Access the full text.
Sign up today, get DeepDyve free for 14 days.
I. Acerbi, L. Cassereau, Ivory Dean, Quanming Shi, A. Au, Catherine Park, Y.-Y. Chen, J. Liphardt, E. Hwang, V. Weaver (2015)
Human breast cancer invasion and aggression correlates with ECM stiffening and immune cell infiltration.Integrative biology : quantitative biosciences from nano to macro, 7 10
CB Moelans (2010)
10.1155/2010/829358Anal. Cell Pathol. (Amst.), 33
T. Russell, S. Jindal, S. Agunbiade, D. Gao, M. Troxell, V. Borges, P. Schedin (2015)
TUMORIGENESIS AND NEOPLASTIC PROGRESSION Myoepithelial Cell Differentiation Markers in Ductal Carcinoma in Situ Progression
Franco Klingberg, Melissa Chow, Anne Koehler, Stellar Boo, L. Buscemi, T. Quinn, M. Costell, B. Alman, E. Génot, B. Hinz (2014)
Prestress in the extracellular matrix sensitizes latent TGF-β1 for activationThe Journal of Cell Biology, 207
C. Moelans, R. Weger, H. Monsuur, Anoek Maes, P. Diest (2010)
Molecular Differences between Ductal Carcinoma In Situ and Adjacent Invasive Breast Carcinoma: A Multiplex Ligation-Dependent Probe Amplification StudyAnalytical Cellular Pathology (Amsterdam), 33
O. Sirka, Eliah Shamir, A. Ewald (2018)
Myoepithelial cells are a dynamic barrier to epithelial disseminationThe Journal of Cell Biology, 217
C. Allegra, D. Aberle, P. Ganschow, S. Hahn, Clara Lee, S. Millon-Underwood, M. Pike, S. Reed, A. Saftlas, S. Scarvalone, Arnold Schwartz, C. Slomski, G. Yothers, R. Zon (2009)
NIH state-of-the-science conference statement: diagnosis and management of ductal carcinoma in situ (DCIS).NIH consensus and state-of-the-science statements, 26 2
M. Sternlicht, P. Kedeshian, Z. Shao, S. Safarians, S. Barsky (1997)
The human myoepithelial cell is a natural tumor suppressor.Clinical cancer research : an official journal of the American Association for Cancer Research, 3 11
TD Russell (2015)
10.1016/j.ajpath.2015.07.004Am. J. Pathol., 185
K. Chin, C. Solórzano, D. Knowles, A. Jones, W. Chou, E. Rodriguez, W. Kuo, B. Ljung, K. Chew, K. Myambo, M. Miranda, S. Krig, J. Garbe, M. Stampfer, P. Yaswen, J. Gray, S. Lockett (2004)
In situ analyses of genome instability in breast cancerNature Genetics, 36
M. Sopel (2010)
The myoepithelial cell: its role in normal mammary glands and breast cancer.Folia morphologica, 69 1
Hugh McDevitt, Sibyl Munson, Rachel Ettinger, Ava Wu (2002)
Multiple roles for tumor necrosis factor-α and lymphotoxin α/β in immunity and autoimmunityArthritis Research, 4
Y. Man, Q. Sang (2004)
The significance of focal myoepithelial cell layer disruptions in human breast tumor invasion: a paradigm shift from the "protease-centered" hypothesis.Experimental cell research, 301 2
M. Allinen, R. Beroukhim, Li Cai, C. Brennan, J. Lahti-Domenici, Haiyan Huang, D. Porter, Min Hu, L. Chin, A. Richardson, S. Schnitt, W. Sellers, K. Polyak (2004)
Molecular characterization of the tumor microenvironment in breast cancer.Cancer cell, 6 1
Li Wang, Cong Xu, Xia Liu, Yang Yang, Lu Cao, Guomin Xiang, Fang Liu, Shu-ling Wang, Jing Liu, Q. Meng, J. Jiao, Y. Niu (2019)
TGF-β1 stimulates epithelial–mesenchymal transition and cancer-associated myoepithelial cell during the progression from in situ to invasive breast cancerCancer Cell International, 19
N. Castro, C. Osorio, C. Torres, E. Bastos, M. Mourão-Neto, F. Soares, H. Brentani, D. Carraro (2008)
Evidence that molecular changes in cells occur before morphological alterations during the progression of breast ductal carcinomaBreast Cancer Research : BCR, 10
C. Moelans, Roel Wegers, H. Monsuurs, A. Maess, P. Diest (2011)
Molecular differences between ductal carcinoma in situ and adjacent invasive breast carcinoma: a multiplex ligation-dependent probe amplification studyCellular Oncology (Dordrecht), 34
L. Steenbergen, A. Voogd, J. Roukema, W. Louwman, L. Duijm, J. Coebergh, L. Poll-Franse (2009)
Screening caused rising incidence rates of ductal carcinoma in situ of the breastBreast Cancer Research and Treatment, 115
M. Deugnier, J. Teulière, M. Faraldo, J. Thiery, M. Glukhova (2002)
The importance of being a myoepithelial cellBreast Cancer Research, 4
Sangjun Lee, S. Stewart, I. Nagtegaal, Jingqin Luo, Yun Wu, G. Colditz, D. Medina, D. Allred (2012)
Differentially expressed genes regulating the progression of ductal carcinoma in situ to invasive breast cancer.Cancer research, 72 17
(2020)
Whole-Exome Sequencing Analysis of the Progression from NonLow-Grade Ductal Carcinoma
M. Sanders, P. Schuyler, W. Dupont, D. Page (2005)
The natural history of low‐grade ductal carcinoma in situ of the breast in women treated by biopsy only revealed over 30 years of long‐term follow‐upCancer, 103
M. Nguyen, Maggie Lee, Jing Wang, J. Tomlinson, Z. Shao, M. Alpaugh, S. Barsky (2000)
The human myoepithelial cell displays a multifaceted anti-angiogenic phenotypeOncogene, 19
Jun Yao, S. Weremowicz, Bin Feng, R. Gentleman, J. Marks, R. Gelman, C. Brennan, K. Polyak (2006)
Combined cDNA array comparative genomic hybridization and serial analysis of gene expression analysis of breast tumor progression.Cancer research, 66 8
Pang-Kuo Lo, Yongshu Zhang, Yuan Yao, Benjamin Wolfson, Justine Yu, Shu-yan Han, Nadire Duru, Qun Zhou (2017)
Tumor-associated myoepithelial cells promote the invasive progression of ductal carcinoma in situ through activation of TGFβ signalingThe Journal of Biological Chemistry, 292
Michael Pickup, Hanane Laklai, I. Acerbi, P. Owens, Agnieszka Gorska, A. Chytil, M. Aakre, V. Weaver, H. Moses (2013)
Stromally derived lysyl oxidase promotes metastasis of transforming growth factor-β-deficient mouse mammary carcinomas.Cancer research, 73 17
Xiao-Jun Ma, R. Salunga, J. Tuggle, J. Gaudet, E. Enright, P. Mcquary, T. Payette, M. Pistone, K. Stecker, Brian Zhang, Yixian Zhou, H. Varnholt, Barbara Smith, M. Gadd, Erica Chatfield, J. Kessler, T. Baer, M. Erlander, D. Sgroi, J. Brugge (2003)
Gene expression profiles of human breast cancer progressionProceedings of the National Academy of Sciences of the United States of America, 100
P. Bankhead, M. Loughrey, José Fernández, Yvonne Dombrowski, D. McArt, P. Dunne, S. McQuaid, R. Gray, L. Murray, H. Coleman, J. James, M. Salto‐Tellez, P. Hamilton (2017)
QuPath: Open source software for digital pathology image analysisScientific Reports, 7
H. Duivenvoorden, A. Spurling, S. O'toole, B. Parker (2018)
Discriminating the earliest stages of mammary carcinoma using myoepithelial and proliferative markersPLoS ONE, 13
S. Lakhani, M. O'hare (2000)
The mammary myoepithelial cell - Cinderella or ugly sister?Breast Cancer Research, 3
K. Kerlikowske (2010)
Epidemiology of ductal carcinoma in situ.Journal of the National Cancer Institute. Monographs, 2010 41
C. Swallow, K. Zee, V. Sacchini, P. Borgen (1996)
Ductal carcinoma in situ of the breast: progress and controversy.Current problems in surgery, 33 7
M. Seijen, E. Lips, A. Thompson, S. Nik-Zainal, A. Futreal, E. Hwang, E. Verschuur, J. Lane, J. Jonkers, D. Rea, J. Wesseling, J. Wesseling, Precision team (2019)
Ductal carcinoma in situ: to treat or not to treat, that is the questionBritish Journal of Cancer, 121
Jl Jones, J. Shaw, J. Pringle, R. Walker (2003)
Primary breast myoepithelial cells exert an invasion‐suppressor effect on breast cancer cells via paracrine down‐regulation of MMP expression in fibroblasts and tumour cellsThe Journal of Pathology, 201
Tyler Risom, David Glass, I. Averbukh, Candace Liu, Alex Baranski, Adam Kagel, E. Mccaffrey, N. Greenwald, Belén Rivero-Gutiérrez, Siri Strand, S. Varma, A. Kong, Leeat Keren, Sucheta Srivastava, Chunfang Zhu, Zumana Khair, D. Veis, K. Deschryver, S. Vennam, C. Maley, E. Hwang, J. Marks, S. Bendall, G. Colditz, R. West, Michael Angelo (2021)
Transition to invasive breast cancer is associated with progressive changes in the structure and composition of tumor stromaCell, 185
MJ Bissell (2005)
10.1101/sqb.2005.70.013Cold Spring Harb. Symp. Quant. Biol., 70
D. Sgroi (2010)
Preinvasive breast cancer.Annual review of pathology, 5
L. Collins, R. Tamimi, H. Baer, J. Connolly, G. Colditz, S. Schnitt (2005)
Outcome of patients with ductal carcinoma in situ untreated after diagnostic biopsyCancer, 103
S. Barsky, N. Karlin (2005)
Myoepithelial Cells: Autocrine and Paracrine Suppressors of Breast Cancer ProgressionJournal of Mammary Gland Biology and Neoplasia, 10
Xi-chen Zhang, S. Hashemi, M. Yousefi, Chunling Gao, Joy Sheng, Jinsong Ni, Wan Wang, J. Mason, Y. Man (2009)
Atypical E-cadherin expression in cell clusters overlying focally disrupted mammary myoepithelial cell layers: implications for tumor cell motility and invasion.Pathology, research and practice, 205 6
B. Nielsen, Fritz Rank, José López, Milagros Balbín, Francisco Vizoso, Leif Lund, K. Danø, Carlos López-Otín (2001)
Collagenase-3 expression in breast myofibroblasts as a molecular marker of transition of ductal carcinoma in situ lesions to invasive ductal carcinomas.Cancer research, 61 19
L. Ding, Ying Su, A. Fassl, Kunihiko Hinohara, X. Qiu, Nicholas Harper, Sung Huh, Noga Bloushtain-Qimron, B. Jovanovic, Muhammad Ekram, Xiaoyuan Zi, William Hines, M. Alečković, Carlos Alcazar, Ryan Caulfield, D. Bonal, Q. Nguyen, Vanessa Merino, Sibgat Choudhury, Gabrielle Ethington, L. Panos, M. Grant, W. Herlihy, A. Au, G. Rosson, P. Argani, Andrea Richardson, D. Dillon, D. Allred, Kirsten Babski, Elizabeth Kim, Charles McDonnell, Jon Wagner, Ron Rowberry, Kristie Bobolis, C. Kleer, E. Hwang, Joanne Blum, Simona Cristea, P. Sicinski, R. Fan, Henry Long, S. Sukumar, So Park, Judy Garber, M. Bissell, Jun Yao, K. Polyak (2019)
Perturbed myoepithelial cell differentiation in BRCA mutation carriers and in ductal carcinoma in situNature Communications, 10
F Pareja (2020)
10.1158/1078-0432.CCR-19-2563Clin. Cancer Res, 26
Min Hu, Jun Yao, Danielle Carroll, S. Weremowicz, Haiyan Chen, D. Carrasco, A. Richardson, S. Violette, T. Nikolskaya, Y. Nikolsky, E. Bauerlein, W. Hahn, R. Gelman, C. Allred, M. Bissell, S. Schnitt, K. Polyak (2008)
