T ansc ip ion Fac o Epip o in Is Essen ial o Too h
Mo phogenesis by Regula ing Epi helial Cell Fa e
and Too h Numbe
*
□
S
Recei ed o publica ion, Oc obe 9, 2007, and in e ised o m, Decembe 19, 2007 Published, JBC Pape s in P ess, Decembe 21, 2007, DOI 10.1074/jbc.M708388200
Takashi Nakamu a
‡
, Susana de Vega
‡1
, Sa oshi Fukumo o
‡§
, Lucia Jimenez
¶2
, Fe nando Unda
¶
, and Yoshihiko Yamada
‡3
F om he
‡
Labo a o y o Cell and De elopmen al Biology, Na ional Ins i u e o Den al and C anio acial Resea ch, Na ional
Ins i u es o Heal h, Be hesda, Ma yland 20892, he
§
Depa men o Pedia ic Den is y, Facul y o Den al Science, Kyushu
Uni e si y, Fukuoka 812-8582, Japan, and he
¶
Depa men o Cell Biology and His ology, Facul y o Medicine and Den is y,
Uni e si y o he Basque Coun y, 48940 Vizcaya, Spain
In oo h mo phogenesis, he den al epi helium and mesen-
chyme in e ac ecip ocally o g ow h and di e en ia ion o
o m he p ope numbe and shapes o ee h. We p e iously
iden i ied epip o in (Ep n), a gene p e e en ially exp essed in
den al epi helia, di e en ia ed ameloblas s, and ce ain ec o-
de mal o gans. To iden i y he ole o Ep n in oo h de elop-
men , we c ea ed Ep n-de icien mice (Ep n
ⴚ/ⴚ
). Ep n
ⴚ/ⴚ
mice
de eloped an excess numbe o ee h, enamel de iciency, de ec s
in cusp and oo o ma ion, and abno mal den in s uc u e.
Mu an oo h ge ms o med mul iple den al epi helial buds in o
he mesenchyme. In Ep n
ⴚ/ⴚ
mola s, apid p oli e a ion and
di e en ia ion o he inne den al epi helium we e inhibi ed,
and he den al epi helium e ained he p ogeni o pheno ype.
Fo ma ion o he enamel kno , a signaling cen e o cusps,
whose cells di e en ia e om he den al epi helium, was also
inhibi ed. Howe e , mul iple p ema u e nonp oli e a ing
enamel kno -like s uc u es we e o med ec opically. These
den al epi helial abno mali ies we e accompanied by dys egula-
ion o Le -1, which is equi ed o he no mal ansi ion om
he bud o cap s age. T ans ec ion o an Ep n ec o p omo ed
den al epi helial cell di e en ia ion in o ameloblas s and ac i-
a ed p omo e ac i i y o he enamel ma ix ameloblas in gene.
Ou esul s sugges ha in Ep n-de icien ee h, ec opic nonp o-
li e a ing egions likely bud o om he sel - enewable den al
epi helium, o m mul iple b anches, and e en ually de elop
in o supe nume a y ee h. Thus, Ep n has mul iple unc ions o
cell a e de e mina ion o he den al epi helium by egula ing
bo h p oli e a ion and di e en ia ion, p e en ing con inuous
oo h budding and gene a ion.
Ec ode mal o gan de elopmen is a complex p ocess ini i-
a ed by induc i e issue in e ac ions (1). The de eloping oo h
is a classic example o such induc i e p ocesses (2–4). Too h
de elopmen is a con inuous p ocess ha can be di ided in o
he ini ia ion, bud, cap, and bell s ages (5, 6). In mice, oo h
de elopmen begins a emb yonic day (E)
4
11.5 by hickening o
he den al epi helium. The den al lamina unde goes u he p o-
li e a ion and subsequen ly de elops in o he oo h bud. The oo h
bud is o med by he in agina ion o he placode and he conden-
sa ion o mesenchyme cells adjacen o he bud. A he bud s age
(E13.5), den al epi helial cells p oli e a e di usely. A he cap
s age (E14.5), den al epi helial cells di e en ia e in o se e al cell
ypes: he ce ical loop, s ella e e iculum, s a um in e me-
dium, ou e den al epi helium, inne den al epi helium, and
enamel kno . Cell dea h by apop osis wi hin he enamel kno is
c i ical o cusp o ma ion in mola s. These epi helial cells p o-
li e a e a di e en di iding a es o o m he unique shape o
he enamel o gan. Fo example, a nonp oli e a ing epi helial
cell mass o ms he enamel kno a he cen e o he oo h bud.
A he bell s age (E17.5), he den al mesenchyme di e en ia es
in o den in ma ix-sec e ing odon oblas s ha o m den in,
and hen he inne den al epi helial cells di e en ia e in o
enamel ma ix-sec e ing ameloblas s ha o m enamel. In
oden s, inciso s con inue o g ow h ough adul hood.
S udies wi h mu an mice ha e iden i ied a numbe o genes
ha egula e oo h de elopmen and mo phology (7). Fo
example, de iciency o Le -1 (8) o p63 (9, 10) a es s oo h
de elopmen a ea ly s ages. De iciency o Msx1 o Pax9 esul s
in a es o oo h de elopmen a he bud s age (11, 12), whe eas
de iciency o Runx2/Cb a1, Sp3, o Shh causes inhibi ion o
cy odi e en ia ion o ameloblas s and/o odon oblas s (13–
15). Gene knock-ou mice o ameloblas in and amelogenin,
enamel ma ix p o eins, de elop se e e enamel hypoplasia wi h
abno mal ameloblas di e en ia ion (16, 17).
Epip o in (Ep n) was p e iously iden i ied by di e en ial
hyb idiza ion using oo h ge m cDNA mic ochips as a zinc
inge ansc ip ion ac o ha is exp essed in ce ain de el-
oping ec ode mal issues such as ee h, hai ollicles, and
*This wo k was suppo ed by g an s om he In amu al Resea ch P og am
o he NIDCR, Na ional Ins i u es o Heal h ( o Y. Y.), g an s-in-aid o
Resea ch Fellows o he Japan Socie y o he P omo ion o Science om
he Minis y o Educa ion, Science and Cul u e o Japan (G an 15689025
and 17689058 o S. F.), and he Uni e si y o he Basque Coun y (G an
IU05/27, o F. U.). The cos s o publica ion o his a icle we e de ayed in
pa by he paymen o page cha ges. This a icle mus he e o e be he eby
ma ked “ad e isemen ” in acco dance wi h 18 U.S.C. Sec ion 1734 solely o
indica e his ac .
□
S
The on-line e sion o his a icle (a ailable a h p://www.jbc.o g) con ains
a supplemen al able.
1
Suppo ed by a ellowship o he Basque Go e nmen , Spain.
2
Suppo ed by a ellowship o he Uni e si y o he Basque Coun y.
3
To whom co espondence should be add essed: Bldg. 30, Rm. 407, NIDCR,
Na ional Ins i u es o Heal h, 30 Con en D ., MSC 4370, Be hesda, MD
20892-4370. Tel.: 301-496-2111; Fax: 301-402-0897; E-mail: yoshi.
[email p o ec ed].
4
The abb e ia ions used a e: E, emb yonic day; ES, emb yonic s em; TUNEL,
e minal deoxynucleo idyl ans e ase-media ed dUTP nick end-labeling;
PCNA, p oli e a ing cell nuclea an igen; B dU d, b omodeoxyu idine; RT,
e e se ansc ip ion; Phos-pRb, phospho yla ed pRb; Dspp, den in sialo-
phosphop o ein; Ambn, ameloblas in.
THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 283, NO. 8, pp. 4825–4833, Feb ua y 22, 2008
P in ed in he U.S.A.
