Original article Strahlenther Onkol 2014 DOI 10.1007/s00066-014-0605-5 Received: 7 October 2013 Accepted: 9 December 2013 © Springer-Verlag Berlin Heidelberg 2014

Andrea Arenz1 · Frank Ziemann1 · Christina Mayer2 · Andrea Wittig1 · Kirstin Dreffke3 · Stefanie Preising1 · Steffen Wagner2 · Jens-Peter Klussmann2 · Rita EngenhartCabillic1 · Claus Wittekindt2 1Department of Radiotherapy and Radiooncology, BMFZ—Biomedical Research Center,

Philipps-University, Marburg, Germany 2Department of Otorhinolaryngology and Head and Neck Surgery,

Justus Liebig University, Giessen, Germany 3Institute for Radiobiology and Molecular Radiooncology, Philipps-University, Marburg, Germany

Increased radiosensitivity of HPV-positive head and neck cancer cell lines due to cell cycle dysregulation and induction of apoptosis

Electronic supplementary material The online version of this article (doi: 10.1007/ s00066-014-0605-5) contains supplementary material, which is available to authorized users.

Head and neck squamous cell carcinoma (HNSCC) represent a heterogeneous group of malignant tumors, developing from the squamous epithelium of the upper aerodigestive tract. In contrast to other subsites, rising incidence rates have been reported for cancer of the oropharynx [1]. While the most important risk factors for developing HNSCC are tobacco and alcohol consumption, recent studies have shown that an increasing subgroup of oropharyngeal cancers are associated with infection of high-risk types of human papillomaviruses (HPV), mainly HPV-16. The majority of HPV-related cancers contains HPV DNA integrated into the host cell genome expressing two viral oncogenes, E6 and E7, which encode proteins necessary for transforming epithelial cells [2, 3]. The HPV-16 E6 oncoprotein suppresses p53 activity through ubiquitin-mediated protein degradation. E6 expressing cells are unable to induce

p53-mediated apoptosis or cell cycle arrest in response to DNA damage and, hence, are susceptible to genomic instability [4]. E7 complexes with the hypophosphorylated pocket proteins pRb, p107, and p130, facilitating the release of the transcription factor E2F, which constitutively activates DNA synthesis and cell proliferation [5]. The Rb tumor suppressor protein inactivates E2F. E7 inactivates the cyclin-dependent kinase (Cdk) inhibitors p21/Cip1 (CDKN1A) and p27/Kip1 (CDKN1B) and complexes with cyclins A and E [6, 7]. These processes are essential for the productive stage of the HPV life cycle because the virus must induce differentiated cells to create an environment supportive of the amplification of viral DNA. Loss of cell cycle control and impaired induction of apoptosis, therefore, constitute central events in HPV-mediated carcinogenesis. Despite known differences in stage distribution at diagnosis with usually lower T but higher N stages, patients with HPV-associated HNSCC have a favorable outcome compared to HPV-negative cancers. However, treatment recommendations in international guidelines do not differentiate between HPVpositive vs. HPV-negative HNSCC [8].

This is partly due to the fact that the reasons for a higher efficiency of radiotherapy in HPV-positive tumors is not understood and sparse data exist to discriminate the molecular mechanisms responsible for the differing treatment response. So far evaluation of biomarker expression of signaling pathways has failed to reveal any statistically strong independent prognostic utility [9]. We therefore investigated intrinsic radiosensitivity and damage response in previously established HPV-positive and HPV-negative HNSCC cell lines after irradiation aiming to identify those mechanisms which illustrate reasons for the increased sensitivity of HPV-positive cancers of the oropharynx.

