Melatonin inhibits the growth of DMBA-induced mammary tumors by decreasing the local biosynthesis of estrogens through the modulation of aromatase activity



Melatonin inhibits the growth of breast cancer cells by interacting with estrogen-responsive pathways, thus behaving as an antiestrogenic hormone. Recently, we described that melatonin reduces aromatase expression and activity in MCF-7 human breast cancer cells, thus modulating the local estrogen biosynthesis. To investigate the in vivo aromatase-inhibitory properties of melatonin in our current study, this indoleamine was administered to rats bearing DMBA-induced mammary tumors, ovariectomized (ovx) and treated with testosterone. In these castrated animals, the growth of the estrogen-sensitive mammary tumors depends on the local aromatization of testosterone to estrogens. Ovariectomy significantly reduced the size of the tumors while the administration of testosterone to ovx animals stimulated tumor growth, an effect that was suppressed by administration of melatonin or the aromatase inhibitor aminoglutethimide. Uterine weight of ovx rats, which depends on the local synthesis of estrogens, was increased by testosterone, except in those animals that were also treated with melatonin or aminoglutethimide. The growth-stimulatory effects of testosterone on the uterus and tumors depend exclusively on locally formed estrogens, since no changes in serum estradiol were appreciated in testosterone-treated rats. Tumors from animals treated with melatonin had lower microsomal aromatase activity than tumors of animals from other groups, and incubation with melatonin decreased the aromatase activity of microsomal fractions of tumors. Animals treated with melatonin had the same survival probability as the castrated animals and significantly higher survival probability than the uncastrated. We conclude that melatonin could exert its antitumoral effects on hormone-dependent mammary tumors by inhibiting the aromatase activity of the tumoral tissue. © 2005 Wiley-Liss, Inc.

Melatonin, the main hormone secreted by the pineal, is an indoleamine that acts as a regulator of neoplastic cell growth, particularly on endocrine-responsive breast cancer.1, 2, 3, 4 In this regard, the most common conclusion is that melatonin, in vivo, reduces the incidence and growth of chemically induced mammary tumors in rodents.3In vitro, melatonin, at concentrations corresponding to the physiologic levels present in human blood during the night, inhibits proliferation, increases expression of p53 and reduces the invasiveness of the estrogen-responsive MCF-7 human breast cancer cells.3, 4, 5, 6, 7, 8, 9 Different hypotheses, including the immunomodulatory actions of melatonin,10 its antioxidative effects11 or the inhibition of telomerase activity,12 have been postulated to explain the oncostatic properties of melatonin. However, the effects of melatonin on mammary cancer have been mostly considered as a consequence of its interaction with the estrogen-signaling pathway13 by 2 different mechanisms: (i) by downregulating gonadal synthesis of steroids and, consequently, decreasing their circulating levels;2, 4 (ii) by interacting with the estrogen receptor, decreasing its expression and inhibiting the binding of the estradiol-estrogen receptor complex to the estrogen-response element on the DNA, thus behaving as a selective estrogen receptor modulator.14, 15

The biosynthesis of estrogens in peripheral tissues depends on the activity of the P450 aromatase enzymatic complex, which catalyzes the conversion of androgens to estrogens.16, 17 The aromatase activity in breast cancer tissue has been demonstrated to be higher than in nonmalignant breast tissue or tissue distal to tumors, thus leading to the hypothesis that an increased production of estrogens within breast tumors may exert a biologic effect and thereby stimulate tumor growth.16, 17 That point is significant in postmenopausal women in which local synthesis due to the aromatization of androgen in breast tissue plays an important role in the pathogenesis of estrogen-dependent breast cancer.17 Therefore, effective inhibition of breast aromatase might be an important modulator of estrogen production in breast cancer cells.18, 19 Recently, our group demonstrated by using MCF-7 human breast cancer cells in culture, which express aromatase20, 21 and MT1 melatonin receptor,22, 23 that melatonin, at physiologic concentrations, reduces aromatase activity in these cells both under basal conditions and when aromatase activity was stimulated by cAMP or cortisol.24 Because this third mechanism by which melatonin could influence estrogen-mediated cancer growth has been studied only in vitro, we considered it of interest to study whether the behavior of melatonin as a selective estrogen enzyme modulator may also be demonstrated in vivo.

