Journal of Pathology J Pathol 2014; 234: 239–252 Published online 30 July 2014 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/path.4390
ORIGINAL PAPER
KPNA2 is overexpressed in human and mouse endometrial cancers and promotes cellular proliferation Kristian Ikenberg,1† Nadejda Valtcheva,1† Simone Brandt,1 Qing Zhong,1 Christine E Wong,1 Aurelia Noske,1 Markus Rechsteiner,1 Jan H Rueschoff,1 Rosmarie Caduff,1 Athanassios Dellas,2 Ellen Obermann,3 Daniel Fink,4 Thomas Fuchs,5 Wilhelm Krek,6 Holger Moch,1 Ian J Frew7 and Peter J Wild1* 1 2 3 4 5 6 7
Institute of Surgical Pathology, University Hospital Zurich, Zurich, Switzerland Krebsliga beider Basel, Basel, Switzerland Institute of Pathology, University Hospital of Basel, Basel, Switzerland Department of Gynecology, University Hospital Zurich, Zurich, Switzerland Department of Electrical Engineering, California Institute of Technology, Pasadena, CA, USA Institute of Cell Biology, ETH Zurich, Zurich, Switzerland Zurich Center for Integrative Human Physiology, University of Zurich, Zurich, Switzerland
*Correspondence to: Peter J Wild, Institute of Surgical Pathology, Schmelzbergstrasse 12, 8091 Zurich, Switzerland. e-mail: [email protected] †
These authors contributed equally to this work.
Abstract Endometrial cancer is the most frequently occurring malignancy of the female genital tract in Western countries. Although in many cases surgically curable, about 30% of the tumours represent an aggressive and untreatable disease. In an attempt to establish a reliable prognostic marker for endometrial carcinomas disregarding their histological diversity, we investigated the expression of KPNA2, a mediator of nucleocytoplasmic transport, and other cell proliferation-associated proteins and their correlation with cancer progression. We analysed patient tissue microarrays (TMAs) assembled from 527 endometrial cancer tissue specimens and uterus samples from a Trp53 knockout mouse model of endometrial cancer. Our data show that KPNA2 expression was significantly up-regulated in human endometrial carcinomas and associated with higher tumour grade (p = 0.026), higher FIGO stage (p = 0.027), p53 overexpression (p < 0.001), activation of the PI3K/AKT pathway, and epithelial–mesenchymal transition. Increased nuclear KPNA2 immunoreactivity was identified as a novel predictor of overall survival, independent of well-established prognostic factors in Cox regression analyses (hazard ratio 1.7, 95% CI 1.13–2.56, p = 0.01). No significant association between KPNA2 expression and endometrial cancer subtype was detected. In the mouse model, KPNA2 showed increased expression levels from precancerous (EmgD, EIC) to far-advanced invasive lesions. We further investigated the cell proliferation capacity after siRNA-mediated KPNA2 knockdown in the human endometrial cancer cell line MFE-296. KPNA2 silencing led to decreased proliferation of the cancer cells, suggesting interplay of the protein with the cell cycle. Taken together, increased expression of KPNA2 is an independent prognostic marker for poor survival. The mechanism of enhanced nucleocytoplasmic transport by KPNA2 overexpression seems a common event in aggressive cancers since we have shown a significant correlation of KPNA2 expression and tumour aggressiveness in a large variety of other solid tumour entities. Introducing KPNA2 immunohistochemistry in routine diagnostics may allow for the identification of patients who need more aggressive treatment regimens. Copyright © 2014 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.
Keywords: KPNA2; importin; endometrial cancer; biomarker; EMT; Snail
Received 16 April 2014; Revised 2 June 2014; Accepted 10 June 2014
No conflicts of interest were declared.
Introduction Endometrial carcinomas of the uterus are the most common tumour type in the Western world that develops in the female genital tract, with over 40 000 new cases and more than 7000 related deaths per year in the United States [1]. Well-established prognostic factors for endometrial carcinoma are tumour grade, Copyright © 2014 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
FIGO stage, and histological subtype [2–4], but these pathological parameters only insufficiently predict patient outcome for currently available therapies. The most frequent form of these tumours is the so-called type I endometrioid carcinomas, which typically have a relatively good prognosis. The rarer but more aggressive type II endometrial carcinomas, comprising serous carcinomas, clear cell carcinomas, and undifferentiated J Pathol 2014; 234: 239–252 www.thejournalofpathology.com
240
carcinomas, account for a disproportionate number of endometrial cancer-associated deaths [3]. The molecular basis of these different forms of cancer remains not completely understood. As most women present early in the course of the disease, the majority of patients have an excellent prognosis after surgery. For patients with advanced metastatic or recurrent endometrial cancer, however, the available therapeutic options are limited. A meta-analysis that evaluated the treatment outcome of advanced endometrial carcinomas by investigating symptom control and impact on the quality of life after cytotoxic chemotherapy showed disappointing results [5]. Reliable biomarkers that are based on the molecular changes of these aggressive tumours are an unmet clinical need and could be potential therapeutic targets. Besides the well-established prognostic parameters, a broad diversity of novel molecular markers has already been evaluated for predictive and prognostic relevance in patients with endometrial carcinomas [6–8]; however, none of them has entered the diagnostic routine yet. We have recently described karyopherin alpha 2 (KPNA2) as a novel prognostic biomarker for disease outcome in patients with breast and prostate cancer [9–12]. The prognostic impact of KPNA2 overexpression has also been described in other solid tumours such as melanoma, gastric, cervical, and urothelial bladder cancer, respectively [13–18]. KPNA2 is an adaptor protein that recognizes the classical nuclear localization signal (NLS) and acts as an important mediator of