Acta Oto-Laryngologica. 2014; 134: 124–134

ORIGINAL ARTICLE

Expression of the epidermal growth factor system in human middle ear cholesteatoma

Acta Otolaryngol Downloaded from informahealthcare.com by Nyu Medical Center on 05/27/15 For personal use only.

METTE BENDIXEN THORUP1,2, MATHIAS MUNK2, STEEN SEIER POULSEN3, MICHAEL GAIHEDE4, EBBA NEXO2, BOE SANDAHL SORENSEN2 & THERESE OVESEN1 1

Department of Otorhinolaryngology, 2Department of Clinical Biochemistry, Aarhus University Hospital (NBG), Aarhus, 3Department of Biomedical Sciences, The Panum Institute, University of Copenhagen, Copenhagen and 4 Department of Otolaryngology, Head and Neck Surgery, Aarhus University Hospital, Aalborg, Denmark

Abstract Conclusion: The detection of the HER4 receptor in 50% of cholesteatomas but never in the reference tissue, and the increased expression of its activating ligand EPI, suggest that EPI-mediated activation of HER4 might play a role in cholesteatoma growth. Objective: To investigate the expression of the epidermal growth factor (EGF) system in human middle ear cholesteatoma. Methods: Forty-seven patients referred for surgery due to cholesteatoma were included in the study. Clinical data were collected. Biopsies of cholesteatoma and skin from the external ear canal were obtained during surgery. mRNA expression was quantified with real-time PCR. The corresponding proteins were visualized using immunohistochemistry. Results: A systematic investigation of all four receptors, HER1, HER2, HER3, and HER4, and the ligands EGF, transforming growth factor (TGF)-a, amphiregulin (AR), heparin-binding EGF-like growth factor (HB-EGF), and epiregulin (EPI) of the EGF system is presented. At the mRNA level, the study demonstrates an up-regulation of mRNA encoding EPI and AR. In contrast HER1 and EGF were down-regulated. HER4 mRNA could be detected in 50% of cholesteatoma and 20% of reference tissues, and the HER4 protein was detectable only in cholesteatoma tissue. HER1 and HER2 were also visualized by immunohistochemistry, whereas the ligands EPI, AR, and EGF were undetectable with our methods.

Keywords: Epiregulin, amphiregulin, HER1, HER4, EGF, Q-PCR

Introduction Cholesteatoma is characterized by the occurrence of keratinizing stratified squamous epithelium within the middle ear cavity, often extending to other parts of the temporal bone, especially the mastoid. The exact pathogenesis is not known, but cholesteatoma development and growth may relate to a hyperproliferative ability of the epithelium [1–4]. The main component of cholesteatoma matrix and of human epidermis is keratinocytes. Among the most important growth factors for human keratinocyte proliferation are members of the epidermal growth factor (EGF) family [5]. Thus, it is tempting to speculate that cholesteatoma is related to a deregulation of the EGF system.

The EGF system comprises a family of four receptors, namely the human epidermal growth factor receptor 1 ((HER1/EGFR/ErbB1), HER2 (ErbB2), HER3 (ErbB3), and HER4 (ErbB4)) and their activating ligands [6]. The receptors are transmembrane receptors with a ligand binding domain, a transmembrane region, and an intracellular domain. Ligand binding induces the formation of receptor homoand heterodimers with the exception that HER2 can dimerize without ligand binding [7]. The ligands of the EGF system can be divided into three groups related to their ability to bind to the receptors: the EGF, transforming growth factor-a (TGF-a), and amphiregulin (AR) are selectively recognized by HER1. Betacellulin (BTC), heparin-binding

Correspondence: Mette Bendixen Thorup MD PhD, Department of Otorhinolaryngology, Aarhus University Hospital (NBG), DK-8000 Aarhus C, Denmark. Tel: +45 78463175. Fax: +45 78463180. E-mail: [email protected]

(Received 16 June 2013; accepted 12 September 2013) ISSN 0001-6489 print/ISSN 1651-2251 online  2014 Informa Healthcare DOI: 10.3109/00016489.2013.847282

EGF-system in cholesteatoma

Acta Otolaryngol Downloaded from informahealthcare.com by Nyu Medical Center on 05/27/15 For personal use only.

