PDF hosted at the Radboud Repository of the Radboud University
Nijmegen
The following full text is a publisher's version.
For additional information about this publication click this link.
http://hdl.handle.net/2066/23019
Please be advised that this information was generated on 2017-06-15 and may be subject to
change.
Acta Otolaryngol (Stockh) 1996; 116: 293-298
Squamous Metaplasia of the Middle Ear Epithelium
W. KUIJPERS,' P. P. C. A. V E N N IX ,1'2 T. A. PETERS' and F. C. S. R A M A E K E R S3
From the Departments of Otorhinolaryngology^ 1University of Nijmegen and 2University of Leiden, and 3Department of Molecular Cell
Biology and Genetics, University of L iniburg, Maastricht, Thc Netherlands
Kuijpers W, Vennix PPCÀ, Peters TA, Ramackcrs FCS. Squamous metaplasia of the middle ear epithelium. Acta
Otolaryngol (Stockh) 1996; 116: 293-298,
I
This study cleats with the expression of cytokeratins (Cks) in squamous cell metaplastic lesions in rat and human middle
ear. In rats, squamous metaplastic lesions could be induced during chronic otitis media. The histological features of these
lesions were similar to those observed in'the human middle ear, Irnmunohistochemistry revealed that squamous cell
metaplasia in both rat and human middle ear is characterised by a loss of simple epithelial cell related Cks and the
appearance of Cks characteristic of stratified and corni tying epithelia, This indicates a true change in the differentiation
of the middle ear epithelium. It is concluded that the Ck profile of the cholesteatoma matrix cannot be used as a variable
to decide whether the origin of cholesteatomas is epidermal or metaplastic. This rat model is suitable Tor studying
squamous cell metaplasia in relation to cholcsteatoma genesis. Key words: rats, hitmans, cytokeratin expression.
INTRODUCTION
Squamous metaplasia is assumed to be a possible
source of cholesteatoma and has been suggested to
account for about 4% of the total number encoun­
tered (1). However, this contribution is difficult to
,
, .
,
prove because these lesions will usually go unnoticed
until the tympanic membrane perforates. This makes
them indistinguishable from cholesteatoma genesis
from ingrowing epidermis.
The metaplastic theory is mainly based on the
presence of stratified epithelium in the middle ear
during chronic otitis media (OM ). However, these
studies chiefly showed the presence of areas of noneornifying epithelium, while cornifying epithelium has
only been observed incidentally (1-6). Also, in experimental animals, squamous metaplasia has been observed after the application of a large variety of
irritating chemicals (7-9), but metaplasia as a consequenee of middle ear infections seems to be very rare.
This investigation is a continuation of a previous
experimental study in which we succeeded in inducing
an infection-related squamous cell metaplasia by Eustachian tube obstruction in rats with an upper res­
piratory tract infection (10). In the present study we
were interested in the relation between cytokeratin
expression and histological changes during squamous
cell metaplasia. Cytokeratins (Cks) are intermediate
filament proteins which are exclusively present in
epithelial cells. The expression of the different types
of Cks (numbered 1 -20) depends on the type of
epithelium and stage of differentiation (11-13). Because of these properties Cks seem very suitable for a
better understanding of squamous cell metaplasia in
the middle ear. The metaplastic lesions in the rat were
studied with a broad panel of monospecific antibod­
ies directed against various cytokeratins. In addition,
C) 1996 Scandinavian University Press. IS S N 0001-6489
squamous metaplastic lesions in biopsies taken from
human cases with chronic OM were included in this
study.
\iA T C D U T A\rn x ^ T U A n c
M A T E R IA L A N D METHODS
For the induction of purulent OM, 100 adult Wistar
rats that displayed clinical symptoms of an upper
respiratory airway disease were used. The animals
were checked otoscopically for the absence of otitis
and the eustachian tube was obstructed on both sides
as described previously (14). All the animals developed a purulent middle ear disease within one week
after operation. After survival times varying from 4
months up to 6 months the animals were killed by an
intracardiac injection of pentobarbital sodium and
decapitated. The specimens were either processed for
embedding in glycol methacrylate (G M A ) for histological studies or for preparing cryosections for immunohistochemistry.
