Abstract
We analyzed human Mut-S-Homologon-2 expression in normal endometrial tissue (n = 15) and malignancies of the uterine corpus (n = 40). Human Mut-S-Homologon-2 protein was investigated immunohistochemically on frozen sections, using a highly sensitive streptavidin-peroxidase technique and a specific mouse monoclonal antibody (clone FE11). Human Mut-S-Homologon-2 labeling pattern was compared with the staining pattern of the proliferation marker Ki-67 in the same tumors. A human Mut-S-Homologon-2 immunoreactivity score (human Mut-S-Homologon-2-IRS: negative 0-1; weak 2-3; moderate 4-6; strong 8-12) for semiquantitative analysis of human Mut-S-Homologon-2 expression is presented. In normal endometrial tissue samples we found weak nuclear immunoreactivity for human Mut-S-Homologon-2 in 67%, whereas the remaining 33% were negative for human Mut-S-Homologon-2 (mean human Mut-S-Homologon-2-IRS 1.25 ± 1.29). All malignancies of the uterine corpus analyzed revealed moderate to strong nuclear immunoreactivity (mean human Mut-S-Homologon-2-IRS 9.00, ± 3.16). Human Mut-S-Homologon-2 staining was heterogeneous, with visual differences among individual tumor cells. Expression of human Mut-S-Homologon-2 protein was consistently and strongly upregulated in tumor cells of malignancies of the uterine corpus compared with normal endometrial tissue (human Mut-S-Homologon-2-PP p<0.001; human Mut-S-Homologon-2-IS p<0.001; human Mut-S-Homologon-2-IRS p<0.001). No statistically significant correlation in comparing the labeling patterns for human Mut-S-Homologon-2 with the labeling patterns for Ki-67 (mean percentage of Ki-67-positive tumor cells 22.00% ± 17.20) was observed in malignancies of the uterine corpus (human Mut-S-Homologon-2-PP p=0.443; human Mut-S-Homologon-2-IS p=0.234; human Mut-S-Homologon-2-IRS p=0.173). Our findings indicate that human Mut-S-Homologon-2 is expressed in normal human endometrial tissue and that expression of human Mut-S-Homologon-2 may be of importance for the genetic stability of malignancies of the uterine corpus in vivo.
Keywords
It was recently demonstrated that microsatellite instability secondary to replication errors can be detected in various malignant human tumors (Duggan et al. 1994; Katabuchi et al. 1995; Quinn et al. 1995; Jovanovic et al. 1996). Microsatellite instability is a recently recognized genetic mechanism which is important in the development of various human cancers and is characterized by length changes at repetitive loci scattered throughout the genome (Koreth et al. 1996). In most patients with hereditary nonpolyposis colon cancer (HNPCC), in which almost every tumor reveals a high incidence of mutations in microsatellite repeat sequences, it was shown that cancer predisposition is attributable to defects in any one of four genes, all of which encode homologues of the microbial mismatch repair proteins mutS and mutL (Fishel et al. 1993). The human Mut-S-Homologon-2 gene specifies a mutS homologue, whereas hMLH-1, hPMS-1, and hPMS-2 encode homologues of mutL. Analysis of all 16 exons of human Mut-S-Homologon-2 in 34 unrelated HNPCC kindreds has revealed a heterogeneous spectrum of mutations (Scherer et al. 1996). Tumor cells that display microsatellite instability are typically defective in mismatch correction, thus providing a direct link between DNA mismatch repair enzymes and genetic stability afforded by this DNA repair system (Prolla et al. 1996).
Endometrial carcinoma is the most common non-rectal carcinoma in women affected by HNPCC, and microsatellite instability has been observed both in the inherited form and in approximately 20% of presumably sporadic endometrial carcinomas (Risinger et al. 1993; Burks et al. 1994).
The aim of this study was to investigate the expression of human Mut-S-Homologon-2 in normal endometrial tissue and in malignancies of the uterine corpus. We asked the following questions: (a) Do we find human Mut-S-Homologon-2 immunoreactivity in normal endometrial tissue, pointing to a functional importance of this DNA repair enzyme for the genetic stability of endometrial cells in vivo? (b) Do we find changes in the human Mut-S-Homologon-2 staining pattern in malignancies of the uterine corpus, pointing to a contribution of this DNA repair enzyme to the carcinogenesis or genetic stability of them in vivo? (c) Do we find indications for mutations of human Mut-S-Homologon-2 gene by immunostaining sporadic endometrial carcinomas using an antibody that is suggested to detect exclusively functionally active human Mut-S-Homologon-2 protein?