FEBRUARY 22, 2008•VOLUME 283•NUMBER 8 JOURNAL OF BIOLOGICAL CHEMISTRY 4825
This is an Open Access a icle unde he CC BY license.
limbs (18). Ep n is a homologue o KLF14/Sp6, belonging o
he Sp ansc ip ion ac o supe amily (19, 20). To da e,
he e a e nine Sp genes (Sp1–Sp9) in he mammalian
genome. These Sp p o eins egula e ansc ip ion o a ious
genes in a posi i e and/o nega i e manne . Al hough all
hese Sp ac o s con ain a high deg ee o s uc u al conse -
a ion o he zinc inge domain, each appea s o ha e dis-
inc i e egula o y speci ici y.
In his epo , we c ea ed Ep n-de icien mice ha ini ially
had delayed oo h de elopmen bu la e de eloped an excess
numbe o ee h and had a comple e loss o enamel and a de ec-
i e den in s uc u e. We ound ha Ep n is an essen ial egu-
la o o b anching o oo h buds o o m he p ope numbe o
ee h and o di e en ia ion o he den al epi helium and
mesenchyme.
MATERIALS AND METHODS
Ta ge ing o Emb yonic S em Cells and Gene a ion o
Chime as—A genomic clone including he epip o in locus was
isola ed om a 129/SV genomic lib a y. The a ge ing ec o
was cons uc ed and elec opo a ed in o R1 ES cells (see Fig. 1).
Th ee independen a ge ed ES cell clones p oduced chime-
as ha ansmi ed he mu an allele h ough he ge mline.
He e ozygous p ogeny we e gene a ed wi h backc ossing o
wild- ype C57BL/6 o a leas i e gene a ions. Animal ca e
was gi en in compliance wi h he Na ional Ins i u es o
Heal h guidelines on he Use o Labo a o y and Expe imen-
al Animals.
Cell Cul u e and T ans ec ion—SF2 cells, a p eameloblas
cell line, we e es ablished as desc ibed (21) and we e cul-
u ed in Dulbecco’s modi ied Eagle’s medium/F12 (In i o-
gen) plus 10% e al bo ine se um. SF2 cells di e en ia ed
in o ameloblas s in he p esence o BMP2 (da a no shown).
SF2 cells we e ans ec ed wi h he Ep n exp ession ec o
(Ep n-pcDNA3.1/Myc-His ec o (18)) using Lipo ec amine
2000 (In i ogen). Fo he ameloblas in (Ambn) p omo e
assay, SF2 cells, NIH3T3, o MC3T3-E1 cells we e co ans-
ec ed wi h he Ep n ec o and he Ambn p omo e epo e
cons uc con aining a 2.7-kb Ambn p omo e in he
pNASSb-basic (P omega) ec o using he Lipo ec amine
ki . The cy omegalo i us p omo e -based pcDNA3.1/

-
galac osidase ec o (In i ogen) was used as a posi i e con-
ol. A e a 48-h ans ec ion,

-galac osidase assays we e
pe o med using he

-galac osidase epo e gene assay
chemiluminescen ki (Roche Applied Science).
Tissue Sec ions, Immunohis ochemis y, B dU d Labeling,
and TUNEL Assays—Mouse heads in se e al s ages o de elop-
men we e dissec ed ou and ixed wi h 4% pa a o maldehyde
in phospha e-bu e ed saline o e nigh a 4 °C. Adul oo h is-
sues we e decalci ied wi h 250 mMEDTA/phospha e-bu e ed
saline and embedded in pa a in. Fo his ological analysis, sec-
ions we e s ained wi h Ha is hema oxylin and eosin Y. Fo
immunohis ochemis y, sec ions we e boiled wi h a ge ed
e ie al solu ion (Dako) and incuba ed in 1% bo ine se um
albumin/phospha e-bu e ed saline (blocking eagen ) o 1 h
p io o incuba ion wi h he p ima y an ibody. We used an i-
bodies o ameloblas in (22); amelogenin and enamelin ( om
D . Simme ); den in sialophosphop o ein ( om D . Fishe );
p63 and E-cadhe in (Pha mingen); PCNA (Zymed Labo a o ies
Inc.); Ki67 (No ocas ); B dU d (Roche Applied Science); and
phospho-Rb (Cell Signaling). Rabbi polyclonal an ibody o he
Ep n pep ide ( esidues 33–51) was aised and pu i ied by a pep-
ide a ini y column. P ima y an ibodies we e de ec ed by Cy-3-
conjuga ed and Cy-5-conjuga ed seconda y an ibodies (Jack-
son ImmunoResea ch). Nuclea s aining was pe o med wi h
Hoechs dye (Sigma). P ima y Ep n and phospho-pRb an ibod-
ies we e de ec ed wi h a bio inyla ed an ibody agains abbi
IgG and a idin-conjuga ed pe oxidase (Zymed Labo a o ies
Inc.) and isualized using diaminobenzidine as a ch omogenic
pe oxidase subs a e. To assess cell p oli e a ion, B dU d (100
mg/kg o body weigh ) was injec ed in ape i oneally in o p eg-
nan emales. The mice we e eu hanized 1 h a e injec ion, and
emb yos we e dissec ed and p epa ed o c yosec ions. To
de ec apop o ic cells, we used he ApopTag威Red in si u apo-
p osis de ec ion ki (Chemicon).
RT-PCR—To al RNA was ex ac ed om SF2 cells ha we e
s ably ans ec ed wi h he Ep n ec o and pooled using he
TRIzol ki (In i ogen). A e DNaseI (Sigma) ea men , 2
g
o o al RNA was used o he e e se ansc ip ion o gene a e
cDNA, which was used as a empla e o PCR eac ions wi h
gene-speci ic p ime s (supplemen al Table 1). Fo he semi-
quan i a i e RT-PCR, cDNA was ampli ied wi h an ini ial dena-
u a ion a 95 °C o 3 min and hen 95 °C o 30 s, 60 °C o 30 s,
and 72 °C o 30 s o 25 cycles and a inal elonga ion s ep a
72 °C o 5 min.
In Si u Hyb idiza ion o Tissue Sec ions—Digoxigenin-11-
UTP-labeled single-s and RNA p obes o Le -1, Shh, and Ep n
(18) we e p epa ed using he digoxigenin RNA labeling ki
(Roche Diagnos ics) acco ding o he manu ac u e ’s ins uc-
ions. In si u hyb idiza ion o mouse sec ions was pe o med as
desc ibed p e iously (18).
Scanning Elec on Mic oscope—Inciso s o 12-week-old mice
we e ixed and decalci ied in 4.13% EDTA, dehyd a ed in
g aded e hanol, and p ocessed o embedding in LR Whi e
esin. Calci ied inciso s we e ac u ed a se e al si es along
hei leng h, dehyd a ed in e hanol, c i ical poin -d ied wi h
CO
2
, and examined in a JEOL JSM 5910 LV a iable-p essu e
scanning elec on mic oscope.
