Download p27Kip1 and development of the inner ear

1581
Development 126, 1581-1590 (1999)
Printed in Great Britain © The Company of Biologists Limited 1999
DEV4116
p27Kip1 links cell proliferation to morphogenesis in the developing organ of
Corti
Ping Chen1 and Neil Segil1,2,*
1Department
2Department
of Cell and Molecular Biology, House Ear Institute, 2100 West Third Street, Los Angeles, CA 90057, USA
of Cell and Neurobiology, University of Southern California Medical School, Los Angeles, CA 90033, USA
*Author for correspondence (e-mail: [email protected])
Accepted 10 February; published on WWW 17 March 1999
SUMMARY
Strict control of cellular proliferation is required to shape
the complex structures of the developing embryo. The
organ of Corti, the auditory neuroepithelium of the inner
ear in mammals, consists of two types of terminally
differentiated mechanosensory hair cells and at least four
types of supporting cells arrayed precisely along the length
of the spiral cochlea. In mice, the progenitors of greater
than 80% of both hair cells and supporting cells undergo
their terminal division between embryonic day 13 (E13)
and E14. As in humans, these cells persist in a nonproliferative state throughout the adult life of the animal.
Here we report that the correct timing of cell cycle
withdrawal in the developing organ of Corti requires
p27Kip1, a cyclin-dependent kinase inhibitor that functions
as an inhibitor of cell cycle progression. p27Kip1 expression
is induced in the primordial organ of Corti between E12
and E14, correlating with the cessation of cell division of
the progenitors of the hair cells and supporting cells. In
wild-type animals, p27Kip1 expression is downregulated
INTRODUCTION
Developmental control of morphogenesis requires that cell
proliferation be coordinated with growth and differentiation.
Recent studies in Drosophila have begun to identify some of
the developmental signals that are able to influence cell cycle
progression (Go et al., 1998; Johnston and Edgar, 1998;
Weigmann et al., 1997), as well as some of their likely cell
cycle targets (Edgar and Datar, 1996). In vertebrates, the
existence of analogous pathways linking development to
control of the cell cycle have been revealed by the occurrence
of developmental defects that result from targeted mutation of
genes involved in cell cycle function, such as cyclin D1 and
the retinoblastoma protein family (Cobrinik et al., 1996;
Sicinski et al., 1995). Nonetheless, the mechanisms that link
developmental events to the cell cycle machinery that controls
cell proliferation remain poorly understood.
Development of the organ of Corti, the auditory sense organ
of mammals, involves the differentiation of two types of
mechanosensory hair cells (inner and outer) and four types of
supporting cells (Deiters’, Hensen’s, Claudius’ and pillar) (Fig.
during subsequent hair cell differentiation, but it persists
at high levels in differentiated supporting cells of the
mature organ of Corti. In mice with a targeted deletion of
the p27Kip1 gene, proliferation of the sensory cell
progenitors continues after E14, leading to the appearance
of supernumerary hair cells and supporting cells. In the
absence of p27Kip1, mitotically active cells are still observed
in the organ of Corti of postnatal day 6 animals, suggesting
that the persistence of p27Kip1 expression in mature
supporting cells may contribute to the maintenance of
quiescence in this tissue and, possibly, to its inability to
regenerate. Homozygous mutant mice are severely hearing
impaired. Thus, p27Kip1 provides a link between
developmental control of cell proliferation and the
morphological development of the inner ear.
Key words: p27Kip1, Organ of Corti, Inner ear, Cell cycle, Cyclindependent kinase inhibitor
1). Each of these cell types has a distinct morphology that
contributes to the complex structural and functional properties
of the organ of Corti. Both the auditory and vestibular sense
organs of the inner ear develop collectively from a thickening
of the epithelial sheet in the dorsolateral region of the head
known as the otic placode (Fritzsch et al., 1998). In the mouse,
after its formation on embryonic day 8.5 (E8.5), the placode
invaginates to form the otocyst on E11. By this time,
specification of the regions that will give rise to sensory
neuroepithelia containing the mechanosensory hair cells and
supporting cells, as well as non- sensory epithelia, which will
give rise to tissues such as Reisner’s membrane and the
endolymphatic tube and sac, have already been specified
(Morsli et al., 1998). By E12, the otocyst has begun to develop
morphologically into the cochlea and vestibule. The organ of
Corti is recognizable at this time as a thickened ridge in the
cochlear portion of the otocyst (Fig. 2B, arrow). At E12, the
progenitors of the hair cells and supporting cells are still
dividing (Ruben, 1967).
Between E12 and E16, the cells of the primordial organ of
Corti exit the cell cycle to form a postmitotic neuroepithelial
1582 P. Chen and N. Segil
sheet suspended in the spiral duct of the cochlea (Ruben,
1967). Greater than 80% of these cells exit the cell cycle
between E13 and E14, with a slightly increased probability of
cells located in the apex of the mature cochlea having exited
the cell cycle earlier than those in the base (Ruben, 1967).
Starting on E15, a gradient of differentiation, which radiates in
both directions from the mid-basal region of the cochlea (Sher,
1971; Lim and Anniko, 1985) turns these newly postmitotic
cells into the various morphologically and functionally
differentiated cell types illustrated in Fig. 1.
These cells remain postmitotic for the life of the animal. In
mammals, the loss of auditory hair cells does not appear to lead
to proliferative regeneration (Chardin and Romand, 1995, but
see Lefebvre et al., 1995) and represents the major cause of
deafness in humans. In the mammalian vestibular system,
damage to hair cells appears to lead to a very limited degree
of repair or regeneration, but whether supporting cell
proliferation leads to the differentiation of new hair cells is still
controversial (Forge et al., 1993, 1995; Warchol et al., 1993,
1995; Rubel et al., 1995). In lower vertebrates, hair cell death
leads to renewed proliferation of supporting cells, which
subsequently leads to the differentiation of both hair cells and
supporting cells (Corwin and Cotanche, 1988; Ryals and
Rubel, 1988).
In higher eukaryotes, the decisions concerning whether to
proceed through another round of cell division, exit the cell
cycle and become quiescent, or re-enter the cell cycle from
quiescence are all regulated by a family of proteins known as
cyclin-dependent kinases (CDK). Exit from the cell cycle
usually occurs during G1 phase and distinct CDKs, CDK4/6
and CDK2, regulate transit through G1 and entry into the
following S phase, respectively (see Elledge, 1996 for review).