Material and methods Cell lines and culture conditions All cell lines were cultured in RPMI 1640 medium supplemented with 10 % FBS, 2 mM L-glutamine, 1 % non-essential amino acids, and gentamicin (0.1 mg ml−1) and were incubated at 37 °C in 5 % CO2 atmosphere. The characteristics of all cell lines are listed in .  Table 1. The identity of cell lines was confirmed using SNP proStrahlentherapie und Onkologie X · 2014 

| 1

Original article Table 1  Characteristics of HPV-positive and HPV-negative cell lines used in this study Cell line (provided by) UD-SCC-1 (a) UM-SCC-6 (b) UM-SCC-11b (b) UT-SCC-33 (c) UM-SCC-47 (b) UM-SCC-104 (b) UPCI:SCC152 (d) 93-VU-147T (e)

Age/ TNM stage sex of donor 64/M T3N2bM0/G3

Treatment of Site of origin HPV p53 References the originatstatus status ing tumor UT Oropharynx neg FS/Wt 11, 12, 13

37/M

T2N0M0/GX

UT

Tongue

neg

Wt

65/M

T2N2aM0/GX

CTx

Larynx

neg

C242S 13, 12

86/F

T2N0M0/G2

UT

Oral cavity

neg

R282W 12, 13

53/M

T3N1M0/G2

nn

HPV16 Wt

12, 13

56/M

T4N2bM0/G2 RTx + CTx

Oral cavity/ tongue Oral cavity

HPV16 Wt

17

47/M

T2N0M0/G2

RTx

Hypopharynx HPV16 Wt

18

58/M

T4N2MX/GX

UT

Oral cavity

14, 15, 16

HPV16 L257R/ 19, 20, 21 Wt

Cell lines were kindly provided by (a) Dr. Thomas Hoffmann, University of Duesseldorf, Germany, (b) Dr. Thomas E. Carey, University of Michigan, United States; (c) Dr. Reidar A. Grenman, Turku University, Finland; (d) Dr. Susanne M. Gollin, University of Pittsburgh, United States; (e) Johan P. de Winter, VU Medical Center, Amsterdam UD University of Düsseldorf; UM University of Michigan; UPCI University of Pittsburgh; UT University of Turku; M male, F female; RTx radiotherapy; CTx chemotherapy; UT untreated; nn not known; Wt wildtype; FS frame shift mutation

files and short tandem repeat testing [10]. HPV status was confirmed by expression of HPV-16 E6 and E7 transcripts in qPCR (Supplementary Fig. 1).

least squares fit (GraphPad Prism 5.0 software). Each experiment was performed in triplicate with a minimum of three independent repetitions.

Irradiation

Flow cytometric analysis of cell cycle phase distribution

Cells were irradiated with 6 MeV photons using a linear accelerator (Elekta Supernova, Elekta, Stockholm, Sweden) with a dose rate of 4 Gy/min at room temperature in a PMMA phantom. Absolute dose measurements confirmed the applied doses.

Clonogenic assay Exponentially growing cells were plated in 6 cm petri dishes (cell densities: 200– 24,000 cells) 14 h prior to irradiation with 1–6 Gy X-rays. After incubation for 14– 18 days, cells were fixed (10 % formaldehyde) and stained (0.1 % crystal violet) for colony counting (colonies ≥ 50 cells). The surviving fraction (SF) was normalized to the plating efficiency of nonirradiated controls and clonogenic surviving fractions calculated [22]. Survival curves were fitted to the linear‒quadratic model (SF = exp−[α × D + β*D2]) according to a

2 |  Strahlentherapie und Onkologie X · 2014

Cells were harvested at the indicated time points. Media and washing buffer were collected and adherent cells gently dissociated with accutase, fixed (70 % ice-cold ethanol), and incubated for 30 min in PBS containing 200 µg/ml RNase A, 0.1 % Triton X-100, and 20 µg/ml propidium iodide. Flow cytometric measurements were performed using a LSR II flow cytometer (Becton Dickinson) by analyzing 20,000 cells per sample. Data were processed using FlowJo 7.6 software (Tree Star Inc., San Carlos, CA, USA).

Immunohistochemistry of γH2AX To accumulate cells in the G1 phase, cells were grown on glass coverslips in petri dishes under serum starvation (0.25 % calf serum, 5 days). Cells were irradiated in the presence of culture medium with 20 % serum supplemented with 5-ethynyl-2′-

deoxyuridine (EdU, 1:1000; Click-iT Assay Kit, Invitrogen) and incubated for the indicated time points. Cells on coverslips were fixed with 4 % paraformaldehyde and stained for EdU incorporation (following the manufacturer’s protocol) and histone H2AX phosphorylation (1:100, anti-phospho-histone H2AX (Ser139), clone JBW310, Millipore) followed by Alexa Fluor-488 conjugated goat-antimouse secondary antibody (1:200, PK-PF488-AK-M1, Promokine). Residual cells in the petri dishes were used to measure cell cycle distribution. γH2AX foci were quantified in EdU-negative and EdU-positive cells (cells having traversed S phase). A minimum of 50 cells in three independent experiments were counted and data are shown as mean values ± SEM.

