Dr. Dan Moore (Rockefeller Institute): I suppose we should follow the previous procedure and be provocative in our statements. Therefore, I will make a few statements that I am sure will be provocative.

Our present work with the milk agent is based on the following, which might be culled a sort of creed: We believe that the Bittner agent is really a virus and that it is carried in a large number of strains of mice. Actually, we believe that most mice carry this agent, and there are really few that do not. There are about nine strains, now reported in the literature, in which the milk agent has been found, even some that are not supposed to have it.

Also, we believe that a primary approach to this kind of investigation should include tissue culture. Dr. Lasfargues has just demonstrated what can be learned with this technique.

We believe that we are able to recognize this agent morphologically wherever we find it in tissue. I am not absolutely sure of this, however, because these tumor viruses can look a whole lot alike, and unless one takes some care in measuring the parameters, it is possible to mistake the virus.

Furthermore, we believe that we are able to trace the development of this agent at the cell membrane, that the entire assembly takes place at the cell membrane, and that the particles are liberated in acinar or intercellular spaces. The cell membrane thickens, protrudes to form a hemisphere, then a horseshoe-shaped spheroid which finally pinches off. The newly formed thick-shelled spheres are of remarkably constant diameter, ca. 95 nut. Particles slightly removed from the cell membrane contain a dense nucleoid within a membranous sac, which indicates a complete rearrangement of dense material soon after the particle is formed.

This process is clearly indicated in the mammary carcinomas of DBA mice kindly given to us by Dr. Anna Goldfeder of the Delafield Hospital: The production of milk agent is greater in these DBA tumors than in any other strain so far studied. The appearance of particles is the same, however, as has been described for other strains including the cultured explants from RIII tumors.

The ubiquity of these milk-agent particles in mammary carcinomas is demonstrated by their occurrence in Heston's (National Institutes of Health) C3Hf mice, a so-called agent-free strain kindly supplied by Dr. B. D. Pullinger of Glasgow. The concentration of particles in these tumors, however, appeared to be relatively very low. This low concentration may account for the difficulty in transmitting the disease from this strain.

Dr. Muhlbock: Everyone concerned with the milk agent has followed with great interest the work done by the Columbia group. I think everyone agrees that their approach to the problem is sound. It would appear from Dr. Lasfargues's report that one of the major difficulties in this field, i.e., the test for the milk agent, has been overcome. Otherwise, one must rely on the biologic test, which takes a long time and is a handicap to be overcome if we want to proceed more rapidly in the future.

Maybe the approach Dr. Lasfargues has found in tissue culture will help in this respect. His conclusions from the work in tissue culture agree with what we see with purely biologic investigations. The role of the connective tissue is certainly very important, and I am glad he stressed this point.

It was gratifying that he emphasized the role of hormones in their influence on the conditions necessary for the modification or multiplication of the milk agent. He indicated, himself, that the evidence for this is not very convincing, and I wonder if he would comment in this respect. First, does the agent multiply in the mammary gland in tin; prelactation stage? After lactation the mammary gland practically disappears by involution, and I would like to ask him what happens to the virus, then, which has multiplied. As another point, we know that the milk agent can multiply in male animals in which there is hardly any mammary gland at all, and we are sure that the mammary-tumor agent can be transferred very easily with the sperm in certain strains of mice.

Dr. Sykes (M. D. Anderson Hospital): Contrary to the findings of Dr. Lasfargues, we in Houston have found that the cells of mammary tumors from agent-free, or supposedly agent-free, mice grow out as beautiful epithelial sheets. The cells are very regular, and no abnormalities are found in them, when they are viewed by phase microscopy.

In contrast, the cells growing out from tumors of mouse strains carrying the virus are bizarre in shape. They are quite unlike normal epithelial cells, tend to be multinucleate, to form giant cells, and to be highly vacuolated, even after 2 years in culture.

The mice providing the spontaneous tumors were the C3H/He, C3H/Bi and C57BL strains. The agent-carrying strains have been the equivalent of these three, and, in addition, the Strong A strain.

Just to confound the issue a little more, one of the spontaneous tumors from a C3H mouse carrying the agent was passed 19 times in the chick-embryo yolk sac and, thereafter, 69 times in tissue culture over a period of 2 years. The cells of this culture have already shown the presence in electron micrographs of characteristic virus particles.

Dr. Alice Moore: (Sloan-Kettering Institute): What I have to say is not related to the milk agent but does point out some of the difficulties of growing these tumor viruses in tissue culture. Dr. Friend described her agent about 4 years ago, and, ever since then, we have been attempting to grow the virus in tissue culture.

A couple of years ago we reported preliminary results indicating that the virus grew in a rather desultory fashion. Since then we have refrained from further publication in the hope that we could better the situation, but this has not been the case.

The results obtained with cultures of mouse embryo, infected either with filtrates or with cells from the mice with leukemia, showed that virus was in the culture supernatant fluid during the 1st month and caused leukemia in 4 of 10 mice inoculated. In the 2d month, no virus was demonstrable, but in the 3d and 4th months, it began to reappear. By the 6th month, there was a little more virus, which caused disease in 6 of 9 mice. Thereafter, it kept appearing and disappearing throughout a period of 21 months. We have had cultures over 2 years, and many of them still produce the virus, but in this rather relaxed fashion. It goes on and on, but we have never succeeded in getting any more virus.

This indicates to me that at least some of these tumor agents are going to be difficult to propagate in tissue culture and that we must either modify our present methods or develop new ones, such as Dr. Lasfargues has done.

Dr. Eddy (National Institutes of Health): An important point of Dr. Lasfargues's paper suggests a solution to Dr. Moore's problem. This is the observation that he was able to use the stroma cells, and I wonder if anything is being done along these lines, whether monolayer cells can be used with the addition, perhaps, of embryo extract or can the stroma cells be preserved in additive cultures? Would not that make possible propagation of the virus, in a fashion, much easier than that employed at present?

May I ask if cortisone or X rays do anything toward making the animals come down a little faster?

Dr. Sabin (Childrens Hospital Research Foundation, Cincinnati): The work presented has required so much courage and physical labor that I hope what I have to say will not be regarded as hypercritical but rather as a resume of the important conclusions that have been reached.