Restoration of BRAK / CXCL14 gene expression by gefitinib is associated with antitumor efficacy of the drug in head and neck squamous cell carcinoma

Abstract

Clinical efficacy of gefitinib (ZD1839, Iressa), which is an inhibitor specific for epidermal growth factor (EGF) receptor tyrosine kinase, has been shown in non‐small‐cell lung carcinoma patients with EGF receptor mutations, so these mutations are useful marker(s) to find a responder for the drug. Recent studies have shown that the EGF receptor gene mutation is rare in squamous cell carcinoma in the esophageal and head and neck regions. We previously reported that the expression of the chemokine BRAK/CXCL14 in head and neck squamous cell carcinoma (HNSCC) cells was down‐regulated by EGF treatment, and that forced expression of BRAK in tumor cells decreased the tumorigenicity of the cells in xenografts. Thus, we investigated the relationship between restoration of BRAK expression by gefitinib and the efficacy of the drug for tumor suppression. We found that EGF down‐regulated BRAK expression through the MEK–extracellular signal regulated kinase pathway and that this down‐regulated expression was restored by gefitinib in vitro. Oral administration of gefitinib significantly (P < 0.001) reduced tumor growth of xenografts of three HNSCC cell lines (HSC‐2, HSC‐3, and HSC‐4), in female athymic nude mice, accompanied by an increase in BRAK expression specifically in tumor tissue. This tumor‐suppressing effect of the drug was not observed in the case of BRAK non‐expressing cells. Furthermore introduction of BRAK shRNA vector reduced both the expression levels of BRAK in HSC‐3 cells and the antitumor efficacy of gefitinib in vivo. Our data showing an inverse relationship between BRAK expression levels in tumor cells and the tumor growth rate indicate that the gefitinib‐induced increase in BRAK expression is beneficial for tumor suppression in vivo. (Cancer Sci 2009)

EGFR signaling plays a crucial role in the aggressive features of human HNSCC.^{( 1 )} Therefore, EGFR has been thought to be a principal therapeutic target for this cancer. Gefitinib (ZD1839, Iressa), which is a TKI for EGFR, has been clinically used for the treatment of terminal patients with NSCLC. Clinical efficacy of gefitinib has been shown in subpopulations of NSCLC patients: adenocarcinoma histology, never‐smoker status, East‐Asian ethnicity, and female gender (see for review^{( 2} , ^{3 )}). Recent studies have shown that these patients have specific mutations in the ATP‐binding pocket of EGFR, which increase binding affinity for gefitinib.^{( 4} , ⁵ , ^{6 )} Although these mutations are thought to be a good predictor for gefitinib sensitivity, it has been shown that 5/59 (8.5%) of NSCLC patients with EGFR mutations do not respond to gefitinib,^{( 7 )} and that 10/90 (11.1%) of NSCLC patients without EGF mutations were diagnosed as having a partial or complete response to gefitinib administration.^{( 8 )} In addition these mutations are rare in HNSCC patients,^{( 9 )} suggesting that other factors are involved in the sensitivity to gefitinib in HNSCC patients. In fact, EGFR gene copy number or downstream effector status has also been thought to be a predictor or biomarkers of gefitinib (see for review^{( 10 )}). Furthermore, Kanazawa et al. ^{( 11 )} have shown that the plasma concentrations of soluble P‐selectin, E‐selectin, soluble interleukin‐2 receptor, and RANTES (regulated upon activation, normal T cell expressed and secreted) in patients with NSCLC increased after treatment with gefitinib. Increases in these factors predict platelet and lymphocyte activation, suggesting that gefitinib has antitumor efficacy through tumor immunity in addition to direct inhibition of tumor cell proliferation.