Materials and Methods
Endometrial Specimens
Freshly excised endometrial specimens (malignancies of the uterine corpus, n = 40; normal endometrial tissue, n = 15) were immediately embedded in OCT Tissue-Tek II (Miles Scientific; Naperville, IL), snap-frozen in liquid nitrogen, and stored at −80C. Most of the neoplasms of the uterine corpus were sporadic endometrial adenocarcinomas (serous adenocarcinoma, n = 4; mucinous adenocarcinoma, n = 2; endometrioid adenocarcinoma, n = 28; clear-cell adenocarcinoma, n = 2). Four tumors were mixed epithelial-nonepithelial tumors (heterologous carcinosarcomas, n = 4). All specimens of normal endometrial tissue were obtained from patients who underwent surgery for leiomyomas of the uterine corpus. All of these patients were in the secretory menstrual phase. The histopathological diagnosis of normal endometrial tissue was confirmed by a specialized histopathologist (K.R.).
Primary Antibody
Human Mut-S-Homologon-2 was detected by mouse monoclonal antibody FE11 (Dianova NA27; Hamburg, Germany). A polyclonal rabbit antibody directed against the Ki-67 antigen (Dako A 047; Hamburg, Germany) was used to immunophenotype proliferating cells in malignancies of the uterine corpus.
Preparation of Sections and Fixation
Serial sections (7 μm) were cut on a cryostat (Reichert-Jung; Heidelberg, Germany) and mounted on pretreated glass slides. Pretreatment of slides with 2% aminopropylmethoxysilane (Sigma; München, Germany) in acetone for 5 min was performed to enhance sticking of sections during the staining procedure. Sections to be stained for human Mut-S-Homologon-2 were fixed in 3.7% paraformaldehyde (Merck 4005; Darmstadt, Germany) in PBS [10 min at room temperature (RT)], incubated in methanol (Merck 6009; 3 min, −20C), and transferred into PBS. Sections to be stained for Ki-67 were air-dried (2 hr, RT), followed by fixation in acetone (10 min, RT), air-drying (a few minutes), chloroform (10 min, RT), air-drying, and rinsing in 0.05 M Tris-buffered saline (TBS), pH 7.6 (10 min, RT).
In Situ Detection of Human Mut-S-Homologon-2 Protein and Ki-67 Antigen
Incubation steps were performed in a moist chamber at RT, covering the sections with 100 μl of the respective reagents. The slides were incubated with the human Mut-S-Homologon-2 antibody at a dilution of 1:20 and with the Ki-67 antibody at a dilution of 1:200. After intermediate washing steps (PBS/TBS twice for 5 min), the sections were incubated with biotin-labeled rabbit anti-mouse IgG or goat anti-rabbit IgG (Dako) at a dilution of 1:400 (30 min, RT) and incubated with streptavidin-peroxidase complexes (Dako; 1:400, 30 min, RT). After rinsing, the sections were incubated with 3-amino-9-ethylcarbazole (AEC; Sigma A5754) as a substrate for the peroxidase reaction, transferred into tapwater, and mounted with Aquatex (Merck).
In control sections, primary antibodies were replaced with polyclonal mouse IgG (Dako) or polyclonal rabbit IgG (Dako). No immunoreactivity was observed in control sections.
