-
Cancer Genes and Genomics
IKKb Overexpression Leads to Pathologic Lesions in
StratifiedEpithelia and Exocrine Glands and to Tumoral
Transformationof Oral Epithelia
Angustias Page1, Jos�e L. Cascallana2, M. Llanos Casanova1,
Manuel Navarro1, Josefa P. Alameda1,Paloma P�erez3, Ana Bravo2, and
Angel Ramírez1
AbstractAlterations in nuclear factor kappaB (NFkB) signaling
have been related with several diseases and importantly
also with cancer. Different animal models with increased or
diminished NFkB signaling have shown that NFkBsubunits and their
regulators are relevant to the pathophysiology of different organs
and tissues. In particular, boththe deletion of the regulatory
subunit b of the kinase of the inhibitor of NFkB (IKKb) and its
overexpression inepidermis lead to the development of skin
inflammatory diseases not associated with tumoral lesions. In this
work,we have studied the consequences of IKKb overexpression in
other organs and tissues. We found that elevatedIKKb levels led to
altered development and functionality of exocrine glands (i.e.,
mammary glands) in transgenicfemale mice. In oral epithelia,
increased IKKb expression produced lichenoid inflammation with
abundantgranulocytes, macrophages, and B cells, among other
inflammatory cells. This inflammatory phenotype wasassociated with
high incidence of tumoral lesions in oral epithelia, contrary to
what was found in skin. Moreover,IKKb also increased the malignant
progression of both spontaneous and experimentally induced oral
tumors.These results highlight the importance of IKKb in epithelial
and glandular homeostasis as well as in oraltumorigenesis and open
the possibility that IKKb activity might be implicated in the
development of oral cancer inhumans. Mol Cancer Res; 9(10);
1329–38. �2011 AACR.
Introduction
Nuclear factor kappaB (NFkB) is a conserved family ofubiquitous
transcription factors formed by homo- andheterodimers of distinct
proteins including RelA/p65, RelB,c-Rel, p50, and p52. It regulates
the expression of dozens ofgenes, many of them involved in the
control of proliferation,apoptosis, and other essential cellular
processes. Impor-tantly, deregulation of this family of
transcription factorsis implicated in the origin of several types
of cancer andinflammatory diseases (1, 2). NFkB members are
usuallymaintained inactive in the cytoplasm by binding to
inhi-bitory proteins of the IkB family. When the cells are
under
activating conditions, IkB proteins are phosphorylated
anddegraded; consequently, NFkB dimers are released, beingable to
migrate to the nucleus, and execute their function astranscription
factors by binding to specific kB sites in theregulatory region of
target genes. Phosphorylation andsubsequent degradation of IkB
proteins are controlledmainly by the IKK complex, which comprises
the catalyticsubunits IKKa and IKKb and the regulatory subunit
IKKg(3, 4). NFkB activity can also be regulated by
posttran-scriptional modifications of NFkB proteins, in
particularp65 phosphorylation (5).Numerous recent reports have
established an important
role for several of the proteins implicated in NFkB activityin
epithelial homeostasis (reviewed in refs. 6, 7). Althoughmouse
models lacking individual members of the NFkBfamily do not show
alterations of skin or other epithelia, thesimultaneous ablation of
different combinations of theseproteins produce epithelial
phenotypes (8), indicating thattheir activity is needed for the
correct function of theepithelia, although there exists some degree
of redundancyamong them. The modification of the expression level
of thevarious components of the IKK complex in animal modelsleads
to different epithelial alterations. IKKa exerts animportant role
in skin function and development, and itis also implicated in skin
cancer (9, 10). Regarding IKKg , ithas been shown that inactivating
mutations of the X-linkedIKKg gene lead to the appearance of the
genodermatosis
Authors' Affiliations: 1Department of Epithelial Biomedicine,
Centro deInvestigaciones Energ�eticas, Medioambientales y
Tecnol�ogicas, Madrid;2Department of Veterinary Clinical Sciences,
Laboratory of PathologyPhenotyping of Genetically Engineered Mice,
Faculty of Veterinary Med-icine, University of Santiago de
Compostela, Lugo; and 3Animal ModelsLaboratory, Instituto de
Biomedicina de Valencia, Valencia, Spain
Note: Supplementary data for this article are available at
Molecular CancerResearch Online (http://mcr.aacrjournals.org/).
Corresponding Author: Angel Ramírez, Department of Epithelial
Biome-dicine, Centro de Investigaciones Energ�eticas,
Medioambientales y Tec-nol�ogicas (CIEMAT), 28040 Madrid, Spain.
Phone: 34-91-346-0882; Fax:34-91-346-6484; E-mail:
[email protected]
doi: 10.1158/1541-7786.MCR-11-0168
�2011 American Association for Cancer Research.
MolecularCancer
Research
www.aacrjournals.org 1329
on June 29, 2021. © 2011 American Association for Cancer
Research. mcr.aacrjournals.org Downloaded from
Published OnlineFirst August 5, 2011; DOI:
10.1158/1541-7786.MCR-11-0168
http://mcr.aacrjournals.org/
-
incontinentia pigmenti in heterozygous females both inhumans and
in animal models (11, 12). IKKb activity isalso important for
epidermal homeostasis, as its skin-specificablation leads to the
development of an inflammatorypsoriasis-like skin disease dependent
on TNFa signalingand macrophage activity (13, 14). We have
describedpreviously that overexpression of IKKb in mouse
skin(K5-IKKb mice) leads, in transgenic (Tg) lines with
highexpression levels, to the development of an inflammatoryskin
disease with histologic features of interface dermatitis(15). Tg
skin keratinocytes, consistent with their increasedIKKb activity,
have augmented signaling of the NFkBsurvival pathway and are less
susceptible to apoptosis thanwild-type (WT) keratinocytes when
treated with TNFa(15). However, less is known about the function of
IKKb instratified epithelia different from that in the skin.