EGF-like growth factor (HB-EGF), and epiregulin (EPI) show dual specificity recognized both by HER1 and HER4. Finally, an additional group of ligands represented by the neuregulins (NRGs) is recognized primarily by HER3 and HER4. Alterations in the expression of the EGF system are a common feature in relation to cancer [6], but they have also been related to other pathological conditions with imbalanced growth such as psoriasis [5]. In the present study we explored the expression of the EGF system in human middle ear cholesteatoma compared to skin from the bony part of the external ear canal (EAC) from the same patient. Material and methods Study population Forty-seven patients referred for surgery due to cholesteatoma at the Departments of Otorhinolaryngology and Head and Neck Surgery, Aarhus and Aalborg, Aarhus University Hospital, Denmark, were recruited during 2006–2008. Clinical data about sex, age, site of origin from the tympanic membrane, type of cholesteatoma (primary/recurrent), spread, and skin disorders (dermatitis, atopic dermatitis, and psoriasis) were collected. Biopsies were obtained during middle ear surgery from three different sites of the cholesteatoma (stem, middle ear, and mastoid) (Figure 1), and for

125

comparison a sample of skin from the bony part of the EAC was obtained from the incision line. Thus, up to four biopsies were obtained from each patient. Infections in the middle ear were treated before patients had cholesteatoma surgery, reducing the influence of infectious processes. After removal, the biopsies were immediately placed in RNase-free tubes (2.2 ml, Eppendorf tubes, Invitrogen, Taastrup, Denmark), snap frozen in dry ice and stored at –80 C until processing. Due to the character of the biopsies, it was not possible to separate the epithelium from the connective tissue, although it would have been preferable. The study was approved by the local scientific ethic committee (journal number 20051044/20050222), and all patients or their parents (for patients below the age of 18 years) gave informed consent before surgery and participation in the study. RNA extraction RNA was extracted using the QIAamp Blood Mini kit (Qiagen, Hilden, Germany). The biopsies were placed in the lysis buffer containing beta-mercaptoethanol and homogenized using a rotor-stator (Heidoplh DIAX 600 disperser, Heidolph Elektro GmbH & KG, Kelheim, Germany). The samples were kept on ice to avoid degradation of RNA. Samples were incubated with DNase (Qiagen) for 15 min at room temperature. The concentration and the purity of the RNA extraction were checked by use of the GeneQuant II instrument (Amersham Biosciences (GE Healthcare), Little Chalfont, UK). Reverse transcription Samples were diluted to an RNA concentration of 0.1 mg/ml. cDNA was made using the high capacity cDNA Reverse Transcription kit (Applied Biosystems, Foster City, CA, USA). A 16mer of T nucleotides was used as primer. Real-time PCR

Figure 1. Schematic drawing of a cholesteatoma arranged as a cyst with a body containing keratin and cellular debris and a stem that attaches the body to the tympanic membrane. The red circles indicate biopsy sites, stem, cholesteatoma in the middle ear, cholesteatoma in the mastoid, and skin from the external ear canal (EAC).

Real-time PCR was conducted on a LightCycler480 (Roche, Basel, Switzerland). For all analyses except HER2 the following mix was used. Five ml LC480 SYBR mastermix, 0.5 ml sense-primer (see Table I for primer sequences and concentration), 0.5 ml anti-sense primer, 3 ml H2O, and 1 ml of target cDNA (prepared as described above) in a total volume of 10 ml. For the analysis of HER2, a hybridization probe was used (5’-CAGATTGCCAAGGGGATGAGCTAC

Acta Otolaryngol Downloaded from informahealthcare.com by Nyu Medical Center on 05/27/15 For personal use only.

126

M. B. Thorup et al.

CTG-3’, labeled with either FAM or TAMRA at the 5’ and 3’ ends, respectively (10 pmol)). The PCR mix consisted of 5 ml LightCycler480 Probes Master (Roche), 0.5 ml sense primer, 0.5 ml anti-sense primer, 0.5 ml TaqMan probe, 2.5 ml H20, and 1 ml target DNA in a total volume of 10 ml. RNA extracted from the bladder cancer cell line HCV29 was used as a calibrator for HER1, HER2, EGF, AR, TGF-a, HB-EGF, and EPI, whereas RNA from the endometrial cell line HEC 1A (ATCC, USA) was used as a calibrator for HER3, and RNA from another endometrial cancer cell line KLE (ATCC) was used as a calibrator for HER4. The calibration curves consisted of a series of two-fold dilutions with the highest concentration being 1.0 mg and the lowest being 121.8 pg, respectively. The calibrators and a positive and a negative control (no RNA added) were included in each run. The PCR program consisted of an initial hot start at 95 C for 10 min, a denaturing step 95 C for 10 s, an annealing temperature specific for each primer set (Table I), and an elongation step 72 C for 5 s. Normalization Using the program NormFinder, the expression of three different genes, namely B2M, YWHAZ, and

