Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Correlation of NQO1 and Nrf2 in Female Genital Tract Cancer and Their Precancerous Lesions (Cervix, Endometrium and Ovary)

Correlation of NQO1 and Nrf2 in Female Genital Tract Cancer and Their Precancerous Lesions... mon and cause significant morbidity and mortality worldwide, whereas vulvar, vaginal and fallopian tube cancers are very rare Background: NAD (P) H/quinone oxidoreductase 1 (NQO1) is a me- [1]. Cervical cancer is the third most common cancer in women tabolizing enzyme that detoxifies chemical stressors and antioxidants. worldwide and the seventh most common cancer overall. Its Nuclear factor erythroid 2-related factor 2 (NrF2) is an important overall mortality incidence ratio is 52% [2]. Endometrial can- transcriptional activator involved in the cellular defense mechanisms cer is the sixth most common cancer in women worldwide. Its against oxidative stress. incidence and mortality rates are higher in more developed re- Methods: The immunohistochemical expression of NQO1 and Nrf2 gions and lowest rates occurring in Asia and Africa. Overall, in 80 cervical, 80 endometrial and 100 ovarian specimens with differ- the mortality incidence ratio of endometrial cancer is 26% [3]. ent lesions was studied. Then we study the relation of both NQO1 and Cancers of the ovary constitutes the eighth most common can- Nrf2 expression and clinicopathological features of carcinoma cases. cers among women worldwide with mortality incidence ratio of 62% [2]. In Middle Egypt with regional registry in Minia, the Results: Immunohistochemical stain showed that NQO1 and Nrf2 incidence of cervical, endometrial and ovarian cancer is 1.06%, were highly expressed in carcinoma compared with normal and pre- 0.67% and 3.75% respectively of cancer sites in females [4]. cancerous lesions. Significant positive correlations were found be- Understanding the mechanisms of carcinogenesis in female re- tween the mean expression of NQO1 and Nrf2 in different lesions. productive organs could contribute to early detection, and will Moreover, there was significant correlation between the high level be helpful in the prevention and treatment of these cancers. of NQO1 and Nrf2 expression and high tumor grade in cervical and NAD (P) H/quinone oxidoreductase-1 (NQO1), also endometrial carcinoma cases. Nrf2 expression was significant with known as DT-diaphorase, is a cytosolic enzyme that uses advanced stage in endometrial and ovarian carcinomas. NADH or NADPH as substrates to catalyze the two-electron reduction of quinones and related compounds, and it is encod- Conclusions: NQO1 and Nrf2 might be new biomarkers for early di- ed by a gene located on chromosome 16q22 [5]. In normal agnosis and prognostic evaluation as well as being targets for therapy cells, NQO1 protects cells against oxidative stress, as well as in patients with tumors in female genital tract. against carcinogenesis by stabilization of the p53 tumor sup- pressor [6]. However, studies on NQO1 expression in cancer Keywords: Cervical carcinoma; Endometrial carcinoma; Ovarian have been contradictory. On the one hand, NQO1 is induced carcinoma; NQO1; Nrf2 along with a battery of defensive genes that provide protection against different stresses to prevent organs from undergoing carcinogen-induced tumorigenesis. On the other hand, reduc- Introduction tive activation of environmental carcinogens including hetero- cyclic amines by NQO1 could contribute to carcinogenesis. Also, the disruption of the NQO1 gene or genetic polymor- Cancers of the female reproductive system include cervical, phism increased the risk of chemical-induced toxicity and can- cers [7]. Comparing normal and malignant tissue, NQO1 was reported to be up-regulated in malignant tissue of the pancreas, Manuscript accepted for publication June 26, 2015 interlobular biliary epithelial cells, breast and lung [8-11] and Department of Pathology, Faculty of Medicine, Minia University, Egypt down-regulated in tumors of the kidney and esophagus [12, Department of Obstetrics and Gynecology, Faculty of Medicine, Minia Uni- 13]. In addition, the high level of NQO1 expression in vari- versity, Maternity Hospital, Egypt ous tumors in combination with its ability to reduce many qui- Corresponding Author: Nisreen Abdel Tawab Abdel Gaber Osman, Depart- nine-containing antitumor drugs has drawn attention to NQO1 ment of Pathology, Faculty of Medicine, Minia University, 61111, Egypt. as a potential molecular target in cancer treatment [14]. The Email: [email protected] molecular mechanism of NQO1 responsible for tumors of the female reproductive system progression remains unclear, and doi: http://dx.doi.org/10.14740/wjon931w Articles © The authors | Journal compilation © World J Oncol and Elmer Press Inc™ | www.wjon.org This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited Osman et al World J Oncol. 2015;6(3):364-374 additional studies are needed to understand its role in female perplasia samples were obtained either by curettage or biopsy tumorigenesis. specimens. Hyperplasia specimens were evaluated according to The transcription factor, nuclear factor erythroid 2-related WHO classification [22]. All EC patients had undergone total factor 2 (Nrf2) is a nuclear transcription factor maintaining abdominal hysterectomy and bilateral salpingo-oophrectomy. intracellular redox homeostasis that induces transcription Ovarian specimens included 10 cases of normal ovarian of a variety of genes through binding to the antioxidant re- tissues, 20 cases of benign ovarian tumors (12 cases serous sponse element in target gene promoters [15, 16]. It is induced and eight cases mucinous), 20 cases of borderline ovarian tu- in response to various agents at the transcriptional level. In mors (12 cases serous and eight cases mucinous) and 50 cases addition, more than 200 gene products including an antioxi- of ovarian carcinoma (35 cases serous and 15 cases mucinous dative enzyme NQO1 are under the transcriptional control carcinoma). Normal ovarian specimens from hysterectomy for Nrf2 [17]. Therefore, activation of Nrf2 confers protec- specimens resected for non-ovarian disease were used. The majority of patients with a diagnosis of primary ovarian cancer tion against cancer [18]. Furthermore, the beneficial effects had undergone radical surgery (staging laparotomy) according of many chemopreventive compounds rely on the activation to standard operating procedures with the primary objective of of the Nrf2-mediated antioxidant response through inhibition maximal tumor reduction. of Nrf2 degradation [19]. Mutations and deregulation of Nrf2 expression levels have been identified in many cancers [18] and lead to chemoresistance to many chemotherapeutic drugs Immunohistochemistry [20, 21]. To date, the association and correlation between NQO1 Paraffin-embedded sections on coated slides were used for and Nrf2 expression in cancers of the female reproductive staining. Sections were cut at 4 μm thick. Immunohistochem- system have not been adequately studied. In this study, we istry was performed using the DAKO LSAB kit (DAKO A/S, aimed to analyze the expression and relationship between both Glostrup, Denmark) as follows: slides were deparaffinized in markers in carcinomas of the cervix, endometrium, ovary and xylene and rehydrated in a graded alcohol series. Antigen re- their precancerous lesions. Also, we studied their relation with trieval was achieved by microwaving in sodium citrate buffer at clinicopathological parameters in carcinoma cases for better pH 6 for 10 min at 95 °C. Endogenous peroxidase was blocked understanding their role in tumorigenesis. with 0.3% hydrogen peroxide in methanol for 30 min. The slides were incubated with the mouse monoclonal primary antibody Materials and Methods for NQO1 (A180, ab28947, Abcam, 1:200), mouse monoclonal primary antibody for human Nrf2 (IgG2a, ab89443, Abcam, 1:100) overnight in a humidity chamber at 4 °C overnight. Then Tissue specimens samples were washed with rinse puffer (PBS) and biotinylated secondary antibody for 30 min at room temperature. Streptavi- Formalin-fixed and paraffin-embedded specimens were col- din was applied for 30 min at room temperature. Visualization lected and prepared for this study from Minia University Hos- was performed using 3,3’-diaminobenzidine (DAB) chromo- pital in collaboration with the cancer unit in the Obstetrics and gen, and Mayer’s hematoxylin was used for counterstaining for Gynecology Department, Minia University between January 10 min. A section of breast carcinoma and lung carcinoma were 2008 and December 2014. The histology of all cases using used as a positive control for NQO1 and Nrf2 proteins respec- hematoxylin-eosin (H&E) stained slides was reviewed. The tively. A negative control was carried out by replacing primary histological grade was assessed according to the World Health antibodies with rise buffer on a section. Organization (WHO) classification standards [22]. Tumors were staged according to the pathologic tumor-node-metas- Evaluation of immunohistochemistry staining tasis (TNM) and FIGO classification according to the Union for International Cancer Control (UICC) criteria seventh Edi- tion and WHO classification [23]. The clinicopathological data Five random fields of vision in each section were selected and were obtained from the pathology reports of cases. The avail- analyzed. The positive area, which was shown in percentage able data include patients’ age, tumor grade and stage. (ratio of positive area to the whole visual field) was calculated. Cervical specimens, include 10 non-neoplastic cervical tissues, 20 squamous intraepithelial lesion (SIL) (eight cases Statistical analysis of low-grade squamous intraepithelial lesion (LSIL) and 12 cases of high-grade squamous intraepithelial lesion (HSIL)) and 50 squamous cell carcinomas (SCCs). All cervical tissue Data were analyzed using the statistical package for the So- specimens were selected from punch biopsies, loop electrosur- cial Sciences version 17.0 (SPSS 17.0). Raw data were used gical excisions, cone biopsies and hysterectomies. to determine means, standard deviations (SDs) and ranges. Endometrial specimens included 10 cyclic endometrium Kruskal-Wallis test was used to compare markers expression (CE) (six cases were proliferative phase (PP) and four cases in different groups followed by Mann-Whitney test which was were secretory phase (SP)), 20 cases endometrial hyperplasia used to compare expression between two groups. Pearson cor- (EH) (eight cases without atypia, 12 cases atypical EH) and relation was used to determine whether there was a positive or 50 were endometrial carcinoma (EC) specimens. CE and hy- negative correlation between each examined marker and each Articles © The authors | Journal compilation © World J Oncol and Elmer Press Inc™ | www.wjon.org 365 NQO1 and Nrf2 in Female Genital Tract Cancer World J Oncol. 