The first 2 authors contributed equally to the current work.
Hsp60 and Hsp10 down-regulation predicts bronchial epithelial carcinogenesis in smokers with chronic obstructive pulmonary disease
Article first published online: 17 OCT 2006
Copyright © 2006 American Cancer Society
Volume 107, Issue 10, pages 2417–2424, 15 November 2006
How to Cite
Cappello, F., Di Stefano, A., David, S., Rappa, F., Anzalone, R., La Rocca, G., D'Anna, S. E., Magno, F., Donner, C. F., Balbi, B. and Zummo, G. (2006), Hsp60 and Hsp10 down-regulation predicts bronchial epithelial carcinogenesis in smokers with chronic obstructive pulmonary disease. Cancer, 107: 2417–2424. doi: 10.1002/cncr.22265
- Issue published online: 8 NOV 2006
- Article first published online: 17 OCT 2006
- Manuscript Accepted: 21 AUG 2006
- Manuscript Revised: 18 AUG 2006
- Manuscript Received: 3 APR 2006
- chaperone expression;
- lung obstruction;
- lung tumors
The relation between smoking, chronic obstructive pulmonary disease (COPD), and lung cancer (LC) is an open field of investigation. A higher frequency of adenocarcinoma has been reported in patients with COPD. Heat shock proteins (Hsps) are implicated in tumoral cell growth and differentiation. The aim of the present study was to investigate the expression of Hsp60 and Hsp10 in bronchial biopsies from smokers with COPD and in 10 lung cancer patients and to evaluate the association between Hsps expression and carcinogenetic steps of LC.
An immunohistochemical study was performed for Hsp60 and Hsp10 in bronchial biopsies from 35 COPD (postbronchodilator forced expiratory volume in 1 second [FEV1]: 53 ± 19% [mean ± SD]) patients with a history of smoking (53 ± 34 pack/years) and in 10 patients with adenocarcinoma or adenosquamous carcinoma (ASC). Immunopositivity was quantified in the bronchial epithelium and in specimens with ASC.
In smokers with COPD, 10 out of 35 patients had a normal bronchial epithelium (NBE), 12 showed basal cell hyperplasia (BCH), 5 squamous metaplasia (SM), and 8 dysplasia (Dy). It was found that 58 ± 23% and 54 ± 23% of NBE and 48 ± 29% and 52 ± 26% of BCH expressed Hsp60 and Hsp10, respectively; in contrast, only 3 ± 3% and 3.6 ± 2% of SM, 1.9 ± 4% and 1.1 ± 2% of Dy expressed Hsp60 and Hsp10, respectively. ASC specimens were negative for Hsps proteins. Interestingly, NBE also present at the edges of ASC specimens was negative for Hsps proteins.
The loss of Hsp60 and Hsp10 immunopositivity is related to the development and progression of bronchial cancer in smokers with COPD. Cancer 2006. © 2006 American Cancer Society.
Lung cancer (LC) is still the leading cause of tumor death worldwide. It is well known that LC is preceded by a sequence of preneoplastic lesions. In 1999, the World Health Organization (WHO) introduced a new classification system1 that defines preneoplastic lesions of LC: basal cells hyperplasia (BCH) and squamous metaplasia (SM) may lead to squamous dysplasia (SD) and, subsequently, adenocarcinoma or adenosquamous carcinoma (ASC).
