BIIB021, a novel Hsp90 inhibitor, sensitizes head and neck squamous cell carcinoma to radiotherapy

Abstract

Heat shock protein 90 (Hsp90) is a molecular chaperone that promotes the conformational maturation of numerous client proteins, many of which play critical roles in tumor cell growth and survival. The ansamycin-based Hsp90 inhibitor 17-allylamino-17-demethoxygeldanamycin (17-AAG) is currently in Phase III clinical testing. However, 17-AAG is difficult to formulate and associated with dose-limited toxicity issues. A fully synthetic and bioavailable Hsp90 inhibitor, BIIB021, was evaluated for antitumor activity in a variety of head and neck squamous cell carcinoma (HNSCC) cell lines and HNSCC xenograft models, either as a single agent or in combination with fractionated radiation and the results were compared with that of 17-AAG. BIIB021 showed strong antitumor activity, comparable with, and in certain instances, superior to 17-AAG. BIIB021 enhanced the in vitro radiosensitivity of HNSCCA cell lines with a corresponding reduction in the expression of key radioresponsive proteins, increased apoptotic cells and enhance G2 arrest. In xenograft studies, BIIB021 exhibited a strong antitumor effect outperforming 17-AAG, either as a single agent and or in combination with radiation, thereby improved the efficacy of radiation. These results suggest that this synthetic and bioavailable Hsp90 inhibitor affects multiple pathways involved in tumor development and progression in the HNSCC setting and may represent a better strategy for the treatment of HNSCC patients, either as a monotherapy or a radiosensitizer. Furthermore, it also demonstrates the benefits of using preclinical models of chemosensitization to radiotherapy to explore clinically relevant radiation dosing schemes.

Head and neck squamous cell carcinoma (HNSCC) is the sixth most common cancer worldwide, with approximately 600,000 new cases reported each year. More than 50% of newly diagnosed patients will eventually relapse with either local recurrence or distant metastases. Unfortunately, in the relapse setting prognosis is poor, with median overall survival limited to about 1 year. Current treatment paradigms depend upon the stage of the disease at presentation. The standard treatment for loco-regional disease involves surgery and/or radiotherapy in either the neo- or adjuvant setting. Radiotherapy almost always employed with locally advanced disease, with a variety of fractionation regimens currently in clinical practice. Efforts to increase the efficacy of radiotherapy have included combination with chemotherapy, initially in the concomitant setting with platinum-containing regimens (chemoradiotherapy) and most recently as adjuvant treatment.1–3 Although the locoregional control and survival rates associated with chemoradiotherapy are higher than those for radiation alone, these intensive regimens are frequently associated with severe toxicities, leading to significant comorbidities. Targeted biological therapies that selectively interfere with cancer cell growth signals may improve patient survival by enhancing the effects of radiation, with the added benefit of reduced systemic toxicity. Current efforts to develop strategies for enhancing tumor radiosensitivity have focused on the use of agents that target a single molecule involved in modulating radiation induced cell death.4, 5 However, the mechanisms employed by cancer cells to overcome growth control are widely heterogeneous, thereby limiting the antitumor activity of inhibitors targeting a single-cellular pathway. Thus, targeting multiple pathways simultaneously should lead to more effective radiosensitization.

Heat shock protein 90 (Hsp90) is a molecular chaperone that promotes the conformational maturation of ‘client’ proteins and protects them from degradation.6 Many of the known clients are protein kinases or transcription factors involved in multiple signal transduction pathways including tyrosine kinases; Bcr-Abl, epidermal growth factor receptor (EGFR) family members, c-Met, insulin-like growth factor-1 receptor (IGF1-R) and pp60^c-src, serine/threonine kinases; Akt, Cdk4, Iκβ kinases α and β and Raf-1 and transcription factors; steroid hormone receptors, p53, Stat3, Mdm2 and telomerase.7–9 Raf-1, Akt and ErbB2 have been reported to be associated with the radioresponse, thereby protecting against radiation-induced cell death.10–12 We and others have previously demonstrated that inhibition of Hsp90 causes degradation of these proteins and enhances tumor cell death in a variety of cell lines and tumor models, including HNSCC.13–15 All these evidences indicate that inhibition of Hsp90 may not only provide a unique therapeutic pathway, but also promote the therapeutic effect of a variety of existing antitumor agents. Several Hsp90 inhibitors are currently under clinical investigation in a variety of oncology indications, spearheaded by 17-allylamino-17-demethoxygeldanamycin (17-AAG), a geldanamycin (GA) derivative. Evidence of biological activity in a variety of in vitro models and clinical efficacy in a range of oncology indications have been observed.16, 17