Detection of radiation-induced apoptosis Phosphatidylserine on the external surface of the plasma membrane (marker for apoptosis) was evaluated by DNA flow cytometry using the annexinV-FITC Detection kit (Promokine, Heidelberg, Germany) according to the manufacturer’s instruction. Unfixed cells were double stained with FITC-conjugated annexin V and propidium iodide (PI) in a Ca2+-enriched binding buffer and a minimum of 20,000 cells analyzed. Early and late apoptotic cells were evaluated by quadrant statistics on PI-negative and annexin V-positive cells using FlowJo 7.6. Results are displayed as the sum of upper right and lower right quadrants as a percentage of annexin V-positive cells normalized to control in three independent experiments.

Western blot analysis Following irradiation, cells were lysed in ice-cold RIPA buffer supplemented with protease inhibitor cocktail and PMSF (AppliChem, Darmstadt, Germany). Lysates were boiled once in SDSPAGE sample buffer (25 mM Tris-HCl, pH 6.8; 10 % glycerol, 2 % SDS, 2.5 % βmercaptoethanol, 0.005 % bromphenol blue) and equivalent amounts of protein were electrophoresed on SDS-PAGE gels. After electroblotting onto a Immobilon-PVDF membrane (Millipore),

Abstract · Zusammenfassung Strahlenther Onkol 2014  DOI 10.1007/s00066-014-0605-5 © Springer-Verlag Berlin Heidelberg 2014 A. Arenz · F. Ziemann · C.  Mayer · A. Wittig · K. Dreffke · S. Preising · S. Wagner · J.-P. Klussmann · R. Engenhart-Cabillic · C. Wittekindt

Increased radiosensitivity of HPV-positive head and neck cancer cell lines due to cell cycle dysregulation and induction of apoptosis Abstract Background and purpose.  Human Papillomavirus (HPV)-related head and neck squamous cell carcinoma (HNSCC) respond favourably to radiotherapy as compared to HPV-unrelated HNSCC. We investigated DNA damage response in HPV-positive and HPVnegative HNSCC cell lines aiming to identify mechanisms, which illustrate reasons for the increased sensitivity of HPV-positive cancers of the oropharynx. Methods.  Radiation response including clonogenic survival, apoptosis, DNA doublestrand break (DSB) repair, and cell cycle redis-

tribution in four HPV-positive (UM-SCC-47, UM-SCC-104, 93-VU-147T, UPCI:SCC152) and four HPV-negative (UD-SCC-1, UM-SCC-6, UM-SCC-11b, UT-SCC-33) cell lines was evaluated. Results.  HPV-positive cells were more radiosensitive (mean SF2: 0.198 range: 0.22–0.18) than HPV-negative cells (mean SF2: 0.34, range: 0.45–0.27; p = 0.010). Irradiated HPVpositive cell lines progressed faster through S-phase showing a more distinct accumulation in G2/M. The abnormal cell cycle checkpoint activation was accompanied by a more

pronounced increase of cell death after x-irradiation and a higher number of residual and unreleased DSBs. Conclusions.  The enhanced responsiveness of HPV-related HNSCC to radiotherapy might be caused by a higher cellular radiosensitivity due to cell cycle dysregulation and impaired DNA DSB repair.

