Ill 



RELATIONSHIP OF GROWTH in vilVO 565 



differentiation of amphibian neural crest cells in tissue culture. The evidence 

 indicates that the a-amino group of this amino acid is necessary for the differentia- 

 tion of pigment cells, and the phenyl ring for differentiation of ectomesenchyme. 

 The character of growth in some cultures may be altered, without varying the 

 composition of the medium, by simply changing the pH. Under conventional 

 conditions in vitro, the mouse mammary tumour studied by Pikovski (1954) lacks 

 organization and produces a strong growth of connective tissue. If the tumour 

 cultures are maintained at pH 9.0-9.3, ducts and acini develop, and the resulting 

 highly differentiated cultures of glandular tissue can be maintained, with trans- 

 plantation every three or four weeks, for many months. They retain their ma- 

 lignancy. 



Part of the environment of any cell is the other cells in the vicinity, and impor- 

 tant studies have been made on the interactions of cell types in tissue culture. 

 When two cell types are grown in apposition, one culture may affect the growth 

 of the other. Normal and malignant fibroblasts have been shown to be stimulated 

 by leucocytes; normal mouse fibroblasts were inhibited by several sarcomas but 

 stimulated by carcinomas (Ludford, 1940; Ludford and Barlow, 1944). On the 

 other hand, the maintenance of some types of tumour cell in vitro requires the 

 presence in the culture of a supporting population of connective tissue cells, 

 perhaps corresponding to the stroma of tumours in vivo. Thus De Bruyn, Korteweg 

 and Van Waveren (1949) and De Bruyn (1949a, b) maintained over a period of 

 years cultures of a mouse lymphosarcoma, and inoculation of these cells into 

 suitable mice produced tumours. It was found possible to make separate cultures 

 of the two types of cells found in the cultures of this tumour, namely lymphoblasts 

 on the one hand and the associated fibroblasts on the other. The fibroblasts pro- 

 duced no tumours. For the first few months in tissue culture, the lymphoblasts 

 were capable of producing tumours, but thereafter this capacity was lost and the 

 cells became "altered" and non-malignant. Recombination of the two strains 

 of cells did not result in restoration of the tumour-producing capacity of the 

 cultures. Another kind of mutual dependence of normal and tumour cells was 

 demonstrated by Bichel (1952). In this case the growing leukaemic cells destroyed 

 the normal cells. The Yoshida ascites tumour of the rat is dependent, not merely 

 for maintenance of neoplastic character but for survival upon the presence of 

 normal cells (Schleich, 1956). Tissue cultures prepared from this tumour after 

 subcutaneous passage contained a mixed population of tumour cells, macrophages 

 and fibroblasts. Tumour cells which lost contact with fibroblasts did not survive; 

 if cultivated independently they ceased to divide within 10 days, and began to 

 degenerate. A dramatic revival occurred if normal rat fibroblasts were then added 

 to the cultures, even if the normal and tumour cells were not in contact. Extracts 

 of fibroblasts were ineffective. A single tumour cell inoculated into a normal 

 fibroblast culture was able to give rise to a luxuriant growth of tumour cells. The 

 interaction and interdependence of normal cells in tissue cultures in relation to 

 tissue differentiation is a field of growing importance. As early as 19 14 Champy 

 had observed that the presence of mesenchyme "prevented dedifferentiation" of 

 organized epithelia. Drew's (1923a, b) experiments in this area were discussed by 

 Strangeways (1924b) and he already at that time concluded that: "although a 



Literature p. §8i 



