Solid-Basaloid Adenoid Cystic Carcinoma of the Breast: An Aggressive Subtype Enriched for Notch Pathway and Chromatin Modifier Mutations With MYB Overexpression

Adenoid cystic carcinoma (AdCC) is a rare triple-negative breast cancer analogous to its extramammary counterparts. Diagnosis of the more aggressive solid-basaloid variant of AdCC (SB-AdCC) can be challenging due to poorly defined histopathologic and molecular features. We characterized 22 invasive and in situ basaloid carcinomas by morphology, immunohistochemistry, genetics, and MYB status using multiple platforms and assessed clinical behavior and neoadjuvant chemotherapy responses. After consensus review, 16/22 cases were classified as SB-AdCC. All SB-AdCC had predominantly solid growth and at least focal myxohyaline stroma and were immune-poor. Eosinophilic squamoid cells (69%, 11/16) and basement membrane-like secretions (69%, 11/16) were common, and intercalated ducts (31%, 5/16) were less frequent. SB-AdCC typically expressed SOX10 (100%, 16/16) and luminal markers (100%, 16/16 CK7; 88%, 14/16 CD117; 93%, 13/14 CAM5.2). SMA (40%, 6/15) expression was less common, and SMM (27%, 3/11), GATA3 (20%, 3/15), and p63 (25%, 4/16) were mostly negative. MYB protein and/or MYB RNA overexpression was universal in evaluable cases (13/13), with RNA in situ hybridization (10/10) more reliable than immunohistochemistry (10/11, plus 4 excisions inconclusive). Fluorescence in situ hybridization and/or next-generation sequencing identified MYB rearrangements (20%, 3/15) and amplifications/copy gains (60%, 9/15) but no MYB::NFIB fusions. SB-AdCC often had aberrations in Notch pathway (60%, including 40% NOTCH1 and 20% NOTCH2) and/or chromatin modifier (60%, including 33% CREBBP) genes, with relatively infrequent TP53 mutations (27%). Unclassified invasive basaloid carcinomas lacking described histologic features of SB-AdCC (n = 4) and basaloid ductal carcinoma in situ (n = 2) showed similar immunoprofiles and genetics as SB-AdCC, including Notch aberrations and MYB overexpression with MYB rearrangements/amplifications. Overall, nodal (22%) and distant (33%) metastases were common, and 23% of patients died of disease (mean follow-up, 35 months; n = 22). Responses were poor in all 7 neoadjuvant chemotherapy-treated patients, without any achieving pathologic complete response. The data highlight the histopathologic spectrum of basaloid carcinomas including SB-AdCC and reveal shared genetics and MYB activation, which can be diagnostically useful. Aggressive behavior and poor treatment responses emphasize a need for additional treatment approaches.

Epimorphic regeneration in the mammalian tympanic membrane

Adult mammals are generally believed to have limited ability to regenerate complex tissues and instead, repair wounds by forming scars. In humans and across mammalian species, the tympanic membrane (TM) rapidly repairs perforations without intervention. Using mouse models, we demonstrate that the TM repairs itself through a process that bears many hallmarks of epimorphic regeneration rather than typical wound healing. Following injury, the TM forms a wound epidermis characterized by EGFR ligand expression and signaling. After the expansion of the wound epidermis that emerges from known stem cell regions of the TM, a multi-lineage blastema-like cellular mass is recruited. After two weeks, the tissue architecture of the TM is largely restored, but with disorganized collagen. In the months that follow, the organized and patterned collagen framework of the TM is restored resulting in scar-free repair. Finally, we demonstrate that deletion of Egfr in the epidermis results in failure to expand the wound epidermis, recruit the blastema-like cells, and regenerate normal TM structure. This work establishes the TM as a model of mammalian complex tissue regeneration.

Specific Oral Microbial Differences in Proteobacteria and Bacteroidetes Are Associated with Distinct Sites When Moving from Healthy Mucosa to Oral Dysplasia-A Microbiome and Gene Profiling Study and Focused Review

