Seminars in Surgical Oncology 8:353-357 (1992)

Cellular and Molecular Biology of Melanoma

GIUSEPPE DELLA PORTA, MD

From the Division of Experimental Oncology A, lstituto Nazionale Turnori, Milan, Italy

Melanoma cells have surface markers that are expressed differently than in normal melanocytes and nevus cells. Monoclonal antibodies may define a phenotypic map of the various melanocytic lesions and can be used in immunohistopathology and immunoscintigraphy. Monoclonal antibodies directed against melanoma-associated glyco- proteins and glycolipids are being tested for therapy. Rearrangements or deletions on chromosome 1, 6, and 7 are the most frequently ob- served cytogenetic abnormalities. Molecular studies have not given a clear picture. A subset of HRAS alleles has been reported to be as- sociated with melanoma. NRAS activation by point mutation has been found in one fourth of the cases. Allele losses at different loci have been reported. Genetic linkage studies have given conflicting results on the presence of a gene for the melanoma-dysplastic nevus syndrome on the short arm of chromosome 1. @ 1992 Wiley-Liss, Inc.

KEY WORDS: surface markers, monoclonal antibodies, immunohistopathology, immunoscintigraphy

INTRODUCTION Melanocytes, the proliferating component of ma-

lignant melanoma of the skin and of its precursor lesions, originate in the embryonic neural crest. Melanoblasts migrate to the epidermis where they differentiate and proliferate in the basal layer. Melanocytes are dendritic cells of various size and shape according to location and stage of activity, have a highly developed Golgi apparatus and rough endoplasmic reticulum, and do not have desmosomes and tonofilaments. Melanogenesis occurs in complex membranous organelles, the melanosomes, where ty- rosinase converts tyrosine to DOPA that is then oxi- dized and polymerized to melanin. Melanocytes con- tain a variable number of melanosomes at various stages of formation. Early melanosomes are located in the perinuclear region from where fully formed melanosomes move along intermediate filaments and microtubules toward the tip of the dendrites, to be transferred to the nearby keratinocytes. Under ul- traviolet irradiation, there is an increased number of active melanocytes and more melanosomes are formed and transferred. The transfer may occur by

In spite of the many studies on culture of normal and neoplastic human melanocytes, the specific regu- lation of their differentiation is still poorly under- stood. In almost all cases, malignant melanoma cells retain the ability to form melanin, although variations in structure and antigenicity of melanosomes in nor- mal, dysplastic, and malignant melanocytes have been documented [l]. Drugs capable of interfering with melanin metabolism were shown to have a selective toxicity against melanoma cells [2].

Basic fibroblast growth factor (bFGF), together with CAMP stimulators, has been shown to be a natu- ral mitogen for melanocytes. Its role in neoplastic transformation of melanocytes is unclear, although the growth of metastatic melanoma cells seems to be independent from this factor suggesting an autocrine mechanism [3]. Interestingly, in vitro studies showed that the production of bFGF by keratinocytes may regulate the growth of normal melanocytes [4].

MELANOMA CELL MARKERS A detailed analysis using polyclonal and mono-

clonal antibodies has recognized numerous cell-sur- direct passage to keratinocytes, extracellular secre- tion, or removal of portions Of dendrites which are engulfed by the keratinocytes.

Address reprint requests to G. Della Porta, M.D., Division of Experimental Oncology A, lstituto Nazionale Turnori, Via Venezian I , 20133 Milan, Italy.

0 1992 Wiley-Liss, Inc.

354 Della Porta

face markers on melanoma cells. The most exten- sively studied is the p97 antigen, a monomeric cell- surface sialoglycoprotein of approximately 97 kD, that has been defined by several monoclonal anti- bodies and has been found in most human melano- mas, in different carcinomas, in certain fetal tissues, and only in trace amounts in normal adult tissues [5]. Structural analysis has shown that p97 is related to transferrin and may have a role in iron metabolism [6]. A recombinant vaccinia virus expressing p97 has been constructed and has been proposed for a clini- cal trial [7].

Among other glycoproteins, a high-molecular- weight proteoglycan of approximately 400 kD com- plexed with a 250-kD glycoprotein [8] has been found highly expressed in more than 90% of human melano- mas. Several monoclonal antibodies against different epitopes of the proteoglycan have been described: their potential application in therapy of melanomas has been demonstrated [9].