Regulation of in situ to invasive breast carcinoma transition.Cancer cell, 13 5
B. Erbas, E. Provenzano, J. Armes, D. Gertig (2006)
The natural history of ductal carcinoma in situ of the breast: a reviewBreast Cancer Research and Treatment, 97
Michael Allen, Gareth Thomas, Sarah Clark, MarwaM. Dawoud, S. Vallath, S. Payne, Jennifer Gomm, S. Dreger, Sarah Dickinson, Dylan Edwards, C. Pennington, I. Šestak, Jack Cuzick, John Marshall, Ian Hart, Louise Jones (2013)
Altered Microenvironment Promotes Progression of Preinvasive Breast Cancer: Myoepithelial Expression of αvβ6 Integrin in DCIS Identifies High-risk Patients and Predicts RecurrenceClinical Cancer Research, 20
Melissa Adriance, J. Inman, O. Petersen, M. Bissell (2005)
Myoepithelial cells: good fences make good neighborsBreast Cancer Research, 7
O. Maller, A. Drain, Alexander Barrett, S. Borgquist, S. Borgquist, B. Ruffell, Igor Zakharevich, Thanh Pham, T. Gruosso, H. Kuasne, J. Lakins, I. Acerbi, J. Barnes, T. Nemkov, Aastha Chauhan, J. Gruenberg, A. Nasir, Ó. Bjarnadóttir, Z. Werb, P. Kabos, Yunni-Yi Chen, E. Hwang, Morag Park, L. Coussens, A. Nelson, K. Hansen, K. Hansen, V. Weaver (2020)
Tumor-associated macrophages drive stromal cell-dependent collagen crosslinking and stiffening to promote breast cancer aggressionNature materials, 20
Alberto Elosegui-Artola, Elsa Bazellières, Michael Allen, Ion Andreu, R. Oria, R. Sunyer, J. Gomm, J. Marshall, J. Jones, X. Trepat, Pere Roca-Cusachs (2014)
Rigidity sensing and adaptation through regulation of integrin typesNature materials, 13
Y. Man, L. Tai, R. Barner, R. Vang, J. Saenger, K. Shekitka, G. Bratthauer, D. Wheeler, Chang Liang, T. Vinh, B. Strauss (2003)
Cell clusters overlying focally disrupted mammary myoepithelial cell layers and adjacent cells within the same duct display different immunohistochemical and genetic features: implications for tumor progression and invasionBreast Cancer Research, 5
D. Porter, J. Lahti-Domenici, A. Keshaviah, Y. Bae, P. Argani, J. Marks, A. Richardson, Amiel Cooper, R. Strausberg, G. Riggins, S. Schnitt, E. Gabrielson, R. Gelman, K. Polyak (2003)
Molecular markers in ductal carcinoma in situ of the breast.Molecular cancer research : MCR, 1 5
M. Sternlicht, S. Barsky (1997)
The myoepithelial defense: a host defense against cancer.Medical hypotheses, 48 1
C. Kretschmer, A. Sterner-Kock, F. Siedentopf, W. Schoenegg, P. Schlag, W. Kemmner (2011)
Identification of early molecular markers for breast cancerMolecular Cancer, 10
www.nature.com/npjbcancer ARTICLE OPEN Mechanostimulation of breast myoepithelial cells induces functional changes associated with DCIS progression to invasion 1 1 1 1 2 2 Mary-Kate Hayward , Michael D. Allen , Jennifer J. Gomm , Iain Goulding , Clare L. Thompson , Martin M. Knight , 1 1 John F. Marshall and J. Louise Jones Women with ductal carcinoma in situ (DCIS) have an increased risk of progression to invasive breast cancer. Although not all women with DCIS will progress to invasion, all are treated as such, emphasising the need to identify prognostic biomarkers. We have previously shown that altered myoepithelial cells in DCIS predict disease progression and recurrence. By analysing DCIS duct size in sections of human breast tumour samples, we identified an associated upregulation of integrin β6 and an increase in periductal fibronectin deposition with increased DCIS duct size that associated with the progression of DCIS to invasion. Our modelling of the mechanical stretching myoepithelial cells undergo during DCIS progression confirmed the upregulation of integrin β6 and fibronectin expression in isolated primary and cell line models of normal myoepithelial cells. Our studies reveal that this mechanostimulated DCIS myoepithelial cell phenotype enhances invasion in a TGFβ-mediated upregulation of MMP13. Immunohistochemical analysis identified that MMP13 was specifically upregulated in DCIS, and it was associated with progression to invasion. These findings implicate tissue mechanics in altering the myoepithelial cell phenotype in DCIS, and that these alterations may be used to stratify DCIS patients into low and high risk for invasive progression. npj Breast Cancer (2022) 8:109 ; https://doi.org/10.1038/s41523-022-00464-4 INTRODUCTION ‘invasion signature’ led to a focus on the breast. microenviron- ment, comprising the myoepithelial, stromal and immune cells, For the majority of invasive breast cancers, progression follows implying key roles for them in the progression to invasive disease. transition through the pre-invasive stage of ductal carcinoma In particular, the presence of a myoepithelial cell layer is in situ (DCIS) . In DCIS, tumour cells proliferate within the lumen of characteristic of DCIS, and disruption to this interface is a hallmark the duct and are retained by a near-continuous myoepithelial cell of invasive progression. Myoepithelial cells play essential roles in layer, which lies in contact with the basement membrane (BM). mammary gland development and function, such that they Progression is marked by tumour cells breaching the myoepithe- maintain luminal epithelial cell polarity and induce ductal lial cell-BM interface and invading into the surrounding stroma. branching and differentiation during mammary gland develop- Prior to the establishment of screening mammography, DCIS 18–21 ment . Studies have indicated that myoepithelial cells play a accounted for <2% of all diagnosed breast cancers, but through 19,22,23 role in tumour suppression by secretion of protease screening programmes DCIS now accounts for ~25% of all breast 2–4 inhibitors and downregulation of matrix metalloproteases (MMPs) cancer diagnoses . Despite the increased detection and treat- that have inhibitory effects on tumour cell growth, invasion and ment of DCIS, there has not been a concurrent decrease in 24–28 angiogenesis . In addition, tumour cells adjacent to a focally invasive breast cancer (IBC) diagnosis , suggesting that many disrupted myoepithelial cell layer display gene expression cases of DCIS are overtreated that would not progress during a changes associated with invasive properties, higher proliferation woman’s lifetime. Indeed, in several small series where DCIS was 29–32 rates and associate with leukocyte infiltration . Due to these left untreated, owing to misdiagnosis, only ~40% progressed to 6–8 tumour suppressive roles, myoepithelial cells are considered invasive disease over 30 years . Thus, there is an urgent clinical gatekeepers to tumour progression. need to identify markers that will predict the progression of DCIS Whereas normal myoepithelial cells have been demonstrated to in order to better direct therapeutic intervention . be tumour suppressive, several studies have identified that DCIS Molecular approaches have been used extensively for decades myoepithelial cells exhibit an altered phenotype, and are to identify such markers that may predict DCIS progression, with 18,32–35 suggested to switch to a tumour promoter function .We most studies focusing on the comparison of tumour epithelial cells previously showed that DCIS myoepithelial cells exhibit de novo from DCIS with their invasive counterpart. Genomic profiling and expression of integrin β6, which is predictive of DCIS progression gene expression analysis has not revealed any specific alterations 10–15 to invasion and disease recurrence. Such that, integrin β6-positive associated with the progression to invasion , and suggests DCIS cases recurred more rapidly than integrin β6-negative DCIS DCIS exhibits a high level of similarity to IBC, supporting DCIS as a 16 36 cases, at 2.3 years compared to 11.4 years, respectively .In precursor to invasion . More recent studies using advanced culture and in vivo studies identified that myoepithelial cell molecular approaches confirm this similarity, but also demonstrate expression of integrin β6 enhanced breast tumour cell invasion through variations in clonal patterns that the mechanism of 17 36 progression of DCIS is likely very diverse . The lack of a clear through TGFβ-mediated upregulation of MMP9 . In support of 1 2 Centre for Tumour Biology, Barts Cancer Institute, Queen Mary University of London, John Vane Science Centre, Charterhouse Square, London EC1M 6BQ, UK. School of Engineering and Materials Science, Queen Mary University of London, Mile End Road, London E1 4NS, UK. email: l.j.jones@qmul.ac.uk Published in partnership with the Breast Cancer Research Foundation 1234567890():,; M.