Abstract Developmental cadmium exposure in vivo disrupts mammary gland differentiation, while exposure of breast cell lines to cadmium causes invasion consistent with the epithelial-mesenchymal transition (EMT). The effects of cadmium on normal human breast stem cells have not been measured. Here, we quantified the effects of cadmium exposure on reduction mammoplasty patient-derived breast stem cell proliferation and differentiation. Using the mammosphere assay and organoid formation in 3D hydrogels, we tested 2 physiologically relevant doses of cadmium, 0.25 and 2.5 µM, and tested for molecular alterations using RNA-seq. We functionally validated our RNA-seq findings with a hypoxia-inducible factor (HIF)-1α activity reporter line and pharmaceutical inhibition of HIF-1α in organoid formation assays. 2.5 µM cadmium reduced primary mammosphere formation and branching structure organoid formation rates by 33% and 87%, respectively. Despite no changes in mammosphere formation, 0.25 µM cadmium inhibited branching organoid formation in hydrogels by 73%. RNA-seq revealed cadmium downregulated genes associated with extracellular matrix formation and EMT, while upregulating genes associated with metal response including metallothioneins and zinc transporters. In the RNA-seq data, cadmium downregulated HIF-1α target genes including LOXL2, ZEB1, and VIM. Cadmium significantly inhibited HIF-1α activity in a luciferase assay, and the HIF-1α inhibitor acriflavine ablated mammosphere and organoid formation. These findings show that cadmium, at doses relevant to human exposure, inhibited human mammary stem cell proliferation and differentiation, potentially through disruption of HIF-1α activity. stem cell, cadmium, breast cancer, mammary gland, RNA-seq, mammosphere Breast cancer is the most common cancer in women, with an estimated 246 660 incident cases in 2016 in the United States alone (Siegel et al., 2016). Approximately 90% of breast cancers are thought to be due to sporadic, rather than hereditary, alterations, potentially arising due to environmental exposures (Rizzolo et al., 2011). Despite decades of research, the environmental risk factors for breast cancer are still not well understood. The toxic heavy metal cadmium is a naturally occurring known human carcinogen which has been strongly linked to lung cancer through occupational health studies (Waalkes, 2003). Its role in breast cancer remains controversial. Multiple case-control studies have reported that urinary cadmium concentrations, a biomarker of cadmium exposure, are higher in breast cancer cases relative to controls (Gallagher et al., 2010; McElroy et al., 2006; Nagata et al., 2013; Strumylaite et al., 2014). Results deriving from prospective cohort studies, are equivocal with some finding positive relationships between cadmium exposure and breast cancer (Julin et al., 2012) and others identifying null associations (Adams et al., 2012, 2014; Eriksen et al., 2014, 2016; Sawada et al., 2012). Experimental work in vivo and in vitro shows that cadmium exposure can alter normal mammary gland development and related developmental pathways. The breast cancer windows of susceptibility hypothesis states that environmental exposures during key developmental time points, particularly in utero, during puberty or pregnancy, can disproportionately increase breast cancer risk later in life (Russo, 2016). These windows represent times when the mammary gland is undergoing substantial remodeling, driven by a population of proliferating and differentiating stem cells (Visvader and Stingl, 2014). Environmental exposures during these time points could alter normal breast stem cell self-renewal, modifying the number of stem cells in the tissue (Ginestier and Wicha, 2007), or otherwise influence differentiation pathways, leading to breast cancer or other forms of breast toxicity. The in vivo data of cadmium exposure during windows of susceptibility show that life stage of exposure is important, where in utero exposures to cadmium display an estrogenic effect leading to an increase in terminal end bud structures (Alonso-Gonzalez et al., 2007; Johnson et al., 2003), while exposures in puberty or adulthood lead to stunted mammary gland development (Davis et al., 2013; Ohrvik et al., 2006). A more limited number of in vitro studies of the effects of cadmium in immortalized breast cell lines highlight that cadmium exposure can lead to the acquisition of an invasive phenotype (Benbrahim-Tallaa et al., 2009; Wei and Shaikh 2017), consistent with dysregulation of the important developmental pathway the epithelial-mesenchymal transition (EMT). Despite a growing body of experimental and epidemiological evidence linking cadmium to altered mammary gland development and potentially breast cancer, very little is known about the effects of cadmium exposure on primary nontransformed and nonimmortalized human breast stem cells. This is an important research gap as breast stem cells are likely a key target for cadmium’s effects on breast cancer or altered mammary gland development. The goal of this study was to test the hypothesis that cadmium induces changes in primary human breast stem cells consistent with alterations in stem cell self-renewal and developmental pathways such at EMT. We established a novel model of the effects of environmental exposures during human breast stem cell proliferation and differentiation, integrating an established technique, the mammosphere formation assay (Dontu et al., 2003) with 3D breast organoid culture (Sokol et al., 2016). Using functional and high-throughput molecular assays, we identify that cadmium, at doses relevant to human exposure, induces significant alterations in adult breast stem cell proliferation and differentiation. MATERIALS AND METHODS Human tissue procurement Nonpathogenic breast tissue was isolated from pre-menopausal women undergoing voluntary reduction mammoplasty at the University of Michigan. Breast tissue was mechanically and enzymatically digested as previously described (Colacino et al., 2016; Dontu et al., 2003) and filtered sequentially through 100 and 40 µm filters to yield a viable single cell suspension of human mammary cells, which was verified microscopically. These cells are a heterogenous mix of the various cell types in the breast. Our previous work has shown that these reduction mammoplasty samples contain cells which express the canonical breast stem cell markers ALDH and CD44+/CD24− (Colacino et al., 2018). In all experiments, cells from each woman were exposed to the various treatment conditions or unexposed as control. Results from experimental treatment were compared with their own control in statistical analysis. There was no pooling of cells from different individuals. This study was reviewed and approved by the University of Michigan Institutional Review Board (HUM00042409). Cadmium dosing and mammosphere formation A general outline of the experimental design is presented in Figure 1. In previous studies, chronic exposure to 2.5 µM cadmium led to transformation of MCF10A cells into cells showing a mesenchymal phenotype with increased invasive capabilities consistent with cancer development (Benbrahim-Tallaa et al., 2009). Studies of cadmium in breast tissue, from breast tumors or benign breast tissues, found mean tissue concentrations