Preclinical studies have shown that 17-AAG can enhance tumor cell sensitivity to radiation.15 Nevertheless, 17-AAG is recognized for its poor pharmaceutical properties which contribute to its dose-limited toxicity and undermine its therapeutic potential.18 Here, we investigated a fully synthetic and bioavailable Hsp90 inhibitor, BIIB021 in a variety of HNSCC cell lines and tumor models, either as single agent or in combination with radiation. We found that BIIB021 showed strong antitumor activity, comparable to, and in certain instances, superior to 17-AAG. BIIB021 enhanced the in vitro radiosensitivity of these cell lines with an associated reduction in the expression of key radioresponse proteins, increased apoptotic cells and enhance G2 arrest. In xenograft studies, BIIB021 exhibited a strong antitumor effect as a single agent and increased the efficacy of radiation.

Material and Methods

Reagents

17-AAG (Fig. 1a) was derived from the Ansamycin antibiotic GA as previously described.19 BIIB021 (Fig. 1b), is a fully synthetic non-GA compound. The compounds were stored as 10 mM stock solutions in DMSO at −20°C, and diluted in the appropriate cell culture medium for use, such that the final DMSO concentration did not exceed 0.01%.

Figure 1. Structure of (a)17-AAG, (b) BIIB021.

Cell culture and cell growth inhibition

Four HNSCC cell lines were used for this study. UM11B (kindly provided by Dr. Gregory Wolf, University of Michigan) and Cal27 (purchased from ATCC) were cultured in DMEM supplemented with 10% fetal bovine serum, 100 U/ml penicillin G and streptomycin, and 1% nonessential amino acids. JHU12 and JHU22 (kindly provided by Dr. David Sidransky, John Hopkins University) were grown in RPMI-1640 with 10% fetal bovine serum and 100 U/ml penicillin G and streptomycin. All cells were cultured in a humidified atmosphere of 5% CO₂ at 37°C.

Logarithmically growing cells were counted and plated at 1,000 or 1,500 (for slower growing cell lines) cells per well (to have 10–15% confluent next day), in triplicate, in 96-well plates and incubated overnight. The next morning, 17-AAG or BIIB021 were added to the cell culture medium at increasing concentrations (0 nM, 65 nM, 125 nM, 250 nM, 500 nM, 1 μM, 2 μM). Cell growth inhibition was examined by MTT assay (Roche Diagnostics, Indianapolis, IN). Briefly, on the 5th day of the culture, 10 μl of MTT labeling reagent was added to each well. After 4 hr of incubation, 100 μl of solubilization buffer was added, and the plates incubated at 37°C overnight. The absorbance of each well was measured at 595 nm using a Vmax kinetics microplate reader (Molecular Devices, Sunnyvale, CA). The concentration of 17-AAG or BIIB021 yielding 50% growth inhibition (IC₅₀) were compared with controls for each cell line and expressed as mean values of at least 3 independent experiments.

Clonogenic assays were used to evaluate the effect of BIIB021 in combination with radiation. Briefly, cells were trypsinized to generate a single cell suspension, and a specific number of cells (to get about 100 clones after 12–14 days) were seeded into 60 mm plates. After allowing cells to attach, BIIB021 was added at about the IC₂₅ of each cell line (JHU12: 125nm; UM11B: 200nm). For BIIB021 in combination with radiation, 16 hr after adding BIIB021, cells were irradiated with a single dose of 2, 5, 7 Gy from a Cesium137 irradiator. Four hours after radiation, all plates were aspirated and fresh media were added. Twelve days after seeding, colonies were stained with crystal violet, and the number of colonies containing at least 50 cells was counted. The colony survival fraction was calculated for each treatment and data were presented as log plot. The results shown were mean values of 3 independent experiments with triplicate setting in each experiment. T-test was used to test the significance of the difference between different treatments.