RESULTS
Epip o in-de icien Mice and Too h De ec s—We p e iously
epo ed ha Ep n is exp essed in he den al epi helium om
he ea ly ini ia ion s age o he enamel ma ix sec e ion s age
and in di e en ia ed odon oblas s (18). Ep n is also exp essed in
ce ain ec ode mal issues such as he ma ix o hai ollicles
and he apical ec ode mal idge in limbs (18). To iden i y he
ole o Ep n in he de elopmen o he oo h, we c ea ed mu an
mice de icien o Ep n by dele ing exon 2, which encodes he
en i e coding sequence o he Ep n gene using he con en ional
gene knock-ou s a egy (Fig. 1). Th ee independen a ge ed
ES cell clones om R1 cells p oduced chime as ha ansmi ed
he mu an allele h ough he ge mline. He e ozygous p ogeny
we e gene a ed wi h backc ossing o wild- ype C57BL/6 o a
leas i e gene a ions. Homozygous mice (Ep n
⫺/⫺
) a e iable,
bu hei body size is signi ican ly smalle han ha o wild- ype
o he e ozygous mice. Abou 20% o Ep n
⫺/⫺
mice die a abou
Roles o Epip o in in Too h De elopmen
4826 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 283•NUMBER 8•FEBRUARY 22, 2008
he age o 2 mon hs mice o an unknown cause, whe eas he
emaining su i e wi h a li e span simila o wild- ype o he -
e ozygous mice. Homozygous mice (Ep n
⫺/⫺
) a e iable and
ha e se e e de ec s in ee h, as desc ibed below. They also ha e
de ec s in skin, hai ollicles, and
digi o ma ion. De elopmen o
ee h in mu an mice was delayed.
Inciso s and mola s o mu an mice
we e no e up ed a he age o 3
weeks, when wild- ype inciso s and
mola s we e isible (Fig. 2, Aand B).
Howe e , a la e s ages, mu an
inciso s had e up ed wi h excess
numbe s (hype don ia), e.g. mo e
han 50 inciso s by he age o 1 yea
(Fig. 2, Cand D). Mu an inciso s
showed a chalky colo , indica ing a
mine al de ec . Radiog aphs also
showed an excess numbe o mola s
(e.g. mo e han eigh mola s on each
side o he mandible when com-
pa ed wi h h ee mola s on each
side o he con ol mandible by he
age o 1 yea ) (Fig. 2, E–H). The
numbe o mola s in Ep n
⫺/⫺
mice
inc eased wi h age (e.g. 4–5 mola s
on each side o he mandible a he
age o 4 weeks; da a no shown). The
mola s had poo ly de eloped cusps,
which cause malocclusion. The e
we e also se e e de ec s in he den al oo s uc u e and al eola
bones (Fig. 2G). Scanning elec on mic oscope imaging o he
inciso su ace o 12-week-old Ep n
⫺/⫺
mice showed ha he
en i e enamel s uc u e was comple ely missing (Fig. 3, Aand
C). He e ozygous Ep n
⫹/⫺
inciso s had enamel, bu i was wide
han wild ype, wi h an i egula and less compac s uc u e
(Fig. 3B). Den inal ubule o ma ion was also i egula in
Ep n
⫺/⫺
mice when compa ed wi h ha o wild- ype mice, sug-
ges ing de ec s in deposi ion o den in ma ix and c ys alliza-
ion in den in (Fig. 3, D–F). In he he e ozygous Ep n
⫹/⫺
inci-
so s, den inal ubules we e also less compac when compa ed
wi h hose o wild- ype mice.
Mul iple B anching o Too h Ge ms and De ec s in Di e en-
ia ion o Den al Epi helium and Mesenchyme—We pe o med
his ological analysis o de eloping mola s and inciso s o de e -
mine he oo h de ec s in de ail (Fig. 4, Aand B). In ea ly mola
de elopmen h ough he bud s age (emb yonic day E13.5),
he e was no signi ican mo phological di e ence be ween
mu an and con ol Ep n
⫹/⫺
mice. In Ep n
⫺/⫺
mice, a single
den al placode (da a no shown) and bud o med a he ini ial
s age. Howe e , mola s we e p o oundly abno mal by E16.5
(Fig. 4A,panels b and ). Mul iple b anches o he den al epi he-
lium in agina ed in o he den al mesenchyme, whe eas con ol
mola s had single b anching and a mesenchyme condensa ion
a ea. In addi ion, cusp o ma ion was inhibi ed in Ep n
⫺/⫺
mola s. A pos na al day 3, in con ol mola s, ameloblas s and
odon oblas s we e di e en ia ed and well pola ized, and
enamel and den in laye s had o med (Fig. 4A,panel c). In con-
as , in Ep n
⫺/⫺
mice, den al epi helial cells o med mul iple
cell laye s in mu an mola s and we e no well pola ized, and
odon oblas di e en ia ion was delayed and no obse ed a
his s age (Fig. 4A,panel g). Consequen ly, he enamel laye was
FIGURE 1. Ta ge ed dis up ion o he mouse Ep n locus. A, pa ial es ic ion maps o he wild- ype Ep n
locus, a ge ing ec o , and a ge ed allele. The en i e coding egion o exon 2 o he Ep n gene, including he
zinc inge domain, was eplaced wi h he pGK-neo a ge ing casse e. Res ic ion enzyme si es a e shown
abo e he line, and he leng hs o DNA agmen s gene a ed by an NdeI diges ha hyb idize wi h he genomic
lanking p obe a e indica ed by he a ow lines.B, Sou he n blo analysis o NdeI-diges ed genomic DNA om
ES cells and ails o wild- ype (WT), he e ozygous (
⫹/⫺
), and homozygous (
⫺/⫺
) mice. Th ee independen
a ge ed ES cell clones om R1 cells p oduced chime as ha ansmi ed he mu an allele h ough he ge m-
line. 10.8kb, a ge ed allele; 6.4kb, wild- ype allele.
+/-E
G
in
in
-/-
F
H
**
*
***
*
** *
*
*
***
*
**
in
al
al
+/-
-/-
-/- +/-AB C -/-D+/-
FIGURE 2. De ec i e oo h o ma ion. A–D, inciso s o 3-week-old (Aand B)
and 12-mon h-old (Cand D) he e ozygous (Ep n
⫹/⫺
)(Aand C) and homozy-
gous (Ep n
⫺/⫺
)(Band D) mice. B, Inciso de elopmen o 3-week-old Ep n
⫺/⫺
mice is delayed. D, 12-mon h-old Ep n
⫺/⫺
KO mice had an excess numbe o
inciso s wi h a mine al de ec . E–H, adiog am analysis o 6-mon h-old mouse
heads om Ep n
⫹/⫺
(Eand F, enla ged) and Ep n
⫺/⫺
mice (Gand H, enla ged).
as e isk, mola s; in, inciso s; a owheads, al, al eola bones (a owheads); ,
oo s.
Roles o Epip o in in Too h De elopmen
FEBRUARY 22, 2008•VOLUME 283•NUMBER 8 JOURNAL OF BIOLOGICAL CHEMISTRY 4827
no o med, and den in o ma ion was delayed. Fu he , he
den al epi helium o med mul iple buds ha esul ed in excess
u u e mola s. In 3-week-old con ol mola s, he enamel o gan
o he oo h ge m was ully de eloped, whe eas in mu an
mola s, he den al lamina was no sepa a ed om he o al epi-
helium laye , enamel was no o med, and he c own o he
mola s had educed den al cusp sizes (Fig. 4A,panels d and h).
In addi ion, se e al mola s had de eloped om a single den al
lamina.
The ini ial s age o inciso de elopmen is simila o ha o
mola s. Howe e , in la e s ages, he epi helial cells only on he
labial side di e en ia e in o ameloblas s and p oduce enamel
ma ix, esul ing in asymme ical enamel o ma ion in inciso s.