-K. Hayward et al. our previous work, others have demonstrated, using in vitro periductal fibronectin revealed that the adjacent normal breast studies, that tumour-associated myoepithelial cells secrete TGFβ ducts had little expression (6%), whereas the percent of positive to promote the invasive progression of DCIS cells due to DCIS ducts was higher in DCIS/IDC (87%) compared with pure enhancing epithelial-to-mesenchymal transition and basal-like DCIS (68%), with no difference between non-high and high-grade phenotypes through activation of the TGFβ/SMAD signalling pure DCIS (Fig. 1g, i and Supplementary Table 3). Matched duct 37,38 pathway . However, the mechanisms inducing such alterations scoring established a correlation between myoepithelial integrin to DCIS myoepithelial cells are largely unknown. β6 and periductal fibronectin expression (Fig. 1j and Supplemen- Here, we show that mimicking the mechanical stretching of tary Table 4). We also confirmed using the gene expression profile myoepithelial cells in DCIS duct expansion induces a DCIS study of human breast tissue samples, a progressive increase in myoepithelial cell phenotype associated with invasive progres- both ITGB6 and FN1 mRNA levels with DCIS progression to IDC (Fig. 1k, l). The data implicate a relationship between integrin β6- sion. We demonstrate expression of myoepithelial cell integrin β6 positive myoepithelial cells and fibronectin deposition surround- and periductal fibronectin in DCIS associates with a higher ing the duct as a function of DCIS progression to invasion. propensity to progress to invasive breast cancer. Using primary and cell line models, we identify that integrin β6-positive myoepithelial cells promote the deposition of fibronectin. Integrin β6-positive myoepithelial cells upregulate fibronectin Together, myoepithelial cell integrin β6 and fibronectin promote expression the activation of TGFβ signalling to induce the secretion of BM- We next investigated the potential relationship between myoe- degrading proteases MMP9 and MMP13 that facilitate breast pithelial integrin β6 and fibronectin deposition. For this purpose, tumour cell invasion. Furthermore, we show that this DCIS we analysed freshly isolated primary DCIS and normal myoepithe- myoepithelial cell phenotype is induced by mechanostimulation lial cells, as well as established myoepithelial cell lines, without in a TGFβ-dependent manner. The findings indicate that 36 and with the expression of integrin β6 . To select appropriate mechanostimulation-mediated alterations to DCIS myoepithelial DCIS samples for further analysis, we first assessed the expression cell function contribute to invasion and that these biomarkers may of integrin β6 using immunohistochemical staining of DCIS breast be used to stratify patients for risk of DCIS progression. tissues with patient-matched DCIS ductal organoid preparations available (Fig. 2a). Two integrin β6-low and two -high DCIS ductal organoid samples were then selected. FACS analysis of these RESULTS samples confirmed there was an increased frequency of integrin DCIS progression is accompanied by increased myoepithelial β6-positive myoepithelial cells in integrin β6-high DCIS compared cell expression of integrin β6 and periductal fibronectin to integrin β6-low DCIS (Fig. 2b). Quantitative reverse deposition transcriptase-PCR (qRT-PCR) analysis revealed higher levels of Tissue fibrosis is a feature of breast cancers that associates with FN1 in myoepithelial cells in integrin β6-high DCIS compared to tumour progression. Fibrotic breast tumours display increased integrin β6-low DCIS (Fig. 2c). Furthermore, induction of integrin abundance of ECM proteins and remodelling enzymes and β6 expression in isolated primary normal myoepithelial cells (β6- 39–41 elevated integrin signalling . We previously showed that 1989 and β6-1492) revealed an increase in fibronectin compared myoepithelial cell expression of integrin β6 promotes the invasion to the control cell populations (N-1989 and N-1492; Fig. 2d, e). of breast tumour cells through TGFβ-driven upregulation of the Similarly, using established myoepithelial cell lines; we identified ECM remodelling enzyme, MMP9 . In a study comparing gene higher levels of fibronectin in the integrin β6-positive myoepithe- lial cell line (β6-1089) compared to the integrin β6-negative expression profiles in human breast tissue samples, DCIS showed a myoepithelial cell line (N-1089; Fig. 2f–h). These findings support significant upregulation of ECM/integrin-related gene categories that integrin β6-positive myoepithelial cells stimulate the deposi- compared to normal breast (Fig. 1a) . Plotting fold change versus tion of fibronectin into the periductal microenvironment. p-values of the gene expression data illustrates several upregu- lated ECM proteins and downregulated BM proteins from the GO_Extracellular_Matrix_Structural_Constituent category in inva- Tumour promoting phenotype of integrin β6-positive sive ductal carcinoma (IDC) compared to DCIS (Fig. 1b). In myoepithelial cells is enhanced by fibronectin-mediated particular, FN1 was one of the most evident and consistent ECM activation of TGFβ signalling alterations in IDC, which led us to speculate fibronectin as a driver Myoepithelial integrin β6 promotes the invasion of breast tumour of DCIS progression. To investigate whether a causal role exists cells in a TGFβ-mediated manner, and integrin β6 activation of between ECM remodelling, myoepithelial integrin signalling and 36,43 TGFβ requires a mechanically resistant fibronectin matrix . This DCIS progression, we first assessed the duct-by-duct expression of raises the possibility that integrin β6-positive myoepithelial cells myoepithelial cell integrin β6 and periductal fibronectin in serial promote breast tumour cell invasion by stimulating TGFβ in a sections of DCIS tissues without (pure DCIS) and with (DCIS/IDC) fibronectin-dependent manner. We therefore examined whether invasive disease, as well as adjacent normal breast tissue (Fig. 1c, fibronectin could modulate the activation of TGFβ in integrin β6- Supplementary Table 1). Haematoxylin and eosin (H&E) staining positive myoepithelial cells, and if this elevated breast tumour cell confirmed the presence of normal and DCIS ducts within these invasion. We observed a decrease in integrin β6-positive tissues, and the presence of invasion in DCIS/IDC tissues was used myoepithelial cell line (β6-1089) migration and adhesion to as a marker of DCIS progression (Fig. 1d). Normal and DCIS ducts latency-associated peptide (LAP) with fibronectin knockdown KD show an intact myoepithelial cell layer, as shown by SMA (FN ) compared to control (CTL) (Fig. 3a, b). qRT-PCR analysis immunohistochemistry (Fig. 1e). Immunohistochemical staining revealed reduced levels of TGFB1 in integrin β6-positive myoe- for myoepithelial integrin β6 revealed that the adjacent normal pithelial cell line with fibronectin knockdown compared to control breast ducts had no expression, whereas DCIS ducts in all patients (Fig. 3c). Immunoblots revealed that the knockdown of fibronectin exhibited some expression of myoepithelial integrin β6 (Fig. 1f, h). in integrin β6-positive myoepithelial cell line reduced the ability of The percent of positive DCIS ducts for myoepithelial integrin β6 TGFβ to stimulate phosphorylation of SMAD2 compared to control was higher in high-grade pure DCIS (45%) compared with non- (Fig. 3d). These results suggest that fibronectin expression by high grade pure DCIS (27%), and increased further in DCIS/IDC integrin β6-positive myoepithelial cells enhances its TGFβ binding (68%), consistent with our previous reports that myoepithelial and activating properties. integrin β6 associates with DCIS progression to invasion (Fig. 1f, h We next explored whether fibronectin could enhance the and Supplementary Table 2) . Immunohistochemical staining for tumour-promoting function of integrin β6-positive myoepithelial npj Breast Cancer (2022) 109 Published in partnership with the Breast Cancer Research Foundation 1234567890():,; M.-K. Hayward et al. a b Gene set terms upregulated in DCIS/Normal GO_ECM_structural_constituent in IDC/DCIS REACTOME_degradation_of_the_extracellular_matrix GO_extracellular_matrix_structural_constituent_conferring_tensile_strength COL12A1 REACTOME_fibronectin_matrix_formation FN1 REACTOME_assembly_of_collagen_fibrils_&_other_multimeric_structures LAMA2 THSB1 REACTOME_integrin_cell_surface_interactions COL10A1 LAMA1 BOWIE_response_to_extracellular_matrix 1 COL11A1 FBN1 NABA_ECM_regulators COL4A1 -log (p-value) 0.0 0.5 1.0 1.5 2.0 -4 -2 0 2 4 logFC Normal DCIS DCIS/IDC d h i j Integrin 6 Fibronectin **** *** 100 100 100 R =0.52 DCIS 80 80 80 DCIS/IDC 60 60 60 40 40 40 20 20 20 0 0 0 0 2040 6080100 % FN-positive ducts k l ITGB6 FN1 15 15 10 10 g 5 5 0 0 Fig. 1 Myoepithelial expression of integrin β6 and periductal fibronectin deposition correlates with human DCIS progression. a GSEA of transcriptional profiles from GEO series GSE21422 showing upregulated extracellular matrix (ECM) remodelling and integrin interaction gene set terms associated DCIS (n= 9) compared to normal breast (n = 5). b Scatter plot of p-value (-log10) vs. log fold change (logFC) for gene expression from the GO_extracellular_matrix_constituent gene set for microarray data for IDC (n = 5) compared to DCIS (n = 9). c Cartoon depicting a normal duct, a DCIS duct without (pure DCIS) and with co-existent invasion (DCIS/IDC). d Haematoxylin and eosin (H&E) staining of human breast tissue samples featuring adjacent normal ducts, DCIS and DCIS/IDC. Scale bar, 100 μm. e–g Representative images of smooth- muscle actin (SMA) (e), integrin β6(f) and fibronectin (g) by immunohistochemical staining in human breast tissue samples featuring adjacent normal ducts, DCIS and DCIS/IDC. h, i Bar graphs showing the mean and individual percentage of ducts positive for integrin β6(h)or fibronectin (i) in adjacent normal (n = 40), DCIS (n = 20) and DCIS/IDC (n = 20) patient samples (error bars, +s.e.m). ****P < 0.0001 (h, ordinary one-way ANOVA and i, Kruskal–Wallis one-way ANOVA). See Supplementary Table 1 for patient information. j, Scatter plot depicting the linear regression of the percentage of positive DCIS ducts for integrin αvβ6(β6) and fibronectin (FN) in serial human tissue sections as in (f,g) (n = 40). k, l, mRNA expression of ITGB6 and FN1 in normal breast tissues (n= 5), DCIS (n = 9) and IDC (n = 5) (error bars, +s.e.m). *P = 0.019 (k) and **P = 0.0023 (l) (ordinary one-way ANOVA). cells. Invasion assays revealed that the knockdown of fibronectin suggest that DCIS myoepithelial cells exhibit a tumour-promoting in the integrin β6-positive myoepithelial cell line led to a decrease phenotype mediated by both integrin β6 and fibronectin. in breast tumour cell invasion in vitro compared to control (Fig. 3e). Consistently, knockdown of fibronectin in the integrin β6- DCIS myoepithelial cells enhance progression to invasion by positive myoepithelial cell line revealed the broad downregulation increasing the expression of MMP13 in protease expression, with the most differentially expressed MMP13 expression by stromal cells adjacent to DCIS has been proteases implicated in promoting breast cancer invasion through 44 implicated in promoting its progression to invasion . We next degradation of the BM, including MMP9 and MMP13 (Fig. 3f). qRT- examined whether myoepithelial MMP13 expression could PCR analysis confirmed reduced levels of MMP9 and MMP13 in promote the progression of DCIS to invasion. We first analysed integrin β6-positive myoepithelial cell line following fibronectin our cohort of DCIS tissue samples by immunohistochemical knockdown compared to control (Fig. 3g, h). These findings staining for MMP13. This revealed that the adjacent normal breast Published in partnership with the Breast Cancer Research Foundation npj Breast Cancer (2022) 109 Cartoon Fibronectin Integrin 6 SMA H&E %positive ducts Normal log2 DCIS DCIS/IDC Normal DCIS -log (p-value) Normal IDC DCIS log2 DCIS/IDC % 6-positive ducts Normal DCIS IDC M.