Letters to the Editor 1407 8 Yin W, Rossin A, Clifford JL, Gronemeyer H. Co-resistance to retinoic acid and 12 Green CL, Evans CM, Zhao L, Hills RK, Burnett AK, Linch DC et al. The prognostic TRAIL by insertion mutagenesis into RAM. Oncogene 2006; 25: 3735 -- 3744. significance of IDH2 mutations in AML depends on the location of the mutation. 9 Jenkins RB, Wrensch MR, Johnson D, Fridley BL, Decker PA, Xiao Y et al. Distinct Blood 2011; 118: 409 -- 412. germ line polymorphisms underlie glioma morphologic heterogeneity. Cancer 13 Boissel N, Nibourel O, Renneville A, Huchette P, Dombret H, Preudhomme C. Genet 2011; 204: 13 -- 18. Differential prognosis impact of IDH2 mutations in cytogenetically normal acute 10 Rice T, Zheng S, Xiao Y, Decker PA, McCoy LS, Smirnov I et al. Associations of myeloid leukemia. Blood 2011; 117: 3696 -- 3697. glioma risk loci by IDH mutation status. NeuroOncology 2011; 13(Suppl 3): iii27. 14 Patnaik MM, Hanson AA, Hodnefield JM, Lasho T, Finke C, Knudson RA et al. 11 Sanson M, Hosking FJ, Shete S, Zelenika D, Dobbins SE, Ma Y et al. Differential prognostic effect of IDH1 versus IDH2 mutations in myelodysplastic Chromosome 7p11.2 (EGFR) variation influences glioma risk. Hum Mol Genet syndromes: a Mayo Clinic study of 277 patients. Leukemia 2011; e-pub ahead of 2011; 20: 2897 -- 2904. print 28 October 2011; doi:10.1038/leu.2011.298.

Dual inhibition of Jak2 and STAT5 enhances killing of myeloproliferative neoplasia cells

