Clinical response to chemotherapy in oesophageal adenocarcinoma patients is linked to defects in mitochondria

Share Embed


Descrição do Produto

Journal of Pathology J Pathol 2013; 230: 410–419 Published online in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/path.4199

ORIGINAL PAPER

Clinical response to chemotherapy in oesophageal adenocarcinoma patients is linked to defects in mitochondria Michaela Aichler,1,# Mareike Elsner,1,# Natalie Ludyga,2 Annette Feuchtinger,1 Verena Zangen,3 Stefan K Maier,1,4 ¨ 1 Ludwig Hierber,3 Herbert Braselmann,3 Stephan Meding,1 Sandra Rauser,1 Benjamin Balluff,1 C´edrik Schone, 5 Hans Zischka, Michaela Aubele,2 Manfred Schmitt,6 Marcus Feith,7 Stefanie M Hauck,8 Marius Ueffing,8 Rupert ¨ 1,2,9 and Axel K Walch1* Langer,9 Bernhard Kuster,4 Horst Zitzelsberger,3 Heinz Hofler 1

Research Unit of Analytical Pathology, Institute of Pathology, Helmholtz Zentrum M¨unchen–German Research Centre for Environmental Health, Neuherberg, Germany 2 Institute of Pathology, Helmholtz Zentrum M¨unchen–German Research Centre for Environmental Health, Neuherberg, Germany 3 Research Unit of Radiation Cytogenetics, Helmholtz Zentrum M¨ unchen–German Research Centre for Environmental Health, Neuherberg, Germany 4 Department of Proteomics and Bioanalytics, Technische Universit¨at M¨unchen, Germany 5 Institute of Toxicology, Helmholtz Zentrum M¨ unchen–German Research Centre for Environmental Health, Neuherberg, Germany 6 Frauenklinik der Technischen Universit¨at M¨unchen, Klinikum Rechts der Isar, Munich, Germany 7 Department of Surgery, Klinikum Rechts der Isar, Technische Universit¨at M¨unchen, Germany 8 Research Unit of Protein Science, Helmholtz Zentrum M¨ unchen–German Research Centre for Environmental Health, Neuherberg, Germany 9 Institute of Pathology, Technische Universit¨at M¨unchen, Germany *Correspondence to: A Walch, Research Unit Analytical Pathology, Institute of Pathology, Helmholtz Zentrum M¨unchen–German Research Centre for Environmental Health, Ingolstaedter Landstrasse 1, 85764 Neuherberg, Germany. E-mail: [email protected] # These

authors contributed equally to this study.

Abstract Chemotherapeutic drugs kill cancer cells, but it is unclear why this happens in responding patients but not in nonresponders. Proteomic profiles of patients with oesophageal adenocarcinoma may be helpful in predicting response and selecting more effective treatment strategies. In this study, pretherapeutic oesophageal adenocarcinoma biopsies were analysed for proteomic changes associated with response to chemotherapy by MALDI imaging mass spectrometry. Resulting candidate proteins were identified by liquid chromatography–tandem mass spectrometry (LC–MS/MS) and investigated for functional relevance in vitro . Clinical impact was validated in pretherapeutic biopsies from an independent patient cohort. Studies on the incidence of these defects in other solid tumours were included. We discovered that clinical response to cisplatin correlated with pre-existing defects in the mitochondrial respiratory chain complexes of cancer cells, caused by loss of specific cytochrome c oxidase (COX) subunits. Knockdown of a COX protein altered chemosensitivity in vitro , increasing the propensity of cancer cells to undergo cell death following cisplatin treatment. In an independent validation, patients with reduced COX protein expression prior to treatment exhibited favourable clinical outcomes to chemotherapy, whereas tumours with unchanged COX expression were chemoresistant. In conclusion, previously undiscovered pre-existing defects in mitochondrial respiratory complexes cause cancer cells to become chemosensitive: mitochondrial defects lower the cells’ threshold for undergoing cell death in response to cisplatin. By contrast, cancer cells with intact mitochondrial respiratory complexes are chemoresistant and have a high threshold for cisplatin-induced cell death. This connection between mitochondrial respiration and chemosensitivity is relevant to anticancer therapeutics that target the mitochondrial electron transport chain. Copyright  2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

Keywords: COX7A2; MALDI imaging mass spectrometry; chemotherapy; oesophageal adenocarcinoma; tumour response

Received 21 December 2012; Revised 25 February 2013; Accepted 28 March 2013

No conflicts of interest were declared.

Introduction The incidence of oesophageal adenocarcinoma, Barrett’s cancer, is rising more rapidly than that of any other tumour in North America and Europe and most of the patients present with already advanced Copyright  2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk

tumour stages [1]. The most common therapeutic approach for these tumours is a multimodal treatment that includes pre-operative application of cis-diamminedichloroplatinum II (cisplatin) and 5-flurouracil (5-FU) chemotherapy, followed by resection [2–8]. Patients who respond to neoadjuvant J Pathol 2013; 230: 410–419 www.thejournalofpathology.com

