Keratinocyte growth factor receptor (KGF-R) in cholesteatoma tissue

Conclusion. Distribution of the receptor for epidermal growth factor (EGF-R) and of the receptor for the keratinocyte growth factor (KGF-R) in cholesteatoma was found to differ in analogy with other epithelial tissues and accordingly to epidermal differentiation and intensity of paracrine stimulation. Moreover, both EGF-R and KGF-R expression was increased, suggesting a fair correlation with aggressiveness and recurrence rate of this pathology. Objectives. To obtain information on the biological behaviour of cholesteatoma by assessing the expression and localization of EGF-R and KGF-R and correlating their tissue distribution with that of cytokeratins as a marker of differentiation. Materials and methods. Cholesteatoma tissue was taken during tympanoplasty surgery and processed for indirect immunofluorescence. Murine monoclonal antibodies were tested for the different growth factor receptors and pancytokeratins analysed. Fluorescence intensity signal was measured on randomly captured digital images, using FISH 2000/HI software, with a pseudocolours generation module. Results. EGF-R was mostly expressed at the level of keratinocytes of the basal layer, while KGF-R signal was mainly distributed on the spinous and granular suprabasal layers that were also highly positive for cytokeratins. Significant correlation between the immunofluorescence signals was found for KGF-R and cytokeratins only, demonstrating that KGF-R expression is increased in more differentiated areas of the cholesteatoma tissue, while EGF-R is associated with proliferative and migratory portions of the lesion.

Keywords: growth factors; immunofluorescence; KGF-R; EGF-R; cytokeratins; Cholesteatoma

Cholesteatoma is a temporal bone pathology that is characterized by multifactorial aetiopathogenesis and clinical expression. It may originate as a congenital form, affecting the middle ear without eardrum perforation as well as the rest of petrous pyramid often involving labyrinthine structures and facial nerve or as an acquired condition, following a retraction pocket at epitympanic level or a marginal eardrum perforation. Another peculiar characteristic of cholesteatoma is represented by the high postoperative recurrency rate, as recurrent or residual cholesteatoma. In this regard, it has also been observed that the recurrence rate is much higher in children than in adults.

Although benign in nature, the clinical behaviour of cholesteatoma often implies its progressive growth, with involvement of the neighbouring middle/inner ear structures (semicircular canals, facial nerve, meningeal wrapping), which may become dehiscent and give rise to life-threatening complications.

Due to these complex clinical manifestations, it has been logical to assume that different growth factors may be involved in favouring the active proliferation of the epithelial cells in cholesteatoma tissue. In particular, in recent decades, considerable interest has been devoted to obtaining evidence of a possible role played by the keratinocyte growth factor (KGF) [1], [2], [3], by cytokines and cytokeratins [4], [5], and by the epidermal growth factor (EGF) [6], [7], [8]. Among them, KGF and its receptors have been particularly investigated by detailed immunohistochemical studies, and their presence has been detected in a large percentage of cholesteatoma patients with recurrent disease, thus suggesting their involvement in epithelial hyperproliferation [2], [3].

This study was designed to shed some light on the expression and localization of KGF-R, EGF-R and cytokeratins within the cholesteatoma tissue, giving some insight as to their possible role in hyperproliferation of cholesteatoma tissue.

Fresh cholesteatoma tissue was taken during 35 consecutive surgical procedures, including closed tympanoplasty (10 cases), open tympanomastoidectomy (8 cases), subtotal petrosectomy (6 cases), planned II stage procedure (8 cases) and congenital middle ear cholesteatoma (3 cases).

The tissue was prepared for immunofluorescence processing, being placed under the stereomicroscope for optimal orientation before OCT-N2O freezing, and 4 µ thick cryosections were obtained. Post-fixation was carried out with 1:1 methanol and acetone for 10 min at 4°C, followed by incubation with murine monoclonal antibody anti-pancytokeratins (1:50) (DAKO, Glostrup, Denmark); murine monoclonal antibody anti-EGFR (1:50) (DAKO); and murine monoclonal antibody anti-KGF-R (1:50) (DAKO). After washing with phosphate-buffered saline (PBS), incubation with rodamine secondary antibody against monoclonal (Texas Red, 1:100, Jackson Immuno Research Laboratories Inc.) was carried out. For nuclei, dark incubation for 5 min with DAPI (4′,6-diamidino-2-phenylindole dehydrochloride) diluted 1:10 000 in PBS was carried out.

A Zeiss Axiovert fluorescence microscope equipped for optical sections (APOTOME®) and digital camera (Zeiss, Oberkochen, Germany) were used for the analysis of fluorescence signal distribution. KGF-R values at the basal and suprabasal layers were individually measured in nine of the whole set of analysed specimens. KGF-R suprabasal/basal ratio values were also assessed. Fluorescence values were statistically analysed by Pearson correlation test.

Analysis of fluorescence signal intensity was carried on randomly captured digital images, using FISH 2000/HI software (Delta Sistemi, Rome, Italy), with a pseudocolours generation module.

Twenty-six of the 35 samples collected were not available for a complete analysis, as they mainly contained corneal debris, which would have shown absence of the different growth factor receptors. The analysis of the results was hence performed on specimens taken from nine patients in whom different surgerical procedures were carried out, such as closed tympanoplasty (five cases), open tympanomastoidectomy (two cases) and subtotal petrosectomy (one case).

