HB-EGF was prepared from U-937 cell conditioned medium as previously described (4,7). EGF was obtained from R&D Systems (Minneapolis, MN, USA). Tyrphostin AG1478, a specific inhibitor of EGF receptor tyrosine kinase, was obtained from Calbiochem (La Jolla, CA, USA). RPMI-1640 was purchased from Nissui Pharmaceutical Co., Ltd., (Tokyo, Japan), fetal bovine serum (FBS) from Dainippon Pharmaceutical Co., Ltd., (Osaka, Japan) and trypsin-EDTA solution and penicillin-streptomycin solution from Gibco-BRL (Gaithersburg, MD, USA). The other chemicals and reagents are described below.

Human thyroid carcinoma cell lines, 8305C and SW579, were obtained from the Japanese Collection of Research Bioresources (JCRB) (HSRRB; Health Science Research Resources Bank, Osaka, Japan) and ATCC (Rockville, MD, USA), respectively. 8305C was derived from an undifferentiated thyroid carcinoma (25) and SW579 from a squamous cell carcinoma of thyroid (26). Cells were maintained in continuous culture at 37°C in a 5% CO₂ humidified atmosphere. Cells were grown in RPMI-1640 medium supplemented with 10% heat-inactivated FBS. Penicillin (100 U/ml) and streptomycin (100 μg/ml) were added to the media.

Thyroid tissues specimens were obtained from the patients undergoing thyroid surgery at the Nara Medical University Hospital (Kashihara, Japan). The present study was approved by the Ethics Committee of the Nara Medical University School of Medicine. Written informed consent for this study was obtained from each patient. The median age of the patients was 59.6±10.6 years (range, 29–81 years). All samples were prepared from the surgical specimens immediately after their removal from the patients and fixed in 10% buffered formalin. The histology was examined by one of the authors (M.N.), and the samples were classified according to the WHO criteria (27). For immunohistochemistry, 33 samples from 24 patients were obtained and included the following cases: 9 normal thyroids, 2 hyperplasias, 2 adenomatous goiters, 4 follicular adenomas, 3 follicular carcinomas, 11 papillary carcinomas and 2 undifferentiated carcinomas.

Cell proliferation was determined by the Cell Counting Kit-8 (Dojindo, Kumamoto, Japan). 8305C cells were grown overnight in RPMI-1640 medium with 10% heat-inactivated FBS onto 96-well plates (5,000 cells/well) at 37°C in a 5% humidified atmosphere. After washing with phosphate-buffered saline (PBS), the cells were incubated in 100 μl/well of serum-free RPMI-1640 medium with 0.1% bovine serum albumin (Fraction V; Sigma). 2-(2-methoxy-4-nitrophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H tetrazolium, monosodium salt (WST-8) was added to the cells (0.5 mM/well), after 48 h of the treatment with varying concentrations of HB-EGF. The absorbance of each well was measured at 455 nm with a reference wavelength at 650 nm with MTP-32 microplate reader (Corona Electric Co., Ltd., Ibaragi, Japan). A strong correlation was confirmed between the cell proliferation by this assay and those as measured by counting the number of the cells (28).

Cell migration was evaluated using a modified Boyden chamber assay (24,29,30). Eight-micron Nucleopore polyvinylpyrrolidine-free polycarbonate filters (Cambridge, MA, USA) were coated with 10.0 μg/ml fibronectin (Iwaki, Chiba, Japan) in PBS for 30 min at room temperature and allowed to air dry. The filter was placed over a 48-well chamber (NeuroProbe, Cabin John, MD, USA) containing varying concentrations of HB-EGF in serum-free RPMI-1640 medium with 0.1% BSA in the lower chamber. After trypsinization, 10,000 cells in 50 μl of serum-free medium were added to the wells in the upper chamber. In the checkerboard assay, varying concentrations of HB-EGF were also added to the upper chamber wells. The chamber was then placed in a 37°C with 5.0% CO₂ humidified incubator for 3 h. Next, the upper surface of the filter was scraped to remove non-migratory cells. The filter was subsequently fixed in 10% buffered formalin for 10 min, washed with PBS and stained with hematoxylin for 10 min. Total cell number per well of the lower surface were counted visually as an index of the cell migration. Inhibition assays were conducted in a similar manner with the addition of tyrphostin AG1478 to the upper and lower chambers. The cells were incubated in serum-free medium with tyrphostin AG1478 for 1 h prior to placement in the chamber.

Cell migration was also assessed by a modified in vitro wound assay (31). Cells were plated in complete medium (serum-free RPMI-1640 medium with 0.1% BSA) on 6-well plates. Initial plating was adjusted to yield subconfluent monolayers at the same cell density after 24 h. The monolayers were then wounded by scratching a line with a plastic scriber, and after washing with PBS, were incubated for the indicated time in the complete medium. The experiment was terminated by fixing the cells, followed by staining with hematoxylin. The distance between the advancing cells on both sides in the controls was compared with that in the presence of HB-EGF and the migratory activity was quantified by counting the cells that had migrated into the cell-free space on photographic enlargements (31–33).

