ALBERT

Description: Background: Peripheral T-cell lymphoma (PTCL) consists of an uncommon and heterogeneous group of lymphomas that are often challenging to diagnose and classify. Since most patients also have a poor survival with standard multiagent chemotherapy, more effective therapeutic approaches are needed to improve patient outcome. Table1: Pathological diagnosis Number of cases profiled AITL 36 ALK(+)ALCL 19 ALK (−)ALCL 08 ATLL 12 T/NK 14 PTCLU 44 Other rare entities 10 Methods: A mRNA profiling study using Affymetrix HGU133+2 arrays on 143 cases of PTCL and NK-cell lymphoma (NKCL) from the International Peripheral T-cell Lymphoma Project, was conducted on pre-treatment biopsies. These included the following pathologically classified cases (Table 1). In addition, we also profiled nine NK cell lines, seven T cell lines, normal resting and activated CD4+ and CD8+ T cells and resting and IL2- activated NK cells from healthy individuals. BRB-ArrayTools was used to develop gene classifiers for the major PTCL entities and survival predictors for AITL based on gene expression data. Results: We have identified key molecular signatures for PTCL and NKCL that have allowed us to construct a robust classifier for AITL (207 transcripts), ALK+ ALCL (94), ATLL (225) and NKCL (127). PTCL-U group may have 3 or 4 molecular subgroups and additional studies with more cases, are necessary to further define this group. Misclassified cases were identified and re-assigned to the molecularly defined entities, including re-assigning of 9/44 PTCL-U to AITL. We have confirmed the enriched expression of genes identified in follicular helper T-cells in AITL, suggesting that AITL is derived from this T-cell subset. A number of oncogenic pathways (e.g. NF-κB, HIF-a,VEGF, IL6) and tumor/host interactions that contributed to local tumor-induced immunosuppression (e.g. TGF-b), were identified in AITL. A molecular predictor of outcome was developed for AITL and validated by leave one-out-cross validation. Since PTCL is an uncommon disease, future studies will require the collaboration of multiple large clinical groups with tissue resources for both discovery and validation. Conclusion: This study has demonstrated that GEP will allow the construction of robust and biologically-meaningful classifiers for PTCL, and prognosticators can be derived for well-defined entities with a sufficient number of cases. GEP will also allow us to identify therapeutically-relevant oncogenic pathways and tumor/host interactions that may lead to improvement in the therapy and outcome of patients with PTCL and NKCL. (This study is a part of the International T-cell Lymphoma Project)

Description: Background: Natural Killer (NK)-cell lymphomas/leukemias (NKL) account for 1–2 % of all non-Hodgkin lymphomas. Although the incidence of NKL is relatively low, the clinical course of these lymphomas is highly aggressive. To elucidate the recurrent genomic abnormalities and the associated changes in the gene expression profile, we performed GEP and array-CGH studies on seven clinically well defined cases and eight well characterized cell lines derived from NKL patients. Methods: Array-CGH was performed on a tiling BAC array and GEP on an Affymetrix 133 plus2 array. The two data sets were correlated to identify functional alterations associated with the genetic abnormalities. Candidate genes on del 6q21 were identified and further studied for mutations and promoter methylation. Results: Our aCGH study identified frequent recurrent gains (〉 25 %) in 1q, 2p, 7q, 13q, 17q and 20pter-qter. Regions of loss with a frequency (〉 25%) included 1p, 6q, 7p, 9p, 9q, 1011q, 12q, 13q, 15q, 16p, 17p, 18p, and 18q consistent with other studies. Only 30%-50% of the genes residing in the gained or deleted regions showed corresponding increased or decreased expression. Many of the genes with increased expression were involved in cell proliferation, growth and energy metabolic processes important for the neoplastic cells. In deleted regions, genes showing decreased expression included transcription factors or repressors (e.g. SP4, PRDM1, NCOR1 and ZNF10), tumor suppressors or negative regulators of the cell cycle (e.g. CDKN2C, CDKN2A, CDKN2B, CHFR and FOXO1A). 6q21 is a frequent deleted region and we found a minimal region of involvement including only three known genes (PRDM1, ATG5 and AIM1), that had consistent low expression. Sequence analysis of these genes revealed mutations in PRDM1 and AIM1 in two out of eight cell lines resulting in truncated proteins of PRDM1 and changes in conserved amino acid sequences of AIM1. Only one cell line with a point mutation in a non-conserved region in ATG5 was detected. DNA methylation analysis showed that the promoter of PRDM1 was highly methylated in NK cell lines having low PRDM1 expression. Reversal of promoter methylation with decitabine treatment resulted in increased PRDM1 transcription consistent with down-regulation of PRDM1 expression on promoter methylation. Finally, we observed progressive up-regulation of PRDM1 expression in primary NK cells in response to IL2 treatment suggestive of a role of PRDM1 on the regulation of NK-cell activation. Conclusion: Combination of high resolution genomic and transcriptional profiling in NK-cell malignancies has provided evidence of a general tumor promoting effect of genomic copy number alterations as well as the identification of candidate genes in a commonly deleted region in 6q. Specifically, we propose PRDM1 as a tumor suppressor gene in NK cell malignancies and loss-offunction of PRDM1 in combination with the haplo-insufficiency of ATG5 might have significant roles in the pathogenesis of NK-cell malignancies.