Mitochondrial defects predict chemotherapy response

chemotherapy have a survival benefit [2,3,9,10]. However, neoadjuvant chemotherapy results in complete downsizing and downstaging of the oesophageal tumour, without any residual tumour in the resection specimen, in only 16% of patients [11,12]. In almost 50% of cases treated with neoadjuvant chemotherapy, the tumour responds only weakly, whereas the sideeffects to the patient are severe [9,13,14]. Therefore, it would be valuable to develop predictive tools to identify patients with worse predicted outcomes, so that such patients could be redirected to more efficacious alternative therapies. To date, no clinical assays exist that predict the response of oesophageal adenocarcinoma patients to neoadjuvant chemotherapy. Several studies have used microarray-based gene expression profiling to generate predictive and/or molecular subtype signatures [15,16], but these previous studies did not include analyses at the proteomic level. In this study we present the relationship between specific protein profiles gained by MALDI imaging mass spectrometry in pretherapeutic biopsies and the patients’ response to neoadjuvant chemotherapy, revealing a mechanism of pre-existing mitochondrial defects and response to chemotherapy.

Materials and methods

411

Table 1. Clinicopathological data of patients Discovery cohort (n = 23) (pretherapeutic biopsies)

Validation cohort (n = 46) (pretherapeutic biopsies)

Responder Non-responder Responder Non-responder Patients (AEG I)

10

13

18

28

Age (years) ≤ 65 ≥ 65

3 7

7 6

12 6

21 7

Histopathological response TRG 1 10 TRG 2 0 TRG 3 0

0 6 7

12 6 0

0 0 28

Sex Female Male

0 13

3 15

4 24

Tumour staging (after neoadjuvant chemotherapy) ypT0 5 0 5 ypT1 2 3 2 ypT2 2 1 6 ypT3 1 8 5 ypT4 0 0 0 ypTx 0 1 0 ypN0 8 7 9 ypN1 2 5 6 ypN2 0 0 0 ypN3 0 0 3 ypNx 0 1 0

0 2 4 21 1 0 5 18 2 5 0

1 9

Study population and tissue samples Pretherapeutic biopsy samples were obtained from patients with locally advanced (cT3-4 N0 or higher without distant metastases) oesophageal adenocarcinoma. Prior to this study, samples had been stored in the archive of the Institute of Pathology, Technische Universit¨at M¨unchen. All patients in this cohort subsequently underwent neoadjuvant chemotherapy with cisplatin in combination with 5-fluorouracil (5-FU) between 1996 and 2008 at the Department of Surgery of the Technische Universit¨at M¨unchen. The tumours were resected 3–4 weeks after completion of chemotherapy. Tumour regression in response to chemotherapy was assessed histopathologically according to a three-grade score, based on an estimation of the proportion of remaining vital tumour tissue relative to the size of the original tumour (the tumour bed): in tumour regression grade 1, residual tumour occupies < 10% of the tumour bed; in grade 2, 10–50%; and in grade 3, > 50% [17]. Patients with TRG1 were classified as responders, TRG2 and TRG3 as non-responders [9]. All tumour tissue specimens analysed in this study were obtained from patients who gave their written informed consent. This study was approved by the Ethics Committee of the Technische Universit¨at M¨unchen. Discovery cohort

Pretherapeutic biopsy samples of 23 oesophageal adenocarcinoma patients were snap-frozen in liquid Copyright  2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk

nitrogen immediately after endoscopy. Samples were stored at −196◦ C. The clinicopathological data regarding this discovery cohort are displayed in Table 1. Validation cohort

Formalin-fixed paraffin-embedded (FFPE) tissue samples from pretherapeutic biopsies of oesophageal adenocarcinoma were obtained from 46 patients. Clinicopathological data regarding this independent validation cohort are included in Table 1. Further patient cohorts

FFPE samples from independent patient cohorts with primary resected, non-neoadjuvant chemotherapytreated oesophageal adenocarcinoma (n = 81), breast cancer (n = 215) or gastric cancer (n = 63) were used for incidence analyses.

MALDI imaging mass spectrometry and identification of individual proteins MALDI imaging mass spectrometry was performed as described previously [18] (see Supporting information, Supporting information and methods). Protein identification in tissue sections was carried out as reported previously [19] (see Supporting information and methods). J Pathol 2013; 230: 410–419 www.thejournalofpathology.com

412

Immunohistochemistry, image analysis and mitochondrial quantitation and morphometry by transmission electron microscopy Immunohistochemical staining was performed on a Discovery XT automated stainer (Ventana, Tucson, AZ, USA) (see Supporting information and methods). Images obtained by immunohistochemistry were quantified using the Definiens Cognition Network Technology (Definiens AG, Munich, Germany) (see Supporting information and methods). Transmission electron microscopy (TEM) was performed according to standard procedures (see Supporting information and methods). Numbers of mitochondria were counted in TEM images of ultrathin sections and expressed relative to the area of cytoplasm present in the same section. The area of cytoplasm and the sizes of mitochondria were determined by defining regions of interest in the Enterprise Image Intelligence Suite.

Cell culture, transfection, and drug treatment For more details, see Supporting information and methods. The human oesophageal adenocarcinoma cell line OE19 (96071721, European Collection of Cell Cultures, Porton Down, Salisbury, UK) was maintained according to the provider’s instructions. Transfection experiments were performed according to standard procedures. Cell proliferation was determined using standard water-soluble tetrazolium WST-1 (05015944001, Roche Diagnostics, Mannheim, Germany) for spectrophotometric assay, according to the manufacturer’s protocol. Clonogenic cell survival of OE19 cells after siRNA transfection and treatment with cisplatin/5-FU was investigated by determining the fraction of surviving cells via standard colony-forming unit (CFU) assay. Slides from the cell culture experiments were washed in PBS and fixed with 4% paraformaldehyde for 15 min at 22◦ C. The slides were stained and processed as described above for immunohistochemistry. For immunoblotting, cells were pelleted and resuspended in extraction buffer, as described previously [20]. Immunoblotting was performed according to standard procedures.

