|
|
|
|
|
|
|
||||
| Home Help Feedback Subscriptions Archive Search Table of Contents | ||||
First published online 22 February 2006
doi: 10.1242/dev.02284
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
1 Vertebrate Development Laboratory, Cancer Research UK London Research
Institute, 44 Lincoln's Inn Fields, London WC2A 3PX, UK.
2 Hereditary Hearing Group, The Wellcome Trust Sanger Institute, Wellcome Trust
Genome Campus, Cambridge, CB10 1SA, UK.
3 Department of Immunology, Tokai University School of Medicine, Bohseidai,
Isehara, Kanagawa 259-1193, Japan.
Author for correspondence (e-mail: julian.lewis{at}cancer.org.uk)
Accepted 16 January 2006
 |
SUMMARY |
|---|
 TOP SUMMARY INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key words: Notch, Jagged1 (Jag1), Delta1 (Dll1), p27Kip1 (Cdkn1b), Ear, Hair cells, Lateral inhibition, Lateral induction, Mouse, Cochlea
|
INTRODUCTION |
|---|
|
TOPSUMMARYINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
;
Lai, 2004). In this way, Notch
signalling gives each cell the means to control gene expression in its
immediate neighbours. Feedback in this control machinery, such that the
strength of the signal a cell receives affects the strength of the signal it
delivers, can generate striking spatial and temporal patterns of gene
expression within a multicellular population
(Lewis, 1998). Thus, in the
phenomenon of lateral inhibition, activation of Notch in a given cell
diminishes the ability of that cell to produce functional ligands that can
activate Notch in the neighbours (Heitzler
and Simpson, 1991). A cell that signals more strongly thereby
causes its neighbours to signal more weakly, and the effect is to amplify
differences between adjacent cells. If the feedback is sufficiently steep, the
predicted outcome is a mosaic of cells in sharply different states
(Collier et al., 1996): cells
that express functional Notch ligands alternate with cells that do not; the
cells that express ligands escape Notch activation because their neighbours do
not express ligands, and vice-versa.
Completely different behaviour is predicted if Notch activation regulates
ligand expression in an opposite way - that is, if Notch activation in a given
cell increases the ability of that cell to produce functional Notch ligands,
as occurs, for example, at the Drosophila wing margin
(Bray, 1998;
de Celis and Bray, 1997) and in
some vertebrate cells (Ross and Kadesch,
2004). In this case, which we have called lateral induction
(Eddison et al., 2000;
Lewis, 1998), the effect is
cooperation between neighbours, instead of competition. A cell that expresses
Notch ligands strongly will make its neighbours do the same. Notch signalling
will tend to intensify, maintain and extend the domain where Notch ligands are
produced and Notch is activated.
The sensory patches in the vertebrate inner ear seem at first sight a
beautiful example of lateral inhibition. They consist of a mosaic of sensory
hair cells and supporting cells, generated from prosensory precursor cells
that express Notch and have the ability to differentiate in either way
(Fekete et al., 1998). At the
onset of differentiation, the nascent hair cells switch on expression of two
Notch ligands, Delta1 (Dll1) and Jagged2 (Jag2, in mammalian terminology)
(Adam et al., 1998;
Lanford et al., 1999;
Morrison et al., 1999),
thereby apparently inhibiting the adjacent prosensory cells from becoming hair
cells and forcing them to become supporting cells instead. Loss of Jag2 in the
mouse allows hair cells to be produced in excess
(Lanford et al., 1999), though
the effect is mild, suggesting that Jag2 and Dll1 cooperate, in
quasi-redundant fashion, to deliver lateral inhibition to the neighbours of
hair cells. Similar but stronger effects are seen in zebrafish with a mutation
in the deltaA gene (Riley et al.,
1999). In the zebrafish mind bomb mutant, where Notch
signalling is still more severely defective
(Haddon et al., 1998;
Jiang et al., 1996;
Schier et al., 1996), all the
cells in the prosensory patches differentiate as hair cells instead of
supporting cells and the expression of the homologs of Dll1 and
Jag2 is greatly increased, as the lateral inhibition hypothesis would
predict.
At the same time, in the same tissue, however, another Notch ligand, Jag1
(Serrate1 in the chick, SerrateA or Jagged1b in the zebrafish), shows
completely opposite behaviour, suggestive of lateral induction rather than
lateral inhibition. Expression of this ligand begins (in mouse and chick)
several days before hair-cell differentiation, and appears to mark the
prosensory patches, within which its level of expression is uniform from cell
to cell (Adam et al., 1998;
Morrison et al., 1999).
Subsequently, Jag1 becomes restricted to the supporting cells, and its
expression in this subpopulation is strong, uniform, and sustained
(Stone and Rubel, 2000), even
though the supporting cells are in contact with one another and exposed to the
influence of Dll1 (transiently) and Jag2 (persistently) from neighbouring hair
cells. All this suggests that Jag1 expression is positively regulated
by activated Notch, so that the effects of Dll1 and Jag2 in neighbouring hair
cells combine with the effects of Jag1 in neighbouring supporting cells to
keep Jag1 levels high in each supporting cell. Experiments in the chick embryo
in which the Notch signalling pathway is artificially blocked
(Eddison et al., 2000) or
overactivated (Daudet and Lewis,
2005) indicate that activated Notch does indeed stimulate
expression of Jag1.
What then is the function of Jag1 in the inner ear? There are at least two
attractive suggestions, and they conflict. The procrastination hypothesis
proposes that Jag1, by keeping the level of Notch activation high in the
prosensory cells and, later, in the supporting cells, serves to prevent them
from differentiating prematurely or inappropriately into hair cells
(Eddison et al., 2000).