nucleocytoplasmic transport [19]. Among other binding partners, KPNA2 interacts with the Nijmegen breakage syndrome (NBS) gene product (NBS1), a component of the MRE11–RAD50–NBS1 (MRN) complex [20]. Since this complex is a central player in double-strand break (DSB) repair [20], the nuclear import of NBS1 is crucial for its tumour suppressor role. In fact, cytoplasmic accumulation of NBS1 may even have a tumour-promoting effect associated with PI3K/AKT pathway activation, underlining the importance of subcellular shuttle proteins for tumourigenesis [21]. In line with the indicated role of the nuclear transport machinery in regulating cell proliferation, there is growing evidence for the involvement of KPNA2 in cell-cycle regulation [22,23]. We also recently demonstrated that mice with endometrium-specific deletion of Trp53 initially exhibited histological changes that are identical to known precursor lesions of type II endometrial carcinomas in humans and later developed carcinomas representing all type II subtypes [24]. The PI3K/AKT signalling pathway was frequently activated in these precursor lesions and tumours, suggesting a genetic cooperation between this pathway and Trp53 deficiency in tumour initiation. Consistent with this idea, analyses of 521 human endometrial carcinomas identified frequent PI3K/AKT pathway activation in type I as well as type II endometrial carcinoma subtypes. PI3K/AKT pathway activation and p53 expression or mutation status each independently predicted poor patient survival. Copyright © 2014 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
K Ikenberg et al
We suggested that molecular alterations in p53 and the PI3K/AKT pathway play different roles in the initiation of the different endometrial cancer subtypes, but that combined p53 inactivation and PI3K/AKT pathway activation are unifying pathogenic features among histologically diverse subtypes of late-stage aggressive endometrial tumours. Thus, this mouse model provides an excellent tool for investigating protein expression changes in the course of endometrial cancer progression. In the present study, we analysed the prognostic relevance of KPNA2 expression in two independent cohorts of patients with endometrial carcinoma and compared this expression pattern with well-established clinicopathological parameters, PI3K/AKT signalling pathway activation, and epithelial–mesenchymal transition (EMT). Furthermore, KPNA2 expression was investigated in our highly accurate mouse model of type II endometrial carcinomas, facilitating the analysis of KPNA2 immunoreactivity in precursor lesions. Finally, we investigated the effect of KPNA2 loss on cell proliferation using in vitro siRNA silencing.
Materials and methods Patients As previously described [24], 527 patients with surgically resected adenocarcinomas and carcinosarcomas of the uterus were identified in the archives of the Institute of Pathology, University of Basel, Basel, Switzerland, and the Institute of Surgical Pathology, University Hospital Zurich, Zurich, Switzerland. Patients with endometrial cancer, who had localized disease, were treated by hysterectomy and bilateral salpingo-oophorectomy (with or without pelvic and para-aortic lymphadenectomy). Adjuvant intravaginal radiation therapy was given post-operatively when invasion of the myometrium or histological grade 3 tumour was found. All carcinomas were subsequently staged according to the seventh edition of the International Union Against Cancer (UICC) and American Joint Committee on Cancer (AJCC) TNM system [25]. The histopathological analysis was performed using H&E-stained sections by two of the authors (HM, RC). Immunohistochemistry was performed in questionable cases to evaluate the histological subtype. Histological grading was carried out according to the FIGO grading system based on the ratio of glandular or papillary structures to solid tumour growth (grade 1, ≤ 5% solid tumour; grade 2, 6–50% solid; and grade 3, > 50% solid) [26]. This grading was undertaken only for type I endometrial cancers, whereas the various type II endometrial carcinomas were automatically regarded as grade 3. Squamous and morular components were excluded from grading. Follow-up information was obtained from the participating Departments of Gynecology at the University Hospital Basel, University Hospital Zurich, Cantonal Hospital in Liestal (Basel-County), and Women’s Hospitals in Loerrach and Rheinfelden, and from the Cancer Registries of J Pathol 2014; 234: 239–252 www.thejournalofpathology.com
High KPNA2 expression promotes endometrial cancer
Basel and Zurich. Overall survival time was calculated from the time point of biopsy diagnosis to the death of patients or the last follow-up appointment.
Human tissue microarrays Tissue microarrays (TMAs) contained 527 formalinfixed, paraffin-embedded (FFPE) endometrial cancer tissues and were constructed as previously described [24]. Selection criteria were based on availability, presence, and size of representative tumour areas in paraffin blocks. Specimens were collected between 1985 and 2005 from the Institute of Surgical Pathology, University Hospital Zurich, Zurich, Switzerland, and the Institute of Pathology, University of Basel, Basel, Switzerland. The Basel TMA included a series of 343 consecutive (non-selected) endometrial carcinomas, containing one tissue core per specimen. The Zurich TMA included a series of 184 consecutive (non-selected) endometrial cancer specimens, comprising two tissue cores per case. Median follow-up of the cohort from Basel was 32 months (1–184 months), and 45 months (1–124 months) for the cohort from Zurich. The study for both cohorts was approved by the Scientific Ethics Committee from Basel and Zurich (approval No KEK-ZH-NR 2010-0358).
Mouse tumour samples For the analysis of the different subtypes of type II endometrial cancer and their precursor lesions, we used our previously described conditional Trp53 knockout mouse model that was generated by crossing Trp53fl/fl mice with mice expressing the Ksp1.3-Cre transgene [24]. Mice were sacrificed at the age of 65–79 weeks and FFPE uteri were subjected to immunohistochemical analysis.