Oral potentially malignant disorders (OPMDs) are a group of conditions that carry a risk of oral squamous cell carcinoma (OSCC) development. Recent studies indicate that periodontal disease-associated pathogenic bacteria may play a role in the transition from healthy mucosa to dysplasia and to OSCC. Yet, the microbial signatures associated with the transition from healthy mucosa to dysplasia have not been established. To characterize oral microbial signatures at these different sites, we performed a 16S sequencing analysis of both oral swab and formalin-fixed, paraffin-embedded tissue (FFPE) samples. We collected oral swabs from healthy mucosa (from healthy patients), histologically normal mucosa adjacent to dysplasia, and low-grade oral dysplasia. Additionally, FFPE samples from histologically normal mucosa adjacent to OSCC, plus low grade and high-grade oral dysplasia samples were also collected. The collected data demonstrate significant differences in the alpha and beta microbial diversities of different sites in oral mucosa, dysplasia, and OSCC, as well as increased dissimilarities within these sites. We found that the Proteobacteria phyla abundance increased, concurrent with a progressive decrease in the Firmicutes phyla abundance, as well as altered levels of Enterococcus cecorum, Fusobacterium periodonticum, Prevotella melaninogenica, and Fusobacterium canifelinum when moving from healthy to diseased sites. Moreover, the swab sample analysis indicates that the oral microbiome may be altered in areas that are histologically normal, including in mucosa adjacent to dysplasia. Furthermore, trends in specific microbiome changes in oral swab samples preceded those in the tissues, signifying early detection opportunities for clinical diagnosis. In addition, we evaluated the gene expression profile of OSCC cells (HSC-3) infected with either P. gingivalis, T. denticola, F. nucelatum, or S. sanguinis and found that the three periodontopathogens enrich genetic processes related to cancer progression, including skin keratinization/cornification, while the commensal enriched processes related to RNA processing and adhesion. Finally, we reviewed the dysplasia microbiome literature and found a significant decrease in commensal bacteria, such as the Streptococci genus, and a simultaneous increase in pathogenic bacteria, mainly Bacteroidetes phyla and Fusobacterium genus. These findings suggest that features of the oral microbiome can serve as novel biomarkers for dysplasia and OSCC disease progression.

An agonistic anti-signal regulatory protein α antibody for chronic inflammatory diseases

Signal regulatory protein (SIRPα) is an immune inhibitory receptor expressed by myeloid cells to inhibit immune cell phagocytosis, migration, and activation. Despite the progress of SIRPα and CD47 antagonist antibodies to promote anti-cancer immunity, it is not yet known whether SIRPα receptor agonism could restrain excessive autoimmune tissue inflammation. Here, we report that neutrophil- and monocyte-associated genes including SIRPA are increased in inflamed tissue biopsies from patients with rheumatoid arthritis and inflammatory bowel diseases, and elevated SIRPA is associated with treatment-refractory ulcerative colitis. We next identify an agonistic anti-SIRPα antibody that exhibits potent anti-inflammatory effects in reducing neutrophil and monocyte chemotaxis and tissue infiltration. In preclinical models of arthritis and colitis, anti-SIRPα agonistic antibody ameliorates autoimmune joint inflammation and inflammatory colitis by reducing neutrophils and monocytes in tissues. Our work provides a proof of concept for SIRPα receptor agonism for suppressing excessive innate immune activation and chronic inflammatory disease treatment.

Improved handling and embedding schemes for cultured murine neuroretinal explants

Traumatic, inherited, and age-related degenerative diseases of the retina, such as retinal detachment, retinitis pigmentosa, and age-related macular degeneration, are characterized by the irreversible loss of retinal neurons. While current treatments aim to prevent neuronal degeneration, there are no available treatments to restore neurons after loss. Cultured murine neuroretinal tissue explants model retinal injury and offer a high throughput approach to identify experimental interventions capable of regenerating neurons. Formalin-fixed paraffin-embedded (FFPE) preparations of murine neuroretinal explants can be used to identify cells throughout the retinal layers to provide information on proliferation and activity following exposure to therapeutics. However, retinal explants are friable, particularly after ex vivo culture, sample handling and FFPE processing steps can result in tissue loss and damage. Friability also prohibits bisecting samples post-culture to display more than one region of interest for analysis. We developed a sample handling and embedding technique for cultured murine neuroretinal explants using HistogelTM in combination with a post-processing trimming step that eliminates tissue loss, increases cross-sectional retinal representation, and captures proximal and central retina on one slide to facilitate analysis of explants subjected to neurotrophic compounds.