Flow cytometry analysis of cell cycle and apoptosis

Cell cycle distribution was measured before and after HNSCC cells were exposed to either 1 μM BIIB021 or 5 Gy radiation for 24 hr. For combination, cells were treated with 1 μM BIIB021 for 16 hr, and then treated with a single fraction of 5 Gy of radiation. Cells were collected 24 hr after exposure to radiation, cells were washed with PBS, fixed with 70% ethanol, incubated with propidium iodide (20 μg/ml) and ribonuclease (200 μg/ml) for 30 min at 37°C, and analyzed by flow cytometry (FACS, Becton Dickson, Franklin Lakes, NJ). A ModFit II software program was used for cell cycle distribution analysis.

To test if cells were undergo apoptosis, cells were stained with annexin V using Annexin V-FITC apoptosis detection kit (BD Biosciences, San Jose, CA). Briefly, cells were plated, treated as above and collected without washing, 1 × 10⁵ cells were incubated with 5 μg/ml FITC-conjugated annexin V in the presence of 5 μg/ml of PI and then screened by flow cytometry. Annexin V positive PI negative cells scored as early apoptotic. Annexin V positive PI positive cells corresponded to late apoptotic cells.

Protein extraction and Western blotting

The drug treatment protocol utilized was the same as for the cell cycle analysis. The cells were harvested 24 hr after radiation treatment and resuspended in NP-40 lysis buffer (50 mM Tris-HCl, 150 mM NaCl, 50 mM NaF, 0.5% NP-40) containing 1 tablet of protease inhibitors cocktail (Boehringer Mannheim, Germany) per 10 ml. For in vivo tumor xenograft samples, tumor tissue was minced and homogenized in lysis buffer and centrifuged to remove debris. The protein extracts were quantitated using a Bio-Rad Bradford protein assay (Bradford Reagent; BioRad, Hercules, CA). A time course for the effect of BIIB021 on client protein expression was also done and cells were collected at 4, 8, 16, 24 and 48 hr after treatment.

Forty micrograms of protein extract was electrophoresed through 12% SDS polyacrylamide gels under denaturing conditions and transferred to nitrocellulose membranes. The membranes were blocked in PBS, 0.1% Tween 20, 5% non-fat dry milk, and incubated with the following primary antibodies: 0.8 μg/ml anti-c-Raf-1 monoclonal antibody (Santa Cruz, Santa Cruz, CA); 0.4 μg/ml anti-EGF-R rabbit polyclonal antibody (kindly provided by Dr. Shelton Earp, University of North Carolina); 0.2 μg/ml anti-phospho-EGFR^Tyr1173 polyclonal antibody, 0.2 μg/ml anti-phospho-Akt^Ser473 polyclonal antibody, 0.2 μg/ml anti-cleaved PARP polyclonal antibody (both Cell Signaling, Beverly, MA) and 0.2 μg/ml anti-Tubulin monoclonal antibody (Roche Molecular Biochemicals, Indianapolis, IN). After washing with 1× PBS-T buffer 3 times, the membrane was incubated with horseradish peroxidase-conjugated goat anti-mouse or anti-rabbit second antibody (Santa Cruz and Amersham Biosciences, Piscataway, NJ, respectively). Specific antigen-antibody interactions were detected with enhanced chemiluminescence (Amersham Biosciences).

In vivo mouse tumor model studies

Experiments were conducted in accordance with the US Public Health Service Policy on Humane Care and Use of Laboratory Animals and the rules and regulations of the University of North Carolina Institutional Animal Care and Use Committee. Seven- to eight-week-old nude mice were obtained from Harlan Horst Laboratory (Indianapolis, IN) and maintained in ventilated caging. The mice were divided into the following 6 groups with 5 mice per group: no treatment control, BIIB021 treated, 17-AAG treated, radiation treated, BIIB021 + radiation treated and 17-AAG + radiation treated.

JHU12 cells were harvested, washed with 1× PBS, suspended in PBS at 3 × 10⁶ cells/100 μl, and then inoculated subcutaneously into the flanks of mice. After the tumor size reached 2–3 mm in diameter (about 6– 8 days), BIIB021 solution or emulsion vehicle were given orally using a mouse feeding tube 3 days/week (M, W, F) for 4– 5 weeks. 17-AAG emulsion was injected intraperitoneally into mice 3 days/week (M, W, F) for 4–5 weeks (multiple doses of BIIB021, were tested and 20 mg/kg/day chosen so as to obtain maximal enhencement).

