Target interaction profiling of midostaurin and its metabolites in neoplastic mast cells predicts distinct effects on activation and growth

Abstract

Proteomic-based drug testing is an emerging approach to establish the clinical value and anti-neoplastic potential of multi-kinase inhibitors. The multikinase inhibitor midostaurin (PKC412) is a promising new agent used to treat patients with advanced systemic mastocytosis (SM). We examined the target interaction-profiles and the mast cell (MC)-targeting effects of two pharmacologically relevant midostaurin metabolites, CGP52421 and CGP62221. All three compounds, midostaurin and the two metabolites, suppressed IgE-dependent histamine secretion in basophils and MC with reasonable IC₅₀ values. Midostaurin and CGP62221 also produced growth-inhibition and dephosphorylation of KIT in the MC leukemia cell line HMC-1.2, whereas the second metabolite, CGP52421, that accumulates in vivo, showed no substantial effects. Chemical proteomic profiling and drug-competition experiments revealed that midostaurin interacts with KIT and several additional kinase-targets. The key downstream-regulator FES was recognized by midostaurin and CGP62221, but not by CGP52421 in MC lysates, whereas the IgE-receptor-downstream target SYK was recognized by both metabolites. Together, our data show that the clinically relevant midostaurin metabolite CGP52421 inhibits IgE-dependent histamine release, but is a weak inhibitor of MC proliferation which may have clinical implications and may explain why mediator-related symptoms improve in SM patients even when disease progression occurs.

Introduction

Systemic mastocytosis (SM) is a hematopoietic neoplasms defined by expansion and pathologic accumulation of mast cells (MC) in internal organs (1–5). In almost all patients, the bone marrow (BM) is affected (1–5). The transforming KIT mutation D816V is expressed in neoplastic cells in most patients (6–10).

Clinical problems in SM result from MC infiltration with consecutive end-organ damage and from various MC-derived mediators (1–3,11–14). Whereas clinically relevant organ damage is seen in aggressive SM (ASM) and MC leukemia (MCL) (1–5,11), mediator-related symptoms can develop in any category of SM (12–14). In patients in whom mediator-related symptoms are severe or even life-threatening, an IgE-mediated allergy may be detected (12–17).

Treatment of SM is based on the subtype of disease and the presence of symptoms (11–17). Patients with mediator-related symptoms are usually treated with MC-stabilizing drugs, histamine receptor antagonists, or immunotherapy (12–17). For severe cases, glucocorticosteroids or IgE-blockage has been proposed (12–17).

In patients with aggressive SM or MCL, cytoreductive agents or targeted drugs are prescribed (1–3,11,13,17–19). A novel promising strategy is to use KIT-targeting tyrosine kinase inhibitors (TKI) (19–25). However, the KIT D816V mutant is resistant against several of these TKI, including imatinib (21,25,26).

PKC412 (midostaurin) is a multi-kinase inhibitor directed against wild type (wt) KIT and several KIT mutants, including D816V (21,27,28). Additional oncogenic kinases are also recognized by midostaurin (27,29). However, the exact profile of molecular targets of midostaurin expressed by neoplastic MC remains unknown. A number of in vitro studies have shown that midostaurin inhibits the proliferation and survival of neoplastic MC exhibiting KIT D816V (21,23,28). In addition, midostaurin can suppress growth of neoplastic MC in vivo in patients with advanced SM (20,24). However, anti-neoplastic effects of midostaurin are often transient (20,24). A remarkable phenomenon is that even in patients who have resistant disease or relapse, mediator-related symptoms improve or disappear during therapy (30).

So far, only a few studies have addressed the potential effects of various TKI on mediator secretion from MC and basophils (31,32). We have shown that midostaurin blocks IgE-dependent release of histamine in MC and basophils (32). However, little is known about targets and mechanisms contributing to this drug-effect.