Numerous studies showed that several gangliosides, sialic acid-containing glycolipids, are expressed in cells of neuroectodermal origin and are present in high con- centration in melanomas. Gangliosides may function as modulators of growth factors receptors and have been reported to be immunogenic in cancer patients [lo]. Monoclonal antibodies directed to epitopes on ganglioside moieties have been found to induce regres- sion of tumor growth in melanoma patients [ 1 1.121. In individual tumors, however, the percentage of mela- noma cells expressing a particular ganglioside is varia- ble [13].

HLA class TI antigens have been shown to be ex- pressed on the surface of neoplastic melanocytes of most human melanomas, whereas normal melano- cytes do not express these antigens [14].

It should be noted that, in general, the expression of melanoma-associated antigens has been found hetero- geneous in tumors from different patients although antigenic heterogeneity was not significant in different tumor specimens from the same patient [15].

Using a large series of monoclonal antibodies, anti- genic markers have been analysed in situ on normal melanocytes, the various types of nevi, primary melanomas, and melanoma metastases allowing the construction of an antigenic phenotype map of the melanocytic lesions [16- 191, in that some markers are common to all normal and proliferating melanocytes, whereas other markers are expressed differently in a proportion of the different lesions. Therefore, these are only quantitative, not absolute differences, since most melanoma antigens are also found in benign melanocytic lesions. However, some are found only on nevus cells but not on melanomas.

CHROMOSOMAL ABNORMALITIES Cytogenetic studies of human melanomas, as of

other solid tumors, were formerly conducted only on a few established cell lines. In recent years, they have been expanded following improved methods that yield good quality metaphases either directly from tumor specimens or after a few days of culturing.

Cytogenetic analyses of more than 100 cases of pri- mary and metastatic melanomas have reported a large variety of nonrandom karyotypic changes involving structural and numerical abnormalities of several chromosomes, most notably chromosomes I , 6 ,7 , and

The most frequent abnormality, observed in about 80% of the cases, consists of rearrangements of chro- mosome 6 with deletions of the long arm as a conse- quence of simple deletions mostly at 6q21-23, non- reciprocal translocations with breakpoints at 6q 1 1 - 13, and isochromosomes of the short arm, in- dicating that a loss of genomic material may condition an early event in the pathogenesis of the disease through the inactivation of a tumor suppressor gene [24]. Moreover, the introduction of a normal chromo- some 6 into melanoma cell lines resulted in the sup- pression of the malignant phenotype [25].

Also chromosome 1 appears to be frequently altered in malignant melanomas with rearrangements in the distal region of the short arm [26]. Changes of chro- mosome 7, including polysomy of the entire chromo- some or i(7q), have been seen in around 50% of analyzed tumors [21]. Early melanocytic lesions have not been extensively analyzed. However, abnormali- ties of the long arm of chromosome 10 have been reported [27].

10 [20-231.

MOLECULAR STUDIES The oncogenes of the ras family have been exten-

sively analyzed in melanomas, particularly after the introduction of the polymerase chain reaction meth- odology using besides cell lines also fresh and paraf- fin-embedded surgical specimens. Mutations almost exclusively of NRAS were found in 5-24% of the cases of primary and metastatic melanomas [28,29]. However, no rus mutations were found in surgical specimens of 30 dysplastic nevi and 15 normal nevi [28]. It appears that ras activation by point mutation does not have a general role in the mutation and pro- gression of melanomas, although it may have a role in the pathogenesis of a subset of melanomas with a less differentiated phenotype [28] or related to sun expo- sure [29].

Further studies to verify whether an enhanced or inappropriate expression of ras genes could be a risk

Cellular and Molecular Biology of Melanoma 355

factor for the development of melanoma, have exam- ined a noncoding region at 3' of the polyadenylation signal of HRASl that consists of variable tandem re- peats (VTR). The VTR region seems to exert an en- hancer-like function, and displays restriction fragment length polymorphisms by changes in the number of its repeat units [30] or by additional restriction sites within the region [31]. The analysis of this polymor- phism in the DNA of peripheral blood lymphocytes of melanoma patients and healthy controls has provided significant evidence that the constitutional presence of certain alleles may be associated with the risk of devel- oping melanomas [32,33].

Following the many reports on chromosomal ab- normalities in melanomas, attempts have been made to identify and characterized the involved genes. So far, no clear evidence has been found of structural alterations of the MYB proto-oncogene that has been mapped to 6q22-23, a region frequently presenting deletions in melanomas [34].