For the radiation protocol, mice were first anesthetized with 125 mg/kg Avertin by intraperitoneal injection, placed into a cone shaped lead shield with a hole on the side to expose only the tumor, and then put into the irradiator. A radiation dose of 1.5 Gy was given per fraction and mice were radiated 4 times a week (T, W, Th, F), 6 Gy/week. The tumor size and mouse weight were measured twice weekly; (T and F). The experiment was ended when some control mice tumor reach to the maximum allowed size (2.0 cm for the longest diameter) or developed ulceration insingle treatment group. At the time of sacrifice, tumor tissue was lysed in NP-40 buffer for the protein expression assay, and a portion fixed in 10% formalin for histology to confirm tumor structure.

Results

17-AAG and BIIB021 inhibited the growth of HNSCC cell lines

Four human HNSCC cell lines were chosen for this study because they all express a high level of EGFR which has been implicated to play a role in radioresistance,20 they exhibit differential sensitivities to 17-AAG, and display differential responses to cisplatin or radiation.13 Cal27 is sensitive to both cisplatin and radiation, while UM11B is relatively resistant to both therapies. In contrast, JHU12 and JHU22 are resistant to cisplatin whilst relative sensitive to radiation. These cell lines were initially treated with increasing concentrations of 17-AAG or BIIB021 for 5 days and the effects on cell viability evaluated by MTT assay. These cell lines exhibited a relatively broad range of sensitivity to 17-AAG, with a mean IC₅₀ value of 690 ± 375 nM, with JHU12 being the most resistant and Cal27 the most sensitive (Fig. 2a). These results are in agreement with those reported previously.13 By contrast, all 4 cell lines were more sensitive to BIIB021, with the mean IC₅₀ of 250 ± 100 nM, falling within a more restricted range (Fig. 2a). Thus, BIIB021 induced growth inhibition in a dose-dependent manner, with a similar biphasic curve for each cell line (Fig. 2b).

Figure 2. Antiproliferation effects of 17-AAG and BIIB021 on HNSCC cell lines. (a) IC50 of 17-AAG and BIIB021 were measured by MTT assay in 4 HNSCC cells and were shown as bar graph. (b) Growth inhibition curve of HNSCC cell lines after BIIB021 treatment. Cells were treated with increasing concentration of BIIB021 for 5 days. Alive cells were quantitated by MTT assay. % of viability = OD595 of treated cells/OD595 of control cells × 100. Data points show the mean of at least 3 experiments and SE in each condition.

BIIB021 enhanced the antitumor effect of radiation in vitro

To determine whether 17-AAG or BIIB021 enhance radiation-induced cell death in HNSCC cells, the relative radiation sensitive cell line JHU12 and relative radiation insensitive cell line UM11B were exposed to 17-AAG or BIIB021 for 16 hr followed by a single dose of radiation (2, 5 or 7 Gy). The impact of the single and combination treatments on cell proliferation was then measured by clonogenic assay. As expected, UM11B cells (Fig. 3b) were relatively resistant to radiation compared to JHU12 (Fig. 3a), 50% killing dose is 4 versus 3 Gy for JHU12 cells. When cells were pretreated with BIIB021 prior to radiation, a strong growth inhibition was observed in UM11B cells at every dose level tested (Fig. 3b; p < 0.05 at 2Gy, p = 0.014 at 5 Gy and p = 0.0125 at 7 Gy when the cell growth inhibition difference between radiation alone were compared to that of BIIB021 plus radiation), This observation suggests that the inhibition of Hsp90 can overcome the radioresistance of this cell line, possibly by causing the degradation of key oncogenic client proteins that mediate the effect. This result is intriguing because it may represents a potential mechanism for overcoming the radiation resistance observed in HNSCC cells, given that radiation is the major therapeutic option for this type of tumor. In JHU12 cells, which is relative sensitive to radiation, the enhancement by BIIB021 is less pronounced than that in UM11B cells, at lower radiation doses (p = 0.06 at 2 Gy, p = 0.05 at 5 Gy) However, at 7 Gy, BIIB021 significantly increased the anti-tumor effect of radiation (p = 0.008) (Fig. 3a). These observations suggest that the most dramatic radiosensitizing effects of an Hsp90 inhibitor are in the radiation resistant setting. Pretreatment of cells with 17-AAG prior to radiation, resulted in an additive effect in both the radiation relative sensitive and relative resistant lines when analyzed with MTT assay (data not shown).

