The Novel JAK Inhibitor CYT387 Suppresses Multiple Signalling Pathways, Prevents Proliferation and Induces Apoptosis in Phenotypically Diverse Myeloma Cells

Home , Janus kinase inhibitor

The novel JAK inhibitor CYT387 suppresses multiple signalling pathways, prevents proliferation and induces apoptosis in phenotypically diverse myeloma cells

KA Monaghan1,2, T Khong1,2, CJ Burns3 and A Spencer1,2

1Myeloma Research Group, Malignant Haematology and Stem Cell Transplantation, Alfred Hospital, Melbourne, Victoria, Australia; 2Australian Centre for Blood Diseases, Department of Clinical Haematology, Monash University, Melbourne, Victoria, Australia and 3YM BioSciences Australia, Melbourne, Victoria, Australia

Janus kinases (JAKs) are involved in various signalling MM. Some human myeloma cell lines (HMCL) cannot pathways exploited by malignant cells. In multiple myeloma proliferate or survive without exogenous IL-6,9,10 and some (MM), the interleukin-6/JAK/signal transducers and activators of conventional drugs are ineffective in the presence of IL-6.11–13 transcription (IL-6/JAK/STAT) pathway has been the focus of The bone marrow microenvironment, known to provide research for a number of years and IL-6 has an established role 14 in MM drug resistance. JAKs therefore make a rational drug supportive signals to MM cells, produces IL-6; hence, target for anti-MM therapy. CYT387 is a novel, orally bioavail- reducing the pro-survival effect of IL-6 may abrogate the drug- able JAK1/2 inhibitor, which has recently been described. This resistant phenotype of MM. CD45 is a phenotypic marker preclinical evaluation of CYT387 for treatment of MM demon- expressed by some myeloma cells, which has been reported to strated that CYT387 was able to prevent IL-6-induced phos- influence IL-6 responsiveness.15 It is therefore likely that CD45 phorylation of STAT3 and greatly decrease IL-6- and insulin-like growth factor-1-induced phosphorylation of AKT and extra- expression might affect sensitivity to JAK inhibitors. cellular signal-regulated kinase in human myeloma cell lines CYT387 is a novel JAK inhibitor that can inhibit JAK1, JAK2, (HMCL). CYT387 inhibited MM proliferation in a time- and dose- JAK3 and TYK2 kinase activity.16,17 The structure and develop- dependent manner in 6/8 HMCL, and this was not abrogated by ment of the compound has recently been described.18 Several the addition of exogenous IL-6 (3/3 HMCL). Cell cycling was JAK inhibitors are currently in various stages of development and inhibited with a G2/M accumulation of cells, and apoptosis was investigation for use in MM including INCB000020,19 induced by CYT387 in all HMCL tested (3/3). CYT387 synergised 20 21,22 (ref.23) 24 in killing HMCL when used in combination with the conven- INCB16562, AG490, AZD1480 and Pyridone 6, tional anti-MM therapies melphalan and bortezomib. Impor- or for other haematological malignancies such as WP1066 in tantly, apoptosis was also induced in primary patient MM cells AML.25 CYT387 has recently undergone Phase I evaluation, (n ¼ 6) with CYT387 as a single agent, and again synergy was demonstrating safe use of low micromolar concentrations, with seen when combined with conventional therapies. no relevant haematological toxicities evident.26 Given the Leukemia (2011) 25, 1891–1899; doi:10.1038/leu.2011.175; tolerance of CYT387 in patients and the putative role of IL-6 published online 26 July 2011 Keywords: multiple myeloma; CYT387; JAK inhibitors; drug in MM drug resistance, JAK inhibitors are being investigated for resistance; STAT signalling their potential use as a single agent or in combination therapy for MM. Furthermore, preliminary in vitro data has demonstrated the potential of JAK/STAT inhibition to sensitise MM cells to conventional therapies.20,22 Here we show that the JAK inhibitor CYT387 can modulate IL-6-stimulated signalling with- Introduction in HMCL, which can sensitise them to various other anti-MM treatments. CYT387 can inhibit proliferation and disrupt the cell Multiple myeloma (MM) is an incurable drug-resistant clonal cycle of MM cells. Finally, CYT387 induces apoptosis as a single B cell malignant neoplasm localised to the bone marrow. Janus agent and synergises with melphalan and bortezomib, when kinases (JAKs) are well-characterised signalling kinases compris- used against either HMCL or primary MM tumour cells. ing four family members JAK1, JAK2, JAK3 and TYK2 that are important in haematological malignancy, as JAK mutations have Materials and methods been shown to contribute to the pathogenesis of both 1–3 4 myeloproliferative disorders and leukaemias. JAKs have an Reagents established role in signalling for many cells (reviewed by Rane 5 The JAK1/2 inhibitor CYT387 was kindly provided by YM and Reddy ), and in MM, JAKs are activated by a variety of BioSciences Australia (Melbourne, Victoria, Australia) and cytokines including interleukin-6 (IL-6),6,7 interferon-a6,8 and 6 dissolved in dimethyl sulfoxide. The proteasome inhibitor epidermal growth factor. Many pathways downstream of JAKs bortezomib (Janssen-Cilag, North Ryde, New South Wales, are exploited by malignant cells; the role of IL-6 and its Australia) was reconstituted in saline. The alkylating agent subsequent activation of JAK/signal transducers and activators of melphalan (Sigma-Aldrich, Sydney, New South Wales, Australia) transcription (STAT) is possibly the most studied pathway in was dissolved in 0.5% HCl.EtOH. All stock drug solutions were diluted in complete RPMI-1640 culture medium to various Correspondence: Professor A Spencer, Myeloma Research Group, concentrations for experimentation. Malignant Haematology and Stem Cell Transplantation, Alfred Hospital and Australian Centre for Blood Diseases, Monash University, AMREP, Commercial Road, Prahran, Victoria, 3181, Australia. E-mail: [email protected] Cell lines and culture conditions Received 2 June 2011; accepted 21 June 2011; published online 26 HMCL LP-1, NCI-H929, OPM2, RPMI-8226 and U266, and the July 2011 human stromal cell line HS5, were obtained from the American Anti-myeloma activity of the JAK inhibitor CYT387 KA Monaghan et al 1892 Type Culture Collection (Manassas, VA, USA). ANBL-6, for 60 min then incubated with mouse monoclonal anti- OCI-MY1 and XG-1 were a kind gift from the Winthrop P phospho-STAT3 (pY705, Santa Cruz, ThermoFisher Scientific, Rockefeller Cancer Institute (Little Rock, AR, USA). HMCL were Scoresby, Victoria, Australia), mouse monoclonal anti-STAT3 grown and treated at densities between 2.0 and 5.0 Â 105 cells/ (Santa Cruz) or mouse monoclonal anti-a-tubulin (Sigma- ml in RPMI-1640 media (Gibco, Invitrogen, Mulgrave, Victoria, Aldrich) for 1–2 h at room temperature or overnight at 4 1C. Australia) supplemented with 10% heat-inactivated foetal The blots were washed three times for 15 min in 0.1% Tween- bovine serum (Lonza, Mt Waverley, Victoria, Australia) and 20/PBS, then incubated with secondary horseradish peroxidase 2mML-glutamine (Gibco, Invitrogen). IL-6-dependent cell lines tagged antibody (swine anti-rabbit Ig HRP or rabbit anti-mouse were cultured with 2–5 ng/ml IL-6 as required. All cells were Ig HRP (Dako, Campbellfield, Victoria, Australia)) for 1–2 h at cultured in a humidified incubator at 37 1C with 5% CO2. All room temperature before washing as above. Blots were HMCL were passaged 24 h before the experimental setup to visualised with Supersignal west pico ECL reagents (Pierce, ensure high viability and cycling. ThermoFisher Scientific).