Immunohistochemical study for HB-EGF, HER1 and HER4 was performed using the avidin-biotin-complex (ABC) method for 9 normal thyroid tissues, 2 hyperplasias, 2 adenomatous goiters, 4 follicular adenomas, 3 follicular carcinomas, 11 papillary carcinomas and 2 undifferentiated carcinomas. Anti-HB-EGF antibody, H-1 antibody, which was generated to synthetic peptides located in cytoplasmic domains, and anti-HER4 polyclonal antibody were established by our coworker (12,19), and used at the concentration of 1:500 and 1:200, respectively. Anti-HER1 polyclonal antibody was purchased from Upstate Biotechnology, Inc., (Lake Placid, NY, USA) and applied at 1:100. Slices (4 μm) of tissue section were deparaffinized and endogenous peroxidase activity was blocked with 0.3% hydrogen peroxide and 0.1% sodium azide in distilled water for 15 min. For immunohistochemistry of HER1, we performed antigen retrieval by incubating the sections with 0.03 mol/l citrate buffer (pH 6.0) and heated to 121°C for 20 min in pressure cooker. After three rinses in PBS pH 7.2 PBS, 10% bovine serum albumin (Wako, Osaka, Japan) was applied for 10 min to block the non-specific reaction. Sections were incubated with the primary antibody for 60 min at room temperature. After rinsing in PBS, they were treated with biotinylated rabbit anti-sheep IgG (Vector Laboratories, Burlingame, CA, USA) at the concentration of 1:200 for anti-HER1 antibody or biotinylated anti-rabbit IgG (Nichirei, Tokyo, Japan) at the concentration of 1:1 for anti-HB-EGF and anti-HER4 antibodies for 15 min. Again after rinsing in PBS, the sections reacted with the ABC (Dako, Copenhagen, Denmark) at the concentration of 1:300 for 15 min. The peroxidase reaction was visualized by incubating the sections with 0.02% 3,3′-diaminobenzidine tetrahydrochloride in 0.05 M Tris buffer (pH 7.6) with 0.01% hydrogen peroxide. The sections were counterstained with hematoxylin. Sections for negative control were prepared by using normal mouse serum instead of primary antibody.

We classified the results into four groups by positive cell rate: (−), 0–5% of the positive cells; (+), 5–50% of positive; (++), 50–75% of positive; (+++), 75–100% of positive.

Immunofluorescence study of the transmembrane form of HB-EGF (proHB-EGF) proteins was performed with indirect immunofluorescence techniques for 8305C cells. Cells were washed with PBS and fixed with 4% paraformaldehyde. After washing in PBS, the cells were incubated with anti-HB-EGF antibody, H-1 antibody, for 30 min at room temperature. After rinsing in PBS, they were stained with fluorescein isothiocyanate washed in PBS. Cells were photographed using a fluorescence microscope (Olympus, Tokyo, Japan).

Total cellular RNA from culture cells was extracted by the acid guanidium-phenol-chloroform technique using the TRIzol (Gibco-BRL) (34). The total RNA was resuspended in diethylpyrocarbonate-treated water and stored at −80°C until use.

RT-PCR was performed as previously described (35). Single-strand cDNA was generated from 10 μg of total cellular RNA in a 25-μl reaction mixture containing 2.5 μM oligo (dT) 18-primer, 5 mM MgCl₂ 10 mM Tris-HCl, 50 mM KCl (pH 8.3), 1 mM d-NTP mixture, 1 U/μl RNase inhibitor and 0.25 U/μl AMV reverse transcriptase (Takara, Kyoto, Japan) for 1 h at 42°C. The reaction was terminated by inactivating the transcriptase at 65°C for 10 min. A 1-μl aliquot of this solution was removed for subsequent first-round PCR by adding each sample to 100 μl of a solution containing 2.5 mM MgCl₂, 10 mM Tris-HCl, 50 mM KCl (pH 8.3), 200 μM dATP, 200 μM dCTP, 200 μM dGTP, 200 μM dTTP, 100 pmol of each of primer A (5′-TCCTCCAAGCCACAAGCACT-3′) and B (5′-AGAAGCCCCACGATGACCAG-3′) for HB-EGF, and 2.5 units of Taq polymerase. The initial step was 94°C for 2 min for denaturing cDNA, and then PCR was carried out for 30 cycles (30 sec at 94°C, 1 min at 55 and 1 min at 72°C). Negative controls were performed without adding the template cDNA. PCR products (10 μl) were stained with ethidium bromide and analyzed by 1.5% agarose gel electrophoresis. To eliminate degraded RNA, amplification of β-actin was also performed as previously described (36).