b-actin was tested. b-Actin turned out to be the best reference gene, presenting with a stability value of 0.418 compared to 0.619 and 0.548 for B2M and YWHAZ, respectively, and was used as an internal standard. Histology Cholesteatoma biopsies from five separate patients were processed for immunohistochemical examination. Frozen tissue samples were cut on a cryostat into 7–10 mm sections and mounted on slides. The sections were fixed in acetone for 10 min at 4 C. The specimens were preincubated for 5 min in 5% bovine serum albumin (A 4503, Sigma-Aldrich, St. Louis, USA) in phosphate-buffered saline (PBS), boiled in a microwave oven at pH 6 at maximum power for 15 min, and thereafter incubated overnight at room temperature with primary antisera. The immunoreactions were visualized by means of biotinylated rabbit anti-mouse immunoglobulins (code E354, Dako, Glostrup, Denmark) for the monoclonal antibodies and biotinylated swine anti-rabbit immunoglobulins (code E353, Dako) for polyclonal antibodies, diluted 1:200 as the second layer, followed by strepavidin-peroxidase complex, Vectastain PK-4000 (Vector Laboratories, Burlingame, CA,

Table I. Primers and analytical conditions for measurement of mRNA coding for the EGF receptors and ligands. Target

Primer sequence

HER1

S: GAG AAC GCC TCC CTC A

HER2

S: AGA TGT TCG GCC CCA

HER3

S: GGT GCT GGG CTT GCT TTT

No. of PCR Calibration Concn Annealing curve (pmol/ml) temperature ( C) cycles product (bp) 5

54

50

261

HCV29

5

59

45

272

HCV29

5

65

50

365

HCV29

5

65

50

265

KLE

10

58

50

266

HCV29

5

70

50

528

HCV29

5

61

50

282

HCV29

5

66

50

90

10

60

50

238

AS: GGT ACT CGT CGG CAT C

AS: GTG GAG CCC CCC GCT

AS: CGT GGC TGG AGT TGG TGT TA HER4

S: TGT GAG AAG ATG GAA GAT GGC AS: GTT GTG GTA AAG TGG AAT GGC

AR

S: GGC TCA GGC CAT TAT GC AS: ACC TGT TCA ACT CTG ACT GA

TGF-a

S: GCC CGC CCG TAA AAT GGT CCC CTC AS: GTC CAC CTG GCC AAA CTC CTC CTC TGG G

HB-EGF S: GGT GGT GCT GAA GCT CTT TC AS: CCC CTT GCC TTT CTT CTT TC EGF

S: CAC TTG GGA GCC TGA GCA GAA

KLE

AS: CAT GCA CAA GTG TGA CTG GAG GT EPI

S: CAA AGT GTA GCT CTG GAC ATG

HCV29

AS: CTG TAC CAT CTG CAG AAA TA A, adenosine; AR, amphiregulin; C, cytosine; EGF, epidermal growth factor; EPI, epiregulin; HB-EGF, heparin-binding EGF-like growth factor; G, guanine; HCV29, cell line originating from bladder mucosa; HER1, human epidermal growth factor receptor 1; KLE, cell line derived from poorly differentiated endometrial carcinoma; T, thymine; TGF-a, transforming growth factor-a.

EGF-system in cholesteatoma

127

A b

a 12 HER2 mRNA expression (HER2/b-action ratio)

10

3 2 1

8 6 4 3 2 1 0

0 l na

al

c ar

e

t

r

e dl

ea

om

t ea

st

le

o ch

al

a

om

St

n er

Ex

a

em

a

st

id

o ch

al

r

e dl

ea

st

o ch

t ea

st

le

id

o ch

to

as

M

id

M

c

om

t ea

le

M

id

d

15 HER4 mRNA expression (HER4/b-action ratio)

10 5 4 3 2 1 0

al

0.4 0.3 0.2 0.1 0.03

0.02

0.01

0.00 l

a an

a

m te

e

r

e dl

id

M

ea

o ch

t ea

st

le

d

oi

t as

o ch

t ea

M

al

e

r

e

i

l dd

M

ea

o ch

om

t ea

st

le

n er

a

om

St

c ar

t

Ex

a

em

an

om

st

le

n er

al

a

om

S

c ar

t

Ex

e

t

a

om

St

c ar

n er

Ex

to

a

em

an

t ea

le

as

M

HER3 mRNA expression (HER3/b-action ratio)

Acta Otolaryngol Downloaded from informahealthcare.com by Nyu Medical Center on 05/27/15 For personal use only.