Indeed, in the zebrafish mind bomb mutant, where Notch signalling is
defective, the whole population of the prosensory patch differentiates into
hair cells prematurely (Haddon et al.,
1998). By contrast, the prosensory induction hypothesis proposes
that activation of Notch by Jag1 serves to put cells into a prosensory state
and is needed to make hair cell differentiation possible
(Adam et al., 1998;
Eddison et al., 2000;
Kiernan et al., 2001). Support
for this latter idea comes from our recent experiments in the chick. By
expressing NICD ectopically, we have found that Notch activity has
two different roles in the inner ear: at early stages it can induce formation
of a sensory patch (converting cells to a prosensory character), whereas at
late stages it mediates lateral inhibition, limiting the proportion of cells
within a prosensory patch that actually differentiate as hair cells
(Daudet and Lewis, 2005).
Homozygous knockout mutations of Jag1 and Dll1 in the
mouse are both lethal at early embryonic stages, before the ear has developed
(Hrabé de Angelis et al.,
1997; Xue et al.,
1999). Heterozygous Dll1 mutants have apparently normal
ears (Morrison et al., 1999).
Heterozygous Jag1 mutants show a mild reduction of the number of
outer hair cells in the cochlea but a mild increase in the number of inner
hair cells, as well as defects in their semicircular canals
(Kiernan et al., 2001;
Tsai et al., 2001).
To discover which, if any, of our hypotheses are correct, we need to see the homozygous knockout phenotypes in the ear. We have used a Cre-Lox strategy to achieve this. Loss of Jag1 does not cause premature differentiation but does lead to a severe deficit of sensory tissue, as predicted by the prosensory induction hypothesis. Loss of Dll1 has effects of an opposite type, though less severe, consistent with the predicted role as a mediator of lateral inhibition. Neither phenotype, however, is quite as we expected; in both cases, in different ways, it seems that the Notch ligands influence patterns of cell division as well as choices of differentiated fate.
|
MATERIALS AND METHODS |
|---|
|
TOPSUMMARYINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
;
Lai, 2004). Details for
Dll1flox are given by Hozumi et al.
(Hozumi et al., 2004). The
same strategy was used for Jag1flox (see Fig. S1 in the
supplementary material). Recombination of Jag1flox by Cre
is predicted to knock out gene function by removing the DSL exons (exons 3 and
4 for Dll1, 4 and 5 for Jag1) and introducing a frame shift
that generates a stop codon shortly after the start of exon 5 (for
Dll1) or 6 (for Jag1).
Mouse breeding
Mice were bred on a C57BL/6J background. The Foxg1Cre
strain, with Cre knocked into the Foxg1 locus in place of
the Foxg1-coding region, was a gift from Jean Hébert and Susan
McConnell, and was maintained as a heterozygous colony, as homozygotes are not
viable; genotyping was by PCR as described by Hébert and McConnell
(Hébert and McConnell,
2000). Both Jag1flox and
Dll1flox mice were homozygous viable. They were genotyped
by PCR as follows: for unrecombined Jag1flox, the primers
were TGAACTCAGGACAGTGCTC and GTTTCAGTGTCTGCCATTGC (flanking the 5' LoxP
site); for the recombined allele,
Jag1
flox, primers were
TGAACTCAGGACAGTGCTC and a primer targeted downstream of the 3' LoxP
site, ATAGGAGGCCATGGATGACT. For unrecombined Dll1flox,
primers were CACACCTCCTACTTACCTGA and GAGAGTACTGGATGGAGCAAG (flanking the
5' LoxP site); for the recombined allele,
Dll1flox, primers were
CACACCTCCTACTTACCTGA and a primer targeted downstream of the 3' LoxP
site, GGCGCTCAAAGGATATGGGA.
Mice carrying the conditional alleles were crossed with Foxg1Cre heterozygotes, and doubly heterozygous progeny were identified by PCR. These were then crossed with mice carrying one copy of a conditional allele, to produce litters containing conditional knockout mice (Foxg1Cre/+;Dll1flox/flox or Foxg1Cre/+;Jag1flox/flox) and sibling controls. For Jag1, out of 499 genotyped embryos collected at E16.5 to P0 from such matings, 37 had the conditional knockout genotype, implying some intrauterine mortality, as Mendelian ratios would give 62. For Dll1, 26 out of 214 embryos had the conditional knockout genotype - close to the Mendelian proportion.
Immunohistochemistry
Embryos were staged by age, taking the day of the vaginal plug as E0.5, and
were checked for morphological stage at the time of collection
(Kaufman, 1992;
Theiler, 1989). For serial
sections at E17.5, embryos were decapitated and their heads were fixed in 4%
paraformaldehyde in PBS for 1 hour, then embedded in 1.5% Lennox agar
(Gibco-BRL) with 3% sucrose and cryosectioned at 15 µm. For whole-mount
immunostaining of the cochlea, inner ears were dissected from the temporal
bone and a small opening was made in the apex of each cochlea before fixation
in 4% formaldehyde in PBS for 1 hour. The tissue was then permeabilised by
immersion for 1 hour in PBS with 0.3% Triton-X and 10% foetal calf serum,
followed by incubation overnight at 4°C with primary antibody in PBS with
0.1% Triton-X and 10% fetal calf serum.
Antibodies and reagents used were: Jag1 (C-20, Santa Cruz; 1/100), p27Kip1 (Cell Signalling; 1/100), calretinin (Chemicon; 1/1000) and Alexa A488-conjugated rabbit anti-goat and goat anti-rabbit as secondaries (Molecular Probes; 1/500). Alexa 594- and 633-congugated phalloidin (Molecular Probes; 1/100) and DAPI were applied to the sections with the secondary antibody.
