Recurring Amplification at 11q22.1-q22.2 Locus Plays an Important Role in Lymph Node Metastasis and Radioresistance in OSCC

A key feature in the pathogenesis of OSCC is genetic instability, which results in altered expression of genes located in amplified/deleted chromosomal regions. In a previous study we have shown that the amplification of the 11q22.1-q22.2 region, encoding cIAP1 and cIAP2, is associated with lymph node metastasis and poor clinical outcome in OSCC. Here, we validate the aCGH results by nuc ish and detect a weak amplification at the 11q22.1-q22.2 locus in 37% of the 182 samples tested. We find positive correlation of 11q22.1-q22.2 amplification with lymph node metastasis, reduced survival, and increased cancer recurrence, and we observe that patients with 11q22.1-q22.2 amplification fail to respond to radiotherapy. We confirm the concurrent overexpression of cIAP1 and cIAP2 and observe differential subcellular localization of the two proteins in OSCC. To ascertain the roles of cIAP1/cIAP2 in lymph node metastasis and radioresistance, we use an in vitro pre-clinical model and confirm the role of cIAP1 in invasion and the role of cIAP2 in invasion and migration. Studies of other tumor types in which cIAP1 is overexpressed suggest that multi-regimen treatments including SMAC mimetics may be effective. Thus, the evaluation of 11q22.1-q22.2 amplifications in OSCC patients may help choose the most effective treatment.


and disease advancement (Supplementary
SCIENTIfIC RepORtS | 7: 16051 | DOI: 10.1038/s41598-017-16247-y Association between 11q alterations and patient survival and tumor recurrence. The strong amplification of 11q22.1-q22.2 was found to be a predictor of poor clinical outcome in terms of recurrence (p = 0.043) and survival (p = 0.004). Specifically, the associated risk was higher in radiation-compliant patients (patients who have completed > 50 Gy radiation treatment) (Table 3A). Kaplan-Meier survival curves for the 11q22.1-q22.2 alterations are shown in Fig. 1B and C. Multivariate Cox proportional hazards models were used to assess the effect of multiple parameters on survival, highlighting that, in addition to node metastasis, the 11q22.1-q22.2 strong amplification is an independent predictor of poor patient survival (Table 3B).
BIRC2 and BIRC3: targets on 11q22.1-q22.2 amplicon. The 11q22.1-q22.2 amplicon includes a cluster of matrix metalloproteinase (MMP) genes and two members of the BIRC family, namely BIRC2 and BIRC3, encoding cIAP1 and cIAP2, respectively (Supplementary Figure S1C). The upregulation of MMP3, MMP7, MMP9 and their association with metastasis have been reported in various cancers 7,43,44 , and our previous analyses reported increased expression of MMP3 in OSCC samples 6,45,46 . Here, we validated this finding using semi-quantitative RT-PCR and IHC (Supplementary Figure S2). Although we observed a significant increase in MMP3 expression in OSCC samples compared to normal tissues, no significant association was observed between MMP3 expression and lymph node metastasis or patient survival (data not shown).
qRT-PCR based analysis revealed a significant increase in the expression of BIRC2 and BIRC3 in OSCCs as compared to normal tissue, and was associated with an increased risk of OSCC development (BIRC2: p = 0.001,   Figure S3). Furthermore, IHC analysis demonstrated concurrent increase in protein expression of cIAP1 and cIAP2 across the transition from normal to OSCC via leukoplakia lesions (Fig. 2C). Interestingly, partially different subcellular localizations of cIAP1 and cIAP2 were observed in both normal and tumor cells, and were further confirmed by a pathologist. cIAP1 localized to the nucleus or the cytoplasm of tumor cells, whereas cIAP2 was seen in the membrane or in the cytoplasm of tumor cells, as depicted in Fig. 3A (a-d). In addition, based on correlation analysis, we found that the probabilities of nuclear cIAP1 and membranous cIAP2 protein expression were higher in normal and leukoplakia samples than in cancerous samples (Fig. 3B). Higher membranous cIAP2 protein expression was observed in well differentiated tumors, as well as in lymph node metastasis negative tumors (Supplementary Table S3). Moreover, cytoplasmic cIAP2 expression strongly correlated with node metastasis (Supplementary Table S3). Polytomous logistic regression with normal as the reference group showed a significant correlation between cIAP1 and cIAP2 cytoplasmic overexpression and the risk of developing OSCC (Supplementary Table S4). Overall, cytoplasmic overexpression of cIAP1 and cIAP2 was positively correlated with progression from normal tissue to OSCC. cIAP1 and cIAP2 expression in an OSCC cell line. We examined both the expression and the localization of cIAP1 and cIAP2 in the buccal mucosa derived OSCC cell line UPCI:SCC029B (SCC29B), which has been reported to be aneuploid for chromosome 11 47 . We observed the upregulation of BIRC2 and BIRC3 in SCC29B when compared to normal gingivobuccal tissue (Fig. 3C). Comparable expression and similar localization of cIAP1 (nuclear and cytoplasmic) and cIAP2 (membranous and cytoplasmic) were observed in SCC29B, as seen in OSCC tumor samples ( Fig. 3A (g,h), C and D). To understand the functional relevance of both these targets in SCC29B, we performed separate shRNA-mediated knockdowns of BIRC2 and BIRC3. As SCC29B is a cell line which is difficult to transfect, we successfully generated three stable lines using only pooled populations (rather than single population cloning) for shcIAP1-1, shcIAP2-1, and pLKO.1-EGFP-f-puro vector (Vec). The observed transfection efficiency for the control Vec is shown in Supplementary Figure

BIRC3 promotes increased migration and invasion in OSCC.
