Acquired differentiation and loss of malignancy of spontaneous pulmonary metastases in human-in-mouse breast cancer models

Jessica Bockhorn1,2, Aleix Prat3, Ya-Fang Chang1, Simo Huang4, Chika Nwachukwu5, Maria J. Gomez-Vega5, Olufunmilayo I. Olopade5, Charles M. Perou6, Geoffrey L. Greene1, Huiping Liu4*

  1. Supplemental legends
  2. Supplemental materials and methods
  3. Supplemental Figures S1-S6
  4. Supplemental Tables S1-S6
  1. Supplemental legends

Supplemental Figure S1. The Differentiation scores (Dscore)of primary tumors and lung met-derived tumors.

Based on gene expression microarrays, a total of 25,422 probes had 100% good data. Of these probes, 22,548 successfully mapped to an entrez gene ID. Probes with a common gene ID were then collapsed and these 13,011 unique genes were median centered. We then utilized a Differentiation Predictor that measures each samples position along a normal mammary developmental axis from mammary stem cell, luminal progenitor, and mature luminal cell states (1).

Supplemental Figure S2.miRNA expression in breast tumors.

Differential expression of candidate miRNAs in triple negative (TN) and non-TN breast tumors from the University of Chicago tumor set (P<0.05).

Supplemental Figure S3. Increased miR-138 expression in metastases versus paired primary tumors.

Differential expression of miR-138 (A), miR-138-1* and miR-138-2* (B), in paired primary tumors versus metastatic tumorsin lymph nodes and distant organs from the GSE37407 dataset (n=21 samples from 14 patients, P<0.05 or 0.01).P-values have been obtained after a paired Student´s t-test.

Supplemental Figure S4. FISH analysis of miR-138 gene copy number

  1. Representative Fish staining images of the miR-138-1 gene (in green) and a centromeric probe for chromosome 3, CEP3 (in red), in the parental (the left panel) and met-derived (the right panel) models.
  2. Table of a manual counting of both green (miR-138-1) and red (CEP3) dots in 50 cells and the average copies of genes per cell for miR-138-1 and CEP3.

Supplemental Figure S5.Ability of the miR-138 activity signature to predict breast cancer-specific survival.

Kaplan-Meier curves of survival outcomes based on miR-138 activity scores signature to predict breast cancer-specific survival in the MDACCC588 (n=588, p=0.00216).

Supplemental Figure S6. Downstream targets of miR-138

  1. Upper panel: top circles show the number of overlapped genes downregulated by miR-138 and predicted direct targets. Bottom table lists predicted 9 direct targets, including EZH2, PPIP5K1, and others.
  2. Histogram bars indicate the up-regulated miR-138 expression level upon transfections.
  3. mRNA levels of representative genes (EZH2, IL-11, VIM and PPIP5K1) inhibited by miR-138 but promoted by anti-miR-138 inhibitor, measured by real-time PCR with specific Taqman primers. ***<0.0001, ****<0.00001
  4. IL-11 protein levels inhibited by miR-138. ***<0.0001
  5. 3’UTR of EZH2, PPIP5K1 and VIM showing predicted miR-138 binding sites between partial complementary sequences.
  6. Luciferase activity assays with inhibitory effects of miR-138 on the luciferase expression by interacting with 3’UTRs of target genes (EZH2, PPIP5K1 and VIM), located downstream of the luciferase gene. *<0.05, **<0.001
  7. EZH2 expression in ER+ and ER- tumors from the U Chicago patient dataset (n=50) as measured by RT-PCR.

Supplemental Table S1.Gene microarray data comparing M1 lung met-derived tumors with the parental tumors.

Supplemental Table S2.IPA gene pathway analysis of 492 differentially expressed genes in M1 lung met-derived tumors compared to the parental tumors.

Supplemental Table S3.Differentially expressed miRNAs between the Met-derived tumors and Parental tumors.

Supplemental Table S4.Gene microarray data comparing MDA-MB-231 cells with enforced expression of miR-138 with control MDA-MB-231 cells.

Supplemental Table S5.Transcription factor predictive binding analyses of hsa-miR-138-1 and has-miR-138-2 promoter regions (-3kb to +1) by JASPAR.

Supplemental Table S6.Ability of the miRNA138 activity signature to predict survival outcome beyond classical clinical-pathological variables (age, tumor size, grade, nodal status, intrinsic subtypes, and the proliferation gene signature in the METABRIC dataset (n=1848, adjusted P<0.05).

