Studies of tumour invasion in organ culture

Studies of tumour invasion in organ culture

Europ. 07. Cancer Vol. 5, pp. 255-263. Pergamon Press 1969. Printed in Great Britain Studies of Tumour Invasion in Organ Culture I. Effects of Basic ...

9MB Sizes 1 Downloads 28 Views

Europ. 07. Cancer Vol. 5, pp. 255-263. Pergamon Press 1969. Printed in Great Britain

Studies of Tumour Invasion in Organ Culture I. Effects of Basic Polymers and Dyes on Invasion and Dissemination MARGARET M. YARNELL and E. J. AMBROSE Department of Cell Biology, Chester Beatty Research Institute, Institute of Cancer Research: Royal Cancer Hospital, Fulham Road, London, S.W.3.

INTRODUCTION ALTHOUO~ the final: justification of studies on invasion in vitro is their reproducibility in the whole animal, it !is nevertheless clear that some of the underlying cell interactions, e.g. contact inhibition of movement [1], density dependent inhibition of mitosis [2], contact guidance [3] can be investigated more readily in tissue culture. The review of Abercrombie and Ambrose [4] showed clearly that surface properties play an important role in invasion. The present and following paper describe an attempt to influLence invasion by altering cell surface properties in various ways. It was decided to try to develop a quantitative technique for the investigation of invasion in vitro using a simpler culture method than those currently aw~ilable, e.g. the technique first described by Wolff in 1961 [5]. It has been suggested that the increase in surface charge demity found in some transformed and tumour cells is due mainly to an increase in the amount of sialic acid in the cell coat [6]. Defendi and Gaffe [7] proposed that this material m a y be responsible for the reduction in contact inhibition often found between these cell~. Since a reduction in

contact inhibition between tumour and normal cells seems necessary for invasion to take place in vitro, it follows that, if the functional component of sialic acid in this situation is the electronegative charge, neutralisation should affect contact inhibition directly, and secondarily, invasion. Neutralisation of the excess charge might also affect invasion by increasing cell adhesions, since some workers [8] have proposed that tumour cells are less adherent because of their abnormally high surface charges. For these reasons, therefore, cells and cultures were treated with the cationic polyelectrolyte poly-L-lysine, and the basic dye toluidine blue. The enzymatic removal of sialic acid, and the subsequent effect on invasion, will form the substance of the second paper.

METHODS In the assay method which was eventually evolved for these experiments, small explant cultures were set up on an agar-based medium, and to these was added a constant known number of potentially invading cells. The combined cultures were fixed after five days, sectioned and stained, and the amount and type of invasion, estimated microscopically and semi-quantitatively, were correlated with various types of treatment.

Submitted for publication 20 September 1968. Accepted 23 January 1969.



Margaret M. lrarnell and E. ~. Ambrose

Details of the tissue cultures Growth medium consisted of Burroughs Wellcome 'BHK' (Eagle's MEM, modified for BHK cells), 10% fresh chick embryo extract, 10% Burroughs Wellcome heated horse serum (No. 5), 275 mg % NaHCO3, and 100 i.u./ ml benzylpenicillin, with a base of 0 . 6 % Difco 'Bactoagar'. The ten times concentrated BHK medium was diluted with tris/HC1 buffer (pH 7"4) instead of with water. This helped to maintain the correct pH. Five ml of this medium were pipetted into a series of 'Esco' (Rubber) disposable 5 cm diameter petri dishes, and 6-8 explant cultures (1 m m s) of 16-20 day foetal C- mouse heart muscle were supported on the solidified agar surface by fragments of Millipore filter (RA, 150 ~ thick, 1 "2 ~t average pore size). BHK cell cultures [9], both untransformed (BHK21) and transformed in vitro with polyoma virus (Py) were provided by Professor Stoker, and were maintained to his specification in the authors' laboratory. Two days before setting up the explant cultures, diluted Indian ink was added to the medium of one or two 24 hr monolayer cultures (in 4 oz medical 'flats') as described by Stoker [10]. It was not thought necessary to test the effect of Indian ink on Py cells, since Stoker reported no change in the attachment, spreading or multiplying of inklabelled Py (and BHK21) cells existing as a monolayer in ink-free medium. On the day an experiment was begun, the cells were detached from the glass with 20 mg %'Versene'. The resulting suspension was washed twice with Hanks saline to remove adsorbed carbon particles, and a cell count was made. The



~g. I.

