Scapular cartilage in calves contains a substance that inhibits the vascularization of solid tumors (1). When this substance was infused into rabbits or mice, no toxic effects were observed in the animals, yet the growth of new blood vessels toward implanted tumors (V2 carcinoma and B16 melanoma) ceased and tumor growth stopped (2, 3). The single factor most limiting to the further study of this substance is its supply. Cartilage is present only in small quantities in mammalian species.

It occurred to us that sharks may be a potential source of this inhibitor because, unlike mammals, sharks have an endoskeleton composed entirely of cartilage. Cartilage composes about 6 percent of the shark's total body weight (4), compared to less than 0.6 percent in calves. In addition, some sharks are very large, about ten times heavier than calves.

Basking sharks (Cetorhinus maximus) 6.1 m long and weighing approximately 409 kg were obtained from Fresh Water Company, Boston. The fins and vertebrae were immediately excised, scraped with a scalpel blade to remove connective tissue, and stored at -20'C. To extract the inhibitor, a modification of the procedures used for calf cartilage (I- 3) was employed. The shark fins were cut into 1-cm' pieces and extracted in a solution containing IM guanidine and 0.02M 2-(N-morpholino) ethanesulfonic acid (MES) for 41 days at room temperature. Extracts were dialyzed exhaustively against water by using membranes with a 3500-dalton cutoff and centrifuged. The supernatant was filtered through Whatman I filter paper and then lyophilized. Five hundred milligrams of cartilage yielded I mg of this extract.

The shark cartilage extract was incorporated into 1-mm³ pellets of ethylene- vinyl acetate copolymer (40 percent vinyl acetate by weight) (5) at a level of 300 µg of extract and 700 mg of polymer. The polymer pellets have been shown to release over I µg of biologically active molecules per day for over 100 days (3, 5). The pellets were implanted into corneal pockets in New Zealand White rabbits. Directly behind the pellets were placed 1.5-MM³ pieces of V2 carcinoma (1). These tumors induced vessels to sprout toward them from the edge of the cornea. The bioassay consisted of measuring the length of the single longest blood vessel with a slit-lamp stereomicroscope.

In experimental corneas, tumors and pellets containing shark extract were used. Control corneas were implanted with tumors and identical-sized pellets containing no extract. Earlier studies have shown that the rate of tumor neovascularization in such controls is statistically indistinguishable from the rate induced by (i) tumors and pellets containing extracts of calf cartilage with no biological activity, (ii) tumors and pellets containing proteins or polysaccharides, and (iii) tumors alone (1, 3, 6).

The extract of basking shark cartilage significantly inhibited tumor neovascularization. Three different tests were conducted., and inhibition was observed in every case (Fig. I and Fig. 2a). After 19 days all control corneas had large, three-dimensional tumors with an average maximum vessel length of 6 mm (half the diameter of the cornea) (Fig. 2b). In contrast, none of the treated corneas had three-dimensional tumors. All treated corneas showed sparse vascularization, with zones of complete inhibition around the pellets (Fig. 2a). Average maximum vessel length was 1.5 mm, 75 percent shorter than in the controls (Fig. 1).

These results demonstrate that basking shark cartilage extract strongly inhibits tumor-induced neovascularization and that significant inhibition can be obtained with the extracts at a crude stage of purification. In contrast, calf cartilage extract must be highly purified by affinity chromatography before inhibition can be observed (1). [Improved purification procedures may exist; however, we have not as yet found one (1, 3).] It took 500 g of calf cartilage to produce I mg of a substance causing 70 percent inhibition of vascular growth (1), slightly less inhibition than observed here with one-thousandth as much starting material. Considering this observation along with the larger size and greater percentage of cartilage body weight in basking sharks as compared to calves, sharks may contain as much as 100,000 times more angiogenesis inhibitory activity on a per animal basis. Thus, with further study, shark cartilage may become a major source of angiogenesis inhibitor. Identification of this factor might provide insights into the tissue development of different species and into why elasmobranchs such as sharks, in contrast to mammals and even bony fish and amphibians, so rarely exhibit neoplasms (7).

The inhibitor does not appear to act directly on the tumor itself. Stereomicroscopic observations showed that in both control and treated corneas, the V2 carcinoma grew slowly in two dimensions in the collagen layers of the  of the cornea, indicating that the tumor cells continue to proliferate even in the presence of the inhibitor. Histologic sections showed healthy tumor cells at the pellet-tumor interface, and the inhibitor did not affect the growth of V2 carcinoma cells in culture. We used the same methods and obtained the same results with respect to tumor cell growth as in previous studies with calf scapular cartilage inhibitor (1-3). Both inhibitors appear to act on capillary advancement rather than on tumor cell growth directly.

It is not known whether all sharks contain angiogenesis inhibitors or whether the same molecule in calf and shark cartilage is responsible for the inhibition of tumor angiogenesis. Shark cartilage contains many of the same biochemical activities as calf cartilage, including lysozyme activity (8), cell growth-promoting activity (9), inhibitory activity against type I collagenase (10), and inhibitory activity against proteases such as  trypsin, chymotrypsin, and plasmin (11) However, when the guanidine extracts of shark and calf cartilage are analy4d, on sodium dodecyl sulfate gels (12), tile patterns are different, with the calf cartilage extract showing more bands, particularly in the molecular weight range ( 20,000 or greater. In addition, the protein content of the calf cartilage extract was 58 percent by weigh", whereas the protein content of the shark extract was 20 percent (13). Direct. comparison of the molecular characteristics of the two inhibitors must await their complete purification.