Supporting Information
Total Page：16
File Type：pdf, Size：1020Kb
Supporting Information Pfeifer et al. 10.1073/pnas.1305656110 SI Materials and Methods we used the Benjamini and Hochberg method to calculate a false Cell Culture, Retroviral Constructs, and Transductions. Human dif- discovery rate (FDR) for every α-significance threshold. Gen- fuse large B-cell lymphoma (DLBCL) cell lines were cultured in erally, more genes were up-regulated (546 genes; P < 0.01 and RPMI (Invitrogen) with 10% (vol/vol) FCS (Sigma), except for FDR < 0.11) than down-regulated (279 genes; P < 0.01 and OCI-Ly1, OCI-Ly2, OCI-Ly4, OCI-Ly7, OCI-Ly10, OCI-Ly19, FDR < 0.11) by PTEN. Using tighter cutoffs, we identified 72 and TMD8, which were cultured in Iscove’s modified Dulbecco genes that were significantly down-regulated (P < 0.0025; FDR < medium supplemented with either 20% human plasma or 10% 0.06) and 82 genes that were up-regulated (P < 0.00025; FDR < 0.02) across all time points after PTEN induction (paired t test FCS. All cell lines were maintained at 37 °C with 5% CO2. For efficient retroviral transductions, cell lines were engineered based on two independent replicates) (Fig. 4B and Table S6). to express the murine ecotropic receptor as previously described To obtain a better understanding of the gene expression changes, (1). Additionally, these cell lines were engineered to express the we performed an unbiased gene set enrichment analysis (GSEA) as bacterial tetracycline repressor, allowing doxycycline-inducible previously described using a previously curated gene expression shRNA or cDNA expression. shRNA-mediated RNA inter- signature database (5, 6). The most enriched gene signatures from > ference was performed as described (1). The targeting sequences this query (enrichment score 0.5) are reported in Table S7. fi of MYC shRNAs 1 and 2 were CGATTCCTTCTAACAGAAATG Additionally, we performed gene expression pro ling of 34 and CCTATGAACTTGTTTCAAATG, respectively. As a negative primary DLBCL patient samples using Affymetrix GeneChip control, we used a previously described shRNA directed against Human Exon 1.0 ST v2 microarrays. Samples from this cohort MSMO1 (2). PTEN (NM_000314.4), MYC (NM_002467.2), and a were later used as the basis for PTEN staining. These microar- constitutive active AKT cDNA (NM_005163.2) (3) were inserted rays were preprocessed with robust multiarray average (RMA) into either a modified version of the inducible pRetroSuper dual background subtraction and quantile normalization. Probes were promoter vector or a noninducible pMSCV vector as previously aggregated by median polish on gene level based on the v14.1 of described (4). PTEN- and mutant PTEN-induced toxicity was the Entrez Gene CDF for Affymetrix GeneChip Human Exon assessed as previously described (4). Briefly, PTEN or mutant 1.0 ST v2 (7). fi PTEN cDNA was transduced using retroviruses. The expression TheDLBCLsampleswereclassied into activated B-cell-like vector coexpressed GFP, allowing us to monitor the proportion of (ABC) and GCB DLBCL as previously described (8). We used two + − GFP vs. GFP cells over time as a measure of toxicity of the co- reference samples from a previous cohort (9) to transfer the molec- expressed PTEN cDNA. Rescue experiments were performed as ular subtype prediction to Affymetrix GeneChip Human Exon arrays. previously described (2). Each experiment was completely repro- PCR Amplification and Sequencing. PCR amplification and sequenc- duced at least two to eight times for each cell line. ing were performed as previously described (10). The sequences PTEN Gene Expression Profiling. Gene expression profiling after reex- for primers applied to amplify exons are summarized in pression of PTEN cDNA was performed in the PTEN-deficient Table S4. germinal center B-cell-like (GCB) DLBCL cell line HT. HT cells Array CGH. Array comparative genomic hybridization (aCGH) in were transduced with PTEN cDNA and selected with puromycin, DLBCL cell lines was performed as previously described (11). and PTEN cDNA expression was induced with doxycycline; 6, 12, aCGH data from 10 DLBCL cell lines were obtained from GEO 18, and 24 h after PTEN induction, total RNA was isolated using through GEO accession no. GSE43272. The aCGH data from the NucleoSpin RNA II Kit (Macherey & Nagel) according to the remaining cell lines have been deposited in the GEO data- the manufacturer’s protocol. RNA was amplified and labeled base (accession no. GSE45495). with the TotalPrep RNA Amplification Kit (Illumina), and la- beled samples were hybridized on HumanHT-12 v4 Expression ’ Quantitative Genomic PCR. To determine the DNA copy number of BeadChips (Illumina) using the manufacturer s protocol. The the PTEN locus, we used a predesigned assay according to the gene expression data have been deposited in the Gene Expres- manufacturer’s protocol (Invitrogen). sion Omnibus (GEO) database (http://www.ncbi.nlm.nih.gov/geo/; accession no. GSE45495). FISH. Primary DLBCL patient samples from cohorts 1 and 2 were