Primary samples Intracellular FACS Primary MM samples were obtained from bone marrow Activation of the JAK/STAT, PI3K/AKT and Ras/MAPK pathways aspirates from relapsed and refractory MM patients, following was investigated using intracellular flow cytometry to measure written informed consent with approval from the Alfred Hospital the phosphorylation of STAT3 at tyrosine 705 (p-STAT3), AKT at Research and Ethics Committee, and isolated and treated as serine 473 (p-AKT), and extracellular signal-regulated kinase previously described.27 Briefly, bone marrow mononuclear cells (ERK)1/2 at threonine 202 and tyrosine 204 (p-ERK). HMCL were (BMMC) were isolated with Ficoll-Paque Plus (Amersham stimulated alone with 10 ng/ml IL-6±200 ng/ml insulin-like Biosciences, Rydalmere, New South Wales, Australia), washed growth factor-1 (IGF-1) or stimulated in CC with HS5 stromal in phosphate-buffered saline (PBS), and red blood cells were cells or primary BMSC, with or without CYT387 treatment. For lysed with NH Cl solution (8.29 g/l ammonium chloride, 4 CC, HS5 and primary BMSC were seeded into a 24-well plate at 0.037 g/l ethylene diamine tetraacetic acid, 1 g/l potassium 2 Â 105 cells/ml and allowed to establish for 4 h, after which bicarbonate). Cells were then washed again in PBS and HMCL prestained with CD38 or CD138 FITC (BD) were added. quantitated by haemocytometer. BMMC samples were then MM cells were stimulated alone (10 ng/ml IL-6, or 5 ng/ml IL-6 cultured in complete RPMI-1640 media (as above for HMCL) for and 100 ng/ml IGF-1), or in CC (direct CC with stroma or 24 h. Subsequently, the BMMC were plated at 5 Â 105 cells/ml transwell CC with stroma), with or without either 60 min of and were treated with CYT387 (5–50 mM), alone or (dependant CYT387 pretreatment or 15 min CYT387 co-treatment. After on cell numbers) in combination with bortezomib (5–40 nM)or stimulation±treatment, MM cells were harvested and fixed with melphalan (50–200 mM) for 24 and/or 48 h. Drug-induced 2% paraformaldehyde for 10–30 min, washed, then permeabi- MM-specific cell apoptosis was then compared with untreated lised with methanol overnight. Methanol was washed off and and vehicle controls by staining for CD45 FITC (BD, North the cells were resuspended in p-STAT3 PE (BD), p-AKT PE (BD) Ryde, New South Wales, Australia), CD38 PerCP-Cy5.5 (BD) or p-ERK (BD) and stained for 45–60 min at room temperature. and Apo 2.7 PE (Immunotech Beckman Coulter, Mt Waverley, Unbound antibody was washed off and the cells resuspended in Victoria, Australia) to determine apoptosis in CD45ÀCD38 þ 2% foetal bovine serum PBS and acquired by FACS. MM cells. Samples were subsequently analysed by fluorescence-activated cell sorting (FACS). Primary bone marrow stromal cells (BMSC) were also Proliferation and viability assays collected from patient BMMC and cells that adhered to the The viability and proliferation of CYT387-treated HMCL and - flask after an initial 24 h culture were cultured with continued untreated/vehicle controls were determined using various selection for adherent cells over several passages. Once cells methods as described previously.28 Proliferation was measured had expanded in culture, they were used in coculture (CC) to first using Celltiter 96 AQeous one solution cell proliferation stimulate MM cells in parallel to experiments utilising the HS5 assay MTS reagent (Promega, South Sydney, New South Wales, stromal cell line. Australia) on a panel of eight HMCL. Cells were cultured at 2.0 Â 105 cells/ml in 100 ml fresh media in 96-well plates for 24, Western blots 48 and 72 h with CYT387 (0.1–5 mM). 20 ml of MTS reagent HMCL were treated with CYT387 (1 or 2 mM) for 60 min and then (Promega) was added for the final 4 h of treatment, and the stimulated with 10 ng/ml IL-6 for 15 min. Protein lysates of plates were read at 490 nm using a Fluostar Optima plate reader CYT387-treated and -untreated HMCL were made with radio- (BMG Labtech, Mornington, Victoria, Australia). The viable cell immunoprecipitation buffer (50 mM Tris.HCl, pH 7.4, 150 mM numbers of a panel of five HMCL that were treated with NaCl, 1 mM phenylmethanesulfonyl fluoride, 1 mM ethylene CYT387, with and without IL-6 co-treatment, was also measured diamine tetraacetic acid, 5 mg/ml Aprotinin, 5 mg/ml Leupeptin, by trypan blue staining and haemocytometer count. 1% Triton X-100, 1% sodium deoxycholate and 0.1% SDS). The HMCL NCI-H929, OCI-MY1 and U266 were then Briefly, cells were incubated in radioimmunoprecipitation buffer selected for further analysis. Apoptosis of CYT387-treated cells on ice for 30–60 min before being centrifuged at 16 100 g for was assessed by FACS with Annexin-V and propidium iodide 20 min at 4 1C and the supernatant collected. Protein concen- (PI) staining. HMCL were treated for 24 or 72 h with 1 or 5 mM tration was quantified using DC Protein Assay (Bio-Rad, CYT387, then harvested and washed in Annexin Buffer (0.01 Gladesville, New South Wales, Australia) as per manufacturer’s HEPES, 0.14 M NaCl, 2.5 mM CaCl2, pH 7.4) and stained with instructions. Subsequently, 100 mg of each protein lysate was Annexin-V FITC (Biosource) made up in Annexin Buffer, for separated by 10% SDS-polyacrylamide gel electrophoresis and 30 min at room temperature. Unbound antibody was then blotted onto nitrocellulose (Hybond ECL, Amersham Bios- washed off with Annexin Buffer and cells were resuspended in ciences) using the Bio-Rad semi-dry transfer system. Membranes Annexin Buffer with 62.5 ng/ml PI (Sigma-Aldrich) and analysed were blocked with 5% skim milk powder 0.1% Tween-20/PBS by FACS.

