IntroductionT-cell acute lymphoblastic leukaemia (T-ALL) is an aggressive haematological malignancy resulting from the transformation of T-cell progenitors and occurs in about 15% of paediatric and 25% of adult ALL cases. The prognosis of this disease has substantially improved due to the introduction of intensified chemotherapy, with cure rates in modern protocols reaching over 75% in children and about 50% in adults (Goldberg, et al 2003, Marks, et al 2009). Nevertheless, the clinical outcome of T-ALL patients with primary resistant or relapsed disease still remains poor (Goldberg, et al 2003, Oudot, et al 2008, Schrappe, et al 2012), and aggressive treatment regimens are often associated with severe acute toxicities and long-term side effects, including the development of secondary tumours later in life. Thus, development of more effective and less toxic anti-leukaemic drugs have been the focus of research efforts in T-ALL (Aifantis, et al 2008).
We have previously demonstrated that tyrosine kinase 2 (TYK2) activation contributes to aberrant survival of human T-ALL cells (Sanda, et al 2013). TYK2 is a member of the Janus-activated kinase (JAK) tyrosine kinase family and our report was the first to implicate TYK2 as an oncogene in T-ALL. Indeed, our gene knockdown experiments showed TYK2 dependency in 14 of 16 (88%) T-ALL cell lines and 5 of 8 (63%) patient-derived T-ALL cells tested (Sanda, et al 2013). In addition, pharmacological inhibition of TYK2 with a small-molecule pan-JAK inhibitor, JAK inhibitor I, induced apoptosis in multiple human T-ALL cell lines (Sanda, et al 2013). Our results also showed that TYK2 promotes the survival of T-ALL cells by upregulating the anti-apoptotic protein BCL2 through phosphorylation and activation of STAT1 (Sanda, et al 2013). These findings establish TYK2 as a promising molecular target for the treatment of T-ALL.
Each of the four JAK family kinases (JAK1, JAK2, JAK3 and TYK2) associates with a distinct set of receptors. Receptor activation triggers receptor intracellular domain phosphorylation, creating docking sites for signal transducers and activators of transcription (STAT) proteins (Ghoreschi, et al 2009, Leonard and O'Shea 1998, Liu, et al 1998). STAT proteins are subsequently phosphorylated by the JAKs, form homodimers or heterodimers, translocate to the cell nucleus, and mediate gene transcription. TYK2 is involved in receptor signalling mediated by inflammatory cytokines including the type-I interferons, interleukin (IL)-12 and IL-23 (Ihle, et al 1995, Leonard and O'Shea 1998, Liu, et al 1998). Therefore TYK2 kinase inhibitors are being developed as potential therapeutics for auto-immune inflammatory diseases, such as psoriasis and inflammatory bowel diseases (Liang, et al 2013a, Liang, et al 2013b). However, generating kinase inhibitors with a high degree of TYK2 selectivity has posed a significant challenge due to the high sequence homology of the active site among the JAK family kinases. TYK2 specificity is important for clinical application of TYK2 kinase inhibitors, because Tyk2 knockout mice are viable with normal blood cell counts (Ghoreschi, et al 2009, Karaghiosoff, et al 2000, Shimoda, et al 2000), whereas deficiency of Jak3 results in severe combined immunodeficiency in mice, and Jak1 or Jak2 knockout mice show perinatal lethality (Ghoreschi, et al 2009). A loss-of-function mutation in the TYK2 gene was identified in a patient with hyperimmunoglobulin E syndrome (Minegishi, et al 2006), a primary immunodeficiency characterized by elevated serum immunoglobulin E. However, an additional seven individuals with homozygous null mutations of TYK2 did not have hyperimmunoglobulin E syndrome, but rather exhibited increased susceptibility to mycobacteria or viral infections due to impaired responses to IL-12 and IFN-α/β (Kreins, et al 2015). Thus, genetic evidence suggests that pharmacological inhibition of TYK2 should not result in acute toxicity in human patients, but careful monitoring for viral or mycobacterial infections would be warranted in patients treated for prolonged periods.