Zusammenfassung Hintergrund.  Für Patienten mit HPV-assoziierten Kopf-Hals-Tumoren (HNSCC) ist im Vergleich zu Patienten mit nicht-HPV-assoziierten Tumoren ein besseres Überleben nach Radiotherapie gesichert. Ziel der Untersuchung war die Identifizierung von Unterschieden in der zellulären DNA-Schadensantwort von HPV-positiven und HPV-negativen Zelllinien, wodurch die bereits in Erprobung stehende Deeskalation einer Radiotherapie bei Patienten mit HPV-assoziierten HNSCC durch experimentelle Daten abgesichert werden könnte. Material und Methoden.  Klonogenes Überleben, Induktion von Apoptose, DNA-Doppelstrang-Reparatur und Zellzyklusverhalten

wurden in vier HPV-positiven (UM-SCC-47, UM-SCC-104, 93-VU-147T, UPCI:SCC152) und vier HPV-negativen (UD-SCC-1, UM-SCC-6, UM-SCC-11b, UT-SCC-33) Kopf- und Halstumorzelllinien nach Bestrahlung untersucht. Ergebnisse.  Die höhere Strahlenempfindlichkeit HPV-assoziierter Zelllinien konnte in vitro bestätigt werden (MW SF2 HPV-positive Zelllinien: 0,198, (range: 0,22–0,18), MW SF2 HPVnegative Zelllinien: 0,34 (0,45–0,27); p = 0,010) (. Fig. 1). Durch Zellzyklusanalysen konnte gezeigt werden, dass HPV-positive Zellen nach einem DNA-Schaden die SPhaseschneller durchschreiten und DNA-Schäden in der G2/M-Phase akkumulieren (. Figs. 2 und 3). Diese abnorme Schadenskontrolle im

Zellzyklus HPV-positiver Zellen geht mit einer gesteigerten Apoptoserate und einer höheren Anzahl nicht reparierter DNA-Strangbrüche (. Fig. 5, 6) einher. Schlussfolgerung.  Das bessere strahlentherapeutische Ansprechen HPV-assoziierter Tumore könnte in der Dysregulierung des Zellzyklus und in einer verminderten Reparaturleistung begründet sein.

the membrane was probed with primary antibodies against cyclin E2, (#4132, 1:1000), cyclin A2 (clone: BF683, #4656, 1:2000, Cell Signaling), p53 (# 554294, 1:1500, BD Pharmingen), and phosphop53 Ser15 (#9284, 1:1000, Cell Signaling), and subsequently incubated with antirabbit/anti-mouse IgG HRP (horseradish peroxidase)-linked antibodies (1:5000, Millipore). Proteins were visualized using the ECL chemiluminescence detection system (Amersham).

Data analysis Statistical comparison of the cohort means of SF2, annexin V staining, subG1 DNA peak analysis, and γH2AXfoci numbers between HPV-positive and

HPV-negative cell lines were calculated using the Mann–Whitney test.

Results Clonogenic survival HPV-positive cells were more radiosensitive (.   Fig.  1a), an effect being pronounced when comparing the cohort means of the surviving fraction at 2 Gy (p = 0.010*) (SF2, .  Fig. 1b) [23]. In the group of HPV-negative cell lines, UMSCC-6 (SF2: 0.446) and UT-SCC-33 (SF2: 0.271) are the most and the least radioresistant cell lines, respectively. In the group of HPV-positive cell lines, UM-SCC-47 showed the highest (0.216) and 93-VU147T the lowest SF2 (0.177).

Keywords Head and neck cancer · HPV-related oropharyngeal cancer · Radiosensitivity · DNA damage repair

Schlüsselwörter HPV-induzierte Tumore im Kopf-HalsBereich · Strahlenempfindlichkeit · DNA Schadensreparatur

Cell cycle regulation In all cell lines, radiation increased the proportion of cells in the G2/M phase to the same extent; however, appearance and partial reversion of the G2/M arrest in HPVnegative cell lines emerged earlier as compared to HPV-positive cells (Supplementary Fig. 2). We performed a comparative grouped analysis between both cohorts, determining relative changes in cell cycle, by dividing the value for an irradiated sample at each time point by a corresponding nonirradiated control (. Fig. 2). Data for each cell line are presented in supplementary Fig 3. HPV-negative cell lines showed the S-phase decrease reaching its minimum 24 h after radiation (. Fig. 2, middle panel). In contrast, in the group of HPV-positive Strahlentherapie und Onkologie X · 2014 

| 3

Original article 0.50

1

UD-SCC-1 UM-SCC-6 UM-SCC-11b UT-SCC-33 UM-SCC-47 UM-SCC-104 93-VU-147 T 93-VU-147T UPCI:SCC152

0.40

0.1

UD-SCC-1 UM-SCC-6 UM-SCC-11b UT-SCC-33 UM-SCC-47 UM-SCC-104 93-VU-147 T 93-VU-147T UPCI:SCC152

0.01

0.001 0

a

SF2

Clonogenic Survival

0.45

1

2

3 4 Dose [Gy]