CDK activity, in turn, is regulated by multiple mechanisms,
including covalent modification of the CDK catalytic subunit;
the relative abundance of required positive cofactors, one of the
several members of the cyclin family of proteins, and by the
activity of negative regulators, the cyclin-dependent kinase
inhibitors (CKI) (see Sherr and Roberts, 1995; Elledge et al.,
1996; Harper and Elledge, 1996 for reviews). In vertebrates,
CKIs are represented by two unrelated families of proteins, the
CIP/KIP family and the INK family. These families interact
with CDKs in different ways, but both work by binding to, and
inhibiting the activity of cyclin-dependent kinases (CDKs)
which can lead to withdrawal from the cell cycle. In spite of
the advances in our knowledge of the regulation of CDK
activity, little is known about how regulation of CKIs is
integrated into specific developmental programs to coordinate
cell proliferation with morphogenesis.
Here we report that between E12 and E14, the cyclindependent kinase inhibitor (CKI) p27Kip1 is induced in the
progenitors of the primordial organ of Corti, correlating with
the time when these cells are withdrawing from the cell cycle.
We also report that, in mice with a targeted deletion of the
p27Kip1 gene (p27−/−) (Fero et al., 1996), the cells of the
developing neuroepithelium undergo a period of prolonged cell
division relative to wild-type littermates, leading to the
appearance of supernumerary hair cells and supporting cells.
This observation suggests that p27Kip1 is responsible for
determining the size of the progenitor population and thus the
morphology of the array of sensory hair cells. These animals
are severely hearing impaired, indicating the importance of
Fig. 1. Hematoxylin-stained section through the organ of Corti of a
newborn mouse. Sensory outer hair cells (OHC) and inner hair cells
(IHC), as well as the non-sensory supporting cell types are indicated.
Deiters’ cells surround each OHC separating them from each other
and their nuclei are located underneath the nuclei of the OHC. The
pillar cells separate the IHCs from the OHCs. Hensen’s and
Claudius’ cells are external to the OHCs as indicated.
precise developmental control of the cell cycle to the normal
development of function. Finally, we have observed that
p27Kip1 is rapidly downregulated in differentiating hair cells,
although its expression persists in postmitotic supporting cells
of the mature organ of Corti, suggesting a role in maintaining
the normally quiescent state of these cells. Consistent with this
hypothesis, we have observed continued proliferation of cells
in the postnatal organ of Corti of p27−/− mice.
MATERIALS AND METHODS
Immunohistochemistry
Antibody raised against p27Kip1 was purchased from NeoMarker and
assayed for specificity by immunoblotting (Fig. 2A). Extracts from
mouse N1E-115 neuroblastoma cells (ATCC) were prepared
according to Kranenburg et al. (1995) before (Fig. 2A, lane 2) and
after (Fig. 2A, lane 3) differentiation and used as controls. Results
indicated that the antibody is monospecific and recognizes a band in
E14 (data not shown) and neonatal (Fig. 2A, lane 1) cochlear extracts,
that comigrates with p27Kip1 from the neuroblastoma cells.
Differences in abundance of p27Kip1 before and after differentiation
of N1E-115 cells are consistent with previously published results
(Kranenburg et al., 1995). The more slowly migrating band visible in
lanes 2 and 3 most likely represents the previously reported
phosphorylated form of p27Kip1. Paraffin sections of temporal
bones were prepared according to Yoho et al. (1997) and
immunohistochemistry was done according to standard procedures
(Harlow and Lane, 1988). Immunohistochemistry of p27Kip1
(NeoMarkers, 1/100 dilution) included an antigen retrieval step
carried out by boiling deparaffinized sections in 10mM Citric acid
buffer, pH 6.0 for 10 minutes. Other antibodies used in this study
included: polyclonal anti-myosin VIIa (courtesy of Christine Petit,
Pasteur Institute) (1/2000 dilution), monoclonal anti-PCNA
(NeoMarker, 1/200 dilution) and monoclonal anti-neurofilament 68
(Sigma, 1/100 dilution). Antibody incubations were done overnight at
4°C and binding was visualized with fluorescein- or rhodamineconjugated secondary antibodies (1/200, Jackson ImmunoResearch).
Organs of Corti from at least three wild-type and three p27Kip1 mutant
animals were examined with these antibodies at each developmental
stage described, except in the PCNA experiment shown in Fig. 6,
where two mutant animals were examined.
p27Kip1 and development of the inner ear 1583
Fig. 2. The temporal and spatial
expression of p27Kip1 in the
developing organ of Corti.
(A) Western blot showing specificity
of antibody to p27Kip1. Protein extract
from neonatal organ of Corti (lane 1)
is compared to extract from
undifferentiated neuroblastoma cells
(N1E-115) (lane 2) known to express
p27Kip1 upon induction of
differentiation (lane 3) (Kranenburg et
al., 1995). A minor protein band,
probably representing phosphorylated
p27Kip1 (Vlach et al., 1997) is detected
in lanes 2 and 3 above the major
p27Kip1 band (arrow). (B) Sections
through the E12 and (C) E14 otocyst
stained with anti-p27Kip1 to show the
time of onset of p27Kip1 expression.
Arrows indicate the thickening of the
dorsal region of the otocyst that gives
rise to the organ of Corti.
(D) Alternate section through E14
otocyst stained with antibody to
neurofilaments to indicate the pattern
of innervation of the primordial organ
of Corti. (E) Section through the midbasal region of the E16 cochlea
double-labeled with p27Kip1 and
(F) myosin VIIa antibodies. Single
arrow indicates the layer of p27Kip1stained supporting cells just below the
hair cells. Clustered arrows indicate
the three outer and one inner hair cells.
(G) Section through the mid-basal
region of the P1 cochlea doublelabeled with anti-p27Kip1 and (H) antimyosin VIIa. Arrows are the same as
in E and F. Size bar, 50 µm.
Confocal imaging
Surface preparations of organ of Corti were prepared from animals
aged P1 to 5 months old. Following fixation in paraformaldehyde,
cochleae were dissected in PBS buffer, stained with rhodaminephalloidin (Molecular Probes) and mounted on slides. Z series scans
were made on a Zeiss LSM 410 inverted laser scanning microscope.
Hair cell counts were made from photographs of these surface
preparations and analyzed with a non-parametric statistic (the MannWhitney U test, StatView).
