The growth of mammalian tumors in fertile eggs; is a filterable cancer virus produced?

In 1936 Bittner demonstrated for the first time that mammary cancer in the laboratory mouse is related in some way to a substance found in the milk. So-called low-tumor strains could be made to develop cancer by feeding the young mice with milk derived from a high-tumor strain mother, whereas the incidence of mammary cancer in high-tumor strain females could be greatly reduced by foster nursing them from birth on low-tumor strain mothers. That these observations are true has been confirmed by many investigators (1). There is no question that the milk influence has a large role to play in the production of mammary cancer in the mouse, and that this agent is propagable in the mouse frorn generation to generation. Its physical properties have been studied extensively by various workers, and it has been found by Bittner to be filterable through Seitz filters; by Bryan and Andervont, through Berkefeld filters; to remain active in 50 per cent glycerine in the icebox for 7 to 9 days but not for as long as 81 days; and to be still effective after lyophilization and resuspension in water 6 months later. The agent is destroyed by heating at 61~ to 66 ~ for 30 minutes, but will survive in mouse milk kept at 8 ~ C. for 14 days. It is active at pH 5.5 to 10.2, but inactive at pH 4.5. Ultracentrifugation shows the milk influence to be sedimented at 110,000 times gravity for 1 hour. Ultraviolet absorption spectrograms reveal a high percentage of nucleic acid in the active fractions of mouse milk, an indication that the material is either an altered ribose nucleic acid complex or something carried along in close association with it. It is not in the ethersoluble fraction of milk, as shown by Hieger. The agent is capable of causing the formation of immune sera when injected into rabbits. In a recent review of the milk influence Andervont (1) concludes that "filtration and uhracentrifugation experiments show that [it] can be transmitted by cell-free material, and its transmission through many generations of mice implies propagation. These features suggest the action of an agent belonging or related to the viruses." Consequently when Taylor (17) published the