Expression of c-Kit (CD117) in Benign and Malignant Human Endometrial Epithelium.
Molecular abnormalities frequently observed in endometrioid carcinoma include early PTEN and K-ras mutations, microsatellite instability, and late p53 mutations.[3-9] While it is well documented that endometrial epithelial cells express estrogen receptors, the mechanism by which this hormone exerts its influence in the development of endometrioid carcinoma is poorly understood. When unopposed, estrogens are believed to serve as a promoting influence, resulting in persistent epithelial proliferation predisposing the endometrium to acquiring significant genetic alterations. Recent data also suggest that estrogen metabolites may be directly and indirectly genotoxic.
The proto-oncogene c-kit encodes a 145-kd transmembrane tyrosine kinase receptor (CD117), the ligand of which is stem cell factor (SCF).[13-15] Expression of c-Kit is weak to moderate in a wide variety of normal, benign, and malignant tissues.[16-18] In a subset of mast cell disorders and gastrointestinal stromal cell tumors,[20,21] over-expression has been linked to either an activating or dominant-inactivating c-kit mutation. Besides a mutational mechanism, it has been proposed that c-kit, together with SCE may also contribute to tumor development via autocrine growth stimulation, paracrine growth stimulation, or both. It is currently unclear whether c-Kit is expressed in endometrial tissues. Contrary to previous data,[23-26] Arber et al have recently demonstrated by immunohistochemistry that endometrial adenocarcinomas (8/8) and hyperplastic endometria (2/2) express CD117. To better define and help resolve these conflicting findings, we evaluated the expression of this growth factor receptor in endometrial tissues using both immunohistochemistry and immunoprecipitation followed by Western blot analysis.
MATERIALS AND METHODS
Formalin-fixed, paraffin-embedded tissues from 7 endometrial hyperplasias, 14 proliferative endometria, and 14 secretory endometria were immunohistochemically analyzed for CD117 expression. Four-micrometer tissue sections were deparaffinized in xylene, rehydrated in graded alcohols, and washed in phosphate-buffered saline. The tissue sections were heated in a microwave in 10 mmol/L citrate buffer solution for antigen retrieval at pH 6.0 in a thermoresistant container for 12 minutes and cooled for 20 minutes. Endogenous peroxidase activity was quenched in 0.3% hydrogen peroxide for 10 minutes. The sections were then incubated for 45 minutes at room temperature with a c-Kit monoclonal antibody that does not bind to the SCF binding site (clone K45; 1:20 dilution, NeoMarkers, Fremont, Calif). Saline was applied instead of primary antibody as a negative control. Antigen-antibody complexes were visualized using a streptavidin-biotin staining technique (Vector Laboratories, Burlingame, Calif) according to the manufacturer's recommendations. Diaminobenzidine was used as a chromogen, and hematoxylin served as a counterstain. Human small cell lung carcinoma cell lines H526 and H526A (generously provided by Geoffrey Krystal, MD) served as positive controls. The immunostaining results were recorded with respect to intensity (1+ to 3+) and distribution of staining (focal indicates [is less than]10% of cells; intermediate, [is greater than]10% but [is less than]50% of cells; and diffuse, [is greater than]50% of cells).
One endometrial polyp, obtained from a case of cryopreserved endometrioid carcinoma, and 9 frozen archived endometrioid endometrial carcinomas were also analyzed by immunohistochemistry in order to examine expression of CD117 in malignant endometria, as well as to identify suitable cases for CD117 immunoprecipitation and Western blot analysis (selection criteria and blotting methodology are described below). Five-micrometer tissue sections were fixed in acetone for 10 minutes, air dried, re-hydrated in graded alcohols, and washed in phosphate-buffered saline. Following a 10-minute incubation in 0.3% hydrogen peroxide, the tissue sections were rinsed in saline and then incubated for 45 minutes with either a c-Kit monoclonal antibody (clone K45, 1:20 dilution) or saline, and processed as described above. As with the formalin-fixed tissues, immunostaining results were recorded with respect to intensity and distribution.