The prognosis for patients with LC is strictly related to the time of diagnosis; indeed, despite advances in therapy the overall survival rate for lung cancer remains only 15% due to the relatively advanced stage of disease at diagnosis.2
It is well accepted that smoking is associated with about 90% of cases of lung cancer3 and that adenocarcinoma or adenosquamous carcinoma has a prevalent association with tobacco consumption.4, 5 Moreover, chronic obstructive pulmonary disease (COPD) is characterized by a slowly progressive and irreversible deterioration of lung function. Cigarette smoking is the major risk factor for the development of COPD, cigarette smokers constituting >90% of all COPD patients.6 Epidemiologic data also show that about 20% of all cigarette smokers develop COPD in industrialized countries.7
The relation between smoking, COPD, and LC is also an open field of investigation. Indeed, epidemiologic studies have shown that the incidence of lung cancer in smokers with COPD is 4 to 5 times that reported in smokers without COPD or with chronic bronchitis alone.8 This finding was more recently confirmed by other authors who showed that in smokers with COPD the rate of development of LC was about 4 times that of the control group of smokers without COPD matched for age, sex, occupation, and smoking history.9 The risk of LC was also increased in proportion to the degree of airways obstruction in moderate to severe COPD patients.10 Because an increased risk of LC was also reported in nonsmokers with impaired lung function,11 the presence of airways obstruction seems to be an independent as well as an additive risk factor for LC development. The shared genetic susceptibility for COPD and lung cancer is under constant investigation.3
Heat shock proteins (Hsps), a class of molecules of fundamental importance for cell life, have been found overexpressed in a wide range of human cancers, due probably to the fact that they are implicated in tumoral cell growth and differentiation. Hence, measurements of Hsp levels may help indicate the presence of abnormal changes during the process of carcinogenesis.12
In particular, Hsp60 and Hsp10 are 2 mitochondrial molecular chaperones that, when overexpressed, protect cells from ischemia/hypoxic/oxidative injury, thus maintaining mitochondrial integrity and function and, as a consequence, suppressing mitochondrial membrane permeability and inhibiting apoptotic and necrotic cell death.13 We previously showed overexpression of Hsp60 and Hsp10 during exocervical,14, 15 colorectal,15, 16 and prostatic17 carcinogenesis. These results were also confirmed by other studies.18, 19 Nevertheless, as Lebret et al.20 first demonstrated, Hsp60 can also down-regulate its expression in vesical cancer.
Some authors studied the relation between Hsp60 levels and the prognosis of tumoral patients, reporting contrasting results; indeed, Hsp60 overexpression was correlated with a worse prognosis in acute myeloid leukemia21 and bladder carcinoma,20 and with a more favorable outcome in esophageal22 and ovarian23 cancer and in osteosarcoma.24 By contrast, no consistent correlation was found between levels of Hsp60 and the behavior of primary prostatic cancer.25 In particular, Faried et al.22 found a correlation between Hsp60 overexpression and the apoptotic index, suggesting that Hsp60 may trigger apoptosis of tumoral cells.
Both pro- and antiapoptotic effects of Hsp60 overexpression have been demonstrated. In particular, Samali et al.26 published a first report suggesting that Hsp60 is involved in apoptosis. Those authors showed that activation of caspase-3 by staurosporine occurs simultaneously with Hsp60 and Hsp10 release from mitochondria. Furthermore, in vitro and in vivo Hsp60 and Hsp10 were shown to associate with pro-caspase3 and possibly accelerate its activation. On the other hand, Kirchhoff et al.27 argued an antiapoptotic role of cytosolic Hsp60 in cardiac myocytes, demonstrating that Hsp60 interacts with proapoptotic Bax and Bak proteins and thus prevents apoptosis onset. However, to date, the exact molecular role of Hsp60 is still far from understood.
We recently proposed that Hsp60 down-regulation may be a novel biomarker during bronchial carcinogenesis,28 although an extensive analysis and the implications of this preliminary observation were not reported.
The aim of the present work was to study the expression of both Hsp60 and Hsp10 in a series of bronchial biopsies (BB) from smokers with a wide range of airways obstruction resembling the carcinogenetic steps of LC. We also aimed to investigate the relation between Hsps expression and LC progression and the association between Hsps and airways obstruction in smokers at higher risk for LC development such as patients suffering from COPD.
MATERIALS AND METHODS
We studied 47 consecutive bronchial biopsies from 35 patients with a history of cigarette smoking (pack/years consumption: 53 ± 34 [mean ± SD]) affected by moderate to very severe COPD5 (postbronchodilator forced expiratory volume in 1 second [FEV-1]: 53 ± 19%). Table 1 summarizes the demographic characteristics of COPD patients.
|Subjects||Age, y||Sex (M/F)||FEV1% predicted||FEV1% Post-β2||Pack/Y||HSP60 (%)||HSP10 (%)||Epithelial diagnosis (n)|
|COPD Patients (n = 35)||67 ± 8||34/1||49 ± 18||53 ± 19||53 ± 34||34 ± 32||34 ± 31||(8) Dy(5)Me(12)Ip(10)No|
Moreover, we selected from our pathology files 10 specimens of adenocarcinoma or adenosquamous (ASC) carcinoma from among smoking patients without a history of COPD and 5 specimens of subjects with normal epithelium (NM) from nonsmoking patients.
The study conformed to the Declaration of Helsinki and informed consent was obtained from each subject. Bronchial biopsies were performed according to the local Ethics Committee Guidelines.