For histology the whole middle ear was dissected
from the skull and fixed in phosphate buffered (0.1
mol/L; pH 7*4) 2,5% glutaraldehyde and subse­
quently decalcified in a solution containing 10%
EDTA (pH 7.4). After dehydration the specimens
were embedded in glycol methacrylate (GMA). Sections (2 /¿m) were stained with toluidine blue,
For immunohistochemistry, the middle cars were
immediately after dissection from the skull stored in a
decalcilication solution containing 10% EDTA and
7.5% polyvinylpyrrolidone in 0,1 mol/L of TRIS-HC1.
buffer (pH 7.2) at 4nC for a period ranging from 4 6
days. After rinsing in the same solution without
EDTA, the specimens were frozen in liquid nitrogen.
Cryosections were placed on poly-L~ly sine coatedslides, air dried and fixed in aeeton. Immunohislo-
294
W. Ruijpers et al.
Acta Otolaryngol (Stockh) 116
^
* “ b :LV^ r>
:•!•:>.>,;.: ; •
:
• V j& S & iï .
. .'. "'tile-' * ••
'••*-.lixv*
t/. • !:\
i 7g, 1. Survey (A) and detail (B) of non-cornifying squamous metaplasia in rat middle ear, 4 months after the induction
of chronic otitis media. Note the abrupt transition of the hyperplastic respiratory epithelium and the squamous epithelium.
G M A section, toluidirv blue stain. Bars represent 200 /*m (A), 100 /nn (B).
chemical staining of these sections was performed as
described previously (10). The following monoclonal
cytokeratin antibodies were used: RC K 103 (broadly
reacting); RCK 105 (Ck 7); CK 18-2 (Ck 18); M20
(Ck 8); 6B10 (Ck 4); 1C7 (Ck 13); R C K 107 (Ck
14); RKSE 60 (Ck 10). For specifications of these
antibodies and references, see (15). All these antibodies with known specificities for human Cks cross-react
with the rat cytokeratins except for M20 (Ck 8).
In addition to these rat ears, 50 biopsy specimens
of human middle ear epithelium from cases of
chronic OM were included in this immunohistochemical study. The biopts were processed in the same way
as the rat specimens, but the décalcification step was
omitted»
RESULTS
Rats
All the rat ears were filled with a purulent mass. The
middle ear mucosa was largely thickened and
Table I.
contained numerous inflammatory cells. The hyper­
trophic epithelial lining was very irregular with cyst­
like formations. It often showed local discontinuities
and contained numerous secretory cells. Two ears
which were embedded in G M A revealed areas of
squamous cell metaplasia. These areas consisted of
stratified non-cornifying epithelium with a regularly
arranged layer of columnar basal cells and flattened
suprabasal cells. There was an abrupt transition be­
tween the hyperplastic pseudostratified epithelium
and the squamous epithelium (Fig. 1).
