expression of modulating factors: Drosophila Fringe
Studies of the Drosophila fringe gene suggest that Fringe
influences the response of cells to ligands for Notch (Irvine and Wieschaus 1994; Kim et al. 1995; Fleming et al. 1997; Panin et al.
1997; for review, see Irvine and Vogt 1997) (see Fig.
4). The role of Fringe is best understood in the
induction of the Drosophila wing margin. The dorsal and
ventral compartments of the wing primordium are destined to form the
dorsal and ventral surfaces of the wing blade. The wing margin forms
where these two compartments meet; both dorsal and ventral cells at the
interface become specialized margin cells. The wing margin cells
secrete a morphogen that organizes wing development and also specifies the distinctive structures of the wing margin, including a highly stereotyped pattern of mechanosensory bristles (for review, see Lawrence and Struhl 1996). Notch-mediated interactions between dorsal
and ventral cells establish the wing margin via a complex interplay of
regulatory interactions among Notch and Fringe, and the Notch ligands
Delta and Serrate.
|
The role of fringe appears to be to cause Notch to respond to
Delta rather than Serrate. Serrate has been implicated as a signal from
dorsal to ventral cells (Diaz-Benjumea and Cohen 1995; Kim et al. 1995;
DeCelis et al. 1996), whereas Delta has been implicated as a signal
from ventral to dorsal cells (Diaz-Benjumea and Cohen 1995; DeCelis et
al. 1996; Doherty et al. 1996). The patterns of Delta and Serrate
expression change over time, dependent on Notch activity. Early in wing
development, Serrate and Fringe are expressed in all cells of the
dorsal compartment under the control of the dorsal selector gene
apterous, whereas Delta is expressed in all cells of both the
dorsal and ventral compartments. However, Delta expression is
dramatically up-regulated in ventral cells along the boundary, in
response to Serrate expressed by dorsal cells. Conversely, high levels
of Delta induced by Serrate in ventral cells signal back across the
boundary and up-regulate Serrate expression in dorsal cells. The end
result is a localized activation of Notch in adjacent stripes of cells
that abut at the dorsal/ventral compartment boundary,
specifying these cells to become wing margin cells.
The presence or absence of Fringe appears to determine to which Notch
ligand wing cells respond. Thus, ventral cells (which normally lack
Fringe) are primed to respond to Serrate (expressed by dorsal cells).
Conversely, wing cells that express Fringe are refractory to Serrate
but primed to respond to Delta (expressed at high levels in ventral
cells in response to Serrate). As argued by Panin et al. (1997), the
dorsally restricted expression of Fringe is therefore responsible for
limiting the induction of Notch by Serrate and Delta to thin stripes of
dorsal and ventral cells that interact across the dorsoventral
compartment boundary. This Fringe-dependent bias is reinforced by an
unexpected property of Delta and Serrate signaling in the wing disc:
for unknown reasons, cells that express high levels of Delta or
Serrate themselves appear unable to receive the same ligand they
express and, hence, are further biased to receive the other ligand (see
also Micchelli et al. 1997). Panin et al. (1997) point out that this
seemingly complicated mechanism may be used to ensure the correct
placement of a source of morphogen to pattern a symmetrical structure.
fringe encodes a novel protein that is predicted to be
secreted but acts cell-autonomously (Irvine and Wieschaus 1994; Panin et al. 1997). It will be of interest to learn if Fringe modifies the
interaction of Notch with its ligands or the response of Notch upon the
binding of a subset of ligands.
Cell signaling: C. elegans vulval precursor cell fate patterning
The patterning of fates of the vulval precursor cells (VPCs) in
C. elegans (for review, see Greenwald 1997; Kornfeld 1997) may
exemplify the use of a heterologous cell signaling mechanism to
superimpose a bias on a LIN-12/Notch-mediated process
(Fig. 5). In wild-type hermaphrodites, three
hypodermal cells, P5.p, P6.p, and P7.p, are VPCs and generate the cells
that form the vulva. These cells lie under the somatic gonad, with P6.p
centered under the anchor cell. Although P5.p, P6.p, and P7.p all have similar developmental potential, they always adopt a pattern of fates
that may be represented as 2°-1°-2° (Sulston and White 1980; Sternberg and Horvitz 1986). This pattern primarily reflects the activity of two different signaling pathways. One pathway involves an
inductive signal from the anchor cell that is transduced by the LET-23
receptor tyrosine kinase in the vulval precursor cells (for review, see
Kornfeld 1997). It is not clear whether the inductive signal is
spatially graded (see Katz et al. 1995) or restricted so that only P6.p
is induced (see Koga and Ohshima 1995; Simske and Kim 1995). The other
pathway involves a LIN-12-mediated lateral signal thought to occur
between neighboring vulval precursor cells (Sternberg 1988).
