Comparative morphological studies on the carcinogenic effect of 7,12-dimethylbenz(A)anthracene (DMBA) in normal or intrasplenic ovarian tissue of C3H mice.

A single intravenous injection of 100 mg/kg body weight (b.w.) of 7, 12 dimethylbenz(a)anthracene (DMBA) induces a high percentage of ovarian granulosa cell tumours in C3H mice. After implantation of ovarian tissue into the spleen of gonadectomized female C3H mice similar tumours were found, resulting from an over-stimulation by pituitary gonadotrophins. In the present study the tumour development in intrasplenic ovarian tissue was observed after an additional single intravenous application of 100 mg/kg b.w. DMBA. It was found that the induction of granulosa cell tumours did not seem to be affected by the carcinogen injection whether 12 weeks before or 12 weeks after ovarian tissue was implanted into the spleen. The morphology of these neoplasms corresponds to the DMBA induced granulosa cell tumours in orthotopic ovaries. A direct carcinogenic effect of DMBA on ovarian cells in mice could not be demonstrated but there are indications that the additional DMBA application accelerated the destruction of the oocytes, which might result in a more rapid intrasplenic tumour induction. ImagesFig. 1Fig. 2Fig. 3Fig. 4Fig. 5Fig. 6

Summary.-A single intravenous injection of 100 mg/kg body weight (b.w.) of 7, 12 dimethylbenz(a)anthracene (DMBA) induces a high percentage of ovarian granulosa cell tumours in C3H mice. After implantation of ovarian tissue into the spleen of gonadectomized female C3H mice similar tumours were found, resulting from an over-stimulation by pituitary gonadotrophins. In the present study the tumour development in intrasplenic ovarian tissue was observed after an additional single intravenous application of 100 mg/kg b.w. DMBA. It was found that the induction of granulosa cell tumours did not seem to be affected by the carcinogen injection whether 12 weeks before or 12 weeks after ovarian tissue was implanted into the spleen. The morphology of these neoplasms corresponds to the DMBA induced granulosa cell tumours in orthotopic ovaries. A direct carcinogenic effect of DMBA on ovarian cells in mice could not be demonstrated but there are indications that the additional DMBA application accelerated the destruction of the oocytes, which might result in a more rapid intrasplenic tumour induction.
OVARIAN tissue was found to develop plication but that implantation of ovarian benign granulosa or granulosa-theca cell tissue into the spleen, followed by intratumours after implantation into the spleen venous DMBA treatment, resulted in the of gonadectomized rats (Biskind and transformation from benign to malignant Biskind, 1944). This is due to an unin-granulosa, a few cases of theca cell hibited stimulation of pituitary gland tumours, and a single androblastoma-like gonadotrophins (Heller and Jungck, 1947; neoplasm. It has been suggested that the Miller and Pfeiffer, 1950; Achilles and change in hormonal balance through the Sturgis, 1951; Lipschutz, Cerisola and implantation of ovarian tissue presupposes Panasevich, 1964) because steroids secreted the necessary conditions for the carcinofrom the implant pass directly through the genic effect of DMBA. portal system to the liver where they are In contrast to rats, normal mice inactivated (Golden and Sevringhaus, developed granulosa cell tumours of the 1938; Lipschutz et al., 1964;Leavitt, ovary after the administration of DMBA Carlson and Meyer, 1971) and therefore (Howell, Marchant and Orr, 1954; Marthe feedback mechanism between pituitary chant, 1957; Mody, 1960; Biancifiori, gland and ovarian tissue is interrupted. Bonser and Caschera, 1961; Kuwahara, Preliminary inVestigations (Hilfrich 1967;Krarup, 1970a). and Mohr, 1973;Hilfrich, 1973Hilfrich, , 1974 have As gonadectomy with implantation of shown that in rats normal, non-implanted ovarian tissue into the spleen of female ovarian tissue is not affected by 7, 12 mice also leads to the development of dimethylbenz(a)anthracene (DMBA) ap-granulosa cell tumours (Furth and Gardner, 1955;Guthrie, 1957), the 10 ml solvent/kg b.w. (controls); (4) 30 present study was undertaken to compare mice given 100 mg DMBA/kg b.w., 12 weeks the influence of DMBA on tumour devel-before implantation of ovarian tissue into the opment in implanted ovarian tissue in spleen; (5) 20 mice, 12 weeks after implantamice with that in rats. tion of ovarian tissue into the spleen, given 100 mg DMBA/kg b.w.
