Figure 1.
Carcinogen exposure during juvenile development does not significantly alter ductal tree growth or morphogenesis during puberty or pregnancy.
(A) Ductal outgrowth. Whole mount preparation of a representative mammary ductal tree in 35 day virgin female BALB/c mice, growing from the nipple (left hand side) rightward past the lymph node (dark inclusion). On this day, mice were injected intraperitoneally with 0.10 µmol DMBA/g mouse or vehicle (tricaprylin). Scale bar = 1 mm. To evaluate the mitotic index of mammary epithelial cells at this stage of development, paraffin sections of glands were stained with anti-Ki67 (or a nuclear counterstain). The multi-layered epithelium shown is typical of terminal end bud. Scale bar = 50 µM. (B, C) Whole mount preparations of mammary ductal trees harvested from mice treated with vehicle or DMBA, 2 or 7 weeks after treatment (as indicated).
Figure 2.
Genotoxin exposure during juvenile development affects differentiation and stem cell frequency in adult ductal trees.
(A) Stem cell assay. Mammary epithelial cells populations were prepared from adults (9–10 weeks old), either exposed to DMBA at 5 weeks or not (administered tricaprylin vehicle), and injected into cleared fat pads at various limiting cell numbers. Four to six weeks later, glands were assessed for colonization and scored as a take (more than 25% colonization), or no take. The data fit the limiting dilution model (see Methods section; likelihood ratio test of single-hit model: P<0.000001) and stem cell frequencies were estimated on the basis of the LimDil statistical program (difference between groups: P<0.0001). (B) Flow cytometric analysis of MEC populations from adults exposed to DMBA as juvenile mice. Mammary epithelial cells were dissociated from 3 mice each (DMBA-treated and control), and stained according to [9], [17], to resolve basal and luminal epithelial cell populations (see Fig. S1 for gating details). Representative flow cytograms are shown. The two principal cell types, luminal and basal, were quantified, and the ratio of luminal/basal cell is shown.
Figure 3.
Development of genotoxin-exposed glands during pregnancy.
(A, B) Analysis of lobulo-alveolar development during pregnancy. Whole mount preparations from timed pregnant mice (6d p.c, 10 week old mice), exposed as juveniles to DMBA (or not). To examine the pattern of growth in more detail, paraffin sections from mammary glands were immunostained with Ki67 (to show cells in cycle; green) and counterstained with a luminal cell-specific stain, CK8 (K8; red; scale bar = 25 µm). Quantitation of the Ki67 index showed no significant difference between genotoxin-exposed mice and the control cohort. (C) Lobulo-alveolar development during pregnancy in Lrp5−/− glands. Samples were processed according to (B), and were similarly stained with Ki67 and CK8, and also with CK5 (K5; blue) to visualize basal epithelial cells.
Figure 4.
Evaluation of the DNA damage response (DDR) in basal and luminal epithelial cells after genotoxin administration in vivo.
(A) DNA damage focus assembly. Mice were treated with either 10 Gy γ-irradiation and harvested 30 minutes later (positive control), or administered DMBA (2 µg/ml) or vehicle (tricaprylin), and their mammary glands were harvested 2 days or 7 weeks later (as shown). Paraffin sections were tested for the formation of nuclear-associated γH2AX foci (green) in basal and luminal epithelial cells (K5, blue and K8, red respectively, counterstained with TOPRO). (B) DDR checkpoint activation. Similar sections were evaluated for their activation of p53, by assaying the nuclear accumulation of pS15-p53. Basement membranes are demarcated by the dotted line, and examples of nuclei positive for pS15-p53 are shown (B, basal; L, luminal). (C) pS15-p53 activation in response to DMBA. To evaluate the activation of DDR in response to DMBA, compared to other more canonical genotoxins, lysates of mammary epithelial cells (HC11 cells) treated with various genotoxins, were evaluated by Western blotting. The genotoxins were: 10 Gy irradiation (1 hour), 2 µg/ml DMBA (24 hours; this is a pro-genotoxin requiring activation by metabolism, or vehicle, DMSO), N-ethyl-N-nitrosourea (1 hour after addition of 500 µg/ml ENU, a direct acting alkylating agent, or control PCB), or 20 µg/ml camptothecin (1 hour, camptothecin is a topoisomerase I inhibitor, or DMSO). Similarly, the