Semiquantitative Analysis of Immunoreactivity
Microscopic analysis was performed by three independent observers (M.F., K.R., and A.W.-H.) including a specialized histopathologist (K.R.). Human Mut-S-Homologon-2 staining intensity (human Mut-S-Homologon-2-SI), percentage of human Mut-S-Homologon-2-positive cells (human Mut-S-Homologon-2-PP), and a resulting human Mut-S-Homologon-2 immunoreactivity score (human Mut-S-Homologon-2-IRS) were assessed as a modification of the technique described previously for estrogen and progesterone receptors (Remmele et al. 1986). In short, this immunoreactivity score (human Mut-S-Homologon-2-IRS: negative 0-1; weak 2-3; moderate 4-6; strong 8-12) was determined by multiplication of the values for human Mut-S-Homologon-2 staining intensity (human Mut-S-Homologon-2 SI: 0, no staining; 1, weak staining; 2, moderate staining; 3, strong staining) and the values for percentage of human Mut-S-Homologon-2-positive tumor cells (human Mut-S-Homologon-2-PP: 1, 0-10%; 2, 11-50%; 3, 51-80%; 4, 81-100%). Sections stained for Ki-67 were assessed by counting the number of Ki-67-positive and -negative cells in the most strongly stained tumor area (magnification X400; at least 200 tumor cells were counted), resulting in a Ki-67 immunoreactivity score (Ki-67-IRS: 0, no cells labeled/tumor; 1, 1-9% of cells labeled/tumor; 2, 10-25% of cells labeled/tumor; 3, 26-100% of cells labeled/tumor).
Statistics
Statistics were performed by using the Kruskal-Wallis and Mann-Whitney U-Wilcoxon Rank Sum W-tests. Statistical significance was defined as p<0.05.
Results
Expression of Human Mut-S-Homologon-2 in Normal Endometrial Tissue
A total of 33% of normal endometrial tissues were human Mut-S-Homologon-2-negative, whereas the remaining 67% revealed weak human Mut-S-Homologon-2 immunoreactivity (mean human Mut-S-Homologon-2-IRS 1.25 ± 1.29; mean human Mut-S-Homologon-2-PP 8.92 ± 12.49; mean human Mut-S-Homologon-2-IS 0.83 ± 0.83). In the endometrial specimens, the connective tissue always gave negative staining results. Fibroblasts, identified by their shape, revealed either negative or very weak staining for human Mut-S-Homologon-2. Endothelial cells of capillaries, arterioles, and venules, as well as pericytes or smooth muscle cells in the microvascular wall, revealed either negative or very weak human Mut-S-Homologon-2-immunoreactivity as well (data not shown).
Expression of Human Mut-S-Homologon-2 in Malignancies of the Uterine Corpus
All analyzed malignancies of the uterine corpus revealed strong nuclear immunoreactivity for human Mut-S-Homologon-2 (mean human Mut-S-Homologon-2-IRS 9.00 ± 3.16; mean human Mut-S-Homologon-2-PP 79.38 ± 16.01; mean human Mut-S-Homologon-2-IS 2.50 ± 0.63). There was no visual difference in comparing the labeling pattern for human Mut-S-Homologon-2 in different types of malignancies of the uterine corpus. Tumor cells revealed consistently heterogeneous and strong human Mut-S-Homologon-2 immunoreactivity (Figure 1; Table 1). In addition, expression of human Mut-S-Homologon-2 protein was consistently and markedly upregulated in tumor cells of malignancies of the uterine corpus compared with normal endometrial tissue (Figure 1; Table 1). This upregulation was statistically significant (human Mut-S-Homologon-2-PP p<0.001; human Mut-S-Homologon-2-IS p<0.001; human Mut-S-Homologon-2-IRS p<0.001).
Comparison of Human Mut-S-Homologon-2 Expression to Expression of Ki-67 Antigen in Malignancies of the Uterine Corpus
All analyzed malignancies of the uterine corpus revealed weak to strong nuclear immunoreactivity for Ki-67 (mean percentage of Ki-67-positive tumor cells 22.00 ± 17.20). Most tumors revealed heterogeneous labeling for Ki-67, with no visual differences between central and peripheral tumor areas (Figure 1). There were no statistically significant correlations between the expression of Ki-67 and the expression of human Mut-S-Homologon-2 (human Mut-S-Homologon-2-PP p=0.443; human Mut-S-Homologon-2-IS p=0.234; human Mut-S-Homologon-2-IRS p=0.173).