Figure 3. Radiosensitising effect of BIIB021 on the survival of HNSCC cells. Specific number of JHU12 (a) and UM11B (b) cells were plated and cells were exposed to IC₂₅ concentration of BIIB021 for each cell line for 16 hr followed by a single radiation dose of 0, 2, 5 or 7 Gy. Cells were then fed with fresh media. The survival fraction was assessed by Colony-formation assay at 12 days after irradiation. Data are shown as the mean value with SE of 3 independent experiments with triplicate setting in each. p < 0.05, for each radiation dose in combination with BIIB021 versus each radiation dose alone for UM11B cells, and p < 0.05 for dose 5 and 7 Gy for JHU12 cells. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

BIIB021 alone induce G1 arrest and enhance radiation induced G2 arrest independent p53 status

To determine the effect of BIIB021 on cell cycle progression in HNSCC cells, either alone or in combination with radiation, HNSCC cell lines with either wild type (JHU12) or mutant p53 (UM11B) were treated with 1 μM BIIB021 and analyzed by FACS 40 hr post treatment. The G1/G0 population increased by 11 and 25%, respectively and that of the S+G2 phase decreased accordingly in both cell lines (Fig. 4). Thus, the G1 arrest observed in both cell lines in response to BIIB021 treatment was independent of p53 status. In contrast, treatment with 5 Gy radiation dramatic increased G2/M accumulation and decrease in the S population in JHU12 cells; while in UM11B cells, only marginal increase in G2/M and but obvious reduction in S phase cells was observed, consistent with the XRT relative resistance phenotype of UM11B. When cells were treated with BIIB021 followed by radiation, an additional increase in the G2/M population was detected. This effect was most dramatic in the JHU12 cell line expressing wild-type p53, with the G2/M population increasing by 27% and the detection of a significant number of aneuploid cells (Fig. 4).

Figure 4. BIIB021 induced cell cycle arrest in HNSCC cell lines. JHU12 and UM11B Cells were treated with 1 μM BIIB021, or 5 Gy of radiation for 24 hr or with BIIB021 16 hours prior to radiation. Cells were harvested 24 hr after radiation and cell cycle distribution was assessed by flow cytometry. Both JHU12 and UM11B cells were arrested at G1/G0 phase after BIIB021 treatment. Combination of BIIB021 and radiation arrested cells in G2/M phase.

BIIB021 induce cell death via apoptosis

To confirm that BIIB021 inhibit cell growth and enhance radiation effect is by inducing apoptosis, Annexin V stain followed by flow cytometry analysis was performed for JHU12 and UM11B cells treated with BIIB021 or radiation alone or the combination. As shown in Figure 5, combination of radiation and BIIB021 induced enhanced apoptosis in both cell lines than either agent alone. This was more manifested in UM11B cells than in JHU12 cells (2.6-fold increase versus 1.6-fold increase) and indicated that BIIB021 sensitized HNSCC cells to radiation by increasing apoptotic cell death.

Figure 5. BIIB021 induce cells undergo apoptosis. Cells were plated and treated as for cell cycle analysis and collected together with supernatant. Annexin V was used to stain for apoptotic cells. Annexin V positive PI negative cells scored as early apoptotic. Annexin V positive PI positive cells corresponded to late apoptotic cells. Annexin V-stain shown that BIIB021 inihibit tumor cell growth is by inducing apoptosis. The percentage of apoptotic cells increased significantly when UM11B cells were treated with both BIIB021 and radiation (2.6-fold increase), this also confirm that the combination of BIIB021 with radiation enhanced the radiation effect on HNSCC cells.

BIIB021 inhibited Hsp90 key client protein expression and activation

The effect of BIIB021 on Hsp90 key client proteins was analyzed in a time course study and was shown in Figure 6a. BIIB021 caused a decrease in EGFR as well as Akt expression at about 16 (UM11B) and 24 hr (JHU12). The effect on the phosphorylation/activation of these two clients occurred at earlier time points (4 hr for UM11B and 8 hr for JHU12) prior to the decline of their protein level.

Figure 6. (a) Time course of BIIB021 induced EGFR and Akt protein degradation in HNSCC cells. JHU12 and UM11B cells were treated with 1 μM BIIB021. Cells were harvested at indicated time. EGFR, pEGFR, Akt and pAkt protein expression level was determined. (b) Suppression of radiation resistant by BIIB021 associated with decrease of key client proteins in HNSCC cell. Cells were either treated with 1 μM BIIB021 for 48 hr, or 5 Gy radiation or BIIB021 plus radiation. Protein expression was determined by Western blot.

