Ординатура / Офтальмология / Английские материалы / Ocular Oncology_Albert, Polans_2003

II.TRANSGENIC MICE

Transgenic mice have recently been described as potentially useful models of human ocular melanoma. These models are based on using the promoter region of the tyrosinase gene to target expression of oncogenes in pigment-producing cells. The tyrosinase promoter was chosen because it drives the expression of downstream genes in melanocytes [3]. Pigmented intraocular tumors that arise in transgenic mice strains represent a spectrum of proliferations of uveal melanocytes and retinal pigment epithelium (RPE) [4–9]. Tanaka and coworkers introduced the murine tyrosinase minigene mg-Tyrs-J into fertilized eggs of albino Balb/c mice and found increased pigment production in the choroid [3]. Bradl and associates reported the regular occurrence of intraocular melanomas in mice expressing a transgene constructed by fusion of the simian virus 40 (SV40) early region, including the coding sequences of the transforming large tumor and small tumor antigens, with a mouse tyrosinase promoter [5]. These transgenic mice developed primary cutaneous melanomas and spontaneous bilateral pigmented tumors of the posterior segment of the eye. A majority of these appeared to originate in the RPE, some were distinctly in the choroid, and some were at the RPE-choroid interface. It is unclear using immunostaining and histologic analysis if these tumors represent a melanoma or a retinal pigment epithelium adenocarcinoma. Distant metastases were reported in lymph nodes, lung, bone, muscle, brain, salivary gland, thymus and subcutaneous tissue, but not in the liver [5].

Similarly, another intraocular tumor from FVB/N mice, which have the SV40 oncogene, has features that suggest an RPE tumor [4]. This tumor, if inoculated into the AC of FVB/N and athymic nude Balb/c mice, does metastasize to the liver [11,12].

Syed and associates characterized Tyr TagA and Tyr TagB, two similar lines of transgenic mice that express the SV40 T antigen, under the control of the tyrosinase promoter. These mice develop spontaneous, bilateral, pigmented intraocular tumors in the absence of primary cutaneous tumors. These tumors were found to arise from the RPE and were composed primarily of epithelioid cells with a variable spindle cell component. Metastases were reported in the subcutaneous tissue, lungs and brain [8].

Another transgenic mouse model was produced by a tyrosinase promoter that targeted expression of the mutated human T24 Ha-ras oncogene [7]. Alterations or activation of the ras oncogene have been studied regarding cutaneous neoplasms; in studies of human cutaneous melanoma, activated ras genes were found in 15–25% of tumors [13]. The T24 Ha-ras gene itself increases proliferation but does not transform primary cultures of melanocytes or rat embryo fibroblasts in tissue culture [14]. Furthermore, in vivo expression of T24 Ha-ras cause functional activation of melanocytes [7]. The resulting transgenic mice exhibited abnormal behavior and morphology, including abnormal movements soon after birth and deafness. Several abnormalities were observed in sections of the eyes and the eyelids. These tissues contained enlarged cells with intracytoplasmic melanin. The uveal tissue was thickened and heavily pigmented [7]. Kramer and associates studied 8 transgenic TPras mice with bilateral pigmented combined uveal melanocytic/RPE proliferations and found benign cytological characteristics in six of the mice (Figs. 1 and 2). The two remaining mice had cytologically malignant, bilateral melanocytic proliferations

Figure 1 TPras transgenic mouse eye with diffuse, heavily pigmented uveal proliferation. (H&E, 610.)

of the uveal tract, which was morphologically consistent with melanoma (Figs. 3 and 4) [15]. It is difficult to ascertain the reason for the alterations in the primary pigmented cells in these various transgenic models. The expression of the SV40 gene and the Ha-ras gene might occur at times during embryological development, not consistent with the production of pigment and proliferation of uveal melanocytes and retinal pigment epithelium. It is possible that expression of different genes at different embryological stages allows for the partially selective proliferation of melanocytes or retinal pigment epithelium [15].

III.SPONTANEOUSLY OCCURRING UVEAL TUMORS IN ANIMALS

A.Ocular Melanoma in Dogs

Tumors of melanocytic origin are the most common form of primary ocular neoplasia in dogs [16,17]. The first melanotic sarcoma (melanoma) of the dog choroid was described in 1919 by Petit [18]. Since then, several cases of uveal melanoma have been described in different breeds of dog [19]. The iris and ciliary body are the most common sites of origin for both benign and malignant ocular melanocytic proliferations in dogs [20–22]. There are only a few reported primary melanocytic tumors of the choroid [16,19,23]. The Callender classification and its modification [24], including the veterinary adaptation [25], are inappropriate for canine tumors, since most canine primary intraocular melanocytic neoplasms are histologically benign and contain a large proportion of plump cells that are not found in human uveal melanoma [17]. There is some indication that, as in humans, tumors with epithelioid-like cells are more locally aggressive than spindle cell tumors

Figure 4 The cytological features of the anterior uveal proliferation shown in Fig. 3 depicting cells with malignant cytological features, including epithelioid-like cells (arrowhead) and spindle cells (arrow). (H&E, 6160.)