Leukemia Anti-myeloma activity of the JAK inhibitor CYT387 KA Monaghan et al 1893 For the synergy experiments, HMCL were treated with CYT387 inhibits PI3K/AKT and Ras/MAPK signalling CYT387 in combination with bortezomib or melphalan for 24 JAK signalling kinases are involved in many cellular pathways, and 48 h before being harvested, and resuspended in FACS which led us to investigate the effect of CYT387 on IL-6 and Buffer (0.5% heat-inactivated foetal bovine serum in PBS) IGF-1-induced PI3K/AKT, and Ras/MAPK signalling in NCI- supplemented with 62.5 ng/ml PI (Sigma-Aldrich). Cells were H929, OCI-MY1 and U266 cells (Figure 1d). OCI-MY1 showed immediately analysed by FACS. The proportion of PI-positive distinct p-AKT activation after IL-6 and IGF-1 stimulation that cells was quantitated by subtracting the background death of was significantly reduced by CYT387 co-treatment. IL-6 and untreated cells. Single drug-treated cells were compared with IGF-1 stimulation induced p-ERK in U266 cells, which was combination-treated cells, and synergism was calculated using significantly inhibited by CYT387. The levels of p-AKT and Calcusyn software (Biosoft, Cambridge, UK). p-ERK showed only a small increase in response to IL-6 and IGF-1 stimulation in NCI-H929. IGF-1 can also induce JAK/ STAT signalling,29,30 because of this, we confirmed that CYT387 co-treatment was able to significantly inhibit p-STAT3 induced Cell cycling by simultaneous IL-6 and IGF-1 stimulation in all three HMCL The effect of CYT387 treatment on HMCL cell cycling was (Supplementary Figure 1C). measured after 24 and 72 h. CYT387 (0.5, 1 or 5 mM)-treated and -untreated HMCL were harvested, washed in PBS and resuspended in 100 ml PBS. Cells were fixed with 1 ml of cold 70% CYT387 inhibits proliferation in HMCL 1 ethanol while being vortexed. Tubes were stored at À20 C until Given that CYT387 can inhibit important signalling pathways in analysis. Once all samples were collected, tubes were HMCL, we next investigated the effect of CYT387 on prolifera- centrifuged at 500 g for 10 min, the supernatant was carefully tion and cycling of the HMCL. The effect of CYT387 (0.1–5 mM) removed and the cells were washed in 5 ml PBS. After the final on cell proliferation/viability was measured by MTS assay at 24, wash, cells were resuspended in 250–500 ml of PI/RNase 48 and 72 h (Figure 2a) on a panel of eight HMCL (IL-6 staining buffer (BD) and incubated in the dark at room non-responsive, predominantly CD45À phenotypeFLP-1, NCI- temperature for 15 min before being analysed by FACS. H929, OPM2, and RPMI-8226, and IL-6 responsive, predominantly CD45 þ phenotypeFANBL-6, OCI-MY1, U266 and XG-1). Six of the eight HMCL had a time- and dose-dependent Data analysis response to CYT387, with inhibition in some HMCL within 24 h. All FACS data was acquired on a BD FACSCalibur and data At 72 h, NCI-H929 and XG-1 were the most sensitive to CYT387 analysis done using Flowjo 7.6. Software (Treestar, Ashland, treatment, demonstrating that CYT387 can be a potent inhibitor OR, USA). All statistical analysis was done using GraphPad of both IL-6-responsive and non-responsive HMCL. Prism 5.03 software (La Jolla, CA, USA). Proliferation of HMCL cultured with CYT387 over 72 h was also assessed by quantitation of viable cell number by haemocytometer (Figure 2b). Three diverse HMCL (NCI-H929, OCI-MY1 and U266) representing both the IL-6-responsive and Results non-responsive phenotype were selected to determine the specific effect of CYT387 on absolute cell number. Because of the obvious JAK/STAT signalling is inhibited by CYT387 relationship between IL-6 signalling and the inhibition of JAK2 by IL-6 signalling through the JAK/STAT pathway is well charac- CYT387, the proliferation of three HMCL was measured with the terised in MM cells, with binding of IL-6 to its receptor inducing addition of IL-6 and/or CYT387. The three HMCL proliferated well JAK2 to phosphorylate STAT3. The ability of CYT387 to inhibit in complete media with or without supplementation with 10 ng/ml JAK2 was first confirmed by measuring the level of STAT3 IL-6, and CYT387 (0.5–1 mM) was able to reduce the proliferation phosphorylation by western blotting and FACS. HMCL (NCI- of HMCL in culture even in the presence of IL-6. A single dose of H929 and U266) were incubated with CYT387 for 1 h before CYT387 inhibited cell proliferation and resulted in a reduction in being stimulated with IL-6 for 15 min to induce p-STAT3. absolute cell number by 50% in NCI-H929 (1 mM), 50% in CYT387 (1–2 mM) inhibited the phosphorylation of STAT3 in IL-6 OCI-MY1 (0.5 mM) and 44% in U266 (1 mM)after72h,compared stimulated samples as demonstrated by FACS (Figure 1a) and with untreated controls. confirmed by western blotting (Figure 1b). As seen in Figures 1a and b, CYT387 was also able to reduce constitutive levels of p-STAT3 in U266 cells, as well as p-STAT3 induced by Treatment of HMCL with CYT387 results in an exogenous IL-6. Overall, the total STAT3 protein was unaffected accumulation of cells in G2/M phase of the cell cycle by IL-6 stimulation or CYT387 treatment. The anti-proliferative effects of CYT387 were further charac- Given the importance of the bone marrow microenvironment terised by evaluating the cell cycle of HMCL treated with in MM growth and survival, it was important to establish if CYT387. HMCL (NCI-H929, OCI-MY1 and U266) treated with CYT387 could similarly modulate signalling in MM cells in CC CYT387 (0.5–5 mM) showed a marked accumulation of cells in with BMSC. This was done with both immortalised BMSC (HS5) the G2/M phase of the cell cycle. This was most pronounced in and primary patient BMSC. In each case, 15 min of CC with NCI-H929 cells, in which there was a 1.5 fold increase in G2/M BMSC (with or without contact) was able to induce phosphor- in cells treated with 1 mM CYT387 and a two fold increase in ylation of STAT3 in the MM cells, whereas contemporaneous cells treated with 5 mM CYT387 (Figure 2c), compared with treatment with CYT387 dramatically reduced the amount of untreated or vehicle-treated controls after 24 and 72 h treatment. p-STAT3 in the HMCL (Figure 1c and Supplementary Figure 1A An additional polyploid population was found in the 5 mM and 1B), thus, demonstrating that CYT387 is able to prevent CYT387 treated samplesFsuggesting that CYT387 treatment STAT3 activation in MM cells induced by soluble factors within causes further aberration from the normal cell cycle of MM the bone marrow microenvironment, as well as the contact- cells. This correlated well with the CYT387-induced cytostasis mediated signalling provided to MM by BMSC. seen in Figure 2b.