For flat-mounting, whole-mount cochleas from E17.5 mice were cut into three, giving apical, middle and basal regions. Those from the younger embryos were left whole. Whole-mount and sectioned specimens were mounted in SlowFade (Molecular Probes) and imaged with a Zeiss LSM510 confocal microscope.
E10.5 embryos were formalin fixed, embedded in wax, serially sectioned at 8 µm and stained with 3A10 neurofilament antibody (generated by T. M. Jessell and J. Dodd, obtained from the Developmental Studies Hybridoma Bank, Iowa), with a biotin-conjugated rabbit anti-mouse secondary antibody (Jackson Immunoresearch Labs) and Haematoxylin counterstain.
Cell counts
Auditory hair cells and supporting cells were counted from whole mounts of
the cochlea stained with fluorescent phalloidin and (except in Jag1
knockouts) with Jag1 antibody. Hair cells were identified by their
phalloidin-stained hair bundles and the apical location of their nuclei in the
epithelium, and supporting cells by their Jag1 staining and the basal location
of their nuclei, as seen in z-series of confocal optical sections
spanning the full depth of the sensory epithelium. For each specimen and each
portion of cochlea, counts were made from three 100 µm lengths of the organ
of Corti, and the average was calculated.
Vestibular hair cells were counted on immunostained cryosections cut perpendicular to the plane of the saccular macula; three samples of 200 µm length were scored per individual. Hair cells were identified by the apical location of their nuclei in the epithelium and by their staining for calretinin.
For measurements of cochleovestibular ganglion volume at E10.5, the outlines of the ganglion were traced and cross-sectional areas measured (using ImageJ) on every second serial section. The areas were then added together and multiplied by twice the section thickness.
|
RESULTS |
|---|
|
TOPSUMMARYINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cre expressed under Foxg1 control recombines Jag1flox early in ear development
For our conditional knockout experiments, we generated lines of mice
carrying floxed alleles of Dll1
(Hozumi et al., 2004) and
Jag1 (see Materials and methods). We crossed these lines with mice
that had the Cre recombinase gene knocked into the Foxg1
locus, to produce Foxg1Cre/+;Jag1flox/flox
embryos. The Foxg1 promoter has been reported to drive expression of
Cre recombinase in the developing inner ear from E8.75, at the early otic
placode stage, 1 day earlier than the earliest reported expression of
Jag1 in the ear, resulting in efficient recombination of floxed
constructs throughout the otic vesicle at subsequent stages
(Hébert and McConnell,
2000; Mitsiadis et al.,
1997; Pirvola et al.,
2002). To check that recombination had occurred as expected, we
stained sections of Foxg1Cre/+;Jag1flox/flox
conditional knockout embryos and their control littermates with antibody
against Jag1 at E10.5, the earliest time at which expression of Jag1 is
normally detected with immunostaining. Complete sets of serial sections
through the left and right otocysts were analysed in this way for three
Jag1 conditional knockout mice and two littermate controls. In all
the conditional knockout mice, the patch of Jag1 antibody staining in the
ventral otocyst was completely lost (Fig.
1A,B). Loss of Jag1 antibody staining was confirmed at later
embryonic stages, at E15.5 and E17.5 (n=3), in wholemount cochlea
preparations (Fig. 1C,D) and in
cryosections of inner ears. We conclude that our conditional knockout strategy
efficiently eliminated expression of the targeted protein in the inner ear,
starting from a very early stage in ear development.
Sensory cells do not differentiate prematurely in the Jag1 conditional knockout cochlea
We wished first of all to test the procrastination hypothesis: that
Jag1 serves to prevent the premature production of hair cells by
activating Notch in all the cells of the prosensory patch. For this, we
focused on the cochlea. Hair-cell differentiation in the normal cochlea begins
around E14.5 in the mid-basal region, and a wave of differentiation proceeds
basally and apically over the next few days
(Lim and Anniko, 1985). By
E17.5 hair cells are seen in the basal and middle regions, but have not yet
differentiated in the apex. Effects on the timing of hair-cell production can
be sensitively detected as changes in this pattern. In addition, the final
arrangement of the auditory hair cells and supporting cells is precise and
predictable, making it easy to detect even a small abnormality in the numbers
of hair cells or supporting cells produced.
|
Outer hair cells are missing but inner hair cells are overproduced in the Jag1 conditional knockout cochlea
Although the timing of hair-cell differentiation appeared normal in the
Jag1 knockouts, the number and distribution of hair cells was greatly
altered. In the middle part of the cochlea at E17.5, the littermate controls
showed the normal stereotyped pattern of three rows of outer hair cells and
one row of inner hair cells (Fig.
2C); by contrast, the Jag1 knockouts showed just two
rather disorganized rows of hair cells
(Fig. 2D). The total number of
hair cells per 100 µm length of cochlea in this middle region was reduced
to 25.8±0.5 (mean±s.e.m., n=8) as compared with
63.5±1.5 (mean±s.e.m., n=10) hair cells in littermate
controls.