The con inuous e up ion o inciso s is main ained by a con in-
uous supply o p ogeni o cells in he apical bud localized in he
ce ical loop (Fig. 4B,panel a,cl). E16.5 con ol inciso s o med
a pola ized inne den al epi helium cell laye on he labial side
bu no he lingual side, and odon oblas s and a hin den in
ma ix laye we e p esen on bo h sides. In mu an inciso s a
his s age, den al epi helia and mesenchyme we e diso ganized
and undi e en ia ed when compa ed wi h con ol Ep n
⫹/⫺
inciso s (Fig. 4B,panel ). The Ep n
⫺/⫺
inciso ge m did no
sepa a e om he o al epi helium (Fig. 4B,panels b and g), and
i de eloped a glo e-like s uc u e illed wi h undi e en ia ed
epi helial cells and mul iple buds su ounding he mesen-
chyme. Simila o mola s, pos na al day 3 mu an inciso s
o med mul iple apical buds, which o igina ed om a single
den al lamina, lacking enamel and
den in laye s (Fig. 4B,panels g and
h). A pos na al day 10, Ep n
⫺/⫺
inciso s e en ually o med den in
su ounded by a hin laye o undi -
e en ia ed epi helial cells (Fig. 4B,
panels i and j).
De ec s in Ameloblas and Odon-
oblas Di e en ia ion—We nex
examined he exp ession o oo h
ma ix p o eins as di e en ia ion
ma ke s in 10-day-old mola s (Fig.
5). Ameloblas in and enamelin, spe-
ci ic o di e en ia ed ameloblas s,
we e exp essed by ameloblas s and
sec e ed in o he enamel ma ix
laye o Ep n
⫹/⫺
mola s, whe eas
nei he p o ein was de ec ed in
Ep n
⫺/⫺
mola s (Fig. 5, A,B,E, and
F). Den in sialophosphop o ein
(Dspp) was exp essed by bo h
ameloblas s and odon oblas s in
Ep n
⫹/⫺
mola s. Howe e , in
Ep n
⫺/⫺
mola s, Dspp was nega-
i e in he epi helium and signi i-
can ly educed in odon oblas s
and he den in laye (Fig. 5, Cand
G). These esul s indica e ha
den al epi helial cell di e en ia-
ion in o ameloblas s was blocked
in Ep n
⫺/⫺
mice and sugges ha
a
e
b
g
d
h
b c
A
B
a d e
gh i j
+/-
mola
+/-
inciso
E13.5 E16.5 P3 3 weeks
P3 P10 P10P3E16.5
o al ca i y (oc), an e io (an ), pos e io (pos), labial (la).
oc
oc
an pos la
*
*
*
*
OC
OC
oc
an pos
oc
an pos
oc
an pos
-/-
mola
-/-
inciso
dm
dm
de
de
en
den
ide
ode
de
ode de den
am
od em
de
am
en
den
od
den
am
de
de
dm
dm
de
dm
dm
de
dm
dm
de
dm
em
cl
cl
c
od
am
en de
dm dm
ocla
oclaocla
FIGURE 4. Abno mal oo h de elopmen . A, his ology o Ep n
⫹/⫺
(panels a–d) and Ep n
⫺/⫺
mola s (panels
e–h). Panels a,b,e, and , on al sec ions, ⫻40; panels c and g,⫻20; panels d and h,⫻10). A ows, mul iple
b anching. B, his ology o Ep n
⫹/⫺
(panels a–e) and Ep n
⫺/⫺
inciso s (panels –j). Panels c,d,e,h,i, and j, on al
sec ions; Panels a,b, , and g, sagi al sec ions. Panels a,c, , and h,⫻20; panels b,d,g, and i,⫻10; panels e and j,
⫻40 (enla ged he boxed a eas in panels d and i). de, den al epi helium; dm, den al mesenchyme; ode, ou e
den al epi helial cells; ide, inne den al epi helial cells; am, ameloblas s; cl, ce ical loop; dn, den in; en, enamel;
em, enamel ma ix; od, odon oblas s; den, den in. Labium (la), o al ca i y (OC), an e io (an ), and pos e io (pos)
indica e ana omical posi ions. The as e isk indica es den al lamina. No e: The enamel laye was de ached om
he den in laye du ing sample p epa a ion in con ol mola s (A,panels c and d) and inciso s (B,panels d and e).
ABC
esin
esin
esin
oe oe
ie
ie
de de
de
D E F
+/+ +/- -/-
FIGURE 3. Absence o enamel and de ec s o den in. A–C, scanning elec on
mic oscope images o he su ace o inciso s o 12-week-old Ep n
⫹/⫹
,Ep n
⫹/⫺
,
and Ep n⫹
⫺/⫺
mice. Wild- ype Ep n
⫹/⫹
mice (A) had well c ys allized ou e (oe)
and inne enamel (ie), whe eas homozygous Ep n
⫺/⫺
mice (C) we e comple ely
missing he en i e enamel s uc u e. He e ozygous Ep n
⫹/⫺
inciso s had enamel,
bu i was wide , wi h an i egula and less compac s uc u e (B). D–F, scanning
elec on mic oscope images o he inciso den in (de). Den inal ubule o ma ion
was i egula in Ep n
⫺/⫺
mice (F) when compa ed wi h ha o wild- ype (D) mice,
sugges ing de ec s in deposi ion o den in ma ix and c ys alliza ion in den in. In
he he e ozygous Ep n
⫹/⫺
inciso s, den inal ubules we e also less compac
when compa ed wi h hose o he wild ype (E). The small squa e on he on al
sec ion o he inciso shows he a ea o enla gemen .
Roles o Epip o in in Too h De elopmen
4828 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 283•NUMBER 8•FEBRUARY 22, 2008
Ep n is equi ed o ameloblas di e en ia ion and ha i
egula es exp ession o Dspp in odon oblas di e en ia ion.
Dys egula ion o Le -1 and Shh—To add ess he molecula
basis o oo h de ec s, we analyzed he exp ession o Le -1 and
Shh, which egula es oo h mo phogenesis (Fig. 6). Dele ion o
Le -1, a downs eam ansc ip ion ac o o Wn /

-ca enin,
esul s in an absence o all ee h (23). Whole-moun in si u
hyb idiza ion o he E11.5 mouse mandible showed Le -1
exp ession in he den al placode o he con ol mandible. In
Ep n
⫺/⫺
mice, Le -1 exp ession was up- egula ed in bo h he
epi helium and he mesenchyme o E11.5 maxilla and mandible
p ocesses (Fig. 6, Aand E). A E14.5, Le -1 exp ession was local-
ized in he enamel kno and su ounding den al mesenchyme in
con ol mice. Howe e , in mu an ee h, Le -1 exp ession was
no localized in he enamel kno bu a he ex ended in o he
inne den al epi helium laye (Fig. 6, Band F). The up- egula-
ion o Le -1 exp ession and i s ex ended exp ession we e also
obse ed in he mesenchyme. Condi ional de iciency o Shh in
he den al epi helium inhibi s di e en ia ion o bo h he den al
epi helium and he den al mesenchyme (25). In E14.5 Ep n
⫹/⫺
mola s, Shh exp ession was es ic ed o he enamel kno . In he
Ep n
⫺/⫺
mola s, Shh exp ession a he enamel kno was signi -
ican ly educed, indica ing ha enamel kno o ma ion was
inhibi ed in Ep n
⫺/⫺
mola s (Fig. 6, Cand G). In con ol E14.5
mola s, Ep n exp ession was obse ed in he inne den al epi-
helium (Fig. 6D). Howe e , in
Ep n
⫺/⫺
mola s, Ep n was com-
ple ely absen as expec ed.