HER1 mRNA expression (HER1/b-action ratio)

8 7 6 5 4 4

st

le

id

o ch

t ea

to

as

M

Figure 2. (A) Expression of mRNA coding for the four human epidermal growth factor (HER) receptors (a, HER1; b, HER2; c, HER3; and d, HER4) in biopsies from patients with cholesteatoma (n = 47). Results from analysis of biopsies taken from different parts of cholesteatoma (stem, middle ear, and mastoid, see Figure 1) and skin from the external ear canal are indicated. All values are expressed relative to the expression of the household gene b-actin to control for differences in mRNA concentration and integrity. Median and interquartile (IQR) ranges are indicated. (B) Expression of mRNA coding for the five ligands (a, epiregulin (EPI); b, amphiregulin (AR); c, epidermal growth factor (EGF); d, heparin-binding EGF-like growth factor (HB-EGF); and d, transforming growth factor-a (TGF-a)) in biopsies from patients with cholesteatoma (n = 47). Results from analysis of biopsies taken from different parts of cholesteatoma (stem, middle ear, and mastoid, see Figure 1) and skin from the external ear canal are indicated. All values are expressed relative to the expression of the household gene b-actin to control for differences in mRNA concentration and integrity. Median and interquartile (IQR) ranges are indicated.

USA), diluted 1:100 for 20 min as the third layer. The sections were finally stained by means of 3,3-diaminobenzidine for 30 min and counterstained with hematoxylin. The following primary antibodies were used: HER1 (1:400; NCl-EGFR, Novocastra, Milton Keynes, UK), HER2 (1:1600; code A 485;

Dakopatts, Glostrup, Denmark), HER4 (Ab-4 Clone HFR-1, monoclonal, Neomarkers, Fremont, CA, USA), EPI (1:1600; MAB 1425, clone 183625, R&D Systems, Minneapolis, USA), AR (1:200; RB-257-p, Neomarkers) and EGF (1:2500; 4554, polyclonal rabbit, produced in-house).

128

M. B. Thorup et al. B

8 6 4 2 0

l

a an

st

r

le

ea

o ch

dd

i

M

d

oi

80 60 40 20 0

al

st

o ch

r

e dl

HB-EGF mRNA expression (HB-EGF/b-action ratio)

EGF mRNA expression (EGF/b-action ratio)

1.0

0.5

l

a

m te

id

to

as

ea

st

t

r

e

l dd

ea

o ch

M

i

0.6 0.4 0.2 0.0

em

St

ea

t

r

e dl

id

M

ea

a

m

to

o ch

a te

s

le

n er

id

to

o ch

s

a te

a

m

to

le

as

M

r

e

l dd

ea

om

om

o ch

t ea

st

le

M

0.8

l

ea

n er

a

a

em

St

t

Ex

e

na

l

na

a rc

al

4 3 2 1 1.0

a rc

0

st

o ch

to

i

M

id

5

t ea

le

as

10

om

t ea

le

n er

100 80 60 40 20 15

a

om

S

rc

Figure 2. (Continued).

st

le

o ch

M

M

1.5

a an

Ex

ea

id

M

30 20 10

al

st

o ch

t ea

d

0.0

Ex

t ea

le

t

Ex

c

al

e

al

n er

om

om

St

c ar

a

a

em

an

t ea

le

t as

2000 1500 1000 500 100

om

t ea

le

t

Ex

TGF-a mRNA expression (TGF-a/b-action ratio)

Acta Otolaryngol Downloaded from informahealthcare.com by Nyu Medical Center on 05/27/15 For personal use only.

ea

n er

a

om

S

rc

al

a

m te

b

Amphiregulin mRNA expression (amphiregulin/b-action ratio)

Epiregulin mRNA expression (epiregulin/b-action ratio)

a 100 80 60 40 20 10

id

to

M

as

st

le

o ch

t ea

EGF-system in cholesteatoma A

B a

a

* P = 0.030

8 HER1 mRNA expression (HER1/b-action ratio)

Epiregulin mRNA expression (epiregulin/b-action ratio)

40

30

20

10

0

Acta Otolaryngol Downloaded from informahealthcare.com by Nyu Medical Center on 05/27/15 For personal use only.

External ear canal

b

20

10

0 External ear canal Mastoid cholesteatoma

* P = 0.023

6

4

2

External ear canal Middle ear cholesteatoma

c 8 HER1 mRNA expression (HER1/b-action ratio)

10

Stem

0

* P = 0.014

20

2

8

External ear canal Middle ear cholesteatoma

c

4

External ear canal

0

30

6

b

* P = 0.048

30

* P = 0.013

0

Stem

HER1 mRNA expression (HER1/b-action ratio)

Epiregulin mRNA expression (epiregulin/b-action ratio)

40

Epiregulin mRNA expression (epiregulin/b-action ratio)

129

*** P < 0.001

6

4

2

0 External ear canal Mastoid cholesteatoma

Figure 3. Paired samples of skin from the external ear canal and cholesteatoma stem (a), middle ear (b), and mastoid (c). mRNA expression is presented relative to the expression of the household gene b-actin to control for differences in mRNA concentration and integrity. (A) Epiregulin/b-actin ratio and (B) HER1/b-actin ratio. Analyses were done using Wilcoxon matched pairs signed rank test. p values are indicated.