Surface and side.views of a culture dish, with enlarged diagram of one culture.

pellet was then resuspended in 0.1 ml saline, and pipetted on to a clean dry sterile glass surface so that it formed a drop. The surfaces of the Millipore fragments were first dried with sterile lens paper to prevent the liquid spreading from the edges of the explant, and then drops of approximately 0.4 ~ml volume containing 40,000-45,000 cells were transferred to each explant in turn, using a very small sterile platinum wire loop. Figure 1 shows several views of a culture at this stage. The culture dishes were then placed in a humidified lunch box, which was permanently fitted with inlet and outlet tubes, and gassed with a mixture containing 5% CO,, before the box was sealed and incubated at 37°C. Cultures were fixed in Bousins at the end of five days, sectioned at 5~t and stained with haematoxylin (Harris') and eosin.

Substances tested (a) Poly-L-lysine. (Bromide salt, Koch Lights, tool. wt. 50,000). i. This was added to the growth medium before gelation to give final concentrations of between 1 and 150 gg/ml. ii. Py cells were incubated at room temperature for 10 rain in serum-free BHK medium containing 75 or 150 gg/ml polylysine. The reaction was stopped with 10% serum. Polylysine treatment killed 30% of the cells at the lower, and 50% at the higher concentration. The cells were added to the explants as above, after washing once. (b) Toluidine blue. (Geo. Gurr Ltd., tool. wt.

305.8.) This was added to the growth medium before gelation, to give final concentration, of 1,5 and 10 ~tg/ml. (c) Control. In each experiment, 6-8 explants received untreated Py cells and were grown on medium containing neither polylysine nor toluidine blue. These were the control cultures: 20% consisted of dead explants at the end of the experimental period, and 1% contained no living Py cells.

Estimation of results This was carried out by microscopic inspection of one or more sections from each culture, at a magnification of X480---more rarely at x 1250. Absolute counts of invading cells per field were impossible to obtain because of the difficulty of deciding whether Py cells at the periphery of the cultures were growing directly on the Millipore, or whether they were in that position after replacing normal fibroblasts outgrowing from the original explant, i.e. it

Studies of Tumour Invasion in Organ Culture I was not possible to recognise invasion in this area. A statistical method based on counts of invading cells in four predetermined areas in the total microscopic field yielded significant results at the beginning, when compared with non-quantitative vi[sual assessment, but this required at least six complete, non-fragmented sections per culture:, which could not always be produced. (It was found by one of the authors, M.M.Y., in later experiments, quite practical to prepare suitable sections personally; future investigations will therefore make use of the statistical method outlined above as well.) The results presented here are thus a preliminary survey of the problem. In default of suitable microscope sections, a less quantitative assay method was devised, by which both the approximate amount and the type of invasion could be estimated. This method consisted of recording deviations from (1) the approximate amount of invasion in untreated explants (termed the 'norm') and (2) the appearance of the invading cells (termed 'alpha') in ,;uch cultures. Normal invasion is defined as that resulting from the addition of between 40,000 and 45,000 Py cells to explants of 16-20 day foetal mouse heart muscle, not more than l m m 3, cultured under standard conditions as described on p. 256, for five days. The histological picture presented by an explant invaded to the normal degree is of replacement of approximately half the explant with Py cells, this replacement having been carried out by the infiltration of streams of cells, issuing forth from that area of the culture which had already (after five days) been replaced completely with invading cells. In some cases these early replacing cells had entered by way of the top surface of the explant (Fig. 9) upon which they had been deposited originally from the: platinum loop. More usually, however, they appeared as the progeny of Py cells deposited, on the Millipore, which had invaded by coming in from underneath the explant (Fig. 6). This latter mode tended to cause the death of the normal tissue, if complete, by separating it from its source of nourishment, and was so described by Leighton [11]. Normal invasion was recorded in 178/220 control cultures (81%) under appropriate conditions. Comparing invasion of BHK21 cells with normal invasion of Py cells, it was noted that at least twice as many cells were required to produce the same amount of invasion, and that infiltration new;r occurred. BHK21 cells invaded as a solid mass of cells (Fig. 7) and streams of single invading cells, so characteristic


