ORIGINAL STUDY
Molecular Portraits of Heterogeneity Related to Cancer Stem Cells in Human Ovarian Cancer Jianfang Zeng, PhD,* Jie Ruan, MD,*Þ Lijing Luo, PhD,*þ Jie Shi, MD,§ Quancai Cui, MD,§ Jiaxin Yang, MD,* and Keng Shen, MD*
Purpose: The aim of this study was to investigate molecular portraits of heterogeneity related to cancer stem cells (CSCs) in human ovarian cancer and to access the value in diagnosis and treatment. Methods: Sixty specimens were collected in both cytoreductive and re-cytoreductive surgeries of 20 serous papillary ovarian adenocarcinoma cases. Expression density and distribution of 3 CSC markers (CD44, CD133, and CD117) and 3 stemness proteins (Bmi1, Nestin, and Oct3/4) were analyzed by immunohistochemical staining. Pairwise comparisons were performed among their expression in primary, metastasis, and relapsing tumors. Results: Some molecules presented different localization in 1 tissue, like CD133 and CD117, and all but Oct3/4 expressed differentially in different specimens of 1 case. Compared to primary or metastatic cancers, recurrent cancers show higher expression of CD133, CD117, and Bmi1, as well as higher histological grades. Conclusions: Our study indicated that there exist extratumoral and intratumoral heterogeneity in ovarian epithelial cancers related to CSCs. And this is worth further studying. Key Words: Ovarian cancer stem cell, Heterogeneity, CD117, Bmi1, Recurrence Received May 3, 2013, and in revised form September 27, 2013. Accepted for publication September 27, 2013. (Int J Gynecol Cancer 2014;24: 29Y35)
the past decades, oncologists focused on studying emI npirical combinations of nonspecific cytotoxic drugs based on typing, grading, and staging. However, overall survivals have changed very little, particularly in advanced situation. As for ovarian cancer, cytoreductive surgery followed by platinum/taxane chemotherapy resulted in a complete response in 70% of advanced patients, but most of them relapsed with resistant disease within 18 months, and the 5-year survival rate remains at 30%.1 The intractable cancers were
usually attributed to the heterogeneity. There are 2 hypotheses that illustrate the correlation between the intractable and the heterogeneity, namely, the ‘‘Darwinian’’ evolution and the cancer stem cells (CSCs) hypotheses. The hypothesis of ‘‘Darwinian’’ evolution holds that treatments threw selection pressures on cancer cells2; as a result, the sensitive subpopulations shrunk, whereas the resistant subpopulations would dominate the regrown tumors, and more dangerous and intractable tumors emerged eventually. Recently, therapy
*Department of Obstetrics and Gynecology, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing; †Department of Obstetrics and Gynecology, The 2nd Affiliated Hospital, Huaxi Medical University, Chengdu; ‡Department of Gynecologic Oncology, Beijing Obstetrics and Gynecology Hospital, Capital Medical University; and §Department of Pathology, Peking Union Medical College Hospital, Chinese Academy of Copyright * 2013 by IGCS and ESGO ISSN: 1048-891X DOI: 10.1097/IGC.0000000000000024
Medical Sciences & Peking Union Medical College, Beijing, China. Address correspondence and reprint requests to Keng Shen, MD, Department of Obstetrics and Gynecology, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, 1st Shuai Fu Yuan, Wang Fu Jing, Dong Cheng District, Beijing 100730, China. E-mail: [email protected], [email protected]. Supported by National Natural Science Foundation of China (No. 30873011). The authors declare no conflicts of interest.
International Journal of Gynecological Cancer
& Volume 24, Number 1, January 2014
Copyright © 2013 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
29
Zeng et al
International Journal of Gynecological Cancer
frustration was attributed to CSCs. Cancer stem cells were described as only a tiny subset of cancer cells characteristic of being able to self-renew, generate heterogeneous cancer cells,3 cause metastasis, and survive from chemotherapy and radiation, so they were named the ‘‘root’’ of cancer. The CSCs were characterized of higher expression of membrane efflux transporters (ie, ABCG2), enhanced DNA repair, and low mitotic index, in addition to escaping from drug attack and thrive after treatment.4 Cancer stem cells were identified based on particular markers. Pioneering work in the identification and isolation of CSCs was undertaken in various cancers including ovarian cancer. CD44+/CD117+, CD133+, CD24+, and CD117+ were identified as CSCs markers in different studies.5Y8 These markers can also be taken as the targets. Some therapeutic treatment targeting CSCs markers are being tested in clinical trials4 but no agreement about the ‘‘best markers’’ were attached yet. Multiple molecules detection might be a better choice in this situation. Besides, considering the hypothesis of ‘‘Darwinian’’ evolution together, CSCs were probably enriched at recurrence which motivated us to investigate the evolution of molecular portrait related to CSCs. Recytoreductive surgery was a common selection for the recurrent patients; therefore, sequential surgical specimens of ovarian cancer could be obtained, which made it possible to investigate the difference between primary and recurrent tumors. For this purpose, we collected 60 specimens of 20 serous papillary adenocarcinoma cases in this study and investigated expression of 6 molecules (CD44, CD133, CD117, Bmi1, Nestin, and Oct3/4) using immunohistochemical (IHC) staining. Here, CD133 was a member of pentaspan transmembrane glycoproteins, which specifically localized to cellular protrusions and reported to be a marker of adult stem cells as well as CSCs. CD117 (C-kit) was a receptor tyrosine kinase type III, which binds to stem cell factor. It was a surface marker for embryonic stem cells, hematopoietic stem cells, and mesenchymal stem cells.9 CD44 was a cell-surface glycoprotein involved in cell-cell interactions, cell adhesion, and migration as was reported to be one of the combined markers of ovarian CSCs.10 Cancer stem cells were usually reported to present high expression of ‘‘stemness genes,’’ such as Bmi1, Oct4, Nestin, and others. Stemness genes function to maintain the ability to self-renew and differentiate. However, they may be shared and independent of the CSC marker expression, then may be another important level to illustrate the molecular portraits related to CSCs. BMI1 (B lymphoma Mo-MLV insertion region 1 homolog) was a member of the polycomb group protein family which function as chromatin-modifying proteins that play important roles in embryo development11 and cancer progression. Oct-4 (octamer-binding transcription factor 4) also known as POU5F1 (POU domain, class 5, transcription factor 1) was a homeodomain transcription factor of the POU family. This protein was critically involved in the self-renewal of undifferentiated embryonic stem cells and frequently used as a marker for human germ cell tumors. Except CD44,12 overexpression of CD133, CD117, Bmi1, Nestin, and Oct3/4 were all reported to be strongly correlated with poor prognosis in cancer.13Y16
30
& Volume 24, Number 1, January 2014
MATERIALS AND METHODS Case Selection and Patient Clinical Information To investigate the clinical molecular evolution under the pressures of treatment, a pairwise study based on chronically self-control was performed. The cases in this study were selected by the following standards: (1) Underwent both primary cytoreduction and secondary cytoreduction in Peking Union Medical College Hospital (PUMCH) from January 2000 to December 2008. (2) No neoadjuvant chemotherapy was performed before the primary cytoreduction. (3) Patients should receive platinum/paclitaxel-containing chemotherapy no less than 4 courses after the primary cytoreductive surgery in PUMCH. (4) The sizes of tumor lesion in the specimens were not smaller than 1 cm3 so that the lesion was observable under a microscope more than 5 middle power fields to guarantee enough for analysis. (5) The first cytoreduction was optimal. (6) All patients signed to approve their sample and tissues to be used in research. In each case, we collected 3 specimens including 1 primary, 1 metastatic, and 1 relapsing specimen). Four-micrometer-thick paraffin-embedded tissue sections were stained with hematoxylin-eosin. Histological typing and grading was reevaluated for this study according to the WHO classification by 2 pathologists independently. All the 20 cases included in this study were diagnosed with ovarian serous papillary adenocarcinoma. In addition, this study and the protocol were approved by the institutional review board of PUMCH. Table 1 summarizes the clinicopathologic characteristics of 20 ovarian cancer patients including ages, FIGO staging, and histological grading.
