This experiment, conducted in C57BL/6 mice, was performed simultaneously with previously reported experiments in BALB/c mice, applying the same protocol originally used to induce mammary carcinomas 2324. A total of 38, 2-month-old mice, were subcutaneously implanted with 40 mg MPA depot (Medrosterona, Gador Laboratory, Buenos Aires, Argentina) every 2 months (four times) for 1 year; 31 control animals received only vehicle. The animals were evaluated twice weekly to detect the appearance of mammary tumors. This experiment was terminated after 1 year because of the emergence of severe skin ulcerations in MPA-treated C57BL/6 mice and in some of the control animals. It has been reported that aged C57BL/6 mice frequently develop spontaneous ulcerative dermatitis 25.

To minimize the possibility of development of skin lesions, the experiment was repeated with a lower MPA dosage, which also proved to be carcinogenic in BALB/c mice 23. Silastic pellets containing 40 mg MPA (Sigma Co, St Louis, MO, ISA) were implanted subcutaneously in 39 virgin, 2-month-old, C57BL/6 female mice, and replaced 6 months later; an equal number of control animals were implanted with control blank silastic pellets. The mice were evaluated daily, and those that developed skin lesions were isolated to prevent further damage, but continued to be evaluated as part of the experiment. The animals were weighed every week and followed for 14 months. Vaginal smears were obtained from selected mice in control and MPA-treated groups to confirm hormone action. After 6 months, MPA-treated mice started to cycle again and MPA pellets were replaced. Every 2 months two mice from both groups were killed for histologic evaluation of mammary gland hormone responsiveness. At the end of the experiments both control and treated animals were killed and mammary glands fixed in formalin (right 4th gland), prepared for whole mounts (left 4th gland), or frozen for Western blot studies.

A total of 24 BALB/c and 24 C57BL/6 virgin females (2-month-old) were implanted subcutaneously with 20 mg MPA, progesterone (Sigma Co.), or control blank silastic pellets (four mice per group). After 4 or 8 weeks the animals were killed, and salivary glands and the 4th right mammary glands were fixed in formalin and embedded in paraffin. Left mammary glands were processed for whole mounts. This experiment was repeated twice. Control animals were killed at diestrus. The morphologic features and the number of ducts, lateral branches, and lobules of the mammary glands of virgin untreated and progestin-treated mice were evaluated in hematoxylin and eosin stained slides and quantified in 40× fields.

The mammary glands were excised together with the attached skin, stretched and pinned down to a cork slab, and fixed in 10% buffered formalin for 24 hours. After fixation, the glands were dissected and extensively washed with distilled water, and lipids were removed by immersing in acetone for 2 days, at room temperature. After hydration with decreasing concentrations of ethanol, the tissues were left overnight in water and stained with toluidine blue (0.025% in water) for 40 min followed by thorough washings with distilled water. The staining was adjusted by destaining, if necessary, with methanol for 30 min and 70% ethanol for other 30 min, under microscopic observation. After washing with distilled water the tissues were fixed in ammonium molybdate for 30 min, washed again, dehydrated in ethanol with increasing concentrations, and transferred to xylene overnight before mounting.

Paraffin-embedded sections were reacted with various antibodies using the avidin biotin peroxidase complex technique (Vectastain Elite ABC kit; Vector Laboratories, Burlingame, CA, USA). Briefly, after dewaxing and hydrating, endogenous peroxidase activity was inhibited using 3% H₂O₂ in distilled water. Blocking solution (2% normal goat serum) was used before specific antibody. Polyclonal antibodies to ER-α (MC-20; Santa Cruz Biotechnology, Santa Cruz, CA, USA), ER-β (chicken antibody kindly provided by Jan-Åke Gustafsson, Karolinska Institutet, Novum, Sweden), and PR A (C-20; Santa Cruz Biotechnology) and the monoclonal antibody to PR-B (Ab 6; Neomarkers, Union City, CA) were used at 1:100 dilution and incubated overnight at 4°C. After incubation with the appropriate secondary antibodies (Vector Laboratories) the reactions were developed with 3-3'diaminobenzidine 0.30% in phosphate-buffered saline with H₂O₂ to a final concentration of 0.5%, under microscopic control. Specimens were counterstained with hematoxylin, dehydrated, and mounted. To quantify ER and PR, 1500 epithelial cells per slide were counted.