To examine the effect of cIAP1 and cIAP2 knockdown on cell migration, we performed scratch wound healing assays. We found that cell migration is decreased when cIAP2 is knocked down in SCC29B cells, but no effect was observed with cIAP1 knockdown ( Fig. 4F and G). Furthermore, matrigel invasion assays showed a decrease in the invasive potential of cells upon either cIAP1 or cIAP2 knockdown ( Fig. 4H and I), without any change in cell proliferation (Fig. 4E). These results suggest that cIAP2 is required for increased migration and invasion, while cIAP1 is required only for increased invasion in SCC29B.
The effect of cIAP1 and cIAP2 on radioresistance. Clinical data showed that 27.9% of the analyzed OSCC patients had poor survival and 40.2% had recurrence, despite complete treatment, i.e., surgery followed by postoperative radiotherapy. Hence, we were interested in understanding the association of cIAP1 and cIAP2 with radioresistance in oral cancers. First, we quantified the expression of both cIAP1 and cIAP2 at the mRNA and protein levels with qRT-PCR and Western blotting and assessed the survival of SCC29B at different radiation doses ( Fig. 5A-D, Supplementary Figure S5A). We observed an insignificant increase in the expression of BIRC2 and BIRC3 with increasing radiation dosage (Fig. 5B-D). Next, we assessed the effects of cIAP1 or cIAP2 knockdown  on radiation response using colony formation assay (Supplementary Figure S5B) and observed no difference in survival of either cIAP1 or cIAP2 knockdown cells compared to vector (Fig. 5E). We speculated that this was a consequence of the increased cIAP1 and cIAP2 expression post irradiation, which was further confirmed by analyzing cIAP1 and cIAP2 expression levels in respective knockdowns at 6 Gy [ Fig. 5F-H]. Specifically, we observed an increase in the expression of cIAP1 and cIAP2 in cells irradiated at 6 Gy, as compared to 0 Gy control, implying that the knockdown system was not sufficient to compensate for the radiation-induced overexpression.

Discussion
Multiple aberrations on chromosome 11 have been strongly correlated with progression, metastasis, and radioresistance in HNSCC 8-10,50-53 . In the present study, we validated and confirmed the association of the 11q13.3 amplification with disease progression. Our data shows a strong correlation between the 11q22.2 amplification and lymph node metastasis and poor survival in HPV negative OSCC patients. Using qRT-PCR and IHC, we demonstrated that the overexpression of BIRC2 (cIAP1) and BIRC3 (cIAP2), both located on 11q22.1-q22.2, was associated with lymph node metastasis in OSCC patients. Consistent with other reports, our study also highlights differences in localization of cIAP1 and cIAP2 in oral tumor tissues. cIAP1 is a tumor necrosis factor receptor (TRAF2)-related protein and localizes either to the nucleus or to the cytosol and shuttles between them [54][55][56] . When in the cytoplasm, it activates caspases in response to the apoptotic signal; while in the nucleus, it regulates cell cycle and cell proliferation 57 . cIAP2 has been reported to localize to the cytosol, peri-nucleus, and mitochondria 58,59 , and this study is the first to report the membranous localization of cIAP2 in OSCC. Cytoplasmic localization of the paralogous protein XIAP has been associated with increased cell survival and reduced apoptosis 60 , which is consistent with the notion that the increased cytoplasmic cIAP1 and cIAP2 expression that we observed in OSCC may promote tumor survival. The mechanisms controlling the distribution of cIAPs into various subcellular locations and their functional relevance in particular locations require further investigation 55 .