K5promoter is also active in other stratified epithelia, suchas
oral epithelia from the mouth and from the nonglandularstomach
(forestomach), as well as in other locations over-expressing IKKb,
as myoepithelial cells of exocrine glandsand epithelial appendages
(16). On the basis of this knowl-edge, we speculated that
transgenic mice overexpressingIKKb under the transcriptional
control of keratin K5regulatory elements could suffer pathologic
alterations alsoin other stratified epithelia where the K5 promoter
is active.In this report, we describe the alterations found in
stratifiedepithelia of different IKKb Tg lines. Most of these
altera-tions were found even in lines that do not suffer from
theserious skin inflammatory phenotype described in the studyof
Page and colleagues (15), therefore allowing the study ofother
epithelia without the interferences caused by anextended
inflammatory skin disease. Our findings indicatethat IKKb plays a
fundamental role in the pathophysiologyof stratified epithelia and
exocrine glands. In addition,alterations in the activity of IKKb
lead to inflammationand tumor development in oral epithelia. Even
more, in oralcarcinogenesis experiments, IKKb induces the
progressionfrom a benign tumor type toward a malignant one.
Materials and Methods
Mice and treatmentsGeneration of K5-IKKb Tg mice, carrying a
VSV-tagged
human IKKb under the control of a 5.2-kbp 50-upstreamfragment of
bovine K5 and a rabbit b-globin intron, waspreviously described
(15). Tg.AC heterozygous mice wereobtained from Taconic Farm Inc.
and mated to hetero-zygous K5-IKKb Tg mice. Tail DNA was used for
micegenotyping by PCR. The primers used were
50-TTCAGGGTGTTGTTTAGAAATGG-30 and 50-CAA-TAAGAATATTTCCACGCCA-30 for
K5-IKKb and 50-TGGCATTTCTTCTGAGCAA-30 and
50-TTGGACA-AACTACCTACAGAGAT-30for the Tg.AC
transgene.4-Nitroquinoline-1-oxide (4-NQO; Sigma) was prepared
and administered following the indications detailed in thework
of Tang and colleagues (17). Briefly, a 5 mg/mL stocksolution of
the carcinogen 4-NQO was prepared in pro-pylenglycol and stored at
4�C. This solution was further
diluted in drinking water at a final concentration of50 mg/mL.
Three-month-old Tg.AC and double K5-IKKb/Tg.AC Tg mice were allowed
ad libitum access tothe drinking water containing 4-NQO during the
6-weektreatment period. Bottles were changed twice a week.Animals
were maintained on standard rodent chow
(SAFE-A04; SAFE) and kept under a 12-hour light–darkcycle. All
experimental procedures were carried out accord-ing to European and
Spanish laws and regulations (Eur-opean Convention ETS 123 on the
use and protection ofvertebrate mammals used in experimentation and
otherscientific purposes; Spanish R.D 1201/2005 of the Ministryof
Agricultural, Food and Fisheries on the protection anduse of
animals in scientific research) and approved by ourInstitution's
Ethics Committee.
Histology and immunohistochemistryUnless other figures are
indicated, we conducted histo-
logic analysis of different stratified epithelia of at least 10
Tgand 4 non-Tg littermates adult mice of lines L1 and L2(with lower
and higher relative skin expression levels,respectively; ref. 15),
as well as 2 additional skin highexpression transgenic founders.
Apart from the glandulardysplasia shown in Figure 2C–F and 2I–J,
specific of themice with higher expression level, the phenotypes
werecommon to all the Tg lines; in fact, the majority of theresults
were obtained in the line with lower expression leveldue to
practical reasons.Mouse tissues were dissected and immediately
frozen in
optimum cutting temperature compound or fixed in 10%buffered
formaldehyde or 70% ethanol and embedded inparaffin. Three- to 5-mm
thick sections were used forimmunohistochemical preparations or for
hematoxylinand eosin (H&E) staining. Cryosections were fixed
for10 minutes at �20�C in acetone. The antibodies usedwere from
Imgenex (IKKb), Covance (K5), CALTAGlaboratories (F4/80), BD
Pharmingen (B220, Gr-1, andCD45), and Dako (CD3). Immunoreactivity
was revealedusing an ABC avidin–biotin–peroxidase system and
ABCsubstrate (Vector Laboratories), and the sections wereslightly
counterstained with hematoxylin. Control experi-ments were carried
out omitting the primary antibodies.
Western blot analysisWhole-cell protein extracts frommouse
organs and tissues
were prepared as described in the work of Perez and collea-gues
(18). The IKKb antibody used inWestern blot analysiswas from
Imgenex; b-actin antibody (Santa Cruz Biotech-nology) was used as a
loading control. The antibody used forphospho-p65 (Ser536)
detection in Supplementary FigureS1 is from Cell Signaling
Technology (#3033).
Analysis of mammary glandsWe analyzed mammary glands from
K5-IKKb Tg and
WT females (2 of each genotype and condition) fromvirgin females
at the ages of 7, 8, and 9 weeks, fromfemales on day 15 of
gestation, from lactating females5 days after delivery, and from
females with involutive
Page et al.
Mol Cancer Res; 9(10) October 2011 Molecular Cancer
Research1330
on June 29, 2021. © 2011 American Association for Cancer
Research. mcr.aacrjournals.org Downloaded from
Published OnlineFirst August 5, 2011; DOI:
10.1158/1541-7786.MCR-11-0168
http://mcr.aacrjournals.org/
-
mammary glands 2, 5, and 8 days after stopping lactation.For the
analysis of involutive mammary glands, femalesfed litters of 5 to 6
mice during 14 days. Then, the pupswere withdrawn and the
involutive mammary glandsnumber 4 or 9 were analyzed at the
indicated times.For whole mount preparations, inguinal number 4
mam-mary glands were fixed in Carnoy's fixative and stainedin
carmine alum, following the protocols described
inhttp://mammary.nih.gov/index.html
Results
Expression of IKKb in K5-IKKb miceWe have previously shown that
Tg mice expressing
IKKb under the transcriptional control of regulatory
elements derived from the keratin K5 gene express thetransgene
in the basal layer of skin and in hair follicles,leading to
activation of the NFkB signaling pathwayin keratinocytes and to the
development of interfacedermatitis (15).Because keratin K5 is
expressed in the basal layer of all
stratified epithelia and in myoepithelial cells of
exocrineglands, these cell types are expected to be targeted by
thekeratin K5–derived regulatory elements included in
thetransgene.We analyzed IKKb expression by Western blotting in
several stratified epithelia and exocrine glands from both L1Tg
and WT animals and found consistently higher IKKblevels in the Tg
samples (Fig. 1A; the lanes for back skin ofTg mice, included in
both Western blot analyses, allow for
Figure 1. Expression of IKKb instratified epithelia and
exocrineglands of K5-IKKb mice. A,Western blot analysis of
IKKbexpression in different organs andtissues containing
stratifiedepithelial or exocrine glands fromline L1 Tg and WT
animals. Prep.gland, preputial gland. B,immunohistochemical
analysis oftransgene expression in stratifiedepithelia. The
genotypes of themice are indicated on the left ofthe figure and the
antibodies usedinside the photographs. Theantibody anti-IKKb gave
specificsignals in the basal cells of the tailskin, foot sole,
palate, andforestomach epithelia of Tg mice(arrows, middle row).
