Cell Research 5.5 (1969) 198-204
A TRANSPLANTABLE IN ORGAN
M. L. ELLISON,’
E. J. AMBROSE
and G. C. EASTY
Chester Beatty Research Institute, Institute of Cancer Research, Royal Cancer Hospital, London S W3, UK
SUMMARY Experiments were carried out to test the hypothesis that tumour cells still retain genetic information coding for specific differentiated characteristics while the characteristics may not themselves be apparent in the tumour tissue. Tumour samples supported on Millipore filter rafts were organ cultured on nutrient agar in combination with embryonic “inducing tissues” for up to 3 weeks, or for up to 12 weeks without embryonic tissue. Characterisable morphological changes were found in one of the tumours tested-a transplantable rat tumour which originated spontaneously in the kidney. Although a highly anaplastic tumour, it formed tubules under the conditions of culture. There was no evidence to suggest that embryonic tissues could “induce” specific differentiations, but it was concluded that the genetic information for a differentiated characteristic was not irreversibly lost in this example of an anaplastic tumour.
The superficial similarity of anaplastic tumour cells to embryonic cells both in proliferative behaviour and in relative lack of specific morphological characteristics has led from time to time [7, 131 to suggestions that there may also be a functional resemblance between these cell types -that tumour cells may represent the results of a faulty cytodifferentiation. Differentiation is considered to be the process whereby different parts of an organism’s genome are selected for use in different cells or groups of cells, and are displayed as the specific characteristics of those cells. A breakdown of differentiative control could be due, either to loss of or damage to the basic genetic information, or to a fault in the expression of that information into the characteristic structure and function of the cell type. If the former were so, in the hypothesis quoted above one would not expect that the tumour could 1 Present address: Department of Anatomy, School of Medicine, University of Pennsylvania, Philadelphia, Pa 19104, USA Exptl Cell Res 55
again show normal differentiation, but in the latter case, the characteristics might reappear if the pathways whereby they are expressed were to be re-established. If the reappearance of characteristics in the cell or in their offspring can be demonstrated there is strong indication that the information is still present, even though it is not being used in the anaplastic tumour tissue. An approach which has been used in attempting to obtain re-differentiation of tumours is to associate them with embryonic tissues or the environment of a regeneration blastema, so that any morphological controlling factors present may act on the cells. Examples of this type of test have been given by Seilern-Aspang and others whose results with chemically induced epithelial tumours in the newt Triturus cristatus seemed to indicate that there was differentiation of malignant cells to normal epidermal cells, particularly in the tail region which has high regenerative capacity [lo, 121. They also used an embryonic system, 5 day chick embryo notochord, injected into chemi-
Tumour differentiating tally induced sarcomas of chickens in vivo, or combined with chicken sarcoma or human sarcoma in vitro [8, 9, 111. Cartilaginous tissue was obtained and was presumed to derive from sarcoma cells, though X-ray pretreatment of the notochord with 5000 r to limit proliferation of the embryonic cells markedly reduced the amount of cartilage. De Lustig and others [l, 23 used the blastoderms of chicks at the primitive streak stage, resting on tumour pieces, as a source of “primary organiser”, and found that after culture in this system Sarcoma 180 showed a tendency to tubular organisation, with a slower growth rate on re-implantation into host animals. Cultivation of 3&-day chick embryonic notochord and spinal cord with mouse osteogenic sarcoma (separated by vitelline membrane) resulted in a reduced number of giant cells and in some cases, abundant osseous trabeculae [l]. In these and other experiments (see Ehrenpreis for review ) there is some evidence that re-differentiation may occur in tumour cells, but it is complicated by the possibility of proliferation in the tumour of normal cells derived from the embryonic or regenerating tissue supplying the “factors”. The experiments reported here represent an attempt to clarify this position-to determine if tumour cells do retain the information necessary for specific morphological characteristics. MATERIALS
Basically, the method used involved the maintenance of tumour tissues in an environment in which they could be isolated from other tissues, and yet retain a certain degree of 3-dimensional architecture favourable to histotypic differentiation. The conditions required for such growth were met by “organ culturing”, using a Millipore filter raft resting on an agar solidified nutrient medium as the substrate for the exnlant. The growth medium consisted of a 1 % solution of Agar (DIFCO) in a 1 :l mixture of NCTC 109 and Simms BSS (Tris buffered and with double quantity of glucose) to which was added one or more of the following proteins: Horse serum (Burroughs Wellcome no. 5) Foetal Calf serum (Microbiological Assoc.); embryo extract (EEEo prepared from 9-l l-day chick embryos by by the method of Paul [a); and an homogenate of fresh hen’s egg. The proteins were added to give either 20% serum, or 16 % serum plus 20 % EE or egg. In a few cases NCTC 109 was omitted. Antibiotics were included at
final concentrations of 100 IU,/ml sodium benzylpenicillin, 37.5 IU/ml streptomycin sulphate, and 2.5 pg/ml
amphoteridin B deoxycholate.