2015;6(3):364-374 Table 1. Positive Expression Rates, Mean Values and SDs for NQO1 in Different Lesions With Examined Sites Site Lesion No. % +ve Mean ± SD Min Max P-value among groups Cervix* Normal 10 20 3.70 ± 5.01 0 15 < 0.001 SIL 20 45 18.35 ± 24.94 0 80 Carcinoma 50 76 44.34 ± 29.52 0 90 Endometrium** Cyclic endometrium 10 10 2.40 ± 3.30 0 10 0.001 Hyperplasia 20 30 10.15 ± 15.61 0 50 Carcinoma 50 60 36.92 ± 31.94 0 85 Ovary*** Normal 10 0 0 0 0 < 0.001 Benign 20 20 6.30 ± 9.88 0 30 Borderline 20 25 12.40 ± 21.08 0 60 Carcinoma 50 56 28.30 ± 27.29 0 80 Test of significance: Kruskal-Wallis, Mann-Whitney test. *Normal vs. SIL, P = 0.091; normal vs. carcinoma, P < 0.001; SIL vs. carcinoma, P = 0.004. **Cyclic endometrium vs. hyperplasia, P = 0.402; cyclic endometrium vs. carcinoma, P = 0.007; hyperplasia vs. carcinoma, P = 0.003. ***Normal vs. benign, P = 0.005; normal vs. borderline, P = 0.003; normal vs. carcinoma, P < 0.001; benign vs. borderline, P = 0.554; benign vs. carcinoma, P = 0.006; borderline vs. carcinoma, P = 0.076. histopathological entity. Kruskal-Wallis test was used to exam- was more pronounced in EC as shown in Figure 2. ine the correlation of NQO1, Nrf2 staining scores in relation to tumor grade and stage. ANOVA test was used to examine the NQO1 expression in distinct tissue types correlation of NQO1, Nrf2 staining scores in relation to age. Statistical significance was set at P ≤ of 0.05. NQO1 expression in cervix Results On studying the mean NQO1 expression in different lesions, we found increased expression in SIL (45%, mean ± SD: 18.35 Positive expression rates, mean values and SDs for NQO1 and ± 24.94) and cervical carcinoma (76%, mean ± SD: 44.34 ± Nrf2 in different lesions for organs examined are listed in Ta- 29.52) cases compared to normal tissue (20%, mean ± SD: 3.70 bles 1, 2. NQO1 expression was detected in the cytoplasm of ± 5.01) and the difference between all examined groups was examined tissue as shown in Figure 1 while Nrf2 was nuclear statistically significant (P < 0.001) as shown in Table 1, Figure and cytoplasmic marker in cervical and ovarian tissue while in 3a. the endometrium its expression was mainly cytoplasmic with As regard SIL, we found in LSIL, NQO1, mean ± SD: little nuclear expression in CE and EH, and nuclear expression 21.37 ± 3.11, while in HLSIL, NQO1: 16.33 ± 2.11, that was Table 2. Positive Expression Rates, Mean Values and SDs for Nrf2 in Different Lesions With Examined Organs Organ Lesion No. % +ve Mean ± SD Min Max P-value among groups Cervix* Normal 10 10 4.60 ± 5.42 0 15 0.001 SIL 20 30 14.50 ± 18.20 0 60 Carcinoma 50 62 39.24 ± 29.25 0 85 Endometrium** Cyclic endometrium 10 10 4.30 ± 6.01 0 20 0.001 Hyperplasia 20 35 12.20 ± 16.91 0 60 Carcinoma 50 64 39.82 ± 32.65 0 85 Ovary*** Normal 10 0 0 0 0 < 0.001 Benign 20 25 10.30 ± 14.50 0 50 Borderline 20 20 14.60 ± 19.33 0 60 Carcinoma 50 72 32.70 ± 29.44 0 85 Test of significance: Kruskal-Wallis, Mann-Whitney test. *Normal vs. SIL, P = 0.155; normal vs. carcinoma, P = 0.002; SIL vs. carcinoma, P = 0.005. **Cyclic endometrium vs. hyperplasia, P = 0.350; cyclic endometrium vs. carcinoma, P = 0.009; hyperplasia vs. carcinoma, P < 0.005. ***Normal vs. benign, P = 0.015; normal vs. borderline, P = 0.003; normal vs. carcinoma, P < 0.001; benign vs. borderline, P = 0.529; benign vs. carcinoma, P = 0.005; borderline vs. carcinoma, P = 0.029. Articles © The authors | Journal compilation © World J Oncol and Elmer Press Inc™ | www.wjon.org 366 Osman et al World J Oncol. 2015;6(3):364-374 Figure 1. NQO1 protein expression in different lesions using IHC (× 400, DAB was used as the chromogen and hematoxylin as counterstain). (a) NQO1 protein was weak in normal cervical epithelium. (b) NQO1 protein staining was moderate positive in LSIL. (c) NQO1 protein showed diffuse and strong cytoplasmic-positive staining in HSIL. (d) NQO1 was strong positive in grade I cervical SCC. (e) NQO1 protein was weak in proliferative endometrium. (f) NQO1 was moderate positive in simple EH. (g) NQO1 was strong positive in atypical EH. (h) NQO1 was strong positive in EC. (i) NQO1 was negative in normal ovarian epithelium. (j) NQO1 was positive mucinous cystadenoma. (k) NQO1 was moderate positive in borderline serous tumor. (l) NQO1 was strong positive in ovarian carcinoma. lower than in LSIL, and this difference did not reach a signifi- 30%, 60% with mean ± SD: 10.15 ± 15.61 and 36.92 ± 31.94 cant level (P = 0.885). respectively, while decrease in CE (10%, mean ± SD: 2.40 Regarding its mean expression in different lesions, statis- ± 3.30), and the difference was statistically significant (P = tically significant differences were seen between normal and 0.001) (Table 1, Fig.3b). carcinoma (P = 0.001) and between SIL and carcinoma (P = For EH, we found in typical EH and atypical EH the mean 0.004). In addition, no significant difference was noticed be- ± SD was 3.00 ± 3.33, 14.91 ± 18.75 respectively, and there tween NQO1 mean expression in normal and SIL (P = 0.091). were no statistically significant differences between both (P = 0.118). Regarding its mean expression in different lesions, sta- NQO1 expression in endometrium tistically significant differences between EC and EH, CE (P = 0.003, P = 0.007 respectively) were noted. There were no statis- We found increased mean expression rate in EH and EC was tically significant differences between EH and CE (P = 0.402). Articles © The authors | Journal compilation © World J Oncol and Elmer Press Inc™ | www.wjon.org 367 NQO1 and Nrf2 in Female Genital Tract Cancer World J Oncol. 2015;6(3):364-374 Figure 2. Nrf2 protein expression in different lesions using IHC (× 400, DAB was used as the chromogen and hematoxylin as counterstain). (a) Nrf2 protein was weak in normal cervical epithelium. (b) Nrf2 protein staining was moderate positive in LSIL. (c) Nrf2 protein showed moderate nuclear-positive staining in HSIL, but weak positive in adjacent normal cervical glands. (d) Nrf2 was strong positive in grade II cervical SCC. (e) Nrf2 protein was positive in proliferative endometrium. (f) Nrf2 was moderate positive in simple EH. (g) Nrf2 was moderate positive in atypical EH. (h) Nrf2 was strong positive in EC. (i) Nrf2 was negative in normal ovarian epithelium. (j) Nrf2 was positive serous cystadenoma. (k) Nrf2 was moderate positive in borderline serous tumor. (l) Nrf2 was strong positive in ovarian carcinoma. NQO1 expression in ovary = 0.006). No statistically significant difference was seen be- tween benign and borderline (P = 0.554), between borderline and carcinoma (P = 0.076). On studying the NQO1 mean expression in different lesions, we found increased expression in carcinoma cases (56%, mean ± SD: 28.30 ± 27.29) than other lesions (20%, mean ± SD: Associations between NQO1 expression and clinicopatho- 6.30 ± 9.88 in benign; and 25%, mean ± SD: 12.40 ± 21.08 in logical data in carcinoma cases borderline tumors), and the difference was statistically signifi- cant (P < 0.001) as shown in Table 1, Figure 3c. Associations between clinicopathological data and NQO1 Regarding NQO1 mean expression in different lesions, mean expression were summarized in Table 3. statistically significant differences were seen between nor - mal and each benign, borderline and malignant tumors (P = As regarding cervical carcinoma, a significant association 0.005, P = 0.003 and P < 0.001 respectively). Statistically sig- between increased NQO1 mean expression and tumor grade nificant difference was seen between benign and carcinoma (P (P = 0.045) was detected. No significant associations were no- Articles © The authors | Journal compilation © World J Oncol and Elmer Press Inc™ | www.wjon.org 368 Osman et al World J Oncol. 2015;6(3):364-374 Figure 3. NQO1 expression box plots (a) in different cervical lesions; (b) in different endometrial lesions; (c) in different ovarian lesions. Horizontal lines in the boxes represent the median value of each group. The top and bottom edges of the boxes indicate the score values from 75th percentile and the 25th percentile respectively. Whiskers represent the highest and lowest values. The range is shown as a vertical line. ticed between its expression and either age or stage. 0.603). In EC, a significant association was noticed between in- As regard Nrf2 mean expression in different lesions, sta- creased NQO1 mean expression and tumor grade (P = 0.011). tistically significant differences between CE and EH, EC (P No significant associations were noticed between its expres- = 0.009, P = 0.005 respectively). There were no statistically sion and either age or stage. significant differences between CE and EH (P = 0.350). Regards ovarian carcinoma, no significant associations were noticed between NQO1 mean expression and any clin- Nrf2 expression in ovary icopathological data. Nrf2 expression in different ovarian lesions, we noticed that Nrf2 expression in distinct tissue types increased mean expression in carcinoma (72%, mean ± SD: 32.70 ± 29.44) than other examined lesions (0% for normal; Nrf2 expression in cervix 25%, mean ± SD: 10.30 ± 14.50 for benign tumors; and 20%, mean ± SD: 14.60 ± 19.33 for borderline tumors), and the dif- ference reached a significant statistical level (P = 0.004) (Table As regard Nrf2 expression in different lesions, we found in- 2, Fig.4c). creased mean expression from normal tissue (10%, mean ± Regarding its mean expression in different lesions, statis- SD: 4.60 ± 5.42) to SIL (30%, mean ± SD: 14.50 ± 18.20) and tically significant differences were identified between normal to cervical carcinoma (62%, mean ± SD: 39.24 ± 29.25). The and carcinoma (P < 0.001). Statistically significant differences difference between all examined groups was statistically sig- were identified between benign and carcinoma (P = 0.005), nificant (P = 0.001) as shown in Table 2, Figure 4a. and between borderline and carcinoma (P = 0.029). Similarly, Regarding Nrf2 expression in SIL, we found that its ex- statistically significant difference was seen between normal pression in LSIL mean ± SD: 11.12 ± 14.21 and in HSIL 16.75 and benign and borderline (P = 0.015, P = 0.003 respectively). ± 20.37, and there was no statistically significant difference In addition, no statistically significant difference was seen be- between both (P = 0.451). tween benign and borderline (P = 0.529). Regarding its mean expression in different lesions, statis- tically significant differences between normal and carcinoma (P = 0.002) and between SIL and carcinoma (P = 0.005) were Associations between Nrf2 expression and clinicopatho- found. No significant difference was noticed between Nrf2 logical data in carcinoma cases mean expression in normal and SIL (P = 0.155). Associations between clinicopathological data and Nrf2 mean Nrf2 expression in endometrium expression were summarized in Table 3. In cervical carcinoma, we found that a significant associa- Nrf2 expression rate was increased in EH (35%, mean ± SD: tion between increased Nrf2 mean expression and high tumor 12.20 ± 16.91) and EC (64%, mean ± SD: 39.82 ± 32.65) than grade (P = 0.016) was detected. No significant associations in CE (10%), and the difference was statistically significant (P were noticed between its mean expression and age or stage. = 0.001) as shown in Table 2, Figure 4b. For Nrf2 overexpression in EC, we found that a significant For EH, we found, in typical EH and atypical EH the mean association between increased its mean expression and tumor ± SD was 10.25 ± 12.72, 13.50 ± 19.66 respectively, and there grade and stage (P = 0.002, P = 0.025 respectively) was detected. were no statistically significant differences between both (P = Regarding ovarian carcinoma, a significant association Articles © The authors | Journal compilation © World J Oncol and Elmer Press Inc™ | www.wjon.org 369 NQO1 and Nrf2 in Female Genital Tract Cancer World J Oncol. 