Antibodies The following antibodies were used: Snail (Abcam, Cambridge, MA, USA; ab17732), MMP-9 (United States Biological, Salem, MA, USA; # M2425-28), periostin (BioVendor R&D, Brno, Czech Republic; RD181045050), E-cadherin (Cell Marque Corporation, Rocklin, CA, USA; 246R-16), p53 (Dako, Carpinteria, CA, USA; M7001), human KPNA2 (Santa Cruz Biotechnology Inc, Santa Cruz, CA, USA; sc-6917), mouse and human cell pellets KPNA2 (Abcam; ab84440), and GAPDH (Sigma-Aldrich, St Louis, MO, USA; #G9545). The immunohistochemical staining of PI3K/AKT pathway members and other markers linked to endometrial cancer progression has been previously described [24] and was performed with the following antibodies: PI3K p110α (Cell Signaling Technology, Beverly, MA, USA; #4255), PTEN (Dako; M3627), phospho-Ser473-AKT (Cell Signaling Technology; #4051), phospho-Ser9-GSK3β (Cell Signaling Technology; #5558), phospho-Ser2448-mTOR (Cell Signaling Technology; #2976), phospho-Thr421/Ser424-p70 S6 kinase (Cell Signaling Technology; #9204), Copyright © 2014 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
241
phospho-Ser240/244-ribosomal S6 protein (Cell Signaling Technology; #2215), phospho-Thr37/46-4E-BP1 (Cell Signaling Technology; #2855), p53 (Dako; M7001), p16 (Santa Cruz Biotechnology; sc-56330), ErbB2 (Novocastra Laboratories, Newcastle upon Tyne, UK; NCL-L-CBE-356), PAX8 (Proteintech Group, Chicago, Illinois, USA; 10336-1-AP), and IMP3 (Dako; M3626).
Immunohistochemical assays Consecutive 3-μm sections were cut from the TMA tissue blocks, the FFPE mouse uteri, and from FFPE cell pellets (embedded after coagulation with plasma and thrombin) of the human endometrial adenocarcinoma cell line MFE-296. After antigen retrieval (in a microwave oven for 10 min at 250 W), immunohistochemistry was carried out in an automated immunostainer (Roche Ventana, Tucson, AZ, USA) according to the manufacturer’s instructions. As a negative control, the primary antibody was omitted. The specificity of the commercial KPNA2 antibodies has been thoroughly validated in former studies [9–11]. Slides were counterstained with haematoxylin prior to dehydration and coverslipping. A TMA containing various tissue types was chosen as an internal control for every staining. Two surgical pathologists (KI, PJW) performed a blinded evaluation of the immunostained slides without knowledge of the clinical data. Causes of non-interpretable results included lack of target tissue and the presence of necrosis or crush artefact. For the analysis of p53 staining, tumours were scored in five groups based on the extent of p53 positivity by determining the frequency of tumour cells displaying strong nuclear staining (0%, 1–10%, 11–50%, 51–90%, 91–100%). For the analysis of KPNA2 staining, the percentage of tumour cells with nuclear immunoreactivity was quantified. Based on initial breast cancer studies [9], KPNA2 positivity was defined as strong nuclear staining in at least 10% of nuclei. HER2 expression was scored according to the DAKO HercepTest [27]. The immunoreactivity of all previously reported markers [24] was evaluated using a stepwise and semi-quantitative scoring system based on staining intensity: 0 (negative): no staining; 1+: weak staining; 2+: moderate staining; 3+: strong staining.
Cell culture, KPNA2 knockdown, and proliferation assays The human endometrial adenocarcinoma cell line MFE-296 was purchased from ECACC (Public Health England Culture Collections, UK) and cultured in Minimal Essential Medium (MEM; Gibco, Carlsbad, CA, USA), supplemented with 10% FBS, 100 U/ml penicillin (Gibco), 100 μg/ml streptomycin (Gibco), and 2 mM L-glutamine (Gibco). At 80% confluency, cells were trypsinized and reverse-transfected with 50 nM siRNA using Lipofectamine RNAiMAX transfection reagent according to the manufacturer’s recommendation (Invitrogen, Carlsbad, CA, USA). AllStars J Pathol 2014; 234: 239–252 www.thejournalofpathology.com
242
K Ikenberg et al
Figure 1. Expression of KPNA2 and epithelial–mesenchymal transition (EMT) markers in endometrial cancer. (A) Immunohistochemical expression of KPNA2 in endometrial cancer. Scale bar = 100 μm. (B) KPNA2 immunoreactivity in different histological subtypes of endometrial cancer. EC, endometrioid adenocarcinoma; MC, mucinous adenocarcinoma; SC, serous adenocarcinoma; CCC, clear cell adenocarcinoma; MMMT, carcinosarcoma; UC, undifferentiated carcinoma. A two-sided Fisher’s exact test was used to compare the frequency of KPNA2 overexpression (>10%) within the histological subtypes of endometrial cancer. (C–F) Immunohistochemical expression of E-cadherin (C), Snail (D), periostin (E), and MMP-9 (F) as markers for EMT in endometrial cancer. Scale bar = 100 μm Copyright © 2014 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
J Pathol 2014; 234: 239–252 www.thejournalofpathology.com
High KPNA2 expression promotes endometrial cancer
negative control siRNA (QIAGEN, Valencia, CA, USA; accession No SI03650318), AllStars Hs Cell Death Control siRNA (QIAGEN; accession No SI04381048), and three siRNAs targeting KPNA2 (QIAGEN; accession Nos KPNA2_si1-3: SI00035525, SI02780631, SI02781051) were used separately. Transfection was carried out in MEM with 10% FBS and antibiotic-free in 96-well plates (7 × 103 cells per well) for MTT assays or in six-well plates (2 × 105 cells per well) for xCELLigence (ACEA Biosciences Inc, San Diego, CA, USA) real-time cell proliferation assays and western blot analyses. Forty-eight hours post-transfection, MTT assays were performed as previously described [12]. For the xCELLigence assay – an electronic impedance-based cell sensing unit [28] – cells were trypsinized 24 h post-transfection from the six-well plate and seeded on E-plates (8 × 103 cells per well). Impedance was measured every 10 min for 55 h. For data normalization, the cell index at time point 1.04 h was subtracted from each value. The thus normalized delta cell indices were plotted against time. The slope was calculated as Cell index = slope × time + intercept, and the doubling time as t
Cell index = A × 2 CI doubling
time
between 1.04 and 50.07 h.