The reduction in the number of hair cells seen in the Jag1 conditional knockout cochlea did not, however, reflect a simple general decrease of hair cell production. Outer hair cells appeared to be lost completely, whereas the number of inner hair cells per 100 µm was almost doubled, to 25.8±0.5 (n=10) compared with 15.3±0.35 (n=8) for the controls. Two criteria indicated that the hair cells in the conditional knockout were all of inner hair cell character. First, they all expressed calretinin, which is normally seen in inner but not outer hair cells at this stage (Fig. 3A,B). Second, like normal inner hair cells, they all lay on the inner side of a distinctive band of supporting cells - the inner pillar cells (Fig. 3C,D). This was clear in all five cases for which appropriate z-series of images were collected.
|
The deficit of cochlear hair cells in the Jag1 conditional knockout reflects a failure of initial production, not subsequent degeneration
The mid-basal region of the cochlea is normally the earliest region of
hair-cell differentiation (Lim and Anniko,
1985). Thus, one interpretation for the gaps seen in the band of
hair cells in this part of the mutant cochlea at E17.5 is that some of the
hair cells initially produced had degenerated. To test this idea, we examined
the pattern of hair cells in the basal region two days earlier in development,
at E15.5 (data not shown). In control cochleas at this stage, when hair cell
differentiation has just begun, a single row of cells with upregulated actin,
the inner hair cells, could be seen in the most basal regions, while multiple
rows of hair cells were already visible in mid-basal regions. In the mutant
cochlea, there was no sign of hair cells in the most basal region, while hair
cells appeared already grouped in islands in the mid-basal region. The deficit
of hair cells is therefore unlikely to be due to hair-cell loss following
normal initial production.
|
Auditory hair cells are produced prematurely and in excess in the Dll1 conditional knockout
We analysed the conditional knockout phenotype for Dll1 in the
same way as for Jag1, focusing mainly on effects in the cochlea. In
the Dll1flox/flox conditional knockout
cochlea, in contrast to Jag1flox/flox, hair
cells differentiated prematurely and were clearly visible already in the apex
at E17.5, although none could yet be seen in this region in control
littermates (Fig. 5A-D).
Moreover, the hair cells in the apical cochlea were present in a huge excess.
Instead of the normal pattern of one row of inner hair cells and three rows of
outer hair cells, which become visible postnatally in the apex of the
wild-type cochlea, between 6 and 8 rows of outer hair cells had already formed
at E17.5 in the mutant, and the inner hair-cell row also appeared to have been
duplicated.
Similar but less extreme effects were seen in the rest of the cochlea. In
the middle and basal regions, there was one extra row of outer hair cells and
an increase in the number of inner hair cells
(Fig. 5E,F). Detailed counts in
the middle turn gave an average of 62±5.1 outer hair cells per 100
µm length of cochlea (mean±s.e.m., n=7) in the
Dll1 conditional knockout, arranged in four orderly rows, compared
with 46±1.7 (mean±s.e.m., n=6) in control mice,
constituting the normal three rows. The number of outer hair cells in this
region per 100 µm was thus increased in the Dll1 conditional
knockout by 
35%.
The premature and excessive production of hair cells was associated with a reduction or delay in the growth of the cochlea, which, viewed as a flat mount at E17.5, showed slightly fewer than the normal number of turns (about 1.5 turns in the mutant, as against 1.75 in the controls) and had a shorter length as measured along its outer curvature [3144±78 µm (n=6) when compared with 4454±56 µm (n=6) for littermate controls]. The extreme broadening of the sensory patch at the apex may thus be a result of premature differentiation, shortening the period of growth preceding differentiation and preventing the prosensory patch from elongating and narrowing in the normal way.
Supporting cells are also produced in excess in the Dll1 conditional knockout
The overproduction and premature differentiation of hair cells in the
Dll1 conditional knockout are effects one would predict if there was
a partial loss of lateral inhibition. One would then also expect the extra
hair cells to be produced at the expense of supporting cells: supporting cell
numbers should be decreased. Yet we saw no obvious deficit
(Fig. 5G,H). To check this
impression, we counted the numbers of hair cells and supporting cells in the
middle region of cochleas from E17.5 individuals (see Materials and methods).
Seven conditional knockout mice and six littermate controls were analysed in
this way. For the sake of precision, we focused on the outer hair cell region,
as this had well-marked borders - the outer one defined by the outer margin of
Jag1 expression, the inner defined by the row of pillar cells. (It
was difficult to make precise counts of supporting cells in the inner hair
cell region because the inner margin of Jag1 expression is rather diffuse.)
The number of Jag1-positive outer supporting cells - Deiters' cells plus outer
pillar cells - was 80±7.4 per 100 µm (mean±s.e.m.,
n=7) in the Dll1 conditional knockout cochleas, compared
with 70 ±4.8 (n=6) in the littermate controls - an increase of
14%. The ratio of supporting cells to hair cells was nevertheless reduced,
from 1.52±0.03 (mean±s.e.m., n=6) for the controls to
1.29±0.04 (mean±s.e.m., n=7) for the knockouts. This
difference is statistically significant (P<0.001,
t-test).
In the apex of the cochlea, where the overproduction of both hair cells and supporting cells was more obvious, the Jag1 stain was expanded to encompass the population of supernumerary hair cells (Fig. 5D). We found 84±7 hair cells per 100 µm and 170±15 supporting cells per 100 µm (mean±s.e.m., n=3) - a much larger proportion of supporting cells to hair cells than in the middle turn of the cochlea in either wild-type or mutant. No comparison could be made with control littermates as hair cell production has not yet occurred in apex of the normal cochlea at E17.5. Nevertheless, these counts imply that the additional hair cells are not produced at the expense of a proportionate reduction in the number of supporting cells.
|
|
This deficit of sensory epithelium could be the result of a failure of lateral inhibition during the genesis of neuroblasts from the early otocyst: loss of Dll1 at an early stage is expected to cause an excessive proportion of vestibular prosensory cells to adopt a neuronal fate and delaminate from the otic epithelium (see Discussion). To investigate this possibility, we examined the size of the rudiment of the cochleovestibular ganglion in Dll1 conditional knockout embryos and littermate controls at E10.5. From serial sections stained with neurofilament antibody, we estimated that this ganglion had a volume of 7.0±1.3 nl (mean±s.d., n=3) in the knockouts, compared with 4.2±0.2 nl (mean±s.d., n=3) in the controls. These figures are rough and should be viewed with caution, because at this stage the rudiments of the facial and cochleovestibular ganglia form a single confluent mass, and it was difficult to distinguish the cochleovestibular component. Nevertheless, they tend to support the suggestion that loss of Dll1 caused excessive numbers of cells to be diverted from an epithelial to a neuronal fate.