Mammary glands were homogenized with a polytron at setting 50 with three bursts of 5 s at a 1:4 ratio of tissue to TEDGS (50 mmol/l Tris [pH 7.4], 7.5 mmol/l EDTA, 0.5 mmol/l dithiothreitol, 10% glycerol, and 0.25 mol/l sucrose) buffer. Protease inhibitors [0.5 mmol/l phenylmethylsulfonylfluoride (PMSF), 0.025 mmol/l CBZ-L-phenylalanine choromethylketone (ZPCK), 0.0025 mmol/l N-alpha-p-tosyl-L-lysine choro-methylketone (TLCK), 0.025 mmol/l N-alpha-p-tosyl-L-phenylalaninechlomethylketone (TPCK), 0.025 mmol/l N-alpha-p-tosyl-L-argininemethylester (TAME)] were added before preparing the extracts. The homogenate was centrifuged for 10 min at 3,300 rpm at 4°C, and the nuclei on the pellet were washed in TEDGS buffer plus 0.01% NP-40. The nuclei were resuspended in TEDGS containing 0.4 mol/l KCl and incubated at 4°C for 30 min, and the nuclear homogenate was centrifuged at 12,000 rpm for 20 min at 4°C. The nuclear extract was diluted 1:2 in TEDGS buffer with 30% glycerol, to reduce the salt concentration. Proteins were quantified using the method proposed by Lowry and coworkers 26.

The samples (100 μg total protein/lane) were separated on 8% (PR) or 10% (ER) discontinuous polyacrylamide gels (SDS-PAGE) using Laemmli's buffer system 27. The proteins were dissolved in sample buffer (6 mmol/l Tris [pH 6.8], 2% SDS, 0.002% bromophenol blue, 20% glycerol, and 5% mercaptoethanol) and boiled for 4 min. After electrophoresis, they were blotted onto a nitrocellulose membrane and blocked overnight in 5% dry skimmed milk dissolved in phosphate-buffered saline-Tween (PBST) 0.1% (0.8% NaCl, 0.02% KCl, 0.144% Na₂PO₄, 0.024% KH₂PO₄ [pH 7.4], 0.1% Tween 20). The membranes were washed several times with PBST and then incubated with PR Ab-7/hPRa 7 (Ab7; Neomarkers, Union City, CA, USA) or C-19 (Santa Cruz), ER-α (MC-20; Santa Cruz), or actin (clone ACTN05; Neomarkers) at room temperature for 2 hours, at a concentration of 2 μg/ml in PBST. Blots were probed with sheep anti-mouse or donkey anti-rabbit immunoglobulin, horseradish peroxidase-conjugated whole antibody (Amersham Life Science, Buckinghamshire, UK). The luminescent signal was generated with the ECL Western blotting detection reagent kit (Amersham Pharmacia Biotech, Buckinghamshire, UK), and the blots were exposed to medical X-ray film (Curix RP1; Agfa Argentina, Florencio Varela, Buenos Aires) for 10 s to 5 min. Band intensity was not quantified if the film was saturated.

Four BALB/c and four C57BL/6 mice were implanted with 20 μg 17β-estradiol (E₂) pellets, or E₂ plus 20 mg MPA, or progesterone, and killed after 2 months. Mammary glands were processed as described above for morphologic studies.

