Functional redundancy between Apc and Apc2 regulates tissue homeostasis and prevents tumorigenesis in murine mammary epithelium

Abstract

Aberrant Wnt signaling within breast cancer is associated with poor prognosis, but regulation of this pathway in breast tissue remains poorly understood and the consequences of immediate or long-term dysregulation remain elusive. The exact contribution of the Wnt-regulating proteins adenomatous polyposis coli (APC) and APC2 in the pathogenesis of human breast cancer are ill-defined, but our analysis of publically available array data sets indicates that tumors with concomitant low expression of both proteins occurs more frequently in the 'triple negative' phenotype, which is a subtype of breast cancer with particularly poor prognosis. We have used mouse transgenics to delete Apc and/or Apc2 from mouse mammary epithelium to elucidate the significance of these proteins in mammary homeostasis and delineate their influences on Wnt signaling and tumorigenesis. Loss of either protein alone failed to affect Wnt signaling levels or tissue homeostasis. Strikingly, concomitant loss led to local disruption of β-catenin status, disruption in epithelial integrity, cohesion and polarity, increased cell division and a distinctive form of ductal hyperplasia with 'squamoid' ghost cell nodules in young animals. Upon aging, the development of Wnt activated mammary carcinomas with squamous differentiation was accompanied by a significantly reduced survival. This novel Wnt-driven mammary tumor model highlights the importance of functional redundancies existing between the Apc proteins both in normal homeostasis and in tumorigenesis.

Figure 1

Deletion of the Apc proteins from murine mammary epithelium. (a) Mammary gland sections from Apc2^+/+ and Apc2^−/− mice were labeled for Apc2 using fluorescent IHC. Although Apc2 is expressed in Wt epithelium, Apc2^−/− glands displayed comprehensive Apc2 loss (scale bar, 50 μm). (b) Virgin mammary gland sections from 10-week-old Blg-Cre⁺Rosa26⁺ mice were labeled for β-galactosidase to assess Cre-mediated recombination. Recombination occurred in a heterogeneous manner, primarily within luminal cells (white arrowheads) but occasionally detectable in apparent non-luminal cells (gray arrowhead) (scale bar, 25 μm). (c) Labeling of Virgin mammary glands section from 10-week-old Blg-Cre⁺Apc^fl/fl mice for Apc using fluorescent IHC revealed a heterogeneous pattern of loss in epithelial cells (scale bar, 50 μm in main image, 25 μm in inlay).

Figure 2

Either Apc or Apc2 is dispensable, however, concurrent loss results in a range of epithelial disruptions. (a) Carmine alum-stained whole mount glands from 10-week-old virgin mice reveals that loss of either Apc protein alone is tolerated, whereas combined loss results in severe defects in ductal branching and epithelial thickening (scale bar, 200 um). (b) Quantification of ductal branching. (c) Quantification of epithelial thickening. (d) H&E-stained sections of mammary tissue from each genotype reveal epithelial disruptions with intraluminal ghost cell nodules in glands deficient for both Apc proteins (Blg-Cre⁺Apc^fl/flApc2^−/−) (scale bar, 50 μm). Labeling for both Ki-67 and caspase-3 exposed an increase in positive cells in epithelium deficient for both Apc proteins (arrows indicate positively labeled cells, scale bar, 50 μm). H&E-stained sections of lactating glands from each genotype revealed Apc2-deficient glands to be indistinguishable from Wt, Blg-Cre⁺Apc^fl/fl glands displayed attenuated alveolar formation with occasional ghost cell nodules (arrow) and Blg-Cre⁺Apc^fl/flApc2^−/−glands displayed a complete lack of differentiated alveoli and vastly perturbed tissue architecture (scale bar, 50 μm). (e) Quantification of Ki-67-positive cells revealed a reduction in proliferation in Apc2^−/− compared with Wt epithelium. A statistical increase was noticed in epithelium deficient for both Apc proteins versus all other genotypes (error bars, s.d., *P⩽0.05 versus Wt, **P⩽0.01 versus all other genotypes, Mann–Whitney U-test, n⩾3). (f) Quantification of caspase-3-positive cells revealed a statistical increase in apoptosis in epithelium deficient for both Apc proteins (error bars=s.d., **P⩽0.001 versus all other genotypes, Mann–Whitney U-test, n⩾3). (g) Sections of 10-week-old virgin glands from Blg-Cre⁺Apc^fl/fl and Blg-Cre⁺Apc^fl/flApc2^−/−were double labeled for cytokeratin 8 (CK8, luminal cell marker) and cytokeratin 5 (CK5, myoepithelial cell marker). In mammary epithelium deficient for both Apc proteins, cells are organized haphazardly indicating disruptions in polarity (scale bar, 50 μm). (h) Sections from these genotypes were also labeled for markers of polarity. Zo-1 staining is almost completely lost along with cells that also display disruptions in E-cadherin (scale bar, 50 μm).