The expression of cIAP1/cIAP2 has been correlated with invasion in esophageal, cervical, and gall bladder cancers [61][62][63] . The current study demonstrates the role of cIAP1 and cIAP2 in invasion or migration in a buccal mucosa-derived OSCC cell line, further confirming the role of these proteins in lymph node metastasis in oral cancers. However, additional studies are required to validate the present cIAP1 and cIAP2 findings in additional OSCC and HNSCC cell lines before implicating these biomarkers in outcome prediction. The overexpression of cIAP1 can lead to genomic alterations due to defects in cell division and is a key regulator of cell proliferation and apoptosis 55 . Similarly, cIAP2 is also involved in regulating apoptotic signals, and the knockout of cIAP1 has been demonstrated to increase cIAP2 expression to compensate for cIAP1 loss, indicating that each of cIAP1 and cIAP2 may be redundant if the other one is functional 11 . Further studies might assess the synergistic effects of cIAP1 and cIAP2 on invasion and metastasis.
Previous reports support the fact that the overexpression of cIAP1/2 proteins may be associated with an unfavorable prognosis after radiotherapy or chemotherapy 14,40,54,64 . Here, we observed an increase in cIAP1 and cIAP2 expression in radiation-treated cells, as compared to their respective controls. This result is consistent with the study by Wang et al. reporting an increased expression of BIRC3 in response to radiotherapy and TMZ treatment in glioblastoma 12 . The knockdown system may not be sufficient to demonstrate the effect of cIAP1 and cIAP2 on radiation response. Further studies are needed to better understand their role on radiation resistance in SCC29B, as well as in other OSCC cell lines, using knockout systems. Since therapy resistance is an important factor in determining patient outcome, the use of SMAC mimetics that inhibit IAPs could improve the efficacy of standard therapy and may emerge as a potential therapy for various cancers including OSCC 14,25,38 .
The current study highlights the importance of 11q22 locus amplification as a clinically relevant marker for predicting lymph node metastasis and poor clinical outcome, i.e., increased recurrence and reduced survival. We also report a correlation between the 11q22.1-22.2 amplification and poor radiation response, which could be further explored to modulate treatment regime for oral cancer patients. For the first time, we demonstrate the localization of cIAP2 to the cell membrane. The cytoplasmic localization of cIAP1 and cIAP2 was frequently observed in tumors, while nuclear cIAP1 and membranous cIAP2 were predominantly observed in normal and well differentiated tumors. Further investigations are needed to understand the molecular mechanisms underlying differential localization and therapeutic potential of cIAP1 and cIAP2 in oral cancers.

Methods
Tissue specimen collection. The study was approved by the Institutional Local Ethics Committee of Tata Memorial Hospital (TMH) and Nair Hospital Dental College, Mumbai, India. Paraffin blocks or frozen tissue samples of neo-primary oral tumor tissues and pre-invasive lesions (leukoplakia) were recruited from ICMR National Tumor Tissue Repository, Tata Memorial Hospital. Non-inflamed gingivobuccal mucosa tissues from clinically healthy individuals with no previous personal history of cancer were obtained from Nair Hospital. The methods were carried out in accordance with the approved guidelines and regulations. Written informed consent was obtained from all the study participants.
Cell line and plasmid constructs. The oral squamous cell carcinoma cell line UPCI:SCC029B (SCC29B), derived from human buccal mucosa cancer, was procured from Dr. Susanne M. Gollin, The University of Pittsburgh, USA 65 . SCC29B cells were cultured in M10 media, as recommended by the supplier. The plasmids containing shRNAs against cIAP1 and cIAP2 were a gift from William Hahn (Addgene plasmid) 66 ; the specific plasmids we used are denoted here as shcIAP1-1 and shcIAP2-1. The pLKO.1-EGFP-f-puro vector without any shRNA was used as vector control (denoted here as Vec). The sequences for the cIAP1 and cIAP2 shRNAs are provided in Supplementary Table S5. The plasmids were purified on a cesium gradient and the presence of the shRNA insert was confirmed by sequencing.

Real-Time PCR (qRT-PCR) Analysis.