Leukemia (2012) 26, 1407--1410; doi:10.1038/leu.2011.338; the principal effect of pimozide in these MPN cell lines is the published online 2 December 2011 inhibition of STAT5 phosphorylation. Three lines of evidence suggest that pimozide is not functioning as a classic kinase inhibitor. First, pimozide does not inhibit Jak family kinases in an isolated in vitro kinase assays (Table 1). Myeloproliferative neoplasms (MPNs) are a group of clonal Second, its pattern of inhibition of substrate phosphorylation disorders that arise from the transformation of hematopoietic catalyzed by extracts from Jak2 V617F-expressing cells using the stem cells. In 2005, several groups reported a single acquired point PamChip tyrosine kinase microarray system (PamGene, Den Bosch, mutation in the Janus kinase 2 (Jak2) gene in the majority of Netherlands)11 is clearly distinct from that mediated by a Jak patients with Philadelphia chromosome (Ph)-negative MPN.1 The inhibitor (including no inhibition of the phopshorylation of the mutation is believed to have a critical role in the pathogenesis of Jak2 peptide by pimozide; Figure 1b). Pimozide did inhibit these disorders,2,3 and it has been suggested that most patients phosphorylation of a peptide derived from Jak1, though this harbor mutations (some of which remain to be identified) that likely occurred through an indirect mechanism. Finally, other potentially serve as a molecular target for selective Jak2 inhibition.4 signaling pathways downstream of Jak2, such as the phosphor- Although small molecule Jak2 inhibitors are entering clinical trials, ylation of extracellular signal-regulated kinase mitogen-activated their ultimate efficacy is unclear.5 In addition to the concern of protein kinases (MAPK) are not inhibited by pimozide (as they are insufficient inhibition of mutated Jak2 in vivo or the emergence of with Jak inhibitor 1) in BAFEJ and HEL cells, and actually show a resistance through activation of complementary pathways, many small but reproducible increased activity (Figure 1c). This finding is MPNs contain other mutational events (for example, mutation in similar to that seen in cells transformed with BCR/ABL,10 and may exon12 of Mpl,6 or the KIT D618V mutation in patients with reflect a loss of a negative regulator of MAPK signaling that is systemic mastocytosis7), and thus are not sensitive to Jak inhibitors. STAT5 dependent.12 Although the exact mechanism by which Therefore, the development of inhibitors to common mediators of pimozide inhibits STAT5 phosphorylation is being elucidated, it diverse signaling pathways in this disease is very desirable. may involve direct effects on STAT5 or on a negative regulator of One convergence point of these pathways is the transcription STAT5 activation rather than by targeting a kinase. factor STAT5. STAT5 regulates the expression of genes controlling Consistent with the central role of STAT5 activation in driving key events such as cell cycle progression and survival. Thus, expression of genes underlying the malignant nature of these continuous STAT5 activation drives increased expression of these cells, inhibition of STAT5 phosphorylation by pimozide also led genes that directly contribute to leukemogenesis. Continued to a decrease in expression of STAT5 target genes (Figure 1d). STAT5 activation appears to be necessary for tumor cell survival,8 Significantly, these include two important pro-survival genes, although normal cells are generally tolerant to the loss of STAT5 Bcl-xl and Mcl1. Bcl-xl appears to be a key mediator in promoting function. Consequently, STAT5 is an attractive target for cancer Jak2-mediated survival in SET2 cells13 as inhibition of Jak2 signaling therapy. Increasing evidence indicates that STAT5 activation is blocked STAT5-mediated regulation of Bcl-xl mRNA level and required for Jak2 V617F-mediated transformation.9 As STAT5 is a reduced its protein expression. Mcl1 is also a critical survival factor critical mediator of the effects of Jak2 V617F, the development of in MPN cells. Specific STAT5-docking domains have a major role in drugs that inhibit this transcription factor holds promise as a inducing Mcl1 expression driven by a mutated FLT3 containing an treatment for MPN. Furthermore, the dual inhibition of both STAT5 internal tandem duplication (FLT3-ITD), a mutation found in about and Jak2 may yield better disease control. 20% of acute myelogenous leukemia patients, independent of We previously identified the neuroleptic drug pimozide as an JAK2 activation.14 inhibitor of STAT5 transcriptional function in a cell-based screen.10 Given that pimozide inhibits STAT5 tyrosine phosphorylation In the present study, we evaluated the effect of pimozide alone and decreases STAT5 target genes expression, we next deter- and in combination with a Jak2 inhibitor on the biology of mined whether pimozide also affects the viability of cells with myeloproliferative cell lines. We examined a model system, Ba/F3 activating mutations of Jak2. Pimozide caused a dose-dependent cells that had been reconstituted with erythropoietin receptor and decrease in viability of the three cell lines, with IC50 (half maximal the mutant form of Jak2 (BAFEJ), as well as the HEL and SET2 inhibitory concentration) values ranging from 4 mM for BAFEJ and human leukemia cell lines that endogenously express Jak2 V617F. SET2 to 10 mM for HEL cells (Figure 1e and data not shown). As Treatment of each cell line with pimozide led to a decrease in expected, pimozide also inhibits survival of parental Ba/F3 cells, tyrosine phosphorylation of STAT5 (Figure 1a and data not which are known to be dependent on STAT5. shown). Reflecting the fact that the phosphorylation of STAT5 is To determine the mechanism by which pimozide decreases mediated by activated Jak2, the Jak2 inhibitor Jak inhibitor 1 also viable cell numbers, BAFEJ and HEL cells were treated with led to decreased phosphorylation of STAT5. STAT3 phosphoryla- pimozide for 24 h, after which they were permeabilized and tion was only minimally affected by pimozide (Figure 1a). Thus, stained with propidium iodide followed by flow cytometric

& 2012 Macmillan Publishers Limited Leukemia (2012) 1402 -- 1448 Letters to the Editor 1408

Figure 1. Pimozide inhibits STAT5 phosphorylation, decreases expression of STAT5 target genes and reduces viability of MPN cells. (a) HEL cells were treated with vehicle, pimozide (10 mM) or Jak inhibitor 1 (1 mM) for 5 h. Immunoblots were performed with the indicated antibodies. (b) Peptide phosphorylation profiles of HEL cells treated with pimozide or JAK inhibitor 1. Relative phosphorylation of each peptide is shown compared with control (orange represents 0.99--0.75 activity, red represents 0.74--0.50 activity). (c) HEL cells were treated with vehicle, pimozide (10 mM) or Jak inhibitor 1 (1 mM) for 5 h, after which immunoblots were performed to phosphorylated MAP kinase and total MAP kinase. Normalized band intensities are indicated. (d) HEL cells were treated with vehicle, pimozide (10 mM) or Jak inhibitor 1 (0.8 mM) for 6 h, after which RNA was harvested, and expression of the indicated genes was measured using quantitative reverse transcriptase PCR and normalized to the expression of HPRT. (e) HEL cells were treated with pimozide or Jak inhibitor 1 at the indicated concentrations for 48 h, after which viable cell number was quantitated by ATP-dependent bioluminescence.