BALB/c (n = 28) and C57BL/6 (n = 27), 2-month-old female mice were implanted subcutaneously with control blank silastic, MPA (40 mg), or progesterone (40 mg) pellets, and after 24 hours the animals were bled by retrorbital puncture under anesthesia. Control mice were bled at diestrus, as evaluated by vaginal smears. Prolactin and growth hormone were measured by radioimmunoassay using a kit provided by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK; Dr AF Parlow, National Hormone and Pituitary Program, Torrance, CA, USA). The assays were performed using 10 μl serum samples in duplicate, and the results were expressed in terms of the reference preparation of mouse prolactin NIDDK-RP3 and mouse growth hormone AFP-107836 standards, respectively. Intra-assay and interassay coefficients of variation were 7.2% and 12.8%, and 8.4% and 13.2%, respectively 28. For insulin-like growth factor (IGF)-I radioimmunoassay, serum samples (15 μl) and IGF-I standards were subjected to the acid-ethanol cryoprecipitation method, as previously described 28. IGF-I was determined using an antibody (UB2-495) provided by Drs L Underwood and JJ Van Wyk, and distributed by the Hormone Distribution Program of the NIDDK. Recombinant human IGF-I (Chiron Corp., Emeryville, CA, USA) was used as radioligand and unlabeled ligand. The assay sensitivity was 6 pg per tube. Intra-assay and interassay coefficients of variation were 8.2% and 14.1%, respectively. Progesterone was determined by radioimmunoassay using an antibody kindly provided by Dr GD Niswender (Colorado State University, Fort Collins, CO, USA), and labeled hormone (progesterone [1,2,6,7 3H(N)]) was purchased from Dupont NEN, Boston, MA, USA. Assay sensitivity was 50 pg, and intra and interassay coefficients of variance were 7.5 and 11.9%, respectively.

Mammary glands from six C57BL/6 and six BALB/c 2-month-old female mice were excised and epithelial cells were isolated using the same protocol that we routinely use to isolate mammary tumor cells 29. An equal number of cells (10⁶ in 5 μl) was inoculated in the right (BALB/c) or left (C57BL/6) cleared fat pads of 21-day-old female nu/nu mice (Swiss background) that had previously been implanted subcutaneously with progesterone (20 mg) or MPA (20 mg) silastic pellet also containing 20 μg E₂; four animals received blank silastic pellets, and four animals were used as a control ensuring that no epithelial cells remained after clearing the fat pads. The mammary glands were cleared according to standardized protocols 3031. Four BALB/c and four C57BL/6 mice were also used as control animals. After 30 days, mammary glands were excised, fixed in 4% buffered formaldehyde, and prepared for histologic evaluation. The experiment was repeated twice. To capture the entire area of interest, several partially overlapping high-definition images were taken from each case with a total magnification of 160×, using a real-time Nikon 1300 digital camera with the QImaging Pro software (MVIA, Inc., Monaca, PA, USA). The images were then aligned together using the PanaVue Image Assembler Software (PanaVue Co., Quebec City, Quebec, Canada). The resulting image was a high-quality panoramic picture of the lesion that allowed us to select the epithelial areas. Images were quantified using the Northern Eclipse Image Analysis Software (Empix Imaging, North Tonawanda, NY, USA) and expressed as the total epithelial area and the number of structures selected in each histologically defined area. To determine ER and PR expression in reconstituted mammary glands, the same procedures were performed in four untreated Swiss nu/nu mice, BALB/c or C57BL/6 mice.

Four to six 6 μm consecutive sections from normal mammary gland were obtained from paraffin-embedded tissue, dewaxed as described above, and stained with hematoxylin and eosin. Approximately 5 × 10³ ductal epithelial cells were then microdissected using a PixCell II LCM system (Arcturus Engineering, Mountain View, CA, USA). Briefly, the cap was inserted into an Eppendorf tube containing digestion buffer (50 μl buffer containing 0.04% Proteinase K*, 10 mmol/l Tris-HCL [pH 8.0], 1 mmol/l EDTA, and 1% Tween-20). The tube was placed in an oven at 37°C to equilibrate and then placed upside down so that the digestion buffer contacts the tissue on the cap. The incubation continued overnight at 37°C and was then centrifuged for 5 min and the cap removed. The reaction was heated to 95°C for 8 min to inactivate the proteinase K. Samples were purified by means of the DNeasy Tissue System kit (Qiagen, Chatsworth, CA, USA) before PCR. We performed tests of the DNA samples by PCR amplifications with four informative microsatellite markers.

