The Effect of Carcinogens. Hormones. and Vitamins on Organ Cultures ILSE LASNITZKI Strangeways Research Laboratory. Cambridge. England Page I. Introduction ...................................................... 80 I1. Principal Culture Methods ....................................... 80 111. The Action of Carcinogens ....................................... 81 1. Effect of 3.CBenzopyrene on Human Fetal Lung ............... 82 83 2. Effect of Tobacco Condensate ............................... 3. Effect of 20-Methylcholanthrene on Mouse Prostate Glands .... 83 4. Effect of Carcinogens on Connective Tissue Growth .......... 86 I V. The Effect of Sex Hormones ..................................... 87 1. Effect on Embryonic Development ............................. 87 2. Effect on Postnatal Organs .................................. 89 3. Effect of Estrogens .......................................... 90 4. Influence of Estrone on the Methylcholanthrene Effect ........ 95 5. Effect of Estradiol on Bone Growth ......................... 96 6. Effect of Androgens ....................................... 96 7. Comparison of the Effects of Estrone. Testosterone Propionate. and 20-Methylcholanthrene on the Mouse Prostate Gland ..... 97 V . The Action of Excess Vitamin A ................................. 98 1. Effect on Bone ............................................. 98 2. Effect of Excess Vitamin A on Epithelium ................. 103 3. Effect of Vitamin A on Other Organs ......................... 106 4. Interaction of Vitamin A with 20-Methylcholanthrene .. ....... 106 V I . The Effect of Vitamin B, on Bone Growth ....................... 107 VII . The Action of Thyroid Hormones ................................. 108 1. Effect of Ambinon on the Chick Thyroid ..................... 108 2. Effect of L-Thyroxine and L-Triiodothyronine on Embryonic 108 Bone ....................................................... VIII . The Effect of Parathyroid Hormone on Bone ..................... 110 I X The Action of Insulin and Growth Hormone ....................... 111 1. Effect of Insulin on Bone Rudiments ......................... 111 2. Effect of Insulin and Growth Hormone on Epidermal Mitosis 112 3. Effect of Insulin on Glycogen Synthesis ....................... 113 4. Effect of Growth Hormone on Embryonic Chick Bone Rudiments ....................................................... 114 X . Action of Corticosteroids ......................................... 114 1. Effect of Cortisone on Lymphocytes ........................... 114 2. Effect of Cortisone on Chick Bone Rudiments ............... 115 XI . Discussion ....................................................... 115 XI1. References ....................................................... 118
I. INTRODUCTION The action of physical and chemical agents in vivo usually sets in motion a complex reaction and interaction of organs and tissues which it is difficult to unravel. The tissue culture technique, on the other hand, allows the effects of such agents to be studied independently of blood or nerve supply. Complicated problems can be broken down into their elements and the question decided whether or not the action of a compound is due to a direct effect on the tissue or mediated through metabolic pathways in vivo. The effects can be detected and the sequence of events followed in detail from an early stage. The comparative ease with which many cultures can be kept under known and identical experimental conditions allows not only a qualitative but also a quantjiative evaluation of effects. In the past, several workers have studied the effect of carcinogens, hormones, and vitamins on cultures consisting of one type of cell. Their results were inconclusive, as the cells used were mainly fibroblasts. Wide variations of concentrations from very low to toxic levels were employed which either elicited no response or produced unspecific damage. For the investigation of specific physiological or pharmacological actions of such compounds organ cultures are much more suitable, because the histological structure and often function of the organ from which they are derived are preserved in vitro and they are thus more closely comparable to tissues in vivo than cultures consisting of one cell type only. The study of the effects of carcinogens, hormones, and vitamins on organ cultures is of very recent date, but it has already yielded much valuable information. I n this chapter the following topics will be reviewed: the direct effects of carcinogenic hydrocarbons on the human fetal lung and mouse prostate gland; the influence of sex hormones on the embryonic development and postnatal behavior of sex and accessory reproductive organs ; the modifications of the carcinogen effect induced by sex hormones ; the changes produced by excess vitamin A in fetal bone and skin and the interaction of this vitamin with a carcinogen; the effect of vitamin BT on fetal bone; the action of thyroid hormones, insulin, and parathyroid hormone on bone growth, on colloid information in the thyroid gland, and on glycogen synthesis; and the effect of cortisone on organ cultures of lymph nodes and bones. 11. PRINCIPAL CULTUREMETHODS The watchglass technique has proved eminently suitable for the cultivation of organized tissue. By the orthodox method devised by Fell and Robison (1929) the explants are grown on the surface of a plasma-extract
VITAMINS AND ORGAN CULTURES
clot in a watchglass enclosed in a petri dish carpeted with damp cotton wool to provide a moist chamber. Modifications of this technique have been introduced by Chen (1954a) and by Shaffer (1956) , who grew cultures attached to lens paper or strips of rayon which are either placed on the clot or float on fluid nutrient medium. These modifications are particularly suitable for the growth of membranes or delicate small organs, as the cultures need not be detached during transfer. Gaillard (1948) and Wolff and Haffen (1952) use embryological watchglasses, the former with a weak coagulum of saline, serum, plasma, and tissue extract, the latter with a mixture of agar with saline, serum, and embryo extract. Gaillard’s and Wolff and Haffen’s methods are sufficient to maintain cultures for short periods, but Fell and Robison’s technique and its modifications provide more favorable conditions for longterm cultivation. 111. THEACTIONOF CARCINOGENS It is not certain whether the induction of tumors by carcinogenic hydrocarbons in susceptible laboratory animals is due to a direct effect or whether these substances have to be metabolized in the organism before exerting their specific action. Various attempts to solve this question by growing fibroblast cultures in the presence of carcinogens have not yielded conclusive results. Larionov et at. ( 1950) observed that 3,4-benzopyrene caused marked inhibition of growth in mouse fibroblast cultures, and when the cultures were subsequently transferred to normal medium they grew more rapidly than their controls. Lebenson and Magat (1937) also noted an acceleration of growth in mouse fibroblasts treated for short periods with 1,2,5,6-dibenzanthraceneand a tendency to invade and destroy normal muscle tissue in culture. They concluded that the cells although not truly malignant had acquired new properties under the influence of the carcinogen. In both experiments no tumors were obtained after injection of treated cultures into mice. Earle et aE. (1943) grew mouse fibroblasts for periods ranging from 6-400 days with I pg. of 20-methylcholanthrene per milliliter of medium, followed by prolonged cultivation in normal medium, and reported a temporary severe growth inhibition and morphological alteration of the cells which assumed features characteristic of tumor cells ; the fibroblasts displayed lateral cohesion, irregular enlargement, and increased granularity of the cytoplasm. The controls, however, showed similar cytological changes, and both sets of cultures induced sarcomas after injection into the same strain of mice from which the original tissue was derived. It was therefore not clear whether the tumors were caused by the carcinogen or were due to a spontaneous malignant transformation of normal cells such as that which has since been reported in Earle’s and
other laboratories after prolonged cultivation of unorganized fibroblasts or epithelial cells (Firor and Gey, 1945; Sanford et al., 1954; Moore et al., 1956). Experiments by Lasnitzki (1951, 1954, 1956) in which organ cultures of the mouse prostate gland and of human fetal lung were grown in the presence of carcinogenic hydrocarbons have yielded more definite results. The choice of these tissues was influenced by the facts that tumors of the prostate are extremely common in man and that in recent years the incidence of human lung cancer has greatly increased. Studies on the response to carcinogens of cultures of these organs are therefore of great practical as well as theoretical interest.
1. Eflect of 3,4-Benzopyrene on Human Fetal Lung The discovery of 3,4-benzopyrene in the condensate of cigarette smoke by Cooper et al. (1954) and by Cooper and Lindsey (1955) has focused the attention on this compound as a possible cause for the increase in lung cancer of heavy cigarette smokers. Various workers have induced lung cancer in laboratory animals by carcinogenic hydrocarbons (Andervont, 1938, 1940 ; Andervont and Shimkin, 1940 ; Shimkin and McClelland, 1949; Horning, 1950), but it is well known that the sensitivity to carcinogens differs widely in various animal species, and it was uncertain whether human lung is susceptible. Lasnitzki (1956) studied the influence of 3,4-benzopyrene on organ cultures of the human fetal lung. Explants grown by the watchglass technique in normal medium showed outgrowth of translucent branching bronchioli in several planes and of unorganized fibroblasts from the periphery of the explant. In sections these cultures resembled the original tissue and consisted of bronchioli embedded in cellular connective tissue (Fig. 1) ; the bronchioli were lined with columnar secretory epithelium which was occasionally ciliated. Cartilage was frequently formed from prechondral areas of the mesenchyme. Addition of 3,4-benzopyrene ( 1 to 6 pg./ml. of medium) for a period of 1 to 4 weeks caused increased proliferation of the lining epithelium after 2 to 4 weeks’ treatment (Fig. 2 ) . This growth was directed toward the lumen and during the early stages was composed of several rows of crowded round or oval cells of the reserve cell type lined by an innermost layer of secretory epithelium. Later the number of cell layers multiplied, often occluding the lumen, and showed irregular nuclear enlargement, polyploidy, abnormal cell divisions, and degeneration and shedding of the secretory epithelium. Treatment with the highest concentration produced not only individual hyperplastic foci but increased general branching of bronchioli
AND ORGAN CULTURES
resembling an adenomatous change. The incidence of hyperplasia expressed as the percentage of treated cultures was 80 to 89% for all three concentrations used, but the first appearance of hyperplastic changes was accelerated and the number of hyperplastic bronchioli and the extent of hyperplasia in them increased with rising dose. With the highest dose, prolongation of treatment beyond 2 weeks caused some epithelial atrophy and the incidence of hyperplasia fell to SO%, indicating that at this dose level the carcinogen has a toxic effect which begins to mask its growthstimulating action. The author considers the hyperplasia together with the cytological abnormalities induced by 3-4 benzopyrene to be a precancerous change.
2. Effect of Tobacco Condensate Findings by Lasnitzki and Kennaway (1955) indicate that other compounds in cigarette smoke may also be carcinogenic. They obtained striking basal cell hyperplasia of the bronchiolar epithelium in lung explants treated for 2 weeks with 500 pg. of tobacco condensate per milliliter of medium, a dose which contains only 0.004 pg. of benzopyrene approximately (Fig. 3). Moreover, tobacco condensate from which the hydrocarbons had been removed induced a similar hyperplasia. Whereas benzopyrene affected the bronchioli independently of size, mainly the larger bronchioli were involved after treatment with the condensate.