B.Ocular Melanoma in Cats

Melanoma is the most frequently reported primary ocular neoplasm in cats [26]. Feline ocular melanoma usually arises on the anterior surface of the iris, and there is usually a diffuse darkening of the entire iris (Fig. 5) [21,27]. The neoplastic melanocytes in diffuse iris melanoma of the cat tend to be large, rounded cells with abundant cytoplasm and centrally positioned, round nuclei. Bizarre, enlarged nuclear forms are commonly encountered in these tumors. A smaller percentage of tumors have localized areas of pigmented spindle cells, and a few are predominantly composed of spindle cells. There is no evidence to suggest that the morphology of the tumor cells is related to prognosis, and longterm follow-up is limited. Intraocular melanomas in cats have been suggested to have a greater malignant potential than those in dogs [21]. There have been reported cases of metastatic disease in cats following enucleation for diffuse iris melanoma [21,28]. Feline ocular melanomas metastasize to the abdominal viscera and particularly the liver [21,27]. The lung is the second most common site of visceral metastasis [21].

C.Pigmented Ocular Tumors in Other Animals

Primary pigmented intraocular tumors—which may represent adenoma, adenocar- cinoma—or uveal melanoma, have also been reported in the horse, sheep, fish, and

Figure 5 Feline ocular melanoma arising on the anterior iris surface.

chicken [19]. Spontaneously occurring cutaneous melanomas in animals, which are utilized in animal models of ocular melanoma, are described further on.

IV. CHEMICAL AND RADIATION INDUCED UVEAL

PROLIFERATIONS

Uveal proliferations have been induced in laboratory animals with various chemical agents. Bensen and Hill reported intraocular tumor development in a rat after intraperitoneal injections of ethione and oral administration of N-2-fluorenylaceta- mide [29]. The tumor was found after 339 days in one of 25 study rats. The tumor arose in and extended around the iris, contained no visible pigment, did not metastasize, and was found to closely resemble a human spindle A type melanocytic proliferation [29]. Patz and collaborators reported experimental production of ocular tumors in mice [30]. Sutures impregnated with methylcholantrene and cholesterol were drawn through and fixed within the anterior chamber (AC) or posterior segment. The resulting iris tumors were not found to be identical to any specific cell type of human ocular melanoma, although they did resemble malignant melanoma of the iris. Tumors in the posterior segment were markedly anaplastic and undifferentiated. The possibility that these tumors originated from the retina or retinal pigment epithelium could not be excluded. No metastases were detected from these tumors [30]. Pigmented intraocular tumors have been produced with radium in dogs [31]. Beagles intravenously injected with 226Ra showed the presence of this material in melanin granules. The lesions induced by the intraocular retention of

radium were dose-dependent and characterized by loss of pigment at high dose levels and melanosis at lower levels. Induction of presumed intraocular melanoma occurred in both eyes in approximately 20% of the affected dogs and arose in the ciliary body. The cell of origin of the radiation-induced tumors was uncertain, although several features supported an origin from the ciliary body pigment epithelium. The tumors were microscopically characterized as consisting of clusters of large, round, densely pigmented cells separated by spindle-shaped cells. The incidence of mitotic figures was low and the tumor grew slowly. Widespread metastases were observed in two cases [31]. Secondary glaucoma occurred in a high percentage of the dogs with melanoma.

Albert and coworkers reported intraocular tumors in rats 6 to 9 months following injections of nickel subsulfide (Ni3S3), a potent carcinogen, into the vitreous cavity. Histologically, these tumors were composed of spindle and epithelioid cells; electron microscopy revealed premelanosomes in tumor cells [32].

An animal model of conjunctival primary acquired melanosis has been induced by the chronic topical application of the chemical carcinogen 7,12-dimethylbenz[a]- anthracene (DMBA) [33]. Chronic topical application of DMBA directly to the rabbit’s sclera is capable of inducing hyperplastic melanocytic lesions of the choroid [34]. Pe’er and coworkers induced primary uveal melanocytic lesions in Dutch (pigmented) rabbits using a two-stage carcinogenesis protocol involving initiation with four weekly topical applications of DMBA in acetone, followed by 12 weekly topical applications of croton oil in acetone. Exposure to DMBA followed by promotion with croton oil was effective in inducing clinically detectable fundus lesions. Lesions that showed unrestricted growth or metastasis were not detected. Clinical regression of the fundus lesions was noted after promotion was discontinued [35].