of Py invasion, were never observed. The procedure for assaying invasion was as follows. By inspection of a randomly chosen microscope section, the explant was judged alive or dead. This was recorded as A (alive) or B (dead). The invading Py cells were likewise judged alive (I) or dead (II). Cultures consisting of a living explant and live invading cells were then examined to determine the amount of invasion. That corresponding to 'normal' invasion (p. 257) was signified as 1, and the whole section (culture) coded as AI1. If fewer than 10-15 cells were seen to have invaded in the section examined, invasion was judged 'reduced' (AI2) and if more than 3/4 of the explant had been replaced by Py cells, invasion was called 'increased' (AI3). Cultures with dead explants and/or Py cells were scanned in the same way to establish the amount of invasion which had taken place. It will be obvious from this description that only gross deviations from the norm were recorded. This, together with the division of degrees of invasion into three categories only, was done to improve the reliability of the method, subject as it was already to many inherent sources of biological variation. Finally, the morphological appearance and arrangement of the invading cells was assessed. This fell into four broad categories (Fig. 2). alpha


Typical of control culture. Cells ovoid in section, with few carbon particles in the cytoplasm. Cells lying in 'sheaves' as BHK21 cells do in monolayer culture, with long axes parallel, cells and nuclei



~ ~ ' ~ . ~ _, (b/ POLARISED




Fig. 2. Diagrammaticrepresentation of (a) alpha, (b) polarised, beta (c) gamma and (d) delta invading cells.

Margaret M. Tarnell and E. J. Ambrose


Table 1. Effects of polylysine and toluidine blue on the viability of standard cultures after 5 days % Cultures with Control Polylysine in culture medium 1-150 Bg/ml Polylysine used to pretreated Py cells 75 Bg/ml 150 og/ml Toluidine blue in culture medium 1 pg/ml 5 pg/ml 10 Bg/ml

flat and elongated, with few carbon particles in the cytoplasm. Also called 'polarised'. gamma Ceils round in section with rounded nuclei, tending to separate, few carbon particles in the cytoplasm. delta Cells rounded in section and deeply stained, quite separate, many carbon particles in the cytoplasm. (No gamma ceils were observed after treatment with polylysine or toluidine blue.) Microscopical examination was carried out 'blind'. The histopathological picture presented by an explant normally invaded by alpha cells can be seen in Fig. 3. Results were calculated by expressing as a percentage the number of cultures showing a particular feature within the total number of available cultures, e.g. reduced invasion. Only cultures in which both explant and invading ceils were still alive at the time of fixation were considered for the results summarised in Tables 1 and 2. (Values for dead cultures could be obtained in the same way when required.) The results obtained by this method of assessment compared quite satisfactorily with those preliminary results obtained from the statistical method (p. 257). Mitotic indices were calculated by counting both the ceils in mitosis and the total number of Py cells in one section of each culture, and Table 2.

dead explants 20

dead Py cel~ only 1

23 32 45 14 45 81

0 16 50 12 44 79

No. cultures 220 46 22 20 28 29 16

adding together the numbers obtained from each sample in the group, before working out the percentages.

Tests in monolayer cultures Monolayers were set up in 'Falcon' disposable plastic flasks, 30 ml capacity, by adding 5 ml organ culture medium minus agar, with 0 . 2 5 × 1 0 e Py cells, and adjusting the gas phase with CO~. Substances to be tested were added after 24 b_r, with complete medium change. R a n d o m counts were made daily on the population, and the final number was counted on the fifth day, after the cells had been removed with trypsin-versene (0.05% trypsin, 20mg% versene). Four test were made with toluidine blue, and two with polylysine. RESULTS From preliminary results it was found that the amount of invasion, under standard conditions, varied directly with the number of untreated Py cells deposited on the explants.