Western blot analyses Cultured cells were lysed on ice from the six-well plates using RIPA buffer (Sigma-Aldrich). Protein extracts (total 15 μg) were run on precasted 4–20% Mini-Protean TGX Stain-Free Gels (Bio-Rad, Hercules, CA, USA; #456-8093). Protein was transferred using the semi-dry Trans-Blot Turbo Transfer System (Bio-Rad) and blocked with 5% milk in TBS-T. Primary antibodies were incubated at 4 ∘ C overnight, both diluted in 5% BSA in TBS-T (α-KPNA2, 1 : 100; α-GAPDH, 1 : 1000). Proteins were visualized with ECL utilizing HRP-coupled secondary antibodies.
Statistical analyses Non-parametric Kaplan–Meier estimators were used to analyse overall survival. Patients were censored at the time of their last clinical follow-up visit. Simultaneous confidence bands [29] at a confidence level of 0.95 were computed for the whole range of time values, capturing the entire true survival curve about 19 out of 20 times. Differences between survival estimates were assessed with the log-rank test (LRT). p values below 0.05 were considered to indicate statistical significance. Statistical analyses were conducted using survival [30] and OIsurv [31] packages in R as well as SPSS version 19.0. A statistical association between clinicopathological and molecular parameters was tested using a two-sided Fisher’s exact test or Pearson chi-square test. In addition Copyright © 2014 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
243
to the univariate Cox regression, a multivariable Cox regression model was adjusted, testing the independent prognostic relevance of the prognostic parameters.
Results KPNA2 protein expression level does not correlate with a distinct histological subtype of endometrial cancer A total of 527 endometrial carcinoma samples were analysed for expression of KPNA2 by immunohistochemistry. We have recently constructed two human endometrial carcinoma TMAs, in total comprising tumour punches from 436 endometrioid adenocarcinomas (type I), six mucinous adenocarcinomas (type I), 34 serous adenocarcinomas (type II), 19 clear cell adenocarcinomas (type II), 16 undifferentiated carcinomas (type II), and 16 carcinosarcomas [24]. 90.5% (n = 477) of the specimens could be evaluated for KPNA2 immunoreactivity, whereas 9.5% (n = 50) were not interpretable due to lack of target tissue, the presence of necrosis or crush artefacts. In 105 of 477 analysed endometrial cancer cases (22%), nuclear staining for KPNA2 was present in at least 10% of the tumour cells (Figure 1A). No significant differences of KPNA2 expression were detected between the different subtypes of endometrial carcinoma (p = 0.355, Figure 1B). KPNA2 expression was significantly associated with high tumour grade (p = 0.026) and high FIGO stage (p = 0.027, Table 1).
KPNA2 is associated with EMT and PI3K pathway activation To investigate whether KPNA2 expression correlates with other molecular alterations in endometrial carcinomas, we stained the TMA for several molecules that are associated with EMT in tumour progression [32–34], including E-cadherin (Figure 1C), Snail (Figure 1D), periostin (Figure 1E), and MMP-9 (Figure 1F). In addition, we took advantage of our previously published analyses on the TMAs [24] which reveal either the activation status of the PI3K/AKT signalling pathways (PI3K p110α, PTEN, phospho-Ser473AKT, phospho-Ser9-GSK3β, phospho-Ser2448-mTOR, phospho-Thr421/Ser424-p70 S6 kinase, phosphoSer240/244-ribosomal S6 protein, phospho-Thr37/464E-BP1) or the expression levels of markers previously proposed to be differentially expressed between different subtypes of endometrial carcinoma (IMP3, PAX8, and p16) [35–37] or shown to predict a poor prognosis (p53, ErbB2) [38–40]. Increased KPNA2 immunoreactivity was associated with higher levels of p53, p16, ErbB2, PAX8, and IMP3; activation of the PI3K/AKT pathway; and EMT. Detailed results showing the association of KPNA2 immunoreactivity with clinicopathological characteristics and immunohistochemical data are depicted in Figures 2A and 2B. J Pathol 2014; 234: 239–252 www.thejournalofpathology.com
244
K Ikenberg et al
Table 1. Clinicopathological and immunohistochemical characteristics in relation to nuclear KPNA2 immunoreactivity and results of univariate survival analyses Nuclear KPNA2 immunoreactivity Variable Clinicopathological characteristics Age at diagnosis ≤ 68 years > 68 years Grade G1 G2 G3 FIGO stage I II III IV Histological subtype Type I Type II MMMT Immunohistochemical data Nuclear p53 (%) 0% ≤ 10% > 10–50% > 50–90% > 90–100% ErbB2 (Dako HercepTest criteria) Negative Score 1+ Score 2+ Score 3+ Cytoplasmic p16 (intensity) Negative Score 1+ Score 2+ Score 3+ Cytoplasmic PTEN (intensity) Negative Score 1+ Score 2+ Score 3+ Cytoplasmic phospho-Ser473-AKT (intensity) Negative Score 1+ Score 2+ Score 3+ Cytoplasmic phospho-Thr37/46-4E-BP1 (intensity) Negative Score 1+ Score 2+ Score 3+ Cytoplasmic phospho-Ser240/244-ribosomal S6 protein (intensity) Negative Score 1+ Score 2+ Score 3+ Nuclear phospho-Thr421/Ser424-p70 S6 kinase (intensity) Negative Score 1+ Score 2+ Score 3+ Cytoplasmic phospho-Ser2448-mTOR (intensity) Negative Score 1+ Score 2+ Copyright © 2014 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
Overall survival
< 10%
≥ 10%
p value*
n
Events
p value†
199 172
51 54
0.377
216 198
44 89
< 0.001
205 89 56
41 33 21
0.026
210 105 66
39 46 29
< 0.001
190 53 49 11
48 9 24 6
0.027
255 63 77 18
63 15 44 10
< 0.001
320 40 12
83 20 2
0.073
344 54 16
96 27 10
< 0.001
202 83 34 18 23
27 40 14 8 16
< 0.001
177 112 36 26 35
35 42 16 12 20
ORIGINAL PAPER
KPNA2 is overexpressed in human and mouse endometrial cancers and promotes cellular proliferation Kristian Ikenberg,1† Nadejda Valtcheva,1† Simone Brandt,1 Qing Zhong,1 Christine E Wong,1 Aurelia Noske,1 Markus Rechsteiner,1 Jan H Rueschoff,1 Rosmarie Caduff,1 Athanassios Dellas,2 Ellen Obermann,3 Daniel Fink,4 Thomas Fuchs,5 Wilhelm Krek,6 Holger Moch,1 Ian J Frew7 and Peter J Wild1* 1 2 3 4 5 6 7
Institute of Surgical Pathology, University Hospital Zurich, Zurich, Switzerland Krebsliga beider Basel, Basel, Switzerland Institute of Pathology, University Hospital of Basel, Basel, Switzerland Department of Gynecology, University Hospital Zurich, Zurich, Switzerland Department of Electrical Engineering, California Institute of Technology, Pasadena, CA, USA Institute of Cell Biology, ETH Zurich, Zurich, Switzerland Zurich Center for Integrative Human Physiology, University of Zurich, Zurich, Switzerland
*Correspondence to: Peter J Wild, Institute of Surgical Pathology, Schmelzbergstrasse 12, 8091 Zurich, Switzerland. e-mail: [email protected] †
These authors contributed equally to this work.