Figure 3

Apc proteins are functionally redundant in control of β-catenin status. (a) Sections of Blg-Cre⁺Apc^fl/fl and Blg-Cre⁺Apc^fl/flApc2^−/−mammary glands were labeled for β-catenin using fluorescent IHC to assess the status of this intracellular Wnt transducer. (i) Epithelium deficient for Apc alone retained a cytoplasmic and membrane associated staining pattern of β-catenin. Staining was at its highest intensity toward the apical surface in a polarized manner. (ii–v) In mammary tissue deficient for both Apc proteins, epithelial β-catenin was disrupted. Certain areas displayed (ii) mis-localized, (iii) strong nuclear or (iv) absent epithelial β-catenin staining. (v) Ghost cells display no DAPI or β-catenin staining (scale bar, 50 μm in top image, 25 μm in i–v). (b) Serial sections of mammary epithelium from each genotype were labeled for Apc, β-catenin, cMyc and CD44. Inactivation of either Apc protein alone did not induce changes in β-catenin localization or status of Wnt targets. Combined loss induced upregulation and nuclear translocation of β-catenin and activated expression of the Wnt target gene cMyc. CD44 expression was undetectable in any genotype (red arrow indicates area deficient for Apc, scale bar, 50 μm).

Figure 4

Functional redundancies exist between Apc proteins in tumor suppression. (a) Wt (yellow line), Apc2^−/−(purple line), Blg-Cre⁺Apc^fl/fl (blue line), Blg-Cre⁺Apc^fl/flApc2^+/−(green line) and Blg-Cre⁺Apc^fl/flApc2^−/− (red line) mice were aged and culled upon signs of ill health (Kaplan–Meier survival curves). Apc2^−/− and Blg-Cre⁺Apc^fl/fl mice displayed no differences in survival compared with Wt mice (log-rank test, P>0.32, n⩾11). Both Blg-Cre⁺Apc^fl/flApc2^+/- and Blg-Cre⁺Apc^fl/flApc2^−/− mice displayed reduced survival compared with other genotypes (asterisks mark statistically different comparisons, P⩽0.01, log-rank test, n⩾11). (b) Mice were examined at time of death. Although Wt, Apc2^−/− and Blg-Cre⁺Apc^fl/fl mice displayed no signs of mammary pathology, 46% of Blg-Cre⁺Apc^fl/fl Apc2^+/- and 80% of Blg-Cre⁺Apc^fl/flApc2^−/− mice exhibited mammary tumors. (c) Representative images of H&E-stained mammary gland sections taken from aged mice. No lesions were found in Wt, Apc2^−/− or Blg-Cre⁺Apc^fl/fl glands, however, small infrequent intraductal aggregates of ghost cells were present in Blg-Cre⁺Apc^fl/fl tissue (arrow). The majority of mammary tumors from Blg-Cre⁺Apc^fl/fl Apc2^+/- and Blg-Cre⁺Apc^fl/flApc2^−/− mice were classified as invasive carcinomas with squamous differentiation (scale bar, 1mm).

Figure 5

Mammary tumors display areas of Wnt activation. Serial sections of tumor tissue were stained with H&E and β-catenin, cMyc and CD44 antibodies. β-Catenin displayed a heterogeneous pattern with numerous cells exhibiting upregulation/nuclear expression. Both cMyc and CD44 were expressed in the majority of cells demonstrating tumors are Wnt activated (top and bottom images scale bar, 50 μm; middle images scale bar, 100 μm).

Figure 6

Characterization of APC and APC2 copy number status in human breast cancer. (a) Human breast cancers from METABRIC and TCGA array data sets, classified according to APC and APC2 copy number status, and subsequent association with breast cancer subtypes. (b) Correlation of Wnt signaling target gene expression levels in samples classified according to APC and APC2 copy number status. (c) Kaplan–Meier of breast cancer patients from METABRIC, classified according to APC and APC2 copy number status.