(a) Lethal effects upon explants and invading cells (Table 1) Polylysine had no adverse effect on either component when present in the medium; toluidine blue, on the other hand, had an increasingly lethal effect on explants and Py

Effects of polylysine and toluidine blue on the amount of invasion and the appearance of the invading cells % Cultures with

Control Folylysine in culture medium Polylysine used to pretreated Py cells Toluidine blue in culture medium

reduced invasion

5 cells

increased invasion

[3 cells

No. cultures






I-5 Bg/ml 75-150 pg/ml

4 7

0 12

12 9

54 12

23 14

75 Bg/ml 1 Bg/ml 5 gg/ml

87 9 7

35 0 0

0 30 79

8 65 63

15 23 14

F/g, 3,

The 'normal' amount of invasion of 'alpha' Py cells into foetal mouse heart (H) after 5 days. H. and E. x 300.

iiii!ii!!ii iii!ii!i 'i!!!i

:ii~!~ !!!!i ~i¸ ii ll¸ i

ii~iii ¸¸iiii/! i¸¸~ ~



iii!iiii;!ii!!!iiiiilfi!d i i ii!/!!~:!

Fig. 4.

T~e effect of including 5lxg/ml toluidine blue in the culture medium: increased invasion. H and E. X 300. (to face p. 258)

Fig. 5.

The effect of pretreating Py cells with 75gtg/ml polylysine: delta cells. H. and E. × 580.


Fig. 6.

ii! !/i/i!!!!iiiiil ili!!ii/ ii} / !! iiil/iiii!ii }

The effect of including l gtg/ml toluidine blue in the culture medium: polarised Py cells. H. and E. × 300.

f!i!!fiilC! : d i!i~!~i!!ii:~i~ii~i~iiii!iiiii~ii, ~ ~~i~ i i i. . . .


~!i~!~d %!£Z ii~!~~~ ~ ~ !~!i ]%~

~ iii~ i ii151~ii5~i ~I~ • i~


Fig. 7.

Invasion 'en bloc' of BHK21 cells labelled with Indian ink into foetal mouse heart after 7 days. H. and E. X 300.

Fig. 8.

Invasioi~of ink-labelled Py cells into foetal mouse heart showing the parallel arrangement of the invading cells. H. and E. x 300.

Fig. 9.

Foetal heart muscle fibroblasts exhibiting parallel alignment within an explant. H. and E. × 580.

Fig. 10. Invasion by infiltration-of BHK21 cells which had spontaneously transformed in monolayer culture to a Py type of growth pattern. H. and E. × 300.

Studies of Turnout Invau'on in O rganCulture--I cells as the concentration in the medium increased from 1 to lOlag/ml. It is interesting to note that toluidine blue was more toxic to the malignant than to tlhe normal element in the cultures. Invading Py cells:, pretreated with 75~tg/ml polylysine before combination with the explants, resulted in cultures of which 15% contained no living Py cells and 12% dead explants after five days. Pretreatment with 1501ag/ml polylysine was associated with an approximately two-fold increase in mortality of both components. (b) Effects upon the amount of invasion (Table 2) Again, the presence of polylysine in the medium at any of the concentrations tried had no effect on the amount of invasion. Toluidine blue at concentrations of 1 and 5 lag/ml was associated with increased invasion in 22% and 71% (Fig. 4) of the cultures, respectively. At the higher concentration, 19% of explants were completely rep]laced by invading cells. Since pretreatmenl: of Py cells with 150 ~tg/ ml polylysine was ,;o lethal to the cultures (Table 1), only the results obtained by the use of 751ag/ml will be considered further. This is necessary because the number of living samples in which the effect of the higher concentration of polylysine could be observed was so reduced. Pretreatment with 75~tg/ml polylysine resulted in reduced invasion in 76% of cultures, though there were no examples of increased invasion. (c) Effect upon the appearance and arrangement of Py cells The change in appearance in Py cells desscribed as 'delta' did not occur in control cultures. It was seen, however, in 12% of cultures maintained on medium containing up to 1501~g/ml polylysine, and 35% when Py cells had been pretreated with 75~g/ml (Fig. 5). Polarised Py cells were observed in 44% of cultures maintained on medium containing 1-5~tg/ml polylysine, and in 55% and 53% of those maintained on 1 (Fig. 6) and 5[,g/ml toluidine blue. Polarisation did not occur after pretreatment with polylysine, nor at higher concentrations when present in the growth medium. (d) Effect on the type of invasion The invasion of polarised Py cells was not of the infiltrating type typical of untreated Py cells (Fig. 3), but rather of the 'en bloc' type seen when BHK21 cells invade (Fig. 7).