Genetic analysis of pleomorphic and florid lobular carcinoma in situ variants: frequent ERBB2/ERBB3 alterations and clonal relationship to classic lobular carcinoma in situ and invasive lobular carcinoma

Pleomorphic (PLCIS) and florid (FLCIS) lobular carcinoma in situ are rare histologic variants of LCIS that are considered more aggressive than classic LCIS (CLCIS), but optimal treatment is controversial. The genetic drivers of these lesions and their clonal relationships to paired CLCIS and ILC have not been characterized. We used capture-based next-generation sequencing to profile 16 LCIS variants (ten PLCIS, six FLCIS), including paired synchronous ILC and CLCIS in 11 and nine cases, respectively. Recurrent pathogenic alterations included CDH1 (9/10 PLCIS, 6/6 FLCIS), PIK3CA (7/10 PLCIS, 2/6 FLCIS), ERBB2 (6/10 PLCIS, 2/6 FLCIS; six mutations, two amplifications), ERBB3 (1/10 PLCIS, 2/6 FLCIS), FOXA1 (4/10 PLCIS, 1/6 FLCIS), TP53 (3/10 PLCIS), and CCND1 (2/10 PLCIS, 1/6 FLCIS). Mutational profiles and mean copy number alterations (CNA) were similar between LCIS variants with and without ILC. Compared with ILC in The Cancer Genome Atlas (TCGA), PLCIS, FLCIS, and associated ILC were enriched for ERBB2 mutations, and PLCIS was enriched for TP53 and FOXA1 mutations. Shared pathogenic mutations and CNA were identified between the LCIS variant and ILC in all cases, and between CLCIS and the LCIS variant/ILC in 89%. CLCIS to PLCIS progression was associated with increased mean nonsynonymous mutations and additional pathogenic alterations and/or CNA in 80%. Mean nonsynonymous mutations and CNA were similar between PLCIS and ILC, although additional pathogenic mutations were associated with invasion in a subset (43%). FLCIS harbored additional clonal pathogenic mutations in only 1/3 cases, and these were not shared with ILC, which was genetically divergent. In another case, ILC was genetically more similar to CLCIS than FLCIS. The results highlight clonal relationships between PLCIS/FLCIS and CLCIS, and implicate PLCIS as a genetically advanced ILC precursor. Frequent ERBB2/ERBB3 alterations in PLCIS and FLCIS are consistent with more aggressive behavior and may have prognostic and therapeutic implications.

Autophagic Degradation of NBR1 Restricts Metastatic Outgrowth during Mammary Tumor Progression

Although autophagy is being pursued as a therapeutic target in clinical oncology trials, its effects on metastasis, the principal cause of cancer mortality, remain unclear. Here, we utilize mammary cancer models to temporally delete essential autophagy regulators during carcinoma progression. Though genetic ablation of autophagy strongly attenuates primary mammary tumor growth, impaired autophagy promotes spontaneous metastasis and enables the outgrowth of disseminated tumor cells into overt macro-metastases. Transcriptomic analysis reveals that autophagy deficiency elicits a subpopulation of otherwise luminal tumor cells exhibiting basal differentiation traits, which is reversed upon preventing accumulation of the autophagy cargo receptor, Neighbor to BRCA1 (NBR1). Furthermore, pharmacological and genetic induction of autophagy suppresses pro-metastatic differentiation and metastatic outgrowth. Analysis of human breast cancer data reveal that autophagy gene expression inversely correlates with pro-metastatic differentiation signatures and predicts overall and distant metastasis-free survival. Overall, these findings highlight autophagy-dependent control of NBR1 as a key determinant of metastatic progression.

De novo leukocyte chemotactic factor 2 amyloidosis in a pediatric renal allograft, 15 years post-transplant

Leukocyte chemotactic factor 2 amyloidosis (ALECT2) is a recently described form of systemic amyloidosis, which most commonly affects the kidney and liver. The LECT2 protein is produced during inflammatory processes, but its precise function in renal diseases in unclear. ALECT2, however, is known to be a relatively common form of renal amyloidosis, after amyloid light chain and serum amyloid A types and is most often seen in patients of Hispanic ethnicity. ALECT2 can occur de novo or as recurrent disease in kidney transplants. We present the first case, to our knowledge, of de novo ALECT2 in a pediatric kidney transplant patient, 15 years post-transplant.