Animals
For production of mutant mice, male (129/Sv) mice, heterozygous for
an induced mutation at the p27Kip1 locus (Fero et al., 1996), were
mated with female (C57BL/6NHSd) mice to produce an F1
generation. p27+/− offspring were identified by polymerase chain
reaction (PCR) as described (Fero et al., 1996). Heterozygotes were
mated and animals homozygous for the induced mutation were
identified by PCR. Wild-type, p27−/− and p27+/− animals used in this
study were from this F2 generation. Genotyping was verified by
western blotting of brain extracts and also by immunohistochemistry
of sections taken through the cochlea, using anti-p27Kip1 antibody
(NeoMarkers) (data not shown). Animal care was in accordance with
institutional guidelines.
Auditory brainstem response (ABR)
ABRs were recorded from two subcutaneous needle electrodes placed
behind the pinna and on the vertex. A third electrode was placed on
the hindleg as a ground. The Tucker Davis ABR Workstation with
TDT SigGen and BioGen software were used for stimulus generation
and data acquisition. Broadband clicks were delivered through an
1584 P. Chen and N. Segil
Fig. 3. Gradient of p27Kip1 downregulation and myosin VIIa
expression in hair cells along the basal-to-apical axis of the cochlea.
A single section through the basal (right) and apical (left) turns of the
E15.5 cochlea double-labeled with antibody to (A) myosin VIIa and
(B) p27Kip1 Scale bar, 50 µm..
Intelligent Hearing System Insert transducer. The broadband click was
calibrated in peak equivalent sound pressure level (peSPL), in situ for
two wild-type animals (ER-7 probe tube microphone system). Clickevoked ABR threshold was defined as the last level (dB peSPL) at
which an ABR waveform could be visually detected.
RESULTS
The onset of expression of p27Kip1 in the developing
organ of Corti
In mice, greater than 80% of the cells of the developing organ
of Corti stop dividing between E13 and E14 of embryonic
development (Ruben, 1967). At this time, the vestibular system
has acquired its gross morphology and the cochlea has made
one and one quarter of its ultimate one and three quarter turns.
Based on morphological criteria, the primordial organ of Corti
develops along the greater epithelial ridge of the otocyst (Fig.
2B, arrow). Just prior to this time, on E12, no p27Kip1 protein
expression is observed in any region of the otocyst (Fig. 2B).
However, 2 days later, correlating with the time of cessation of
cell division, p27Kip1 is strongly expressed at this site (Fig. 2C,
arrow). Confirmation that this is the site of organ of Corti
development was obtained by staining alternate sections with
a neurofilament antibody (Fig. 2D) to visualize the nerve fibers
from the spiral ganglion cells that are known to reach the
Fig. 4. p27Kip1 is expressed in the supporting cells of the vestibular
sensory epithelia in P6 animals. Sections were labeled with antibody
to p27Kip1 as in Fig. 2. Brackets indicate the hair cell layer,
arrowhead indicates the supporting cell layer. (A) Hematoxylinstained section through sacculus. (B) Anti-p27Kip1-stained section
through sacculus; (C) utriculus; (D) crista ampullaris. Note
stereocilia visible on hair cells in hematoxylin-stained section (A)
and unstained dark nuclei in the hair cell layer of p27Kip1-stained
sections (B-D). Size bar, 50 µm.
primordial organ of Corti before the morphological
differentiation of hair cells (Pujol et al., 1998). At this time,
supporting cells and hair cells are not distinguishable
morphologically or by antibodies to one of the earliest markers
of hair cell differentiation, myosin VIIa (Sahly et al., 1997)
p27Kip1 and development of the inner ear 1585
Fig. 5. A comparison of the structure of organ of Corti from p27+/+ wild-type (A-C), p27+/− heterozygous (D-F) and p27−/− homozygous mutant
(G-I) mice, age P6. Loss of p27Kip1 expression due to a targeted gene deletion causes the development of supernumerary hair cells and
supporting cells in the organ of Corti. (A,D,G) Confocal images of surface preparations used to compare the overall arrangement of sensory
cells in the organ of Corti. They were stained with rhodamine-conjugated phalloidin to visualize the actin-rich stereocilia of the sensory hair
cells. (B,E,H) Cross sections through the organ of Corti in the mid-cochlear region stained with antibody to the hair-cell-specific antigen,
myosin VIIa. (C,F,I) Alternate sections stained with hematoxylin to reveal the cellular architecture of the organ of Corti. Brackets mark the
multiple rows of outer hair cells, three in the p27+/+ and p27+/− animals, four in the p27−/− animals. Arrowheads point to the single row of inner
hair cells present in the wild-type organ of Corti, as well as to the rows of inner hair cells containing supernumerary hair cells in the p27+/− and
p27−/− mutant animals. Sections from wild-type animals contain the normal number of inner and outer hair cells (B,C). The occasional presence
of supernumerary inner hair cells in heterozygotes is illustrated in E but not in F. Note supernumerary cells present in the pillar cell region of
the organ of Corti from p27−/− animals where normally one inner pillar and one outer pillar cells are present (compare C with I, asterisks).
Fig. 6. Anti-PCNA staining of E16
organ of Corti from p27−/− (A,B) and
p27+/+ embryos (C,D) reveals actively
cycling cells in the mutant organ of
Corti. Sections through the mid-basal
region of the developing cochlea of E16
p27−/− and p27+/+ animals were doublelabeled with antibody against PCNA
(A,C, red) to stain cycling cells, and
myosin VIIa (B,D, green) to stain
differentiating hair cells. The hair cell
region is indicated by brackets, while
the region of Deiters’ cells is indicated
by clustered arrows. Note PCNAstained nuclei (red) in a position above
the arrows in the p27−/− organ of Corti
(A) indicating the continued
proliferation at this site, as well as the
absence of PCNA staining in the
comparable region in p27+/+ animal (C).
Size bar, 50 µm.
1586 P. Chen and N. Segil
(data not shown). The same pattern of p27Kip1 expression is
still seen 1 day later, on E15 (data not shown). Thus, p27Kip1
is one of the first molecular markers for the primordial organ
of Corti, allowing its identification prior to morphological or
biochemical differentiation of hair cells and supporting cells.
The sharp boundaries of p27Kip1 expression within the
developing otocyst (Fig. 2C) reflect the precise regional
specification of the sensory regions in the otocyst that occurred
at earlier embryonic times (Morsli et al., 1998).