(e) Effect on mitosis in Py cells The mitotic index of 25,700 invading nonpolarised Py cells maintained on control medium, was 1.1%. That of 2500 polarised Py cells maintained on polylysine-containing medium was 0.6%, and of 8000 similar cells on toluidine blue-containing medium (1 and 51ag/ml) was 0.7%.

DISCUSSION Lethal effects in general Considering that Kornguth et al. [12] found that polylysine at concentrations less than 200~tg/ml had no adverse effect on Ehrlich ascites turnout cells, it was unexpected to find considerable cell death, under similar conditions, with Py cells in the present study. Nero et al. [13] reported that polylysine did not penetrate mammalian red blood cells but, in the absence of serum, was adsorbed on the cell surface. It is clear that polylysine, in the experiments under review, caused no cell damage when serum was present in the medium; from this it may be assumed that the molecule became adsorbed preferentially on the serum molecule. Other results suggest, however, that polylysine did penetrate the treated cells, thereby causing their death when serum was absent. From experiments in monolayer cultures, it was observed that toluidine blue penetrated Py cells even when serum was present, and that this had a lethal effect. This would seem to explain the progressive damage to cultures (p. 258) with increasing concentrations of the dye, and the lack of effect of polylysine under similar conditions.

Effects of polylysine on invasion If the reduction in invasion which followed pretreatment of invading cells with 75~tg/ml polylysine was due to another cause besides the death of 30% of the treated ceils; this other cause may be characterised by the presence of cells, since the 3 cells were only found after pretreatment. If this is then so, the 8 cells have a reduced capacity for invasion. There is an interesting parallel with untreated Py cells during the first 48 hr after deposition on an explant. These cells also have the appearance, and invade scarcely at all; the lack of invasion is obviously due to the fact that the cells had not, during the first 48 hr, begun to attach, spread or move about on the explant. This is what happens, of course, when Py cells are seeded on a glass surface and incubated. It may be that polylysine reduced the capacity of the cells to move. Our cinemato-


Margaret M. Yarnell and E. J. Ambrose

graphic observations (unpublished) confirm that polylysine-treated cells exhibit no directional movement at all, in comparison with untreated cells, which move freely. Nevo et al. [13] showed that the polylysine molecule is adsorbed on the mammalian cell surface, and that this is associated with a reduction in the surface charge. Korohoda and Ambrose (in press) found that polylysine, in concentrations of 100 ~tg/ml and above, caused total cessation of movement in the protozoan Naegleria gruberi. This paralysis was associated with a decrease in net negative surface charge. Richardson [14] reported mitotic arrest in prophase in mouse ascites tumour cells after polylysine had been administered intraperitoneally, but no significant increase in the number of treated cells at any stage of the mitotic cycle was observed in the present study.

Polarisation The appearance and arrangement of invading Py cells which has been termed 'polarised' was observed in 10% of the organ cultures grown under control conditions. An explanation must be found, therefore, not for the development of a new phenomenon, but for the increase of the occurrence of an already existing one, under certain conditions. Not only did the percentage of cultures showing polarisation increase after toluidine blue was included in the medium, but also cultures treated with 5~tg/ml had approximately three times as many polarised cells as cultures treated with 1~tg/ml. The presence of polarised cells when polylysine was present in the medium will not be considered, since it only occurred at the lowest range of concentrations. Polarisafion could have been caused by the presence of more Py cells than usual within the', explant, since it was associated, in many cases, with increased invasion (Table 2). This was due apparently to a stimulation by toluidine blue of mitosis in Py cells; l llg/ml toluidine blue in the medium of Py monolayer cultures was associated with a two-fold increase in mitotic rate at the end of five days. Since Py cells tend to invade heart muscle in parallel lines (Fig. 8), presumably under contact guidance, the typical polarised appearance would appear if enough Py cells were present in one area. This would also explain t h e greater incidence at 5 ~tg/ml than at 1 ~tg/ml, for the stimulatory effect increased in that order. The reduction in mitotic rate to approximately half in cells already polarised would then be secondary, and be due to the inhibition of growth ultimately seen in crowded Py cultures.