Lung Function Tests, Fiberoptic Bronchoscopy, Collection and Processing of Bronchial Biopsies in Patients With COPD
Recruitment of patients with COPD, lung function tests, fiberoptic bronchoscopy, and collection and processing of bronchial biopsies in patients with COPD were performed at the Division of Pulmonary Diseases, Rehabilitation Center of Veruno (NO), Italy, from 1997 to 2005 as part of a wider study project involving the patients included in the present study. Archival specimens with bronchial carcinoma and specimens with normal epithelium were obtained from the Pathology Unit of Palermo, Italy.
Pulmonary function tests, indicating the degree of bronchial obstruction, included measurements of FEV1 and forced vital capacity (FVC) before and after the inhalation of 400 μg of albuterol (6200 Autobox Pulmonary Function Laboratory; Sensormedics, Yorba Linda, CA). The predicted normal values used were those from the European Community for Steel and Coal.29 Fiberoptic bronchoscopy and bronchial biopsies were performed and processed as previously described.30, 31
Histochemistry and Immunohistochemistry
Paraffin-embedded or formalin-fixed frozen sections were stained with histochemical and immunohistochemical methods. Hematoxylin-eosin (H&E) staining was applied for pathologic assessment and identification of normal epithelium and epithelial hyperplasia, metaplasia, and dysplasia. Moreover, 4-μg thick sections were stained with a streptavidin-peroxidase kit (LSAB2 system peroxidase, Cat. No. K0675, Dako, Carpinteria, CA), using primary antibodies versus Hsp60 (Mouse Anti-Hsp60 monoclonal antibody, Sigma, St. Louis, MO, Cat. No. H4149, dilution 1:300) and Hsp10 (Rabbit Anti-Cpn10 Polyclonal Antibody, StressGen Biotechnologies, Victoria BC, Canada, Cat. No. SPA-110, dilution 1:300). Negative controls and human keratin (Mouse Anti-AE1/AE3 Keratin Monoclonal, Dako, Cat. No. M3515, dilution 1:50) as positive controls were run simultaneously.
Hsp60 and Hsp10 proteins were quantified in the epithelium of 47 disposable bronchial biopsies from 35 patients with COPD (in 12 patients, 2 biopsies each were analyzed) composed of basal and columnar cells. The final value was expressed as a percentage of total epithelial cells for each patient.
Samples for Western blotting were subjected to SDS-PAGE electrophoresis on a 12% polyacrylamide minigel (Bio-Rad, Hercules, CA). Total proteins extracted from paraffin-embedded sections were loaded at 30 μg/lane. After electrophoresis, proteins were blotted onto PVDF membrane (Amersham, Arlington Heights, IL) according to the manufacturer's instructions. Incubation of filters with the primary antibody was performed overnight at 4°C; primary antibodies used were mouse Anti HSP-60 (Sigma, H 4149) and rabbit anti Cpn-10 (StressGen, SPA-110). Antibodies were diluted following the manufacturer's instructions in antibody buffer (1% bovine serum albumin in 0.05% Tween-20 [TBS]). Antibody binding to blotted antigens was revealed by incubation with ECL detection reagents (Amersham) and exposure to an autoradiographic film (Kodak BioMax, Rochester, NY).
Group data are expressed as mean ± standard deviation (Table 1) for functional data or percentage for morphologic data (Table 2). Differences between groups obtained on the basis of the epithelial diagnosis were analyzed using the Student unpaired t-test for functional data. Morphologic data were analyzed by the Kruskall-Wallis test for multiple comparisons followed by a Mann-Whitney U-test for comparison between groups. Correlation coefficients were evaluated using the Spearman rank method. Probability values of P < .05 were considered significant. Data analysis was performed using the Stat View SE Graphics program (Abacus Concepts, Berkeley, CA).
