Targeting MCL‐1 and MAPK overcomes venetoclax resistance in FLT3‐ITD‐positive AML cells harbouring activating PTPN11 (SHP‐2) mutations
Maximilian Fleischmann, Ole Hansen, Diana Voigtländer, Julia Bechwar, Lenny‐Joseph Schwietzer, Sanja Bahr, Ulf Schnetzke, Mike Fischer, Florian H. Heidel, Tina M. Schnöder, Jörg P. Müller, Andreas Hochhaus, Sebastian Scholl
TL;DR
This study shows that combining venetoclax with MCL-1 and MEK inhibitors can overcome resistance in AML cells with specific mutations.
Contribution
The study identifies a novel therapeutic strategy for overcoming venetoclax resistance in AML with PTPN11 and FLT3-ITD mutations.
Findings
PTPN11-E76K mutation causes resistance to venetoclax by increasing MCL-1 and BCL(x)L levels.
Combining venetoclax with MCL-1 and MEK inhibitors significantly enhances apoptosis in resistant cells.
PTPN11 and FLT3 mutations are key factors in venetoclax resistance and patient risk stratification.
Abstract
Venetoclax (VEN)‐based therapies have improved the treatment of acute myeloid leukaemia (AML); however, the emergence of resistance remains a major limitation. Mutations in protein tyrosine phosphatase (PTP) non‐receptor type 11 (PTPN11) and FMS like tyrosine kinase 3 with internal tandem duplication (FLT3‐ITD) are common in resistant patients and are linked to activation of mitogen‐activated protein kinase (MAPK) signalling and increased expression of anti‐apoptotic proteins such as myeloid cell leukaemia 1 (MCL‐1) and b‐cell lymphoma‐extra large (BCL(x)L). Murine Ba/F3 cells with different FLT3‐ITD variants were lentiviral transduced to express either wild‐type PTPN11 (Src‐homology 2 containing PTP) or the activating PTPN11‐E76K mutation. Cells were treated with VEN, the MCL‐1 inhibitor S63845 and the mitogen‐activated protein kinase (MEK) inhibitor trametinib (TRA), alone or in…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
FIGURE 1
FIGURE 2
FIGURE 3
FIGURE 4
FIGURE 5
FIGURE 6
FIGURE 7- —Deutsche Forschungsgemeinschaft10.13039/501100001659
- —Interdisziplinäres Zentrum für Klinische Forschung, Universitätsklinikum Jena
- —José Carreras Leukämie‐Stiftung10.13039/501100005677
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsProtein Tyrosine Phosphatases · Chronic Myeloid Leukemia Treatments · Chronic Lymphocytic Leukemia Research
INTRODUCTION
The current treatment strategy for acute myeloid leukaemia (AML) includes the B‐cell lymphoma 2 (BCL‐2) inhibitor venetoclax (VEN), which has shown high clinical efficacy and is well established in elderly patients.1 Nevertheless, up to 30% of patients exhibit resistance to combination therapy with VEN and hypomethylating agents, posing a significant treatment challenge.2 Molecular analyses have identified activating mutations in the protein tyrosine phosphatase (PTP) non‐receptor type 11 (PTPN11), encoding for SH2 containing PTP (SHP‐2), with a reported prevalence of approximately 8% in AML cases, conferring resistance. These mutations are associated with markedly lower complete response rates and reduced overall survival. Moreover, PTPN11 mutations frequently co‐occur with a monocytic disease phenotype and are found alongside mutations in the FMS like tyrosine kinase 3 with internal tandem duplication (FLT3‐ITD) and rat sarcoma (RAS) mutations in 17% and 21% of cases respectively.3 These co‐mutations are known to contribute to primary or secondary resistance mechanisms, leading to notably impaired clinical outcomes and limiting the efficacy of targeted therapies.4, 5, 6, 7 Therefore, co‐targeting downstream signalling pathways may represent a promising strategy to overcome drug resistance and improve therapeutic efficacy.