2015;6(3):364-374 Table 3. Associations Between NQO1 and Nrf2 Expression Scores and Clinicopathological Data in Carcinoma Cases NQO1 Nrf2 Organ Clinicopathological parameter No. of cases Mean ± SD P-value Mean ± SD P-value Cervix Age Mean 55.20 ± 6.54 44.34 ± 29.52 0.445 39.24 ± 29.25 0.526 Grade I 14 25.14 ± 33.26 0.045 25.14 ± 33.03 0.016 II 20 26.80 ± 29.65 28.00 ± 29.11 III 16 33.50 ± 32.61 23.00 ± 23.59 Stage I 8 40.62 ± 37.74 0.461 36.88 ± 32.28 0.594 II 28 48.32 ± 29.80 35.11 ± 29.01 III 10 42.40 ± 25.57 48.50 ± 28.67 IV 4 28.75 ± 20.15 49.75 ± 29.89 Endometrium Age Mean 51.90 ± 5.77 36.92 ± 31.94 0.802 39.82 ± 55.00 0.923 Grade I 16 17.25 ± 27.14 0.011 18.44 ± 28.61 0.002 II 21 45.38 ± 31.39 48.48 ± 31.35 III 23 47.46 ± 28.99 52.15 ± 28.06 Stage I 10 12.60 ± 23.98 0.082 13.40 ± 24.73 0.025 II 13 40.69 ± 32.61 41.54 ± 31.31 III 19 42.00 ± 31.28 49.11 ± 33.18 IV 8 49.13 ± 30.82 48.00 ± 29.21 Ovary Age Mean 55.74 ± 9.92 28.30 ± 27.29 0.734 32.70 ± 29.44 0.395 Grade I 10 22.70 ± 30.31 0.712 30.80 ± 37.00 0.576 II 23 26.87 ± 26.72 29.26 ± 30.14 III 17 33.53 ± 27.02 18.47 ± 24.10 Stage I 8 12.50 ± 21.66 0.167 16.88 ± 28.77 0.020 II 15 22.53 ± 30.07 19.80 ± 31.06 III 15 33.47 ± 25.14 45.27 ± 25.06 IV 12 39.58 ± 25.71 43.67 ± 23.50 Test of significance: Kruskal-Wallis, ANOVA tests. P-value < 0.05 is considered significant. was noticed between increased Nrf2 mean expression and ad- positive correlation was noted between NQO1 and Nrf2 mean vanced tumor stage (P = 0.020). No significant associations expression in all examined cases (r = 0.734, P < 0.001). A sig- were noticed between its expression and either age or grade. nificant positive correlation was noted between NQO1/Nrf2 (r = 0.818, P < 0.001) in SIL. Similarly, a significant positive correlation was noted (r = 0.615, P < 0.001) in carcinoma. No Correlations between immunohistochemical markers significant correlations were noted between them in normal expression cervical tissue (r = 0.615, P = 0.051). Correlations between immunohistochemical markers ex- pression in different endometrial lesions were found. A signifi- Correlations between immunohistochemical markers expres- sion in different cervical lesions were found. A significant cant positive correlation was noted between NQO1 and Nrf2 Articles © The authors | Journal compilation © World J Oncol and Elmer Press Inc™ | www.wjon.org 370 Osman et al World J Oncol. 2015;6(3):364-374 Figure 4. NrF2 expression box plots (a) in different cervical lesions; (b) in different endometrial lesions; (c) in different ovarian lesions. Horizontal lines in the boxes represent the median value of each group. The top and bottom edges of the boxes indicate the score values from 75th percentile and the 25th percentile respectively. Whiskers represent the highest and lowest values. The range is shown as a vertical line. mean expression in all examined cases (r = 0.922, P < 0.001). rate of NQO1 protein was slightly higher in well-differentiated A significant positive correlation was noted (r = 0.909, P < SCC (43.75%) than in CIN3 (40.74%) [26]. In addition, pre- 0.001) in CE. Similarly, a significant positive correlation was vious results noted more NQO1 overexpression in carcinoma noted (r = 0.637, P = 0.003) in EH. Also a significant correla - than in normal or precancerous lesions in breast [10], colon tion was in EC (r = 0. 925, P < 0.001). [27] and liver [28], and the difference reached a significant Correlations between immunohistochemical markers ex- level. Our results with previous results indicate that NQO1 up- pression in different ovarian lesions were found. A significant regulation may be an early event in cancer progression. These negative correlation was noted between NQO1 and Nrf2 (r = findings suggest that NQO1 protein level might be used as an 0.741, P < 0.001) in all cases. Similarly, a significant positive early diagnostic indicator of this disease. correlation was noted in benign tumors (r = 0.623, P = 0.003) To further illustrate that NQO1 may be an effective predic- and in carcinoma (r = 0.740, P < 0.001). No significant correla - tor of poor prognosis, the correlation between NQO1 expres- tions were noted in borderline tumors (r = 0.6432, P = 0.057). sion and clinicopathological features of cervical, endometrial and ovarian carcinomas was analyzed. We found that high-lev- el expression of the NQO1 protein was significantly correlated Discussion with poor differentiation in cervical and EC and not associated with advanced stage. NQO1 overexpression was reported to be associated with high tumor grade in carcinoma of cervix and NQO1 flavoprotein has been found to be expressed in many breast [10, 26], with advanced stage cervical [26], breast [10], body tissues [9-11]. It is conceivable that NQO1 is primarily colon [27] and liver [28] carcinomas and with nodal metasta- involved in protecting normal cells from oxidant stress. Such ses [10]. NQO1 overexpression induced tumor cell prolifera- finding has led to the suggestion that NQO1 can be impor - tion via the up-regulation of cyclins [29] and was accompanied tant in cancer chemoprevention. However, polymorphism in by an increase in other antioxidant enzymes, such as HMOX-1 the NQO1 gene has been reported to be associated with an and GST, providing tumors with increased protection against increased risk of various cancers such as breast [10], lung [11], cytotoxic agents allowing for rapid cancer progression [30]. gastric [24] and head and neck cancer [25]. These results indicated that NQO1 played a predictive role in In the present study, we found that staining of NQO1 is tumor progression and might be useful as a poor prognostic mainly localized in the cytoplasm and these observations were biomarker of cancer. in agreement with previous studies [8-10, 26, 27]. NQO1 has NQO1 overexpression in tumors but not normal tissue positive cytoplasmic expression in 76%, 60% and 65% in cer- has made it an attractive target for treatment of lung cancer. It vical, endometrial and ovarian carcinoma respectively. Previ- is the main activator of quinone-containing alkylating agents ous studies demonstrated that NQO1 immunopositivity rate such as mitomycins [31]. So that, patients with KRAS muta- ranged between 21% and 80% [10, 11, 28] in various tumors. tions may utilize quinone-containing alkylating agents more We found increased NQO1 expression from normal tissue to efficiently due to increased NQO1 expression [11]. On the SIL and cervical carcinoma and then from SIL to carcinoma. contrary, loss of NQO1 expression appeared to be candidates The difference between all examined groups was statistically significant. This finding was also observed in the endometrial for adjuvant chemotherapy in patients with cholangiocarcino- and ovarian lesions. In a previous study, a strong positive rate ma [32]. Therefore, the role of NQO1 and related inhibitors in of NQO1 protein expression in both SCCs (54.80%) and cervi- chemotherapy appears questionable. A comprehensive similar cal intraepithelial neoplasia (CINs) (27.59% in CIN-1, 34.21% analysis of the relationship between NQO1 enzyme activity in CIN-2 and 40.74% in CIN-3) was significantly higher than and chemosensitivity in female tract cancer is essential. in the normal cervix (4%). Interestingly, the strongly positive In this study, we found that Nrf2 is nuclear, and cytoplas- Articles © The authors | Journal compilation © World J Oncol and Elmer Press Inc™ | www.wjon.org 371 NQO1 and Nrf2 in Female Genital Tract Cancer World J Oncol. 2015;6(3):364-374 mic marker by immunohistochemistry in cervical and ovarian tracellular antioxidant response and is negatively regulated by tissue, while in the endometrium its expression was mainly cy- Keap1, may be partially responsible to the aggressive biologi- toplasmic with little nuclear expression and these observations cal behavior and poor clinical outcome due to its known effect were in agreement with previous studies [33-35]. Other reports of increased resistance to chemotherapeutic drugs as cisplatin detect Nrf2 expression mainly in the nucleus [27, 36, 37]. It in both endometrial and ovarian cancer cells [18, 20, 21]. These is well established that oxidative stress is the primary signal findings may also provide an opportunity for therapeutic inter - that causes cytoplasmic Nrf2 to accumulate within the nucleus vention against chemoresistance via applications of either Nrf2 [34]. It has been documented that persistent nuclear expression inhibitors or gene knockdown approaches [33]. Therefore, a of Nrf2 results in the production of antioxidants that protect new chemotherapeutic protocol that includes antioxidant ther- cancer cells from reactive oxygen species. Higher concentra- apy may be a useful method for solving chemoresistance [37]. tion of Nrf2 in the nucleus may reflect upstaging of cancer, ag- Regarding correlation between NQO1 and Nrf2 mean ex- gressive tumor behavior and poor clinical outcome [26, 33, 37]. pression in different lesions, we found positive correlations Half of ovarian carcinomas with positive nuclear Nrf2 staining between the two proteins especially in carcinomas. Similar had either Keap1 mutations or absent Keap1 mRNA expression correlations were reported [27]. This Nrf2-NQO1/MRP1 sig- resulting in platinum resistance [38]. So, nuclear Nrf2 expres- nal pathway may be attributed to the stress response and self- sion in cancer cells would have a higher malignant potential. protective effort of the cells during malignant transformation. Therefore, it is essential to evaluate nuclear expression of Nrf2. Considering the role of Nrf2 in regulating genes as NQO1 and We found Nrf2 expression was 62%, 64% and 72% in MRP1, which act to detoxify drugs or attenuate drug-induced cervical, endometrial and ovarian carcinoma respectively. Pre- oxidative stress, it is possible that highly expressed nuclear vious studies demonstrated that Nrf2 immunoreactivity was Nrf2 plays a role in increasing treatment resistance and results frequently detected in various human malignancies, such as in short survival [27]. To the best of our knowledge, this was intrahepatic chorangiocellular [32], endometrial [33], breast the first study examining NQO1 and Nrf2 in cervical, endome - [36], gastric [35, 37], ovarian [38], lung [39], pancreatic [40] trial and ovarian tissue and the change of their expression in and gallbladder [41] carcinomas, and its rate of immunoposi- different lesions in each tissue type. tivity ranged between 26% and 76% in these studies. We found increased expression from normal tissue to SIL Conclusions and cervical carcinoma and then from SIL to carcinoma. The difference between all examined groups was statistically sig- nificant. In the endometrium, we found Nrf2 expression was NQO1 and Nrf2 play a key role in the progression of female increased from CE to EH to EC, and the difference was statisti- tract tumors, and high level of both proteins were strongly as- cally significant. Similar results were reported in the endome - sociated with high grade and advanced stage. The high pro- trium with lower expression of Nrf2 in atypical EH and higher portion of NQO1 expression suggests that NQO1 may be a in endometrial cancer [33, 42-44]. Finally in ovarian tissue we significant biomarker and a potential therapeutic target for found the same results as Nrf2 expression was increased from carcinoma patients. Nrf2 expression in cancer may be useful normal to benign tumors and from