Article Figures & Data
Figures
- FIG. 1.
Overexpression of HA-tagged IRS-1 or IRS-2 disrupts MCF-10A acinus morphogenesis. (A) Representative Western blot illustrating the generation of MCF-10A stable clones overexpressing either HA-tagged IRS-1 or IRS-2 compared to endogenous levels expressed in empty vector controls (pLXCN2). Immunoblot showing IGF-I-induced phosphorylation of ERK and Akt in both IRS-1- and IRS-2-overexpressing cell lines. (B) Bar chart depicting the growth rates (expressed in cell numbers) of MCF-10A-HA-IRS-1 and MCF-10A-HA-IRS-2 cells grown in SFM with or without IGF-I compared to that of the pLNCX2 controls. (C) Confocal images of 3D Matrigel cultures of MCF-10A-pLNCX2 (top panels), MCF-10A-HA-IRS-1 (middle panels), and MCF-10A-HA-IRS-2 (bottom panels) clones stained with nuclear stain TOPRO3 (panels 1 to 3), Ki67/TOPRO3 (panels 4 to 6), GM130/TOPRO3 (panels 7 to 9), and laminin V/TOPRO3 (panels 10 to 12).
- FIG. 2.
Development of MMTV-HA-IRS-1 and MMTV-HA-IRS-2 transgenic mice. (A) Illustration of the MMTV-HA-IRS expression vector. bGH polyA, bovine growth hormone polyadenylation sequence. (B) Representative Western blots depicting HA, IRS-1, or IRS-2 in mammary gland lysates from transgenic (Tg) and wild-type (Wt) mice at 10 days of lactation. (C) HA immunoblotting of multiple tissue lysates from transgenic (+) and wild-type (−) mice. Kid, kidney; Liv, liver; Spln, spleen; Mamm, mammary gland; Sal g, salivary gland; Uter., uterus. (D to K) IHC staining of HA (brown cells) resulted in similar patterns of expression in mammary glands from IRS-1 (left panel) and IRS-2 (right panel) transgenic mice. Images are TEBs) (D and E) and mammary glands from 6-week-old virgin mice (F and G), mice midpregnancy (day 8) (H and I), and 4-week involuted mice (J and K). The scale bar (black) in panels D and E represents 200 μm, and that in panels F to K represents 10 μm. In panels D and E, the arrows point to the cap cell layer.
- FIG. 3.
Progressive mammary hyperplasia in MMTV-HA-IRS-2 transgenic mice. (A to D) Representative images showing early mammary gland hyperplasia in virgin 24-week-old MMTV-HA-IRS-2 transgenic mice. Whole-mount mammary glands (A and B) and H&E staining (C and D) illustrate the early ductal lesions and unique epithelial cell proliferation across the width of the duct (bridging of cells) highlighted by arrows (A). In panel C, arrows point to mitotic figures. The scale bar represents 10 μM. (E and F) Representative images illustrating the increase in ductal hyperplasia in virgin HA-IRS-2 transgenic mice at 52 weeks of age. H&E images of mammary gland lesions from multiparous MMTV-HA-IRS-2 transgenic mice at 24 (G) or 52 (H) weeks of age are shown. (G) Representative squamous metaplasia. (H) Representative mammary intraepithelial neoplasia.
- FIG. 4.
Mammary tumorigenesis in MMTV-HA-IRS-1 and MMTV-HA-IRS-2 transgenic mice. Wt, wild type. (A) Whole-mount images of multiple mammary gland lesions in 18-month-old multiparous MMTV-HA-IRS-2 transgenic mice (top panel and bottom right panel) and 18-month-old MMTV-HA-IRS-1 transgenic mice (bottom left panel). Arrows point to lesions within the gland, and the open diamonds indicate the lymph nodes. (B to D) Kaplan-Meier plots illustrate the times to palpate tumors as percentages of multiparous female IRS-1 (B) or IRS-2 (C) transgenic mice compared to their respective wild-type control littermates. In addition, mammary tumor formation was significantly increased due to parity in IRS-2 transgenic mice compared to age-matched virgin controls (D). The P value for MTTF was calculated by the log rank test by using the Prism computer program. There were 20 mice in each group.
- FIG. 5.
