When the skin from the abdominal wall of mouse embryos taken at mid-pregnancy is explanted on a nutritive gel as described by Hardy (fig. 1), primary and secondary mammary ducts differentiate from the skin epidermis. This type of culture is necessarily complex, since a variety of tissues are growing in an organized fashion. The mammary ducts expand in the underlying connective and adipose tissues where they appear, after histological section, as tubes limited by 2 or 3 rows of cu-boidal cells. Besides the mammary epithelium, which is our main concern, and its loose connective environment, the hair follicles, the germinal layer and the keratinized cells of the skin still represent an important part of the culture. The ratio of mammary epithelium thus obtained is therefore extremely small. The maintenance period of such cultures is also relatively short since, after 10 to 12 days of explantation, foci of necrosis develop rapidly. In our first two attempts this procedure was followed:

Fragments of embryonic abdominal skin from agent-free C57BL mice are explanted on a nutritive gel containing defatted milk from the RIII strain in the final dilution of 10"3. After 10 days, the fragments are ground in their medium and centrifuged. The supernatant is in part inoculated intraperitoneally into C57BL baby mice while another part is diluted a thousand times in the medium of a new set of cultures. In this manner "passages" of the primary inoculum are made with 1000-fold dilutions for each 10-day period of maintenance in vitro. The results of ' the bioassays which became apparent 12 months later are summarized in table 1.

Table 1. Tumor Induction With Organ-Culture Medium

Concentrations of MTA* inoculums

Number of mice inocu-lated**

Number of induced tumors

Time of cultivation (days)

Time of appearance of first tumor (months)

Controls, acellular

MTA 10^-3

62

37

10

medium

MTA 10^-12

40

2

11

IMTA++ 10^-3

60

0

Culture extracts

CMTA§ 10^-3

62

5

4

14

(relative dilution

CMTA 10^-6

63

2

14

13

of primary

CMTA 10^-9

62

8

24

10

inoculum)

CMTA 10^-12

64

10

35

13

*MTA: mammary-tumor agent (RIII milk).

**C57BL strain, no spontaneous tumor in 12 years; Intraperitoneal inoculation, 0.1 ml. ++IMTA: incubated mammary-tumor agent (4 days at 37° C). §CMTA: cultivated mammary-tumor agent.

One can see that, while a dilution to 10^-3 of fresh RIII milk induces tumors in 37 of the inoculated mice, only 2 appear after a simple dilution to 10^-12. Furthermore, the same milk loses all its carcinogenic activity after 4 days of incubation at 37° C. This we could expect, since other authors have reported a loss of activity after 1 hour at 37° C. Nevertheless the milk agent is retained in the presence of living tissues, as shown by the induction of 5 tumors at the end of the first passage. This number climbed to 8 on the third passage and to 10 on the fourth. By then the primary inoculum had been diluted to 10~12 and kept at the incubation temperature of 37° C. for 35 days. A comparison with the above acellular controls suggests a definite multiplication of the agent in the cultures.