benign to borderline tumors for evaluation of biological malignant potential. Overall, our and from borderline tumors to carcinomas with a significant present work implies that NQO1 and Nrf2 might be new bio- difference between the examined groups. This elevated Nrf2 markers for early diagnosis and tumorigenesis in patients with expression may be induced by gonadotropins and sex-steroid tumors of female genital tract. hormones, which suggest that these hormones are involved in ovarian cancer development via modulation of Nrf2 signaling. Therefore, its inhibition may represent an effective therapeutic Competing Interests strategy for treatment [45]. Previous studies in different organs reported that Nrf2 ex- The authors declare that they have no competing interests. pression was more overexpressed in carcinoma than normal and precancerous lesions in pancreatic [34, 40], gastric [35] and breast [36] carcinomas. Nrf2 expression may represent Author Contributions one of the early molecular events in the neoplastic transforma- tion of several tumors. Regarding association of Nrf2 overexpression and clin- N. A. Osman performed project development, data collection, icopathological data, we found that its expression was signifi- immunostaining, data analysis and manuscript writing. N. M. cantly associated with high tumor grade in cervical and EC and Abd El-Maqsoud performed project development, data collec- with advanced tumor stage in endometrial and ovarian carci- tion, immunostaining, data analysis and manuscript writing. S. noma and no significant association with the age. Our findings A. El Gelany performed data collection and manuscript writ- are in accordance with those reported by [27, 35-37, 41]. No ing. All authors read and approved the final manuscript. differences were noted in age, grade and stage in ovarian car- cinoma [38, 45]. Overexpression of Nrf2 in gallbladder adeno- References carcinoma was correlated with tumor differentiation, staging, metastasis and shorter overall survival [41]. 1. Weiderpass E, Labreche F. Malignant tumors of the female Furthermore, overexpression of Nrf2, a regulator of an in- Articles © The authors | Journal compilation © World J Oncol and Elmer Press Inc™ | www.wjon.org 372 Osman et al World J Oncol. 2015;6(3):364-374 reproductive system. Saf Health Work. 2012;3(3):166- 17. Kwak MK, Wakabayashi N, Itoh K, Motohashi H, 180. Yamamoto M, Kensler TW. Modulation of gene expression 2. Ferlay J, Shin HR, Bray F, Forman D, Mathers C, Parkin by cancer chemopreventive dithiolethiones through the DM. Cancer incidence and mortality worldwide: IARC Keap1-Nrf2 pathway. Identification of novel gene clusters CancerBase No. 10. Lyon (France): International Agency for cell survival. J Biol Chem. 2003;278(10):8135-8145. for Research on Cancer; 2010. 18. Jiang T, Huang Z, Lin Y, Zhang Z, Fang D, Zhang DD. 3. Merritt MA, Cramer DW. Molecular pathogenesis of en- The protective role of Nrf2 in streptozotocin-induced dia- dometrial and ovarian cancer. Cancer Biomark. 2010;9(1- betic nephropathy. Diabetes. 2010;59(4):850-860. 6):287-305. 19. Kobayashi A, Kang MI, Okawa H, Ohtsuji M, Zenke Y, 4. Ibrahim AS, Khaled HM, Mikhail NN, Baraka H, Kamel Chiba T, Igarashi K, et al. Oxidative stress sensor Keap1 H. Cancer incidence in egypt: results of the national pop- functions as an adaptor for Cul3-based E3 ligase to regu- ulation-based cancer registry program. J Cancer Epide- late proteasomal degradation of Nrf2. Mol Cell Biol. miol. 2014;2014:437971. 2004;24(16):7130-7139. 5. Lajin B, Alachkar A. The NQO1 polymorphism C609T 20. Cho JM, Manandhar S, Lee HR, Park HM, Kwak MK. (Pro187Ser) and cancer susceptibility: a comprehensive Role of the Nrf2-antioxidant system in cytotoxicity me- meta-analysis. Br J Cancer. 2013;109(5):1325-1337. diated by anticancer cisplatin: implication to cancer cell 6. Hu X, Zhang Z, Ma D, Huettner PC, Massad LS, Nguyen resistance. Cancer Lett. 2008;260(1-2):96-108. L, Borecki I, et al. TP53, MDM2, NQO1, and suscepti- 21. Wang XJ, Sun Z, Villeneuve NF, Zhang S, Zhao F, Li Y, bility to cervical cancer. Cancer Epidemiol Biomarkers Chen W, et al. Nrf2 enhances resistance of cancer cells to Prev. 2010;19(3):755-761. chemotherapeutic drugs, the dark side of Nrf2. Carcino- 7. Su XL, Yan MR, Yang L. NQO1 C609T polymor- genesis. 2008;29(6):1235-1243. phism correlated to colon cancer risk in farmers from 22. Clarke BA, Gilks B. Ovarian Carcinoma: Recent Devel- western region of Inner Mongolia. Chin J Cancer Res. opments in Classification of Tumour Histological Sub- 2012;24(4):317-322. type. Canadian Journal of Pathology. 2011;33-42. 8. Awadallah NS, Dehn D, Shah RJ, Russell Nash S, Chen 23. Edge SB, Byrd DR, Compton CC, Fritz AG, Greene FL, YK, Ross D, Bentz JS, et al. NQO1 expression in pan- Trotti A (eds). Ovary and primary peritoneal carcinoma. creatic cancer and its potential use as a biomarker. Appl In: AJCC Cancer Staging Manual. 7th ed. New York, NY: Immunohistochem Mol Morphol. 2008;16(1):24-31. Springer: 419-428; 2010. 9. Buranrat B, Chau-in S, Prawan A, Puapairoj A, Zeekpud- 24. Lin L, Qin Y, Jin T, Liu S, Zhang S, Shen X, Lin Z. sa P, Kukongviriyapan V. NQO1 expression correlates Significance of NQO1 overexpression for prognostic with cholangiocarcinoma prognosis. Asian Pac J Cancer evaluation of gastric adenocarcinoma. Exp Mol Pathol. Prev. 2012;13 (Suppl):131-136. 2014;96(2):200-205. 10. Yang Y, Zhang Y, Wu Q, Cui X, Lin Z, Liu S, Chen L. 25. Soucek P, Susova S, Mohelnikova-Duchonova B, Gro- Clinical implications of high NQO1 expression in breast madzinska J, Moraviec-Sztandera A, Vodicka P, Vod- cancers. J Exp Clin Cancer Res. 2014;33:14. ickova L. Polymorphisms in metabolizing enzymes and 11. Yilmaz A, Mohamed N, Patterson KA, Tang Y, Shilo K, the risk of head and neck squamous cell carcinoma in Villalona-Calero MA, Davis ME, et al. Increased NQO1 the Slavic population of the central Europe. Neoplasma. but not c-MET and survivin expression in non-small cell 2010;57(5):415-421. lung carcinoma with KRAS mutations. Int J Environ Res 26. Ma Y, Kong J, Yan G, Ren X, Jin D, Jin T, Lin L, et al. Public Health. 2014;11(9):9491-9502. NQO1 overexpression is associated with poor prognosis 12. Zappa F, Ward T, Pedrinis E, Butler J, McGown A. in squamous cell carcinoma of the uterine cervix. BMC NAD(P)H: quinone oxidoreductase 1 expression in kid- Cancer. 2014;14:414. ney podocytes. J Histochem Cytochem. 2003;51(3):297- 27. Ji L, Wei Y, Jiang T, Wang S. Correlation of Nrf2, NQO1, 302. MRP1, cmyc and p53 in colorectal cancer and their rela- 13. Marjani HA, Biramijamal F, Rakhshani N, Hossein- tionships to clinicopathologic features and survival. Int J Nezhad A, Malekzadeh R. Investigation of NQO1 genetic Clin Exp Pathol. 2014;7(3):1124-1131. polymorphism, NQO1 gene expression and PAH-DNA 28. Chiu MM, Ko YJ, Tsou AP, Chau GY, Chau YP. Analy- adducts in ESCC. A case-control study from Iran. Genet sis of NQO1 polymorphisms and p53 protein expression Mol Res. 2010;9(1):239-249. in patients with hepatocellular carcinoma. Histol Histo- 14. Bey EA, Reinicke KE, Srougi MC, Varnes M, Ander- pathol. 2009;24(10):1223-1232. son VE, Pink JJ, Li LS, et al. Catalase abrogates beta- 29. Garate M, Wani AA, Li G. The NAD(P)H:Quinone lapachone-induced PARP1 hyperactivation-directed pro- Oxidoreductase 1 induces cell cycle progression and grammed necrosis in NQO1-positive breast cancers. Mol proliferation of melanoma cells. Free Radic Biol Med. Cancer Ther. 2013;12(10):2110-2120. 2010;48(12):1601-1609. 15. Motohashi H, Yamamoto M. Nrf2-Keap1 defines a physi- 30. Lau A, Villeneuve NF, Sun Z, Wong PK, Zhang DD. ologically important stress response mechanism. Trends Dual roles of Nrf2 in cancer. Pharmacol Res. 2008;58(5- Mol Med. 2004;10(11):549-557. 6):262-270. 16. Zhang DD. Mechanistic studies of the Nrf2-Keap1 sign- 31. Beall HD, Winski SI. Mechanisms of action of quinone- aling pathway. Drug Metab Rev. 2006;38(4):769-789. containing alkylating agents. I: NQO1-directed drug de- Articles © The authors | Journal compilation © World J Oncol and Elmer Press Inc™ | www.wjon.org 373 NQO1 and Nrf2 in Female Genital Tract Cancer World J Oncol. 2015;6(3):364-374 velopment. Front Biosci. 2000;5:D639-648. lial ovarian cancer. Cancer Res. 2011;71(15):5081-5089. 32. Wakai T, Shirai Y, Sakata J, Matsuda Y, Korita PV, Taka- 39. Solis LM, Behrens C, Dong W, Suraokar M, Ozburn mura M, Ajioka Y, et al. Prognostic significance of NQO1 NC, Moran CA, Corvalan AH, et al. Nrf2 and Keap1 ab- expression in intrahepatic cholangiocarcinoma. Int J Clin normalities in non-small cell lung carcinoma and asso- Exp Pathol. 2011;4(4):363-370. ciation with clinicopathologic features. Clin Cancer Res. 33. Chen N, Yi X, Abushahin N, Pang S, Zhang D, Kong 2010;16(14):3743-3753. B, Zheng W. Nrf2 expression in endometrial serous 40. Hong YB, Kang HJ, Kwon SY, Kim HJ, Kwon KY, Cho carcinomas and its precancers. Int J Clin Exp Pathol. CH, Lee JM, et al. Nuclear factor (erythroid-derived 2)- 2010;4(1):85-96. like 2 regulates drug resistance in pancreatic cancer cells. 34. Lister A, Nedjadi T, Kitteringham NR, Campbell F, Cos- Pancreas. 2010;39(4):463-472. tello E, Lloyd B, Copple IM, et al. Nrf2 is overexpressed 41. Wang J, Zhang M, Zhang L, et al. Correlation of Nrf2, in pancreatic cancer: implications for cell proliferation HO-1, and MRP3 in gallbladder cancer and their relation- and therapy. Mol Cancer. 2011;10:37. ships to clinicopathologic features and survival. J Surg 35. Hu XF, Yao J, Gao SG, Wang XS, Peng XQ, Yang YT, Res. 2010;164: 99-105. 42. Liang SX, Chambers SK, Cheng L, Zhang S, Zhou Y, Feng XS. Nrf2 overexpression predicts prognosis and Zheng W. Endometrial glandular dysplasia: a putative pre- 5-FU resistance in gastric cancer. Asian Pac J Cancer Prev. 2013;14(9):5231-5235. cursor lesion of uterine papillary serous carcinoma. Part 36. Onodera Y, Motohashi H, Takagi K, Miki Y, Shibahara Y, II: molecular features. Int J Surg Pathol. 2004;12(4):319- Watanabe M, Ishida T, et al. NRF2 immunolocalization in 331. human breast cancer patients as a prognostic factor. En- 43. Fadare O, Liang SX, Ulukus EC, Chambers SK, Zheng docr Relat Cancer. 2014;21(2):241-252. W. Precursors of endometrial clear cell carcinoma. Am J 37. Kawasaki Y, Ishigami S, Arigami T, Uenosono Y, Yanagi- Surg Pathol. 2006;30(12):1519-1530. ta S, Uchikado Y, Kita Y, et al. Clinicopathological sig- 44. Yi X, Zheng W. Endometrial glandular dysplasia and en- nificance of nuclear factor (erythroid-2)-related fac- dometrial intraepithelial neoplasia. Curr Opin Obstet Gy- tor 2 (Nrf2) expression in gastric cancer. BMC Cancer. necol. 2008;20(1):20-25. 2015:155. 45. Liao H, Zhou Q, Zhang Z, Wang Q, Sun Y, Yi X, Feng Y. 38. Konstantinopoulos PA, Spentzos D, Fountzilas E, Fran- NRF2 is overexpressed in ovarian epithelial carcinoma coeur N, Sanisetty S, Grammatikos AP, Hecht JL, et al. and is regulated by gonadotrophin and sex-steroid hor- Keap1 mutations and Nrf2 pathway activation in epithe- mones. Oncol Rep. 2012;27(6):1918-1924. Articles © The authors | Journal compilation © World J Oncol and Elmer Press Inc™ | www.wjon.org http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png World Journal of Oncology Pubmed Central

Correlation of NQO1 and Nrf2 in Female Genital Tract Cancer and Their Precancerous Lesions (Cervix, Endometrium and Ovary)

Osman, Nisreen Abdel Tawab Abdel Gaber; Abd El-Maqsoud, Nehad M. R.; El Gelany, Saad Abdelnaby A.
