The Effects of Some Oestrogenic Substances and Cyclic Hydrocarbons On the Rate of Growth of Cells

In view of the fact that the oestrogenic substances play a part in the aetiology of mammary cancer in mice it seems important to investigate the effects of these substances on the growth of cells in general, apart from their action on the reproductive organs.

The existing literature is scanty and may be summarized as follows:

Portes, Lantz & Krajevitch (1939) found that oestradiol enhanced the growth of colon bacilli in vitro. Tavastcherna (1936) obtained stimulation of the rate of growth of heart-tissue cultures with follicular hormone. This is supported by Von Haam & Cappel (1939), who report that oestrin in a dilution of 10⁻⁷ caused slight stimulation of the growth of mouse-heart fibroblasts, as indicated by an increase in the area of the culture and a rise in the mitotic index. Higher concentrations were found to have an inhibitory effect. Maeda (1936), using a preparation called ovaglandol (exact nature unknown), obtained an increased rate of growth of iris epithelium cultured in vitro. Yagi (1937), using rabbit tissues grown in vitro, found ovahormon (nature unknown) to stimulate the growth of uterine, ovarian and testicular tissues, but to inhibit the growth of dental pulp, spleen and bone marrow.

Any comparison of these results is impossible, since dosages, methods of application and the agents used have varied.

The present work was undertaken in order to examine the effects of these substances on a different cell type, namely, a free-swimming Protozoan, and if possible to afford confirmatory evidence to some of the above work.

The animal used was Colpidium sp. It was grown in pure culture in a medium consisting of equal parts of glass-distilled water and Peters’s solution containing 15×10⁸Staphylococcus albus per c.c. The stock cultures were maintained in small glass pots. Throughout the experiments precautions to ensure sterility of the apparatus and reagents used were taken, as it was found that bacterial infection, with consequent effects on growth rate, occurs very easily.

As the basis of each experiment a healthy and rapidly dividing culture was taken and spun at low speed in the centrifuge for a ‘few minutes to remove any debris. Within a short time the animals had swum to the surface and were then removed with a pipette. From this culture samples were then taken.

The method of sampling was to allow the culture to run up a capillary tube until the level of a mark on the tube was reached. The outside was wiped dry and then the contents delivered into a well slide down to the level of a second mark by holding the tube vertically. By this means it was found possible to achieve a high standard of accuracy of sampling.

For each experiment ten samples were taken from the same stock culture into well slides. All received two drops of the Staphylococcus emulsion in Peters’s solution. Five, to be used as the controls, received two drops of glass-distilled water each, and the other five, as the experimentals, received two drops of the oestrogen each, the concentration of the oestrogen being twice that required in the final medium.

The slides were placed in pairs in Petri dishes which contained a layer of moistened blotting paper. All cultures were kept in the incubator at a temperature of 20°C. After 24 hr. the animals in both the control and experimental cultures were killed by the addition of one drop of Susa per slide. This caused the animals to sink to the bottom of the slide. Then the total number of animals per slide was counted, a low-power microscope fitted with a moving stage being used for the purpose.

The oestrogens used were oestrone (Organon), oestradiol benzoate (Organon) and stilboestrol (B.D.H.). They were initially made into 1: 5000 crystalline suspensions in glass-distilled water by dropping an acetone solution into water and then pumping off the acetone. Further dilutions were then made of these suspensions. The oestradiol benzoate and the stilboestrol were found to remain in suspension even at extreme dilution. Oestrone was in solution at a dilution of 10⁻⁵. Higher concentrations always contained some crystals.

Attempts to obtain figures using oestrone in dilutions of less than 10⁻⁵ were unsuccessful. Growth was very erratic, and in many cases the substance proved to be lethal. This was probably due to the fact that crystals were present, and whilst some animals ingested these and received very heavy doses, others only received the effects of that part in solution.

Results for lower concentrations of oestrone, where solution was complete, are given in Table 1. The data have been examined according to the method of Student’s ^‘t’ test. The probability values obtained are shown in the table. (For details of the method see Fisher (1941).)