-K. Hayward et al. Isolated DCIS myos a b c 6-low 6-high FACS FN1 25 3 **** 0 0 Isolated primary myos Myo cell lines d e f g h 1989 1492 FN1 1089 FN1 *** N- N- 6- N- 6- 4 6- **** 6 *** 6 FN FN HSC70 HSC70 Fig. 2 Integrin β6-positive myoepithelial cells upregulate fibronectin expression. a Representative images of integrin β6by immunohistochemical staining in human DCIS tissue samples classified as low and high expression of integrin β6. Scale bar, 100 μm. + + - b Bar graph showing the percent integrin β6-positive (β6 ) myoepithelial (myo) cells (β4 EpCAM ) in human DCIS samples (n = 4) (error bars, +s.e.m). c Bar graphs showing qRT-PCR analysis for FN1 using RNA isolated from DCIS myoepithelial cells with integrin β6-low (n = 2) and β6- high (n = 2) expression (error bars, +s.e.m). 3 technical replicates; ****P < 0.0001 (two-tailed t-test). d Immunoblots of integrin β6 and fibronectin (FN) in primary normal myoepithelial cells; 1989 and 1492, transfected with an empty vector (N-) or integrin β6 expression construct (β6-). HSC70 serves as loading control. Images are representative of three experiments. e Bar graphs showing qRT-PCR analysis for FN1 using RNA isolated from primary normal myoepithelial cells, generated as in (d) (error bars, +s.e.m). n = 3 biological replicates, 3 technical replicates; 1989 ***P = 0.0003 and ***P = 0.0009 (two-tailed t-test). f Immunoblots of integrin β6 and fibronectin in myoepithelial cell lines; N-1089 and β6-1089. HSC70 serves as loading control. Images are representative of three experiments. g Bar graphs showing qRT-PCR analysis for FN1 using RNA isolated from myoepithelial cell lines as in (f). n = 3 biological replicates, 3 technical replicates; ****P < 0.0001 (two-tailed t- test). h Representative images of integrin β6 (green) and fibronectin (FN, magenta) by immunofluorescent staining in myoepithelial cell lines as in (f). Nuclei were counterstained with DAPI (blue). Scale bar, 20 μm. Images are representative of three experiments. tissue had no expression, whereas the percentage of positive DCIS which results in the expansion of the duct and as a consequence, ducts was higher in DCIS/IDC (72%) compared with pure DCIS stretching of the myoepithelial cell layer (Fig. 5a, Supplementary Fig. 1a–e). We assessed DCIS duct size and identified that integrin (40%), with no difference between non-high and high grade pure β6-positive DCIS ducts (460 μm) on average were larger than DCIS (Fig. 4a, b and Supplementary Table 5). We also confirmed, integrin β6-negative DCIS ducts (380 μm) (Fig. 5b, c, Supplemen- using the gene expression profile study of human breast tissue tary Table 6). To mimic the stretching of the myoepithelial cell samples, a progressive increase in MMP13 mRNA levels with the layer as seen in DCIS expansion, we applied a 10% static, progression of DCIS (Fig. 4c). Consistently, qRT-PCR analysis equibiaxial stretch to isolated primary and cell line models of revealed that levels of MMP13 were increased in all integrin β6- normal myoepithelial cells. Consistent with our tissue study, positive primary and cell line models of myoepithelial cells mechanical stretching of the normal myoepithelial cell line (N- examined, compared to their integrin β6-negative counterparts 1089) led to an increase in the expression of integrin β6 and (Fig. 4d–g). Furthermore, invasion assays revealed that the fibronectin (Fig. 5d, g, j, m). Isolated primary normal myoepithelial knockdown of MMP13 in integrin β6-positive myoepithelial cell cells (N-1989 and N-1492) exposed to mechanical stretching line led to a decrease in breast tumour cell invasion in vitro similarly showed a significant increase in integrin β6 and compared to control (Fig. 4h). These findings identify MMP13 as a fibronectin expression (Fig. 5e, h, k, n and f, i, l, o, respectively). key protease elevated in DCIS progression, and implicate the These data suggest that the DCIS myoepithelial cell phenotype integrin β6-fibronectin-MMP13 axis in the pro-tumourigenic observed here may be regulated, at least in part, by evolving properties of DCIS myoepithelial cells. mechanics in the development of DCIS. We next explored whether mechanical stretching could induce Mechanical stretching of normal myoepithelial cells induces a the associated tumour-promoting phenotype seen in DCIS DCIS phenotype associated with integrin β6 expression myoepithelial cells. Invasion assays revealed both normal primary We next investigated potential mechanisms whereby a DCIS and cell line models of myoepithelial cells exposed to stretch- myoepithelial cell phenotype could be induced. DCIS is char- enhanced breast tumour cell invasion in vitro, compared to the acterised by the proliferation of tumour cells within the duct, unstretched controls (Fig. 6a, Supplementary Fig. 2a, b). npj Breast Cancer (2022) 109 Published in partnership with the Breast Cancer Research Foundation N-1089 6-1089 Integrin 6 Fold change N-1989 6-1989 6 -Myos N-1492 + - (% of 4 , EpCAM ) 6-1492 6-low 6-high Fold change Fold change 6-low 6-high FN/DAPI 6/DAPI N-1089 6-1089 M.-K. Hayward et al. Integrin v6+ myo cell line a b c d e KD Migration to LAP TGFB1 FN Adhesion to LAP CTL Invasion 0.8 1.5 1.5 0 5 1530 0 5 1530 **** CTL 1.5 CTL CM *** **** KD **** **** FN KD FN CM pSMAD2 0.6 1.0 1.0 1.0 0.4 SMAD2 0.5 0.5 0.5 0.2 HSC70 0.0 0.0 0.0 0.0 KD KD CTL FN CTL FN 231 MCF-7 f gh Protease array MMP9 MMP13 **** 1.5 2.0 1.5 CTL CM KD **** FN CM 1.5 1.0 1.0 1.0 0.5 0.5 0.5 0.0 0.0 0.0 KD KD CTL FN CTL FN Fig. 3 Integrin β6-positive myoepithelial cells promote breast cancer cell invasion by fibronectin-mediated activation of TGFβ signalling. KD a Migration assay for myoepithelial cell line; β6-1089 transfected with a control (CTL) or fibronectin-targeted (FN ) siRNA, to BSA and LAP (error bars, +s.e.m). n= 3 biological replicates, 3 technical replicates; ****P < 0.0001 (two-tailed t-test). b Adhesion assay for myoepithelial cell lines generated as in (a) to LAP (error bars, +s.e.m). n = 3 biological replicates, 3 technical replicates; ***P = 0.0001 (two-tailed t-test). c Bar graph showing qRT-PCR analysis for TGFB1 using RNA isolated from myoepithelial cell lines generated as in (a). n = 3 biological replicates, 3 technical replicates; ****P < 0.0001 (two-tailed t-test). d Immunoblots of phospho-SMAD2 (pSMAD2) and SMAD2 in myoepithelial cell lines generated as in (a), and stimulated with 5 ng/mL TGFβ (time in min). HSC70 serves as loading control. Images are representative of three experiments. e Invasion assay for MDA-MB-231 (231) and MCF-7 cells using conditioned media (CM) isolated from myoepithelial cell lines generated as in (a) (error bars, +s.e.m). n = 3 biological replicates, 4-6 technical replicates; ****P < 0.0001 (two-tailed t-test). f Array analysis for proteases using conditioned media isolated from myoepithelial cell lines generated as in (a) (error bars, +s.e.m). n = 1 biological replicate, 2 technical replicates. g, h Bar graphs showing qRT-PCR analysis for MMP9 (g) and MMP13 (h) using RNA isolated from myoepithelial cell lines generated as in (a) (error bars, +s.e.m). n = 3 biological replicates, 3 technical replicates; ****P < 0.0001 (two-tailed t-test). Consistently, mechanical stretching of all myoepithelial cells myoepithelial phenotype associated with progression to invasion, revealed a broad upregulation in protease expression compared and that these markers may be used to identify patients at higher to unstretched controls (Fig. 6b; Supplementary Fig. 2c, d). Indeed, risk for invasive progression. stretched normal primary and cell line models showed an increase in MMP9 and MMP13 (Fig. 6b; Supplementary Fig. 2c, d). Gelatin DISCUSSION zymography confirmed elevated MMP9 activity in all of the stretched primary and cell line models of normal myoepithelial Here we demonstrate that physical cues induce a tumour- cells examined, compared to the unstretched controls (Fig. 6c–e). promoting phenotype in myoepithelial cells. Our findings support Furthermore, qRT-PCR analysis confirmed enhanced levels of the critical role played by myoepithelial cells in DCIS progression MMP9 and MMP13 in all normal myoepithelial cells exposed to and are consistent with our prior data implicating myoepithelial stretch compared to unstretched controls (Fig. 6f–k). These results integrin β6 in driving invasion . We elaborate upon these prior show an association between duct expansion and induction of studies by identifying a molecular mechanism whereby tissue myoepithelial integrin β6 expression, and functionally link mechanics increases myoepithelial integrin β6 and periductal myoepithelial cell stretch to generation of a pro-tumourigenic fibronectin expression through the tension-dependent enhance- phenotype. These data suggests that evolving tissue mechanics ment of TGFβ signalling that promotes BM-degrading proteases during DCIS development could induce the tumour-promoting that facilitate invasive progression. Accordingly, our study phenotype of DCIS myoepithelial cells. identifies phenotypic changes in myoepithelial cells that may stratify women who are at