Fbxw7 is an F-box protein that contributes to regulation of cell proliferation and cell fate determination as well as to tumor suppression in various tissues. In this study, we generated mice with mammary gland–specific ablation of Fbxw7 (Blg-Cre/Fbxw7F/F mice) and found that most neonates born to mutant dams die soon after birth as a result of defective maternal lactation. The mammary gland of mutant dams was markedly atrophic and manifested both excessive cell proliferation and apoptosis in association with the accumulation of Notch1 and p63. Despite the hypoplastic nature of the mutant mammary gland, Blg-Cre/Fbxw7F/F mice spontaneously developed mammary tumors that resembled basal-like carcinoma with marked intratumoral heterogeneity. Additional inactivation of Trp53 in Blg-Cre/Fbxw7F/F mice further promoted onset and development of mammary tumors, suggesting that spontaneous mutation of Trp53 may facilitate transition of hypoplastic mammary lesions to aggressive cancer in mice lacking Fbxw7. RNA-sequencing analysis of epithelial- and mesenchymal-like cell lines from a Blg-Cre/Fbxw7F/F mouse tumor revealed an increased mutation rate and structural alterations in the tumor and differential expression of upstream transcription factors including known targets of Fbxw7. Together, our results implicate Fbxw7 in the regulation of cell differentiation and in tumor suppression in the mammary gland. Loss of Fbxw7 increases mutation rate and chromosome instability, activates signaling pathways governed by transcription factors regulated by Fbxw7, and triggers the development of mammary tumors with prominent heterogeneity.

Significance:

Mammary gland–specific ablation of Fbxw7 in mice results in defective gland development and spontaneous mammary tumor formation reminiscent of human basal-like carcinoma with intratumoral heterogeneity.

The mammary gland epithelium comprises inner luminal cells and outer basal cells. The former include differentiated hormone-sensing cells as well as different classes of progenitors, whereas the latter include a large fraction of myoepithelial cells with specialized contractile functions and a small subpopulation of stem cells (1). The stem cells of the mammary gland give rise to differentiated epithelium via a series of committed intermediates in a hierarchical manner, with the final cell fate decisions being irreversible. Multipotent or lineage-restricted stem cells thus produce progenitors that lack self-renewal capacity and whose descendants undergo terminal differentiation toward either the myoepithelial or luminal epithelial compartments (1, 2).

Breast cancer is the most common female malignancy (3), and substantial advances in the diagnosis and treatment of this condition have been achieved. Microarray analysis of gene expression profiles in a large number of breast cancer specimens resulted in the categorization of such tumors into five groups: luminal subtypes A and B, both of which are positive for the estrogen receptor (ER); Her2 (also known as Neu or Erbb2) positive; basal-like carcinoma (BLC); and normal breast–like carcinoma (4). An additional molecular subtype, designated claudin low, was subsequently also identified on the basis of gene expression (5). Claudin-low carcinoma (CLC) shows no or only a low level of expression of luminal differentiation markers and manifests features of epithelial-to-mesenchymal transition (EMT) as well as distinctive cancer stem cell (CSC)–like properties (6). BLC and CLC are subtypes of breast cancer with an ER, progesterone receptor (PR)-negative (PR), and Her2 phenotype, and these triple-negative tumors (TNT) have a poor clinical prognosis (7). In contrast to ER+ and Her2+ tumors, there is no effective therapeutic strategy for TNTs, mainly as a result of their intratumoral heterogeneity with regard to histology, gene expression patterns, and genetic alterations. Intratumoral heterogeneity is associated with chemoresistance, invasiveness, and metastatic capability, with these characteristics rendering treatment difficult. Although clonal evolution and the CSC model have been proposed to explain the development of intratumoral heterogeneity, the mechanism of its development in breast cancer remains relatively uncharacterized (8, 9).

Mutations of various genes have been shown to contribute to the development of BLC. BRCA1, TP53, PTEN, and RB1 are commonly lost in familial and sporadic forms of TNTs (10). Indeed, conditional ablation of Brca1 in the mouse mammary gland resulted in the development of BLC on a background of p53 deficiency (11). Mutant mice lacking either PTEN or Rb1 in the mammary gland also develop BLC (12). Luminal ER cells are able to give rise to various molecular subtypes of breast cancer including BLC and CLC (13), but the molecular mechanisms underlying the development of these tumors have remained unclear.

Alterations of the gene for F box– and WD repeat domain–containing protein 7 (Fbxw7, also known as Fbw7, SEL-10, hCdc4, and hAgo) were recently identified in a subset of breast cancers (14, 15). Furthermore, expression of FBXW7 was found to be significantly reduced in tumors of high histologic grade or in hormone receptor–negative tumors, and a low level of FBXW7 expression appeared to be correlated with poor prognosis (16, 17). Another study showed that FBXW7 mutations are frequent in human TNTs (18). Fbxw7 is a member of the F-box protein family that was initially identified as a negative regulator of LIN-12 (Notch)–mediated signaling in Caenorhabditis elegans (19). It functions as the substrate recognition subunit of an SCF (Skp1–Cullin–F-box protein) ubiquitin ligase complex that targets specific proteins for ubiquitylation and consequent degradation by the proteasome. In addition to Notch, the targets of Fbxw7 include various proteins that control cell-cycle progression (20) including cyclin E, c-Myc, c-Jun, mTOR, and KLF5 (21–23). Fbxw7 has therefore been considered to serve as an oncosuppressor protein. Indeed, mutations in FBXW7 have been detected in many types of human malignancy including T-cell acute lymphoblastic leukemia, lung squamous cell carcinoma, as well as colorectal, bladder, endometrial, and cervical cancer (24) in addition to breast cancer.

Mice deficient in Fbxw7 die in utero as a result of defective vascular development (25). To avoid such early embryonic mortality, we and others have generated mice in which Fbxw7 is conditionally inactivated in T cells (26), hematopoietic stem cells (27), liver (28), intestine (29), or brain (30). Studies with these mutant mice have shown that Fbxw7 regulates cell proliferation and cell fate decisions in various organs by targeting a variety of substrates for degradation. However, the physiologic role of Fbxw7 and the consequences of its inactivation in the mammary gland have remained unknown, despite the clinical data suggesting that Fbxw7 expression status might be a determinant of breast carcinogenesis.

We have now examined the consequences of Fbxw7 deletion in the mouse mammary gland by crossing mice that harbor a “floxed” Fbxw7 allele with those that express a transgene for Cre recombinase under the control of the mammary gland–specific promoter for the β-lactoglobulin gene (Blg). We found that deletion of Fbxw7 in mammary epithelial cells resulted in both excessive proliferation and massive apoptosis during late pregnancy, resulting in marked gland atrophy. Female mice of this strain eventually show a high incidence of breast carcinoma development, with some of them developing TNTs that mimic BLC. We established both epithelial- and mesenchymal-type cell lines from one of these Fbxw7-deficient mammary tumors, and we found that both types of cell line contained tumor-initiating cells. Our results thus suggest that Fbxw7 plays a pivotal role in mammary gland development and also functions as a tumor suppressor in these glands.

Generation of mammary gland–specific Fbxw7-mutant mice and spontaneous mammary tumor formation

Mice homozygous for a floxed allele of Fbxw7 (Fbxw7F/F mice; ref. 26) were crossed with Blg-Cre transgenic mice (kindly provided by A.R. Clarke, Cardiff School of Biosciences; ref. 31). Deletion of exon 5 of the floxed Fbxw7 allele was confirmed by PCR analysis of genomic DNA as described previously (28). Blg-Cre/Fbxw7F/F mice were also crossed with mice lacking Trp53 (Trp53−/−) mice. Expression of the Blg-Cre transgene in mammary epithelial cells increases from midgestation to lactation, and, to ensure maximal loss of Fbxw7 function in the mammary gland, we allowed female mice to go through three rounds of pregnancy. The mice were then set aside for tumor development, and they were monitored for palpable tumors weekly. We analyzed inguinal glands exclusively in mice of each genotype. All mouse experimental protocols were approved by the Institutional Animal Care and Use Committee of Kyushu University (Fukuoka, Japan).

Cell lines and cell culture

HEK293T cells were purchased and authenticated from ATCC, and cultured in DMEM supplemented with 10% FBS. Established mammary tumor cell lines were authenticated by RNA-sequencing. Mycoplasma contamination for all cell lines was checked using MycoAlert (Lonza) when frozen cultures were recovered, and cells were not passaged more than 30 times.

Whole-mount and histologic analyses of mammary glands

The inguinal mammary gland was excised, fully spread on a glass slide, and fixed in Carnoy solution for 8 hours. The tissue was then immersed in 70% ethanol for 15 minutes, gradually transitioned to distilled water, stained overnight with carmine alum, dehydrated with a graded series of ethanol solutions, and cleared in xylene. Each gland was finally mounted with the use of Permount (Thermo Fisher Scientific). For histologic analysis, tissue was fixed with 4% paraformaldehyde in PBS, embedded in paraffin, sectioned at a thickness of 4 μm, and stained with hematoxylin and eosin.

Immunofluorescence microscopy

Mammary glands were fixed with 4% paraformaldehyde in PBS and sectioned at a thickness of 4 μm for immunostaining with antibodies to Ki-67 (Novocastra), to Her2 (29D8, Cell Signaling Technology), to mouse CK5 (AF 138, Covance), to mouse CK14 (AF 64, Covance), to CK18 (RGE53, Santa Cruz Biotechnology), to p53 (FL-393, Santa Cruz Biotechnology), to ERα (MC-20, Santa Cruz Biotechnology), and to PR (hPRa 7, Thermo Fisher Scientific). Immune complexes were detected with Alexa Fluor 488– or Alexa Fluor 546–conjugated goat antibodies to mouse or rabbit IgG (Invitrogen). Nuclei were counterstained with Hoechst 33342 (Sigma-Aldrich).

TUNEL assay

The TUNEL (terminal deoxynucleotidy transferase–mediated dUTP nick-end labeling) assay was performed as described previously (32). In brief, paraffin-embedded sections of mammary gland were treated with H2O2, permeabilized for 15 minutes at 37°C with proteinase K (20 μg/mL, Sigma-Aldrich), and then incubated for 1 hour at 37°C with a reaction mixture containing terminal deoxynucleotidyl transferase (Invitrogen) and biotinylated dUTP (Boehringer Ingelheim). Labeled DNA was visualized with an ABC Kit (Vector Laboratories) and diaminobenzidine.