From the ears which were processed for immunohistochemistry two specimens revealed areas of cornifying squamous metaplasia. Immunohistochcmical
staining showed a homogeneous expression of the
simple eithelial cell Cks 18 and 19 in the middle ear
epithelium whereas Ck 7 was expressed heteroge­
neously, From the stratification markers, Ck 4 was
expressed sparse to heterogeneously and Ck 13 was
absent. Expression of the basal cell Ck 14 was limited
to the basal cells. These data do not fundamentally
Cytokeratin expression in middle ear epithelium of rats and hitmans
-1- Homogeneous positive reaction, — Negative reaction, ± Sparse to heterogeneous reaction, B Staining of basal cells,
!Duta from references 10 and 15, 2Antibody M20 does not cross-react with rat Ck 8
Rats
Normal1
Ck
7
8
18
19
4
13
10
14
+
2
+
+
±
—
—
+B
Humans
Metaplasia
Normal1
2
+
H-
—
—
+
±
+
Metaplasia
non-cornifying
Metaplasia
cornifying
—
—
—
—
+
dh
±
+
+
—
—
-f-B
+
—
—
—
-h
+
Acta Otolaryngol (Stockh) 116
Metaplasia in the middle ear
295
Fig. 2. Cytokeratin expression in metaplastic eornilying squamous epithelium in the rat middle ear, 4 months after the
induction of chronic otitis media. (A ) Survey showing expression of Ck 4 in metaplastie epithelium (m), but not in the
middle ear epithelium (e). (B) Survey showing expression of Ck 14 in both metaplastie epithelium (m) and middle ear
epithelium (e). (C) Detail of metaplastie epithelium in (A ) showing expression of Ck 4 in all suprabasal cell layers. (D)
of Ck 14 in all cell layers. (E) Metaplastie epithelium showing
scattered suprabasal expression of Ck 10. Bars represent 500 //m (A , B), 100 pm (C E).
differ from those observed in the normal middle ear
epithelium (Table I, (10, 15)). The immunohistochemical data of the metaplastie epithelium are sum­
marised in Table I and depicted in Fig. 2. This
showed a sparse to heterogeneous expression in the
cells (Fig. 2E). No expression of the
markers for simple and complex epithelia (Cks 7, IS,
19) was observed.
and 13 in all suprabasal cell layers (Fig. 2A, C), while
the basal cell marker Ck 14 was expressed in all cell
layers ( Fig. 2B, D). The corniiicattion marker Ck 10
Hitman specimens
The mucosa of the biopsy specimens from chronic
OM invariably showed the presence of numerous
296
W. Kuijpers et al.
Acta Otolaryngol (Stockh) 116
Fig. 3. Differential expression of various cytokeratins in metaplastic squamous epithelium in human middle ear. (A) Survey
shows a gradual transition from hyperplastic middle ear epithelium (e) to non-cornifying stratified epithelium (nc), which
passes into cornifying stratified epithelium (c) Note exclusive expression of Ck 8 in the hyperplastic middle ear epithelium
and the abrupt transition between cornifying epithelium and middle ear epithelium. (B) Expression of Ck 19 in middle ear
epithelium and non-cornifying stratified epithelium. (C) Expression of Ck 4 in middle ear epithelium and non-cornifying
stratified epithelium. (D) Expression of Ck 14 in the basal cells of the middle ear epithelium and in all cells of the stratified
epithelia. (E) Exclusive expression of Ck 10 in the suprabasal cells of the cornifying stratified epithelium. Bars represent 500
/¿m (A), 200 /tm (B-E).
inflammatory cells. The epithelium was hypertrophic.