DT sensitivity was performed for 8305C cells, since proHB-EGF is uniquely the receptor for DT (16). Approximately 10,000 cells/well were plated in a 24-well plate and incubated for 16 h. After washing each well with cold PBS, 0.5 ml/well of Ham’s F12 medium was added and the cells were exposed to DT (3.3–1,000 ng/ml) for 5 h. Subsequently, 10 μl/well of ³H-Leu (100 μCi/ml) was added to each well and the cells were incubated for 1 h. The cells were harvested with trypsin and the extent of radiolabel incorporation was measured by beta-counter (Walla, Turku, Finland).

EGF and TGF-α have been demonstrated to be potent mitogenic factors for thyroid carcinoma cells (23,37–39). To determine whether the cell growth of thyroid cancer 8305C and SW579 cells is modulated by HB-EGF, we studied the effects of HB-EGF on cell proliferation. HB-EGF enhanced the growth of both cell lines in a dose-dependent manner (Fig. 1). In 8305C and SW579 cells, half-maximal growth stimulation occurred at 15 and 1 ng/ml, respectively.

It has previously been reported that HB-EGF is a potent chemotactic factor as well as a powerful mitogenic factor for SMC (12). To test whether the cell migration of 8305C and SW579 cells is modulated by HB-EGF, a wound assay and a modified Boyden chamber assay were performed.

In a wound assay, wounds of 1 mm width were made in subconfluent monolayers of 8305C cells on 10 μg/ml fibronectin-coated plates, and the cells were allowed to migrate into the cell-free area over a 24-h period (Fig. 2A). Cell migration in this assay was quantified by counting the cells that had advanced into the wounded cell-free area from a number of randomly selected 1-mm segments of the initial edge. As shown in Fig. 2B, at 6 and 12 h after the wounding, no significant differences were observed in the cell migration between the presence and the absence of HB-EGF. However, at 18 and 24 h, 10 ng/ml HB-EGF significantly stimulated the cell migration as compared to HB-EGF-free condition (P<0.05). These results indicate that HB-EGF enhances the cell migration of 8305C cells in a time-dependent manner.

In a modified Boyden chamber assay, HB-EGF induced chemotaxis of 8305C and SW579 cells at the concentrations with half-maximal stimulation of 1 and 0.3 ng/ml, respectively (Fig. 2C). Furthermore, the checkerboard assay, in which varying concentrations of HB-EGF were placed in the upper and lower wells of Boyden chamber apparatus, was performed for 8305C cells, to verify whether HB-EGF stimulates chemotaxis or chemokinesis. Cell migration to HB-EGF was predominantly consistent with chemotactic response as seen in Fig. 2D. Furthermore, when HB-EGF was present only in the upper wells, the increase in cell migration was minimal. These results indicate that HB-EGF stimulates directional chemotaxis rather than significant stimulation of random cell motility.

HB-EGF has been reported to activate HER1 tyrosine phosphorylation and then to stimulate proliferation and chemotaxis in cells expressing HER1 (21). To verify whether HB-EGF stimulates HER1-mediated chemotaxis in cancer cells, we tested the inhibitory effect of tyrphostin AG1478 in HB-EGF-induced chemotaxis. In a modified Boyden chamber chemotaxis assay, the chemotactic effects of HB-EGF for 8305C and SW579 cells were markedly inhibited by the pretreatment with 100 nM tyrphostin AG1478 for 60 min prior to this assay (Fig. 2C). Furthermore, the migration with the pretreatment of tyrphostin AG1478 was more suppressed even without HB-EGF (Fig. 2C). In a wound assay, the migration was also inhibited by tyrphostin AG1478 (data not shown). These data suggest that HB-EGF-induced chemotactic effects could be mediated by HER1.

To verify whether HB-EGF mRNA and proHB-EGF protein are expressed in 8305C cells, RT-PCR and immunofluorescence study were performed. HB-EGF mRNA expression in 8305C cells was detected with RT-PCR (Fig. 3A), and the protein expression was also observed in the cell surface by immunofluorescence staining with FITC-conjugated anti-rabbit immunoglobulin (Fig. 3B). Moreover, the functional presence of proHB-EGF protein was also confirmed by demonstration of specific sensitivity to DT (Fig. 3C), indicating that proHB-EGF can be available for DT-binding and toxin internalization as previously reported in prostate cancer cell line LNCaP cells (40). In addition, HER1 protein and mRNA expression in 8305C and SW579 cells (data not shown) were detected with both immunofluorescence study using FACSCalibur (Becton-Dickinson, San Jose, CA, USA) and RT-PCR, but HER4 was not.

The results of HB-EGF, HER1 and HER4 in thyroid tissues are presented in Table I. HB-EGF and HER4 staining in differentiated thyroid carcinoma tissues, such as papillary carcinomas, were localized in cytoplasm and/or cell membrane of the cells, whereas HER1 immunoreactivity was observed predominantly in cell membrane as shown Fig. 4. The intensity of HB-EGF and HER4 immunostaining in carcinomas was stronger and the number of positive cells was higher than in normal tissues. In some undifferentiated thyroid carcinoma tissues, however, HB-EGF staining was negative, although HER4 staining was strongly positive. On the other hand, HER1 was expressed widely in most malignant and benign tissues of the thyroid (Table I).