World Journal of Oncology , Volume 6 (3) – Jun 12, 2015
Download PDF
11 pages

Loading next page...
 
/lp/pubmed-central/correlation-of-nqo1-and-nrf2-in-female-genital-tract-cancer-and-their-l4Fm09R81p

References (53)

Publisher
Pubmed Central
Copyright
Copyright 2015, Osman et al.
ISSN
1920-4531
eISSN
1920-454X
DOI
10.14740/wjon931w
Publisher site
See Article on Publisher Site

Abstract

mon and cause significant morbidity and mortality worldwide, whereas vulvar, vaginal and fallopian tube cancers are very rare Background: NAD (P) H/quinone oxidoreductase 1 (NQO1) is a me- [1]. Cervical cancer is the third most common cancer in women tabolizing enzyme that detoxifies chemical stressors and antioxidants. worldwide and the seventh most common cancer overall. Its Nuclear factor erythroid 2-related factor 2 (NrF2) is an important overall mortality incidence ratio is 52% [2]. Endometrial can- transcriptional activator involved in the cellular defense mechanisms cer is the sixth most common cancer in women worldwide. Its against oxidative stress. incidence and mortality rates are higher in more developed re- Methods: The immunohistochemical expression of NQO1 and Nrf2 gions and lowest rates occurring in Asia and Africa. Overall, in 80 cervical, 80 endometrial and 100 ovarian specimens with differ- the mortality incidence ratio of endometrial cancer is 26% [3]. ent lesions was studied. Then we study the relation of both NQO1 and Cancers of the ovary constitutes the eighth most common can- Nrf2 expression and clinicopathological features of carcinoma cases. cers among women worldwide with mortality incidence ratio of 62% [2]. In Middle Egypt with regional registry in Minia, the Results: Immunohistochemical stain showed that NQO1 and Nrf2 incidence of cervical, endometrial and ovarian cancer is 1.06%, were highly expressed in carcinoma compared with normal and pre- 0.67% and 3.75% respectively of cancer sites in females [4]. cancerous lesions. Significant positive correlations were found be- Understanding the mechanisms of carcinogenesis in female re- tween the mean expression of NQO1 and Nrf2 in different lesions. productive organs could contribute to early detection, and will Moreover, there was significant correlation between the high level be helpful in the prevention and treatment of these cancers. of NQO1 and Nrf2 expression and high tumor grade in cervical and NAD (P) H/quinone oxidoreductase-1 (NQO1), also endometrial carcinoma cases. Nrf2 expression was significant with known as DT-diaphorase, is a cytosolic enzyme that uses advanced stage in endometrial and ovarian carcinomas. NADH or NADPH as substrates to catalyze the two-electron reduction of quinones and related compounds, and it is encod- Conclusions: NQO1 and Nrf2 might be new biomarkers for early di- ed by a gene located on chromosome 16q22 [5]. In normal agnosis and prognostic evaluation as well as being targets for therapy cells, NQO1 protects cells against oxidative stress, as well as in patients with tumors in female genital tract. against carcinogenesis by stabilization of the p53 tumor sup- pressor [6]. However, studies on NQO1 expression in cancer Keywords: Cervical carcinoma; Endometrial carcinoma; Ovarian have been contradictory. On the one hand, NQO1 is induced carcinoma; NQO1; Nrf2 along with a battery of defensive genes that provide protection against different stresses to prevent organs from undergoing carcinogen-induced tumorigenesis. On the other hand, reduc- Introduction tive activation of environmental carcinogens including hetero- cyclic amines by NQO1 could contribute to carcinogenesis. Also, the disruption of the NQO1 gene or genetic polymor- Cancers of the female reproductive system include cervical, phism increased the risk of chemical-induced toxicity and can- cers [7]. Comparing normal and malignant tissue, NQO1 was reported to be up-regulated in malignant tissue of the pancreas, Manuscript accepted for publication June 26, 2015 interlobular biliary epithelial cells, breast and lung [8-11] and Department of Pathology, Faculty of Medicine, Minia University, Egypt down-regulated in tumors of the kidney and esophagus [12, Department of Obstetrics and Gynecology, Faculty of Medicine, Minia Uni- 13]. In addition, the high level of NQO1 expression in vari- versity, Maternity Hospital, Egypt ous tumors in combination with its ability to reduce many qui- Corresponding Author: Nisreen Abdel Tawab Abdel Gaber Osman, Depart- nine-containing antitumor drugs has drawn attention to NQO1 ment of Pathology, Faculty of Medicine, Minia University, 61111, Egypt. as a potential molecular target in cancer treatment [14]. The Email: [email protected] molecular mechanism of NQO1 responsible for tumors of the female reproductive system progression remains unclear, and doi: http://dx.doi.org/10.14740/wjon931w Articles © The authors | Journal compilation © World J Oncol and Elmer Press Inc™ | www.wjon.org This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited Osman et al World J Oncol. 2015;6(3):364-374 additional studies are needed to understand its role in female perplasia samples were obtained either by curettage or biopsy tumorigenesis. specimens. Hyperplasia specimens were evaluated according to The transcription factor, nuclear factor erythroid 2-related WHO classification [22]. All EC patients had undergone total factor 2 (Nrf2) is a nuclear transcription factor maintaining abdominal hysterectomy and bilateral salpingo-oophrectomy. intracellular redox homeostasis that induces transcription Ovarian specimens included 10 cases of normal ovarian of a variety of genes through binding to the antioxidant re- tissues, 20 cases of benign ovarian tumors (12 cases serous sponse element in target gene promoters [15, 16]. It is induced and eight cases mucinous), 20 cases of borderline ovarian tu- in response to various agents at the transcriptional level. In mors (12 cases serous and eight cases mucinous) and 50 cases addition, more than 200 gene products including an antioxi- of ovarian carcinoma (35 cases serous and 15 cases mucinous dative enzyme NQO1 are under the transcriptional control carcinoma). Normal ovarian specimens from hysterectomy for Nrf2 [17]. Therefore, activation of Nrf2 confers protec- specimens resected for non-ovarian disease were used. The majority of patients with a diagnosis of primary ovarian cancer tion against cancer [18]. Furthermore, the beneficial effects had undergone radical surgery (staging laparotomy) according of many chemopreventive compounds rely on the activation to standard operating procedures with the primary objective of of the Nrf2-mediated antioxidant response through inhibition maximal tumor reduction. of Nrf2 degradation [19]. Mutations and deregulation of Nrf2 expression levels have been identified in many cancers [18] and lead to chemoresistance to many chemotherapeutic drugs Immunohistochemistry [20, 21]. To date, the association and correlation between NQO1 Paraffin-embedded sections on coated slides were used for and Nrf2 expression in cancers of the female reproductive staining. Sections were cut at 4 μm thick. Immunohistochem- system have not been adequately studied. In this study, we istry was performed using the DAKO LSAB kit (DAKO A/S, aimed to analyze the expression and relationship between both Glostrup, Denmark) as follows: slides were deparaffinized in markers in carcinomas of the cervix, endometrium, ovary and xylene and rehydrated in a graded alcohol series. Antigen re- their precancerous lesions. Also, we studied their relation with trieval was achieved by microwaving in sodium citrate buffer at clinicopathological parameters in carcinoma cases for better pH 6 for 10 min at 95 °C. Endogenous peroxidase was blocked understanding their role in tumorigenesis. with 0.3% hydrogen peroxide in methanol for 30 min. The slides were incubated with the mouse monoclonal primary antibody Materials and Methods for NQO1 (A180, ab28947, Abcam, 1:200), mouse monoclonal primary antibody for human Nrf2 (IgG2a, ab89443, Abcam, 1:100) overnight in a humidity chamber at 4 °C overnight. Then Tissue specimens samples were washed with rinse puffer (PBS) and biotinylated secondary antibody for 30 min at room temperature. Streptavi- Formalin-fixed and paraffin-embedded specimens were col- din was applied for 30 min at room temperature. Visualization lected and prepared for this study from Minia University Hos- was performed using 3,3’-diaminobenzidine (DAB) chromo- pital in collaboration with the cancer unit in the Obstetrics and gen, and Mayer’s hematoxylin was used for counterstaining for Gynecology Department, Minia University between January 10 min. A section of breast carcinoma and lung carcinoma were 2008 and December 2014. The histology of all cases using used as a positive control for NQO1 and Nrf2 proteins respec- hematoxylin-eosin (H&E) stained slides was reviewed. The tively. A negative control was carried out by replacing primary histological grade was assessed according to the World Health antibodies with rise buffer on a section. Organization (WHO) classification standards [22]. Tumors were staged according to the pathologic tumor-node-metas- Evaluation of immunohistochemistry staining tasis (TNM) and FIGO classification according to the Union for International Cancer Control (UICC) criteria seventh Edi- tion and WHO classification [23]. The clinicopathological data Five random fields of vision in each section were selected and were obtained from the pathology reports of cases. The avail- analyzed. The positive area, which was shown in percentage able data include patients’ age, tumor grade and stage. (ratio of positive area to the whole visual field) was calculated. Cervical specimens, include 10 non-neoplastic cervical tissues, 20 squamous intraepithelial lesion (SIL) (eight cases Statistical analysis of low-grade squamous intraepithelial lesion (LSIL) and 12 cases of high-grade squamous intraepithelial lesion (HSIL)) and 50 squamous cell carcinomas (SCCs). All cervical tissue Data were analyzed using the statistical package for the So- specimens were selected from punch biopsies, loop electrosur- cial Sciences version 17.0 (SPSS 17.0). Raw data were used gical excisions, cone biopsies and hysterectomies. to determine means, standard deviations (SDs) and ranges. Endometrial specimens included 10 cyclic endometrium Kruskal-Wallis test was used to compare markers expression (CE) (six cases were proliferative phase (PP) and four cases in different groups followed by Mann-Whitney test which was were secretory phase (SP)), 20 cases endometrial hyperplasia used to compare expression between two groups. Pearson cor- (EH) (eight cases without atypia, 12 cases atypical EH) and relation was used to determine whether there was a positive or 50 were endometrial carcinoma (EC) specimens. CE and hy- negative correlation between each examined marker and each Articles © The authors | Journal compilation © World J Oncol and Elmer Press Inc™ | www.wjon.org 365 NQO1 and Nrf2 in Female Genital Tract Cancer World J Oncol. 