The Potential Role of Natural Tumor Promoters in Marine Turtle Fibropapillomatosis

—Fibropapillomatosis (FP) in green turtles Chelonia mydas is a debilitating, neoplastic disease that has reached worldwide epizootic levels. The etiology of FP is unknown but has been linked to oncogenic viruses. Toxic benthic dinoflagellates (Prorocentrum spp.) are not typically considered tumorigenic agents, yet they have a worldwide distribution and produce a tumor promoter, okadaic acid (OA). Prorocentrum spp. are epiphytic on macroalgae and seagrasses that are normal * Corresponding author: [email protected] 1 Center for Marine Science Research, University of North Carolina at Wilmington, 7205 Wrightsville Avenue, Wilmington, North Carolina 28403, USA. Received May 13, 1998; accepted January 16, 1999 200 LANDSBERG ET AL. components of green turtle diets. Here we show that green turtles in the Hawaiian Islands consume Prorocentrum and that high-risk FP areas are associated with areas where P. lima and P. concavum are both highly prevalent and abundant. The presence of presumptive OA in the tissues of Hawaiian green turtles further suggests exposure and a potential role for this tumor promoter in the etiology of FP. Marine turtle fibropapillomatosis (FP) is a neoplastic disease that has been documented from the Atlantic Ocean (Florida, Brazil), Indo-Pacific region (Australia, Sri Lanka, Sarawak, Malaya, Bonin Islands [Japan]), Pacific Ocean (Hawaiian Islands), and Caribbean Sea (Cayman Islands, Puerto Rico, Dominican Republic, Virgin Islands, Barbados, Antigua and Barbuda, Central America) (Balazs 1991; Ehrhart 1991; Herbst 1994; Williams et al. 1994). Fibropapillomas are commonly found in green turtles Chelonia mydas, although loggerhead turtles Caretta caretta and olive ridley turtles Lepidochelys olivacea are also known to be affected (Jacobson et al. 1991; Herbst 1994). The tumors are benign, but growths of fibropapillomas can adversely affect locomotion, vision, swallowing, and breathing. Visceral fibromas can disrupt normal organ function to the extent that death ensues (Herbst 1994). Fibropapillomatosis was first reported from a captured turtle in 1938, when 1.5% (3/200) of free-ranging green turtles in Key West, Florida, were affected (Smith and Coates 1938). In the last 20 years, there has been a dramatic increase in the prevalence of FP in green turtles in the Hawaiian Islands, Florida, and in the Caribbean Sea. In the Hawaiian Islands, there is a clearly defined distributional pattern of FP with up to 92% prevalence in Keneohe Bay, Oahu (Balazs 1991). In 1995 in Florida Bay, prevalences of up to 60% were reported in green turtles and 11% in loggerheads (B. Schroeder, National Marine Fisheries Service, personal communication). In the Caribbean, increases in FP have also been noted since the mid-1980s, particularly in Puerto Rico and Colombia (Williams et al. 1994). Possible causes of FP include infectious agents such as oncogenic viruses (Jacobson et al. 1991; Herbst et al. 1995; Casey et al. 1997; Quackenbush et al. 1998) and parasites (Dailey and Morris 1995; Aguirre et al. 1998), genetic predisposition (Herbst 1994), toxicants (Aguirre et al. 1994), ultraviolet radiation (Smith and Coates 1938), and other as yet undefined synergistic environmental factors (Herbst 1994). Herpesvirus was detected in fibropapillomas from Florida green turtles by the use of electron microscopy (Jacobson et al. 1991), and experimental transmission of fibropapillomas was achieved by inoculating disease-free green turtles with cell-free tumor homogenate (Herbst et al. 1995, 1996). Recent findings include the presence of retrovirus in both tumored and nontumored green turtles and the presence of three closely related herpesviruses in association with fibropapillomas in green turtles from Hawaii and Florida and in olive ridleys from Costa Rica (Casey et al. 1997; Quackenbush et al. 1998). Attempts to culture virus have not yet been successful, nor have Koch’s postulates been fulfilled. The fact that tumors have been experimentally induced in animals using oncogenic viruses or cell-free filtrates in laboratory studies does not necessarily infer a single cause–effect relationship in what may be a multifactored or multistep neoplastic process occur-