Feldman, Andrew L.

Abstract Introduction: Identification of causal DNA copy number variations (CNVs) is critical to the understanding of cancer development and progression, and is important clinically for accurate prognosis and efficient treatment. T-cell lymphomas are replete with amplifications and deletions that result in over- and under-expression of genes. Array comparative genomic hybridization (aCGH) has been the technology of choice to detect CNVs but recently, it has been proposed that mate-pair Next Generation Sequencing (mpNGS) can produce similar results. Here, we report a comparison between the two techniques. Methods: We have developed an algorithm based on extreme value theory to identify CNVs across the genome at high resolution from mate-pair NGS data. The algorithm is based on plotting the distribution of read-counts in windows across the genome, and detecting deletions and gains by finding the windows that differ significantly from the distribution of the normal part of the genome. We compared this algorithm with aCGH data to detect CNVs. We ran five cases of T-cell lymphoma with both techniques and compared the results using concordance statistics. Results: With 50kb lower cutoff for a CNV, mpNGS detected 98% of the cumulative deleted genome and 94% of the amplified/gained genome that were called by aCGH. Most of the discordant CNVs that were detected by only aCGH were in regions of repetitive areas where mapping fails. Conversely, aCGH detected 75% of the cumulative deleted genome and 74% of the amplified/gained genome that were called by mpNGS. Conclusion: A detailed comparison of these techniques showed that both have pros and cons. In general, mpNGS can detect most CNVs detected by aCGH, as long as they are more than 50KB. Additionally, mpNGS offers confirmatory data, particularly for deletions, based on detection of aberrant mate pairs that map to the rejoined ends of deleted regional. Finally, the mate-pair technology can also detect balanced translocations and can also predict fusion genes. Citation Format: George Vasmatzis, Andrew L. Feldman, Sarah H. Johnson, Erik C. Thorland, Rafael Fonseca, Esteban Braggio, Troy J. Gliem. Copy number detection using genomics technologies: A comparison between aCGH and NGS. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 3572. doi:10.1158/1538-7445.AM2014-3572

Abstract Background: Fusion genes, transcripts, and proteins are key drivers of hematopoietic oncogenesis and can serve as effective therapeutic targets. Among peripheral T-cell lymphomas (PTCLs), little is known about fusion genes except for those involving ALK. Discovery of fusion genes may provide therapeutic targets for PTCLs, which have poor outcomes with conventional therapy. We have developed methods for detecting candidate fusions using (1) mate-pair DNA sequencing (MPseq), an approach that spans the entire genome at a fraction of the cost of traditional whole-genome sequencing, and (2) RNA sequencing (RNAseq). Each technique has strengths, limitations, and false discovery rates. Here, we integrated results of both methods to discover and orthogonally validate expressed fusion genes in PTCLs. Methods: DNA and RNA were extracted from 49 frozen, OCT-embedded clinical PTCL specimens, including 25 anaplastic large cell lymphomas; 16 PTCLs, not otherwise specified; and 8 other PTCLs. MPseq and RNAseq were performed using Illumina library protocols and a HiSeq 2000. Detection of fusion genes and transcripts followed our published methods (Blood 2012;120:2280 and Nucleic Acids Res 2011;39:e100, respectively). Functional annotation of genes involved in fusions was performed using DAVID. Results: Of 2818 candidate chromosomal rearrangements identified by MPseq and 90 candidate fusion transcripts identified by RNAseq, 45 events were present in both datasets (mean/case, 0.9; range, 0-7). Fusions were inter-chromosomal in 16 and intra-chromosomal in 29. Annotation clusters with the highest enrichment scores related to SH2 domain-containing proteins and tyrosine kinases (including FER and MAP2K3). Seven genes were recurrently involved in fusions, including known fusion partners such as ALK and previously unreported fusion partners such as PPP1R8, a phosphatase inhibitor important for the activity of Polycomb group proteins such as EZH2. Detection of candidate events only by MPseq could be explained in part by breakpoints mapping to non-genic regions (79% of events). Detection only by RNAseq could be explained in part by unannotated transcripts spanning neighboring named genes (median intra-chromosomal distance, 0.09 MB vs. 12.7 MB for intra-chromosomal candidates identified by MPseq; p<0.0001, Wilcoxon test). Conclusions: Integrated DNA/RNA sequencing is an effective, clinically feasible tool for simultaneous discovery and orthogonal validation of expressed fusion genes. This approach yielded fusions in PTCLs involving genes with key functions in lymphocytes, including novel and recurrent fusions, some of which might be targetable with compounds currently in clinical trial. Candidate intra-chromosomal fusion transcripts not identified by MPseq may represent unrecognized isoforms or the results of novel read-through mechanisms, and merit further study. Citation Format: Andrew L. Feldman, George Vasmatzis, Yan W. Asmann, Sarah H. Johnson, Bruce W. Eckloff, Sumit Middha, Jaime I. Davila, Paul J. Kurtin, Brian K. Link, Stephen M. Ansell, James R. Cerhan. Integrated DNA/RNA sequencing for discovery and orthogonal validation of expressed fusion genes in peripheral T-cell lymphomas. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 5585. doi:10.1158/1538-7445.AM2014-5585