Loss of p27Kip1 does not underlie the Dll1 conditional knockout phenotype
The pattern of four rows of outer hair cells and almost two rows of inner
hair cells seen in the Dll1 conditional knockout mice is strikingly
similar to the pattern of supernumerary hair cells reported in cochleas with
null mutations of p27Kip1
(Chen and Segil, 1999;
Lowenheim et al., 1999).
p27Kip1 is a cyclin-dependent kinase inhibitor (CKI) that regulates
progress through the cell cycle (Sherr and
Roberts, 1999). It is expressed in the cochlea in the region of
the developing sensory patch before hair cell differentiation, as cells of the
patch exit the cell cycle (Chen et al.,
2002; Chen and Segil,
1999; Ruben,
1967). When p27Kip1 is deleted, cells of the
sensory patch continue to proliferate beyond E14.5, after the time of their
normal terminal mitosis (Chen and Segil,
1999; Lowenheim et al.,
1999). We wondered, therefore, whether a failure of
p27Kip1 expression, secondary to a defect of Notch
signalling, might explain the overproduction of hair cells and supporting
cells in the Dll1 knockout. In the wild-type cochlea at E14.5, a
broad band of p27Kip1-positive cells is seen by immunofluorescence
in the future hair-cell region. In the Dll1 conditional knockout
cochlea at this stage, the pattern of p27Kip1 immunofluorescence,
in all three specimens that we analysed, appeared almost exactly the same
(Fig. 7A,B). The abnormality of
hair-cell and supporting-cell numbers is therefore not likely to be the
consequence of a loss of p27Kip1. As we discuss below, it remains
possible nevertheless that altered patterns of proliferation play some part in
the Dll1 knockout phenotype.
|
|
DISCUSSION |
|---|
|
TOPSUMMARYINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
). Our
findings described here indicate that Jag1 is responsible for activating the
prosensory function of Notch, supporting previous suggestions from Jag1
heterozygotes (Kiernan et al.,
2001), while Dll1 mediates the function of Notch in lateral
inhibition. Moreover, it appears that the disturbances of Notch signalling may
also alter the pattern of sensory cell proliferation.
Partial failure of lateral inhibition explains the Dll1 knockout phenotype
In the case of Dll1, the finding that cochlear hair cells are
produced early and in excess is strongly reminiscent of the mind bomb
phenotype in zebrafish, where Notch signalling is defective and all cells in
the sensory patch develop as hair cells, and do so prematurely
(Haddon et al., 1998). This is
what one would expect if lateral inhibition is reduced
(Goriely et al., 1991). Taken
with what we know about Dll1 and Notch signalling in other systems, a failure
of lateral inhibition seems much the most likely cause of the Dll1
knockout phenotype. The overproduction of hair cells is relatively slight, but
this is easily accounted for by the presence of Jag2, which has been shown to
have similar expression in hair cells and similar action, creating redundancy
in the lateral inhibition mechanism
(Lanford et al., 1999).
More puzzling, however, is the finding that supporting cells are also produced in excess. This seems to clash with the simple lateral inhibition hypothesis. On the other hand, the ratio of supporting cells to hair cells is reduced, as the lateral inhibition hypothesis would predict. There are several possible interpretations.
), and it
could be followed by some cell rearrangement
(Goodyear and Richardson,
1997). ) have
found evidence of this in the cochlea of mice that are homozygous for a
knockout mutation of Jag2 combined with a hypomorphic allele of Dll1:
they see a phenotype that is similar to the one we have described above, but
more extreme, and they find by BrdU labelling that cell proliferation is
prolonged abnormally in the supporting-cell population. Thus, in both our
mutants and theirs, it seems likely that an initial failure of lateral
inhibition is followed by some compensatory proliferation of the residual
supporting cells.
The above suggestions are not mutually exclusive, and all of them postulate
that the primary effect of loss of Dll1 is a reduction of lateral inhibition.
How, then, are we to explain the loss of sensory maculae in the vestibular
part of the inner ear? In the zebrafish mind bomb mutant, hair cells
in the absence of supporting cells are extruded from the otic epithelium and
rapidly disappear (Haddon et al.,
1999); a similar process could have occurred in the Dll1
mutants, leading to loss of one or both of the maculae. Another interpretation
invokes the known early role of Dll1 in otic neurogenesis
(Adam et al., 1998;
Alsina et al., 2003). Early in
ear development, a subset of cells in the anteroventral region of the otocyst
become determined as neuroblasts and delaminate from the epithelium; when
lateral inhibition fails, as a result of failure of Notch signalling, an
excessive proportion of cells undergo this fate
(Haddon et al., 1998). Our
measurements of the volume of the cochleovestibular ganglion suggest that this
indeed occurred in our Dll1 knockouts. The diversion of cells towards
a neuronal fate would be expected to deplete the prosensory population in the
anteroventral otic epithelium, which is thought to give rise to the maculae
(Adam et al., 1998;
Morsli et al., 1998;
Satoh and Fekete, 2005).