2015;6(3):364-374 Table 1. Positive Expression Rates, Mean Values and SDs for NQO1 in Different Lesions With Examined Sites Site Lesion No. % +ve Mean ± SD Min Max P-value among groups Cervix* Normal 10 20 3.70 ± 5.01 0 15 < 0.001 SIL 20 45 18.35 ± 24.94 0 80 Carcinoma 50 76 44.34 ± 29.52 0 90 Endometrium** Cyclic endometrium 10 10 2.40 ± 3.30 0 10 0.001 Hyperplasia 20 30 10.15 ± 15.61 0 50 Carcinoma 50 60 36.92 ± 31.94 0 85 Ovary*** Normal 10 0 0 0 0 < 0.001 Benign 20 20 6.30 ± 9.88 0 30 Borderline 20 25 12.40 ± 21.08 0 60 Carcinoma 50 56 28.30 ± 27.29 0 80 Test of significance: Kruskal-Wallis, Mann-Whitney test. *Normal vs. SIL, P = 0.091; normal vs. carcinoma, P < 0.001; SIL vs. carcinoma, P = 0.004. **Cyclic endometrium vs. hyperplasia, P = 0.402; cyclic endometrium vs. carcinoma, P = 0.007; hyperplasia vs. carcinoma, P = 0.003. ***Normal vs. benign, P = 0.005; normal vs. borderline, P = 0.003; normal vs. carcinoma, P < 0.001; benign vs. borderline, P = 0.554; benign vs. carcinoma, P = 0.006; borderline vs. carcinoma, P = 0.076. histopathological entity. Kruskal-Wallis test was used to exam- was more pronounced in EC as shown in Figure 2. ine the correlation of NQO1, Nrf2 staining scores in relation to tumor grade and stage. ANOVA test was used to examine the NQO1 expression in distinct tissue types correlation of NQO1, Nrf2 staining scores in relation to age. Statistical significance was set at P ≤ of 0.05. NQO1 expression in cervix Results On studying the mean NQO1 expression in different lesions, we found increased expression in SIL (45%, mean ± SD: 18.35 Positive expression rates, mean values and SDs for NQO1 and ± 24.94) and cervical carcinoma (76%, mean ± SD: 44.34 ± Nrf2 in different lesions for organs examined are listed in Ta- 29.52) cases compared to normal tissue (20%, mean ± SD: 3.70 bles 1, 2. NQO1 expression was detected in the cytoplasm of ± 5.01) and the difference between all examined groups was examined tissue as shown in Figure 1 while Nrf2 was nuclear statistically significant (P < 0.001) as shown in Table 1, Figure and cytoplasmic marker in cervical and ovarian tissue while in 3a. the endometrium its expression was mainly cytoplasmic with As regard SIL, we found in LSIL, NQO1, mean ± SD: little nuclear expression in CE and EH, and nuclear expression 21.37 ± 3.11, while in HLSIL, NQO1: 16.33 ± 2.11, that was Table 2. Positive Expression Rates, Mean Values and SDs for Nrf2 in Different Lesions With Examined Organs Organ Lesion No. % +ve Mean ± SD Min Max P-value among groups Cervix* Normal 10 10 4.60 ± 5.42 0 15 0.001 SIL 20 30 14.50 ± 18.20 0 60 Carcinoma 50 62 39.24 ± 29.25 0 85 Endometrium** Cyclic endometrium 10 10 4.30 ± 6.01 0 20 0.001 Hyperplasia 20 35 12.20 ± 16.91 0 60 Carcinoma 50 64 39.82 ± 32.65 0 85 Ovary*** Normal 10 0 0 0 0 < 0.001 Benign 20 25 10.30 ± 14.50 0 50 Borderline 20 20 14.60 ± 19.33 0 60 Carcinoma 50 72 32.70 ± 29.44 0 85 Test of significance: Kruskal-Wallis, Mann-Whitney test. *Normal vs. SIL, P = 0.155; normal vs. carcinoma, P = 0.002; SIL vs. carcinoma, P = 0.005. **Cyclic endometrium vs. hyperplasia, P = 0.350; cyclic endometrium vs. carcinoma, P = 0.009; hyperplasia vs. carcinoma, P < 0.005. ***Normal vs. benign, P = 0.015; normal vs. borderline, P = 0.003; normal vs. carcinoma, P < 0.001; benign vs. borderline, P = 0.529; benign vs. carcinoma, P = 0.005; borderline vs. carcinoma, P = 0.029. Articles © The authors | Journal compilation © World J Oncol and Elmer Press Inc™ | www.wjon.org 366 Osman et al World J Oncol. 2015;6(3):364-374 Figure 1. NQO1 protein expression in different lesions using IHC (× 400, DAB was used as the chromogen and hematoxylin as counterstain). (a) NQO1 protein was weak in normal cervical epithelium. (b) NQO1 protein staining was moderate positive in LSIL. (c) NQO1 protein showed diffuse and strong cytoplasmic-positive staining in HSIL. (d) NQO1 was strong positive in grade I cervical SCC. (e) NQO1 protein was weak in proliferative endometrium. (f) NQO1 was moderate positive in simple EH. (g) NQO1 was strong positive in atypical EH. (h) NQO1 was strong positive in EC. (i) NQO1 was negative in normal ovarian epithelium. (j) NQO1 was positive mucinous cystadenoma. (k) NQO1 was moderate positive in borderline serous tumor. (l) NQO1 was strong positive in ovarian carcinoma. lower than in LSIL, and this difference did not reach a signifi- 30%, 60% with mean ± SD: 10.15 ± 15.61 and 36.92 ± 31.94 cant level (P = 0.885). respectively, while decrease in CE (10%, mean ± SD: 2.40 Regarding its mean expression in different lesions, statis- ± 3.30), and the difference was statistically significant (P = tically significant differences were seen between normal and 0.001) (Table 1, Fig.3b). carcinoma (P = 0.001) and between SIL and carcinoma (P = For EH, we found in typical EH and atypical EH the mean 0.004). In addition, no significant difference was noticed be- ± SD was 3.00 ± 3.33, 14.91 ± 18.75 respectively, and there tween NQO1 mean expression in normal and SIL (P = 0.091). were no statistically significant differences between both (P = 0.118). Regarding its mean expression in different lesions, sta- NQO1 expression in endometrium tistically significant differences between EC and EH, CE (P = 0.003, P = 0.007 respectively) were noted. There were no statis- We found increased mean expression rate in EH and EC was tically significant differences between EH and CE (P = 0.402). Articles © The authors | Journal compilation © World J Oncol and Elmer Press Inc™ | www.wjon.org 367 NQO1 and Nrf2 in Female Genital Tract Cancer World J Oncol. 2015;6(3):364-374 Figure 2. Nrf2 protein expression in different lesions using IHC (× 400, DAB was used as the chromogen and hematoxylin as counterstain). (a) Nrf2 protein was weak in normal cervical epithelium. (b) Nrf2 protein staining was moderate positive in LSIL. (c) Nrf2 protein showed moderate nuclear-positive staining in HSIL, but weak positive in adjacent normal cervical glands. (d) Nrf2 was strong positive in grade II cervical SCC. (e) Nrf2 protein was positive in proliferative endometrium. (f) Nrf2 was moderate positive in simple EH. (g) Nrf2 was moderate positive in atypical EH. (h) Nrf2 was strong positive in EC. (i) Nrf2 was negative in normal ovarian epithelium. (j) Nrf2 was positive serous cystadenoma. (k) Nrf2 was moderate positive in borderline serous tumor. (l) Nrf2 was strong positive in ovarian carcinoma. NQO1 expression in ovary = 0.006). No statistically significant difference was seen be- tween benign and borderline (P = 0.554), between borderline and carcinoma (P = 0.076). On studying the NQO1 mean expression in different lesions, we found increased expression in carcinoma cases (56%, mean ± SD: 28.30 ± 27.29) than other lesions (20%, mean ± SD: Associations between NQO1 expression and clinicopatho- 6.30 ± 9.88 in benign; and 25%, mean ± SD: 12.40 ± 21.08 in logical data in carcinoma cases borderline tumors), and the difference was statistically signifi- cant (P < 0.001) as shown in Table 1, Figure 3c. Associations between clinicopathological data and NQO1 Regarding NQO1 mean expression in different lesions, mean expression were summarized in Table 3. statistically significant differences were seen between nor - mal and each benign, borderline and malignant tumors (P = As regarding cervical carcinoma, a significant association 0.005, P = 0.003 and P < 0.001 respectively). Statistically sig- between increased NQO1 mean expression and tumor grade nificant difference was seen between benign and carcinoma (P (P = 0.045) was detected. No significant associations were no- Articles © The authors | Journal compilation © World J Oncol and Elmer Press Inc™ | www.wjon.org 368 Osman et al World J Oncol. 2015;6(3):364-374 Figure 3. NQO1 expression box plots (a) in different cervical lesions; (b) in different endometrial lesions; (c) in different ovarian lesions. Horizontal lines in the boxes represent the median value of each group. The top and bottom edges of the boxes indicate the score values from 75th percentile and the 25th percentile respectively. Whiskers represent the highest and lowest values. The range is shown as a vertical line. ticed between its expression and either age or stage. 0.603). In EC, a significant association was noticed between in- As regard Nrf2 mean expression in different lesions, sta- creased NQO1 mean expression and tumor grade (P = 0.011). tistically significant differences between CE and EH, EC (P No significant associations were noticed between its expres- = 0.009, P = 0.005 respectively). There were no statistically sion and either age or stage. significant differences between CE and EH (P = 0.350). Regards ovarian carcinoma, no significant associations were noticed between NQO1 mean expression and any clin- Nrf2 expression in ovary icopathological data. Nrf2 expression in different ovarian lesions, we noticed that Nrf2 expression in distinct tissue types increased mean expression in carcinoma (72%, mean ± SD: 32.70 ± 29.44) than other examined lesions (0% for normal; Nrf2 expression in cervix 25%, mean ± SD: 10.30 ± 14.50 for benign tumors; and 20%, mean ± SD: 14.60 ± 19.33 for borderline tumors), and the dif- ference reached a significant statistical level (P = 0.004) (Table As regard Nrf2 expression in different lesions, we found in- 2, Fig.4c). creased mean expression from normal tissue (10%, mean ± Regarding its mean expression in different lesions, statis- SD: 4.60 ± 5.42) to SIL (30%, mean ± SD: 14.50 ± 18.20) and tically significant differences were identified between normal to cervical carcinoma (62%, mean ± SD: 39.24 ± 29.25). The and carcinoma (P < 0.001). Statistically significant differences difference between all examined groups was statistically sig- were identified between benign and carcinoma (P = 0.005), nificant (P = 0.001) as shown in Table 2, Figure 4a. and between borderline and carcinoma (P = 0.029). Similarly, Regarding Nrf2 expression in SIL, we found that its ex- statistically significant difference was seen between normal pression in LSIL mean ± SD: 11.12 ± 14.21 and in HSIL 16.75 and benign and borderline (P = 0.015, P = 0.003 respectively). ± 20.37, and there was no statistically significant difference In addition, no statistically significant difference was seen be- between both (P = 0.451). tween benign and borderline (P = 0.529). Regarding its mean expression in different lesions, statis- tically significant differences between normal and carcinoma (P = 0.002) and between SIL and carcinoma (P = 0.005) were Associations between Nrf2 expression and clinicopatho- found. No significant difference was noticed between Nrf2 logical data in carcinoma cases mean expression in normal and SIL (P = 0.155). Associations between clinicopathological data and Nrf2 mean Nrf2 expression in endometrium expression were summarized in Table 3. In cervical carcinoma, we found that a significant associa- Nrf2 expression rate was increased in EH (35%, mean ± SD: tion between increased Nrf2 mean expression and high tumor 12.20 ± 16.91) and EC (64%, mean ± SD: 39.82 ± 32.65) than grade (P = 0.016) was detected. No significant associations in CE (10%), and the difference was statistically significant (P were noticed between its mean expression and age or stage. = 0.001) as shown in Table 2, Figure 4b. For Nrf2 overexpression in EC, we found that a significant For EH, we found, in typical EH and atypical EH the mean association between increased its mean expression and tumor ± SD was 10.25 ± 12.72, 13.50 ± 19.66 respectively, and there grade and stage (P = 0.002, P = 0.025 respectively) was detected. were no statistically significant differences between both (P = Regarding ovarian carcinoma, a significant association Articles © The authors | Journal compilation © World J Oncol and Elmer Press Inc™ | www.wjon.org 369 NQO1 and Nrf2 in Female Genital Tract Cancer World J Oncol. 