BRAK, which is a CXC type of chemokine^{( 12 )} and is also known as CXCL14, contributes B‐cell, monocyte, and dendritic cell infiltration into tumor tissues.^{( 13} , ^{14 )} Generally, BRAK mRNA expression is abundant in normal tissues, but it is very low level in certain carcinomas and carcinoma cell lines.^{( 12} , ^{15 )} Earlier we reported that BRAK is an EGF‐suppressed molecule in HNSCC cells, and has the potential for antitumor activity in a xenograft model.^{( 16 )} These observations prompted us to investigate the relationship between BRAK expression levels and gefitinib efficacy. Here, we show that restoration of BRAK expression after administration of gefitinib is strongly associated with antitumor efficacy of gefitinib in HNSCC cell xenografts.

Materials and Methods

Reagents.  Gefitinib (ZD1839, Iressa) was kindly provided by AstraZeneca (Macclesfield, UK). Other reagents and their sources used were as follows: human EGF, gentamicin sulfate, G418 sulfate, HEPES, trypsin, EDTA, Immunostar chemiluminescence assay kit, and Apoptosis in situ Detection kit from Wako (Tokyo, Japan); DMEM, Fungizone, Lipofectamine, SuperScript II reverse transcriptase, and TRIzol from Invitrogen (Carlsbad, CA, USA); PD98059, U0126, FR180204, and LY‐294002 from Calbiochem (La Jolla, CA, USA); FBS from Thermo Electron (Melbourne, Vic., Australia); Ex Taq DNA polymerase from Takara (Otsu, Japan); Brilliant SYBR Green qPCR Master Mix from Stratagene (La Jolla, CA, USA); A mAb against BRAK from Nozaki Monoclonal Antibody (Kanagawa, Japan), against Akt from Upstate (Billerica, MA, USA); Histofine staining kit and mAb against Ki‐67 from Nichirei (Tokyo, Japan); polyclonal antibody against EGFR from Upstate; against pEGFR from Cell Signaling Technology (Boston, MA, USA), against ERK, pERK, pAKT, and biotin‐conjugated swine antirabbit IgG and goat antimouse IgG from Santa Cruz (CA, USA); and N102 blocking reagent from NOF (Tokyo, Japan).

Cells and cell culture.  Head and neck squamous cell carcinoma (HNSCC) cell lines used in this study and their origins were: HSC‐3, HSC‐4, YCU‐T891, YCU‐T862, KCC‐T873, KCC‐TCM903, KCC‐TCM902, KCC‐TCM901, and KCC‐T871 (tongue); YCU‐M862 and KCC‐M871 (oropharynx); YCU‐MS861 and KCC‐MS871 (maxillary sinus); KCC‐L871 and YCU‐L891 (larynx); HSC‐2 and YCU‐OR891 (oral floor); and YCU‐H891 (hypopharynx). HSC‐2, HSC‐3, and HSC‐4 were obtained from the Japanese Collection of Research Bioresources Cell Bank. Other HNSCC cell lines were established in our laboratory and previously described elsewhere.^{( 17} , ¹⁸ , ^{19 )} Cells were cultured with DMEM supplemented with 50 μg/mL gentamicin sulfate, 250 ng/mL fungizone, 12.6 mM HEPES, and 10% FBS at 37°C under 95% air and 5% CO₂ and routinely subcultured by treatment with 0.25% trypsin. Cell numbers were counted with a Coulter Counter (Beckman Coulter, Fullerton, CA, USA).

Polymerase chain reaction (PCR), RT‐PCR, and qPCR analyses.  Total RNA (50 ng), extracted by using TRIzol reagent, was reverse‐transcribed with SuperScript II and amplified with Ex Taq DNA polymerase. For qPCR analysis, we used a Brilliant SYBR Green qPCR Master Mix. Primer sets used were as follows: for RT‐PCR and qPCR of BRAK, 5′‐AAT GAA GCC AAA GTA CCC GC‐3′ (forward) and 5′‐AGT CCT TTG CAC AAG TCT CC‐3′ (reverse), which yielded a 232 bp product; for RT‐PCR of BMAC, 5′‐TGA AGC TCT TCT TAC CAG GG‐3′ (forward) and 5′‐AAG CTC AGA TGT GTG GTG AC‐3′ (reverse), which yielded a 210 bp product; for qPCR of β–actin, 5′‐AAA GAC CTG TAC GCC AAC AC‐3′ (forward) and 5′‐CTC GTC ATA CTC CTG CTT GC‐3′ (reverse), which yielded a 224 bp product; and for RT‐PCR of β‐actin, 5′‐AGC CAT GTA CGT TGC TA‐3′ (forward) and 5′‐AGT CCG CCT AGA AGC A‐3′ (reverse), which yielded a 745 bp product as described previously.^{( 20} , ^{21 )} PCR thermal cycle conditions were as follows: denaturation at 94°C for 30 s, annealing at 58°C for BRAK or 60°C for β‐actin for 30 s, and extension at 72°C for 40 s. For the standard PCR, the products were separated on 1.5% agarose gels and visualized by ethidium bromide staining.