  1. Supplemental Materials and Methods

Independent gene expression microarray datasets:The various signatures and genes were evaluated in multiple publicly available datasets previously reported: UNC337 (3) and MDACC588) (1,4,5), METABRIC(6) and GSE37407(7). The UNC337 dataset (GSE18229) includes microarray dataset with 320 primary breast tumors and 17 true normal breast tissues. Raw data was obtained from the UMD website. The MDACC588 dataset includes distant relapse-free survival data from 588 primary tumors that were treated with neoadjuvantanthracycline/taxane-based chemotherapy and endocrine therapy (if the tumor was estrogen receptor positive by IHC). The METABRIC dataset is a UK/Canadian-based set of 1848 patients with primary breast cancer with follow-up data. Finally, the GSE37407 dataset includes global miRNA expression profiling on 47 tumor samples from 14 patients with paired samples from primary breast tumors and corresponding lymph node and distant metastases. The miRNA and mRNA expression profiles of the University of Chicago tumor dataset (n=46) were previously deposited to GSE39543 and GSE22049 respectively (8).

Vectors and cloning: The lentiviral gateway vectors pFU-L2G and pFU-L2T were used to label human-in-mouse breast tumor xenografts(2). MiRNA precursor entry clones were subcloned to make pFU-miR-PGK-L2G using LR clonase II (Invitrogen). Human EZH2 cDNA was subcloned into pDEST40 (Invitrogen) from a donor vector (ORFeome collection, University of Chicago). Luciferase vector containing the 3’UTR of EZH2, PPIP5K1 and VIM were obtained from Switchgear Genomics. The luciferase reporter vector pGL4 (Invitrogen) was used to clone the promoter regions (-3kb) of miR-138-1 and miR-138-2 genes on chromosome 3 and 16 respectively, which were PCR amplified from BAC clonesRP11-147A10 and RP11-833C20(distributed by Children's Hospital Oakland Research Institute, CHORI). Sequences of the designed primers with cloning insertion sites are: 138-1-3kbForXhoI forward with an XhoI site (ATCTCGAGTCACAGGGGACACAAAGAACAGATT), 138-1-3kbRevBglII reverse primer with a Bgl II site (GTAGATCTGGCAACGGCCTGATTCACAAC),138-2-3kbForXhoI forward primer with an XhoI site (ATCTCGAGCTGTGTCTTCTTGCATCTGAACC), and 138-2-3kbRevHindIII reverse with a Hind III site (GTAAGCTTTGCGCTGCTCGTCGG). Amplified PCR All vectors were verified by sequencing.

Luciferase assay: Hek293T cells were plated at 50,000 cells per well in 48 well plates one day prior to transfection. Cells were subsequently transfected with miRNA mimics (Dharmacon) as described earlier. After six hours, the cells were then transfected overnight with 3’UTR luciferase (Switchgear genomics) and control renilla vectors (Promega) using Fugene (Roche-Promega) as per manufacturer’s instructions. For promoter luciferase assays, the upstream putative promoter region (-3kb) of miR-138 was cloned into pGL4 (Invitrogen) as described above. 50,000 MDA-MB-231 cells were plated in 48 well plates one day prior to transfection. Cells were then transfected with the GATA3 expression vector pCMV6XL5 (or pcDNA vector control), pGL4-138-3kB and renilla luciferase vector. Cells were then read as described earlier. The cells were lysed 48 hours later using 5X passive lysis buffer (Promega) and read using a dual luciferase kit (Promega) on a microplate reader (Perkin Elmer). The luciferase signal was normalized to the renilla signal and compared to scramble control.

Lentivirus production and transduction: High-titer lentivirus was produced as described(2).Second-generation packaging vectors (VPR and VSVG) were used to produce miRNA precursor lentiviruses, and third-generation packaging vectors (pRSV-rev, VSVG, and pMDLg/RRE) were used to produce other lentiviruses.Viable CD44+ breast cancer cells were isolated from TN1 xenograft tumors(2) and transduced with lentiviruses at 10 or 20 multiplicities of infection (MOI 10 or 20).