of cadmium ranging from 17.5 to 37 ng/g (Strumylaite et al., 2008, 2011). Assuming a tissue density equal to that of water, these concentrations would correspond to a dose of 0.156 and 0.33 µM, respectively. Based on these previously published results and our dose-response experiments, we chose 2 relevant doses of cadmium for further study, 0.25 and 2.5 µM. The primary mammosphere formation assay is surrogate readout of breast stem cell proliferation capacity, whereas the secondary mammosphere formation assay is readout of breast stem cell self-renewal capacity. Single primary human breast cells were plated in mammosphere formation conditions, in the presence of 0.25 or 2.5 µM cadmium chloride or vehicle control following our previously established protocol (Colacino et al., 2016). Briefly, cells were plated at a concentration of 100 000/mL in one 96-well and 2 sets of 6-well ultralow attachment plates (Corning) conditions in MammoCult (StemCell) media. Additional media (either control or cadmium containing) was added at 3 days. Primary mammospheres formed for 7 days in the 96-well plate, at which point mammospheres >40 µm were counted manually using an EVOS XL Cell Imaging System (AMG, Bothell, Waltham) which has an onscreen scale bar for quantifying mammosphere size. Primary mammospheres from 1 set of the 6-well plates were collected via centrifugation and dissociated into a single cell suspension using TrypLE Express (Life Technologies). The viability and number of these cells was assessed using the Trypan Blue Exclusion Assay (Life Technologies). These cells were then plated in a 96-well non-adherent plate at 100 000 cells/ml of MammoCult for quantification of secondary mammosphere formation. The number of secondary mammospheres formed was counted quantified manually after 7 days using the same criteria as described earlier for primary mammosphere formation. At least 3 technical replicates were quantified per condition across 8 different study participants. Figure 1. View largeDownload slide Conceptual diagram of the experimental design. Epithelial cells isolated from voluntary reduction mammoplasty tissues were cultured in the presence or absence of 2 doses of cadmium in mammosphere forming conditions. Primary mammospheres were either dissociated into single cells or replated in secondary mammosphere formation conditions, plated in collagen-based hydrogels to induce breast organoid formation, or RNA was extracted and sequenced. Figure 1. View largeDownload slide Conceptual diagram of the experimental design. Epithelial cells isolated from voluntary reduction mammoplasty tissues were cultured in the presence or absence of 2 doses of cadmium in mammosphere forming conditions. Primary mammospheres were either dissociated into single cells or replated in secondary mammosphere formation conditions, plated in collagen-based hydrogels to induce breast organoid formation, or RNA was extracted and sequenced. Cell viability assessment Control or treated primary mammospheres were dissociated into a single cell suspension using TrypLE Express (Life Technologies). Viability of these cells was assessed using the Trypan Blue exclusion assay, quantified in an unbiased manner using a Countess automated cell counter (ThermoFisher). Mammary organoid formation At the end of the 7–10 days of primary mammosphere formation in the presence or absence of cadmium, mammospheres were embedded in floating 3D collagen-based hydrogel scaffolds to induce branching morphogenesis, a process which models mammary gland development. Hydrogels were prepared as previously described (Sokol et al., 2016). Briefly, Laminin (Roche), Fibronectin (Roche), 100 and 700 KDa Sodium Hyaluronate (Lifecore Biomedical), MEGM growth media (Lonza), and primary mammospheres were added to rat tail Collagen Type I (EMD Millipore). We added 0.2 M sodium hydroxide (Sigma-Aldrich) to induce hydrogel polymerization, and plated the gels in 24-well plates for 1 h at 5%CO2 and 37°C. After polymerization, media was added and the hydrogels were detached from the plate using a metal spatula. The hydrogels were plated in either MEGM media as a control, or MEGM containing 0.25 or 2.5 µM cadmium chloride. These were cultured at 37°C at 5% CO2 for 2–3 weeks with media changes every 3–4 days. The number of branching structures that formed in each gel was quantified manually after 10–14 days. At least 3 technical replicates were quantified per condition across 6 different study participants. RNA sequencing of primary mammospheres After 7 days of growth in primary mammosphere formation conditions, we extracted RNA from the control or cadmium treated mammospheres using the AllPrep DNA/RNA micro kit (Qiagen), including an on-column DNase treatment for RNA extraction. RNA concentration and quality was determined using a Nanodrop (Thermo) and Bioanalyzer (Agilent). We depleted ribosomal RNAs with Ribominus (Life) and prepared sequencing libraries utilizing the SMARTer Stranded RNA-Seq kit (Clontech) following the manufacturer’s recommended protocol. Libraries were multiplexed (6 per lane) and sequenced using paired end 50 cycle reads on a HiSeq 4000 (Illumina). Library preparation and sequencing took place at the University of Michigan DNA Sequencing Core Facility following their standard protocols. RNA-seq data analysis The RNA-seq libraries were aligned to the GRCh38.p10 human genome (GRCh38_GencodeV26 https://www.gencodegenes.org/releases/current.html; last accessed May 11, 2018). Quality control of raw fastq files was performed using FastQC v 0.11.5 (Andrews, 2012) and MultiQC v0.9 (Ewels et al., 2016) to identify features of the data that may indicate quality problems (eg, low-quality scores, over-represented sequences, and inappropriate GC content). Mapping of raw sequences were performed using STAR-2.5.3.a (Dobin et al., 2013) with default parameters. The RNA-seq library sizes ranged from 25.4 to 66.7 million reads (average 46.3 million) an alignment rate is 66.5%–75.1% (average 71.2%). MultiQC, RSeQC v2.6.4 (Wang et al., 2012) and QoRTS v1.2.26 (Hartley and Mullikin, 2015) were used for a second round of quality control (postalignment), to ensure that only high quality data would be input to expression quantitation and differential expression analysis. Gene expression levels were quantified using Subread v 1.5.2 package FeatureCounts, performing strand-specific pair-end reads counting (Liao et al., 2014). Normalization and transforming read counts for diagnostic plots, including PCA and hierarchical clustering, were performed using DESeq2 v1.14.1 Bioconductor package (Love et al., 2014). The differential expression testing was conducted with the Bioconductor package edgeR v 3.16.5 using edgeR-robust and the quasi-likelihood functionality of edgeR (Chen et al., 2016; Robinson et al., 2010; Zhou et al., 2014). We conducted filtering of the low expressed genes keeping only genes that are expressed at least 1 CPM (counts per million reads) in at least 3 samples. The comparisons were performed for the following groups: (1) cells treated with 2.5 µM cadmium chloride versus control and (2) cells treated with 0.25 µM cadmium chloride versus control. The p-value adjustments for multiple testing were made using the Benjamini-Hochberg false discovery rate (FDR) approach. Genes and transcripts satisfying