BIIB021 alone and in combination with radiation induced degradation of key oncogenic proteins and was associated with apoptosis.

Although Hsp90 has numerous client proteins,21 EGFR, c-Raf-1 and Akt were of particular interest because they have been reported to play a role in radioresistance.22–24 Furthermore, the reduction of either total proteins or their active/phosphorylated form has been used as biomarkers of Hsp90 inhibition.9 To explore the degradation of these key client proteins, western blot analysis was performed on extracts of JHU12 and UM11B cells following 24 hr exposure to either 1 μM BIIB021 or 5 Gy radiation alone, or in combination. As shown in Figure 6b, radiation alone had a limited effect on the level of the proteins examined. However BIIB021 treatment alone resulted in a significant reduction of EGFR, c-RAF and pAKT in both cell lines. This effect was enhanced in a client protein—specific manner when radiation was coadministered.

All treatments induced apoptosis to a certain degree, based on the cleavage of the Caspase-3 substrate, PARP (Fig. 6b). Under the conditions employed, radiation alone only marginally increased PARP cleavage when compared to control in both cell lines. Furthermore, the BIIB021 alone also induced moderative PARP cleavage, and was comparable between the 2 cell lines. Not surprisingly, the most significant effect was seen with the co-treatment of BIIB021 and radiation. These results indicated that BIIB021 mediated degradation of Hsp90 client proteins could restore sensitivity of HNSCC cells to radiation especially in cell lines that are relative resistant to radiation.

BIIB021 significantly enhanced antitumor growth effect of radiation in HNSCC xenografts

The radiosensitizing effect of BIIB021 was tested in an in vivo HNSCC xenograft model. In this study, we employed JHU12 tumor-bearing mice, as UM11B grew slow in vivo. Radiotherapy regimens currently in clinical practice for the treatment of HNSCC, involve fractionated daily radiation often in combination with concurrent chemotherapy, for the reason that the biologic basis of chemosensitization may vary with single dose radiation. Therefore, in our xenograft studies, we applied the same approach and treated mice with fractionated radiation 4 times a week (T, W, Th, F) for 4–5 weeks. BIIB021 was given 3 days a week (M, W, F) for a total of 4–5 weeks. As shown in Figure 7, both BIIB021 and radiation alone exhibited dose-dependent antitumor activity. BIIB021 was particularly active in this tumor model, where 83 and 94% tumor growth inhibition was observed at 40 and 80 mg/kg, respectively.

Figure 7. Both BIIB021 and fractionated radiation inhibited tumor growth in JHU012 xenografts BIIB021 was dosed 1 week after flank injection with 3 × 10⁶ JUH012 cells, when tumor size were palpable (about 2 mm in diameter). Increasing doses of BIIB021 was given orally Monday, Wednesday and Friday. 0.5–5.0 Gy radiation was given Tuesday through Friday. Tumor volume was determined twice a week using calipers. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

For the combination study, to maximize the potential for observing an optimal synergistic effect, sub-optimal doses of each therapy were used specifically; 10 and 20 mk/kg of BIIB021 and 1.5 Gy of radiation. As shown in Figure 8a, the 2 doses of BIIB021 (10 and 20 mk/kg) and radiation alone inhibited tumor growth at 30, 49 and 56%, respectively, whilst the combination therapy elicited 91% inhibition (p = 0.028 compared with radiation alone and p = 0.02 compared to BIIB021 alone). In this experiment, BIIB021 showed comparable, if not better, efficacy than 17-AAG as a single agent at equivalent dosages. However, in the combination arm, the size of the tumors treated with 17-AAG and radiation stabilized after 3 weeks of treatment (14% of the control tumor volume), whilst those treated with BIIB021 and radiation grew at a slower rate at the same time point, even showing evidence of regression after 4 weeks (16% in 10 mg/kg and 8% in 20 mg/kg plus radiation of control tumor volume) (Fig. 8a). Weight loss in the single modality groups was less than 5% relative to controls and 8–10% in the combination treatment groups, suggesting that both drugs were well tolerated (data not shown). Five weeks after tumor cell inoculation, most tumors in control group reached size of about 15 mm in diameter, some mice tumor in single treatment groups also grown to over 10 mm in diameter and developed ulceration, while tumors in combination treatment grow were much smaller. These data demonstrate that BIIB021, an orally available Hsp90 inhibitor, has strong antitumor activity in a HNSCC tumor model and sensitizes the tumors to radiation according to a fractionated treatment that mimics current clinical regimens.

