(A) Schematic of the experimental protocol. Wholemounts of (B) cleared mammary gland; (C) 6 week Trp53 null outgrowth; (D) 10 week Trp53 null outgrowth; (E) tumor bearing Trp53 null outgrowth. (F) Examples of tumor genotype defined by PCR of wildtype and null allele. (G) Tumor growth rate as a function of host irradiation. Tumors that arose in irradiated hosts grew significantly faster compared to those in the sham group (top panel). Tumor doubling time was approximately 2 days in the irradiated host group compared to 8 days in the sham group (bottom panel). (H) Histopathology of Trp53 mouse mammary tumors: left, adenocarcinoma, middle, squamous cell carcinoma, and right, spindle cell carcinoma; scale bar=100 µm. Wholemounts from Trp53 null epithelium transplanted to mice that were sham-irradiated (I) or irradiated with (J) 100 cGy, (K) 200 cGy, or (L) 400 cGy before transplantation. Doses of 200 cGy and above exhibit reduced branching, thinner ducts (arrows), and lack of alveolar buds (arrow heads), indicative of ovarian insufficiency. Scale bar=1 mm.

(A) Analyses of the time-to-tumor occurrence of tumors in sham (black) and hosts irradiated with 10 (blue), 50 (grey), or 100 (red) cGy. Significance was calculated by the log-rank test. (B) Tumor occurrence in transplants pooled from all radiation doses groups (purple, n=45) compared to sham irradiated controls (black, n=29) was accelerated (p<0.0005, log-rank test). (C) Tumor frequency at experiment termination in each dose group (sham, 20/29; 10, 14/14; 50, 17/17 ; 100 cGy, 14/14; p<0.05, Chi-square test). (D) Trp53 null tumor growth rate was increased in hosts previously irradiated with 100 cGy (open symbols) compared to sham (closed symbols) hosts (bars, SEM). Host irradiation at lower doses showed a similar trend but with wider variance. See also Table S1.

(A) UHC of Trp53 null mouse tumors based on SD of 1.0 from sham (red) or irradiated (purple) hosts that were either spindle cell carcinoma (gold) or adenocarcinoma (turquoise). Latency for each tumor is listed below the column. (B) Supervised hierarchical clustering of permutation analysis using SAM with a threshold of 2-fold change identified 24 genes that classified tumors that arose in irradiated (purple) hosts versus sham-irradiated (red). The genes of the irradiated host core (24-IHC) are listed at the right. IPA networks of gene interactions among the 24-IHC include cell-to-cell signaling and interaction, cellular development, hematopoiesis, and cellular assembly and organization. (C,D) IPA network of the top two gene networks generated from the 156-IHC. Note that TGFβ is a central node in both networks (yellow circle). IPA of 156-IHC also revealed enrichment for genes involved (E) leukocyte chemo-attraction and binding (p=0.007), (F) monocyte maturation (p=0.006), and (G) proliferation of tumor cell lines (p=0.0007). Red ovals, induced; green ovals, suppressed. (H) Dendrogram of tumor expression profiles based on the 24–IHC genes indicates that unsupervised hierarchical clustering did not segregate tumors from sham-irradiated (red) versus irradiated (purple) Tgfb1 +/− mice. See also Figure S1 and Table S2.

(A) Heat map based on PTM (p<0.05 and threshold of 1.25-fold) for radiation-induced genes common to mammary gland from Tgfb1 wildtype (black) and heterozygote (grey) littermates at 1 (orange) and/or 4 (blue) weeks after sham irradiation (yellow) or 10 cGy (green) exposure. (B) Heat map based on PTM (p<0.05) and threshold of 1.25-fold change for genes that are down regulated (blue) or up regulated (red) in mammary gland from irradiated wildtype (black) but not Tgfb1 heterozygote (grey) littermates at 1 (orange) or 4 (blue) weeks after sham (yellow) or 10 cGy (green) radiation exposure. (C) IPA networks of the genes up-regulated by radiation in both genotypes invoked cellular growth and proliferation, reproductive system development and function, and organismal development. Note TGFβ is a node (yellow circle). (D) IPA network of the genes induced by radiation only in wildtype hosts included functions involved in hematological disease, metabolic disease, and connective tissue development and function. See also Tables S3, S4, and S5.