V.VIRAL-INDUCED UVEAL PROLIFERATIONS

A.Feline Uveal Melanoma Induced by Gardner Strain of Feline Sarcoma Virus

McCullough and coworkers determined that feline sarcoma virus can induce malignant melanoma associated with fibrosarcoma in nonocular tissues of the cat [36]. Additionally, feline leukemia virus injected into the eyes of newborn kittens or into fetal kittens resulted in severe developmental abnormalities and the production of a retinal tumor resembling retinoblastoma [37]. Albert and coworkers injected the Gardner strain of feline sarcoma virus into the inferior iris root of kittens [38,39]. Tumors were clinically detectable in approximately 80% of the injected eyes, were typically present by 40 days after injection, and were always recognizable by 60 days [38,40]. Two types of lesion were observed: (1) discrete areas of increased pigmentation of the iris or a mass enlarging within the iris and ciliary body and

(2) diffuse tumors, clinically characterized by thickening of the iris. Both the discrete and diffuse types continued to enlarge and eventually filled the anterior chamber and, in some cases, extended beyond the sclera [38]. Hypertrophy and hyperplasia of uveal melanocytes were histologically noted as early as 14 days after injection. In progressively enlarging tumors, pigmented epithelioid and pigmented or nonpig-

mented spindle-shaped cells invaded the iris stroma. In the large tumors, epithelioid cells were common and often predominated. The ultrastructural appearance of the melanoma cells was similar to that described for human ocular melanoma. A significant difference from human melanoma was the constant finding of virus particles budding from the cell membrane [38].

There are several advantages of this model of virally induced uveal proliferations. It uses a virus that is among the most thoroughly studied of the tumor viruses [41,42]. Transformation of normal melanocytes into tumor cells occurs in this model and may be studied. Disadvantages include the high mortality among infected animals as a result of feline infectious peritonitis. The presence of easily identifiable virus within most tumor cells is in striking contrast to human uveal melanoma. Injected animals develop disseminated viral infections, and secondary tumors produced are second primary tumors induced by virus shed from the primary intraocular melanoma, not metastases [43]. Most of the second primary tumors, ultrastructurally and in tissue culture, resemble fibrosarcoma [43]. Additionally, caution must be used in working with oncogenic viruses.

B.Uveal Melanoma Induced by SV40

In 1962, the simian virus 40 (SV40), a papovavirus found as a contaminant in rhesus monkey kidney tissue culture, was noted to induce neoplastic transformation in hamster kidney in vitro [44]. In 1968, Albert and coworkers introduced SV40 to hamster retina, choroid, and iris tissue grown in vitro [45]. These transformed tissue culture cells were subcutaneously injected into hamsters, resulting in firm, white neoplasms. Spindle and epithelioid cell types were observed [45]. SV40 tumors derived from retina, iris, and choroid tissue were similar in histological appearance, which was not unexpected, since an oncogenic virus may produce tumors with similar morphology in many different tissues [46]. The tumor cells observed by Albert and coworkers gradually lost their pigmentation and lacked ultrastructural evidence of premelanosomes.

VI. INTRAOCULAR INOCULATION OF TISSUE CULTURE MELANOMA CELLS IN ANIMAL EYES

A.Ocular Immune Privilege

The value of the anterior chamber (AC) of the animal eye as a privileged environment for heterologous tumor implantation was studied by Greene in 1949 [47–49]. He succeeded in transplantating rabbit tumors and several human tumors into the AC of animals of various species, including rabbit, mouse, rat, hamster, guinea pig, pig, sheep, and goat. It has subsequently been found that, after preliminary transfer of a new species of tumor cells into the AC, the tumor cells grown in the AC may be successfully transferred to and grown in different body regions of naive animals [48]. Investigators have found that introduction of tumor cells [50] or viruses [51] into the AC induces antigen-specific suppressor T cells that suppress cellular immunity to that antigen [52].

The immune privilege of the AC has been termed anterior chamber-associated immune deviation (ACAID) and has been confirmed in many studies [52]. In ACAID, T-suppressor cells cause suppression of delayed-type hypersensitivity (DTH), whereas humoral immunity is preserved. This explains long-term survival of the allograft in the AC, although destruction of tumor in the AC after a prior subcutaneous graft indicates that some antigenicity exists. It is possible that a humoral response plays a role in this form of AC tumor destruction [53]. The posterior segment of the eye, especially the subretinal space and vitreous, contains many immunomodulatory factors that maintain immune privilege [54]. The ocular milieu may thus provide immunological privilege for uveal melanoma [55].

B.Cell Lines

1.Greene Melanoma

In 1958, Greene isolated a spontaneous cutaneous hamster melanoma [56]. Homologous subcutaneous transfer of this tumor tissue resulted in a 100% growth rate. All the transplanted tumors grew progressively, and all the recipient animals died within 5 months of diffuse metastases.