IHC Staining and Antibodies The primary antibodies: Bmi-1 (1.T.21) were purchased from Abcam, USA; CD133 (C24B9) and ct-3/4(C52G3) were from Cell Signaling Technology; and CD117 (2E4), CD44 (156-3C11), and Nestin (10C2) were from Invitrogen, USA. Formalin-fixed, paraffin-embedded tissue blocks were cut into 4-Km-thick serial sections, which were mounted on polyL-lysineYcoated slides, and dried at 60-C for 1 hour. Then, the slides were deparaffinized in xylene and rehydrated conventionally. Antigen retrieving was performed by microwave (high level) in citrate buffer solution (pH 6.0) or EDTA buffer solution (pH 8.0) for 30 minutes. The endogenous peroxidase was blocked with 3% H2O2 for 10 minutes. After overnight incubation of first antibody with proper dilution at 4-C and 30 minutes incubation of antimouse and rabbit ELivision System-HRP (Maixin, Fuzhou, China) at room temperature. Diaminobenzidine was used as a chromogen (DAB substrate System; Dako, Milan, Italy). Sections were counterstained with hematoxylin, dehydrated and cleared in xylene, and then mounted and examined with a light microscope (Eclipse 80i, Nikon, Inc, Japan). Images were taken by microscope equipped with a digital camera (Nikon DMX1200).
Evaluation of the Staining All samples were analyzed by 2 histotechnicians (L-J. Luo and J. Ruan) who were both unaware of the clinical data. Image-Pro Plus Version 6.0 was used to quantify the staining. * 2013 IGCS and ESGO
Copyright © 2013 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
49 43 50 52 68 53 43 61 54 55 49 51 42 46 57 44 53 44 66 62
IIIC IIIC IIC IIIC IIIC IIIC IIIC IIIC IIIC IIIC IIIA IIIC IIIC IIIC IIIC IIIC IIIC IIIC IIIC IIIC
FIGO Staging* 2-
G2 G3 G2 G3 G2 G2 G3 G3 G3 G3 G3 G3 G2 G3 G2 G2 G2 G2 G2 G2 G3 G3 G1 G2 G2 G3 G3 G3 G2 G3 G2 G3 G2 G3 G2 G2 G3 G3 G3 G3 G3: 7/20 G3: 14/20
1-
Grading M j j + ++ ++ + ++ + j +++ +++ + ++ +++ +++ + ++ j ++ +
P j +++ + ++ ++ ++ ++ + j +++ + j +++ +++ + j + +++ ++ +
CD44 P
M
R
P
M
CD117 R
P
M
Bmi1 R
P
M
Nestin R
P
M
OCT-4 R
j + + +++ j + ++ ++ ++ +++ + + + j j j +++ j j j j j +++ + j +++ + j +++ j j j +++ j j j j j j ++ ++ +++ ++ ++ j j j j j ++ ++ +++ + ++ +++ + ++ +++ j j j j j j +++ j j j j j j +++ +++ +++ j ++ + + + + + j j j j ++ +++ + +++ +++ j j ++ + + + j j j j j j j +++ + +++ + + ++ j j j ++ j j j j j +++ + ++ ++ +++ + ++ j j j +++ ++ j +++ j +++ +++ ++ ++ +++ +++ +++ j + + + +++ j j j j j j j j +++ + j +++ j j j + j j j + + +++ ++ ++ +++ + + + j j j +++ + + ++ j j +++ ++ +++ +++ +++ +++ ++ j j j + j j j +++ +++ +++ ++ +++ +++ + + + j j j +++ j j ++ j j j + ++ ++ j + ++ j j j ++ j j j j ++ +++ ++ ++ +++ + ++ + + + + ++ + + +++ + ++ +++ + ++ +++ +++ + j j j j + j j j j + +++ + +++ +++ + j +++ j j j ++ j j j j j j ++ + +++ j + + j j j j j j j j j j ++ j ++ + j ++ j j j +++ j j j + ++ +++ + ++ ++ ++ ++ ++ j j j
R
CD133
& Volume 24, Number 1, January 2014
1-, 2- referred to the first surgery and the repeated cytoreduction surgery, respectively. *FIGO staging in the first surgery. CC, Number of platinum-paclitaxel containing chemotherapy courses; FSI, first surgery interval; M, metastatic tumors; P, primary tumors; PS, partially sensitive; R, recurrent tumors; R, resistant; S, sensitive.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Case
Age of Onset, y
TABLE 1. Expression of CSCs markers in 20 ovarian cancer patients
International Journal of Gynecological Cancer Heterogeneity of Ovarian Cancer to CSCs
* 2013 IGCS and ESGO
Copyright © 2013 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
31
International Journal of Gynecological Cancer
Zeng et al
Disagreement in the assessment of staining was found in 9 of the slides examined, and 2 pathologists further reviewed these slides and analyzed the reason of the disagreement, then a consensus was reached after. Positive reactions were identified by the presence of a brown precipitate. As for CD44, CD133, and CD117, the expression was scored as a fraction of positive cancer cells in the whole tumor area. Because the CSC hypotheses hold that the proportions of CSCs in the bulk of tumor were occasionally rare or minority, the samples were considered positive if any positive staining was detected, even if there were only rare, isolated single cells. For further statistical analysis, the expression was categorized into 4 groups according to the positive percentage of the expression: j, no positive staining was observed; +, 1% or less of the tumor cells was positive staining; ++, greater than 1% and 10% or less of the tumor cells were positive staining; and +++, greater than 10% of the tumor cells were positive staining. As for the 3 stemness gene (Oct3/ 4, Nestin, and Bmi1), the staining intensity was scored as negative, 1; weak (primrose yellow), 2; moderate (yellowish brown), 3; or strong (dark brown), 4. The proportion of immunopositive cells of interest was scored as 10% or less, 1; greater than 10% to 50% or less, 2; greater than 50% to e75% or less, 3; and greater than 75%, 4. The corresponding scores from the same sample were then multiplied. Scores 4 or less were designated as (j); 4 to 8 as (+); 8 to 12 as (++); and 12 to 16 as (+++). The intensity of staining was categorized into 4 grades as follows: j, negative; +, weak; ++, moderate; or +++, strong. With regard to the minority of positive staining for Oct3/4, CD133, and CD117, we performed staining of the 3 in 3 successive sections, respectively, and the evaluations of staining were summarized from all the 3 sections.
Statistical Analysis Statistical analysis was carried out with the SPSS statistical software package (SPSS Standard version 18.0, SPSS Inc). Wilcoxon signed rank tests were used to analyze pairwise comparisons among the expression of the primary, metastasis, and recurrent specimens. A P value less than 0.05 was considered statistically significant.
RESULTS Tumor Histological Grade Progressed in Recurrent Invasive Ovarian Carcinoma Table 1 summarizes the clinicopathologic characteristics of 20 advanced ovarian cancer patients. Age of the cases varied from 42 to 68 years and the average age was 52.95 years, close to the reported average onset age of ovarian cancer (22) (3). Among the 20 cases, 19/20 primary tumors were FIGO stage III and 1/20 FIGO stage IIC. With regard to histological grading, 12/20 cases were diagnosed as grade 2 (G2) and 7/20 were grade 3 (G3) in the initial diagnosis; whereas 7/12 of G2 cases progressed to G3 and the 1 G1 case progressed to G2 at recurrence and none retrogressed. Histological grading progressing in recurrent ovarian carcinoma indicates that recurrent tumors are more primitive and less differentiated.
