S100A6 Protein Expression is Different in Spitz Nevi and Melanomas Adriana Ribé, M.D., Ph.D., N. Scott McNutt, M.D. Dermatopathology Division, Department of Pathology, New York Presbyterian Hospital—Cornell University Weill Medical College, New York, New York and melanomas, without differences. In summary, The Spitz nevus is a benign melanocytic lesion that a simple immunohistochemical test for S100A6 pro- can be identified reliably in many cases by conven- tein differentiated between Spitz nevi, melanomas, tional histopathological criteria. However, there are and melanocytic nevi. This marker could be used subsets of Spitz nevi and of malignant melanoma when the distinction is very difficult or controver- that closely resemble each other and represent di- sial in routine studies, especially when there is a agnostic challenges. S100 proteins are of interest junctional component. Further molecular analyses because of their involvement in neoplastic pro- of the S100A6 protein and gene should be per- cesses and their genes are clustered in chromosome formed to study the underlying genetic bases for 1q21. Chromosome 1 contains mutations in several such differences. types of tumors, including melanomas. The expres- sion of different S100 proteins (A2, A6 and A8/A9 or KEY WORDS: Melanocyte, Melanoma, S100, A12) was examined in 42 Spitz nevi, 105 melano- S100A6, Spindle and epithelioid cell nevus, Spitz mas, and 73 melanocytic nevi to test the hypothesis nevus. that their expression differs among these entities Mod Pathol 2003;16(5):505–511 and may contribute to the distinction between these entities. The results showed an up-regulation of The epithelioid and spindle cell nevus was first S100A6 protein in Spitz nevi, melanomas, and mela- described in 1948 by Spitz and named juvenile mel- nocytic nevi but with a different percentage of pos- anoma (1). It is a benign melanocytic lesion found itivity and pattern of immunoreactivity. The differ- predominantly in children and adolescents and ences between these three entities were statistically that can be identified reliably in many cases by significant (P < .001). All 42 Spitz nevi (100%) previously well-described major and minor his- showed strong and diffuse S100A6 protein expres- topathological criteria (2–5). However, there are sion, both in junctional and in dermal components subsets of Spitz nevi and of malignant melanoma of the nevi. Thirty-three percent of melanomas ex- that closely resemble each other and represent di- pressed S100A6 (35/105). The expression was agnostic challenges. Additionally, some lesions are mainly weak (30/35) and patchy in the dermal com- only partially excised (shave biopsies), thus often ponent and was negative or minimal in the junc- preventing the assessment of the lateral edges and tional component. Fifty-six percent of different sub- deep components. The importance of distinguish- types of melanocytic nevi (41/73) expressed S100A6, ing between these two entities lies in their different almost all of them weakly (40/41) and in the dermal prognosis and treatment. Multiple studies using component. Normal intraepidermal melanocytes several approaches (immunohistochemistry, flow were negative. The melanocytic cells in these three cytometry and molecular analysis) have been per- entities did not express S100A2, S100A8/A9 or A12. formed to identify potential differences between However, an up-regulation of S100A2 and Spitz nevi and melanomas, but the results have S100A8/A9 or A12 proteins was observed in normal been controversial or difficult to apply to routine keratinocytes in the epidermis overlying Spitz nevi material (6–38). S100 protein is not a single protein; instead there are a group of S100 proteins with Copyright © 2003 by The United States and Canadian Academy of diverse functions. The S100 protein family, a sub- Pathology, Inc. 2ϩ VOL. 16, NO. 5, P. 505, 2003 Printed in the U.S.A. class of low molecular weight Ca -binding pro- Date of acceptance: January 24, 2003. teins, regulate a variety of cellular processes via Address reprint requests to: N. Scott McNutt, M.D., Dermatopathology (F- 309), Weill Medical Center of Cornell University, 1300 York Avenue, New interaction with different target proteins in a York, NY 10021; fax: 212-746-8570; e-mail: [email protected]. calcium-dependent manner. They are of interest DOI: 10.1097/01.MP.0000071128.67149.FD because of their involvement in neoplastic pro- 505 cesses and their genes are clustered in chromosome of positive lesional cells. Appropriate positive and 1q21. Chromosome 1 is frequently altered in several negative controls were also included. tumors, including melanomas (28, 39, 40). The The ␹2 test was used for comparison of groups. A most widely known commercially available anti- P value of Ͻ.05 was considered significant. body “anti-S100” mainly reacts with cells contain- ing S100B polypeptide chains. This antibody was used as an early marker for melanocytic neoplasms RESULTS and remains sensitive but not specific (41–43). Re- cently, monoclonal antibodies