Abstract Endometrial cancer is the most frequently occurring malignancy of the female genital tract in Western countries. Although in many cases surgically curable, about 30% of the tumours represent an aggressive and untreatable disease. In an attempt to establish a reliable prognostic marker for endometrial carcinomas disregarding their histological diversity, we investigated the expression of KPNA2, a mediator of nucleocytoplasmic transport, and other cell proliferation-associated proteins and their correlation with cancer progression. We analysed patient tissue microarrays (TMAs) assembled from 527 endometrial cancer tissue specimens and uterus samples from a Trp53 knockout mouse model of endometrial cancer. Our data show that KPNA2 expression was significantly up-regulated in human endometrial carcinomas and associated with higher tumour grade (p = 0.026), higher FIGO stage (p = 0.027), p53 overexpression (p < 0.001), activation of the PI3K/AKT pathway, and epithelial–mesenchymal transition. Increased nuclear KPNA2 immunoreactivity was identified as a novel predictor of overall survival, independent of well-established prognostic factors in Cox regression analyses (hazard ratio 1.7, 95% CI 1.13–2.56, p = 0.01). No significant association between KPNA2 expression and endometrial cancer subtype was detected. In the mouse model, KPNA2 showed increased expression levels from precancerous (EmgD, EIC) to far-advanced invasive lesions. We further investigated the cell proliferation capacity after siRNA-mediated KPNA2 knockdown in the human endometrial cancer cell line MFE-296. KPNA2 silencing led to decreased proliferation of the cancer cells, suggesting interplay of the protein with the cell cycle. Taken together, increased expression of KPNA2 is an independent prognostic marker for poor survival. The mechanism of enhanced nucleocytoplasmic transport by KPNA2 overexpression seems a common event in aggressive cancers since we have shown a significant correlation of KPNA2 expression and tumour aggressiveness in a large variety of other solid tumour entities. Introducing KPNA2 immunohistochemistry in routine diagnostics may allow for the identification of patients who need more aggressive treatment regimens. Copyright © 2014 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.
Keywords: KPNA2; importin; endometrial cancer; biomarker; EMT; Snail
Received 16 April 2014; Revised 2 June 2014; Accepted 10 June 2014
No conflicts of interest were declared.
Introduction Endometrial carcinomas of the uterus are the most common tumour type in the Western world that develops in the female genital tract, with over 40 000 new cases and more than 7000 related deaths per year in the United States [1]. Well-established prognostic factors for endometrial carcinoma are tumour grade, Copyright © 2014 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
FIGO stage, and histological subtype [2–4], but these pathological parameters only insufficiently predict patient outcome for currently available therapies. The most frequent form of these tumours is the so-called type I endometrioid carcinomas, which typically have a relatively good prognosis. The rarer but more aggressive type II endometrial carcinomas, comprising serous carcinomas, clear cell carcinomas, and undifferentiated J Pathol 2014; 234: 239–252 www.thejournalofpathology.com
240
carcinomas, account for a disproportionate number of endometrial cancer-associated deaths [3]. The molecular basis of these different forms of cancer remains not completely understood. As most women present early in the course of the disease, the majority of patients have an excellent prognosis after surgery. For patients with advanced metastatic or recurrent endometrial cancer, however, the available therapeutic options are limited. A meta-analysis that evaluated the treatment outcome of advanced endometrial carcinomas by investigating symptom control and impact on the quality of life after cytotoxic chemotherapy showed disappointing results [5]. Reliable biomarkers that are based on the molecular changes of these aggressive tumours are an unmet clinical need and could be potential therapeutic targets. Besides the well-established prognostic parameters, a broad diversity of novel molecular markers has already been evaluated for predictive and prognostic relevance in patients with endometrial carcinomas [6–8]; however, none of them has entered the diagnostic routine yet. We have recently described karyopherin alpha 2 (KPNA2) as a novel prognostic biomarker for disease outcome in patients with breast and prostate cancer [9–12]. The prognostic impact of KPNA2 overexpression has also been described in other solid tumours such as melanoma, gastric, cervical, and urothelial bladder cancer, respectively [13–18]. KPNA2 is an adaptor protein that recognizes the classical nuclear localization signal (NLS) and acts as an important mediator of nucleocytoplasmic transport [19]. Among other binding partners, KPNA2 interacts with the Nijmegen breakage syndrome (NBS) gene product (NBS1), a component of the MRE11–RAD50–NBS1 (MRN) complex [20]. Since this complex is a central player in double-strand break (DSB) repair [20], the nuclear import of NBS1 is crucial for its tumour suppressor role. In fact, cytoplasmic accumulation of NBS1 may even have a tumour-promoting effect associated with PI3K/AKT pathway activation, underlining the importance of subcellular shuttle proteins for tumourigenesis [21]. In line with the indicated role of the nuclear transport machinery in