Dead animals were autopsied and all MATERIALS AND METHODS organs fixed in 4% buffered formali4; Ninety female, 3-month old C3H mice paraplast sections were stained with haema-(Laboratory Animals Breeding and Research toxylin and eosin and van Gieson stain; the Centre, Bomholtg&rd, Denmark) of 25-30 g ovarian tumours were also stained with PAS body weight (b.w.) were kept in groups of 3 and Alcian blue, Gomori, Masson-Goldner in Makrolon cages Type II (E. Becker & and Sudan III. The effective number of Co., GMBH, Castrop-Rauxel, FRG) under animals is based on the number of mice standard laboratory conditions (room tem-surviving after the first tumour of any site perature 22 : 2°C; relative humidity had been observed. A few mice were lost 55 ± 5%; air exchange 8 times/h), with through cannibalism or the adverse effects of Hope Farms RMH-TMB pelleted diet the surgical operation. (Woerden, The Netherlands) and water ad For statistical comparison of intrasplenic libitum. Sixty mice were ovariectomized ovarian tumour incidence, the chi-square under ether anaesthesia (Pronarcosi, Hoechst test, and for the average survival times, the AG, Frankfurt, FRG), and a piece of ovary U-test after Mann and Whitney (1947) were 1-2 mm in diameter was implanted into the performed. spleen according to the method described by Biskind and Biskind (1949). The animals received a single intravenous injection of 100 mg DMBA (special 15% fat emulsion Table I summarizes the average with 7,12-dimethylbenz(a)anthracene 5 mg/ survival rates as well as the incidence of g; The Upjohn Company, Kalamazoo, Michi-ovarian and other tumours in Groups 1 gan, USA) per kg b.w. or 10 ml/kg b.w. of the and 2 (orthotopic ovaries). In control solvent (intravenous fat emulsion without animals no ovarian tumours were found; dextrose; The Upjohn Company, Kalamazoo, however, after treatment with a single Michigan, USA). doseof DMBA 7 o th a showe Treatment groups.-These consisted of dose of DMBA 78.9% of the mice showed (1) 10 mice given 10 ml solvent/kg b.w. an ovarian tumour, in one animal bi-(controls); (2) 20 mice given 100 mg DMBA/ laterally (Fig. 1). The first neoplasm of kg b.w.; (3) 10 mice, 12 weeks after implanta-the ovary was observed 29 weeks after tion of ovarian tissue into the spleen, given DMBA application.
Macroscopically,    (10) intraspl. Adenoma or adenocarcinoma of ov.) lung (7) Adenocarcinoma of mammary gland (4) Adenoma or carcinoma of adrenal gland cortex (4) Stromal sarcoma (1), haemangioma (1) and fibrovascular polyp (1) of uterus Retroperitoneal neuroblastoma (1) Keratoacanthoma of skin (1) Hepatoma (1)  these neoplasms were up to 25 mm in with blood filled cysts, haemorrhages and diameter with grey-white tumour masses, necroses in larger neoplasms (Fig. 2). and most also exhibited extensive haem-These observations were comparable with orrhages and necroses. In one case lung those detected in tumours of orthotopic metastases were found. In the carcinogen ovaries. Liver metastases were found, one treated Group 2 a large number of mice case each in Groups 3 and 4, and 2 in with leukaemias, as well as neoplasms Group 5. other than the ovary, were detected; The induced ovarian tumours demonthese were mainly of the uterus, fore-strated similar morphological patterns, stomach, vagina and lung (Table I).
