Nuclear survivin is associated with cell proliferative advantage in uterine cervical carcinomas during radiation therapy
- 1Department of Pathology, Kitasato University School of Medicine, Kanagawa, Japan
- 2Department of Obstetrics and Gynecology, Kitasato University School of Medicine, Kanagawa, Japan
- Correspondence to Professor Makoto Saegusa, Department of Pathology, Kitasato University School of Medicine, 1-15-1 Kitasato, Minami-ku, Sagamihara, Kanagawa 252-0374, Japan;
Contributors PC and MS: data collection, statistical analysis, writing, and discussion. SK: sample collection, cell culture, and discussion. MH: sample collection and immunohistochemical examination.
- Accepted 3 January 2012
- Published Online First 3 March 2012
Background Although the anticancer effects of radiation therapy for patients with uterine cervical squamous cell carcinoma (U-SCC) are widely acknowledged, little is known about the resultant morphological alterations in tumour tissue kinetics.
Aims To make a detailed assessment of possible roles of survivin expression in apoptosis and cell proliferation in U-SCC during radiation therapy.
Methods 181 biopsy specimens from 55 consecutive U-SCCs of patients receiving radiation therapy were studied using a combined morphological (apoptosis) and immunohistochemical (MIB-1 and survivin) approach. The intracellular distribution of various splice variants of the survivin gene was also examined.
Results Tumour cell proliferation, determined as MIB-1 labelling indices (LIs), as well as nuclear survivin (N-Surv) LIs, were inversely correlated with irradiation dosage, in contrast to relatively minor changes in apoptotic indices, suggesting a shift in tumour tissue kinetics towards a relative predominance of cell deletion. In addition, the low N-Sur LI category showed significant stepwise decrease in MIB-1 LIs during therapy, in contrast to no changes in the high category. Exogenous overexpression of three variants of the survivin gene resulted in different expression patterns, showing cytoplasmic staining with or without dot formation for survivin and survivin-2B and distinct nuclear accumulation for survivin-deled exon 3 (∆Ex3).
Conclusions Results showed that nuclear survivin, including survivin itself and the survivin-∆Ex3 splice variants, may participate in modulation of altered cell kinetics of U-SCC during radiation therapy.
Uterine cervical squamous cell carcinoma (U-SCC) is still a problematic disease in under-developed and developing countries. Although there are many effective screening methods, including human papillomavirus (HPV) DNA testing, the incidence is still high. In GLOBOCAN 2008, it was the second most common cancer in women, with an age-standardised rate (ASR) (per 100 000 person-years) of 17.8, and about half of the patients (ASRs=9.8) dying from the disease.1
Radiation is the major therapy for advanced U-SCC, triggering apoptosis of tumour cells, but failure can occur due to resistance.2 Various techniques have been developed for detection of apoptotic cells,3 but simple examination using H&E-stained sections is the easiest and most reproducible method. In addition, analysis of cell proliferation is important to predict response to treatment and prognosis.4 5
Survivin is a member of the inhibitor of apoptosis proteins family, with expression in the nuclei, cytoplasm or both.6 The molecule is found in many fetal, and some proliferating but not resting adult tissues.7 In uterine cervical tissues, cytoplasmic survivin is important in normal squamous differentiation, while the nuclear forms are involved in HPV-mediated disruption of normal cellular maturation.8
Although the anticancer effects of irradiation for U-SCC are widely acknowledged, little is known about resultant morphological alterations in tumour tissue kinetics. In this study, to clarify functional roles of survivin in U-SCC during radiation therapy, expression was immunohistochemically investigated, and compared with the status of apoptosis and cell proliferation.
Materials and methods
A total of 181 biopsy specimens (4–7 sections per case) from 55 consecutive patients with stage IIb–IVa U-SCC, according to the patient records at Kitasato University Hospital, were selected during the periods 1994–1998 and 2005–2006, according to the criteria of the FIGO system. Mean age of the patients was 58.4 (range 28–84) years. Uterine cervical tissues were taken before initiation of treatment and at each time of insertion of instruments for brachytherapy to evaluate changes in cancer cells. Small and insufficient tumour tissues were excluded from the study. All samples were routinely fixed in 10% formalin, and embedded in paraffin. In addition, five U-SCC tissue samples were snap-frozen in liquid nitrogen for mRNA analysis.
The surveillance after completion of treatment was at 3-monthly intervals in the first 2 years, at 4-monthly intervals in the third year, every 6 months in the fourth and fifth years, and thereafter once a year. Mean survival time was 60.2 (range 2.5–149.3) months. At the time of this analysis, 17 patients had died from the disease and 38 patients were still alive without apparent disease.
Definitive radiotherapy with conventional techniques was given to all patients using a linear accelerator (LINAC) device with 10-MV x-rays for external beam radiotherapy (EBRT) and a remote after loading system for brachytherapy, being scheduled to total doses of 50.4 Gy combined with remote after loading system brachytherapy started after 15 fractions of EBRT.