Description: Killer immunoglobulin-like receptors (KIRs) play an essential role in the regulation of natural killer cell functions. KIR genes are highly polymorphic in nature, showing both haplotypic and allelic variations among people. We demonstrated in both in vitro and in vivo models a significant heterogeneity in function among different KIR2DL1 alleles, including their ability to inhibit YT-Indy cells from degranulation, interferon γ production, and cytotoxicity against target cells expressing the HLA-Cw6 ligand. Subsequent experiments showed that the molecular determinant was an arginine residue at position 245 (R245) in its transmembrane domain that mechanistically affects both the efficiency of inhibitory signaling and durability of surface expression. Specifically, in comparison with R245-negative alleles, KIR2DL1 that included R245 recruited more Src-homology-2 domain-containing protein tyrosine phosphatase 2 and β-arrestin 2, showed higher inhibition of lipid raft polarization at immune synapse, and had less down-regulation of cell-surface expression upon interaction with its ligand. Thus, our findings provide novel insights into the molecular determinant of KIR2DL1 and conceivably a fundamental understanding of KIR2DL1 allelic polymorphism in human disease susceptibility, transplant outcome, and donor selection.

Description: Abstract 313 Background: Natural Killer (NK) cell lymphomas (NKCL) are rare with aggressive clinical behavior. The majority of these cases belong to extra-nodal NK/T-cell lymphoma of nasal type (ENKTL) of the current World Health Organization (WHO) classification scheme. ENKTL also includes peripheral T-cell lymphomas (PTCL) that are similar in many respects to the NK cell counterpart. Due to rarity of the disease and difficulty in obtaining adequate biopsy specimens, the molecular mechanisms underlining ENKTL are largely unknown. We profiled a series of NK-cell lymphoma cases and many well- characterized cell lines of NK- and T-cell lineages to define molecular classifiers that can distinguish NKCL from PTCL, including lymphomas of cytotoxic T-cells. We also evaluated oncogenic pathways in these tumors and the therapeutic potential of a novel inhibitor of a cell cycle regulator (aurora kinase A). Patients and Methods: The gene expression profiling (GEP) of ENKTL (n=21) and PTCL-U (n=50) cases were performed using HG U133 plus 2 arrays (Affymetrix Inc, CA). GEP of other PTCL subtypes (n=90), normal NK and T cells (resting and activated), NK and T cell lines (n=14) and indolent NK- cell/large granular lymphocytic proliferation (NK-LGLP) (n=5) were used for comparative analysis. Immunohistochemistry (IHC) was used to validate the GEP findings. A novel aurora-kinase-A inhibitor (MK-8745) was obtained from Merck & Co (Merck & Co., Inc. NJ, USA) and incubated with the cell lines for 2 -24 hours at 0.1-1 μM concentrations. Results: The ENKTL showed a male predominance (2:1) with a median age of 55 years at diagnosis and aggressive clinical behavior [5-year OS (

Description: The PR (PRDI-BF1-RIZ) domain zinc finger protein 1 (PRDM1) is a transcription repressor with a pivotal role in plasma-cell differentiation. We identified clonal inactivating mutations in PRDM1 in the diffuse large B-cell lymphoma (DLBCL) cell line OCI-Ly3 and in 8 of 35 de novo clinical DLBCL samples. The mutational spectrum consists predominantly (7 cases) of single-nucleotide mutations affecting consensus splice donor sites, some of which are recurrent, that lead to splicing aberrations and premature translation termination. In 2 of these cases, point mutations appear to be caused by RNA editing with G-to-A and U-to-G conversions. Other mutations include frame-shift deletion and chromosomal inversion. Except for one mutant, which may act as a dominant-negative, all mutations are associated with either deletion or silencing of the paired PRDM1 allele. This study identifies PRDM1 inactivation as a recurrent genetic defect in DLBCL cells and establishes PRDM1 as a potential tumor suppressor gene in DLBCL. Moreover, it implies inhibition of terminal differentiation as a pathogenetic pathway in DLBCL, particularly for the activated B-cell–like DLBCL. It also demonstrates for the first time the potential role of RNA editing in lymphomagenesis.