De ec s in Rapid P oli e a ion o
he Inne Den al Epi helium and
Fo ma ion o Mul iple P ema u e
Enamel Kno -like S uc u es—Since
we demons a ed ha Ep n exp es-
sion p omo es cell p oli e a ion in
cul u e (18), we examined cell p o-
li e a ion by immunos aining o
p oli e a ion ma ke s Ki67 and
PCNA. In con ol E16.5 mola s,
Ki67 and PCNA we e s ained
s ongly in he ou e and inne epi-
helial laye s and condensing mes-
enchyme cells (Fig. 7, A,panels a
and c). In Ep n
⫺/⫺
mola s, s aining
in hese cells was dec eased and dis-
pe sed in he oo h ge m (Fig. 7A,
panels b and d). In con ol Ep n
⫹/⫺
mola s, Ep n exp ession was
es ic ed o he inne den al epi he-
lial cells and he enamel kno cells a
his s age (Fig. 7A,panels e and ). A
E17.5, in Ep n
⫹/⫺
mola s, he inne
epi helial cell laye was expanded,
and i s ongly exp essed Ki67 (Fig.
7B,panel a). In Ep n
⫺/⫺
mola s,
Ki67-posi i e cells ( ed) we e dis-
pe sed in he whole oo h ge m and
also in su ounding mesenchyme
cells (Fig. 7B,panel b). I is di icul
o dis inguish apid and slow p oli e a ing cells by Ki67 and
PCNA immunos aining because hese p o eins a e exp essed
om la e G
1
and ha e a ela i ely long hal -li e. To de ec ap-
idly p oli e a ing cells, we pe o med a sho e m B dU d
inco po a ion assay and ound ha he inne den al epi helial
cells we e apidly p oli e a ing (Fig. 7B,panel c, yellow). In
Ep n
⫺/⫺
mola s, B dU d-posi i e cells we e signi ican ly
educed (Fig. 7B,panel d). The enamel kno epi helial cells no -
mally do no p oli e a e and e en ually die by apop osis (26). In
E17.5 Ep n
⫹/⫺
mola s, he e we e TUNEL-posi i e cells in sec-
onda y enamel kno s (Fig. 7B,panel e,a owheads). In Ep n
⫺/⫺
mola s, mul iple ec opic TUNEL-posi i e a eas we e obse ed
in he enamel o gan (Fig. 7B,panel ). Unlike he enamel kno s
o Ep n
⫹/⫺
mola s, Shh, one o he enamel kno ma ke s, did
no exp ess in hese a eas, so hese TUNEL-posi i e spo s likely
ep esen mul iple p ema u e enamel kno s. The pRb p o ein
plays a c i ical ole in G
1
-S checkpoin con ol (27). I is phos-
pho yla ed (Phos-pRb) a he end o G
1
and emains so
h oughou S-G
2
. Unphospho yla ed pRb p e en s cell p oli -
e a ion, whe eas phospho yla ion o pRb esul s in he ac i a-
ion o a numbe o genes equi ed o S phase ansi ion (28).
We ound ha in con ol Ep n
⫹/⫺
E17.5 mola s, mos o he
inne den al epi helial cells we e posi i e o Phos-pRb immu-
nos aining, and enamel kno epi helial cells we e nega i e o
Phos-pRb, co esponding o he p oli e a i e s a e and nonp o-
+/-
-/-
B
Ambn Enam
Enam
en
F
Dspp
den
A
Ambn
en
am
E
Dspp
en
den
am
den
en
am
Gden
H
CD
FIGURE 5. De ec s in oo h ma ix p o ein exp ession. A–H, exp ession o oo h ma ix p o eins in 10-day-old
Ep n
⫹/⫺
(A–D) and Ep n
⫺/⫺
mola s (E–H). Dand H, his ology o he same se ial sec ions. Ambn, Enam, and Dspp
we e no exp essed in den al epi helial cells o Ep n
⫺/⫺
mola s. Dspp exp ession in odon oblas s was educed
in Ep n
⫺/⫺
mola s. en, enamel; den, den in; enam, enamelin. Magni ica ion, ⫻10.
GH
ABC D
+/-
-/-
Le -1 Shh Ep nLe -1
EF
FIGURE 6. Dys egula ion o Le -1 and Shh. A–G,in si u hyb idiza ion o Le -1, Shh, and Ep n mRNA in Ep n
⫹/⫺
(A–D) and Ep n
⫺/⫺
mola s (E–H): Le -1 in E11.5 heads (Aand E) and E14.5 mola s (Band F); Shh in E14.5 (Cand G);
Ep n in E14.5 (Dand H). Le -1 exp ession was up- egula ed in he epi helium and mesenchyme o E11.5 maxilla
and mandible p ocesses and was no localized in he enamel kno bu ex ended in o he inne den al epi he-
lium laye in Ep n
⫺/⫺
mice. Shh exp ession o he enamel kno in E14.5 Ep n
⫺/⫺
mola s was educed.
Roles o Epip o in in Too h De elopmen
FEBRUARY 22, 2008•VOLUME 283•NUMBER 8 JOURNAL OF BIOLOGICAL CHEMISTRY 4829
li e a i e s a e o hese cells, espec i ely (Fig. 7B,gand i). In
Ep n
⫺/⫺
mola s, Phos-pRb le els we e lowe , and he Phos-
pRb-posi i e egions we e pa chy and dis up ed wi h Phos-
pRb-nega i e egions (Fig. 7B,panels h and j). The Phos-pRb-
nega i e egions con ained nonp oli e a ing cells and likely
ep esen mul iple p ema u e enamel kno -like s uc u es.
Ep n was exp essed in bo h he inne den al epi helial laye and
he seconda y enamel kno s in Ep n
⫹/⫺
mola s (Fig. 7B,panel
k). Ou esul s sugges ha Ep n is equi ed o di e en ia ion
o he den al epi helium in o inne and enamel kno epi helia.
Ep n P omo es Di e en ia ion o Den al Epi helial Cells in
Cul u e—In Ep n
⫺/⫺
mice, ameloblas di e en ia ion was
blocked, and no enamel o med, sugges ing ha Ep n is
equi ed o ameloblas di e en ia ion. To de e mine he ac i -
i y o Ep n in den al epi helial di e en ia ion, we ans ec ed
he Ep n exp ession ec o in o p eameloblas cell line SF2,
which has di e en ia ion po en ial, and analyzed he exp es-
sion o ameloblas ma ke s by RT-PCR (Fig. 8A). We ound ha
Ep n ans ec ion inc eased he exp ession o Ambn and Dspp.
Le -1 is a ma ke o p eameloblas cells, and i s exp ession
dec eases when he cells di e en ia e in oo h de elopmen .
Le -1 exp ession was educed in he Ep n- ans ec ed cells,
whe eas ib onec in and hypoxan hine phospho ibosyl ans-
e ase exp ession emained he same. To u he con i m he
di e en ia ion-p omo ion ac i i y o Ep n, we analyzed he
Ambn p omo e ac i i y o an ameloblas in epo e cons uc
in cells co ans ec ed wi h he Ep n exp ession ec o (Fig. 8B).
Ep n ans ec ion enhanced Ambn p omo e ac i i y in SF2
cells, bu i did no enhance he p omo e ac i i y in NIH3T3 o
MC3T3-E1 cells, indica ing cell ype-speci ic ac i i y o Ep n.