Here we describe the identification and characterization of the novel highly potent and selective TYK2 inhibitor, NDI-031301. We show that this inhibitor has significant anti-leukaemic activity against human T-ALL cell lines in vitro due to its ability to efficiently induce apoptosis in these cells. The inhibitor is orally bioavailable and xenograft studies with a human T-ALL cell line KOPT-K1 showed that it has significant anti-tumour activity in vivo without appreciable side effects. Thus, our preclinical findings warrant further testing of NDI-031301 or related compounds as promising drugs for targeted treatment of T-ALL, with the expectation that effective inhibition of the TYK2 pathway would eventually be combined with other anti-leukaemic agents in novel strategies of combination therapy.
Materials and methodsReagentsNDI-031301 was provided by Nimbus Therapeutics (Cambridge, MA, USA). Tofacitinib baricitinib, trametinib, SP600125, and SB203580 were purchased from Selleck Chemicals (Houston, TX, USA).
Cell cultureAll human T-ALL cell lines (KOPT-K1, DU.528, HPB-ALL and SKW-3) were obtained from the American Type Culture Collection (ATCC; Manassas, VA, USA) or Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH (DSMZ; Braunschweig, Germany). Ba/F3 derivatives expressing various oncogenic fusion kinases, namely, TEL-ABL (ETV6-ABL1), TEL-JAK1 (ETV6-JAK1), TEL-JAK2 (ETV6-JAK2), and TEL-JAK3 (ETV6-JAK3), were obtained from Dr. Richard Morrigl and were described previously (Lacronique, et al 2000). Ba/F3 cells transformed by TEL-TYK2 were generated with the pBabe-Neo retroviral vector expressing ETV6 (TEL)-TYK2 cDNA (Lacronique, et al 2000), which was obtained from Dr. Olivier A. Bernard. These cells were maintained in RPMI-1640 medium (GIBCO, Waltham, MA, USA) supplemented with 10% fetal bovine serum (Sigma-Aldrich, St. Louis, MO, USA) and 1% penicillin/streptomycin (Invitrogen, Waltham, MA, USA).
shRNA knockdown experimentsAll shRNA constructs cloned into the lentiviral vector pLKO.1-puro were obtained from the RNAi Consortium (Broad Institute, Cambridge, MA, USA). Target sequences for each shRNA are listed in Table SI. Each construct was cotransfected into HEK293T cells with the packaging plasmid psPAX2 and envelope plasmid pMD2.G using FuGENE 6 reagent (Roche, Basel, Switzerland). Supernatants containing the lentivirus were collected and filtered through a 0.45-μm cellulose acetate membrane filter. T-ALL cell lines were then infected with lentivirus in the presence of polybrene (8 μg/ml) and HEPES (10 mM) by centrifugation at 2,500 rpm for 1.5 h at 30°C, and the infected cells were selected by puromycin for at least 36 h.
Cell viability analysisThe Cell Titer Glo assay (Promega, Fitchburg, WI, USA) was used to assess relative cell viability upon treatment. Cells were plated at a density of 5000 - 10000 cells per well in a 96-well plate and incubated with dimethylsulfoxide (DMSO) or increasing concentrations of inhibitor. The relative cell viability was measured after different treatment intervals and reported as a percentage of the DMSO control. The concentration of inhibitor required for 50% inhibition of cell viability (IC50) was determined with GraphPad Prism 6.02 (GraphPad Software, La Jolla, CA, USA).
Apoptosis and cell cycle analysisBoth the terminal deoxynucleotidyl transferase (TdT) dUTP nick-end labeling (TUNEL) assay and propidium iodide (PI) staining were performed with the APO-BrdU™ TUNEL assay kit (Invitrogen, Waltham, MA, USA), according to the manufacturer’s recommendation. Briefly, 2 × 106 cells of each treated sample were fixed with 1% paraformaldehyde in phosphate-buffered saline (PBS) for 15 min on ice, washed in PBS and incubated in 70% ethanol at −20°C overnight. After washing and incubation in DNA labeling solution containing TdT and bromodeoxyuridine (BrdU) triphosphates for 4 h at 37°C, the cells were washed and incubated in staining buffer containing an Alexa Fluor 488 dye-labelled anti-BrdU antibody for 30 min of incubation at room temperature, followed by addition of a PI/RNase mixture. After 30 min incubation at room temperature, TUNEL positivity and cell-cycle distribution were analysed with a BD FACSCalibur instrument (BD Biosciences, San Jose, CA, USA). The threshold for TUNEL positivity was determined as the maximal TUNEL signal observed in the untreated cells.