Immunofluorescence analysis of EGF-R showed that the receptor-specific signal was mainly expressed on the basal layer of cholesteatoma epithelium (Figure 1). Quantitative analysis of the immunofluorescence intensity revealed that the mean distribution of the signal for EGF-R appeared more intense at the basal layer, with values 2.4 times greater than in the spinous layer, and 4.1 times greater than in the granular layer (Figure 1).

Graph: Figure 1. Immunofluorescence: EGF-R distribution in cholesteatoma epithelium – signal appears to be more intense in the basal layer. The graph shows the different distribution of the signal of EGF-R in the different layers of cholesteatoma matrix. The basal layer appears to have a greater concentration of EGF-R than the suprabasal layers. X-axis, arbitrary unit values of fluorescence.

Immunofluorescence analysis of KGF-R showed that the receptor staining was mostly distributed on the suprabasal layers, with prevalence on the spinous layer (Figure 2). The quantitative analysis of the signal intensity showed that mean distribution of KGF-R was found more intensely in the suprabasal layers, with values 2.5 times greater in the spinous layer and 2.9 times greater in the granular layer than in the basal layer (Figure 2). Single values for basal, suprabasal and suprabasal/basal ratio are shown in Table I.

Graph: Figure 2. Immunofluorescence: KGF-R distribution in cholesteatoma epithelium – signal appears to be more intense in the suprabasal layers. The graph shows the different distribution of the signal of KGF-R in the different layers of cholesteatoma matrix. The suprabasal layers appear to have a greater concentration of KGF-R than the basal layer. X-axis, arbitrary unit values of fluorescence.

Table I.  Single KGF-R distribution values found in separate specimens at the basal and suprabasal layer levels, and suprabasal/basal ratio.

1289 CI, confidence interval.

Staining of pancytokeratins was then performed to evaluate the extent of epithelial differentiation. As expected, the cytokeratin signal was clearly evident all along the epithelial layers, but was higher in intensity over the suprabasal layers, giving rise to different correlation patterns with EGF-R and KGF-R distribution (Figure 3). EGF-R expression was found to be inversely correlated with pancytokeratin expression (r= − 0.73). A greater inverse correlation (r = −0.94) was found in areas with less differentiated epithelium, characterized by elongated basal keratinocytes, one-two layered suprabasal keratinocytes and poor corneal production. In these latter areas, EGF-R expression on the basal layer was 1.6 greater than on pluristratified areas. In contrast, KGF-R expression appeared to be directly correlated with cytokeratin expression (r= + 0.68) and keratinocyte differentiation, as they were both shown to be mainly expressed in the suprabasal layers.

Graph: Figure 3. Immunofluorescence: pancytokeratins signal distribution is prevalently located at the suprabasal layer.

The aggressive pattern that usually characterizes the clinical behaviour of temporal bone cholesteatoma has generated several studies that aimed to elucidate a possible role played by different growth factors. In this regard, KGF has been considered an important marker for the identification of a possible dysregulatory proliferative activity of cholesteatoma matrix. KGF is a member of the fibroblast growth factor family and is also known as FGF-7 [9]. It is synthesized and secreted by stromal cells only and is considered to be a paracrine effector of epithelial cell growth [9], and also a significant role in stimulation of epithelial cells proliferation and differentiation [10] in cholesteatoma tissue [1], [2].

Assuming the importance that could be played by involvement of different growth factors, it could be hypothesized that the dysregulatory hyperproliferative behaviour of the epithelium could also depend upon their different distribution pattern in cholesteatoma tissue. Both in situ hybridization [1] and reverse transcription polymerase chain reaction (PCR) [2] techniques have previously allowed identification of the expression of KGF and KGF-R in cholesteatoma tissue. More recently, Yamamoto-Fukuda et al. [3] carried out an immunohistochemical evaluation of paraffin-embedded specimens, also trying to correlate laboratory data with the clinical recurrence of the disease.

A rather consistent pattern of KGF-R distribution was obtained in the present study. Overall mean values ranged from 16.2 to 24.6, but suprabasal layer distribution was statistically higher than basal distribution. These values have also generated different suprabasal versus basal layer ratios. Taking into account the paucity of samples in the present study, it is difficult to make any statement on the importance that it could have for defining a more or less aggressive pattern of cholesteatoma tissue.

As far as EGF-R expression is concerned, similar to results previously described by Kojima et al. [6] by using a non-radioactive in situ hybridization method, and in slight contrast with results reported by Li et al. [7] by an immunostaining method, the present study enabled detection of EGF-R expression mainly distributed in the basal layer of cholesteatoma tissue. More recently, Chi et al. [8] supported the hypothesis that the presence of EGF in cholesteatoma tissue was devoid of any specific role in proliferation of epithelial cells.

In the present study it was not possible to reproduce the results previously shown by Olszewska and Sudhoff [5] in terms of a specific distribution pattern of cytokeratins in cholesteatoma tissue, with prevalence of CK10, CK14 and CK34βE12 subtypes, as anti-pancytokeratin antibodies were used and found to be expressed all along the epithelium, in a positive correlation with KGF-R expression in the suprabasal layers.

It is necessary to perform a gentle and careful dissection of pathological tissue so as not to be forced to discard potentially interesting material, as occurred with our study material.

Collection and analysis of further samples as well as correlation with clinical data (primary or revision surgery, age, congenital or acquired forms) would surely be helpful for a better definition of specific roles played by different growth factors in this ear pathology.