Statistics Data were analyzed and figures were produced using STATA 9 and GraphPad Prism version 5. The Wilcoxon matched pairs signed rank test was used for

comparison of expression levels of receptors and ligands of the EGF system in cholesteatoma and reference tissue from each patient. The Wilcoxon stratified test was used to calculate a summary p value for a within-patient comparison of cholesteatoma sites

130

M. B. Thorup et al.

versus the reference site. Associations between continuous variables were assessed by Spearman’s rank correlation. A two-tailed p value £ 0.05 was considered significant. Results

Acta Otolaryngol Downloaded from informahealthcare.com by Nyu Medical Center on 05/27/15 For personal use only.

Patient population The study population consisted of 47 patients, 23 (49%) females and 24 (51%) males. Median age was 28 years (range 2–73 years). At the time of surgery 32 (68%) of the patients presented with primary and 15 (32%) with recurrent cholesteatoma. In patients with primary cholesteatoma, 12 (38%) originated from the attic, 13 (41%) from the tympanic sinus, 3 (9%) from the pars tensa, and 4 (13%) from various sites (mixed cholesteatoma). The cholesteatoma was located in the middle ear in 10 (21%), in both middle ear and the antrum in 10 (21%), and in the middle ear, antrum extending to the mastoid in 27 (57%) of the cases. Six (13%) of 47 patients had a history of skin disorders (dermatitis, atopic dermatitis, and psoriasis). Expression of mRNA encoding the EGF system BiopsiesofcholesteatomaandEACskinfrom47patients were analyzed. In six cases reference tissue could not be obtained due to surgical considerations and in six cases it was not possible to purify RNA from the biopsy because of its limited size. The results are expressed relative to the expression of b-actin for each sample. A significant up-regulation of mRNA encoding the ligand EPI was observed in all parts of the cholesteatoma tissue (stem, p = 0.030, n = 23;

middle ear, p = 0.048, n = 23; mastoid, p = 0.014, n = 23) with median concentrations 4.8-fold (stem), 4.2-fold (middle ear), and 3.0-fold (mastoid) higher than in skin from the EAC (Figures 2B and 3A). The results were supported by the results obtained from the Wilcoxon stratified test (p = 0.029) (Table II). A significant up-regulation of mRNA encoding the ligand AR (p = 0.026, n = 23) was found in the stem of cholesteatoma compared with skin from the EAC (2.1-fold). This difference was not apparent for cholesteatoma in the middle ear and mastoid (Figure 2B). Applying the Wilcoxon stratified test, AR showed a borderline significant p value (p = 0.051) (Table II). This indicates a common trend towards an upregulation of AR in all parts of the cholesteatoma, although significant p values were only found in the stem when comparing the individual sites to the reference tissue. The median level for EGF mRNA was 18.5-fold lower in cholesteatoma from the mastoid than in the reference tissue (p = 0.028, n = 23), while no difference was observed in the stem and middle ear (Figure 2B). A significant outcome of the Wilcoxon stratified test (p = 0.021) (Table II) indicates a common trend towards a down-regulation of mRNA encoding EGF in all parts of cholesteatoma tissue compared with reference tissue. There were no differences between cholesteatoma tissue and reference tissue concerning the expression of the other ligands (HB-EGF and TGF-a). The EGF receptors showed a significant reduction of mRNA encoding HER1 in all sites (stem, p = 0.013, n = 23; middle ear, p = 0.023, n = 23; mastoid, p < 0.001, n = 23) with median

Table II. Results of the Wilcoxon stratified test. Variable

Stratified Wilcoxon’s test

p value

Up- or down- regulated

HER1

16.229

< 0.001*

HER2

0.005

0.943

...

HER3

0.591

0.442

...

HER4

1.411

0.235

...

EGF

3.805

0.021*

Down

EPI

4.760

0.029*

Up

AR

3.805

0.051*

Up

TGF-a

2.042

0.153

...

HB-EGF

1.374

0.241

...

Down

The table displays results of the Wilcoxon stratified test that summarizes p value for a within-patient comparison of cholesteatoma sites versus the reference site. A significant outcome (asterisk) indicates a common trend towards an up- or down-regulation of the component despite only one or two of the individual sites presenting a significant p value between the site and the reference sample. AR, amphiregulin; EGF, epidermal growth factor; EPI, epiregulin; HB-EGF, heparin-binding EGF-like growth factor; HER1, human epidermal growth factor receptor 1; TGF-a, transforming growth factor-a.