Identification of high-risk human papillomavirus and Rb/E2F pathway genomic alterations in mutually exclusive subsets of colorectal neuroendocrine carcinoma

Colorectal neuroendocrine carcinomas, both small cell and large cell types, are highly aggressive tumors with poor prognosis compared with colorectal adenocarcinoma. The molecular drivers of neuroendocrine carcinoma are best defined in small cell lung cancer, which shows near-universal genomic alterations in TP53 and RB1. The genetics of colorectal neuroendocrine carcinoma remain poorly understood; recent studies demonstrated infrequent RB1 alterations and genetics closely resembling colorectal adenocarcinoma. To better define the molecular pathogenesis of colorectal neuroendocrine carcinoma, we performed capture-based next-generation sequencing on 25 cases and evaluated for expression of p53, Rb, p16, and high-risk human papillomavirus (HR-HPV) subtypes using immunohistochemistry, in situ hybridization, and polymerase chain reaction. Rb/E2F pathway dysregulation was identified in nearly all cases (23/25, 92%) and occurred via three distinct mechanisms. First, RB1 genomic alteration was present in 56% (14/25) of cases and was accompanied by Rb protein loss, high p16 expression, and absence of HR-HPV; these cases also had frequent genomic alterations in TP53, the PI3K/Ras and Wnt pathways, as well as in DNA repair genes, with 4/14 cases being hypermutated. Second, 16% (4/25) of cases, all left-sided, had TP53 alteration without RB1 alteration; half of these harbored high-level amplifications in CCNE1 and MYC or MYCN and arose in patients with ulcerative colitis. Finally, 28% (7/25) of cases, all rectal or anal, lacked genomic alterations in RB1 or TP53 but were positive for HR-HPV. Our data demonstrate that Rb/E2F pathway dysregulation is essential in the pathogenesis of colorectal neuroendocrine carcinoma, akin to neuroendocrine carcinomas in other anatomic sites. Moreover, colorectal neuroendocrine carcinomas stratify into three distinct molecular subgroups, which can be differentiated based on Rb protein and HR-HPV status. HR-HPV infection represents a distinct mechanism for Rb and p53 inactivation in cases lacking genomic alterations in either gene. Differential treatment strategies for hypermutated and HPV-driven cases could improve patient outcomes.

Myoepithelial cells are a dynamic barrier to epithelial dissemination

Myoepithelial cells function collectively as a dynamic barrier to the invasion and dissemination of Twist1⁺ luminal epithelial cells and both luminal and basal phenotype breast cancer cells. Barrier function depends on myoepithelial abundance and both smooth muscle contractility and intercellular adhesion within the myoepithelium.

The mammary epithelium is composed of an inner luminal and surrounding myoepithelial cell layer. The presence of cancer cells beyond the myoepithelium defines invasive breast cancer, yet the role of the myoepithelium during invasion remains unclear. We developed a 3D organotypic culture assay to model this process through lineage-specific expression of the prometastatic transcription factor Twist1. We sought to distinguish the functional role of the myoepithelium in regulating invasion and local dissemination. Myoepithelial-specific Twist1 expression induced cell-autonomous myoepithelial cell escape. Remarkably, luminal-specific Twist1 expression was rarely sufficient for escape. Time-lapse microscopy revealed that myoepithelial cells collectively restrain and reinternalize invading Twist1⁺ luminal cells. Barrier function correlated with myoepithelial abundance and required the expression of α-smooth muscle actin and P-cadherin. We next demonstrated that myoepithelial cells can restrain and recapture invasive cancer cells. Our data establish the concept of the myoepithelium as a dynamic barrier to luminal dissemination and implicate both smooth muscle contractility and intercellular adhesion in barrier function.

Twist1-positive epithelial cells retain adhesive and proliferative capacity throughout dissemination

Dissemination is the process by which cells detach and migrate away from a multicellular tissue. The epithelial-to-mesenchymal transition (EMT) conceptualizes dissemination in a stepwise fashion, with downregulation of E-cadherin leading to loss of intercellular junctions, induction of motility, and then escape from the epithelium. This gain of migratory activity is proposed to be mutually exclusive with proliferation. We previously developed a dissemination assay based on inducible expression of the transcription factor Twist1 and here utilize it to characterize the timing and dynamics of intercellular adhesion, proliferation and migration during dissemination. Surprisingly, Twist1(+) epithelium displayed extensive intercellular junctions, and Twist1(-) luminal epithelial cells could still adhere to disseminating Twist1(+) cells. Although proteolysis and proliferation were both observed throughout dissemination, neither was absolutely required. Finally, Twist1(+) cells exhibited a hybrid migration mode; their morphology and nuclear deformation were characteristic of amoeboid cells, whereas their dynamic protrusive activity, pericellular proteolysis and migration speeds were more typical of mesenchymal cells. Our data reveal that epithelial cells can disseminate while retaining competence to adhere and proliferate.