Consecutiveparaffin sections were incubatedwith an antibody
specific forkeratin K5 (bottom row). Thestomach plica is indicated
by asurrounding line. Under thesedetection conditions, almost
noIKKb stainingwas seen inWTmice(top row). C,immunohistochemical
analysis oftransgene expression in exocrineglands. IKKb and K5
expression inmyoepithelial cells of Meibomianglands are shown.
D,immunohistochemical analysis ofK5 and IKKb expression inlactating
mammary glands.Myoepithelial cells were positivefor K5 but not for
IKKb in WTfemales, but both proteins weredetected in myoepithelial
cells ofTg samples (arrows). Bar: 200 mmin B and C and 50 mm in
D.
IKKb Leads to Oral Tumorigenesis
www.aacrjournals.org Mol Cancer Res; 9(10) October 2011 1331
on June 29, 2021. © 2011 American Association for Cancer
Research. mcr.aacrjournals.org Downloaded from
Published OnlineFirst August 5, 2011; DOI:
10.1158/1541-7786.MCR-11-0168
http://mcr.aacrjournals.org/
-
the comparison of the expression level between Tg and
WTsamples).Immunohistochemical analysis of the expression
pattern
of K5-IKKb transgene with antibodies specific for IKKbshowed
that it was expressed, as expected, in the basal celllayer of the
stratified epithelia studied (tail skin, foot soleskin, and oral
epithelia of the palate and forestomach;Fig. 1B, arrows). Of note,
IKKb is expressed in the stomachplica, that is, the region where
the glandular and thenonglandular parts of the murine stomach join
(arrowheadsand arrows, respectively, in the stomach panel of Fig.
1B).IKKb was also expressed in myoepithelial cells of
exocrineglands [as mammary, preputial, salivary, and
eyelidMeibomian glands; see Fig. 1B for IKKb expression insweat
glands (arrowheads in foot sole panel); Fig. 1C forexpression
inMeibomian glands; and Fig. 1D for expressionin myoepithelial
cells of mammary glands]. The morerestricted basal staining of IKKb
relative to that of keratinK5 in the tissues shown in Figure 1B
could be because ofdifferences in the stability of these proteins
or to differentsensitivity of the antibodies used. IKKb
overexpression wasassociated with increased phosphorylation of p65
at serine536 in stratified epithelia and exocrine glands of Tg
mice(Supplementary Fig. S1). This posttranscriptional
modifi-cation, which enhances p65 transactivation potential, is
amarker of NFkB activation (5, 19). It is interesting to notethat
IKKb overexpression did not lead to changes in theexpression level
of IKKa or IKKg (Supplementary Fig. S2).
Dysplasia associated with inflammation in ectodermalderivatives
and exocrine glands in K5-IKKb miceIn those Tg lines with higher
expression level (different
from the line L1 shown in Fig. 1), we observed,
evenmacroscopically, phenotypic consequences of K5-IKKbtransgene
expression in ectodermal derivatives, includingeyelids, nails, and
teeth, and in exocrine glands. The processof eyelid fusion that
usually takes place in the mouse embryobefore day 18.5 of
development frequently failed in Tg mice
and they were born with one or both eyes opened (Fig. 2Aand B,
arrows). Some adult animals showed thickening ofeyelids as a
consequence of proliferation of dysplasticmyoepithelial and/or
reserve cells and substitution of thenormal exocrine cells from
Meibomian glands associatedwith discrete inflammation (Fig. 2C–F,
arrows). The major-ity of transgenic mice also had nail
malformations, some-times accompanied by fragility; they showed
abnormallybended and overgrown nails, lacking the usual sharp tip
ofthe nails in WT mice (Fig. 2G and H, arrowheads). Thisphenotype
was likely the consequence of dysplasia andlichenoid inflammation
of the stratified epithelia of thenail bed (not shown). Many Tg
animals showed hyperpig-mentation of the foot pads (Fig. 2G and H,
arrows) becauseof the presence of macromelanophages as a
consequence of apreceding damage in epidermis and hair follicles
(see ref.15); this hyperpigmentation was usually more marked
inolder mice. In addition to these phenotypic alterations,some Tg
males showed atrophic preputial glands (Fig. 1Iand J) as a
consequence of proliferation of dysplasticmyoepithelial and/or
reserve cells that provoked atrophyof the secretory acini (not
shown). Some Tg mice alsodeveloped atrophy of major salivary glands
(especially par-otid) and minor salivary glands of the tongue by
thedysplastic proliferation of myoepithelial cells (not
shown).These alterations were observed mainly in Tg lines withhigh
expression level. Nevertheless, we also found epithelialand
glandular alterations in Tg lines with lower expressionlevel (i.e.,
line L1), as well as effects on oral cancer devel-opment (see
later).
Mammary gland development and functionality areimpaired in
K5-IKKb miceWe analyzed mammary glands from both Tg females
from line L1 and nontransgenic littermates at different ages.In
56-day-old young virgin animals, WT females showed awell-developed
tree of ducts, filling the majority of thefat pad of the gland. In
contrast, Tg females showed a
A C
B D F H J
G IEWT
Tg
Figure 2.Macroscopic alterations found in K5-IKKb Tg mice. A and
B, K5-IKKb Tg mice frequently were born with 1 or the 2 eyes
opened. C–F, transformationof Meibomian glands in adult mice from
the highest expressing lines. Glandular cells are transformed into
dysplastic and basophilic epithelial cellslacking secretory
differentiation and associated with periacinar infiltration by
mononuclear inflammatory cells (see white arrows in E and F). G and
H,K5-IKKb Tg mice also showed altered nails (arrowheads) and
hyperpigmentation of the fat pads (arrows). I and J, preputial
glands were frequently of smallersize in adult K5-IKKb male mice
than in WT mice. Bar: 500 mm in C and D and 50 mm in E and F.
Page et al.