higher risk of invasive progression and DCIS myoepithelial phenotype induced by mechanical are therefore ideal candidates for more aggressive therapies. Our stretching is mediated by TGFβ signalling findings also suggest that treatments aimed at inhibiting DCIS myoepithelial cell function would constitute novel treatment We next examined the relationship between mechanostimulation, modalities and could potentially be used to lower the number of TGFβ signalling and induction of the DCIS myoepithelial cell AB women undergoing invasive surgery. phenotype. Inhibition of the TGFβRII (RII ) with a blocking Previous work examining DCIS progression has attributed the antibody in stretched normal myoepithelial cell line (N-1089) transition to invasion to specific features of the breast micro- abrogated the upregulation of integrin β6 and fibronectin AB environment. Such prior studies have identified genes differen- compared to control (CTL ) (Fig. 7a-d). Invasion assays revealed tially expressed between DCIS and invasion that encode for cell inhibition of the TGFβRII reduced breast tumour cell invasion 15,42 in vitro, compared to control (Fig. 7e). Consistently, qRT-PCR adhesion and ECM-related proteins . In support of this, recent analysis showed reduced MMP9 and MMP13 levels in these cells high-dimensional analysis using multiplex ion beam imaging by (Fig. 7f, g). These data suggest mechanical stimulation of TGFβ time of flight (MIBI-TOF) to examine the histological stages of DCIS signalling could be essential in promoting, further, the DCIS progression identified the most distinctive property delineating Published in partnership with the Breast Cancer Research Foundation npj Breast Cancer (2022) 109 Relative expression Relative migration BS A MMP1 LAP MMP2 BS A MMP3 LAP MMP7 MMP8 MMP9 Relative adhesion MMP10 MMP12 MMP13 Fold change Fold change Fold change Relative invasion M.-K. Hayward et al. a b c Normal DCIS DCIS/IDC MMP13 MMP13 **** P = 0.1475 Isolated DCIS myos Myo cell lines Isolated primary myos Integrin v6+ myo cell line d ef g h Invasion MMP13 MMP13 MMP13 MMP13 **** *** *** 4 1.5 *** 5 **** 3 4 CTL CM *** KD MMP13 CM 2 1.0 0.5 1 1 0 0 0 0 0.0 231 MCF-7 Fig. 4 MMP13 expression correlates with human DCIS progression to invasion and promotes breast cancer cell invasion in vitro. a Representative images of MMP13 by immunohistochemical staining in human breast tissue samples featuring adjacent normal ducts, DCIS and DCIS/IDC. Scale bar, 20 μm. b Bar graphs showing the mean and individual percentage of ducts positive for MMP13 in adjacent normal (n = 40), DCIS (n= 20) and DCIS/IDC (n = 20) patient samples (error bars, +s.e.m). ****P < 0.0001 (Kruskal–Wallis one-way ANOVA). c mRNA expression of MMP13 in normal breast tissues (n = 5), DCIS (n = 9) and IDC (n = 5) (error bars, +s.e.m). * P = 0.0373 (Kruskal–Wallis one-way ANOVA). d–g Bar graphs showing qRT-PCR analysis for MMP13 using RNA isolated from DCIS myoepithelial cells with integrin β6-low and β6- high expression (d), myoepithelial cell lines; N-1089 and β6-1089 (e) and from primary normal myoepithelial cells; 1989 (f) and 1492 (g) transfected with an empty vector (N-) or integrin β6 expression construct (β6-) (error bars, +s.e.m). n = 3 biological replicates, 3 technical replicates; ****P < 0.0001 (d,e), ***P = 0.0001 (f) and ***P = 0.0006 (g) (two-tailed t-test). h Invasion assay for MDA-MB-231 (231) and MCF-7 cells KD using conditioned media (CM) isolated from myoepithelial cell line; β6-1089 transfected with a control (CTL) or MMP13-targeted (MMP13 ) siRNA (error bars, +s.e.m). n = 3 biological replicates, 4-6 technical replicates; ***P = 0.0001 (two-tailed t-test). DCIS from invasive disease was an increase in stromal fibrosis support the finding that MMP13 is strongly associated with DCIS associated with collagen deposition and remodelling, and cancer- progression , thereby linking this pathway to patient outcome. Our tissue analysis identified an increased DCIS duct size associated fibroblast frequency and proliferation . This study associated with myoepithelial integrin β6 positivity. We hypothe- further identified that DCIS recurrence risk is heavily influenced by sized that physical cues in DCIS due to proliferative expansion of myoepithelial morphology and phenotype . Here, we quantified ducts may trigger these phenotypic and functional changes to higher myoepithelial integrin β6 and periductal fibronectin myoepithelial cells. Indeed, by modelling the physical tension expression in DCIS ducts associated with invasive disease in seen in myoepithelial cells during DCIS expansion, we identified a tissues, supporting the role for perturbed integrin signalling, ECM consistent and significant upregulation of integrin β6 and remodelling and an altered myoepithelial cell phenotype in fibronectin. These changes were associated with the induction driving DCIS progression. Interestingly, these observed changes in of our observed tumour-promoting phenotype- upregulation of integrin β6 and fibronectin were more frequently observed in MMP9 and MMP13. Furthermore, we identified a tension- high-grade DCIS ducts than in non-high grade DCIS ducts, which dependent requirement of TGFβ signalling that facilitates the could explain their progression to invasion more quickly, although 8 induction of this phenotype in normal myoepithelial cells. These all grades have equal potential to progress . Matched duct findings imply that physical cues in DCIS regulate myoepithelial analysis in tissues implicated a relationship between myoepithelial cell phenotype and function that critically regulates progression integrin β6 and periductal fibronectin expression. Consistently, our to invasion. In support of mechanical regulation of myoepithelial analysis of isolated primary DCIS and normal myoepithelial cells, cells, we have previously shown that increasing substrate rigidity, as well as established myoepithelial cell lines, corroborates our as seen in DCIS progression, resulted in the loss of homeostasis clinical findings, revealing significantly increased fibronectin force generation by integrin β6-positive myoepithelial cells, expression by integrin β6-positive myoepithelial cells. Together which could alter their ability to function as a barrier to invasive these data suggest that DCIS myoepithelial cells exhibit an ECM dissemination into the stroma . Indeed, studies have shown that remodelling phenotype. Using our myoepithelial cell line models, myoepithelial cells act as a dynamic barrier to tumour cell we implicate a functional relationship between myoepithelial invasion that relies upon their contractility and adhesion .We integrin β6 and fibronectin in driving breast tumour cell invasion suggest that this function is compromised in integrin β6-positive in vitro, through TGFβ-mediated protease activity. Our studies myoepithelial cells, that along with their ability to degrade the npj Breast Cancer (2022) 109 Published in partnership with the Breast Cancer Research Foundation Fold change MMP13 6-low 6-high Fold change N-1089 6-1089 Fold change N-1989 %positive ducts 6-1989 Normal Fold change DCIS DC IS / IDC N-1492 6-1492 log2 Normal Relative invasion DCIS IDC M.-K. Hayward et al. - + Normal duct v6 DCIS duct v6 DCIS duct bc Duct size ** Unstretched Stretched d g j m Unstr Str 1089 ITGB6 Unstr **** Str 4 4 2 2 FN 1 1 HSC70 Unstr Str Unstr Str h k n Unstr Str 1989 ITGB6 Unstr Str **** 4 5 **** FN 2 HSC70 0 0 Unstr Str Unstr Str f i l o 1492 ITGB6 Unstr Str Unstr Str *** **** FN 2 HSC70 0 0 Unstr Str Unstr Str BM, facilitates the breaching of tumour cells into the surrounding stimulate the expression of BM-degrading MMPs that promote stroma. Our studies highlight the importance of future studies invasion of tumour cells into the stroma. Our findings thus utilizing mechanical models to investigate mechanisms by which provide new evidence supporting a mechanism regulating DCIS the myoepithelial cell barrier is lost. In summary, we show here myoepithelial cell phenotype that facilitates invasive progression, that tension to myoepithelial cells induces integrin β6 and and identifies biomarkers that may be used to stratify women fibronectin expression, which in turn activates TGFβ signalling to with DCIS. Published in partnership with the Breast Cancer Research Foundation npj Breast Cancer (2022) 109 FN1 **** FN1 FN1 Primary-myo; N-1492 Primary-myo; N-1989 Myo cell line; N-1089 Integrin v6 FFN/DAPI 6/DAPI FN/DAPI 6/DAPI FN/DAPI 6/DAPI Fold change Fold change Fold change Duct diameter (m) 6-neg Fold change Fold change Fold change 6-pos M.