Molecular studies were also conducted on the short arm of chromosome 1, where rearrangements were frequently observed; moreover, because a linkage analysis in families with hereditary melanoma and dysplastic nevi has given evidence that a pleiotropic gene may be located on the region lp36 near the Rhesus blood group locus [35]. This evidence was con- firmed by a recent multipoint linkage analysis of 26 polymorphic loci on l p in the DNA from normal lymphocytes or fibroblasts of 120 members of 6 North American families with hereditary melanoma and dys- plastic nevi [36]. However, a similar study on 6 Dutch families [37] and another study on North American families [38] gave results that seem to exclude the pres- ence of a gene for the melanoma-dysplastic nevus syndrome on the short arm of chromosome 1 .

Also the loss of constitutional heterozygosity at 24 loci on l p was analyzed on the DNA of primary and metastatic melanomas and of the autologous lym- phocytes.

Allelic losses were observed approximately in 50% of informative cases but only in metastatic melano- mas, examined either as direct biopsies or as cell lines, providing evidence that loss of heterozygosity at loci on distal l p is a late event in tumor progression [39]. Another similar study, however, found a random dis- tribution of allelic losses on several chromosomes sug- gesting a significant genetic instability of melanoma cells [40].

CONCLUSIONS The available evidence indicates that human mela-

noma cells are characterized by a complex of antigens, none of which is melanoma specific, although some of

them are melanoma associated, which may be respon- sible of the frequent spontaneous regressions and may be exploited for immunotherapeutic approaches. It should be noted that the precise biological function of most antigens, and even the biochemical nature of some of them, is still unknown. However, our knowl- edge is progressing rapidly, using also gene cloning procedures. A more precise characterization of the sequential acquisition or loss of expression of mole- cules related to the malignant phenotype will allow a better definition of the multiple stages of progression from the normal melanocyte to the metastatic mela- noma. The availability of a large number of mono- clonal antibodies to be used with immunohistochemi- cal techniques for tumor diagnosis will have an increasing role in early diagnosis and in the definition of prognostic markers.

No sufficient molecular data are available to try to construct a melanoma model. According to other models, tumorigenesis is a multistep process con- trolled by the interplay of several genomic events that cause the development of the early neoplastic pheno- type and the progression to the invasive and meta- static disease. The known genomic events include the activation of proto-oncogenes, through mutations or rearrangements, into dominant oncogenes and the inactivation of regulatory genes, that are deleted or mutated at both alleles and have therefore been classi- fied as recessive oncogenes and tumor-suppressor genes. The large number of dominant oncogenes which have been identified, probably reflects the great variety of functions that their normal counterparts have in normal physiology of the many different tis- sues and cell types and in the sequence of differentia- tion steps of each cell prototype. Also, the many dif- ferent carcinogenic factors, that may act preferably on a given sequence, could condition the differential acti- vation of oncogenes. Accordingly, it should be ex- pected that even a single tumor type may be character- ized by the alternative activation of several oncogenes. In melanomas, only the oncogenes of the ras family have been studied to a certain extent and NRAS has been found activated by point mutation in up to one- fourth of cases. NRAS activation seems to be related to tumor progression and it remains to be established whether it corresponds to a particular phenotype or etiology. HRASl has been found activated only in a few cases although certain alleles seem to act as a risk factor for melanoma development.

The most frequent chromosomal abnormalities oh- served in melanomas consist of rearrangements or de- letions on chromosomes 1, 6, and 7. In other tumors, following the model of retinoblastoma, the study of deletions has led to the identification of recessive

356 Della Porta

tumor suppressor genes. In melanomas, in spite of the fact that approximately 10% of the cases occur in predisposed families together with dysplastic nevi, a clear picture pointing to a suppressor gene has not emerged so far. Contrasting evidence has been pre- sented for a melanoma gene on the distal part of the short arm of chromosome 1 . Most of the observations are recent and limited, and further studied are needed, also to clarify the contrasting results.

It is interesting to note that a long and complex series of investigations conducted on the malignant melanomas that develop in Xiphophorus fish hybrids [41] has led to the identification and cloning of the dominant Tu oncogene that encodes for a transmem- brane growth factor receptor tyrosine kinase and is activated in the absence of a regulatory gene 1421.

1

2

3

4

5

6

7

8

9

10

11

12.

13

14.