The combination of VEN and hypomethylating agents has been shown to effectively target leukaemic stem cells (LSCs) by inhibiting oxidative phosphorylation (OXPHOS).8 It has been previously demonstrated that AML cells harbouring an activating PTPN11 mutation exhibit significantly increased OXPHOS activity, which results in reduced VEN sensitivity and decreased targeting of critical LSCs. This effect is likely mediated by a primary or adopted upregulation of myeloid cell leukaemia 1 (MCL‐1), representing a key resistance mechanism against VEN.9 Notably, the PTPN11‐E76K mutant is commonly observed in AML and is known to preferentially upregulate MCL‐1, potentially leading to reduced sensitivity to MCL‐1 inhibition.10 Furthermore, FLT3‐ITD mutations strongly upregulate MCL‐1 expression through constitutive activation of the transducer and activator of transcription 5 (STAT5).11 The co‐occurrence of FLT3‐ITD and PTPN11 mutations suggests a pivotal role of MCL‐1 in AML cell survival and therapy resistance. Several studies have investigated combination therapies utilizing novel MCL‐1 inhibitors (MCL1is) and demonstrating their ability to selectively induce apoptosis even in VEN‐resistant AML cells. These findings suggest MCL‐1 inhibition as a promising strategy to overcome resistance mechanisms and improve treatment efficacy.12, 13, 14
Previous studies have demonstrated that PTPN11 plays a crucial role in FLT3‐ITD signalling, promoting leukaemogenesis. For instance, knockdown of PTPN11 in murine FLT3‐ITD cells resulted in reduced extracellular signal‐regulated kinase (ERK) phosphorylation, indicating a functional impact of PTPN11 on the mitogen‐activated protein kinase (MEK) signalling pathway and inducing aberrant growth in multiple haematopoietic compartments.15, 16, 17 The anti‐apoptotic rapidly accelerated fibrosarcoma (RAF)/MEK/ERK pathway is frequently activated in AML patients, often driven by mutated oncogenic tyrosine kinases such as RAS or FLT3. This pathway also plays a crucial role in regulating BCL‐2 family members, including MCL‐1, further contributing to drug resistance. Meanwhile, clinical trials have shown that MEK inhibitor monotherapy lacks significant efficacy, highlighting the potential of combination strategies to effectively target this pathway and overcome resistance.18, 19, 20, 21 The combination of the MEK inhibitors cobimetinib or alvocidib with VEN has shown promising therapeutic responses in various AML cell line and xenograft models. This effect is primarily attributed to the downregulation of MCL‐1, suggesting a potential strategy to overcome VEN resistance and enhance treatment efficacy.22, 23 A recent study demonstrating VEN resistance in RAS‐mutated LSCs further underscores the critical role of this pathway in mediating resistance. These findings highlight the RAF/MEK/ERK signalling axis as a key target for overcoming VEN resistance in AML.24
The presented study sought to characterize the molecular mechanisms underlying VEN resistance in PTPN11‐ and FLT3‐ITD mutated AML cell line models and primary AML samples and aimed to develop pharmacological strategies to overcome VEN resistance.
METHODS
Cell lines
The study utilized two Ba/F3 cell lines: FLT3‐ITD 598/599 and FLT3‐ITD G613E. These stable cell lines were generated by retroviral transduction with particles encoding either the FLT3‐ITD 598/599 mutation or the FLT3 tyrosine kinase mutation G613E, which conferred interleukin 3 (IL‐3)‐independent growth.25 Subsequently, both cell lines were transduced with viral particles encoding either wild‐type (WT) PTPN11, the PTPN11 E76K mutant or a green fluorescent protein (GFP)‐positive control. The PTPN11 and GFP‐encoding plasmids (pMSCV‐PTPN11‐IRES‐GFP for both WT and E76K variants) were generously provided by the working group of Dr. Kevin Shannon, University of California, San Francisco, CA, United States.17 Finally, GFP‐positive cells were isolated by flow cytometric sorting.
The human AML cell lines U937 and MOLM13 were purchased from ‘Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH’ (DSMZ, Braunschweig, Germany). Figure 1 provides an overview of the cell lines used in this study. Detailed information about the cultivation of primary cells is shown in Supporting Information S1.
Outline of the generation of Ba/F3 cell line models with distinct FMS like tyrosine kinase 3 with internal tandem duplication (FLT3‐ITD) and protein tyrosine phosphatase (PTP) non‐receptor type 11 (PTPN11) variants used in this study. Ba/F3 cells stably expressing FLT3 ITD 598/599 or G613E were transduced with indicated PTPN11 or GFP encoding viral particles and subsequently sorted for similar GFP levels.
Reagents
ABT‐199 (VEN), gilteritinib, trametinib (TRA), S6385 (MCL1i) and ABT‐737 (dual BCL‐2 family inhibitor) were purchased from Selleck Chemicals (Houston, TX, USA) and dissolved in dimethylsulfoxide (DMSO). A list of the used antibodies is shown in Supporting Information S1.
PrestoBlue assay
Cell metabolic activity was assessed using the PrestoBlue reagent (Thermo Fisher Scientific, Waltham, MA, USA). A 9x10^‐5^L cell suspension was plated into 96‐well plates, followed by the addition of 1x10^‐5^ L PrestoBlue reagent. After a 30‐min incubation period, fluorescence was measured using a Clariostar plate reader. Data analysis was conducted with Mars Data Analysis Software (Labtech, Ortenberg, Germany).