Western blot analysis.  We determined BRAK, pAKT, total Akt, pERK, total ERK, pEGFR, and total EGFR levels by using western blotting as previously reported.^{( 22} , ^{23 )} Cells were lysed with Laemli buffer consisting of 62.5 mM Tris‐HCl (pH 6.8) containing 2% SDS, 8% glycerol, 0.05% bromophenol blue, and 5% 2‐mercaptoethanol. The proteins were electrophoresed in an SDS–10% polyacrylamide gel and transferred onto an Immobilon–P membrane at a constant voltage of 50 volts for 2 h. After blocking with 20% N102 blocking reagent in TBS solution [20 mM Tris‐HCl (pH 7.6) containing 137 mM NaCl and 0.05% Tween‐20], the membrane was incubated overnight at 4°C with appropriate primary antibodies in TBS solution containing 0.05% Tween‐20 and 20% N102 blocking reagent. It was then subsequently incubated with biotin‐conjugated swine antimouse or antirabbit IgG, as the secondary antibody, for 20 min at 37°C. After a 20‐min incubation at 37°C with horseradish peroxidase–conjugated streptavidin, immunoreactive bands were visualized by using an Immunostar chemiluminescence assay kit.

Construction of and transfection with shRNA expressing vector targeting BRAK.  We constructed a 21‐mer shRNA expressing vector targeting BRAK (sh‐BRAK) and its scrambled sequence‐expressing vector (Sh‐Scramble) as a control by using a pSINsi‐hU6 vector (Takara). The insert sequence for BRAK shRNA was 5′‐GCA CCA AGC GCT TCA TCA ACT GTG AAG CCA CAG ATG GGT TGA TGA AGC GCT TGG TGC‐3′ and that for the scramble control shRNA was 5′‐GCC ATA CGC GAC ATA ACC TCT GTG AAG CCA CAG ATG GGA GGT TAT GTC GCG TAT GGC‐3′. The underlined nucleotides indicate the 19‐bp hairpin loop sequence. HSC‐3 cells were transfected with those vectors by use of a Lipofectamine and the stably expressed cells were cloned according to limiting dilution method after pharmacological selection with the antibiotics G418 (400 μg/mL).

Tumor growth in vivo.  Cells (1 × 10⁷ cells/site) were injected subcutaneously into both sides of female BALBc nu/nu mice (Clea Japan, Tokyo, Japan). When the tumor burden was established under the skin 10 days after the injection of tumor cells, tumor‐bearing mice were daily administered oral gefitinib (50 mg/kg) which was suspended in 0.5% Tween 80. Tumor volume was calculated according to the following formula: (a × b ²)/2, where a is the longer and b is the shorter dimension.^{( 24 )}

Immunohistochemical staining.  Tumor tissues were taken under anesthesia, fixed in 4% paraformaldehyde for 24 h, and embedded in paraffin wax. The labeling index of Ki‐67, a cell proliferation maker, was determined by immunohistochemistry using a Histofine staining kit. Apoptosis was determined by using an apoptosis in situ detection kit.

Sequence for exons 19 and 21 of EGFR.  Region at exon 19 and exon 21 of EGFR were amplified for direct sequencing. PCR products were purified by gel extraction using a QIAquick gel extraction kit (Quiageh, Tokyo, Japan). Sequence analysis was done on a CEQ2000 sequencer using a dye terminator cycle sequencing kit (both Beckman Coulter).