have become avail- All Spitz nevi cases (42/42) strongly expressed able that mainly react with S100aa and have iden- S100A6 protein in a diffuse pattern, regardless of tified several variants of the S100a chains (28). The the subtype (compound, dermal, or junctional; expression of different S100 proteins (A2, A6, and Figs. 1, 2). The melanocytic cells in all Spitz nevi A8/A9 or A12) was examined in Spitz nevi, melano- cases did not react with anti-S100A2 or MAC387 mas, and melanocytic nevi cases to test the hypoth- antibodies. However, the keratinocytes in the epi- esis that their expression differs among these enti- dermis overlying Spitz nevi reacted with anti- ties and may contribute to the distinction between S100A2 antibody in some cells in the basal layer and these entities. with the MAC387 antibody in some cells in the upper third of the epidermis. The HMB45 antibody was weakly positive on the junctional component, MATERIALS AND METHODS whereas the dermal component usually was nega- The cases were obtained from the files at the tive with only scattered positive cells. Dermatopathology Division of the New York Pres- S100A6 protein was expressed in 35 of 105 mela- byterian Hospital-Cornell University Weill Medical nomas (33%), 30 of them with weak positivity College, between 1997 and 2002, based on the avail- (86%). The distribution of positive cases in the mel- ability of adequate tissue material for study. Forty- anoma subtypes was as follows: 10 of 12 nodular two cases of Spitz nevi, 105 cases of melanomas, (83%), 3 of 7 metastatic (43%), 16 of 54 superficial and 73 cases of different types of melanocytic nevi spreading (30%), 2 of 7 lentigo maligna (28.5%), 3 of were selected. The diagnosis was made on multiple 21 in situ (14%), and 1 acral (Figs. 3, 4). Only 3 H&E routine-stained sections before this study. Af- nodular melanomas and 1 superficial spreading ter review of newly cut routine microscopic sec- and 1 acral melanoma were strongly positive. The tions, additional unstained sections were cut, expression in the positive melanoma cases was deparaffinized, pretreated with pepsin (0.25% w/v, patchy throughout the tumor and negative or min- pH 2.0) for 5 minutes at 45° C, and then reacted imal in the junctional component. Three desmo- with the monoclonal antibodies: anti-S100A2, anti- plastic melanomas were negative. In all melanoma S100A6, MAC387, and HMB45 (Table 1). A cases, the melanocytes did not stain with anti- streptavidin-alkaline phosphatase detection system S100A2 or MAC387 antibodies. However, S100A2 with a red chromogen (Vector Lab) was used. The protein was expressed in the keratinocytes located reaction time for the red chromogen was 20 min- in the basal or parabasal layer of the epidermis utes. Two investigators verified the results of the adjacent to melanoma cells and MAC387 antibody immunostains without references to the original was often observed in the full thickness of the over- diagnosis, although in some cases the diagnosis was lying epidermis. HMB45 was positive in some of the obvious. The staining was evaluated semiquantita- tively and read as positive (strong or weak) and negative. In the cases regarded as strongly positive, staining was observed in Ͼ50% of the lesional cells whereas the weakly, positive cells were between 1 and 50% positive. The negative cases showed Ͻ1% TABLE 1. Monoclonal Antibodies Used in this Study Antibody Clone Source Dilution Anti-S-100A2 (S100L) SH-L1 Sigma 1:1000 Anti-S-100A6 CACY-100 Sigma 1:1000 (calcyclin) Myeloid histiocyte MAC387 Dakopatts 1:100 antigen FIGURE 1. Spitz nevus. This photograph shows a strong and uniform HMB45, HMB45 DAKO 1:100 reactivity with the monoclonal antibody against S100A6 protein in the melanosome nevus cells. 506 Modern Pathology architectural disorder (88%), 6 of 12 combined nevi (50%), 3 of 6 blue nevi (50%), 5 of 13 compound nevi (38%), 4 of 12 intradermal nevi (33%), and 1 of 3 acral nevi (33%; Fig. 5). Only 1 nevus with archi- tectural disorder was strongly positive. The immu- noreactivity in the positive cases was restricted to the deep dermal component. Two pigmented spin- dle cell nevi were negative. Neither anti-S100A2 nor MAC387 antibodies reacted with the nevus cells. However, anti-S100A2 and MAC387 antibodies were observed in a few cells in the basal and upper layers of the epidermis, respectively. HMB45 was positive in the junctional component of the nevi FIGURE 2. Spitz nevus. As in Figure 1, this photograph shows a strong and uniform reactivity with the monoclonal antibody against and in scattered nevus cells in the dermis. Blue nevi S100A6 protein in the nevus cells. were strongly positive throughout. The differences observed in the number of posi- tive cells for anti-S100A6 antibody in Spitz nevi, melanomas, and melanocytic nevi cases were sta- tistically significant, either comparing the three en- tities as whole groups (P Ͻ .001) or comparing more similar histologic subgroups as junctional Spitz ne- vi–melanoma in situ (P Ͻ .001) and Spitz nevi– nodular melanoma (P Ͻ .001).