Protein isolation and western blot analysis
Following incubation, protein lysates were prepared according to established protocols. The lysates were separated using sodium dodecyl sulphate‐polyacrylamide gel electrophoresis and transferred onto a polyvinylidene fluoride membrane. After a 1‐h blocking step, the membrane was incubated with the primary antibody overnight at 4°C or for 1 h at 37°C with continuous shaking. Following 1‐h incubation with the secondary antibody at room temperature (RT), the chemiluminescence reagent was applied to visualize the membrane. Digital images were captured using the Vilber Fusion FX system (FX7, Vilber Lourmat, France).
Flow cytometry
Following incubation, cells were washed twice with cold phosphate‐buffered saline (PBS) and resuspended in 2x10^‐5^L of staining solution containing 2x10^‐5^L Annexin V binding buffer containing 1x10^‐6^L allophycocyanin (APC)‐labelled Annexin and were incubated for 30 min in the dark at RT. After the addition of 400x10^‐4^ L Annexin V binding buffer per sample, the cells were analysed by flow cytometry using a FACSCalibur (Becton Dickinson, Heidelberg, Germany).
Statistics
Statistical analysis was performed using two‐sided anova and two‐sided t‐tests, with p‐values <0.05 considered statistically significant (****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05, ns indicating no significance). Data were analysed using Prism 9 Software (GraphPad Software, Inc., La Jolla, CA, USA). Unless stated otherwise, values represent the mean ± standard deviation from biological triplicates. Flow cytometry data were analysed with FlowJo Software Version 9 (Ashland, OR, USA), and immunoblot quantification was carried out using ImageJ software (1.52v). Determination of synergy is explained in Supporting Information S1.
RESULTS
Activating PTPN11‐E76K mutation confers multi‐drug resistance
To assess the impact of the PTPN11 mutation on apoptosis and metabolic activity under VEN treatment, flow cytometry and metabolic assays were performed. These revealed a significantly reduced induction of apoptosis. Specifically, 18% of Ba/F3 FLT3‐ITD PTPN11‐E76K cells and 17% of Ba/F3 FLT3‐ITD PTPN11‐WT cells underwent cell death following VEN therapy (Figure 2A–D). Non‐linear regression analysis showed half maximal inhibitory concentration (IC_50_) values for Ba/F3 FLT3‐ITD 598/599 GFP‐control, PTPN11 WT and PTPN11‐E76K mutants as 4.1, 5.8 and 6.4 μM respectively. For FLT3‐ITD G613E‐expressing cells, the corresponding IC_50_ values were 3.8, 7.0 and 8.4 μM (Figure 2C,D). The most pronounced effects on cell viability were observed at VEN concentrations between 3 and 7 μM. Furthermore, a non‐significant trend towards reduced sensitivity to the FLT3 inhibitor gilteritinib was observed in PTPN11‐E76K–expressing cells, supporting FLT3‐independent signalling driven by the activated phosphatase (Figure 2E,F).
*Apoptosis in response to venetoclax (VEN) and gilteritinib treatment and impact on signal transduction. (A, B) Apoptosis was assessed by Annexin V staining and flow cytometry after 48 h incubation with the indicated concentrations of VEN in Ba/F3 FMS like tyrosine kinase 3 with internal tandem duplication (FLT3‐ITD) 598/599 (A) and Ba/F3 FLT3‐ITD G613E (B) cells expressing GFP (control), protein tyrosine phosphatase (PTP) non‐receptor type 11 (PTPN11) wild‐type (WT), or the PTPN11 E76K mutant. Bars and symbols represent mean ± standard deviation (SD) from three independent experiments performed in technical triplicates. Statistical analysis was performed using two‐way anova (*p < 0.05, **p = 0.01, ***p < 0.001). (C–F) Cell viability was measured using PrestoBlue assay and normalized to the DMSO control (0.01% v/v) for VEN (C, D) and gilteritinib (E, F). Data are presented as mean ± SD from three independent experiments performed in technical duplicates. Statistical significance was determined using two‐way anova (*p < 0.05, **p = 0.01, **p = 0.001). (G, H, J, K) Western blot analysis of Ba/F3 FLT3‐ITD 598/599 and G613E cells expressing GFP, PTPN11 WT, or E76K. Panels show phosphorylation and total expression of indicated proteins. One representative Western blot is shown while fold‐changes represent values of biological triplicates. Statistical significance was determined using one‐sample t‐test compared with GFP control. (L, M) Changes in myeloid cell leukaemia 1 and BCL(x)L expression after 48‐h treatment with 3 μM VEN are shown. One representative Western blot is presented, with fold‐change quantification based on biological duplicates.