-K. Hayward et al. Fig. 5 Mechanical stretching normal myoepithelial cells induces integrin β6 expression. a Cartoon depicting duct sizes for a normal duct, and a DCIS duct without and with integrin β6 expression (not to scale). b Representative images of integrin β6 by immunohistochemical staining in human breast tissue samples featuring adjacent normal ducts (40 μm), and an integrin β6-negative and β6–positive DCIS duct. Scale bar, (top) 200 μm and (bottom) 100 μm. c Box and whiskers plot showing DCIS duct sizes with (n = 656) or without (n = 713) integrin β6 expression. Box plot represents the median (central line) and interquartile range (IQR; box), and whiskers represent the maximum and minimum. **P = 0.0312 (two-tailed t-test). d–f Representative images of integrin β6 (green) and fibronectin (FN, magenta) by immunofluorescent staining in unstretched (Unstr) or stretched (Str) myoepithelial cell line; N-1089 (d) and primary normal myoepithelial cells; N-1989 (e) and N-1492 (f). Nuclei were counterstained with DAPI (blue). Scale bar, 20 μm. Images are representative of three experiments. g–i Immunoblots of integrin β6 and fibronectin (FN) in unstretched or stretched myoepithelial cells generated as in (d–f). HSC70 serves as loading control. Images are representative of three experiments. j–l Bar graphs showing qRT-PCR analysis for ITGB6 using RNA isolated from unstretched or stretched myoepithelial cell line; N-1089 (j) and primary normal myoepithelial cells; N-1989 (k) and N-1492 (l) (error bars, +s.e.m). n = 3 biological replicates, 3 technical replicates; ****P < 0.0001 (j, k) and ***P = 0.0001 (l) (two-tailed t-test). m–o, Bar graphs showing qRT-PCR analysis for FN1 using RNA isolated from unstretched or stretched myoepithelial cells generated as in (j–l) (error bars, +s.e.m). n = 3 biological replicates, 3 technical replicates; ****P < 0.0001 (two-tailed t-test). METHODS Human DCIS and normal myoepithelial cell isolation Fresh breast tissues were digested into ductal organoids by manual Gene expression microarray analysis chopping followed by digestion with 5% FBS in RPMI containing 1 mg/ml Array dataset from GEO (Gene Expression Omnibus) database with the collagenase 1 A (Roche Life Science) and 1 mg/ml hyaluronidase (Sigma), accession ID #GSE21422 was selected to assess expression profile changes overnight at 37 °C. Ductal organoids were then digested to a single-cell in an independent cohort of human breast tissues. The samples consist of suspension with 0.05%/0.02% trypsin/EDTA solution (Hyclone) containing normal healthy breast (n = 5), DCIS (n = 9) and IDC (n = 5) tissues. Gene 0.4 mg/mL DNase (Roche Life Science) for 15 minutes at 37 °C for ontology was performed using Gage (version 2.36.0) with gene lists from subsequent cell isolation. Normal and DCIS myoepithelial cells were MsigDB version 7.2. isolated using fluorescence-activated cell sorting (FACS). Cells were incubated for 45 min at 4 °C with the following human-specific primary Human breast tissue collection and processing antibodies at 1 μg/1 million cells: for normal cell suspensions EpCAM-FITC Tissue specimens were donated by women undergoing surgery for DCIS, (BD Biosciences, EBA-1) and CD10-APC, (BD Biosciences, HI10a) and for and tissues were collected as formalin-fixed and paraffin-embedded (FFPE) DCIS cell suspensions; EpCAM-PE (BD Biosciences, EBA-1), integrin β4- or as fresh tissue for FACS-mediated DCIS myoepithelial cell isolation. Alexa-Fluor 488 (Invitrogen, 422325) and integrin β6-APC (R&D Systems, Samples of DCIS without evidence of invasion (pure DCIS; n = 20) and DCIS 437211). Cells were then incubated with 4′,6-diamidino-2-phenylindole with co-existent invasion (DCIS/IDC; n = 20) were selected for immunohis- (DAPI) for 4 °C for 10 min to distinguish live/dead cells. EpCAM cells were tochemical analyses. DCIS was classified as low, intermediate, and high gated out to avoid epithelial contamination, and CD10 normal grade based on nuclear characteristics, and grouped as non-high grade myoepithelial cells or integrin β4 + /β6 + DCIS myoepithelial cells were (low and intermediate) and high grade for analysis. Clinicopathologic sorted into RPMI (Sigma) with 10% foetal bovine serum (FBS, Sigma). BD details of tissues analysed by immunohistochemistry are provided in FACSAria II cell sorters were used to conduct cell sorting using FACSDiva Supplementary Table 1. Tissue specimens were donated by women software (BD Biosciences). Data were analysed using FlowJo software (Tree undergoing reduction mammoplasty, and tissues were collected as fresh Star). Isolated myoepithelial cells were collected and used for primary cell tissue for FACS-mediated human normal myoepithelial cell isolation and culture or RNA isolation. subsequent cell culture. All human breast tissues were collected from consenting patients undergoing surgery at Barts Health NHS Trust London Primary and cell line culture between 2000 and 2015. Samples were stored and analysed with All cell lines were tested for mycoplasma contamination by PCR-based deidentified labels to protect patient data in accordance with data under method and confirmed negative for mycoplasma before experiments and the terms of the Barts Cancer Institute Breast Tissue Bank (REC no: 15/EE/ were maintained at 37 °C in a humidified 5% CO atmosphere. Isolated 0192). 2 normal myoepithelial cells were cultured in HuMEC Ready Medium (Thermo Fisher Scientific) supplemented with 50 μg/mL bovine pituitary Immunohistochemical staining extract (BPE, Invitrogen), 0.5 μg/mL hydrocortisone, 10 ng/mL EGF, 5 μg/mL Tissue sections were dewaxed in xylene, rehydrated in graded alcohols, insulin, 0.5 μg/mL fungizone (Invitrogen) and 10 μg/mL gentamicin (Sigma) endogenous peroxidases were blocked with 3% H O in methanol, antigen 2 2 and cultured on plates coated with 10 µg/cm type I collagen (Corning). retrieved in pepsin solution (Life Technologies) and blocked in 5% BSA in Myoepithelial cell lines; N-1089 and β6-1089 were grown in Nutrient PBS, followed by incubation with primary antibodies specific to SMA (Dako, Mixture Ham’s F-12 (Sigma) supplemented with 10% FBS, 1 μg/mL 1A4, 1:500), integrin β6 (Calbiochem, 442.5C4, 1:800), fibronectin (Sigma, hydrocortisone (Sigma), 10 ng/mL Epidermal Growth Factor (EGF, Sigma), IST-4, 1:400), MMP13 (abcam, VIIIA2, 1:100) or p63 (abcam, 4A4, 1:50) and 10 μg/mL insulin (Sigma). Breast cancer cell lines; MDA-MB-231 and overnight at 4 °C. Sections were washed with PBS prior to incubation with MCF-7 were obtained from American Type Culture Collection (ATCC), anti-mouse biotinylated F(ab’)2, developed using ABC reagent and DAB verified with STR profiling (LGC Standards, Teddington, UK, tracking (Vector Laboratories), counterstained with haematoxylin, dehydrated in number 710081047), and grown in DMEM (Sigma) supplemented with graded alcohols and mounted with DPX. 10% FBS. Immunohistochemical analysis DNA and siRNA transfection Stained sections were imaged using a 3DHISTECH Panoramic digital slide Cells were reverse transfected with 10 μg integrin β6 pcDNA3.1 neo, a gift scanner and analysed using QuPath (version 0.2.3) open-source software . from Dean Sheppard (Addgene, plasmid 13580), or pcDNA3.1 empty Immunohistochemical analysis was performed on a duct-by-duct basis. vector (Invitrogen) using the jetPRIME transfection reagent (PolyPlus). Cells Ducts were numbered and identified as either; normal, benign or DCIS were reverse transfected with 9 nM fibronectin, MMP13 or non-targeting within each case, by an expert breast pathologist. For expression analysis; control (CTL) siRNA (Dharmacon) using interferin transfection reagent each duct was then scored as negative or positive for integrin β6, (Polyplus). Functional assays were carried out 48 hr post-transfection. fibronectin and MMP13 in serial sections. For samples stained with fibronectin, periductal staining was measured, which was defined as a TGFβ stimulation 50 µm region bordering DCIS lesions. For DCIS duct size analysis, only cross-sectional ducts were included. For cell and nuclear size analysis; Cells were cultured in serum-free media for 24 h prior to stimulation. Cells myoepithelial cells positive for SMA or p63 were segmented by semi- were then washed in PBS to remove residual media and were then automated detection, and cell and nuclear morphology features were stimulated with 5 ng/mL recombinant human active TGFβ1 (PeproTech) in extracted. serum-free media for 5, 15 and 30 min. npj Breast Cancer (2022) 109 Published in partnership with the Breast Cancer Research Foundation M.-K. Hayward et al. ab Invasion Protease array ** 2.0 Unstr CM Unstr CM Str CM Str CM ** 1.5 1.0 0.5 0.0 231 MCF-7 Myo cell lines Isolated primary myos cd e 1089 1989 1492 Unstr Str Unstr Str Unstr Str MMP9 MMP9 MMP9 f gh MMP9 MMP9 MMP9 *** 3 3 3 ** 2 2 2 1 1 1 0 0 0 Unstr Str Unstr Str Unstr Str ij k MMP13 MMP13 MMP13 *** 5 ** 3 3 4 *** 2 2 1 1 0 0 0 Unstr Str Unstr Str Unstr Str Fig. 6 Mechanical stretching normal myoepithelial cells activates a tumour promoting phenotype. a Invasion assay for MDA-MB-231 (231) and MCF-7 cells using conditioned media (CM) isolated from unstretched (Unstr) or stretched (Str) normal myoepithelial cell line; N-1089 (error bars, +s.e.m). n = 3 biological replicates, 5 technical replicates; ****P < 0.0001 (two-tailed t-test). b Array analysis for proteases using conditioned media isolated from unstretched or stretched myoepithelial cell line; N-1089 (error bars, +s.e.m). n= 1 biological replicate, 2 technical replicates. c–e Gelatin zymography for MMP9 activity using conditioned media isolated from unstretched or stretched myoepithelial cell line; N-1089 (c) and primary normal myoepithelial cells; N-1989 (d) and N-1492 (e). Images are representative of three experiments. f–h Bar graphs showing qRT-PCR analysis for MMP9 using RNA isolated from unstretched or stretched myoepithelial cell line; N-1089 (f) and primary normal myoepithelial cells; N-1989 (g) and N-1492 (h) (error bars, +s.e.m). n = 3 biological replicates, 3 technical replicates; **P = 0.0035 (f), ***P = 0.0005 (g) and *P = 0.0232 (h) (two-tailed t-test). i–k Bar graphs showing qRT-PCR analysis for MMP13 using RNA isolated from unstretched or stretched myoepithelial cell line; N-1089 (i) and primary normal myoepithelial cells; N-1989 (j) and N-1492 (k) (error bars, +s.e.m). n = 3 biological replicates, 3 technical replicates; **P = 0.0017 (i), ***P = 0.0007 (j) and ***P = 0.0009 (k) (two-tailed t-test). TGFβRII inhibition Transwell® migration and invasion assays Cells were incubated with 10 μg/mL TGFβRII-blocking antibody (R&D Motility assays were performed using Transwell® migration inserts (8 μm Systems) or IgG isotype control antibody (Merck Millipore, GC270) in pore size, polycarbonate membrane, Corning). For migration assays the serum-free media for 20 min at 4 °C on a rotating-wheel before plating. underside of inserts were coated with 0.5 μg/mL recombinant human Functional assays were carried out 48 h postantibody treatment. latency-associated peptide of TGFβ1 (LAP, R&D Systems) or 0.1% BSA; for invasion assays the top of each insert was coated with Matrigel (BD Biosciences) diluted 1:3 in DMEM. Migrating and invading MDA-MB-231 and MCF-7 cells to the lower chamber were counted after 8 h, and 24 or Conditioned media 48 h incubation, respectively, using a CASY counter (Schärfe System). For Conditioned media (CM) was generated by culturing cells in serum-free migration, total cell count for each sample was calculated by adding the media for 48 h. CM was concentrated 24-fold with centrifugal filters (Fisher) counts of the upper and lower chambers. Relative cell migration was then with 3 K molecular weight cut off (MWCO) at 4000 g for 45 min at 4 °C. Published in partnership with the Breast Cancer Research Foundation npj Breast Cancer (2022) 109 Fold change Fold change R elat ive invas ion Fold change Fold change R elat ive expres s ion MMP1 MMP2 MMP3 Fold change Fold change MMP7 MMP8 MMP9 MMP10 MMP12 MMP13 M.