All P values quoted refer to both tails of the distribution. In the case of 10⁻⁷, since there is an expectation we may use the half value of P, 0·0194, as a test for significance. This indicates that the difference between the two sets of figures is not due to random, sampling errors. In the case of 5×10⁻⁸ there is obviously no significant difference between the sets of figures.

Using oestradiol benzoate and stilboestrol no detailed results could be obtained. In both cases dilutions down to 10⁻⁷ were found to be lethal. Lower concentrations gave anomalous results. Here again this was probably due to the presence of crystals in the medium.

Results of some tests to see whether the proliferative effect of the oestrone was maintained over periods longer than 24 hr. are shown in Table 2. Cultures were seeded with forty organisms picked from a rapidly growing culture, and at the end of 24 hr. subcultures of the same number of animals were made from each slide. The remainder were then fixed and counted. This was repeated day by day. Stimulation was found to occur for a time and then retardation set in, became progressively more noticeable, and eventually led to the death of the animals.

From the foregoing it is seen that oestrone in dilutions of 10⁻⁵ to 10⁻⁷ is effective in stimulating the rate of growth of Colpidium sp. when contact with the reagent is not maintained for more than about 24 hr. No conclusions can be drawn as to the action of oestradiol benzoate or stilboestrol in this respect.

From the present work and the existing literature it is apparent that the oestrogenic substances must be regarded as growth stimulators, this growth-stimulating effect being a general one and not restricted to the tissues of the reproductive organs.

The significance of the reversal of the stimulating effect of oestrone when the observations are extended is not clear. It may be that the oestrone is retained in the cytoplasm of the animals and that a toxic and eventually lethal dose is accumulated. This effect of stimulation followed by retardation may account for some of the discrepancies in other workers’ results.

Many authors have drawn attention to chemical similarities between the cyclic hydrocarbons and the oestrogenic substances. Further, there are similarities in biological behaviour and both have been shown to play a part in the aetiology of cancer. Evidence of the oestrogenic properties of the cyclic hydrocarbons has been obtained by Cook & Dodds (1933) and Perry (1938). Waddington & Needham (1935) and Waddington (1938) have found that carcinogenic and oestrogenic substances are some of the most active compounds in causing the induction of a neural tube in Amphibian gastrulae.

Here an attempt is made to compare the effects of the oestrogenic substances and the cyclic hydrocarbons on the rate of growth of the lower organisms and the isolated tissues of the higher animal. With regard to lower organisms, thirteen papers have been found. In three cases (Goldstein, 1937; Hopper & Clapp, 1939; and Spencer & Melroy, 1940) bacteria were used, but graphs and data are given in only one paper (Goldstein, 1937). Increased proliferation with carcinogenic hydrocarbons was in all cases observed, but not with non-carcinogens.

Wright & Anderson (1938) used the mould F. lini, exposing it to ultra-violet irradiated dibenzanthracene; they found an initial inhibitory period followed by increased proliferation, as measured by increase in weight and utilization of glucose. In three papers yeast was employed, in one (Hollaender, Cole & Brackett, 1939) neither dosage nor data were given; in the other two papers (Cook, Hart & Joly, 1938, 1939) data were given; at 3·6 ×10⁻⁸M of dibenzanthracene for 24 hr. retarded growth was observed, and with 3·6×10⁻⁴ increased rate of proliferation. Two observers (Mottram, 1939 and Wolman, 1939) used Paramecium: increased proliferation based upon good data was found with the carcinogens in dilutions of about 1 ×10⁻⁶, but not with four noncarcinogens. Hammett & Reimann (1935) found increased rate of the production of buds and an increased metabolic rate, when methylcholanthrene or dibenzanthracene in saturated solutions was applied to Obelia geniculata. Owen, Weiss & Prince (1938) observed an enhancement of the rate of regeneration of new heads in the case of the planarian, E. dorotocephala, using saturated solutions of the hydrocarbons, as well as an increase in the number of the regenerated animals. The same workers (Owen et al. 1939) record increase of the rate of growth of rootlet transplants of Pisum sativum. There is therefore complete unanimity in the records of the action of the carcinogenic hydrocarbons on lower organisms, and increased proliferation of the cells under suitable dosage.