To assess the impact of the PTPN11 mutation on signalling transduction, western blot analysis was performed on untreated cell lines, revealing a distinct effect of the activating PTPN11‐E76K mutation on key anti‐apoptotic signalling pathways (Figure 2). Although no changes were observed in the expression of upstream targets such as FLT3 or phosphorylatedSTAT5 (pSTAT5) (data not shown), pERK levels were significantly elevated, showing up to a fourfold increase compared to DMSO controls, indicating enhanced MAPK pathway activation (Figure 2G). Additionally, higher levels of anti‐apoptotic proteins, including MCL‐1 and BCL(x)L, known mediators of VEN resistance, were detected (Figure 2J,K). Notably, even PTPN11‐WT overexpression led to increased phosphorylation of PTPN11, albeit to a lesser extent than the E76K mutant (Figure 2H).
Following 48‐h incubation with 3 μM VEN, a compensatory upregulation of MCL‐1 and BCL(x)L was observed across all Ba/F3 cell lines (Figure 2L,M). To overcome this, the MCL‐1‐inhibitor (MCL1i) S6385 was introduced, showing no effect on cell viability up to a concentration of 10 μM in monotherapy (Figure 3). However, the combination of VEN and MCL1i induced strong synergistic effects, promoting apoptosis. Despite this, a significantly reduced apoptosis induction was still observed in both PTPN11 E76K‐expressing Ba/F3 FLT3‐ITD cell lines (Figure 3A,B). Additionally, targeting both BCL‐2 and BCL(x)L with the dual inhibitor ABT‐737 failed to overcome the observed resistance (Figure 3D).
*Apoptosis induction and viability of Ba/F3 FMS like tyrosine kinase 3 with internal tandem duplication (FLT3‐ITD) cells treated with venetoclax (VEN), myeloid cell leukaemia 1 (MCL‐1) inhibitor (MCL1i), or ABT‐737. (A, B) Apoptosis was assessed by Annexin V staining and flow cytometry after 24 h treatment of Ba/F3 FLT3‐ITD 598/599 (A) and Ba/F3 FLT3‐ITD G613E (B) cells expressing GFP (control), protein tyrosine phosphatase (PTP) non‐receptor type 11 (PTPN11) wild‐type or the PTPN11 E76K mutant. Cells were treated with DMSO (0.01% v/v), VEN (3 μM), MCL1i (1 μM) or a combination of both (combi). (C) Cell viability of Ba/F3 FLT3‐ITD 598/599 cells after 48 h treatment with increasing concentrations of MCL1i, measured by PrestoBlue assay and normalized to DMSO control. (D) Dose–response analysis of Ba/F3 FLT3‐ITD 598/599 and G613E cells treated with the indicated concentrations of ABT‐737 for 24 h measured by Presto Blue viability assay. Bars and symbols represent mean ± SD from two independent experiments performed in technical triplicates. Statistical analysis was performed using two‐way anova (***p < 0.0001).
Additional MEK inhibition induces strong apoptotic effects in Ba/F3 FLT3‐ITD cell lines
Based on the pronounced ERK activation observed in western blot analysis, we investigated whether MEK inhibition with TRA could sensitize PTPN11‐mutant cells to apoptosis when combined with BCL‐2 and MCL1is (Figure 4). TRA (0.3 μM) was combined with VEN (0.05 μM) and the MCL1i (1 μM), each at concentrations that induced minimal apoptosis when used alone. In Ba/F3 FLT3‐ITD 598/599 cells (Figure 4A), this triple combination triggered robust apoptosis, reaching 64% in GFP controls, 76% in PTPN11‐WT and 59% in E76K‐overexpressing cells—compared to ≤5% apoptosis with any single agent. Similarly, in Ba/F3 FLT3‐ITD G613E cells, apoptosis rates were 58% (GFP), 72% (WT) and 46% (E76K) after combination treatment (Figure 4B). While overall highly effective, the E76K mutant exhibited consistently reduced apoptotic responses in both cell models. Formal synergy calculation by zero interaction potency (ZIP)‐ and highest single agent (HSA)‐scoring confirmed the synergistic effect of the triple combination in the GFP and WT‐control cells, while E76K‐overexpressing cells showed only additive or even antagonistic effects (Tables S2 and S3; Figures S1 and S2).
*Apoptosis induction in Ba/F3 FMS like tyrosine kinase 3 with internal tandem duplication (FLT3‐ITD) protein tyrosine phosphatase (PTP) non‐receptor type 11 (PTPN11)‐mutated cells following combined treatment with venetoclax (VEN), myeloid cell leukaemia 1 (MCL‐1)‐inhibitor (MCL1i) and trametinib (TRA). (A, B) Apoptosis was assessed by flow cytometry using Annexin V staining after 24 h treatment of Ba/F3 FLT3‐ITD 598/599 (A) and Ba/F3 FLT3‐ITD G613E (B) cells expressing GFP (control), PTPN11 wild‐type (WT) or the PTPN11 E76K mutant. Cells were treated with DMSO (0.03% v/v), VEN (0.05 μM), MCL1i (1 μM) and/or mitogen‐activated protein kinase inhibitor TRA (0.3 μM) as indicated, alone or in combination. Bars represent mean ± SD from four independent experiments, each conducted in technical triplicates. Statistical analysis was performed using two‐way anova (***p < 0.0001). (C, D) Western blot analysis of Ba/F3 FLT3‐ITD 598/599 and G613E cells expressing GFP, PTPN11 WT or E76K. Panels show basal phosphorylation and total expression of indicated proteins after incubation with DMSO (0.01% v/v) or 0.3 μM TRA for 24 h. One representative western blot is shown while fold changes were calculated from biological duplicates.