roles performed both by mutated Jak2 and STAT5 in MPNs, we Table 1. Pimozide does not directly inhibit Jak tyrosine kinase activity hypothesized that dual inhibition of both Jak2 and STAT5 may lead to enhanced effects on myeloproliferative cells. When HEL Kinase % Inhibition of kinase activity cells were treated with pimozide or Jak inhibitor 1, each of the drugs alone led to a 50% reduction in tyrosine phosphorylation JAK1 6 JAK2 À3 of STAT5. However, when cells were treated with the combina- JAK2 JH1 JH2 À2 tion, there was near-complete loss of STAT5 phosphorylation JAK2 JH1 JH2 V617F À3 (Figure 2a). Reflecting this combinatorial effect on STAT5 JAK3 À7 phosphorylation, treatment of HEL cells with the combination of pimozide and Jak inhibitor 1 also led to a greater reduction in the The effect of pimozide on the activity of selected tyrosine kinases was level of the pro-survival protein Mcl1 (Figure 2b). It also led to a analyzed using the SelectScreen Kinase Profiling service. greater reduction in viable cells over a range of combinations in HEL cells as well as with the BAFEJ and SET2 cell lines (Figure 2c analysis. In both cell lines, pimozide led to a prominent decrease and data not shown). Thus, this dual inhibition of the Jak--STAT in cells in S phase with a concomitant increase in cells in the pathway leads to enhanced cytotoxicity of myeloproliferative cells. G0/G1 phase of the cell cycle (Supplementary Figure 1a). To determine the mechanism for this combinatorial effect As this change did not appear sufficient to explain the decrease between pimozide and a Jak2 inhibitor, we analyzed the induction in viable cell number observed, we treated the cells with pimozide of apoptosis as measured by staining with annexin V-FITC and for 48 h, after which they were stained with annexin V-FITC and propidium iodide followed by flow cytometric analysis. In HEL cell propidium iodide followed by flow cytometric analysis. Pimozide line, the percentage of apoptotic cells increased from 8.2% at led to an approximate threefold increase in apoptotic cells in both baseline to 23.1% with pimozide alone, and to 35.6% with Jak the lines, indicating that this drug induces both a cell cycle arrest inhibitor 1. The combination of both drugs led to an increase of and apoptosis, consistent with its effect on both pro-survival the apoptotic cells to almost 70% (Figure 2d). Similar results were genes (Bcl-xl and Mcl-1) and genes promoting cell cycle found in the BAFEJ cell line as well, over a range of concentration progression (Cyclin D1; Supplementary Figure 1b). of both drugs (data not shown). Thus pimozide, particularly in Given that pimozide and Jak inhibitor 1 both decrease STAT5 combination with the Jak2 inhibitor Jak inhibitor 1, leads to a phosphorylation through distinct mechanisms, and the central potent induction of apoptosis.

Leukemia (2012) 1402 -- 1448 & 2012 Macmillan Publishers Limited Letters to the Editor 1409

Figure 2. The combination of pimozide and Jak inhibitor 1 leads to enhanced effects on MPN cell lines. (a) HEL cells were treated with pimozide (7.5 mM), Jak inhibitor 1 (0.2 mM), or both for 5 h after which immunoblot analysis was performed for phosphorylated STAT5 and total STAT5. Normalized band intensities are displayed below. (b) HEL cells were treated with vehicle, pimozide (10 mM), Jak inhibitor 1 (0.8 mM) or the combination of both drugs for 8 h, after which Mcl1 was measured by immunoblot analysis. (c) HEL cells were treated with pimozide (10 mM), Jak inhibitor 1 (0.8 mM) or the combination of both drugs for 48 h, after which viable cell number was measured using an ATP-dependent bioluminescence assay. (d) HEL cells were treated with pimozide (10 mM), Jak inhibitor 1 (0.8 mM), or the combination for 48 h, after which annexin V/propidium iodide staining and flow cytometry was performed. The graphs represent the percentage of annexin V-positive apoptotic cells in each of the treatment conditions.