By E16, the differentiating sensory hair cells become
recognizable morphologically, coincident with the appearance
of one of the earliest known hair cell markers, myosin VIIa
(Sahly et al., 1997) (Fig. 2F, clustered arrows). Interestingly,
p27Kip1 is no longer detected in the hair cells, although
expression can still be seen in the nuclei of surrounding
supporting cells of the sensory epithelium (Fig. 2E, arrow). In
newborn and adult mice (Fig. 2G,H and data not shown),
p27Kip1 continues to be expressed in the differentiated
supporting cells including Deiters’, Hensen’s, Claudius’ and
pillar cells.
Based on morphological observations, it has been reported
that, between E15 and E16, a gradient of hair cell
differentiation develops starting in the mid-basal regions of the
cochlea and spreading in both the apical and basal directions
(Sher, 1971; Lim and Anniko, 1985; see Rubel, 1978, for
review). To determine more precisely when p27Kip1 is
downregulated relative to hair cell differentiation and to test
whether this gradient is reflected in the pattern of myosin VIIa
expression, cross sections through a single cochlea from an
embryo between E15 and E16 were stained with antibody to
myosin VIIa (Fig. 3A) and p27Kip1 (Fig. 3B). These sections
show both apical and basal turns and indicate the presence of
a gradient of expression of the hair cell marker myosin VIIa.
In a turn of this cochlea from the basal region (Fig. 3A), we
found that inner hair cells (arrowhead) are labeled relatively
strongly with myosin VIIa (green) while outer hair cells
(bracket) are more weakly labeled. In contrast, no staining of
myosin VIIa is observed in the apical turn present in the same
cochlea at this stage. p27Kip1 staining is present in the
supporting cell layer of the basal turn, but absent from all the
hair cells (Fig. 3B), as it is throughout the cochlea at slightly
later embryological times. However, in the apical turn, where
no myosin VIIa staining is observed (Fig. 3A), p27Kip1 appears
to stain all of the cells in the primordial sensory epithelium.
These results suggest that the inner hair cells mature before the
outer hair cells and show the close temporal correlation
between p27Kip1 downregulation and the differentiation of hair
cells as seen by the onset of myosin VIIa expression. They also
demonstrate a gradient of differentiation within the cochlea,
consistent with that previously reported in morphological
studies (cited in Rubel, 1978).
p27Kip1 is also expressed in the vestibular system
In the vestibular sensory organs, p27Kip1 is also localized in the
non-sensory supporting cells of neonatal (P6) animals (Fig.
4A-D). As shown in Fig. 4B, occasional p27Kip1-positive cells
are seen in the sensory cell layer as well, although the type and
origin of these cells is unknown. In contrast to their
counterparts in the cochlea, the cells of the vestibular sensory
organs withdraw from the cell cycle over a longer period of
time, ranging from E14 to P3 (Ruben, 1967). Developmentally,
we observed p27Kip1 expression in subregions of the sensory
component of the differentiating vestibular epithelia as early as
E14 (data not shown). This suggests that a similar sequence of
developmental events involving p27Kip1 in exit from the cell
cycle takes place in the vestibular system. However, the levels
of p27Kip1 expressed in the developing and mature vestibular
system appear to be lower than in comparable sections of organ
of Corti making their analysis more difficult. Consequently, we
have not precisely correlated the expression of p27Kip1 with the
cessation of cell division in the developing vestibular epithelia.
Hyperplasia of the organ of Corti in p27Kip1 knockout
mice
The restricted pattern of p27Kip1 expression in the developing
sensory epithelium led us to examine the developmental
consequences of an induced mutation in the p27Kip1 gene (Fero
et al., 1996; Kiyokawa et al., 1996; Nakayama et al., 1996). In
the absence of p27Kip1, animals are larger than wild-type
littermates as a result of multiorgan hyperplasia (Fero et al.,
1996; Kiyokawa et al., 1996; Nakayama et al., 1996). Wildtype mice normally contain a single row of inner hair cells
(arrowhead) and three rows of outer hair cells (brackets, Fig.
5A-C). In mice homozygous for a p27Kip1 mutation (p27−/−),
supernumerary hair cells developed in both the inner and outer
rows of hair cells (Fig. 5G-I). The inner row contains a partly
disorganized line of at most two hair cells (Fig. 5G-I,
arrowheads), while the outer hair cells show a pattern of four,
partly disorganized, lines of cells (Fig. 5G-I, brackets).
Representative samples of outer and inner hair cells in the midbasal region of the cochlea were counted and revealed
signifcantly more cells in mutant (n=4) versus wild-type (n=4)
mice (P<0.05). Outer hair cell numbers in the mutant animals
were increased by a mean of 36% and inner hair cell numbers
were increased by a mean of 23%. In addition, an excess
number of supporting cells, including pillar cells separating
inner from outer hair cells and occupying the area of the tunnel
of Corti, are present in p27−/− animals (Fig. 5I, asterisks).
Normally, only one inner pillar cell and one outer pillar cell
are present in this region (Figs 2B, 5C, asterisks). These results
indicate that there is an increase in both hair cells and
supporting cells in the absence of p27Kip1.
In mice heterozygous for a p27Kip1 mutation (p27+/−),
the outer hair cell population of the heterozygous animals
appears normal (Fig. 5D-F). However, we observed occasional
supernumerary inner hair cells throughout the organ of
Corti in 6-day-old heterozygous animals (Fig. 5D,E). This
phenomenon was never observed in wild-type littermates (Fig.
5A-C). Thus, the severity of this abnormality appears to be
dose dependent in that homozygotes are more severely affected
than heterozygotes (Fig. 5), although the degree of dose
dependence has not been quantified.
On the basis of neuronal fiber staining, hair cells in P6
animals appear to have a normal pattern of innervation (data
not shown). However, analysis of auditory brainstem responses
(ABR) indicated that mutant animals were severely hearing
impaired. 10-week-old p27−/− animals had a significantly
elevated mean, click-evoked, ABR threshold (77 db SPL, n=3)
relative to comparably aged wild-type animals (20 db SPL,
n=6) (P<0.05). Homozygous mutant animals showed no
obvious behavioral defects related to vestibular function such
as circling behavior or balance problems. The cause of the
p27Kip1 and development of the inner ear 1587
hearing deficit in homozygous animals could lie in the
peripheral abnormalities that we have described or it could lie
in central abnormalities that result either from the p27Kip1
mutation or as a consequence of the peripheral anatomical
abnormalities. These possibilities are currently under
investigation.