DISCUSSION
We ha e iden i ied se e al oo h de ec s in Ep n
⫺/⫺
mice:
comple e absence o enamel, de ec in cusps, abno mal den in,
and supe nume a y ee h. In Ep n
⫺/⫺
ee h, di e en ia ion o
he den al epi helium in o he enamel kno and inne den al
epi helia was impai ed. A he bud s age o oo h de elopmen ,
Ep n exp ession is es ic ed o he inne den al epi helium, and
a he cap s age, when he enamel kno s a s o o m, i s exp es-
sion con inues in he expanding inne den al epi helial laye
and also s a s in he enamel kno , a signaling cen e esponsi-
ble o cusp o ma ion. Thus, Ep n exp ession pa e ns in no -
PCNA
+/-
-/-
Ki67 Ecad/Ep n
+/-
-/-
Ki67 B dU TUNEL
a
b
c
d
e
A
Bdl
dl
dl
dl
ide
ide
dm
dm
dm
dm
a
b
c
d
e
g
h
Phos-pRb Ep n
dl
ide
**
dm
**
i
j
k
l
*
*
*
*
+/-
-/-
FIGURE 7. De ec s in apid den al epi helial cell p oli e a ion and enamel
kno o ma ion. A, cell p oli e a ion and apop osis. Panels a– , immuno-
s aining o Ki67 (panels a and b), PCNA (panels c and d), and Ep n (panels e and
) in E16.5 Ep n
⫹/⫺
(panels a,c, and e) and Ep n
⫺/⫺
mola s (panels b,d, and ). In
con ol mola s, Ki67 and PCNA s aining was s ong in ou e and inne den al
epi helial laye s (panels a and c), whe eas in Ep n
⫺/⫺
mola s, Ki67- and PCNA-
posi i e cells we e dispe sed, and he e we e no clea epi helial cell bound-
a ies. In con ol mola s, Ep n was es ic ed o he inne den al epi helial cells.
Ep nwas absen in Ep n
⫺/⫺
mola s.Panels eand , double-immunos aining o
E-cadhe in (g een) o s aining epi helial cells and Ep n ( ed). Magni ica ion,
⫻20. dl, den al lamina; ide, inne den al epi helial cells; dm, den al mesen-
chyme. B, Ki67 s aining (panels a and b), B dU d (B dU) labeling (g een); ed,
nuclea (panels c and d); TUNEL assays ( ed, TUNEL-posi i e; blue, nuclea )
(panels e and ); Phos-pRb (panels g and h); and enla ged images (panels i and
h)o heboxed a eas in panels g and hand Ep n (panels k and l) s aining o
E17.5 Ep n
⫹/⫺
(panels a,c,e,g,i, and k) and Ep n
⫺/⫺
mola s (panels b,d, ,h,j,
and l). The inne den al epi helial cells in Ep n
⫹/⫺
mola s we e s ongly
labeled wi h B dU d bu no hose in Ep n
⫺/⫺
mola s. Seconda y enamel
kno s (a owheads)inEp n
⫹/⫺
mola s we e TUNEL-posi i e, and in Ep n
⫺/⫺
mola s, he e we e mul iple posi i e spo s. In Ep n
⫹/⫺
mola s, he inne den al
epi helial cells bu no enamel kno epi helial cells (as e isk) we e posi i e o
Phos-pRb, whe eas in Ep n
⫺/⫺
mola s, Phos-pRb-posi i e egions we e dis-
up ed by unphospho yla ed pRb egions. Bo h he inne den al and he
enamel kno epi helia exp essed Ep n in Ep n
⫹/⫺
mola s. as e isk, seconda y
enamel kno .
None
Epip o in
Vec o
Epip o in
Ameloblas in
Hp
Fib onec in
Dspp
Le 1
A
Den al epi helium
NIH3T3
MC3T3-E1
0
50
100
75
25
Epip o in
Ambn p omo e ac i i y (%)
Con ol
B
FIGURE 8. P omo ion o p eameloblas di e en ia ion in o ameloblas s
by Ep n. A, RT-PCR analysis o SF2 cells, a p eameloblas cell line, ans ec ed
wi h he Ep n exp ession ec o (Ep n-pcDNA3.1/Myc-His ec o ). Exp ession
o mRNA o ameloblas in and Dspp, ma ke s o ameloblas s, was inc eased
by Ep n o e exp ession. In con as , exp ession o Le -1 mRNA, a ma ke o
p eameloblas s, was dec eased by Ep n o e exp ession. Hp , hypoxan hine
phospho iboxyl ans e ase. B, ameloblas in p omo e ac i i y o he Ep n
epo e cons uc in cells co ans ec ed wi h he Ep n exp ession ec o . Ep n
p omo e ac i i y was inc eased in SF2 cells bu no in NIH3T3 o MC3T3-E1
cells.
Roles o Epip o in in Too h De elopmen
4830 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 283•NUMBER 8•FEBRUARY 22, 2008
mal oo h de elopmen co ela e wi h enamel hypoplasia and
cusp de ec s obse ed in Ep n
⫺/⫺
ee h. In Ep n
⫺/⫺
mola s,
Le -1 and Shh, ma ke s o he enamel kno , we e educed in he
egion co esponding o he enamel kno , indica ing ha he
o ma ion o he enamel kno was inhibi ed. These esul s sug-
ges ha Ep n is equi ed o di e en ia ion o di e en cell
ypes o he enamel o gan.
One o he cha ac e is ic ea u es o he inne den al epi he-
lium is apid p oli e a ion, as shown by sho e m B dU d
labeling (Fig. 7B,panel c). Howe e , in Ep n
⫺/⫺
mola s, he
numbe o apidly p oli e a ing cells was signi ican ly less han
ha o con ol mola s (Fig. 7B,panel d). These esul s sugges
ha Ep n is equi ed o apid cell p oli e a ion o he inne
den al epi helium, which is consis en wi h ou p e ious epo
ha Ep n ans ec ion in o p eameloblas s p omo es cell p oli -
e a ion (18). Fu he mo e, a he cap s age o Ep n
⫺/⫺
mola s,
Le -1 and BMP-4 (da a no shown) con inued o be exp essed in
he cell laye co esponding o he inne den al epi helium and
in mesenchyme condensa ion a eas, whe eas in wild- ype
mola s, Le -1 and BMP-4 exp ession is es ic ed o he enamel
kno and a mesenchyme condensa ion. Phos-pRb immuno-
s aining o Ep n
⫺/⫺
mola s showed i egula cell laye s in he
uppe a ea o he enamel o gan (Fig. 7B,panels h and j). Phos-
pho yla ion o pRb a he G
1
/S ansi ion phase dissocia es E2F
om he inac i e E2F-pRB complex and p omo es cell cycle
p og ession. In con ol Ep n
⫹/⫺
mola s, s ong phospho yla-
ion o pRb was obse ed in he inne den al epi helial laye , bu
pRB was no phospho yla ed in he enamel kno , which is con-
sis en wi h he B dU d labeling esul s. In Ep n
⫺/⫺
mola s,
Phos-pRB-posi i e laye s ha e mul iple dis up ions by non-
phospho yla ed pRB-con aining, nonp oli e a i e cell egions.
The nonp oli e a ion p ope y o he unique cell masses in
Ep n
⫺/⫺
mola s is simila o ha o he enamel kno o con ol
Ep n
⫹/⫺
mola s, and hese s uc u es likely ep esen mul iple
p ema u e enamel kno -like s uc u es. The cell cycle-exi
enamel kno epi helial cells and pos mi o ic di e en ia ed
ameloblas s exp ess Ep n, and hey e en ually die by apop osis.
The e o e, Ep n unc ions no only o p omo e p oli e a ion
and di e en ia ion o he inne den al epi helium, bu a la e
s ages, also p omo e and main ain he di e en ia ion s a e o
ameloblas s and enamel kno epi helia. Ou da a sugges ha
Ep n ac i i y o apid cell p oli e a ion is media ed in pa
h ough phospho yla ion o pRb, which leads o he elease o
E2F om a pRB-E2F complex o ac i a e genes o cell p oli e -
a ion. In addi ion, ou esul s sugges ha Ep n is equi ed o
cell ype speci ica ion o he den al epi helium. I has been
shown ha pRb no only supp esses cell cycle p og ession bu
also egula es cell lineage speci ica ion o se e al ypes o s em
cells in myogenesis, adipogenesis, and hema opoiesis by coop-
e a ing wi h cell ype-speci ic ansc ip ion ac o s (29). Fo
example, pRb coope a es wi h MyoD o ac i a e muscle-spe-
ci ic genes (30, 31). G ow h a es is appa en ly no su icien o
induce ameloblas di e en ia ion, simila o hese issues. Ep n
may wo k oge he wi h pRb as a cell ype-speci ic ac o o cell
ype speci ica ion in odon ogenesis.