Immunoblot analysis

Total protein extracts were prepared from mammary gland with RIPA buffer. The extracts (30 μg) were subjected to immunoblot analysis as described previously (33) with antibodies to cyclin E (M-20), to p53 (Pab 240), to Mdm2 (SMP14), to c-Myc (N-262), to β-casein (B-5), to Notch2 (M-20), to Notch3 (M-20), and to Notch4 (H-225), all of which were obtained from Santa Cruz Biotechnology; with those to Aurora A and to cyclin B1 (BD Transduction Laboratories); with that to p63 (4A4, Thermo Fisher Scientific); with those to NICD1, to Snail (C15D3), and to Slug (C19G7), all from Cell Signaling Technology; with that to hemagglutinin (HA; HA11, Babco); and with that to FLAG (M2, Sigma-Aldrich). As a control, each membrane was stripped and then probed with antibodies to Hsp90 (BD Transduction Laboratories) or to GAPDH (Santa Cruz Biotechnology).

Reverse transcription and real-time PCR analysis

Total RNA was extracted from mammary gland by the guanidinium thiocyanate–phenol-chloroform method, purified, and subjected (1 μg) to reverse transcription (RT) with random hexanucleotide primers and ReverTra Ace α (Toyobo). The resulting cDNA was subjected to real-time PCR analysis in a reaction mixture that contained 1 × SYBR Green PCR Master Mix (Applied Biosystems) and 200 nmol/L gene-specific primers. Assays were performed in triplicate with an ABI StepOne Plus Real-Time PCR System (Applied Biosystems). The PCR protocol comprised 40 cycles of incubation at 60°C for 30 seconds and 95°C for 5 seconds. The sequences of the PCR primers (sense and antisense, respectively) were 5′-AAAATGCTGCACACTGCAGG-3′ and 5′-CGAGTCCTTCAATGATGCTCAG-3′ for Hey1, 5′-AAACGACCTCCGAAAGCGA-3′ and 5′-CGGTGAATTGGACCTCATCACT-3′ for Hey2, 5′-GAGGAACATGCAGGACCTGG-3′ and 5′-TTCATGTAGGCAGCATCCAC-3′ for Krt5, 5′-TGGCTGCCGATGACTTCC-3′ and 5′-GGCTCTCAATCTGCATCTCC-3′ for Krt14, 5′-ATGACACCAACATCACAAGG-3′ and 5′-AGAGCTGGCAATCTGGGC-3′ for Krt18, 5′-CATGTGTGTGACTGTGAAGG-3′ and 5′-TCCGTAGAAACAGTAGGAGC-3′ for Cdh1, 5′-CAGGAGTCAAACGAGTACCG-3′ and 5′-GTGTCCTGGTAGTTAGCAGC-3′ for Vim, 5′-GTGCTTGGGCTTTGGCAGTG-3′ and 5′-GAGGCATGTGCAGCTCATCC-3′ for Fn1, and 5′-GGAACATAGCCGTAAACTGC-3′ and 5′-TCACTGTGCCTGAACTTACC-3′ for the β-tubulin gene. Reactions for β-tubulin mRNA were performed concurrently on the same plate as those for the test mRNAs, and results were normalized by the corresponding amount of β-tubulin mRNA.

Preparation of single-cell suspensions from mammary tumor tissue

Mammary tumors were dissected and digested for 8 hours at 37°C in EpiCult-B supplemented with 5% FBS, collagenase (300 U/mL), and hyaluronidase (100 U/mL). After lysis of red blood cells by the addition of NH4Cl, a single-cell suspension was obtained through sequential dissociation of the tissue fragments by gentle pipetting for 1 to 2 minutes in the presence of 0.25% trypsin and then for 2 minutes in the presence of dispase II (5 mg/mL) and DNase I (0.1 mg/mL, Sigma-Aldrich) followed by filtration through a 40-μm mesh. All reagents were from StemCell Technologies unless otherwise specified.

Establishment of mammary tumor cell lines

Single-cell suspensions of mammary tumor cells were maintained in DMEM supplemented with 10% FBS, and 10 cell lines were cloned by limiting dilution. Deletion of exon 5 of the floxed Fbxw7 allele was confirmed by PCR analysis of genomic DNA after cloning. For generation of Notch1- and p63-deficient cells, complementary oligonucleotides containing the mouse Notch1 and p63 sgRNA target sequences were annealed and cloned into the BbsI site of pX330 (Addgene). The target sequences of the sgRNAs (sense and antisense, respectively) were 5′-CACCCCAAGTGGGACCTGCCTGAA-3′ and 5′-AAACTTCAGGCAGGTCCCACTTGG-3′ for Notch1, and 5′-CACCCCACGCACAGAATAGCGTGA-3′ and 5′-AAACTCACGCTATTCTGTGCGTGG-3′ for p63. For generation of the targeting vector for homologous recombination, puromycin resistance gene driven by phosphoglycerate kinase promoter flanked by 5′ and 3′ regions of homology containing 1-kb upstream or downstream of the second exon of Notch1 or forth exon of p63 were incorporated into pBSII-SK+ (Addgene). The sgRNA vector and the targeting vector were introduced into cloned tumor cells with the use of the Lipofectamine 2000 (Thermo Fisher Scientific), and the cells were then subjected to selection in medium containing puromycin (5 μg/mL).

Flow cytometry

Single-cell suspensions of mammary tumor cell lines were stained with antibodies to CD44 and to CD24 (clones IM7 and 30-F1, respectively, both from eBioscience) and were then analyzed with a FACSCalibur instrument (BD).

Tumorigenesis assay

Cells (1 × 107) were injected subcutaneously into 8-week-old female nude mice (Kyudo). Tumor size was measured at the indicated times after cell injection.

Gene expression in cultured cells by retroviral infection

A cDNA encoding mouse Fbxw7 was subcloned into the retroviral vector pMX-puro (provided by T. Kitamura, University of Tokyo, Tokyo, Japan). The resulting construct was used to prepare a recombinant ecotropic retrovirus for infection of target cells as described previously (28).

RNA sequencing and data analysis

RNA sequencing was performed as described previously (34), with minor modifications. Prepared libraries were sequenced with a NovaSeq 6000 system (Illumina). Paired-end reads were mapped to mm10 (UCSC) with the use of HISAT (v2.2.0) and the mm10 GTF file. Data were processed for calculation of FPKM (fragments per kilobase of exon model per million reads mapped) values with the use of Cuffdiff (v2.2.1). Gene ontology (GO) analysis was performed with the use of PANTHER (v15.0; ref. 35). Somatic variant calling was performed as described previously (36), with minor modifications. The aligned reads in BAM format were sorted, and duplicate reads were flagged (MarkDuplicates, Picard v2.9.2). Variant calling was conducted with the use of Mutect2 (GATK v4.1.4.1), and the resulting VCF files were annotated with SNPdat (v1.0.5). Variants annotated as “within CDS” and “Nonsynonymous” were counted. Normalized expression data from Cuffdiff were analyzed with gene set enrichment analysis (GSEA) software (v1.8.0, Broad Institute) and the Molecular Signatures Database (v6.0). Transcriptomic structural variations (TSV) were detected with SQUID (v1.5; ref. 37) and visualized with ClicO FS (38). Enrichment analysis of upstream transcription factors was performed with TRRUST, a manually curated database of human and mouse transcriptional regulatory networks, through the Metascape Web server (39). All sequence data have been deposited in the DDBJ sequence read archive under the accession number DRA 010579.

Clinical breast cancer data

We downloaded the largest public breast cancer cohort, METABRIC (40), from cBioPortal (https://www.cbioportal.org). Mouse differentially expressed genes (DEG) were converted to their human orthologs with the use of the NCBI Gene database, and expression profiles for the 858 human genes were extracted from the METABRIC cohort. Hierarchical clustering analysis was performed with the use of R (v3.6.1). Mutation analysis was performed with precomputed files provided in the repository.

Coimmunoprecipitation

A cDNA encoding mouse DNp63 was cloned from mouse mammary cells, and subcloned into pcDNA3.1(+) with HA epitope (pcDNA3-HA-mDNp63). For the coimmunoprecipitation assay, HEK293T cells were cotransfected with p3XFLAG-mFbxw7α or p3XFLAG-mFbxw7ΔF and pcDNA3-HA-mDNp63. Cell lysis, immunoprecipitation, and immunoblot analysis were performed as described previously (32).

Cycloheximide chase

HEK293T cells were cotransfected with p3XFLAG-mFbxw7α or p3XFLAG-mFbxw7ΔF and pcDNA3-HA-mDNp63. Cells were incubated in the presence of cycloheximide (100 μg/mL) for the indicated times and harvested. Cell lysates were subjected to immunoblot analysis.

Statistical analysis

Data are presented as means ± SD and were analyzed with the two-tailed Student t test unless otherwise noted. The log-rank test was applied to Kaplan–Meier analysis. For extraction of DEGs, GO analysis, and GSEA, we used an adjusted P value (Q value, FDR) of <0.05. The HR was calculated with the Cox proportional hazard model. Fisher exact test was performed for the data in Figs. 1B and 7E. A P value of <0.05 was considered statistically significant, with the exception of enrichment analysis for transcription factors (Fig. 6H and I), for which we used the default criterion of the established method (P <0.01).

Figure 1.

Fbxw7 deletion in mammary epithelial cells results in severe atrophy of mammary glands and a lactation defect. A, Genomic PCR analysis for the mammary gland of mice of the indicated genotypes. The positions of amplified fragments corresponding to wild-type (WT), floxed, and ΔE5 (exon 5–deleted) alleles are indicated. B, Suckling behavior of the offspring of Blg-Cre/Fbxw7F/F dams (left) and frequency of successful nursing by Blg-Cre/Fbxw7F/F dams (right). The number of deliveries by females of each genotype and the number of successful weanings were determined without taking into account the number of pups in each delivery. A successful weaning was thus defined as one in which more than three pups survived. *, P < 0.05 (Fisher exact test). C, Gross appearance of neonates born to Blg-Cre/Fbxw7+/F (control) or Blg-Cre/Fbxw7F/F females. Dilated stomachs (arrowheads) were apparent in neonates born to control females, whereas stomachs (arrowheads) were barely detectable for those born to Blg-Cre/Fbxw7F/F females. Skin turgor was also decreased for pups of the Blg-Cre/Fbxw7F/F dams. Bar, 10 mm. D, Stomach of neonates born to control or Blg-Cre/Fbxw7F/F dams. Bar, 2 mm.E, Whole-mount analysis of mammary glands stained with carmine alum. Normal (Blg-Cre/Fbxw7+/F) mammary glands begin to develop from midgestation, with acini undergoing marked expansion from 18.5 dpc to delivery. In contrast, mammary glands of Blg-Cre/Fbxw7F/F females failed to develop throughout pregnancy and delivery. Bars, 400 μm (top), 200 μm (middle), or 100 μm (bottom). F, Cell number for inguinal mammary glands of Blg-Cre/Fbxw7+/F (control) or Blg-Cre/Fbxw7F/F mice. Single cells were isolated from glands immediately after delivery and counted. Data are means + SD for five mice of each genotype. *, P < 0.05 (two-tailed Student t test). G, Immunoblot (IB) analysis of β-casein in extracts of mammary glands of Blg-Cre/Fbxw7+/F or Blg-Cre/Fbxw7F/F mice at the indicated times. Hsp90 was analyzed as a loading control. Numbers represent the relative β-casein/Hsp90 ratio determined by densitometry.