Upon chronic oral dosing to patients, midostaurin displays a dose- and time-dependent pharmacokinetic profile, with plasma concentrations increasing during the first 3-6 days of treatment, followed by a significant decrease, before reaching a stable plateau after 3-4 weeks (33–37). The drug has two active metabolites (Supplementary Figure S1), which result from cytochrome CYP3A4-mediated oxidation of the parent-drug (36,37). Upon repeated dosing of midostaurin, CGP52421 accumulates significantly, whereas CGP62221 does not. Thus, in a study in patients with diabetes mellitus, it was found that following multiple oral dosing (25-75 mg b.i.d. or 75 mg t.i.d.), day-28 plasma exposure as assessed by area under the plasma-concentration time curve (AUC), for CGP62221 was slightly higher than that for midostaurin, whereas for CGP52421, the AUC52421/AUCmidostaurin ratio was greater than 5 (37).

Clinically important questions in the context of mastocytosis are i) whether CGP62221 and CGP52421 retain activity against neoplastic MC compared to midostaurin, ii) whether the less active metabolite would act as an antagonist neutralizing midostaurin effects, and iii) whether the metabolites block IgE-dependent cell activation in the same way as midostaurin.

Materials and Methods

Reagents

Reagents used in this study are described in Supplementary Material and Methods.

Isolation of human blood basophils and MC

Human blood basophils, lung MC, cord blood progenitor cell-derived MC, and mononuclear cells (MNC) freshly obtained from SM patients, were isolated as described (38,39). A detailed description is provided in Supplementary Material and Methods. The patients´ characteristics are shown in Table 1.

Table 1.

Patients´ characteristics and growth-inhibitory effects of midostaurin, CGP52421 and CGP62221 on primary neoplastic cells (MC-containing cells)

Patient	Age	Gender	Diagnosis	KIT Mutation	Serum Tryptase (µg/l)	% of MC in BM*	% of MC in Sample	cell source	IC50 (µM) midostaurin	IC50 (µM) CGP52421	IC50 (µM) CGP62221
#1	55	m	ISM	D816V	124.0	20%	3%	BM	0.1-0.25	>1	0.1-0.25
#2	36	f	ISM	D816V	19.5	5%	5%	BM	0.5-1	>1	0.1-0.25
#3	24	m	ISM	D816V	53.7	5%	5%	BM	0.1-0.25	0.5	0.05
#4	63	f	ISM	D816V	77.8	25%	10%	BM	0.1	0.25-0.5	0.01-0.05
#5	37	m	ISM	D816V	4.5	5%	n.a.	PB	n.a.	n.a.	n.a.
#6	47	f	ISM	D816V	32.5	n.a.	n.a.	PB	n.a.	n.a.	n.a.
#7	72	m	ISM-AHNMD	D816V	191.0	20%	7%	BM	0.5	>1	0.1-0.25
#8	57	m	ASM	D816V	152.0	10%	10%	BM	0.25-0.5	>1	0.075-0.1
#9	71	m	ASM	D816V	113.0	15%	5%	BM	0.1-0.5	>1	0.05-0.1
#10	73	m	ASM	D816V	152.0	25%	3%	BM	0.1	0.5-1	0.01-0.05
#11	73	m	ASM	D816V	789.0	95%	40%	BM	0.25-0.5	>1	0.075-0.1
#12	55	m	ASM	D816V	200.0	60%	3%	BM	0.5	>1	0.1-0.25
#13	61	f	MCL	D816Y	1,690.0	99%	85%	BM	0.1-0.25	1	0.075-0.1
#14	54	f	MCL	D816H	904.0	50%	>95%	BM	0.25-0.5	0.5-1	0.01-0.05
#14	54	f	MCL	D816H	904.0	50%	97%	PB	0.5-1	>1	0.1

MC, mast cells; BM, bone marrow; m, male; f, female; ISM, indolent systemic mastocytosis; ASM, aggressive systemic mastocytosis; MCL, mast cell leukemia; ISM-AHNMD, indolent systemic mastocytosis- associated clonal haematological non-mast cell lineage diseases; BM, bone marrow; PB, peripheral blood ; n.a., not available.