PTEN-induced changes in gene expression were measured in investigated for the presence of MYC translocations using FISH two completely independent biologic replicates, and gene ex- as previously described (12, 13). pression changes in PTEN-transduced cells were compared with cells that were transduced with the empty vector alone. The in- Cell Viability Assay, Annexin-V Staining, and SNARF-1 Proliferation dependent measurements were preprocessed and normalized Assay. GCB DLBCL cells were incubated with DMSO or differ- in the following manner. Data were imported on raw bead level, ent concentrations of the pan-PI3K inhibitor Ly294002 (Cayman and subsequently, a bead-level spot filter was applied for each Chemicals). Cell viability was measured after incubation for 4 d microarray based on the fitted density mode for the background using the Cell Proliferation Kit II (Roche) according to the intensities. Afterward, bead intensities of all measured micro- manufacturer’s recommendations. arrays were quantile-normalized, and beads were grouped by Annexin-V PE (BD Pharmingen) staining was performed in the measured sequence to form bead sets. For genes having more GCB DLBCL cell lines BJAB, HT, and K422 2 d after PTEN than one measured sequence, additional merged bead sets were cDNA transduction according to the manufacturer’s protocol. created. Bead sets with more than 50% of their beads excluded Annexin-V staining of GFP-positive cells was measured by flow by the spot filter were also excluded. Additional analyses was cytometry. performed on gene level using median aggregation. SNARF-1 (carboxylic acid, acetate, and succinimidyl ester; Life Differentially expressed genes were identified in the following Technologies) staining was performed in the DLBCL cell lines manner. A one-tailed paired t test was used to calculate P values BJAB, HT, and K422 before PTEN cDNA transduction. SNARF-1 for every gene based on the eight microarray pairs. Additionally, dilutions were measured 2 and 6 d after PTEN transduction. Pfeifer et al. www.pnas.org/cgi/content/short/1305656110 1of13 SNARF staining was assessed by flow cytometry according to the transferred to polyvinylidene difluoride membranes (Millipore). manufacturer’s protocol. All antibodies used in this study were obtained from Cell Signal- ing, except for PTEN (Santa Cruz), MYC (Epitomics), p-MYC Western Blotting. Western blotting was performed as previously (Abcam), actin, and α-tubulin (Sigma). described (2). In brief, whole-cell protein lysates were harvested Western blotting for p-AKT was performed in DLBCL cell from DLBCL cell lines or primary patient samples in Phospho- lines and 16 primary GCB DLBCL samples from cohort 1. In safe lysis buffer (EMD Millipore), and protein was quantified addition, four GCB DLBCL samples with known PTEN sta- using the BCA assay (Thermo Scientific). Subsequently, lysates tus, for which frozen material was available, were analyzed for were separated by SDS/PAGE on 12% polyacryalmide gels and p-AKT levels. 1. Ngo VN, et al. (2006) A loss-of-function RNA interference screen for molecular targets 8. Wright G, et al. (2003) A gene expression-based method to diagnose clinically distinct in cancer. Nature 441(7089):106–110. subgroups of diffuse large B cell lymphoma. Proc Natl Acad Sci USA 100(17): 2. Wenzel SS, et al. (2013) MCL1 is deregulated in subgroups of diffuse large B-cell 9991–9996. lymphoma. Leukemia 27(6):1381–1390. 9. Hummel M, et al. (2006) A biologic definition of Burkitt’s lymphoma from 3. Kharas MG, et al. (2010) Constitutively active AKT depletes hematopoietic stem cells transcriptional and genomic profiling. N Engl J Med 354(23):2419–2430. and induces leukemia in mice. Blood 115(7):1406–1415. 10. Lenz G, et al. (2008) Oncogenic CARD11 mutations in human diffuse large B cell 4. Hailfinger S, et al. (2011) Malt1-dependent RelB cleavage promotes canonical NF- lymphoma. Science 319(5870):1676–1679. kappaB activation in lymphocytes and lymphoma cell lines. Proc Natl Acad Sci USA 108 11. Lenz G, et al. (2008) Molecular subtypes of diffuse large B-cell lymphoma arise by (35):14596–14601. distinct genetic pathways. Proc Natl Acad Sci USA 105(36):13520–13525. 5. Shaffer AL, et al. (2006) A library of gene expression signatures to illuminate normal 12. Obermann EC, Csato M, Dirnhofer S, Tzankov A (2009) Aberrations of the MYC gene and pathological lymphoid biology. Immunol Rev 210:67–85. in unselected cases of diffuse large B-cell lymphoma are rare and unpredictable by 6. Subramanian A, et al. (2005) Gene set enrichment analysis: A knowledge-based morphological or immunohistochemical assessment. J Clin Pathol 62(8):754–756. approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci USA 13. Kwanhian W, et al. (2012) MicroRNA-142 is mutated in about 20% of diffuse large B- 102(43):15545–15550. cell lymphoma. Cancer Med 1(2):141–155. 7. Dai M, et al. (2005) Evolving gene/transcript definitions significantly alter the interpretation of GeneChip data. Nucleic Acids Res 33(20):e175.