Material and methods

Animals and housing conditions

Female Sprague-Dawley rats, born in our vivarium, were housed 3–4 per polycarbonate cage, maintained at a room temperature of 22°C ± 1°C, exposed to a 12 hr light (300 lux)/12 hr darkness photoperiod and with access to food and water ad libitum.

Tumor induction

At 55 days of age, all animals received a single intragastric dose of dimethylbenzanthracene (DMBA) (20 mg in 1 ml of sesame oil). From that moment on, the animals were examined on a weekly basis to detect the appearance of palpable mammary tumors. Animals that failed to develop tumors by 12 weeks after the administration of the carcinogen were discarded.

Experimental design

To investigate the aromatase-inhibitory properties of melatonin on the growth of estrogen-dependent DMBA-induced mammary tumors, we used ovariectomized (ovx) rats treated or not with testosterone. In these castrated animals, the growth of the estrogen-sensitive mammary tumors depends on the aromatization of androgens (testosterone administered s.c. to the animals) to estrogens in the mammary tumor tissue. Thus, rats bearing mammary tumors (1 cm in diameter) were randomly assigned to different experimental groups (15–16 rats/group): Group O, ovx animals without subsequent treatment with androgens; Group OT, ovx rats treated with testosterone; Group OTM, ovx and testosterone-supplemented rats receiving melatonin in drinking water; Group OTA, ovx and testosterone-supplemented rats treated with the aromatase inhibitor aminoglutethimide. Group C (control) consisted of uncastrated tumor-bearing animals without any subsequent treatment.

Surgical treatments, tumor size and number, survival rates and autopsy procedures

Rats were ovariectomized or sham operated under i.p.-administered tribromoethanol anesthesia (25 mg/100 g body weight) by standard procedures. The evolution of the size of the tumors was controlled by a weekly measuring of the 2 perpendicular tumor axes with calipers and calculating the tumor area as in Rose and Noonan.25 The number of tumors as well as the number of surviving animals in each group were also recorded weekly.

At the end of the experiment (9 weeks after ovx), the animals were sacrificed and blood samples collected for determination of serum estradiol (ELISA kits from Diametra, Foligno, Italy). Mammary tumors as well as mammary glands without macroscopic lesions were collected. One half of each tumor was fixed in 10% formalin, embedded in paraffin and stained with hematoxylin-eosin for histologic examination; the other half was frozen in liquid nitrogen and stored at −70°C for the measurement of aromatase activity. Body and uterine weights were also measured in all animals.

Testosterone, melatonin and aminoglutethimide treatments

Testosterone (Sigma-Aldrich Química, Madrid, Spain) was dissolved in ethanol, diluted in sesame oil, and 20 mg/Kg body weight given s.c., 3 times per week, for 9 weeks starting at 2 days after ovariectomy.

Melatonin (Sigma-Aldrich Química) was dissolved in ethanol and added to the drinking water at a concentration of 25 μg/ml. The final ethanol concentration was 0.01% for all groups. The water bottles were covered with aluminum foil and fresh solutions prepared every 2 days. Since adult rats drank about 20 ml/day with 90–95% of this total daily water taken up during the dark period, the melatonin dosage in our experiment was approximately 500 μg melatonin/day. Melatonin was given for 9 weeks from the second day after ovariectomy.

Aminoglutethimide (Sigma-Aldrich Química) was dissolved in a small volume of ethanol, diluted in sesame oil and 50 mg/Kg BW injected s.c. 3 times per week, for 9 weeks starting at 2 days after ovariectomy.

Measurement of microsomal aromatase activity

Microsomal aromatase activity was measured in frozen tumors. To obtain the microsomal fraction, each tumor was cut into fragments and homogenized (1:10, wet weight/vol) in chilled 50 mM potassium phosphate buffer (pH 7.4), containing 250 mM sucrose, 10 mM dithiothreitol and 3 mM magnesium chloride. Homogenization was performed on ice using three 5 sec bursts of the Polytron homogenizer, each burst being separated by 10 sec intervals. The homogenates were then centrifuged for 10 min at 850g. Pellets were discarded and supernatants were subjected to centrifugation at 100,000g for 60 min. Microsomal pellets were resuspended in the above described homogenization buffer.