Western blot analysis following TRA treatment demonstrated complete inhibition of ERK phosphorylation, accompanied by a compensatory increase in phosphorylated PTPN11. TRA monotherapy modestly affected MCL‐1 and BCL(x)L levels, showing a trend towards BCL(x)L downregulation (Figure 4C,D).
VEN‐resistant and PTPN11‐mutated human AML cell line U937 shows sensitivity towards dual MCL‐1 and BCL‐2 inhibition
Dose–response analyses demonstrated significant resistance of the PTPN11‐mutated, FLT3 WT, human U937 cell line to VEN monotherapy, with an IC_50_ >10 μM (Figure 5A). In contrast, the FLT3‐ITD‐mutated, PTPN‐11 WT, human MOLM13 cell line was sensitive to VEN treatment (IC_50_ = 40.7 nM; Figure 5B) and TRA, while being resistant towards treatment with the MCL1i (IC_50_ 64 = nM; above 10 μM respectively). Additionally, U937 cells demonstrated reduced sensitivity to single‐agent MCL‐1 inhibition (IC_50_ = 174 nM) and were resistant to TRA treatment (IC_50_ >10 μM), consistent with a drug‐resistant phenotype.
Drug sensitivity in human U937 and MOLM13 cell lines. (A) Dose–response curves showing cell viability of U937 cells after 24 h treatment with venetoclax (VEN), trametinib (TRA) or myeloid cell leukaemia 1 (MCL‐1) inhibitor (MCL1i). Viability was normalized to DMSO control and measured by absorbance at 490 nm. IC50 values are indicated. (B) Dose–response curve of MOLM13 cells treated with increasing concentrations of VEN, MCL1i and TRA for 24 h. (C) Apoptosis rates in U937 cells after 24 h treatment with indicated single agents or their combinations: VEN (100 nM), MCL1i (50 nM) and TRA (100 nM). (D) Apoptosis rates in MOLM13 cells after 24 h treatment with indicated single agents and their combinations and concentrations for VEN, MCL1i and TRA. Apoptosis was assessed by Annexin V staining and flow cytometry. (E) Corresponding western blot to figure (C) for U937 cells incubated for 6 h with the indicated drugs. Down: Quantification of protein expression normalized to the DMSO control. GAPDH served as a loading control. One representative blot from three independent biological replicates is shown. (F) Western blot analysis of indicated proteins in untreated MOLM13 and U937 cells. Glyceraldehyde 3‐phosphate dehydrogenase (GAPDH) was used as a loading control. Bars and data points represent mean ± SD from at least two independent experiments.
Despite minimal apoptosis induction by single agents or other dual treatments, the combination of VEN (100 nM) and the MCL1i (50 nM) led to robust apoptosis in 68% of U937 cells after 24 h, compared to 10%–39% for other dual combinations (Figure 5C). The addition of TRA did not further increase this effect (71% apoptosis).
Mechanistically, western blot analysis of untreated MOLM13 and U937 cells showed higher basal levels of phosphorylated protein kinase B (pAKT), phosphorylated myeloid cell leukemia 1 pMCL‐1 and BCL(x)L in U937 cells, which may contribute to their VEN resistance (Figure 5F). After 6 h of treatment, cells exhibited MCL‐1 upregulation and effective ERK phosphorylation suppression following TRA treatment (Figure 5E). Accordingly, MOLM13 cells showed a differential response: they were more sensitive to the VEN+TRA combination than to VEN+MCL‐1i, consistent with their greater BCL‐2‐ and lower MCL‐1 dependence and their higher intrinsic sensitivity to TRA. The triple combination increases apoptosis rate from 57.3% to 70.4% while monotreatment with MCL1i did not induce any apoptosis (Figure 5D).
