Immunohistochemical analysis of hMLH1 and hMSH2 proteins in serous ovarian tumours

Introduction

Ovarian cancer takes a first place as a cause of death from gynaecological malignancies in Poland and it is the fifth leading cause of cancer deaths among women [1]. The poor ratio of survival characteristic of ovarian cancers results from the high percentage of cases diagnosed at an advanced stage. Ovarian cancer in its early stages (I/II) is difficult to diagnose until it spreads and advances to later stages (III/IV). The early detections of ovarian cancer does not exist because of the lack of well-defined prognostic markers for prescreening. However, there is significant heterogenity within the epithelial ovarian cancer group. Most primary ovarian carcinomas are of four morphological types: serous, endometrioid, mucinous and clear cell. Histologically defined subtypes and low- and high-grade malignancies all have variable clinical manifestation and underlying molecular signatures. Relatively little is known about the molecular pathogenesis of ovarian cancer. The best recognized causes of ovarian cancer susceptibility are germline mutation in the BRCA1 and BRCA2 tumours suppressor genes [2, 3]. The role of the DNA mismatch repair (MMR) genes in this malignancy is not clearly understood.
The human DNA mismatch repair family is a highly conserved group of proteins that function in genome stabilization and mutation avoidance. The DNA mismatch repair system consists of many genes, including: hMSH2, hMLH1, hMSH3, hPMS1, hPMS2 and hMSH6. These genes are very important in distinguishing and repairing misparing and slippage errors in DNA synthesis. MMR inactivation leads to the occurrence of unrepaired deletions in mono- and dinucleotide repeats resulting in variable lenghts of these repeats. This is called microsatellite instability (MSI). MSI can be caused by genetic or epigenetic inactivation of several genes involved in MMR and is used as a marker for MMR deficiency [4, 5]. In most studies MSI is evaluated using different kinds of microsatellite markers. An MSI criterion was standardized with the use of two mononucleotide repeats (BAT26 and BAT25), and three dinucleotide repeats (D5S346, D2S123 and D17S250) known as the National Cancer Institute panel [6].
Defective DNA mismatch repair gene function is thought to promote tumourigenesis by accelerating mutations in oncogenes and tumour suppressor genes. The lack of functional hMLH1 and hMSH2 is associated with the presence of microsatellite instability found in some human tumours. Germline mutations in MMR genes are associated with colorectal, endometrial, or gastric carcinomas that develop in individuals with hereditary non-polyposis colon carcinoma (HNPCC). Some patients with HNPCC also have increased risk of other cancers, including ovarian cancer [7]. There is evidence that MSI may play a role in the mechanism of carcinogenesis.
The use of PCR-based methods to detect MSI is relatively expensive and time-consuming. Immuno-histochemical analysis offers an alternative method for estimation of MSI status as a result of the inactivation of the hMLH1 and hMSH2 genes [8]. Tumours showing high frequency of microsatellite instability are immunohistochemically characterized by lack of hMLH1 and hMSH2 protein expression. Many researchers have demonstrated that immunohisto-chemistry can be used to identify MSI [9-11]. The sensitivity and the specificity of this test were 97% and 100%, respectively [9, 11]. The use of immunohistochemistry offers a relatively rapid method for prescreening tumours for defects in the expression of MMR genes therefore, the aim of the present study was to evaluate the immunoexpression of hMLH1 and hMSH2 proteins in serous ovarian tumours.

Materials and methods

Patients
The analysis comprised 60 women diagnosed and treated for epithelial ovarian tumours at the Gynaecology and Obstetrics Institute of Medical University of Lodz between 1997 and 2002. Women were aged from 19 to 80 years (mean ±SD = 55.6 ±12.8). Twenty one women aged from 38 to 80 years (mean ±SD = 57.2 ±11.6) were treated for G3 serous cancer. Ten women aged from 40 to 73 years (mean ±SD = 57.9 ±10.1) had G2 serous cancer. Six women aged from 44 to 73 years (mean ±SD = 53.8 ±10.6) were treated for G1 serous cancer. Eight women aged from 36 to 68 years (mean ±SD = 55.4 ±10.5) had borderline serous ovarian tumours. Fifteen women aged from 19 to 75 years (mean ±SD = 52.5 ±17.7) suffered from benign tumours (cystadenomas) of the ovary.

Light microscopy
Paraffin-embedded tissue sections taken from postoperative material were diagnosed using standard haematoxylin and eosin staining.