Discussion
To our knowledge, we have extensively analyzed for the first time the expression of the DNA mismatch repair enzyme human Mut-S-Homologon-2 in malignancies of the uterine corpus by immunohistochemistry. All endometrial carcinomas analyzed revealed strong nuclear immunoreactivity for human Mut-S-Homologon-2. In addition, our results revealed upregulation of human Mut-S-Homologon-2 protein in malignancies of the uterine corpus compared with normal endometrial tissue. It has previously been demonstrated that the majority of inactivating mutations of human Mut-S-Homologon-2 lead to a lack of expression or the expression of a truncated protein not detectable by the antibody that we used in this study (Burks et al. 1994; Katabuchi et al. 1995). The results of our immunohistochemical study therefore have several important implications. First, they suggest that all tumors we analyzed were not affected by inactivating mutations of the human Mut-S-Homologon-2 gene, adding to the body of evidence that mutations of the human Mut-S-Homologon-2 gene are not common in sporadic endometrial carcinomas (Katabuchi et al. 1995; Papadopoulos et al. 1995). Preliminary studies suggested that microsatellite instability in sporadic endometrial carcinomas is not caused by mutations in the human Mut-S-Homologon-2 gene (Katabuchi et al. 1995).
Second, our results suggest that the expression of functionally active human Mut-S-Homologon-2 is upregulated in sporadic endometrial carcinomas. This human Mut-S-Homologon-2 upregulation may be caused by the neoplastic process driven by an increase in the rate of mutations in oncogenes and tumor suppressor genes. When we compared expression of human Mut-S-Homologon-2 protein with that of Ki-67 antigen in sequential sections, tumor specimens revealed no visual correlation. These findings suggest that expression of human Mut-S-Homologon-2 is not exclusively regulated by the proliferative activity of these tumor cells but instead is influenced by different unknown mechanisms. It can be speculated whether upregulation of human Mut-S-Homologon-2 in malignancies of the uterine corpus may be induced by other genes involved in postreplication recovery.
Expression of human Mut-S-Homologon-2 (
Human Mut-S-Homologon-2 immunoreactivity (human Mut-S-Homologon-2-IRS), human Mut-S-Homologon-2 immunostaining (human Mut-S-Homologon-2-IS), and number of human Mut-S-Homologon-2-positive cells (human Mut-S-Homologon-2-PP) in malignancies of the uterine corpus (n = 40) and normal endometrial tissues (n = 15)
Recently, it has been shown that mutations in distinct mutS and mutL genes can block nucleotide excision repair, indicating a direct contribution of DNA mismatch repair enzymes to nucleotide excision repair pathways (Mellon and Champe 1996; Mellon et al. 1996).
Because nucleotide excision repair removes damaged DNA (i.e., cross linked DNA after many anticancer treatments), upregulation of biologically and functionally active human Mut-S-Homologon-2 in malignancies of the uterine corpus may enhance DNA repair capacity and increase the resistance of tumor cells to radiotherapy or chemotherapy. Upregulation of functionally active human Mut-S-Homologon-2 may be a mechanism at least in part responsible for the lack of effectiveness of various chemotherapy regimens in sporadic endometrial carcinomas. The analysis of tumors for DNA repair capacity, including human Mut-S-Homologon-2 status, might be of value in predicting the therapeutic response to chemotherapy or radiation. Both the treatment and the prognosis of endometrial carcinoma are strongly related to the stage of the disease (5-year survival rate for Stage I endometrial carcinoma 82-95%; for Stage II endometrial carcinoma 50-60%; for Stage III endometrial carcinoma 15−25%; for Stage IV endometrial carcinoma <10%) (Grigsby et al. 1987; Berek et al. 1988). Standard treatment for patients with early stage disease is surgery with or without adjuvant radiation therapy (Taghian et al. 1988). It is still debatable whether all patients should undergo pelvic or para-aortic lymph node dissection or whether this treatment should be reserved for patients with high risk factors (Creasman et al. 1987).
It was recently shown that the human Mut-S-Homologon-2 gene is directly regulated by p53 (Scherer et al. 1996). The promoter region of the human Mut-S-Homologon-2 gene was sequenced and a p53 recognition site displaying strong wild-type p53 binding activity was identified (Scherer et al. 1996). Transient co-transfection experiments using a human Mut-S-Homologon-2 promoter-reporter construct and a p53 expression plasmid confirmed the functional importance of this finding. Therefore, these data demonstrated for the first time direct regulation of a DNA mismatch repair gene by p53. We have now begun to analyze the correlation of human Mut-S-Homologon-2 and p53 expression in malignancies of the uterine corpus to answer the question of whether upregulation of human Mut-S-Homologon-2 may be induced by an increase in p53 protein in malignancies of the uterine corpus cells.