Cell proliferation in the organ of Corti of p27Kip1
knockout mice
The abnormalities observed in the organ of Corti of p27−/−
mice suggest that p27Kip1 functions as a growth inhibitor
during development, and its ablation leads to prolonged or
unscheduled proliferation of hair and supporting cell
progenitors, ultimately leading to the differentiation of
supernumerary cells. We tested this hypothesis by comparing
the expression of Proliferating Cell Nuclear Antigen (PCNA),
a marker for cells that are in, or about to enter, the DNA
replication phase of the cell cycle and thus are scheduled to
undergo another round of cell division (Galand, 1989), in
p27+/+ and p27−/− animals. Since the onset of p27Kip1
expression (Fig. 2C) correlates with the cessation of cell
division of organ of Corti progenitor cells on E14 (Ruben,
1967), we studied the incorporation of PCNA into replicating
DNA in p27−/− and p27+/+ embryos between E15 and E16
when cell division should have ceased. In p27−/− embryos,
PCNA-positive cells are observed in the Deiters’ cell region
(Fig. 6A, arrows) beneath the newly differentiated hair cells
stained with anti-myosin VIIa (Fig. 6B, bracket), indicating
that proliferation continues in these animals past the normal
time of cell cycle withdrawal. In contrast, in wild-type embryos
PCNA-labeled cells are absent from the Deiters’ cell region
(Fig. 6C, arrows) of the organ of Corti, beneath the myosin
VIIa-stained hair cells (Fig. 6D, brackets).
To test the duration of this aberrant proliferative activity, we
stained P6 cochleas from p27−/− and p27+/+ animals for
expression of the PCNA marker (Fig. 7). In p27−/− animals,
PCNA-positive cells are no longer seen in Deiters’ cell region
as they are at E16, but appear in clusters in the region of
Hensen’s cells, lateral to the outermost row of outer hair cells
(Fig. 7A, arrow), as well as in the pillar cell regions separating
inner and outer hair cells (Fig. 7B, arrows). As expected, no
PCNA staining is apparent in the organ of Corti (Fig. 7C,
bracket and arrow) from p27+/+ animals, and no PCNA-positive
cells were observed in comparable sections through the organ
of Corti of heterozygous animals (data not shown). These
results suggest that p27Kip1 is required for maintaining the
supporting cells of the mature organ of Corti in their normally
quiescent state.
DISCUSSION
While the control of cell proliferation, growth and
differentiation are separable phenomena, coordination between
these phenomena is required if correct morphogenetic
patterning is to occur (see Skaer, 1998; Johnston, 1998 for
reviews). By studying the effects of an induced mutation in the
cell cycle regulator p27Kip1, we have identified one element
required for coordinating the cessation of cell division with
differentiation and morphogenetic patterning in the developing
organ of Corti. Our results show that p27Kip1 activity is
Fig. 7. Cell division persists in the basal region of the organ of Corti
of postnatal (P6) homozygous mutant animals. (A,B) Anti-PCNA
staining of sections from early postnatal (P6) organ of Corti from
homozygous mutant and wild-type animals and (C) wild-type
control. (A) Arrow indicates clusters of PCNA-positive cells in
Hensen’s cell region; (B) double arrow indicates positive cells in the
region of the pillar cells; (C) bracket and arrowhead indicate outer
and inner hair cell regions respectively. No PCNA-positive cells are
seen in the organ of Corti from wild-type animals at this stage. Scale
bar, 50 µm.
required for the timely withdrawal from the cell cycle of the
cells of the developing sensory epithelium and its absence leads
to the appearance of supernumerary cells throughout the
mature organ of Corti. Additional studies of the pattern of
expression of p27Kip1 during normal development of the inner
ear, suggest that p27Kip1 may play a subsequent role in cellular
differentiation and homeostasis of the various cell types in the
sensory epithelium.
Coordinating cell proliferation with morphogenetic
events
The ability of p27Kip1 to coordinate cell proliferation with
morphogenetic events has been suggested by its well-known
capacity to block cell cycle progression by inhibiting the
activity of CDK4/6 and CDK2 (Elledge et al., 1996), by its
ability to respond in vitro to a variety of known cellular growth
regulators (Sherr and Roberts, 1995), and finally by the
developmental abnormalities that are caused by its targeted
deletion in mice (Fero et al., 1996; Kiyokawa et al., 1996;
Nakayama et al., 1996). Regulation of p27Kip1 protein levels
1588 P. Chen and N. Segil
in vitro is governed at the transcriptional (Kawasaki et al.,
1998), translational (Hengst and Reed, 1996; Millard et al.,
1997) and, perhaps most importantly, the post-translational
(Pagano et al., 1995; Vlach et al., 1997) level. However, to our
knowledge, the specific mechanisms used to regulate p27Kip1
levels in vivo have not been studied, and the molecular
machinery underlying the temporally and spatially restricted
induction of p27Kip1 protein that we have described (Fig. 2) is
currently unknown.
Morsli et al. (1998) have shown that the gene lunatic fringe,
a modulator of the Notch signaling pathway, is expressed in
the sensory regions of the developing cochlea in a restricted
region that appears to presage p27Kip1 expression. While it is
not clear whether the expression of this gene in the cochlea has
the exact same boundaries as p27Kip1, it is possible that the
Notch signaling pathway affects morphogenesis of the organ
of Corti through the regulation of cell proliferation in a manner
similar to that shown for wing development in Drosophila
(Johnston and Edgar, 1998). Although the specific stimulus for
p27Kip1 induction is not known, the sharp boundaries of
p27Kip1 expression that arise in the otocyst between E12 and
E14 (Fig. 2C) are likely to reflect boundaries that have been
set up by morphogenetic gradients analogous to those seen
during the patterning of other embryonic anlage and, as such,
may serve as a downstream marker for the specification of the
boundaries of the sensory epithelia.