|Epithelium||Age, y||FEV1% predicted||FEV1% Post-β||Pack/y||Hsp60 (%)||Hsp10 (%)|
|Dysplasia (n = 8)||66 ± 11||43 ± 18||46 ± 19||30 ± 13||1.9 ± 4*,†||1.1 ± 2‖,¶|
|Metaplasia (n = 5)||70 ± 8||58 ± 17||62 ± 14||59 ± 48||3.0 ± 3‡,§||3.6 ± 2#,**|
|Iperplasia (n = 12)||66 ± 8||47 ± 19||51 ± 20||53 ± 34||48 ± 29||52 ± 26|
|Normal (n = 10)||68 ± 5||53 ± 18||57 ± 18||69 ± 32||58 ± 23||54 ± 23|
The pathologic assessment of bronchial biopsies from all COPD patients showed that this study group was composed of 10 subjects with normal epithelium (NM), 12 with epithelial hyperplasia (BCH), 5 with metaplasia (SM), and 8 with dysplasia (SD). In all COPD patients, immunopositivity for Hsp60 and Hsp10 proteins was 34 ± 32% and 34 ± 31%, respectively (Table 1). Concerning Hsp60 and Hsp10 expression on the basis of epithelial diagnosis, 58 ± 23% and 54 ± 23% of epithelial cells were immunostained for Hsp60 and Hsp10, respectively, in the 10 subjects with normal epithelium. The 12 subjects with epithelial hyperplasia showed 48 ± 29% and 52 ± 26% of immunopositive cells for Hsp60 and Hsp10, respectively. In the presence of epithelial metaplasia, these percentages were reduced to 3.0 ± 3 and 3.6 ± 2, respectively; subjects with epithelial dysplasia showed the lowest percentages of immunopositivity: 1.9 ± 4 and 1.1 ± 2, respectively, for Hsp60 and Hsp10 (see Table 2) (Kruskall-Wallis test: P = .0007 and P < .0001 for Hsp60 and Hsp10, respectively; Mann-Whitney U-test for comparison between groups is reported in Table 2). Bronchial biopsies from 2 smokers with normal lung function and epithelial dysplasia were negative for Hsp60 and Hsp10 immunoexpression (data not shown).
Concerning the pattern of immunostaining, Hsp60 and Hsp10 immunopositivity in NM and in BCH was largely confined to ciliary and goblet elements. In addition, only 1 in 10 adenocarcinoma or ASC specimens were positive (<5%) for Hsp60, whereas 0 in 10 were immunostained for Hsp10. Interestingly, normal epithelium present at the edges of ASC specimens (NME) was negative for both Hsp60 and Hsp10 proteins.
Hsp-60 and Hsp-10 proteins were below the detection limits for Western blotting (data not shown). This result is probably explained by the lower sensitivity of Western blotting compared with immunohistochemistry when analyzing proteins from paraffin-embedded sections.
Structure-Structure and Structure-Function Correlations
In all COPD patients a positive significant correlation was observed between Hsp60 and Hsp10 epithelial immunoexpression (R = 0.80, P = .0001) (Fig. 3). No significant correlations were observed between FEV1 values or pack/year consumption and Hsp60 or Hsp10 protein expression.
This study shows, for the first time, that in specimens with bronchial ASC and in smokers with COPD and concomitant epithelial dysplasia, Hsp60 and Hsp10 proteins in the bronchial epithelium are down-regulated in comparison with smokers with COPD and with normal epithelium or epithelial hyperplasia. Moreover, in the present study we show through analysis of the epithelial changes in a group of well-characterized smokers suffering from moderate to very severe COPD that more than 20% (8 of 35 patients) had epithelial dysplasia in their bronchial biopsies. We found that Hsp60 and Hsp10 are highly down-regulated in epithelial dysplasia compared with normal epithelium and hyperplasia in our smokers with COPD. This down-regulation is also largely present in specimens with bronchial ASC, suggesting that progressive loss of these proteins may play a role in bronchial carcinogenesis. Interestingly, we found that normal mucosa at the edges of bronchial adenocarcinoma specimens (NME) was negative for both Hsp60 and Hsp10; as a consequence, we postulate that 1 of the steps in bronchial carcinogenesis may consist of molecular changes favorable to tumoral development that determine Hsps loss in airway epithelium. The specific molecular changes, however, that cause these modifications in Hsp60 and Hsp10 expression remain to be determined.