^*Percent values of mast cells are based on immunohistochemical examinations using BM sections and an antibody against tryptase. Patients IC₅₀ values were determined by ³H-thymidine uptake experiments. Patients used for histamine release experiments: #5, #6, #9, #10 (PB after dextran isolation) and #11 (BM after Ficoll isolation).

Cell lines

Two subclones of the human MCL line HMC-1 were used, HMC-1.1 harbouring KIT V560G, and HMC-1.2 harbouring KIT V560G and KIT D816V (21,26,40). A detailed description of cell lines is provided in the Supplement.

Histamine release experiments

Histamine release experiments were performed on basophils and MC as described previously (31,32). A detailed decription is provided in the Supplement.

Proliferation experiments and analysis of drug-exposed cells

Neoplastic MC were incubated with control medium or various concentrations of midostaurin and its metabolites at 37°C for 24-48 hours and analysed for proliferation and survival (apoptosis). A detailed description of the methods is provided in Supplementary Material and Methods. A list of antibodies used in this study is shown in Supplementary Table S1.

Western blotting

Western blotting was performed essentially as described previously (39). A detailed description of the method is provided in Supplementary Material and Methods.

Chemical Proteomic profiling (CPP), drug competition experiments, and liquid chromatography mass spectrometry

In order to determine the target spectrum of midostaurin in neoplastic MC, we employed an affinity-chromatography based strategy using a coupleable analogue of midostaurin (c-midostaurin) that allows bead-based immobilization. CPP and drug affinity chromatography experiments were performed essentially as described (41–44) using HMC-1 cells and primary MNC obtained from a patient with ASM (5% MC) and one with MCL (>95% MC). The resulting affinity matrix was incubated with protein lysates of neoplastic MC and interacting proteins were identified by mass spectrometry. A detailed description is provided in Supplementary Material and Methods.

Knock-down of FES in HMC-1.2 cells

Lentiviral transfection of HMC-1.2 cells with shRNA against FES was performed as described (45). A detailed description is provided in the Supplement.

Flow cytrometry analyses

Flow cytometry analyses were performed on primary blood basophils and HMC-1.2 cells according to published methods (23,32,46). A detailed description is provided in Supplementary Material and Methods.

Statistical analysis

To determine the level of significance in drug inhibition experiments, the paired Student´s t test was applied. A p-value of less than 0.05 was considered to indicate statistical significance.

Results

Effects of midostaurin and its metabolites on proliferation of neoplastic MC

Midostaurin was found to inhibit proliferation in HMC-1.1 and HMC-1.2 cells in a dose-dependent manner. A similar growth-inhibitory effect was seen with CGP62221 in both HMC-1 subclones (IC₅₀ 50-250 nM) (Figure 1A). However, the second metabolite, CGP52421, did not produce comparable anti-proliferative effects (Figure 1A). Similar results were obtained when metabolite-effects were compared in freshly obtained BM or peripheral blood MNC from patients with various subtypes of SM (Figure 1B-1C). In fact, midostaurin and CGP62221 produced stronger growth-inhibitory effects in these cells compared to CGP52421 (Figure 1B and 1C, Table 1). However, the potency of CGP52421 (IC₅₀ value) varied from patient to patient, and in some of them, the metabolite´s IC₅₀ reached a “submicromolar” range (Figure 1B and 1C, Table 1).

Figure 1. Effects of midostaurin, CGP52421 and CGP62221 on growth of neoplastic mast cells.

A-C: HMC-1.1 cells (A, left panel) and HMC-1.2 cells (A, right panel) as well as primary bone marrow (BM) mononuclear cells obtained from 2 patients with indolent systemic mastocytosis (ISM; mast cell purity: 10% and 3%) (B) and 2 with aggressive SM (ASM; mast cell purity: 10% and 3%) (C); and peripheral blood MNC from one patient with mast cell leukemia (MCL; mast cell purity: 97%) (C), were incubated in control medium (CTR) or various concentrations of midostaurin, CGP52421 or CGP62221 as indicated at 37°C for 48 hours. After incubation, uptake of ³H-thymidine was measured. Results shown in “A” are expressed as percent of medium control (Co) and represent the mean±S.D. from 7 independent experiments. Results shown in “B” and “C” are expressed as percent of control and represent the mean±S.D. of triplicates.