Western blotsWhole-cell lysates were prepared in radioimmunoprecipitation assay (RIPA) buffer (Cell Signaling Tech, Danvers, MA, USA) with FOCUS™ ProteaseArrest™ (G-Biosciences, St. Louis, MO, USA) and Phosphatase Inhibitor Cocktail Set II (EMD Millipore, Billerica, MA, USA). Immunoblotting was performed with specific antibodies to TYK2, STAT1, phospho-STAT1 (Tyr701), STAT3, phospho-STAT3 (Tyr705), STAT5, phospho-STAT5 (Tyr694), STAT6, phospho-STAT6 (Tyr641), AKT, phospho-AKT (Thr308), ERK1/2 (also termed MAPK3/MAPK1), phospho-ERK1/2 (Thr202/Tyr204), SAPK/JNK (also termed MAPK9/MAPK8) (MAPK14), phospho-SAPK/JNK (Thr183/Tyr185), p38 MAPK (also termed MAPK14), phospho-p38 MAPK (Thr180/Tyr182), BCL2, BCLXL (also termed BCL2L11), MCL1, PARP and α-tubulin (all from Cell Signaling Tech).
T-ALL xenograft model2 × 106 luciferase-expressing KOPT-K1 cells were injected into the tail vein of 7-week-old female NOD-SCID-IL2Rcγnull (NSG) mice (The Jackson Laboratory, Bar Harbor, ME, USA). The tumour burden was measured by bioluminescence imaging (BLI) every 7 days, using an IVIS Spectrum system (Caliper Life Sciences, Santa Clara, CA, USA). For each BLI time point, cancer bioluminescence was visualized by subcutaneous injection of d-luciferin (Promega) in PBS at 75 mg/kg. After tumour engraftment was established by BLI measurement, mice were divided into 2 groups and treated by oral gavage either with vehicle control (20% HP-β-CD) or NDI-031301 at 100 mg/kg bid (twice daily) for 29 days. After completion of the treatment, 4 mice from each treatment group were sacrificed to obtain femur and spleen. The bone marrow cells were extracted from the femurs by crushing the bone in medium supplemented with 10% fetal bovine serum. The presence of human leukaemia cells was assessed by flow cytometric analysis following staining with fluorescein isothiocynate (FITC)-conjugated anti-human CD45 antibody (Biolegend, San Diego, CA, USA). The femur tissues were also fixed in 10% formalin, sectioned, paraffin-embedded and stained with haematoxylin and eosin. Stained slides were viewed and photographed using an Olympus BX41 microscope and Q-color5 digital camera (Olympus, Center Valley, PA, USA) and imaged with Adobe Photoshop CS4 software (Adobe, San Jose, CA, USA). Survival of the vehicle and NDI-031301 treated mice was defined from initiation of therapy until a moribund state was reached, and was analysed by the Kaplan-Meier method.
Statistical analysisStatistical significance in assays with identical cell lines was assessed with Student’s t test (two-tailed). GraphPad Prism 6.02 (GraphPad Software) was used for all statistical analyses. The Calcusyn 2.0 software (Biosoft, Cambridge, UK) was used to analyse combined drug effects and produce the isobolograms normalized to the IC50 of each inhibitor.
ResultsNDI-031301 is a potent and selective inhibitor of the TYK2 tyrosine kinaseUsing a series of proprietary JAK family co-crystal structures in combination with a quantitative computational model to predict selectivity over the JAK family members, we designed a series of selective TYK2 inhibitors (Wang, et al 2015). Elaboration of multiple hit series led to the identification of NDI-031301 as an inhibitor that is preferentially active against TYK2 compared to other JAK family kinases. Preclinical studies have illustrated that NDI-031301 potently inhibits targeted immune cell cytokine signalling and demonstrates outstanding efficacy in ameliorating disease in mouse models of psoriasis (Miao, et al 2015). The profile of NDI-031301, including the results of the biochemical TYK2 kinase assay, biochemical JAK family selectivity, plasma protein binding, and pharmacokinetic parameters in mice after oral administration are shown in Table I.