Acta Otolaryngol Downloaded from informahealthcare.com by Nyu Medical Center on 05/27/15 For personal use only.

EGF-system in cholesteatoma

A

B

C

D

E

F

G

131

H

Figure 4. Localization of HER1 and HER2 in cholesteatoma tissue (stem, middle ear, and mastoid) and skin from the external ear canal (reference tissue) identified by immunohistochemistry. HER1 and HER2 showed the same staining pattern in cholesteatoma and reference tissue and were localized to the cell surface in the stratum basale and lower part of the stratum spinosum. There was no difference in staining intensity between the three biopsy sites or between cholesteatoma and reference tissue. The heterogeneic nature of cholesteatoma is evident, with varying thickness of epithelium and connective tissue. (A) HER1, skin from the bony part of the external ear canal (reference tissue) (235); (B) HER1, cholesteatoma from the stem (370); (C) HER1, cholesteatoma from the middle ear (370); (D) HER1, cholesteatoma from the mastoid (370); (E) HER2, skin from the bony part of the external ear canal (370); (F) HER2, cholesteatoma from the stem (235); (G) HER2, cholesteatoma from the middle ear (235); and (H) HER2 cholesteatoma from the mastoid (370).

Acta Otolaryngol Downloaded from informahealthcare.com by Nyu Medical Center on 05/27/15 For personal use only.

132

M. B. Thorup et al.

concentrations 2.5-fold (stem), 3.1-fold (middle ear), and 4.0-fold (mastoid) lower than in the reference tissue (Figures 2A and 3B). The results were supported by the results obtained from the Wilcoxon stratified test (p < 0.001) (Table II). No differences were displayed in the mRNA expression encoding HER2, HER3, and HER4 between cholesteatoma tissue and reference tissue. However, HER4 was only detected in 20% of reference tissue samples as opposed to 50% of cholesteatoma samples (Figure 2A). Finally, there were no significant associations between the levels of mRNA expression of the EGF system, age, sex, and the origin, type, and extent of the cholesteatoma. Immunohistochemical examinations Immunohistochemical analysis was performed for components of the EGF system displaying altered expression at the mRNA level between cholesteatoma and reference tissue. Furthermore, the receptors HER2 and HER4 were investigated because of their involvement in EPI-mediated signaling. HER1 and HER2 showed similar expression patterns and were located as a membrane staining on the surface of keratinocytes in the stratum basale and in the lower part of the stratum spinosum in both cholesteatoma and reference tissue (Figure 4). HER4 showed a

A

B

C

D

Figure 5. Localization of HER4 in cholesteatoma tissue (stem, middle ear, and mastoid) and skin from the external ear canal (reference tissue) identified by immunohistochemistry. HER4 was only expressed in cholesteatoma tissue. It was localized to the cell surface in the stratum basale and in the lower part of the stratum spinosum. The staining intensity was the same for all three biopsy sites. (A) Skin from the bony part of the external ear canal (reference tissue) (143); (B) cholesteatoma from the stem (228); (C) cholesteatoma from the middle ear (228); and (D) cholesteatoma from the mastoid (143).

similar expression in cholesteatoma to HER1 and HER2, but the receptor was not expressed in the reference tissue (Figure 5). The ligands EPI, AR, and EGF could not be detected with the antibodies applied.

Discussion To our knowledge, our data represent the first systematic investigation of the mRNA expression of all four HER receptors and their activating ligands in cholesteatoma tissue. Interestingly, immunohistochemistry showed that HER1, HER2, and surprisingly also HER4, are expressed in cholesteatoma tissue whereas only HER1 and HER2 are present in EAC skin. This finding allows signaling through both HER1 and HER4 and dimers formed with these receptors in cholesteatoma tissue. Analysis of mRNA revealed a down-regulation of HER1 in cholesteatoma (2.5-fold, stem; 3.1-fold, middle ear; and 4.0-fold, mastoid) but despite this down-regulation the up-regulation of EPI was higher (4.8-fold, stem; 4.2-fold, middle ear; and 3.0-fold, mastoid) and an up-regulation was also found for AR (2.1-fold), suggesting that the net effect is increased HER1 signaling in cholesteatoma tissue. Due to the limited amount of cholesteatoma tissue available, investigation of the HER1 phosphorylation was not possible. EPI is a very potent activator of mitogenesis and unique among EGF ligands, because it is the ligand expressing the broadest specificity by stimulating homodimers of both HER1 and HER4, as well as all possible ligand-stimulatable HER complexes [8]. EPI and AR are the most important growth factors for human keratinocyte proliferation. The ligands are produced by keratinocytes and act as autocrine growth factors. Furthermore, they interact with each other via a cross-induction mechanism, where members of the EGF system induce the expression of other members, resulting in a strong proliferative response [9]. In addition, EPI and AR are up-regulated in psoriasis, a condition proposed to show resemblance to cholesteatoma. Our results may suggest a role for components of the EGF system in the growth and development of human middle ear cholesteatoma and support the theory of an increased proliferative ability of the epithelium, as originally proposed by Anniko and Mendel in 1981 [1]. Furthermore, the increased expression of growth factors in the stem (EPI and AR) may indicate that this part of the cholesteatoma possesses an increased proliferative ability, which is in accordance with the results reported by Cheshire et al. [10]. This finding is further supported by Macias et al.