Leukemia Anti-myeloma activity of the JAK inhibitor CYT387 KA Monaghan et al 1894

Figure 1 CYT387 prevents signalling downstream of IL-6 or CC stimulation. HMCL were incubated with or without CYT387 (1–2 mM) for 1 h before stimulation with 10 ng/ml IL-6 for 15 min. Cells were then harvested and p-STAT3 (pY705) was measured. (a) By intracellular FACS with the geometric mean fluorescence intensity measured and graphed (n ¼ 3, mean±s.e., stimulated cells±CYT387 were analysed using a one-way ANOVA with Tukey post-test *Po0.05, **Po 0.01). (b) By western blot for p-STAT3 (pY705), total STAT3 and a-tubulin as a loading control. (c)p- STAT3 was also induced in HMCL using IL-6, direct CC with HS5 immortalised BMSC or primary BMSC or transwell (TW) ‘soluble only’ CC with HS5. HMCL were fluorescently labelled with CD38 or CD138 and stimulated for 15 min with or without co-treatment with 2 mM CYT387. Representative plots of NCI-H929, (n ¼ 3, for NCI-H929, OCI-MY1 and U266). (d) NCI-H929, OCI-MY1 and U266 cells were starved overnight and stimulated with 5 ng/ml IL-6 and 100 ng/ml IGF-1 for 15 min, with or without co-treatment with 2 mM CYT387. p-AKT (pS473) and p-ERK1/2 (pT202/pY204) were measured by intracellular FACS with the geometric mean fluorescence intensity normalised to the untreated (UT) control and averaged (n ¼ 4, mean±s.e., stimulated cells±CYT387 were analysed using a one-way ANOVA with Tukey post-test *Po0.05).