Combination of MCL‐1 and BCL‐2 inhibition is effective in primary AML samples
Characterization of five primary AML patient samples (#1–5) revealed distinct patterns in key pro‐survival signalling pathways and apoptotic regulators (Figure 6A). Especially, patient #4, harbouring high‐risk mutations including FLT3‐ITD, PTPN11 and TP53, showed elevated levels of MCL‐1 and BCL(x)L, alongside increased phosphorylation of ERK and AKT, indicating increased pro‐survival signalling. In contrast, samples #1–3, classified within the favourable molecular Prognostic Risk Score (mPRS) group, showed lower pathway activation, while sample #5 was intermediate and #4 unfavourable. Consistent with their FLT3‐ITD mutations, patients #4 and #5 also displayed pronounced pSTAT5 phosphorylation. Detailed clinical and molecular data are provided in Table S1. Assessing four additional VEN‐resistant patient samples, western blotting showed again the strongest ERK phosphorylation in the two cases with PTPN11 mutations (#6, #7). All four exhibited elevated MCL‐1 and/or BCL‐(x)L, consistent with a mechanism of VEN resistance. Apoptosis assays of the combination treatments yielded results comparable to Figure 6 (Figure S2A–C).
*Venetoclax (VEN) sensitivity, signalling profiles and apoptosis induction in primary AML patient samples. (A) Western blot analysis of indicated proteins with GAPDH serving as a loading control in five primary acute myeloid leukaemia (AML) samples (#1–#5). Right panels show densitometric quantification of phosphorylated‐to‐total protein ratios and protein expression levels normalized to GAPDH and patient #1. (B) Clinical and molecular characteristics of patient samples (#1–5), including mutation status and molecular Prognostic Risk Score‐based risk classification. Below: Dose–response curves of selected samples (#1, #2, #4, #5) after 24‐h treatment with increasing concentrations of VEN. IC50 values are indicated and calculated by non‐linear regression. (C) Apoptosis induction in samples #1, #2, #4 and #5 following 24 h treatment with the indicated agents: VEN (30 nM), myeloid cell leukaemia 1 inhibitor (30 nM), trametinib (30 nM) or combination. Apoptosis was measured by Annexin V staining and flow cytometry. (D) Summary of apoptosis induction across all patient samples (n = 4). Data are presented as box‐and‐whisker plots. Bars and curves represent mean ± SD from at least two biological replicates. Statistical significance was determined using one‐way anova (****p < 0.0001; ***p < 0.001; *p < 0.01).
These molecular patterns correlated with VEN sensitivity assessed by metabolic assays: patient #2 (favourable mPRS) showed pronounced sensitivity (IC_50_ = 25.8 nM), patient #1 (favourable mPRS) exhibited moderate sensitivity (IC_50_ = 129 nM) and patient #4 (unfavourable mPRS) was highly resistant (IC_50_ >10 μM) (Figure 6B). Apoptosis assays strongly correlated with the sensitivity data obtained by viability tests using PrestoBlue (#4: 13.5%, #5: 25.6%) compared to favourable patient #1 and #2 (37.3% and 56.3% respectively). Dual treatment with an MCL1i enhanced apoptosis in patients with favourable mPRS‐status (e.g. #2: 97.5%, #1: 77.1%; Figure 6C). Aggregate analysis across samples #1–5 verified that the combination of VEN and MCL‐1 inhibition significantly increased apoptosis compared to single‐agent treatment (p < 0.01) (Figure 6D).
DISCUSSION
Resistance to VEN remains a major clinical challenge in AML treatment. Among downstream effectors of RAS/MAPK signalling, upregulation of the anti‐apoptotic protein MCL‐1 is a well‐established driver of VEN resistance.12 Preclinical data have demonstrated that dual targeting of BCL‐2 and phosphoinositide 3 kinase (PI3K) can downregulate MCL‐1 and restore apoptotic priming via BCL‐2 associated x protein (BAX) activation.26 Additionally, direct MCL‐1 inhibition has shown efficacy in AML models with RAS pathway activation, including PTPN11 mutations.9 Our data corroborate these findings, showing that VEN induces compensatory upregulation of MCL‐1 and BCL(x)L, which is effectively countered by co‐inhibition of MCL‐1, leading to synergistic apoptosis in AML cell lines and primary samples. Comparable results from Luedtke et al. emphasize the dependence on apoptotic effectors Bim, Bax and Bak for the success of dual MCL‐1/BCL‐2 targeting. Nonetheless, they also raised concerns regarding the impact of this combination on normal haematopoietic stem cells and proposed indirect MCL‐1 downregulation—such as through cyclin‐dependent kinase 9 (CDK9), exportin 1 (XPO1) or PI3K inhibitors—as a strategy to enhance VEN sensitivity while sparing toxicity.27, 28, 29
Despite promising preclinical data, initial clinical studies with selective MCL1is (e.g. AZD5991) have shown limited efficacy and raised safety concerns such as cardiac toxicity. Similarly, early clinical trials evaluating MEK inhibition by TRA in combination with VEN and hypomethylating agents in relapsed/refractory AML failed to demonstrate substantial benefit and were associated with significant toxicity.30, 31 In the VEN‐resistant monocytic cell line U937, harbouring the PTPN11 G60D mutation but lacking the FLT3‐ITD, TRA did not further enhance the already effective VEN and MCL1i combination. Similar results were observed in the primary AML samples, suggesting that TRA efficacy is cell model dependent. Although patient numbers are very limited and culture conditions may have influence, this preclinical testing aligns molecular VEN sensitivity with mPRS‐based clinical risk stratification and may support resistance mechanisms driven by MCL‐1 and MAPK activation.