3. Eflect of ZO-Methykhohnthrene on Mouse Prostate Glands Lasnitzki (1951) also obtained precancerous changes in mouse prostate glands from adult mice grown in the presence of 20-methylcholanthrene. The living explants in normal medium show outgrowth of differentiated alveoli and a halo of unorganized fibroblasts and in sections are seen to consist of alveoli lined by a row of cuboidal or columnar epithelium surrounded by fibromuscular stroma and cellular connective tissue (Fig. 4). The carcinogen was applied (2 to 4 pg./ml. of medium) for 10 days, and the cultures maintained in normal medium for the same period. Increased proliferation of the lining epithelium was seen after only 5 to 10 days’ treatment, but the hyperplasia became more extensive as cultivation proceeded even when the carcinogen was withdrawn, and after 3 weeks’ growth most alveoli were lined by eight to twelve rows of epithelial cells (Fig. 5 ) . The hyperplasia was usually accompanied by squamous metaplasia of the epithelium which became stratified into a layer of peripheral basal cells followed by an inner layer of prickle cells. The secretory epithelium de-
FIG.1. Section through a culture of human fetal lung grown for 3 weeks in normal medium. It consists of bronchioli lined with one row of epithelium, embedded in cellular connective tissue. FIG.2. Section through a culture of human fetal lung grown in medium containing 3,4-benzopyrene ( 5 pg./ml.) . Note hyperplasia of the bronchiolar epithelium and paucity of the connective tissue. (From Lasnitzki, 1956.) FIG. 3. Section through a culture of human fetal lung grown for 2 weeks in medium containing tobacco condensate (500 VgJml.). (Lasnitzki, unpublished.) FIG.4. Section through a culture of a mouse prostate gland grown for 3 weeks in normal medium, showing alveoli surrounded by fibromuscular stroma. (From Lasnitzki, 1955b.) FIG. 5. Section through a culture of a mouse prostate gland grown for 11 days in medium containing 20-methylcholanthrene (4 pg./rnl.) and carried on a further 10 days in normal medium. Note extensive hyperplasia of the alveolar epithelium and paucity of stroma. (Lasnitzki, 1955b). FIG.6. Section through a mouse prostate culture grown for 11 days in medium containing 20-methylcholanthrene (4 pg./ml.) and carried on a further 10 days in medium containing excess vitamin A. The hyperplasia is much reduced as compared with that shown in Fig. 5. (From Lasnitzki, 1955b.)
CARCINOGENS, HORMONES, VITAMINS AND ORGAN CULTURES
generated, was shed, and was replaced by flat cornifying elements. Mitosis in such alveoli was confined to the basal layer. Apart from this orderly stratification, changes of a more anaplastic nature were observed, viz. the appearance of foci or irregularly arranged crowded cells showing nuclear enlargement and polyploid divisions. Mitosis was increased three- to fourfold after 4 days’ treatment and remained constant until the withdrawal of the carcinogen after which it fell to the control level. A second wave of mitotic activity followed, possibly owing to an increased growth potential of the now-altered epithelium. Lasnitzki and Pelc (1957) correlated the morphological changes and stimulation of mitosis with DNA synthesis by means of autoradiography. Comparison of uptake of CI4-adenine in epithelial nuclei of treated and untreated explants, respectively, showed that the uptake was slightly increased in treated cultures, but this increase followed rather than preceded mitotic stimulation. The authors suggest that the carcinogen may not primarily influence DNA synthesis but affects mitosis directly by forcing cells into division which would otherwise have proceeded to differentiation.
4. Eflect of Carcinogens on Connective Tissue Growth Whereas epithelial growth in lung and prostate cultures is stimulated by benzopyrene and methylcholanthrene, that of connective tissue is inhibited (Lasnitzki, 1951, 1956). The migration of fibroblasts from the explant is suppressed after a few days’ treatment, and the stroma within becomes gradually poorer in cells and fibers which finally disappear. Cartilage is not formed in treated lung explants. Autoradiographs of explants treated with C14-adenine (Lasnitzki and Pelc, 1957) show that the fibroblasts cease to incorporate the tracer after a week’s growth in the presence of methylcholanthrene. EarIe and Voegtlin (1938, 1940) also report inhibition of growth and the appearance of cell degenerations in cultures of rat and mouse fibroblasts grown with the same carcinogen, but Creech (1940) obtained mitotic stimulation in mouse fibroblasts treated with a dose of 0.015 pg. methylcholanthrene per milliliter. These results suggest that the differential response of epithelium and connective tissue may represent not a qualitative difference in effect but a difference of degree, and that the stimulating dose for fibroblasts may be roughly one-hundredth that for epithelium if the concentrations used by Lasnitzki are taken as the growth-stimulating dose for epithelium. Even so, the striking difference in susceptibility of these two elements in both human and mouse tissue are very interesting and may be important factors in the histogenesis of epithelial tumors.
A N D ORGAN CULTURES
IV. THE EFFECT OF SEX HORMONES
1. EBect on Embryonic Development Experiments suggest that the development and maintenance of the male accessory reproductive organs during fetal life depend on testicular hormone. Thus Jost and Bergerard (1949) and Jost and Bozic (1951) describe retrogression of the Wolffian ducts in cultured fragments of fetal castrated rats, and Raynaud and Frilley (1946, 1947, 1950) report that the accessory reproductive glands of the rat are smaller or absent after destruction of the testes by X-rays. This theory was put to test by Price and Pannabecker (1956), who studied the role of the testis and sex hormones on retention or loss of Wolffian ducts grown as organ cultures. Reproductive tracts from rat embryos removed 15 to 18 days post copulation, which consisted of genital tubercle, posterior urogenital sinus, and Wolffian and Miillerian ducts (Fig. 8A), were explanted with either both testes or one, without testes, with ovaries, or with addition of either testosterone propionate or estradiol to the culture medium (Fig. 7). The culture period ranged from 1 to 6 days. In cultures containing both testes or one which in some experiments had been detached and replaced in its original position the Wolfian ducts persisted as continuous tubes and developed primordia of seminal vesicles and prostatic buds (Fig. 8 B ) . Removal of testes or their substitution by ovaries was followed by reduction of the diameter of the duct, loss of lumen, shedding of epithelial components, and degeneration of the surrounding sheath of mesenchyme (Fig. 8C). Addition of testosterone propionate (8 to 40 pg./ml. of medium) prevented this retrogression. With one exception this effect was similar in explants from all ages, but in tracts derived from older embryos explanted with one testis and spread out into a V shape the continuity and diameter of the duct depended on its distance from the testis. Addition of estradiol to the culture medium (0.02 pg./ml., approximately) led to a retrogression of the Wolffian ducts in explants derived from younger fetuses; in several older explants a partial retention or cystic dilatation was observed. Seminal vesicles were developed only in explants from older fetuses and were larger on the testis side in tracts containing one testis only. Testosterone propionate stimulated their growth and became more efficient with advancing age of the fetus from which the explants were taken. In the absence of testes or the presence of ovaries the seminal vesicles failed to develop. Addition of estradiol, on the other: hand, stimulated their growth
in four out of eleven explants. In all series of experiments the prostate glands or prostatic buds were seen in older explants at advanced stages of embryonic development, but in general they were more numerous and larger in the presence of testes or testosterone. These results indicate that the fetal testes elaborate a hormone which in culture reaches the other parts of the tract by diffusion, maintains the Wolffian ducts, and stimulates the formation of primordia of the seminal vesicle and prostatic buds; it further suggests that in vivo these organs depend for their maintenance and development on testicular hormones.
DAYS POST COITUM 17.5
FIG.7. Diagrammatic representation of rat reproductive tracts explanted at
174- days. A, tract at time of explantation; B, C, D, E, F, tracts cultured for 4 days. B, with both testes present; seminal vesicles and prostate glands developed.
C, without testes ; the Wolffian ducts regressed and no seminal vesicles appeared. D, with one testis; results as in B. E, with one testis, placed at a greater distance from the opposite side of the tract; the Wolffian duct on the gonadless side regressed slightly and seminal vesicles were smaller or absent. F, with no testes but testosterone added to the clot; the wolfian ducts were retained and seminal vesicles and prostatic buds developed. S, seminal vesicle ; M,Miillerian duct ; W,Wolffian duct, P, prostate. (From Price and Pannabecker, 1956.)
Evidence that the female gonad also gives off a diffusible hormone is presented by Wolff (1953) and by Wolff and Haffen (1952), who explanted gonads of the duck before they had reached sexual differentiation. If gonads of opposite sexes were grown in close contact, the female gonad developed into an ovary but the male gonad formed an organ of intermediate character. A similar feminizing influence on the undifferentiated male gonad was seen after addition of estradiol benzoate.
CARCINOGENS, HORMONES, VITAMINS A N D ORGAN CULTURES
Wolff (1953) demonstrated the role of testosterone in the regression of the male Miillerian ducts of the chick. Male ducts explanted after sexual differentiation usually regress in vitro, an effect attributed by the author to their previous exposure in vivo to the male sex hormone. Undifferentiated ducts, on the other hand, follow the neutral development and differentiate into female structures (Fig. 9), but contact with testicles in culture or, addition of testosterone propionate cause their rapid necrosis (Fig. 10). Scheib-Pfleger ( 1955) found a striking loss of the total nitrogen content of regressing Mullerian ducts treated with testosterone and claims that the hormone directly increases the activity of proteolytic enzymes. These enzymes can be inactivated by exposure of the ducts to ultrasonic waves, according to Lutz and Lutz-Osterlag ( 1956).
FIG.8. A . normal rat fetus at 17+ days with Wolffian duct dilatations and median Miillerian ducts B. 19+-day-old explant with testes showing Wolffian duct dilatations and median utriculus prostaticus. C. 21-t-day-old explant without testes ; the wolffian ducts have retrogressed and no seminal vesicles developed. (From Price and Pannabecker, 1956.)
Wolff and Wolff (1953) describe the effect of estradiol and testosterone on the development of the duck syrinx. This organ explanted without addition of hormones differentiates according to the male type; it becomes asymmetrical and forms broad rings of cartilage on the left and slender rings on the right side which are joined by a ventral cartilaginous protuberance (Fig. 11). Both estradiol and testosterone induce female differentiation ; the organs retain their symmetrical shape, the cartilage remains at the mesenchymal or precartilaginous stage or only develops thin translucent rings, and the ventral protuberance is missing (Fig. 12).
on Postnatal Organs
Tissue culture experiments have shown that the sex hormones not only influence the development of fetal organs but also modify directly the growth and differentiation of postnatal sex and accessory reproductive organs.
3. Effect of Estrogens In most sexually mature rodents the vaginal epithelium undergoes cyclic changes under the influence of endogenous estrogens and progesterone. During diestrus or in the prepuberal state the epithelium consists of a row of cuboidal cells beneath an outer layer of mucified cells. During estrus or after injections or local application of estrogen the basal cells multiply and differentiate into squamous keratinizing epithelium and the superficial mucified cells are shed.
FIG.9. Development of the male Miillerian duct of the chick explanted before sexual differentiation and grown in vitro for 5 days. The duct is lined by a row of columnar epithelium and surrounded by healthy cellular connective tissue. FIG. 10. Comparable explant of male miillerian duct grown with testosterone. Note necrosis of epithelium and connective tissue and disappearance of the lumen. (From Wolff, 1953.) FIG.11. Transverse section through a duck syrinx following male development in vitro. Note proliferation of cartilage on ventral side. FIG.12. Transverse section through a duck syrinx which developed according to the female type in vitro. The cartilage is much thinner than shown in Fig. 11. (From Wolff and Wolff, 1953.)
Vaginal explants from prepuberal rats and mice have, with one exception, been found to form stratified keratinizing epithelium on explantation in normal medium, a response which resembles the precocious keratinization of skin explants in culture. Addition of estrogens to the medium hastens the squamous changes.