Mol Cancer Res; 9(10) October 2011 Molecular Cancer
Research1332
on June 29, 2021. © 2011 American Association for Cancer
Research. mcr.aacrjournals.org Downloaded from
Published OnlineFirst August 5, 2011; DOI:
10.1158/1541-7786.MCR-11-0168
http://mcr.aacrjournals.org/
-
more rudimentary ductal design, which hardly exceeded thelymph
node; in addition, the terminal end buds wereirregularly shaped
(arrows in Fig. 3A and B). When westudied lactating females 5 days
after delivery, we found thatthe mature alveoli of Tg mammary
glands were smaller thanthat of WT females (compare the diameter of
alveoli inFig. 3C and D), suggesting deficient functionality of
theglands in Tg females. In addition, the luminal secretory cellsin
Tg females showed less vacuoles in their apical polesduring
lactation, which probably results in low milk pro-duction (Fig. 3E
and F, arrows).The remodeling program that takes place at the end
of
lactation was also altered in Tg animals. At 2 days
ofinvolution, Tg mammary glands showed a macroscopicappearance
clearly different from that in WT glands: manyof them were swollen
and whitish, apparently still contain-ing milk. Microscopically, 2
days after the cessation oflactation, WT mammary glands showed
massive apoptosisof milk-producing alveolar cells. In contrast, Tg
glandsshowed a retarded involution process, with scarce
apoptotic
cells inside the mammary ducts (Fig. 3G–J, arrows). At thistime,
the alveoli showed an increased diameter in Tg mice,also indicative
of a delay in the involution process. Thesephenotypes are
consistent with the triggering of survivalsignals as a consequence
of increased NFkB activation incells of Tg mice. In summary, IKKb
overexpression affectsmammary gland at different levels; it leads
to reduceddevelopment of mammary glands in female Tg mice,
toaltered milk production in lactating females, and to
delayedregression of the gland in the involution process.To
evaluate the functional effect of these alterations, we
carried out crosses between WT females and Tg males, andvice
versa, and compared several reproductive parameters inthe 2 types
of crosses. As can be seen in SupplementaryTable S1, the percentage
of pups that reached the age of9 days was 100% in WT females and
less than 70% whenthe mother was Tg. Furthermore, in this last
group, thesurviving pups had reduced weight, reaching only 76.5%
ascompared with animals fed by WT females. This impair-ment in
newborn survival and fitness is not directly relatedto the presence
of the K5-IKKb transgene in the pups, as wedid not find these
alterations in the opposite breeding, thatis, WT females crossed
with Tg males, see SupplementaryTable S1. In summary, Tg females
bring up their litters lessefficiently than WT females, resulting
in higher death ratein their pups and less growth of the
survivors.
Inflammatory phenotype in oral and
forestomachepitheliaMacroscopically, many Tg mice also showed gross
altera-
tions of other ectodermal derivatives such as teeth andpalate.
Both supernumerary incisors and tooth fragility wererecorded. A
consistent finding during the examination ofthe oral cavity of Tg
mice was the thickening of the oralmucosa of the palate, even in
young mice (Fig. 4A, arrows).Microscopically, the palate epithelium
often showed pseu-docarcinomatous hyperplasia in Tg mice,
characterized byextreme irregular proliferation of solid epithelial
cordsmainly at the base of the rete ridges, bearing a
superficialresemblance to invasive squamous cell carcinomas
(SCC;see Fig. 4B palate, arrows). This was accompanied by
severelichenoid inflammation, basal cell damage, and
exocytosis(inset in Fig. 4B). All these features were also observed
inoral epithelia of the forestomach at the level of the plica(Fig.
4B stomach, arrow, and Fig. 4C). More than 70% ofthe animals
analyzed showed inflammation of the palate andforestomach (in some
cases associated with other alterations,see below).In the palate
(not shown) and forestomach plica, infil-
trating cells were frequently observed. In these inflamma-tory
lesions, eosinophil granulocytes were abundant, asshown by H&E
staining of histologic sections (Fig. 4C,white arrows). The
immunostaining against CD45 antigen(characteristic of hematopoietic
inflammatory cells) showedthat the inflamed oral epithelia were
clearly infiltrated by alarge amount of positive cells. Staining
with anti-Gr-1antibody, specific for granulocytes, confirmed the
presenceof this inflammatory cell type (black arrows in
Fig.4C).
A B
C D
E F
G H
I J
Figure 3. Developmental and functional alterations in mammary
glands ofK5-IKKb mice. Mammary glands of K5-IKKb females and WT
littermateswere analyzed at different ages. A and B, whole mount
staining ofmammary glands of 56-day-old virgin females. Arrows
indicate terminalend buds. C–F, low- and high-power amplification
of day 5 lactatingmammary glands. Arrows indicate vacuoles
containing milk. G–J, low- andhigh-power amplification of
involutive mammary glands 2 days after thecessation of lactation.
Arrows indicate apoptotic cells. Bar: 500 mm in Aand B; 300 mm in
C, D, G, and H; and 100 mm in E, F, I, and J.
IKKb Leads to Oral Tumorigenesis
www.aacrjournals.org Mol Cancer Res; 9(10) October 2011 1333
on June 29, 2021. © 2011 American Association for Cancer
Research. mcr.aacrjournals.org Downloaded from
Published OnlineFirst August 5, 2011; DOI:
10.1158/1541-7786.MCR-11-0168
http://mcr.aacrjournals.org/
-
F4/80- and B220-positive cells (macrophages and B
cells,respectively) were also more abundant in the inflammatoryfoci
of Tg mice than in WT forestomach (Fig. 4C). CD3staining also
showed more T cells in Tg than in WT oralepithelia (not shown). To
unravel the contribution oflymphocytes to the overall phenotype
found in the oralepithelia of Tg mice, we studied K5-IKKb
transgenic micelacking lymphocytes. We generated K5-IKKb Tg mice
indifferent immunodeficient genetic backgrounds: K5-IKKb/nude
(lacking T lymphocytes) or K5-IKKb/severe com-bined immunodeficient
mice (lacking T, B, and naturalkiller cells). The oral phenotype
obtained in both immu-nodeficient models was similar to the
phenotype obtained inK5-IKKb immunocompetent mice (not shown),
indicatingthat T, B, and natural killer cells are not responsible
for the
oral phenotype. We also analyzed the implication of TNFsignaling
in the oral epithelia phenotype, as TNF signalingis essential for
the pathogenesis of skin lesions in transgenicmice lacking IKKb or
IKKg in their skin (13, 20). Westudied histologic preparations of
palate and forestomachepithelia from K5-IKKb Tg mice treated with
etanercept toinhibit signaling through TNFa receptor. The
phenotypicalterations observed in these mice were similar to
untreatedTg mice (not shown), indicating that these phenotypes
areindependent of TNFa signaling.