Figure 1.

Fbxw7 deletion in mammary epithelial cells results in severe atrophy of mammary glands and a lactation defect. A, Genomic PCR analysis for the mammary gland of mice of the indicated genotypes. The positions of amplified fragments corresponding to wild-type (WT), floxed, and ΔE5 (exon 5–deleted) alleles are indicated. B, Suckling behavior of the offspring of Blg-Cre/Fbxw7F/F dams (left) and frequency of successful nursing by Blg-Cre/Fbxw7F/F dams (right). The number of deliveries by females of each genotype and the number of successful weanings were determined without taking into account the number of pups in each delivery. A successful weaning was thus defined as one in which more than three pups survived. *, P < 0.05 (Fisher exact test). C, Gross appearance of neonates born to Blg-Cre/Fbxw7+/F (control) or Blg-Cre/Fbxw7F/F females. Dilated stomachs (arrowheads) were apparent in neonates born to control females, whereas stomachs (arrowheads) were barely detectable for those born to Blg-Cre/Fbxw7F/F females. Skin turgor was also decreased for pups of the Blg-Cre/Fbxw7F/F dams. Bar, 10 mm. D, Stomach of neonates born to control or Blg-Cre/Fbxw7F/F dams. Bar, 2 mm.E, Whole-mount analysis of mammary glands stained with carmine alum. Normal (Blg-Cre/Fbxw7+/F) mammary glands begin to develop from midgestation, with acini undergoing marked expansion from 18.5 dpc to delivery. In contrast, mammary glands of Blg-Cre/Fbxw7F/F females failed to develop throughout pregnancy and delivery. Bars, 400 μm (top), 200 μm (middle), or 100 μm (bottom). F, Cell number for inguinal mammary glands of Blg-Cre/Fbxw7+/F (control) or Blg-Cre/Fbxw7F/F mice. Single cells were isolated from glands immediately after delivery and counted. Data are means + SD for five mice of each genotype. *, P < 0.05 (two-tailed Student t test). G, Immunoblot (IB) analysis of β-casein in extracts of mammary glands of Blg-Cre/Fbxw7+/F or Blg-Cre/Fbxw7F/F mice at the indicated times. Hsp90 was analyzed as a loading control. Numbers represent the relative β-casein/Hsp90 ratio determined by densitometry.

Close modal

Severe developmental defects of the Fbxw7-deficient mammary gland

We previously generated mice that harbor a floxed Fbxw7 allele in which exon 5 (which encodes the F-box domain) is flanked by loxP sites (26). To analyze Fbxw7 function in mammary epithelial cells, we crossed these Fbxw7F/F mice with mice harboring a Cre transgene under the control of the Blg promoter (Blg-Cre mice). This promoter was shown to activate the transgene in secretory epithelial cells of the mammary gland (41). The expression of the Cre transgene is low in the gland of virgin mice, but it increases gradually during gestation, birth, and lactation (31). Development of the mammary gland appeared to proceed normally in young Blg-Cre/Fbxw7F/F females (Supplementary Fig. S1). To ensure maximal deletion of the floxed Fbxw7 allele by Cre in mammary epithelial cells, we subjected Blg-Cre/Fbxw7F/F females to three rounds of pregnancy. This procedure resulted in almost complete loss of the floxed allele in the mammary gland of Blg-Cre/Fbxw7F/F mice (Fig. 1A).

We examined lactation by Blg-Cre/Fbxw7F/F females after crossing them with C57BL/6 (Fbxw7+/+) males. Most neonates (Fbxw7+/F) born to the mutant females died within 2 to 3 days after birth (Fig. 1B). Although suckling behavior of the newborns appeared normal (Fig. 1B), they had dry skin and contained little milk in the stomach (Fig. 1C and D). This morbidity of the neonates was rescued by their transfer to wild-type foster mothers, with all three such neonates transferred being weaned successfully, suggesting that the Blg-Cre/Fbxw7F/F dams were responsible for the death of their offspring. Consistent with these observations, whole-mount analysis revealed that the acini of Blg-Cre/Fbxw7F/F mammary glands were markedly atrophic and appeared to store no milk, even immediately after delivery (Fig. 1E). In addition, the cell number for (Fig. 1F) and amount of β-casein produced by (Fig. 1G) inguinal mammary glands on the day of delivery were markedly reduced for Blg-Cre/Fbxw7F/F mice compared with Blg-Cre/Fbxw7+/F mice. These findings thus suggested that Blg-Cre/Fbxw7F/F dams are defective in lactation, and that the mutant mammary gland does not develop normally after midgestation, when the Blg promoter becomes active.

Excessive proliferation and massive apoptosis in the Fbxw7-deficient mammary gland

Given that Fbxw7 targets many regulators of the cell cycle and apoptosis including cyclin E, c-Myc, Notch, c-Jun, and Mcl-1 for degradation (22, 23), we examined cell proliferation and apoptosis in the Fbxw7-deficient mammary gland. Microscopic examination of mammary histological sections revealed that overall development and maturation of the mammary gland in Blg-Cre/Fbxw7F/F females were disturbed and delayed compared with those in control mice at 16.5 days postcoitum (dpc; Fig. 2A). In control mice, the number of mammary epithelial cells was increased at 16.5 dpc, and acini of the mammary gland had begun to dilate, with each cell containing many droplets in the cytoplasm, reflecting functional maturation for the production of milk. In contrast, mammary epithelial cells showed only a moderate level of proliferation in Blg-Cre/Fbxw7F/F females at 16.5 dpc. The defect in gland maturation in Fbxw7-deficient females was most pronounced immediately after delivery (Fig. 2A), with the immature glands containing only a small number of acini and the acini being abnormally dilated. Three weeks after weaning or neonatal death, most acini in the glands of Blg-Cre/Fbxw7F/F females remained abnormally dilated, although some of them had shrunk to normal size (Fig. 2A).

Figure 2.

Fbxw7 deletion induces excessive proliferation and massive apoptosis in mammary glands. A, Hematoxylin and eosin staining of mammary gland sections from Blg-Cre/Fbxw7+/F (control) and Blg-Cre/Fbxw7F/F female mice during pregnancy at 16.5 dpc as well as immediately after delivery and at 3 weeks after weaning. Bars, 50 μm. B, Immunofluorescence staining of Ki-67 in mammary gland sections of Blg-Cre/Fbxw7+/F and Blg-Cre/Fbxw7F/F mice at the indicated times (left). Bars, 20 μm. The proportion of Ki-67–positive cells was determined from three randomly selected images (right), with the data being means + SD for the technical triplicates. N.D., not detected. *, P < 0.05 (two-tailed Student t test). C, TUNEL staining for mammary gland sections of Blg-Cre/Fbxw7+/F and Blg-Cre/Fbxw7F/F mice immediately after delivery (left). Arrowheads, TUNEL-positive cells. Bars, 20 μm. The proportion of TUNEL-positive cells was determined from three randomly selected images (right), with the data being means + SD for the technical triplicates. N.D., not detected. *, P < 0.05 (two-tailed Student t test). D, Immunoblot analysis of the indicated proteins in protein extracts prepared from the mammary glands of Blg-Cre/Fbxw7+/F and Blg-Cre/Fbxw7F/F female mice immediately after delivery. The relative intensity of each band normalized by that of Hsp90 is shown. E, The relative indicated each protein/Hsp90 band intensity ratio (mean + SD) determined as in D for three mice of each genotype. *, P < 0.05 (two-tailed Student t test).

Figure 2.

Fbxw7 deletion induces excessive proliferation and massive apoptosis in mammary glands. A, Hematoxylin and eosin staining of mammary gland sections from Blg-Cre/Fbxw7+/F (control) and Blg-Cre/Fbxw7F/F female mice during pregnancy at 16.5 dpc as well as immediately after delivery and at 3 weeks after weaning. Bars, 50 μm. B, Immunofluorescence staining of Ki-67 in mammary gland sections of Blg-Cre/Fbxw7+/F and Blg-Cre/Fbxw7F/F mice at the indicated times (left). Bars, 20 μm. The proportion of Ki-67–positive cells was determined from three randomly selected images (right), with the data being means + SD for the technical triplicates. N.D., not detected. *, P < 0.05 (two-tailed Student t test). C, TUNEL staining for mammary gland sections of Blg-Cre/Fbxw7+/F and Blg-Cre/Fbxw7F/F mice immediately after delivery (left). Arrowheads, TUNEL-positive cells. Bars, 20 μm. The proportion of TUNEL-positive cells was determined from three randomly selected images (right), with the data being means + SD for the technical triplicates. N.D., not detected. *, P < 0.05 (two-tailed Student t test). D, Immunoblot analysis of the indicated proteins in protein extracts prepared from the mammary glands of Blg-Cre/Fbxw7+/F and Blg-Cre/Fbxw7F/F female mice immediately after delivery. The relative intensity of each band normalized by that of Hsp90 is shown. E, The relative indicated each protein/Hsp90 band intensity ratio (mean + SD) determined as in D for three mice of each genotype. *, P < 0.05 (two-tailed Student t test).

Close modal

Immunofluorescence analysis with antibodies to Ki-67 revealed a higher level of cell proliferation for Fbxw7-deficient mammary epithelial cells than for control cells before the day of delivery (Fig. 2B). We also examined the frequency of apoptosis with the TUNEL assay and found that it was also increased in Fbxw7-deficient mammary epithelial cells at the time of delivery (Fig. 2C). Cells positive for Ki-67 or TUNEL were not detected in mammary glands of control or mutant mice at 3 weeks after weaning. These results thus suggested that the loss of Fbxw7 results in transient promotion of cell-cycle progression followed by induction of apoptosis in the mammary gland, similar to the effects of Fbxw7 ablation in T lymphocytes, hematopoietic stem cells, and hepatocytes (26–28). Immunoblot analysis revealed that the abundance of neither cyclin E nor c-Myc, two major targets of Fbxw7, was affected by the loss of Fbxw7 in mammary glands (Fig. 2D). In contrast, the expression of other cell-cycle promoters such as Aurora A and cyclin B as well as that of p53 and its downstream target Mdm2 were increased in the mammary gland of Blg-Cre/Fbxw7F/F females immediately after delivery (Fig. 2D and E). These changes are likely secondary effects of the transient promotion of the cell cycle by Fbxw7 deletion, however, and such abnormal activation of the cell cycle may trigger p53-dependent apoptosis, resulting in the defective maturation of the mammary gland apparent in Fbxw7-deficient mice.