Adhesion in mammary development: novel roles for E-cadherin in individual and collective cell migration

Epithelial tissues are essential for barrier function, secretion, and regulation of fluid transport. Their function requires cell polarity and cell-cell adhesion, mediated through intercellular junctions. Conversely, disruption of adhesion and polarity is thought to drive cancer progression. The mammary gland is an important model for cell adhesion due to its postnatal hormonally regulated development; ducts undergo branching morphogenesis in response to steroid hormones during puberty. These hormonal signals induce a transition from simple to stratified architecture, initiated by asymmetric luminal cell divisions. Ductal elongation is accomplished by this multilayered, low-polarity epithelium, and polarity is reestablished as elongation ceases. The requirement for cell adhesion has been tested in 3D culture and in vivo, using gene deletion, knockdown, and misexpression in both developmental and homeostatic contexts. Attention has focused on E-cadherin, the major classical cadherin in luminal epithelial cells. Classic studies revealed a requirement for E-cadherin during lactation, and E-cadherin loss is widely posited to promote metastasis. However, recent findings demonstrated a broader requirement for E-cadherin during branching morphogenesis and homeostasis and also, surprisingly, in epithelial dissemination. These studies suggest that long-standing models of the role of adhesion in epithelial biology need to be revisited. Advances in inducible gene expression and knockdown, CRISPR/Cas9 technology, and fluorescent labeling of genetically modified cells offer the opportunity to test the roles of diverse adhesion systems and to develop a mechanistic understanding of how cell adhesion regulates development and cancer.

3D culture assays of murine mammary branching morphogenesis and epithelial invasion

Epithelia are fundamental tissues that line cavities, glands, and outer body surfaces. We use three-dimensional (3D) embedded culture of primary murine mammary epithelial ducts, called "organoids," to recapitulate in days in culture epithelial programs that occur over weeks deep within the body. Modulating the composition of the extracellular matrix (ECM) allows us to model cell- and tissue-level behaviors observed in normal development, such as branching morphogenesis, and in cancer, such as invasion and dissemination. Here, we describe a collection of protocols for 3D culture of mammary organoids in different ECMs and for immunofluorescence staining of 3D culture samples and mammary gland tissue sections. We illustrate expected phenotypic outcomes of each assay and provide troubleshooting tips for commonly encountered technical problems.

Three-dimensional organotypic culture: experimental models of mammalian biology and disease

Mammalian organs are challenging to study as they are fairly inaccessible to experimental manipulation and optical observation. Recent advances in three-dimensional (3D) culture techniques, coupled with the ability to independently manipulate genetic and microenvironmental factors, have enabled the real-time study of mammalian tissues. These systems have been used to visualize the cellular basis of epithelial morphogenesis, to test the roles of specific genes in regulating cell behaviours within epithelial tissues and to elucidate the contribution of microenvironmental factors to normal and disease processes. Collectively, these novel models can be used to answer fundamental biological questions and generate replacement human tissues, and they enable testing of novel therapeutic approaches, often using patient-derived cells.

Twist1-induced dissemination preserves epithelial identity and requires E-cadherin

Dissemination of epithelial cells is a critical step in metastatic spread. Molecular models of dissemination focus on loss of E-cadherin or repression of cell adhesion through an epithelial to mesenchymal transition (EMT). We sought to define the minimum molecular events necessary to induce dissemination of cells out of primary murine mammary epithelium. Deletion of E-cadherin disrupted epithelial architecture and morphogenesis but only rarely resulted in dissemination. In contrast, expression of the EMT transcription factor Twist1 induced rapid dissemination of cytokeratin-positive epithelial cells. Twist1 induced dramatic transcriptional changes in extracellular compartment and cell-matrix adhesion genes but not in cell-cell adhesion genes. Surprisingly, we observed disseminating cells with membrane-localized E-cadherin and β-catenin, and E-cadherin knockdown strongly inhibited Twist1-induced single cell dissemination. Dissemination can therefore occur with retention of epithelial cell identity. The spread of cancer cells during metastasis could similarly involve activation of an epithelial motility program without requiring a transition from epithelial to mesenchymal character.

Abstract C39: Building a novel molecular model for Twist1-induced epithelial dissemination

Most solid tumors originate from epithelial cells, and their metastasis represents the major cause of death among cancer patients. Because each tumor contains many mutations, it has been difficult to isolate how individual genetic events contribute to a tumor cell's ability to disseminate. Two posited mechanisms for metastasis are loss of cell-cell adhesion and activation of an epithelial-mesenchymal transition (EMT). These molecular models converge on the cell adhesion gene E-cadherin (Cdh1). E-cadherin is frequently mutated in human cancers, and it can also be transcriptionally repressed during EMT. In turn, expression of EMT transcription factors in multiple epithelial cancers has been correlated with increased invasion and metastasis. However, the acute consequences of E-cadherin loss or EMT activator expression in otherwise normal epithelial tissues remain incompletely understood.