Fluorescence in situ hybridization (FISH): Dual-color FISH assays were conducted using following probes: mir138-1/CEP3 probe mixture containing homebrewed mir138-1 DNA (BAC clone RP11-147A10; 3p21.33) labeled with SpectrumGreen and the SpectrumOrangeCEP3 (centromere enumeration probe for chromosome 3, Vysis/Abbott Molecular (Des Plaines, IL).). Homebrewed probes were directly labeled using Nick Translation Kit (Vysis/Abbott Molecular) according to the manufacturer instructions. Chromosomal mapping and hybridization efficiency for the probe mixture was verified in metaphase spreads of a normal lymphoblastoid cell line. Metaphase cell preparations from normal peripheral blood lymphocytes were done according to routine protocols. (The AGT Cytogenetics Laboratory Manual.Vol. 3.Lippincott & Raven; 1997.) In all FISH experiments, the hybridization procedure and post-hybridization washes were done as described by Vysis/Abbott Molecular.

FISH results interpretation: In each specimen, 50 well-defined malignant nuclei with abnormal signal patterns were scored. The absolute number of each signal, the mean copy number of signal per cell, the ratios mir138-1:CEP3and the percentage of cells with given copy number of each signal per cell were calculated.

Statistical Analysis: For all assays and analyses in vitro, if not specified, a student T test was used to evaluate the significant difference or p-values and standard deviations (SD) of mean values were depicted as error bars in figures. For animal studies in vivo, tumor growth curves were analyzed using a linear mixed model in R software(9) with the tumor volume (bioluminescence signals) or its fold change as a response variable. Lung metastases were analyzed using Wilcoxon rank sum test in R software. Log transformation was taken on both tumor volume and its fold change. As two tumors were planted and observed from each mouse, mouse ID (ear tag number) was included in the model as random effect. To allow autocorrelation among the tumor size measurements from the same mouse, AR (1) (autoregressive model of order 1) was added to the error term.

All microarray cluster analyses were displayed using Java Treeview version 1.1.4. Average linkage hierarchical clustering was performed using Cluster v2.12 (10). Biologic analysis of microarray data was performed with the DAVID annotation tool (11). Survival curves were calculated by the Kaplan-Meier method and compared by the log-rank test. A Cox model was used to evaluate the association of the miR-138 signature with survival outcome.

1.Prat A, Parker JS, Karginova O, Fan C, Livasy C, Herschkowitz JI, et al. Phenotypic and molecular characterization of the claudin-low intrinsic subtype of breast cancer. Breast Cancer Res 2010;12(5):R68.

2.Liu H, Patel MR, Prescher JA, Patsialou A, Qian D, Lin J, et al. Cancer stem cells from human breast tumors are involved in spontaneous metastases in orthotopic mouse models. Proc Natl Acad Sci U S A 2010;107(42):18115-20.

3.Prat A, Parker JS, Karginova O, Fan C, Livasy C, Herschkowitz JI, et al. Phenotypic and molecular characterization of the claudin-low intrinsic subtype of breast cancer. Breast Cancer Res;12(5):R68.

4.Usary J, Zhao W, Darr D, Roberts PJ, Liu M, Balletta L, et al. Predicting drug responsiveness in human cancers using genetically engineered mice. Clin Cancer Res 2013;19(17):4889-99.

5.Hatzis C, Pusztai L, Valero V, Booser DJ, Esserman L, Lluch A, et al. A genomic predictor of response and survival following taxane-anthracycline chemotherapy for invasive breast cancer. JAMA : the journal of the American Medical Association 2011;305(18):1873-81.

6.Curtis C, Shah SP, Chin SF, Turashvili G, Rueda OM, Dunning MJ, et al. The genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups. Nature 2012;486(7403):346-52.

7.Gravgaard KH, Lyng MB, Laenkholm AV, Sokilde R, Nielsen BS, Litman T, et al. The miRNA-200 family and miRNA-9 exhibit differential expression in primary versus corresponding metastatic tissue in breast cancer. Breast Cancer Res Treat 2012;134(1):207-17.

8.Bockhorn J, Dalton R, Nwachukwu C, Huang S, Prat A, Yee K, et al. MicroRNA-30c inhibits human breast tumour chemotherapy resistance by regulating TWF1 and IL-11. Nat Commun 2013;4:1393.

9.Team RDC. R: A language and environment for statistical computing. . Vienna, Austria: R Foundation for Statistical Computing; 2011.

10.Eisen MB, Spellman PT, Brown PO, Botstein D. Cluster analysis and display of genome-wide expression patterns. Proceedings of the National Academy of Sciences of the United States of America 1998;95(25):14863-8.

11.Dennis G, Jr., Sherman BT, Hosack DA, Yang J, Gao W, Lane HC, et al. DAVID: Database for Annotation, Visualization, and Integrated Discovery. Genome Biol 2003;4(5):P3.