Accumulation of Notch1 and p63 associated with abnormal differentiation of Fbxw7-deficient mammary glands

Mature mammary gland acini are composed predominantly of two types of mammary epithelial cells—luminal cells and myoepithelial cells—both of which are required for proper gland development and lactation. We therefore investigated whether Fbxw7 deficiency influences the differentiation of mammary epithelial cells by immunofluorescence analysis of cytokeratin (CK) 14 and CK18, markers for myoepithelial cells and luminal cells, respectively (Fig. 3A). At 16.5 dpc, CK14-positive myoepithelial cells surrounded CK18-positive luminal cells in the acini of mammary glands from control pregnant female mice, whereas the number of CK18-positive cells was substantially decreased and the extent of CK14 expression was markedly increased in Fbxw7-mutant mice. At the time of delivery, these abnormalities were even more prominent in the mutant mice, indicating that Fbxw7-deficient luminal cells did not differentiate or expand. Involution of the mammary gland was complete at 3 weeks after weaning in control females, but it remained incomplete in the mutant mice at this time, with some acini still being slightly dilated. We also confirmed these changes in CK14 and CK18 expression in Fbxw7-mutant mice by immunoblot analysis (Fig. 3B). These observations suggested that Fbxw7 is indispensable for the appropriate differentiation, expansion, and involution of mammary epithelial cells associated with the lactation period.

Figure 3.

Abnormal differentiation of Fbxw7-deficient mammary glands associated with accumulation of Notch1 and p63. A, Immunofluorescence staining for CK14 (green) and CK18 (red) in mammary glands of Blg-Cre/Fbxw7+/F (control) and Blg-Cre/Fbxw7F/F female mice at 16.5 dpc, immediately after delivery, and at 3 weeks after weaning. Bars, 50 μm. B, Immunoblot analysis of CK14 and CK18 in protein extracts of mammary glands of mice as in A. Numbers represent the relative CK14/Hsp90 or CK18/Hsp90 ratio determined by densitometry.C, Immunoblot analysis of NICD1 and p63 in protein extracts of mammary glands of mice as in A. Numbers represent the relative NICD1/Hsp90 or p63/Hsp90 ratio determined by densitometry. D, The relative p63/Hsp90 band intensity ratio immediately after delivery (mean + SD) determined as in C for three mice of each genotype. *, P < 0.05 (two-tailed Student t test). E, RT and real-time PCR analysis of Notch target gene (Hey1, Hey2), p63 target gene (Krt5, Krt14), and Krt18 expression in mammary glands of mice as in A. Normalized data are expressed relative to the corresponding value for control mice and are means + SD from biological triplicates. *, P < 0.05 (two-tailed Student t test).

Figure 3.

Abnormal differentiation of Fbxw7-deficient mammary glands associated with accumulation of Notch1 and p63. A, Immunofluorescence staining for CK14 (green) and CK18 (red) in mammary glands of Blg-Cre/Fbxw7+/F (control) and Blg-Cre/Fbxw7F/F female mice at 16.5 dpc, immediately after delivery, and at 3 weeks after weaning. Bars, 50 μm. B, Immunoblot analysis of CK14 and CK18 in protein extracts of mammary glands of mice as in A. Numbers represent the relative CK14/Hsp90 or CK18/Hsp90 ratio determined by densitometry.C, Immunoblot analysis of NICD1 and p63 in protein extracts of mammary glands of mice as in A. Numbers represent the relative NICD1/Hsp90 or p63/Hsp90 ratio determined by densitometry. D, The relative p63/Hsp90 band intensity ratio immediately after delivery (mean + SD) determined as in C for three mice of each genotype. *, P < 0.05 (two-tailed Student t test). E, RT and real-time PCR analysis of Notch target gene (Hey1, Hey2), p63 target gene (Krt5, Krt14), and Krt18 expression in mammary glands of mice as in A. Normalized data are expressed relative to the corresponding value for control mice and are means + SD from biological triplicates. *, P < 0.05 (two-tailed Student t test).

Close modal

Among the various molecules implicated in the control of cell fate decisions during mammary epithelial development, we focused on Notch and p63. These proteins have been found to be targeted by Fbxw7 (19, 42) in addition to contributing to fate determination in mammary epithelial cells (43–45). Immunoblot analysis revealed that the abundance of the Notch1 intracellular domain (NICD1) in the mammary gland of Blg-Cre/Fbxw7F/F females was moderately increased compared with that in Blg-Cre/Fbxw7+/F females both at 16.5 dpc and immediately after delivery (Fig. 3C). At 3 weeks after weaning, the signal for NICD1 was almost undetectable in control mice, whereas it remained at a level similar to that apparent on the day of delivery in Blg-Cre/Fbxw7F/F mice. The abundance of Notch2, Notch3, or Notch4 in the mammary gland was unaffected by the loss of Fbxw7 (Supplementary Fig. S2). The expression of p63 was also increased in the mammary gland of Blg-Cre/Fbxw7F/F mice compared with that of control mice at 16.5 dpc and immediately after delivery (Fig. 3C and D), whereas it was slightly decreased in the mutant mice at 3 weeks after weaning. This pattern of p63 expression may therefore not be attributable solely to impairment of protein degradation in the absence of Fbxw7; it might instead also reflect a secondary effect of the abnormal differentiation of the Fbxw7-mutant mammary gland.

We also examined the expression of downstream target genes of Notch and p63 by RT and real-time PCR analysis. The expression of both Notch target genes (Hey1 and Hey2) and p63 target genes (Krt5 and Krt14) was markedly increased in the mutant mammary glands compared with the control mammary glands at 16.5 dpc and immediately after delivery (Fig. 3E). Expression of the Krt18 gene (which encodes CK18) was decreased in the Fbxw7-mutant mammary gland, consistent with the loss of luminal cells. Collectively, these results thus suggested that the accumulation of Notch1 and p63 proteins due to Fbxw7 ablation might result in the abnormal differentiation of mammary epithelial cells in a manner dependent on activation of their downstream target genes.

Predisposition to mammary tumor development conferred by loss of Fbxw7

Despite severe atrophy of the mammary gland, Blg-Cre/Fbxw7F/F female mice developed mammary tumors spontaneously within a 2-year observation period (Fig. 4A). Biallelic excision of floxed Fbxw7 in the mammary tumors was confirmed by genomic PCR analysis (Fig. 4B). Although most of the mice with mammary tumors did not manifest gross metastasis, a few of the mutant animals had multiple metastatic lesions in the lung and liver (Fig. 4C). Biallelic excision of floxed Fbxw7 was also confirmed in these metastatic lesions (Fig. 4D). Histopathologic examination revealed that some of the mammary tumors had a morphology typical of adenocarcinoma with apparent progressive proliferation of the atypical glands (Fig. 4E). However, other mammary tumors were composed predominantly of spindle-shaped cells or manifested squamous metaplasia, geographic necrosis, or an invasive border (Fig. 4E), all of which are morphologic characteristics of BLC (46, 47). We then examined expression of the myoepithelial/basal cell markers CK5 and CK14 as well as of Her2, ERα, and PR in the tumors by immunofluorescence analysis. As expected, all mammary tumors that developed in the Fbxw7-mutant mice were negative for Her2, ERα, and PR, and most of them contained lesions positive for CK5 and CK14 (Fig. 4F; Supplementary Table S1; Supplementary Fig. S3). These results thus indicated that the mammary tumors developed by Blg-Cre/Fbxw7F/F female mice have features of BLC.

Figure 4.

Blg-Cre/Fbxw7F/F female mice develop spontaneous mammary tumors. A, Gross appearance of a mammary tumor (arrowheads) developed by a Blg-Cre/Fbxw7F/F mouse at 78 weeks of age. B, Genomic PCR analysis of mammary tumors developed by Blg-Cre/Fbxw7F/F mice. The positions of amplified fragments corresponding to floxed and ΔE5 alleles of Fbxw7 are indicated. Tail DNA was examined as a control. C, Gross appearance of lung (left) and liver (right) of Blg-Cre/Fbxw7F/F mice with multiple metastatic lesions (arrowheads). D, Genomic PCR analysis of normal tissue (N) and metastatic lesions (M) of the lung and liver of Blg-Cre/Fbxw7F/F mice. E, Hematoxylin and eosin staining of mammary tumors developed by Blg-Cre/Fbxw7F/F mice. The tumors were highly heterogeneous, showing typical adenocarcinoma (left), spindle-shaped tumor cells arranged in short fascicles (middle; arrowheads), or squamous metaplasia and geographic necrosis (right; arrowheads and arrow, respectively). Bars, 50 μm (left and middle) or 100 μm (right). F, Immunofluorescence staining for CK5 or CK14 (green) in mammary tumor and normal mammary gland sections. Nuclei were stained with Hoechst 33342 (blue). Bars, 50 μm. G, Immunofluorescence staining for p53 (green) in mammary tumors developed by Blg-Cre/Fbxw7F/F mice. Nuclei were stained with Hoechst 33342 (blue). Tumor 3 was negative and tumor 5 positive for nuclear expression of p53. Bars, 50 μm. H, Mammary tumor–free rate for cohorts of the indicated numbers of Blg-Cre/Fbxw7+/F (control), Blg-Cre/Fbxw7F/F, Trp53+/−, and Blg-Cre/Fbxw7F/F/Trp53+/− mice. *, P = 0.0342; †, P < 0.0001 (log-rank test). I, Hematoxylin and eosin staining of mammary tumors developed by Blg-Cre/Fbxw7F/F/Trp53+/− mice showing squamous metaplasia (left), spindle-shaped tumor cells (middle), and geographic necrosis (right; arrows). Bars, 50 μm. J, Immunofluorescence staining for CK5 or CK14 (green) in mammary tumors developed by Blg-Cre/Fbxw7F/F/Trp53+/− mice. Nuclei were stained with Hoechst 33342 (blue). Bars, 50 μm. K, Immunoblot analysis of NICD1, p63, CK14, and CK18 in protein extracts of mammary tumors developed by Blg-Cre/Fbxw7F/F or Blg-Cre/Fbxw7F/F/Trp53+/− mice. GAPDH was analyzed as a loading control.

Figure 4.