-K. Hayward et al. Stretched myo cell line a b c AB AB CTL RII 1089 ITGB6 FN1 AB AB **** CTL RII 2.0 1.5 **** 1.5 1.0 1.0 FN 0.5 0.5 HSC70 0.0 0.0 AB AB AB AB CTL RII CTL RII ef g Invasion MMP9 MMP13 *** AB 2.0 1.5 CTL CM 1.5 ** *** AB RII CM ** 1.5 1.0 1.0 1.0 0.5 0.5 0.5 0.0 0.0 0.0 AB AB AB AB 231 MCF-7 CTL RII CTL RII Fig. 7 TGFβ signalling regulates a mechanically induced tumour promoting phenotype in normal myoepithelial cells. a Representative images of integrin β6 (green) and fibronectin (FN, magenta) by immunofluorescent staining in stretched normal myoepithelial cell line; N-1089 AB AB treated with a control (CTL )or TGFβRII (RII ) blocking antibody. Nuclei were counterstained with DAPI (blue). Scale bar, 20 μm. b Immunoblots of integrin β6 and fibronectin in stretched normal myoepithelial cell lines generated as in (a). HSC70 serves as loading control. Images are representative of three experiments. c, d Bar graphs showing qRT-PCR analysis for ITGB6 (c) and FN1 (d) using RNA isolated from stretched normal myoepithelial cell lines generated as in (a) (error bars, +s.e.m). n = 3 biological replicates, repeated 3 times; ****P < 0.0001 (two-tailed t-test). e Invasion assay for MDA-MB-231 (231) and MCF-7 cells using conditioned media (CM) isolated from stretched normal myoepithelial cell lines generated as in (a) (error bars, +s.e.m). n = 3 biological replicates, repeated 4 times; MDA-MB-231 ***P = 0.0001 and MCF-7 ****P < 0.0001 (two-tailed t-test). f, g Bar graphs showing qRT-PCR analysis for MMP9 (f) and MMP13 (g) using RNA isolated from stretched normal myoepithelial cell lines generated as in (a) (error bars, +s.e.m). n = 3 biological replicates, repeated 3 times; ***P = 0.0001 (f) and **P = 0.0041 (two-tailed t-test). calculated by using the lower chamber count versus total cell count. For Immunofluorescence invasion, for each sample only the lower chamber was counted. Relative Cells were fixed in 4% formaldehyde for 10 min and permeabilised with cell invasion was then calculated by normalising to the control. 0.1% Triton X-100 in PBS for 5 min. Cells were then blocked with 5% BSA in PBS for 10 min prior to incubation with primary antibodies specific for integrin β6 (Merck, 10D5, 1:100) and fibronectin (FN, Sigma, IST-4, 1:100) Adhesion Assay overnight at 4 °C. Cells were then washed with PBS and incubated with 96-well plates were coated with 0.5 μg/mL recombinant human LAP (R&D goat anti-mouse Alexa-Fluor 488 secondary antibody (Invitrogen, 1:200), Systems) or 0.1% BSA and incubated for 1 h at 37 °C. Cells were then followed by additional washing and then mounted and counterstained seeded and allowed to adhere for 1 hr at 37 °C before fixing in methanol with ProLong Gold Antifade reagent containing 4′,6-diamidino-2-pheny- and staining with 0.1% crystal violet. Stained cells were dissolved with 30% lindole (DAPI, Invitrogen). Images were viewed on a Zeiss LSM 710 Meta acetic acid and absorbance read at 550 nm. Background binding to BSA microscope. was subtracted from LAP, and relative adhesion was calculated by normalising to the control. Quantitative reverse transcriptase-PCR (qRT-PCR) analysis Total RNA was isolated from cells using the Quick-RNA MiniPrep Kit (Zymo Immunoblotting Research). cDNA was synthesized using Moloney-Murine Leukemia Virus Total cellular protein was isolated using RIPA buffer (50 mM Tris-HCl pH 7.4, (M-MLV) reverse transcriptase with random nucleotide primers (Sigma). 150 mM NaCl, 1% IPEGAL CA-630, 0.1% Na-DOC, 1 mM EDTA) supplemen- Quantitative reverse transcriptase-PCR (qRT-PCR) was performed on cDNA ted with protease and phosphatase inhibitor cocktails (EMD Millipore). using SYBR Green Master Mix (Thermo Fisher Scientific) on a StepOnePlus Lysates containing equal amounts of protein (30 μg) were electrophoresed Real-Time PCR System (Applied Biosystems). Gene expression was in 6-10% SDS-PAGE gel, electroblotted to nitrocellulose membrane quantified using the following primers: 18 S forward: CACGGGAAACCTC (Amersham). Membranes were then blocked with 0.1% Tween-20 in TBS ACCCGGC; 18 S reverse: AACGGCCATGCACCACCACC; ITGB6 forward: (TBS-T) supplemented with 5% milk for 1 hr prior to incubation with GAAGGAATGATCACGTACAAG; ITGB6 reverse: AGCAGGGAGTCTTCACAGGT; primary antibodies specific for integrin β6 (Santa Cruz, C-19, 1:500), FN1 forward: AACAAACACTAATGTTAATTGCCC; FN1 reverse: TCGGGAA fibronectin (FN, Sigma, IST-4, 1:500), phospho-SMAD2 S465/467 (CST, TCTTCTCTGTCAGC; TGFB1 forward: GGAAATTGAGGGCTTTCGCC; TGFB1 138D4, 1:500), SMAD2 (CST, 86F7, 1:500) and HSC70 (Santa Cruz, B-6, reverse: CCGGTAGTGAACCCGTTGAT; MMP9 forward: GAACCAATCTCACC- 1:1000) overnight at 4 °C. Membranes were then washed with TBS-T and GACAGG; MMP9 reverse: GCCACCCGAGTGTAACCATA; MMP13 forward: incubated with appropriate species-specific HRP-conjugated secondary TCTACACCTACACCGGCAAA; MMP13 reverse: GGTTGGGGTCTTCATCTCCT. antibodies (Dako, 1:1000). Signals were visualised using Enhanced Chemiluminescence (ECL) reagents (Amersham) and exposure to film. Fold changes in mRNA expression were calculated by the ΔΔCt method Films were developed in a Konica Film Processor (SRX-101A). All blots were using 18 S as an endogenous control. Results are expressed as fold change processed in parallel and derive from the same experiment. by normalizing to the controls. npj Breast Cancer (2022) 109 Published in partnership with the Breast Cancer Research Foundation Relative invasion N-1089 FN/DAPI 6/DAPI Fold change Fold change Fold change Fold change M.-K. Hayward et al. Human Proteome Protease Array 12. Chin, K. et al. In situ analyses of genome instability in breast cancer. Nat. Genet 36, 984–988 (2004). Human proteome protease arrays (R&D Systems) were processed 13. Ma, X. J. et al. Gene expression profiles of human breast cancer progression. Proc. according to the manufacturer’s instructions using concentrated CM Natl Acad. Sci. USA 100, 5974–5979 (2003). (250 μg). Signals were visualised using ECL reagents and exposure to film. 14. Porter, D. et al. Molecular markers in ductal carcinoma in situ of the breast. Mol. Films were developed in a Konica Film Processor. Cancer Res 1, 362–375 (2003). 15. Lee, S. et al. Differentially expressed genes regulating the progression of Zymography ductal carcinoma in situ to invasive breast cancer. Cancer Res 72,4574–4586 Concentrated CM (100 μg) was resolved on a 10% Tris-Glycine gel (2012). supplemented with 0.1% gelatin (Invitrogen). Gels were renatured in a 16. Moelans, C. B., de Weger, R. A., Monsuur, H. N., Maes, A. H. & van Diest, P. J. buffer (2.5% (v/v) Triton X-100), followed by incubation in a developing Molecular differences between ductal carcinoma in situ and adjacent invasive buffer (5 mM Tris-Base, 4 mM HCl, 20 mM NaCl, 0.5 mM CaCl ) overnight at breast carcinoma: a multiplex ligation-dependent probe amplification study. 37 °C and visualised with Coomassie R-250 (Thermo Fisher Scientific). Anal. Cell Pathol. (Amst.) 33, 165–173 (2010). Images of stained gels were captured under illumination using the UVP 17. Pareja, F. et al. Whole-Exome Sequencing Analysis of the Progression from Non- Imagestore 5000 (Ultra-Violet Products). Low-Grade Ductal Carcinoma. Clin. Cancer Res 26, 3682–3693 (2020). 18. Adriance, M. C., Inman, J. L., Petersen, O. W. & Bissell, M. J. Myoepithelial cells: good fences make good neighbors. Breast Cancer Res 7, 190–197 (2005). Mechanical Stretch 19. Deugnier, M. A., Teulière, J., Faraldo, M. M., Thiery, J. P. & Glukhova, M. A. The Cells were seeded in flexible-bottomed BioFlex culture plates coated with importance of being a myoepithelial cell. Breast Cancer Res 4, 224–230 (2002). type IV collagen (Dunn Lab), and grown for 72 h. Immediately prior to 20. Sopel, M. The myoepithelial cell: its role in normal mammary glands and breast stretching, cells were removed from the periphery of the well, and the cancer. Folia Morphol. (Warsz.) 69,1–14 (2010). media was replaced. Cells were then exposed to a static stretch with 10% 21. Bissell, M. J., Kenny, P. A. & Radisky, D. C. Microenvironmental regulators of tissue elongation of the flexible surface using a computerised vacuum-operated structure and function also regulate tumor induction and progression: the role of instrument (Flexcell strain unit FX-5000 Tension Plus; Flexcell International) extracellular matrix and its degrading enzymes. Cold Spring Harb. Symp. Quant. maintained in a cell culture incubator, for 48 h. Unstretched controls cells Biol. 70, 343–356 (2005). were plated on BioFlex culture plates for an equivalent time but were not 22. Sternlicht, M. D., Kedeshian, P., Shao, Z. M., Safarians, S. & Barsky, S. H. The human subjected to stretch. myoepithelial cell is a natural tumor suppressor. Clin. Cancer Res 3, 1949–1958 (1997). 23. Lakhani, S. R. & O’Hare, M. J. The mammary myoepithelial cell–Cinderella or ugly Statistical Analysis sister? Breast Cancer Res 3,1–4 (2001). GraphPad Prism (version 9.1.2) was used to perform all statistical analyses 24. Nguyen, M. et al. The human myoepithelial cell displays a multifaceted anti- and statistical significance was determined using the appropriate tests as angiogenic phenotype. Oncogene 19, 3449–3459 (2000). noted in the corresponding figure legends. 