Two variations of culture sequence were employed. The first (“short-term-culturin0 consisted of nlatine out 2 mms‘pieces of tumour supported by squares-of Tb; tvne Millioore filter (25 u thick. 0.8 u nore size1 onto the igir surface and incibaiing for’up tb 5 weeks at j7’C in 5 % CO% in air. The second involved a transfer of each tumour explant on its Millipore raft to freshly prepared medium after every 7 days incubation, the gas phase being renewed at each transfer. This “long-term” method permitted culture of tumour samples in an organotypic state for at least the 12 weeks attempted. Initially tumour samples were cultured (by the shortterm method) in combination with embryonic tissues of the “cartilage induction” system ; i.e. Notochord and/or spinal cord isolated from 3-5-day chicks embryos (as in Seilem-Aspang et al. [8, 9, 111).
RESULTS When Sarcoma 180 (a mouse transplanted tumour) was cultured in direct contact with 5-day chick notochord, cartilage was formed in the explant (see fig. 1); but all the cells in the cartilage matrix appeared to be chick cells. Since, in the 5-day chick embryo pre-cartilaginous condensations have already formed closely adherent to the notochord, making it impossible to remove completely all somite tissue from it, it was concluded that these cells incorporated into the explant were responsible for the cartilage seen. When younger embryos, in which complete isolation was possible, were used, no cartilage formed. In another approach the notochord was irradiated with 5000 r to prevent proliferation of the adherent somite cells, but, unless the irradiation was carried out in oxygen-saturated medium, embryonic cells were not prevented from differentiating into cartilage. A similar experiment was carried out on a Dimethyl 1,2,benzanthracene-induced chicken sarcoma, using irradiated 4-day chick notochord in direct contact with the tumour, and unirradiated notochord across a TA Millipore filter barrier from the tumour. These cultures were maintained living for up to 8 weeks, but showed no cartilage. However, these experiments involved an embryonic system which did not correspond specifically with the tissue of origin of the tumour. Exptl
Fig. I. Sarcoma 180, short-term-cultured C, cartilage; S-180, Sarcoma 180.
in a rat at the 73rd
Fig. 5. Rat renal
Fig. 6. Rat renal Fvntl
Fig. 7. Rat renal tumour
7 weeks on embryo extract containing medium.