(A) Kaplan-Meier analyses of the time-to-tumor occurrence in Tgfb1 heterozygote hosts irradiated with sham (black), 10 (blue), 50 (grey), and 100 (red) cGy. Host irradiation did not decrease tumor latency. Significance was calculated by the log-rank test. (B) Tumor occurrence in transplants into Tgfb1 heterozygote hosts pooled from all radiation doses groups (purple, n=86) compared to sham irradiated controls (black, n=26). (C) Tumor incidence of Trp53 null outgrowths does not significantly increase in irradiated Tgfb1 heterozygote hosts compared to sham hosts at 365 days post transplantation. Sham, n=15/26; 10 cGy, n=21/31; 50 cGy, n=16/22; and 100 cGy, n=20/33 (ns, not significant). (D) Tumor growth rate was not affected by host irradiation (bars, SEM). (E) TGFβ treatment significantly (p<0.0001) increased mammary tumor incidence (black) compared to control parental CDβGeo cells (grey) transplanted to cleared mammary glands. (F) Most CDβGeo cells give rise to ductal outgrowths, as shown in a representative tissue section (H&E, bar=50 µm). (G) A few CDβGeo injections give rise to nodular tumors (H&E, bar=50 µm). (H) CDβGeo cells exposed to prolonged TGFβ in vitro rapidly generate solid tumors (H&E, bar=50 µm). See also Table S6.

(A) The frequency ER-negative tumors was significantly greater (p<0.002) in irradiated hosts compared to sham hosts. (B) The frequency of ER-negative Trp53 null tumors arising of hosts irradiated with 10 cGy was significantly increased in either host genotype (black, p<0.05; grey, Tgfb1 +/−, p<0.05). (C) ER immunohistochemistry in 5 week old outgrowths of Trp53 null mammary outgrowths; scale bar=100 µm. (D) The frequency ER-positive cells in outgrowths was not affected by host irradiation (sham hosts, 34±6% SEM, n=3 vs irradiated hosts, 31±2% SEM, n=9). (E) The ER-115 profile clusters ER-negative tumors that arose in irradiated (purple) from sham-irradiated (red) hosts. (F) Dendrogram showing the ER-115 does not cluster ER-positive tumors. See also Table S7.

(A) Notch ligand, Jag1, is increased at 1 wk and a transducer of Notch signaling, Rpbj, is increased at 4 wk after irradiation as measured by qRT-PCR (bars, SEM). (B) Notch ligand, Jag1, and a transducer of Notch signaling, Rpbj, are increased at 4 wk in irradiated Tgfb1 heterozygote mammary tissue as measured by qRT-PCR (bars, SEM). (C, D) Dual immunostaining of Notch (green) and β-catenin (red) in mammary epithelium in which nuclei are stained with DAPI (blue; Bar=25 µm). Arrowheads indicate cells that have high nuclear Notch immunoreactivity and β-catenin that are increased in irradiated tissues (D) compared to sham-irradiated tissue (C). (E,F) Multiscale in situ sorting of nuclear Notch and β-catenin immunoreactivity shows that radiation (F; n=486 cells) significantly increased the frequency of Notch positive cells (p<0.05, Fisher’s exact test) and dual stained cells (p<0.001, Fisher’s exact test) compared to sham-irradiated tissues (E; n=424 cells).

(A) The overlap between the MaSC signature (Lim et al., 2010), IHC-156 and ER-115 is indicated within the Venn diagram and the p-value for enrichment determined with ConceptGen is shown outside the regions of interest. (B) Venn diagram showing the overlap between the MaSC signature and the genes regulated by radiation in the Tgfb1 wildtype (WT) and heterozygote (HT) mammary gland as described for A. (C) Radiation significantly (p<0.01) increased the proportion of lin-/Cd24^med/Cd49^hi cells determined by FACS analysis of mammary epithelial cells isolated from tissue of mice irradiated 6 weeks before compared to sham-irradiated mice (bars, SEM). Dose was not associated with the degree of response. (D) The mammary repopulating capacity of cells from mice irradiated as in C is significantly increased (p<0.05) as determined by limiting dilution estimation (+95% C.I.). (E) Schematic of distinct mechanisms by which host irradiation affects tumor latency and type.