CYT387 induces apoptosis in HMCL mononuclear cells was also investigated after 72 h, with only After demonstrating the potent effect of CYT387 on cell low levels of apoptosis seen (mean 14% death after 72 h with signalling and proliferation, we next investigated the induction 5 mM CYT387, data not shown). of apoptosis by Annexin-V/PI FACS staining in three HMCL (NCI-H929, OCI-MY1 and U266). An increased proportion of apoptotic cells was detected in all three HMCL, which was most CYT387 synergises with other anti-myeloma agents evident in NCI-H929 (Figure 3a), with a 21 and 52% increase in in HMCL apoptotic cells detected at 24 and 72 h, respectively, after To assess CYT387 as part of a combination therapy, NCI-H929, treatment with 5 mM CYT387 (Figure 3b). The toxicity OCI-MY1 and U266 were treated with a range of doses of of CYT387 (0.5–10 mM) on normal donor peripheral blood CYT387, bortezomib and melphalan to establish dose effect

Leukemia Anti-myeloma activity of the JAK inhibitor CYT387 KA Monaghan et al 1895

Figure 2 CYT387 inhibits HMCL proliferation. (a) CYT387 inhibits HMCL in a time- and dose-dependant manner. IL-6 phenotype HMCL (ANBL-6, OCI-MY1, U266 and XG-1) and non-IL-6 phenotype HMCL (LP-1, NCI-H929, OPM2 and RPMI-8226) were cultured for 24, 48 and 72 h untreated (UT), with CYT387 (0.1, 0.5, 1, 2.5 or 5 mM) or with vehicle (dimethyl sulfoxide). Cell proliferation was then determined by MTS assay (72 h data shown, n ¼ 3, mean±s.e.). (b) Treatment with CYT387 inhibits myeloma cell proliferation even in the presence of IL-6. Absolute cell numbers of viable cells were determined by haemocytometer counts of HMCL cultured alone (UT) with IL-6 (10 ng/ml), with CYT387 (0.5–1 mM), or with IL-6 and CYT387. Culture with CYT387 greatly decreased the proliferation of the HMCL over 72 h (treatment at time 0 only). Results represent the mean of three independent experiments±s.e. (c) CYT387 prevents cell cycling. HMCL were treated with CYT387 (1 mM or 5 mM) for 24 and 72 h then they were harvested and ﬁxed, and cell cycle analysed by FACS. Representative cell cycle plots of NCI-H929 UT or 5 mM CYT387 for 24 or 72 h, with mean of four independent experiments±s.e. of cycling cells in G2/M phase of the cell cycle.

curves for each drug before combining with CYT387 and CD38 þ CD45À MM cell populations were assessed for measuring the synergy using Calcusyn software. Dose effect apoptosis by flow cytometry. Six patients were treated and curves generated for each compound (melphalan and bortezo- apoptosis was seen in between 5 and 59% of MM cells treated mib data not shown) show CYT387 (0.5–10 mM) induced with 20 mM CYT387 after 48 h (Figure 5a). The effect of CYT387 apoptosis in three HMCL in a time- and dose-dependent manner (10–20 mM), in combination with melphalan (50–200 mM) and (Figure 4a). CYT387 displayed synergism with both bortezomib bortezomib (10–40 nM), was also investigated. CYT387 was seen (1–4 nM) and melphalan (25–50 mM), but with variations in the to synergise with melphalan in two of three patients and synergy degree of synergy seen with different drug dosages and analysis was also observed with bortezomib in some patients/doses timepoints (Figure 4b). No distinct pattern of specific drug dose, (Figure 5b). timepoint or cell line was evident. CYT387 was also combined with the second-generation proteasome inhibitor NPI-0052, as well as other conventional therapies cisplatin, dexamethasone Discussion and etoposide, which showed mixed synergistic and antagonis- tic results (data not shown). Multiple lines of evidence have confirmed the role of IL-6/JAK/ STAT signalling in MM, including experiments demonstrating the IL-6 dependence of some MM cells, the upregulation of CYT387 induces apoptosis as a single agent and proliferation of MM cells with IL-6, the inhibition of drug- synergises with bortezomib and melphalan in primary induced apoptosis by IL-6, and most important to this MM cells investigation, the direct induction of apoptosis in MM cells by The effect of CYT387 on ex-vivo primary MM cells was also inhibition of the IL-6/JAK/STAT pathway. These data and the investigated, both as a single agent and as part of combination abundance of IL-6 in the bone marrow microenvironment therapy. MM patient BMMC were cultured with various doses of makes JAK/STAT signalling a rational target for inhibition with CYT387 (5–50 mM) for 24 and 48 h, after which the new chemotherapeutics and is supported by preliminary in vitro