Immunohistochemistry
Paraffin-embedded tissue sections were mounted onto SuperFrost slides, deparaffinized, then treated in a microwave oven in a solution of TRS (Target Retrieval Solution, pH 6.0, Dako) for 30 minutes (2 × 6 minutes 360 W, 2 × 5 180 W, 2 × 4 minutes 90 W) and transferred to distilled water. Endogenous peroxidase activity was blocked by 0.3% hydrogen peroxide in distilled water for 30 minutes, and then sections were rinsed with Tris-buffered saline (TBS, Dako, Denmark) and incubated all night with mouse monoclonal anti-human antibodies: hMLH1 (PharMingen International, San Diego, USA; dilution 1 : 200) and hMSH2 (abcam Inc, Cambridge, UK; dilution 1 : 250). Immunoreactive proteins were visualized using EnVision-horseradish peroxidase kit (Dako, Carpinteria, CA, USA) according to the instructions of the manufacturer. Visualisation was performed by incubating the sections in a solution of 3,3’-diaminobenzidine (Dak^oCytomation, Denmark). After washing, the sections were counter-stained with haematoxylin and coverslipped. For each antibody and for each sample a negative control was processed. Negative controls were carried out by incubation in the absence of the primary antibody and always yielded negative results.

Morphometry
The hMLH1 and hMSH2 immunoexpression was assessed by means of an image analysis system consisting of a personal computer equipped with a Pentagram graphical tablet, Indeo Fast card (frame grabber, true-color, real-time), produced by Indeo (Taiwan), and Panasonic colour TV camera (Japan) coupled with a Carl Zeiss microscope (Germany). This system was controlled by MultiScan 8.08 software, produced by Computer Scanning Systems, Poland, working under macroinstructions written specially for this analysis. The percentage of immunopositive cells in a 1000 tumour cells for each slide was measured.

Statistical analysis
Differences between groups were tested using one-way ANOVA with post-hoc RIR Tukey test, preceded by evaluation of normality. The Kruskall-Wallis test was used where appropriate. Results were considered statistically significant if p < 0.05.

Results

hMLH1 and hMSH2 protein expression was exclusively nuclear. Loss of expression of hMLH1 was detected in 3 of 37 (8.1%) cases of ovarian cancer (one case in G2 and two cases in G3), and in 4 of 8 (50%) cases of borderline tumour. hMSH2 immuno-reactivity was observed in all examined cases of ovarian cancer. Lack of hMSH2 immunoexpression was observed in 5 of 8 (62.5%) borderline tumours. Positive staining of hMLH1 and hMSH2 was detected in all cases of ovarian cystadenomas.
The morphometric data of the immunoexpression of hMLH1and hMSH2 in ovarian cancer, borderline tumours and cystadenomas are shown in Table I.
The immunoexpression of hMLH1 was lower in borderline patients as compared to all other groups however, only the difference between borderline and G2 cases (Fig. 1, Fig. 2) was statistically significant (p = 0.047). The immunoexpression of hMSH2 protein (Fig. 3, Fig. 4) was significantly lower in borderline group than in G1, G2, G3 cancers (p = 0.01, p = 0.037, p = 0.014 respectively) and benign ovarian tumours (p = 0.02). No statistically significant differences were found among particular groups of cancers and ovarian cystadenomas for hMSH2 protein.