This enhanced cytotoxic effect on myeloproliferative cells of brief exposures to pimozide were sufficient to induce some loss in pimozide in conjunction with a Jak inhibitor likely reflects both a viability. This may reflect the fact that loss of expression of STAT5- more complete inhibition of STAT5 function, as well as inhibition of dependent pro-survival genes, such as Bcl-xl and Mcl1, may independent pathways downstream of activated Jak2. By inhibiting trigger an irreversible cell death pathway as occurs with oncogene distinct targets, combination treatments may allow a greater addiction. cytotoxic effect on the tumor cells, and may decrease the emergence Even if it is not possible to achieve adequate levels of pimozide of drug resistance. Also, by allowing the use of lower concentrations in patients to be used as a STAT5 inhibitor, these findings have of each drug, this strategy may minimize side effects. several implications. They establish the utility of a direct STAT5 Pimozide is approved for the treatment of Tourette’s syndrome, inhibitor in MPNs; they provide a chemical scaffold that may be and has a well described safety profile in humans. However, it can modified to develop more efficacious and less toxic STAT5 cause cardiac or neurologic toxicity, and thus it has a limited inhibitors; and they provide a probe to allow the identification therapeutic window. Pimozide is not known to cause hematologic of unique cellular targets for STAT5 inhibition. toxicity, and has very little effect on the survival of peripheral In conclusion, pimozide inhibits STAT5 activation in MPN cells, blood mononuclear cells from healthy donors or colony formation decreases STAT5-dependent gene expression and induces apop- of CD34 þ cells from healthy individuals.10 To determine the tosis in these cells. Notably, these effects are enhanced when minimum exposure necessary to decrease viability of MPN cells, pimozide is combined with Jak2 inhibition. These data suggest we exposed cells to pimozide for varying times, washed out the that directly inhibiting STAT5, as well as simultaneously inhibiting drug and then assessed survival in a standard 72 h viability assay both STAT5 and Jak2, may be effective strategies for the treatment (Supplementary Figure 1c). These experiments showed that even of MPN.

& 2012 Macmillan Publishers Limited Leukemia (2012) 1402 -- 1448 Letters to the Editor 1410 CONFLICT OF INTEREST 4 Pardanani A, Hood J, Lasho T, Levine RL, Martin MB, Noronha G et al. The authors declare no conflict of interest. TG101209, a small molecule JAK2-selective kinase inhibitor potently inhibits myeloproliferative disorder-associated JAK2V617F and MPLW515L/K mutations. Leukemia 2007; 21: 1658 -- 1668. ACKNOWLEDGEMENTS 5 Goldman JM, Green AR, Holyoake T, Jamieson C, Mesa R, Mughal T et al. Chronic This work was supported by the Kittredge Foundation (Dana-Farber Cancer Institute), myeloproliferative diseases with and without the Ph chromosome: some the Brent Leahey Fund (Dana-Farber Cancer Institute), Gabrielle’s Angel Foundation unresolved issues. Leukemia 2009; 23: 1708 -- 1715. (New York, NY), the Claudia Adams Barr Program in Innovative Basic Cancer Research 6 Pikman Y, Lee BH, Mercher T, McDowell E, Ebert BL, Gozo M et al. MPLW515L is a (Dana-Farber Cancer Institute) and Friends of the Dana-Farber Cancer Institute. novel somatic activating mutation in myelofibrosis with myeloid metaplasia. PLoS Med 2006; 3: e270. M Bar-Natan1,2, EA Nelson1, SR Walker1, Y Kuang3, 7 Grimwade LF, Happerfield L, Tristram C, McIntosh G, Rees M, Bench AJ et al. Phospho-STAT5 and phospho-Akt expression in chronic myeloproliferative RJ Distel3 and DA Frank1,2 1 neoplasms. Br J Haematol 2009; 147: 495 -- 506. Department of Medical Oncology, Dana-Farber 8 Nelson EA, Frank DA. Developing pharmacological modulators of STAT signaling. Cancer Institute, Boston, MA, USA; In: Stephanou A (ed). JAK-STAT Pathway in Disease. Landes Bioscience: Austin, 2 Departments of Medicine, Brigham and Women’s 2009, pp 164 -- 172. Hospital, and Harvard Medical School, 9 Funakoshi-Tago M, Tago K, Abe M, Sonoda Y, Kasahara T. STAT5 activation is Boston, MA, USA and critical for the transformation mediated by myeloproliferative disorder-associated 3Translational Research Laboratory, The Center for JAK2 V617F mutant. J Biol Chem 2010; 285: 5296 -- 5307. Clinical and Translational Research, Dana-Farber 10 Nelson EA, Walker SR, Weisberg E, Bar-Natan M, Barrett R, Gashin LB et al. The Cancer Institute, Boston, MA, USA STAT5 inhibitor pimozide decreases survival of chronic myelogenous leukemia cells resistant to kinase inhibitors. Blood 2011; 117: 3421 -- 3429. E-mail: [email protected] 11 Sikkema AH, Diks SH, den Dunnen WF, ter Elst A, Scherpen FJ, Hoving EW et al. Kinome profiling in pediatric brain tumors as a new approach for target discovery. Cancer Res 2009; 69: 5987 -- 5995. REFERENCES 12 Rane SG, Reddy EP. Janus kinases: components of multiple signaling pathways. 1 Levine RL, Gilliland DG. Myeloproliferative disorders. Blood 2008; 112: 2190 -- 2198. Oncogene 2000; 19: 5662 -- 5679. 2 Delhommeau F, Pisani DF, James C, Casadevall N, Constantinescu S, Vainchenker 13 Gozgit JM, Bebernitz G, Patil P, Ye M, Parmentier J, Wu J et al. Effects of the JAK2 W. Oncogenic mechanisms in myeloproliferative disorders. Cell Mol Life Sci 2006; inhibitor, AZ960, on Pim/BAD/BCL-xL survival signaling in the human JAK2 V617F 63: 2939 -- 2953. cell line SET-2. J Biol Chem 2008; 283: 32334 -- 32343. 3 Wernig G, Mercher T, Okabe R, Levine RL, Lee BH, Gilliland DG. Expression of 14 Yoshimoto G, Miyamoto T, Jabbarzadeh-Tabrizi S, Iino T, Rocnik JL, Kikushige Y Jak2V617F causes a polycythemia vera-like disease with associated myelofibrosis et al. FLT3-ITD up-regulates MCL-1 to promote survival of stem cells in acute myeloid in a murine bone marrow transplant model. Blood 2006; 107: 4274 -- 4281. leukemia via FLT3-ITD-specific STAT5 activation. Blood 2009; 114: 5034--5043.