Immunoprecipitation and Western Blot Analyses
Cell lysates were prepared from thick (about 60-[micro]m) sections of endometrial tissue using the same frozen tissue blocks previously analyzed by immunohistochemistry for CD117 expression. Cases were selected that consisted predominantly of tumor with relatively few mast cells and histiocytes (both c-Kit-expressing cell types). Tissue sections were lysed in 20 mmol/L Tris-HCl (pH 7.4), 50 mmol/L sodium chloride, 1% Nonidet P-40, and a cocktail of protease inhibitors (Boehringer-Mannheim, Mannheim, Germany) for 30 minutes on ice. Following centrifugation for 30 minutes at 14 000 g, protein concentrations were measured using a Bradford-based colorimetric assay (Bio-Rad, Hercules, Calif). Equal amounts of protein (1 mg) from 2 endometrial carcinomas and 1 endometrial polyp were immunoprecipitated overnight at 4 [degrees] C with either 1 [micro]g c-Kit monoclonal antibody (clone K45) or saline, and then incubated with protein A/G PLUS agarose (Santa Cruz Biotechnology, Santa Cruz, Calif) for 1 hour at 4 [degrees] C. Antigen-antibody complexes were extensively washed, separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, electrotransferred onto Hybond ECL nitrocellulose membrane (Amersham, Arlington Heights, Ill), and then immunoblotted with c-Kit monoclonal antibodies (clone K45 monoclonal and clone 1.D9.3D6 monoclonal, Roche/Boehringer Mannheim, Indianapolis, Ind) followed by horseradish peroxidase-labeled anti-mouse IgG (Vector). A luminol-based chemiluminescent reaction (ECL Reagent; Amersham) allowed for the identification of horseradish peroxidase-labeled secondary antibody. For positive controls, H526 cell lysates were either immunoprecipitated (100 [micro]g protein) as described or directly loaded onto the gel (40 [micro]g protein) along with the immunoprecipitated endometrial proteins.
The immunohistochemical results for CD117 expression in formalin-fixed, paraffin-embedded benign endometrial tissue are summarized in Table 1. Thirteen (93%) of the proliferative endometria were positively immunostained for CD117, of which 5 cases had intense, focal cytoplasmic staining within epithelial cells (Figure 1), 6 cases exhibited intense membrane staining correlating with the terminal bar of isolated ciliated epithelial cells (Figure 2), and 2 cases had both cytoplasmic and terminal bar staining. Eleven (79%) of the secretory endometria expressed CD117; however, cytoplasmic staining of epithelial cells was less intense (Table 1), and only 1 case had membrane staining similar to that noted in ciliated cells of the proliferative endometria. Four (57%) of the endometrial hyperplasias stained positively for CD117. In all 4 cases, 3 simple hyperplasias and 1 complex atypical hyperplasia, an intense cytoplasmic staining reaction was observed (Figure 3). The cytoplasmic staining in both the hyperplasias and proliferative endometria was characterized by a paranuclear reaction, predominantly in the subnuclear/ basal compartment, but also in the supranuclear region. Consistent with previous data,[18,28] mast cells and to a lesser degree endothelial cells within the endometrium and myometrium were positively immunostained, thereby providing an internal positive control for each case. Histiocytes also stained positively for CD117. No staining reaction was ever detected following omission of the primary antibody.
[Figure 1-3 ILLUSTRATION OMITTED]
[TABULAR DATA 1 NOT REPRODUCIBLE IN ASCII]
We then immunohistochemically analyzed 9 randomly selected, frozen endometrioid carcinomas and 1 endometrial polyp that arose in association with endometrioid carcinoma (the carcinoma was not represented in the cryopreserved tissues) to determine whether malignant epithelia also expressed CD117. As shown in Table 2, the polyp and 4 of 9 carcinomas expressed CD117. Immunostaining was localized to the cytoplasm of endometrial epithelial cells and varied in intensity from moderate to intense (Figure 4, A). We then analyzed 3 of these same cryopreserved tissues by immunoprecipitation followed by Western blotting. Our intent was to evaluate the molecular weight and amount of CD117 in endometrial tissue to confirm antibody specificity, as well as to potentially identify aberrant forms (ie, truncations) of this growth factor receptor in endometria. One case of endometrioid carcinoma (case 2) and the endometrial polyp (case 3) were selected for analysis based on their intense CD117 immunostaining reaction and relatively low numbers of mast cells and histiocytes. Also included was a single case of endometrioid carcinoma (case 1), which had undetectable CD117 expression within the tumor by immunohistochemistry and relatively low numbers of mast cells and histiocytes. As shown in Figure 4, B, a 145-kd band corresponding to full length c-CD117 is present in all samples, with substantially more in carcinoma case 2 than in carcinoma case 1, a finding highly consistent with our immunostaining data (Figure 4, A). The endometrial polyp (case 3) had a more prominent 145-kd band than the negatively stained carcinoma; however, the level of expression was less than that of the CD117-overexpressing carcinoma. The faint 145-kd band in carcinoma case 1 likely represents expression of c-Kit by residual mast cells, histiocytes, and endothelial cells. Nonspecific bands with varying degrees of intensity were present in all lanes of CD1107-immunoprecipitated samples, including a cell lysate prepared from the positive control cell line, H526.