criteria (FDR < 0.10 and absolute value of fold change |FC| >1.5) were considered differentially expressed genes (DEGs). For the set of genes from the comparison 2.5 µM cadmium chloride versus control (141 genes with FDR < 0.1) the heatmap visualization with hierarchical clustering with complete linkage and dendrogams by genes and samples were performed to identify the main expression profiles observed across the 2 treatment conditions and control. Normalized by DESeq2 read counts were log-transformed with pseudo-count (=1) and then data were normalized again by subtracting the overall average expression of each gene from each expression value. Pathway analysis Enriched gene ontology (GO) terms, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways and transcription factors for each of the analyzed comparisons were tested using RNA-Enrich (Lee et al., 2016) (http://lrpath-db.med.umich.edu/; last accessed May 11, 2018). RNA-Enrich tests for gene sets that have higher significance values (eg, for differential expression) than expected at random. By not requiring a cutoff for significance, RNA-Enrich is able to detect both pathways with a few very significant genes and pathways with many only moderate DEGs. A directional RNA-Enrich test, which tests for significantly up- versus down- regulated gene sets, was run for each comparison using default settings. Only concepts with <500 genes were considered for this analysis. Custom code was implemented to reduce redundancy (remove less significant, closely related GO terms) for presenting the top enriched terms by cadmium treatment. Hypoxia-inducible factor-1α inhibition experiments Embryonic kidney cell line HEK293T were seeded into a 24-well plate at a cell density of 5 × 104 cells per well. hypoxia-inducible factor (HIF)1α response element driven luciferase was co-transfected with β-galactosidase into cells with polyethylenimine (Polysciences Inc., Warrington, Pennsylvania) for 24 h. After 24 h of transfection, cells were treated with 250 nM cadmium chloride in the presence or absence of FG4592 (100 µM, HIF prolyl hydroxylase inhibitor) for another 24 h. Cells were lysed in reporter lysis buffer (Promega, Madison, Wisconsin) and luciferase assay was performed and normalized to β-galactosidase activity. Primary human breast epithelial cells grown in mammosphere forming conditions were also treated with the pharmaceutical inhibitor of HIF-1α activity acriflavine at concentrations of 1 and 5 µM. Primary and secondary mammosphere formation rates and organoid formation were measured as described earlier, quantifying at least 3 technical replicates across cells from 3 study participants. Statistical analysis Differences of primary mammospheres, secondary mammospheres, and hydrogel structures were analyzed across treatments within individuals first using ANOVA and next using a paired t test, comparing control to the treatments and adjusting for multiple comparisons using the Bonferonni correction. We divided the p-value of .05 by the number of comparisons in each experiment to establish a cutoff for statistical significance. RESULTS Effects of Cadmium Exposure on Primary and Secondary Mammosphere Formation To quantify the effects of cadmium on normal breast stem cell proliferation and selfrenewal capacity, human mammary cells were plated in mammosphere forming conditions in the presence or absence of the 2 cadmium doses. Neither dose of cadmium was found to alter cellular viability, as assessed by trypan blue staining (Supplementary Figure 1). Across 8 independent biological replicates, we did not observe a significant alteration in primary mammosphere formation at the lower cadmium dose, but 2.5 µM cadmium chloride exposures decreased primary sphere formation by approximately 33% (Figure 2A). Similar results were observed for secondary sphere formation, where 0.25 µM cadmium chloride had no significant effect and the 2.5 µM dose inhibited secondary mammosphere formation by approximately 40%, although this result was not statistically significant after adjusting for multiple comparisons (p = .047) (Figure 2C). Representative images of the mammospheres formed in the various experimental conditions are displayed in Figures 2B and 2D. Figure 2. View largeDownload slide The effects of cadmium on mammosphere formation. A, The effects of 0.25 or 2.5 µM cadmium chloride on primary mammosphere formation (n = 8 unique individuals). Data are displayed relative to control. Representative images are shown in (B). C, The effects of 0.25 or 2.5 µM cadmium chloride, relative to control, on secondary mammosphere formation (n = 8 unique individuals), with representative images are shown in (D). *p < .007. Figure 2. View largeDownload slide The effects of cadmium on mammosphere formation. A, The effects of 0.25 or 2.5 µM cadmium chloride on primary mammosphere formation (n = 8 unique individuals). Data are displayed relative to control. Representative images are shown in (B). C, The effects of 0.25 or 2.5 µM cadmium chloride, relative to control, on secondary mammosphere formation (n = 8 unique individuals), with representative images are shown in (D). *p < .007. Effects of Cadmium Exposure on 3D Branching Organoid Formation After 7–10 days, primary mammospheres are plated in 3D collagen-based hydrogels which mimic the composition of the extracellular matrix (ECM) of the human breast. Over the course of 2–3 weeks, the cells grow into complex branching organoids (Figure 3A), which closely recapitulate mammary gland structures in situ in the human mammary gland (Sokol et al., 2016). Unlike in the mammosphere formation experiments, 0.25 µM cadmium chloride treatment reduced the number of branching organoid structures formed by approximately 73%, while the 2.5 µM exposure led to a reduction of approximately 87% (Figure 3B). Imaging of hydrogels showed changes to the extent of the branching ductal growth after cadmium exposure (Figure 3C). Cells exposed to 0.25 µM cadmium chloride and grown in hydrogels demonstrated ductal growth, but was significantly less pronounced and it did not extend through the hydrogel as extensively as in the control conditions. The 2.5 µM cadmium-exposed hydrogels had very little growth overall. These data demonstrate that cadmium exposure significantly decreases branching structure formation. Figure 3. View largeDownload slide The effects of cadmium on organoid formation. A, Mammospheres plated in floating collagen-based hydrogels will form complex branching structures over the course of 2 weeks. B, The effects of 0.25 or 2.5 µM cadmium chloride on branching structure formation in 3D hydrogels, relative to control (n = 6 unique individuals). C, Representative images of structures formed in each experimental condition. *p < .006. Figure 3. View largeDownload slide The effects of cadmium on organoid formation. A, Mammospheres plated in floating collagen-based hydrogels will form complex branching structures over the course of 2 weeks. B, The effects of 0.25 or 2.5 µM cadmium chloride on branching structure formation in 3D hydrogels, relative to control (n = 6 unique individuals). C, Representative images of structures formed in each experimental condition. *p < .006. RNA-Sequencing of Exposure Treated Primary Mammospheres The significant decreases in branching morphogenesis caused by cadmium