Limits...
ELF5 Drives Lung Metastasis in Luminal Breast Cancer through Recruitment of Gr1+ CD11b+ Myeloid-Derived Suppressor Cells.

Gallego-Ortega D, Ledger A, Roden DL, Law AM, Magenau A, Kikhtyak Z, Cho C, Allerdice SL, Lee HJ, Valdes-Mora F, Herrmann D, Salomon R, Young AI, Lee BY, Sergio CM, Kaplan W, Piggin C, Conway JR, Rabinovich B, Millar EK, Oakes SR, Chtanova T, Swarbrick A, Naylor MJ, O'Toole S, Green AR, Timpson P, Gee JM, Ellis IO, Clark SJ, Ormandy CJ - PLoS Biol. (2015)

Bottom Line: Myeloid-derived suppressor cells, a group of immature neutrophils recently identified as mediators of vasculogenesis and metastasis, were recruited to the tumor in response to ELF5.Depletion of these cells using specific Ly6G antibodies prevented ELF5 from driving vasculogenesis and metastasis.Thus, in the MMTV-PyMT mouse mammary model, increased ELF5 levels drive metastasis by co-opting the innate immune system.

View Article: PubMed Central - PubMed

Affiliation: Garvan Institute of Medical Research and The Kinghorn Cancer Centre, Darlinghurst, New South Wales, Australia.

ABSTRACT
During pregnancy, the ETS transcription factor ELF5 establishes the milk-secreting alveolar cell lineage by driving a cell fate decision of the mammary luminal progenitor cell. In breast cancer, ELF5 is a key transcriptional determinant of tumor subtype and has been implicated in the development of insensitivity to anti-estrogen therapy. In the mouse mammary tumor virus-Polyoma Middle T (MMTV-PyMT) model of luminal breast cancer, induction of ELF5 levels increased leukocyte infiltration, angiogenesis, and blood vessel permeability in primary tumors and greatly increased the size and number of lung metastasis. Myeloid-derived suppressor cells, a group of immature neutrophils recently identified as mediators of vasculogenesis and metastasis, were recruited to the tumor in response to ELF5. Depletion of these cells using specific Ly6G antibodies prevented ELF5 from driving vasculogenesis and metastasis. Expression signatures in luminal A breast cancers indicated that increased myeloid cell invasion and inflammation were correlated with ELF5 expression, and increased ELF5 immunohistochemical staining predicted much shorter metastasis-free and overall survival of luminal A patients, defining a group who experienced unexpectedly early disease progression. Thus, in the MMTV-PyMT mouse mammary model, increased ELF5 levels drive metastasis by co-opting the innate immune system. As ELF5 has been previously implicated in the development of antiestrogen resistance, this finding implicates ELF5 as a defining factor in the acquisition of the key aspects of the lethal phenotype in luminal A breast cancer.

Show MeSH

Related in: MedlinePlus

Effects of forced Elf5 expression on tumor growth and cell autonomous pro-tumorigenic traits.Panel A, PyMT mammary tumors showing heterogeneous expression of ELF5 visualized by EGFP expression. Panel B, percent primary tumor burden for each experimental group showing equivalence. Panel C, survival analysis of time to ethical endpoint (10% tumor burden), time to tumor detection, and time from detection to the ethical endpoint, in animals of the indicated genotypes carrying tumors that developed in situ. Panel D, survival analysis, as above, in animals carrying tumors that developed from an intraductal transplant of EGFP+ tumor cells derived from animals treated long term with DOX. Log-rank p-values are shown for +/- DOX comparison. N.S.; not significant. Panel E, representative image of cell proliferation measured by BrdU incorporation (red cells) in EGFP high (bright green) compared to EGFP low/no areas (dark green) of primary tumors, quantified by counting cells in random fields (bar chart). Panels F and G, Boyden chamber assays of motility and invasion through matrigel of EGFP high cells (PyMT ELF5) compared to wild-type PyMT cells separated by flow sorting. Panel H, lung colonies per field that developed from EGFP high or wild-type cells injected through the tail vein. Panel I, changes in gene expression of a set of genes involved in epithelial and mesenchymal characteristics. Significant increase in expression in red (UP), decrease in blue (DN), and nonsignificant changes in grey. Labels are EGFP- (G-), EGFP+ (G+) and WT (W). Raw data for panels E, F, G, and H can be found in S1 Data.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4696735&req=5