Control of cell proliferation in the developing organ
of Corti
During development of the organ of Corti, p27Kip1 levels rise
abruptly between E12 and E14 (Fig. 2) and the majority of the
cells that will make up the mature sensory epithelium exit the
cell cycle over the period of 24 -48 hours that follow (E13 and
E15, Ruben, 1967). In animals that lack p27Kip1, proliferation
within the E16 sensory epithelium continues (as seen by the
presence of PCNA-positive cells in the Deiters’ cell region,
Fig. 6A), even as some of the newly generated cells are
beginning to differentiate into hair cells (Fig. 6B). The
supernumerary cells that we observe (Fig. 5) are the
consequence of this abnormal proliferation. At later times (P6,
Fig. 7) PCNA-positive cells are restricted to the region of
Hensen’s, Claudius’ and pillar cells and are absent from the
Deiters’ cell region. The reason for this change is not known,
but is likely to be related to the mechanism governing hair cell
and supporting cell differentiation discussed below.
The fact that both hair cells and supporting cells are able to
stop dividing and differentiate in the organ of Corti of p27−/−
animals indicates that there are multiple levels of cell cycle
control at work. The nature of these other controls on cell
proliferation in the organ of Corti is not clear. In studies of
developing oligodendrocytes in vitro, it was found that the
absence of p27Kip1 leads to a prolonged period of cell division
(Casaccia-Bonnefil et al., 1997; Durand et al., 1998),
correlating with an increased number of cells in vivo in p27−/−
animals (Casaccia-Bonnefil et al., 1997). While p27−/−
oligodendrocytes are able to undergo cell cycle arrest and
differentiation in vivo, under specific in vitro growth
conditions, differentiation of these cells is partially impaired
(Casaccia-Bonnefil et al., 1997; Tikoo et al., 1998). These
observations, like our own, are consistent with the overall
phenotype displayed by p27−/− animals, namely the
generalized hyperplasia observed in many, if not all tissues,
along with a range of abnormalities in tissue organization (Fero
et al., 1996; Kiyokawa et al., 1996; Nakayama et al., 1996). In
spite of the increase in cell number, the function of most tissues
is not grossly perturbed (Fero et al., 1996; Kiyokawa et al.,
1996; Nakayama et al., 1996) and homozygous mutant animals
are viable. The presence in some tissues of more than one CKI
(Franklin et al., 1998; Zhang et al., 1998), as well as the
existence of potential CKI-independent pathways able to
downregulate CDK activity (Elledge et al., 1996), suggests that
overlapping pathways regulating the cell cycle during
development may be the norm. However, as our results
indicate, p27−/− animals are severely hearing impaired,
probably reflecting a strict requirement for precise cellular
organization in the organ of Corti.
Specification of cell type in the organ of Corti
The primary mechanism specifying hair cell versus supporting
cell differentiation remains an important question for the field.
While alternative models exist, several investigators have
suggested that at the time of their terminal mitosis, the cells of
the sensory epithelium are not determined to be either hair cells
or supporting cells, and only sometime after cell cycle exit do
they differentiate into one or the other cell type, possibly
through a process of lateral inhibition (see Fekete, 1996 for
review). The specification of hair cells versus supporting cells
following terminal mitosis is consistent with results from
recent studies of terminal differentiation in the basilar papilla
of the chicken using retroviral tracing methods (Fekete et al.,
1998), as well as studies on the differentiation of sensory hair
cells during in vitro organ culture of the mouse organ of Corti
(Kelley et al., 1993, 1995). These workers show that for a
limited period of time following terminal cell division, cells of
the sensory epithelium can be induced to become
supernumerary hair cells, either through retinoic acid treatment
(Kelley et al., 1993) or by the death of a neighboring, already
differentiated hair cell (Kelley et al., 1995). This is
accomplished without additional cell division, suggesting that
the new cells are likely to have been derived from cells that
would otherwise have differentiated into supporting cells or
from a population of cells that would normally have undergone
apoptosis.
If these arguments are correct, the initial induction of
p27Kip1 in the otocyst, which occurs prior to hair cell and
supporting cell differentiation, is likely to be responsible for
regulating the correct number of postmitotic precursor cells on
which these later specification processes can act. By this
argument, the supernumerary cells produced because of the
absence of p27Kip1 are the result of the normal patterning
within an increased population of progenitors. In contrast, the
supernumerary cells that are produced in response to retinoic
acid, in the absence of new cell division (Kelley et al., 1993),
likely result from a more direct influence on the process of
postmitotic differentiation.
While the very existence of both hair cells and supporting
cells in p27−/− animals argues strongly against a required,
positive role for p27Kip1 in the terminal differentiation of either
cell type, it leaves open the possibility of a negative role in
regulating the differentiation of one or the other. In this case, the
rapid downregulation of p27Kip1 that occurs in differentiating
hair cells (Fig. 2) could be a necessary consequence of the
p27Kip1 and development of the inner ear 1589
initiation of differentiation in these cells. We have observed that
the induction of myosin VIIa, a marker of hair cell
differentiation, occurs after the last cell division and is
temporally correlated with the downregulation of p27Kip1. Just
how tightly correlated is perhaps best seen by our demonstration
of the gradient of differentiation within the E15 cochlea, where
basal sections reveal hair cells expressing myosin VIIa and
lacking p27Kip1 expression, at the same time as postmitotic
progenitor cells in more apical regions of the cochlea have yet
to begin expression of myosin VIIa and still express p27Kip1
(Fig. 3). At no time do we observe myosin VIIa and p27Kip1
double-labeled cells in our preparations. Nor do we observe cells
in the postmitotic organ of Corti that lack both markers. This
suggests that the expression of these two markers may be
mutually exclusive. Although no causative relationship has been
demonstrated, a mutually exclusive pattern of expression exists
between one of the other CIP/Kip family members (p21Cip1), and
markers of cell differentiation in postmitotic keratinocytes (Di
Cunto et al., 1998) suggesting that p21Cip1 has a direct role in
the differentiation of these cells, beyond its role in regulating cell
proliferation. The downregulation of p27Kip1, which appears to
be part of the program of hair cell differentiation, could serve an
analogous function.
Quiescence and regeneration
The presence of proliferating cells in the postnatal organ of
Corti of p27−/− animals (Fig. 7), as well as the persistence of
p27Kip1 expression in mature supporting cells (Fig. 2G,H, and
data not shown), reveals a second role for p27Kip1, that of
maintaining these cells in a quiescent state. While its
persistence may be required in mature supporting cells to
maintain this state, it may also offer a possible explanation for
the lack of regeneration in mammals. In lower vertebrates such
as chickens, the loss of hair cells due to acoustic trauma or
ototoxic shock leads to re-entry of supporting cells into the cell
cycle and the subsequent differentiation of their progeny into
hair cells (Corwin and Cotanche, 1988; Ryals and Rubel, 1988).
We hypothesize that, if supporting cells in chickens also contain
high levels of a p27Kip1 homologue, it is rapidly degraded upon
loss of hair cells, in order to allow them to re-enter the cell cycle.
The difference between mammals and birds could lie in the
level of p27Kip1 in mature supporting cells or the efficiency with
which it is downregulated following injury. Alternatively, other
cell cycle regulators besides p27Kip1 may be involved.
Regardless, the implication is that the high levels of p27Kip1 that
persist in supporting cells of mammals could be an impediment
to re-entry into the cell cycle following loss of hair cells.
We wish to thank Andres Collazo, Yun-Shain Lee, David Lim,
Federico Kalinec, Greta Segil and Howard Worman for invaluable
comments on the manuscript and Ed Rubel for helpful discussion of
our results. Butch Welch provided expert graphics assistance and Xi
Lin provided expert assistance with confocal imaging. Caroline
Abdala generously assisted with the ABR analysis. We also thank
Matthew Fero and James Roberts for the p27Kip1 knockout mice,
Christine Petit (Pasteur Institut) who generously supplied the antimyosin VIIa antibody used in this study, and Yun-Shain Lee who
helped with the surface preparation shown in Fig. 5A, D and G.
REFERENCES
Casaccia-Bonnefil, P., Tikoo, R., Kiyokawa, H., Friedrich, J. V., Chao, M.
V. and Koff, A. (1997). Oligodendrocyte precursor differentiation is
perturbed in the absence of the cyclin-dependent kinase inhibitor p27Kip1.
Genes Dev. 11, 2335-2346.
Chardin, S. and Romand, R. (1995). Regeneration and mammalian auditory
hair cells [letter; comment]. Science 267, 707-711.
Cobrinik, D., Lee, M. H., Hannon, G., Mulligan, G., Bronson, R. T., Dyson,
N., Harlow, E., Beach, D., Weinberg, R. A. and Jacks, T. (1996). Shared
role of the pRB-related p130 and p107 proteins in limb development. Genes
Dev. 10, 1633-1644.
Corwin, J. T. and Cotanche, D. A. (1988). Regeneration of sensory hair cells
after acoustic trauma. Science 240, 1772-1774.
Di Cunto, F., Topley, G., Calautti, E., Hsiao, J., Ong, L., Seth, P. K. and
Dotto, G. P. (1998). Inhibitory function of p21Cip1/Waf1 in differentiation
of primary mouse keratinocytes independent of cell cycle control. Science
280, 1069-1072.
Durand, B., Fero, M. L., Roberts, J. M. and Raff, M. C. (1998). p27Kip1
alters the response of cells to mitogen and is part of a cell-intrinsic timer
that arrests the cell cycle and initiates differentiation. Curr. Biology 8, 431440.
Edgar, B. A. and Datar, S. A. (1996). Zygotic degradation of two maternal
Cdc25 mRNAs terminates Drosophila’s early cell cycle program. Genes
Dev. 10, 1966-1977.
Elledge, S. J. (1996). Cell cycle checkpoints: preventing an identity crisis.
Science 274, 1664-1672.
Elledge, S. J., Winston, J. and Harper, J. W. (1996). A question of balance:
the role of cyclin-kinase inhibitors in development and tumorigenesis.
Trends in Cell Biol. 6, 388-397.
Fekete, D. M. (1996). Cell fate specification in the inner ear. Curr. Opin.
Neurobiol. 6, 533-541.
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.
Fero, M. L., Rivkin, M., Tasch, M., Porter, P., Carow, C. E., Firpo, E.,
Polyak, K., Tsai, L. H., Broudy, V., Perlmutter, R. M., Kaushansky, K.
and Roberts, J. M. (1996). A syndrome of multiorgan hyperplasia with
features of gigantism, tumorigenesis, and female sterility in p27(Kip1)deficient mice. Cell 85, 733-744.
Forge, A., Li, L. and Nevill, G. (1995). [Comments on] Mammalian vestibular
hair cell regeneration. Science 267, 706-707.
Forge, A., Li, L., Corwin, J. T. and Nevill, G. (1993). Ultrastructural
evidence for hair cell regeneration in the mammalian inner ear [see
comments]. Science 259, 1616-1619.
Franklin, D. S., Godfrey, V. L., Lee, H., Kovalev, G. I., Schoonhoven, R.,
Chen-Kiang, S., Su, L. and Xiong, Y. (1998). CDK inhibitors p18(INK4c)
and p27(Kip1) mediate two separate pathways to collaboratively suppress
pituitary tumorigenesis. Genes Dev. 12, 2899-2911.
Fritzsch, B., Barald, K. F. and Lomax, M. I. (1998). Early Development of
the Vertebrate Ear. In Development of the Auditory System. (ed. E. W. Rubel,
A. N. Popper and R. R. Fay). pp. 80-145. New York: Springer-Verlag.
Galand, P. e. a. (1989). Cyclin/PCNA immunostaining as an alternative to
tritiated thymidine pulse labelling for marking S phase cells in paraffin
sections from animal and human tissues. Cell Tissue Kinet. 22, 383-392.
Go, M. J., Eastman, D. S. and Artavanis-Tsakonas, S. (1998). Cell
proliferation control by Notch signaling in Drosophila development.
Development 125, 2031-2040.
Harlow, E. and Lane, D. (1988). Antibodies: A Laboratory Manual. Cold
Spring Harbor: Cold Spring Harbor Press.
Harper, J. W. and Elledge, S. J. (1996). Cdk inhibitors in development and
cancer. Curr. Opin. Genet. Dev. 6, 56-64.
Hengst, L. and Reed, S. I. (1996). Translational control of p27Kip1
accumulation during the cell cycle. Science 271, 1861-1864.
Johnston, L. A. (1998). Uncoupling growth from the cell cycle. BioEssays 20,
283-286.
Johnston, L. A. and Edgar, B. A. (1998). Wingless and Notch regulate cellcycle arrest in the developing Drosophila wing. Nature 394, 82-84.
Kawasaki, H., Eckner, R., Yao, T. P., Taira, K., Chiu, R., Livingston, D.