[Figure 4 ILLUSTRATION OMITTED]
[TABULAR DATA 2 NOT REPRODUCIBLE IN ASCII]
In this study we demonstrated in situ expression of CD117 in benign endometria. Although these data are consistent with a study by Arber et al, we did not detect immunostaining in endometrial stromal cells, as previously described in normal endometrial tissue and decidua.[24,25] We are aware of 4 independent studies that have been unable to detect CD117 by immunohistochemistry in normal endometrial epithelial tissue.[16,24-26] A close examination of the methodologies employed in the different studies reveals that when a citrate antigen-retrieval step was used with formalin-fixed, paraffin-embedded tissues, CD117 staining was detectable (our data and reference 18), whereas when cryopreserved tissues were immunostained using a standard protocol (ie, no antigen retrieval with frozen samples), staining was scored as negative.[16,24-26] Therefore, in benign endometria in which levels of c-Kit expression may be lower relative to the carcinomas, antigen retrieval may be necessary to increase the sensitivity of the assay. Moreover, tissue architecture is often compromised and background levels are higher in cryopreserved tissues, making immunohistochemical interpretation difficult if expression levels are low.
Perhaps one of the most interesting findings in this study was that in benign endometrial tissue, CD117 immunostaining tended to be more intense and/or more frequently observed in epithelial cells of hyperplastic and proliferative endometria when compared with secretory endometria. Because CD117 is generally regarded as a plasma membrane-bound protein, we expected the staining to be localized predominantly on the cell surface. However, membrane staining was less common than cytoplasmic staining in benign endometria, as reported previously by us and others. Predominantly cytoplasmic c-Kit immunostaining has also been described in a number of other normal human tissues. The significance of both the spatial focality and cellular localization of CD117 immunostaining in benign endometria remains to be determined. The specific and distinctive staining pattern of ciliated epithelial cells, known to increase in prominence under the influence of estrogen, and the relatively more intense staining of proliferative and hyperplastic endometrium suggest a possible relationship between estrogens and CD117 expression in endometrial tissue. Although merely speculative at this juncture, we hypothesize that c-Kit expression in endometrial epithelium may be up-regulated in an hyperestrogenic milieu. Together with its ligand (SCF), which is expressed in endometrial stromal cells and present in sera of blood, c-kit could then potentially allow for growth regulation in either a paracrine or endocrine manner, resulting in increased proliferation of glandular epithelial cells and placing that epithelium at increased risk of acquiring neoplasia-fomenting mutations. Further investigation into our hypothesis of an estrogen-driven, c-kit/SCF model of endometrial tumorigenesis is pending the availability of reliable SCF antibodies and the conduction of in vitro studies.
Here we have also demonstrated by immunohistochemistry the expression of CD117 in a subset of human endometrial carcinomas (4/9) and a carcinoma-related polyp. As with benign endometria, immunostaining was; principally located in the cytoplasm of the carcinomas, consistent with the findings of Arber et al and our previous experience. In contrast, 2 independent studies failed to detect CD117 expression in endometrial carcinomas. In an immunohistochemical study by Matsuda et al, a very small sample set (n = 4) could be the explanation. In the other, failure to detect c-kit mRNA by Northern blotting could have resulted owing to the analysis of total RNA rather than poly-A-enriched RNA. As expected, data indicate that reverse-transcription polymerase chain reaction is more sensitive at detecting c-kit mRNA transcripts in gastric carcinoma cell lines, suggesting low levels of c-kit message. We cannot, however, explain why 10 frozen endometrial carcinomas analyzed by Natali et al had undetectable CD117 immunostaining. Recognizing the contradictory data regarding expression of CD117 in endometrial carcinomas, we felt it was critical to also perform Western blot analysis with some of the same cryopreserved tumor samples analyzed by immunohistochemistry. Using this approach, we have correlated levels of CD117 immunoprecipitated from lysed endometrial tissue with immunostaining intensity, thereby confirming expression of CD117 in some malignant endometria.
Our Western blot analysis demonstrated that CD117, expressed in the endometrial polyp and carcinoma, is about 145 kd, which is consistent with the size of the full-length CD117 protein. It is most likely that the smaller, fainter bands on the blot (ranging in size front about 80 to 120 kd) represent nonspecific antibody-protein interactions. However, based on the data generated from this study, we cannot exclude the possibility that some of the bands may represent CD117 truncations, as previously described in gastric and colon cancer cells. Moreover, we cannot conclude that full-length CD117 expressed in these endometrial tissues is a functionally normal gene product. Clearly, there are numerous examples, most notably in gastrointestinal stromal cell tumors[20,21] and mastocytosisy,[19,34] of mutations resulting either in a constitutively active or nonfunctional receptor. Examining those endometrioid carcinoma cases with positive immunostaining for c-kit gene mutations is certainly warranted.