K5-IKKb mice develop spontaneous tumors in the oralepitheliaA
causal relationship between inflammation and cancer
has been known for years, as many human tumors develop
Figure 4. Histologic phenotypeand inflammation in oral
epitheliain K5-IKKb mice. A, low-powermagnification of palate. Note
thethickening of palate epithelium inTg mice, resulting in less
markedwaves of the hard palate thanin WT mice (arrows).
B,pseudocarcinomatoushyperplasia in adult palate andforestomach
epithelia, showingaberrant proliferation of branchedepithelial
cords (arrows),associated with band-likeinfiltration of
inflammatory cellsand vacuolar degeneration of thebasal layer (see
inset in palatepanel). C, histologic sections fromWT and Tg stomach
plica werestained with H&E (white arrows ininset indicate
eosinophilgranulocytes) or immunostainedwith antibodies specific
forhematopoietic infiltrating cells(CD45), granulocytes
(Gr-1),macrophages (F4/80), or B cells(B220), as indicated. In
thestomach plica, all types ofinflammatory cells areoverrepresented
in K5-IKKbanimals. Bar: 500 mm in A, 300 mmin B (palate), and 200
mm in B(stomach) and C.
Page et al.
Mol Cancer Res; 9(10) October 2011 Molecular Cancer
Research1334
on June 29, 2021. © 2011 American Association for Cancer
Research. mcr.aacrjournals.org Downloaded from
Published OnlineFirst August 5, 2011; DOI:
10.1158/1541-7786.MCR-11-0168
http://mcr.aacrjournals.org/
-
in organs that suffer chronic inflammatory processes (seeref.
21). Because K5-IKKb mice showed persistent inflam-mation in oral
epithelia, they could provide a good modelfor studying in vivo the
relationship between tumor inci-dence in oral epithelia and the
presence of inflammatorycells. We studied the incidence of tumors
in oral stratifiedepithelia from the palate and from the
forestomach in Tgmice. Histologic analysis of K5-IKKb mice
frequentlyrevealed foci of carcinoma in situ in those regions
withpseudocarcinomatous hyperplasia associated with inflam-mation;
these foci were characterized by high-grade atypia ofkeratinocytes,
even presenting epithelial pearls characteristicof
well-differentiated SCCs but limited to the oral mucosawith the
integrity of the basal membrane (Fig. 5A andTable 1). The
spontaneous development of SCCs thatinfiltrated subjacent tissues
was also sporadically seen in9.5% of the analyzed animals in the
palate and 14.7% in theforestomach (Fig. 5B and Table 1; see
malignant epithelialcells infiltrating the deep submucosa of the
palate andstomach, arrows). These results indicate that IKKb
over-expression, inflammation, and the risk of neoplastic
trans-
formation are closely interrelated in oral stratified
epithelia,being a high level of IKKb sufficient to trigger the
devel-opment of spontaneous tumors in the oral epithelia of
Tgmice.
IKKb provokes malignant transformation ofexperimentally induced
oral tumorsTo further study the potential of IKKb as a tumor
promoter in oral cancer, we subjected populations of K5-IKKb and
control animals to oral carcinogenesis experi-ments. We used
K5-IKKb mice that also carried a Tg.ACtransgene, which directs the
expression of an active form ofv-Ha-Ras to epithelial cells (22)
and thus represent a strainof initiated mice suited for the study
of tumor developmentin skin and other stratified epithelia. As a
control popula-tion, we used Tg.AC mice that did not carry the
K5-IKKbtransgene. Animals of both genotypes were treated with4-NQO
in the drinking water, following administrationregimens that result
in hyperplasia of oral epithelia and theoccurrence of tumors or
preneoplastic lesions (17). Themajority of Tg.AC control animals (6
of 7) developedpapillomas in the epithelia of the forestomach.
Thesetumors, although sometimes of great size, were alwaysbenign,
showing conjunctive papillary projections coveredby neoplastic
proliferation of typical squamous epitheliumwith net limits of the
basal membrane (see Fig. 6A and B).All the treated K5-IKKb/Tg.AC
transgenic animals (5 of 5)developed tumoral lesions in the
forestomach. Strikingly,we found that the majority of them (4 of 5
cases) containednumerous malignant foci of infiltrating carcinoma,
wherestrands of squamous malignant cells broke down the
basalmembrane and infiltrated the submucosal conjunctive layer(Fig.
6C and D, arrows). These invasive foci were highlydysplastic, with
loss of the typical multilayered organizationof the epithelium, and
conformed by epithelial cells with ahigh degree of atypia. Similar
results were obtained when weanalyzed tumor incidence in the palate
epithelia, as malig-nant SCCs were detected in some of the
K5-IKKb/Tg.ACmice (2 of 5) whereas Tg.AC mice did not developedany
SCC (see Fig. 6E and F). Overall, these results highlightthe
importance of IKKb in the development and progres-sion toward
malignancy of oral tumors.
Figure 5. Spontaneous tumoral lesions in oral epithelia in
K5-IKKb mice.The palate and forestomach showed foci of premalignant
lesions(carcinoma in situ, A) as well as infiltrating SCCs (B).
Bar: 200 mm in A(stomach) and B and 100 mm in A (palate).
Table 1. Incidence of inflammation and spontaneous premalignant
or tumoral lesions in oral epithelia ofK5-IKKb Tg mice
n Percentage of micewithout inflammation
Percentage of mice with inflammation
Nontumoral lesions Carcinoma in situ SCC
Palate 21 19.0 0 71.4 9.5Forestomach 34 26.5 44.1 14.7 14.7
NOTE: Each animal was classified according to the absence or
presence of inflammation; animals scoring positive for
inflammationwere further classified according to the absence or
presence of premalignant or tumoral lesions. The total number of
mice analyzed isindicated by n.