The presented results here demonstrate that even combined BCL‐2/BCL(x)L inhibition fails to overcome resistance in PTPN11‐mutated contexts, underscoring the dominant role of MCL‐1 in RAS‐driven AML.32, 33 Although monotherapies targeting MCL‐1 or MEK show limited clinical benefit, our data support combination strategies as a promising approach to overcome resistance. Importantly, synergistic apoptosis was achieved at sub‐cytotoxic doses, highlighting the potential for low‐dose, sequential or intermittent regimens to minimize haematological and cardiac toxicity.
Recent studies highlight the critical role of RAS pathway activation in mediating this resistance. Especially LSCs harbouring RAS mutations drive altered BCL‐2 family gene expression and promote monocytic differentiation, which correlates with diminished clinical responses.24 In relapsed or refractory AML post‐VEN treatment, RAS/MAPK pathway activation rapidly emerges through the clonal selection of RAS‐mutated subclones.34 In this setting, targeting upstream modulators of RAS signalling, such as son of sevenless homolog 1 (SOS1), has emerged as a promising strategy, with preclinical data demonstrating synergy between SOS1 inhibition and VEN.35 Similarly, in janus kinase 2 (JAK2)‐mutated neoplasms, MAPK‐driven anti‐apoptotic signalling via Y‐box binding protein 1 (YBX1) contributes to treatment resistance.36, 37
Mention must be made of the limitationsin our study. Our work relies mainly on engineered cell lines with a small set of primary AML samples. While these models enable clean mechanistic evaluation, they do not capture full patient heterogeneity. Moreover, overexpression systems and the focus on the E76K allele may not generalize to different PTPN11 variants or co‐mutation contexts.
Drug studies were performed in ex vivo assays without in vivo efficacy and safety (including the known cardiac toxicity of MCL‐1 inhibition). Thus, the results are primarily hypothesis generating and require further optimization in larger primary cohorts.
Nonetheless, the presented findings corroborate that PTPN11 mutations, particularly the activating E76K variant, contribute significantly to VEN resistance by sustaining MAPK signalling and increasing MCL‐1 and BCL(x)L expression. In the co‐mutated Ba/F3 cell line model, resistance was only overcome by simultaneous inhibition of BCL‐2 and MCL‐1. Addition of MEK inhibition via TRA enhanced apoptosis synergistically, though this effect was attenuated in E76K‐overexpressing cells. Notably, the low VEN dose (50 nM) was ineffective alone, emphasizing the critical synergy of the triple combination.
Ongoing clinical trials investigating triplet regimens, such as VEN+5‐azacitidine in combination with FLT3 (e.g. gilteritinib) or isocitrate dehydrogenase 1 and 2 (IDH1/2) inhibitors (e.g. ivosidenib), have reported high response rates, albeit with increased myelosuppression.32, 33 The presented findings suggest that analogous triplet strategies, incorporating MEK and MCL‐1 inhibition, may be particularly beneficial in a subset of AML with RAS pathway activation (Figure 7).
Schematic representation of FLT3‐ITD signalling and mechanisms of treatment resistance. FLT3‐ITD activates downstream signalling pathways, including RAS/RAF/MEK/ERK, PI3K/AKT/mTOR (mechanistic target of rapamycin) and JAK/STAT pathways, promoting cell survival and resistance to apoptosis. The PTPN11 mutation enhances RAS activation, leading to increased signalling through MEK/ERK. Key inhibitors targeting these pathways are indicated: Gilteritinib (FLT3 inhibitor), trametinib (MEK inhibitor), venetoclax (BCL‐2 inhibitor) and S63845 (MCL‐1 inhibitor).
Taken together, the provided data emphasize the mechanistic relevance of PTPN11 and FLT3 mutations in conferring VEN resistance and provide rationale for combinatorial targeting of MAPK signalling and anti‐apoptotic BCL‐2 family proteins in AML.
AUTHOR CONTRIBUTIONS
MF, OH, DV, JB, L‐JS and SB were responsible for data collection and performing experiments. US, MFi, FHH, TMS and JPM were responsible for methodology and resources; JPM, AH and SS supervised the project and edited the manuscript. MF wrote the manuscript.