Jag1 mediates a prosensory function of Notch
At the outset, we considered the hypothesis that Jag1, by keeping Notch
activated throughout each prosensory patch, served to prevent premature
differentiation. We can now rule out this possibility: where hair cells were
produced in the Jag1 knockout cochlea, they were produced at the
normal time. This contrasts with the behaviour seen in Dll1
knockouts, where hair cells at the apex of the cochlea differentiated at least
a day prematurely. It seems, therefore, that the role of Jag1, in the cochlea
at least, is not the inhibition of hair cell differentiation, but something
else.
Previous studies have described the heterozygous Jag1
loss-of-function phenotype in the ear: numbers of outer hair cells were
somewhat reduced, and numbers of inner hair cells somewhat increased
(Kiernan et al., 2001;
Tsai et al., 2001). Our
findings for homozygous Jag1 knockouts are similar but more extreme:
outer hair cells are lost entirely, while the number of inner hair cells is
roughly doubled. How is this curious combination of effects to be
explained?
We have given evidence against interpretations in terms of hair-cell degeneration or conversion of outer hair cells to an inner hair-cell character. The obvious suggestion, rather, is that the observed loss of outer hair cells in the Jag1 knockout reflects a need for Jag1 as activator of the prosensory function of Notch in the outer hair cell region. The effects of loss of Jag1 in the vestibular system can be explained in a similar way. Some sensory patches, the anterior and posterior cristae, were entirely missing in the Jag1 conditional knockout. Others, the horizontal crista and the utricular macula, appeared to be present as epithelial thickenings but lacked hair cells. Finally, the saccular macula was apparently unaffected by loss of Jag1. This pattern of vestibular defects is quite different from that seen in the Dll1 conditional knockouts, and it can be easily explained if we consider that the role of Jag1 is to activate the prosensory function of Notch and thereby maintain or extend the set of cells competent for sensory differentiation. Some prosensory regions are evidently dependent on this action of Jag1, others are not.
Jag1, like Notch, may have distinct early and late functions
If Jag1 contributes to the induction of prosensory patches, how are we to
explain the increase of inner hair cell numbers when Jag1 is lost? One
possibility is that Jag1, like Notch, has a late function in lateral
inhibition that is distinct from its early, inductive, prosensory function: by
activating Notch at later stages, during hair cell determination, it may
contribute to lateral inhibition in concert with Dll1 and Jag2. Although loss
of the early (inductive) effect would lead to loss of sensory epithelium, loss
of the late (inhibitory) effect could lead to overproduction of hair cells in
the patches that remain.
Another interpretation is suggested, however, by our unexpected finding
that loss of Jag1 leads to a failure of p27Kip1 expression
in the cochlea. As p27Kip1 is an inhibitor of cell cycling, its
loss is predicted to permit increased numbers of cell divisions, leading to
increased production of both hair cells and supporting cells. Indeed, the
p27Kip1 knockout (Chen
and Segil, 1999; Lowenheim et
al., 1999) has an excess of both inner and outer hair cells. Thus,
our observations may perhaps be most plausibly explained by the hypothesis
that Jag1 is needed, first, to extend the prosensory state by lateral
induction beyond the inner hair cell region into the outer hair cell region,
and, second, to enable expression of p27Kip1, thereby
helping to terminate cell proliferation within the prosensory patch when the
time comes for differentiation.
One of the key issues for future research will be to determine whether the
early (prosensory induction) and late (lateral inhibition and proliferation
control) functions of Jag1 and Dll1, with their very different consequences
for hair-cell production, are mediated by the same or different members of the
Notch protein family (Shimizu et al.,
1999; Shimizu et al.,
2000). Other Notch genes besides Notch1 may be expressed
in the mouse ear - indeed, we have preliminary evidence that Notch3
is expressed there (data not shown) - and they may be specialized as between
these two roles. Certainly at least two different members of the Hes family of
downstream mediators of Notch signalling - Hes1 and Hes5 -
are expressed in the ear, in subtly different patterns, and with different
mutant phenotypes that could reflect differential involvement in signals
delivered by Jag1 and Delta; the Hes1 knockout, for example, has an
excess of inner hair cells in the cochlea, while the Hes5 knockout
has an excess of outer hair cells (Zheng
et al., 2000; Zine et al.,
2001).
Our observations show that the different Notch ligands have radically different but interlocking functions in the ear, and that these functions can be interpreted in terms of the different roles of Notch signalling at successive stages in the induction and internal patterning of the sensory patches and the different rules by which Notch activity regulates the expression of each ligand. In all sorts of other tissues, from the central nervous system to the vasculature and the epidermis, multiple Notch ligands are expressed in closely correlated but distinctive patterns. The lessons learnt from the ear may thus be important for Notch signalling in other systems.
|
ACKNOWLEDGMENTS |
|---|
|
Footnotes |
|---|
Supplementary material for this article is available at http://dev.biologists.org/cgi/content/full/133/7/1277/DC1
|
REFERENCES |
|---|
|
TOPSUMMARYINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Adam, J., Myat, A., Le Roux, I., Eddison, M., Henrique, D., Ish-Horowicz, D. and Lewis, J. (1998). Cell fate choices and the expression of Notch, Delta and Serrate homologues in the chick inner ear: parallels with Drosophila sense-organ development. Development 125,4645 -4654.[Abstract]
Alsina, B., Giraldez, F. and Varela-Nieto, I. (2003). Growth factors and early development of otic neurons: interactions between intrinsic and extrinsic signals. Curr. Top. Dev. Biol. 57,177 -206.[Medline]
Baron, M. (2003). An overview of the Notch signalling pathway. Semin. Cell Dev. Biol. 14,113 -119.[CrossRef][Medline]
Bray, S. (1998). Notch signalling in Drosophila: three ways to use a pathway. Semin. Cell Dev. Biol. 9,591 -597.[CrossRef][Medline]
Chen, P. and Segil, N. (1999). p27(Kip1) links cell proliferation to morphogenesis in the developing organ of Corti. Development 126,1581 -1590.[Abstract]
Chen, P., Johnson, J. E., Zoghbi, H. Y. and Segil, N. (2002). The role of Math1 in inner ear development: uncoupling the establishment of the sensory primordium from hair cell fate determination. Development 129,2495 -2505.