Considering, now, the effects upon tissue cultures of mice, rats and fowls, it will be seen that there is little agreement. In many papers neither data nor statistical analysis was given; they are therefore of little importance. Bisceglie & di Grazia (1936) used 3 ·4-benzpyrene on chick-heart tissues in dilution ofiin500 to 1 in 10,000 and found that growth remained luxurious. Larionow, Ivachentzova &.Tchertkova (1938) used 0 ·02 −0 ·05% and obtained transitory stimulation; in another paper (1941) 1·5×10⁻⁶ and 6×10⁻⁶, and observed secondary areas in the cultures proliferating more rapidly than the original tissues, this after $112$ –6 months’ exposure. Lebenzon (1938a, 1938b) used saturated solutions and obtained retardation in rate of growth, but with small concentrations (1940), not specified, acceleration. Lewis (1935), with concentrations up to 0·1 %, found no interference with growth during 24−48 hr.; Magat, Lebenzon & Volkeson (1937) found no effect on either growth or metabolism, but the concentration of dibenzanthracene used was not given. Mauer (1938), in concentration of 1 in 40,000 to 1 in 400,000, found degeneration and death after 4−5 transplants, as did also Timofeevski & Bene-volenska (1940) in concentration of 1 in 1000 to 1 in 1,000,000 for 3 days to 3 months. It is doubtful if any sound conclusion can be drawn from these experiments when no measure of effect was made, other than visual impressions. Degeneration and death as compared with controls is perhaps valuable and for the most part was obtained only with high concentrations of the hydrocarbons.

Dealing now with those observers who give data and statistical analysis, four papers are available: Timofeevski & Benevolenska (1939) obtained no evidence of the stimulation of the rate of growth of rat tissues and bone marrow and rabbit leucocytes during 1 month; the concentration used varied from 1 ×10⁻⁸ for dibenzanthracene to 1×10⁻⁸ for methylcholanthrene. Earle & Voegtlin (1938) likewise obtained no stimulation using methylcholanthrene, 2×10⁻³ to 2×10⁻⁷, on tissues of rats and mice. They continued for 4 months, and at the end found only degeneration and death of the cells. Creech, in two papers (1939, 1940), used fibroblasts from mouse embryos and applied choleic acid compounds of dibenzanthracene, 1 × 10⁻⁶, and of methylcholanthrene, 1×10⁻⁸, for periods up to 70 hr. He obtained increased proliferation as measured by mitotic index and outgrowth measurements; in concentration of ten times the above, retardation was obtained. With the non-carcinogen phenanthrene, 1×10⁻⁴ to 1×10⁻⁶, only retardation occurred.

It must be concluded, therefore, that in the case of tissue cultures increased proliferation, though sometimes found, is by no means observed regularly; however, conditions of dosage varied greatly both in concentration and time of application; for the most part prolonged exposure led to degeneration and death, whilst stimulation was a short-time effect under very low concentrations. In tissue cultures, the cells are suspended in a fluid rich in proteins and lipoids with which the cyclic hydrocarbons are known to combine, and this may well greatly modify the dosage to the cells, whereas in the case of the lower animals, the medium used was usually water or one of simple composition.

Taking into consideration experiments on both the lower organisms and tissue cultures it seems reasonable to draw the conclusion that both the oestrogenic substances and the cyclic hydrocarbons will, for a short time, act as stimulators of growth provided that they are only present in very low concentrations. Higher concentrations are toxic, leading to degeneration and death. Where contact with the chemical is maintained for more than a few days the evidence indicates that there is a divergence between the effects with the lower organisms and with tissue cultures.

The author wishes to express his thanks to Dr J. C. Mottram, Director of Research at the Mount Vernon Hospital, for suggesting this work and for encouragement and advice in carrying it out.