0.35 0.30 0.25 0.20

5

0.15

6

p = 0.010 HPV-negative

HPV-positive

b

Fig. 1 8 Clonogenic cell survival after exposure to increasing doses of X-rays. a HPV-positive HNSCC cell lines (open symbols) are more radiosensitive than HPV-negative cell lines (closed symbols). Data points represent the mean survival fraction of three independent experiments. b Radiation sensitivity expressed as a function of SF2 (surviving fraction at 2 Gy). Statistically significance between the cohort means of SF2 values of HPV-positive and HPV-negative cell lines was confirmed using Mann–Whitney test (p = 0.01)

cell lines, increase of S-phase cells was unimpaired relative to controls until 10 h after irradiation. Only over time did the proportion of S-phase cells also decrease. Correspondingly, the proportion of cells in G1 decreased to a higher extent in HPV-positive cells (. Fig. 2, left panel). Radiationinduced accumulation in the G2/M phase occurred later and more sustained in HPVpositive cell lines compared to HPV-negative cells (. Fig. 2, right panel). Quantification of grouped western blot analysis revealed higher S-phase promoting cyclin E2 levels 6 h after irradiation in HPV-positive cell lines than in HPVnegative cells (.   Fig. 3a). At later time points, levels of cyclin E2 declined below nonirradiated control levels in HPV-positive cell lines, whereas HPV-negative cell lines converged to baseline expression. In both groups, cyclin A2 protein expression remained high until 24 h after irradiation, displaying, on average, comparable protein levels.

Activation of p53 following radiation We used radiation-induced phosphorylation of p53 at serine 15 to investigate the functional status of p53 (.  Fig. 4). Four of the p53 mutant cell lines [UD-SCC-1, UM-SCC-11b, UT-SCC-33 (HPV-negative) and, 93-VU-147T (HPV-positive)] expressed high levels of p53 protein. UD-SCC-1 and 93-VU-147T harbor

4 |  Strahlentherapie und Onkologie X · 2014

mutated as well as wild-type p53 [21, 24], UM-SCC-11b contains a missense mutation [25]. In contrast to that, HPV-positive and wt p53 cell lines UM-SCC-104, UPCI:SCC152, and UM-SCC-47 showed only weak p53 protein levels. An increase of total p53 levels following irradiation was detectable in UD-SCC-1 only, whereas in all cell lines, except HPV-negative UM-SCC-6, activated radiation-induced p53 Ser15 phosphorylation was increased. Neither p53 nor p53 Ser15 phosphorylated protein was induced in p53 wild-type UM-SCC-6 at levels detectable by western blot, which is in accordance with the literature [26–28].

Influence of cell cycle transition defects on double strand break repair: γH2AX foci quantification We used serum deprivation in UMSCC-47 (HPV-positive) and UD-SCC-1 (HPV-negative) to arrest cells in the G0/ G1 phase. Simultaneously to 1 Gy X-rays, we added full growth medium supplemented with 5-ethynyl-2′-deoxyuridine (EdU) to measure S-phase cell cycle progression. Kinetics of residual numbers of γH2AX foci per nucleus are shown in EdU-positive (representing cells having passed G1/S transition after irradiation) and EdU-negative cells (. Fig. 5). The mean level of γH2AX foci per cell in nonirradiated HPV-positive and HPVnegative control cells was comparable

(1.4 ± 0.5 and 1.3 ± 0.3, data not shown). In both cell lines, the number of γH2AX foci in EdU-labeled cells was higher than in EdU-negative cells. Ten minutes after irradiation with 1 Gy, similar numbers of γH2AX foci per nucleus were visible in both cell lines, demonstrating equal efficiencies of foci formation. The kinetics of foci disappearance at early time points was comparable in both cell lines. Strikingly, 24 h after irradiation, a significantly higher fraction of foci persisted in UMSCC-47 cells (mean 7.5 ± 0.9, p