Immunohistochemistry and TdT-mediated dUTP-biotin nick end labelling assays
Immunohistochemistry (IHC) was performed using a combination of microwave-oven heating and polymer immunocomplex (Envision; Dako, Copenhagen, Denmark) methods. Anti-Ki-67 antigen (MIB-1) mouse monoclonal (Dako), anti-survivin rabbit polyclonal (R&D Systems, Minneapolis, Minnesota, USA) and anti-single-stranded (ss) DNA polyclonal antibodies (Dako) were applied. As a positive control, a case of primary antibody-positive colon cancer was used, while rabbit or mouse sera were applied as negative controls.
For evaluation of IHC findings, nuclei immunopositive for MIB-1 and survivin were counted in at least 500 cells, and labelling indices (LIs) were then calculated as percentages of immunopositive cells in the total number of tumour cells counted, under high power (×40 object and ×10 ocular lenses). On the basis of the percentage of cytoplasmic survivin-immunopositive cells, the lesions were also subdivided into four categories: 0, no or <5% positive cells; 1+, 5–10%; 2+, 10–30%; 3+, >30%. Cases were defined as immunopositive when over 5% of the cells were stained in each section, independent of the nuclear status.
TUNEL (TdT-mediated dUTP-biotin nick end labelling) assays were carried out using an in situ cell death detection kit (Roche, Mannheim, Germany), according to the manufacturer's instructions.
Assessment of radiation-therapeutic efficacy and apoptosis
Sections were evaluated for radiation-therapeutic efficacy (R-TE), using the criteria of Shimosato's histological classification with minor modifications,9 as follows: grade (G)1, absence or minimal cell damage; G2, the presence of considerable cell damage, including G2A, viable cells present over a quarter of the tumour area, and G2B, less than a quarter; G3, only non-viable cells; and G4, no tumour elements. Biopsy specimens diagnosed as G3 and G4 stages were excluded from the study, because of limited tumour components in the sections.
Apoptotic cells were identified in H&E-stained sections, according to the criteria of Kerr et al.10 At least 500 cancer cells per slide in areas remote from necrosis and fibrosis were counted and calculated for apoptotic indices (AIs), in a manner similar to the analysis of LIs. Positive cells for TUNEL and ssDNA IHC assays in 30 samples of U-SCC were also examined, using the same approach.
PCR and immunofluorescence for the survivin gene
Total cellular RNAs were extracted from two endometrial carcinoma cells, HeLa cells, and five frozen U-SCC samples. Full length cDNA of the survivin gene cDNAs were amplified by PCR using specific primers.11 The PCR products were subcloned into pcDNA3.1 (Invitrogen, Carlsbad, California, USA). The identity of all constructs was confirmed by sequencing prior to use. The GAPDH gene was amplified as an internal control.12
After transfection of survivin-expression vector into Ishikawa cells, using LipofectAMINE PLUS (Invitrogen), cells were incubated with anti-survivin antibodies. FITC-labelled anti-rabbit IgG (Molecular Probes, Leiden, The Netherlands) was used as the secondary antibody.
Comparative data were analysed using Pearson's correlation coefficient, the χ2 test, and the Mann–Whitney U test. The cut-off for statistical significance was set as p<0.05. The prognostic significance of parameters, categorised on the basis of mean LIs or immunopositivity in before-treatment specimens, was assessed by the log-rank test, and survival was measured from time of initial biopsy for U-SCC.
Changes in apoptosis, cell proliferation and survivin expression in U-SCC during radiation therapy
With a total dosage of irradiation at over 30 Gy, most of the samples were categorised into G2 stage, demonstrating marked enlargement of nuclei with or without increase of eosinophilia in the cytoplasm, in contrast to G1 stage with the minimal alterations at doses less than 15 Gy (figure 1). In addition, most samples with over 40–50 Gy were categorised into G3 and G4 stages.
As shown in figure 2A, apoptotic cells were readily detectable in H&E-stained sections, on the basis of characteristic features,13 14 the AIs being positively correlated with those detected by TUNEL and ssDNA IHC assays (figure 2B). Immunoreactivity for survivin was observed in both or either of the nuclear and cytoplasmic compartments of tumour cells, as well as in mitotic cells, in contrast to only nuclear MIB-1 immunostaining (figure 3).
The irradiation dose was inversely correlated with LIs of MIB-1, as well as nuclear survivin (N-Surv), but there was no relation with AIs (figure 4). No significant differences in AIs were observed across R-TE stages, while MIB-1 LIs showed a significant stepwise decrease from before treatment (BT) to G2B stages. N-Surv LIs, as well as the cytoplasmic positivity (Cyto-Surv), were also significantly decreased in G2B as compared to BT and G1 stages (figure 5A and table 1). In addition, N-Surv LIs showed a strong positive correlation with MIB-1 LIs, and a weak one with AIs, but such an association was not evident between AIs and MIB-1 LIs (figure 5B). There were no significant differences in N-Surv LIs for each R-TE stage between Cyto-Surv-positive and -negative categories (figure 6).