exposure suggested that cadmium treatment was downregulating key pathways involved in stem cell differentiation or invasion. To comprehensively characterize the transcriptional alterations caused by cadmium, RNA was sequenced from 0.25 to 2.5 µM cadmium chloride treated cells, from 4 independent biological replicates, grown in primary mammosphere conditions for 7 days. Primary mammosphere formation yields a population of cells enriched for stem and early progenitor cells (Dontu et al., 2003). Visualization through multidimensional scaling showed that the samples cluster clearly based on individual rather than treatment (Figure 4A). This suggests that the majority of the variance is explained by interindividual differences in breast gene expression. Consistent with the effects on primary and secondary mammosphere formation, 2.5 µM cadmium chloride treatment lead to more significant alterations in gene expression compared with 0.25 µM treatment (101 and 5 genes altered at FDR < 0.05, respectively, Figures 4B and 4C; alterations for all genes presented in Supplementary Tables 1 and 2). All of the significant changes observed at 0.25 µM treatment was also found at 2.5 µM treatment (Figure 4D). Hierarchical clustering of the samples based on the expression of genes identified as differentially expressed (FDR < 0.10) in the 2.5 µM cadmium treatment showed that the 2.5 µM cadmium treatment samples clustered distinctly from the control and 0.25 µM samples (Figure 4E). Figure 4. View largeDownload slide Transcriptomic profiling of the effects of cadmium exposure in primary mammospheres. A, Multidimensional scaling plot of RNA expression in control, 0.25 or 2.5 µM cadmium chloride treated mammospheres (n = 4 unique individuals). B, Volcano plot representing differential gene expression in the 0.25 µM cadmium treatment versus control. Red color indicates a significantly DEG. C, Volcano plot representing differential gene expression in the 2.5 µM cadmium treatment versus control. D, Venn diagram representing the overlap in DEGs between the 0.25 µM cadmium treatment versus control and 2.5 µM cadmium treatment versus control comparisons. E, Unbiased clustering analysis of the 12 samples based on expression of the genes identified as differentially expressed between 2.5 µM cadmium treatment versus control. Colorbar: Green = 2.5 µM cadmium, Orange = 0.25 µM cadmium, Gray = Control. F, Gene sets enriched for DEGs in the 2.5 µM cadmium treatment versus control comparison. Figure 4. View largeDownload slide Transcriptomic profiling of the effects of cadmium exposure in primary mammospheres. A, Multidimensional scaling plot of RNA expression in control, 0.25 or 2.5 µM cadmium chloride treated mammospheres (n = 4 unique individuals). B, Volcano plot representing differential gene expression in the 0.25 µM cadmium treatment versus control. Red color indicates a significantly DEG. C, Volcano plot representing differential gene expression in the 2.5 µM cadmium treatment versus control. D, Venn diagram representing the overlap in DEGs between the 0.25 µM cadmium treatment versus control and 2.5 µM cadmium treatment versus control comparisons. E, Unbiased clustering analysis of the 12 samples based on expression of the genes identified as differentially expressed between 2.5 µM cadmium treatment versus control. Colorbar: Green = 2.5 µM cadmium, Orange = 0.25 µM cadmium, Gray = Control. F, Gene sets enriched for DEGs in the 2.5 µM cadmium treatment versus control comparison. Cadmium Exposure Induces Changes in Metal Response and ECM-Related Pathways At the 0.25 µM dose, the 5 genes that were significantly altered were all upregulated, and represent known metal response genes: MT1E, MT1X, and MT1M are 3 subtypes of the metal binding protein metallothionein 1 while SLC30A1 and SLC30A2 are zinc transporters. These results show that even at low doses of cadmium, stem cell enriched populations of breast epithelial cells are sensitive to the presence of divalent cations and can upregulate biological processes to detoxify or eliminate cadmium. Similar changes were observed at the 2.5 µM dose of cadmium chloride, where the top upregulated genes were associated with response to metal ions (Figure 4F). An increase in expression of known oxidative response genes, including HMOX1, TXNRD1, and SPP1, was also observed. Intriguingly, the pathways most enriched for downregulated genes at the 2.5 µM dose were involved in the formation of the ECM or interaction with the ECM (e.g. focal adhesion) (Figure 4F). Specifically, 2.5 µM cadmium chloride lead to a decrease in expression of many ECM genes: COL1A2, COL6A1, COL6A2, COL6A3, FBN1, FN1, and NOV. Further, genes associated with EMT were also downregulated with cadmium treatment, including ZEB1, VIM, and TGFBI. No changes in known estrogen receptor alpha target genes were noted, including PGR, C3, GREB1, NRIP1, or ABCA3 (Johnson et al., 2003; Lin et al., 2004) (Supplementary Tables 1 and 2), suggesting that at these doses and this time point, cadmium is not activating estrogen signaling as a metalloestrogen. Overall, these results show that 2.5 µM cadmium chloride exposure increase the expression of metal ion response genes and downregulate the expression of genes involved in ECM production and interaction as well as EMT. Cadmium, HIF-1α, and Mammary Stem Cell Growth A closer examination of the genes and pathways dysregulated by 2.5 µM cadmium chloride exposure revealed a consistent downregulation of genes known as targets of the transcription factor HIF-1α. Specifically, we observed a significant downregulation of known HIF-1α targets LOXL2, VIM, ZEB1, IGFBP2, and PDGFRA (Figure 5A). HIF-1α is required for mammary stem cell expansion and branching morphogenesis (Seagroves et al., 2003). We first tested whether cadmium can inhibit HIF-1α transcription factor activity using a luciferase reporter assay. The 0.25 µM cadmium chloride dose reduced the amount of HIF-1α activity (Figure 5B). Additionally, 0.25 µM cadmium chloride significantly attenuated the effects of the pharmaceutical hypoxia mimic and HIF-1α stabilizer FG-4592 (Beuck et al., 2012; Figure 5B). To test the role that HIF-1α plays in mammary stem cell growth and differentiation, we treated primary mammary epithelial cells with 1 or 5 µM of the pharmaceutical HIF-1α inhibitor acriflavine (Lee et al., 2009). The 1 µM dose of acriflavine did not alter cellular viability after 1 week, while 5 µM acriflavine treatment decreased cellular viability by approximately 15%, although this effect was non-significant after adjustment for multiple comparisons (p = .048) (Supplementary Figure 1). Acriflavine treatment inhibited primary and secondary mammosphere formation almost completely and eliminated the formation of any branching organoid structures, a phenotype similar to that observed with cadmium treatment (Figs. 5C–E). Data are presented as average number of mammospheres for branching organoids formed per condition, rather that proportion of mammospheres formed relative to control, due to the almost complete ablation of mammosphere and organoid formation with acriflavine treatment. These results show that HIF-1α activity is important for human mammary stem cell proliferation and branching morphogenesis, and that cadmium, at the doses tested, significantly inhibits the