FUNDING INFORMATION
FHH was supported by the German Research Council under grant numbers HE6233/8‐1 (project number 449291356), HE6233/9‐1 (project number 453491106), HE6233/10‐1 (project number 505859092), and in part by HE6233/15‐1 (project number 517204983). DV and OH were supported by a grant from the ‘Institute and Centrum for Clinical Research’ (IZKF), Jena. MF received funding from the Deutsche José Carreras Leukämie Stiftung (DJCLS 02 CS/2024) and the ‘Institute and Centrum for Clinical Research’ (IZKF), Jena (grant CSP‐07). JPM was supported by the German Research Foundation (DFG) under grant numbers MU955/14‐2 and MU955/15‐1.
CONFLICT OF INTEREST STATEMENT
FHH has served on advisory boards for Novartis, CTI, BMS/Celgene, Janssen, AbbVie, GSK, Merck, Silence Therapeutics, Prelude Therapeutics and AOP, and has received research funding from Novartis, Celgene/BMS and CTI. MF has received travel support from AbbVie, Jazz Pharmaceuticals, AOP and Lilly. SS has served on advisory boards for Amgen, Novartis and Daiichi Sankyo; received honoraria from Novartis; and travel support from AbbVie. AH received research support from Novartis, BMS, Pfizer, Incyte, Enliven and TERNS, and honoraria from Novartis and Incyte. US received travel support and honoraria from AbbVie, AstraZeneca, BeiGene, BMS, Gilead/Kite, Janssen, Novartis, SOBI and Takeda. All other authors declare no conflicts of interest.
Supporting information
Figure S1. ZIP synergy analysis of venetoclax, MCL‐1 inhibitor, and trametinib combinations in Ba/F3‐FLT3(598/599) cells.
Figure S2. Venetoclax sensitivity, signalling profiles, and apoptosis induction in additional four primary AML patient samples.
Table S1. Patient characteristics of primary AML samples showed in Figure 6 and Figure S2. Table S2. HSA Synergy score calculation for apoptosis assay of the triple combination showed in Figure 4A,B. Table S3. Calculation of indicated Synergy Scores of experiments shown in Figure S1.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Shahswar R , Gabdoulline R , Krueger K , Wichmann M , Götze KS , Braitsch K , et al. A novel prognostic risk model for patients with refractory/relapsed acute myeloid leukemia receiving venetoclax plus hypomethylating agents. Leukemia. 2025;39(3):614–622.39779979 10.1038/s 41375-024-02501-6PMC 11879869 · doi ↗ · pubmed ↗
- 2Di Nardo CD , Jonas BA , Pullarkat V , Thirman MJ , Garcia JS , Wei AH , et al. Azacitidine and venetoclax in previously untreated acute myeloid leukemia. N Engl J Med. 2020;383(7):617–629.32786187 10.1056/NEJ Moa 2012971 · doi ↗ · pubmed ↗
- 3Scholl S , Fleischmann M , Schnetzke U , Heidel FH . Molecular mechanisms of resistance to FLT 3 inhibitors in acute myeloid leukemia: ongoing challenges and future treatments. Cells. 2020;9(11):2493.33212779 10.3390/cells 9112493 PMC 7697863 · doi ↗ · pubmed ↗
- 4Stasik S , Eckardt JN , Kramer M , Röllig C , Krämer A , Scholl S , et al. Impact of PTPN 11 mutations on clinical outcome analyzed in 1529 patients with acute myeloid leukemia. Blood Adv. 2021;5(17):3279–3289.34459887 10.1182/bloodadvances.2021004631 PMC 8525221 · doi ↗ · pubmed ↗
- 5Alfayez M , Issa GC , Patel KP , Wang F , Wang X , Short NJ , et al. The clinical impact of PTPN 11 mutations in adults with acute myeloid leukemia. Leukemia. 2021;35(3):691–700.32561839 10.1038/s 41375-020-0920-z · doi ↗ · pubmed ↗
- 6Ravikumar VI , Leonard JT , Newell LF , Gandhi A , Meyers G , Hayes‐Lattin B , et al. PTPN 11 mutations are associated with shorter survival in patients with acute myeloid leukemia treated with hypomethylating agents and venetoclax. Blood. 2024;144:4312.
- 7Bataller A , Bazinet A , Di Nardo CD , Maiti A , Borthakur G , Daver NG , et al. Prognostic risk signature in patients with acute myeloid leukemia treated with hypomethylating agents and venetoclax. Blood Adv. 2024;8(4):927–935.38113472 10.1182/bloodadvances.2023011757 PMC 10877112 · doi ↗ · pubmed ↗
- 8Chen X , Glytsou C , Zhou H , Narang S , Reyna DE , Lopez A , et al. Targeting mitochondrial structure sensitizes acute myeloid leukemia to venetoclax treatment. Cancer Discov. 2019;9(7):890–909.31048321 10.1158/2159-8290.CD-19-0117 PMC 6606342 · doi ↗ · pubmed ↗