Expression And Clinical Significance Of PML And Relative Regulators In Lung Cancer

BackgroundSince 70's in 20 centuries, cancer morbidity and mortality are increasing significantly, and now have become the leading cause of death in both town and countryside in China. The morbidity and mortality of lung cancer is particularly going up. Despite of many progressions in treating the deadly disease, which including multimodality and individual therapy, the fatality rate of lung cancer ranks as the No. 1 in malignancies of human beings. Early diagnosis and treatment is still seen as the fundamental strategy for curing lung cancer.At first, promyelocytic leukemia (PML) gene received attention due to its role in the pathogenesis of acute promyelocytic leukemia (APL), and was deemed for being aberrant in all cases with APL. PML displays important effects of regulating cell differentiation, inducing apoptosis and premature senescence in the studies with APL, liver cancer, breast cancer, bladder cancer, prostate cancer, et al. Following transfection of four liver cancer cell lines with replication-deficient recombinant PML adenovirus, the overexpressed PML initially induced a substantial G1 cell cycle arrest and triggered massive cell death in all tested cell lines, irrespective of their p53 status. The cell death effect of PML was higher than that induced by p53 over a longer period of time. The relation of PML to growth and development of lung cancer, however, has not been determined. By searching literature from databases of Pubmed in USA and of CNKI in China, we found only three studies on both lung cancer and PML, which being lack of enough samples (total 87 cases), clinical and prognostic analysis. We, therefore, constructed the tissue microarrays that containing normal lung tissues, benign lung tumours and lung cancers, then probed into expression and clinical significance of PML, P53, P16^INK4A, Bcl-2, Survivin and Cox-2, at last explored the transcriptional status of PML gene in lung cancer.Methods1,PatientsFrom January 1, 1998 to June 30, 2000, there are 148 cases with lung cancer received surgical resection in the Thoracic Deparment of First Hospital of China Medical University, and owned usable formalin-fixed and paraffin-embedded (FPE) tissue, and without preoperative chemotherapy or radiotherapy.2,Constructing and examining the tumor tissue microarraysThe tumor tissue microarrays (TTMs) were constructed with samples of lung carcinomas (n=148), pulmonary benign tumors (n=5), and pulmonary normal tissues (n=7). There were four cases of lung cancer discharged for lacking of enough tissue through erroneous sampling or immunohistochemical staining. The coincidence between the whole tissue sections and 1 to 4 array cores were validated by immunohistochemical staining with monoclonal antibody of COX-2.3,Immunohistochemical stainingDetecting the expressions of PML, P53, P16^INK4A, Bcl-2, Survivin and Cox-2 using on the tissue microarrays, and analysing the correlation with the clinical features and prognosis.4,Extracting RNA and RT-PCRSelecting 40 cases of lung cancer totally without the expression of PML protein; the RNA being extracted from these FPE lung cancer tissues (saved from 60 to 85 months, mean 69 months) by using Trizol one-step method, and being compared with that from fresh lung cancer tissues; then PML mRNA being checked up by utilizing RT-PCR.5,The statistical analysisThe statistical analysis was performed using SPSS 13.0 statistical software. A P-value of less than 0.05 was considered to indicate statistical significance. The crosstabs method was used to compare the statistical difference of positive ratios between groups. Fisher's exact test, Continuity Corrected Chi-square and Pearson Chi-square were used, according to the total cases and the minimum expected count, to assess the correlation between the presence of oncoprotein expression and clinicopathological parameters (gender, tumor status, node status, pTNM-stage, histology, et al.). Statistical correlation was assessed by using the Spearman correlation coefficient. The Kaplan-Meier product-limit estimator was used to estimate cancer specific survival and general survival. The log rank test and Breslow test were used to compare survival between groups. The Cox proportional hazards model was used for monovariate and multivariate analyses to estimate the simultaneous effects of prognostic factors on survival, and was stratified by pathologic TNM-stage (I , II, III).Results1,The quality of the tissue microarraysRatio of punch loss during sectioning and staining is less than 3% in the tissue microarrays. Rates of COX-2 expression are 55% (33/60) in whole sections, 53.3% (32/60) in both 4 and 3 cores of the tissue microarrays, 50% (30/60) in 2 cores, and 46.7% (28/60) in 1 core. The difference between rates of COX-2 expression is not statistically significant.2,The expression and clinical significance of these factorsPML expressions were observed in the nuclei of 40 cases with lung cancer, in the cytoplasm of 26 cases, and in the both of 12 cases. There were higher levels of PML nuclear staining in NSCLC than in SCLC (31.4% vs. 8.7%, P=0.026), and lower levels of PML cytoplasm staining (14% vs. 39.1%, P=0.01). PML expression was strong in nuclei of two pulmonary leiomyomas, however, often weak in lung cancer (46/54). The 5-year survival rate of SCLC was higher in PML positive patients than in PML negative ones (50% vs. 23%, mean: 54 months vs. 13 months). PML expression was the independent factor influencing the prognosis of SCLC by Cox univariate and multivariate analyses.P53 expression was found in 33.3% (48/144) lung cancer, and absent in benign tumors and normal tissues of lung (P=0.038). Expression rate of P53 was 33.1% (40/121) in NSCLC, and 34.8% (8/21) in SCLC. There is a significant difference between the expression rates of NSCLC, and benign tumors and normal tissues of lung (P=0.040), and with critical state between SCLC and benign tumors and normal tissues of lung (P=0.057). P16^INK4Aexpression rate was 28.5% (41/144) in lung cancer, and 8.3% (1/12) in benign tumors and normal tissues of lung (only a pulmonary leiomyoma with weak stain). P16^INK4A was strongly stained in SCLC (69.6%, 16/23), however, in NSCLC with a rate of 20.7% (25/121). Poor or non-differentiated tumors had a greater retention of P16^ink4a expression tnan well or moderate differentiated did (36.5% vs.20%, P=0.028).Survivin expression ratio was 90.3% (130/144) in lung cancer, and 58.3% (7/12) in benign tumors and normal tissues of lung (P=0.005); the positive ratio was 91.7% in NSCLC, and 82.6% in SCLC; younger patients had higher expression ratio than older patients did (95.3 % vs. 83.1%, P=0.015). Bcl-2 staining was strong in lung cancer with a rate of 52.8% (76/144); however, was weak in benign tumors and normal tissues of lung with a rate of 16.7% (2/12, P=0.016). There was stronger expression in squamous cell lung cancer than in adenocarcinoma (86.7 % vs. 27.4%, P < 0.001).The positive rate of COX-2 in lung cancer was 61.8% (89/144), however, 25% (3/12) in benign tumors and normal tissues of lung (P=0.029); and 66.9% in NSCLC, 34.8% in SCLC (P=0.004); a greater expression of COX-2 was found in female, non-smoker, and moderate or well differentiated cancers. COX-2 positive rate was higher in adenocarcinoma than in squamous carcinoma (79% vs. 48.9%, P=0.001).Of 144 cases of lung cancer, there was positive correlation between PML and Survivin expression (P<0.001, r=0.313), inverse correlation between PML expression on the nuclei and p16 expression, positive correlation between P53 and either P16 or Bcl-2, positive correlation between COX-2 and Survivin, and inverse correlation between COX-2 and Bcl-2. The positive correlation between PML and Survivin expression was seen in NSCLC, squamous cell cancer and adenocarcinoma. There was positive correlation between P53 and Bcl-2, also between COX-2 and Survivin, but inverse correlation between COX-2 and Bcl-2 in NSCLC. The positive correlation existed between P53 and either P16 or Bcl-2, also between COX-2 and Survivin in adenocarcinoma. In squamous carcinoma, there was inverse correlation between PML and either P53 or Bcl-2, but positive correlation between COX-2 and Survivin. In SCLC, there was no correlation among the factors.3,The results of extracting RNA and RT-PCRRNA derived from fresh tissue showed the distinct 18S and 28S ribosomal RNA bands, and RNA isolated from FPE displayed a broad molecular weight distribution ranging from 50bps to 450bps (mean 190bps) by electrophoresis; RNA from both tissues had the ratios of OD260/OD280 ranging from 1.9 to 2.1. RT-PCR results displayed that PML mRNA (295bps) was not absent in 40 cases, in which PML protein was absent completely. But the bars of PML from FPE were smaller and weaker than that ofβ-actin, also than that from fresh tissues.Conclusion1,The tissue microarray is a credible tool for molecular biology study, and owns the characteristics of high-throughout, saving time, and convenience for designing a cohort with a large sample.2,PML had some expression in the cytoplasm of lung cancer. In SCLC, PML was seldom found in nuclei, however, was upregulated in cytoplasm of some cases, which might be the characteristics of SCLC; meanwhile, PML was one of hazardous factors that influenced long-term survival of postoperative SCLC patients. As an important suppressor of SCLC, PML could be useful as a biomarker, or for the targeted therapy.3,In all lung cancers, PML expression was correlated with Survivin, Bcl-2, P16, and P53. For the first time, COX-2 and PML was studied together in lung cancer, despite of no relation being displayed.4,The RNA isolated from FPE tissue by Trizol one-step method had a high level of purity, but was suitable for amplifying gene with a size less than 400bp. Loss of PML protein in lung cancer was not caused by downward adjustment of mRNA.5,The development of a lung cancer involves a series of complex and interwoven mechanisms, relating to a chaos of several paths of cellular regulation. Lung cancer owns varieties of molecule expression paterns, which promoting identification of biological characteristics and targeted therapy to incline to the principle of combining integration and individualization.