32
& Volume 24, Number 1, January 2014
Heterogeneous Expression of 6 Stem Cell Markers in Serous Ovarian Cancer Immunohistochemical staining of the 6 proteins (CD44, CD117, CD133, Bmi1, Oct4, and Nestin) in positive and representative pictures in ovarian cancer were displayed in Figure 1. The positive specimens for each molecule are summarized in Table 1. We called it a positive case if we found positive expression in any of the 3 specimens. CD44, CD117, and CD133 were the 3 surface markers. The serous ovarian carcinoma is frequently positive for CD44 (19/20). In most of the cases, CD44+ epithelial cancer cells were located sporadically or in clusters (Fig. 1A), and only in a few cases (5/20) were thoroughly CD44 positive (Fig. 1G). Lymphocytes infiltrating the stromal areas were always CD44 strong positive as previously reported (indicated in hollow arrows). CD133 expression was seen at the luminal surface of cancer glands mainly with cribriform features both in colon cancer (positive control, Fig. 1B) and serous ovarian carcinoma (Fig. 1H). Expression was detected in only 6 (30%) of the 20 cases with low positive rates. CD117 is characteristic positive in GIST (Fig. 1C) and was detected positive in 13/20 serous ovarian carcinoma here (Fig. 1I). Mast cells were strong positive for CD117 (Fig. 2, indicated by hollow arrows). CD44+ and CD117+ cells were found distributed sporadically or in clusters, especially in the areas adjacent to stroma, which might support a previous finding of the CD44+/CD117+ phenotype in ovarian CSCs. CD133+ cells were usually found on the membrane of the glandular cavity (located in the middle of the adenoid structure or displayed in a scattered distribution). Distinct tissue localization of the CSC markers could be conferred that there are different CSC populations. Although Oct4 was reported usually expressed in germ cell tumor, as is indicated in Figure 1D, we had just observed weak expression in cytoplasm of the cancer cells in 4/20 cases with only a few positive nuclei (Fig. 1J). Nestin was positive in microblood vessel of all cases (Fig. 1E, indicated by hollow arrows) and cancer epithelia of 19/20 cases (Fig. 1K). Bmi1 was detected positive in nuclei of all 20 cases with high positive proportions of cancer cells simultaneously (Fig. 1F, indicated by black arrows). Sometimes, some cancer stromal cells were positive for Bmi1 too (Fig. 1L, indicated by hollow arrows). In this part, we concluded that each case has its specific molecular portrait related to CSCs, that is, individual heterogeneity.
Compared to Primary or Metastatic Cancer, Expression of CD133, CD117, and Bmi1 Increased in Recurrent Tumor Cancer stem cells were hypothesized to be the ‘‘root’’ of cancer metastasis and recurrence, and we speculated that more stemness traits would be present at relapse. Here, we respectively scored the expression of the 6 proteins in the primary, metastasis, and recurrent lesions of each patient (Table 1) and conducted pairwise comparison among the 3 groups. Except Oct3/4 expressed consistently in 3 groups, all of the other 5 markers scored differentially among the 3 groups (Table 1). When comparing their expression in metastasis tumors with primary tumors, only CD117 was * 2013 IGCS and ESGO
Copyright © 2013 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
International Journal of Gynecological Cancer
& Volume 24, Number 1, January 2014
found significantly up-regulated in metastasis tumors (P = 0.01, Tables 1 and 2). There seemed no significant
Heterogeneity of Ovarian Cancer to CSCs
difference between CD133 and Bmi1 expression (P = 0.317 and 0.363, respectively). Compared with the primary tumor, the recurrent tumors scored higher expression of CD133, CD117, and Bmi1 (P = 0.024, 0.002, and G0.001, respectively). This also occurred when compared to the metastasis tumors (P = 0.026, 0.026, and 0.003, respectively). The representative pictures of CD133, CD117, and Bmi1 are displayed in Figure 2. CD44, Oct-4, and Nestin found no significant difference among the 3 groups (Table 1), so their representative pictures were not displayed. The differential molecular portrait between primary and metastatic tumor provided further support of intratumoral heterogeneity and, that in recurrent tumors, confirmed heterogeneous progression in malignant ovarian cancer evolution.
DISCUSSION In the past decade, oncology was witnessing a targeted therapies revolution brought about by beneficial effects on some tumor types, notably monoclonal antibodies against surface molecules such as CD20 or epidermal growth factor receptor. Some other more advanced technology (ie, RNA interfering, microRNA, and gene therapy) and targeted drugs is under investigation and clinical trials.17 Cancer stem cells were speculated to be the ‘‘root’’ of cancer; therefore eliminating CSCs were thought to be prosperous. Many and varied therapeutic approaches against these cells are being evaluated and some being tested in clinical trials.4 Literally, as previously mentioned, higher expression of CSC markers and/or stemness genes usually stands for more resistance to traditional therapy and sooner revival. There is currently no agreement about the ‘‘best marker’’ for the identification of ovarian CSCs. Previous study confirmed the intratumoral heterogeneity of ovarian CSCs based on 5 clonal cell lines from 1 patient18; therefore, targeting CSCs were speculated to be more complicated than expected, and FIGURE 1. The localization of 6 markers is shown by IHC staining. Photographs of each marker (CD44, CD117, CD133, Bmi1, Oct4, and Nestin) were arranged in rows. The leftmost column showed the photographs of positive control of each marker, and other column were listed the representative photographs of ovarian serous carcinoma. A and B, CD44 expressed in stromal cell of ovarian cancer (A, indicated by hollow arrows) and in cancer cells (B, indicated by black arrows). C and D, Gastrointestinal stromal tumor (GIST) showing high expression of CD117 (C) and a cluster of cells positive for CD117 in serous ovarian carcinoma (D). E and F, CD133 positive on the glandular surface of colon cancer (E) and with luminal-like clusters in serous ovarian carcinoma (F) (indicated by black arrows). G and H, Nuclear expression of Bmi1 in colon cancer (G) and serous ovarian carcinoma (H) (indicated by black arrows); some stromal cells positive for Bmi1 in serous ovarian carcinoma (H). I and J, OCT3/4 expressed in spermatocytoma (I) and serous ovarian carcinoma (J). K and L, Nestin expressed in glioma cells (K), serous ovarian carcinoma cells (K, indicated by black arrows) and microblood vessels (L, indicated by hollow arrows). Black bar represents 50 Km. * 2013 IGCS and ESGO
Copyright © 2013 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
33
Zeng et al
International Journal of Gynecological Cancer
& Volume 24, Number 1, January 2014
FIGURE 2. CD133, CD117, and Bmi1 expressions were different in primary, metastasis, and recurrent lesions. The upper row showed CD133 staining in case 9. There is a nearly 5% positive staining in the primary tumor and more than 10% positive staining in the recurrent tumor, whereas no positive staining were observed in the metastatic tumor. The middle row listed the representative photographs CD117 staining in case 6. CD117 was negative in the primary tumor but more than 10% positive in the metastatic and relapse tumor. The bottom row listed the representative photographs of Bmi1 staining in case 4. The intensity of Bmi1 staining was weak in primary tumor, moderate in metastatic tumor, and strong in relapse tumor. Black bar represents 100 Km. multimolecular detection was indispensable. Tissue-based analysis by IHC was widely used to investigate cell-specific expression and to discriminate tumor cells from bystander cells. In this study, CD44+ and CD117+ cells were found to occur sporadically or in clusters, especially in the areas adjacent to stroma, whereas CD133+ cells were located in the middle of the adenoid structure or displayed in a multipoint distribution, so it could be conferred that CD133 and CD117 were usually localized in different clusters; we call this intratumoral heterogeneity (1). Because of intratumoral heterogeneity, eradicating CSCs would rely on multiple targets rather than single target. Multiple staining might be useful to investigate the coexpression of several molecules in parallel, but