Microsomal aromatase activity was measured by using the 3H2O release method, based on the formation of tritiated water during aromatization of a labeled androgenic substrate.26 Briefly, 100 nM [1β-3H(N)]-androst-4-ene-3,17-dione (NEN Life Science Products, Boston, MA) (25–30 Ci/mM) was first preincubated for 30 min at 37°C, with a NADPH-generating system consisting of 5 mM NADP sodium salt, 50 mM glucose-6-phosphate dipotassium salt and 2 IU/ml glucose-6-phosphate dehydrogenase. Then microsomes (40 mg/tube) were added and incubated for 24 hr at 37°C, in a water bath with gentle shaking. After the 24 hr of incubation, the tubes were placed on ice, and 30% (wt/vol) ice-cold trichloroacetic acid was added, vortexed and centrifuged at 1,700g for 5 min. The supernatants were extracted with chloroform, vortexed and then centrifuged at 800g for 5 min. The resulting aqueous supernatants were adsorbed with 5% dextran-coated charcoal, vortexed centrifuged at 1,700g for 10 min and the supernatant added to vials with scintillation cocktail and counted in a beta counter. The amount of radioactivity measured in [3H] water was corrected by subtracting the blank values from each sample, obtained by incubating tubes containing buffer with the tritiated androgen but no microsomes. The values were also corrected by taking into account the fractional retention of tritium in medium water throughout the procedure of incubation and processing, utilizing parallel tubes containing buffer plus known amounts of [3H] water (NEN Life Science Products) through incubation and assay. The fractional retention of tritium in medium water throughout the incubation and processing of samples was always higher than 85%.

In some experiments, microsomes obtained from the tumors of control animals (Group O) were incubated for 24 hr at 37°C, in presence of 1 nM melatonin or its diluent (ethanol at a final concentration lower than 0.0001%). Then, aromatase activity of microsomes incubated with and without melatonin was measured as above.

Statistical analysis

Differences in body weight, uterine weight, serum estradiol, tumor surface and number of tumors were analyzed by one-way ANOVA followed by the Student-Newman-Keuls multiple comparisons test when appropriated. Survival data were analyzed by the Kaplan-Meier method, using the log-rank and the Breslow statistic tests for comparing survival curves. Tumor microsomal aromatase activity was analyzed by nonparametric Kruskal-Wallis test followed by the Dunn's multiple comparison test. The nonparametric Wilcoxon signed-rank was used to compare the aromatase activity of tumor incubated with melatonin or the diluent. Results from all statistical tests were considered as statistically significant at p < 0.05.


Histopathology of mammary tumors

All mammary tumors induced by DMBA were adenocarcinomas of ductal origin with different architectural patterns: solid-tubule, cystic and papillary, with or without secretions (figures not shown).

Evolution of body weight

Figure 1 shows the time-course changes in body weight on rats from the 5 experimental groups. Ovariectomy significantly increased body weight (p < 0.01) in relation to the noncastrated controls. The addition of testosterone, testosterone plus melatonin or testosterone plus aminoglutethimide to the ovx animals did not modify the evolution of their body weight.

Figure 1.

Effects of ovariectomy and treatments with testosterone, melatonin or aminoglutethimide on body weight of rats bearing DMBA-induced mammary tumors. Values are means ± S.E.M. a, p < 0.01 vs. other groups.

Tumor growth

Figure 2 shows the time-course changes in tumor size depending on the different treatments. Ovariectomy was followed by a significant reduction in tumor size (p < 0.001), and up to 85% of the castrated animals experienced a complete regression of the tumor. The administration of testosterone to ovx animals was able to maintain the rate of tumor growth at similar values to those in intact animals (Fig. 3). Ovx and testosterone-supplemented animals treated with aminoglutethimide exhibited a tumor growth similar to ovx rats, and 77% of the animals showed a complete remission of the tumor. Melatonin treatment of ovx testosterone-supplemented animals was able to counteract the stimulatory effect of testosterone (p < 0.001) on tumor growth in a similar way to aminoglutethimide (Figs. 2 and 3).

Figure 2.

Time course of changes in DMBA-induced tumor size in rats after ovariectomy and treatment with testosterone, melatonin or aminoglutethimide. Values are means ± S.E.M. a, p < 0.05 vs. O and OTA; b, p < 0.05 vs. O, OTA and OTM; c, p < 0.001 vs. O, OTA and OTM.