Discussion

Ovarian cancer is one of the most common forms of hereditary cancer in adults females and the most common gynaecological malignancy in perimenopausal and postmenopausal women. It has been recently suggested that ovarian serous cancer follows a dualistic pathway with low-grade carcinomas arising from borderline tumours and high-grade carcinomas originating de novo [12]. Numerous data suggest that microsatellite instability and loss of heterozygosity are implicated in ovarian carcinogenesis [13, 14].
Ovarian cancer is a heterogeneous disease characterized by various histological types which have different MSI frequencies. Only a few reports have analyzed the immunoexpression of hMLH1 and hMSH2 proteins. Tumours are usually identified using molecular methods. The reported frequency of MSI in ovarian tumours varies, ranging from 0 to 50% [15-18]. MSI was found in 30-50% of cases of endometrioid ovarian cancer, in 38% of mucinous adenocarcinomas, in 14% of clear cell carcinomas and 0-13% of serous carcinoma [8, 16, 18-20]. Helleman et al. [21] observed that the frequency of MSI is higher in mucinous and endometrioid adeno-carcinoma compared to clear cell and serous adenocarcinoma (the overall frequencies of MSI were 22%, 16%, 9% and 8%, respectively). The same authors hypothesize that histology of cancer might be correlated with different MSI frequency.
Serous carcinoma is the most common epithelial ovarian cancer, and MSI studies have focused mainly on this subtype. In our study, positive staining of hMSH2 was detected in all examined cases of serous cancer. We observed the loss of immunoexpression of hMLH1 in 8.1% of examined cases of serous cancer. Helleman et al. [21] using PCR methods, detected MSI in 8% of cases of serous adenocarcinoma. Our results are in concordance with previous findings reporting a relatively low frequency of MSI in ovarian serous carcinomas [16, 18, 22, 23].
Our study did not reveal statistically significant differences in hMLH1 and hMSH2 immuno-expression between G1, G2 and G3 cancer or between serous carcinomas and serous cystadenomas. According to Giarnieri et al. [24] in carcinoma of the uterine cervix no significant correlation of immunoexpression of hMLH1 and hMSH2 was observed among histopathological grades, but the number of cases for G1, G2 and G3 cancer was 8, 12 and 3, respectively. To our knowledge the data concerning the immunoexpression of hMLH1 and hMSH2 protein in histopathological grade of ovarian serous cancer are rather scanty. For this reason, verification of our results in a much larger number of cases is necessary.
The role of MSI in pathogenesis of borderline tumours is unclear. In our study the loss of immunoexpression of hMLH1 was observed in 50% and MSH2 in 62.5% of cases of borderline tumour. We also observed statistically significant differences of immunoexpression of hMSH2 between G1, G2, G3 serous carcinomas and borderline tumours, as well as hMLH1 immunoexpression between G1 serous carcinomas and borderline tumours. The mechanism underlying development of ovarian borderline tumours is still not well understood. Wolf et al. [25] in microdissected samples from serous borderline tumours, using comparative genomic hybridization and/or fluorescence in situ hybridization, showed that 3 of 13 ovarian borderline tumours had detectable numerical abnormalities. Yoon et al. [23], using the PCR method and 5 conventional MSI markers, described MSI in 4 of 46 (8.6%) ovarian tumours, including 2 of 21 (9.5%) borderline ovarian tumours and 2 of 25 (8%) malignant ovarian tumours. On the other hand, Wolf et al. [26] suggest that some borderline tumours may develop through mechanisms other than chromosomal imbalances or microsatellite instability.
In conclusion, the low immunoexpression of hMSH2 and hMLH1 in borderline tumours revealed in our study suggests that the examined proteins may be involved in the development of early stages of ovarian carcinogenesis.
This study was supported by the Medical University of Lodz, grant 503-6038-1.