|
lin-12 transcription appears to remain uniform during VPC
specification, suggesting that a transcription-based feedback mechanism analogous to that used in the AC/VU decision is not used
to establish the pattern of vulval fates (Wilkinson and Greenwald
1995). Instead, reception of the inductive signal appears to be an
important influence on lateral signaling: In let-23 genetic
mosaics, a let-23(
) VPC adopts the 2° fate when
adjacent to a let-23(+) VPC (Koga and Ohshima 1995; Simske
and Kim 1995). This observation suggests that inductive signaling
(whether spatially graded or not) appears likely to govern the
expression or activity of the ligand for LIN-12. Furthermore, inductive
signaling appears to cause a specific down-regulation of LIN-12 protein
accumulation (D. Levitan and I. Greenwald, unpubl.). Thus, a
combination of ligand up-regulation (transcriptional or
post-transcriptional) and receptor down-regulation (post-transcriptional) may bias P6.p to adopt the 1° fate.
It is interesting to note that reduction in the activity of
rhomboid, which potentiates the activity of the
Drosophila EGF receptor, causes reduced expression of
Delta during wing vein development (Sturtevant et al. 1993).
This appears to be another example of a linkage of EGF receptor
activity to DSL ligand expression, although the effect of EGF receptor
activity in this case is the opposite effect of what appears to occur
during C. elegans VPC specification.
| |
Other developmental roles for LIN-12/Notch proteins |
|---|
Most work on LIN-12/Notch proteins has been
concerned with their roles in mediating cell fate decisions. However,
in Drosophila, Notch and its ligand Delta have been
implicated in axon extension in defined neurons (Giniger et al. 1993).
Recently, Giniger (1998) has shown that axon aberrations associated
with reduced Notch activity do not appear to reflect
underlying changes in cell identity. He also showed that Notch is
expressed in growth cones in primary Drosophila neurons
cultured in vitro (Giniger 1998). LIN-12/Notch proteins
are also found in mature neurons in Drosophila (Fehon et al.
1991) and C. elegans (D. Levitan and I. Greenwald, unpubl.), raising the possibility of other functions in neural development.
Another proposed role for Notch is to sequester Dishevelled, a protein
that functions in the Wingless/Wnt signal transduction pathway (Axelrod et al. 1996). In Drosophila, the Wingless
(Wnt) and Notch signaling pathways both have input into many of the same cell fate decisions (e.g., Couso and Martinez Arias 1994; Gonzalez-Gaitan and Jaeckle 1995; Rulifson and Blair 1995). The dishevelled gene displays genetic interactions with
Notch as well as with wingless. Because there is some
evidence that Dishevelled may interact with the intracellular domain of
Notch, it has been proposed that Notch functions to sequester
Dishevelled; this role for Notch is therefore different from that
proposed for Notch signal transduction in cell fate decisions (see below).
| |
The mechanism of LIN-12/Notch signal transduction |
|---|
It is now generally accepted that LIN-12/Notch
proteins function as receptors that are activated by the binding of DSL
ligands. Genetic studies have suggested that receptor self-association may be involved in receptor activation (Greenwald and Seydoux 1990;
Heitzler and Simpson 1993).
The mechanism by which activated LIN-12/Notch proteins
transduce signals and alter gene expression has been unresolved. Recent results (Schroeter et al. 1998; Struhl and Adachi 1998) support an
unusual mechanism for signal transduction described below.
Three important observations are pertinent to models for the mechanism
of signal transduction: (1) Expression of just the intracellular domain
of LIN-12 and Notch causes phenotypes associated with
LIN-12/Notch activation (Lieber et al. 1993; Struhl et
al. 1993); (2) the free Notch intracellular domain is
nuclearly localized (Fortini et al. 1993; Lieber et al. 1993; Struhl et
al. 1993; Kopan et al. 1994; Nye et al. 1994); (3) the intracellular
domains of LIN-12/Notch proteins physically interact with
proteins that function in the nucleus to regulate gene expression. The
best characterized of these proteins are the site-specific DNA-binding protein Su(H) in Drosophila, LAG-1 in C. elegans, and
CBF1 (or RBPJ
) in vertebrates. Genetic studies in
Drosophila and C. elegans have established that
Su(H)/LAG-1 is involved in
LIN-12/Notch-mediated reception of intercellular signals
(see Schweisguth and Posakony 1994; Bailey and Posakony 1995;
Lecourtois and Schweisguth 1995; Christensen et al. 1996). There is
also evidence that LIN-12 interacts physically and functionally with
EMB-5, a C. elegans protein that is similar in sequence to a
yeast protein that controls chromatin structure (Hubbard et al. 1996).