regulating cell proliferation, there is growing evidence for the involvement of KPNA2 in cell-cycle regulation [22,23]. We also recently demonstrated that mice with endometrium-specific deletion of Trp53 initially exhibited histological changes that are identical to known precursor lesions of type II endometrial carcinomas in humans and later developed carcinomas representing all type II subtypes [24]. The PI3K/AKT signalling pathway was frequently activated in these precursor lesions and tumours, suggesting a genetic cooperation between this pathway and Trp53 deficiency in tumour initiation. Consistent with this idea, analyses of 521 human endometrial carcinomas identified frequent PI3K/AKT pathway activation in type I as well as type II endometrial carcinoma subtypes. PI3K/AKT pathway activation and p53 expression or mutation status each independently predicted poor patient survival. Copyright © 2014 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
K Ikenberg et al
We suggested that molecular alterations in p53 and the PI3K/AKT pathway play different roles in the initiation of the different endometrial cancer subtypes, but that combined p53 inactivation and PI3K/AKT pathway activation are unifying pathogenic features among histologically diverse subtypes of late-stage aggressive endometrial tumours. Thus, this mouse model provides an excellent tool for investigating protein expression changes in the course of endometrial cancer progression. In the present study, we analysed the prognostic relevance of KPNA2 expression in two independent cohorts of patients with endometrial carcinoma and compared this expression pattern with well-established clinicopathological parameters, PI3K/AKT signalling pathway activation, and epithelial–mesenchymal transition (EMT). Furthermore, KPNA2 expression was investigated in our highly accurate mouse model of type II endometrial carcinomas, facilitating the analysis of KPNA2 immunoreactivity in precursor lesions. Finally, we investigated the effect of KPNA2 loss on cell proliferation using in vitro siRNA silencing.
Materials and methods Patients As previously described [24], 527 patients with surgically resected adenocarcinomas and carcinosarcomas of the uterus were identified in the archives of the Institute of Pathology, University of Basel, Basel, Switzerland, and the Institute of Surgical Pathology, University Hospital Zurich, Zurich, Switzerland. Patients with endometrial cancer, who had localized disease, were treated by hysterectomy and bilateral salpingo-oophorectomy (with or without pelvic and para-aortic lymphadenectomy). Adjuvant intravaginal radiation therapy was given post-operatively when invasion of the myometrium or histological grade 3 tumour was found. All carcinomas were subsequently staged according to the seventh edition of the International Union Against Cancer (UICC) and American Joint Committee on Cancer (AJCC) TNM system [25]. The histopathological analysis was performed using H&E-stained sections by two of the authors (HM, RC). Immunohistochemistry was performed in questionable cases to evaluate the histological subtype. Histological grading was carried out according to the FIGO grading system based on the ratio of glandular or papillary structures to solid tumour growth (grade 1, ≤ 5% solid tumour; grade 2, 6–50% solid; and grade 3, > 50% solid) [26]. This grading was undertaken only for type I endometrial cancers, whereas the various type II endometrial carcinomas were automatically regarded as grade 3. Squamous and morular components were excluded from grading. Follow-up information was obtained from the participating Departments of Gynecology at the University Hospital Basel, University Hospital Zurich, Cantonal Hospital in Liestal (Basel-County), and Women’s Hospitals in Loerrach and Rheinfelden, and from the Cancer Registries of J Pathol 2014; 234: 239–252 www.thejournalofpathology.com
High KPNA2 expression promotes endometrial cancer
Basel and Zurich. Overall survival time was calculated from the time point of biopsy diagnosis to the death of patients or the last follow-up appointment.
Human tissue microarrays Tissue microarrays (TMAs) contained 527 formalinfixed, paraffin-embedded (FFPE) endometrial cancer tissues and were constructed as previously described [24]. Selection criteria were based on availability, presence, and size of representative tumour areas in paraffin blocks. Specimens were collected between 1985 and 2005 from the Institute of Surgical Pathology, University Hospital Zurich, Zurich, Switzerland, and the Institute of Pathology, University of Basel, Basel, Switzerland. The Basel TMA included a series of 343 consecutive (non-selected) endometrial carcinomas, containing one tissue core per specimen. The Zurich TMA included a series of 184 consecutive (non-selected) endometrial cancer specimens, comprising two tissue cores per case. Median follow-up of the cohort from Basel was 32 months (1–184 months), and 45 months (1–124 months) for the cohort from Zurich. The study for both cohorts was approved by the Scientific Ethics Committee from Basel and Zurich (approval No KEK-ZH-NR 2010-0358).
Mouse tumour samples For the analysis of the different subtypes of type II endometrial cancer and their precursor lesions, we used our previously described conditional Trp53 knockout mouse model that was generated by crossing Trp53fl/fl mice with mice expressing the Ksp1.3-Cre transgene [24]. Mice were sacrificed at the age of 65–79 weeks and FFPE uteri were subjected to immunohistochemical analysis.