Polarisation could also have come about by a return, for some reason, of mutual contact inhibition between toluidine blue-treated Py cells, with or without the help of contact guidance. The phenomenon was never observed in dye-treated Py monolayers, suggesting that neither mutual interactions, nor increased growth rate were enough by themselves to produce polarisation. Therefore, it seems most probable that polarisation was due to a degree of contact inhibition exerted upon the invading cells by surrounding heart cells, as described by Stoker in a monolayer situation with similar cells [15]. If the normal cells were arranged in a parallel fashion--and Fig. 9 shows that they can b e - the invading cells, if under contact influence, would also line up. Invading and invaded cells have been seen to lie side-by-side and parallel after toluidine blue-treatment in the organ cultures, but it was not possible to determine whether the interaction was one of inhibition, or guidance or whether it was in fact fortuitous. The reduced mitotic rate of the polarised cells corresponds with this interpretation, since Stoker reported that Py cells aligned on nondividing mouse fibroblasts were inhibited from dividing, and the fibroblasts in the present study were rarely seen to divide. The progressive increase in polarisation observed when cultures were grown in the presence of increasing amounts of the dye may thus be due to a reduction in surface charge following the adsorption of toluidine blue on the Py cell membrane, which had allowed the cells to approach more nearly and to align themselves more completely on the normal fibroblasts. The untransformed BHK21 cell usually invades in a polarised manner. One other possible cause of polarisation should be mentioned briefly. Polylysine has been shown to link red blood cells, despite their electrostatic mutual repulsion [16]. However, since this effect was traced to the size of the molecule concerned as well as to its charge, it seems unlikely to have occurred with toluidine blue, which has a relatively small molecular weight. There is no evidence to suggest that toluidine blue was polymerised at the cell surface in the present study, thereby increasing its cationic effect. This limitation, and the failure to observe polarisation in homologous cultures of Py cells render the explanation untenable.

Invasion of BHK21 cells The difference in invasion pattern between Py and BHK21 cells (p. 257) must be due to

Studies of Tumour Invasion in Organ Culture~I some difference in the manner in which the two types of cells invade. From the experiments in the present study, it would seem that Py ceils invade by means of free cell movements, cell division, and some capacity to nourish themselves on the normal tissue which they invade. There are suggestions that BHK21 cells are not capable of living off the explant to the same degree as Py ceils are, but the possibility has yet to be fully explored. If BHK21 cells were completely normal, it would be surprising that they invaded at all. Ir~ fact, this line is considered premalignant by several workers [17], in part because the cells are capable of forming multilayered monolayers. It seems unlikely, however, that the difference in invasive behaviour is due to a radical difference be.tween mitosis control in Py and BHK21 cells, as this appears to be a matter rather of degree. The situation as regards growth-control in transformed and untramformed BHK cells is too unsettled at present to allow really categorical statements. It seems more likely that the difference is due to a reduction of invasive movements in BHK21 cells. If intercellular control of movement operates in three dimensions, BHK21 cells, which are supposed to operate it in monolayer culture, might also do so in the invasive situation. As Abercrombie has pointed out [18], retention ot" mutual adhesions is not necessarily a barrier to invasion. In fact,


Leighton suggested that aggregation into broad 'fronts' is necessary for some types of cells to invade, since their individual capacities to respond to metabolic and nutritional gradients are so small [19]. This would explain the 'en bloc' adherent nature of BHK21 invasion, in comparison with t h e infiltrating Py type. The probability exists that BHK21 cells are subject to contact inhibition (of movement) from the normal cells of the explant, thereby dissemination is reduced. If this is so, and is also the cause of polarisation in invading Py cells (p. 260), the similarity between the two types of invasion would be explained satisfactorily. It is of interest to note that BHK21 cells, kept in monolayer culture long enough (approximately 6 weeks) for the inevitable 'spontaneous' transformation to a Py type of growth and appearance to occur, invaded under standard conditions in the typical Py manner (Fig. 10). In conclusion, invasion in this in vitro system seems to be largely a matter of movement, variations were probably due to differences in intercellular controls, and toluidine blue reduced the dissemination of the invading cells. No evidence was found to support the proposal (p. 255) that the degree of contact inhibition is governed by the presence of charged molecules at the surface, and that reduction of this charge will prevent invasion in vitro.