Statistical analysis.  The Student’s t‐test was used to assess statistically significant differences between two groups, with P < 0.05 being considered statistically significant.

Results

Gefitinib restored EGF‐induced BRAK mRNA repression through the MEK–ERK pathway.  We have shown that EGF down‐regulated BRAK mRNA expression.^{( 16 )} Moreover, this down regulation was restored by the co‐presence of PD98059, a specific inhibitor for MEK, suggesting that the EGFR–MEK–ERK signaling pathway is involved in EGF‐induced BRAK down‐regulation. In order to examine this possibility, we investigated dose‐dependent effects of several enzyme inhibitors on BRAK mRNA expression. In the presence of 10 ng of EGF, an EGFR inhibitor, gefitinib (Fig. 1a), MEK inhibitors that act with different mechanisms, PD98059l (Fig. 1b) and U0126 (Fig. 1c), and an ERK inhibitor FR180204 (Fig. 1d), dose‐dependently restored expression of BRAK mRNA indicating that the EGFR–MEK–ERK pathway regulates BRAK mRNA expression. Next we examined whether modulation of BRAK mRNA expression by EGF and/or gefitinib is reflected in protein levels and whether gefitinib treatment attenuates the EGF effect of elevating the BRAK protein level. In these experiments 1 μM of gefitinib was employed, because the concentration was the lowest concentration that induced nearly maximum effect (Fig. 1a). Western blot analysis clearly showed EGF‐induced BRAK repression and the restoration of this down‐regulation by gefitinib at the protein level (Fig 2a,b) in concordance with the results of RT‐PCR (Fig. 1a).

Figure 1.

 Dose‐dependent effects of inhibitors for epidermal growth factor receptor (EGFR), MEK, or extracellular signal‐regulated kinase (ERK) on BRAK mRNA expression. Nearly confluent HSC‐3 cells were incubated with EGF (10 ng/mL) in the presence or absence of varying concentrations of an inhibitor for EGFR, gefitinib (a); for MEK, PD98059 (b) or U0126 (c); or for ERK, FR180204 (d). Total RNA was extracted 24 h after EGF stimulation and BRAK mRNA expression was analyzed by standard RT‐PCR. PCR products were visualized by ethidium bromide staining after electrophoresis in agarose gel. Steady state levels of BRAK mRNA were quantified by real‐time quantitative PCR and values of BRAK mRNA/β‐actin in the absence of respective inhibitors were set at 100%. **P < 0.001 (Student’s t‐test) compared with control, where respective inhibitors are absent; values are presented as mean ± SD (n = 3 in all panels).

Figure 2.

 Gefitinib restores epidermal growth factor (EGF)‐induced BRAK repression with a concomitant decrease in the level of phosphorylated EGFR and extracellular signal‐regulated kinase (ERK). Nearly confluent HSC‐3 cells were incubated with or without gefitinib (1 μM) and/or EGF (10 ng/mL). (a) Cells were incubated with EGF for 15 min to detect phosphorylated EGF receptor (pEGFR), EGFR, phosphorylated ERK (pERK), ERK phosphorylated Akt (pAKT), and Akt levels for 24 h to detect the BRAK protein level. Their protein levels were determined by western blotting after treatment with respective antibodies. Relative intensities for BRAK (b) and phosphorylated levels of EGFR (pEGFR, c), ERK (pERK, d), and Akt (pAKT, e), were normalized by β ‐tubulin and their total proteins, respectively. **P < 0.001 (Student’s t‐test); values are presented as mean ± SD (n = 3 in all panels).