Antibodies The following antibodies were used: Snail (Abcam, Cambridge, MA, USA; ab17732), MMP-9 (United States Biological, Salem, MA, USA; # M2425-28), periostin (BioVendor R&D, Brno, Czech Republic; RD181045050), E-cadherin (Cell Marque Corporation, Rocklin, CA, USA; 246R-16), p53 (Dako, Carpinteria, CA, USA; M7001), human KPNA2 (Santa Cruz Biotechnology Inc, Santa Cruz, CA, USA; sc-6917), mouse and human cell pellets KPNA2 (Abcam; ab84440), and GAPDH (Sigma-Aldrich, St Louis, MO, USA; #G9545). The immunohistochemical staining of PI3K/AKT pathway members and other markers linked to endometrial cancer progression has been previously described [24] and was performed with the following antibodies: PI3K p110α (Cell Signaling Technology, Beverly, MA, USA; #4255), PTEN (Dako; M3627), phospho-Ser473-AKT (Cell Signaling Technology; #4051), phospho-Ser9-GSK3β (Cell Signaling Technology; #5558), phospho-Ser2448-mTOR (Cell Signaling Technology; #2976), phospho-Thr421/Ser424-p70 S6 kinase (Cell Signaling Technology; #9204), Copyright © 2014 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
241
phospho-Ser240/244-ribosomal S6 protein (Cell Signaling Technology; #2215), phospho-Thr37/46-4E-BP1 (Cell Signaling Technology; #2855), p53 (Dako; M7001), p16 (Santa Cruz Biotechnology; sc-56330), ErbB2 (Novocastra Laboratories, Newcastle upon Tyne, UK; NCL-L-CBE-356), PAX8 (Proteintech Group, Chicago, Illinois, USA; 10336-1-AP), and IMP3 (Dako; M3626).
Immunohistochemical assays Consecutive 3-μm sections were cut from the TMA tissue blocks, the FFPE mouse uteri, and from FFPE cell pellets (embedded after coagulation with plasma and thrombin) of the human endometrial adenocarcinoma cell line MFE-296. After antigen retrieval (in a microwave oven for 10 min at 250 W), immunohistochemistry was carried out in an automated immunostainer (Roche Ventana, Tucson, AZ, USA) according to the manufacturer’s instructions. As a negative control, the primary antibody was omitted. The specificity of the commercial KPNA2 antibodies has been thoroughly validated in former studies [9–11]. Slides were counterstained with haematoxylin prior to dehydration and coverslipping. A TMA containing various tissue types was chosen as an internal control for every staining. Two surgical pathologists (KI, PJW) performed a blinded evaluation of the immunostained slides without knowledge of the clinical data. Causes of non-interpretable results included lack of target tissue and the presence of necrosis or crush artefact. For the analysis of p53 staining, tumours were scored in five groups based on the extent of p53 positivity by determining the frequency of tumour cells displaying strong nuclear staining (0%, 1–10%, 11–50%, 51–90%, 91–100%). For the analysis of KPNA2 staining, the percentage of tumour cells with nuclear immunoreactivity was quantified. Based on initial breast cancer studies [9], KPNA2 positivity was defined as strong nuclear staining in at least 10% of nuclei. HER2 expression was scored according to the DAKO HercepTest [27]. The immunoreactivity of all previously reported markers [24] was evaluated using a stepwise and semi-quantitative scoring system based on staining intensity: 0 (negative): no staining; 1+: weak staining; 2+: moderate staining; 3+: strong staining.
Cell culture, KPNA2 knockdown, and proliferation assays The human endometrial adenocarcinoma cell line MFE-296 was purchased from ECACC (Public Health England Culture Collections, UK) and cultured in Minimal Essential Medium (MEM; Gibco, Carlsbad, CA, USA), supplemented with 10% FBS, 100 U/ml penicillin (Gibco), 100 μg/ml streptomycin (Gibco), and 2 mM L-glutamine (Gibco). At 80% confluency, cells were trypsinized and reverse-transfected with 50 nM siRNA using Lipofectamine RNAiMAX transfection reagent according to the manufacturer’s recommendation (Invitrogen, Carlsbad, CA, USA). AllStars J Pathol 2014; 234: 239–252 www.thejournalofpathology.com
242
K Ikenberg et al
Figure 1. Expression of KPNA2 and epithelial–mesenchymal transition (EMT) markers in endometrial cancer. (A) Immunohistochemical expression of KPNA2 in endometrial cancer. Scale bar = 100 μm. (B) KPNA2 immunoreactivity in different histological subtypes of endometrial cancer. EC, endometrioid adenocarcinoma; MC, mucinous adenocarcinoma; SC, serous adenocarcinoma; CCC, clear cell adenocarcinoma; MMMT, carcinosarcoma; UC, undifferentiated carcinoma. A two-sided Fisher’s exact test was used to compare the frequency of KPNA2 overexpression (>10%) within the histological subtypes of endometrial cancer. (C–F) Immunohistochemical expression of E-cadherin (C), Snail (D), periostin (E), and MMP-9 (F) as markers for EMT in endometrial cancer. Scale bar = 100 μm Copyright © 2014 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
J Pathol 2014; 234: 239–252 www.thejournalofpathology.com
High KPNA2 expression promotes endometrial cancer
negative control siRNA (QIAGEN, Valencia, CA, USA; accession No SI03650318), AllStars Hs Cell Death Control siRNA (QIAGEN; accession No SI04381048), and three siRNAs targeting KPNA2 (QIAGEN; accession Nos KPNA2_si1-3: SI00035525, SI02780631, SI02781051) were used separately. Transfection was carried out in MEM with 10% FBS and antibiotic-free in 96-well plates (7 × 103 cells per well) for MTT assays or in six-well plates (2 × 105 cells per well) for xCELLigence (ACEA Biosciences Inc, San Diego, CA, USA) real-time cell proliferation assays and western blot analyses. Forty-eight hours post-transfection, MTT assays were performed as previously described [12]. For the xCELLigence assay – an electronic impedance-based cell sensing unit [28] – cells were trypsinized 24 h post-transfection from the six-well plate and seeded on E-plates (8 × 103 cells per well). Impedance was measured every 10 min for 55 h. For data normalization, the cell index at time point 1.04 h was subtracted from each value. The thus normalized delta cell indices were plotted against time. The slope was calculated as Cell index = slope × time + intercept, and the doubling time as t
Cell index = A × 2 CI doubling
time
between 1.04 and 50.07 h.