Effects of midostaurin and its metabolites on cell survival of HMC-1 cells

Midostaurin and CGP62221 induced apoptosis in HMC-1.1 and HMC-1.2 cells at pharmacologically meaningful concentrations, as evidenced by light microscopy (Figure 2A) and active caspase-3 staining (Figure 2B). Apoptosis-inducing effects of midostaurin and CGP62221 were also confirmed by Tunel assay (Figure 2C). By contrast, the second metabolite, CGP52421, did not induce substantial apoptosis in HMC-1 cells (Figure 2A-C). Moreover, midostaurin and its metabolites showed no stubstantial effects on survival of normal BM MNC (Supplementary Figure S2A).

Figure 2. Effects of midostaurin, CGP52421 and CGP62221 on viability of HMC-1 cells.

A: HMC-1 cells were incubated in control medium (CTR) or in various concentrations of midostaurin, CGP52421 or CGP62221 at 37°C for 24 hours. Thereafter, the percentage of apoptotic cells was quantified by light microscopy. Results represent the mean±S.D. of 4 independent experiments. Asterisk: p<0.05. B: HMC-1 cells were incubated in control medium (CTR) or various concentrations of midostaurin or its metabolites at 37°C for 24 hours. Thereafter, cells were stained with an antibody against active caspase 3 and analyzed by flow cytometry. Results show the percentages of active caspase 3-positive cells and represent the mean±S.D. of 3 independent experiments. Asterisk: p<0.05. C: HMC-1 cells were incubated in control medium, midostaurin, CGP52421 or CGP62221 at 37°C for 24 hours. In HMC-1.2 cells (right panel), drugs were applied at 1 µM and in HMC-1.1 cells (left panel), drugs were applied at 0.5 µM. After incubation, cells were examined for viability and apoptosis by Tunel assay as described in the text.

The midostaurin metabolite CGP52421 does not block phosphorylation of KIT at 1 µM

We next asked whether the midostaurin metabolites are capable of targeting the KIT kinase. Whereas midostaurin and CGP62221 were found to block constitutive phosphorylation of KIT in HMC-1.2 cells at 1 µM, no comparable inhibitory (deactivating) effect was seen with CGP52421 (Supplementary Figure S2B). Similar data were obtained with the major KIT D816V-downstream kinase FES. Whereas midostaurin and CGP62221 (1 µM) suppressed FES kinase-activity in HMC-1.2 cells, CGP52421 failed to completely block FES activity at 1 µM (Supplementary Figure S2C).

CGP52421 does not interfere with midostaurin-induced or CGP62221-induced inhibition of growth of KIT-mutated HMC-1 cells

A clinically important question is whether the less active midostaurin metabolite CGP52421 that accumulates during therapy, would competitively interfere with midostaurin-induced or CGP62221-induced growth inhibition. To address this question, we preincubated HMC-1.1 cells and HMC-1.2 cells with CGP52421 (1 µM, 1 hour) and then added midostaurin or CGP62221. However, preincubation with CGP52421 did not alter the growth-inhibitory effects of midostaurin or CGP62221 in HMC-1 cells (Supplementary Figure S3A).

Midostaurin as well as its metabolites cooperate with cladribine (2CdA) in inducing growth inhibition in HMC-1.2 cells

We have previously shown that midostaurin and 2CdA produce synergistic growth-inhibitory effects on neoplastic MC expressing KIT D816V (21). In the current study, we asked whether the less effective midostaurin metabolite, CGP52421, that accumulates in vivo, would also cooperate with 2CdA to producing growth-inhibition. Indeed, our data show that CGP52421 cooperates with 2CdA in inhibiting the growth of HMC-1.2 cells (Supplemtary Figure S3B) and the second midostaurin metabolite, CGP62221, was also found to cooperate with 2CdA to cause growth inhibition in HMC-1.2 cells (Supplementary Figure S3B).