It is known that EGFR signaling has two major pathways: Ras–Raf‐1–MEK–ERK signaling and PI3 kinase–Akt signaling.^{( 25 )} Our model also showed that EGF treatment stimulated these two major downstream pathways (Fig. 2a,c–e). Our data showed that the MEK–ERK signaling pathway is involved in EGF‐induced BRAK repression. We also tested whether the PI3 kinase–Akt signaling pathway is responsible for the down‐regulation of BRAK expression by using western blotting. As shown in Figure 3a,c, BRAK mRNA expression was suppressed concomitant with EGF‐induced ERK activation and inhibition of ERK activation by MEK inhibitor PD98059–restored expression of BRAK mRNA. Treatment with PD98059 alone elevated the BRAK mRNA level more than its basal level (Fig. 3a), suggesting that the endogenous activator of the MEK–ERK signaling pathway is also present in the cells. In contrast, although inhibition of PI3 kinase activity by treatment with LY294002 reduced EGF‐induced down‐regulation of phosphorylated Akt (pAKT), LY294002 treatment did not restore EGF suppression of BRAK mRNA expression (Fig. 3b,d). These results indicate that MEK–ERK was a major downstream pathway of EGFR in the down‐regulation of BRAK mRNA expression in HNSCC cells.

Figure 3.

 An MEK inhibitor (PD98059) reverses epidermal growth factor (EGF)‐induced BRAK mRNA repression with a concomitant decrease in the level of phosphorylated ERK but no reversal occurs with a phosphatidylinositol 3‐phosphate (PI3) inhibitor (LY‐294002). Nearly confluent HSC‐3 cells were incubated with or without EGF (10 ng/mL) in the presence or absence of PD98059 (50 μM) or LY‐294002 (25 μM). Cells were lysed 15 min after EGF stimulation for detection of phosphorylated Akt (pAKT) (c) and phosphorylated ERK (pERK) (d) or 24 h after the stimulation for the levels of BRAK mRNA (a,b). Total ERK and Akt and their phosphorylated form were determined by western blotting. Columns show relative intensity for phosphorylated forms against their total protein expression. pERK and pAKT were normalized by total ERK and Akt. (c,d) The steady state level of BRAK mRNA was quantified by real‐time quantitative PCR. β ‐Actin mRNA level was used for the normalization. **P < 0.001 (Student’s t‐test); values are presented as mean ± SD (n = 3 in all panels).

Gefitinib restored EGF‐induced BRAK mRNA down‐regulation in the case of BRAK‐expressing carcinoma cells.  Because a previous study has shown that BRAK expression was lower in clinical specimens and cell lines of HNSCC than in adjacent normal tissues,^{( 14} , ¹⁵ , ^{26 )} we also determined the frequency of BRAK expression in HNSCC cell lines. Fig. 4a showed that eight cell lines out of 18 (44.4%) expressed a detectable amount of BRAK mRNA under serum‐free culture conditions, as determined by RT‐PCR. Such an expression was reduced by EGF treatment and this down‐regulation was restored or the expression level was even elevated by the addition of gefitinib higher than its constitutive basal level under these serum‐free culture conditions (Fig. 4b,c).

Figure 4.

 Gefitinib attenuates the inhibitory effect of epidermal growth factor (EGF) on BRAK mRNA expression in eight out of 18 kinds of HNSCC cell lines. (a) BRAK mRNA expression in 18 kinds of head and neck squamous cell carcinoma (HNSCC) cell lines. Nearly confluent cells were serum‐starved overnight. Total RNA was extracted and BRAK mRNA expression was analyzed by RT‐PCR. (b) Effect of gefitinib and/or EGF on HNSCC cell lines. Nearly confluent cells were serum‐starved overnight and further incubated with or without gefitinib (1 μM) and/or EGF (10 ng/mL). Total RNA was extracted 24 h after EGF stimulation and BRAK mRNA expression was analyzed by standard RT‐PCR. PCR products were visualized by ethidium bromide staining after electrophoresis in agarose gel. (c) Steady state levels of BRAK mRNA were quantified by real‐time quantitative PCR. **P < 0.001 (Student’s t‐test); values are presented as the mean ± SD (n = 3 in panel c).