Figure 3.

Time course of changes in DMBA-induced tumor size in rats after ovariectomy and treatment with testosterone, melatonin or aminoglutethimide, expressed in relation to tumor size in control (uncastrated) animals.

Figure 4 shows the evolution of average number of tumors per rat. In a way similar to that for tumor size, the number of tumors per animal in ovx rats was significantly lower than in control (uncastrated) animals (p < 0.001). Among ovx animals, those treated with testosterone showed the highest number of tumors. When melatonin or aminoglutethimide were given to ovx testosterone-treated rats, the number of tumors were significantly lower than in ovx rats treated only with testosterone. No differences were found between the number of tumors in ovx testosterone-treated animals receiving melatonin or aminoglutethimine.

Figure 4.

Evolution of the number of DMBA-induced tumors in rats after ovariectomy and treatment with testosterone, melatonin or aminoglutethimide. Values are means ± S.E.M. a, p < 0.01 vs. O and OTA; b, p<0.05 vs. O; c, p<0.001 vs. O, OTA and OTM; d, p<0.01 vs. C and OT; e, p<0.01 vs. C.

Serum estradiol concentration and uterine weight

Figure 5 shows the absolute (mg) and relative (mg/100 g body weight) uterine wet weight. As expected, the lowest uterine weight corresponded to ovx rats (p < 0.001). Ovx rats receiving testosterone showed uterine weights similar to those of uncastrated (control) animals. The aromatase inhibitor aminoglutethimide, as well as melatonin, partially counteracts the stimulatory effects of testosterone on uterine weight.

Figure 5.

Effects of different treatments (ovariectomy and administration of testosterone, melatonin or aminoglutethimide) on absolute (mg) (black bars) and relative (mg/100 g body weight) (white bars) uterine wet weight of rats bearing DMBA-induced mammary adenocarcinomas. Values are means ± S.E.M. a, p<0.001 vs. C, OT, OTM and OTA; b, p<0.05 vs. OT.

At the end of the experiment, 17β-estradiol concentration was measured in serum samples. Ovx rats exhibited low levels of estradiol. Neither treatment with testosterone nor testosterone + aminoglutethimine or testosterone + melatonin modified serum 17β-estradiol levels compared to ovx rats (data not shown).

Survival probability

The lowest survival probability corresponded to control (uncastrated) animals (p < 0.01), whereas none of the ovx rats (Group O) died during the experimental period. Among ovx animals, the treatment with testosterone or testosterone plus aminoglutethimide reduced, though not significantly, their survival probability. However, ovx rats treated with testosterone and melatonin in drinking water had a 100% survival rate (p < 0.01), as described for ovx untreated animals (data not shown).

Microsomal aromatase activity

Microsomal aromatase activity was measured on each tumor at the end of the experiment. Figure 6 shows how tumors from animals treated with melatonin had lower aromatase activity than tumors of animals from other groups. From 7 tumors corresponding to control animals, microsomal fractions were incubated for 24 hr with 1 nM melatonin and aromatase activity measured. Figure 7 shows the effects of melatonin on microsomal aromatase activity in these mammary tumors from intact animals; each bar represents a single tumor and expresses the quotient between the aromatase activity with melatonin or with the diluent. The incubation with melatonin decreased significantly (p < 0.05) the microsomal aromatase activity (quotient < 1).

Figure 6.

Microsomal aromatase activity on DMBA-induced mammary tumors in rats. The tumors were obtained from rats bearing DMBA-induced mammary adenocarcinomas 9 weeks after ovariectomy and treatment with testosterone, melatonin or aminoglutethimide. Data are expressed as the percentage of the control group (means ± S.E.M.). a, p<0.05 vs. C.

Figure 7.

Effects of melatonin on microsomal aromatase activity in mammary tumors from control (uncastrated) animals incubated for 24 hr with 1 nM melatonin or the diluent. Each bar represents a single tumor and expresses the ratio between the aromatase activity in the presence or not of melatonin. Wilcoxon rank test p<0.05.