transcriptional activity of HIF-1α. Figure 5. View largeDownload slide The effects of cadmium on HIF-1α activity and HIF-1α activity on breast stem cell growth. A, Relative expression of known HIF-1α target genes between 2.5 µM cadmium treatment versus control. B, Effects of FG 4592, a known HIF-1α activator, 0.25 µM cadmium, or a combination of the 2 on HIF-1α activity as assessed by a luciferase reporter assay. *p < .0005 difference from control, **p < .05 difference from control, ^p < .0005 difference between FG 4592 treatment alone. C, Primary mammosphere formation rates in 1 µM acriflavine, or 5 µM acriflavine treatment, relative to control (n = 3 unique individuals). D, Secondary mammosphere formation rates in acriflavine treatments relative to control (n = 3 unique individuals). E, Branching organoid formation rates in acriflavine treatments relative to control (n = 3 unique individuals). Figure 5. View largeDownload slide The effects of cadmium on HIF-1α activity and HIF-1α activity on breast stem cell growth. A, Relative expression of known HIF-1α target genes between 2.5 µM cadmium treatment versus control. B, Effects of FG 4592, a known HIF-1α activator, 0.25 µM cadmium, or a combination of the 2 on HIF-1α activity as assessed by a luciferase reporter assay. *p < .0005 difference from control, **p < .05 difference from control, ^p < .0005 difference between FG 4592 treatment alone. C, Primary mammosphere formation rates in 1 µM acriflavine, or 5 µM acriflavine treatment, relative to control (n = 3 unique individuals). D, Secondary mammosphere formation rates in acriflavine treatments relative to control (n = 3 unique individuals). E, Branching organoid formation rates in acriflavine treatments relative to control (n = 3 unique individuals). DISCUSSION Cadmium is a known human carcinogen, however, its role in human breast carcinogenesis and altered mammary gland development remains controversial. Here, we used 3D tissue culture methods of primary adult human breast cells to model the effects of cadmium exposure during times of stem cell expansion and differentiation. We identified that cadmium exposure, at concentrations in the range of those previously detected in human breast tissue, significantly inhibits primary and secondary mammosphere formation, which are functional readouts of stem cell proliferation and selfrenewal capacity (Dontu et al., 2003). Further, cadmium exposure significantly inhibits breast epithelial cell ductal elongation and branching morphogenesis in collagen-based hydrogel cultures which model the human breast ECM (Sokol et al., 2016). To identify pathways and genes dysregulated by cadmium that explain this inhibition of growth, we sequenced RNA from cadmium-exposed primary mammospheres, which are enriched for stem and progenitor cells. Cadmium significantly upregulates genes associated with metal response, while downregulating ECM-production and genes involved in focal adhesion, including known targets of the transcription factor HIF-1α. We validated that cadmium exposure at the doses tested can inhibit HIF-1α activity and showed that pharmacologic HIF-1α inhibition ablates mammary growth in our model. Overall, we present a novel approach to assay the effects of environmental toxicants on breast stem cell proliferation and differentiation using patient-derived breast cells. We show that cadmium can inhibit mammary gland branching morphogenesis at doses relevant to human exposure and identify HIF-1α dysregulation as an important mechanism of toxicity during mammary gland growth and differentiation. Studies assaying the effects of cadmium exposure on altered breast development or cancer in vivo have identified that the life stage timing of exposure is essential in defining effects. These findings are in line with the breast cancer windows of susceptibility hypothesis, where exposures during key developmental stages, specifically in utero, during puberty, and during pregnancy, disproportionately influence mammary gland development and breast cancer risk (Russo, 2016). The physiological changes during these developmental windows are driven by stem cells undergoing rapid proliferation and differentiation, which potentially increases their vulnerability to environmental influence. In vivo studies of in utero exposure show that cadmium alters mammary gland development in a manner consistent with estrogen exposure, leading to an increase in the number of terminal end buds (Davis et al., 2013; Johnson et al., 2003), a stem cell enriched population which drives adult mammary gland differentiation (Paine and Lewis, 2017). Interestingly, however, the increase in terminal end buds was found to be dose dependent, occurring at the lower (0.5 μg/kg), rather than higher (5 μg/kg) dose tested (Johnson et al., 2003), and were not associated with an increase in DMBA-induced carcinogenesis, unlike estrogen treatment (Davis et al., 2013). Mice exposed to cadmium in the prepubertal window, instead, experienced a decrease in the number of mammary gland terminal end buds and decreased overall mammary gland ductal development (Alonso-Gonzalez et al., 2007). Mice treated with cadmium during pregnancy displayed a remodeled mammary gland characterized by an increase in fat content and less active alveolar epithelial cells (Ohrvik et al., 2006). Similarly, exposure of lactating Holstein cows to cadmium lead to a decrease in milk production (Miller et al., 1967). Here, we found that exposure of breast cells derived from adults to low dose cadmium lead to impaired branching morphogenesis. Unlike the studies of cadmium exposure in utero, we did not observe an estrogenic effect, as reflected by increased expression of estrogen responsive genes. In aggregate, these results suggest that the timing of cadmium exposure during mammary gland development is likely important, where in utero cadmium exposure could lead to an increase in terminal end bud formation, while exposure during adulthood could inhibit ductal morphogenesis and proper mammary gland function. Through an unbiased approach, we identified biological pathways significantly altered with cadmium treatment in normal breast cells. The most significantly upregulated pathways following cadmium exposure included metal response, where we saw upregulation of metallothionine genes MT1E, MT1X, and MT1M at both cadmium doses. Metallothionines are well described response proteins to cadmium toxicity in the liver and lung, where they sequester cadmium and increase resistance to cadmium toxicity (Klaassen et al., 2009). Interestingly, increased metallothionein expression is associated with poor prognosis in breast cancer (Goulding et al., 1995). Our results show that these proteins likely play an important role in cadmium detoxification in noncancerous breast cells. We also observed significant downregulation of genes involved in the production of, or interaction with, the ECM. Our findings are consistent with those from a study of non-tumorigenic lung cells treated with nontoxic doses of cadmium, where downregulation of genes coding for collagens I, IV, and V was observed (Baroni et al., 2015). As ECM development and interaction are key regulators of both mammary gland development and breast cancer (Hu et al., 2017), future investigation into the mechanisms by which cadmium influences ECM