IKKb Leads to Oral Tumorigenesis
www.aacrjournals.org Mol Cancer Res; 9(10) October 2011 1335
on June 29, 2021. © 2011 American Association for Cancer
Research. mcr.aacrjournals.org Downloaded from
Published OnlineFirst August 5, 2011; DOI:
10.1158/1541-7786.MCR-11-0168
http://mcr.aacrjournals.org/
-
Discussion
In this study, we have focused on the involvement ofIKKb in the
pathophysiology of stratified epithelia andexocrine glands.
Previously published data (13–15, 23)indicate an important role for
IKKb in the developmentand functionality of skin. The diverse
phenotypes observedin other stratified epithelia and exocrine
glands of K5-IKKbTg mice indicate the importance of IKKb also in
theseorgans and tissues. Furthermore, increased level of IKKb ledto
the appearance of an inflammatory disease in oralepithelia of the
palate and the stomach and to a highincidence of tumoral lesions in
oral epithelia of transgenicmice.The activity of IKK complex is
crucial for the correct
function of mammary epithelial cells. Thus, it is knownthat the
kinase activity of IKKa plays an important role incyclin D1
induction and mammary epithelial cells pro-liferation (24). In
addition, a proapoptotic function ofIKKb in mammary epithelial
cells has been described, asfemale mice with conditional deletion
of IKKb directed bythe promoter of the b-lactoglobulin milk protein
geneshow delayed apoptosis of epithelial cells in the
involutionprocess (25). However, we have also observed a
delayed
apoptosis in involutive mammary glands of miceoverexpressing
IKKb. This intriguing concordance inthe phenotypes due to IKKb loss
of function in milk-producing epithelial cells (25) and IKKb
overexpression inmyoepithelial cells (this report) is likely
related to thedifferent cell populations targeted in both
experiments.Furthermore, it also highlights the complex
organizationof the mammary gland and the delicate interrelation
andbalance that exist between the different cell types thatbuild up
this gland.IKKb, besides acting as an NFkB regulator by
phosphor-
ylating IkB proteins, also regulates other cellular
responsesthrough phosphorylation of other physiologic targets.Among
others, we could point out several proteins relatedto cancer as the
forkhead transcription factor FoxO3a, thehistone deacetylase HDAC1,
Aurora A, and the tumorsuppressors p16, p53, or Tsc1 (refs. 26–29
and referencestherein). Recently, it has been described that IKKb
can exerta novel nuclear function as an adaptor protein for
IkBadegradation in a manner independent of its kinase activity(30).
Taking into account the variety of phenotypes foundin K5-IKKb mice,
including its propensity to oral carcino-genesis, some of them
could be partly independent ofNFkB. In this context, it would be of
interest to analyzewhether the tendency to develop oral tumors in
K5-IKKbTg mice is exerted through NFkB and whether it isdependent
on the function of IKKb as part of the IKKcomplex. This knowledge
could be clinically relevant, as thepotential therapeutic
interventions would be different.Thus, the generation and study of
animals expressingmutant forms of IKKb unable to interact with the
othermembers of the IKK complex or lacking its kinase
activityshould provide interesting information.NFkB is an important
mediator of inflammation and
tumorigenesis. It is recognized that persistent inflamma-tion
may promote susceptibility to carcinogenesis, partlyby creating a
stroma enriched with cytokines and growthfactors (31). Although the
response to altered NFkBactivity varies among different cell types,
increased NFkBactivity is usually associated with tumor
development(1, 32). IKKb also acts as a tumor promoter in
differentcell types, as it happens in colitis-associated cancer
andmelanoma development in animal models, as well as inhuman breast
cancer patients (33–35). In the case of oralepithelia, we have
observed inflammation and basal celldamage as a consequence of
targeted overexpression ofIKKb. The subsequent epithelial
hyperplasia is accompa-nied by cellular atypia and the development
of high-gradedysplastic changes of the epithelia. These alterations
couldbe ascribed to carcinoma in situ when limited to theepithelium
by the basal membrane or to SCC wheninfiltrated the subjacent
conjunctive tissue. As IKKboverexpression leads to the development
of pretumoralor tumoral lesions in oral epithelia of transgenic
mice andpromotes their malignant conversion in a
carcinogenesisassay in the presence of activating forms of
Ha-ras,deregulation of IKKb activity could underlie some casesof
oral cancer in humans. In this regard, it would be of
AA B
C D
E F
A B
E F
Figure 6.Malignant lesions in oral epithelia of
K5-IKKb/Tg.ACmice treatedwith 4-NQO. A–D, tumoral lesions in
forestomach epithelia of 4-NQO–treated Tg.AC (A and B) and
K5-IKKb/Tg.AC (C and D) mice. B and D,magnifications of the frames
indicated in A and C. White lines indicate thebasal membrane, that
is discontinuous in D. E and F, H&E-stained sectionsof the
palate of Tg.AC (E) and K5-IKKb/Tg.AC (F) mice after 6 weeks
oftreatment with 4-NQO. White lines indicate the border between
epithelialcells and subjacent conjunctive tissue (E) and islets of
malignant epithelialcells that have infiltrated the submucosa (F);
other invasive foci areindicated with arrows. Bar: 500 mm in A and
C; 200 mm in B; and 100 mmin D–F.
Page et al.
Mol Cancer Res; 9(10) October 2011 Molecular Cancer
Research1336
on June 29, 2021. © 2011 American Association for Cancer
Research. mcr.aacrjournals.org Downloaded from
Published OnlineFirst August 5, 2011; DOI:
10.1158/1541-7786.MCR-11-0168
http://mcr.aacrjournals.org/
-
interest to test whether the inhibition of IKKb in
oralmalignancies has therapeutic value.Back skin appears to be more
resistant than oral epithelia
to carcinogenesis mediated by increased activity of IKKb/NFkB
pathway because we have not observed spontaneoustumor development
in the skin of K5-IKKb Tg mice (15).This is compatible with the
results obtained by otherauthors using other animal models in which
the inhibitionof NFkB signaling leads to skin hyperplasia and even
tumorformation (36, 37). Our study highlights the complexity ofthe
regulatory networks controlled by IKKb and NFkB, astheir
manipulation renders completely different outcomesin different cell
types. The marked differences in tumordevelopment susceptibility
that we have found between skinand oral epithelia when subjected to
IKKb overexpression,chronic inflammation, and basal cell damage are
surprising.In this regard, K5-IKKb Tg mice represent an
interestingmodel for the study of the signaling pathways and
regulatorymechanisms underlying tumor formation that could
explainthe different tumoral susceptibility in oral epithelia
andskin.In summary, the present work indicates that IKKb plays
a
central role in epithelial homeostasis and in exocrine
glandfunction; IKKb also promotes the formation of tumorallesions
in oral epithelia and their malignancy in cooperation
with activating oncoproteins such as viral Ha-ras.