CGP52421 and CGP62221 inhibit IgE-dependent release of histamine

Midostaurin has recently been described to suppress IgE-dependent histamine release (32). In the present study, CGP52421, CGP62221 and midostaurin were found to inhibit IgE-dependent histamine secretion in normal blood basophils with pharmacologically relevant IC₅₀ values (0.01-1 µM) (Figure 3A). In addition, midostaurin was found to inhibit IgE-mediated histamine release in blood basophils obtained from SM patients (Figure 3B). Moreover, we were able to show that oral administration of midostaurin (2x100 mg/day) results in a decrease in IgE-dependent histamine release in ex vivo-recovered blood basophils (Figure 3C). Finally, we were able to show that midostaurin and its metabolites inhibit IgE-dependent histamine release from primary bone marrow MC from a patient with ASM (Figure 3D), human lung MC (Supplementary Figure S4A) and cord blood progenitor cell-derived MC (Supplementary Figure S4B).

Figure 3. Effects of midostaurin on histamine release in human basophils and MC.

A: Primary blood basophils (healthy donors) were incubated in control medium (CTR) with or without midostaurin, CGP52421 or CGP62221 (0.01-10 µM) for 30 minutes. Thereafter, cells were exposed to histamine release buffer (HRB) with or without anti-IgE antibody E-124.2.8 (1 µg/ml) at 37°C for 30 minutes. After incubation, cells were centrifuged at 4°C, and cell-free supernatants and cell suspensions recovered and examined for histamine-content by RIA. Histamine release was calculated as percent of total histamine and is expressed as percent of control. Results represent the mean±S.D. of 5 independent experiments. Asterisk: p<0.05. B: Primary blood basophils (ISM=2, ASM=1) were incubated in control medium (CTR) with or without midostaurin (0.01-10 µM) for 30 minutes. Then, histamine release was measured as described above. Histamine release was calculated as percent of total histamine and is expressed as percent of control. Results represent the mean±S.D. of 3 independent experiments. Asterisk: p<0.05. C, left panel: Basophils from a patient with ASM (before therapy) were incubated in medium or medium containing 1 µM midostaurin at 37°C for 30 minutes. Thereafter, cells were incubated in HRB in the absence or presence of anti-IgE antibody E-124.2.8 (0.001-10 µg/ml) at 37°C for 30 minutes. After incubation, cells were centrifuged at 4°C, and cell-free supernatants and cell suspensions analyzed for histamine content. Histamine release is expressed as percent of total histamine; results represent the mean±S.D. of triplicates. C, right panel: Basophils obtained from the same patient with ASM before(?-?) and 8 days after (?-?) treatment with midostaurin (100 mg twice daily) were incubated in HRB in the absence or presence of anti-IgE antibody E-124.2.8 (0.001-10 µg/ml) for 30 minutes. Then, histamine release was measured as described above. Results represent the mean±S.D. of triplicates. D: Primary bone marrow MNC from a patient with ASM (purity of MC: 40%) were incubated in control medium (CTR) or medium containing midostaurin, CGP52421 or CGP62221 (0.01-10 µM) for 30 minutes. Thereafter, histamine release was measured. Histamine release was expressed as percent of control. Results represent the mean±S.D. of triplicates.