not only in orthotopic ovaries (Group 2) Table II lists the average survival but also (with or without the carcinogen rates as well as the number of animals application) in implanted tissue (Groups with ovarian tumours in the spleen and 3-5). In all cases granulosacelltumours those with other neoplasms occurring in were found, with clusters and cords or Groups 3, 4 and 5. All these 3 groups parenchymatous growth of granulosa cells; demonstrated neoplastic growths in the luteinized theca cells and fibre formation spleen that ranged in size from small were present only sparsely (Fig. 3). nodules only a few mm in diameter to Follicular structures of varying sizes were tumours of up to 40 mm. Upon cut also seen, the larger frequently being section yellowish or predominantly grey-blood filled (Fig. 4). Furthermore, tubular white tumour masses were seen, together adenoid formations (Fig. 5)  luteinized areas (Fig. 6) resembling (when genic effect of DMBA on mouse ovaries is extensive) a luteoma were detected; these related to the immediate destruction of morphological alterations mostly preceded small oocytes, whereas the prolonged the described granulosa cell proliferations. retention of DMBA after intraperitoneal In the granulosa cell tumours numerous injection is of no significance for neoplastic mitoses and infiltrative growth could be development. observed, in addition to necroses and haem-Further, the induction of granulosa orrhages in the more extensive neo-cell tumours in intrasplenic ovarian implasms.
plants of mice (Furth and Sobel, 1947; In the DMBA treated Groups 4 and 5 Gardner, 1955;Guthrie, 1957) and by the average survival time after implantax-ray irradiation (Furth and Butterworth, tion of ovarian tissue into the spleen was 1936; Lick, Kirschbaum and Mixer, 1949; significantly lower (P < 0.01, P < 0 05 Kaplan, 1950;Guthrie, 1958) preceded respectively) than in the control Group 3 the destruction of the oocytes. The (Table II). Furthermore, the survival developed ovarian tumours in all the time of intrasplenic granulosa cell tumour aforementioned induction methods were bearing animals was significantly similar and of granulosa cell origin. shortened in Groups 4 and 5 (P = 0.005) Therefore, the authors assumed that the compared with Group 3. However, there over-stimulation by pituitary gonadotrowas no statistically significant difference phins might be decisive for the tumour (0 9 > P > 0 8) in the incidence ofovarian induction, not only in intrasplenic ovarian tumours in the spleen (Table II) among implants but also in degeneratively untreated mice (Group 3), the mice changed ovaries after DMBA treatment or treated with DMBA 12 weeks before irradiation because these gonads lose (Group 4), or 12 weeks after (Group 5) the ability to produce sufficient steroids implantation of ovarian tissue into the to control the output of gonadotrophic spleen.
hormones. The retention of a normal functioning ovary (Marchant, 1960;Jull, DISCUSSION 1969), hypophysectomy (Marchant, 1961) The present study has shown that the or the application of an antigonadotrophic development of granulosa cell tumours serum (Ely, 1959) prevented the ovarian after implantation of ovarian tissue into tumour development. the spleen of C3H mice is not affected by The results of the present study additional DMBA application; the mor-confirm the hormone theory of ovarian phology of these neoplasms corresponds to tumour induction in mice after application the DMBA induced granulosa cell tumours of DMBA; in contrast to rats, mice in orthotopic ovaries. Similar findings demonstrate no direct carcinogenic effect were described by Li, Gardner and Kaplan of DMBA on ovarian cells, even under the (1947) when they used x-ray irradiated conditions of implantation into the spleen. ovaries for intrasplenic implantation. Since in the DMBA treated Groups 4 and 5 Howell et al. (1954) reported the induction the survival time of intrasplenic granulosa of granulosa cell tumours in orthotopic cell tumour bearing animals was signiovaries of mice after repeated skin painting ficantly lower compared with control mice with DMBA. Marchant (1957), Mody (Group 3), it is possible that the additional (1960) as well as Krarup (1969Krarup ( , 1970a DMBA application accelerated the destruccould show that DMBA application to tion of the oocytes, which might result in a mice led to a rapid destruction mainly of more rapid intrasplenic tumour induction. the small oocytes which, therefore, seem to be the primary target cells of DMBA in the ovaries of mice. Furthermore, Krarup I am grateful to Christine Murphy for and Loft (1971) observed that the carcino-assisting with the manuscript.