Supplementary Information accompanies the paper on the Leukemia website (http://www.nature.com/leu)

Long-term WT1 peptide vaccination for patients with acute myeloid leukemia with minimal residual disease

Leukemia (2012) 26, 1410--1413; doi:10.1038/leu.2011.343; adjuvant (Freund Corporation, Tokyo, Japan) was i.d. injected, and published online 13 December 2011 the vaccination was performed three times at 2-week intervals. The terms of both clinical and molecular responses were clearly defined as respectively, a decrease in leukemic blast cells Acute myeloid leukemia (AML) still has a rather poor prognosis, in BM and a decrease in WT1 mRNA levels in BM and/or PB. which is mainly ascribed to relapse in the majority of patients Molecular CR was defined as undetectable levels of WT1 mRNA despite achievement of complete remission (CR). To prevent in PB (o50 copies/mg RNA). Molecular relapse was defined as relapse, minimal residual disease (MRD) remaining in CR has to be abnormal levels of WT1 mRNA (X50 copies/mg RNA in PB and/or completely eradicated. The graft-versus-leukemia effect induced X1000 copies/mg RNA in BM). Vaccination was extended for by allogeneic hematopoietic stem cell transplantation is the most patients who showed clinical response or no progression within powerful mechanism for eradication of MRD, thus strongly 2 weeks from the end of three WT1 vaccinations. As for immuno- þ þ þ suggesting the therapeutic potential of immunotherapy. logical monitering, WT1 tetramer CD8 T cells and IFN-g WT1 þ þ The Wilms’ tumor gene WT1 is highly expressed in various types tetramer CD8 T cells were analyzed. Of the eight AML of malignancies, and the WT1 protein is one of the most promising patients with hematological CR (leukemic blast o5%) but with leukemia and tumor-associated antigens.1--6 Based on findings of MRD who were WT1-vaccinated, five patients who showed a pre-clinical studies, we performed a phase I clinical study of WT1 decrease in WT1 mRNA levels (molecular response) were peptide immunotherapy for patients with AML, myelodysplastic repeatedly WT1-vaccinated, but two relapsed later. The remaining syndromes, breast or lung cancer, and were able to confirm the three patients received WT1 vaccine for X8 years and have safety and clinical efficacy of this therapy.7 In this report, we survived in molecular CR for X8 years without significant adverse describe the long-term outcome for AML patients enrolled in this effects except for local erythema at the vaccine injection sites study. (Table 1). The major inclusion criteria for vaccination were that patients had HLA-A*2402 and that WT1 mRNA levels, which reflect the amount of MRD,8,9 were 41 Â 103 or 41 Â 102 copies/mg RNA in CASE REPORTS bone marrow (BM) or peripheral blood (PB), respectively. A natural Case 1 WT1 peptide (amino acid 235--243 CMTWNQMNL) or the modified A 32-year-old male was diagnosed as AML (M4Eo) in January 1997 WT1 peptide (CYTWNQMNL)10 emulsified with Montanide ISA51 and achieved hematological CR by induction chemotherapy. After

Leukemia (2012) 1402 -- 1448 & 2012 Macmillan Publishers Limited