Blg-Cre/Fbxw7F/F female mice develop spontaneous mammary tumors. A, Gross appearance of a mammary tumor (arrowheads) developed by a Blg-Cre/Fbxw7F/F mouse at 78 weeks of age. B, Genomic PCR analysis of mammary tumors developed by Blg-Cre/Fbxw7F/F mice. The positions of amplified fragments corresponding to floxed and ΔE5 alleles of Fbxw7 are indicated. Tail DNA was examined as a control. C, Gross appearance of lung (left) and liver (right) of Blg-Cre/Fbxw7F/F mice with multiple metastatic lesions (arrowheads). D, Genomic PCR analysis of normal tissue (N) and metastatic lesions (M) of the lung and liver of Blg-Cre/Fbxw7F/F mice. E, Hematoxylin and eosin staining of mammary tumors developed by Blg-Cre/Fbxw7F/F mice. The tumors were highly heterogeneous, showing typical adenocarcinoma (left), spindle-shaped tumor cells arranged in short fascicles (middle; arrowheads), or squamous metaplasia and geographic necrosis (right; arrowheads and arrow, respectively). Bars, 50 μm (left and middle) or 100 μm (right). F, Immunofluorescence staining for CK5 or CK14 (green) in mammary tumor and normal mammary gland sections. Nuclei were stained with Hoechst 33342 (blue). Bars, 50 μm. G, Immunofluorescence staining for p53 (green) in mammary tumors developed by Blg-Cre/Fbxw7F/F mice. Nuclei were stained with Hoechst 33342 (blue). Tumor 3 was negative and tumor 5 positive for nuclear expression of p53. Bars, 50 μm. H, Mammary tumor–free rate for cohorts of the indicated numbers of Blg-Cre/Fbxw7+/F (control), Blg-Cre/Fbxw7F/F, Trp53+/−, and Blg-Cre/Fbxw7F/F/Trp53+/− mice. *, P = 0.0342; †, P < 0.0001 (log-rank test). I, Hematoxylin and eosin staining of mammary tumors developed by Blg-Cre/Fbxw7F/F/Trp53+/− mice showing squamous metaplasia (left), spindle-shaped tumor cells (middle), and geographic necrosis (right; arrows). Bars, 50 μm. J, Immunofluorescence staining for CK5 or CK14 (green) in mammary tumors developed by Blg-Cre/Fbxw7F/F/Trp53+/− mice. Nuclei were stained with Hoechst 33342 (blue). Bars, 50 μm. K, Immunoblot analysis of NICD1, p63, CK14, and CK18 in protein extracts of mammary tumors developed by Blg-Cre/Fbxw7F/F or Blg-Cre/Fbxw7F/F/Trp53+/− mice. GAPDH was analyzed as a loading control.

Close modal

When combined with deletion of Trp53, targeted disruption of Brca1 or Rb1 in mice was shown to result in the formation of tumors with BLC characteristics and either metaplastic elements or EMT features, respectively, although mammary-specific deletion of Trp53 alone leads to the formation of tumors of multiple molecular subtypes (48, 49). We therefore investigated the Trp53 status of the mammary tumors developed by Blg-Cre/Fbxw7F/F mice. Immunofluorescence analysis revealed that three out of the four tumors examined manifested extensive expression of p53 in the nucleus (Fig. 4G; Supplementary Table S1; Supplementary Fig. S3). Given that expression of p53 detected by immunostaining often results from mutations that stabilize the protein (50), we sequenced exons 2 to 9 of Trp53 in these tumors. However, among the three p53-expressing tumors, we identified only one missense mutation of p53 (E282K, in the DNA-binding domain) in tumor 5. We speculated that the high intratumoral heterogeneity of the mammary tumors developed by Blg-Cre/Fbxw7F/F mice might make it difficult to detect Trp53 mutations. To determine whether loss of p53 function might influence the development of BLC induced by the deletion of Fbxw7, we crossed Blg-Cre/Fbxw7F/F mice with Trp53−/− mice. Although Blg-Cre/Fbxw7F/F/Trp53−/− mice were viable, most of the animals died 2 to 6 months after birth with other malignancies such as lymphoma before development of mammary tumors. To overcome this problem, we studied Blg-Cre/Fbxw7F/F/Trp53+/− mice. Although some mice of this genotype also died of malignancies other than mammary tumors, we were able to evaluate tumorigenicity in the mammary gland. After three rounds of pregnancy, Blg-Cre/Fbxw7F/F/Trp53+/− mice developed mammary tumors with significantly shorter latencies (log-rank test, P < 0.0001) compared with Blg-Cre/Fbxw7F/F mice (Fig. 4H; Supplementary Table S1). Histopathologic analysis of three independent tumors revealed that they all showed characteristics of BLC with squamous metaplasia, spindle-shaped tumor cells, and geographic necrosis (Fig. 4I). Indeed, immunofluorescence analysis showed that all three tumors expressed CK5 and none of them expressed ERα, PR, or Her2 (Fig. 4J; Supplementary Table S1; Supplementary Fig. S3). There is no obvious difference in expression of NICD1, p63, CK14, and CK18 between tumors developed in Blg-Cre/Fbxw7F/F mice and Blg-Cre/Fbxw7F/F/Trp53+/− mice (Fig. 4K). Loss of p53 function was thus found to promote earlier development and to increase the incidence of mammary tumors in Blg-Cre/Fbxw7F/F mice. Trp53 mutations may therefore also contribute to the development of BLC associated with Fbxw7 deletion.

BLC characteristics and intratumoral heterogeneity of mammary tumors developed by Fbxw7-deficient mice

To investigate further the intratumoral heterogeneity of Fbxw7-null mammary tumors, we established and cloned cell lines from one BLC-like tumor (tumor no. 3, in which p53 was not detected; Supplementary Table S1) developed by a Blg-Cre/Fbxw7F/F female mouse. Ten independent immortalized cell lines were obtained without the use of any specific procedures such as oncogene induction, and six representative lines (clones A–F) were subjected to further characterization. On the basis of their morphologic features, the cell lines were classified into two groups: epithelial type (clones A–C) and mesenchymal type (clones D–F; Fig. 5A). Immunoblot analysis revealed that expression of the EMT markers Snail and Slug (51) was markedly increased in the mesenchymal-type cell lines compared with the epithelial-type cell lines, in which these proteins were virtually undetectable (Fig. 5B). Consistent with these results, the abundance of mRNAs for the mesenchymal markers vimentin and fibronectin was increased in the mesenchymal-type cell lines, whereas the mRNA for E-cadherin, a marker for epithelial cells, was not detected (Fig. 5C; Supplementary Fig. S4). Only epithelial-type cell lines expressed CK14 (Fig. 5D). We also examined the expression of CD44 and CD24, two cell surface markers whose expression in the CD44high/CD24low configuration is associated with EMT and tumor-initiating cells (52). Whereas the epithelial-type cells manifested a CD44high/CD24high phenotype, the mesenchymal-type cells were CD44high/CD24low (Fig. 5E), suggesting that the latter cells have undergone EMT and may have tumor initiation ability.

Figure 5.

Characterization of basal-like mammary carcinoma developed by Blg-Cre/Fbxw7F/F mice. A, Phase-contrast images of epithelial-type and mesenchymal-type cell lines established and cloned from a mammary tumor developed by a Blg-Cre/Fbxw7F/F mouse. B, Immunoblot analysis of NICD1, p63, c-Myc, cyclin E, Snail, and Slug in protein extracts of six such epithelial-type (clones A–C) and mesenchymal-type (clones D–F) cell lines. C, RT and real-time PCR analysis of E-cadherin (Cdh1), vimentin (Vim), and fibronectin (Fn1) gene expression in clones A and F. Normalized data are expressed relative to the corresponding value for clone A and are means + SD from technical triplicates. *, P < 0.05 (two-tailed Student t test). D, Immunoblot analysis of CK14 and CK18 in protein extracts of the cell lines shown in A. E, Flow cytometric analysis of the cell surface markers CD44 and CD24 in the cell lines shown in A. The proportion of cells in left-upper region and in right-upper region was determined with the data being means + SD for the technical triplicates. F, Representative tumors formed 10 days after subcutaneous transplantation of cells of clone A or clone F into nude mice. G, Tumor volume measured 20 days after cell transplantation as in F. Data are means + SD for four tumors formed in two mice. H, Hematoxylin and eosin staining of tumors at 20 days after cell transplantation as in F. Bars, 50 μm.

Figure 5.

Characterization of basal-like mammary carcinoma developed by Blg-Cre/Fbxw7F/F mice. A, Phase-contrast images of epithelial-type and mesenchymal-type cell lines established and cloned from a mammary tumor developed by a Blg-Cre/Fbxw7F/F mouse. B, Immunoblot analysis of NICD1, p63, c-Myc, cyclin E, Snail, and Slug in protein extracts of six such epithelial-type (clones A–C) and mesenchymal-type (clones D–F) cell lines. C, RT and real-time PCR analysis of E-cadherin (Cdh1), vimentin (Vim), and fibronectin (Fn1) gene expression in clones A and F. Normalized data are expressed relative to the corresponding value for clone A and are means + SD from technical triplicates. *, P < 0.05 (two-tailed Student t test). D, Immunoblot analysis of CK14 and CK18 in protein extracts of the cell lines shown in A. E, Flow cytometric analysis of the cell surface markers CD44 and CD24 in the cell lines shown in A. The proportion of cells in left-upper region and in right-upper region was determined with the data being means + SD for the technical triplicates. F, Representative tumors formed 10 days after subcutaneous transplantation of cells of clone A or clone F into nude mice. G, Tumor volume measured 20 days after cell transplantation as in F. Data are means + SD for four tumors formed in two mice. H, Hematoxylin and eosin staining of tumors at 20 days after cell transplantation as in F. Bars, 50 μm.

Close modal

We next assessed the ability of the cell lines to form tumors in xenograft models based on subcutaneous transplantation into nude mice. Unexpectedly, both the epithelial-type and the mesenchymal-type cells were able to form new tumors, although the mesenchymal-type cells showed a tendency (not statistically significant) to form larger tumors compared with the epithelial-type cells (Fig. 5F and G). Histologic analysis revealed that the epithelial-type cells formed tumors with epithelial characteristics, showing atypical gland formations, whereas the mesenchymal-type cells formed tumors composed of spindle-shaped cells (Fig. 5H). These observations together were indicative of intratumoral heterogeneity for the mammary carcinomas developed by Blg-Cre/Fbxw7F/F mice, with more than one cell lineage possessing the characteristics of tumor-initiating cells.

Molecular basis for development of heterogeneous tumors in the Fbxw7-deficient mammary gland

To explore the mechanism underlying the development of tumors with prominent heterogeneity in Fbxw7-deficient mammary glands, we performed RNA-sequencing analysis for the normal mammary epithelium, an Fbxw7-deficient tumor (tumor no. 3; Supplementary Table S1), and the cell clones established from this tumor. The mutation frequency was almost twice as high for tumor cells than for cells of the normal mammary epithelium (Fig. 6A), and transcriptomic structural alterations were detected in the Fbxw7-deficient tumor cells (Fig. 6B). These results suggested that Fbxw7 loss increases the mutation rate and gives rise to chromosome instability, consistent with previous observations (53). Clustering analysis based on gene expression for the six independent tumor cell clones revealed that they could be divided into two groups consistent with the morphologic characterization: group 1 (epithelial type) and group 2 (mesenchymal type; Fig. 6C). We also identified 937 DEGs between epithelial- and mesenchymal-type clones (Fig. 6D and E). GO analysis of these DEGs for each group revealed that genes related to epithelial features were enriched in group 1 whereas those related to mesenchymal features were enriched in group 2 (Fig. 6F). GSEA showed that gene sets specifically expressed in human basal- or mesenchymal-type breast cancers were enriched in groups 1 and 2, respectively (Fig. 6G). We next explored upstream transcription factors for the DEGs, and found that they included many known Fbxw7 substrates in both group 1 and group 2 (Fig. 6H). Genes targeted by Notch1 or p63 were upregulated in group 1, whereas such a tendency was not apparent in group 2. These results suggested that hyperactivation of Notch1- or p63-dependent pathways might confer epithelial-type characteristics on Fbxw7-deficient mammary tumor cells. Of note, the Nfkb1 subunit of the transcription factor NFκB, a known target of Fbxw7, was included in both groups (Fig. 6H) and might therefore contribute to the heterogeneity of Fbxw7-deficient mammary tumors by lowering the barrier for cell fate transitions (see Discussion).

Figure 6.

Heterogeneity of mammary tumor cells formed as a result of Fbxw7 deficiency. A, Somatic variant calls determined by RNA-sequencing analysis of cells isolated from a mammary tumor developed by a Blg-Cre/Fbxw7F/F mouse (tumor no. 3) and those isolated from a normal mammary gland of an 8-week-old C57BL/6J female mouse. B, Circus plot of TSVs for tumor cells as in A. Scores represent the number of reads for each TSV. C, Hierarchical clustering of gene expression patterns for six independent cell clones derived from the tumor in A. The six clones were separated into two groups: group 1 (epithelial type) and group 2 (mesenchymal type). D, Volcano plot of DEGs for group 1 versus group 2 (pseudocount of 1 added to FPKM values). E, Heatmap of expression levels for DEGs in each clone. F, GO analysis of DEGs. Representative findings for each group are presented. G, GSEA of gene sets specifically expressed in human basal- or mesenchymal-type breast tumors for group 1 versus group 2. H, Enrichment analysis of transcription factors (TF) in group 1 and group 2 as in C. Venn diagram showed enriched transcription factors that are known substrates of Fbxw7 (top left). Enrichment analysis of transcription factors whose target genes were upregulated in group 1 compared with group 2 (bottom left), and transcription factors whose target genes were upregulated in group 2 compared with group 1 (right).

Figure 6.

Heterogeneity of mammary tumor cells formed as a result of Fbxw7 deficiency. A, Somatic variant calls determined by RNA-sequencing analysis of cells isolated from a mammary tumor developed by a Blg-Cre/Fbxw7F/F mouse (tumor no. 3) and those isolated from a normal mammary gland of an 8-week-old C57BL/6J female mouse. B, Circus plot of TSVs for tumor cells as in A. Scores represent the number of reads for each TSV. C, Hierarchical clustering of gene expression patterns for six independent cell clones derived from the tumor in A. The six clones were separated into two groups: group 1 (epithelial type) and group 2 (mesenchymal type). D, Volcano plot of DEGs for group 1 versus group 2 (pseudocount of 1 added to FPKM values). E, Heatmap of expression levels for DEGs in each clone. F, GO analysis of DEGs. Representative findings for each group are presented. G, GSEA of gene sets specifically expressed in human basal- or mesenchymal-type breast tumors for group 1 versus group 2. H, Enrichment analysis of transcription factors (TF) in group 1 and group 2 as in C. Venn diagram showed enriched transcription factors that are known substrates of Fbxw7 (top left). Enrichment analysis of transcription factors whose target genes were upregulated in group 1 compared with group 2 (bottom left), and transcription factors whose target genes were upregulated in group 2 compared with group 1 (right).

Close modal

We then performed clustering analysis for 1,904 human patients with breast cancer in the METABRIC cohort on the basis of the expression profiles for 858 human orthologs of the mouse DEGs identified above. Such analysis divided the patients into two clusters (Fig. 7A). Patients with luminal A or B subtypes of breast cancer were found predominantly in cluster 1, whereas those with BLC or CLC were enriched in cluster 2 (Fig. 7B). Analysis of the overall survival rate showed that the prognosis of cluster 2 was worse than that of cluster 1 (Fig. 7C). Of note, most breast tumors with FBXW7 mutations alone were of the luminal type, whereas most of those harboring mutations in both FBXW7 and TP53 were BLC, CLC, or Her2+ (Fig. 7B), suggesting that mutations in TP53 in breast cancers with mutations in FBXW7 may lead to changes in tumor traits in humans. Collectively, our findings suggest that the loss of Fbxw7 might increase the mutation rate and chromosome instability and thereby induce tumorigenesis in the mammary gland. In addition, activation of signaling pathways governed by transcription factors that are targeted by Fbxw7, including NFκB, may give rise to prominent intratumoral heterogeneity.

Figure 7.

Fbxw7-deficient mammary tumors in mice resemble human breast cancers. A, Scheme for clustering analysis of patients with breast cancer. Hierarchical clustering based on the expression of human orthologs of the DEGs identified in epithelial versus mesenchymal cell clones derived from an Fbxw7-deficient mouse mammary tumor revealed two separate clusters in the METABRIC cohort of human patients with breast cancer. B, Most patients with luminal A or B tumors in the METABRIC cohort were categorized in cluster 1, whereas almost all (94%) patients with BLC or CLC were in cluster 2 (top). TP53 mutation status and tumor subtype for patients with breast cancer in the METABRIC cohort harboring FBXW7 mutations (middle). Relation between TP53 mutation status and tumor subtype in patients with FBXW7 mutations as evaluated with Fisher exact test. WT, wild type; Mut, mutant (bottom). C, Kaplan–Meier curves of 5-year overall survival rate for METABRIC patients in cluster 1 or cluster 2. The HR, its 95% confidence interval, and the log-rank P value are shown.

Figure 7.

Fbxw7-deficient mammary tumors in mice resemble human breast cancers. A, Scheme for clustering analysis of patients with breast cancer. Hierarchical clustering based on the expression of human orthologs of the DEGs identified in epithelial versus mesenchymal cell clones derived from an Fbxw7-deficient mouse mammary tumor revealed two separate clusters in the METABRIC cohort of human patients with breast cancer. B, Most patients with luminal A or B tumors in the METABRIC cohort were categorized in cluster 1, whereas almost all (94%) patients with BLC or CLC were in cluster 2 (top). TP53 mutation status and tumor subtype for patients with breast cancer in the METABRIC cohort harboring FBXW7 mutations (middle). Relation between TP53 mutation status and tumor subtype in patients with FBXW7 mutations as evaluated with Fisher exact test. WT, wild type; Mut, mutant (bottom). C, Kaplan–Meier curves of 5-year overall survival rate for METABRIC patients in cluster 1 or cluster 2. The HR, its 95% confidence interval, and the log-rank P value are shown.

Close modal

We here generated mice with mammary epithelial cell–specific ablation of Fbxw7 and thereby identified key roles for Fbxw7 in normal development of the mammary gland from midgestation to lactation as well as in tumor suppression. Almost all neonates born to dams with mammary epithelial cell–specific Fbxw7 deficiency did not survive for longer than 2 to 3 days after birth as a result of a lack of maternal milk availability. Pathologic analysis revealed that the impaired lactation was attributable to a defect in the development of mammary luminal cells. Our results suggest that certain substrates of Fbxw7 accumulate and promote transient proliferation of luminal cells followed by the induction of cell-cycle arrest and apoptosis mediated by p53 activation. Indeed, we detected activation of the cell cycle, p53 induction, and apoptosis in the Fbxw7-deficient mammary glands. Similar effects of Fbxw7 ablation were previously described for mature T cells, which showed accumulation of c-Myc and subsequent p53 activation leading to the promotion of apoptosis (26). Our present findings also resemble those for Fbxw7-deficient mature T cells in that Trp53 deletion reversed the fate of the T cells from one of suppressed proliferation to unrestrained growth and tumor formation. Unlike Fbxw7-deficient mature T cells, however, Fbxw7 substrates such as c-Myc and cyclin E that contribute to cell-cycle progression did not accumulate in Fbxw7-deficient mammary glands. Furthermore, additional ablation of p53 did not rescue the differentiation defect of Fbxw7-deficient mammary glands (Supplementary Table S2). We therefore propose that other effects of Fbxw7 ablation in addition to promotion of the cell cycle are responsible for this phenotype. It is also possible that loss of Fbxw7 might skew the differentiation of progenitor cells toward myoepithelial cells rather than luminal cells in a manner dependent on the accumulation of Fbxw7 substrates, and that the resulting cells might eventually undergo apoptosis.

Among previously identified Fbxw7 substrates, the abundance of Notch1 and p63 was found to be increased in Fbxw7-deficient mammary glands, with both of these proteins being known as key regulators of mammary epithelial cell fate (43–45). Consistent with our findings, p63 was previously shown to promote differentiation of mammary myoepithelial cells (54). In addition, mice transgenic for Notch1 were found to be unable to nurse their offspring (43). The similarity of these phenotypes to those of our Fbxw7-deficient mice suggests that the accumulation of Notch1 and p63 in the mutant mammary gland accounts, at least in part, for the defect in nursing of offspring.

With regard to p63, we detected the interaction of Fbxw7 and p63 by coimmunoprecipitation analysis of normal mammary glands, but overexpression of Fbxw7 did not promote the degradation of p63 in HEK293T cells under the conditions of our experiments (Supplementary Fig. S5), leaving open the question of whether p63 is indeed a direct target of Fbxw7. Although p63 was previously shown to be a target of the β isoform of Fbxw7 (Fbxw7β; ref. 42), female mice specifically lacking Fbxw7β (55) show no apparent abnormality in mammary development and are apparently able to nurse their pups normally. In addition, the Fbxw7β-deficient mice do not appear to develop mammary tumors (55). It is thus likely that the increased expression of p63 in the mammary gland of Blg-Cre/Fbxw7F/F mice is a result of the abnormal differentiation of mammary epithelial cells induced by the accumulation of another substrate of Fbxw7 such as Notch1. It is of note that loss of the Notch cofactor RBP-Jκ in mammary glands resulted in an increase in the number of myoepithelial cells (44), an effect similar to that of Fbxw7 ablation in this study. These results suggest that both excess and loss of Notch1 signaling might lead to defective mammary gland differentiation, and that Notch1 accumulation in Fbxw7-deficient mammary glands is primarily responsible for the differentiation defect and tumorigenesis in the mutant mice—although it remains possible that the accumulation of multiple known substrates of Fbxw7, or that of an unknown substrate, gives rise to the observed phenotypes.

Blg-Cre/Fbxw7F/F female mice spontaneously developed mammary tumors after a long latency period. In addition, given that heterozygous deletion of Trp53 in Blg-Cre/Fbxw7F/F mice shortened the latency and increased the incidence of mammary tumor development, the onset of mammary tumorigenesis on the Fbxw7-null background appears to be dependent on p53 status. The transient hyperproliferation followed by apoptosis of mammary epithelial cells as well as the mammary tumorigenesis dependent on p53 loss apparent in Blg-Cre/Fbxw7F/F mice are similar to effects on T cells observed in mice with T cell–specific Fbxw7 ablation (Lck-Cre/Fbxw7F/F mice; ref. 26). On the other hand, neither brain- nor liver-specific Fbxw7-deficient mice show increased tumorigenesis (28, 30). Mammary epithelial cells and thymocytes may therefore be more susceptible to oncogenic transformation than neuronal cells and hepatocytes in the absence of Fbxw7. This increased susceptibility of mammary epithelial cells and thymocytes may be due to their active proliferation and a consequent increased mutation rate of Trp53.