CARCINOGENS, HORMONES, VITAMINS A N D ORGAN CULTURES
Dux (1941) was the first to demonstrate this direct effect of the hormone. H e grew vaginal explants from young rats, either adding estrogen to the culture medium or using serum from animals which had received injections of estrogen. Untreated cultures developed squamous cornifying epithelium, including a layer of flat cells containing keratohyalin, beneath the keratin. In explants treated with estrogen there was a striking increase in the number and size of the keratohyalin granules accompanied by precocious but incomplete keratinization. Kahn (1954) found that the addition of estrone to the medium of vaginal explants from rats 3 to 4 weeks old (1.3 pg./ml.) considerably hastened the stratification and cornification of the prepuberal epithelium. Thus, within 1 to 2 days the explants formed stratified squamous epithelium under the influence of the hormone, whereas the control cultures underwent the same process between the fourth and fifth days in culture. Similar results were obtained by Lasnitzki (unpublished experiments) in mouse vaginas derived from mice 2 to 3 weeks old. She observed, however, that mitosis was usually increased in the basal layer of estrone-treated explants as compared with that of the controls. Biggers et al. (1956) studied the effect of eleven different estrogens including estrone, diethylstilbestrol, equilin, and 3,17p-estradiol on the vaginal epithelium of very young mice, 1 to 9 days old. I n contrast to the findings of the other investigators the epithelium of their control explants preserved the typical prepuberal state (Fig. 13) throughout the 3day period of the experiment and consisted of a row of cuboidal cells adjoining the basal membrane with an outer layer of mucified elements. This difference in behavior may be related to the age of the animals from which the cultures were derived. Thus in the older explants the keratinization in vitro may be promoted by a longer exposure to estrogenic hormones before explantation. In cultures treated with estrogens for 4 hours the stratum germinativum had multiplied to four to eight layers of cuboidal cells beneath a row of swollen mucified cells, giving a faintly positive reaction with the periodic acid-Schiff method ( P A S ) . In living explants the basal cells in the treated explants showed some birefringence at this stage. A t 48 and 72 hours the cell layers had increased to eight to twelve and formed keratin at the periphery (Fig. 14), and the outermost mucified epithelium was sloughed. All eleven estrogens produced a similar response, but there was a marked gradation of changes in different explants in each group and a variation in response along the length of the explant. Dose-response curves showed that the median effective dose was between 3.9 x 10-6 and
FIG.13. Section through a vagina explant from a 6-day-old mouse grown for 4 hours in normal medium showing simple epithelium. FIG.14. Section through a similar explant after 48 hours’ growth in medium containing 10-2 pg. of estrone, showing superficial cells, thick keratinized layer, stratified layers, and stratum germinativum. (From Biggers et al., 1956.) FIG.15. Section through an explant of a prostate gland from a 6-week-old mouse grown for 3 weeks in medium containing estrone (2 w / m l . ) showing hyperplasia and squamous metaplasia of the alveolar epithelium. (From Lasnitrki, 1954.)
FIG.16. Section through an explant from a prostate gland from a 6-month-old mouse grown for 3 weeks with the same dose of estrong as explant shown in Fig. 15, showing atrophy of the epithelium and increase of the fibromuscular stroma. (From Lasnitzki, 1955a.) FIG.17. Section through an explant of a prostate gland from a 6-week-old mouse grown for 10 days in medium containing testosterone propionate (50 pg./ml.). The structural differentiation of the gland is retained. FIG.18. Section through a prostate gland from a 6-month-old mouse grown for 10 days with the same dose of testosterone propionate as the gland shown in Fig. 17. Note hyperplasia of the alveolar epithelium. (From Lasnitzki, 1955a).
CARCINOGENS, HORMONES, VITAMINS AND ORGAN CULTURES
7.8 X pg., which is one-tenth of the median effective dose required to produce comparable changes in the animal by intravaginal application. The effects of estrogen on cultures of the uterus of rats and mice were investigated by Verne (1935) and Gaillard and de Jongh (1938). Verne reports increased proliferation of rat uterine epithelium grown in plasma from rats previously injected with estrogen. Gaillard and de Jongh found a differential response of muscle, endometrium, and epithelium which depended on the concentration of the hormone used. After a dose of estrone of 1/5000 i.u. the muscle layer thickened appreciably, and the epithelium, which in untreated cultures became necrotic, was preserved. After 1 i.u. the muscle layers diminished in size but epithelium and endometrial connective tissue became hypertrophic. Price and Pannabecker (1956) found that estradiol stimulates the growth of the seminal vesicle of the rat during embryonic development. Gaillard and de Jongh (1939) obtained a similar stimulating effect of estrone on the seminal vesicle of young mice. The optimum growthpromoting concentration differed for the various components of the gland ; a dose of 1/50 i.u. caused maximal thickening of the smooth muscle layer to three to four times the control size, but higher doses were necessary to stimulate growth of the connective tissue of the tunica propria and of the epithelium which occasionally became multilayered. The role of the sex hormones in the production of prostatic tumors had been recognized in recent years. Both the benign enlargement of the gland and prostatic cancer have been attributed to a disturbance of the androgen-estrogen balance during advancing age. The question arises whether the changes in the hormonal balance are a primary cause or whether mature cells respond in a different manner from young cells to alterations in their environment. Lasnitzki (1954), who studied the direct effect of estrone on prostate glands derived from mice of two different ages, obtained different effects under identical experimental conditions. Estrone (2 to 4 pg./ml.) added to the medium of glands from young sexually immature mice induced mitotic stimulation, multiplication of basal cells leading to hyperplasia, and squamous metaplasia of the alveolar epithelium (Fig. 1.5) after 10 to 20 days’ treatment. In cultures kept in normal medium after withdrawal of the hormone, mitosis fell to a normal level and the hyperplasia was gradually reversed. Abnormal mitotic figures could often be distinguished in meta- and anaphase, showing clumping, breakage, and dislocation of chromosomes. Similar abnormalities were described by Mollendorff (1942) in rabbit fibroblasts treated with large doses of estradiol and are probably caused by an unspecific toxic effect of the hormone.
CARCINOGENS, HORMONES, VITAMINS AND ORGAN CULTURES
Addition of estrone to the medium of glands derived from mice 5 to 6 months old, on the other hand, caused flattening and atrophy of the alveolar epithelium accompanied by hypertrophy of the fibromuscular stroma, particularly after the higher dose ( 4 pg./ml.) (Fig. 16). Most of the alveoli became surrounded by a wide, dense zone of collagenous and muscle fibers, and stromal elements filled the interalveolar spaces, absorbing and replacing necrotic alveoli in the process. The reversal of the estrone effect in older glands may be related to the length of exposure to androgenic hormones prior to explantation ; thus epithelial atrophy and stromal growth in the older glands may represent an antagonistic response due to a sensitization of the prostatic epithelium by longer exposure to androgenic hormones in contrast to glands derived from young animals.
4. Influence of Estrone on the ~ e t ~ ~ Z c h 5 ~ n t hEfiect rene Estrogens used in the therapy of prostatic cancer in man usually cause a temporary regression of primary and secondary tumors (Huggins et al., 1941; Kahle et al., 1942, 1943). In experimentally induced glandular carcinomas of the mouse prostate the administration of estrogen is likewise followed by a retardation of growth (Horning, 1949). Two different modes of action may be responsible for this effect : (1) Estrogen restricts the production of androgenic hormones by inhibiting the output of gonadotropic hormone in the pituitary, and the cancer cells still dependent on androgenic hormones break down; or (2) estrogen damages the cancer cells directly (Burrows, 1949). Lasnitzki (1954) obtained a direct modification of the effects of methylcholanthrene by estrone added either simultaneously with the carcinogen or after its withdrawal. The influence of the hormone differed according to the age of the glands at the time of explantation. Simultaneous addition of estrone and 20-methylcholanthrene to the medium of young glands induced a similar incidence and degree of hyperplasia as with the carcinogen alone, but the cultures underwent squamous metaplasia earlier and more extensively. . In older explants the hormone counteracted the effects of the carcinogen by reducing the incidence of hyperplasia and stimulating stromal growth. Administration of estrone to young glands previously treated with the carcinogen was followed by a reduction of hyperplasia and an increase in squamous metaplasia. Frequently, widespread destruction of the hyperplastic epithelium either before or after squamous transformation took place. In the former, the cells enlarged, the cytoplasm became vacuolated, and the nuclei pycnotic; in the latter, squamous cells and cornifying elements were shed
as one mass, leaving an almost empty alveolus behind. This degeneration resembles that described by Fergusson and Franks (1953) in cases of prostatic cancer following estrogen therapy, a similarity which suggests a direct effect of the hormone on the cancer cells apart from a possible indirect action via the pituitary gland.
5. Effect of Estradiol on Bone Growth I n vivo estrogens inhibit the growth in length of bones. They cause a decrease in thickness of the cartilage, an increase of the subepiphyseal osseous spiculae, and early union of the epiphyses by precocious ossification. Goyena (1955), who grew femurs of 7- to %day chick embryos with various concentrations of estradiol, found that the addition of the hormone either made no difference, or in the higher concentrations ( 1 : 32,000) slightly but significantly increased the lengths of the bones during the first 5 days in culture. H e concluded that the growth inhibition seen in Vivo is not due to a direct effect of the hormone but may be operated via the pituitary gland. 6 . Effect of Androgens The direct action of male sex hormones has been examined by Demuth ( 1941) and Lasnitzki ( 1955a). Demuth reports a twofold increase in area of epithelial sheets from the rat seminal vesicles treated with testosterone propionate (50 pg./ml.) , but the connective tissue was not influenced. Lasnitzki grew mouse prostate glands from mice 4 to 6 weeks old and 5 to 6 months old in the presence of testosterone propionate, and as with estrone obtained a different response depending on the age of the glands. Unlike the organ in situ, young prostate glands grown in normal medium showed some dedifferentiation in culture. The lining epithelium of the ducts was reduced in height and the folding lost; new alveoli formed at the periphery of the explant were straight and lined with cuboidal epithelium. The addition of testosterone propionate (50 pg./ml.) to the medium of such young organs preserved their structural differentiation, and the explants closely resembled glands of the same age in vivo (Fig. 17). Prolongation of the treatment, however, led to considerable proliferation of the lining epithelium. At 3 weeks in most alveoli the epithelium consisted of four to six rows of densely packed cells showing many normal cell divisions. The growth proceeded inward toward the lumen, and the cells gradually increased in size as a result of cytoplasmic enlargement ; the nuclei remained constant in size and often became indented. The interalveolar stroma was not affected by the hormone, and squamous changes were absent.
CARCINOGENS, HORMONES, VITAMINS AND ORGAN CULTURES
Addition of the same dose of the hormone to the medium of the older glands was followed by epithelial hyperplasia shortly after the beginning of treatment (Fig. IS). After 10 days’ growth with the hormone the epithelial cells lining the alveoli had multiplied intensively ; in most alveoli the growth was directed outward away from the lumen which rarely became occluded. As in young glands squamous metaplasia was absent and the connective tissue was not affected by the hormone. Increase in cell size could also be distinguished, but unlike that seen in the young glands this was caused by an irregular enlargement of the nuclei. Cell division was abundant and was often abnormal : breakage and aberration of chromosomes, bridge formation in anaphase, and polyploidy were common. Prolongation of the treatment caused atrophy and widespread necrosis of the alveolar epithelium. These results suggest that explants from young mice require a continuous supply of androgenic hormones for their maintenance which is not available in sufficient quantities from the culture medium, but older explants are either independent of it or need a smaller concentration to preserve differentiation. The finding that a dose of hormone which was sufficient to maintain young glands in culture induced hyperplasia in the older organs indicates that the mature epithelium has not become refractory but rather more sensitive to the hormone. This is supported by the appearance of degenerative changes in the older organs after periods of application exceeding 10 days. Apart from the mitotic abnormalities and the irregular increase in cell size which were observed in the older glands only, the difference in response between the older and younger organs is mainly one of degree, since the latter undergo epithelial hyperplasia if the concentration of the hormone is increased by prolonged treatment. This is in contrast to the effects of estrone which vary qualitatively with age.