specificity of the nuclear stain was cross-checked to Western blot data of lysates of primary mammary epithelial cells in culture, exposed to DMBA (as indicated). Non-treated cells (NT) are also shown for comparison. (D) Induction of proliferation is equally allocated to basal and luminal cells after exposure to DMBA. Genotoxin-exposed and control-treated mice (1, 2 and 7 weeks after treatment) were administered BrdU (intraperitoneally), and mammary glands harvested 2 hours later. Data that describes the lineage-specific mitotic index is shown for samples 2 weeks after DMBA treatment. Thus, the fraction of BrdU-positive basal and luminal cells was assayed (n = 3, >2000 cells each), and illustrated here as a fraction of each cell type (top panel) or as a fraction of total cells (basal cells are a minority). Statistically different values are indicated * (p<0.05). (E) Induction of proliferation is focalized after exposure to DMBA. Genotoxin-exposed and control-treated mice (1, 2 and 7 weeks after treatment) were administered BrdU (intraperitoneally), and mammary glands harvested 2 hours later. Mitotic cells were widely distributed in vehicle-treated glands. Though this pattern was also evident in DMBA-administered glands, in addition, there were “bursts” of mitotic activity, defined as five or more BrdU positive cells in a 20-cell radius (B). Bursts were measured per length of duct (n = 3; arbitrary unit of cm after image capture).
Figure 5.
DMBA-induced DDR is accompanied by loss of Wnt signaling in vitro and in vivo.
(A) DMBA exposure reduces LRP activation in vivo, and this reduction in active Wnt signaling is durable. Mammary glands from DMBA-treated (and control) mice were snap-frozen, ground and lysed as described. Proteins were analyzed by Western blotting, and assayed for activation of LRP (p-LRP). Blots were normalized using a vinculin internal loading control. The relative signal was calculated for each pair of treated/control samples, for the timepoints indicated post-exposure. (B) DMBA exposure of MECs in vitro reduces Wnt ligand -induced (and basal) activation of Wnt signaling. MECs were transferred to culture for 24 h, treated with Wnt3a (100 ng/ml, or not) as described in the Methods section, and lysed for analysis of p-LRP, 24 and 48 hours later. (C, D) DMBA exposure of HC11 mammary epithelial cells reduces Wnt ligand -induced (and basal) activation of Wnt signaling. (C) HC11 cells were treated with DMBA and assayed for the appearance of γH2AX 24 hours later (by immunostaining as detailed for Fig. S3). (D) Lysates of similarly treated cells were analyzed by Western blotting for LRP activation (phosphor-LRP).
Figure 6.
The genotoxic response ablates Wnt-induced basal cell accumulation.
(A) Immunohistochemical analysis of lineage-specific responses to DMBA and Wnt3a. Cultured MECs were transferred to culture for 24 hours, and treated with DMBA (2 µg/ml), recombinant mouse Wnt3a (100 ngs/ml) or both for 24 hours, as indicated (Wnt alone, data not shown). Cells were either incubated with BrdU (for 60 mins) or not, then fixed and stained as indicated. (B) Examples of images used for quantitative analysis. To obtain the fraction of cells of each lineage that showed activation of DNA damage, or cell division, the image of the lineage-specific marker was overlaid on H2AX or BrdU. (C) Quantitation of immunohistochemical analysis of lineage-specific responses to DMBA and Wnt3a. The fraction of each of K5- or K8-positive cells was scored for expression of H2AX, or mitotic index, 24 or 48 hours after exposure to DMBA or Wnt3a (as indicated). (D) Assay of Wnt- and DDR-dependent transcriptional reporters. The Wnt reporter, Axin2, and the p53 target, p21, were measured by qPCR assay of RNA extracts of cell lysates. (E). Visualization of luminal and basal cells in Wnt-treated cultures. Cell cultures were fixed and stained with lineage-specific markers (K8, luminal; K5, basal) together with a nuclear counterstain. (F) Quantitation of lineage-specific responses to DMBA and Wnt3a. Results were quantified (at least 1500 cells were counted in each sample; 3 experiments), and the relative differentiation of cultures expressed as % K5 and K8-positive cells, and the luminal/basal ratio. * indicates significant difference from non-Wnt3a treated cultures, and ** indicates significant difference from both non-Wnt3a treated, and Wnt-treated cultures.