Description: Follicular lymphoma (FL) is genetically characterized by the presence of the t(14;18)(q32;q21) chromosomal translocation in approximately 90% of cases. In contrast to FL carrying the t(14;18), their t(14;18)-negative counterparts are less well studied about their immunohistochemical, genetic, molecular, and clinical features. Within a previously published series of 184 FLs grades 1 to 3A with available gene expression data, we identified 17 FLs lacking the t(14;18). Comparative genomic hybridization and high-resolution single nucleotide polymorphism (SNP) array profiling showed that gains/amplifications of the BCL2 gene locus in 18q were restricted to the t(14;18)-positive FL subgroup. A comparison of gene expression profiles showed an enrichment of germinal center B cell–associated signatures in t(14;18)-positive FL, whereas activated B cell–like, NFκB, proliferation, and bystander cell signatures were enriched in t(14;18)-negative FL. These findings were confirmed by immunohistochemistry in an independent validation series of 84 FLs, in which 32% of t(14;18)-negative FLs showed weak or absent CD10 expression and 91% an increased Ki67 proliferation rate. Although overall survival did not differ between FL with and without t(14;18), our findings suggest distinct molecular features of t(14;18)-negative FL.

Description: Mantle cell lymphoma (MCL) is an aggressive B-cell malignancy and is characterized by the chromosome translocation t(11;14)(q13;q32), which results in aberrant expression of cyclin D1. The t(11;14)(q13;q32) alone is not sufficient to result in lymphoma, and additional genetic alterations are necessary. Secondary genomic alterations are frequently detected in MCL, of which chromosome 13q31–q32 gain/amplification is one of the most frequent. Amplification at chromosome 13q31–q32 targets a microRNA cluster, miR-17~92, which resides within intron 3 of c13orf25, a non-protein-coding gene at 13q31.3. The miR-17~92 cluster consists of seven miRNAs, and overexpression of the cluster accelerates the development of MYC-induced lymphoma in mice and contributes to carcinogenesis in the lung. Nevertheless, the functional role of miR-17~92 in MCL has not been investigated. Based on gene expression profiling study of 82 primary MCL biopsy specimens using whole genome Affymetrix U133 2.0 plus arrays, we found that overexpression of miR-17~92 was associated with worse overall survival in patients with MCL (p=0.021). We further demonstrated that enforced expression of miR-17~92 reduced chemotherapy-induced apoptosis in retrovirally transduced MCL cell lines. Immunoblot analysis showed that PTEN was down-modulated in MCL cells with overexpression of miR-17~92, and the level of PTEN was restored with reduction of miR-20a levels using an antisense oligonucleotide or a “sponge” plasmid with multiple tandem miR-20a binding sites. The finding was further confirmed by luciferase assays using a reporter plasmid containing the PTEN 3′UTR, and the effect was abrogated by mutating the miR-20 binding site in reporter plasmid, indicating that PTEN was downregulated through the binding of miRNA to the 3′UTR of the transcript. Furthermore, we demonstrated that overexpression of miR-17~92 in MCL cells increased the phosphorylation of AKT and its downstream targets. Interestingly, we found that the protein phosphatase PHLPP2, a negative regulator of the PI3K/Akt pathway, was also a direct target of the miR-17~92 cluster. Moreover, we found that BIM, a BH3-only proapoptotic protein, was a direct target of the miR-17~92 and down-modulation of BIM negatively impacted the intrinsic pathway of apoptosis. In summary, we demonstrated that overexpression of miR-17~92 in MCL cells downmodulates multiple proteins involved in PI3K/Akt signaling and apoptosis, and that downregulation of these targets collaboratively enhances cell growth and chemoresistance in tumor cells. As a consequence, overexpression of miR-17~92 may be associated with poorer survival in MCL patients. Our findings disclose a novel oncogenic pathway in MCL and suggest that targeting the miR-17~92 cluster may provide a novel therapeutic approach for this disease, which is incurable with current chemotherapeutic regimens.