Figure 8. BIIB021 sensitized JHU12 xenograft to radiation therapy and reduced the protein levels associated with radiation resistant in mice tumor. (a) BIIB021 or 17-AAG was dosed 1 week after flank injection with 3 × 106 JUH012 cells. BIIB021 was given orally at indicated dosage, while 17-AAG (20 mg/kg/day) was injected intraperitoneally Monday, Wednesday and Friday. 1.5 Gy × 4 fractionated Radiation was given Tuesday through Friday. For combination therapeutic group, radiation was given 16–20 hr after compound treatment. Tumor volume was determined using calipers. Tumor volume in combination therapeutic group is significantly reduced compared with that of radiation or BIIB021 treatment alone (p = 0.028 and 0.02). (b) Reduction of EGF-R and Raf-1 in JHU12 xenograft tumors. Mice from treatment groups or control group were sacrificed 16 hr after the final treatment. Tumor lysate was obtained and measured for the level of EGF-R, Raf-1 by Western blot. This experiment was repeated with a second set of animals with equivalent results.

We also looked at the degradation of specific Hsp90 client proteins, including EGFR, pAkt and Raf-1 in the tumor tissue. Despite the fact that radiation alone had no effect, BIIB021 treatment led to a marked reduction in the expression of these proteins (Fig. 8b). Furthermore, this reduction was enhanced in combination with radiation, suggesting that BIIB021 contributed to tumor growth inhibition via degradation of key client proteins.

Discussion

The goal of radiotherapy is to kill tumor cells efficiently with as little damage to normal tissue as possible. Unfortunately, radiotherapy alone often fails to eliminate tumors in advanced squamous cell carcinoma, leading to locoregional recurrence and the development of distant metastasis. Concurrent radiation and chemotherapy with cisplatin-based regimens has improved tumor control significantly. However, a subset of patients do not respond well to cisplatin-based chemoradiotherapy, and cisplatin-related toxicity is of major concern.

Hsp90 inhibitors are a group of promising antitumor agents that lead to the selective degradation of proteins involved in multiple oncogenic processes.16, 25 Hsp90 inhibitors hinder the growth of many types of tumors in both in vitro and in vivo tumor models.26, 27 We have previously described a novel intraperitoneally administered, ansamycin-based Hsp90 inhibitor, EC5, which effectively inhibited the growth of HNSCC cells in vitro and in in vivo tumor models.13 Several Hsp90 client proteins including EGFR, ErbB2, Raf-1, vascular epithelial growth factor (VEGF) and protein kinase B (PKB0/Akt) are known to play a role in radiation resistance,22–24 and this positions Hsp90 inhibitors as good candidates for therapeutic radiosensitization.

In this report, we describe the effect of the orally available non-ansamycin-based Hsp90 inhibitor BIIB021, alone and in combination with radiation, on the growth of 4 HNSCC cell lines and in an in vivo xenograft model. As reported previously for other Hsp90 inhibitors both in HNSCC and in other cell types,13, 28 BIIB021 alone induced anti-proliferative effects in all of the 4 cell lines tested. The average IC₅₀ of BIIB021 in the 4 cell lines tested here was lower than that of 17-AAG, the lead Hsp90 inhibitor in clinical trials. In combination with radiation, BIIB021 exhibited strong antitumor effect and sensitized the tumor cells to radiation. Annexin V stain followed by FACS analysis demonstrated that the enhanced effect of BIIB021 on radiation occurs by induction of apoptosis. Cell cycle analysis demonstrated that BIIB021 induced G1/G0 growth arrest and significantly enhanced radiation-induced G2 arrest in both JHU012 and UM11B cells. In addition, aneuploidy was also observed in the relative radiosensitive JHU012 cells, but not in the relative radioresistant UM11B cells, suggesting that BIIB021 might potentiate radiosensitization by different mechanisms in different cell lines. Using glioma, pancreatic carcinoma and prostate carcinoma cell lines, Bull et al. have found that 17-DMAG, a water-soluble form of 17-AAG, can abrogate radiation induced G2 arrest14 and concluded that this may be the mechanism responsible for the enhancement of radiosensitization effect of 17-DMAG. This observation suggests that the cotreatment may overcome the self-checking ability of the cells forcing them to continue to proliferate whilst carrying mutations, finally culminating in cell death. Interestingly, we found that the BIIB021 alone induces cells arrest at G1, but in combination with radiation enhances G2 arrest by activating G2 checkpoint, followed by apoptotic cell death. This is supported by the enhanced apoptosis observed in response to the combination treatment. Further investigation as to the impact of BIIB021 and radiation on CHK1, CHK2 and other check-point associated proteins will shed light on this proposed mechanism of radiosensitization.