Collier, J. R., Monk, N. A., Maini, P. K. and Lewis, J. H. (1996). Pattern formation by lateral inhibition with feedback: a mathematical model of delta-notch intercellular signalling. J. Theor. Biol. 183,429 -446.[CrossRef][Medline]
Daudet, N. and Lewis, J. (2005). Two
contrasting roles for Notch activity in chick inner ear development:
specification of prosensory patches and lateral inhibition of hair-cell
differentiation. Development
132,541
-551.
de Celis, J. F. and Bray, S. (1997). Feed-back mechanisms affecting Notch activation at the dorsoventral boundary in the Drosophila wing. Development 124,3241 -3251.[Abstract]
Eddison, M., Le Roux, I. and Lewis, J. (2000).
Notch signaling in the development of the inner ear: lessons from Drosophila.
Proc. Natl Acad. Sci. USA
97,11692
-11699.
Fekete, D. M., Muthukumar, S. and Karagogeos, D.
(1998). Hair cells and supporting cells share a common progenitor
in the avian inner ear. J. Neurosci.
18,7811
-7821.
Goodyear, R. and Richardson, G. (1997). Pattern
formation in the basilar papilla: evidence for cell rearrangement.
J. Neurosci. 17,6289
-6301.
Goriely, A., Dumont, N., Dambly-Chaudiere, C. and Ghysen, A. (1991). The determination of sense organs in Drosophila: effect of the neurogenic mutations in the embryo. Development 113,1395 -1404.[Abstract]
Haddon, C., Jiang, Y. J., Smithers, L. and Lewis, J. (1998). Delta-Notch signalling and the patterning of sensory cell differentiation in the zebrafish ear: evidence from the mind bomb mutant. Development 125,4637 -4644.[Abstract]
Haddon, C., Mowbray, C., Whitfield, T., Jones, D., Gschmeissner, S. and Lewis, J. (1999). Hair cells without supporting cells: further studies in the ear of the zebrafish mind bomb mutant. J. Neurocytol. 28,837 -850.[CrossRef][Medline]
Hébert, J. M. and McConnell, S. K. (2000). Targeting of cre to the Foxg1 (BF-1) locus mediates loxP recombination in the telencephalon and other developing head structures. Dev. Biol. 222,296 -306.[CrossRef][Medline]
Heitzler, P. and Simpson, P. (1991). The choice of cell fate in the epidermis of Drosophila. Cell 64,1083 -1092.[CrossRef][Medline]
Hozumi, K., Negishi, N., Suzuki, D., Abe, N., Sotomaru, Y., Tamaoki, N., Mailhos, C., Ish-Horowicz, D., Habu, S. and Owen, M. J. (2004). Delta-like 1 is necessary for the generation of marginal zone B cells but not T cells in vivo. Nat. Immunol. 5, 638-644.[CrossRef][Medline]
Hrabé de Angelis, M., McIntyre, J., 2nd and Gossler, A. (1997). Maintenance of somite borders in mice requires the Delta homologue DII1. Nature 386,717 -721.[CrossRef][Medline]
Jacobsen, T. L., Brennan, K., Arias, A. M. and Muskavitch, M. A. (1998). Cis-interactions between Delta and Notch modulate neurogenic signalling in Drosophila. Development 125,4531 -4540.[Abstract]
Jiang, Y. J., Brand, M., Heisenberg, C. P., Beuchle, D., Furutani-Seiki, M., Kelsh, R. N., Warga, R. M., Granato, M., Haffter, P., Hammerschmidt, M. et al. (1996). Mutations affecting neurogenesis and brain morphology in the zebrafish, Danio rerio. Development 123,205 -216.[Abstract]
Kaufman, M. H. (1992). The Atlas of Mouse Development. London: Academic Press.
Kiernan, A. E., Ahituv, N., Fuchs, H., Balling, R., Avraham, K.
B., Steel, K. P. and Hrabé de Angelis, M. (2001). The
Notch ligand Jagged1 is required for inner ear sensory development.
Proc. Natl Acad. Sci. USA
98,3873
-3878.
Kiernan, A. E., Cordes, R., Kopan, R., Gossler, A. and Gridley,
T. (2005). The Notch ligands DLL1 and JAG2 act
synergistically to regulate hair cell development in the mammalian inner ear.
Development 132,4353
-4362.
Lai, E. C. (2004). Notch signaling: control of
cell communication and cell fate. Development
131,965
-973.
Lanford, P. J., Lan, Y., Jiang, R., Lindsell, C., Weinmaster, G., Gridley, T. and Kelley, M. W. (1999). Notch signalling pathway mediates hair cell development in mammalian cochlea. Nat. Genet. 21,289 -292.[CrossRef][Medline]
Lewis, J. (1998). Notch signalling and the control of cell fate choices in vertebrates. Semin. Cell Dev. Biol. 9,583 -589.[CrossRef][Medline]
Lim, D. J. and Anniko, M. (1985). Developmental morphology of the mouse inner ear. A scanning electron microscopic observation. Acta Otolaryngol. Suppl. 422, 1-69.[Medline]
Lowenheim, H., Furness, D. N., Kil, J., Zinn, C., Gultig, K.,
Fero, M. L., Frost, D., Gummer, A. W., Roberts, J. M., Rubel, E. W. et al.