Une m/thode exp/rimentale a/td mise au point pour/tudier in vitro le rt$le de la charge de surface sur l'invasion. Des hombres connus de ceUules BHK21 Py furent ajout/s g~ des cultures organotypiques de coeurfoetal de souris, et maintenus, pendant 5 jours, dans un milieu comportant de l'agar, le milieu de Eagle modifi/, du sdrum de cheval et de l'extrait emb!yonnaire. Les effets des 2 substances basiques, le bleu de toluidine et la polyL-lysine, furent /tudi/s. Les concentrations du colorant, d/passant 1 ~tg[ml du milieu de culture/taient l/thales, tandis que la polylysine n'avait pas d'effet toxique, mgme ~ la plus forte concentration utilis/e, 150 lag[ml. Les cellules prdtraitges par des concentrations plus basses de polylysine, en l'absence de s/rum, subissaient cependant un certain dommage; de plus, le pouvoir invasif de ces ceUules gtait diminu/. On suppose que l'adsorption de polylysine rgduit le pouvo~" invasif, en diminuant la motilit/ des cellules. La pr/sence de bleu de toluidine, ~ des concentrations non I/thales, provoquait, de fafon typique, un pMnomkne que l' on a appel/ la 'polarisation', c' est-~-dire une disposition des cellules ,,.nvahissant le tissu normal semblable ~ celle des fibroblastes diplo~des normaux, poussant en couche monocellulaire sur le verre. Les raisons possibles de la polarisation sont discut/es, en insistant particulikrement sur les modes d'invasion diff/rents manifest/spar les cellules BHK21 Py et par les fibroblastes non transform/s BHK21 dont elles sont issues. SUMMARY A culture and assay system was developed to investigate the role of surface charge on


Margaret M. YarneU and E. J. Ambrose invasion in vitro. Known numbers of BHK21 Py cells were added to organ cultures of foetal mouse heart, and maintainedfor five days on a medium consisting of agar, modified Eagles medium, horse serum and embryo extract. The effects of treating with the basic substances toluidine blue and poly-L-lysine were then studied. Concentrations of the dye in the culture medium in excess of 1 ~tg[ml were progressively lethal, though polylysine in the medium had no toxic effects even at the highest level tested 150~tg/ml. Cells pretreated with lower concentrations of polylysine in the absence of serum, however, showed a certain amount of damage; in addition, subsequent invasive powers were shown to be impaired. It is suggested that the reduced invasion was due to adsorption of the polylysine molecule, which decreased the capacity of the cells to move. The presence of toluidine blue in the culture medium when not lethal, typically produced a phenomenon termed 'polarisation'--the arrangement of the invading transformed cells within the normal tissue in the manner of diploid normalfibroblasts, growing as a monolayer on glass. Possible reasonsfor polarisation are discussed, with particular reference to the differing modes of invasion displayed by BHIf21 Py cells and by the BHK21 (untransformed) parent fibroblast. ZUSAMMENFASSUNG Es wurde ein System der Zellkultur entwickelt, um zu untersuchen, welche Rolle die Oberflgchenspannung bei der Invasion in vitro spielt. Eine bestimmte Anzahl yon BHK21 Py-Zellen wurde Organkulturen yon J'6talem M~useherz beigegeben und fiir 5 Tage mit einem Medium, welches aus Agar, ver~ndertem Eagles Medium, Pferdeserum und embryonalem Extrakt bestand, gehalten. AnschlieJ3endwurde die Wirkung der Behandlung mit den basischen Substanzen Toluidin-blau und Poly-T-lysin untersucht. Die Farbstoffkonzentrationen in Kulturmedien von mehr als 1 ~tg[ml fiihrten in zunehmendem Marie zum Tode der yellen, w~hrend die st~rkste Konzentration yon 150 ttg/ml Poly-L-lysin keine toxische Wirkung zeigte. ZeUen, die bei Weglassen des Serums mit einer niedrigeren Poly-L-lysin-Ifonzentration vorbehandelt wurden, wiesen einen gewissen Sch~digungsgrad auf ; weiter zeigte sich, daft die Intensit~t der anschlieoSenden Invasion beeintr~chtigt war. Das liiflt darauf schlieflen, daft die Beeintr~chtigung der Invasion auf eine Adsorption des Poly-L-lysinMolekiils zuriickzufiihren ist, wodurch die Bewegungsfi~higkeit der Zellen herabgesetzt wird. Die Anwesenheit yon Toluidin-blau im Kulturmedium bei einer nicht tiidlichen Ifonzentration rief ein typisches Ph~nomen hervor, die sogennante 'Polarisation'. Die ver~nderten Zellen, welche in das normale Gewebe eindringen, sind iihnlich diploiden Bindegewebszellen angeordnet, wie sie in einfacher Schicht auf Glas wachsen. Die Ursachenfiir die Polarisation werden diskutiert, bei besonderer Beriicksichtigung der unterschiedlichen Artender Invasion yon BHK21 Py-Zellen und yon BHK21 (unverdnderten) Bindegewebezellen. REFERENCES I. M. AB~.RCRO~mm and J. E. M. HEAYSMAN, Observations on the socialbchaviour of cells in tissue culturc--I. Speed of rnovcmcnt of chick heart fibroblasts in relation to their mutual contacts. Exp. Cell Res. 5, III (1953). 2. M . G . P . STOKER and H. RUBIN, Density dependent inhibition of cell growth in culture. Nature (Lond.) 215, 171 (1967). 3. P. Wsiss, In vitro experiments on the factors determining the course of the outgrowing nerve fibre. 07. exp. Zool. 68, 393 (1934). 4. M. AB~RCRO~mm and E. J. A~mRomz, The surface properties of cancer ceils. A review. CancerRes. 22, 525 (1962). 5. ET. WOLFF and EM. WOLFF, Le role du rndsondphros de rembryon de Poulet dam la nutrition de cellules cancdreuses--II. Etude par la mdthode de la membrane vitelline. 07. Embryol. exp. Morph. 9, 678 (1961). 6. E.J. A~BRos~, Electrophoretic behaviour of cells. Progr.Biophys. molec.Biol. 16, 243 (1966). 7. V. D B ~ D I and G. GASm, Surface mucopolysaccharides of polyorna virus transformed cells, o7. cell comp. Physiol. 62, 23 (1963).