Three classes of signal transduction mechanisms have been proposed. One
model is that LIN-12/Notch proteins act as a passive tether to keep LAG-1/Su(H) out of the nucleus and that
ligand binding induces release and nuclear import of these factors
(Fortini and Artavanis-Tsakonas 1994). This model accounts for the
physical interaction between Notch and Su(H) as well as evidence that
Notch can sequester Su(H) at the membrane under some conditions
(Fortini and Artavanis-Tsakonas 1994). However, the passive tether
model is incompatible with the genetic data. For example, the
Notch phenotype is similar to the Su(H)
phenotype; however, acccording to the passive tether model, Su(H) should enter the nucleus and be active in the absence of Notch. Furthermore, there is evidence that the subcellular localization of
Su(H) does not change upon Notch activation in vivo (Gho et al. 1996).
A more conventional mechanism of signal transduction could also be
invoked. The JAK-STAT signaling system, which involves the
modification of a transcription factor by a ligand-activated receptor
tyrosine kinase, is a useful analogy (Darnell 1997). Thus,
LIN-12/Notch may interact with LAG-1/Su(H)
at the cell surface; in response to ligand binding, Su(H) (or another
associated factor) is modified, dissociates from the receptor,
translocates to the nucleus, and regulates expression of downstream
target genes. At this time, there is no evidence for any enzymatic
activity of the LIN-12/Notch intracellular domain or for
a physically associated protein that might provide such a modifying activity.
A different model for the mechanism of signal transduction is that
ligand binding leads to cleavage and nuclear translocation of the
LIN-12/Notch intracellular domain in a complex with
transcription factors (Fig. 6). This model was first
proposed to account for the observation that the intracellular domain
of Notch is nuclearly localized and behaves like an activated receptor
(Lieber et al. 1993; Struhl et al. 1993). It is consistent with the
findings that the intracellular domains of LIN-12/Notch
interact with transcription factors and studies in cultured cells
suggesting that the Notch intracellular domain potentiates the
transcriptional activation of target genes by Su(H)/CBF1
(Jarriault et al. 1995; Chen et al. 1997; Eastman et al. 1997).
However, under normal in vivo circumstances, the intracellular domain
has not been visualized by antibody staining in the nuclei of cells
undergoing LIN-12/Notch-mediated signaling.
|
Recent work in Drosophila has provided evidence in support of
ligand-dependent nuclear access and a role for the intracellular domain
in regulating transcription of target genes in vivo. Struhl and Adachi
(1998) reasoned that if the mechanism of Notch signal transduction
depends on nuclear import of the intracellular domain, the amount of
the intracellular domain that accumulates in the nucleus must be very
low (or else it would be visible using conventional methods). They
therefore devised a sensitive assay for nuclear access by inserting a
GAL4-VP16 transcription factor domain into the intracellular domain of
Notch and used this method to demonstrate ligand-dependent nuclear
access as assayed by both UAS-lacZ reporter gene output and
phenotypic rescue. Furthermore, they showed that nuclear access is
important for signal transduction by manipulating the subcellular
localization of the intracellular domain with sequences that target to
the membrane (eliminating activity) or nucleus (potentiating activity).
Finally, they showed that the intracellular domain appears to
participate in a transcription complex because adding a transcriptional
activator domain to the intracellular domain of the full-length
receptor promotes Notch activity, whereas adding
transcriptional repressor domains blocks Notch activity.