Conse-quently, IKKb could be related to the pathogenesis ofhuman
tumoral diseases affecting oral epithelia. In thiscontext, it would
be interesting to test the potential ther-apeutic value of IKKb
inhibition in the treatment of humanoral cancers triggered by
inflammation.
Disclosure of Potential Conflicts of interest
No potential conflicts of interest were disclosed.
Acknowledgments
We thank I. de los Santos and F. Sanchez-Sierra for excellent
technical assistanceand J. Martinez, E. Almeida and the personnel
of the CIEMAT Animal Unit for thecare of the mice used in this
work. We also thank Dr. V. Lafarga for advice in theanalysis of
mammary glands and Dr. E. Radaelli for critical reading of the
manuscript.
Grant Support
This research was supported by grants from the Spanish
government(Ministerio de Ciencia e Innovaci�on; SAF2010-22156 to A.
Ramírez and PI-10/01480 to M.L. Casanova).
The costs of publication of this article were defrayed in part
by the paymentof page charges. This article must therefore be
hereby marked advertisement inaccordance with 18 U.S.C. Section
1734 solely to indicate this fact.
Received April 14, 2011; revised July 29, 2011; accepted August
1, 2011;published OnlineFirst August 5, 2011.
References1. Karin M. Nuclear factor-kappaB in cancer
development and progres-
sion. Nature 2006;441:431–6.2. Pasparakis M. Regulation of
tissue homeostasis by NF-kappaB sig-
nalling: implications for inflammatory diseases. Nat Rev
Immunol2009;9:778–88.
3. Hayden MS, Ghosh S. Shared principles in NF-kappaB signaling.
Cell2008;132:344–62.
4. Israel A. The IKK complex, a central regulator of NF-kappaB
activa-tion. Cold Spring Harb Perspect Biol 2010;2:a000158.
5. Perkins ND. Post-translational modifications regulating the
activityand function of the nuclear factor kappa B pathway.
Oncogene2006;25:6717–30.
6. Sur I, Ulvmar M, Toftgard R. The two-faced NF-kappaB in the
skin. IntRev Immunol 2008;27:205–23.
7. Wullaert A, Bonnet MC, Pasparakis M. NF-kappaB in the
regulation ofepithelial homeostasis and inflammation. Cell Res
2011;21:146–58.
8. Gerondakis S, Grumont R, Gugasyan R, Wong L, Isomura I, Ho
W,et al. Unravelling the complexities of the NF-kappaB signalling
path-way using mouse knockout and transgenic models.
Oncogene2006;25:6781–99.
9. Descargues P, Sil AK, Karin M. IKKalpha, a critical regulator
ofepidermal differentiation and a suppressor of skin cancer. EMBO
J2008;27:2639–47.
10. Moreno-Maldonado R, Ramirez A, Navarro M,
Fernandez-AceneroMJ, Villanueva C, Page A, et al. IKKalpha enhances
human keratino-cyte differentiation and determines the histological
variant of epider-mal squamous cell carcinomas. Cell Cycle
2008;7:2021–9.
11. Makris C, Godfrey VL, Krahn-Senftleben G, Takahashi T,
RobertsJL, Schwarz T, et al. Female mice heterozygous for IKK
gamma/NEMO deficiencies develop a dermatopathy similar to the
humanX-linked disorder incontinentia pigmenti. Mol Cell
2000;5:969–79.
12. Schmidt-Supprian M, Bloch W, Courtois G, Addicks K, Israel
A,Rajewsky K, et al. NEMO/IKK gamma-deficient mice model
incon-tinentia pigmenti. Mol Cell 2000;5:981–92.
13. Pasparakis M, Courtois G, Hafner M, Schmidt-Supprian M,
Nenci A,Toksoy A, et al. TNF-mediated inflammatory skin disease in
mice withepidermis-specific deletion of IKK2. Nature
2002;417:861–6.
14. Stratis A, Pasparakis M, Rupec RA, Markur D, Hartmann K,
Scharf-fetter-Kochanek K, et al. Pathogenic role for skin
macrophages in amouse model of keratinocyte-induced psoriasis-like
skin inflamma-tion. J Clin Invest 2006;116:2094–104.
15. Page A, Navarro M, Garin M, Perez P, Casanova ML, Moreno R,
et al.IKKbeta leads to an inflammatory skin disease resembling
interfacedermatitis. J Invest Dermatol 2010;130:1598–610.
16. Ramirez A, Bravo A, Jorcano JL, Vidal M. Sequences 50 of the
bovinekeratin 5 gene direct tissue- and cell-type-specific
expression of alacZ gene in the adult and during development.
Differentiation1994;58:53–64.
17. Tang XH, Knudsen B, Bemis D, Tickoo S, Gudas LJ. Oral cavity
andesophageal carcinogenesis modeled in carcinogen-treated mice.
ClinCancer Res 2004;10:301–13.
18. Perez P, Page A, Bravo A, Del Rio M, Gimenez-Conti I,
Budunova I,et al. Altered skin development and impaired
proliferative and inflam-matory responses in transgenic mice
overexpressing the glucocorti-coid receptor. FASEB J
2001;15:2030–2.
19. Viatour P, Merville MP, Bours V, Chariot A. Phosphorylation
of NF-kappaB and IkappaB proteins: implications in cancer and
inflamma-tion. Trends Biochem Sci 2005;30:43–52.
20. Nenci A, HuthM, Funteh A, Schmidt-SupprianM, BlochW,Metzger
D,et al. Skin lesion development in a mouse model of
incontinentiapigmenti is triggered by NEMO deficiency in epidermal
keratinocytesand requires TNF signaling. Hum Mol Genet
2006;15:531–42.
21. Maeda S, Kamata H, Luo JL, Leffert H, Karin M. IKKbeta
coupleshepatocyte death to cytokine-driven compensatory
proliferation thatpromotes chemical hepatocarcinogenesis. Cell
2005;121:977–90.
22. Leder A, Kuo A, Cardiff RD, Sinn E, Leder P. v-Ha-ras
transgeneabrogates the initiation step in mouse skin tumorigenesis:
effects ofphorbol esters and retinoic acid. Proc Natl Acad Sci U S
A 1990;87:9178–82.