34
it is not essential in clinical diagnoses. There are still disadvantages in IHC analysis: standardization is difficult because the key step, antigen retrieval, is difficult to be standardized; the success of IHC depends on the first antibodies; it is difficult to quantitate because of noise signals; and finally, it need well-trained pathologists and technicians. Nowadays, other efficient techniques were carried out in multiple diagnoses. For example, high-throughput analysis (ie, microarray analysis of mRNA and/or microRNA) could reveal complex and vast alterations of the transcriptomes in cancer cells.19 In this way, more related information, including typing and even therapy response predictions could be obtained when combined with bioinformatics analysis. * 2013 IGCS and ESGO
Copyright © 2013 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
International Journal of Gynecological Cancer
& Volume 24, Number 1, January 2014
TABLE 2. Statistical significance of Wilcoxon rank test of paired comparison of gene expressions among the primary, the metastatic, and the relapse tumors
CD44 CD133 CD117 Bmi1 Nestin Oct3/4
Primary and Metastasis
Primary and Relapse
Metastasis and Relapse
0.952 0.317 0.010 0.363 0.490 1.000
0.502 0.024 0.002 G0.001 0.565 1.000
0.413 0.026 0.026 0.003 0.338 1.000
In this study, 3 serial sections of each block were stained to confirm exact scores of CD133, CD117, and Oct4 because of the very low positive rates. We found that not all the sections were detected with positive signals (data not shown). Although Bmi1 and Nestin were thoroughly expressed and indicated, the stemness networks maintaining the CSC population could be independent of their marker expression. Furthermore, Bmi1 and CD117 did not express consistently in both primary and metastatic tumors, despite no statistical differences. Therefore, for comprehensive and accurate molecular typing related to CSCs, multiple biopsies were essential. According to CSC hypothesis, recurrent tumor followed by traditional treatment might selectively be a characteristic of increasing self-renew or stemness. Histological grade stands for the degree of tumor dedifferentiation. We found that 7/12 of G2 cases progressed to G3 in recurrence and none of the 20 cases retrogressed in histological grading, which is consistent with previous discovery that grading progress at recurrence.20 Correspondingly, the expression score of stem cell markers (CD133 and CD117) increased, and stemness gene (Bmi1) up-regulated remarkably in recurrent tumors versus in primary or metastatic tumors (Tables 1 and 2, Fig. 2). The increasing expression of stem cell markers or stemness genes in recurrent lesion could be interpreted that tumors gained more stemness at recurrence and might be more dangerous. Our findings here indicated that more complicated targeted therapy concerning multiple targets should be considered. In summary, we tested 6 molecules related to ovarian CSCs by IHC in 3 groups of specimens of 20 patients with advanced ovarian serous carcinoma. First, we found that the extratumoral and intratumoral heterogeneities in these cases urged multiple target therapies; second, we found that expression of CD133, CD117, and Bmi1 up-regulated remarkably in recurrent tumors versus in primary and metastatic tumors (Tables 1 and 2, Figure 2), indicating that tumors gained more stemness at recurrence and more complicated targeted therapy were preferred. As of the analysis on the basis of tumor tissues, detection of multiple CSCs markers and/or stemness genes and multiple biopsies might be helpful with the information about the molecular portraits related to CSCs. Further investigation on this point with more sophisticated and specific tools, for example, direct genetic analysis, RNA, or protein microarray studies might be valuable.
Heterogeneity of Ovarian Cancer to CSCs
REFERENCES 1. Jemal A, Siegel R, Ward E, et al. Cancer Statistics, 2009. CA Cancer J Clin. 2009;59:225Y249. 2. Nowell PC. The clonal evolution of tumor cell populations. Science. 1976;194:23Y28. 3. Clarke MF, Dick JE, Dirks PB, et al. Cancer stem cellsVperspectives on current status and future directions: AACR Workshop on Cancer Stem Cells. Cancer Res. 2006;66:9339Y9344. 4. Dean M, Fojo T, Bates S. Tumour stem cells and drug resistance. Nat Rev Cancer. 2005;5:275Y284. 5. Baba T, Convery PA, Matsumura N, et al. Epigenetic regulation of Cd133 and tumorigenicity of Cd133+ ovarian cancer cells. Oncogene. 2009;28:209Y218. 6. Gao MQ, Choi YP, Kang S, et al. Cd24+ cells from hierarchically organized ovarian cancer are enriched in cancer stem cells. Oncogene. 2010;29:2672Y2680. 7. Luo L, Zeng J, Liang B, et al. Ovarian cancer cells with the Cd117 phenotype are highly tumorigenic and are related to chemotherapy outcome. Exp Mol Pathol. 2011;91:596Y602. 8. Zhang M, Song T, Yang L, et al. Nestin and Cd133: valuable stem cell-specific markers for determining clinical outcome of glioma patients. J Exp Clin Cancer Res. 2008;27:85. 9. Lu M, Glover CH, Tien AH, et al. Involvement of tyrosine kinase signaling in maintaining murine embryonic stem cell functionality. Exp Hematol. 2007;35:1293Y1302. 10. Zhang S, Balch C, Chan MW, et al. Identification and characterization of ovarian cancer-initiating cells from primary human tumors. Cancer Res. 2008;68:4311Y4320. 11. Lee TI, Jenner RG, Boyer LA, et al. Control of developmental regulators by polycomb in human embryonic stem cells. Cell. 2006;125:301Y313. 12. Rodrı´guez-Rodrı´guez L, Sancho-Torres I, Mesonero C, et al. The CD44 receptor is a molecular predictor of survival in ovarian cancer. Med Oncol. 2003;20(3):255Y263. 13. Boldrini L, Ursino S, Gisfredi S, et al. Expression and mutational status of C-Kit in small-cell lung cancer: prognostic relevance. Clin Cancer Res. 2004;10:4101Y4108. 14. Kim HS, Yoo SY, Kim KT, et al. Expression of the stem cell markers Cd133 and Nestin in pancreatic ductal adenocarcinoma and clinical relevance. Int J Clin Exp Pathol. 2012;5:754Y761. 15. Wang Q, He W, Lu C, et al. Oct3/4 and Sox2 are significantly associated with an unfavorable clinical outcome in human esophageal squamous cell carcinoma. Anticancer Res. 2009; 29:1233Y1241. 16. Yang GF, He WP, Cai MY, et al. Intensive expression of Bmi-1 is a new independent predictor of poor outcome in patients with ovarian carcinoma. 2010;133. 17. Kreipe HH, von Wasielewski R. Beyond typing and grading: target analysis in individualized therapy as a new challenge for tumour pathology. Recent Results Cancer Res. 2007; 176:3Y6. 18. Choi YP, Shim HS, Gao MQ, et al. Molecular portraits of intratumoral heterogeneity in human ovarian cancer. Cancer Lett. 2011;307:62Y71. 19. West L, Vidwans SJ, Campbell NP, et al. A novel classification of lung cancer into molecular subtypes. PLoS One. 2012; 7:e31906. 20. Cserni G. Tumour histological grade may progress between primary and recurrent invasive mammary carcinoma. J Clin Pathol. 2002;55:293Y297.