Drug profiling of midostaurin in neoplastic MC reveals a unique spectrum of kinase targets and IgE receptor-downstream signaling molecules

As assessed by chemical proteomic profiling and mass spectrometry, c-midostaurin was found to bind to a number of oncogenic signaling molecules in neoplastic MC. A summary of drug-binding results is shown in Supplementary Table S2 and Figure 4A-4B. Major kinase targets recognized by c-midostaurin in HMC-1.1 and HMC-1.2 cells were KIT, SYK, FES, GSK3B, AAK1, BIKE, TBK1, PKN1, AMPK1 and MARK2 (Supplementary Table S2; Figure 4A). Interestingly, several targets such as RSK1-3, were only identified by midostaurin in HMC-1.1 but not in HMC-1.2 cells. c-midostaurin did not bind SRC, LYN and FGR in HMC-1 lysates (Supplementary Table S2; Figure 4A). A similar target spectrum was found in primary MC-containing MNC. Again, several kinases, including SYK, BIKE and FES, were recognized by c-midostaurin (Supplementary Table S2, Figure 4B). Interestingly, in primary cell lysates, c-midostaurin recognized BTK (Supplementary Table S2, Figure 4B). Most of these kinase targets were also identified by midostaurin in a cell-free system (29). In control experiments, we were able to show that midostaurin and c-midostaurin inhibit growth of HMC-1.2 cells with comparable IC₅₀ values (Supplementary Figure S5A). The chemical structures of midostaurin and c-midostaurin are shown in Supplementary Figure S5B.

Figure 4. Midostaurin kinase target signature in HMC-1 cells and primary neoplastic mast cells (MC) as determined by chemical proteomic profiling (CPP).

Kinase target profiles obtained by CPP using c-midostaurin and lysates of HMC-1 cells (A), primary neoplastic cells (5% MC) of a patient with aggressive systemic mastocytosis (ASM) (B, left panel) and neoplastic MC (>95% purity) of a patient with MC leukemia (MCL) (B, right panel). HMC-1.1 cells, HMC-1.2 cells, and primary MC were processed and analyzed by CPP and liquid chromatography mass spectrometry (LCMS) as described in the text. The figure shows major kinase binders identified in form of a kinome-map adapted from Cell Signaling Technology (www.cellsignal.com). D: Drug affinity experiments were performed using HMC-1.2 cell lysates and various drugs applied to compete with c-midostaurin in its binding to FES. Expression of FES was determined by Western Blotting as described in the text.

Dissection of the target spectrum of CGP52421 and CGP62221 as determined by drug competition experiments

To learn whether CGP52421 and CGP62221 also bind to midostaurin targets, we performed quantitative drug-competition experiments using c-midostaurin. As expected, binding of c-midostaurin to its targets was successfully displaced by midostaurin in HMC-1.2 lysates (Table 2 and Supplementary Table S3) and primary neoplastic MC (Supplementary Table S4). Whereas CGP62221 and midostaurin showed comparable target-binding properties, CGP52421 was unable to compete with c-midostaurin in binding to several kinase targets in HMC-1.2 cells (Table 2 and Supplementary Table S3). Most intriguingly, CGP52421 failed to displace the binding of c-midostaurin to FES in HMC-1.2 lysates, suggesting that FES is not recognized by this midostaurin metabolite (Table 2; Supplementary Table S3, Figure 4C). Binding of SYK was displaced by CGP62221 and midostaurin, and less effectively by CGP52421 (Table 2 and Supplementary Table S3). In flow cytometry experiments, midostaurin and both metabolites were found to downregulate the phosphorylation of SYK in HMC-1.2 cells (Supplementary Figure S6).

Table 2.

Dissection of the target spectrum of midostaurin, CGP52421 and CGP62221 in HMC-1.2 cells as determined by drug competition experiments

Kinase	average spectral counts*	midostaurin competition	CGP62221 competition	CGP52421 competition
AAK1	29	0.30	0.49	0.98
BIKE	24	0.34	0.81	1.06
FES	19.5	0.58	0.55	1.03
AMPK1	15	0.68	0.97	1.23
SYK	14.5	0.65	0.64	0.77
PRKAG1	13	0.66	0.89	1.00
TBK1	12	0.19	0.31	0.61
PKN1	8.5	0.58	0.60	0.61
GSK3B	7.5	0.88	1.19	1.21
CAMKK2	6.5	0.56	1.08	1.15
KIT	5.5	0.61	0.57	0.50
AURKA	5.5	0.84	0.83	1.19
FER	4.5	0.83	0.57	0.86
PIM1	3	0.89	0.80	0.82
MLK3	3	0.53	1.25	1.21

Results represent the mean of 2 biological replicates. No competition with c-midostaurin: ≥1; competition with c-midostaurin <1.