pbio.1002330.g001: Effects of forced Elf5 expression on tumor growth and cell autonomous pro-tumorigenic traits.Panel A, PyMT mammary tumors showing heterogeneous expression of ELF5 visualized by EGFP expression. Panel B, percent primary tumor burden for each experimental group showing equivalence. Panel C, survival analysis of time to ethical endpoint (10% tumor burden), time to tumor detection, and time from detection to the ethical endpoint, in animals of the indicated genotypes carrying tumors that developed in situ. Panel D, survival analysis, as above, in animals carrying tumors that developed from an intraductal transplant of EGFP+ tumor cells derived from animals treated long term with DOX. Log-rank p-values are shown for +/- DOX comparison. N.S.; not significant. Panel E, representative image of cell proliferation measured by BrdU incorporation (red cells) in EGFP high (bright green) compared to EGFP low/no areas (dark green) of primary tumors, quantified by counting cells in random fields (bar chart). Panels F and G, Boyden chamber assays of motility and invasion through matrigel of EGFP high cells (PyMT ELF5) compared to wild-type PyMT cells separated by flow sorting. Panel H, lung colonies per field that developed from EGFP high or wild-type cells injected through the tail vein. Panel I, changes in gene expression of a set of genes involved in epithelial and mesenchymal characteristics. Significant increase in expression in red (UP), decrease in blue (DN), and nonsignificant changes in grey. Labels are EGFP- (G-), EGFP+ (G+) and WT (W). Raw data for panels E, F, G, and H can be found in S1 Data.

Mentions: Induction of ELF5 was measured in whole tumors by imaging EGFP fluorescence. A heterogeneous pattern of expression was observed (Fig 1A), which may have resulted from a chimeric expression pattern of the rtTA transgene, a feature of older MTB mice [33]. We used Kaplan-Meier survival plots to analyze primary tumor growth. Only mice that showed a tumor burden of ~10% (7%–13%) of body weight at autopsy were included in the analysis (Fig 1B). Overall survival at ~10% tumor burden showed no significant difference (Fig 1C LHS), however, forced expression of Elf5 produced tumors that were detected earlier (Fig 1C middle), but which took longer to then reach the ethical endpoint (Fig 1C right-hand side [RHS]). To overcome the effects of heterogeneous ELF5 induction (Fig 1A), we performed intraductal allografts of Fluorescence-Activated Cell Sorting (FACS)-sorted (Lin- and CD24+) tumor cells that were either EGFP (ELF5) positive or negative. Purified cells were injected into the mammary ducts of FVB/N host animals pretreated with DOX and maintained on DOX. EGFP+ transplants resulted in longer overall survival, longer time to tumor detection and longer time to the ethical endpoint, than transplants originated from EGFP- cells (Fig 1D). To demonstrate that EGFP/ELF5 was not only expressed in a particular subset within the mammary epithelium, we performed a similar experiment including allografts made from cells that were sorted (Lin- and CD24+) from excised tumors not carrying the ELF5 transgene (PyMT/wild type [WT]) or cells that were purified from tumors (PyMT/ELF5) made fluorescent by a short 7 d pulse of DOX administration, to allow flow capture of EGFP+ cells as before, but then injected into the mammary ducts of hosts either pretreated and maintained on DOX, or not ever treated with DOX (S2A Fig). As before, EGFP+ allografts maintained on DOX produced slower growing tumors. The two control groups (WT and EGFP+ with no DOX after transplant) produced tumors that expanded at indistinguishable rates.