Western blot analyses Cultured cells were lysed on ice from the six-well plates using RIPA buffer (Sigma-Aldrich). Protein extracts (total 15 μg) were run on precasted 4–20% Mini-Protean TGX Stain-Free Gels (Bio-Rad, Hercules, CA, USA; #456-8093). Protein was transferred using the semi-dry Trans-Blot Turbo Transfer System (Bio-Rad) and blocked with 5% milk in TBS-T. Primary antibodies were incubated at 4 ∘ C overnight, both diluted in 5% BSA in TBS-T (α-KPNA2, 1 : 100; α-GAPDH, 1 : 1000). Proteins were visualized with ECL utilizing HRP-coupled secondary antibodies.
Statistical analyses Non-parametric Kaplan–Meier estimators were used to analyse overall survival. Patients were censored at the time of their last clinical follow-up visit. Simultaneous confidence bands [29] at a confidence level of 0.95 were computed for the whole range of time values, capturing the entire true survival curve about 19 out of 20 times. Differences between survival estimates were assessed with the log-rank test (LRT). p values below 0.05 were considered to indicate statistical significance. Statistical analyses were conducted using survival [30] and OIsurv [31] packages in R as well as SPSS version 19.0. A statistical association between clinicopathological and molecular parameters was tested using a two-sided Fisher’s exact test or Pearson chi-square test. In addition Copyright © 2014 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
243
to the univariate Cox regression, a multivariable Cox regression model was adjusted, testing the independent prognostic relevance of the prognostic parameters.
Results KPNA2 protein expression level does not correlate with a distinct histological subtype of endometrial cancer A total of 527 endometrial carcinoma samples were analysed for expression of KPNA2 by immunohistochemistry. We have recently constructed two human endometrial carcinoma TMAs, in total comprising tumour punches from 436 endometrioid adenocarcinomas (type I), six mucinous adenocarcinomas (type I), 34 serous adenocarcinomas (type II), 19 clear cell adenocarcinomas (type II), 16 undifferentiated carcinomas (type II), and 16 carcinosarcomas [24]. 90.5% (n = 477) of the specimens could be evaluated for KPNA2 immunoreactivity, whereas 9.5% (n = 50) were not interpretable due to lack of target tissue, the presence of necrosis or crush artefacts. In 105 of 477 analysed endometrial cancer cases (22%), nuclear staining for KPNA2 was present in at least 10% of the tumour cells (Figure 1A). No significant differences of KPNA2 expression were detected between the different subtypes of endometrial carcinoma (p = 0.355, Figure 1B). KPNA2 expression was significantly associated with high tumour grade (p = 0.026) and high FIGO stage (p = 0.027, Table 1).
KPNA2 is associated with EMT and PI3K pathway activation To investigate whether KPNA2 expression correlates with other molecular alterations in endometrial carcinomas, we stained the TMA for several molecules that are associated with EMT in tumour progression [32–34], including E-cadherin (Figure 1C), Snail (Figure 1D), periostin (Figure 1E), and MMP-9 (Figure 1F). In addition, we took advantage of our previously published analyses on the TMAs [24] which reveal either the activation status of the PI3K/AKT signalling pathways (PI3K p110α, PTEN, phospho-Ser473AKT, phospho-Ser9-GSK3β, phospho-Ser2448-mTOR, phospho-Thr421/Ser424-p70 S6 kinase, phosphoSer240/244-ribosomal S6 protein, phospho-Thr37/464E-BP1) or the expression levels of markers previously proposed to be differentially expressed between different subtypes of endometrial carcinoma (IMP3, PAX8, and p16) [35–37] or shown to predict a poor prognosis (p53, ErbB2) [38–40]. Increased KPNA2 immunoreactivity was associated with higher levels of p53, p16, ErbB2, PAX8, and IMP3; activation of the PI3K/AKT pathway; and EMT. Detailed results showing the association of KPNA2 immunoreactivity with clinicopathological characteristics and immunohistochemical data are depicted in Figures 2A and 2B. J Pathol 2014; 234: 239–252 www.thejournalofpathology.com
244
K Ikenberg et al
Table 1. Clinicopathological and immunohistochemical characteristics in relation to nuclear KPNA2 immunoreactivity and results of univariate survival analyses Nuclear KPNA2 immunoreactivity Variable Clinicopathological characteristics Age at diagnosis ≤ 68 years > 68 years Grade G1 G2 G3 FIGO stage I II III IV Histological subtype Type I Type II MMMT Immunohistochemical data Nuclear p53 (%) 0% ≤ 10% > 10–50% > 50–90% > 90–100% ErbB2 (Dako HercepTest criteria) Negative Score 1+ Score 2+ Score 3+ Cytoplasmic p16 (intensity) Negative Score 1+ Score 2+ Score 3+ Cytoplasmic PTEN (intensity) Negative Score 1+ Score 2+ Score 3+ Cytoplasmic phospho-Ser473-AKT (intensity) Negative Score 1+ Score 2+ Score 3+ Cytoplasmic phospho-Thr37/46-4E-BP1 (intensity) Negative Score 1+ Score 2+ Score 3+ Cytoplasmic phospho-Ser240/244-ribosomal S6 protein (intensity) Negative Score 1+ Score 2+ Score 3+ Nuclear phospho-Thr421/Ser424-p70 S6 kinase (intensity) Negative Score 1+ Score 2+ Score 3+ Cytoplasmic phospho-Ser2448-mTOR (intensity) Negative Score 1+ Score 2+ Copyright © 2014 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
Overall survival
< 10%
≥ 10%
p value*
n
Events
p value†
199 172
51 54
0.377
216 198
44 89
< 0.001
205 89 56
41 33 21
0.026
210 105 66
39 46 29
< 0.001
190 53 49 11
48 9 24 6
0.027
255 63 77 18
63 15 44 10
< 0.001
320 40 12
83 20 2
0.073
344 54 16
96 27 10
< 0.001
202 83 34 18 23
27 40 14 8 16
< 0.001
177 112 36 26 35
35 42 16 12 20