To assess the cellular activity and selectivity of NDI-031301 for TYK2, we tested its ability of this drug to inhibit the proliferation of Ba/F3 cells transformed by each of a series of constitutively active JAK family kinases (TEL-JAK1, TEL-JAK2, TEL-JAK3 and TEL-TYK2) or by a more distantly related tyrosine kinase, TEL-ABL (Fig 1A). NDI-031301 most strongly inhibited the growth of TYK2-transfomed cells, with the IC50 of 1.583 μM following 72 h of exposure, whereas Ba/F3 cells transformed by other tyrosine kinases showed decreased sensitivity to this drug (Fig 1D). Tofacitinib (JAK1, JAK2 and JAK3 inhibitor) and baricitinib (JAK1 and JAK2 inhibitor) were less effective inhibitors than NDI-031301 on the proliferation of TYK2-transformed cells, with IC50 values of 23.89 μM (tofacitinib) and 3.98 μM (baricitinib) after 72 h of treatment, while these drugs were highly active against Ba/F3 cells transformed by the other JAK family kinases (Fig 1B, C and D). These results demonstrate that NDI-031301 is a selective inhibitor of TYK2 tyrosine kinase with activity in cells transduced and rendered IL-3-independent by an activated form of this kinase.
NDI-031301 induces growth inhibition and apoptosis in human T-ALL cell linesTo assess the anti-tumour potency of NDI-031301 against T-ALL, we tested the inhibitory effect of this drug on the proliferation of multiple T-ALL cell lines (Fig 2A). This agent induced remarkable cytotoxicity in each of 4 human T-ALL cell lines representing different molecular subtypes of the disease (DU.528, KOPT-K1, HPB-ALL and SKW-3), with IC50 values of 0.8186 – 2.38 μM after 72 h of exposure (Fig 2D). This inhibitory effect was more potent than that of either tofacitinib or baricitinib, except that DU.528 cells were highly sensitive to all JAK kinase inhibitors tested in this study (Fig 2B, C and D). To gain insight into the cytotoxic mechanism triggered by NDI-031301, we next monitored changes in the relative numbers of KOPT-K1 cells treated with NDI-031301 over a 5-day period. As shown in Figure 2E, DMSO-treated KOPT-K1 cells demonstrated exponential growth while NDI-031301 treatment demonstrated dose-dependent growth suppression, with 1 μM NDI-031301 causing minor growth suppression of KOPTK1 cells and 3 μM NDI-031301 treatment reducing total cell number 8-fold on the fifth day of treatment, as compared to cell numbers at Day 0. Moreover, we demonstrated that NDI-031301 caused a profound apoptotic response in the KOPT-K1 cells, as shown by the appearance of 31.1% apoptotic cells following treatment with 3μM NDI-031301 for 48 h in a TUNEL apoptotic assay (Fig 2F). An increased level of apoptosis in KOPT-K1 cells was also observed at 24 h after treatment with 3 μM NDI-031301 by flow cytometric analysis following Annexin V-FITC and PI double staining (Fig S1).
TYK2 inhibition leads not only to suppression of STAT1 phosphorylation, but also activation of MAPK signalling pathways in KOPT-K1 cellsOur previous results indicated that TYK2 acts to upregulate levels of the anti-apoptotic protein BCL2 in T-ALL cells through phosphorylation and activation of STAT1 (Sanda, et al 2013). However, this is not the only pathway involved in TYK2-mediated promotion of survival, because a specific inhibitor of BCL2, ABT-199, is not active in the killing of most T-ALL cells (Anderson, et al 2014, Chonghaile, et al 2014, Peirs, et al 2014). Rather, only T-ALL cells with an early T-cell progenitor (ETP) phenotype show enhanced sensitivity to ABT-199 (Anderson, et al 2014, Chonghaile, et al 2014, Peirs, et al 2014), implying that other mechanisms in addition to the downregulation of the STAT1-BCL2 signalling contribute to the induction of apoptosis by TYK2 inhibition. Thus, we investigated cellular signalling pathways that are associated with cell survival and specifically affected by TYK2 inhibition with NDI-031301. We treated KOPT-K1 cells with NDI-031301, tofacitinib, or baricitinib for 24 h, and then examined the levels of total and phosphorylated protein of known cell survival signalling pathways (Fig 3A). Two drug concentrations, 1 μM and 3 μM, of NDI-031301 were selected for this assay because this compound had produced a dose-dependent inhibitory effect on TYK2-transformed Ba/F3 cells (Fig 1A) as well as KOPT-K1 cells (Fig 2A) over this concentration range. The concentrations of tofacitinib (1 μM) and baricitinib (300 nM) were determined from the results of the Ba/F3 assay, in which these concentrations efficiently inhibited the proliferation of Ba/F3 cells transformed by corresponding JAK kinases, without affecting the viability of TYK2-transformed cells (Fig 1B and C).