REFERENCES Jimbow K, Miyake Y, Homma K, et al: Characterization of melanogenesis and morphogcnesis of melanosomes by physio- chemical properties of melanin and melanosomes in malignant melanoma. Cancer Res 44: 1 128 - 1 134, 1984. Wick MM: An experimental approach to chemotherapy of melanoma. J Invest Dermatol 74: 63-65. 1980. Halaban R, Know BS. Ghosh S. et al: bFGF as an autocrine growth factor for human melanomas. Oncogene Res 3: 177- 186, 1988. Halaban R, Langdon R, Birchall N. et al: Basic fibroblast growth factor from human keratinocytes is a natural mitogen for melanocytes. J Cell Biol 107: 161 1-1619, 1988. Woodbury R G, Brown J P. Loop S M. et al: Analysis of normal and neoplastic human tissues for the tumor-associated protein p97. Int J Cancer 27: 145-149, 1981. Brown JP. Hewick RM. Hellstrom I, et al: Human melanoma- associated antigen p97 is structurally and functionally related to transferrin. Nature 296: 171-173, 1982. Estin CD, Stevenson US, Plowman G D , et al: Recombinant vaccinia virus vaccine against the human melanoma antigen p97 for use in immunotherapy. Proc Natl Acad Sci USA 85: 52 56, 1988. Bumol TF, Reisfeld RA: Unique glycoprotein -proteogly- can complex defined by monoclonal antibody on human melanoma cells. Proc Natl Acad Sci USA 79: 1245-1249. 1982. OIdham RK, Foon KA, Morgan AC Jr. et al: Monoclonal antibody therapy of malignant melanoma: In vivo localization in cutaneous metastasis after intravenous administration. J Clin Oncol 2: 1235-1244, 1984. Tai T, Paulson JC, Cahan LD, Irie RF: Ganglioside GM2 as a human tumor antigen (OFA-1-1). Proc Natl Acad Sci USA 80: 5392-5396, 1983. Houghton A N. Mintzer D, Cordon-Cardo C. et al: Mouse monoclonal IgG3 antibody detecting GD3 ganglioside: A phase 1 trial in patients with malignant melanoma. Proc Nail Acad Sci USA 82: 1242, 1985. Irie RF. Morton DL: Regression of cutaneous metastatic mel- anoma by intralesional injection with human monoclonal anti- body to ganglioside GD,. Proc Natl Acad Sci USA 83: 8695- 8698, 1986. Hcrsey P, Jamal 0, Henderson C, et al: Expression of the gangliosides GM3, GD3 and GD2 in tissue sections of normal skin, naevi, primary and metastatic melanoma. Int J Cancer 41: 336-343. 1988. Taramelli D, Fossati G , Mazzocchi A, et al: Classes I and 11 HLA and melanoma-associated antigen expression and modu-

lation on melanoma cells isolated from primary and metastatic lesions. Cancer Res 46: 433-439, 1986.

15. Berd D, Herlyn M, Koprowski H, Mastrangelo MJ: Flow cytometric determination of the frequency and heterogeneity of expression of human melanoma-associated antigens. Cancer Res 49: 6840-6844, 1989.

16. Herlin M, Thurin J, Balaban G, et al: Characteristics of cul- tured human melanocytes isolated from different stages of tumor progression. Cancer Res 46: 5670-5676, 1985.

17. Holzmann B, Brocker EB, Lehmann JM, el al: Tumor progres- sion in human malignant melanoma: Five stages defined by their antigenic phenotypes. Int J Cancer 39: 466-471, 1987.

18. Houghton AN, Real FX, Davis LJ, et al: Phenotypic heterogc- neity of melanoma: Relation to the differentiation program of melanoma cells. J Exp Med 164: 812-829, 1989.

19. Jimbow K. Takahashi H, Akutsu Y, Meada K: Fine structural and immunohistochemical properties of dysplastic melano- cytic nevi: Comparison with malignant melanoma. J Invest Dermatol 92: 304-309, 1989.

20. Becher R. Gibas Z, Karakousis C, Sandberg AA: Nonrandom chromosome changes in malignant melanoma. Cancer Res 43: 5010 - 5016. 1983.

2 1. Balaban GB, Herlyn M, Clark WH Jr, Nowell PC: Karyotypic evolution in human malignant melanoma. Cancer Genet Cyto- genet 19: 113-122, 1986.