Studies of Tumour Invasion in Organ Culture--I 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.

A. S. G. CUR~S, Cell contacts: some physical considerations. Amer. Nat. 94, 37 (11960). M. STOKSR and I. MACPHERSON, Syrian hamster fibroblast cell line BHK21 and its derivatives. Nature (Lond.) 2039 1355 (1964). M. STOKER,A simple marker technique for cells in culture. Exp. Cell Res. 35, 429 (1964). J. I,EmHTON, 'Neoplastic blockade', a proposed mechanism for the destruction of normal tissue by tumour cells in vitro. Abstracts of Papers, 17th Meeting of the American Tissue Culture Association, p. 21 (1966). S.E. KORNUTH,M. A. STAHMANNand J. W. ANDERSON,Effect of polylysine on the cytology of Ehrlich's ascites tumour cells. Exp. Cell Res. 24, 484 (1961). A. ]~TEvo, A. DE VRIES and A. KATCHALSKY,Interaction of basic poly-amino acids with the red blood cell--I. Combination of polylysine with single cells. Biochim. biophys. Acta (Amst.) 17, 536 (1955). T. RmHARDSON,J. HODGETT, A. LINDm~R and M. A. STAHMANN,Action of polylysine on some ascites tumours in mice. Proc. Soc. exp. Biol. (N.Y.) 101, 382 (1959). M . G . P . STOKER,M. SHEARERand C. O'NEILL,Growth inhibition of polyomatransformed cells by contact with static normal fibroblasts. 07. Cell Sci. 1~ 297 (1966). A. KATCHALSKY,D. ])ANON, A. N~.vo and A. DR VRIES, Interactions of basic polyelectrolytes with the red blood cell--II. Agglutination of red blood cells by polymeric bases. Biochim. biophys. Acta (Amst.) 33, 120 (1959). V. DEFENDI,J. LEHMANand P. KRAEMER,'Morphologically normal' hamster cells with malignant properties. Virology 19~ 592 (1963). M. ABERCROMBIE,J. E. M. HEAYSMANand H. M. KARTHAUSER,Observations on the social behaviour of ceils in tissue culture---Ill. Mutual influence of sarcoma cells and fibroblasts. Exp. Cell Res. 13~ 276 (1957). J. LEIOHTON, Contributions of tissue culture studies to an understanding of the biology of cancer. A review. Cancer Res. 17, 929 (1957).