A total of 10 µl RNA (1.5 µg) were converted to cDNA using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, USA), as described by the manufacturer. Following conversion, 6 ng of cDNA were used for TaqMan qRT-PCR analysis using fluorescent TaqMan probes obtained from Applied Biosystems for BIRC2 (Hs01112284_m1), BIRC3 (Hs00985031_g1), and RNA18S5 (Hs99999901_s1) as a control. All the experiments were performed in duplicate and the results were analyzed using QuantStudio 12 K Flex software v1.2.2 (Applied Biosystems, USA), as described previously 6 . Figure 5. cIAP1 or cIAP2 knockdown had no effect on radioresistance. SCC29B cells were γ-irradiated with increasing dose, cells were harvested 24 hours post radiation, and expression of BIRC2 and BIRC3 were analyzed by (A and B) qRT-PCR and (C) Western blotting, as well as (D) clonogenic survival assay to calculate median lethal dose (LD50) of SCC29B. (E) The survival fraction of Vec and knockdowns was evaluated by clonogenic assay post γ-irradiation. (F-G) Protein expression of cIAP1 and cIAP2 was analyzed by Western blot and Immunocytochemistry (ICC) original magnification 630X at different radiation doses in Vec, shcIAP1-1 and shcIAP2-1 cells. Cropped blots are displayed (full-length blots and the histogram plots of the protein expressions after normalization with β-actin are presented in Supplementary Figure S6). The mean ± standard deviations of three independent experiments are plotted.
Interphase/Nuclear in-situ Hybridization (nuc ish). Nuc ish was performed using 1) BAC probes (BACPAC resource center, Children's Hospital Oakland Research Institute, USA) for the 11q13.3 locus (RP11-300I6) and the Chr11 centromere (CEP) (RP11-135H8) and 2) SureFISH probes (Agilent Technologies, USA) for the 11q22.1-11q22.2 locus (G101223G-8) and the Chr11 CEP (G101083R-8). The specificities of all the BAC-derived probes and sureFISH probes were confirmed on metaphase target slides (Vysis, CA, US) before hybridizations. Visualization and enumeration of nuc ish signals was done as described previously 6 . A weak amplification was defined as having the gene probe counts between 3 and 8. Strong amplification was defined as having either gene probe counts greater than 8 or innumerable clusters for the gene locus. For statistical analyses, two types of comparisons were done: 1) no change (diploid nuclei) vs. change (includes cases with either weak amplification or with strong amplification); 2) no change vs. strong amplification (excluding cases with weak amplification). Immunohistochemistry (IHC). Immunohistochemical staining, grading, and H-Score analysis were performed as described previously 6 . In this study, IHC was performed for cIAP1 and cIAP2 using Vectastain Universal Elite ABC Kit (Vector Labs, USA) or Dako EnVision ™ FLEX Mini Kit, High pH kit (DAKO, Agilent Technologies, Denmark). Details on antibodies and heat-based antigen retrieval are provided in Supplementary  Table S6.

Immunofluorescence (IF) and Immunocytochemistry (ICC). Immunofluorescence analysis was
performed on tissue samples and SCC29B cells. For tissue preprocessing and antigen retrieval, the steps were performed as described for IHC, followed by tissue fixing in cold methanol for 10 minutes, followed by blocking. Tissues were next incubated for 16 hours at 4 °C with the respective antibody at a dilution provided in Supplementary Table S6. On the following day, the slides were incubated with an Alexa Fluor 488 anti-rabbit secondary antibody (Life Technologies, USA) at 1:200 dilutions, in a humid chamber for 1 hour at room temperature.
For ICC, SCC29B cells were grown on a coverslip; after the cells reached 80% confluency, the cells were fixed in 4% paraformaldehyde and then permeabilized using 0.3% Triton-X100 for 20 minutes at room temperature. Following blocking, the cells were treated with primary antibodies inside a humidified chamber for 16 hours (overnight) at 4 °C. On the following day, the cells were treated with the Alexa 488 or Alexa 568 conjugated secondary antibody anti rabbit IgG (Life Technologies, USA) at a dilution of 1:100 and incubated for 1 hour at room temperature in a humidifying container. Both the tissues and the cells were treated with nuclear counterstain 4′,6-Diamidine-2′-phenylindole dihydrochloride (DAPI) and then mounted using the Vectashield mounting agent (Vector Laboratories, USA). Confocal images were obtained by using an LSM 780 Carl Zeiss Confocal system with an Argon 488 nm, 568 nm, and 405 nm lasers (Carl Zeiss, Germany), and image analysis was performed using the LSM image browser.
Transfection of SCC29B cells. Transfection was done in a 12-well plate with 1 μg of plasmid DNA, using Lipofectamine ™ 3000 Reagent (Invitrogen, USA), and carried out as per manufacturer's protocol, at 70-90% cell confluency. Stable lines including each of the two shRNAs to knock down cIAP1 and cIAP2 were generated by using a pool of all cell populations that survived puromycin selection (0.75 μg/ml).