7. Comparison of the Effects of Estrone, Testosterone Propionate, and Z U - ~ e t h ~ Z c h o ~ u n ~ on h r ethe ~ e Mouse Prostate Gland
All three substances stimulate mitosis of the prostatic epithelium leading to the formation of a multilayered stratified epithelium. After methylcholanthrene this change occurs independently of the age of the explanted organ, whereas after estrone the effect is seen in young glands only. With testosterone propionate larger doses are required to obtain hyperplasia in the young gland than in the older organ. Squamous metaplasia takes place after estrone and methylcholanthrene but is absent after testosterone. Withdrawal of the carcinogen is followed by further proliferation with high mitotic activity, but after withdrawal of estrone the hyperplasia is gradually reversed.
Methylcholanthrene always inhibits the growth of the connective tissue. Estrone, on the other hand, does not influence the stroma in young glands and enhances its proliferation in older explants ; and testosterone does not affect it at either age. The hyperplastic epithelium induced by estrone or testosterone in young glands shows orderly stratification and a gradual and regular increase in cell size owing to enlargement of the cytoplasmic area ; the carcinogen also produces foci of irregularly arranged crowded cells. The cellular enlargement seen after methylcholanthrene and testosterone in the older organs is irregular and due to an increase in nuclear size. Mitotic abnormalities after estrone show dislocation and breakage of chromosomes but not polyploidy which is prevalent in carcinogen-treated explants and also occurs in older glands after application of testosterone. The cytological abnormalities which are characteristic of precancerous changes are found both after application of methylcholanthrene and in older glands after testosterone treatment. This resemblance suggests that under certain conditions the hormone itself may act as a carcinogen. Thus in cases where the usual senile diminution in the androgen level is missing malignancy may be caused by prolonged exposure of sensitive senescent epithelium to doses of androgen which are no longer physiological for the altered epithelium. V.
THEACTIONOF EXCESS VITAMINA 1. EjJect on Bone
Experiments on young animals have shown (Mellanby, 1938, 1939, 1947) that vitamin A determines the shape and texture of certain bones by controlling the activity of osteoblasts and osteoclasts. Spontaneous fractures of the long bones were observed in young rats fed on a diet containing excessive amounts of vitamin A (Davies and Moore, 1934; Strauss, 1934 ; van Metre, 1947) or after large doses of pure vitamin A (Wolbach and Bessey, 1942) ; Moore and Wang, 1943 ; Pavcek et aE., 1945). These gross effects were associated with striking histological changes. Strauss (1934) observed reduction of endochondrial ossification, and Irving (1949) a reduced deposition of bone. Wolbach and Bessey (1942) and Wolbach (1947) showed that in the limb bones of young rats and guinea pigs niaturation and degeneration of the cartilage cells were accelerated, leading to an early closure of the epiphyses. Experiments by Barnicot (1950) suggested a direct effect of the vitamin. H e reported advanced osteoclastic absorption in parietal bone grafts from young mice implanted with pellets of crystalline vitamin A into the cerebrum of littermates. The conditions, however, were not closely com-
VITAMINS AND ORGAN CULTURES
parable to those in the hypervitaminotic animals, as the grafts were exposed to very large amounts of the vitamin. I n tissue culture, in vivo conditions can be imitated more easily by growing organ cultures of bone in a medium containing concentrations of the vitamin similar to those present in the blood of hypervitaminotic animals. Early bone rudiments of the chick grow and differentiate in vitro in a fairly normal way (Fell and Robison, 1929; Fell, 1951). They increase to three to four times their original length within 7 to 10 days’ cultivation and differentiate into hypertrophic cartilage in the shaft, followed distally by areas of proliferative flattened cells and smaller epiphyseal cells. The chondroblasts in the shaft are separated by thick partitions of matrix which stains metachromatically with toluidine blue. Osteoblasts are arranged in regular rows at the inner side of the periosteum. Fell and Mellanby (1952) explanted the humerus, radius, ulna, femur, and tibia of 5- to 7-day chick embryos and late fetal and infant mice in media to which had been added doses of vitamin A ranging from 900 to 3200 i.u. per 100 ml. of plasma and found that the vitamin profoundly interfered with growth and differentiation of the bone explants. The growth of the rudiments was slowed down and finally arrested; constrictions appeared between shaft and epiphyses leading to a detachment of the terminal ends of most rudiments (Figs. 19, 20). The vitamin severely affected the cartilage matrix which shrank and softened into a meshwork of loosely arranged fibers. It lost its metachromasia and basophilia, and its affinity for van Gieson stain was increased. The changes spread from the periphery toward the center, probably owing to the diffusion of the vitamin into the tissue, and were more severe in the shaft than in the epiphyses, The chondroblasts, however, remained healthy but were often smaller than in the controls, and several cells became enclosed in one capsule ; periosteal ossification was only slightly inhibited. Mouse bone rudiments also enlarge in normal medium but to a smaller degree than those from the chick. Endochondrial ossification usually ceases in vitro, but some periosteal bone is deposited. The vitamin produced a much more severe effect on the mouse bone than on those of the chick; the cartilage matrix dwindled rapidly, leaving the chondroblasts naked. The extent of damage in both chick and mouse rudiments depended on the length of exposure to the vitamin and on the stage of development at the time of explantation. Thus the younger primordia were more drastically changed than older bones, but there was also a difference in susceptibility of the various rudiments at the same stage of development. Herbertson (1955), who confirmed Fell and Mellanby’s findings on
FIG. 19. Femur from a 6-day chick embryo cultivated in normal medium for 9 days. Note the differentiation of the cartilage into hypertrophic region in the shaft, zone of flattened cells, and epiphyses. The cartilage matrix is strongly basophilic. FIG.20. Femur from the opposite side of the same chick as shown in Fig. 19 after 9 days’ cultivation in medium containing excess vitamin A. Note the small size of the explant and the loss of basophilia from the matrix; the condylar end is nearly detached from the shaft. (From Fell and Mellanby, 1952.)
FIG.21. Control skin from a 13-day chick embryo grown for 6 days in normal medium. A fairly thick stratum corneum has developed. FIG.22. Explant from the same experiment as the control shown in Fig. 21 after 6 days’ cultivation in medium containing excess vitamin A. The superficial cells have degenerated and are being sloughed; the rest of the epidermis is being reorganized into secretory epithelium. FIG.23. Similar skin explant to that shown in Fig. 22 after 12 days’ growth in +A medium followed by 4 days in normal medium. Note the intense secretory activity of the cells which have now developed into typical goblet cells. (From Fell, 1957.)
FIG.24. Epiphyseal cartilage of a 7-day chick embryo femur grown for 4 days in normal medium. FIG.25. Comparable area of a femur to that shown in Fig. 24, grown in medium containing triiodothyronine. The cartilage is better developed, the cells are larger, and the matrix is more profuse. (From Fell and Mellanby, 1956.) FIG.26. Parietal bone from nearly full-term mouse cultivated in Vitro for 6 days. FIG.27. Similar explant to that shown in Fig. 26 grown in contact with chick embryo parathyroid gland tissue showing intense absorption of bone. (From Gaillard, 1955b.)
CARCINOGENS, HORMONES, VITAMINS AND ORGAN CULTURES
chick bones, found that the femur was most sensitive to the action of the vitamin and the ulna and radius were least affected. She reports that the changes caused by the vitamin were arrested after transfer to normal medium but were not entirely reversed. There was no further elongation of the shaft, but sometimes it increased slightly in width. The epiphyses, on the other hand, enlarged again, although they had been shrinking in +A medium. In the shaft the cartilage matrix was not re-formed, nor was its metachromasia restored, and its hypertrophic cartilage came to resemble osteoid tissue. There was new formation of periosteal bone which merged imperceptibly with the remaining ground substance of the cartilage which, like the bone, took up the pink of chromotrope. The cells showed all gradations from young osteocytes in the osseous layer to recognizable hypertrophic cells in the shaft interior. Vitamin A added to normal plasma had a much more rapid and potent action than plasma from hypervitaminotic animals, a difference which is correlated with the extractability of the vitamin in the two plasmas. All the added vitamin could be recovered by shaking the plasma with petrol ether, but in the natural high-A plasma the protein had first to he denatured with ethanol before it could be extracted (Fell, 1954). It is uncertain how the action of the vitamin on the bone is brought about, but findings by Fell and Mellanby (1952) indicate that the presence of living cartilage cells is necessary to mediate its effect. Thus, mouse bones in which the cells had been killed by heating to 45" C. were not affected by the vitamin. Fell et al. (1956) showed that the vitamin modified the sulfate metabolism of the cartilage. Using autoradiography, they compared the uptake of labeled sulfate (NA2S0436) in chick bone rudiments grown in normal medium and with high doses of vitamin A. In the control explants, a few the tracer was concentrated in the cells ; hours after the addition of at later times large amounts were found in the matrix. Transfer from labeled to unlabeled medium did not cause any loss of the sulfate. I n +A medium the uptake was either diminished or absent in the peripheral regions of the shaft and epiphyses, depending on the duration of growth in the presence of excess vitamin A ; transfer to unlabeled medium was followed by appreciable loss in the shaft which preceded the disappearance of metachromasia. The authors tentatively suggest that the vitamin causes the cells either to elaborate a new enzyme or to produce larger amounts of one which is already present and which converts the chondroitin sulfate of the ground substance into a more soluble form. Only when sufficient quantities have been released from the cells into the matrix will the exist-
AND ORGAN CULTURES
ing chondroitin sulfate affected ; hence there is a time lag between the disappearance of the tracer and of metachromatic staining.