The results of Struhl and Adachi (1998) provide strong evidence for in
vivo nuclear access and action of the intracellular domain of Notch. In
principle, nuclear access could be afforded by ligand-dependent
cleavage and release of the intracellular domain, analogous to the
events that release sterol regulatory element binding proteins (SREBPs)
from a transmembrane protein in response to cholesterol depletion (for
review, see Brown and Goldstein 1997). Alternatively, an unprecedented
trafficking event might enable the intracellular domain of activated
Notch to gain access to the nucleus while still attached to the
remainder of the protein. Recent work using mammalian tissue culture
cells has provided biochemical evidence for a ligand-dependent cleavage event that releases the Notch intracellular domain to function in the
nucleus (Schroeter et al. 1998). The cleavage model presented in Figure
6 is a synthesis of the available functional and biochemical data.
| |
Other influences on LIN-12/Notch activity |
|---|
Genetic analysis in C. elegans and Drosophila has identified many genes that influence LIN-12/Notch activity. Some of these genes have been defined by mutations that result in phenotypes associated with defects in LIN-12/Notch signaling. Other genes have been defined by mutations that suppress or enhance mutations in the LIN-12/Notch signaling pathway. Some genes have been defined by both approaches.
Many Drosophila genes that influence Notch activity
have been characterized. Several encode nuclear proteins that may be
involved in signal transduction, including mastermind (Smoller
et al. 1990), neuralized (Boulianne et al. 1991),
groucho (see Paroush et al. 1994; Fisher et al. 1996), and
strawberry notch (Majumdar et al. 1997). The big
brain gene encodes a protein that resembles an ion channel and
potentiates Notch activity (Doherty et al. 1997), and
deltex encodes a cytoplasmic protein that interacts with the intracellular domain of Notch (Matsuno et al. 1995). The scabrous gene encodes a secreted glycoprotein that influences Notch activity during eye
development (Baker and Zitron 1995; Lee et al. 1996).
As described above, Su(H) is a transcription factor and a key signal
transducing element in many cell fate decisions. Hairless influences
Notch signaling by negatively regulating the activity of the
Su(H) (Schweisguth and Posakony 1994; Bang and Posakony 1995). Hairless
interacts with Su(H) via a direct protein-protein interaction and
interferes with the DNA-binding activity of Su(H) (Brou et al. 1994).
The bHLH proteins of the E(Spl) complex appear to be direct
targets of Su(H) (Bailey and Posakony 1995; Lecourtois and Schweisguth 1995).
In some cases, the results of genetic and molecular characterization suggest other factors that influence cell fate decisions mediated by LIN-12/Notch signaling. Here, potential effects of protein turnover are considered, as suggested by studies of the C. elegans sel-1 and sel-10 genes. Also considered are aspects of protein processing and trafficking, based on studies of the Drosophila kuzbanian (kuz) gene and its C. elegans homolog, sup-17, and by studies of the C. elegans sel-12 and hop-1 presenilin genes.
Protein turnover
Studies of the C. elegans sel-1 and sel-10
genes, both negative regulators of lin-12 activity, have
suggested that protein turnover may be important in
LIN-12/Notch mediated cell fate decisions (Grant and
Greenwald 1996, 1997; Hubbard et al. 1997). SEL-1 is similar to the
Saccharomyces cerevisiae HRD3 protein, which has been
implicated in turnover of the membrane protein HMG CoA reductase (Hampton et al. 1996). SEL-10 is a member of the
F-box/WD40 repeat-containing protein family, members of
which have been shown to target proteins for ubiquitination and
turnover (for review, see Hoyt 1997).
Constitutive turnover or ligand-induced down-regulation of
LIN-12/Notch proteins may be important for cell fate
decisions to occur normally (discussed in Grant and Greenwald 1997;
Hubbard et al. 1997). For example, in the AC/VU decision,
Z1.ppp and Z4.aaa initially signal each other: in the absence of
turnover or down-regulation of LIN-12, this initial signaling might
create sufficient activated receptor so that both cells would achieve
the threshold value of effector activity; in effect, rapid turnover
would limit the output from a single ligand-receptor interaction.
Furthermore, in this decision, signaling from activated receptor would
persist in the absence of down-regulation, masking the effects of the differential transcription underlying the feedback mechanism. Receptor
turnover may also be necessary in cases where Notch is used
for successive decisions, such as in the specification of fates during
Drosophila eye development (Cagan and Ready 1989), so that
activated receptors are cleared before the next signaling event occurs.
Protein processing and trafficking
LIN-12/Notch proteins (and their ligands) are
transmembrane proteins that must be properly processed, modified, and
sorted. LIN-12/Notch proteins appear to be
proteolytically cleaved in the extracellular domain to yield a
heterodimer composed of the amino- and carboxy-terminal cleavage
products (Crittenden et al. 1994; Blaumueller et al. 1997; Pan and
Rubin 1997). This proteolytic cleavage event is proposed to occur in
intracellular vesicles and to be important for Notch trafficking to the
cell surface (Blaumueller et al. 1997).