References
1. Jemal A, Siegel R, Ward E, et al. Cancer statistics 2009. CA Cancer J Clin 2009; 59: 225-249.
2. Ferla R, Calo V, Cascio S, et al. Founder mutations in BRCA1 and BRCA2 genes. Ann Oncol 2007; 18: 93-98.
3. Ratajska M, Brozek I, Senkus-Konefka E, et al. BRCA1 and BRCA2 point mutations and large rearrangements in breast and ovarian cancer families in Northern Poland. Oncol Rep 2008; 19: 263-268.
4. Kolas NK, Cohen PE. Novel and diverse functions of the DNA mismatch repair family in mammalian meiosis and recombination. Cytogenet Genome Res 2004; 107: 216-231.
5. Laghi L, Bianchi P, Malesci A. Differences and evolution of the methods for the assessment of microsatellite instability. Oncogene 2008; 27: 6313-6321.
6. Boland CR, Thibodeau SN, Hamilton SR, et al. A National Cancer Institute Workshop on Microsatellite Instability for cancer detection and familial predisposition: development of international criteria for the determination of microsatellite instability in colorectal cancer. Cancer Res 1998; 58: 5248-5257.
7. Brown GJ, St John DJ, Macrae FA, Aittomaki. Cancer risk in young women at risk of hereditary nonpolyposis colorectal cancer: implications for gynecologic surveillance. Gynecol Oncol 2001; 80: 346-349.
8. Cai KQ, Albarracin C, Rosen D, et al. Microsatellite instability and alteration of the expression of hMLH1 and hMSH2 in ovarian clear cell carcinoma. Hum Pathol 2004; 35: 552-559.
9. Marcus VA, Madlensky L, Gryfe R, et al. Immunohisto-chemistry for hMLH1 and hMSH2: a practical test for DNA mismatch repair-deficient tumors. Am J Surg Pathol 1999; 23: 1248-1255.
10. Plevová P, Krepelová A, Papezová M, et al. Immunohisto-chemical detection of the hMLH1 and hMSH2 proteins in hereditary non-polyposis colon cancer and sporadic colon cancer. Neoplasma 2004; 51: 275-284.
11. Thibodeau SN, French AJ, Roche PC, et al. Altered expression of hMSH2 and hMLH1 in tumors with microsatellite instability and genetic alterations in mismatch repair genes. Cancer Res 1996; 56: 4836-4840.
12. Geyer JT, López-García MA, Sánchez-Estevez C, et al. Pathogenetic pathways in ovarian endometrioid adeno-carcinoma: a molecular study of 29 cases. Am J Surg Pathol 2009; 33: 1157-1163.
13. Plisiecka-Hałasa J, Dansonka-Mieszkowska A, Kraszewska E, et al. Loss of heterozygosity, microsatellite instability and TP53 gene status in ovarian carcinomas. Anticancer Res 2008; 28: 989-996.
14. Shen XJ, Ali-Fehmi R, Weng CR, et al. Loss of heterozygosity and microsatellite instability at the Xq28 and the A/G heterozygosity of the QM gene are associated with ovarian cancer. Cancer Biol Ther 2006; 5: 523-528.
15. Allen HJ, DiCioccio RA, Hohmann P, et al. Microsatellite instability in ovarian and other pelvic carcinomas. Cancer Genet Cytogenet 2000; 117: 163-166.
16. Fujita M, Enomoto T, Yoshino K, et al. Microsatellite instability and alterations in the hMSH2 gene in human ovarian cancer. Int J Cancer 1995; 64: 361-366.
17. Gras E, Catasus L, Arguelles R, et al. Microsatellite instability, MLH-1 promoter hyper-methylation, and frameshift mutations at coding mononucleotide repeat microsatellites in ovarian tumors. Cancer 2001; 92: 2829-2836.
18. King BL, Carcangiu ML, Carter D, et al. Microsatellite instability in ovarian neoplasms. Br J Cancer 1995; 72: 376-382.
19. Krajinovic M, Richer C, Gorska-Flipot I, et al. Genomic loci susceptible to replication errors in cancer cells. Br J Cancer 1998; 78: 981-985.
20. Ohwada M, Suzuki M, Saga Y, Sato I. DNA replication errors are frequent in mucinous cystadenocarcinoma of the ovary. Cancer Genet Cytogenet 2000; 117: 61-65.
21. Helleman J, van Staveren IL, Dinjens WN, et al. Mismatch repair and treatment resistance in ovarian cancer. BMC Cancer 2006; 31: 201.
22. Haas CJ, Diebold J, Hirschmann A, et al. Microsatellite analysis in serous tumors of the ovary. Int J Gynecol Pathol 1999; 18: 158-162.
23. Yoon BS, Kim YT, Kim JH, et al. Clinical significance of microsatellite instability in sporadic epithelial ovarian tumors. Yonsei Med J 2008; 49: 272-278.
24. Giarnieri E, Mancini R, Pisani T, et al. Msh2, Mlh1, Fhit, p53, Bcl-2, and Bax expression in invasive and in situ squamous cell carcinoma of the uterine cervix. Clin Cancer Res 2000; 6: 3600-3606.
25. Wolf NG, Abdul-Karim FW, Farver C, et al. Analysis of ovarian borderline tumors using comparative genomic hybridization and fluorescence in situ hybridization. Genes Chromosomes Cancer 1999; 25: 307-315.
26. Wolf NG, Farver C, Abdul-Karim FW, Schwartz S. Analysis of microsatellite instability and X-inactivation in ovarian borderline tumors lacking numerical abnormalities by comparative genomic hybridization. Cancer Genet Cytogenet 2003; 145: 133-138.

Address for correspondence
Olga Stasikowska-Kanicka MSc , PhD
Department of Nephropathology
Medical University of Lodz
ul. Czechosłowacka 8/10
92-216 Łódź
phone\fax +48 42 679 01 91
e-mail: olgast@op.pl