It contained many secretory cells and crypts lined
with a varying number of secretory and non-secretory
cells. In three biopts squamous metaplasia was ob­
served. It consisted of non-cornifying squamous ep­
ithelium in two specimens. The third specimen
showed a gradual change from hyperplastic pseudos­
tratified epithelium into non-cornifying stratified ep­
ithelium, which gradually passed into cornifying
epithelium with keratohyalin granules (Fig.3). There
was an abrupt transition between the cornifying ep­
ithelium and the middle ear epithelium (Fig. 3A). The
Ck profile of the middle ear epithelium in all the
biopts was similar to that observed in the normal
epithelial lining (Table I, (10)). It showed homoge­
neous expression of the simple epithelial cell Cks 7, 8,
18, 19 and a scarce to heterogeneous expression of
the stratification markers Cks 4 and 13. Expression of
the basal cell marker Ck 14 was limited to the basal
cells. The same expression profile was observed in the
hyperplastic pseudostratified epithelium in the meta­
plastic area (Fig. 3) although the expression of Cks 4
Metaplasia in the middle ear
Acta Otolaryngol (Stockh) 116
and 13 was more marked. In the non-cornifying
metaplastic squamous epithelium Cks 4 and 13 were
homogeneously expressed in all suprabasal cell layers
(Fig. 3C), whereas Ck 14 was expressed in all cell
layers (Fig. 3D). No expression was observed for the
cornification marker Ck 10 (Fig, 3E) and the markers
of simple epithelia (Fig. 3A), except for Ck 19 which
was expressed in all cell layers (Fig. 3B). The cornifying epithelium showed suprabasal expression of the
cornification marker Ck 10 (Fig. 3E). The basal cell
marker Ck 14 extended into all suprabasal cell layers
(Fig. 3D). No expression of Cks 4, 13, 7, 8, 18 and 19
was observed in this area (Fig. 3A-C), In the area of
non-cornifying epithelium adjacent to the cornifying
epithelium Cks 10, 4 and 13 were co-expressed,
DISCUSSION
This study demonstrates that the middle ear epithe­
lium of the rat can undergo squamous cell metaplasia
during chronic OM. This finding is in line with
comparable observations in the human middle ear (1)
and in other areas of the respiratory tract (16, 17),
which shows the susceptibility of this epithelium for
squamous cell metaplasia.
The present immunohistochemical observations
demonstrate that squamous metaplasia of the middle
ear epithelium of the rat is associated with a loss of
simple epithelial cell related Cks. This is paralelled by
the appearance of the stratification markers of noncornifying epithelia Cks 4 and 13 and to a lesser
extent of the cornification marker CK 10, In addi­
tion, the basal cell marker Ck 14 extends into all cell
layers. This altered Ck profile must be due to a
modification of the differentiation character of the
progenitor cells of the middle ear epithelium.
A similar change of the Ck profile is observed in
the human specimens. However, in these specimens
the simple epithelial cell marker Ck 19 is retained in
the non-cornifying lesion, whereas in one specimen
squamous cell metaplasia had developed further into
a lesion which phenotypically resembled epidermis. In
this lesion Ck 10 expression substituted completely
for Cks 4 and 13, which makes the epithelium indis­
tinguishable from normal epidermis. Remarkably,
also in the human lesions expression of the basal cell
Ck 14 extended into the suprabasal cells. This upregulated Ck 14 expression indicates a state of hyperpro­
liferation.
These immunohistochemical findings are very simi­
lar to those observed in squamous cell metaplasia of
the lung (16) and the nose (17), except for the
expression of the basal cell Cks, which remain limited
to the basal cells. Extension of the basal cell Cks into
the suprabasal cell layers has only been reported in
297
squamous cell metaplasia induced by vitamin A de­
pletion (18).
It has been suggested that the presence of non-epidermal Cks in cholesteatoma matrix refers to a meta­
plastic origin (19). However, the similarity of the Ck
profile in the epidermal type of differentiation of a
metaplastic lesion (exclusive expression of Ck 10)
with that of the cholesteatoma matrix originating
from ingrowing epidermis (10) is in conflict with this
suggestion, In addition, expression of non-epidermal
Cks has also been observed in dysplastic areas of the
advancing front of cholesteatomas originating from
ingrowing epidermis (10, Vennix, impubi, data).
Summarizing, this study shows that infection in­
duced squamous cell metaplasia in the rat middle ear
appears to be-an appropriate model for studying
metaplasia in relation to cholesteatoma genesis.
REFEREN CES
1. Sade J, Babiacki A, Pinkus G. The metaplastic and
congenital origin of cholesteatoma. Acta Otolaryngol
(Stockh) 1983; 96: 119-29.
2. Birrel JF. Cholesteatosis of the attic. J Laryngol Otol
1958; 72; 620-5.
3. Bendek GA. Histopathology of transudatory secretory
otitis media. Arch Otolaryngol 1963; 78; 33-8.