We therefore utilized a combination of organotypic culture, Cre-lox based gene deletion, inducible expression systems, and time-lapse imaging to test the sufficiency of single genetic perturbations to induce dissemination in primary murine mammary epithelium. E-cadherin deletion induced loss of simple epithelial architecture and inhibited branching morphogenesis in 3D culture and in vivo. Furthermore, E-cadherin loss acutely induced collective invasion past basement membrane. However, E-cadherin- cells remained non-protrusive, adhered to one another, and rarely disseminated. Conversely, expression of the EMT transcription factor Twist1 induced rapid dissemination of individual epithelial cells into the extracellular matrix (ECM), and disseminated cells proliferated to form secondary sites. Surprisingly, Twist1+ cells were observed to have membrane-localized E-cadherin during dissemination and during migration in the ECM. Our data thus provide distinct roles for E-cadherin and Twist1 in the early cellular processes of metastasis. We are currently testing whether shRNA knockdown of E-cadherin synergizes with Twist1 to further enhance dissemination.

To date, we have demonstrated the sufficiency of constitutive Twist1 expression to induce dissemination of both major mammary cell types: inner luminal epithelial cells and outer myoepithelial cells. Moreover, we observe abnormal localization of myoepithelial cells and basement membrane to the inside of the tissue. However, it remains unclear why some Twist1-expressing cells disseminate while many others remain in the epithelium. We hypothesize that intrinsic, cell type-specific factors influence disseminative capacity. We next seek to test the sufficiency of Twist1 to induce cell dissemination and secondary site formation in distinct mammary epithelial subpopulations. We have developed mouse models that use cell-type specific promoters to drive Twist1 expression in myoepithelial, luminal progenitor, and differentiated luminal lineages. We anticipate that differences across cell types may help identify the subpopulations within a heterogeneous tumor with the highest metastatic potential. Furthermore, we expect that extrinsic factors, such as normal myoepithelial cells, may restrict dissemination of Twist1-expressing luminal cells. Understanding such modulation of Twist1's phenotypic outcomes could suggest new approaches to suppressing metastasis without necessitating direct inhibition of Twist1, itself an “undruggable” target.

Citation Format: Eliah R. Shamir, Phuoc T. Tran, Andrew J. Ewald. Building a novel molecular model for Twist1-induced epithelial dissemination. [abstract]. In: Proceedings of the Third AACR International Conference on Frontiers in Basic Cancer Research; Sep 18-22, 2013; National Harbor, MD. Philadelphia (PA): AACR; Cancer Res 2013;73(19 Suppl):Abstract nr C39.

ECM microenvironment regulates collective migration and local dissemination in normal and malignant mammary epithelium

Breast cancer progression involves genetic changes and changes in the extracellular matrix (ECM). To test the importance of the ECM in tumor cell dissemination, we cultured epithelium from primary human breast carcinomas in different ECM gels. We used basement membrane gels to model the normal microenvironment and collagen I to model the stromal ECM. In basement membrane gels, malignant epithelium either was indolent or grew collectively, without protrusions. In collagen I, epithelium from the same tumor invaded with protrusions and disseminated cells. Importantly, collagen I induced a similar initial response of protrusions and dissemination in both normal and malignant mammary epithelium. However, dissemination of normal cells into collagen I was transient and ceased as laminin 111 localized to the basal surface, whereas dissemination of carcinoma cells was sustained throughout culture, and laminin 111 was not detected. Despite the large impact of ECM on migration strategy, transcriptome analysis of our 3D cultures revealed few ECM-dependent changes in RNA expression. However, we observed many differences between normal and malignant epithelium, including reduced expression of cell-adhesion genes in tumors. Therefore, we tested whether deletion of an adhesion gene could induce sustained dissemination of nontransformed cells into collagen I. We found that deletion of P-cadherin was sufficient for sustained dissemination, but exclusively into collagen I. Our data reveal that metastatic tumors preferentially disseminate in specific ECM microenvironments. Furthermore, these data suggest that breaks in the basement membrane could induce invasion and dissemination via the resulting direct contact between cancer cells and collagen I.