production and interaction will likely provide important insight into breast biology. Mechanistic studies of the effects of cadmium on immortalized breast cell lines, both nontumorigenic and cancer, have identified that cadmium can also induce cancer-related alterations in key developmental processes, including the EMT. During EMT, epithelial cells lose their tight junctions to their neighboring cells and gain invasive and migratory properties, which are important for the development of metastatic cancer (Wang and Zhou 2011). A 40-week treatment of the immortalized, but nontumorigenic, breast cell line MCF10A with 0.25 µM cadmium lead to the acquisition of an invasive, aggressive phenotype consistent with basal breast cancers through a mechanism independent of ER-α (Benbrahim-Tallaa et al., 2009). Recently, treatment of MDAMB231 cells, an already aggressive and invasive triple negative breast cancer cell line, with 1 or 3 μM cadmium lead to a more invasive and metastatic phenotype (Wei and Shaikh 2017). Although the results of these studies lead us to initially hypothesize that we would observe a similar EMT-like phenotype in cadmium treated primary breast stem cells, we instead observed the opposite. Both our functional and molecular data showed that cadmium treatment of primary cells lead to decreased invasion and growth in hydrogel cultures, further supported by a downregulation of key genes involved in EMT including VIM and ZEB1. One potential explanation for the difference in our findings and those previous could lie in the different response of various cell types in the breast to environmental stress. We, and others, have recently shown that breast stem cells exist in at least 2 phenotypes, including an epithelial phenotype, which are characterized by aldehyde dehydrogenase activity (ALDH+), and a mesenchymal phenotype, which are characterized by expression of surface markers CD44+/CD24− (Colacino et al., 2018). Some of the major biological pathways which distinguish these 2 types of stem cells include those which we identify here as modulated by cadmium, including focal adhesion and ECM-receptor interactions. Although ALDH+ cells are plentiful in the normal breast, and drivers of normal mammosphere formation (Colacino et al., 2018), ALDH+ cells are rare in MCF10A and MDAMB231 cells (Charafe-Jauffret et al., 2009). Thus, the different cell type distributions in the normal breast and in these cell lines may have a significant impact in the overall response to cadmium exposure. Future research should focus on understanding these differences, particularly in light of the inconsistent association between cadmium and breast cancer risk in epidemiological studies. Through our unbiased approach, we found that 2.5 µM cadmium exposure led to the downregulation of multiple known targets of the HIF-1α. HIF-1α is an oxygen sensing transcription factor that plays an essential role in mammary gland development. For example, while mammary glands from nulliparous targeted HIF-1α knockout mice do not display phenotypic differences from control mice, during pregnancy HIF-1α knockout mice have stunted mammary gland differentiation, ultimately leading to lactation failure (Seagroves et al., 2003). This finding suggests that HIF-1α activity is essential for the mammary gland expansion during pregnancy and lactation, but not mammary gland development in utero. This aligns with the findings of the effects of cadmium exposure during puberty and pregnancy significantly disrupting mammary gland formation and leading to different phenotypic effects compared with in utero exposure. Using a pharmacologic inhibitor of HIF-1α, acriflavine (Lee et al., 2009), we found that inhibition of HIF-1α ablates mammosphere formation and branching morphogenesis, confirming the importance of this transcription factor in adult mammary gland differentiation. A number of the HIF-1α target genes identified as downregulated here, including ZEB1, LOXL2, SERPINE1, and VIM have been implicated as important players in key pathways in both development and cancer, including angiogenesis, EMT, and invasion (Dengler et al., 2014). An essential role for HIF-1α has also been reported in branching morphogenesis across multiple other organ systems, including the lung (Shimoda and Semenza 2011), the vascular network (Krock et al., 2011), and the kidney (Tsuji et al., 2014). Others have previously reported that cadmium exposure can inhibit HIF-1α transcriptional activity in other organ systems and cell types. Cadmium exposure caused known HIF-1α target genes to not be induced during hypoxia, disrupted HIF-1α DNA binding in Hep3B cells, and decreased luciferase expression in a HIF-1α activity reporter system similar to the one used here (Chun et al., 2000). Others have reported that cadmium exposure in rat lung fibroblasts lead to a decrease in HIF-1α binding to promoter region of the lysyl oxidase gene, a known HIF-1α target (Gao et al., 2013). Taken in aggregate, these results suggest that exposure to cadmium at doses relevant to human exposures can inhibit normal HIF-1α function during human mammary stem cell proliferation and differentiation, leading to an inhibition of branching morphogenesis. This study has a number of strengths relative to the existing literature. First, this is, to the best of our knowledge, the first study to test the effects of cadmium on patient-derived primary breast cells, rather than immortalized cell lines or in animal models. We developed and used a novel model of human mammary stem cell proliferation and differentiation, including a physiologically relevant 3D hydrogel based organoid system (Sokol et al., 2016), to characterize the effects of cadmium exposure. We incorporated an unbiased whole transcriptome analysis to comprehensively quantify the effects of cadmium in a human stem cell enriched population, and functionally validated the findings of HIF-1α inhibition. This study also has a number of weaknesses. We used primary breast cells from adult reduction mammoplasty patients, which may not represent a truly “normal” breast cell population (Degnim et al., 2012), or may not recapitulate the biological effects of cadmium exposure in utero. Although our cadmium treatments ranged from 7 to 21 days total, we did not attempt to recapitulate the extended cadmium exposure (up to 40 weeks) of other studies, which may better model the effects of chronic exposure to cadmium (Benbrahim-Tallaa et al., 2009). Due to the relatively short time course used, we are also unable to conclude whether differentiation is fully interrupted or if the tissue has fallen behind in differentiation. Additionally, for this study, we were blinded to the study participant characteristics besides age, including parity or family history of breast cancer. Thus, we were unable to make any conclusions about epidemiological factors that may predict increased or decreased susceptibility to the effects of cadmium. Future studies incorporating the life stage or other breast cancer risk factors of the donors, while in parallel conducting a longer-term cadmium exposure, may be able to better model human population-relevant effects of cadmium on mammary gland toxicity or cancer risk. Here, we show that cadmium exposure, at doses relevant to human health, can significantly inhibit adult mammary