Constitutive phosphorylation and p85 binding of HA-IRS-1 in transgenic mice. (A) Mammary gland lysates from MMTV-HA-IRS-1 transgenic (TG) and wild-type littermate (WT) mice at 10 days of lactation. Lysates from four independent mice (50 μg) were immunoblotted for anti-pY, total IRS-1, or p85 (left panels). Immunoprecipitation was performed on 500 μg of lysates with 5 μg of HA antibody and then probed with the same antibodies. (B) Mammary tumors (T) and mammary glands that did not contain a tumor (Mg) from the same transgenic mouse were immunoblotted for p-Akt and p-ERK1/2. Total levels were also immunoblotted as a control.
- FIG. 6.
Histologic characteristics of MMTV-HA-IRS mammary tumors and metastases. Representative H&E images highlighting the characteristics of IRS-1, IRS-2, or IGF-IR pathway tumors are shown. (A) IRS-2 undifferentiated solid nodular mammary tumor with sparse stroma and a lack of myoepithelial cells and squamous metaplasia. (B) IRS-2 papillary tumor with defined cords of branched ductal architecture. (C) IRS-2 pilar tumor with inflammatory infiltrates in dense, well-developed stroma (black arrow). (D) IRS-2 tumor defined by glandular acinus formation with lactation (black arrow). (E) IRS-1 spindle cell tumor. (F) Highly differentiated IRS-2 pilar tumor with dense stroma containing lymphocytic infiltrates. Red arrows show cells undergoing squamous transdifferentiation. (G) Constitutively active IGF-I receptor (CD8-IGF-IR) mammary tumor with keratinization. (H) CD8-IGF-IR mammary tumor with branched ductal architecture, glandular formation, and lactation. (I) Representative image of macroscopic lung metastases in IRS transgenic animals. (J) H&E image of lung metastasis showing characteristics of the primary mammary tumor. The scale bar (black) represents 50 μm.
- FIG. 7.
MMTV-HA-IRS tumors contain luminal and myoepithelial cells and express markers of putative mammary progenitor cells. (A to C) Representative IHC for epithelial markers in a tumor from an MMTV-HA-IRS-1 transgenic mouse. (A) Anti-alpha smooth muscle actin. (B) Anti-K14. (C) Anti-K8. (D) Anti-K6 IHC on an IRS-2 tumor. (E) SCA-1 IF on an MMTV-HA-IRS-2 mammary tumor. The counterstain is DAPI (blue). (F) SCA-1 IF (red) on a mammary tumor from MMTV-HER2 transgenic mice (note the absence of SCA-1-positive staining in the tumor; positive staining is noted in the normal mammary gland periphery [white arrow]).
- FIG. 8.
MMTV-HA-IRS tumors show increased β-catenin levels and downstream signaling. (A) Representative immunoblotting of IRS-2 tumors (T) and mammary glands from the same transgenic mouse that did not contain a palpable or visible tumor upon dissection (Mg). Membranes were blotted for β-catenin, cyclin D1, and c-myc proteins. Keratin 8 served as a loading control for epithelial content, and β-actin served as a protein loading control. (B) Immunoprecipitation of HA or β-catenin followed by reciprocal HA or β-catenin immunoblotting from three HA-IRS-1 tumor lysates or three HA-IRS-2 tumor lysates. Input represents 75 μg of lysates, and 750 μg was used for the IP. (C) Immunoprecipitation of HA followed by β-catenin immunoblotting from MCF-10A cells stimulated with or without IGF-I (5 nM) for 24 h. Input represents 75 μg of lysates and 750 μg was used for the IP. (D) Immunoblotting of cell lysates from MCF-10A-HA-IRS-1 and MCF-10A-HA-IRS-2 clones for cyclin D1, A, or B.
Tables
- TABLE 1.
Histologies of IRS mammary tumors
Tumor type No. of tumors (% of total) HA-IRS-1 HA-IRS-2 Undifferentiated (solid adenocarcinoma) 2 (40) 4 (20) Differentiated Adenosquamous 1 (20) 5 (25) Squamous cell 1 (20) 3 (15) Spindle cell 1 (20) 0 Glandular 0 1 (5) Papillary 0 4 (20) Pilar 0 3 (15)