2. Eflect of Excess Vitamin A on Epithelium Experiments ilz vivo suggest that vitamin A influences the type and direction of differentiation in epidermis and mucous epithelium. Lack of the vitamin was found to produce squamous keratinizing epithelium in organs which are normally lined by mucus-producing ciliated cells as in the respiratory tract of rats and guinea pigs (Wolbach and Howe, 1925, 1928), O n the other hand, high doses of the vitamin given either systematically (Studer and Frey, 1949) or locally (Sabella et al., 1951) caused a thickening of the epidermis in rats. Fell and Mellanby (1953) studied the direct effect of the vitamin on organ cultures of ectoderm of 7-day chick embryos. At this stage the epidermis in vivo is composed of one basal layer of columnar cells and one superficial row of flat cells. Cultivation in vitro on normal medium caused precocious stratification and keratinization (Fig. 21), but addition of the vitamin (ZOO0 to 3000 i.u. per lo0 ml. of plasma) interfered with this process and produced a mucus-secreting, often ciliated epithelium resembling that of the nasal mucosa (Figs. 22,23). After 9 days with excess vitamin A the superficial flattened cells were replaced by cuboidal or columnar cells showing red granules within their free border and vacuoles staining blue with Azan and a brilliant red with mucicarmine. In the lower strata cavities appeared, owing to the retraction or shrinkage of cells. These became filled with periodic acidSchiff (PAS) positive material by secretion from the surrounding healthy cells. Mitosis, which in squamous epithelium was confined to the basal cells, could be found in all cell layers including that of the secretory epithelium. Transfer to normal medium accelerated the changes at first; after 4 days the superficial cells formed a typicz! mucous membrane consisting of ciliated and goblet cells which secreted profusely. But the metaplastic changes were not stable, and soon no more secretory cells were formed and new squamous keratinizing epithelium differentiated beneath the mucous membrane which was finally shed. The connective tissue of the dermis was adversely affected by the vitamin. It was less fibrous, was poorer in cells, and often became separated from the epithelium by spaces filled with fluid and traversed by thin fibrous strands. This study was extended to human fetal skin by Lasnitzki (1956). Epidermis from 3- to 4-month fetuses explanted on normal medium differ-
entiates like that of the chick into squamous keratinizing epithelium. It develops also a keratohyalin layer which is absent in chick ectoderm. In vitamin A-treated cultures the epidermis formed several strata of cuboidal swollen cells showing an irregular outline at the cuticular surface ; the keratohyalin layer and keratin were missing. In the upper layers and at the margin, small cavities filled with PAS-positive material and individual cells containing granules staining bright red with mucicarmine could be distinguished. The basal membrane was often edematous and lost its continuity, and the dermis became severely disorganized. Transfer to normal medium did not accentuate the metaplastic changes ; instead the whole of the mucoid epithelium including the stratum germinativum was sloughed and new squamous keratinizing epithelium re-formed from intact hair follicles. The influence of the developmental stage on the response to vitamin A was studied by Fell (1957) in the ectoderm from 13- and 18-day chick embryos. At these ages the epidermis consisted of a stratified squamous keratinizing epithelium before explantation. The author found that the 13-day skin was affected in the same way as that of a 7-day chick. In the 18-day explants keratinization was suppressed, but mucous metaplasia was less frequent and less extensive than in the younger explants. The type of change depended on the stage of differentiation the cells had reached when exposed to the vitamin. Cells near the stratum corneum prevented from forming keratin produced a material which stained with P A S but not with mucicarmine. Less differentiated elements became distended with vacuoles and showed PAS-positive but mucicarmine-negative material in their peripheral cytoplasm ; and the deeper undifferentiated layer frequently, but not always, formed areas of columnar cells secreting mucus which stained with both P A S and mucicarmine. The metaplastic change was correlated with a striking difference in the uptake of radioactive sulfate (Fell et at., 1954). Glucksmann, Howard, and Pelc (unpublished experiments) found that in mice injected with radioactive sulfate autoradiographs of the various organs showed an active uptake of S35 in the mucous epithelia and very little in the epidermis. Siinilarly, autoradiographs of explants of chick ectoderm treated with S35 showed intense blackening over the mucus-secreting epithelium of cultures grown with excess vitamin A and very little deposit over the squamous keratinizing epithelium of the controls. In both types of cultures there was marked uptake by the connective tissue surrounding the epithelium. On the other hand, the uptake of radioactive cystine was clearly diminished over the mucoid epithelium of +A-treated cultures a s com-
CARCINOGENS, HORMONES, VITAMINS AND ORGAN CULTURES
pared with that of squamous keratinizing epidermis of the control cultures (personal communication). Weiss and James (1955) demonstrated that brief exposure to a high concentration of the vitamin produced changes in chick ectoderm similar to those produced by continuous administration of lower doses. Cell suspensions of ectoderm from young chick embryos were exposed for brief periods to a concentration of vitamin A fifty times as high (0.06 i.u.) as the continuous dose used by Fell and Mellanby and were then explanted in normal medium. Untreated suspensions reaggregated to form hollow cysts or solid pearls built of concentric layers of squamous keratinizing epithelium, the outermost stratum corresponding to the basal layer, the innermost to the free cuticular order. Cultures exposed briefly to the vitamin also developed cysts or pearls but showed no trace of concentric stratification ; instead of the usual progressive flattening, all cells remained cuboidal, their nuclei were well preserved, and no keratin was formed. In cultures receiving “booster” treatment, i.e. immersion in vitamin A between and preceding transfers, the cysts became lined with typical mucus-secreting goblet cells with large masses of mucus in their apical cytoplasm. Since brief exposure to the vitamin produces the same changes as continuous treatment, the authors contend that the vitamin does not alter the metabolism of the cells gradually during their growth in vitro but acts as an inductive agent which switches the mechanism of differentiation into an alternate pathway. The changes are interpreted as “the after effect and indicator of a crucial event of relatively short duration,” probably due to surface changes. Lasnitzki and Greenberg ( 1956), however, demonstrated the presence of the vitamin in similarly treated chick ectoderm. Minute amounts of vitamin A can be detected by the method of Popper and Greenberg (1941), which is based on the fact that when exposed to an intense source of ultraviolet at long wavelengths the vitamin shows a green fading fluorescence. Organ cultures of chick ectoderm were immersed in Tyrode solution containing 0.05% vitamin A alcohol, a concentration slightly lower than that used by Weiss and James, and after thorough rinsing were explanted onto normal medium. Examination of the cultures after 5 days’ growth under the ultraviolet microscope at 3400 to 3800 A. revealed the green fading fluorescence characteristic of vitamin A. I n addition the amount of the vitamin in the tissue was determined by means of Beckman’s spectrophotometer, and it was found that 20 mg. of tissue contained approximately 50 pg. of vitamin A. It may, therefore, be concluded that the vitamin is
taken up rapidly and stored as such in the tissue and that its effect is due rather to a continuous action than to an “induction.”
3. Efect of Vitamin A on Other Organs Fell (1954) studied the action of vitamin A on the differentiation and secretion of the mucous membranes of the embryonic gut and nose and on two other ectodermal derivatives, the retina and the otocyst. Explants of gut and nose were not affected by the vitamin; both control and treated cultures differentiated well and secreted profusely. The retina usually grew better than its control and differentiated normally in a concentration of the vitamin which caused mucous metaplasia of the ectoderm. The otocyst also differentiated normally but was much smaller than its control. This retardation of growth appeared to be a secondary effect due to the complete suppression of the cartilaginous capsule by the vitamin. Lasnitzki (1955b) found that the vitamin did not appreciably influence the growth and differentiation of the adult mouse prostate but caused a slight increase in secretion and deposition of PAS-positive material in the secretory cells. 4. Interaction of Vitamin A with 2U-Methylcholunthrene This carcinogen directly stimulates the proliferation of the alveolar epithelium in mouse prostate glands in vitro (Lasnitzki, 1951). The hyperplastic epithelium always undergoes squamous metaplasia, and the secretory elements at the lumen of the alveolus degenerate and are shed. The addition of vitamin A simultaneously with methylcholanthrene did not interfere with the increased proliferation of the alveolar epithelium, caused by the carcinogen, but it prevented the squamous changes and preserved the secretory cells (Lasnitzki, 1955b). The layers of prickle and flattened cells were replaced by columnar and cuboidal elements with basophilic cytoplasm and connected by tonofibrils with each other and with the innermost layer of secretory elements ; they thus represented a “hybrid” between columnar epithelium and a stratum spinosum. In contrast to the squamous alveoli where mitosis is confined to the basal layer, these hybrid cells remained potentially dividing and mitotic figures couId be observed in all strata. Hence the hyperplasia was usually more extensive than in explants treated with the carcinogen alone. Administration of the vitamin after withdrawal of methylcholanthrene not only suppressed the squamous changes but antagonized the action of the carcinogen and brought about a partial reversal of the hyperplasia, so that there were fewer hyperplastic alveoli per treated explant and in these the
VITAMINS AND ORGAN CULTURES
hyperplasia was less extensive. The remainder of the alveoli appeared normal and were lined by one row of columnar secretory cells (Fig. 6). This reduction of the hyperplasia by the vitamin may have been due to its increased uptake by the cells in the absence of competition from the carcinogen. It was not caused by a direct damaging effect on the epithelium such as that seen after estrone (Lasnitzki, 1954), nor was it due to an inhibition of cell division, since it was found (Lasnitzki, 1 9 5 5 ~ )that excess vitamin A increased the mitotic rate in chick fibroblasts ilz vitro. It is suggested that the effect is an indirect one and that the vitamin controls the ratio of cell multiplication and cell differentiation and thus restores the balance of the two processes. These results indicate that the vitamin A level may influence the growth rate of epithelial tumors. A slight increase above the normal may accelerate tumor growth, but higher concentrations may retard it. OF VITAMINBT VI. THE EFFECT
Carnitine, first discovered in meat extract ( GulewiFsch and Krimberg, 1905), has since been indentified as vitamin BT, present in yeast extract and many animal tissues (Fraenkel and Blewett, 1947; Fraenkel, 1951, 1953,1954). The naturally occurring compounds (CH3) 3
N + -CH2 - C H O H - CH2 - COO -
as well as the synthetic razemic dicarnitine (CH3)3
N - CH2 - C H O H - CH2 -CO
HOOC - CH2 - CH - CH2 - N
Cl are essential metabolites for the growth and development of the larva of Tenebrio molitur (Carter et aE., 1952a, b ; Leclercq, 1954). LiPbecq-Hutter (1956) studied the effect of dicarnitine on the growth of tibias and femurs of 7-day chick embryos, explanted by the watchglass method in medium depleted in embryo extract. Addition of dicarnitine to the culture medium in concentrations of 2.5 or 5 nig./l. promoted the growth of both rudiments. The bones increased in length by one- to twothirds over that of the controls, and the periosteum of the treated cultures was usually thicker. The degree of growth promotion varied in inverse proportion with the initial size of the rudiments on explantation; i.e., the stimulation was greater in smaller explants, and vice versa.
ILSE L A S N I T Z K I
Mitotic counts in control explants fell to 7% of those seen in the rudiments in Vivo, probably owing to the depletion of the medium in embryo extract. Addition of dicarnitine greatly attenuated this fall ; in treated cultures the number of cell divisions was 45% of those found in vivo. This rise in mitotic activity was due to a true stimulation and not to interference by the compound with the distribution of mitotic phases which remained normal. The results suggest that dicarnitine partially replaces the growthpromoting substances missing in the depleted medium and that the vitamin in its naturally occurring form may in fact be one of the growth-promoting factors present in embryo extract. This is supported by results by Fraenkel (1953), who found that 12-day chick embryos contain 44 to 88 pg. of carnitine per gram of dry weight.
VII. THE ACTIONOF THYROID HORMONES
1. Effect of Ambinon on the Chick Thyroid Gaillard (1955a) examined the influence of a thyrotropic hormone, Ambinon, on the devqlopment of the chick thyroid in culture. In normal medium, explants of the thyroid from &day chick embryos formed juxtanuclear colloid after 2 days’ growth. Follicle formation was slightly retarded in vitro as compared with the development in Viv5. Addition of Ambinon stimulated colloid production but did not influence follicle formation, a result which, according to the author, indicates that the two are independent processes.