Epithelial DNA was typed by PCR using a panel of four microsatellites (also known as simple sequence length polymorphisms) polymorphic between C57BL/6 and the outbred Swiss recipient nude females. The simple sequence length polymorphism primers (Mouse MapPairs) were purchased from Invitrogen (Carlsbad, CA, USA). An aliquot of 100 ng genomic DNA (5 μl of 20 ng/μl) was amplified in a 25 μl PCR reaction containing 2.5 μl 10× PCR buffer (15 mmol/l MgCl₂), 0.5 μl 10 mmol/l dNTP, 1 μl of each primer (final concentration 180 μmol/l), 0.1 μl 5 U/μl Taq polymerase, and 14.9 μl sterile water. Hot start PCR (in which crucial reaction components are withheld until reaction temperatures are reached) was performed by means of FastStart Taq polymerase (Roche, Indianapolis, IN, USA). Initial denaturation (94°C for 6 min), 35 cycles (45 s at 94°C, 45 s at 55°C, and 30 s at 72°C), and final incubation at 72°C for 7 min were carried out using a Perkin-Elmer Model 9600 DNA thermal cycler (Applied Biosystems, Foster City, CA, USA). PCR products were separated by electrophoresis in 4% agarose gels, and stained with 0.5 μg/ml ethidium bromide.

From a total of 47 BALB/c mice initially treated with MPA, 34 developed ductal infiltrating carcinomas, from which 13 arose during the first 12 months (27.6%). The mean latency was 52 weeks, and no mammary tumors were observed in untreated control animals 24. MPA-treated C57BL/6 female mice did not develop mammary tumors, and neither did the control untreated animals (treated BALB/c versus treated C57BL/6 for 1 year; P < 0.001; data not shown). C57BL/6 mice treated with MPA developed severe skin ulcerations, and this arm of the experiment had to be stopped after a year.

In a second experiment, only C57BL/6 mice were treated using a protocol that had also proven to be carcinogenic in BALB/c animals, consisting of implanting 40 mg MPA pellets at the beginning of the experiment and replacing them after 6 months 23. None of the animals in this group developed any mammary lesions, and after 14 months the experiment was terminated (Figure 1a).

MPA induces proliferative ductal lesions in BALB/c mice that are characterized by preneoplastic intraductal solid, papillary, or cribiform growth associated with an intense stromal reaction, with periductal fibrosis, an increase in mammary connective tissue, and moderate to intense mononuclear infiltration 32. In whole mount studies, mammary glands from MPA-treated animals exhibited a gross distortion of the mammary gland architecture and developed hyperplasia in the form of paraductal branching (Figure 1b, top). Ductal hyperplasia was also a common finding. None of these lesions were observed in C57BL/6 MPA-treated mice (Figure 1b, bottom).

We previously observed a significant anabolic effect of MPA on the body weight of BALB/c mice 3334 and, as shown in this study, a similar effect was observed in C57BL/6 mice (Figure 1c; two-way ANOVA, P < 0.001).

To evaluate early morphologic responses to MPA or progesterone on the mammary gland, C57BL/6 and BALB/c mice were treated with MPA or progesterone, or implanted with blank silastic pellets for 4 or 8 weeks, after which the animals were killed and the mammary glands excised and processed for whole mount or hematoxylin and eosin studies. When untreated mammary glands from control virgin mice of both strains, without any treatment and in the same phase of the estrous cycle, were compared, BALB/c untreated mammary glands exhibited more ductal side branching than did C57BL/6 glands (Figure 2a,b, left panels), and only BALB/c mice exhibited a significant increase in MPA-induced ductal branching in mammary glands (Figure 2a,b, middle panels; P < 0.001). The effect of progesterone was different from that of MPA; lobular differentiation was observed in both strains, although the effect in BALB/c mice was more prominent (Figures 2a,b, right panels). Quantification of these morphologic differences is shown in Figure 2c. These experiments confirm that MPA and progesterone have different effects in C57BL/6 and BALB/c animals, with significantly lower hormone responsiveness in the C57BL/6 strain.

