Title: Downregulation of glutathione transferase π sensitizes lymphoma/leukaemia cells to platinum‐based chemotherapy
Abstract: Over ten years ago, oxaliplatin was substituted for cisplatin in a DHAX regimen for patients with refractory high-grade non-Hodgkin lymphoma and significant activity was observed in patients resistant to conventional therapy (Chau et al, 2001). Oxaliplatin also had a favourable single agent activity in previously treated patients with refractory lymphoma (Oki et al, 2005) and produced a response in mucosa-associated lymphoid tissue lymphomas (Raderer et al, 2005). The enzyme glutathione-s-transferase π (GTPπ) (Townsend & Tew, 2003) and the anti-apoptoic factors [BCL2/BCL2L1 (bcl-xl)] have been implicated in cisplatin/carboplatin resistance in ovarian cancer (Jian & Meyer-Hermann, 2010). AKR1C1 [aldo-keto reductase family 1, member C1, also known as dihydrodiol dehydrogenase (DDH)] has also been shown to produce resistance to cisplatin in cell lines derived from different primary sites (Chen et al, 2008). GTPπ levels appeared to be related to prognosis in diffuse large B cell lymphoma (DLBCL) (Ribrag et al, 2003) and mantle cell lymphoma (MCL) (Thieblemont et al, 2008). A study was therefore undertaken with six lymphoma/leukaemia cell lines from different subclasses of lymphoma/leukaemia in order to determine whether sensitivity to platinum drugs could be increased by GTPπ downregulation. Cisplatin and oxaliplatin were obtained from Sigma Chemical Company (St. Louis, MO, USA), cell culture reagents and gentamicin were obtained from Cellgro (Herndon, VA, USA) and RNA-zol B from Tel-test (Friendsworth, TX, USA) and Taq DNA polymerase and Omniscript reverse transcriptase were from Qiagen (Valencia, CA, USA). The lymphoma/leukaemia cell lines were purchased from the American Type Culture Collection (ATCC; Manassas, VA, USA). The cell culture suspensions were run on a FACS-Caliber flow cytometer (Becton Dickenson, San Jose, CA, USA). Formalin-fixed de-paraffinized cell blocks were reacted with antibodies (BCL2, ALK, CD30 and the Epstein-Barr virus in situ hybridization (EBV ISH) kit were purchased from Ventana, Phoenix, AZ, USA) followed by detection of the primary antibody - horseradish peroxidase complex with DAB (3, 3′-diaminobenzidine). MTT [3-(4,5-dimethythiazol-. 2-yl)-2,5-diphenyl tetrazolium bromide] assays and Western blots were performed as previously described (Chen et al, 2008); the antibodies included monoclonals to BCL2, MCL1 (Santa Cruz Biochemistry, Santa Cruz, CA, USA) rabbit polyclonal to GTP-π (Enzolife Science, Plymouth Mtg, PA, USA), AKR1C1 and AKR1C2 mouse polyclonal antibodies (Abnova, Walnut, CA, USA) and a mouse monoclonal to AKR1C3 (Abcam, Cambridge, MA, USA). Knockdown experiments were performed as described by the manufacturer (Ambion, Austin, TX, USA). The Toledo cells were comprised of intermediate-sized lymphocytes that expressed CD10, CD19, CD20 and BCL2 with kappa light chain restriction and a (14, 18) translocation i.e. follicular cell origin (ATCC - diffuse-large cell lymphoma) (Table 1). The 3008 cell line expressed CD19, CD20, CD5, BCL2 and cyclin D1 with lambda light chain restriction and a (11, 14) translocation consistent with MCL (ATTC-mantle cell). The 2262 cells expressed CD25, CD30, ALK-1 with a (2, 5) translocation corresponding to an anaplastic large cell lymphoma (ALK positive) (ATCC - large cell immunoblastic lymphoma.) The 2289 cell line was composed of large cells (with cleaved nuclei) positive for CD10, CD20, BCL6, occasional BCL2 and cytoplasmic CD22 with lambda light chain restriction and a (14,18) translocation - DLBCL of the G-C subtype (consistent with ATCC). The CEM cell line expressed the T-cell antigens- CD2, CD3, CD5, CD7, cytoplasmic CD3 and CD10, but showed no light chain overexpression, consistent with adult T-cell leukaemia/lymphoma ('acute lymphoblastic leukaemia' – ATCC). The Daudi cell line expressed CD19, 20, 10 and EBV small RNA (EBER) with kappa light chain restriction and a 8,14 (MYC) translocation - Burkitt lymphoma (Burkitt - ATCC). The cytotoxicity profiles of these cell lines to cisplatin and oxaliplatin showed poor correlation (Fig 1 C, D). There was a striking difference in sensitivity of the 2289 (DLBCL) cell line to cisplatin and adriamycin as compared to oxaliplatin – it was relatively sensitive to the latter drug. Interestingly, the MCL cell line (3008) was relatively resistant to adriamycin and oxaliplatin but sensitive to cisplatin. Western blots (Fig 1A) of a series of proteins implicated in cisplatin resistance showed little correlation with DDH, BCL2, MCL1 or thioredoxin levels. As GTPπ was present in most of the resistant cell lines, the effect of its downregulation was investigated. In Fig 1B, the Western blots of the GTPπ knockdowns and bar graphs (Fig 1C–E) documenting the effect of its downregulation on the 50% inhibitory concentration (IC50) of each drug is shown (Fig 1 C–E). Significant sensitization to cisplatin (decrease in IC50 of 50% or more) was observed in all cell lines except 2289 (DLBCL) and surprisingly to oxaliplatin in 3008, 2262 and 2289, but to a lesser degree (30–50%). The IC50 of adriamycin was also decreased in the MCL line 3008 (a cell line resistant to adriamycin) but was unchanged or even increased in the other cell lines. In multiple myeloma (Petrini et al, 1995) a relationship between GTPπ and adriamycin resistance appeared to be related to co-expression of the p glycoprotein, however, reverse transcription polymerase chain reaction showed minimal ABCB1 expression in all the cell lines except Toledo (data not presented). Interestingly, the MCL line GTPπ knockdown showed sensitization to both cisplatin and oxaliplatin and the follicular lymphoma cell line (Toledo) knockdown exhibited a marked increase in sensitivity to cisplatin. Clinical studies with lymphomas suggest that GTPπ levels may have prognostic significance in DLBCL and MCL (Ribrag et al, 2003; Thieblemont et al, 2008). GTPπ inhibitors have also shown robust cytotoxic activity in some haematological cell lines, perhaps related to dissociation of a JNK-GTPπ complex (Turella et al, 2005). This study has shown that in a variety of different lymphoma cell lines, down-regulation of GTPπ resulted in increased sensitivity to either cisplatin or oxaliplatin (or both) which may have future therapeutic implications i.e. employment of antisense derivatives of GTPπ or employing enzyme inhibitors. It should be noted that GTPπ knockdowns sensitized the cells to cisplatin to a far greater extent than oxaliplatin (except for 2289 (DLCBL) - minimal effect). However, this cell line is sensitive to oxaliplatin and GTPπ knockdown increased further its sensitivity to this drug. The work was supported by the National Institutes of Health [Grant R01-CA098804, Funding Agency: National Cancer Institute].