Our results suggest that loss of Fbxw7 may be a key event that contributes to the induction of BLC. Consistent with our observations in mice, a low expression level of FBXW7 was shown to be correlated with poor prognosis in human breast cancer (16). However, forced expression of Fbxw7 in mammary tumor cell lines established from a Blg-Cre/Fbxw7F/F mouse did not reverse their malignant phenotype (Supplementary Fig. S6), suggesting that the parent tumor had already lost its addiction to Fbxw7 deficiency. Given that loss of Fbxw7 induces chromosome instability in part as a result of the accumulation of cyclin E (53), tumor cells without functional Fbxw7 might readily develop other gene alterations. Consistent with this notion, our RNA-sequencing analysis revealed an increased mutation rate and transcriptomic structural alterations in Fbxw7-deficient mammary tumor cells. We recently showed that Fbxw7 contributes to the maintenance of disseminated tumor cells (56). The role of Fbxw7 as a brake of the cell cycle thus appears to be important for both normal cells and dormant cancer cells, with Fbxw7 ablation leading to carcinogenesis in the former cells and allowing eradication of the latter cells. Dormant cancer cells of mammary tumors developed by Blg-Cre/Fbxw7F/F mice likely rely on an Fbxw7-independent mechanism for their maintenance.

Our RNA-sequencing analysis of six independent tumor cell clones established from a BLC tumor that developed in an Fbxw7-deficient mammary gland revealed that the clones could be divided into two groups that coincided with epithelial or mesenchymal morphology and were characterized by 937 DEGs. Upstream transcription factors for these DEGs were found to include many known Fbxw7 substrates. Genes targeted by Notch1 or p63 were upregulated in epithelial-type clones, suggesting that hyperactivation of Notch1- or p63-dependent pathways might confer epithelial-type characteristics in these cells. To examine Notch1 and p63 dependency in tumor cell clones, genes encoding Notch1 and p63 were ablated. However, RNA-sequencing data showed that p63 mRNA was not detectable in these cells, and ablation of the gene for Notch1 did not affect the expression of downstream targets (Supplementary Fig. S7). These results suggest that these cell lines are no longer dependent on Notch1 or p63 for maintenance of their epithelial characteristics. Of note, genes targeted by the Nfkb1 subunit of NFκB, a substrate of Fbxw7 (57), were substantially enriched among DEGs in both epithelial- and mesenchymal-type clones. Given that NFκB promotes cell state transitions by lowering the barriers between states (58), we speculate that the loss of Fbxw7 results in the accumulation of NFκB, which promotes cell plasticity and thereby increases tumor heterogeneity.

New therapies for BLC and other TNTs are urgently needed, given that these tumors do not respond to current treatments. However, most preclinical studies of such tumors have relied on culture models or on xenograft tumor models based on cell lines. Such experimental systems do not necessarily recapitulate the process of human cancer development in a precise manner. In contrast, genetically engineered mouse models of breast cancer have important advantages for investigation of the molecular mechanisms underlying tumorigenesis and for the development of new targeted therapies. We have now shown that loss of Fbxw7 in mammary epithelial cells predisposes mice to the development of BLC. We cloned not only epithelial-type tumor cells but also mesenchymal-type tumor cells from one BLC tumor, with both cell types having the ability to serve as tumor-initiating cells. Indeed, many human breast tumors contain multiple types of tumor cells, and such tumors are often difficult to treat (8, 9). Our mouse model should thus provide a useful tool for investigation of the molecular basis of and for validation of new targeted therapies for breast cancer.

I. Onoyama reports grants from Japan Society for the Promotion of Science (JSPS) during the conduct of the study. S. Nakayama reports grants from Japan Society for the Promotion of Science (JSPS) during the conduct of the study. H. Shimizu reports grants from Japan Society for the Promotion of Science (JSPS) during the conduct of the study. K.I. Nakayama reports grants from Japan Society for the Promotion of Science (JSPS) during the conduct of the study. No other disclosures were reported.

I. Onoyama: Conceptualization, resources, data curation, formal analysis, supervision, investigation, methodology, writing-original draft, writing-review and editing. S. Nakayama: Conceptualization, data curation, software, formal analysis, investigation, visualization, writing-original draft, writing-review and editing. H. Shimizu: Conceptualization, data curation, software, formal analysis, funding acquisition, investigation, methodology, writing-original draft, writing-review and editing. K.I. Nakayama: Conceptualization, supervision, funding acquisition, writing-original draft, writing-review and editing.