Acta Otolaryngol Downloaded from informahealthcare.com by Nyu Medical Center on 05/27/15 For personal use only.

EGF-system in cholesteatoma [11], who report AR to be involved in the onset of cholesteatoma pathology. We found no significant associations between the mRNA expressions of the EGF system age, sex, and the origin, type, and extent of the cholesteatoma, which is in accordance with the results presented by Hoppe [12]. Only selected parts of the EGF system have previously been investigated in cholesteatoma tissue and studies show conflicting data. A single study has been published describing the expression of EPI in cholesteatoma. Yoshikawa et al. [13] analyzed gene expression profiles in fibroblasts cultured from cholesteatoma samples. Interestingly, they found an overexpression of EPI. Using immunohistochemistry, Buija et al. [14] found overexpression of HER1 in cholesteatoma samples compared with EAC skin. The results were confirmed using ELISA and in situ hybridization [15]. Likewise, Ergün et al. [16] reported an increased expression of HER1 and TGF-a in cholesteatoma compared with retroauricular skin. Kojima et al. [17] used in situ hybridization to demonstrate a strong signal of mRNA expression of HER1 in all cell layers of cholesteatoma, whereas the expression was restricted to the basal cell layer of EAC skin. On the contrary, Hsu et al. [18] found no difference in the HER1 distribution between cholesteatoma and retroauricular skin samples using immunohistochemistry. Several studies describe an overexpression of TGF-a [19] and HER2 [20] in cholesteatoma tissue. No reports exist describing the expression of HB-EGF, HER3, and HER4. The discrepancy between our results and the results presented in the literature could be explained by various factors such as the application of different methods, the biopsy site, the heterogeneity of cholesteatoma, and the choice of reference tissue. The majority of studies presented in the literature measured protein expression using immunohistochemistry or in situ hybridization; thus many factors may influence the translation of mRNA to protein. Therefore, caution should be used when comparing protein and mRNA measurements. Immunohistochemistry and in situ hybridization are both semi-quantitative methods, whereas Q-PCR is highly sensitive and specific and a true quantitative method. Cholesteatoma tissue is extremely heterogeneous, consisting of a sparse amount of keratinocytes comprising the epithelium resting on a connective tissue and a relatively large amount of cellular debris and keratin. Additionally, the epithelium varies in thickness. As the receptors are located in the epithelium, quantification using immunohistochemical methods and in situ hybridization will be highly dependent on where in a cholesteatoma the section is

133

positioned. On the other hand mRNA results may be diluted because the biopsies consisted of both epithelium and connective tissue. The choice of reference tissue may also influence the results and previously EAC skin and retroauricular skin have been used randomly. Since cholesteatoma most likely originates from either EAC skin or the tympanic membrane, EAC skin seems to be the appropriate reference material and was used in this study. Our results demonstrate that the EGF system is deregulated in cholesteatoma. It is most remarkable that the potent ligands EPI and AR are up-regulated in cholesteatoma and particularly that these ligands potentially will be able to signal through both HER1 and HER4 in cholesteatoma tissue, as both receptor proteins were detected, as opposed to the reference tissue where only HER1 was identified. Additional studies are needed to demonstrate whether these ligands and receptors are involved in cholesteatoma growth. Also, our findings indicate that the stem of cholesteatoma expresses the highest levels of potential growth-promoting ligands and receptors. Thus we speculate that the stem might possess an increased proliferative capacity, which implies that complete removal of the stem during middle ear surgery is of utmost importance.

Acknowledgments The authors are grateful to the patients who donated tissue samples to the present study. We wish to thank Lars Vendelbo Johansen MD DMSc and Kjell Tveterås MD for helping with the sampling procedure. Expert technical assistance provided by laboratory technicians Birgit Mortensen, Birgit Westh Mortensen, Lene Dabelstein, Alice Willemoes, and Jette Shousboe is greatly appreciated. Professor Michael Væth MSci PhD is thanked for help with statistical analysis. We thank the Oticon Foundation, the Oda Pedersen Research Foundation, and the Institute of Clinical Medicine for financial support. Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

References [1] Anniko M, Mendel L. Cholesteatoma. A clinical and morphological analysis. Acta Otolaryngol 1981;91:275–83. [2] Kim HJ, Tinling SP, Chole RA. Increased proliferation and migration of epithelium in advancing experimental cholesteatomas. Otol Neurotol 2002;23:840–4.