2. Effect of L-Thyroxine and L-Triiodothyronine on Embryonic Bone It has long been recognized that the thyroid influences the growth and structure of bone and that both deficiency and excess of thyroxine produce skeletal abnormalities. Clinical evidence shows that osteoporosis is frequently found in hyperthyroidism (Bernard, 1927 ; Plummer, 1928 ; Aub et al., 1929; Hunter, 1930). Skeletal growth was influenced experimentally in laboratory animals by administration of thyroxine. Dott ( 1923) described acceleration of epiphyseal activity in hyperthyroidic kittens and puppies. H e found an increased rate of proliferation in the cartilage and in the matrix evidence of hastened maturity. Silberberg and Silberberg ( 1938, 1940) reported that the hormone accelerated the age changes in the skeleton of young guinea pigs and mice. Results on the effect of thyroxine on the development of chick bones in ovo have been conflicting. Willier (1924) found that grafting of thyroid into 7- to 10-day chick embryos considerably inhibited the growth of the limb bones, while Beyer (1952) obtained an in-
CARCINOGENS, HORMONES, VITAMINS AND ORGAN CULTURES
crease in the wet and dry weight of chick embryos from eggs injected with thyroxine before incubation. Previous studies on the effect of thyroxine on tissue cultures were mainly concerned with the action of the hormone on fibroblasts and yielded inconsistent and variable results which did not clarify the mechanism of action of the hormone (Vogelaar and Ehrlichmann, 1936 ; Ebeling, 1924 ; van Haam and Cappel, 1940). Fell and Mellanby (1955) grew chick long-bone rudiments with a concentration of thyroxine of the same order as that present in hyperthyroidic human serum and found a direct effect of the hormone under conditions where all systemic factors had been eliminated. The rudiments were treated with the hormone at different stages of development; in one series the primordia were at the blastematous or precartilaginous stage; in another, older series cartilage had already been formed. The hormone accelerated the normal hypertrophy of the cartilage cells but also caused regressive changes. The same concentration was stimulatory or harmful, depending on the stage of development and also on the rate of differentiation of the particular bone used. Young rudiments were more easily stimulated than the older primordia; the humerus responded best at all ages and showed acceleration of hypertrophy leading to rapid extension of the bone and temporary rise in growth rate over that of the controls. The radius, which differentiates more slowly than the other bones, was occasionally stimulated at the blastematous stage but always in the older explants. In the tibia and femur the maturation of the cartilage was sometimes accelerated in the young bones but only very slightly and temporarily at the older stage. The older bones were more liable to the toxic action of the hormone, which caused degeneration of the cartilage cells leading to inhibition of growth, Experiments in which a much higher dose than that present in hyperthyroidic serum was used demonstrated the great difference in susceptibility between the various rudiments. Thus tibia and femur were more adversely affected than the wing bones of which the radius was the least susceptible. Although all other bones showed a varying degree of growth inhibition, the radius actually increased in length. The dependence of the effect on the developmental stage probably explains, at least in part, the conflicting results by Willier (1924) and Beyer (1952) on chick embryos in o m . Beyer, who obtained accelerated growth of the embryo, treated it at an early stage when the tissue was more easily stimulated, whereas Willier used older embryos which were adversely affected by the hormone. Fell and Mellanby (1956) also studied the direct effect of triiodo-
thyronine ( T 3 ) , discovered by Gross and Leblond ( 1951), and synthesized and identified in human blood by Gross and Pitt-Rivers (1952, 1953). Like thyroxine it accelerated -the normal hypertrophy of embryonic chick bones in culture (Figs. 24, 25). The precocious hypertrophy spread abnormally fast from the diaphysis to the cells of the proliferative zone which ceased to divide. This cessation of mitosis in the growing zone combined with degeneration of the hypertrophic cells in the shaft led to a marked inhibition of growth in certain rudiments. As with thyroxine, the extent and degree of the effect depended on the age and rate of development of the individual bones. Thus the rapidly developing leg bones were more severely affected than the humerus ; the slowly differentiating ulna and radius were stimulated in the young explants and only slightly retarded in bones from older chicks. A direct comparison of the effect of T 3 with that of L-thyroxine (T4) showed that T 3 was four times as potent in reducing the growth of the femur and tibia. This is in good agreement with results on animals and human beings. To mention only two examples: T3 was found to be four to seven times as effective in increasing the O2consumption in rats (Tomich and Woollett, 1954) and ten times as active in depressing the uptake of radioactive iodine in human beings (Starr and Liebhold-Stoeck, 1953). Lawson ( 1956) correlated the developmental differences of chick tibia and radius and their difference in response to T3 with their rate of accumulation of total nitrogen. During a 6-day culture period in normal medium the tibia elongated more rapidly than the radius and showed a faster rate of accumulation of total N. Treatment with T 3 induced an increase in length of the tibia during the first 24 hours followed by retardation of growth, whereas the growth of the radius was stimulated for 4 days and then fell to the control level. Wet weight and total nitrogen in the treated tibia were less and those of the radius more than in the controls. OF PARATHYROID HORMONE O N BONE VIII. THEEFFECT Gaillard (1955b,c) has shown that a hormone released by the parathyroid gland in vitro directly affected the resorption of bone rudiments in culture. He grew parietal bone from nearly full-term mouse fetuses in close contact with parathyroid gland tissue from newly born human beings, from young mice, and from 9- to 12-day chick embryos, and observed intense resorption of bone. After the first day the ground substance became highly PAS-positive, which according to Engel (1952) is due to a polymerization of glycoproteins. After 3 to 4 days’ growth the PAS reaction in the ground substance changed and varied from highly positive to negative, indicating further chemical alterations in the matrix.
CARCINOGENS, HORMONES, VITAMINS AND ORGAN CULTURES
During the first 4 days of cultivation the changes took place without the aid of osteoclasts, but after 5 to 9 days osteoclasts appeared and speeded up the resorption of the remaining bone (Figs. 26, 27). New bone formation was often seen at the inner side of the periosteum, suggesting that the primary steps of bone development were not influenced by the hormone. Cinematographic observations of living bone explants grown in contact with parathyroid fragments showed increased motility of connective tissue cells between the bone trabeculae. Both resorption and reconstruction of bone occurred during the first 3 days of combined cultivation, but gradually resorption predominated. O n the fourth day large osteoclasts appeared and settled down close to dense parts of the matrix which then began to disappear. Cell-free culture fluid from parathyroid cultures when added to the medium of parietal mouse bone induced a similar resorption of the ground substance as that seen in combined cultures, but the changes appeared earlier. This result indicates that a humoral agent which is gradually given off by the parathyroid tissue is responsible for the effect seen.
IX. THEACTION OF INSULINAND GROWTH HORMONE I . The Effect of Insulin on Bone Rudiments Chen (1954b) studied the direct effect of insulin on developing chick bone rudiments in culture. The hormone (0.16 to 0.0016 i.u./ml. of medium) interfered with the normal differentiation of the cartilaginous shaft and at the same time promoted the growth of the epyphysis and the surrounding soft tissue. After 5 to 6 days’ treatment the ends of the rudiments greatly enlarged and the shaft became considerably bent. This effect was due to imperfect differentiation of the cartilage ; the zone of flattened cells was poorly defined or absent so that the shaft merged into the epyphysis, the cartilage cells were smaller than normal and irregular in shape, and their capsules were ill-defined. The amount of periosteal bone was variable and seemed to depend on the stage of development of the particular bones at the time of explantation. Some rudiments were completely devoid of bone and remained a uniform mass of cartilage. The absence of embryo extract in the medium accelerated and enhanced the effect of the hormone. The reduction in shaft length became apparent after 2 days’ treatment, and the disproportion of shaft and epyphyseal growth resulted in a striking distortion of shape. This result suggested that the extract may partially inactivate the hormone or its active principle. This theory was put to the test by Chen. H e compared the response of bone rudiments grown with insulin and extract which had been incubated together prior to cultivation with that of control explants grown with
the two substances incubated separately and found that insulin incubated with extract had become ineffective. In the author's view this inactivation may be due to an insulinase which had been identified in tissue extracts by Mirsky et al. (1949) and may also be present in embryo extract. The mode of action of insulin on bone is obscure. Duraiswami (1950) suggested that the hormone affects the production of chondroitin sulfate, but it seems likely that the action is far more complex, since insulin not only interferes with the differentiation of cartilage but at the same time promotes the growth of the epyphysis.
2. Eflect of Insulin and Growth Horwone on Epidermal Mitosis The mitotic rate in mouse epidermis was found to depend on the rate of entry of glucose into the cells and on its subsequent conversion to energy (Bullough, 1951). Cori ( 1950) demonstrated that insulin stimulates carbohydrate metabolism by facilitating the glucokinase reaction, whereas the pituitary growth hormone acts as inhibitor. Bullough (1954b) studied the influence of both hormones on short-term cultures of mouse epidermis kept in phosphate-buffered saline. Mitotic activity in vitro was independent of factors which in the animal depress the rate of cell division, such as poor nutrition, overcrowding, or severe shock. Neither did the effect of the diurnal cycle on mitotic activity, normally seen in vivo, persist in vitro. The estrogen level of the animals, was, however, reflected in the mitotic rate' in vitro ; thus, ear fragments from mice in proestrus, at a time when the estrogen level was high, showed twice as many divisions as fragments from mice in diestrus and three times as many as ear fragments from castrated animals. The androgen level, on the other hand, did not significantly influence the rate of mitosis in vitro (Bullough, 1954a). The addition of insulin (1.25 to 12.5 pg./ml.) to the medium of ear fragments of male mice almost doubled the mitotic rate if glucose was used as substrate but was ineffective when either L-lactate or fructose was substituted for glucose. L-Lactate or fructose alone, on the other hand, induced a significantly higher mitotic rate than glucose. From this evidence the author concluded that insulin stimulates a reaction in glucose metabolism which takes place before the pyruvate stage and is most probably the reaction : glucose + glucose-6 -phosphate. Addition of pituitary growth hormone to the medium depressed the mitotic rate in the ear fragments if glucose was used as substrate but did not inhibit mitosis if either L-lactate or fructose was used. According to the author this suggests that the growth hormone selectively inhibits the glucokinase reaction.
AND ORGAN CULTURES
Growth hormone and insulin proved to be antagonists under the experimental conditions used. The lowering of the mitotic rate by the growth hormone was partially offset by large doses of insulin, and mitotic stimulation due to insulin could be inhibited by large amounts of the growth hormone. 3. Eflect of Insulin on Glycogen Synthesis More evidence that insulin directly influences carbohydrate metabolism in culture was presented by Sidman (1956b), who grew adipose tissue of the rat in the presence of this hormone. Brown adipose tissue differentiates in culture in the absence of nervous or endocrine influences and is determined while still part of the loose mesenchyme (Sidman, 1956a). The author explanted mesenchyme of the interscapular region of late fetal rats and found that after a few days the cells withdrew their amoeboid processes and aggregated into islands clearly demarcated from the surrounding mesenchyme. The cells increased in size and deposited small droplets of lipid in their cytoplasm which enlarged and coalesced ; eventually typical multilocular adipose cells were formed. This occurred faster and more profusely in mesenchyme grown in serum than in Parker’s synthetic medium No. 770. The addition of insulin to the medium (0.04 to 4.0 yg./ml.) stimulated the synthesis of glycogen, hastened the rate of lipid deposition, and increased the survival time of the tissue in culture. Under otherwise similar conditions the effect was greater in cultures grown in serum than in synthetic medium, and the author contends that the serum may contain substances which potentiate the action of the hormone. After 3 to 4 days glycogen synthesis decreased even in the presence of freshly added insulin, but this fall could be deferred by addition of glucose. The author claims that the decrease in glycogen synthesis was not caused by exhaustion of essential metabolites but represented a shift in the pathways of synthetic activity in the direction of lipogenesis. That the tissue retains the ability to synthesize glycogen for longer periods in vitro is borne out by the fact that insulin stimulates glycogen synthesis when added to cultures after various periods of growth in normal medium. This also indicates that the hormone acts independently of cofactors bound in the fresh tissue. Stadie and his colleagues (Stadie, 1954; Haugaard et al., 1954; Haugaard and Marsh, 1952) demonstrated the rapid binding of isotope-labeled insulin and Chayen and Smith ( 1954) that of fluorescein-labeled insulin to many tissues, including adipose tissue. Sidman confirmed their results and obtained evidence that the hormone was not only rapidly bound but
also remained effective for several days in vitro. Thus, cultures immersed for short periods in Tyrode solution containing insulin and subsequently grown in normal medium showed a stimulation of glycogen synthesis similar to that produced in explants exposed continuously to the hormone. The effects of insulin in vitro were identical with those observed in vivo in the brown adipose tissue of the rat after denervation (Sidman and Fawcett, 1954) a finding which suggests that denervation increases the sensitivity of the adipose tissue to circulating insulin.