IKKb Leads to Oral Tumorigenesis
www.aacrjournals.org Mol Cancer Res; 9(10) October 2011 1337
on June 29, 2021. © 2011 American Association for Cancer
Research. mcr.aacrjournals.org Downloaded from
Published OnlineFirst August 5, 2011; DOI:
10.1158/1541-7786.MCR-11-0168
http://mcr.aacrjournals.org/
-
23. Stratis A, PasparakisM,Markur D, Knaup R, Pofahl R,Metzger
D, et al.Localized inflammatory skin disease following inducible
ablation of Ikappa B kinase 2 in murine epidermis. J Invest
Dermatol 2006;126:614–20.
24. Cao Y, Bonizzi G, Seagroves TN, Greten FR, Johnson R,
Schmidt EV,et al. IKKalpha provides an essential link between RANK
signaling andcyclin D1 expression during mammary gland development.
Cell2001;107:763–75.
25. Baxter FO, Came PJ, Abell K, Kedjouar B, Huth M, Rajewsky K,
et al.IKKbeta/2 induces TWEAK and apoptosis in mammary epithelial
cells.Development 2006;133:3485–94.
26. Vashisht Gopal YN, Arora TS, Van Dyke MW. Tumour necrosis
factor-alpha depletes histone deacetylase 1 protein through IKK2.
EMBORep 2006;7:291–6.
27. Xia Y, Padre RC, De Mendoza TH, Bottero V, Tergaonkar VB,
VermaIM. Phosphorylation of p53 by IkappaB kinase 2 promotes
itsdegradation by beta-TrCP. Proc Natl Acad Sci U S A
2009;106:2629–34.
28. Chariot A. The NF-kappaB-independent functions of IKK
subunits inimmunity and cancer. Trends Cell Biol
2009;19:404–13.
29. Guo Y, Yuan C, Weghorst CM, Li J. IKKbeta specifically binds
to P16and phosphorylates Ser8 of P16. Biochem Biophys Res
Commun2010;393:504–8.
30. Tsuchiya Y, Asano T, Nakayama K, Kato T Jr, Karin M, Kamata
H.Nuclear IKKbeta is an adaptor protein for IkappaBalpha
ubiquitinationand degradation in UV-induced NF-kappaB activation.
Mol Cell2010;39:570–82.
31. Schottenfeld D, Beebe-Dimmer J. Chronic inflammation: a
commonand important factor in the pathogenesis of neoplasia. CA
Cancer JClin 2006;56:69–83.
32. Aggarwal BB. Nuclear factor-kappaB: the enemy within. Cancer
Cell2004;6:203–8.
33. Hu MC, Lee DF, Xia W, Golfman LS, Ou-Yang F, Yang JY, et
al.IkappaB kinase promotes tumorigenesis through inhibition of
fork-head FOXO3a. Cell 2004;117:225–37.
34. Greten FR, Karin M. The IKK/NF-kappaB activation pathway-a
targetfor prevention and treatment of cancer. Cancer Lett
2004;206:193–9.
35. Yang J, Splittgerber R, Yull FE, Kantrow S, Ayers GD, Karin
M, et al.Conditional ablation of Ikkb inhibits melanoma tumor
development inmice. J Clin Invest 2010;120:2563–74.
36. van Hogerlinden M, Rozell BL, Ahrlund-Richter L, Toftgard R.
Squa-mous cell carcinomas and increased apoptosis in skin with
inhibitedRel/nuclear factor-kappaB signaling. Cancer Res
1999;59:3299–303.
37. Seitz CS, Lin Q, Deng H, Khavari PA. Alterations in
NF-kappaBfunction in transgenic epithelial tissue demonstrate a
growth inhibitoryrole for NF-kappaB. Proc Natl Acad Sci U S A
1998;95:2307–12.
Page et al.
Mol Cancer Res; 9(10) October 2011 Molecular Cancer
Research1338
on June 29, 2021. © 2011 American Association for Cancer
Research. mcr.aacrjournals.org Downloaded from
Published OnlineFirst August 5, 2011; DOI:
10.1158/1541-7786.MCR-11-0168
http://mcr.aacrjournals.org/
-
2011;9:1329-1338. Published OnlineFirst August 5, 2011.Mol
Cancer Res Angustias Page, José L. Cascallana, M. Llanos Casanova,
et al. Oral EpitheliaEpithelia and Exocrine Glands and to Tumoral
Transformation of
Overexpression Leads to Pathologic Lesions in StratifiedβIKK
Updated version
10.1158/1541-7786.MCR-11-0168doi:
Access the most recent version of this article at:
Material
Supplementary
http://mcr.aacrjournals.org/content/suppl/2011/08/05/1541-7786.MCR-11-0168.DC3
Access the most recent supplemental material at:
Cited articles
http://mcr.aacrjournals.org/content/9/10/1329.full#ref-list-1
This article cites 37 articles, 7 of which you can access for
free at:
Citing articles
http://mcr.aacrjournals.org/content/9/10/1329.full#related-urls
This article has been cited by 1 HighWire-hosted articles.
Access the articles at:
E-mail alerts related to this article or journal.Sign up to
receive free email-alerts
Subscriptions
Reprints and
[email protected]
To order reprints of this article or to subscribe to the
journal, contact the AACR Publications Department at
Permissions
Rightslink site. Click on "Request Permissions" which will take
you to the Copyright Clearance Center's (CCC)
.http://mcr.aacrjournals.org/content/9/10/1329To request
permission to re-use all or part of this article, use this link
on June 29, 2021. © 2011 American Association for Cancer
Research. mcr.aacrjournals.org Downloaded from
Published OnlineFirst August 5, 2011; DOI:
10.1158/1541-7786.MCR-11-0168
http://mcr.aacrjournals.org/lookup/doi/10.1158/1541-7786.MCR-11-0168http://mcr.aacrjournals.org/content/suppl/2011/08/05/1541-7786.MCR-11-0168.DC3http://mcr.aacrjournals.org/content/9/10/1329.full#ref-list-1http://mcr.aacrjournals.org/content/9/10/1329.full#related-urlshttp://mcr.aacrjournals.org/cgi/alertsmailto:[email protected]://mcr.aacrjournals.org/content/9/10/1329http://mcr.aacrjournals.org/