The histological features of the tumor had changed after the tumor had been maintained by serial transfer through six generations over a 2-year period. A variation in pigment content was noted in one animal after the third passage; transfer of this tumor resulted in growth varying in color from the original black to pale white. The white tumors differed from the parental cell line with respect to pigment content and biological behavior [56]. The most distinctive biological changes associated with the amelanotic transformation consisted of an enhancement in heterologous transplantability and the attainment of the ability to metastasize [57].

In 1966, Greene and Harvey described another method of producing amelanotic melanomas in the hamster [57]. This new method was derived from the observation that intra-aortic inoculation of cells from melanotic melanoma produced amelanotic metastases, whereas intravenous inoculation of melanotic melanoma or vascular invasion from subcutaneous transplants resulted in pigmented metastases, similar to the primary tumor. Ultrastructurally, the cells in the amelanotic metastases were primitive and could not be differentiated from other anaplastic tumors [57]. The pigmented parental melanoma is rarely transplantable in foreign species and does not grow if placed into the AC of the rabbit eye. In contrast, amelanotic metastases readily grow in foreign species. The amelanotic variant fills the AC of the rabbit eye in less than 2 weeks, and large, widespread metastases involving the skin, contralateral eye, brain, pituitary, mammary tissue, skeletal muscle, liver, kidney, spleen, ovaries, adrenals, gastric and intestinal mucosa, diaphragm, lungs, heart, and thymus are observed within 1 month [57]. With continued passages of tissue culture cells, tumor growth in the AC and metastases are enhanced.

An advantage of using the hamster Greene melanoma model is that light and electron microscopic features of the amelanotic Greene melanoma are similar to human uveal melanoma, especially the epithelioid cell type [58]. Although Greene melanoma is a popular model for ophthalmic researchers, it has shortcomings. At

least one retrovirus associated with Greene melanoma is apparently able to transform adjacent ocular tissues—for instance, the retinal pigment epithelium [59]. This retrovirus may cause distant viral infection and neoplasia, which may be confused with metastases [59], similar to what occurs in the feline leukemia-sarcoma virus model. Another disadvantage of this model is that the biological behavior of Greene melanoma differs from human uveal melanoma. Even without therapy, Greene melanoma exhibits necrosis and hemorrhage 8–10 days after inoculation into the AC of the rabbit eye [53]. These changes could interfere with interpretation of the results of experimental therapies.

2.B16 Melanoma

The B16 melanoma arose spontaneously in 1954 in skin at the base of the ear of a C57BL mouse at the Jackson Laboratory [60]. Since its origin, it has been carried continuously in mice and used in many investigations. Pigmented cell strains were established in vitro from this tumor in the early 1960s [61]. Fidler developed a system of animal transplantation and tissue culture to select B16 tumor lines that possess enhanced metastatic properties [62]. The original cell line was designated as line number 26 (Fidler’s melanoma clone 26), and its daughter lines designated as lines 27, 28, 29, and 30. Later, Fidler renamed them to be B16 line F1 for the first passage line and then numbered them consecutively (i.e., lines F2, F3, etc.) [63]. Subcultures of B16 cells with different metastatic rates, such as the B16-F10, B16-F10 Queens [62–64], B16-LS8, and B16-LS9 [65,66] cell lines were developed after serial passages and isolation of metastatic tumors.

The B16 melanoma cell line has been used extensively for investigating the metastatic process and the immunological parameters that influence metastatic tumor development in mice [67]. The B16 melanoma model, although not ideal, offers several important advantages over other animal models. The origin and genetic background of the tumor is well documented [67]. Intraocular B16 melanoma metastasizes spontaneously in a predictable time frame [64,68]. B16 melanoma has been used extensively in a wide variety of studies of the biology and immunology of tumor metastasis and is among the most commonly used tumors in metastasis research [67].

3.Human Uveal Melanoma

In 1969, the discovery that the mutant athymic nude mouse was unable to reject a heterotransplant of human tumor tissue (human adenocarcinoma) [69] opened a new area for experimental study of human tumors, including the analysis of metastatic properties [70]. In contrast to human cutaneous melanoma, human uveal melanoma cells do not readily grow in tissue culture [71]. Culturing of uveal melanoma was first attempted in 1929 by Kirby [72]. Kirby’s studies as well as subsequent observations of in vitro growth of uveal melanoma consisted of relatively short-term experiments [73–76]. It was demonstrated that uveal melanoma cells can survive in culture for periods of several months or longer if transfer of the cells is not attempted. Although, until recently, no continuous cell lines of uveal melanoma have been established, some interesting features of this tumor were observed, including the spontaneous transformation of spindle cells to epithelioid cells and vice versa [74]. The first continuous cell lines of human uveal melanoma were reported by Albert