Abstract Objectives and Rationale: Structural variants (SV) including large copy number aberrations (CNV), translocations, inversions, and large insertions and deletions (INDEL) play a critical role in tumorigenesis and progression. In fact, we now know that tumors can be categorized according to the size of mutations harbored. In several cancers, including ovarian and breast cancer, the large structural mutations, rather than single site mutations, play a dominate role in tumor etiology. Therefore, it's critical to implement reliable algorithm for the detection of structural variants in DNA sequencing data. Mate-pair sequencing is a protocol specifically implemented for detection of the whole-genome level structural variants. It requires less sequencing depth therefor is cost effective, and enables the detection of CNVs, translocations, and inversions simultaneously. However, so far there has been no reliable bioinformatics pipeline for the analyses of the mate-pair sequencing data. Methods: Our novel algorithm, the SnowShoes-SV, is an exhaustive search algorithm designed specifically for mate-pair DNA sequencing data analyses. It calls the SVs based on disconcordant read pairs. The false SVs are filtered according to the following criteria: (i) the number of the supporting read pairs; (ii) the lack of reads from control data that implicate SV at the same region; (iii) the mappability and uniqueness of the region based on data from the ENCODE project; (iv) consistencies of the mapping orientations of the supporting read pairs; (v) the similar sizes between sequencing library and the two end read clusters. Results: Using a set of samples previously genotyped by aCGH, the SnowShoes-SV successfully detected all known CNVs and other SVs. It also identified copy number neutral translocations and inversions previously not identified by aCGH. In addition, the algorithm nominated novel SVs which are to be validated by PCR. Conclusions: SnowShoes-SV is a highly sensitive and specific algorithm for SV detection from the mate-pair DNA sequencing data. Citation Format: Yan W Asmann, Chen Wang, Brian M Necela, Xianfeng Chen, Jean-Pierre A Kocher, Matthew J Maurer, Thomas M Habermann, Susan L Slager, Andrew L Feldman, Anne J Novak, James R Cerhan, Edith A Perez, E Aubrey Thompson. An exhaustive algorithm for detecting copy number aberrations and large structural variants in whole-genome mate-pair sequencing data [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr P6-07-03.

Abstract Interferon regulatory factor-4 (IRF4) is a tightly-regulated transcription factor involved in growth and differentiation of normal lymphocytes. IRF4 is expressed in multiple lymphoid malignancies and its expression drives tumor growth. Thus, characterization of the molecular events regulating IRF4 levels is key to design any targeting strategy. Here, we define a novel regulatory mechanism controlling IRF4 expression in peripheral T-cell lymphoma (PTCL), an aggressive non-Hodgkin lymphoma with poor overall survival rates and high levels of this transcription factor. In normal T cells, IRF4 was induced upon external activation of protein kinase C (PKC) by phorbol myristate acetate and ionomycin, an effect reversed by the PKC inhibitor, enzastaurin. Unlike normal T cells, PTCL cells constitutively expressed IRF4 in the absence of external stimulation, and this expression was resistant to enzastaurin. To identify regulators of IRF4 expression in PTCL, we performed a pharmacological screen using inhibitors that targeted PKC-associated pathways, including mitogen-activated protein kinases, PI3K/AKT/mTOR, and NF-κB. Of the inhibitors tested, only the proteasome inhibitor, bortezomib, showed a decrease in IRF4 expression in a dose- and time-dependent manner. To confirm this result we used a more specific NF-κB inhibitor blocking IκB kinase, BMS-345541. This inhibitor dramatically reduced IRF4 expression by 95.6% (P<0.0001). We used cell-based assays to investigate further the role of NF-κB in the transcriptional regulation of IRF4. A reporter assay for the IRF4 promoter demonstrated reduced promoter activity by siRNA knockdown of RelB and p52 NF-κB subunits to 31.4% and 36.9% of control (P=0.009). IRF4 promoter activity also was inhibited 67.9% by IκBαΔN, a dominant negative inhibitor of NF-κB activity (P=0.002), and was increased 71% by siRNA knockdown of IκBα (P<0.05). Chromatin immunoprecipitation assays confirmed the occupancy of NF-κB subunits in the IRF4 promoter, and this interaction was diminished in the presence of BMS-345541. Finally, knockdown of RelB and p52 subunits had a tendency to decrease IRF4 gene expression by 24.4% and 27.3% (P=0.1). Taken together, these data demonstrate transcriptional regulation of IRF4 by the NF-κB family in PTCL, and suggest the NF-κB pathway may represent a therapeutic target for PTCLs with concurrent IRF4 and NF-κB expression. Citation Format: Rebecca L. Boddicker, N. Sertac Kip, Luciana L. Almada, Julie C. Porcher, Deanna M. Grote, Stephen M. Ansell, Martin E. Fernandez-Zapico, Andrew L. Feldman. Interferon regulatory factor-4 is transcriptionally regulated by NF-κB in peripheral T-cell lymphomas. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 2323. doi:10.1158/1538-7445.AM2014-2323

Key PointsThe NF-κB subunits p52 and RelB increase IRF4 promoter activity and expression in PTCL cells. A positive feedback loop involving CD30, NF-κB, and IRF4 drives PTCL cell proliferation and can be blocked by NF-κB inhibitors.