All receptors were evaluated by immunohistochemistry using validated antibodies 35363738. The number of positive cells was expressed as the percentage of stained nuclei per number of epithelial cells/400× fields. There were significant differences in the number of ER-α-positive cells between virgin mammary glands of both strains (41.2 ± 10% in BALB/c versus 19 ± 7.5% in C57BL/6 mice; P < 0.001; Figures 3a and 4c). MPA treatment resulted in a significant decrease in ER-α expression only in BALB/c mice (Figure 3a; P < 0.05). ER-β expression was low in control virgin mice of both strains (Figures 3b and 4c) that increased in MPA (P < 0.05) and in progesterone-treated BALB/c animals (P < 0.001; Figure 3b) but not in C57BL/6 mice.

The pattern of PR-A expression was very similar to that of ER-α; again, significantly higher levels were observed in untreated BALB/c mice (Figure 4a,c), which were more efficiently downregulated by MPA (P < 0.001) than by progesterone (P < 0.05; Figure 5a). Virgin untreated mice of both strains exhibit almost no PR-B expression, as previously reported 35 (Figures 4b,c). Interestingly, PR-B was highly expressed in MPA and progesterone treated BALB/c (P < 0.001 and P < 0.05, respectively) and C57BL/6 mice (P < 0.01; Figure 4b), and followed a similar pattern of expression as ER-β.

Nuclear extracts from progestin-treated or virgin control mice were used for Western blots (Figure 5a). Extracts were normalized by protein content. Because mammary glands from virgin mice have less epithelial cells than progestin-treated mice, and on the other hand both stroma and epithelium may express steroid receptors, which may be differentially regulated, we only quantified the ratio between both isoforms in the same blot (Figure 5b). The PR-A/PR-B ratio was significantly higher in virgin as compared with progestin treated mice of both strains, supporting the immunohistochemical data. In addition, virgin BALB/c mammary glands expressed higher levels of PR-A and ER-α (data not shown) than did C57BL/6 glands (P < 0.01). Significant differences between MPA and progesterone treatment were not detected in Western blots.

MPA, because of its androgenic action 39, has a specific effect on salivary gland morphology in that it induces the development of convoluted granular ducts 24, which was observed in both strains (Figure 6a). Progesterone had no notable effects on salivary glands.

ER knockout mice develop a hypoplastic mammary gland phenotype as a consequence of a pituitary gland dysfunction associated with a decrease in prolactin levels 40. To explore the possibility of pituitary dysfunction as an explanation for the strain differences in hormone responsiveness, we evaluated serum levels of prolactin in control and progestin-treated mice, but we failed to observe any difference between C57BL/6 and BALB/c mice. In addition, prolactin levels were increased in progestin-treated animals of both strains (P < 0.001, two-way ANOVA; Figure 6b). Similarly, no differences in progesterone, growth hormone, or IGF-I levels between strains were observed. Furthermore, serum levels of progesterone reached after the implantation of 20 mg progesterone pellets was also similar in both strains (Figure 6b).

We were now interested in determining whether the differences in ER-α and PR-A expression observed between both strains would even out when epithelial mammary cells from both strains were inoculated in the same host. Because the differences were evident only in untreated animals, we performed the same recombination experiments in untreated mice. Figure 9 shows ducts that originated from C57BL/6 or BALB/c cells transplanted in the same nu/nu mouse; no significant differences in ER-α or PR-A expression were observed. These results indicate that host factors are responsible for the strain differences observed in mammary gland hormone receptor expression and that the level of these receptors is related to the degree of hormone responsiveness.