* 2013 IGCS and ESGO
Copyright © 2013 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
35
Molecular Portraits of Heterogeneity Related to Cancer Stem Cells in Human Ovarian Cancer Jianfang Zeng, PhD,* Jie Ruan, MD,*Þ Lijing Luo, PhD,*þ Jie Shi, MD,§ Quancai Cui, MD,§ Jiaxin Yang, MD,* and Keng Shen, MD*
Purpose: The aim of this study was to investigate molecular portraits of heterogeneity related to cancer stem cells (CSCs) in human ovarian cancer and to access the value in diagnosis and treatment. Methods: Sixty specimens were collected in both cytoreductive and re-cytoreductive surgeries of 20 serous papillary ovarian adenocarcinoma cases. Expression density and distribution of 3 CSC markers (CD44, CD133, and CD117) and 3 stemness proteins (Bmi1, Nestin, and Oct3/4) were analyzed by immunohistochemical staining. Pairwise comparisons were performed among their expression in primary, metastasis, and relapsing tumors. Results: Some molecules presented different localization in 1 tissue, like CD133 and CD117, and all but Oct3/4 expressed differentially in different specimens of 1 case. Compared to primary or metastatic cancers, recurrent cancers show higher expression of CD133, CD117, and Bmi1, as well as higher histological grades. Conclusions: Our study indicated that there exist extratumoral and intratumoral heterogeneity in ovarian epithelial cancers related to CSCs. And this is worth further studying. Key Words: Ovarian cancer stem cell, Heterogeneity, CD117, Bmi1, Recurrence Received May 3, 2013, and in revised form September 27, 2013. Accepted for publication September 27, 2013. (Int J Gynecol Cancer 2014;24: 29Y35)
the past decades, oncologists focused on studying emI npirical combinations of nonspecific cytotoxic drugs based on typing, grading, and staging. However, overall survivals have changed very little, particularly in advanced situation. As for ovarian cancer, cytoreductive surgery followed by platinum/taxane chemotherapy resulted in a complete response in 70% of advanced patients, but most of them relapsed with resistant disease within 18 months, and the 5-year survival rate remains at 30%.1 The intractable cancers were
usually attributed to the heterogeneity. There are 2 hypotheses that illustrate the correlation between the intractable and the heterogeneity, namely, the ‘‘Darwinian’’ evolution and the cancer stem cells (CSCs) hypotheses. The hypothesis of ‘‘Darwinian’’ evolution holds that treatments threw selection pressures on cancer cells2; as a result, the sensitive subpopulations shrunk, whereas the resistant subpopulations would dominate the regrown tumors, and more dangerous and intractable tumors emerged eventually. Recently, therapy
*Department of Obstetrics and Gynecology, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing; †Department of Obstetrics and Gynecology, The 2nd Affiliated Hospital, Huaxi Medical University, Chengdu; ‡Department of Gynecologic Oncology, Beijing Obstetrics and Gynecology Hospital, Capital Medical University; and §Department of Pathology, Peking Union Medical College Hospital, Chinese Academy of Copyright * 2013 by IGCS and ESGO ISSN: 1048-891X DOI: 10.1097/IGC.0000000000000024
Medical Sciences & Peking Union Medical College, Beijing, China. Address correspondence and reprint requests to Keng Shen, MD, Department of Obstetrics and Gynecology, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, 1st Shuai Fu Yuan, Wang Fu Jing, Dong Cheng District, Beijing 100730, China. E-mail: [email protected], [email protected]. Supported by National Natural Science Foundation of China (No. 30873011). The authors declare no conflicts of interest.
International Journal of Gynecological Cancer
& Volume 24, Number 1, January 2014
Copyright © 2013 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
29
Zeng et al
International Journal of Gynecological Cancer
frustration was attributed to CSCs. Cancer stem cells were described as only a tiny subset of cancer cells characteristic of being able to self-renew, generate heterogeneous cancer cells,3 cause metastasis, and survive from chemotherapy and radiation, so they were named the ‘‘root’’ of cancer. The CSCs were characterized of higher expression of membrane efflux transporters (ie, ABCG2), enhanced DNA repair, and low mitotic index, in addition to escaping from drug attack and thrive after treatment.4 Cancer stem cells were identified based on particular markers. Pioneering work in the identification and isolation of CSCs was undertaken in various cancers including ovarian cancer. CD44+/CD117+, CD133+, CD24+, and CD117+ were identified as CSCs markers in different studies.5Y8 These markers can also be taken as the targets. Some therapeutic treatment targeting CSCs markers are being tested in clinical trials4 but no agreement about the ‘‘best markers’’ were attached yet. Multiple molecules detection might be a better choice in this situation. Besides, considering the hypothesis of ‘‘Darwinian’’ evolution together, CSCs were probably enriched at recurrence which motivated us to investigate the evolution of molecular portrait related to CSCs. Recytoreductive surgery was a common selection for the recurrent patients; therefore, sequential surgical specimens of ovarian cancer could be obtained, which made it possible to investigate the difference between primary and recurrent tumors. For this purpose, we collected 60 specimens of 20 serous papillary adenocarcinoma cases in this study and investigated expression of 6 molecules (CD44, CD133, CD117, Bmi1, Nestin, and Oct3/4) using immunohistochemical (IHC) staining. Here, CD133 was a member of pentaspan transmembrane glycoproteins, which specifically localized to cellular protrusions and reported to be a marker of adult stem cells as well as CSCs. CD117 (C-kit) was a receptor tyrosine kinase type III, which binds to stem cell factor. It was a surface marker for embryonic stem cells, hematopoietic stem cells, and mesenchymal stem cells.9 CD44 was a cell-surface glycoprotein involved in cell-cell interactions, cell adhesion, and migration as was reported to be one of the combined markers of ovarian CSCs.10 Cancer stem cells were usually reported to present high expression of ‘‘stemness genes,’’ such as Bmi1, Oct4, Nestin, and others. Stemness genes function to maintain the ability to self-renew and differentiate. However, they may be shared and independent of the CSC marker expression, then may be another important level to illustrate the molecular portraits related to CSCs. BMI1 (B lymphoma Mo-MLV insertion region 1 homolog) was a member of the polycomb group protein family which function as chromatin-modifying proteins that play important roles in embryo development11 and cancer progression. Oct-4 (octamer-binding transcription factor 4) also known as POU5F1 (POU domain, class 5, transcription factor 1) was a homeodomain transcription factor of the POU family. This protein was critically involved in the self-renewal of undifferentiated embryonic stem cells and frequently used as a marker for human germ cell tumors. Except CD44,12 overexpression of CD133, CD117, Bmi1, Nestin, and Oct3/4 were all reported to be strongly correlated with poor prognosis in cancer.13Y16
30
& Volume 24, Number 1, January 2014
MATERIALS AND METHODS Case Selection and Patient Clinical Information To investigate the clinical molecular evolution under the pressures of treatment, a pairwise study based on chronically self-control was performed. The cases in this study were selected by the following standards: (1) Underwent both primary cytoreduction and secondary cytoreduction in Peking Union Medical College Hospital (PUMCH) from January 2000 to December 2008. (2) No neoadjuvant chemotherapy was performed before the primary cytoreduction. (3) Patients should receive platinum/paclitaxel-containing chemotherapy no less than 4 courses after the primary cytoreductive surgery in PUMCH. (4) The sizes of tumor lesion in the specimens were not smaller than 1 cm3 so that the lesion was observable under a microscope more than 5 middle power fields to guarantee enough for analysis. (5) The first cytoreduction was optimal. (6) All patients signed to approve their sample and tissues to be used in research. In each case, we collected 3 specimens including 1 primary, 1 metastatic, and 1 relapsing specimen). Four-micrometer-thick paraffin-embedded tissue sections were stained with hematoxylin-eosin. Histological typing and grading was reevaluated for this study according to the WHO classification by 2 pathologists independently. All the 20 cases included in this study were diagnosed with ovarian serous papillary adenocarcinoma. In addition, this study and the protocol were approved by the institutional review board of PUMCH. Table 1 summarizes the clinicopathologic characteristics of 20 ovarian cancer patients including ages, FIGO staging, and histological grading.