Therefore, a homologous system was tried, using dorsal spinal cord of l&13 day mouse and rat embryos (postulated as kidney tubule inducer ) cultured transfilter to a transplantable rat renal tumour. The tumour which originated spontaneously, was in its 52nd transplant passageat the start of these experiments. Of a total of 28 tumours short-term-cultured in this combination only one clear case of differentiation of the tumour was obtained, large tubules being formed in the explant transfilter to 11 day mouse spinal cord. However, in this experiment there was also some differentiation in the control explant grown without inducing tissue, and indications of tubules in the sample taken from the rat at the time of explantation. Although this renal tumour was a highly anaplastic form, not normally showing any tubules at all, during the 20 transplant passagesover which it was studied (6 animals/passage)tubules did appear in three individual instances. In each
of these cases there was much necrotic tissue and the tumour grew very slowly in the animal. Thus it seemsthat this tumour has a latent ability to differentiate, which was enhanced but not “induced” by exposure to embryonic “inducing tissues”. A series of long-term-cultures was therefore undertaken to see if this capacity for differentiation could be brought out by prolonged maintenance in culture without embryonic tissue. Figs 2-6 illustrate stages in the culturing of one tumour taken from a rat in which it had been growing for 14 days (fig. 2). Fig. 3 shows the reaction to medium containing Foetal Calf serum as the only protein nutrient, after 2 weeks. Survival on this medium was for about 6-7 weeks during which time there was no indication of any differentiation. However, medium containing embryo extract in addition to either foetal calf serum or horse serum permitted survival, with differentiation, of Expti Cell Res 55
M. L. Ellisom, E. J. Ambrose & G. C. Easty
the same tumour for 7 weeks. Figs 4, 5, 6, and 7 taken after 2, 5, 6, and 7 weeks respectively, show the sequence of tumour morphology on this medium. It should be noted that representative samples of this rat renal tumour were the only tumours among a range tested in long-term-cultures which showed any differentiation. Samples of Walker carcinoma, Sarcoma 180, DMBA-induted chicken sarcoma, and a human renal secondary tumour, maintained in the same way for 6-10 weeks did not differentiate.
DISCUSSION Although no evidence has been obtained to suggest that tumour cells can be “induced” to differentiate characteristics for which they are not already determined (i.e. re-differentiate along a developmental pathway not in use in the cells of the tissue of origin), the results of the renal tumour cultures do suggest that this tumour has still got the necessary genetic information for differentiation-notably for tubule information. This indicates that at least in this tumour it is possible for tumour cells to lose overt differentiation for many cell generations, yet retain the gene codes for them, and equally important, retain the ability to use the codes so that the characteristics re-appear when conditions are again favourable. This redifferentiation seems to be in response to stabilising environmental factors, but the actual nature of the extrinsic stimuli which are stabilising expression of information is not clear. In these experiments an association of overall reduced growth rate with
Cell Res 55
differentiation was noted, of differentiation prevents whether re-expression of naturally from a slower further analysis.
but whether the onset rapid cell division, or characteristics results division rate awaits
We would like to thank the following members of the Chester Beatty Research Institute. and their assistants. for their co-operation in this stud;: Mr E. Woollard fo; the histology. -_, Mr B. V. C. Mitchlev for the exoerimental tumours which he made available,-and Mr K.-G. Moreman for the photomicrography. We are grateful to Dr R. L. Carter for his comments and advice. This investigation has been supported by grants to the Chester Beatty Research Institute (Institute of Cancer Research: Royal Cancer Hospital) from the Medical Research Council and the British Empire Cancer Campaign for Research, and by the Public Health Service Research grant no. CA-03188-08 from the National Cancer Institute, US Public Health Service.
REFERENCES 1. De Lustig, E S & Lustig, L, Rev sot Argentinia biol 40 (1964) 207. 2. De Lustig, E S & De Matrajt, H A, Rev sot Argentinia biol 37 (1961) 180. 3. Ehrenpreis, G Ya, Israel program for scientific translations (1964). 4. Grobstein, C, Exptl cell res 10 (1956) 424. 5. L&h. J W, Cytodifferentiation and macromolecular synthesis (id ti Locke) p. 35 (21st symp sot develop and growth). Academic Press, New York (1963). 6. Paul, J, Cell and tissue culture, 3rd edn. Livingstone, Edinburgh (1965). 7. - Cell differentiation (eds A V S de Reuck &J Knight) p. 196. Churchill, London (1967). 8. Seilern-Asuane. F. Honus. E & Kratochwil. / K. , Acta biol med germ’lO’(1963) ;147. 9. - Ibid. 11 (1963) 281. 10. Seilern-Aspang, F & Kratochwil, K, J embryo1 exptl morph01 10 (1962) 337. 11. - Acta biol med germ 10 (1963) 443. 12. - Regeneration in animals and related problems (ed V Kiortsis & H A L Tramusch) p. 452. Amsterdam (1965). 13. Waddington, C H, Nature (Lond) 135 (1935) 606. Received October 1, 1968