Leukemia Anti-myeloma activity of the JAK inhibitor CYT387 KA Monaghan et al 1896 logical focus for preliminary investigation, it must be stressed that such target cell populations represents only a subset of MM cells. Furthermore, patients may demonstrate mixed populations of both CD45À and CD45 þ MM cells.15 Importantly, we have demonstrated the effectiveness of CYT387 against NCI-H929, a HMCL considered to have an IL-6 non-responsive phenotype, suggesting that CYT387 will be effective against a range of MM phenotypes, whereas others20 have reported only limited success against CD45À MM, using alternative JAK inhibitors. Our data demonstrating synergy between CYT387 and bortezomib, or melphalan against several HMCL and primary MM tumour cells confirms previously published work.20 Available data suggests that the inhibition of JAK in MM cells may have downstream effects other than the direct inhibition of the JAK/STAT pathway. The importance of IL-6 in MM cell survival has been well characterised in terms of the JAK/STAT pathway, but there is now increasing evidence of IL-6-induced PI3K/AKT activation19,33 and Ras/MAPK activation21,33–36 in various HMCL. Given the established importance of both the PI3K/AKT and Ras/MAPK pathways in addition to the JAK/STAT pathway in MM, small molecule inhibitors that could modulate all three could have enormous clinical potential. Investigating the effects of JAK inhibition on the PI3K/AKT and Ras/MAPK pathways has yielded contrasting results. AZD1480 and AG490 have been shown to reduce IL-6-induced activation of Ras/ MAPK,21,23 but AG490 in the hands of others showed no decrease in IL-6-induced Ras/MAPK activation.34 The JAK inhibitor INCB20 could inhibit IL-6-induced Ras/MAPK activation in MM.1S cells, but did not affect constitutive Ras/MAPK activation in INA-6 cells.19 The inhibition of JAK with AG490 could also abrogate AKT activation in phosphatase and tensin homolog-mutated OPM2 MM cells.33 Similarly, INCB20 could also inhibit IL-6-induced p-AKT, but had no effect on IGF-1- induced p-AKT in INA-6 cells.19 Variability in the available data may be the result of differences in inhibitors and heterogeneity amongst HMCL that are commonly studied. In our study, there was a significant reduction in IL-6 and IGF-1-induced p-ERK in U266 cells, as well as a dramatic reduction of IL-6 and IGF-1- induced p-AKT in OCI-MY1 cells, supporting the data of others Figure 3 CYT387 induces apoptosis in HMCL. (a) Representative Annexin-V and PI plots of NCI-H929. UT, vehicle (dimethyl sulfoxide) that suggests JAK inhibition may have broader anti-MM activity treated, 24 h 5 mM CYT387 treatment and 72 h 5 mM CYT387 treatment. than would be initially expected. The inhibition of IL-6-induced (b) Proportion of apoptotic (Annexin-V þ or PI þ ) cells after CYT387 JAK/STAT and PI3K/AKT signalling may also result in a reduction treatment compared with UT. Data shown is the mean of four in IL-6 receptor expression on the surface of MM cells,33 ± independent experiments s.e. which could also lead to a reduction in the pro-survival effects of IL-6. The profound anti-proliferative effect of CYT387 on various studies that have demonstrated MM cell apoptosis induction via HMCL after a single dose is an important demonstration of its siRNA targeting of the JAK/STAT pathway.31 Furthermore, the effectiveness. Furthermore, the ability of CYT387 to inhibit signal transduction role of JAKs in other important pathways HMCL growth even in the presence of exogenous IL-6, which has (reviewed by Rane and Reddy5) and the pro-survival effects of been shown many times as a mediator of drug resistance11–13 is JAK mutations in other haematological malignancies1–4 have noteworthy. This significant effect of JAK inhibition may be the already demonstrated the therapeutic potential of JAK inhibition. result of HMCL having some dependence on JAK/STAT for The therapeutic challenge is to inhibit MM cells in the presence proliferative signals, or more likely, the involvement and of IL-6 or BMSC. Here, we have expanded upon previous work subsequent inhibition of alterative signalling pathways men- evaluating JAK inhibition in MM by studying a broader range of tioned above. The effect on proliferation is also seen in the cell HMCL, by demonstrating that CYT387 can inhibit JAK/STAT cycling analysis, which demonstrates that CYT387 can prevent activation in the context of CC models and by demonstrating the cell cycling. Also of interest was the additional polyploidy impact of CYT387 on primary MM tumour cells. population (8n) induced by CYT387 treatment, which may It has been hypothesised that in the earlier stages of MM, the suggest a role for JAK signalling leading to cell cycle regulation malignant cells are predominantly CD45 þ , whereas in more or may be the result of CYT387 inhibition of other kinases advanced, drug-resistant disease, CD45À MM cells predomi- involved in cell cycle, such as aurora B, aurora C, cyclin A or nate.32 Moreover, CD45À MM cells are considered less IL-6 cyclin B as described by Pardanani et al.17 A direct or responsive and express fewer IL-6 receptors than CD45 þ MM downstream effect of JAK inhibitors on cell cycling proteins is cells.15 Other studies of JAK inhibition have focused predomi- supported by studies on other JAK inhibitors, Scuto et al.23 found nantly on CD45 þ IL-6 responsive HMCL, and although this is a AZD1480 inhibited cyclin D2 in two HMCL.

Leukemia Anti-myeloma activity of the JAK inhibitor CYT387 KA Monaghan et al 1897

Figure 4 CYT387 synergises with melphalan and bortezomib in HMCL. (a) Dose effect curves of NCI-H929, OCI-MY1 and U266 after 24 and 48 h treatment CYT387 treatment (0.5–10 mM) as determined by the proportion of PI þ cells minus background death (untreated). Mean of four independent experiments±s.e. (b) CYT387 (1–5 mM) in combination with melphalan (25–50 mM) or bortezomib (1–4 nM) shows synergism. Synergy was measured using a combination index calculated by Calcusyn software, where values less than 1 represent synergism, plotted against the fraction of cells killed with various dose/ratios of the drugs. Synergy is seen between melphalan and CYT387 at a range of doses/ratios/cell lines and timepoints. Bortezomib and CYT387 demonstrated synergistic or nearly additive in 18/24 combinations. Synergism was calculated from dose effect curves of the mean of four independent experiments.

Figure 5 CYT387 induces apoptosis in primary samples as a single agent or in combination with melphalan and bortezomib. (a) Proportion of apoptotic (Apo 2.7 þ ) CD38 þ CD45À primary patient myeloma cells after 48 h CYT387 treatment (n ¼ 6). (b) Synergy between CYT387 (10–20 mM) and melphalan (50–200 mM) or bortezomib (10–40 nM) after 24 h treatment as determined by Calcusyn software in primary patient CD38 þ CD45À cells.