Much evidence has accumulated in the last few years stressing that mitochondrial Hsps may play a unique role in cell transformation and cancer development. These molecules function as chaperones in regulating cellular homeostasis, promoting cell survival, and influencing apoptosis.32, 33 A number of studies demonstrate that Hsp-induced cytoprotection is due in part to the suppression of apoptosis.34 Moreover, disruption of the delicate balance in the chaperone network may result in the association of insoluble proteins, in enzymatic activity being triggered at an inappropriate time, in incorrect localization within the tumoral cell, or any combination of these.32, 34
We also show that epithelial metaplasia is associated with a much lower but significant down-regulation of Hsp60 and Hsp10 proteins compared with normal epithelium and hyperplasia. This finding suggests that profound morphologic changes that take place during the development of epithelial metaplasia are associated with significant molecular modifications inducing down-regulation of Hsps, as we have documented here. It has been recognized that the sequence of precursor lesions for bronchial ASC is hyperplasia-metaplasia-dysplasia.35, 36 This proposed sequence is based on both circumstantial and genetic evidence. In fact, preneoplastic lesions have been found in association with malignancies or with a high risk to develop lung cancer, whereas genetic similarities have been observed between preneoplastic lesions and diagnosed tumors.36
We recently showed that Hsp60 may be down-regulated during bronchial carcinogenesis,28 but we did not provide an extensive analysis and interpretation of these data in our preliminary report. Hsp60 expression during cancer development and progression may depend on the tissues of origin of the neoplasm. Because it has been supposed that Hsp60 may have both pro- and antiapoptotic roles in tumoral cells,26, 27, 37 we postulated that its levels may be dependent on the expression of other proteins involved in the activation of the apoptotic pathway. A similar functional explanation can be proposed for Hsp10. We observed a highly significant positive correlation between Hsp60 and Hsp10 immunoexpression (see Fig. 3, R = 0.80) in our patients. This finding is also in line with the genetic pattern of Hsp60 and Hsp10 production. In fact, the biosynthesis of both these proteins comes under a common genetic control.38
We did not observe any correlation between severity of bronchial obstruction (FEV1 values) and degree of Hsp60 and Hsp10 immunoexpression, suggesting that the presence of COPD more than the degree of severity may influence the development of epithelial changes, such as dysplasia, observed during bronchial carcinogenesis. Alternatively, a relatively small range of FEV1 values or the limited sample size of moderate-to-very-severe COPD patients studied here may account for the lack of significant correlation between Hsps immunoexpression and degree of bronchial obstruction.
For comparative purposes, we also analyzed the epithelial expression of Hsp60 and Hsp10 in bronchial biopsies from 2 smokers with normal lung function and epithelial dysplasia, and did not find these proteins in association with epithelial dysplasia (data not shown). We speculate that epithelial dysplasia in the general population of smokers, with and without COPD, could be associated with loss of Hsp60 and Hsp10 protein immunoexpression. Whether epithelial dysplasia is, like LC, 4 to 5 times higher in smokers with COPD remains to be determined.
To the best of our knowledge, this is the first study demonstrating epithelial Hsp60 and Hsp10 down-regulation during lung carcinogenesis in vivo. The Hsp60/Hsp10 complex may protect cells from ischemia-induced cell death, acting not only as a molecule involved in protein stabilization and folding, but also as a protein modulator.39–43
We previously showed the overexpression of Hsp60 and Hsp10 in a variety of tumors and pretumoral lesions, such as large bowel cancer, exocervical cancer, and prostate cancer.14–17 Although in normal cells these chaperones are localized mostly in the mitochondrial matrix, we reported their accumulation within cancer cells. Their roles in the cytoplasm are not yet known, although they probably influence the process of programmed cell death.
On the basis of their action in protecting cells, avoiding protein denaturation, and inhibiting apoptosis, one could argue that Hsp60 and Hsp10 overexpression favors tumoral growth. Nevertheless, most prognostic and predictive studies have revealed that Hsp60 and Hsp10 overexpression is correlated with a better outcome,22–24 possibly due to the immunity antitumoral response that they are able to induce.44 Indeed, it was well demonstrated that some Hsps, among them Hsp60 and Hsp10, are potent immune modulators that can lead to activation of both innate and adaptive immune responses to tumor.45–47 As a consequence, we may postulate that Hsp60 and Hsp10 down-regulation during bronchial carcinogenesis could be related to a reduced antitumoral immune response toward neoplastic cells. Nevertheless, this hypothesis needs to be confirmed by other studies.
In conclusion, we postulate that the loss of Hsp60 and Hsp10 immunopositivity is related to bronchial cancer development and progression in smokers with COPD and, hence, that Hsp60 and Hsp10 could be useful new biomarkers for diagnostic and prognostic purposes in lung cancer management. Moreover, we believe that a better understanding of the molecular basis of Hsp60 and Hsp10 expression during bronchial carcinogenesis will also expand our therapeutic targets against this cancer.
- 1Histological typing of lung and pleural tumours. In: World Health Organization International Histological Classification of Tumours. Geneva: WHO, 1999., , in collaboration with pathologists from 14 countries.
- 7Global Initiative for Chronic Obstructive Lung Disease. Global strategy for the diagnosis, management and prevention of chronic obstructive pulmonary disease. NHLBI/WHO workshop report. Bethesda, MD: National Heart, Lung and Blood Institute, April 2001; Update of the management sections, GOLD website (www.goldcopd.com). Date updated: July 2003. Last update 2004.