IHC Staining and Antibodies The primary antibodies: Bmi-1 (1.T.21) were purchased from Abcam, USA; CD133 (C24B9) and ct-3/4(C52G3) were from Cell Signaling Technology; and CD117 (2E4), CD44 (156-3C11), and Nestin (10C2) were from Invitrogen, USA. Formalin-fixed, paraffin-embedded tissue blocks were cut into 4-Km-thick serial sections, which were mounted on polyL-lysineYcoated slides, and dried at 60-C for 1 hour. Then, the slides were deparaffinized in xylene and rehydrated conventionally. Antigen retrieving was performed by microwave (high level) in citrate buffer solution (pH 6.0) or EDTA buffer solution (pH 8.0) for 30 minutes. The endogenous peroxidase was blocked with 3% H2O2 for 10 minutes. After overnight incubation of first antibody with proper dilution at 4-C and 30 minutes incubation of antimouse and rabbit ELivision System-HRP (Maixin, Fuzhou, China) at room temperature. Diaminobenzidine was used as a chromogen (DAB substrate System; Dako, Milan, Italy). Sections were counterstained with hematoxylin, dehydrated and cleared in xylene, and then mounted and examined with a light microscope (Eclipse 80i, Nikon, Inc, Japan). Images were taken by microscope equipped with a digital camera (Nikon DMX1200).
Evaluation of the Staining All samples were analyzed by 2 histotechnicians (L-J. Luo and J. Ruan) who were both unaware of the clinical data. Image-Pro Plus Version 6.0 was used to quantify the staining. * 2013 IGCS and ESGO
Copyright © 2013 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
49 43 50 52 68 53 43 61 54 55 49 51 42 46 57 44 53 44 66 62
IIIC IIIC IIC IIIC IIIC IIIC IIIC IIIC IIIC IIIC IIIA IIIC IIIC IIIC IIIC IIIC IIIC IIIC IIIC IIIC
FIGO Staging* 2-
G2 G3 G2 G3 G2 G2 G3 G3 G3 G3 G3 G3 G2 G3 G2 G2 G2 G2 G2 G2 G3 G3 G1 G2 G2 G3 G3 G3 G2 G3 G2 G3 G2 G3 G2 G2 G3 G3 G3 G3 G3: 7/20 G3: 14/20
1-
Grading M j j + ++ ++ + ++ + j +++ +++ + ++ +++ +++ + ++ j ++ +
P j +++ + ++ ++ ++ ++ + j +++ + j +++ +++ + j + +++ ++ +
CD44 P
M
R
P
M
CD117 R
P
M
Bmi1 R
P
M
Nestin R
P
M
OCT-4 R
j + + +++ j + ++ ++ ++ +++ + + + j j j +++ j j j j j +++ + j +++ + j +++ j j j +++ j j j j j j ++ ++ +++ ++ ++ j j j j j ++ ++ +++ + ++ +++ + ++ +++ j j j j j j +++ j j j j j j +++ +++ +++ j ++ + + + + + j j j j ++ +++ + +++ +++ j j ++ + + + j j j j j j j +++ + +++ + + ++ j j j ++ j j j j j +++ + ++ ++ +++ + ++ j j j +++ ++ j +++ j +++ +++ ++ ++ +++ +++ +++ j + + + +++ j j j j j j j j +++ + j +++ j j j + j j j + + +++ ++ ++ +++ + + + j j j +++ + + ++ j j +++ ++ +++ +++ +++ +++ ++ j j j + j j j +++ +++ +++ ++ +++ +++ + + + j j j +++ j j ++ j j j + ++ ++ j + ++ j j j ++ j j j j ++ +++ ++ ++ +++ + ++ + + + + ++ + + +++ + ++ +++ + ++ +++ +++ + j j j j + j j j j + +++ + +++ +++ + j +++ j j j ++ j j j j j j ++ + +++ j + + j j j j j j j j j j ++ j ++ + j ++ j j j +++ j j j + ++ +++ + ++ ++ ++ ++ ++ j j j
R
CD133
& Volume 24, Number 1, January 2014
1-, 2- referred to the first surgery and the repeated cytoreduction surgery, respectively. *FIGO staging in the first surgery. CC, Number of platinum-paclitaxel containing chemotherapy courses; FSI, first surgery interval; M, metastatic tumors; P, primary tumors; PS, partially sensitive; R, recurrent tumors; R, resistant; S, sensitive.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Case
Age of Onset, y
TABLE 1. Expression of CSCs markers in 20 ovarian cancer patients
International Journal of Gynecological Cancer Heterogeneity of Ovarian Cancer to CSCs
* 2013 IGCS and ESGO
Copyright © 2013 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
31
International Journal of Gynecological Cancer
Zeng et al
Disagreement in the assessment of staining was found in 9 of the slides examined, and 2 pathologists further reviewed these slides and analyzed the reason of the disagreement, then a consensus was reached after. Positive reactions were identified by the presence of a brown precipitate. As for CD44, CD133, and CD117, the expression was scored as a fraction of positive cancer cells in the whole tumor area. Because the CSC hypotheses hold that the proportions of CSCs in the bulk of tumor were occasionally rare or minority, the samples were considered positive if any positive staining was detected, even if there were only rare, isolated single cells. For further statistical analysis, the expression was categorized into 4 groups according to the positive percentage of the expression: j, no positive staining was observed; +, 1% or less of the tumor cells was positive staining; ++, greater than 1% and 10% or less of the tumor cells were positive staining; and +++, greater than 10% of the tumor cells were positive staining. As for the 3 stemness gene (Oct3/ 4, Nestin, and Bmi1), the staining intensity was scored as negative, 1; weak (primrose yellow), 2; moderate (yellowish brown), 3; or strong (dark brown), 4. The proportion of immunopositive cells of interest was scored as 10% or less, 1; greater than 10% to 50% or less, 2; greater than 50% to e75% or less, 3; and greater than 75%, 4. The corresponding scores from the same sample were then multiplied. Scores 4 or less were designated as (j); 4 to 8 as (+); 8 to 12 as (++); and 12 to 16 as (+++). The intensity of staining was categorized into 4 grades as follows: j, negative; +, weak; ++, moderate; or +++, strong. With regard to the minority of positive staining for Oct3/4, CD133, and CD117, we performed staining of the 3 in 3 successive sections, respectively, and the evaluations of staining were summarized from all the 3 sections.
Statistical Analysis Statistical analysis was carried out with the SPSS statistical software package (SPSS Standard version 18.0, SPSS Inc). Wilcoxon signed rank tests were used to analyze pairwise comparisons among the expression of the primary, metastasis, and recurrent specimens. A P value less than 0.05 was considered statistically significant.
RESULTS Tumor Histological Grade Progressed in Recurrent Invasive Ovarian Carcinoma Table 1 summarizes the clinicopathologic characteristics of 20 advanced ovarian cancer patients. Age of the cases varied from 42 to 68 years and the average age was 52.95 years, close to the reported average onset age of ovarian cancer (22) (3). Among the 20 cases, 19/20 primary tumors were FIGO stage III and 1/20 FIGO stage IIC. With regard to histological grading, 12/20 cases were diagnosed as grade 2 (G2) and 7/20 were grade 3 (G3) in the initial diagnosis; whereas 7/12 of G2 cases progressed to G3 and the 1 G1 case progressed to G2 at recurrence and none retrogressed. Histological grading progressing in recurrent ovarian carcinoma indicates that recurrent tumors are more primitive and less differentiated.