25. Jones, J. L., Shaw, J. A., Pringle, J. H. & Walker, R. A. Primary breast myoepithelial cells exert an invasion-suppressor effect on breast cancer cells via paracrine down-regulation of MMP expression in fibroblasts and tumour cells. J. Pathol. DATA AVAILABILITY 201, 562–572 (2003). The array dataset analysed to assess expression profiles changes in human breast 26. Barsky, S. H. & Karlin, N. J. Myoepithelial cells: autocrine and paracrine suppressors tissues is available under accession number GSE21422 . All further datasets of breast cancer progression. J. Mammary Gland Biol. Neoplasia 10, 249–260 generated and analysed in this study are available from the corresponding author (2005). upon reasonable request. 27. Hu, M. et al. Regulation of in situ to invasive breast carcinoma transition. Cancer Cell 13, 394–406 (2008). 28. Sternlicht, M. D. & Barsky, S. H. The myoepithelial defense: a host defense against Received: 20 October 2021; Accepted: 13 July 2022; cancer. Med Hypotheses 48,37–46 (1997). 29. Zhang, X. et al. Atypical E-cadherin expression in cell clusters overlying focally disrupted mammary myoepithelial cell layers: implications for tumor cell motility and invasion. Pathol. Res Pr. 205, 375–385 (2009). 30. Man, Y. G. et al. Cell clusters overlying focally disrupted mammary myoepithelial REFERENCES cell layers and adjacent cells within the same duct display different immuno- 1. Sgroi, D. C. Preinvasive breast cancer. Annu Rev. Pathol. 5, 193–221 (2010). histochemical and genetic features: implications for tumor progression and 2. Swallow, C. J., Van Zee, K. J., Sacchini, V. & Borgen, P. I. Ductal carcinoma in situ of invasion. Breast Cancer Res 5, R231–R241 (2003). the breast: progress and controversy. Curr. Probl. Surg. 33, 553–600 (1996). 31. Man, Y. G. & Sang, Q. X. The significance of focal myoepithelial cell layer dis- 3. van Steenbergen, L. N. et al. Screening caused rising incidence rates of ductal ruptions in human breast tumor invasion: a paradigm shift from the “protease- carcinoma in situ of the breast. Breast Cancer Res Treat. 115, 181–183 (2009). centered” hypothesis. Exp. Cell Res 301, 103–118 (2004). 4. Kerlikowske, K. Epidemiology of ductal carcinoma in situ. J. Natl Cancer Inst. 32. Allinen, M. et al. Molecular characterization of the tumor microenvironment in Monogr. 2010, 139–141 (2010). breast cancer. Cancer Cell 6,17–32 (2004). 5. van Seijen, M. et al. Ductal carcinoma in situ: to treat or not to treat, that is the 33. Russell, T. D. et al. Myoepithelial cell differentiation markers in ductal carcinoma question. Br. J. Cancer 121, 285–292 (2019). in situ progression. Am. J. Pathol. 185, 3076–3089 (2015). 6. Collins, L. C. et al. Outcome of patients with ductal carcinoma in situ untreated 34. Duivenvoorden, H. M., Spurling, A., O’Toole, S. A. & Parker, B. S. Discriminating the after diagnostic biopsy: results from the Nurses’ Health Study. Cancer 103, earliest stages of mammary carcinoma using myoepithelial and proliferative 1778–1784 (2005). markers. PLoS One 13, e0201370 (2018). 7. Erbas, B., Provenzano, E., Armes, J. & Gertig, D. The natural history of ductal 35. Ding, L. et al. Perturbed myoepithelial cell differentiation in BRCA mutation carcinoma in situ of the breast: a review. Breast Cancer Res Treat. 97, 135–144 carriers and in ductal carcinoma in situ. Nat. Commun. 10, 4182 (2019). (2006). 36. Allen, M. D. et al. Altered microenvironment promotes progression of preinvasive 8. Sanders, M. E., Schuyler, P. A., Dupont, W. D. & Page, D. L. The natural history of breast cancer: myoepithelial expression of αvβ6 integrin in DCIS identifies high- low-grade ductal carcinoma in situ of the breast in women treated by biopsy only risk patients and predicts recurrence. Clin. Cancer Res 20, 344–357 (2014). revealed over 30 years of long-term follow-up. Cancer 103, 2481–2484 (2005). 37. Lo, P. K. et al. Tumor-associated myoepithelial cells promote the invasive pro- 9. Allegra, C. J. et al. NIH state-of-the-science conference statement: diagnosis and gression of ductal carcinoma. J. Biol. Chem. 292, 11466–11484 (2017). management of ductal carcinoma in situ (DCIS). NIH Consens. State Sci. Statements 38. Wang, L. et al. TGF-β1 stimulates epithelial-mesenchymal transition and cancer- 26,1–27 (2009). associated myoepithelial cell during the progression from in situ to invasive 10. Yao, J. et al. Combined cDNA array comparative genomic hybridization and serial breast cancer. Cancer Cell Int 19, 343 (2019). analysis of gene expression analysis of breast tumor progression. Cancer Res 66, 39. Pickup, M. W. et al. Stromally derived lysyl oxidase promotes metastasis of 4065–4078 (2006). transforming growth factor-β-deficient mouse mammary carcinomas. Cancer Res 11. Castro, N. P. et al. Evidence that molecular changes in cells occur before mor- 73, 5336–5346 (2013). phological alterations during the progression of breast ductal carcinoma. Breast 40. Acerbi, I. et al. Human breast cancer invasion and aggression correlates with ECM Cancer Res 10, R87 (2008). stiffening and immune cell infiltration. Integr. Biol. (Camb.) 7, 1120–1134 (2015). Published in partnership with the Breast Cancer Research Foundation npj Breast Cancer (2022) 109 M.-K. Hayward et al. 41. Maller, O. et al. Tumour-associated macrophages drive stromal cell-dependent analysis and immunoblotting. M.M.K provided access to Flexcell instrument. C.L.T collagen crosslinking and stiffening to promote breast cancer aggression. Nat. managed the Flexcell instrument. J.F.M. provided intellectual input on study design Mater. 20, 548–559 (2021). and analysis. J.L.J. and M.K.H. wrote the manuscript with editorial input from all 42. Kretschmer, C. et al. Identification of early molecular markers for breast cancer. authors. Mol. Cancer 10, 15 (2011). 43. Klingberg, F. et al. Prestress in the extracellular matrix sensitizes latent TGF-β1 for activation. J. Cell Biol. 207, 283–297 (2014). CONFLICT OF INTEREST 44. Nielsen, B. S. et al. Collagenase-3 expression in breast myofibroblasts as a The authors declare no competing interests. molecular marker of transition of ductal carcinoma in situ lesions to invasive ductal carcinomas. Cancer Res 61, 7091–7100 (2001). 45. Risom, T. et al. Transition to invasive breast cancer is associated with progressive ADDITIONAL INFORMATION changes in the structure and composition of tumor stroma. Cell 185, Supplementary information The online version contains supplementary material 299–310.e218 (2022). available at https://doi.org/10.1038/s41523-022-00464-4. 46. Elosegui-Artola, A. et al. Rigidity sensing and adaptation through regulation of integrin types. Nat. Mater. 13, 631–637 (2014). Correspondence and requests for materials should be addressed to J. Louise Jones. 47. Sirka, O. K., Shamir, E. R. & Ewald, A. J. Myoepithelial cells are a dynamic barrier to epithelial dissemination. J. Cell Biol. 217, 3368–3381 (2018). Reprints and permission information is available at http://www.nature.com/ 48. Bankhead, P. et al. QuPath: Open source software for digital pathology image reprints analysis. Sci. Rep. 7, 16878 (2017). Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. ACKNOWLEDGEMENTS The authors wish to acknowledge the role of the Breast Cancer Now Tissue Bank in collecting and making available the samples used in the generation of this publication, and all the patients who donated. We also thank the BCI pathology core Open Access This article is licensed under a Creative Commons for tissue processing and the BCI flow cytometry core for machine assistance. This Attribution 4.0 International License, which permits use, sharing, work was supported by the Pathological Society of Great Britain and Ireland PhD adaptation, distribution and reproduction in any medium or format, as long as you give studentship, the Breast Cancer Now Tissue Bank Cell Culture Programme, a Cancer appropriate credit to the original author(s) and the source, provide a link to the Creative Research UK (CRUK) Core Service Grant (C16420/A18066) and a CRUK Centre Grant to Commons license, and indicate if changes were made. The images or other third party BCI (C355/A25137). material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory AUTHOR CONTRIBUTIONS regulation or exceeds the permitted use, you will need to obtain permission directly J.L.J. and M.K.H. conceived, designed and directed the study. M.K.H. performed gene from the copyright holder. To view a copy of this license, visit http:// set enrichment analyses. M.K.H. performed H&E and IHC staining of human breast creativecommons.org/licenses/by/4.0/. tissues. J.L.J. assessed human tissue pathology. M.K.H, J.J.G and I.G. performed FACS isolation of myoepithelial cells. M.D.A. designed and developed the myoepithelial cell © Crown 2022 lines and provided technical support. M.K.H. performed cell culture assays, qRT-PCR npj Breast Cancer (2022) 109 Published in partnership with the Breast Cancer Research Foundation
npj Breast Cancer – Springer Journals
Published: Sep 20, 2022
You can share this free article with as many people as you like with the url below! We hope you enjoy this feature!
Read and print from thousands of top scholarly journals.
Already have an account? Log in
Bookmark this article. You can see your Bookmarks on your DeepDyve Library.
To save an article, log in first, or sign up for a DeepDyve account if you don’t already have one.
Copy and paste the desired citation format or use the link below to download a file formatted for EndNote
Access the full text.
Sign up today, get DeepDyve free for 14 days.
All DeepDyve websites use cookies to improve your online experience. They were placed on your computer when you launched this website. You can change your cookie settings through your browser.