Western blotting. For Western blots, 50 μg of the whole cell lysate was resolved on 10% SDS-PAGE gel and transferred to PVDF membranes (Amersham Hybondtm-P, GE Healthcare, USA), followed by Western blotting with antibodies for cIAP1 and cIAP2 (antibody details are provided in Supplementary Table S6). The blots were developed using Amersham ECL Prime Western Blotting Detection Reagent (GE Healthcare, USA), according to the manufacturer's instructions and quantified with Image J software.
Scratch wound healing assays. Cells were grown to 90% confluency in a 6-well plate, followed by treatment with 10 μg/ml of mitomycin C (Sigma-Aldrich, USA) for 3 hours. Subsequently, the cells were washed and a linear scratch wound was made in each well. The cells were maintained at 37 °C and 5% CO 2 on an Axiovert 200 M Inverted microscope (Carl Zeiss, Germany) with a cell incubator stage. Cells were observed by time lapse microscopy and images were taken every 10 minutes for 20 hours, using a AxioCamMRm Camera (Carl Zeiss, Germany) with a 10X phase I objective.
Matrigel cell invasion assay. Matrigel invasion assays were performed in a 24-well plate using 2 × 10 5 cells re-suspended in 200 μl of serum-free media. These cells were added to the upper chambers and 600 μl of conditioned media (1:1 ratio of conditioned and fresh serum containing media) were added in the lower chamber. The inner side of the insert with 0.8 mm membrane (BD Falcon, USA) was pre-coated with 15 μl of Matrigel (Corning, USA). After 24 hours, the inserts with cells on the outer side of the membrane were fixed with 4% para-formaldehyde, stained with 1% crystal violet (Sigma-Aldrich, USA), and mounted on slides using D.P.X mounting reagent (Merck, USA). Images were taken using an Olympus SZ61 stereo microscope with a 10X objective.
Cell proliferation assay. An MTT-based assay was used to measure cell proliferation in 96-well plates on 5000 cells for eight days. At 24 hour intervals, cells were treated with 20 µl MTT (3-(4, 5-dimethylthiazolyl-2)-2, 5-diphenyltetrazolium bromide) (Sigma-Aldrich, USA) for 4 hours at 37 °C in a CO 2 incubator. The reaction was stopped using 100 µl stop solution (10% SDS in 0.01 N HCL), followed by overnight (16 hours) incubation at 37 °C. This assay measures cell metabolic activity by comparing optical densities between the end and the start of the experiment. Absorbance was read at 530 nm and 690 nm using a microplate reader, SPECTROstar Nano (BMG Labtech, Germany).
SCIENTIfIC RepORtS | 7: 16051 | DOI:10.1038/s41598-017-16247-y Clonogenic cell survival assay. Clonogenic survival was determined at radiation doses of 0, 2, 4, 6, and 8 Gy for the original SCC29B and 0, 2 and 6 Gy for the three derived clones with a control vector (Vec) plasmid, a cIAP1 knockdown plasmid, and a cIAP2 knockdown plasmid. Cells from the exponential growth phase were γ-irradiated using 60 Co-γ Linear Accelerator (Bhabhatron-2, ACTREC, Tata Memorial Centre), and then seeded in triplicate in 60 mm plates. Colonies were allowed to grow for 10 days before being stained with 0.05% crystal violet. The survival fraction at a given dose was calculated as the number of colonies formed divided by the product of the number of cells plated and plating efficiency.
Statistical analyses. The statistical analyses were performed using IBM SPSS version 21. Chi-square tests or Spearman correlation tests were used to determine the correlation between protein expression levels or locus amplification and disease progression, as well as clinicopathological characteristics such as lymph node metastasis, tumor stage, and grade. Multinomial logistic regression was used to evaluate the relationship between protein expression scores and the risk of leukoplakia and OSCC development, with normal tissue as a reference; odds ratios (OR) were computed by adjusting for age and gender 67,68 . Disease-specific survival (DSS) and recurrence free survival (RFS) were examined visually with Kaplan-Meier curves and analyzed by log rank tests. Binomial logistic regression was performed to analyze the dependence of 11q13.3 and 11q22.1-q22.2 amplifications on lymph node metastasis using either 1) cases with both loci altered, 2) cases with only 11q13.3 altered and no change at 11q22.1-q22.2, or 3) cases with only 11q22.1-q22.2 altered and no change at 11q13.3. Multivariate Cox regression was performed to identify the degree of association between various parameters such as lymph node metastasis, tumor stage, grade, and clinical outcome. All in vitro assays were performed in triplicate. For cell line experiments, paired t-tests were used to analyze the relationships between vector control and cIPA1 and cIAP2 knockdown clones. All p-values < 0.05 were considered statistically significant.