^*the average of the maximal number of spectra that matched uniquely to the protein has been used, only kinases with a spectral count of at least 2 were considered as ´detected kinases´.

Validation of major midostaurin targets in neoplastic MC and normal basophils

Since FES was identified as a major target of midostaurin and CGP62221, but was not recognized by CGP52421, we asked whether silencing of FES by shRNA is associated with reduced growth of HMC-1.2 cells. Indeed, we found that a lentiviral mediated knock-down of FES in HMC-1.2 cells is followed by reduced growth compared to a random control shRNA (Supplementary Figure S7A). This observation suggests that FES is an important KIT-downstream target of midostaurin that is not recognized by CGP52421. We were also interested to learn whether pharmacologic SYK inhibition results in a decreased IgE-mediated histamine release. We found that the SYK inhibitor P505-15 counteracts IgE-dependent release of histamine in basophils and cord blood cell-derived MC (Figure 5A and B). In addition, P505-15 was found to inhibit IgE-dependent upregulation of CD63 and CD203c in basophils (Figure 5C). By contrast, P505-15 did not inhibit growth of HMC-1 cells up to 1 µM (Supplementary Figure S7B) and showed only a weak effect on CD63 expression in HMC-1.2 cells, whereas midostaurin and its metabolites decreased CD63 expression in these cells (Supplementary Figure S7C).

Figure 5. Effects of SYK inhibition on histamine release on MC and CD63 and CD203c upregulation in basophils.

A and B: Primary blood basophils (A) and cord blood derived MC (B) were incubated in control medium (CTR) with or without P505-15 (0.01-10 µM) at 37°C for 30 minutes. Thereafter, cells were exposed to HRB with or without anti-IgE antibody E-124.2.8 (1 µg/ml for basophils; 10 µg/ml for MC) at 37°C (30 minutes). After incubation, cells were centrifuged at 4°C, cell-free supernatants and cell suspensions recovered and examined for histamine content by RIA. Histamine release was calculated as percent of total histamine and is expressed as percent of control. Results represent the mean±S.D. of 4 (A) or 3 (B) independent experiments. Asterisk: p<0.05. C: Human basophils in whole blood were preincubated with medium (CTR) or medium containing P505-15 (0.01-10 µM) at 37°C for 30 minutes. Then cells were challenged with anti-IgE (1 µg/ml) at 37°C for 15 minutes. Thereafter, expression of CD63 and CD203c was analyzed by flow cytometry as described in the text. Upregulation of CD63 and CD203c is expressed as stimulation index (SI). Results represent the mean±S.D. of 3 independent experiments. Asterisk: p<0.05.

Discussion

In advanced SM, patients usually suffer from mediator-related symptoms as well as organ damage produced by local MC infiltrates (1–3,12–14). During the past few years, several promising new drugs, potentially useful as cytoreductive agents in ASM and MCL, have been developed and tested in preclinical studies. One of these agents is the multi-kinase blocker midostaurin (20–24). However, although mediator-related symptoms often improve, hematologic responses are usually short-lived (20,24,30). Some of these patients relapse with KIT D816V-negative disease (20). We examined the effects of two pharmacologically relevant midostaurin metabolites, CGP52421 and CGP62221, on MC proliferation, survival, and IgE-dependent mediator secretion. We show that all 3 agents inhibit IgE-dependent histamine release. However, whereas growth of neoplastic MC was suppressed by midostaurin and CGP62221, the second metabolite, CGP52421, that accumulates in vivo, showed only weak or no effects on MC growth and survival which may be relevant clinically.