ELF5 Drives Lung Metastasis in Luminal Breast Cancer through Recruitment of Gr1+ CD11b+ Myeloid-Derived Suppressor Cells.

Gallego-Ortega D, Ledger A, Roden DL, Law AM, Magenau A, Kikhtyak Z, Cho C, Allerdice SL, Lee HJ, Valdes-Mora F, Herrmann D, Salomon R, Young AI, Lee BY, Sergio CM, Kaplan W, Piggin C, Conway JR, Rabinovich B, Millar EK, Oakes SR, Chtanova T, Swarbrick A, Naylor MJ, O'Toole S, Green AR, Timpson P, Gee JM, Ellis IO, Clark SJ, Ormandy CJ - PLoS Biol. (2015)

Effects of forced Elf5 expression on tumor growth and cell autonomous pro-tumorigenic traits.Panel A, PyMT mammary tumors showing heterogeneous expression of ELF5 visualized by EGFP expression. Panel B, percent primary tumor burden for each experimental group showing equivalence. Panel C, survival analysis of time to ethical endpoint (10% tumor burden), time to tumor detection, and time from detection to the ethical endpoint, in animals of the indicated genotypes carrying tumors that developed in situ. Panel D, survival analysis, as above, in animals carrying tumors that developed from an intraductal transplant of EGFP+ tumor cells derived from animals treated long term with DOX. Log-rank p-values are shown for +/- DOX comparison. N.S.; not significant. Panel E, representative image of cell proliferation measured by BrdU incorporation (red cells) in EGFP high (bright green) compared to EGFP low/no areas (dark green) of primary tumors, quantified by counting cells in random fields (bar chart). Panels F and G, Boyden chamber assays of motility and invasion through matrigel of EGFP high cells (PyMT ELF5) compared to wild-type PyMT cells separated by flow sorting. Panel H, lung colonies per field that developed from EGFP high or wild-type cells injected through the tail vein. Panel I, changes in gene expression of a set of genes involved in epithelial and mesenchymal characteristics. Significant increase in expression in red (UP), decrease in blue (DN), and nonsignificant changes in grey. Labels are EGFP- (G-), EGFP+ (G+) and WT (W). Raw data for panels E, F, G, and H can be found in S1 Data.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4696735&req=5