M. and Yokoyama, K. K. (1998). Distinct roles of the co-activators p300
and CBP in retinoic-acid-induced F9-cell differentiation. Nature 393, 284289.
Kelley, M. W., Talreja, D. R. andCorwin, J. T., (1995). Replacement of hair
cells after laser microbeam irradiation in cultured organs of corti from
embryonic and neonatal mice. J. Neuroscience 15, 3013-3026.
Kelley, M. W., Xu, X. M., Wagner, M. A., Warchol, M. E. and Corwin, J.
T. (1993). The developing organ of Corti contains retinoic acid and forms
1590 P. Chen and N. Segil
supernumerary hair cells in response to exogenous retinoic acid in culture.
Development 119, 1041-1053.
Kiyokawa, H., Kineman, R. D., Manova-Todorova, K. O., Soares, V. C.,
Hoffman, E. S., Ono, M., Khanam, D., Hayday, A. C., Frohman, L. A.
and Koff, A. (1996). Enhanced growth of mice lacking the cyclin-dependent
kinase inhibitor function of p27(Kip1). Cell 85, 721-732.
Kranenburg, O., Scharnhorst, V., Van der Eb, A. J. and Zantema, A.
(1995). Inhibition of cyclin-dependent kinase activity triggers neuronal
differentiation of mouse neuroblastoma cells. J. Cell Biol. 131, 227-234.
Lefebvre, P. P., Malgrange, B., Moonen, G. and Van de Water, T. R. (1995).
[Comments on] Regeneration and mammalian auditory hair cells. Science
267, 709-711.
Lim, D. J. and Anniko, M. (1985). Developmental morphology of the mouse
inner ear. Acta Otolaryngol. [Suppl] (Stockh) 422, 1-69.
Millard, S. S., Yan, J. S., Nguyen, H., Pagano, M., Kiyokawa, H. and Koff,
A. (1997). Enhanced ribosomal association of p27(Kip1) mRNA is a
mechanism contributing to accumulation during growth arrest. J. Biol.
Chem. 272, 7093-7098.
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.
Neuroscience 18, 3327-3335.
Nakayama, K., Ishida, N., Shirane, M., Inomata, A., Inoue, T., Shishido,
N., Horii, I., Loh, D.Y. and Nakayama, K. (1996). Mice lacking p27(Kip1)
display increased body size, multiple organ hyperplasia, retinal dysplasia,
and pituitary tumors. Cell 85, 707-720.
Pagano, M., Tam, S. W., Theodoras, A. M., Beer-Romero, P., Del Sal, G.,
Chau, V., Yew, P. R., Draetta, G. F. and Rolfe, M. (1995). Role of the
ubiquitin-proteasome pathway in regulating abundance of the cyclindependent kinase inhibitor p27. Science 269, 682-685.
Pujol, R., Lavigne-Rebillard, M. and Lenoir, M. (1998). Development of
Sensory and Neural Structures in the Mammalian Cochlea. In Development
of the Auditory System. (ed. E. W. Rubel, A. N. Popper and R. R. Fay). pp.
146-192. New York: Springer-Verlag.
Rubel, E. W. (1978). Ontogeny of Structure and Function in the Vertebrates
Auditory System. In Handbook of Sensory Physiology. (ed. M. Jacobson).
pp. 135-220. Berlin: Springer Verlag.
Rubel, E. W., Dew, L. A. and Roberson, D. W. (1995). Mammalian vestibular
hair cell regeneration. Science 267, 701-703.
Ruben, R. J. (1967). Development of the inner ear of the mouse: a
radioautographic study of terminal mitoses. Acta Oto-Laryngologica,
supplementum 220, 1-44.
Ryals, B. M. and Rubel, E. W. (1988). Hair cell regeneration after acoustic
trauma in adult Coturnix Quail. Science 240, 1774-1776.
Sahly, I., El-Amraoui, A., Abitbol, M., Petit, C. and Dufier, J. L. (1997).
Expression of myosin VIIa during mouse embryogenesis. Anat. Embryol.
196, 159-170.
Sher, A. E. (1971). The embryonic and postnatal development of the inner ear
of the mouse. Acta Otolaryngol [Suppl] (Stockh) 285, 1-77.
Sherr, C. J. and Roberts, J. M. (1995). Inhibitors of mammalian G1 cyclindependent kinases. Genes Dev. 9, 1149-1163.
Sicinski, P., Donaher, J. L., Parker, S. B., Li, T., Fazeli, A., Gardner, H.,
Haslam, S. Z., Bronson, R. T., Elledge, S. J. and Weinberg, R. A. (1995).
Cyclin D1 provides a link between development and oncogenesis in the
retina and breast. Cell 82, 621-630.
Skaer, H. (1998). Who pulls the string to pattern cell division in Drosophila?
Trends in Genetics 14, 337-339.
Tikoo, R., Osterhout, D. J., Casaccia-Bonnefil, P., Seth, P., Koff, A. and
Chao, M. V. (1998). Ectopic expression of p27Kip1 in oligodendrocyte
progenitor cells results in cell-cycle growth arrest. J. Neurobiol. 36, 431440.
Vlach, J., Hennecke, S. and Amati, B. (1997). Phosphorylation-dependent
degradation of the cyclin-dependent kinase inhibitor p27. EMBO J 16, 53345344.
Warchol, M. E. and Corwin, J. T. (1995). [Comments on] Mammalian
vestibular hair cell regeneration. Science 267, 704-706.
Warchol, M. E., Lambert, P. R., Goldstein, B. J., Forge, A. and Corwin,
J. T. (1993). Regenerative proliferation in inner ear sensory epithelia from
adult guinea pigs and humans. Science 259, 1619-1622.
Weigmann, K., Cohen, S. M. and Lehner, C. F. (1997). Cell cycle
progression, growth and patterning in imaginal discs despite inhibition of
cell division after inactivation of Drosophila Cdc2 kinase. Development 124,
3555-3563.
Yoho, E. R., Thomopoulos, G. N., Thalmann, I., Thalmann, R. and
Schulte, B. A. (1997). Localization of organ of Corti protein II in the adult
and developing gerbil cochlea. Hearing Research 104, 47-56.
Zhang, P., Wong, C., DePinho, R. A., Harper, J. W. and Elledge, S. J.
(1998). Cooperation between the Cdk inhibitors p27(KIP1) and p57(KIP2)
in the control of tissue growth and development. Genes Dev. 12, 3162-3167.