The induction of apoptosis in MM cells in primary marrow studies inhibiting IL-6 signalling have failed to demonstrate any CCs using clinically relevant low-micromolar doses is a critical convincing evidence of apoptosis when MM cells were treated demonstration of the potential of CYT387. In contrast, other in the presence of BMSC.35 The investigators suggested that this

Leukemia Anti-myeloma activity of the JAK inhibitor CYT387 KA Monaghan et al 1898 might represent MM cell independence from IL-6 in the rearrangement: clinical implications. Cancer Res 1993; 53: presence of BMSC. Our findings with CYT387 could be 5320–5327. interpreted as either refuting the latter or alternatively being 10 Zhang XG, Gaillard JP, Robillard N, Lu ZY, Gu ZJ, Jourdan M et al. consistent with the capacity of JAK signalling inhibition to Reproducible obtaining of human myeloma cell lines as a model for tumor stem cell study in human multiple myeloma. Blood interrupt non-IL6-mediated survival pathways. Consistent with 1994; 83: 3654–3663. the latter was that the primary MM samples treated with CYT387 11 Cheung WC, Van Ness B. The bone marrow stromal microenvir- were representative of autologous whole marrow CCs with the onment influences myeloma therapeutic response in vitro. CD38 þ CD45À MM cells making up between 4 and 67% of Leukemia 2001; 15: 264–271. treated cells. Therefore, the demonstration that CYT387 was 12 Moreaux J, Legouffe E, Jourdan E, Quittet P, Reme T, Lugagne C able to induce apoptosis of these heavily treated MM cells was et al. BAFF and APRIL protect myeloma cells from apoptosis induced by interleukin 6 deprivation and dexamethasone. Blood particularly encouraging. 2004; 103: 3148–3157. Our pre-clinical evaluation of CYT387 has demonstrated 13 Perez LE, Parquet N, Shain K, Nimmanapalli R, Alsina M, promising anti-myeloma effects of an orally bioavailable Anasetti C et al. Bone marrow stroma confers resistance to Apo2 compound already through phase I clinical trials. We show ligand/TRAIL in multiple myeloma in part by regulating c-FLIP. the inhibition of important signalling pathways in MM cells, the J Immunol 2008; 180: 1545–1555. subsequent reduction in cell viability, proliferation and corre- 14 Klein B, Zhang XG, Jourdan M, Content J, Houssiau F, Aarden L et al. Paracrine rather than autocrine regulation of myeloma-cell sponding inhibition of the cell cycle. Furthermore, we present growth and differentiation by interleukin-6. Blood 1989; 73: 517–526. specific evidence of the induction of apoptosis in both primary 15 Hata H, Xiao H, Petrucci MT, Woodliff J, Chang R, Epstein J. MM cells and HMCL. The synergy demonstrated by CYT387 Interleukin-6 gene expression in multiple myeloma: a character- with common myeloma therapies bortezomib and melphalan istic of immature tumor cells. Blood 1993; 81: 3357–3364. make it a very attractive compound for further study in 16 Tyner JW, Bumm TG, Deininger J, Wood L, Aichberger KJ, Loriaux the clinic. MM et al. CYT387, a novel JAK2 inhibitor, induces hematologic responses and normalizes inflammatory cytokines in murine myeloproliferative neoplasms. Blood 2010; 115: 5232–5240. Conflict of interest 17 Pardanani A, Lasho T, Smith G, Burns CJ, Fantino E, Tefferi A. CYT387, a selective JAK1/JAK2 inhibitor: in vitro assessment of kinase selectivity and preclinical studies using cell lines and Katherine A Monaghan, Tiffany Khong and Andrew Spencer primary cells from polycythemia vera patients. Leukemia 2009; 23: declare no potential conflict of interest. Christopher J Burns is an 1441–1445. employee of YM BioSciences, and holds equity in the company. 18 Burns CJ, Bourke DG, Andrau L, Bu X, Charman SA, Donohue AC et al. Phenylaminopyrimidines as inhibitors of Janus kinases (JAKs). Bioorg Med Chem Lett 2009; 19: 5887–5892. Acknowledgements 19 Burger R, Le Gouill S, Tai YT, Shringarpure R, Tassone P, Neri P et al. Janus kinase inhibitor INCB20 has antiproliferative and CYT387 was kindly provided by YM BioSciences, Australia. This apoptotic effects on human myeloma cells in vitro and in vivo. Mol Cancer Ther 2009; 8: 26–35. study was supported by part funding from the Multiple Myeloma 20 Li J, Favata M, Kelley JA, Caulder E, Thomas B, Wen X et al. Research Foundation. INCB16562, a JAK1/2 selective inhibitor, is efficacious against multiple myeloma cells and reverses the protective effects of cytokine and stromal cell support. Neoplasia (New York, NY) 2010; 12: 28–38. References 21 De Vos J, Jourdan M, Tarte K, Jasmin C, Klein B. JAK2 tyrosine kinase inhibitor tyrphostin AG490 downregulates the mitogen- 1 Kralovics R, Passamonti F, Buser AS, Teo SS, Tiedt R, Passweg JR activated protein kinase (MAPK) and signal transducer and et al. A gain-of-function mutation of JAK2 in myeloproliferative activator of transcription (STAT) pathways and induces apoptosis disorders. N Engl J Med 2005; 352: 1779–1790. in myeloma cells. Br J Haematol 2000; 109: 823–828. 2 Levine RL, Wadleigh M, Cools J, Ebert BL, Wernig G, Huntly BJ 22 Alas S, Bonavida B. Inhibition of constitutive STAT3 activity et al. Activating mutation in the tyrosine kinase JAK2 in sensitizes resistant non-Hodgkin’s lymphoma and multiple mye- polycythemia vera, essential thrombocythemia, and myeloid loma to chemotherapeutic drug-mediated apoptosis. Clin Cancer metaplasia with myelofibrosis. Cancer cell 2005; 7: 387–397. Res 2003; 9: 316–326. 3 Scott LM, Tong W, Levine RL, Scott MA, Beer PA, Stratton MR et al. 23 Scuto A, Krejci P, Popplewell L, Wu J, Wang Y, Kujawski M et al. JAK2 exon 12 mutations in polycythemia vera and idiopathic The novel JAK inhibitor AZD1480 blocks STAT3 and FGFR3 erythrocytosis. N Engl J Med 2007; 356: 459–468. signaling, resulting in suppression of human myeloma cell growth 4 Mullighan CG, Zhang J, Harvey RC, Collins-Underwood JR, and survival. Leukemia 2011; 25: 538–550. Schulman BA, Phillips LA et al. JAK mutations in high-risk 24 Pedranzini L, Dechow T, Berishaj M, Comenzo R, Zhou P, Azare J childhood acute lymphoblastic leukemia. Proc Natl Acad SciUSA et al. Pyridone 6, a pan-Janus-activated kinase inhibitor, induces 2009; 106: 9414–9418. growth inhibition of multiple myeloma cells. Cancer Res 2006; 66: 5 Rane SG, Reddy EP. Janus kinases: components of multiple 9714–9721. signaling pathways. Oncogene 2000; 19: 5662–5679. 25 Ferrajoli A, Faderl S, Van Q, Koch P, Harris D, Liu Z et al. WP1066 6 French JD, Walters DK, Jelinek DF. Transactivation of gp130 in disrupts Janus kinase-2 and induces caspase-dependent apoptosis myeloma cells. J Immunol 2003; 170: 3717–3723. in acute myelogenous leukemia cells. Cancer Res 2007; 67: 7 Berger LC, Hawley TS, Lust JA, Goldman SJ, Hawley RG. Tyrosine 11291–11299. phosphorylation of JAK-TYK kinases in malignant plasma cell lines 26 Pardanani A, George G, Lasho T, Hogan WJ, Litzow MR, Begna K growth-stimulated by interleukins 6 and 11. Biochem Biophys Res et al. A Phase I/II Study of CYT387, An Oral JAK-1/2 Inhibitor, Commun 1994; 202: 596–605. In Myelofibrosis: Significant Response Rates In Anemia, Spleno- 8 Gomez-Benito M, Balsas P, Carvajal-Vergara X, Pandiella A, megaly, and Constitutional Symptoms. Blood; 52nd ASH Annual Anel A, Marzo I et al. Mechanism of apoptosis induced by IFN- Meeting and Exposition 6/12/2010; Orlando, Florida, USA, 116. alpha in human myeloma cells: role of Jak1 and Bim and Abstract 460. p. 206. potentiation by rapamycin. Cell Signal 2007; 19: 844–854. 27 Khong T, Sharkey J, Spencer A. The effect of azacitidine on 9 Jelinek DF, Ahmann GJ, Greipp PR, Jalal SM, Westendorf JJ, interleukin-6 signaling and nuclear factor-kappaB activation and Katzmann JA et al. Coexistence of aneuploid subclones within a its in vitro and in vivo activity against multiple myeloma. myeloma cell line that exhibits clonal immunoglobulin gene Haematologica 2008; 93: 860–869.