The results in Figure 3A indicate that treatment with 3 μM of NDI-031301 leads to reduction of STAT1 Tyr-701 phosphorylation and BCL2 protein levels in KOPT-K1 cells, consistent with our previous finding that TYK2 phosphorylates STAT1 and upregulates BCL2 expression in most T-ALL cells (Sanda, et al 2013). Decreased levels of STAT1 Tyr-701 phosphorylation were also observed in cells treated with tofacitinib or baricitinib, demonstrating that this activity is not specific to TYK2 inhibition and that other JAK kinases are involved in the regulation of STAT1 Tyr-701 phosphorylation in KOPT-K1 cells. The phosphorylation levels of STAT3 (Tyr-705), STAT5 (Tyr-694) and STAT6 (Tyr-641) were not affected by treatment with either NDI-031301 or other JAK kinase inhibitors (Fig 3A). The levels of AKT Thr-308 phosphorylation were also unchanged after each of the treatments (Fig 3A). We found that each of the mitogen-activated protein kinase (MAPK) pathways were activated specifically by treatment with 3 μM of NDI-031301, as indicated by increased phosphorylation levels of ERK1/2 (Thr-202/Tyr-204), SAPK/JNK (Thr-183/Tyr-185), and p38 MAPK (Thr-180/Tyr-182) (Fig 3A). Importantly, efficient induction of apoptosis was observed only in the cells treated with 3 μM of NDI-031301, as indicated by the presence of cleaved PARP (Fig 3A). These results indicate that downregulation of STAT1 Tyr-701 phosphorylation is not sufficient by itself to induce apoptosis in KOPT-K1 cells, suggesting instead that activation of the MAPK pathways might also play a role in NDI-031301-induced apoptosis in these cells.
We then asked whether the effect of NDI-031301 on cellular signalling could be recapitulated by gene knockdown of TYK2. Silencing of TYK2 by lentiviral shRNA knockdown (shTYK2 2 and 3) resulted in elimination of STAT1 Tyr-701 phosphorylation in KOPT-K1 cells (Fig 3B), consistent with the result in the cells treated with NDI-031301 (Fig 3A). The increased phosphorylation levels of SAPK/JNK (Thr-183/Tyr-185) and p38 MAPK (Thr-180/Tyr-182) were clearly observed in the cells transfected with TYK2-targeting shRNAs, while the levels of ERK1/2 phosphorylation (Thr-202/Tyr-204) were not upregulated (Fig 3B). Thus, TYK2 signalling is required for the down-modulation of the SAPK/JNK and p38 MAPK pathways in KOPT-K1 cells, as indicated both genetically and by NDI-031301-mediated inhibition of TYK2 tyrosine kinase activity.
Activation of p38 MAPK is involved in NDI-031301-induced apoptosis in KOPT-K1 cellsPrevious studies have shown that the SAPK/JNK and p38 MAPK pathways play key roles in the negative selection of CD4+/CD8+ thymocytes (Rincon and Davis 2009, Sohn, et al 2003, Sohn, et al 2007, Sugawara, et al 1998, Tanaka, et al 2002), an essential process by which self-reactive T-cells are removed from the host. Thus, we hypothesized that activated MAPK pathways are also involved in NDI-031301-mediated apoptosis in T-ALL cells. Time-course analyses with KOPT-K1 cells demonstrated that a dramatic increase in levels of p38 MAPK phosphorylation (Thr-180/Tyr-182) and a decrease in levels of STAT1 phosphorylation (Tyr-701) occurred at 1 h after treatment with 3 μM of NDI-031301, while increased levels of ERK1/2 Thr-202/Tyr-204 and SAPK/JNK Thr-183/Tyr-185 phosphorylation were observed after 8 h of treatment (Fig 4A). Cleaved PARP was first detected at 12 h of treatment (Fig 4A). These results indicate that NDI-031301 mediates activation of the MAPK pathways before the induction of apoptosis.