4. Paparella M M , Dito WR, Enzyme studies in serous
otitis media. Arch Otolaryngol 1964; 79: 393-8.
5. Palva T, Palva A, Dammert K. Middle ear mucosa and
chronic middle car disease. Arch Otolaryngol 1968; 87:
3.11.
6 . Bernstein JM, Hayes ER, Ishikawa T, Tomasi TB,
Herd JK. Secretory otitis media: a histochemieal and
immunochemical report. Trans Am Acad Ophthalmol
Otol 1972; 76: 1305-18.
7. Schroer A, Tierexperimentelle Studien zu genetischen
Problemen des Mittelohr-Cholestcatoms. Zschr Laryn­
gol Rhinot Otol 1962; 41: 104-8.
8. Greenberg SD, Dickey JR, Neely JG. Effects of to­
bacco on the ear. Ann Otol Rhinol Laryngol 1973; 82:
311.7.
9. Wright CG, MeycrholT W L Burns DK. Middle ear
cholesteatoma. An animal model. Am J Otolaryngol
1985; 6; 327 41.
10. Vennix PPCA, Kuijpers W, Tonnaer ELGM, Peters
TA, Ram ackers FCS. Cy to keratins in induced epider­
moid formations and cholesteatoma lesions. Arch Oto­
laryngol Head Neck Surg 1990; 116: 560-5.
11. Moll R, Franke WW, Schiller DL, Geiger B, Krepler
R. The catalogue of human cyto keratins: patterns of
expression in normal epithelia, tumors and cultured
cells. Cell 1982; 31: 11 24.
12. Sun TT, Tseng SCG, Huary AJW. Monoclonal anti­
body studies of mammalian epithelial keratins. A re­
view. Ann NY Acad Sci 1985; 455: 307 29.
13. Ramaekers FCS, Vooijs GP, Smedts l\ Keratins as
differentiation markers in tumor biology and surgical
pathology. In: Spa nd id os D, ed. Current perspectives
on molecular and surgical oncology, London: JAI
Press, 1992: Vol 1 (part A) 285 318.
298
W. Knijpers ei a i
14. Kuijpers W, Beek van der JM H. The effect of eustachian tube obstruction on middle ear mucosal
transformation. In: Bernstein JM, Ogra PL, eds. Im­
munology of the ear. New York: Raven Press, 1986:
205-29.
15. Vennix PPCA, Kuijpers W, Peters TA, Tonnaer
ELGM» Ramaekers FCS. Growth and differentiation
of meatal skin grafts in the middle ear of the rat.
Arch Otolaryngol Head Neck Surg 1994; 120; 110211.
16. Leube RE, Rustad TJ. Squamous cell metaplasia in
the human lung: molecular characteristics of epithelial
stratification Virchows Arch B Cell Pathol 1991; 61:
227-53.
17. Stosiek P, Kasper M, Moll R. Changes in cytokeratin
expression accompany squamous metaplasia of the
human respiratory epithelium Virchows Arch A
Pathol Anat 1992; 421: 133-41.
Acta Otolaryngol (Stockh) 116
18. Lancilotti F, Darwiche N, Celli G, DeLuca LM. Retinoid
status and the control of keratin expression and adhesion
during the histogenesis of squamous metaplasia of tra­
cheal epithelium. Cancer Res 1992; 52: 6144-52.
19. Blitterswijk van CA, Grote JJ, Lutgert RW. Cyto­
keratin patterns of tissues related to cholesteatoma
pathogenesis. Ann Otol Rhinol Laryngol 1989; 98:
635-40.
Address for correspondence:
Dr. W, Kuijpers
Dept, of Otorhinolaryngology
University of Nijmegen
Ph. van Leydenlaan 15
NL-6500 HB Nijmegen
The Netherlands
Tel: +31 24-3615144
Fax: +31 24-3540251