During treatment of patients with midostaurin, two pharmacologically relevant metabolites are generated, CGP52421 and CGP62221 (33–37). The non-linear pharmacokinetics of CGP62221 follows the same pattern as that of midostaurin, whereas the second metabolite, CGP52421, rises over time until reaching steady-state concentrations after one month of daily treatment (33–37). Our data show that CGP52421 is less effective in producing growth inhibition and apoptosis in neoplastic MC when compared to midostaurin. By contrast, the other metabolite, CGP62221, showed almost the same growth-inhibitory effects on neoplastic MC. Similar results have been described with AML cells (35). This is of particular interest as the percentage of MC in the primary samples varied among donors, and even if most non-MC lineage cells in advanced SM express KIT D816V, the response to TKI may be different in MC compared to other (clonal) cell types. On the other hand, IC₅₀ values were comparable among donors, and responses were also seen in those samples where a vast majority of cells were MC. Since CGP52421 achieved an exposure up to 5 to 7-fold higher compared to midostaurin in patients (37), a clinically important question was whether this metabolite would act as a competitor of midostaurin. However, our data show that CGP52421 has no influence on midostaurin-induced growth-inhibition in neoplastic MC.

KIT D816V is considered a major target of midostaurin. However, midostaurin also interacts with other target-antigens, such as PDGFR, FLT3 and PKC (27,29). We asked whether the target-spectrum of CGP52421 differs from that of midostaurin. Our data show that CGP52421 is a less potent inhibitor of KIT compared to midostaurin and CGP62221 in HMC-1.2 cells, which may explain at least in part the differential effects on cell growth. However, in drug-competition experiments, we also found that additional targets of midostaurin are less potently bound by CGP52421. Among these were FES, AAK1, and BIKE. FES is of particular interest, because this target has recently been implicated in KIT D816V-driven MC proliferation (47). In the present study, we were able to confirm that FES is a relevant KIT-downstream kinase target in HMC-1.2 cells. From these data, it is tempting to speculate that despite of its high plasma concentrations in vivo, CGP52421 may not contribute to hematologic efficacy, due to its less potent effects on KIT and FES.

We have previously shown that midostaurin inhibits not only the proliferation of neoplastic MC but also IgE-dependent mediator secretion (32). In this study we were able to confirm that midostaurin is a potent inhibitor of IgE-dependent histamine release in basophils (32). We also found that both drug metabolites inhibit anti-IgE-dependent histamine release in basophils and MC. Thus, whereas CGP52421 is a weak inhibitor compared to midostaurin regarding proliferation, no such differences were seen when examining IgE-dependent histamine release. This observation is of particular interest as mediator-related symptoms in midostaurin-treated patients often improve even if no hematologic response is obtained (30).

We next asked what targets of midostaurin might be responsible for blocking histamine release in MC and basophils. We found that several different targets such as SYK, PDK1, TBK1 or PKN1 are recognized by those compounds. Of these, SYK was identified as a functionally important target mediating histamine secretion. In fact, midostaurin as well as its metabolites were found to downregulate SYK-phosphorylation in HMC-1.2 cells. In addition, we were able to show that pharmacologic SYK-inhibition is associated with reduced histamine release. Together, these data show that SYK is a relevant target of midostaurin in MC and basophils.

Both midostaurin and cladribine (2CdA) can inhibit growth of neoplastic MC in patients with ASM or MCL (20,21,24,46,48). Previous data have shown that both drugs, when combined, exert growth-inhibitory effects on neoplastic MC in vitro (21). In the present study, we asked whether 2CdA would also produce cooperative growth-inhibitory effects when combined with midostaurin metabolites. Our results show that both metabolites cooperate with 2CdA in producing growth inhibition in HMC-1.2 cells, which would be in favor of clinical trials combining midostaurin and 2CdA for treatment of advanced SM.

In summary, our data show that the pharmacologically relevant midostaurin metabolite CGP52421 that accumulates in vivo, exerts little if any growth-inhibitory effects on neoplastic MC, but retains histamine release-blocking activity. These observations may explain why mediator-related symptoms improve in midostaurin-treated patients even if no hematologic remission is obtained.

Supplementary Material

Supplementary Information is available at Leukemia´s website.