We obse ed a delay o oo h de elopmen in Ep n
⫺/⫺
mice.
This is likely in pa due o delayed and impai ed di e en ia ion
o he den al mesenchyme. In Ep n
⫺/⫺
mola s, he e we e mul-
iple mesenchyme condensa ion a eas, and he exp ession o
Le -1 (Fig. 6F) and BMP-4 (da a no shown) was up- egula ed in
hese a eas. Since in ea ly s ages Ep n is no exp essed in he
den al mesenchyme, hese abno mali ies o he den al mesen-
chyme mus be caused by impai ed signaling om Ep n-de i-
cien den al epi helium. Howe e , la e , Ep n is exp essed in
di e en ia ed odon oblas s. In Ep n
⫺/⫺
mice, exp ession o
Dspp by odon oblas s was signi ican ly educed (Fig. 5G), and
den inal ubules we e abno mal. Thus, Ep n is equi ed o co -
ec di e en ia ion o he odon oblas ic mesenchyme cells.
One o he s iking pheno ypes o Ep n-de icien mice is
supe nume a y ee h in bo h mola s and inciso s. In Ep n
⫺/⫺
mice, ini ial pa e ning o oo h de elopmen was no mal, and a
single den al placode o med. A la e s ages, he den al epi he-
lium begins o in agina e in o he mesenchyme a mul iple si es
wi hin he bud and o ms mul iple b anches ha esul in
hype don ia. Canonical Wn /

-ca enin signaling and i s down-
s eam molecule Le -1 a e essen ial o oo h de elopmen (8).
Le -1 exp ession in he den al epi helium is equi ed o he
bud-cap ansi ion and no mal oo h de elopmen , sugges ing
he c i ical ole o Le -1 exp ession in he inne den al epi he-
lium o he induc ion o he mesenchyme o o m a den al
papilla (8, 23). O e exp ession o Le -1 unde he con ol o he
epi helial cell-speci ic K14 p omo e in ansgenic mice de el-
ops abno mal in agina ions o he den al epi helium in he
mesenchyme, sugges ing ha he pe u ba ions in he epi he-
lium and mesenchyme a e caused by Le -1 exp ession in he
epi helium (32). Mo e ecen ly, i was ound ha cons i u i e
ac i a ion o

-ca enin by an exon 3 dele ion o he loxed-

-
ca enin gene wi h K14-C e p oduces supe nume a y ee h (33).
Mu an mice de elop mul iple ec opic buds in o he mesen-
chyme wi h supe nume a y enamel kno s, sugges ing ha
excess ee h de elop by a enewal p ocess in which enamel
kno s bud o om he den al epi helium. In E16

-ca enin
mu an mola s, Ep n is exp essed in spo y a eas and appa en ly
no in he inne den al epi helium. Some o hese spo y Ep n-
posi i e a eas a e hough o be ec opic enamel kno s because
o he ma ke s o he enamel kno a e exp essed in hese a eas.
These pheno ypes seem o ha e some simila i ies o hose o
Ep n
⫺/⫺
mola s whe e Ep n is absen in he inne den al epi he-
lium, and mul iple p ema u e enamel kno -like s uc u es o m
in Ep n
⫺/⫺
mola s, sugges ing ha supe nume a y ee h in
hese mu an mice de elop in pa h ough a common mecha-
nism. In he absence o Ep n, den al epi helial cells ail o di e -
en ia e in o apidly p oli e a ing inne den al epi helium and
e ain a slow-p oli e a ing bu sel - enewable p ogeni o phe-
no ype, and mul iple nonp oli e a ing and imma u e enamel
kno -like s uc u es a e o med. These nonp oli e a ing a eas
likely bud o om he undi e en ia ed and sel - enewable den-
al epi helium, leading o supe nume a y ee h, which is simila
o he hypo hesis p oposed by Ja inen e al. (33) in a ecen
epo . The con inuous inc ease in he numbe o mola s wi h
age in Ep n
⫺/⫺
mice is unique and is likely a ibu able in pa o
he sus ained p ogeni o pheno ype o he den al epi helium.
Le -1 may play a ole in sus aining he p ogeni o pheno ype
since i s exp ession is inhibi ed in he inne den al epi helium a
he cap s age in wild- ype mola s. This is consis en wi h ou
da a and p e ious epo s showing ha Le -1 and

-ca enin a e
Roles o Epip o in in Too h De elopmen
FEBRUARY 22, 2008•VOLUME 283•NUMBER 8 JOURNAL OF BIOLOGICAL CHEMISTRY 4831
dys egula ed and ha mul iple oo h budding occu s (33). In
Ep n
⫺/⫺
mice, ee h a e no coa ed a all by enamel bu ha e
den in. Ep n
⫺/⫺
den al mesenchyme cells, which a e loca ed
adjacen o he p ema u e inne den al epi helium, can di e -
en ia e in o den in-sec e ing odon oblas s, al hough wi h some
delays. These esul s sugges ha odon oblas s can be gene -
a ed om den al mesenchyme cells when he den al epi helium
mig a es and in e ac s wi h hem. F om he clinical ansla-
ional pe spec i e, hese indings a e signi ican o issue engi-
nee ing o egene a e ee h since in adul ee h, den al mesen-
chymal cell a e a ailable bu no den al epi helial cells.
Mammalian den i ion is complex wi h di e ences be ween
species in oo h numbe , shape, and enewal sys em, bu in
mos mammalians, he inciso s, canines, and p emola s
appea in wo gene a ions (i.e. p ima y and pe manen den-
i ion). The ini ia ion o he pe manen den al laminas
begins a he lingual side o he p ima y den al laminas by
b anching. The pe manen den al lamina mig a es in o he
den al mesenchyme and main ains a mul ipo en den al epi-
helial p ogeni o cell s a e and subsequen ly o ms a pe ma-
nen oo h bud. The mechanism o he wo gene a ions in
den i ion is no ully unde s ood. Ep n is an e olu iona ily
conse ed gene in mammals, and we belie e ha Ep n may
play an impo an ole in den al pa e ning. Elucida ing Ep n
unc ion will lead o he de elopmen o new issue enginee -
ing o oo h egene a ion and he iden i ica ion o human
diseases caused by Ep n mu a ions.
Ep n exp ession con inues om he ea ly s ages o den al
epi helial cells o di e en ia ed ameloblas s. A he E14.5 cap
s age, Ep n exp ession is es ic ed o he inne den al epi he-
lium. Ou co ans ec ion s udies using a den al epi helial cell
line and he Ep n exp ession ec o showed ha Ep n p omo es
he di e en ia ion o den al epi helial cells in o ameloblas phe-
no ypes. Ep n ac i a es he exp ession o mRNA o amelo-
blas in and Dspp. We also ound ha he ameloblas in p o-
mo e epo e cons uc was ac i a ed in den al epi helial
cells when he cells we e co ans ec ed wi h he Ep n exp es-
sion ec o . These esul s sugges ha Ep n is a key an-
sc ip ion ac o o he induc ion o ameloblas di e en ia-
ion and is essen ial o ameloblas in exp ession. The
ac i a ion o he ameloblas in p omo e by Ep n is cell ype-
speci ic. These esul s sugges ha he p omo e ac i a ion
is likely equi ed o an addi ional ac o (s) o be p esen in
p eameloblas s o in ameloblas s. The ac o (s) may wo k
coope a i ely wi h Ep n o independen ly o ac i a e he
ameloblas in p omo e . The coope a i e ac i i ies o Ep n
wi h o he ac o s may explain he mul iple unc ions o Ep n
(e.g. i p omo es apid cell p oli e a ion o he den al epi he-
lium and he di e en ia ion o enamel kno epi helium and
ameloblas ). I is also possible ha Ep n exe s di e en
unc ions depending on i s exp ession le els.