2015;6(3):364-374 Table 3. Associations Between NQO1 and Nrf2 Expression Scores and Clinicopathological Data in Carcinoma Cases NQO1 Nrf2 Organ Clinicopathological parameter No. of cases Mean ± SD P-value Mean ± SD P-value Cervix Age Mean 55.20 ± 6.54 44.34 ± 29.52 0.445 39.24 ± 29.25 0.526 Grade I 14 25.14 ± 33.26 0.045 25.14 ± 33.03 0.016 II 20 26.80 ± 29.65 28.00 ± 29.11 III 16 33.50 ± 32.61 23.00 ± 23.59 Stage I 8 40.62 ± 37.74 0.461 36.88 ± 32.28 0.594 II 28 48.32 ± 29.80 35.11 ± 29.01 III 10 42.40 ± 25.57 48.50 ± 28.67 IV 4 28.75 ± 20.15 49.75 ± 29.89 Endometrium Age Mean 51.90 ± 5.77 36.92 ± 31.94 0.802 39.82 ± 55.00 0.923 Grade I 16 17.25 ± 27.14 0.011 18.44 ± 28.61 0.002 II 21 45.38 ± 31.39 48.48 ± 31.35 III 23 47.46 ± 28.99 52.15 ± 28.06 Stage I 10 12.60 ± 23.98 0.082 13.40 ± 24.73 0.025 II 13 40.69 ± 32.61 41.54 ± 31.31 III 19 42.00 ± 31.28 49.11 ± 33.18 IV 8 49.13 ± 30.82 48.00 ± 29.21 Ovary Age Mean 55.74 ± 9.92 28.30 ± 27.29 0.734 32.70 ± 29.44 0.395 Grade I 10 22.70 ± 30.31 0.712 30.80 ± 37.00 0.576 II 23 26.87 ± 26.72 29.26 ± 30.14 III 17 33.53 ± 27.02 18.47 ± 24.10 Stage I 8 12.50 ± 21.66 0.167 16.88 ± 28.77 0.020 II 15 22.53 ± 30.07 19.80 ± 31.06 III 15 33.47 ± 25.14 45.27 ± 25.06 IV 12 39.58 ± 25.71 43.67 ± 23.50 Test of significance: Kruskal-Wallis, ANOVA tests. P-value < 0.05 is considered significant. was noticed between increased Nrf2 mean expression and ad- positive correlation was noted between NQO1 and Nrf2 mean vanced tumor stage (P = 0.020). No significant associations expression in all examined cases (r = 0.734, P < 0.001). A sig- were noticed between its expression and either age or grade. nificant positive correlation was noted between NQO1/Nrf2 (r = 0.818, P < 0.001) in SIL. Similarly, a significant positive correlation was noted (r = 0.615, P < 0.001) in carcinoma. No Correlations between immunohistochemical markers significant correlations were noted between them in normal expression cervical tissue (r = 0.615, P = 0.051). Correlations between immunohistochemical markers ex- pression in different endometrial lesions were found. A signifi- Correlations between immunohistochemical markers expres- sion in different cervical lesions were found. A significant cant positive correlation was noted between NQO1 and Nrf2 Articles © The authors | Journal compilation © World J Oncol and Elmer Press Inc™ | www.wjon.org 370 Osman et al World J Oncol. 2015;6(3):364-374 Figure 4. NrF2 expression box plots (a) in different cervical lesions; (b) in different endometrial lesions; (c) in different ovarian lesions. Horizontal lines in the boxes represent the median value of each group. The top and bottom edges of the boxes indicate the score values from 75th percentile and the 25th percentile respectively. Whiskers represent the highest and lowest values. The range is shown as a vertical line. mean expression in all examined cases (r = 0.922, P < 0.001). rate of NQO1 protein was slightly higher in well-differentiated A significant positive correlation was noted (r = 0.909, P < SCC (43.75%) than in CIN3 (40.74%) [26]. In addition, pre- 0.001) in CE. Similarly, a significant positive correlation was vious results noted more NQO1 overexpression in carcinoma noted (r = 0.637, P = 0.003) in EH. Also a significant correla - than in normal or precancerous lesions in breast [10], colon tion was in EC (r = 0. 925, P < 0.001). [27] and liver [28], and the difference reached a significant Correlations between immunohistochemical markers ex- level. Our results with previous results indicate that NQO1 up- pression in different ovarian lesions were found. A significant regulation may be an early event in cancer progression. These negative correlation was noted between NQO1 and Nrf2 (r = findings suggest that NQO1 protein level might be used as an 0.741, P < 0.001) in all cases. Similarly, a significant positive early diagnostic indicator of this disease. correlation was noted in benign tumors (r = 0.623, P = 0.003) To further illustrate that NQO1 may be an effective predic- and in carcinoma (r = 0.740, P < 0.001). No significant correla - tor of poor prognosis, the correlation between NQO1 expres- tions were noted in borderline tumors (r = 0.6432, P = 0.057). sion and clinicopathological features of cervical, endometrial and ovarian carcinomas was analyzed. We found that high-lev- el expression of the NQO1 protein was significantly correlated Discussion with poor differentiation in cervical and EC and not associated with advanced stage. NQO1 overexpression was reported to be associated with high tumor grade in carcinoma of cervix and NQO1 flavoprotein has been found to be expressed in many breast [10, 26], with advanced stage cervical [26], breast [10], body tissues [9-11]. It is conceivable that NQO1 is primarily colon [27] and liver [28] carcinomas and with nodal metasta- involved in protecting normal cells from oxidant stress. Such ses [10]. NQO1 overexpression induced tumor cell prolifera- finding has led to the suggestion that NQO1 can be impor - tion via the up-regulation of cyclins [29] and was accompanied tant in cancer chemoprevention. However, polymorphism in by an increase in other antioxidant enzymes, such as HMOX-1 the NQO1 gene has been reported to be associated with an and GST, providing tumors with increased protection against increased risk of various cancers such as breast [10], lung [11], cytotoxic agents allowing for rapid cancer progression [30]. gastric [24] and head and neck cancer [25]. These results indicated that NQO1 played a predictive role in In the present study, we found that staining of NQO1 is tumor progression and might be useful as a poor prognostic mainly localized in the cytoplasm and these observations were biomarker of cancer. in agreement with previous studies [8-10, 26, 27]. NQO1 has NQO1 overexpression in tumors but not normal tissue positive cytoplasmic expression in 76%, 60% and 65% in cer- has made it an attractive target for treatment of lung cancer. It vical, endometrial and ovarian carcinoma respectively. Previ- is the main activator of quinone-containing alkylating agents ous studies demonstrated that NQO1 immunopositivity rate such as mitomycins [31]. So that, patients with KRAS muta- ranged between 21% and 80% [10, 11, 28] in various tumors. tions may utilize quinone-containing alkylating agents more We found increased NQO1 expression from normal tissue to efficiently due to increased NQO1 expression [11]. On the SIL and cervical carcinoma and then from SIL to carcinoma. contrary, loss of NQO1 expression appeared to be candidates The difference between all examined groups was statistically significant. This finding was also observed in the endometrial for adjuvant chemotherapy in patients with cholangiocarcino- and ovarian lesions. In a previous study, a strong positive rate ma [32]. Therefore, the role of NQO1 and related inhibitors in of NQO1 protein expression in both SCCs (54.80%) and cervi- chemotherapy appears questionable. A comprehensive similar cal intraepithelial neoplasia (CINs) (27.59% in CIN-1, 34.21% analysis of the relationship between NQO1 enzyme activity in CIN-2 and 40.74% in CIN-3) was significantly higher than and chemosensitivity in female tract cancer is essential. in the normal cervix (4%). Interestingly, the strongly positive In this study, we found that Nrf2 is nuclear, and cytoplas- Articles © The authors | Journal compilation © World J Oncol and Elmer Press Inc™ | www.wjon.org 371 NQO1 and Nrf2 in Female Genital Tract Cancer World J Oncol. 2015;6(3):364-374 mic marker by immunohistochemistry in cervical and ovarian tracellular antioxidant response and is negatively regulated by tissue, while in the endometrium its expression was mainly cy- Keap1, may be partially responsible to the aggressive biologi- toplasmic with little nuclear expression and these observations cal behavior and poor clinical outcome due to its known effect were in agreement with previous studies [33-35]. Other reports of increased resistance to chemotherapeutic drugs as cisplatin detect Nrf2 expression mainly in the nucleus [27, 36, 37]. It in both endometrial and ovarian cancer cells [18, 20, 21]. These is well established that oxidative stress is the primary signal findings may also provide an opportunity for therapeutic inter - that causes cytoplasmic Nrf2 to accumulate within the nucleus vention against chemoresistance via applications of either Nrf2 [34]. It has been documented that persistent nuclear expression inhibitors or gene knockdown approaches [33]. Therefore, a of Nrf2 results in the production of antioxidants that protect new chemotherapeutic protocol that includes antioxidant ther- cancer cells from reactive oxygen species. Higher concentra- apy may be a useful method for solving chemoresistance [37]. tion of Nrf2 in the nucleus may reflect upstaging of cancer, ag- Regarding correlation between NQO1 and Nrf2 mean ex- gressive tumor behavior and poor clinical outcome [26, 33, 37]. pression in different lesions, we found positive correlations Half of ovarian carcinomas with positive nuclear Nrf2 staining between the two proteins especially in carcinomas. Similar had either Keap1 mutations or absent Keap1 mRNA expression correlations were reported [27]. This Nrf2-NQO1/MRP1 sig- resulting in platinum resistance [38]. So, nuclear Nrf2 expres- nal pathway may be attributed to the stress response and self- sion in cancer cells would have a higher malignant potential. protective effort of the cells during malignant transformation. Therefore, it is essential to evaluate nuclear expression of Nrf2. Considering the role of Nrf2 in regulating genes as NQO1 and We found Nrf2 expression was 62%, 64% and 72% in MRP1, which act to detoxify drugs or attenuate drug-induced cervical, endometrial and ovarian carcinoma respectively. Pre- oxidative stress, it is possible that highly expressed nuclear vious studies demonstrated that Nrf2 immunoreactivity was Nrf2 plays a role in increasing treatment resistance and results frequently detected in various human malignancies, such as in short survival [27]. To the best of our knowledge, this was intrahepatic chorangiocellular [32], endometrial [33], breast the first study examining NQO1 and Nrf2 in cervical, endome - [36], gastric [35, 37], ovarian [38], lung [39], pancreatic [40] trial and ovarian tissue and the change of their expression in and gallbladder [41] carcinomas, and its rate of immunoposi- different lesions in each tissue type. tivity ranged between 26% and 76% in these studies. We found increased expression from normal tissue to SIL Conclusions and cervical carcinoma and then from SIL to carcinoma. The difference between all examined groups was statistically sig- nificant. In the endometrium, we found Nrf2 expression was NQO1 and Nrf2 play a key role in the progression of female increased from CE to EH to EC, and the difference was statisti- tract tumors, and high level of both proteins were strongly as- cally significant. Similar results were reported in the endome - sociated with high grade and advanced stage. The high pro- trium with lower expression of Nrf2 in atypical EH and higher portion of NQO1 expression suggests that NQO1 may be a in endometrial cancer [33, 42-44]. Finally in ovarian tissue we significant biomarker and a