The high incidence of hormone-dependent breast cancer in postmenopausal women suggests an important role of extragonadal steroids on mammary carcinogenesis.27 The local estrogen synthesis in normal and neoplasic breast tissue depends on the aromatization of androgens by the activity of enzymes of the aromatase complex.16, 17, 28, 29 This is the reason for interest in developing drugs able to interfere with the synthesis of steroid hormones by inhibiting the enzymes controlling the interconversion from androgenic precursors, the so-called selective estrogen enzyme modulators (SEEMs).30

In a previous experiment, by using as a model the ER+ human breast cancer cell line MCF-7, we demonstrated for the first time to our knowledge that melatonin, at doses in the range of the nocturnal serum concentration in most mammals, not only inhibits the expression of aromatase but also the activity of the enzyme.24 In our present experiment, our objective was to assess whether melatonin exerts antiaromatase effects in vivo. In this case, we used as a model the DMBA-induced rat mammary tumor. The administration of this chemical carcinogen to 55-day-old rats induces the development of estrogen-dependent mammary adenocarcinomas, with a latency of approximately 8 weeks and an incidence higher than 90%.2, 3, 31 Since the DMBA-induced tumors are estrogen-dependent,32 ovariectomy causes a cessation of its growth and, in most cases, a complete regression of the tumor. However, if tumor-bearing castrated animals are treated with testosterone, the growth of the tumors continues in a similar way to that in control (uncastrated) animals. The testosterone-induced growth of the tumors depends on the estrogens locally formed because of the action of aromatase enzyme on the androgenic precursor.33 This fact is easily demonstrable because the aromatase inhibitor aminoglutethimide counteracts the growth-stimulating effects of testosterone, whereas the androgen receptor inhibitor cyproterone acetate does not.24 The effects of melatonin in this model are similar to that of the aromatase inhibitor, that is to say, counteracting the effects of testosterone and thus supporting a possible antiaromatase role of melatonin in vivo. The weight of the uterus also served to demonstrate the antiaromatase role of melatonin in vivo, since this indolamine inhibits the growth stimulatory effects of testosterone in a similar way to aminoglutethimide. We must emphasize that the growth-stimulatory effects of testosterone can be attributed only to locally formed estrogens, since no changes in circulating level could be appreciated in testosterone-treated rats.

The measuring of aromatase activity in tumors (Fig. 6) indicates that it is lowest in those tumors from animals treated with melatonin. This result, which supports our hypothesis, must be analyzed by taking into consideration some experimental circumstances. One is the heterogeneity of the tumor mass, which introduces a source of variability into the composition of the fragment used to measure the aromatase activity. To try to reduce this fact of variability, we repeated the measurement of microsomal aromatase activity several times in each tumor. Another fact is that we collected the tumors at the end of the experiment, 9 weeks after ovariectomy and application of the different treatments; in those treatments such as melatonin and aminoglutethimide administration, which causes tumors to regress, only a few of the initial tumors, presumably those with the highest aromatase activity, were collected for analysis. When we studied the effect of melatonin on microsomal fractions of tumors from intact (control) animals, in 6 of the 7 tumors analyzed, melatonin reduced aromatase activity in relation to the tumors incubated with the diluent, whereas in only 1 case was the contrary effect found; analyzed as a whole, the differences between aromatase activities with or without melatonin were significant (Fig. 7).

The aromatase gene (CYP19) is, in mammary cancer cells, under the control of promoters II and I.3, regulated by cAMP.34, 35, 36 Melatonin, through a membrane-bound Gi protein-coupled receptor (MT1), downregulates cAMP in different cell types.37, 38, 39 In MCF-7 cells, it has been demonstrated that melatonin at nanomolar concentration reduces the forskolin-induced increase of cAMP39 and, in murine mammary tissue, our group described that melatonin decreases cAMP and increases cGMP in a dose- and time-dependent way.40 Taken together, these data suggest that melatonin could modulate aromatase through its modulatory activity on cAMP.

We conclude that melatonin could exert its antitumoral effects on hormone-dependent mammary tumors in ovariectomized animals by inhibiting the aromatase activity of the tumoral tissue. The description of this new mechanism of interaction of melatonin with the estrogen pathway, which adds to its well-known effects on synthesis of gonadal estrogens, downregulation of the expression of the estrogen receptor and inhibition of the binding of the estradiol-estrogen receptor complex to the estrogen response element in DNA, makes melatonin an interesting compound to be tested for its possible therapeutic value in breast cancer.


We thank Ms. G.Viar-Ruiz for her technical assistance.