Gefitinib reduced tumor growth of xenografts accompanied by an increase in BRAK mRNA level.  Previously we showed that an increase in BRAK expression by transfection of HNSCC cells with the vector encoding BRAK dramatically reduced tumorigenicity,^{( 16 )} suggesting that the pharmacological agent that induces BRAK expression might show BRAK‐mediated antitumor activity. Gefitinib, an EGFR TKI, is the most appropriate agent for this purpose. Oral administration of gefitinib (50 mg/kg/day; an amount generally used for animal studies) significantly reduced the size of tumors produced not only by the HSC‐3 (tongue), but also by the HSC‐2 (oral floor) and HSC‐4 (tongue) cell lines (Fig. 5a–c). We further determined whether BRAK expression was induced in the xenograft. As shown in Figure 5e, BRAK mRNA levels dramatically increased after the administration in three cell lines (HSC‐2, 3, 4), whose growth was reduced by the treatment. On the other hand, gefitinib did not increase BRAK mRNA expression in YCU‐H891 cells in vivo (Fig. 5e) as well as in in vitro culture (Fig. 4b), whose growth in vivo was not affected by this drug (Fig. 5d). Interestingly, gefitinib administration did not affect mouse BRAK (BMAC) mRNA levels in host tissues such as lung, liver, brain, and kidney (Fig. 5f).

Figure 5.

 Effects of oral gefitinib on the tumor volume and BRAK mRNA expression. Head and neck squamous cell carcinoma (HNSCC) cells (1 × 10⁷), HSC‐2 (a), HSC‐3 (b), HSC‐4 (c), and YCU‐H891 (d), were subcutaneously injected into both flanks of 10 female athymic nude mice, and allowed to form a tumor burden for 10 days. The mice were daily administered oral gefitinib (50 mg/kg/day). Tumor volume was measured daily. In some mice treated with gefitinib for 4 days, tumor tissues or host organs were taken to extract total RNA. BRAK mRNA expressions in transplants (e) and host organs (f) were determined by RT‐PCR. PCR products were visualized by ethidium bromide staining after electrophoresis in 1.5% agarose gel. Arrows indicate time‐point of administration of gefitinib. *P < 0.05; **P < 0.001 (Student’s t‐test); values are expressed as the mean ± SD (n = 10).

ShRNA‐mediated inhibition of BRAK expression reduced efficacy of gefitinib.  Because gefitinib did not have a growth reduction efficacy for the tumor cells that were unable to have expression of BRAK mRNA induced (Fig. 5d,e), we further investigated whether the gefitinib‐mediated increase in BRAK expression was associated with efficacy of gefitinib by inhibiting BRAK expression of cells. We constructed BRAK shRNA‐expressing vector to knockdown BRAK mRNA expression and established stable transformants from parental HSC‐3 cells: sh‐BRAK No3 and sh‐BRAK No12. The scrambled sequence‐containing vector was used as the control shRNA (Sh‐Scramble). Daily administration of oral gefitinib at a dose of 50 mg/kg/day significantly reduced the tumor volume in Sh‐Scramble introduced xenografts, as well as for parental HSC‐3 cells with a concomitant increase in BRAK mRNA expression (Fig. 6a,b), whereas it did not show any antitumor efficacies for BRAK‐knockdowned cell xenografts (sh‐BRAK No3 and sh‐BRAK No12) (Fig. 6a,b). Immunohistochemical analyses showed that gefitinib reduced the Ki‐67 labeling index in cell clones both of BRAK shRNA and scrambled control transfectants irrespective of levels of BRAK mRNA expression (Fig. 6c). We further analyzed the effect of gefitinib on the apoptotic index in tumor tissue. Figure 6d shows that the apoptotic rate of scrambled shRNA‐transfected control cells was induced by gefitinib treatment, but that of BRAK shRNA‐transfected cells was not, suggesting that the antitumor activity of gefitinib involves at least two mechanisms: inhibition of cell growth stimulation signaling and stimulation of BRAK‐mediated apoptosis.

Figure 6.