BIIB021 either alone or in combination with radiation reduced both the expression and activation of EGFR, Raf-1 and pAkt, proteins that are important for tumor cell survival and radioresistance in HNSCC. These results indicate that the radiosensitizing effect of BIIB021 on HNSCC cells is carried out through inhibition of the expression of radioresistance-associated oncoproteins, thereby sensitizing the cells to radiation and resulting in cell death. These results are consistent with a previous report which showed that GA can sensitize cells lines from a variety of human tumors to radiation.29 In the study, Machida et al. showed that GA was cytotoxic and potentiated radiation in SQ-5, an EGFR overexpressing, lung squamous cell carcinoma cell line, and in DLD-1, a human colon adenocarcinoma cell line. The effect was more pronounced in SQ-5 cells than in DLD-1 cells, indicating that the overexpression of a Hsp90 client, in this instance EGFR, sensitized the cells even further to the combined treatment of Hsp90 inhibitor and radiation. In a further study, using T-24 (a bladder cancer cell line) and HMV-1 (a melanoma cell line), they demonstrated that GA preferentially sensitized these tumor cell lines to radiation compared to normal cells.30 The combination of GA and radiation was shown to abolish Akt activity and strongly enhance the induction of apoptosis in these tumor cell lines. In an experiment with 17-AAG in 3 HNSCC cell lines (SQ20B, SCC61 and SCC13), they found that 17-AAG enhanced radiosensitivity more effectively in radioresistant tumor cells than in radiosensitive cells.31 This effect was observed in both monolayer culture and spheroids and again was achieved by inhibiting the PI3K-Akt pathway. Taken together, these data indicate a general ability of Hsp90 inhibitors to potentiate the effects of radiation in HNSCC cells, given that the tumor cell lines used in each study have diverse genetic backgrounds, including differences in p53 status.

Fractionated radiation regimens are used therapeutically in order to minimize associated side effects, by reducing the size of each fraction, whilst increasing the overall dose and thereby maintaining the therapeutic potential. Therefore, a fractionated regimen was used in our in vivo efficacy studies, which showed that BIIB021 at doses lower than those required as montherapy (10 and 20 mg/kg) can effectively inhibit tumor growth when combined with fractionated radiation. The therapeutic effect of the combined treatment was much more effective compared to either treatment alone and was more pronounced than that in vitro. After 3 weeks of combined treatment, tumor stasis was achieved at the lower BIIB021 dose, and tumor regression at the higher dose. 17-AAG demonstrates a similar radio-sensitizing effect, but is not as effective as BIIB021 at the same dose level. Previous studies from Bisht et al.32 also showed that as a Hsp90 inhibitor, 17-AAG can sensitize the response to either a single dose (12 Gy) or a fractioned course (2 Gy given over 5 consecutive days) of radiation therapy as demonstrated by growth inhibition in both HeLa and SCC VII xenograft models.

Overall, our studies have shown that BIIB021, a fully synthetic, orally available Hsp90 inhibitor demonstrated strong anti-tumor effects in HNSCC cell lines and tumor models. Furthermore BIIB021 synergizes with radiation, a commonly used therapeutic modality in the treatment of HNSCC, significantly enhancing the efficacy of the latter. We have demonstrated that the therapeutic effect of BIIB021 is comparable to that of 17-AAG in in vitro models, but is superior in in vivo models, both as a single agent and in combination with radiation. In addition, BIIB021 possesses superior pharmaceutical properties as shown by its water solubility and good bioavailability. Furthermore, unlike ansamycin derivatives, BIIB021 is not a substrate for p-glycoprotein, and hence multidrug resistance (MDR) mechanisms will not compromise its therapeutic application.33 All these properties make BIIB021 a promising antitumor agent for the treatment of HNSCC patients, by promoting the therapeutic activity of radiation.

Acknowledgements

This work was supported by a funding from American Academy of Otolaryngology/Head & Neck Surgery to Carol Shores.