Using a monoclonal antibody, we detected CD117 expression in a high percentage of benign endometrium, with the most intense immunostaining in hyperplastic and proliferative endometrium. Our study also demonstrates that CD117 is expressed in a subset of endometrial carcinomas using both immunohistochemistry and Western blot analyses. We feel these findings provide a solid foundation for now addressing an important question: Does c-kit play a role in endometrial carcinogenesis? Based on data generated in this study, we can now focus on a number of relevant issues, namely, the effects of estrogen on levels of CD117 in primary human endometrial epithelial cell cultures, a mutational analysis of c-kit in endometrial carcinomas, and the functionality of this tyrosine kinase receptor in human endometrial carcinoma cell lines.
We thank Geoffrey Krystal, MD, McGuire Virginia Medical Center, Richmond, Va, for kindly providing the small cell lung carcinoma cell lines. We are also indebted to Renee Workman, BS, Tracey Lamb, BA, and Teresa Gainey, BS, for their fine technical support in the immunohistochemical studies.
[1.] Landis SH, Murray T, Bolden S, Wingo PA. Cancer statistics. CA Cancer J Clin. 1999;49:8-31, 1.
[2.] Practice guidelines: uterine corpus--sarcomas: Society of Gynecologic Oncologists clinical practice guidelines. Oncology (Huntingt.) 1998;12:284-286.
[3.] Burks RT, Kessis TD, Cho KR, Hadrick L. Microsatellite instability in endometrial carcinoma. Oncogene. 1994;9:1163-1166.
[4.] Boyd J, Risinger JI. Analysis of oncogene alterations in human endometrial carcinoma: prevalence of ras mutations. Mol Carcinog. 1991;4:189-195.
[5.] Ignar-Trowbridge D, Risinger II, Dent GA, et al. Mutations of Ki-ras oncogene in endometrial carcinoma. Am J Obstet Gynecol. 1992;167:227-232.
[6.] Obata K, Morland SJ, Watson RH, et al. Frequent PTEN/MMAC mutations in endometrioid but not serous or mucinous epithelial ovarian tumors. Cancer Res. 1998;58:2095-2097.
[7.] Tashiro H, Blazes MS, Wu R, et al. Mutations in PTEN are frequent in endometrial carcinoma but rare in other common gynecological malignancies. Cancer Res. 1997;57:3935-3940.
[8.] Duggan BD, Felix JC, Muderspach LI, Tourgeman D, Zheng J, Shibata D. Microsatellite instability in sporadic endometrial carcinoma. J Natl Cancer Inst. 1994;86:1216-1221.
[9.] Duggan BD, Felix JC, Muderspach LI, Tsao JL, Shibata DK. Early mutational activation of the c-Ki-ras oncogene in endometrial carcinoma. Cancer Res. 1994; 54:1604-1607.
[10.] Moutsatsou P, Sekeris CE. Estrogen and progesterone receptors in the endometrium. Ann N Y Acad Sci. 1997;816:99-115.
[11.] Preston-Martin S, Pike MC, Ross RK, Jones PA, Henderson BE. Increased cell division as a cause of human cancer. Cancer Res. 1990;50:7415-7421.
[12.] Mobley JA, Bhat AS, Brueggemeier RW. Measurement of oxidative DNA damage by catechol estrogens and analogues in vitro. Chem Res Toxicol. 1999; 12:270-277.
[13.] Lev S, Blechman JM, Givol D, Yarden Y. Steel factor and c-kit protooncogene: genetic lessons in signal transduction. Crit Rev Oncog. 1994;5:141-168.
[14.] McCulloch EA, Minden MD. The cell surface receptor encoded by the proto-oncogene KIT and its ligand. Cancer Treat Res. 1993;64:45-77.
[15.] Vliagoftis H, Worobec AS, Metcalfe DD. The protooncogene c-kit and c-kit ligand in human disease. J Allergy Clin Immunol. 1997;100:435-440.
[16.] Laramie A, Drobnjak M, Gerald W, Saad A, Cote R, Cordau-Cardo C. Expression of c-kit and kit ligand proteins in normal human tissues. J Histochem Cytochem. 1994;42:1417-1425.
[17.] Tsuura Y, Hiraki H, Watanabe K, et al. Preferential localization of c-kit product in tissue mast cells, basal cells of skin, epithelial cells of breast, small cell lung carcinoma and seminoma/dysgerminoma in human: immunohistochemical study on formalin-fixed, paraffin-embedded tissues. Virchows Arch. 1994; 424:135-141.