pbio.1002330.g001: Effects of forced Elf5 expression on tumor growth and cell autonomous pro-tumorigenic traits.Panel A, PyMT mammary tumors showing heterogeneous expression of ELF5 visualized by EGFP expression. Panel B, percent primary tumor burden for each experimental group showing equivalence. Panel C, survival analysis of time to ethical endpoint (10% tumor burden), time to tumor detection, and time from detection to the ethical endpoint, in animals of the indicated genotypes carrying tumors that developed in situ. Panel D, survival analysis, as above, in animals carrying tumors that developed from an intraductal transplant of EGFP+ tumor cells derived from animals treated long term with DOX. Log-rank p-values are shown for +/- DOX comparison. N.S.; not significant. Panel E, representative image of cell proliferation measured by BrdU incorporation (red cells) in EGFP high (bright green) compared to EGFP low/no areas (dark green) of primary tumors, quantified by counting cells in random fields (bar chart). Panels F and G, Boyden chamber assays of motility and invasion through matrigel of EGFP high cells (PyMT ELF5) compared to wild-type PyMT cells separated by flow sorting. Panel H, lung colonies per field that developed from EGFP high or wild-type cells injected through the tail vein. Panel I, changes in gene expression of a set of genes involved in epithelial and mesenchymal characteristics. Significant increase in expression in red (UP), decrease in blue (DN), and nonsignificant changes in grey. Labels are EGFP- (G-), EGFP+ (G+) and WT (W). Raw data for panels E, F, G, and H can be found in S1 Data.
Mentions: Induction of ELF5 was measured in whole tumors by imaging EGFP fluorescence. A heterogeneous pattern of expression was observed (Fig 1A), which may have resulted from a chimeric expression pattern of the rtTA transgene, a feature of older MTB mice [33]. We used Kaplan-Meier survival plots to analyze primary tumor growth. Only mice that showed a tumor burden of ~10% (7%–13%) of body weight at autopsy were included in the analysis (Fig 1B). Overall survival at ~10% tumor burden showed no significant difference (Fig 1C LHS), however, forced expression of Elf5 produced tumors that were detected earlier (Fig 1C middle), but which took longer to then reach the ethical endpoint (Fig 1C right-hand side [RHS]). To overcome the effects of heterogeneous ELF5 induction (Fig 1A), we performed intraductal allografts of Fluorescence-Activated Cell Sorting (FACS)-sorted (Lin- and CD24+) tumor cells that were either EGFP (ELF5) positive or negative. Purified cells were injected into the mammary ducts of FVB/N host animals pretreated with DOX and maintained on DOX. EGFP+ transplants resulted in longer overall survival, longer time to tumor detection and longer time to the ethical endpoint, than transplants originated from EGFP- cells (Fig 1D). To demonstrate that EGFP/ELF5 was not only expressed in a particular subset within the mammary epithelium, we performed a similar experiment including allografts made from cells that were sorted (Lin- and CD24+) from excised tumors not carrying the ELF5 transgene (PyMT/wild type [WT]) or cells that were purified from tumors (PyMT/ELF5) made fluorescent by a short 7 d pulse of DOX administration, to allow flow capture of EGFP+ cells as before, but then injected into the mammary ducts of hosts either pretreated and maintained on DOX, or not ever treated with DOX (S2A Fig). As before, EGFP+ allografts maintained on DOX produced slower growing tumors. The two control groups (WT and EGFP+ with no DOX after transplant) produced tumors that expanded at indistinguishable rates.

Bottom Line: Myeloid-derived suppressor cells, a group of immature neutrophils recently identified as mediators of vasculogenesis and metastasis, were recruited to the tumor in response to ELF5.Depletion of these cells using specific Ly6G antibodies prevented ELF5 from driving vasculogenesis and metastasis.Thus, in the MMTV-PyMT mouse mammary model, increased ELF5 levels drive metastasis by co-opting the innate immune system.

View Article: PubMed Central - PubMed

Affiliation: Garvan Institute of Medical Research and The Kinghorn Cancer Centre, Darlinghurst, New South Wales, Australia.

ABSTRACT
During pregnancy, the ETS transcription factor ELF5 establishes the milk-secreting alveolar cell lineage by driving a cell fate decision of the mammary luminal progenitor cell. In breast cancer, ELF5 is a key transcriptional determinant of tumor subtype and has been implicated in the development of insensitivity to anti-estrogen therapy. In the mouse mammary tumor virus-Polyoma Middle T (MMTV-PyMT) model of luminal breast cancer, induction of ELF5 levels increased leukocyte infiltration, angiogenesis, and blood vessel permeability in primary tumors and greatly increased the size and number of lung metastasis. Myeloid-derived suppressor cells, a group of immature neutrophils recently identified as mediators of vasculogenesis and metastasis, were recruited to the tumor in response to ELF5. Depletion of these cells using specific Ly6G antibodies prevented ELF5 from driving vasculogenesis and metastasis. Expression signatures in luminal A breast cancers indicated that increased myeloid cell invasion and inflammation were correlated with ELF5 expression, and increased ELF5 immunohistochemical staining predicted much shorter metastasis-free and overall survival of luminal A patients, defining a group who experienced unexpectedly early disease progression. Thus, in the MMTV-PyMT mouse mammary model, increased ELF5 levels drive metastasis by co-opting the innate immune system. As ELF5 has been previously implicated in the development of antiestrogen resistance, this finding implicates ELF5 as a defining factor in the acquisition of the key aspects of the lethal phenotype in luminal A breast cancer.

Show MeSH
Related in: MedlinePlus