32
& Volume 24, Number 1, January 2014
Heterogeneous Expression of 6 Stem Cell Markers in Serous Ovarian Cancer Immunohistochemical staining of the 6 proteins (CD44, CD117, CD133, Bmi1, Oct4, and Nestin) in positive and representative pictures in ovarian cancer were displayed in Figure 1. The positive specimens for each molecule are summarized in Table 1. We called it a positive case if we found positive expression in any of the 3 specimens. CD44, CD117, and CD133 were the 3 surface markers. The serous ovarian carcinoma is frequently positive for CD44 (19/20). In most of the cases, CD44+ epithelial cancer cells were located sporadically or in clusters (Fig. 1A), and only in a few cases (5/20) were thoroughly CD44 positive (Fig. 1G). Lymphocytes infiltrating the stromal areas were always CD44 strong positive as previously reported (indicated in hollow arrows). CD133 expression was seen at the luminal surface of cancer glands mainly with cribriform features both in colon cancer (positive control, Fig. 1B) and serous ovarian carcinoma (Fig. 1H). Expression was detected in only 6 (30%) of the 20 cases with low positive rates. CD117 is characteristic positive in GIST (Fig. 1C) and was detected positive in 13/20 serous ovarian carcinoma here (Fig. 1I). Mast cells were strong positive for CD117 (Fig. 2, indicated by hollow arrows). CD44+ and CD117+ cells were found distributed sporadically or in clusters, especially in the areas adjacent to stroma, which might support a previous finding of the CD44+/CD117+ phenotype in ovarian CSCs. CD133+ cells were usually found on the membrane of the glandular cavity (located in the middle of the adenoid structure or displayed in a scattered distribution). Distinct tissue localization of the CSC markers could be conferred that there are different CSC populations. Although Oct4 was reported usually expressed in germ cell tumor, as is indicated in Figure 1D, we had just observed weak expression in cytoplasm of the cancer cells in 4/20 cases with only a few positive nuclei (Fig. 1J). Nestin was positive in microblood vessel of all cases (Fig. 1E, indicated by hollow arrows) and cancer epithelia of 19/20 cases (Fig. 1K). Bmi1 was detected positive in nuclei of all 20 cases with high positive proportions of cancer cells simultaneously (Fig. 1F, indicated by black arrows). Sometimes, some cancer stromal cells were positive for Bmi1 too (Fig. 1L, indicated by hollow arrows). In this part, we concluded that each case has its specific molecular portrait related to CSCs, that is, individual heterogeneity.
Compared to Primary or Metastatic Cancer, Expression of CD133, CD117, and Bmi1 Increased in Recurrent Tumor Cancer stem cells were hypothesized to be the ‘‘root’’ of cancer metastasis and recurrence, and we speculated that more stemness traits would be present at relapse. Here, we respectively scored the expression of the 6 proteins in the primary, metastasis, and recurrent lesions of each patient (Table 1) and conducted pairwise comparison among the 3 groups. Except Oct3/4 expressed consistently in 3 groups, all of the other 5 markers scored differentially among the 3 groups (Table 1). When comparing their expression in metastasis tumors with primary tumors, only CD117 was * 2013 IGCS and ESGO
Copyright © 2013 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
International Journal of Gynecological Cancer
& Volume 24, Number 1, January 2014
found significantly up-regulated in metastasis tumors (P = 0.01, Tables 1 and 2). There seemed no significant
Heterogeneity of Ovarian Cancer to CSCs
difference between CD133 and Bmi1 expression (P = 0.317 and 0.363, respectively). Compared with the primary tumor, the recurrent tumors scored higher expression of CD133, CD117, and Bmi1 (P = 0.024, 0.002, and G0.001, respectively). This also occurred when compared to the metastasis tumors (P = 0.026, 0.026, and 0.003, respectively). The representative pictures of CD133, CD117, and Bmi1 are displayed in Figure 2. CD44, Oct-4, and Nestin found no significant difference among the 3 groups (Table 1), so their representative pictures were not displayed. The differential molecular portrait between primary and metastatic tumor provided further support of intratumoral heterogeneity and, that in recurrent tumors, confirmed heterogeneous progression in malignant ovarian cancer evolution.
DISCUSSION In the past decade, oncology was witnessing a targeted therapies revolution brought about by beneficial effects on some tumor types, notably monoclonal antibodies against surface molecules such as CD20 or epidermal growth factor receptor. Some other more advanced technology (ie, RNA interfering, microRNA, and gene therapy) and targeted drugs is under investigation and clinical trials.17 Cancer stem cells were speculated to be the ‘‘root’’ of cancer; therefore eliminating CSCs were thought to be prosperous. Many and varied therapeutic approaches against these cells are being evaluated and some being tested in clinical trials.4 Literally, as previously mentioned, higher expression of CSC markers and/or stemness genes usually stands for more resistance to traditional therapy and sooner revival. There is currently no agreement about the ‘‘best marker’’ for the identification of ovarian CSCs. Previous study confirmed the intratumoral heterogeneity of ovarian CSCs based on 5 clonal cell lines from 1 patient18; therefore, targeting CSCs were speculated to be more complicated than expected, and FIGURE 1. The localization of 6 markers is shown by IHC staining. Photographs of each marker (CD44, CD117, CD133, Bmi1, Oct4, and Nestin) were arranged in rows. The leftmost column showed the photographs of positive control of each marker, and other column were listed the representative photographs of ovarian serous carcinoma. A and B, CD44 expressed in stromal cell of ovarian cancer (A, indicated by hollow arrows) and in cancer cells (B, indicated by black arrows). C and D, Gastrointestinal stromal tumor (GIST) showing high expression of CD117 (C) and a cluster of cells positive for CD117 in serous ovarian carcinoma (D). E and F, CD133 positive on the glandular surface of colon cancer (E) and with luminal-like clusters in serous ovarian carcinoma (F) (indicated by black arrows). G and H, Nuclear expression of Bmi1 in colon cancer (G) and serous ovarian carcinoma (H) (indicated by black arrows); some stromal cells positive for Bmi1 in serous ovarian carcinoma (H). I and J, OCT3/4 expressed in spermatocytoma (I) and serous ovarian carcinoma (J). K and L, Nestin expressed in glioma cells (K), serous ovarian carcinoma cells (K, indicated by black arrows) and microblood vessels (L, indicated by hollow arrows). Black bar represents 50 Km. * 2013 IGCS and ESGO
Copyright © 2013 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
33
Zeng et al
International Journal of Gynecological Cancer
& Volume 24, Number 1, January 2014
FIGURE 2. CD133, CD117, and Bmi1 expressions were different in primary, metastasis, and recurrent lesions. The upper row showed CD133 staining in case 9. There is a nearly 5% positive staining in the primary tumor and more than 10% positive staining in the recurrent tumor, whereas no positive staining were observed in the metastatic tumor. The middle row listed the representative photographs CD117 staining in case 6. CD117 was negative in the primary tumor but more than 10% positive in the metastatic and relapse tumor. The bottom row listed the representative photographs of Bmi1 staining in case 4. The intensity of Bmi1 staining was weak in primary tumor, moderate in metastatic tumor, and strong in relapse tumor. Black bar represents 100 Km. multimolecular detection was indispensable. Tissue-based analysis by IHC was widely used to investigate cell-specific expression and to discriminate tumor cells from bystander cells. In this study, CD44+ and CD117+ cells were found to occur sporadically or in clusters, especially in the areas adjacent to stroma, whereas CD133+ cells were located in the middle of the adenoid structure or displayed in a multipoint distribution, so it could be conferred that CD133 and CD117 were usually localized in different clusters; we call this intratumoral heterogeneity (1). Because of intratumoral heterogeneity, eradicating CSCs would rely on multiple targets rather than single target. Multiple staining might be useful to investigate the coexpression of several molecules in parallel, but