22. Limon J, Dal Cin P, Sait SNJ, et al: Chromosome changes in metastatic human melanoma. Cancer Genet Cytogenet 30:

23. Sozzi G. Miozzo M, Calderone C, et al: Chromosome abnor- malities and fragile sites in human melanoma. Cancer Genet Cytogenet 44: 61 -67, 1990.

24. Trent JM, Thompson FH, Meyskens FL Jr: Identification of a recurring translocation site involving chromosome 6 in human malignant melanoma. Cancer Res 49: 420-423, 1989

25. Trent JM. Stanbridge EJ, McBride HL, et al: Tumorigenicity in human melanoma cell lines controlled by introduction of human chromosome 6. Science 247: 568-571, 1990.

26. Balaban G , Herlyn M, Guerry D, et al: Cytogenetics of human malignant melanoma and premalignant lesions. Cancer Genet Cytogenet 1 1 : 429-439, 1984.

27. Parmiter AH, Balaban G, Clarck WH Jr, Nowell PC: Possible involvement of the chromosome region 10q24 +q26 in early stages of melanocytic neoplasia. Cancer Genet Cytogenet 30:

28. Albino AP, Nanus DM, Mentle IR, et al: Analysis of ras oncogenes in malignant melanoma and precursor lesions: Cor- relation of point mutations with differentiation phenotype. Oncogene 4: 1363- 1374, 1989.

29. van't Veer LJ, Burgering BMT, Versteeg R, et al: N-ras muta- tions in human cutaneous melanoma from sun-exposed body sites. Mol Cell Biol 9: 3114-31 16, 1989.

30. Capon DJ, Chen EY, Levinson AD, et al: Complete nucleotide sequences of the T24 human bladder carcinoma oncogene and its normal homologue. Nature 302: 33-37, 1983.

3 1. Pierotti MA, Radice P, Biunno I, et al: Detection of two TaqI polymorphisms in the VTR region of the human Ha-rasl onco- gene. Cytogenet Cell Genet 43: 174-180, 1986.

32. Radice P. Pierotti MA, Borrello MG, et al: HRASI proto- oncogene polymorphisms in human malignant melanoma: Taql defined alleles significantly associated with the disease. Oncogene 2: 91 -95, 1987.

33. Hayward NK. Nancarrow DJ, Parsons PG, ct al: TaqI poly- morphism within the c-Ha-ras-l VTR is associated with mela- noma. Hum Genet 83: 395-396, 1989.

34. Mecse E. Meltzer PS, Witkowski CM, Trent JM: Molecular mapping of the oncogene M Y B and rcarrangcmcnts in malig- nant melanoma. Gene Chromosomc Cancer 1: 88-94, 1989.

35. Green MH, Goldin LR, Clarck WH Jr, et al: Familial cutane- ous malignant melanoma: Autosomal dominant trait possibly linked to the Rh locus. Proc Natl Acad Sci USA 80: 6071 ~

607.5. 1983.

901 -21 1. 1988.

313-317. 1988.

Cellular and Molecular Biology of Melanoma 357

36. Bale SJ, Dracopoli NC, Tucker MA, et al: Mapping the gene for hereditary cutaneous malignant melanoma-dysplastic nevus to chromosome lp. N Engl J Med 320: 1367-1372, 1989.

37. van Haeringen A, Bergman W, Nelen MR, et al: Exclusion of the dysplastic nevus syndrome (DNS) locus from the short arm of chromosome 1 by linkage studies in Dutch families. Genom- ics 5: 61-64, 1989.

38. Cannon-Albright LA, Goldgar DE, Wright EC, et al: Evidence against the reported linkage of the cutaneous melanoma-dys- plastic nevus syndrome locus to chromosome lp36. Hum Genet 46: 912-918, 1990.

39. Dracopoli NC, Harnett P, Bale SJ, et al: Loss of alleles from

the distal short arm of chromosome 1 occurs late in melanoma tumor progression. Proc Natl Acad Sci USA 86: 4614-4618, 1989.

40. Nordenskjold M, Werelius B, Ringborg U: Allele losses in malignant melanoma reflect random events. Anticancer Res 10: 883-886, 1990.

41. Anders F, Scharti M, Barnekow A, Anders A: Xiphophorus as an in vivo model for studies on normal and defective control of oncogenes. Adv Cancer Res 42: 191 -275, 1984.

42. Wittbrodt J, Adam D, Malitschek B, et al: Novel puta- tive receptor tyrosine kinase encoded by the melanoma- inducing Tu locus in Xiphophorus. Nature 341: 415-421, 1989.