We thank T. Kitamura for pMX-puro; S. Aishima, Y. Nishihara, M. Sakamoto, and R. Irie for discussion; N. Kitajima, Y. Yamada, and K. Takeda, T. Akinaga for technical assistance; members of our laboratory for comments on the manuscript; and A. Ohta and M. Kimura for help in preparation of the manuscript. Computations were performed in part on the NIG supercomputer at ROIS National Institute of Genetics. This work was supported by Japan Society for the Promotion of Science (JSPS) KAKENHI grants (JP18H05215 to K.I. Nakayama, JP20K09646 to I. Onoyama, JP19J10033 to S. Nakayama, and JP19K20403 to H. Shimizu).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Sleeman
KE
,
Kendrick
H
,
Ashworth
A
,
Isacke
CM
,
Smalley
MJ
. 
CD24 staining of mouse mammary gland cells defines luminal epithelial, myoepithelial/basal and non-epithelial cells
.
Breast Cancer Res
2006
;
8
:
R7
.
2.
Lloyd-Lewis
B
,
Harris
OB
,
Watson
CJ
,
Davis
FM
. 
Mammary stem cells: premise, properties, and perspectives
.
Trends Cell Biol
2017
;
27
:
556
67
.
3.
Siegel
RL
,
Miller
KD
,
Jemal
A
. 
Cancer statistics, 2019
.
CA Cancer J Clin
2019
;
69
:
7
34
.
4.
Perou
CM
,
Sorlie
T
,
Eisen
MB
,
van de Rijn
M
,
Jeffrey
SS
,
Rees
CA
, et al
Molecular portraits of human breast tumours
.
Nature
2000
;
406
:
747
52
.
5.
Herschkowitz
JI
,
Simin
K
,
Weigman
VJ
,
Mikaelian
I
,
Usary
J
,
Hu
Z
, et al
Identification of conserved gene expression features between murine mammary carcinoma models and human breast tumors
.
Genome Biol
2007
;
8
:
R76
.
6.
Anders
CK
,
Abramson
V
,
Tan
T
,
Dent
R
. 
The evolution of triple-negative breast cancer: from biology to novel therapeutics
.
Am Soc Clin Oncol Educ Book
2016
;
35
:
34
42
.
7.
Prat
A
,
Parker
JS
,
Karginova
O
,
Fan
C
,
Livasy
C
,
Herschkowitz
JI
, et al
Phenotypic and molecular characterization of the claudin-low intrinsic subtype of breast cancer
.
Breast Cancer Res
2010
;
12
:
R68
.
8.
Sinha
VC
,
Piwnica-Worms
H
. 
Intratumoral heterogeneity in ductal carcinoma in situ: chaos and consequence
.
J Mammary Gland Biol Neoplasia
2018
;
23
:
191
205
.
9.
Zhang
M
,
Lee
AV
,
Rosen
JM
. 
The cellular origin and evolution of breast cancer
.
Cold Spring Harb Perspect Med
2017
;
7
:
a027128
.
10.
Cancer Genome Atlas Network
. 
Comprehensive molecular portraits of human breast tumours
.
Nature
2012
;
490
:
61
70
.
11.
Liu
X
,
Holstege
H
,
van der Gulden
H
,
Treur-Mulder
M
,
Zevenhoven
J
,
Velds
A
, et al
Somatic loss of BRCA1 and p53 in mice induces mammary tumors with features of human BRCA1-mutated basal-like breast cancer
.
Proc Natl Acad Sci U S A
2007
;
104
:
12111
6
.
12.
Saal
LH
,
Gruvberger-Saal
SK
,
Persson
C
,
Lovgren
K
,
Jumppanen
M
,
Staaf
J
, et al
Recurrent gross mutations of the PTEN tumor suppressor gene in breast cancers with deficient DSB repair
.
Nat Genet
2008
;
40
:
102
7
.
13.
Melchor
L
,
Molyneux
G
,
Mackay
A
,
Magnay
FA
,
Atienza
M
,
Kendrick
H
, et al
Identification of cellular and genetic drivers of breast cancer heterogeneity in genetically engineered mouse tumour models
.
J Pathol
2014
;
233
:
124
37
.
14.
Lawrence
MS
,
Stojanov
P
,
Mermel
CH
,
Robinson
JT
,
Garraway
LA
,
Golub
TR
, et al
Discovery and saturation analysis of cancer genes across 21 tumour types
.
Nature
2014
;
505
:
495
501
.
15.
Meric-Bernstam
F
,
Frampton
GM
,
Ferrer-Lozano
J
,
Yelensky
R
,
Perez-Fidalgo
JA
,
Wang
Y
, et al
Concordance of genomic alterations between primary and recurrent breast cancer
.
Mol Cancer Ther
2014
;
13
:
1382
9
.
16.
Ibusuki
M
,
Yamamoto
Y
,
Shinriki
S
,
Ando
Y
,
Iwase
H
. 
Reduced expression of ubiquitin ligase FBXW7 mRNA is associated with poor prognosis in breast cancer patients
.
Cancer Sci
2011
;
102
:
439
45
.
17.
Wei
G
,
Wang
Y
,
Zhang
P
,
Lu
J
,
Mao
JH
. 
Evaluating the prognostic significance of FBXW7 expression level in human breast cancer by a meta-analysis of transcriptional profiles
.
J Cancer Sci Ther
2012
;
4
:
299
305
.
18.
Santarpia
L
,
Qi
Y
,
Stemke-Hale
K
,
Wang
B
,
Young
EJ
,
Booser
DJ
, et al
Mutation profiling identifies numerous rare drug targets and distinct mutation patterns in different clinical subtypes of breast cancers
.
Breast Cancer Res Treat
2012
;
134
:
333
43
.
19.
Sundaram
M
,
Greenwald
I
. 
Suppressors of a lin-12 hypomorph define genes that interact with both lin-12 and glp-1 in Caenorhabditis elegans
.
Genetics
1993
;
135
:
765
83
.
20.
Nakayama
KI
,
Nakayama
K
. 
Ubiquitin ligases: cell-cycle control and cancer
.
Nat Rev Cancer
2006
;
6
:
369
81
.
21.
Yada
M
,
Hatakeyama
S
,
Kamura
T
,
Nishiyama
M
,
Tsunematsu
R
,
Imaki
H
, et al
Phosphorylation-dependent degradation of c-Myc is mediated by the F-box protein Fbw7
.
EMBO J
2004
;
23
:
2116
25
.
22.
Yumimoto
K
,
Nakayama
KI
. 
Recent insight into the role of FBXW7 as a tumor suppressor
.
Semin Cancer Biol
2020
. DOI: 10.1016/j.semcancer.2020.02.017.
23.
Yumimoto
K
,
Yamauchi
Y
,
Nakayama
KI
. 
F-Box proteins and cancer
.
Cancers
2020
;
12
:
1249
.
24.
Iwatsuki
M
,
Mimori
K
,
Ishii
H
,
Yokobori
T
,
Takatsuno
Y
,
Sato
T
, et al
Loss of FBXW7, a cell cycle regulating gene, in colorectal cancer: clinical significance
.
Int J Cancer
2010
;
126
:
1828
37
.
25.
Tsunematsu
R
,
Nakayama
K
,
Oike
Y
,
Nishiyama
M
,
Ishida
N
,
Hatakeyama
S
, et al
Mouse Fbw7/Sel-10/Cdc4 is required for notch degradation during vascular development
.
J Biol Chem
2004
;
279
:
9417
23
.
26.
Onoyama
I
,
Tsunematsu
R
,
Matsumoto
A
,
Kimura
T
,
de Alboran
IM
,
Nakayama
K
, et al
Conditional inactivation of Fbxw7 impairs cell-cycle exit during T cell differentiation and results in lymphomatogenesis
.
J Exp Med
2007
;
204
:
2875
88
.
27.
Matsuoka
S
,
Oike
Y
,
Onoyama
I
,
Iwama
A
,
Arai
F
,
Takubo
K
, et al
Fbxw7 acts as a critical fail-safe against premature loss of hematopoietic stem cells and development of T-ALL
.
Genes Dev
2008
;
22
:
986
91
.
28.
Onoyama
I
,
Suzuki
A
,
Matsumoto
A
,
Tomita
K
,
Katagiri
H
,
Oike
Y
, et al
Fbxw7 regulates lipid metabolism and cell fate decisions in the mouse liver
.
J Clin Invest
2011
;
121
:
342
54
.
29.
Sancho
R
,
Jandke
A
,
Davis
H
,
Diefenbacher
ME
,
Tomlinson
I
,
Behrens
A
. 
F-box and WD repeat domain-containing 7 regulates intestinal cell lineage commitment and is a haploinsufficient tumor suppressor
.
Gastroenterology
2010
;
139
:
929
41
.
30.
Matsumoto
A
,
Onoyama
I
,
Sunabori
T
,
Kageyama
R
,
Okano
H
,
Nakayama
KI
. 
Fbxw7-dependent degradation of Notch is required for control of “stemness” and neuronal-glial differentiation in neural stem cells
.
J Biol Chem
2011
;
286
:
13754
64
.
31.
Selbert
S
,
Bentley
DJ
,
Melton
DW
,
Rannie
D
,
Lourenco
P
,
Watson
CJ
, et al
Efficient BLG-Cre mediated gene deletion in the mammary gland
.
Transgenic Res
1998
;
7
:
387
96
.
32.
Nishiyama
M
,
Oshikawa
K
,
Tsukada
Y
,
Nakagawa
T
,
Iemura
S
,
Natsume
T
, et al
CHD8 suppresses p53-mediated apoptosis through histone H1 recruitment during early embryogenesis
.
Nat Cell Biol
2009
;
11
:
172
82
.
33.
Kamura
T
,
Hara
T
,
Matsumoto
M
,
Ishida
N
,
Okumura
F
,
Hatakeyama
S
, et al
Cytoplasmic ubiquitin ligase KPC regulates proteolysis of p27(Kip1) at G1 phase
.
Nat Cell Biol
2004
;
6
:
1229
35
.
34.
Nakayama
S
,
Yumimoto
K
,
Kawamura
A
,
Nakayama
KI
. 
Degradation of the endoplasmic reticulum-anchored transcription factor MyRF by the ubiquitin ligase SCF(Fbxw7) in a manner dependent on the kinase GSK-3
.
J Biol Chem
2018
;
293
:
5705
14
.
35.
Mi
H
,
Muruganujan
A
,
Ebert
D
,
Huang
X
,
Thomas
PD
. 
PANTHER version 14: more genomes, a new PANTHER GO-slim and improvements in enrichment analysis tools
.
Nucleic Acids Res
2019
;
47
:
D419
D26
.
36.
Muto
Y
,
Moroishi
T
,
Ichihara
K
,
Nishiyama
M
,
Shimizu
H
,
Eguchi
H
, et al
Disruption of FBXL5-mediated cellular iron homeostasis promotes liver carcinogenesis
.
J Exp Med
2019
;
216
:
950
65
.
37.
Ma
C
,
Shao
M
,
Kingsford
C
. 
SQUID: transcriptomic structural variation detection from RNA-seq
.
Genome Biol
2018
;
19
:
52
.
38.
Cheong
WH
,
Tan
YC
,
Yap
SJ
,
Ng
KP
. 
ClicO FS: an interactive web-based service of Circos
.
Bioinformatics
2015
;
31
:
3685
7
.
39.
Zhou
Y
,
Zhou
B
,
Pache
L
,
Chang
M
,
Khodabakhshi
AH
,
Tanaseichuk
O
, et al
Metascape provides a biologist-oriented resource for the analysis of systems-level datasets
.
Nat Commun
2019
;
10
:
1523
.
40.
Curtis
C
,
Shah
SP
,
Chin
SF
,
Turashvili
G
,
Rueda
OM
,
Dunning
MJ
, et al
The genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups
.
Nature
2012
;
486
:
346
52
.
41.
Whitelaw
CB
,
Harris
S
,
McClenaghan
M
,
Simons
JP
,
Clark
AJ
. 
Position-independent expression of the ovine beta-lactoglobulin gene in transgenic mice
.
Biochem J
1992
;
286
:
31
9
.
42.
Galli
F
,
Rossi
M
,
D'Alessandra
Y
,
De Simone
M
,
Lopardo
T
,
Haupt
Y
, et al
MDM2 and Fbw7 cooperate to induce p63 protein degradation following DNA damage and cell differentiation
.
J Cell Sci
2010
;
123
:
2423
33
.
43.
Hu
C
,
Dievart
A
,
Lupien
M
,
Calvo
E
,
Tremblay
G
,
Jolicoeur
P
. 
Overexpression of activated murine Notch1 and Notch3 in transgenic mice blocks mammary gland development and induces mammary tumors
.
Am J Pathol
2006
;
168
:
973
90
.
44.
Buono
KD
,
Robinson
GW
,
Martin
C
,
Shi
S
,
Stanley
P
,
Tanigaki
K
, et al
The canonical Notch/RBP-J signaling pathway controls the balance of cell lineages in mammary epithelium during pregnancy
.
Dev Biol
2006
;
293
:
565
80
.
45.
Candi
E
,
Cipollone
R
,
Rivetti di Val Cervo
P
,
Gonfloni
S
,
Melino
G
,
Knight
R
. 
p63 in epithelial development
.
Cell Mol Life Sci
2008
;
65
:
3126
33
.
46.
Diaz
LK
,
Cryns
VL
,
Symmans
WF
,
Sneige
N
. 
Triple negative breast carcinoma and the basal phenotype: from expression profiling to clinical practice
.
Adv Anat Pathol
2007
;
14
:
419
30
.
47.
Fadare
O
,
Tavassoli
FA
. 
The phenotypic spectrum of basal-like breast cancers: a critical appraisal
.
Adv Anat Pathol
2007
;
14
:
358
73
.
48.
Lin
SC
,
Lee
KF
,
Nikitin
AY
,
Hilsenbeck
SG
,
Cardiff
RD
,
Li
A
, et al
Somatic mutation of p53 leads to estrogen receptor alpha-positive and -negative mouse mammary tumors with high frequency of metastasis
.
Cancer Res
2004
;
64
:
3525
32
.
49.
Herschkowitz
JI
,
Zhao
W
,
Zhang
M
,
Usary
J
,
Murrow
G
,
Edwards
D
, et al
Comparative oncogenomics identifies breast tumors enriched in functional tumor-initiating cells
.
Proc Natl Acad Sci U S A
2012
;
109
:
2778
83
.
50.
Muller
PA
,
Caswell
PT
,
Doyle
B
,
Iwanicki
MP
,
Tan
EH
,
Karim
S
, et al
Mutant p53 drives invasion by promoting integrin recycling
.
Cell
2009
;
139
:
1327
41
.
51.
Peinado
H
,
Olmeda
D
,
Cano
A
. 
Snail, Zeb and bHLH factors in tumour progression: an alliance against the epithelial phenotype?
Nat Rev Cancer
2007
;
7
:
415
28
.
52.
Al-Hajj
M
,
Wicha
MS
,
Benito-Hernandez
A
,
Morrison
SJ
,
Clarke
MF
. 
Prospective identification of tumorigenic breast cancer cells
.
Proc Natl Acad Sci U S A
2003
;
100
:
3983
8
.
53.
Rajagopalan
H
,
Jallepalli
PV
,
Rago
C
,
Velculescu
VE
,
Kinzler
KW
,
Vogelstein
B
, et al
Inactivation of hCDC4 can cause chromosomal instability
.
Nature
2004
;
428
:
77
81
.
54.
Wuidart
A
,
Sifrim
A
,
Fioramonti
M
,
Matsumura
S
,
Brisebarre
A
,
Brown
D
, et al
Early lineage segregation of multipotent embryonic mammary gland progenitors
.
Nat Cell Biol
2018
;
20
:
666
76
.
55.
Matsumoto
A
,
Tateishi
Y
,
Onoyama
I
,
Okita
Y
,
Nakayama
K
,
Nakayama
KI
. 
Fbxw7beta resides in the endoplasmic reticulum membrane and protects cells from oxidative stress
.
Cancer Sci
2011
;
102
:
749
55
.
56.
Shimizu
H
,
Takeishi
S
,
Nakatsumi
H
,
Nakayama
KI
. 
Prevention of cancer dormancy by Fbxw7 ablation eradicates disseminated tumor cells
.
JCI Insight
2019
;
4
:
e125138
.
57.
Fekrmandi
F
,
Wang
TT
,
White
JH
. 
The hormone-bound vitamin D receptor enhances the FBW7-dependent turnover of NF-kappaB subunits
.
Sci Rep
2015
;
5
:
13002
.
58.
Ramirez
D
,
Kohar
V
,
Lu
M
. 
Toward modeling context-specific EMT regulatory networks using temporal single cell RNA-Seq data
.
Front Mol Biosci
2020
;
7
:
54
.