34
it is not essential in clinical diagnoses. There are still disadvantages in IHC analysis: standardization is difficult because the key step, antigen retrieval, is difficult to be standardized; the success of IHC depends on the first antibodies; it is difficult to quantitate because of noise signals; and finally, it need well-trained pathologists and technicians. Nowadays, other efficient techniques were carried out in multiple diagnoses. For example, high-throughput analysis (ie, microarray analysis of mRNA and/or microRNA) could reveal complex and vast alterations of the transcriptomes in cancer cells.19 In this way, more related information, including typing and even therapy response predictions could be obtained when combined with bioinformatics analysis. * 2013 IGCS and ESGO
Copyright © 2013 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
International Journal of Gynecological Cancer
& Volume 24, Number 1, January 2014
TABLE 2. Statistical significance of Wilcoxon rank test of paired comparison of gene expressions among the primary, the metastatic, and the relapse tumors
CD44 CD133 CD117 Bmi1 Nestin Oct3/4
Primary and Metastasis
Primary and Relapse
Metastasis and Relapse
0.952 0.317 0.010 0.363 0.490 1.000
0.502 0.024 0.002 G0.001 0.565 1.000
0.413 0.026 0.026 0.003 0.338 1.000
In this study, 3 serial sections of each block were stained to confirm exact scores of CD133, CD117, and Oct4 because of the very low positive rates. We found that not all the sections were detected with positive signals (data not shown). Although Bmi1 and Nestin were thoroughly expressed and indicated, the stemness networks maintaining the CSC population could be independent of their marker expression. Furthermore, Bmi1 and CD117 did not express consistently in both primary and metastatic tumors, despite no statistical differences. Therefore, for comprehensive and accurate molecular typing related to CSCs, multiple biopsies were essential. According to CSC hypothesis, recurrent tumor followed by traditional treatment might selectively be a characteristic of increasing self-renew or stemness. Histological grade stands for the degree of tumor dedifferentiation. We found that 7/12 of G2 cases progressed to G3 in recurrence and none of the 20 cases retrogressed in histological grading, which is consistent with previous discovery that grading progress at recurrence.20 Correspondingly, the expression score of stem cell markers (CD133 and CD117) increased, and stemness gene (Bmi1) up-regulated remarkably in recurrent tumors versus in primary or metastatic tumors (Tables 1 and 2, Fig. 2). The increasing expression of stem cell markers or stemness genes in recurrent lesion could be interpreted that tumors gained more stemness at recurrence and might be more dangerous. Our findings here indicated that more complicated targeted therapy concerning multiple targets should be considered. In summary, we tested 6 molecules related to ovarian CSCs by IHC in 3 groups of specimens of 20 patients with advanced ovarian serous carcinoma. First, we found that the extratumoral and intratumoral heterogeneities in these cases urged multiple target therapies; second, we found that expression of CD133, CD117, and Bmi1 up-regulated remarkably in recurrent tumors versus in primary and metastatic tumors (Tables 1 and 2, Figure 2), indicating that tumors gained more stemness at recurrence and more complicated targeted therapy were preferred. As of the analysis on the basis of tumor tissues, detection of multiple CSCs markers and/or stemness genes and multiple biopsies might be helpful with the information about the molecular portraits related to CSCs. Further investigation on this point with more sophisticated and specific tools, for example, direct genetic analysis, RNA, or protein microarray studies might be valuable.
Heterogeneity of Ovarian Cancer to CSCs
REFERENCES 1. Jemal A, Siegel R, Ward E, et al. Cancer Statistics, 2009. CA Cancer J Clin. 2009;59:225Y249. 2. Nowell PC. The clonal evolution of tumor cell populations. Science. 1976;194:23Y28. 3. Clarke MF, Dick JE, Dirks PB, et al. Cancer stem cellsVperspectives on current status and future directions: AACR Workshop on Cancer Stem Cells. Cancer Res. 2006;66:9339Y9344. 4. Dean M, Fojo T, Bates S. Tumour stem cells and drug resistance. Nat Rev Cancer. 2005;5:275Y284. 5. Baba T, Convery PA, Matsumura N, et al. Epigenetic regulation of Cd133 and tumorigenicity of Cd133+ ovarian cancer cells. Oncogene. 2009;28:209Y218. 6. Gao MQ, Choi YP, Kang S, et al. Cd24+ cells from hierarchically organized ovarian cancer are enriched in cancer stem cells. Oncogene. 2010;29:2672Y2680. 7. Luo L, Zeng J, Liang B, et al. Ovarian cancer cells with the Cd117 phenotype are highly tumorigenic and are related to chemotherapy outcome. Exp Mol Pathol. 2011;91:596Y602. 8. Zhang M, Song T, Yang L, et al. Nestin and Cd133: valuable stem cell-specific markers for determining clinical outcome of glioma patients. J Exp Clin Cancer Res. 2008;27:85. 9. Lu M, Glover CH, Tien AH, et al. Involvement of tyrosine kinase signaling in maintaining murine embryonic stem cell functionality. Exp Hematol. 2007;35:1293Y1302. 10. Zhang S, Balch C, Chan MW, et al. Identification and characterization of ovarian cancer-initiating cells from primary human tumors. Cancer Res. 2008;68:4311Y4320. 11. Lee TI, Jenner RG, Boyer LA, et al. Control of developmental regulators by polycomb in human embryonic stem cells. Cell. 2006;125:301Y313. 12. Rodrı´guez-Rodrı´guez L, Sancho-Torres I, Mesonero C, et al. The CD44 receptor is a molecular predictor of survival in ovarian cancer. Med Oncol. 2003;20(3):255Y263. 13. Boldrini L, Ursino S, Gisfredi S, et al. Expression and mutational status of C-Kit in small-cell lung cancer: prognostic relevance. Clin Cancer Res. 2004;10:4101Y4108. 14. Kim HS, Yoo SY, Kim KT, et al. Expression of the stem cell markers Cd133 and Nestin in pancreatic ductal adenocarcinoma and clinical relevance. Int J Clin Exp Pathol. 2012;5:754Y761. 15. Wang Q, He W, Lu C, et al. Oct3/4 and Sox2 are significantly associated with an unfavorable clinical outcome in human esophageal squamous cell carcinoma. Anticancer Res. 2009; 29:1233Y1241. 16. Yang GF, He WP, Cai MY, et al. Intensive expression of Bmi-1 is a new independent predictor of poor outcome in patients with ovarian carcinoma. 2010;133. 17. Kreipe HH, von Wasielewski R. Beyond typing and grading: target analysis in individualized therapy as a new challenge for tumour pathology. Recent Results Cancer Res. 2007; 176:3Y6. 18. Choi YP, Shim HS, Gao MQ, et al. Molecular portraits of intratumoral heterogeneity in human ovarian cancer. Cancer Lett. 2011;307:62Y71. 19. West L, Vidwans SJ, Campbell NP, et al. A novel classification of lung cancer into molecular subtypes. PLoS One. 2012; 7:e31906. 20. Cserni G. Tumour histological grade may progress between primary and recurrent invasive mammary carcinoma. J Clin Pathol. 2002;55:293Y297.
* 2013 IGCS and ESGO
Copyright © 2013 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
35