Several genes that influence LIN-12/Notch activity appear to influence processing and trafficking. With the exception of the putative ligand-induced cleavage event, it is not known whether various steps in LIN-12/Notch processing and trafficking occur constitutively or are points of regulation during cell fate decisions.
A metalloprotease of the ADAM family
Kuz in Drosophila and
SUP-17 in C. elegansfacilitates
lin-12/Notch signaling (Rooke et al. 1996; Pan
and Rubin 1997; Sotillos et al. 1997; Tax et al. 1997; Wen et al.
1997). The genetic interaction between sup-17/kuz and
lin-12/Notch requires the
LIN-12/Notch extracellular domain (Pan and Rubin 1997;
Sotillos et al. 1997; Wen et al. 1997). Pan and Rubin (1997) have
presented evidence that Kuz is required in Drosophila for the
proteolytic processing of Notch into two fragments consistent with a
cleavage event in the extracellular domain. It is not known whether the
LIN-12/Notch extracellular domain is a substrate of the
SUP-17/Kuz protease or whether SUP-17/Kuz is
involved in maturation or activation of LIN-12/Notch proteins.
Proteins of the presenilin family have also been identified as
influences on LIN-12/Notch signaling. In humans,
presenilins were identified by mutations that cause Alzheimer's
disease (for review, see Schellenberg 1995). A connection between
presenilin activity and LIN-12/Notch signaling was first
suggested by the finding that the C. elegans sel-12 presenilin
facilitates the activity of lin-12 and glp-1 (Levitan
and Greenwald 1995), and supported by the observation that reducing the
activity of both C. elegans presenilins causes phenotypes
associated with the absence of lin-12 and glp-1
activity (Li and Greenwald 1997). The functional relationship between
presenilin and Notch activity appears to have been evolutionarily
conserved, because targeted disruption of the mouse PS1 gene
causes striking phenotypes associated with reduced Notch
activity (Shen et al. 1997; Wong et al. 1997).
Presenilins are multipass transmembrane proteins, and their mechanism
of function is not known. A recent study has suggested that PS1 is
required for a cleavage event during amyloid precursor protein
processing (DeStrooper et al. 1998), although whether presenilins
promote cleavage per se or a trafficking event necessary for cleavage
is not known. The effect of presenilins on amyloid precursor protein
processing is intriguing in view of the proteolytic processing events
that LIN-12/Notch proteins undergo. Perhaps presenilins
facilitate LIN-12/Notch activity by promoting one or more
proteolysis events associated with LIN-12/Notch
maturation or activation.
| |
Conservation of LIN-12/Notch form and function in vertebrates |
|---|
There is every reason to believe that precedents established from
studies in invertebrates will be directly applicable to vertebrates.
Many of the components that have been identified in invertebrates,
principally by genetic methods, have also been found in vertebrates
(see Table 1). Furthermore, the available functional and expression
data indicate that the vertebrate Notch pathway plays similar roles in
mediating cell-cell interactions that specify cell fate (for review,
see Gridley 1997). For example, activity of the Xenopus ligand
X-Delta-1 appears to control lateral specification of neurons, as
ectopic X-Delta-1 activity inhibits the formation of primary neurons,
and a dominant-negative form of X-Delta-1 promotes the formation of
excess primary neurons (Chitnis et al. 1995). It is likely that further
identification of components of the signal transduction mechanism and
of other influences on LIN-12/Notch signaling in
invertebrates will continue to inform studies of Notch signaling in
vertebrate development.
| |
Acknowledgments |
|---|
I thank Gary Struhl, Gerald Siu, and members of my laboratory for comments on this manuscript. I am an Associate Investigator of the Howard Hughes Medical Institute.
| |
Footnotes |
|---|
1 E-MAIL greenwald{at}cuccfa.ccc.columbia.edu; FAX (212) 305-1721.
| |
References |
|---|
|
Top
Introduction
References
|
|---|
, by direct protein-protein interaction with Drosophila Hairless.
Genes & Dev.
8:
2491-2503 and NOTCH.
J. Biol. Chem.
272:
14110-14114
and m 
expression.
Mol. Cell. Biol.
17:
5620-5628[Abstract].