Relationships of survivin expression with apoptosis and cell proliferation
To examine the relationships of survivin expression with apoptosis, cell proliferation and survival during radiation therapy, cases were subdivided into categories, with the average N-Surv LI as cut-off (mean±SD: 26.5±12.3%) and cytoplasmic positivity, as well as AI (2.2±1.9%) and MIB-1 LIs (51.4±16.7%), in BT sections. Average AIs appeared to be increased in G2A as compared to BT in the high N-Surv LI category, while MIB-1 LIs showed significantly stepwise decreases from BT to G2B stages in the low category. Such associations were not observed in cases of both positive and negative Cyto-Surv categories (figure 7). In addition, both nuclear and cytoplasmic survivin status, as well as AIs and MIB-1 LIs, did not have any prognostic value (data not shown).
Splice variants of the survivin gene
Knowing the presence of several splice variants for survivin gene,11 we examined associations between the variants and intracellular localisation. As shown in figure 8A, three different variants, survivin, survivin-deled exon 3 (∆Ex3) and survivin-2B isoforms, were assessed in two endometrial carcinomas and HeLa cells, as well as three of five U-SCC cases.
Exogenous overexpression of the three variants resulted in different expression patterns, showing cytoplasmic staining with or without dot formation for survivin and survivin-2B, in contrast to survivin-∆Ex3 with distinct nuclear accumulation, along with several dots (figure 8B).
The present study clearly demonstrated dramatic changes in morphology of U-SCCs of patients receiving over 30 Gy, in contrast to relatively minor events at doses less than 15 Gy. Moreover, differences with the dosage were not observed between G2A and G2B stages, indicating the possible existence of a critical threshold for triggering effects between 15 and 30 Gy. In addition, most samples receiving over 40 Gy were categorised into G3 and G4 stages, but it was difficult to conduct further examinations due to limited or lacking residual tumour components.
Although survivin is considered to serve as a regulator for the apoptotic process through inhibition of procaspase-3 and -7 and activation of caspase-915–17 our results demonstrated a very weak correlation between the two during radiation therapy. One possible reason for the anomalous results may include the presence of HPV in U-SCC cells affecting apoptosis susceptibility, since cells expressing the E6 and E7 viral proteins fail to undergo apoptosis following DNA damage through alteration in function of p53 or the retinoblastoma gene product.18–20 In addition, the findings of relatively minor changes in apoptosis, in contrast to significant decreases in cell proliferation during therapy, allow us to speculate a shift in tumour tissue kinetics towards a relative predominance of cell deletion.
Interestingly, nuclear survivin expression was positively correlated with tumour cell proliferation, in line with the reports that expression peaks in the G2/M phase, followed by a rapid decline in the G1 phase through an active transport mechanism between the nucleus and cytoplasm.21–23 Moreover, cells exogenously overexpressing survivin and survivin-∆Ex3 variants showed dot formation in both nuclear and cytoplasmic compartments, in line with evidence that they are able to interact with mitotic spindle microtubules, centrosomes and the mid-bodies of dividing cells.15 16 With regard to lack of a direct association between nuclei and cytoplasm during radiation therapy, it appears that the two pools may be independent modulators of tumour kinetics.
It is widely known that alternative splice variants of survivin gene, survivin, survivin-2B and survivin-∆Ex3, may be important for fine tuning of its action, demonstrating that antiapoptotic survivin and survivin-∆Ex3 are counterbalanced by the non-antiapoptotic survivin-2B variant.24 25 In our in vitro assays, exogenous overexpression of survivin-∆Ex3 exhibited a preferential localisation in the nuclear compartment, in contrast to the predominantly cytoplasmic distribution observed for survivin and survivin-2B in cell lines. Moreover, low nuclear survivin expression was significantly associated with a significant stepwise decrease in MIB-1 LIs during therapy, in contrast to no change in the high expression category, as well as the cytoplasmic status. Given the fact of functional antagonism between cytoplasmic survivin and survivin-2B variants through competitive binding of heterologous interaction partners,26 27 the results suggest that nuclear survivin, including both survivin itself and the ∆Ex3 splice variants, may play an important role in the maintenance of U-SCC tumour cell viability during radiation therapy.
In conclusion, the present study clearly demonstrated that nuclear survivin may participate in modulation of altered cell kinetics of U-SCC during radiation therapy.
Uterine cervical squamous cell carcinoma (U-SCC) cell proliferation and nuclear survivin expression were inversely correlated to irradiation dosage.
The low nuclear survivin expression group showed significant stepwise decrease in cell proliferation during radiation therapy, in contrast to no changes in the high nuclear survivin expression group.
Nuclear survivin may participate in modulation of altered U-SCC cell kinetics during radiation therapy.
The authors thank Dr Miwa Kawaguchi for her contribution to this study.
Competing interests None.
Ethics approval This study was conducted with the approval of the local research ethics committee.
Provenance and peer review Not commissioned; externally peer reviewed.