(1999). Gene disruption of p27(Kip1) allows cell proliferation in
the postnatal and adult organ of corti. Proc. Natl Acad. Sci.
USA 96,4084
-4088.
Mitsiadis, T. A., Henrique, D., Thesleff, I. and Lendahl, U. (1997). Mouse Serrate-1 (Jagged-1): expression in the developing tooth is regulated by epithelial-mesenchymal interactions and fibroblast growth factor-4. Development 124,1473 -1483.[Abstract]
Morrison, A., Hodgetts, C., Gossler, A., Hrabé de Angelis, M. and Lewis, J. (1999). Expression of Delta1 and Serrate1 (Jagged1) in the mouse inner ear. Mech. Dev. 84,169 -172.[CrossRef][Medline]
Morsli, H., Choo, D., Ryan, A., Johnson, R. and Wu, D. K.
(1998). Development of the mouse inner ear and origin of its
sensory organs. J. Neurosci.
18,3327
-3335.
Pirvola, U., Ylikoski, J., Trokovic, R., Hebert, J. M., McConnell, S. K. and Partanen, J. (2002). FGFR1 is required for the development of the auditory sensory epithelium. Neuron 35,671 -680.[CrossRef][Medline]
Riley, B. B., Chiang, M., Farmer, L. and Heck, R. (1999). The deltaA gene of zebrafish mediates lateral inhibition of hair cells in the inner ear and is regulated by pax2.1. Development 126,5669 -5678.[Abstract]
Ross, D. A. and Kadesch, T. (2004). Consequences of Notch-mediated induction of Jagged1. Exp. Cell Res. 296,173 -182.[CrossRef][Medline]
Ruben, R. J. (1967). Development of the inner ear of the mouse: a radioautographic study of terminal mitoses. Acta Otolaryngol. Suppl. 220, 1-44.
Satoh, T. and Fekete, D. M. (2005). Clonal
analysis of the relationships between mechanosensory cells and the neurons
that innervate them in the chicken ear. Development
132,1687
-1697.
Schier, A. F., Neuhauss, S. C., Harvey, M., Malicki, J., Solnica-Krezel, L., Stainier, D. Y., Zwartkruis, F., Abdelilah, S., Stemple, D. L., Rangini, Z. et al. (1996). Mutations affecting the development of the embryonic zebrafish brain. Development 123,165 -178.[Abstract]
Sherr, C. J. and Roberts, J. M. (1999). CDK
inhibitors: positive and negative regulators of G1-phase progression.
Genes Dev. 13,1501
-1512.
Shimizu, K., Chiba, S., Kumano, K., Hosoya, N., Takahashi, T.,
Kanda, Y., Hamada, Y., Yazaki, Y. and Hirai, H. (1999). Mouse
jagged1 physically interacts with notch2 and other notch receptors. Assessment
by quantitative methods. J. Biol. Chem.
274,32961
-32969.
Shimizu, K., Chiba, S., Saito, T., Kumano, K. and Hirai, H. (2000). Physical interaction of Delta1, Jagged1, and Jagged2 with Notch1 and Notch3 receptors. Biochem. Biophys. Res. Commun. 276,385 -389.[CrossRef][Medline]
Stone, J. S. and Rubel, E. W. (2000). Cellular
studies of auditory hair cell regeneration in birds. Proc. Natl
Acad. Sci. USA 97,11714
-11721.
Theiler, K. (1989). The House Mouse Altas of Embryonic Development. New York: Springer-Verlag.
Tsai, H., Hardisty, R. E., Rhodes, C., Kiernan, A. E., Roby, P.,
Tymowska-Lalanne, Z., Mburu, P., Rastan, S., Hunter, A. J., Brown, S. D. et
al. (2001). The mouse slalom mutant demonstrates a role for
Jagged1 in neuroepithelial patterning in the organ of Corti. Hum.
Mol. Genet. 10,507
-512.
Woods, C., Montcouquiol, M. and Kelley, M. W. (2004). Math1 regulates development of the sensory epithelium in the mammalian cochlea. Nat. Neurosci. 7,1310 -1318.[CrossRef][Medline]
Xue, Y., Gao, X., Lindsell, C. E., Norton, C. R., Chang, B.,
Hicks, C., Gendron-Maguire, M., Rand, E. B., Weinmaster, G. and Gridley,
T. (1999). Embryonic lethality and vascular defects in mice
lacking the Notch ligand Jagged1. Hum. Mol. Genet.
8, 723-730.
Zheng, J. L., Shou, J., Guillemot, F., Kageyama, R. and Gao, W. Q. (2000). Hes1 is a negative regulator of inner ear hair cell differentiation. Development 127,4551 -4560.[Abstract]
Zine, A., Aubert, A., Qiu, J., Therianos, S., Guillemot, F.,
Kageyama, R. and de Ribaupierre, F. (2001). Hes1 and Hes5
activities are required for the normal development of the hair cells in the
mammalian inner ear. J. Neurosci.
21,4712
-4720.
Related articles in Development:
This article has been cited by other articles:
|
|
 |
|
  D. Zou, C. Erickson, E.-H. Kim, D. Jin, B. Fritzsch, and P.-X. Xu Eya1 gene dosage critically affects the development of sensory epithelia in the mammalian inner ear Hum. Mol. Genet., November 1, 2008; 17(21): 3340 - 3356. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 |