4. Efect of Growth Hormone on Embryonic Chick Bone Rudiments As stated above, Bullough (1954a) reported that the pituitary growth hormone inhibited mitosis in fragments of mouse ears. Hay (personal communication), who studied the effect of this hormone on bone growth in vitro, on the other hand, obtained a slight but significant growth promotion of chick embryo tibia and femur explanted by the watchglass method in serum without embryo extract. Addition of 0.18 mg. of growth hormone per milliliter of medium increased the wet weight by 8.5% and the dry weight by 6.5% over those of the controls.
X. ACTION OF CORTICOSTEROIDS 1 . Effect of Cortisone on Lymphocytes It has long been known that exposure of animals to stress or shock reduces the number of lymphocytes in the body. This effect is mediated through the adrenal cortex, since it fails to appear in adrenalectomized animals but can be reproduced by injection of adrenal cortical extract. Schrek ( 1949, 1951) demonstrated that small doses of 17-hydrocortisone and cortisone significantly shortened the survival time of thymus lymphocytes in suspension but that deoxycorticosterone was inactive ; increasing the concentration did not increase the effect substantially. Trowell (1953) confirmed and extended this work. H e studied the action of cortisone on organ cultures of lymph nodes from young rats, cultured in a serum-saline medium, and used the percentage of pycnotic lymphocytes as the criterion of damage. Normal 2-day cultures of lymph nodes usually showed 1 to 2% of pycnotic cells. Addition of cortisone to the medium significantly increased this percentage but did not affect the reticulocytes or macrophages. The number of pycnotic cells rose with the concentration of cortisone and with the length of exposure, and increased proportionally to the log of time during the first 24 hours. Thus a concentration of 0.1 pg. had a very slight effect, but 10 pg. killed 45% of lymphocytes within 2 days. Purine and pyrimidine bases added simultaneously with cortisone slightly
CARCINOGENS, HORMONES, VITAMINS A N D ORGAN CULTURES
diminished the cytotoxic action, while potassium, insulin, and testosterone had no influence. Lymph nodes cultured for 3% hours with cortisone (3.3 pg./ml.) and transferred to normal medium for the remainder of the 48-hour culture period showed only a very slight increase of pycnosis over the normal. This indicates that cortisone does not cause immediate and irreversible damage but that the length of exposure is important.
2. Eflect of Cortisone on Chick Bone Rudiments Cortisone inhibits growth in most laboratory animals including chick embryos, and this is associated with striking developmental abnormalities in cartilage and bone. Buiio and Goyena (1955) showed that this effect was due to a direct action of the hormone on the bone. They explanted femurs of 7-day chick embryos in the presence of hydrocortisone and cortisone (0.025 pg./ml.). After 48 hours’ growth with either hormone, the treated rudiments showed a growth inhibition of 37% of the controls. The difference between experimental and control cultures increased until the fifteenth day and then stayed constant. The hormones caused no changes in the ground substance and no decrease in the size of the cartilage cells, and the authors claim that the inhibition is due to a reduced rate of proliferation of chondro- and osteoblasts. As with lymphocytes the deleterious effect on the bone was reversible. Cultures maintained in normal medium after a 96 hours’ cultivation with the hormones recovered rapidly and reached the control size after 2% days in normal medium. XI. DISCUSSION The substances studied so far may be divided into two main categories: those which produce functional changes and merely reproduce in vitro the happenings inside the organism, and those which interfere with normal development, e.g. by disturbing the balance between cell proliferation and cell differentiation or by misdirecting differentiation. The action of sex hormones on their target organs, that of the parathyroid gland on bone rudiments, and the influence of insulin on glycogen synthesis belong to the first group ; carcinogens which induce increased epithelial proliferation and squamous metaplasia, and excess vitamin A which severely affects the cartilage matrix or causes mucous metaplasia of the skin, to the second category. Under certain circumstances the sex hormones also cause unphysiolo$ical changes resembling those of the carcinogenic hydrocarbons, and the effects of thyroxine and T 3 can either be physiological or toxic, according to the dose and developmental stage
of the bone rudiments used. All the compounds examined produce their specific effects by direct action. This does not, however, preclude the possibility that in vivo, in addition to exerting a direct action, they may be metabolized, potentiated, or partially inactivated before reaching their target. The work of Wolff and his colleagues and of Price and Pannabecker demonstrates clearly that both testes and ovaries produce hormones during fetal life which remain effective in culture and produce effects identical with those of synthetic androgens or estrogens. Similarly, the parathyroid gland continues to function after explantation and secretes a hormone which causes bone resorption in rudiments grown either in contact with the gland or in the culture fluid of parathyroid fragments. Most compounds induce their specific effects shortly after the beginning of treatment, an indication that they are taken up rapidly by the growing tissues. The results of Weiss and James and of Lasnitzki and Greenberg with vitamin A and those of Sidman with insulin show that both substances are stored in the tissue and remain effective for several days in culture. It is interesting to speculate as to which cells in an organized tissue are primarily affected by the various compounds investigated. Analysis of the changes induced in epithelia such as those of skin, vagina, or prostate suggest that the stratum germinativum in skin and vagina and the undifferentiated reserve cells in the prostate are the primary targets. Carcinogenic hydrocarbons and the sex hormones, for instance, stimulate cell division in the basal cells and thus induce abnormally high proliferation. In the prostate niethylcholanthrene and estrogen in addition force the newly formed cells into an abnormal pattern of differentiation and transform mucus-secreting elements into squamous keratinizing epithelium. This change does not occur in the cells which are already differentiated but only in generations of newly formed cells. The opposite effect, viz. mucous metaplasia of a squamous keratinizing epithelium is seen in the embryonic skin after exposure to excess vitamin A. Again the full effect is observed in the undifferentiated basal cells, but experiments by Fell in which explants form older embryos were used showed that cells already in the process of keratinization but not completely cornified could still be deflected from their normal course and formed a material which had certain staining properties of both keratin and mucin. The mucous change seemed to be irreversible ; this was demonstrated by Fell and Mellanby, who found that on return to normal medium the mucous cells were not converted into squamous elements but were shed, while new keratinizing epithelium was regenerated from the stratum
CARCINOGENS, HORMONES, VITAMINS A N D ORGAN CULTURES
germinativum. It is tempting to interpret the metaplastic changes as proof that the basal cells remain permanently bipotential and capable of modulation (Weiss, 1949). Although this concept may be new for the skin, it must be remembered that the cyclic changes of the vaginal epithelium occurring normally under the influence of the sex hormones give another striking example of the flexibility of the stratum germinativum and its capacity for modulation. The role of the connective tissue in the production of physiological or pharmacological changes has been little explored, and experimental evidence on this point is, so far, rather scanty. Coujard (1943) and Champy et al. (1950) found that epithelia from the uterus, vagina, and seminal vesicle were stimulated in culture by the addition of estrogens and other hormones only if the connective tissue was present and that isolated epithelial cells failed to respond. They maintained that the effect is mediated through nerve endings of the autonomic nervous system present and functioning in the intact connective tissue. Other observations also point to the importance of the connective tissue for the survival and maintenance of epithelial cells and their ability to respond to environmental changes. Thus, in organ cultures of skin or vagina, detachment of the epithelium from its connective tissue base invariably led to its degeneration, and the degree of response in the intact “organ” was usually correlated with the condition of the connective tissue. It was found, for instance, that in vaginal explants the response to hormones was most marked in areas where the underlying connective tissue was well developed and rich in cells (Lasnitzki, unpublished observations). The reason for this beneficial influence is not clear and is well worth investigating. Apart from a possible role of autonomic nerve endings as mediators of stimuli put forward by Coujard and his colleagues, other mechanisms may play a part. The connective tissue may store the compounds added to the medium, release them gradually to their targets, and thus act as barrier and prevent or mitigate overdosage. On the other hand connective tissue cells may facilitate the nutrition of the epithelium by conditioning the medium and breaking down proteins to smaller metabolic units which can be utilized more readily. So far, the criteria of effect have been mainly morphological ones. It must be remembered, however, that structural changes are only the final manifestations of chemical changes, and future work should try to correlate the two. Application of cytochemical methods and autoradiography to organ cultures have already yielded much valuable information and are being extended to an increasing number of problems. The introduction of modern microchemical methods now allows us to
study metabolic changes in tissue cultures with greater accuracy than was hitherto possible. Thus the effect of carcinogens, hormones, or vitamins on the uptake and utilization of amino acids can be ascertained by chromatography and by the addition of labeled compounds to the culture medium. The Warburg technique and its modifications have so far been employed to measure respiration and glycolysis in cultures consisting of one cell type only, but it is hoped that the near future will see the adoption of these methods to the study of organized growth. The use of synthetic chemically defined media for organ cultures will greatly facilitate such metabolic studies by the more rigid control of experimental conditions which would thus be ensured. XII. REFERENCES Andervont, H. B. (1938) Public Health Repts. (U.S.) 63, 1647. Andervont, H. B. (1940) J . Natl. Cancer Inst. 1, 135. Andervont, H. B., and Shimkin, M. B. (1940) I. Natl. Cancer Inst. 1,225. Aub, J. C., Bauer, W., Heath, C., and Ropes, M. (1929) I . Cltn. Invest. 7 , 97. Barnicot, N. A. (1950) J . Anat. 64, 374. Bernard, A. (1927) Miinch. med. Wochschr. 74, 432. Beyer, R. F. (1952) Endocrinology 60,497. Biggers, J. D., Claringbold, P. J., and Hardy, M. H. (1956) J. Physiol. (London) 191, 497. Bullough, W. S. (1951) Proc. Roy. SOC.B138, 562. Bullough, W. S. (1954a) Exptl. Cell Research 17, 176. Bullough, W. S. (1954b) Exptl. Cell Research 17, 186. Buiio, W.and Goyena, H. (1955) Proc. SOC.Exptl. Biol. Med. 89, 622. Burrows, H. (1949) “The Biological Action of Sex Hormones,” p. 356. Cambridge Univ. Press, London. Carter, H. E., Bhattacharyya, P. K., Weidman, K. R., and Fraenkel, G. (1952a) Arch. Biochem. Biophys. 36, 241. Carter, H. E., Bhattacharyya, P. K., Weidman, K. R., and Fraenkel, G. (1952b) Arch. Biochem. Biophys. 38, 405. Champy, C., Coujard, R., and Demay, M. (1950) Ann. endocrinol. ( P a r i s ) 11, 195. Chayen, J., and Smith, R. H. (1954) Biochem. I . 68, 8. Chen, J. M. (1954a) Exptl. Cell Research 7 , 518. Chen, J. M. (1954b) I. Physiol. (London) 126, 148. Cooper, R. L., and Lindsey, A. J. (1955) Brit. J . Cancer 9, 304. Cooper, R. L., Lindsey, A. J., and Waller, R. E. (1954) Chem. 6 Ind. (London) (Rev.) 1418. Cori, C. R. (1950) Intern. Congr. Biochem. 1st Cong. Cambridge, Engl. Abstr. Communs.