Sp3-de icien mice de elop enamel hypoplasia (14). The
inne den al epi helium appa en ly di e en ia es o o m a
single laye o pola ized ameloblas s in Sp3-de icien ee h.
Howe e , Sp3-de icien mice do no p oduce he enamel
ma ix p o eins ameloblas in o amelogenin. We ound ha
Sp3 is s ongly exp essed by he den al epi helium in Ep n-
de icien ee h (da a no shown). The e o e, Sp3 is no a
di ec a ge gene o Ep n. These esul s indica e ha Sp3
alone is no su icien o ameloblas in exp ession and sug-
ges ha Sp3 wo ks wi h ano he p o ein ac o (s) o amelo-
blas in gene ac i a ion.
In his s udy, we ha e unco e ed no el di e se oles o Ep n
in oo h de elopmen . We demons a e ha Ep n is essen ial
o oo h mo phogenesis by egula ing oo h numbe s and den al
epi helial cell a e. This p ocess appea s o in ol e coo dina ed
signaling du ing he in e ac i e di e en ia ion o he den al epi-
helium and mesenchyme. A la e s ages, Ep n ac s as a key egu-
la o o ameloblas di e en ia ion and also p obably o main e-
nance o he di e en ia ion s a e. Thus, ou s udies p o ide new
dimensions o unde s anding o he molecula mechanism go -
e ning he complex p ocesses in oo h mo phogenesis.
Acknowledgmen s—We hank La y Fishe o he Dspp an ibody and
Glenn Longenecke and Ashok Kulka ni o helping wi h he injec ion
o Ep n a ge ed ES clones in o blas cys s. We also hank Thomas
Bugge, Ka in Lis , Roman Szabo, and Masahi o Iwamo o o hei
aluable sugges ions.
REFERENCES
1. Pispa, J., and Thesle , I. (2003) De . Biol. 262, 195–205
2. Je n all, J., and Thesle , I. (2000) Mech. De . 92, 19–29
3. Fuchs, E., and Ragha an, S. (2002) Na . Re . Gene . 3, 199–209
4. Alonso, L., and Fuchs, E. (2006) J. Cell Sci. 119, 391–393
5. Mile ich, I., and Sha pe, P. T. (2003) Hum. Mol. Gene . 12, Spec. No. 1,
R69–R73
6. Thesle , I. (2003) J. Cell Sci. 116, 1647–1648
7. Thesle , I. (2006) Am. J. Med. Gene . A 140, 2530–2535
8. K a ochwil, K., Dull, M., Fa inas, I., Galce an, J., and G osschedl, R. (1996)
Genes De . 10, 1382–1394
9. Mills, A. A., Zheng, B., Wang, X. J., Vogel, H., Roop, D. R., and B adley, A.
(1999) Na u e 398, 708–713
10. Yang, A., Schwei ze , R., Sun, D., Kaghad, M., Walke , N., B onson, R. T.,
Tabin, C., Sha pe, A., Capu , D., C um, C., and McKeon, F. (1999) Na u e
398, 714–718
11. Sa oka a, I., and Maas, R. (1994) Na . Gene . 6, 348–356
12. Pe e s, H., Neubuse , A., K a ochwil, K., and Balling, R. (1998) Genes De .
12, 2735–2747
13. Abe g, T., Ca ende , A., Gaikwad, J. S., B oncke s, A. L., Wang, X., Wal-
imo-Si en, J., Thesle , I., and D’Souza, R. N. (2004) J. His ochem. Cy o-
chem. 52, 131–139
14. Bouwman, P., Gollne , H., Elsasse , H. P., Eckho , G., Ka is, A., G os eld,
F., Philipsen, S., and Suske, G. (2000) EMBO J. 19, 655–661
15. Dassule, H. R., Lewis, P., Bei, M., Maas, R., and McMahon, A. P. (2000)
De elopmen (Camb.)127, 4775–4785
16. Fukumo o, S., Kiba, T., Hall, B., Ieha a, N., Nakamu a, T., Longenecke , G.,
K ebsbach, P. H., Nanci, A., Kulka ni, A. B., and Yamada, Y. (2004) J. Cell
Biol. 167, 973–983
17. Gibson, C. W., Yuan, Z. A., Hall, B., Longenecke , G., Chen, E., Thyaga a-
jan, T., S eena h, T., W igh , J. T., Decke , S., Pidding on, R., Ha ison, G.,
and Kulka ni, A. B. (2001) J. Biol. Chem. 276, 31871–31875
18. Nakamu a, T., Unda, F., de-Vega, S., Vilaxa, A., Fukumo o, S., Yamada,
K. M., and Yamada, Y. (2004) J. Biol. Chem. 279, 626–634
19. Kaczynski, J., Cook, T., and U u ia, R. (2003) Genome Biol. 4, 206
20. Suske, G., B u o d, E., and Philipsen, S. (2005) Genomics 85, 551–556
21. Kawano, S., Mo o omi, T., Toyono, T., Nakamu a, N., Uchida, T., Ohishi,
M., Toyoshima, K., and Ha ada, H. (2002) Connec . Tissue Res. 43,
409–412
22. K ebsbach, P. H., Lee, S. K., Ma suki, Y., Kozak, C. A., Yamada, K. M., and
Yamada, Y. (1996) J. Biol. Chem. 271, 4431–4435
23. an Gende en, C., Okamu a, R. M., Fa inas, I., Quo, R. G., Pa slow, T. G.,
B uhn, L., and G osschedl, R. (1994) Genes De . 8, 2691–2703
Roles o Epip o in in Too h De elopmen
4832 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 283•NUMBER 8•FEBRUARY 22, 2008
24. Fjeld, K., Ke unen, P., Fu manek, T., K innsland, I. H., and Luukko, K.
(2005) De . Dyn. 233, 161–166
25. G i li-Linde, A., Bei, M., Maas, R., Zhang, X. M., Linde, A., and McMahon,
A. P. (2002) De elopmen (Camb.)129, 5323–5337
26. Vaah oka i, A., Abe g, T., Je n all, J., Ke anen, S., and Thesle , I. (1996)
Mech. De . 54, 39–43
27. She , C. J., and Robe s, J. M. (1995) Genes De . 9, 1149–1163
28. Qin, X. Q., Chi enden, T., Li ings on, D. M., and Kaelin, W. G., J . (1992)
Genes De . 6, 953–964
29. Galde isi, U., Cipolla o, M., and Gio dano, A. (2006) Oncogene 25,
5250–5256
30. Li, F. Q., Coon od, A., and Ho wi z, M. (2000) Mol. Cell. Biol. 20,
5129–5139
31. No i ch, B. G., Mulligan, G. J., Jacks, T., and Lassa , A. B. (1996) J. Cell Biol.
135, 441–456
32. Zhou, P., By ne, C., Jacobs, J., and Fuchs, E. (1995) Genes De . 9, 700–713
33. Ja inen, E., Salaza -Ciudad, I., Bi chmeie , W., Take o, M. M., Je n all, J.,
and Thesle , I. (2006) P oc. Na l. Acad. Sci. U. S. A.
Roles o Epip o in in Too h De elopmen
FEBRUARY 22, 2008•VOLUME 283•NUMBER 8 JOURNAL OF BIOLOGICAL CHEMISTRY 4833