Description: Follicular lymphoma (FL) is a B-cell non-Hodgkin lymphoma that is characterized in approximately 85% of cases by the chromosomal translocation t(14;18) involving BCL2. While FL3b generally lack the t(14;18), this translocation is also absent in 15% of FL grades 1, 2 and 3a. The current study was designed to identify the frequency of t(14;18)-negative FL in a series of 166 cases of FL1, 2 and 3a in which global gene expression profiles had been established previously (Dave et al., NEJM351:2159–69, 2004). Furthermore, we sought to compare genetic alterations and gene expression profiles between FL with and without the t(14;18). Combined polymerase chain reaction (PCR) and tissue microarray-based fluorescence in situ hybridization (FISH) identified 17 t(14;18)-negative FL cases in this series (9%). Virtually all FL cases carrying the t(14;18) showed BCL2 expression by immunohistochemistry (Dako, clone 124), whereas 11 of the FL cases without a t(14;18) were BCL2-negative at the protein level. Clinically, there was no difference between the t(14;18)-negative and -positive FL subgroups regarding age and gender distribution as well as in median survival times. Comparative genomic hybridization (CGH) in the 166 FL cases revealed a characteristic pattern of chromosomal gains and losses, as previously described. However, significant differences were observed between the t(14;18)-negative and -positive FL subgroups. Specifically, the t(14;18)-positive FL subgroup showed gains of chromosomes 18q (18%), 8q (12%) and X (13%), as well as losses of 13q (16%) and 10q (16%), whereas none of these aberrations were observed in the t(14;18)-negative FL cases. To compare gene expression between the two groups, we used gene set enrichment analysis (GSEA), BRB array tools and a two-sided t-test. Cell cycle-associated genes were found to be enriched in the t(14;18)-negative FL subset. These differences were even more pronounced in FL cases that lacked both the t(14;18) and BCL2 expression at the protein level. Importantly, genes expressed in non-malignant bystander cells appeared also differentially enriched and a cytotoxic gene expression signature was found to be more prominent in t(14;18)-negative FL. These findings point to a different composition of the non-neoplastic cells in t(14;18)-positive and -negative FL and could indicate subtle differences in the immunological microenvironment of t(14;18)-negative FL.

Description: Diffuse large B-cell lymphoma (DLBCL) is a heterogeneous diagnostic category with at least three different molecular subtypes distinguishable by gene expression profiling, termed germinal center B cell-like (GCB) DLBCL, activated B cell-like (ABC) DLBCL, and primary mediastinal B cell lymphoma (PMBL). We performed array comparative genomic hybridization (aCGH) in patient samples and cell lines representing different DLBCL subtypes to determine if they utilize distinct pathogenetic mechanisms. Using an array consisting of 386, 165 oligonucleotides (NimbleGen), we performed aCGH on 203 untreated de novo DLBCL samples and 30 DLBCL cell lines, and the same samples were profiled for gene expression using Affymetrix U133 plus arrays. Patient samples included 72 GCB DLBCLs, 74 ABC DLBCLs, 31 PMBLs, and 26 unclassified DLBCLs. Following segmentation of the aCGH data into intervals with a uniform copy number, segments were combined into minimal common regions (MCRs) that were recurrently altered in more than one sample. Statistical differences in MCR frequency between DLBCL subtypes were corrected for multiple hypothesis testing using a false discovery rate (FDR) calculation. The DLBCL subtypes differed in the frequency of MCRs residing at many chromosomal loci, and we used gene expression data to define potential target genes in these MCRs. The INK4a/ARF tumor suppressor locus on 9p21 was selectively lost in ABC DLBCL: homozygous deletions of INK4a/ARF was observed in 20% of ABC DLBCLs but in only 3% of GCB DLBCLs and never in PMBLs (FDR=4.5 E-3). Among ABC DLBCLs, loss of INK4a/ARF was associated with increased proliferation rate, as measured by a proliferation gene expression signature, and adverse survival (p=0.007, log rank test). 16% of ABC DLBCL cases had gain/amplification and overexpression of SPIB, a gene on 19q13 encoding an ETS family transcription factor that is characteristically expressed in ABC DLBCL. This copy number alteration was observed much less frequently in GCB DLBCL (3%) and never in PMBL (FDR=2.6 E-2). GCB DLBCLs had recurrent amplification and overexpression of C13orf25, which encodes the mir-17-92 polycistronic cluster of microRNAs that is transcriptionally activated by c-myc and cooperates with c-myc to accelerate tumor development. C13orf25 amplification was detected in 16% of GCB DLBCLs but in only 3% of PMBLs and never in ABC DLBCL (FDR=3.8 E-3). Recurrent amplification and overexpression of JAK2 on 9p24 was observed in 35% of PMBL cases, but only in 5% of GCB DLBCLs and 4% of ABC DLBCLs respectively (FDR=6.2 E-4). In summary, aCGH revealed copy number abnormalities in DLBCL that had strikingly different frequencies in the three DLBCL subtypes, supporting the hypothesis that these subtypes represent distinct diseases that utilize different oncogenic mechanisms. Our analysis specifically implicated the INK4a/ARF locus as a tumor suppressor and SPIB as an oncogene in ABC DLBCL, the mir-17-92 microRNA cluster as an oncogene in GCB DLBCL, and JAK2 as an oncogene in PMBL.