 Introduction of shRNA targeting BRAK reduces efficacy of gefitinib in HSC‐3 cells. HSC‐3 cells were transfected with shRNA‐expressing vector targeting BRAK (shBRAK) or their scrambled control vector (Sh‐Scramble). Stable transfectants (1 × 10⁷ cells) were subcutaneously injected into both flanks of six female athymic nude mice, and allowed to form a tumor burden for 10 days. The mice were daily administered oral gefitinib (50 mg/kg/day). (a) Relative tumor volume was normalized by the vehicle control 6 days after administration of gefitinib. (b) BRAK mRNA expression in xenografts were determined by RT‐PCR. PCR products were visualized by ethidium bromide staining after electrophoresis in agarose gel. Tumor tissues were taken after administration of gefitinib for six days. They were fixed and embedded in paraffin wax. Immunohistochemical analysis of tumor specimens from Sh‐Scramble, shBRAK No3, and shBRAK No12 xenografts were performed for Ki‐67 and for apoptosis. Labeling indices for Ki‐67 (c) and apoptosis (d) were counted and expressed as the ratios normalized by the each vehicle control. Ratios were normalized by the vehicle control. *P < 0.05; **P < 0.001 (Student’s t‐test); values are expressed as the mean ± SD (n = 10 in panel a, n = 3 in panels c and d).

Determination of EGFR mutations in gefitinib responsiveness carcinoma cell lines HSC‐2, ‐3, and ‐4.  Because it has been shown that gefitinib‐administered NSCLC patients with EGFR mutations such as an in‐frame deletion in exon 19 (746–750, 747–751, or 747–753) or point mutation in exon 21 L858R transition gave a higher response rate diagnosed as a complete response,^{( 4} , ⁵ , ^{7 )} we determined whether the high sensitivity of HSC‐2, ‐3, and ‐4 cell lines to gefitinib was due to EGFR mutations. We could not detect any mutations in exon 19 (Fig. S1a) or exon 21 (Fig. S1b) of the EGFR gene in HSC‐2, ‐3, and ‐4 cell lines (Table 1), in agreement with our previously reported 16 HNSCC cell lines.^{( 17 )}

Table 1.

 Comparison of EGFR exon 19 and 21 sequences of HNSCC cell lines with the mutations leading to amino acid changes in the ATP‐binding domain of the tyrosine kinase observed in patients with non‐small‐cell lung carcinoma

Cell lines	Exon 19			Exon 21
	In‐frame deletion (746–750)	In‐frame deletion (747–751)	In‐frame deletion (747–753)	Amino acid substitution (L858R)
HSC‐2	ND	ND	ND	ND
HSC‐3	ND	ND	ND	ND
HSC‐4	ND	ND	ND	ND

ND, not detected.

Taken together, our data indicated that increase in BRAK expression after administration of gefitinib was strongly associated with the antitumor efficacy of gefitinib itself, independently from the EGFR mutations in HNSCC xenografts.

Discussion

It has been shown that EGFR is highly expressed in several types of carcinoma including HNSCC and that high expression levels of EGFR reduce recurrence‐free or overall survival rates in 70% and act as a strong prognostic indicator.^{( 27 )} Thus, EGFR TKIs have been developed for “Molecular Targeted Cancer Therapy”. Gefitinib and erlotinib are EGFR TKIs and have been tested in a clinical trial for patients with NSCLC (see for review^{( 28 )}), breast cancer,^{( 29 )} esophageal cancer,^{( 30 )} and HNSCC.^{( 31} , ³² , ³³ , ^{34 )} Although the efficacy of these agents has been strongly associated with specific EGFR mutations in NSCLC,^{( 4} , ^{5 )} some clinical trials have shown the exceptional cases.^{( 7} , ^{8 )} Besides, in HNSCC, EGFR mutations leading to amino acid change, which affect gefitinib’s binding affinity for the ATP binding pocket of the EGFR tyrosine kinase domain, are rare: 3/41 (7.3%) in Korean patients^{( 8 )} and none in 16 cell lines derived from Japanese patients.^{( 17 )} These observations suggest that another factor(s) is involved in the clinical benefit of gefitinib. On the other hand, we have shown that CXC chemokine BRAK mRNA expression forced by transfection with its expression vector with a CMV promoter dramatically reduced the tumorigenicity of HSC‐3 cells.^{( 16 )} In addition, BRAK expression was down‐regulated by EGF stimulation. These observations prompted us to determine the possibility that pharmacological induction of BRAK expression using EGFR TKI would be associated with antitumor activity. The experimental data showed that gefitinib inhibited tumorigenicity and also reduced the volume of tumor xenografts with concomitant stimulation of BRAK mRNA expression in tumor tissue whose stimulation was not seen in normal mouse tissue. Besides, this high efficacy of gefitinib was attenuated by inhibition of BRAK mRNA by transfection with BRAK shRNA expression vector (see Fig. 6). These data show that the antitumor efficacy of gefitinib was strongly associated with the increase in BRAK gene expression, which is in good agreement with our previous data, i.e. vector‐mediated forced expression of BRAK reduced tumorigenicity in nude mice.^{( 16 )}