Statistical analysis For data processing and statistical analysis of MALDI imaging mass spectrometry data, tumour-associated spectra were defined and extracted using the FlexImaging 3.0 software (Bruker Daltonik GmbH). Of these tumour-specific spectra, 100 were randomly picked for each sample and submitted to recalibration, normalization based on their total ion count in the observation mass range, and peak detection, utilizing ClinProTools 2.2 software (Bruker Daltonic GmbH). Significant differences in peak intensities of the m/z species between responders and non-responders were evaluated by the Wilcoxon rank-sum test, with a significance cut-off of p < 0.05. Hierarchical clustering of the expression profiles was carried out using the R software package Copyright  2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk

M Aichler et al

(v 2.13.0, R Project for Statistical Computing, Vienna, Austria) on the standardized, exported data. R is a free software environment for statistical computing and graphics. Statistical analysis for immunohistochemical and in vitro assay results are provided in Supporting information and methods.

Results Proteomic signature associated with response to neoadjuvant chemotherapy Pretherapeutic biopsy material from patients with oesophageal adenocarcinoma who subsequently underwent neoadjuvant chemotherapy with cisplatin/5-FU, followed by tumour resection, was analysed by MALDI imaging mass spectrometry (Table 1, Discovery Cohort). In total, 150–250 proteins in the mass-tocharge (m/z ) range 2000–23 000 were present in the proteomic signature of each biopsy. Specificity for tumour cells was ensured by defining regions of interest prior to analysis. The proteomic signatures of biopsies from responding and non-responding patients revealed discriminative proteomic spectra (Figure 1). Statistical analyses revealed 22 proteins that were significantly differentially expressed between responders and nonresponders (p < 0.05). Hierarchical clustering was performed on the eight proteins that contributed the most to discriminatory power (p < 0.05, AUC > 0.9). This analysis yielded a 98% separation between responding and non-responding patients (Figure 2A). For protein identification via LC–MS/MS, proteins were extracted from cryosections of the same tissue and separated by SDS–PAGE, followed by a tryptic in-gel digestion (Figure S2). Molecular weights of the database entries of the identified proteins were compared with the m/z values from the imaging experiment, where six masses could successfully be matched: cytochrome c oxidase polypeptide 7A2 (COX7A2); 40S ribosomal protein S27 (RPS27); FXYD domain-containing ion transport regulator 3 (FXYD3); cytochrome c oxidase subunit 6C (COX6C); cytochrome c oxidase subunit 6B1 (COX6b1); and complex I–MLRQ.

Reduced expression of mitochondrial proteins in responding patients Of the proteins identified, four were mitochondrial proteins. Three of these proteins, COX7A2, COX6B1 and COX6C, are subunits of complex IV of the mitochondrial respiratory chain, whereas complex I–MLRQ is a subunit of complex I. All four mitochondrial proteins exhibited significantly reduced expression in patients who responded to neoadjuvant chemotherapy. These results support a model in which reduced expression of mitochondrial respiratory chain proteins in solid tumours has a strong impact on the response of cancer cells to treatment with cisplatin (Figure 3). Reduction in the levels of mitochondrial respiratory chain J Pathol 2013; 230: 410–419 www.thejournalofpathology.com

Mitochondrial defects predict chemotherapy response

413

Figure 1. Discriminative protein spectra from two individual patients who responded or did not respond to neoadjuvant chemotherapy, detected by MALDI imaging mass spectrometry of pretherapeutic oesophageal adenocarcinoma biopsies. (A) Average spectrum and haematoxylin and eosin (H&E) staining of a sample of non-responding patient tissue; scale bar = 25 µm. (B) Average spectrum and H&E staining of a sample of responding patient tissue; *significantly differentially expressed proteins. In this example, two proteins (m/z = 2756 and 4156) are more strongly expressed in non-responding patients.

proteins, resulting in defective mitochondria, lowers the threshold for cell death and thereby increases the sensitivity of tumour cells to chemotherapy.

Effects of reduced expression of the mitochondrial protein COX7A2 To obtain more insight into the mechanism proposed above, the most discriminating protein, COX7A2 (m/z = 6722), was further functionally characterized by in vitro analyses. Using electron microscopy (EM), the impact of COX7A2 expression on mitochondrial integrity in pretherapeutic biopsies of oesophageal adenocarcinoma patients (n = 6) was investigated. In patients with reduced expression of COX7A2 (Figure 2B, C), mitochondria of oesophageal adenocarcinoma cancer cells were swollen, rounded in shape and exhibited loss of cristae. By contrast, in patients with normal expression of COX7A2, the mitochondria were round and electron-dense, with lamellar or plate-like cristae (Figure 2D). The oesophageal adenocarcinoma cell line OE19 was chosen as a model system for functional analysis. This cell line is characterized by normal levels of COX7A2 expression, which we confirmed by immunoblot analysis and immunocytology (see supplementary material, Figure S1). COX7A2 expression in OE19 could be knocked down by transfection with specific siRNA (see supplementary material, Figure S1). In EM analysis, OE19 cells with normal COX7A2 expression showed mitochondria that contained an electron-dense matrix and lamellar cristae (Figure 4A). Upon knockdown of COX7A2 by RNA interference, mitochondria exhibited Copyright  2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk

an abnormal cup-shaped structure (Figure 4A). The number of mitochondria in siRNA-transfected cells increased significantly relative to the number in control cells (Figure 4B), whereas the size of mitochondria remained unchanged (Figure 4C).