We next determined whether pharmacological inhibition of each MAPK pathway could protect T-ALL cells from proliferative suppression and apoptosis induced by NDI-031301. We treated KOPT-K1 cells for 72 h with 0 to 3 μM of NDI-031301 in combination with each of the following MAPK pathway inhibitors: SB203580 (p38 MAPK inhibitor; 0, 5 and 10 μM), trametinib (MEK1/2 inhibitor; 0, 1.25 and 2.5 nM) and SP600125 (SAPK/JNK inhibitor; 0 and 5 μM), and then assessed the cell proliferative responses. We found that combined treatment with SB203580 partially rescued the NDI-031301-induced decrease of viability in KOPT-K1 cells (Fig 4B). Isobologram analysis indicated that this combination produced remarkable antagonism between the two inhibitors in decreasing cell viability, as indicated by the average combination index (CI) value of 2.06 (Fig 4C). Combination with trametinib or SP600125 showed no rescue effects on the viability of KOPT-K1 cells treated with NDI-031301 (Fig S2A and C) with only modest antagonism for combination with trametinib (average CIs: 1.39) (Fig S2B) or SP600125 (average CIs: 1.15) (Fig S2D). SB203580 also partially rescued NDI-031301-induced apoptosis in KOPT-K1 cells, as indicated by a decrease in TUNEL positivity (Fig 4D), and an absence or decreased levels of cleaved PARP induced by 3 μM of NDI-031301 (Fig 4E). We also found that SB203580 efficiently inhibited NDI-031301-mediated increase of levels of p38 MAPK Thr-180/Tyr-182 phosphorylation without downregulating the levels of ERK1/2 Thr-202/Tyr-204 and SAPK/JNK Thr-183/Tyr-185 phosphorylation (Fig 4E). These results demonstrate that activation of the p38 MAPK pathway, but not the ERK or SAPK/JNK pathways, is at least partially responsible for NDI-031301-induced apoptosis in KOPT-K1 cells, indicating involvement of the p38 MAPK pathway in TYK2-mediated survival in T-ALL cells. Further investigation to assess NDI-031301 sensitivity and the effect of the drug treatment on cellular signalling pathways in an additional 3 T-ALL cell lines indicated variable levels of increased phosphorylation of p38 MAPK and downregulation of BCL2 that did not clearly correlate with the IC50 values for NDI-031301 in these cell lines (Fig S3A-D). Thus, there appears to be considerable heterogeneity in altered signal transduction in the different cell lines after treatment with NDI-031301 in the timing and magnitude of the response, which is expected given the wide heterogeneity of oncogenic mutations and chromosomal abnormalities among these T-ALL cell lines.
NDI-031301 suppresses the proliferation of T-ALL cells in vivoWe then tested the efficacy of NDI-031301 in vivo using a murine xenograft model. By injecting NSG mice with 2 × 106 KOPT-K1 cells expressing luciferase, we were able to quantify leukaemia growth by BLI. Once tumour engraftment was established by BLI (Fig 5A), treatment with either vehicle or 100 mg/kg NDI-031301 twice daily was initiated and continued for 29 days. This dosage of NDI-031301 was well tolerated by the mice without any appreciable side effects and body weight loss (Fig S4). We found that treatment with NDI-031301 significantly delayed tumour progression in this model (Fig 5B and C): mean bioluminescence on day 42 was 3.126 × 1010/photons (ph)/s/cm2/steradian (sr) for vehicle control, versus 1.277 × 1010/ph/s/cm2/sr for NDI-031301-treated animals (P < 0.0001; n = 6 in each group). As a result, NDI-031301 significantly increased overall survival compared with vehicle treatment (P = 0.0064; n = 6 in each group) (Fig 5D).
After 29 days of treatment, mice receiving NDI-031301 had marked reductions in splenomegaly by comparison with controls (Fig 6A and B). Flow cytometric analysis to detect human CD45-positive cells showed a significant decrease in the extent of leukaemic cell infiltration in the femur bone marrow of NDI-031301-treated mice versus vehicle-treated mice (Fig 6C and D). Decreased numbers of human leukaemia cells after NDI-031301 treatment were also observed in the peripheral blood (Fig S5). Histopathological analysis of the bone marrow of the vehicle-treated mice showed massive infiltration by leukaemic cells (Fig 6E), while NDI-031301 treatment suppressed leukaemic cell infiltration (Fig 6F). Taken together, our results demonstrate that NDI-031301 is efficacious in delaying the growth of human T-ALL cells in vivo and in extending the survival of mice bearing human T-ALL cells.
DiscussionHere we report that TYK2 inhibition by a small-molecule inhibitor, NDI-031301, results in robust anti-tumour activity in T-ALL cells. We found that NDI-031301 shows selective inhibitory activity against TYK2 as compared to other JAK kinases in living cells, as indicated by transformed Ba/F3 assays. This inhibitor induced cytotoxicity in 4 representative human T-ALL cell lines, consistent with our previous result showing the growth and survival inhibition of these cells by silencing TYK2 with shRNA (Sanda, et al 2013). Finally, the anti-leukaemic activity of NDI-031301 was recapitulated in a KOPT-K1 in vivo xenograft study, indicating that this inhibitor is able to efficiently suppress the growth of human T-ALL cells in vivo as well as in vitro.
Our results also indicated that in the KOPT-K1 T-ALL cell line, activation of three different MAPK signalling pathways was observed after treatment with NDI-031301. However, of these pathways, only the activation of p38 MAPK contributed to apoptosis induced by this inhibitor. Increased phosphorylation levels of p38 MAPK (Thr-180/Tyr-182) were also observed in the cells transfected with each of two different shRNAs targeting TYK2 (Fig 3B), indicating that NDI-031301-mediated activation of p38 MAPK is probably mediated through inhibition of TYK2. Time-course analyses with KOPT-K1 cells demonstrated elevated levels of p38 MAPK phosphorylation (Thr-180/Tyr-182), as observed at 1 h post-treatment with NDI-031301, suggesting that TYK2 negatively regulates proteins that control the activation of p38 MAPK in T-ALL cells. Previous studies have shown a critical role for p38 MAPK pathway activation in the anti-leukaemic effects of dasatinib (Dumka, et al 2009), imatinib (Parmar, et al 2004) and IFNα (Mayer, et al 2001) in BCR-ABL-transformed cells; however, our finding is the first to demonstrate involvement of p38 MAPK pathway activation in T-ALL cell death occurring after TYK2 inhibition. We did not find clear association between variable levels of phosphorylation of p38 MAPK and differential NDI-031301 sensitivities in human T-ALL cell lines (Fig 4A and Fig S3A-D), suggesting that the consequences of p38 MAPK activation are highly context-dependent and probably differ based on the wide genetic and phenotypic heterogeneity among individual T-ALL cases. Importantly, activation of p38 MAPK is associated with the deletion of CD4+CD8+ double-positive thymocytes by negative selection (Rincon and Davis 2009, Sohn, et al 2003, Sohn, et al 2007, Sugawara, et al 1998, Tanaka, et al 2002), although the mechanism of this effect still remains to be established. It is conceivable that TYK2 normally promotes the survival of a subset of developing thymocytes, and that this pathway becomes essential for the survival of transformed T-ALL cells.
Our findings clearly support TYK2 inhibition with NDI-031301 or a related compound as a potential therapeutic strategy for patients with T-ALL, and also raise the possibility that enhancing p38 MAPK activation in T-ALL cells may be an approach to accentuate its anti-leukaemic activity. Further preclinical studies with selective TYK2 inhibitors, both alone and in combination with other targeted inhibitors, are needed to delineate an optimal strategy for targeting this pathway, which will be essential for the design of early stage clinical trials to incorporate TYK2 inhibition into available regimens for the treatment of patients with T-ALL.