Acta Otolaryngol Downloaded from informahealthcare.com by Nyu Medical Center on 05/27/15 For personal use only.

134

M. B. Thorup et al.

[3] Bujia J, Schilling V, Holly A. Hyperproliferation-associated keratin expression in human middle ear cholesteatoma. Acta Otolaryngol 1993;113:364–8. [4] Olszewska E, Wagner M, Bernal-Sprekelsen M, Ebmeyer J, Dazert S, Hilmann H. Ethiopathogenesis of cholesteatoma. Eur Arch Otorhinolaryngol 2003;261:6–24. [5] Shirakata Y, Kishimoto J, Tokumaru S, Yamasaki K, Hanakawa Y, Tohyama M, et al. Epiregulin, a member of the EGF family, is over-expressed in psoriatic epidermis. J Dermatol Sci 2007;45:69–72. [6] Yarden Y, Sliwkowski MX. Untangling the ErbB signalling network. Nat Rev Mol Cell Biol 2001;2:127–37. [7] Citri A, Skaria KB, Yarden Y. The deaf and the dumb: the biology of ErbB-2 and ErbB-3. Exp Cell Res 2003;54–65. [8] Shelly M, Pinkas-Kramarski R, Guarino BC, Waterman H, Wang LM, Lyass L, et al. Epiregulin is a potent pan-ErbB ligand that preferentially activates heterodimeric receptor complexes. J Biol Chem 1998;273:10496–505. [9] Shirakata Y, Komurasaki T, Toyoda H, Hanakawa Y, Yamasaki K, Tokumaru S, et al. Epiregulin, a novel member oftheepidermalgrowthfactorfamily,isanautocrinegrowthfactor in normal human keratinocytes. J Biol Chem 2000;275:5748–53. [10] Cheshire IM, Blight A, Proops DW. An in vitro growth study on cholesteatoma and normal skin. Clin Otolaryngol Allied Sci 1995;20:453–60. [11] Macias MP, Gerkin RD, Macias JD. Increased amphiregulin expression as a biomarker of cholesteatoma activity. Laryngoscope 2010;120:2258–63. [12] Hoppe F. Proliferation behavior of cholesteatoma. HNO 1995;43:710–15.

[13] Yoshikawa M, Kojima H, Wada K, Tsukidate T, Okada N, Saito H, et al. Identification of specific gene expression profiles in fibroblasts derived from middle ear cholesteatoma. Arch Otolaryngol Head Neck Surg 2006; 132:734–42. [14] Bujia J, Holly A, Schilling V, Negri B, Pitzke P, Schulz P. Aberrant expression of epidermal growth factor receptor in aural cholesteatoma. Laryngoscope 1993;103:326–9. [15] Bujia J, Kim C, Holly A, Sudhoff H, Ostos P, Kastenbauer E. Epidermal growth factor receptor (EGFR) in human middle ear cholesteatoma: an analysis of protein production and gene expression. Am J Otol 1996; 17:203–6. [16] Ergün S, Zheng X, Carlsöö B. Expression of transforming growth factor alpha and epidermal growth factor receptor in middle ear cholesteatoma. Am J Otol 1996;17:393–6. [17] Kojima H, Shiwa M, Kamide Y, Moriyama H. Expression and localization of mRNA for epidermal growth factor and epidermal growth factor receptor in human cholesteatoma. Acta Otolaryngol 1994;114:423–9. [18] Hsu YC, Ho KY, Chai CY, Lee KW, Wang LF, Wu SC, et al. Expression of epidermal growth factor receptor in human middle ear cholesteatoma. Kaohsiung J Med Sci 2003;19: 497–502. [19] Shiwa M, Kojima H, Moriyama H. Expression of transforming growth factor-alpha (TGF-alpha) in cholesteatoma. J Laryngol Otol 1998;112:750–4. [20] Sakamoto T, Kondo K, Yamasoba T, Suzuki M, Sugasawa M, Kaga K. Overexpression of ErbB-2 protein in human middle ear cholesteatomas. Laryngoscope 2004;114:1988–91.

Expression of the epidermal growth factor system in human middle ear cholesteatoma.

The detection of the HER4 receptor in 50% of cholesteatomas but never in the reference tissue, and the increased expression of its activating ligand E...
874KB Sizes 0 Downloads 0 Views