stem cell proliferation and differentiation. These findings are in line with animal studies of cadmium exposure during puberty and pregnancy, whose results stand distinct from studies of in utero exposure. The effects we observed are likely mediated, at least in part, by cadmium’s disruption of HIF-1α activity, identifying HIF-1α as an important target for human mammary gland toxicity. Although our results do not support the hypothesis that cadmium acts as a breast cancer initiator through the induction of adult stem cell proliferation, our results suggest that cadmium exposure in adult women may significantly inhibit stem cell proliferation and differentiation. SUPPLEMENTARY DATA Supplementary data are available at Toxicological Sciences online. FUNDING This work was supported by the Ravitz Family Foundation, grants (R01 ES028802 and P30 ES017885) from the National Institute of Environmental Health Sciences (to J.A.C) and (CA130810) from National Cancer Institute (to Y.M.S), National Institutes of Health. REFERENCES Adams S. V., Newcomb P. A., White E. ( 2012). Dietary cadmium and risk of invasive postmenopausal breast cancer in the vital cohort. Cancer Causes Control 23, 845– 854. Google Scholar CrossRef Search ADS PubMed Adams S. V., Quraishi S. M., Shafer M. M., Passarelli M. N., Freney E. P., Chlebowski R. T., Luo J., Meliker J. R., Mu L., Neuhouser M. L., et al. , ( 2014). Dietary cadmium exposure and risk of breast, endometrial, and ovarian cancer in the women’s health initiative. Environ Health Perspect 122, 594– 600. Google Scholar PubMed Alonso-González C., González A., Mazarrasa O., Güezmes A., Sánchez-Mateos S., Martínez-Campa C., Cos S., Sánchez-Barceló E. J., Mediavilla M. D. ( 2007). Melatonin prevents the estrogenic effects of sub-chronic administration of cadmium on mice mammary glands and uterus. J. Pineal Res. 42, 403– 410. Google Scholar CrossRef Search ADS PubMed Andrews S. 2012. Fastqc: A quality control tool for high throughput sequence data. Available at: http://www.bioinformatics.babraham.ac.uk/projects/fastqc/, last accessed May 11, 2018. Baroni T., Lilli C., Bellucci C., Luca G., Mancuso F., Fallarino F., Falabella G., Arato I., Calvitti M., Marinucci L., et al. ( 2015). In vitro cadmium effects on ecm gene expression in human bronchial epithelial cells. Cytokine 72, 9– 16. Google Scholar CrossRef Search ADS PubMed Benbrahim-Tallaa L., Tokar E. J., Diwan B. A., Dill A. L., Coppin J. F., Waalkes M. P. ( 2009). Cadmium malignantly transforms normal human breast epithelial cells into a basal-like phenotype. Environ. Health Perspect. 117, 1847– 1852. Google Scholar CrossRef Search ADS PubMed Beuck S., Schanzer W., Thevis M. ( 2012). Hypoxia-inducible factor stabilizers and other small-molecule erythropoiesis-stimulating agents in current and preventive doping analysis. Drug Test. Anal. 4, 830– 845. Google Scholar CrossRef Search ADS PubMed Charafe-Jauffret E., Ginestier C., Iovino F., Wicinski J., Cervera N., Finetti P., Hur M.-H., Diebel M. E., Monville F., Dutcher J., et al. ( 2009). Breast cancer cell lines contain functional cancer stem cells with metastatic capacity and a distinct molecular signature. Cancer Res. 69, 1302– 1313. Google Scholar CrossRef Search ADS PubMed Chen Y., Lun A. T. L., Smyth G. K. ( 2016). From reads to genes to pathways: Differential expression analysis of rna-seq experiments using rsubread and the edger quasi-likelihood pipeline. F1000Research 5, 1438. Google Scholar PubMed Chun Y. S., Choi E., Kim G. T., Choi H., Kim C. H., Lee M. J., Kim M. S., Park J. W. ( 2000). Cadmium blocks hypoxia-inducible factor (hif)-1-mediated response to hypoxia by stimulating the proteasome-dependent degradation of hif-1alpha. Eur. J. Biochem. 267, 4198– 4204. Google Scholar CrossRef Search ADS PubMed Colacino J., Azizi E., Brooks M., Fouladdel S., McDermott S. P., Lee M., et al. , ( 2018). Heterogeneity of normal human breast stem and progenitor cells as revealed by transcriptional profiling. Stem Cell Reports. 10, 1596– 1609. Google Scholar CrossRef Search ADS PubMed Colacino J. A., McDermott S. P., Sartor M. A., Wicha M. S., Rozek L. S. ( 2016). Transcriptomic profiling of curcumin-treated human breast stem cells identifies a role for stearoyl-coa desaturase in breast cancer prevention. Breast Cancer Res. Treat . 158, 29– 41. Google Scholar CrossRef Search ADS PubMed Davis J., Khan G., Martin M. B., Hilakivi-Clarke L. ( 2013). Effects of maternal dietary exposure to cadmium during pregnancy on mammary cancer risk among female offspring. J. Carcinogenesis 12, 11. Google Scholar CrossRef Search ADS Degnim A. C., Visscher D. W., Hoskin T. L., Frost M. H., Vierkant R. A., Vachon C. M., Shane Pankratz V., Radisky D. C., Hartmann L. C. ( 2012). Histologic findings in normal breast tissues: Comparison to reduction mammaplasty and benign breast disease tissues. Breast Cancer Res. Treat. 133, 169– 177. Google Scholar CrossRef Search ADS PubMed Dengler V. L., Galbraith M., Espinosa J. M. ( 2014). Transcriptional regulation by hypoxia inducible factors. Crit. Rev. Biochem. Mol. Biol. 49, 1– 15. Google Scholar CrossRef Search ADS PubMed Dobin A., Davis C. A., Schlesinger F., Drenkow J., Zaleski C., Jha S., Batut P., Chaisson M., Gingeras T. R. ( 2013). Star: Ultrafast universal rna-seq aligner. Bioinformatics 29, 15– 21. Google Scholar CrossRef Search ADS PubMed Dontu G., Abdallah W. M., Foley J. M., Jackson K. W., Clarke M. F., Kawamura M. J., Wicha M. S. ( 2003). In vitro propagation and transcriptional profiling of human mammary stem/progenitor cells. Genes Dev. 17, 1253– 1270. Google Scholar CrossRef Search ADS PubMed Eriksen K. T., Halkjær J., Sørensen M., Meliker J. R., McElroy J. A., Tjønneland A., Raaschou-Nielsen O. ( 2014). Dietary cadmium intake and risk of breast, endometrial and ovarian cancer in danish postmenopausal women: A prospective cohort study. Plos One 9, e100815. Google Scholar CrossRef Search ADS PubMed Eriksen K. T., McElroy J. A., Harrington J. M., Levine K. E., Pedersen C., Sorensen M., et al. ( 2016). Urinary cadmium and breast cancer: A prospective danish cohort study. J. Natl. Cancer Inst. 109, djw204. Ewels P., Magnusson M., Lundin S., Kaller M. ( 2016). Multiqc: Summarize analysis results for multiple tools and samples in a single report. Bioinformatics 32, 3047– 3048. Google Scholar CrossRef Search ADS PubMed Gallagher C. M., Chen J. J., Kovach J. S. ( 2010). Environmental cadmium and breast cancer risk. Aging (Albany NY) 2, 804– 814. Google Scholar CrossRef Search ADS PubMed Gao S., Zhou J., Zhao Y., Toselli P., Li W. ( 2013). Hypoxia-response element (hre)–directed transcriptional regulation of the rat lysyl oxidase gene in response to cobalt and cadmium. Toxicol. Sci. 132, 379– 389. Google Scholar CrossRef Search ADS PubMed Ginestier C., Wicha M. S. ( 2007). Mammary stem cell number as a determinate of breast cancer risk. Breast Cancer Res. 9, 109. Google Scholar CrossRef Search ADS PubMed Goulding H., Jasani B., Pereira H., Reid A., Galea M., Bell J. A., Elston C. W., Robertson J. F., Blamey R. W., Nicholson R. A., et al. ( 1995). Metallothionein expression in human breast cancer. Br. J. Cancer 72, 968– 972. Google Scholar CrossRef Search ADS PubMed Hartley S. W., Mullikin J. C. ( 2015). Qorts: A comprehensive toolset for quality control and data processing of rna-seq experiments. BMC Bioinformatics 16, 224. Google Scholar CrossRef Search ADS PubMed Hu G., Li L., Xu W. ( 2017). Extracellular matrix in mammary gland development and breast cancer progression. Front. Lab. Med. 1, 36– 39. Google Scholar CrossRef Search ADS Johnson M. D., Kenney N., Stoica A., Hilakivi-Clarke L., Singh B., Chepko G., Clarke R., Sholler P. F., Lirio A. A., Foss C., et al. , ( 2003). Cadmium mimics the in vivo effects of estrogen in the uterus and mammary gland. Nat. Med. 9, 1081– 1084. Google Scholar CrossRef Search ADS PubMed Julin B., Wolk A., Bergkvist L., Bottai M., Akesson A. ( 2012). Dietary cadmium exposure and risk of postmenopausal breast cancer: A population-based prospective cohort study. Cancer Res. 72, 1459– 1466. Google Scholar CrossRef Search ADS PubMed Klaassen C. D., Liu J., Diwan B. A. ( 2009). Metallothionein protection of cadmium toxicity. Toxicol. Appl. Pharmacol. 