Sensitivity of cancer cells with mitochondrial alterations OE19 cells were transfected with COX7A2-specific siRNA, and then treated with the chemotherapeutic drugs. In electron micrographs, the mitochondria of treated cells appeared swollen and rounded in shape, with broken cristae. The cytoplasm of these cells contained autolysosomes bearing mitochondrial residue, indicating mitochondrial degradation (Figure 4A). The number of mitochondria significantly decreased in treated cells (Figure 4B), whereas the size of the mitochondria significantly increased due to swelling (Figure 4C). Mitochondria of oesophageal adenocarcinoma cells directly treated with cisplatin/5-FU only exhibited reduced cristae and formation of vacuoles, whereas mitochondria of control cells remained ultrastructurally unchanged (Figure 4A). Treatment with cisplatin/5-FU changed neither the number nor the size of mitochondria relative to control cells (Figure 4B, C). OE19 cells transfected with COX7A2-specific siRNA or untransfected cells treated with cisplatin/5FU exhibited a statistically significant decrease in proliferation. Treatment of OE19 cells with cisplatin/5FU after knockdown of COX7A2 caused a further decrease in proliferation relative to that of COX7A2silenced cells or cisplatin/5-FU-treated untransfected J Pathol 2013; 230: 410–419 www.thejournalofpathology.com

414

Figure 2. Association of reduced expression of mitochondrial respiratory chain proteins with the response to cisplatin-based neoadjuvant chemotherapy. (A) Hierarchical clustering of patientspecific MALDI imaging mass spectrometry profiles of the eight proteins (p < 0.05, AUC > 0.9) that most accurately discriminated responders from non-responders. (B) The protein whose profile best discriminated responders from non-responders was identified as COX7A2; MALDI imaging mass spectrometry revealed that COX7A2 is expressed in non-responding patients but absent in responding patients. (C) COX7A2 in pretherapeutic biopsies of oesophageal adenocarcinoma visualized by immunohistochemistry. The immunohistochemical validation of m/z 6722 (corresponding to COX7A2 protein) expression confirmed the results of MALDI imaging mass spectrometry of responding and non-responding patients; scale bar = 100 µm. (D) Electron microscopy revealed mitochondrial defects in patients with low COX7A2 expression, whereas mitochondria were normal in cases with high COX7A2 expression; scale bar = 700 nm.

cells (Figure 5A). Transfected OE19 cells with COX7A2 siRNA and/or treated with cisplatin/5-FU resulted in a statistically significant decrease in cell viability in colony formation assay. Treatment of OE19 cells with cisplatin/5-FU after silencing of COX7A2 caused a further decrease in colony formation relative to that of COX7A2-silenced cells or cisplatin/5-FU only-treated untransfected cells (Figure 5B, Table S1). Copyright  2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk

M Aichler et al

Figure 3. Schematic model of the hypothesized mechanism underlying the predisposition to chemosensitivity in patients with pre-existing defects in mitochondria. Reduced expression of mitochondrial respiratory chain proteins causes defective mitochondria. These defects reduce the threshold for cell death in cancer cells and thus increase the sensitivity of tumour cells to chemotherapy.

Validation of COX7A2 for response prediction The robustness of the proteomic signatures obtained with MALDI imaging mass spectrometry and the impact of the results of the in vitro studies was clarified by studying pretherapeutic samples from an independent patient cohort with oesophageal adenocarcinoma (n = 46, Table 1). Pretherapeutic biopsies were immunohistochemically stained for COX7A2 (Figure 6A) and the intensity of staining was determined by image analysis. Statistical analysis of the COX7A2 staining intensities in responding and non-responding patients revealed that expression of COX7A2 was strongly associated with the response to neoadjuvant chemotherapy (Figure 6A).

Incidence of reduced expression of COX7A2 in other types of cancer The incidence of reduced expression of COX7A2 was evaluated in independent cohorts of patients with J Pathol 2013; 230: 410–419 www.thejournalofpathology.com

Mitochondrial defects predict chemotherapy response

415

Figure 4. Influence of COX7A2 on mitochondria. (A) OE19 control cells exhibited normal mitochondria with an electron-dense matrix. Knockdown of COX7A2 resulted in cup-shaped mitochondria, whereas treatment with cisplatin/5-FU alone resulted in mitochondria with a reduced concentration of cristae and formation of vacuoles. COX7A2-depleted OE19 cells treated with chemotherapeutic drugs exhibited extensively swollen and rounded mitochondria containing distorted cristae. These cells also contained autolysosomes bearing mitochondrial residue (arrow); scale bar = 200 nm. (B) Knockdown of COX7A2 significantly increased the number of mitochondria per µm2 cytoplasm (p = 0.007), whereas subsequent treatment with cisplatin/5-FU significantly reduced the number of mitochondria (p < 0.001) (C) Knockdown of COX7A2 did not affect the size of mitochondria, whereas subsequent treatment with cisplatin significantly increased mitochondrial size.

different types of tumours (oesophageal adenocarcinoma, n = 81 primary resected tumours without treatment with neoadjuvant chemotherapy); breast cancer, n = 215; gastric cancer, n = 63) (Figure 6B). Immunohistochemical staining revealed that 31–68% of patients, depending on cancer type, exhibited reduced expression of COX7A2 (Figure 6B). Thus, the mechanism we initially discovered in oesophageal adenocarcinoma may be of general relevance to cisplatinbased treatment of many types of solid tumours. Copyright  2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk

Discussion Neoadjuvant chemotherapy has been widely adopted as a practical means to improve surgical outcomes for patients with locally advanced oesophageal adenocarcinomas [2–10]. Little is known, however, about the factors that influence the response to cisplatin-based neoadjuvant chemotherapy, especially in regard to post-therapeutic pathological staging. Knowledge of such factors would contribute to more accurate J Pathol 2013; 230: 410–419 www.thejournalofpathology.com

416

M Aichler et al

Figure 5. Influence of COX7A2 knockdown on cell proliferation and viability. OE19 cells were transfected with siRNA targeting COX7A2 or mock-transfected and then treated or not with cisplatin/5-FU. (A) Proliferation: the graph represents the number of cells relative to the number in the control (untransfected, untreated) sample (mean ± SD, n = 3); representative pictures of typical cell growth are depicted on the left. (B) Colony-forming unit (CFU) assay: the graph represents the number of colonies relative to the control sample (mean ± SD, n = 5); representative pictures of the CFU assay are shown on the left.

prognoses and aid decisions regarding alternative treatment strategies. We integrated pathological staging parameters of response after neoadjuvant chemotherapy with patient-specific protein expression profiles generated from pretherapeutic biopsy samples by MALDI imaging mass spectrometry (Figure 1). We were able to identify a proteomic signature that is predictive for response to neoadjuvant chemotherapy. Of particular note, patients responding to neoadjuvant chemotherapy exhibited significantly reduced expression of specific mitochondrial respiratory chain proteins (Figure 2A, B). By striking contrast, nonresponding patients did not exhibit such proteomic alterations. In responding patients, reduced expression of these proteins was associated with mitochondrial defects (Figure 2C, D). Likewise, siRNA-mediated knockdown of these proteins in an oesophageal adenocarcinoma cell line caused similar mitochondrial abnormalities in vitro (Figures 4, 5). Proteomic expression profiling has a high potential to improve prediction of patient outcomes. The identified signature reveals the impact of mitochondria on the response to chemotherapy. We chose the most relevant Copyright  2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk

of the mitochondrial proteins we identified, COX7A2 (cytochrome c oxidase subunit 7A2), for characterization in functional studies. COX7A2 is a member of mitochondrial respiratory chain complex IV [21,22]. COX7A2 is encoded by a nuclear gene [22–26] and plays roles in the dimerization and structural integrity of the complex IV, thereby regulating its activity [27–30]. To date, COX7A2 has not been well characterized, neither has it been evaluated in the context of cancer or the response to anticancer therapy. In this study, reduced expression of COX7A2 was observed in a subpopulation of oesophageal adenocarcinoma patients (Figures 2A–D, 6A) who responded to neoadjuvant chemotherapy. Pretherapeutic biopsy material derived from these patients exhibited alterations in mitochondrial morphology (Figure 2D). Therefore, we hypothesized that reduced expression of mitochondrial respiratory chain proteins causes these alterations in mitochondria and that these alterations predispose the patients to respond to neoadjuvant chemotherapy (Figure 3). This hypothesis is in accordance with the results of an in vitro study showing that response to chemotherapy correlates with, and may be partially J Pathol 2013; 230: 410–419 www.thejournalofpathology.com

Mitochondrial defects predict chemotherapy response

Figure 6. Validation of COX7A2 as a predictor of the response to chemotherapy. (A) Immunohistochemical detection of COX7A2 in pretherapeutic biopsies of an independent cohort of patients with oesophageal adenocarcinoma; scale bar = 100 µm. COX7A2 expression was significantly associated with the chemotherapy response in the validation patient cohort (p = 0.0015). (B) Immunohistochemical COX7A2 protein expression in pretherapeutic biopsies of another independent cohort of patients with oesophageal, breast and gastric cancer.

governed by, the pretreatment proximity of tumour cell mitochondria to the apoptotic threshold (mitochondrial priming) [31]. The effect of mitochondrial preconditioning before a subsequent treatment could be a major determinant of whether cells live or die. The status of mitochondria prior to treatment might therefore serve as a powerful predictor of cancer response to chemotherapy [32]. Mitochondrial damage induced by cisplatin reduces the proliferation and viability of cells [33], consistent with the results of our in vitro experiments: cisplatin/5FU treatment altered the integrity of mitochondria and significantly decreased both the proliferation and viability of cells (Figures 4A–C, 5A, B). Pretreating the mitochondria of oesophageal adenocarcinoma cells by knocking down COX7A2 before cisplatin/5-FU treatment had a cumulative effect on cells, causing more severe mitochondrial alterations and further reducing proliferation and viability (Figures 4A–C, 5A, B). These results are in line with the theory of mitochondrial priming [31,32] and observations of enhanced Copyright  2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk

417

cell death response after mitochondrial pretreatment [31,32]. Several studies have shown that the respiratory chain of mitochondria, of which COX7A2 is a member, plays a key role in the survival of cancer cells [34,35]. Indeed, several anti-cancer compounds target sites in the mitochondria [36,37]. There are even pharmacological compounds, eg prophyrin photosensitizers, N -(4-hydroxyphenyl) retinamide (fenretinid) or doxorubicin, that specifically target mitochondrial respiratory chain complex IV, of which COX7A2 is a component, inducing cancer cells to undergo death [38–47]. Such compounds might be useful in mitochondrial priming in order to sensitize cancer cells to neoadjuvant chemotherapy. Such a strategy is very promising, especially in the case of patients with chemotherapy-resistant tumours, who could be sensitized with such compounds and thereby respond successfully to cisplatin-based neoadjuvant chemotherapy. Further studies will be required to determine whether pretreatment of oesophageal adenocarcinoma cells with pharmacological compounds targeting complex IV, and the resulting effects on the condition of mitochondria, can improve the outcomes of neoadjuvant chemotherapy strategies in non-responding patients. We also investigated other cancer types that are commonly treated with cisplatin-based neoadjuvant chemotherapy. In populations of patients with oesophageal adenocarcinoma, breast or gastric cancer, we observed reduced expression levels of COX7A2 (Figure 6B). A recent study of breast cancer models reported that combinations of mitochondria-targeting drugs act synergistically to induce cancer cells to undergo cell death [48]. These results support our hypothesis regarding the potential value of pretreatment strategies targeting mitochondria, in terms of reducing the survival of cancer cells. Previous investigations of the response of oesophageal adenocarcinoma to neoadjuvant chemotherapy did not reveal COX7A2 as a predictive marker. However, all of these studies investigated this response at the level of mRNA expression, which is not necessarily reflected at the protein level. However, the expression programme for certain COX genes is fixed: they exhibit unaltered mRNA levels, even when cytochrome c oxidase is knocked down [49]. For that reason, it is possible that studies at the mRNA expression level would have been unable to detect the specific mitochondrial alterations we describe here, because these changes only can be detected at the proteomic level. Furthermore, our approach has the additional advantage of high sensitivity: the MALDI imaging mass spectrometry technique we used enabled us to analyse pretherapeutic biopsy samples that are too small for use in other proteomic analytical techniques, allowing us to reveal a discriminating mitochondrial signature that is beneath the detection threshold of earlier studies. Thus, MALDI imaging mass spectrometry might represent a novel tool for tissue-based research, eg for the analysis of animal J Pathol 2013; 230: 410–419 www.thejournalofpathology.com

418

models, drug candidates or endoscopic biopsies from patients [50]. The identified proteomic signature is associated with response of oesophageal adenocarcinoma to cisplatin-based neoadjuvant chemotherapy. This signature was validated in an independent dataset. Several proteins of the mitochondrial respiratory chain, including COX7A2, were differentially regulated between responding and non-responding patients. The identification of mitochondrial proteins provided novel insights into mechanisms underlying response to cisplatin-based neoadjuvant chemotherapy, and the role of mitochondrial priming in lowering the threshold for sensitivity to anti-cancer drugs in solid tumours.

Acknowledgements The authors would like to thank Ulrike Buchholz, Claudia-Mareike Pfl¨uger, Luise Jennen, Michaela H¨ausler and Andreas Voss for excellent technical assistance. The authors gratefully acknowledge the financial support of the BMBF (Grant Nos 01EZ0803, 0315508A and 01IB10004E) and the Deutsche Forschungsgemeinschaft (Grant Nos SFB 824 TP B1, SFB 824 TP Z02 and WA 1656/3-1) to AW.

Author contributions MAi and AW conceived the study, performed the data analyses and wrote the manuscript; ME, NL, AF, VZ, CS, LH, SM and SR assisted in generation, analysis and interpretation of the data; MAu, HZis, MS, HZit and HH participated in interpretation of the data; MF and RL collected the clinical data and contributed to interpretation; BB and HB assisted in generating statistical analysis; and SM, SMH, MU and BK were responsible for the protein identification and interpretation. All authors were involved in writing the manuscript and had final approval of the submitted version.

References 1. Bollschweiler E, Wolfgarten E, Gutschow C, et al. Demographic variations in the rising incidence of esophageal adenocarcinoma in white males. Cancer 2001; 92: 549–555. 2. Fink U, Stein HJ, Siewert JR. [Multimodal therapy of tumors of the upper gastrointestinal tract]. Chirurgie 1998; 69: 349–359. 3. Zacherl J, Sendler A, Stein HJ, et al. Current status of neoadjuvant therapy for adenocarcinoma of the distal esophagus. World J Surg 2003; 27: 1067–1074. 4. Swisher SG, Pisters PW, Komaki R, et al. Gastroesophageal junction adenocarcinoma. Curr Treat Options Oncol 2000; 1: 387–398. 5. Schauer M, Stein H, Lordick F, et al. Results of a multimodal therapy in patients with stage IV Barrett’s adenocarcinoma. World J Surg 2008; 32: 2655–2660. 6. van Hagen P, Hulshof MC, van Lanschot JJ, et al. Preoperative chemoradiotherapy for esophageal or junctional cancer. N Engl J Med 2012; 366: 2074–2084. Copyright  2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk

M Aichler et al

7. Cunningham D, Allum WH, Stenning SP, et al. Perioperative chemotherapy versus surgery alone for resectable gastroesophageal cancer. N Engl J Med 2006; 355: 11–20. 8. Ychou M, Boige V, Pignon JP, et al. Perioperative chemotherapy compared with surgery alone for resectable gastroesophageal adenocarcinoma: an FNCLCC and FFCD multicenter phase III trial. J Clin Oncol 2011; 29: 1715–1721. 9. Becker K, Fumagalli U, Mueller JD, et al. Neoadjuvant chemotherapy for patients with locally advanced gastric carcinoma: effect on tumor cell microinvolvement of regional lymph nodes. Cancer 1999; 85: 1484–1489. 10. Walsh TN, Noonan N, Hollywood D, et al. A comparison of multimodal therapy and surgery for esophageal adenocarcinoma. N Engl J Med 1996; 335: 462–467. 11. Torres C, Turner JR, Wang HH, et al. Pathologic prognostic factors in Barrett’s associated adenocarcinoma: a follow-up study of 96 patients. Cancer 1999; 85: 520–528. 12. Siewert JR, Lordick F, Ott K, et al. Induction chemotherapy in Barrett cancer: influence on surgical risk and outcome. Ann Surg 2007; 246: 624–628; discussion, 628–631. 13. Lordick F, Ott K, Krause BJ, et al. PET to assess early metabolic response and to guide treatment of adenocarcinoma of the oesophagogastric junction: the MUNICON phase II trial. Lancet Oncol 2007; 8: 797–805. 14. Ott K, Weber WA, Lordick F, et al. Metabolic imaging predicts response, survival, and recurrence in adenocarcinomas of the esophagogastric junction. J Clin Oncol 2006; 24: 4692–4698. 15. Langer R, Specht K, Becker K, et al. Association of pretherapeutic expression of chemotherapy-related genes with response to neoadjuvant chemotherapy in Barrett carcinoma. Clin Cancer Res 2005; 11: 7462–7469. 16. Langer R, Specht K, Becker K, et al. Comparison of pretherapeutic and posttherapeutic expression levels of chemotherapyassociated genes in adenocarcinomas of the esophagus treated by 5-fluorouracil- and cisplatin-based neoadjuvant chemotherapy. Am J Clin Pathol 2007; 128: 191–197. 17. Langer R, Ott K, Feith M, et al. Prognostic significance of histopathological tumor regression after neoadjuvant chemotherapy in esophageal adenocarcinomas. Mod Pathol 2009; 22: 1555–1563. 18. Rauser S, Marquardt C, Balluff B, et al. Classification of HER2 receptor status in breast cancer tissues by MALDI imaging mass spectrometry. J Proteome Res 2010; 9: 1854–1863. 19. Elsner M, Rauser S, Maier S, et al. MALDI imaging mass spectrometry reveals COX7A2, TAGLN2 and S100-A10 as novel prognostic markers in Barrett’s adenocarcinoma. J Proteomics 2012; 75: 4693–4704. 20. Ludyga N, Anastasov N, Gonzalez-Vasconcellos I, et al. Impact of protein tyrosine kinase 6 (PTK6) on human epidermal growth factor receptor (HER) signalling in breast cancer. Mol Biosyst 2011; 7: 1603–1612. ˚ resolution 21. Iwata S, Ostermeier C, Ludwig B, et al. Structure at 2.8 A of cytochrome c oxidase from Paracoccus denitrificans. Nature 1995; 376: 660–669. 22. Kadenbach B, Jarausch J, Hartmann R, et al. Separation of mammalian cytochrome c oxidase into 13 polypeptides by a sodium dodecyl sulfate–gel electrophoretic procedure. Anal Biochem 1983; 129: 517–521. 23. Kuhn-Nentwig L, Kadenbach B. Isolation and characterization of human heart cytochrome c oxidase. J Bioenerg Biomembr 1986; 18: 307–314. 24. Capaldi RA. Structure and function of cytochrome c oxidase. Annu Rev Biochem 1990; 59: 569–596. 25. Cooper CE, Nicholls P, Freedman JA. Cytochrome c oxidase: structure, function, and membrane topology of the polypeptide subunits. Biochem Cell Biol 1991; 69: 586–607. J Pathol 2013; 230: 410–419 www.thejournalofpathology.com

Mitochondrial defects predict chemotherapy response

26. Taanman JW. Human cytochrome c oxidase: structure, function, and deficiency. J Bioenerg Biomembr 1997; 29: 151–163. 27. Huttemann M, Lee I, Pecinova A, et al. Regulation of oxidative phosphorylation, the mitochondrial membrane potential, and their role in human disease. J Bioenerg Biomembr 2008; 40: 445–456. 28. Fontanesi F, Soto IC, Barrientos A. Cytochrome c oxidase biogenesis: new levels of regulation. IUBMB Life 2008; 60: 557–568. 29. Helling S, Vogt S, Rhiel A, et al. Phosphorylation and kinetics of mammalian cytochrome c oxidase. Mol Cell Proteom 2008; 7: 1714–1724. 30. Lazarou M, Smith SM, Thorburn DR, et al. Assembly of nuclear DNA-encoded subunits into mitochondrial complex IV, and their preferential integration into supercomplex forms in patient mitochondria. FEBS J 2009; 276: 6701–6713. 31. Ni Chonghaile T, Sarosiek KA, Vo TT, et al. Pretreatment mitochondrial priming correlates with clinical response to cytotoxic chemotherapy. Science 2011; 334: 1129–1133. 32. Deng J, Carlson N, Takeyama K, et al. BH3 profiling identifies three distinct classes of apoptotic blocks to predict response to ABT-737 and conventional chemotherapeutic agents. Cancer Cell 2007; 12: 171–185. 33. Park MS, De Leon M, Devarajan P. Cisplatin induces apoptosis in LLC-PK1 cells via activation of mitochondrial pathways. J Am Soc Nephrol 2002; 13: 858–865. 34. Rohlena J, Dong LF, Kluckova K, et al. Mitochondrially targeted α-tocopheryl succinate is antiangiogenic: potential benefit against tumor angiogenesis but caution against wound healing. Antioxid Redox Signal 2011; 15: 2923–2935. 35. Mathupala SP, Ko YH, Pedersen PL. The pivotal roles of mitochondria in cancer: Warburg and beyond and encouraging prospects for effective therapies. Biochim Biophys Acta 2010; 1797: 1225–1230. 36. Neuzil J, Dong LF, Ramanathapuram L, et al. Vitamin E analogues as a novel group of mitocans: anti-cancer agents that act by targeting mitochondria. Mol Aspects Med 2007; 28: 607–645. 37. Neuzil J, Tomasetti M, Zhao Y, et al. Vitamin E analogs, a novel group of ‘mitocans’, as anticancer agents: the importance of being redox-silent. Mol Pharmacol 2007; 71: 1185–1199. 38. You KR, Wen J, Lee ST, et al. Cytochrome c oxidase subunit III: a molecular marker for N -(4-hydroxyphenyl)retinamise-induced oxidative stress in hepatoma cells. J Biol Chem 2002; 277: 3870–3877.

419

39. Slapke J, Schewe T, Hummel S, et al. Lung strips from guinea pigs as test system for lipoxygenase inhibitors. Inhibition of arachidonic acid-induced contractions by 3-t-butyl-4-hydroxyanisole and nordihydroguaiaretic acid. Biomed Biochim Acta 1983; 42: 1309–1318. 40. Hilf R. Mitochondria are targets of photodynamic therapy. J Bioenerg Biomembr 2007; 39: 85–89. 41. Kelso GF, Porteous CM, Coulter CV, et al. Selective targeting of a redox-active ubiquinone to mitochondria within cells: antioxidant and antiapoptotic properties. J Biol Chem 2001; 276: 4588–4596. 42. Dong LF, Jameson VJ, Tilly D, et al. Mitochondrial targeting of vitamin E succinate enhances its pro-apoptotic and anti-cancer activity via mitochondrial complex II. J Biol Chem 2011; 286: 3717–3728. 43. Dong LF, Jameson VJ, Tilly D, et al. Mitochondrial targeting of α-tocopheryl succinate enhances its pro-apoptotic efficacy: a new paradigm for effective cancer therapy. Free Radic Biol Med 2011; 50: 1546–1555. 44. Biasutto L, Mattarei A, Marotta E, et al. Development of mitochondria-targeted derivatives of resveratrol. Bioorg Med Chem Lett 2008; 18: 5594–5597. 45. Mattarei A, Biasutto L, Marotta E, et al. A mitochondriotropic derivative of quercetin: a strategy to increase the effectiveness of polyphenols. Chem Biochem 2008; 9: 2633–2642. 46. Asayama S, Kawamura E, Nagaoka S, et al. Design of manganese porphyrin modified with mitochondrial signal peptide for a new antioxidant. Mol Pharm 2006; 3: 468–470. 47. Cuchelkar V, Kopeckova P, Kopecek J. Novel HPMA copolymerbound constructs for combined tumor and mitochondrial targeting. Mol Pharm 2008; 5: 776–786. 48. Cheng G, Zielonka J, Dranka BP, et al. Mitochondria-targeted drugs synergize with 2-deoxyglucose to trigger breast cancer cell death. Cancer Res 2012; 72: 2634–2644. 49. Fornuskova D, Stiburek L, Wenchich L, et al. Novel insights into the assembly and function of human nuclear-encoded cytochrome c oxidase subunits 4, 5a, 6a, 7a and 7b. Biochem J 2010; 428: 363–374. 50. Balluff B, Rauser S, Ebert MP, et al. Direct molecular tissue analysis by MALDI imaging mass spectrometry in the field of gastrointestinal disease. Gastroenterology 2012; 143: 544–549.

SUPPORTING INFORMATION ON THE INTERNET The following Supporting information may be found in the online version of this article: Supporting information and methods Figure S1. siRNA knockdown of COX7A2 in OE19 cells. Figure S2. Protein identification from cryo-sections. Table S1. Interaction coefficients of cell viability assay.

Copyright  2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk

J Pathol 2013; 230: 410–419 www.thejournalofpathology.com

Lihat lebih banyak...

Comentários

Copyright © 2017 DADOSPDF Inc.