Coujard, R. (1943) Bull. biol. France et Belg. 77, 120. Creech, E. M. H. (1940) Am. J. Cancer 39, 149. Davies, A. W., and Moore T. (1934) Biochem. I . 28, 288. Demuth, F. (1941) A c f a Brevia Neerl. Physiol. Pharmacol. Microbial. 11, No. 61, 116. Dott, N. M. (1923) Quart. J. Exptl. Physiol. 28, 241.
CARCINOGENS, HORMONES, V I T A M I N S AND ORGAN CULTURES
Duraiswami, P. K. (1950) Brit. Med. J . 11, 384. Dux, C. (1941) Ann. endocrinol. (Paris) 2,39. Earle, W.R., Nettleship, A., Schilling, E. W., Stark, Th. H., Straus, N. R., Brown, M. F., and Shelton, E. (1943) J. Natl. Cancer Inst. 4,213. Earle, W.R., and Voegtlin, C. (1938) Am. I. Cancer S4,373. Earle, W.R., and Voegtlin, C. ( 1940) Public Health Repts. (U.S.) 66, 303. Ebeling, A. H. (1924) Compt. rend. SOC. biol. 90, 1449. Engel, M. B. (1952) A.M.A. Arch. Pathol. SS, 115. Fell, H.B. (1951) Methods in Med. Research 4,235. Fell, H . B. (1954) Ann. N.Y. Acad. Sci. 68, Art. 7, 1183. Fell, H.B. (1957) Proc. Roy. SOC.Bl40,242. Fell, H.B., and Mellanby, E. (1952) J. Physiol. (London) 116,320. Fell, H. B., and Mellanby, E. (1953) I. Physiol. (London) 119,470. Fell, H. B., and Mellanby, E. (1955) J. Physiol. (London) 127, 427. Fell, H.B., and Mellanby, E. (1956) J , Physiol. (London) 159,89. Fell, H. B., and Robison, R. (1929) Biochem. J. 2S, 767. Fell, H.B., Mellanby, E., and Pelc, S. R. (1954) Brit. Med. J . 22, 611. Fell, H. B., Mellanby, E., and Pelc, S. R. (1956) J. Physiol. (London) lS4, 179. Fergusson, J. D., and Franks, L. M. (1953) Brit. J. Surg. 40,2. Firor, W.M.,and Gey, G. 0. (1945) Ann. Surg. l21,700. Fishrnan, W.H. (1951) Vitamins and Hormones 9,213. Fraenkel, G. (1951) Arch. Biochem. Biophys. 34,457. Fraenkel, G. (1953) Biol. Bull. 104,359. Fraenkel, G. (1954) Arch. Biochem. Biophys. 60,486. Fraenkel, G., and Blewett, M. (1947) Biochem. J. 41, 469. Gaillard, P.J. (1948) Symposia SOC.Exptl. Biol. No. 2, 139. Gaillard, P.J. (1955a) Acta Physiol. et Pharmacol. Neerl. 4, 108. Gaillard, P. J. (1955b) Koninkl. Ned. Akad. Wetenschap. Proc. C68, 279. Gaillard, P.J. (1955~) Exptl. Cell Research Suppl. 3, 154. Gaillard, P. J., and De Jongh, S. E. (1938) Acta Brevia Neerl. Physiol. Pharmacol. Microbiol. 8, 172. Gaillard, P. J., and De Jongh, S. E. (1939) Acta Brevia Neerl. Physiol. Pharmacol. Microbiol. 9, 159. Goyena, H. (1955) Acta physiol. Latinoam. 6,215. Gross, J., and Leblond, C. P. (1951) Proc. SOC.Exptl. Biol. Med. 76,686. Gross, J., and Pitt-Rivers, R. (1952) Lancet 262, 439. Gross, J., and Pitt-Rivers, R. (1953) Biochem. I. 63,645. Gulewitsch, W., and Krimberg, R. (1905) Z.physiol. Chem. 46,326. Haam, E. van, and Cappel, L. (1940) Am. J. Cancer S9, 354. Haugaard, N.,Haugaard, E. S., and Stadie, W. C. (1954) 1.Biol. Chem. 211,289. Haugaard, N., and Marsh, J. (1952) J. Biol. Chem. 194,33. Herbertson, M.A. (1955) 1.Embryol. & Exptl. Morphol. 3, Part 4, 355. Horning, E. S. (1949) Brit. I. Cancer S, 211. Horning, E.S. (1950) Brit, 1.Cancer 4,235. Huggins, C., Scott, W. W., and Hodges, C. W. (1941) 1. Urol. 46,997. Hunter, D. (1930) Lancet I, 947. Irving, J. T. (1949) J. Physiol. (London) 108,92. Jost, A., and Bergerard, Y. (1949) Compt. rend. soc. biol. 149,608. Jost, A.,and Bozic, B. (1951) Compt. rend. soc. biol. 146,647.
Kahle, P. J., Ogden, H. D., and Getzoff. P. L. (1942) J. Urol. 48,83. Kahle, P. J., Schenken, J. R., and Burns, E. L. (1943) J . Urol. 60,711. Kahn, R. H. (1954) Nature 174, 317. Larionov, L. F., Chertkova, M. A., and Samokvalova, A. S. (1950) Bull. biol. mid. exptl. U.R.S.S. 9, 515. Lasnitzki, I. (1951) Brit. J. Cancer 6, 345. Lasnitzki, I. (1954) Cancer Research 14, 632. Lasnitzki, I. (1955a) J . Endocrinol. 12, 236 Lasnitzki, I. (1955b) Brit. J . Cancer 9,434. Lasnitzki, I. (1955~) Exptl. Cell Research 8, 121. Lasnitzki, I. (1956) Brit. J . Cancer 10, 510. Lasnitzki, I., and Greenberg, R. (1956) “Argonne National Laboratory Quarterly Report.” Lasnitzki, I., and Kennaway, E. L. (1955) “Annual Report, Strangeways Research Laboratory,” p. 22. Lasnitzki, I., and Pelc, S. R. (1957) Exptl. Cell. Research 13, 140. Lawson, K. (1956) “Annual Report Strangeways Research Laboratory.” Lebenson, E., and Magat, M. (1937) J. mid. acad. sci. R.S.R. (Ukraine) 7, 381. Leclercq, J. (1954) Arch. intern. physiol. 62, 101. Likbecq-Hutter, S. (1956) J . Embryol. Exptl. Morphol. 4, 279. Lutz, H.,and Lutz-Ostertag, I. (1956) Arch. anat. microscop. morphol. exptl. 46,
Mellanby, E. (1938) J. Physiol. (London) 94,380. Mellanby, E. (1939) J. Physiol. (London) B6, 36P. Mellanby, E. (1947) J. Physiol. (London) 106, 382. Metre, T.E.van (1947) Bull. Johns Hopkins Hosp. 81, 305. Mirsky, I. A., Broh-Kahn, R. H., Perisutti, G., and Brand, J. (1949) Arch. Biochem. 20, 1.
Mollendorff, W. von (1942) Chem. Zentr. 2, 793. Moore, A.F., Southam, C. M., and Sternberg, S. S. (1956) Science 124, 127. Moore, T.,and Wang, Y. L. (1943) Biochem. J. 87, viii. Pavcek, P. L., Herbst, E. J., and Elvehjem, C. A. (1945) J. Nutrition SO, 1. Plummer, W. A. (1928) Proc. Staff Meetings Mayo Clinic 8, 119. Popper, H., and Greenberg, R. (1941) A . M . A . Arch. Pathol. 32, 11. Price, D.,and Pannabecker, R. (1956) Ciba Foundation Colloquium Ageing 2, 3. Raynaud, A.,and Frilley, M. (1946) Compt. rend. 22S, 1187. Raynaud, A.,and Frilley, M. (1947) Ann. endocrinol. ( P a r i s ) 8,400. Raynaud, A.,and Frilley, M. (1950) Ann. endocrinol. (Paris) 11, 32. Sabella, J. D.,Bern, H. A., and Kahn, R. H. (1951) Proc. SOC.Exptl. Biol. Med. 76, 499.
Sanford, K. K., Likely, G. D., and Earle, W. R. (1954) J. Nutl. Cancer Znst. 16,215. Scheib-Pfleger, D. (1955) Bull. biol. France et Belg. 89, 475. Schrek, R. (1949) Endocrinology 46,317. Schrek, R. (1951) Proc. SOC. Exptl. Biol. Med. 76, 557. Shaffer, B. M. (1956) Exptl. Cell Research 11,244. Shimkin, M. B., and McClelland, J. N. (1949) J. Natl. Cancer Znst. 10, 597. Sidman, R. L. (1956a) Anat. Record l24, 581. Sidman, R. L. (1956b) Anat. Record 124, 723. Sidman, R. L., and Fawcett, D. W. (1954) Anat. Record 118, 487.
CARCINOGENS, HORMONES, VITAMINS AND ORGAN CULTURES
Silberberg, M., and Silberberg, R. (1938) Growth 2, 327. Silberberg, M., and Silberberg, R. (1940) Growth 4, 305. Stadie, W. C. (1954) Physiol. Revs. 34, 52. Exptl. Biol. Med. 83, 52. Starr, P., and Liebhold-Scoeck,R. (1953) Proc. SOC. Strauss, K. S. (1934) Beitr. pathol. A n a f . u. allgem. Pathol. 94, 345. Studer, A., and Frey, J. R. (1949) Schweiz. med. Wochschr. 79, 382. Tomich, E. G., and Woollett, E. A. (1954) J. Endocrinol. 11, No.2, 131. Trowell, 0. A. (1953) J. Physiol. (London) 119,274. Verne, J. (1935) Compt. rend. assoc. anata. XXX Congr. 514. Volgelaar, J. P. M., and Ehrlichmann, E. (1936) A m . J. Cancer 26, 358. Weiss, P. (1949) Intern. Conf. Cancer. 50. Weiss, P., and James, R. (1955) Exptl. Cell Research Suppl. 3, 381. Willier, B. H. (1924) Am. J. Anat. 98, 67. Wolbach, S. B. (1947) J. Bone and Joint Surg. 29, 171. Wolbach, S.B., and Bessey, 0. A. (1942) Physiol. Revs. 22,233. Wolbach, S. B., and Howe, P. R. (1925) J. Exptl. Med. 42, 753. Wolbach, S. B., and Howe, P. R. (1928) Arch. Pathol. Lab. Med. 6,239. Wolff, Etienne (1953) Schweiz. med. Wochschr. 83, 171. Wolff, Etienne, and Haffen, K. (1952) J. Exptl. 2001.119,381. Wolff, Etienne, and Wolff, Emilienne (1953) Poultry Sci. 32, 348.