Leukemia Anti-myeloma activity of the JAK inhibitor CYT387 KA Monaghan et al 1899 28 Monaghan K, Khong T, Smith G, Spencer A. CYT997 Causes myeloma: clinical and biological correlations. Leukemia 2005; apoptosis in human multiple myeloma. Invest New Drugs 2011; 19: 1466–1470. 29: 232–238. 33 Thabard W, Collette M, Mellerin MP, Puthier D, Barille S, 29 Zong CS, Chan J, Levy DE, Horvath C, Sadowski HB, Wang LH. Bataille R et al. IL-6 upregulates its own receptor on some human Mechanism of STAT3 activation by insulin-like growth factor I myeloma cell lines. Cytokine 2001; 14: 352–356. receptor. J Biol Chem 2000; 275: 15099–15105. 34 Puthier D, Bataille R, Amiot M. IL-6 up-regulates mcl-1 in human 30 Gual P, Baron V, Lequoy V, Van Obberghen E. Interaction of Janus myeloma cells through JAK / STAT rather than ras/MAP kinase kinases JAK-1 and JAK-2 with the insulin receptor and the insulin- pathway. Eur J Immunol 1999; 29: 3945–3950. like growth factor-1 receptor. Endocrinology 1998; 139: 884–893. 35 Chatterjee M, Honemann D, Lentzsch S, Bommert K, Sers C, 31 Chatterjee M, Stuhmer T, Herrmann P, Bommert K, Dorken B, Herrmann P et al. In the presence of bone marrow stromal cells Bargou RC. Combined disruption of both the MEK/ERK and the human multiple myeloma cells become independent of the IL-6/ IL-6R/STAT3 pathways is required to induce apoptosis of multiple gp130/STAT3 pathway. Blood 2002; 100: 3311–3318. myeloma cells in the presence of bone marrow stromal cells. 36 Ferlin M, Noraz N, Hertogh C, Brochier J, Taylor N, Klein B. Blood 2004; 104: 3712–3721. Insulin-like growth factor induces the survival and proliferation of 32 Kumar S, Rajkumar SV, Kimlinger T, Greipp PR, Witzig TE. myeloma cells through an interleukin-6-independent transduction CD45 expression by bone marrow plasma cells in multiple pathway. Br J Haematol 2000; 111: 626–634.

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

Leukemia