potential therapeutic target for found the same results as Nrf2 expression was increased from carcinoma patients. Nrf2 expression in cancer may be useful normal to benign tumors and from benign to borderline tumors for evaluation of biological malignant potential. Overall, our and from borderline tumors to carcinomas with a significant present work implies that NQO1 and Nrf2 might be new bio- difference between the examined groups. This elevated Nrf2 markers for early diagnosis and tumorigenesis in patients with expression may be induced by gonadotropins and sex-steroid tumors of female genital tract. hormones, which suggest that these hormones are involved in ovarian cancer development via modulation of Nrf2 signaling. Therefore, its inhibition may represent an effective therapeutic Competing Interests strategy for treatment [45]. Previous studies in different organs reported that Nrf2 ex- The authors declare that they have no competing interests. pression was more overexpressed in carcinoma than normal and precancerous lesions in pancreatic [34, 40], gastric [35] and breast [36] carcinomas. Nrf2 expression may represent Author Contributions one of the early molecular events in the neoplastic transforma- tion of several tumors. Regarding association of Nrf2 overexpression and clin- N. A. Osman performed project development, data collection, icopathological data, we found that its expression was signifi- immunostaining, data analysis and manuscript writing. N. M. cantly associated with high tumor grade in cervical and EC and Abd El-Maqsoud performed project development, data collec- with advanced tumor stage in endometrial and ovarian carci- tion, immunostaining, data analysis and manuscript writing. S. noma and no significant association with the age. Our findings A. El Gelany performed data collection and manuscript writ- are in accordance with those reported by [27, 35-37, 41]. No ing. All authors read and approved the final manuscript. differences were noted in age, grade and stage in ovarian car- cinoma [38, 45]. Overexpression of Nrf2 in gallbladder adeno- References carcinoma was correlated with tumor differentiation, staging, metastasis and shorter overall survival [41]. 1. Weiderpass E, Labreche F. Malignant tumors of the female Furthermore, overexpression of Nrf2, a regulator of an in- Articles © The authors | Journal compilation © World J Oncol and Elmer Press Inc™ | www.wjon.org 372 Osman et al World J Oncol. 2015;6(3):364-374 reproductive system. Saf Health Work. 2012;3(3):166- 17. Kwak MK, Wakabayashi N, Itoh K, Motohashi H, 180. Yamamoto M, Kensler TW. Modulation of gene expression 2. Ferlay J, Shin HR, Bray F, Forman D, Mathers C, Parkin by cancer chemopreventive dithiolethiones through the DM. Cancer incidence and mortality worldwide: IARC Keap1-Nrf2 pathway. Identification of novel gene clusters CancerBase No. 10. Lyon (France): International Agency for cell survival. J Biol Chem. 2003;278(10):8135-8145. for Research on Cancer; 2010. 18. Jiang T, Huang Z, Lin Y, Zhang Z, Fang D, Zhang DD. 3. Merritt MA, Cramer DW. Molecular pathogenesis of en- The protective role of Nrf2 in streptozotocin-induced dia- dometrial and ovarian cancer. Cancer Biomark. 2010;9(1- betic nephropathy. Diabetes. 2010;59(4):850-860. 6):287-305. 19. Kobayashi A, Kang MI, Okawa H, Ohtsuji M, Zenke Y, 4. Ibrahim AS, Khaled HM, Mikhail NN, Baraka H, Kamel Chiba T, Igarashi K, et al. Oxidative stress sensor Keap1 H. Cancer incidence in egypt: results of the national pop- functions as an adaptor for Cul3-based E3 ligase to regu- ulation-based cancer registry program. J Cancer Epide- late proteasomal degradation of Nrf2. Mol Cell Biol. miol. 2014;2014:437971. 2004;24(16):7130-7139. 5. Lajin B, Alachkar A. The NQO1 polymorphism C609T 20. Cho JM, Manandhar S, Lee HR, Park HM, Kwak MK. (Pro187Ser) and cancer susceptibility: a comprehensive Role of the Nrf2-antioxidant system in cytotoxicity me- meta-analysis. Br J Cancer. 2013;109(5):1325-1337. diated by anticancer cisplatin: implication to cancer cell 6. Hu X, Zhang Z, Ma D, Huettner PC, Massad LS, Nguyen resistance. Cancer Lett. 2008;260(1-2):96-108. L, Borecki I, et al. TP53, MDM2, NQO1, and suscepti- 21. Wang XJ, Sun Z, Villeneuve NF, Zhang S, Zhao F, Li Y, bility to cervical cancer. Cancer Epidemiol Biomarkers Chen W, et al. Nrf2 enhances resistance of cancer cells to Prev. 2010;19(3):755-761. chemotherapeutic drugs, the dark side of Nrf2. Carcino- 7. Su XL, Yan MR, Yang L. NQO1 C609T polymor- genesis. 2008;29(6):1235-1243. phism correlated to colon cancer risk in farmers from 22. Clarke BA, Gilks B. Ovarian Carcinoma: Recent Devel- western region of Inner Mongolia. Chin J Cancer Res. opments in Classification of Tumour Histological Sub- 2012;24(4):317-322. type. Canadian Journal of Pathology. 2011;33-42. 8. Awadallah NS, Dehn D, Shah RJ, Russell Nash S, Chen 23. Edge SB, Byrd DR, Compton CC, Fritz AG, Greene FL, YK, Ross D, Bentz JS, et al. NQO1 expression in pan- Trotti A (eds). Ovary and primary peritoneal carcinoma. creatic cancer and its potential use as a biomarker. Appl In: AJCC Cancer Staging Manual. 7th ed. New York, NY: Immunohistochem Mol Morphol. 2008;16(1):24-31. Springer: 419-428; 2010. 9. Buranrat B, Chau-in S, Prawan A, Puapairoj A, Zeekpud- 24. Lin L, Qin Y, Jin T, Liu S, Zhang S, Shen X, Lin Z. sa P, Kukongviriyapan V. NQO1 expression correlates Significance of NQO1 overexpression for prognostic with cholangiocarcinoma prognosis. Asian Pac J Cancer evaluation of gastric adenocarcinoma. Exp Mol Pathol. Prev. 2012;13 (Suppl):131-136. 2014;96(2):200-205. 10. Yang Y, Zhang Y, Wu Q, Cui X, Lin Z, Liu S, Chen L. 25. Soucek P, Susova S, Mohelnikova-Duchonova B, Gro- Clinical implications of high NQO1 expression in breast madzinska J, Moraviec-Sztandera A, Vodicka P, Vod- cancers. J Exp Clin Cancer Res. 2014;33:14. ickova L. Polymorphisms in metabolizing enzymes and 11. Yilmaz A, Mohamed N, Patterson KA, Tang Y, Shilo K, the risk of head and neck squamous cell carcinoma in Villalona-Calero MA, Davis ME, et al. Increased NQO1 the Slavic population of the central Europe. Neoplasma. but not c-MET and survivin expression in non-small cell 2010;57(5):415-421. lung carcinoma with KRAS mutations. Int J Environ Res 26. Ma Y, Kong J, Yan G, Ren X, Jin D, Jin T, Lin L, et al. Public Health. 2014;11(9):9491-9502. NQO1 overexpression is associated with poor prognosis 12. Zappa F, Ward T, Pedrinis E, Butler J, McGown A. in squamous cell carcinoma of the uterine cervix. BMC NAD(P)H: quinone oxidoreductase 1 expression in kid- Cancer. 2014;14:414. ney podocytes. J Histochem Cytochem. 2003;51(3):297- 27. Ji L, Wei Y, Jiang T, Wang S. Correlation of Nrf2, NQO1, 302. MRP1, cmyc and p53 in colorectal cancer and their rela- 13. Marjani HA, Biramijamal F, Rakhshani N, Hossein- tionships to clinicopathologic features and survival. Int J Nezhad A, Malekzadeh R. Investigation of NQO1 genetic Clin Exp Pathol. 2014;7(3):1124-1131. polymorphism, NQO1 gene expression and PAH-DNA 28. Chiu MM, Ko YJ, Tsou AP, Chau GY, Chau YP. Analy- adducts in ESCC. A case-control study from Iran. Genet sis of NQO1 polymorphisms and p53 protein expression Mol Res. 2010;9(1):239-249. in patients with hepatocellular carcinoma. Histol Histo- 14. Bey EA, Reinicke KE, Srougi MC, Varnes M, Ander- pathol. 2009;24(10):1223-1232. son VE, Pink JJ, Li LS, et al. Catalase abrogates beta- 29. Garate M, Wani AA, Li G. The NAD(P)H:Quinone lapachone-induced PARP1 hyperactivation-directed pro- Oxidoreductase 1 induces cell cycle progression and grammed necrosis in NQO1-positive breast cancers. Mol proliferation of melanoma cells. Free Radic Biol Med. Cancer Ther. 2013;12(10):2110-2120. 2010;48(12):1601-1609. 15. Motohashi H, Yamamoto M. Nrf2-Keap1 defines a physi- 30. Lau A, Villeneuve NF, Sun Z, Wong PK, Zhang DD. ologically important stress response mechanism. Trends Dual roles of Nrf2 in cancer. Pharmacol Res. 2008;58(5- Mol Med. 2004;10(11):549-557. 6):262-270. 16. Zhang DD. Mechanistic studies of the Nrf2-Keap1 sign- 31. Beall HD, Winski SI. Mechanisms of action of quinone- aling pathway. Drug Metab Rev. 2006;38(4):769-789. containing alkylating agents. I: NQO1-directed drug de- Articles © The authors | Journal compilation © World J Oncol and Elmer Press Inc™ | www.wjon.org 373 NQO1 and Nrf2 in Female Genital Tract Cancer World J Oncol. 2015;6(3):364-374 velopment. Front Biosci. 2000;5:D639-648. lial ovarian cancer. Cancer Res. 2011;71(15):5081-5089. 32. Wakai T, Shirai Y, Sakata J, Matsuda Y, Korita PV, Taka- 39. Solis LM, Behrens C, Dong W, Suraokar M, Ozburn mura M, Ajioka Y, et al. Prognostic significance of NQO1 NC, Moran CA, Corvalan AH, et al. Nrf2 and Keap1 ab- expression in intrahepatic cholangiocarcinoma. Int J Clin normalities in non-small cell lung carcinoma and asso- Exp Pathol. 2011;4(4):363-370. ciation with clinicopathologic features. Clin Cancer Res. 33. Chen N, Yi X, Abushahin N, Pang S, Zhang D, Kong 2010;16(14):3743-3753. B, Zheng W. Nrf2 expression in endometrial serous 40. Hong YB, Kang HJ, Kwon SY, Kim HJ, Kwon KY, Cho carcinomas and its precancers. Int J Clin Exp Pathol. CH, Lee JM, et al. Nuclear factor (erythroid-derived 2)- 2010;4(1):85-96. like 2 regulates drug resistance in pancreatic cancer cells. 34. Lister A, Nedjadi T, Kitteringham NR, Campbell F, Cos- Pancreas. 2010;39(4):463-472. tello E, Lloyd B, Copple IM, et al. Nrf2 is overexpressed 41. Wang J, Zhang M, Zhang L, et al. Correlation of Nrf2, in pancreatic cancer: implications for cell proliferation HO-1, and MRP3 in gallbladder cancer and their relation- and therapy. Mol Cancer. 2011;10:37. ships to clinicopathologic features and survival. J Surg 35. Hu XF, Yao J, Gao SG, Wang XS, Peng XQ, Yang YT, Res. 2010;164: 99-105. 42. Liang SX, Chambers SK, Cheng L, Zhang S, Zhou Y, Feng XS. Nrf2 overexpression predicts prognosis and Zheng W. Endometrial glandular dysplasia: a putative pre- 5-FU resistance in gastric cancer. Asian Pac J Cancer Prev. 2013;14(9):5231-5235. cursor lesion of uterine papillary serous carcinoma. Part 36. Onodera Y, Motohashi H, Takagi K, Miki Y, Shibahara Y, II: molecular features. Int J Surg Pathol. 2004;12(4):319- Watanabe M, Ishida T, et al. NRF2 immunolocalization in 331. human breast cancer patients as a prognostic factor. En- 43. Fadare O, Liang SX, Ulukus EC, Chambers SK, Zheng docr Relat Cancer. 2014;21(2):241-252. W. Precursors of endometrial clear cell carcinoma. Am J 37. Kawasaki Y, Ishigami S, Arigami T, Uenosono Y, Yanagi- Surg Pathol. 2006;30(12):1519-1530. ta S, Uchikado Y, Kita Y, et al. Clinicopathological sig- 44. Yi X, Zheng W. Endometrial glandular dysplasia and en- nificance of nuclear factor (erythroid-2)-related fac- dometrial intraepithelial neoplasia. Curr Opin Obstet Gy- tor 2 (Nrf2) expression in gastric cancer. BMC Cancer. necol. 2008;20(1):20-25. 2015:155. 45. Liao H, Zhou Q, Zhang Z, Wang Q, Sun Y, Yi X, Feng Y. 38. Konstantinopoulos PA, Spentzos D, Fountzilas E, Fran- NRF2 is overexpressed in ovarian epithelial carcinoma coeur N, Sanisetty S, Grammatikos AP, Hecht JL, et al. and is regulated by gonadotrophin and sex-steroid hor- Keap1 mutations and Nrf2 pathway activation in epithe- mones. Oncol Rep. 2012;27(6):1918-1924. Articles © The authors | Journal compilation © World J Oncol and Elmer Press Inc™ | www.wjon.org

Journal

World Journal of Oncology – Pubmed Central

Published: Jun 12, 2015

There are no references for this article.