Key Points BCL2 mutations in FL correlate with activation-induced cytidine deaminase expression and frequently alter the amino acid sequence of the protein. Mutations in the BCL2 coding sequence at diagnosis are associated with shortened time to transformation and earlier death due to lymphoma.

Abstract Motivation: RNA-seq has become the method of choice to quantify genes and exons, discover novel transcripts and detect fusion genes. However, reliable variant identification from RNA-seq data remains challenging because of the complexities of the transcriptome, the challenges of accurately mapping exon boundary spanning reads and the bias introduced during the sequencing library preparation. Method: We developed RVboost, a novel method specific for RNA variant prioritization. RVboost uses several attributes unique in the process of RNA library preparation, sequencing and RNA-seq data analyses. It uses a boosting method to train a model of ‘good quality’ variants using common variants from HapMap, and prioritizes and calls the RNA variants based on the trained model. We packaged RVboost in a comprehensive workflow, which integrates tools of variant calling, annotation and filtering. Results: RVboost consistently outperforms the variant quality score recalibration from the Genome Analysis Tool Kit and the RNA-seq variant-calling pipeline SNPiR in 12 RNA-seq samples using ground-truth variants from paired exome sequencing data. Several RNA-seq–specific attributes were identified as critical to differentiate true and false variants, including the distance of the variant positions to exon boundaries, and the percent of the reads supporting the variant in the first six base pairs. The latter identifies false variants introduced by the random hexamer priming during the library construction. Availability and implementation: The RVboost package is implemented to readily run in Mac or Linux environments. The software and user manual are available at http://bioinformaticstools.mayo.edu/research/rvboost/. Supplementary information: Supplementary data are available at Bioinformatics online.

Abstract Background: Peripheral T-cell lymphomas (PTCLs) represent a heterogeneous group of generally aggressive malignancies, with 5-year overall survival (OS) rates of ∼35%. Only ALK translocations, which define ALK-positive anaplastic large T-cell lymphoma (ALCL), have emerged as a genetic biomarker predicting favorable outcome following conventional chemotherapy (e.g. CHOP). Using next-generation sequencing, we discovered additional recurrent chromosomal translocations in PTCLs involving either the DUSP22 locus on 6p25.3 or TP63 on 3q28. The purpose of this study was to evaluate the prognostic utility of combined testing for both these events in PTCL patients. Methods: We examined the clinicopathologic features of 134 systemic PTCLs using WHO criteria, including 49 ALCLs, 23 angioimmunoblastic PTCLs, 50 PTCLs of “not otherwise specified” type, and 12 PTCLs of other subtypes. Immunohistochemistry was used to determine ALK expression, an established surrogate for ALK translocations in ALCL. Fluorescent in situ hybridization (FISH) was performed to detect translocations of DUSP22 or TP63 in all cases. OS was plotted using the Kaplan-Meier method. Comparisons were made using the log-rank test. Results: There were 83 males and 51 females, with a mean age of 58 years (range, 6 to 90 years). Median OS for all patients was 22 months (5-year OS, 36%) with a median follow-up of 24 months (range, 0.1 to 354 months). Translocations of DUSP22, TP63, and ALK were mutually exclusive. Patients with ALK-positive ALCL (n=17) had a median OS of 176 months (median follow-up, 68 months; 5-year OS, 69%). Despite being ALK-negative, cases with translocations of DUSP22 or TP63 shared pathologic features with ALCL. Median OS was not reached in patients with DUSP22 translocations (n=7; median follow-up, 88 months; 5-year OS, 100%). Patients with tumors harboring TP63 translocations (n=9) had a median OS of 13 months (all dead at 76 months; 5-year OS, 11%). Patients with none of these abnormalities (n=101) had a median OS of 17 months (median follow-up, 13 months; 5-year OS, 29%). Differences in OS among these groups were highly significant (p=0.0002). Conclusions: PTCLs with DUSP22 translocations had a favorable prognosis similar to ALK-positive ALCLs. In contrast, TP63 were associated with very poor prognosis. FISH for these biomarkers is a promising tool for risk stratification, particularly since both translocations occurred in cases with overlapping pathologic features. PTCLs with DUSP22 translocations may represent a genetically defined subset of ALK-negative ALCLs. Like patients with ALK-positive ALCLs, these patients may not derive additional benefit from intensified up-front strategies such as early transplantation. Patients with none of these abnormalities comprised 75% of patients studied and had an intermediate prognosis, indicating the need for additional biomarkers in this group to improve risk stratification. Citation Format: Edgardo R. Parrilla Castellar, George Vasmatzis, Ryan A. Knudson, Rhett P. Ketterling, Karen L. Grogg, William R. Macon, Mark E. Law, Stephen M. Ansell, Ahmet Dogan, Andrew L. Feldman. Genetic biomarkers for risk stratification of patients with peripheral T-cell lymphoma. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 3538. doi:10.1158/1538-7445.AM2013-3538