[18.] Arber DA, Tamayo R, Weiss LM. Paraffin section detection of the c-kit gene product (CD117) in human tissues: value in the diagnosis of mast cell disorders. Hum Pathol. 1998;29:498-504.
[19.] Boissan M, Feger F, Guillosson JJ, Arock M. c-Kit and c-kit mutations in mastocytosis and other hematological diseases. J Leukoc Biol. 2000;67:135-148.
[20.] Sakurai S, Fukasawa T, Chong JM, Tanaka A, Fukayama M. c-kit gene abnormalities in gastrointestinal stromal tumors (tumors of interstitial cells of Cajal. Jpn J Cancer Res. 1999;90:1321-1328.
[21.] Lux ML, Rubin BP, Biase TL, et al. KIT extracellular and kinase domain mutations in gastrointestinal stromal tumors. Am J Pathol. 2000;156:791-795.
[22.] Turner AM, Zsebo KM, Martin F, Jacobsen FW, Bennett LG, Broudy VC. Nonhematopoietic tumor cell lines express stem cell factor and display c-kit receptors. Blood. 1992;80:374-381.
[23.] Matsuda R, Takahashi T, Nakamura S, et al. Expression of the c-kit protein in human solid tumors and in corresponding fetal and adult normal tissues. Am J Pathol. 1993;142:339-346.
[24.] Natali PG, Nicotra MR, Sures I, Santaro E, Bigotti A, Ullrich A. Expression of c-kit receptor in normal and transformed human nonlymphoid tissues. Cancer Res. 1992;52:6139-6143.
[25.] Kauma S, Huff T, Krystal GW, Ryan J, Takacs P, Turner T. The expression of stem cell factor and its receptor, c-kit in human endometrium and placental tissues during pregnancy. J Clin Endocr Metab. 1996;81:1261-1266.
[26.] Inoue M, Kyo S, Fujita M, Enomoto T, Kondoh G. Coexpression of the c-kit receptor and the stem cell factor in gynecological tumors. Cancer Res. 1994; 54:3049-3053.
[27.] Brown RW, Chirala R. Utility of microwave-citrate antigen retrieval in diagnostic immunohistochemistry. Mod Pathol. 1995;8:515-520.
[28.] Horie K, Fujita J, Takakura K, et al. The expression of c-kit protein in human adult and fetal tissues. Hum Reprod. 1993;8:1955-1962.
[29.] Domson K, Kornstein MJ, Burks RT. c-kit proto-oncogene expression in benign endometrium and endometrial carcinoma [abstract]. Mod Pathol. 1996; 9:89A.
[30.] Manegold C, Jablonowski H, Armbrecht C, Strohmeyer G, Pietscht T. Serum levels of stem cell factor are increased in asymptomatic human immunodeficiency virus-infected patients and are associated with prolonged survival. Blood. 1995;86:243-249.
[31.] Hassan S, Kinoshita Y, Kawanami C, et al. Expression of protooncogene c-kit and its ligand stem cell factor (SCF) in gastric carcinoma cell lines. Dig Dis Sci. 1998;43:8-14.
[32.] McCulloch EA, Minden MD. The cell surface receptor encoded by the proto-oncogene KIT and its ligand. Cancer Treat Res. 1993;64:45-77.
[33.] Toyota M, Hinoda Y, Itoh F, Takoka A, Imai K, Yachi A. Complementary DNA cloning and characterization of truncated form of c-kit in human colon carcinoma cells. Cancer Res. 1994;54:272-275.
[34.] Akin C, Kirshenbaum AS, Semere T, Worobec AS, Scott LM, Metcalf DD. Analysis of the surface expression of c-kit and occurrence of the c-kit Asp816Val activating mutation in T cells, B cells, and myelomonocytic cells in patients with mastocytosis. Exp Hematol. 2000;28:140-147.
Accepted for publication August 8, 2000.
From the Department of Pathology, Medical College of Virginia at Virginia Commonwealth University, Richmond, Va (Drs Elmore and Kornstein and Mr Moore); the Department of Pathology, winchester of Greenville, PA, Greenville, SC (Dr Burks).
Presented in part at the US and Canadian Academy of Pathology annual meetings, Washington, DC, March 23-29, 1996, and Orlando, Fla, March 1-7, 1997.
Reprints: Lynne W. Elmore, PhD, Deaprtment of Pathology, Medical College of Virginia at Virginia Commonwealth Universit, Box 980662, Richmond, VA 23298.