Although gefitinib has been reported to cause a decrease in the Ki‐67 labeling index but not in the apoptotic index,^{( 35 )} here, we showed that BRAK shRNA‐mediated BRAK suppression abolished the growth‐suppressing efficacy of gefitinib in vivo due to the decrease in the apoptotic index, without affecting the Ki‐67 labeling index. Similar results were reported by Piechocki et al.,^{( 36 )} whose report also showed that the antitumor activity of gefitinib was due to increase in the apoptotic index rather than the proliferative index in acinic cell adenocarcinoma overexpressing human epidermal growth factor receptor 2 (HER2)/neu. In such a case, an increase in BRAK might be involved. EGFR mutations are known to be associated with selective Akt activation to promote cell survival. We could find no mutations in the HSC‐2, HSC‐3, and HSC‐4 cell lines (Fig. S1a,b and Table 1). We have previously reported that heterozygous and synonymous transition in EGFR exon 20 was correlated with an IC₅₀ (MTT assay) of gefitinib, but there was no correlation between BRAK mRNA expression and the IC₅₀. This may be explained by the fact that BRAK expression alone did not affect the proliferation rate or apoptosis in vitro.^{( 16 )}

We determined the frequency of BRAK mRNA expression in 18 kinds of HNSCC cell lines cultured under serum‐free conditions. As a result, BRAK mRNA expression was detected in 44.4% of the HNSCC cell lines, a frequency higher than that previously reported,^{( 12} , ¹⁴ , ^{15 )} when cells were cultured in the presence of serum.

In our experiments, BRAK mRNA levels were higher under serum‐free culture conditions than in serum‐supplemented culture conditions (data not shown). Besides, BRAK mRNA expression in these cells was inhibited by EGF treatment and this down‐regulation was recovered by gefitinib treatment in all cell lines tested.

In conclusion, our data indicate that the gefitinib‐induced restoration in BRAK expression is beneficial for tumor suppression in vivo. These data also suggest that stimulation of BRAK expression in serum‐free cultures by gefitinib treatment in vitro may be a useful new approach to determine the gefitinib responsiveness of patients with HNSCC.

Abbreviations

DMEM

  Dulbecco’s modified Eagle’s medium

EGF

 epidermal growth factor

EGFR

  epidermal growth factor receptor

ERK

extracellular signal‐regulated kinase

FBS

fetal bovine serum

HNSCC

 head and neck squamous cell carcinoma

mAb

monoclonal antibody

NSCLC

non‐small‐cell lung cancer

pAKT

phosphorylated Akt

PCR

polymerase chain reaction

pEGFR

phosphorylated EGFR

pERK

phosphorylated ERK

PI3

phosphatidylinositol 3‐phosphate

qPCR

real‐time quantitative PCR

RT‐PCR

reverse transcriptase–PCR

shRNA

short hairpin RNA

TKI

tyrosine kinase inhibitor

Supporting information

Fig. S1. No mutation at the regions of exon19 and 21 of the EGFR gene in HSC‐2, ‐3 and ‐4 cell lines. Mutations of EGFR in the exon 19 (a) and 21 (b) were analyzed by the direct sequence of RT‐PCR products.

Please note: Wiley‐Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

Supporting info item