238, 215– 220. Google Scholar CrossRef Search ADS PubMed Krock B. L., Skuli N., Simon M. C. ( 2011). Hypoxia-induced angiogenesis: Good and evil. Genes Cancer 2, 1117– 1133. Google Scholar CrossRef Search ADS PubMed Lee C., Patil S., Sartor M. A. ( 2016). Rna-enrich: A cut-off free functional enrichment testing method for rna-seq with improved detection power. Bioinformatics 32, 1100– 1102. Google Scholar CrossRef Search ADS PubMed Lee K., Zhang H., Qian D. Z., Rey S., Liu J. O., Semenza G. L. ( 2009). Acriflavine inhibits hif-1 dimerization, tumor growth, and vascularization. Proc. Natl. Acad. Sci.U.S.A. 106, 17910– 17915. Google Scholar CrossRef Search ADS PubMed Liao Y., Smyth G. K., Shi W. ( 2014). Featurecounts: An efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics 30, 923– 930. Google Scholar CrossRef Search ADS PubMed Lin C.-Y., Ström A., Vega V., Li Kong S., Li Yeo A., Thomsen J. S., Chan W., Doray B., Bangarusamy D. K., Ramasamy A., et al. ( 2004). Discovery of estrogen receptor α target genes and response elements in breast tumor cells. Genome Biol. 5, R66. Google Scholar CrossRef Search ADS PubMed Love M. I., Huber W., Anders S. ( 2014). Moderated estimation of fold change and dispersion for rna-seq data with deseq2. Genome Biol. 15, 550. Google Scholar CrossRef Search ADS PubMed McElroy J. A., Shafer M. M., Trentham-Dietz A., Hampton J. M., Newcomb P. A. ( 2006). Cadmium exposure and breast cancer risk. J. Natl Cancer Inst. 98, 869– 873. Google Scholar CrossRef Search ADS PubMed Miller W. J., Lampp B., Powell G. W., Salotti C. A., Blackmon D. M. ( 1967). Influence of a high level of dietary cadmium on cadmium content in milk, excretion, and cow performance1. J. Dairy Sci. 50, 1404– 1408. Google Scholar CrossRef Search ADS Nagata C., Nagao Y., Nakamura K., Wada K., Tamai Y., Tsuji M., Yamamoto S., Kashiki Y. ( 2013). Cadmium exposure and the risk of breast cancer in japanese women. Breast Cancer Res. Treat. 138, 235– 239. Google Scholar CrossRef Search ADS PubMed Ohrvik H., Yoshioka M., Oskarsson A., Tallkvist J. ( 2006). Cadmium-induced disturbances in lactating mammary glands of mice. Toxicol. Lett. 164, 207– 213. Google Scholar CrossRef Search ADS PubMed Paine I. S., Lewis M. T. ( 2017). The terminal end bud: The little engine that could. J. Mammary Gland Biol. Neoplasia 22, 93– 108. Google Scholar CrossRef Search ADS PubMed Rizzolo P., Silvestri V., Falchetti M., Ottini L. ( 2011). Inherited and acquired alterations in development of breast cancer. The Application of Clinical Genetics 4, 145– 158. Google Scholar PubMed Robinson M. D., McCarthy D. J., Smyth G. K. ( 2010). Edger: A bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26, 139– 140. Google Scholar CrossRef Search ADS PubMed Russo J. 2016. The windows of susceptibility to breast cancer. In: The Pathobiology of Breast Cancer (J. Russo, Ed.). Springer International Publishing, Cham, 1– 20. Google Scholar CrossRef Search ADS Sawada N., Iwasaki M., Inoue M., Takachi R., Sasazuki S., Yamaji T., Shimazu T., Endo Y., Tsugane S. ( 2012). Long-term dietary cadmium intake and cancer incidence. Epidemiology 23, 368– 376. Google Scholar CrossRef Search ADS PubMed Seagroves T. N., Hadsell D., McManaman J., Palmer C., Liao D., McNulty W., et al. ( 2003). Hif1alpha is a critical regulator of secretory differentiation and activation, but not vascular expansion, in the mouse mammary gland. Development 130, 1713– 1724. Google Scholar CrossRef Search ADS PubMed Shimoda L. A., Semenza G. L. ( 2011). Hif and the lung: role of hypoxia-inducible factors in pulmonary development and disease. Am. J. Respir. Crit. Care Med. 183, 152– 156. Google Scholar CrossRef Search ADS PubMed Siegel R. L., Miller K. D., Jemal A. ( 2016). Cancer statistics, 2016. Cancer J. Clin. 66, 7– 30. Google Scholar CrossRef Search ADS Sokol E. S., Miller D. H., Breggia A., Spencer K. C., Arendt L. M., Gupta P. B. ( 2016). Growth of human breast tissues from patient cells in 3d hydrogel scaffolds. Breast Cancer Res. 18, 19. Google Scholar CrossRef Search ADS PubMed Strumylaite L., Bogusevicius A., Ryselis S., Pranys D., Poskiene L., Kregzdyte R., et al. ( 2008). Association between cadmium and breast cancer. Medicina 44, 415– 420. Google Scholar PubMed Strumylaite L., Bogusevicius A., Abdrachmanovas O., Baranauskiene D., Kregzdyte R., Pranys D., Poskiene L. ( 2011). Cadmium concentration in biological media of breast cancer patients. Breast Cancer Res. Treat. 125, 511– 517. Google Scholar CrossRef Search ADS PubMed Strumylaite L., Kregzdyte R., Bogusevicius A., Poskiene L., Baranauskiene D., Pranys D. ( 2014). Association between cadmium and breast cancer risk according to estrogen receptor and human epidermal growth factor receptor 2: Epidemiological evidence. Breast Cancer Res. Treat. 145, 225– 232. Google Scholar CrossRef Search ADS PubMed Tsuji K., Kitamura S., Makino H. ( 2014). Hypoxia-inducible factor 1alpha regulates branching morphogenesis during kidney development. Biochem. Biophys. Res. Commun. 447, 108– 114. Google Scholar CrossRef Search ADS PubMed Visvader J. E., Stingl J. ( 2014). Mammary stem cells and the differentiation hierarchy: Current status and perspectives. Genes Dev. 28, 1143– 1158. Google Scholar CrossRef Search ADS PubMed Waalkes M. P. ( 2003). Cadmium carcinogenesis. Mut. Res. 533, 107– 120. Google Scholar CrossRef Search ADS Wang L., Wang S., Li W. ( 2012). Rseqc: Quality control of rna-seq experiments. Bioinformatics 28, 2184– 2185. Google Scholar CrossRef Search ADS PubMed Wang Y., Zhou B. P. ( 2011). Epithelial-mesenchymal transition in breast cancer progression and metastasis. Chin. J. Cancer 30, 603– 611. Google Scholar CrossRef Search ADS PubMed Wei Z., Shaikh Z. A. ( 2017). Cadmium stimulates metastasis-associated phenotype in triple-negative breast cancer cells through integrin and beta-catenin signaling. Toxicol. Appl. Pharmacol. 328, 70– 80. Google Scholar CrossRef Search ADS PubMed Zhou X., Lindsay H., Robinson M. D. ( 2014). Robustly detecting differential expression in rna sequencing data using observation weights. Nucleic Acids Res. 42, e91. Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of the Society of Toxicology. All rights reserved. For permissions, please e-mail: email@example.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)
Toxicological Sciences – Oxford University Press
Published: May 7, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera