JMC14: a novel dual PI3Kδ/CSF1R inhibitor with potent antitumor activity in hematological and solid tumors

Abstract

PI3Kδ, predominantly expressed in immune cells and markedly dysregulated in B-cell malignancies, emerges as a promising and well-validated therapeutic target in hematologic cancers. Meanwhile, CSF1R regulates the formation and polarization of tumor-associated macrophages (TAMs), facilitating immune suppression and tumor progression in various solid tumors. Although targeting PI3Kδ or CSF1R has shown promise, the clinical application is often constrained by off-target effects, toxicity, and limited efficacy, particularly in solid malignancies. In this study, we identified JMC14, a novel dual inhibitor targeting PI3Kδ and CSF1R with a distinct structure and favorable selectivity among human kinome, yielding IC₅₀ values of 12 nM against PI3Kδ and 143 nM against CSF1R, respectively. JMC14 preferentially inhibited PI3Kδ-mediated signaling at the cellular level and exhibited robust antiproliferative activity across 10 lines of diffuse large B-cell lymphoma (DLBCL) cells, outperforming the approved PI3Kδ inhibitor idelalisib. Notably, its efficacy negatively correlated with the PI3Kα expression among the cell lines tested, suggesting a compensatory pathway mediated by PI3Kα. Daily oral administration of JMC14 (10, 30, or 100 mg/kg, for 21 days) dose-dependently suppressed tumor progression in xenografts derived from TMD8 cells and DLBCL patients, accompanied by good tolerance. Additionally, M-NFS-60 myeloid leukemia cells, which are dependent on the CSF-1-CSF1R axis for survival and proliferation, were effectively inhibited by JMC14 both in vitro and in vivo, further validating its inhibitory activity targeting CSF1R. Furthermore, JMC14 demonstrated potent antitumor activity in murine triple-negative breast cancer (TNBC), which was associated with its activity to reshape the immune microenvironment by reducing M2-like TAMs, enhancing CD8⁺ T cell infiltration. Collectively, these findings establish JMC14 as a potent dual PI3Kδ/CSF1R inhibitor with remarkable efficacy against both hematologic and solid malignancies with hyperactivation of PI3Kδ and/or CSF1R, highlighting the potential of JMC14 as a useful probe to dissect the interaction of PI3Kδ and CSF1R in tumor progression and immune reprogramming.

Introduction

Phosphoinositide 3-kinase delta (PI3Kδ) plays a critical role in B-cell receptor (BCR)-mediated signaling, which is essential for activation, proliferation, survival, and migration of B cells [1]. Overexpression or hyperactivation of PI3Kδ is often found in B-cell malignancies, validating it a therapeutic target. Meanwhile, PI3Kα is ubiquitously expressed across various tissues and has been shown to participate in compensatory signaling with PI3Kδ [2, 3]. Over the past few decades, substantial efforts have focused on developing PI3Kδ inhibitors for the treatment of lymphomas [4–9]. To date, four PI3Kδ inhibitors have been approved by the US Food and Drug Administration (FDA) for treating relapsed or refractory indolent lymphomas: idelalisib (PI3Kδ), duvelisib (PI3Kδ/γ), umbralisib (PI3Kδ/CK1ε), and copanlisib (PI3Kα/δ) [10–13]. Although durable clinical responses have been yielded with these agents, a broad spectrum of toxicities, including hepatotoxicity, severe diarrhea, colitis, pneumonitis, infections, and intestinal perforations, have been observed in clinical settings [14, 15]. In particular, first-generation PI3Kδ inhibitors have been linked to liver toxicity, likely attributable to their chemical structures [16]. In contrast, the second-generation PI3Kδ inhibitors such as parsaclisib, with a distinct structure, have shown a significant reduction in toxicity [17], highlighting the critical need for novel inhibitors with improved safety profiles.

In addition to hematologic malignancies, PI3Kδ inhibitors are being explored for the treatment of solid tumors, where they can directly target tumor cells and modulate immune suppressive tumor microenvironment (TME). Notably, PI3Kδ is overexpressed in various solid tumors such as breast and liver cancers, and its expression frequently increases during tumor progression [18]. Preclinical data indicated that pharmacologic inhibition of PI3Kδ exerted significant antitumor effects [19], partly by enhancing effector T-cell (Teff) functions through direct inactivation of regulatory T-cells (Treg) [20] and myeloid-derived suppressive cells (MDSC) [21]. While the precise molecular mechanisms remain unclear, these findings highlight the therapeutic potential of targeting PI3Kδ in solid tumors, where the immunosuppressive microenvironment particularly favors the activity of suppressive cell populations.

CSF1R, a member of class III receptor tyrosine kinase (RTK III) family, shares structural similarities with other receptors such as Flt3 (FMS-like tyrosine kinase 3), c-Kit (stem cell factor receptor), and PDGFR (platelet-derived growth factor receptors) α and β [22]. It is primarily activated by two ligands: colony-stimulating factor 1 (CSF-1) and interleukin 34 (IL-34), which drive proliferation and differentiation of monocyte and macrophage precursors. Aberrant expression or activation of CSF1R has been implicated in the progression of various cancers, including myeloid leukemia [23]. In solid tumors, elevated CSF1R expression facilitates the recruitment and polarization of M2-like macrophages, commonly known as tumor-associated macrophages (TAMs). TAMs are associated with poor prognosis by promoting tumor growth, angiogenesis, invasion, and resistance to therapy [24]. Consequently, targeting the CSF1-CSF1R signaling axis has emerged as a promising therapeutic strategy [25]. Several CSF1R inhibitors, including both small molecules and monoclonal antibodies, are currently undergoing clinical evaluation for the treatment of both solid tumors and hematologic malignancies [26–28]. Turalio (PLX3397), targeting CSF1R, c-Kit, and mutant Flt3, is the only CSF1R inhibitor approved by the FDA for glioblastoma [29]. Additionally, the selective CSF1R inhibitor sotuletinib (BLZ945) is in phase I/II trials for solid tumors [30]. Despite these advancements, there remains an urgent need for the development of novel CSF1R inhibitors with unique chemical scaffolds and distinct mechanisms of action to overcome current limitations and expand therapeutic options.

Given the critical roles of PI3Kδ and CSF1R in tumor progression and immune regulation, a dual-targeted approach may offer synergistic benefits by simultaneously disrupting key oncogenic signals and immunosuppressive pathways. Up to date, no small-molecule inhibitors have been reported that concurrently target both PI3Kδ and CSF1R. Here, we present the characterization of JMC14, a first-in-class dual PI3Kδ/CSF1R inhibitor with a distinct structural profile. JMC14 demonstrates potent antitumor activity against B-cell lymphomas and triple-negative breast cancer (TNBC) in both in vitro and in vivo studies. By concurrently targeting PI3Kδ and CSF1R, JMC14 effectively reprograms the TME to potentiate antitumor immunity. Collectively, our study identified JMC14 not only as a promising lead compound for cancer therapy but also as a valuable molecular probe for elucidating the interaction of PI3Kδ and CSF1R in tumor progression and immune reprogramming.

Materials and methods

Compounds

JMC14 was synthesized as previously described [31]. Idelalisib and sotuletinib were purchased from MeilunBio (Dalian, China). Stock solutions of all compounds were prepared at a concentration of 10 mM in dimethyl sulfoxide (DMSO, Sigma-Aldrich, St.Louis, MO, USA), and stored in aliquots at −20 °C. Prior to each experiment, the compounds were diluted to the concentration needed. For the in vivo experiments, JMC14 was formulated in 5% (v/v) polyethylene glycol (PEG) 300 (Sangon Biotech, Shanghai, China). Idelalisib was prepared in normal saline containing 30% PEG-400, 5% propylene glycol, and 0.5% Tween 80 (v/v, Sangon Biotech, Shanghai, China). Sotuletinib was dissolved in 20% (v/v) 1,3-cyclodextrin (MeilunBio, Dalian, China).

In vitro kinase assay

The PI3K kinase activity of the compounds was evaluated using the PI3-Kinase HTRF™ Assay kit (#33-016, Sigma-Aldrich, St.Louis, MO, USA), following a previously described protocol [32]. Briefly, each enzyme at its EC₈₀ concentration was incubated with 10 μM PIP2 in assay buffer containing the diluted test compounds. The reaction was initiated by the addition of ATP and terminated after incubation with the stop solution at room temperature for 30 min. Light emission intensity was measured using a Synergy H1 microplate reader (Biotek Instruments, Winooski, VT, USA). IC₅₀ values were determined by applying a logistic curve fit to the data with GraphPad Prism 8 (GraphPad Software, Inc., San Diego, CA, USA).

The enzyme activity of CSF1R (#14-551, Millipore, MA, USA) was evaluated using an Enzyme-Linked Immunosorbent Assay (ELISA) as previously reported [8]. In brief, 96-well ELISA plates were pre-coated overnight with 20 µM Poly (Glu, Tyr)_4:1 substrate provided by Sigma-Aldrich (St.Louis, MO, USA). Active CSF1R was incubated with serially diluted test compounds in reaction buffer at 37 °C for 1 h. Following incubation, wells were rinsed with phosphate-buffered saline and then treated with the anti-phosphotyrosine (PY99) antibody and a secondary antibody conjugated to horseradish peroxidase (HRP). HRP activity was determined using o-phenylenediamine (OPD) as the substrate, and the absorbance at 490 nm was measured with a multi-well spectrophotometer (Molecular Devices, Sunnyvale, CA, USA).

The in vitro kinase profiling was conducted by Eurofins Discovery (Dundee, UK).

Cell lines and cell culture

The cell lines Z-138, SU-DHL-6, Jurkat, Pfeiffer, Molt-4, KG-1, HH, JEKO-1, Ramos, MM.1S, THP-1, HEL, PC-3, NIH-3T3, RAW264.7, 4T1, and EMT6 were sourced from the American Type Culture Collection (ATCC, Manassas, VA, USA). WIL-2-NS, KMS-11, and MOLM-13 were obtained from the Japanese Collection of Research Bioresources Cell Bank (JCRB, Osaka, Japan). KARPAS-422, KARPAS-299, U2932, OCI-LY-19, SU-DHL-4, L-428, RL, DB, WSU-DL-CL2, and HT were purchased from the German Collection of Microorganisms and Cell Cultures (DSMZ, Braunschweig, Germany). OCI-LY-10, TMD8, and MV-4-11 originated from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China), while Eol-1 was sourced from the European Collection of Authenticated Cell Cultures (ECACC, Salisbury, UK). The PY8119 cell line was kindly provided by Dr Su-ling Liu from Shanghai Cancer Center and Institutes of Biomedical Sciences, Fudan University. Authentication of all relevant cell lines was confirmed via short-tandem repeat (STR) analysis. Cells were maintained following the guidelines provided by the suppliers.

Cell proliferation assay

For suspended cells, proliferation was evaluated using the Cell Counting Kit-8 (CCK-8, Dojindo, Kumamoto, Japan) according to the manufacturer’s protocol. Briefly, cells were seeded into 96-well plates and treated in triplicate with serial dilutions of test compounds for 72 h. Following the addition of the CCK-8 reagent, optical density (OD) at 450 nm was recorded using a multi-well spectrophotometer. The inhibition rate was calculated as (OD_control – OD_treated – OD_day0)/(OD_control– OD_day0) × 100%, where OD_day0 represents the OD of untreated cells at treatment initiation. The calculation of GI₅₀ values was determined using GraphPad Prism 8.

For adherent cells, proliferation was assessed by sulforhodamine B (SRB) assay as previously described [33]. Inhibition rates were calculated, and IC₅₀ values were determined by fitting four-parameter logistic regression models to dose-response curves using SoftMax Pro (Molecular Devices, Sunnyvale, CA, USA).

Western blot analysis

Cell lysates were collected and subjected to Western blot analysis. Antibodies targeting phospho- or total CSF1R (#3155 and #3152), AKT (#13038, #4060 and #4691), p70S6K (#9205 and #9202), S6 (#2211, #2215 and #2217), 4E-BP1 (#2855 and #9644), Erk1/2 (#4370 and #4695) and the loading control GAPDH (#97166) were purchased from Cell Signaling Technology (Danvers, MA, USA). β-Actin (#A5441) was obtained from Sigma-Aldrich (St.Louis, MO, USA).

Animal experiments

All experimental protocols and procedures were approved and conducted in compliance with the guidelines of the Institutional Animal Care and Use Committee (IACUC) of Shanghai Institute of Materia Medica, Chinese Academy of Sciences (No: 2018-10-MLH-01 and 2021-09-MLH-05). Female Balb/c SCID and C57BL/6 mice, sourced from the Shanghai Institute of Materia Medica (Shanghai, China), were subcutaneously inoculated with tumor cells into their right flanks. Once tumor volumes reached approximately 100 mm³, the mice were randomized into groups to receive oral administration of either the vehicle or test compounds. Patient-derived xenograft (PDX) experiments were performed by WuXi AppTec (Shanghai, China). Body weight and tumor volume were monitored twice weekly. The formula for calculating Tumor volume (V) was V = 0.5 × length × width². The ratio of the tumor volume at a particular time to the initial volume was used to determine the relative tumor volume (RTV). To calculate the treatment-to-control ratio (T/C), the formula T/C (%) = (RTV_treated/RTV_control) ×100% was used, with RTV_treated and RTV_control representing the RTV of the treatment and vehicle control groups.

Pharmacokinetic and pharmacodynamic studies were conducted to examine the distribution and biological effects of JMC14. Mice with TMD8 tumors were randomly assigned to groups (n = 3) and treated with either a vehicle or JMC14 at doses of 30 or 100 mg/kg. Tumor tissues were collected at 1, 2, 4, 8, and 24 h after treatment. JMC14 concentrations in tumor tissues were assessed through high-performance liquid chromatography paired with tandem mass spectrometry (HPLC-MS/MS). Tumor homogenates were prepared and analyzed by Western blot.

Immunohistochemical analysis

Tumor samples were preserved in 4% paraformaldehyde and subsequently embedded in paraffin. Hematoxylin and eosin (H&E) staining and immunohistochemical staining of CD45, CD4, CD8, F4/80, CD206, or CD31 were performed by Shanghai ZuoCheng Bio Company (Shanghai, China). The prepared slides were examined using a Leica DM6 B microscope integrated with a sCMOS camera system (Leica, Wetzlar, Germany).

Statistical analysis

The data shown were from at least two independent experiments. Statistical analyses were performed using GraphPad Prism 8.0. An unpaired two-tailed Student’s t-test was applied to compare differences between two groups, while one-way or two-way analysis of variance (ANOVA) was used for comparisons among multiple groups. A P value of less than 0.05 was considered statistically significant. Unless stated otherwise, data are expressed as mean ± SD. The significance of difference was denoted as follows: * P < 0.05, ** P < 0.01, *** P < 0.001, and **** P < 0.0001.

Results

JMC14 is a potent and selective inhibitor of PI3Kδ and CSF1R

To identify potent and selective PI3K inhibitors, we rationally developed a series of pyrrolo[2,1-f] [1,2,4]triazine-based compounds [34, 35], with JMC14 emerging as a leading candidate (Fig. 1a). As shown in Fig. 1b, JMC14 exhibited potent inhibition against PI3Kδ (IC₅₀ = 12.0 ± 4.6 nM), while displaying minimal activity against other PI3K isoforms, particularly PI3Kγ (IC₅₀ > 10,000 nM). Idelalisib, the first PI3Kδ inhibitor approved by the FDA, demonstrated a comparable inhibitory profile across the four class I PI3K isoforms. To confirm JMC14’s selectivity at the cellular level, we examined its capacity to inhibit AKT phosphorylation at S473, an indicator of PI3K activity, in cells upon selective activation of each PI3K isoform. The PI3K-AKT pathway is selectively activated by platelet-derived growth factor BB (PDGF-BB) via PI3Kα in NIH-3T3 murine fibroblasts. In PC-3 prostate cancer cells, lysophosphatidic acid (LPA) stimulates PI3Kβ through G-protein-coupled receptor signaling, whereas C5a triggers PI3Kγ activation in RAW264.7 macrophages. In contrast, PI3Kδ is exclusively activated by anti-IgM stimulation in Ramos lymphoma cells. As shown in Fig. 1c, JMC14 at 10 nM significantly inhibited AKT phosphorylation in anti-IgM–stimulated Ramos cells, whereas approximately 30-fold higher concentration was required to achieve comparable inhibition in cells with activated PI3Kα, PI3Kβ or PI3Kγ (Fig. 1d). A similar pattern emerged with idelalisib, further supporting the selective targeting of PI3Kδ by JMC14. To further explore the selectivity of JMC14 among human kinome, a comprehensive kinase screening was performed. At a concentration of 10 µM, JMC14 inhibited PI3K isoforms and their mutants as well as CSF1R kinase by over 90% (Fig. 1e, Supplementary Table S1). Notably, JMC14 exhibited an IC₅₀ value of 143.4 nM against CSF1R, which was greater than sotuletinib (IC₅₀ = 1.7 nM), a known CSF1R inhibitor included as a reference compound (Fig. 1f, Supplementary Fig. S1), indicating the inhibitory activity of JMC14 against CSF1R. Taken together, these results established JMC14 as a potent and selective dual inhibitor of PI3Kδ and CSF1R.

Fig. 1. JMC14 is a potent and selective inhibitor of PI3Kδ and CSF1R.

a The chemical structure of JMC14. b Inhibitory activity of JMC14 and idelalisib against isoforms of human class I PI3K. IC₅₀ values shown are mean ± SD from two independent experiments. c JMC14 selectively inhibited PI3Kδ at cellular level. NIH-3T3, PC-3, RAW264.7 and Ramos cells were starved in serum-free medium overnight and then pretreated with JMC14 or idelalisib for 5 to 30 min prior to stimulation with PDGF-BB (20 ng/mL), LPA (10 μM), C5a (200 ng/mL) or anti-IgM (3 μg/mL). Phosphorylated AKT at Ser473 was assessed by Western blot. Representative images are shown. d Relative inhibition of AKT phosphorylation (%) by JMC14 and idelalisib across different PI3K isoforms was quantified. Band intensity was measured using ImageJ. Data are mean ± SD from two independent experiments. e Kinase profiling map of JMC14 against more than 300 human kinases was generated using the TREEspot Software Tool from DiscoveRx KinomeScan™. f The inhibitory activity of JMC14 against CSF1R kinase.

JMC14 inhibits DLBCL cell proliferation via blocking PI3K-mediated signaling

Given the pivotal involvement of PI3Kδ signaling in the growth and proliferation of hematological cancers, we evaluated the anti-proliferative effects of JMC14 in 28 distinct cancer cell lines representing diverse hematological malignancies. JMC14 demonstrated potent anti-proliferative activity, with GI₅₀ values below 1 μM in 35.7% (10/28) of the tested cell lines, 7 of which were derived from diffuse large B-cell lymphoma (DLBCL) (Fig. 2a), indicating that JMC14 may be particularly effective against DLBCL cells. Notably, the GI₅₀ values varied significantly among the tested cell lines, ranging from 54.0 nM in TMD8 cells to more than 10 μM in cell lines such as DB and KARPAS-299. Furthermore, JMC14 exhibited a similar but more potent profile compared to idelalisib, with GI₅₀ values approximately 2–3 times lower than those of idelalisib, indicating that JMC14 demonstrated superior efficacy, though both compounds target similar pathways. To explore the relationship between JMC14’s activity and the expression of PI3K or CSF1R, we examined the correlation between its GI₅₀ values and the protein levels of the four class I PI3K isoforms [32]. A considerable positive correlation (R² = 0.88) was detected between PI3Kα levels and JMC14 GI₅₀ values, suggesting that cell lines with lower PI3Kα expression exhibited greater sensitivity to JMC14 compared to those with higher PI3Kα expression (Supplementary Fig. S2a). Although cells with elevated PI3Kδ expression were generally more sensitive to JMC14, no significant correlation was observed between PI3Kδ protein levels and GI₅₀ values (Supplementary Fig. S2b). Similarly, no correlation was observed between CSF1R mRNA levels and JMC14 activity (Supplementary Fig. S2c).

Fig. 2. JMC14 inhibits DLBCL cell proliferation via blocking PI3K-mediated signaling.

a The anti-proliferative activity of JMC14 and idelalisib against a panel of hematological malignancy cell lines was evaluated using the CCK-8 assay. GI₅₀ values were calculated and plotted. GI₅₀ was assigned as 10 μM in cells where the inhibition rate at 10 μM was less than 50%. Data are presented as mean ± SD from two independent experiments. Abbreviations: AML acute myeloid leukemia, HL Hodgkin’s lymphoma, MM multiple myeloma, TCL T-cell lymphoma. b, c TMD8, SU-DHL-6, and Pfeiffer cells were treated with JMC14 for 1 h at the indicated concentrations, followed by Western blot analysis with indicated antibodies. Band intensity was quantified using ImageJ (c). Data shown are representative (b) or mean ± SEM from two independent experiments (c).

To further investigate the mechanism of action of JMC14, we examined its effects on PI3K signaling in ABC-DLBCL TMD8 cells and GCB-DLBCL SU-DHL-6 and Pfeiffer cells, characterized by persistent PI3K/AKT activation driven by PTEN loss or dysregulated BCR signaling [36, 37]. As illustrated in Fig. 2b, JMC14 suppressed AKT phosphorylation at S473 and T308 in a concentration-dependent manner across all analyzed DLBCL cell lines. In TMD8 and Pfeiffer cells, treatment with JMC14 at 1 µM or 10 µM reduced AKT phosphorylation levels by approximately 40% (Fig. 2c). Interestingly, lower concentrations of JMC14 effectively inhibited the phosphorylation of downstream targets such as p70S6K1, 4E-BP1, and S6, suggesting a cascade amplified effect in signal blockade. In contrast, this amplification was absent in SU-DHL-6 cells, where comparable inhibition on AKT and its downstream targets was observed upon JMC14 treatment. Specifically, treatment with JMC14 at 10 μM reduced AKT phosphorylation at S473 by 44.7%, while phosphorylation of p70S6K1 at T389 and S6 at S240/242 or S235/236 decreased by 39.7%, 55.4% or 50.6%, respectively (Fig. 2c). However, at the same concentration, JMC14 exhibited weaker inhibitory effects on 4E-BP1 phosphorylation at T37/46 in SU-DHL-6 cells, potentially due to alternative upstream regulation of mTOR in this cell line. In summary, these findings demonstrate that JMC14 potently inhibits the proliferation of DLBCL cells via blockade of PI3K-mediated signaling, while differential effects on downstream targets may be influenced by cell-specific signaling contexts.

JMC14 suppresses tumor growth in TMD8 xenografts and DLBCL PDX

Encouraged by the promising activity of JMC14 observed in DLBCL cells, we next evaluated its therapeutic potential in vivo using TMD8 xenografts. Oral administration of JMC14 for 21 days resulted in a dose-dependent inhibition of tumor growth (Fig. 3a). At a dose of 100 mg/kg, JMC14 significantly inhibited the growth of xenografts with a T/C value of 17.6%. In comparison, idelalisib demonstrated a higher T/C value of 60.4%, indicating less efficacy than JMC14 in TMD8 xenografts. Although sotuletinib (100 mg/kg) also attenuated tumor growth (T/C = 27.4%), its effect was less pronounced than that of JMC14 at the same dose. Notably, no significant changes in body weight were observed in any of the tested animals (Fig. 3b).

Fig. 3. JMC14 suppresses tumor growth in TMD8 xenografts and DLBCL PDX.

a, b Efficacy of JMC14 on the growth of TMD8 xenografts. BalB/c SCID mice bearing TMD8 xenografts were treated with vehicle control, sotuletinib (100 mg/kg, once daily), idelalisib (50 mg/kg, twice daily), or JMC14 (10, 30, or 100 mg/kg, once daily) for 21 days (n = 6). Relative tumor volumes are shown as mean ± SEM (a). Body weight of mice was measured twice a week in each group during the treatment (b). c, d Mice bearing TMD8 tumor received a single dose of JMC14 (30 or 100 mg/kg) and were sacrificed at the indicated times. Tumor tissues were collected, and lysates were analyzed by Western blot to detect indicated proteins (c). The concentration of JMC14 in tumor tissues was measured (d). e, f Efficacy of JMC14 on the growth of DLBCL PDX LY-24-0063. PDX-bearing mice were administered with vehicle control, idelalisib (50 mg/kg, twice daily) or JMC14 (10, 30, or 100 mg/kg, once daily) for 21 days (n = 6). Relative tumor volumes are shown as mean ± SEM (e). Body weight of mice was measured twice a week in each group during the treatment (f).

To assess the effect of JMC14 on PI3K signaling in tumor tissues, mice bearing TMD8 xenografts were administered a single oral dose of JMC14 (30 or 100 mg/kg). Tumors were harvested at 1, 2, 4, 8, or 24 h post-treatment. As shown in Fig. 3c, phosphorylated AKT at S473 was reduced significantly 1 h after administration of JMC14 at both 30 and 100 mg/kg. Phosphorylated AKT began to recover 8 h after treatment of JMC14 at 30 mg/kg, while inhibition of AKT phosphorylation persisted up to 8 h and was reversed up to 24 h upon treatment. Pharmacokinetic analysis revealed peak concentrations (C_max) of 3.9 µM or 30.1 µM in tumor tissues after dosing of JMC14 at 30 mg/kg or 100 mg/kg, respectively (Fig. 3d). The concentrations of JMC14 declined in a time-dependent manner, which well correlated with the inhibition on AKT phosphorylation. These findings demonstrate that JMC14 inhibits PI3K pathway in a dose- and time-dependent manner, reflecting its efficacy to inhibit the growth of TMD8 xenografts.

Patient-derived xenografts (PDXs) preserve the histological characteristics and heterogeneity of the original tumors, making them highly relevant models for preclinical studies. To further evaluate the therapeutic potential of JMC14, the LY-24-0063 PDX was established from a patient with DLBCL. As demonstrated in Fig. 3e, JMC14 at 100 mg/kg significantly suppressed the growth of LY-24-0063 PDX, resulting in a T/C value of 35.2%. Lower dose of JMC14 at 10 mg/kg or 30 mg/kg exhibited marginal effect, with T/C values of 94.0% or 82.5%, respectively. Idelalisib at 50 mg/kg administered twice a day achieved a T/C value of 44.0%. Though JMC14 and idelalisib demonstrated comparable efficacy in this model, improved pharmacokinetic profile of JMC14 allowed daily administration. JMC14 treatment showed no significant impact on body weight in all dose levels (Fig. 3f), further confirming its favorable safety profile. These findings underscore the potent and dose-dependent activity of JMC14 against DLBCL.

JMC14 displays potent activity against myeloid leukemia in vitro and in vivo via blocking CSF1R and PI3K signaling

To assess the cellular activity of JMC14 against CSF1R, we employed M-NFS-60 myeloid leukemia cells, known for their strong dependence on CSF-1 for proliferation and survival [38, 39]. JMC14 concentration-dependently inhibited the proliferation of M-NFS-60 cells stimulated by CSF-1 (Fig. 4a). The inhibitory potency of JMC14 on M-NFS-60 cell proliferation was comparable to that of sotuletinib, with IC₅₀ values of 289 nM and 221 nM, respectively. Accordingly, CSF1 triggered the activation of CSF1R and its down-stream AKT and Erk1/2 in M-NFS-60 cells, while this process was abrogated by JMC14 in a dose-dependent manner (Fig. 4b). Similar observation was obtained in the presence of sotuletinib, while idelalisib failed to inhibit the phosphorylation of Erk1/2 irrespective blockade of AKT activation (Fig. 4b). These results demonstrated that JMC14 exerted potent anti-proliferative activity in M-NFS-60 cells by disrupting the CSF1/CSF1R signaling axis and its downstream pathways.

Fig. 4. JMC14 displays potent activity against myeloid leukemia in vitro and in vivo via blocking CSF1R and PI3K signaling.

a Cells were treated with JMC14 or sotuletinib for 72 h, and cell viability was determined by CCK-8 assay. b M-NFS-60 cells were starved in serum-free medium for 2 h and then treated with JMC14, sotuletinib, or idelalisib at the indicated concentration for 30 min prior to stimulation with CSF1 (100 ng/mL). Cell lysates were collected, and Western blot was performed with indicated antibodies. c, d Efficacy of JMC14 on the growth of M-NFS-60 xenografts. SCID mice bearing M-NFS-60 xenografts were treated with vehicle control, sotuletinib (200 mg/kg), or JMC14 (100 mg/kg), once daily for 13 days (n = 7). Relative tumor volumes are shown as the mean ± SEM (c). Body weight of mice was measured twice a week during the treatment (d). e Mice bearing M-NFS-60 tumors received a single dose of JMC14 at 100 mg/kg and were sacrificed at the indicated times. Tumor tissues were collected, and lysates were analyzed by Western blot to detect indicated proteins.

We next evaluated the therapeutic efficacy of JMC14 in mice bearing M-NFS-60 xenografts. Mice were administered daily with either JMC14 or sotuletinib. As shown in Fig. 4c, treatment with JMC14 at 100 mg/kg significantly suppressed tumor growth, yielding a T/C value of 25.4%, which was slightly higher than the 18.8% achieved with sotuletinib at 200 mg/kg. No significant changes in body weight were observed between treatment and control groups, indicating that JMC14 and sotuletinib were well-tolerated (Fig. 4d). Furthermore, the anti-tumor effect correlated with the suppression of CSF1R and PI3K signaling pathways, as evidenced by a pronounced decrease in the phosphorylation of CSF1R, Erk1/2, AKT, S6K and S6 in tumor tissues collected 2 h after a single dose (Fig. 4e). Notably, the phosphorylation of S6K and S6 was partially restored 8 h post-treatment (Fig. 4e), suggesting that sustained inhibition of CSF1R/PI3K signaling may be crucial for the efficacy of JMC14.

JMC14 suppresses the proliferation of triple-negative breast cancer cells

PI3Kδ and CSF1R are significantly associated with tumor progression and immune evasion in breast cancer [40, 41]. To explore the potential therapeutic response by dual targeting PI3Kδ and CSF1R in triple-negative breast cancer (TNBC), three lines of murine TNBC 4T1, PY8119 and EMT6 were employed. As shown in Fig. 5a, JMC14 demonstrated moderate antiproliferative activity against all three cell lines, with IC₅₀ values of 7.9 µM, 5.5 µM, or 6.5 µM, respectively. In contrast, both sotuletinib and idelalisib displayed minimal effects, with IC₅₀ values exceeding 10.0 µM, suggesting that simultaneous targeting of CSF1R and PI3Kδ may enhance the inhibition of cell proliferation. We next examined the effect of JMC14 on CSF1R and PI3K signaling in these cells. As shown in Fig. 5b, JMC14 inhibited the phosphorylation of AKT, S6K, and S6, key signaling components downstream of PI3K, in a concentration-dependent manner. Notably, lower concentration of JMC14 was required to achieve similar levels to inhibit the phosphorylation of S6K and S6 compared to that of AKT, indicating a signal amplification effect within the pathway (Fig. 5c). JMC14 exerted lower potency in inhibiting Erk1/2 phosphorylation, achieving only 20% inhibition at a concentration of 10 µM (Fig. 5c). A similar effect was observed with sotuletinib, although sotuletinib was much less active than JMC14 to block PI3K/Akt signaling cascade. In contrast, idelalisib at 10 µM effectively downregulated the phosphorylation of AKT, S6K, and S6, yet had minimal impact on Erk1/2 phosphorylation. These findings suggest that the simultaneous inhibition of PI3Kδ and CSF1R may enable JMC14 to attenuate AKT- and Erk1/2-mediated signaling, thereby contributing to its superior antiproliferative activity compared to the mono-targeting of either PI3Kδ or CSF1R.

Fig. 5. JMC14 suppresses the proliferation in triple-negative breast cancer cells.

a 4T1, PY8119, or EMT6 cells were treated with JMC14, sotuletinib, or idelalisib for 72 h, and cell proliferation was assessed by SRB assay. The data were from two independent experiments. b, c 4T1, PY8119, or EMT6 cells were exposed to JMC14, sotuletinib, or idelalisib for 1 h at the indicated concentrations. Western blot was performed with the indicated antibodies. Band intensity was quantified using ImageJ (c). Data shown are representative (b) or mean ± SEM from two independent experiments (c).

JMC14 attenuates the growth of PY8119 allografts and reprograms the TME

To investigate the antitumor efficacy of JMC14 in an immune-competent model, PY8119 cells were orthotopically inoculated into C57BL/6 mice. JMC14 inhibited the tumor growth in a dose-dependent manner (Fig. 6a). Notably, JMC14 at 100 mg/kg exhibited pronounced tumor growth inhibition with a T/C value of 26.9%, displaying improved efficacy compared to that of 68.1% for idelalisib (50 mg/kg, BID) and 46.8% for sotuletinib (100 mg/kg), respectively (Fig. 6a). Moreover, no significant changes in body weight were observed in all treatment groups (Fig. 6b).

Fig. 6. JMC14 attenuates the growth of PY8119 allografts and reprograms the TME.

a, b Efficacy of JMC14 on the growth of PY8119 allografts. C57 mice bearing PY8119 allografts were administered with vehicle control, idelalisib (50 mg/kg, twice daily), sotuletinib (100 mg/kg, once daily), or JMC14 (25, 50, 100 mg/kg, once daily) for 18 days (n = 9). Relative tumor volumes are shown as mean ± SEM (a). Body weight of mice was measured twice a week. c Mice bearing PY8119 tumors received a single dose of JMC14 at 100 mg/kg and were sacrificed at the indicated times. Tumor tissues were collected, and lysates were analyzed by Western blot to detect indicated proteins. d, e PY8119 tumor sections were collected after JMC14 treatment for 18 days and immunostained with the indicated antibodies. Representative images are shown (d). Scale bar: 50 µm. The positive staining area (%) was quantified using ImageJ (e).

To further investigate the effect of JMC14 on CSF1R and PI3K signaling in tumor tissue, a single dose of JMC14 (100 mg/kg) was administered to mice bearing PY8119 allografts. Tumor tissues were collected at 2, 4, and 8 h after treatment. As shown in Fig. 6c, JMC14 effectively down-regulated the phosphorylation of CSF1R, AKT, S6K, and S6 within 2 h after administration, while a longer time was needed to suppress the Erk1/2 phosphorylation. Phosphorylation of AKT and S6 partially recovered 8 h post-treatment. To elucidate the effect of JMC14 on TME, immunohistochemical staining was performed with tumor tissues from mice treated with JMC14 for 18 days. As shown in Fig. 6d, the staining CD45 markedly increased after JMC14 treatment, indicating enhanced immune cell infiltration. A significant reduction in CD4 staining, accompanied by an increase in CD8 staining, was observed (Fig. 6e), suggesting a shift toward a cytotoxic immune response. Given the critical role of CSF1R signaling in the survival, proliferation, and differentiation of macrophages, we next evaluated macrophages infiltration. A marked decrease in the F4/80 staining was observed (Fig. 6d, e), indicating reduced infiltration of total macrophages following JMC14 treatment. Particularly, the staining of CD206 representing immunosuppressive M2 macrophages significantly decreased (Fig. 6d, e), suggesting the effect of JMC14 to abrogate immune evasion. Moreover, JMC14 treatment significantly reduced CD31 staining, indicating a pronounced inhibitory effect on tumor vascularization (Fig. 6d, e). Collectively, these results demonstrate that JMC14 exerts potent activity against TNBC via direct inhibition of tumor cell proliferation and reprogramming of TME.

Discussion

PI3Kδ and CSF1R are not only critical for the proliferation of certain types of malignancy but also key mediators of immune suppressive TME. In this study, we characterized JMC14, a novel dual PI3Kδ/CSF1R inhibitor featuring an innovative structure, and demonstrated its potent efficacy in malignancies with activated PI3Kδ and/or CSF1R including B-cell lymphoma, myeloid leukemia, and TNBC. JMC14 exhibited remarkable selectivity, displaying a 34-fold preference for PI3Kδ over other isoforms. JMC14 exerted its antitumor effects through multiple mechanisms, including dual inhibition of PI3Kδ and CSF1R signaling as well as modulation of the TME.

Our findings revealed that JMC14 achieved superior efficacy compared with idelalisib, which may be attributed to its distinct structure enabling dual targeting PI3Kδ and CSF1R. The development of PI3Kδ inhibitors has progressed significantly since the approval of idelalisib for the treatment of relapsed chronic lymphocytic leukemia (CLL), small lymphocytic lymphoma (SLL), and follicular lymphoma (FL). Despite its initial success, the clinical utility of idelalisib has been limited by dose-limiting toxicities and the emergence of resistance, highlighting the urgent need for next-generation inhibitors with improved selectivity, efficacy, and safety. Similarly, early CSF1R inhibitors frequently displayed multi-target activity, leading to increased toxicity [42]. Our kinase profiling data indicated that JMC14 selectively inhibited PI3Kδ and CSF1R with negligible activity against other kinases, including other family members of RTK III. These observations underscore that JMC14 may act as a distinct molecular probe to dissect the interaction of PI3Kδ and CSF1R. While further structural optimization is warranted, our results also establish JMC14 as a promising lead compound for the development of next-generation PI3Kδ or CSF1R inhibitors with improved efficacy and safety.

We found that JMC14 possessed potent efficacy in malignancies with activated PI3Kδ and/or CSF1R including B-cell lymphoma, myeloid leukemia, and TNBC, which was consistent with its selective targeting PI3Kδ and CSF1R. JMC14 demonstrated potent anti-proliferative activity across a panel of 28 lines of hematological cancer cells, with particularly notable efficacy against DLBCL cells, suggesting that JMC14 may be especially effective for treating DLBCL, a subtype of lymphoma characterized by aberrant PI3Kδ signaling and poor prognosis in many cases. Consistent with our previous findings, the efficacy of JMC14 was strongly associated with PI3Kα expression in DLBCL cell lines, as evidenced by a robust positive correlation between relative PI3Kα levels and JMC14 sensitivity. These data suggest a compensatory role for PI3Kα in maintaining PI3K/AKT pathway activity upon PI3Kδ inhibition, emphasizing the notion of functional crosstalk among PI3K isoforms. Although PI3Kδ is predominantly expressed in immune cells, and high CSF1R expression has been linked to poor prognosis in certain lymphoma subtypes [43], our study did not detect a significant correlation between JMC14 efficacy and either PI3Kδ or CSF1R expression. This discrepancy may be attributable to our limited sample size or an alternative mechanism wherein JMC14’s antitumor activity extends beyond direct inhibition of these pathways, possibly involving broader TME modulation. It should be noted that JMC14 demonstrated superior antitumor activity against TNBC compared to the single-targeting PI3Kδ by idelalisib or CSF1R by sotuletinib. TAMs and Foxp3⁺ regulatory T cells (Tregs) are key components of the immunosuppressive TME. Elevated CSF1R expression in TAMs promotes immune suppression and tumor progression, whereas PI3Kδ regulates the function of CD4⁺ T cells by modulating their differentiation, activation, and response to TME [44], underscoring the potential combination regimen to target the TME [45]. Indeed, JMC14 not only inhibited cell proliferation through blockade of the CSF1R/PI3K signaling but also reshaped the immune microenvironment by reducing infiltration of M2 macrophages and CD4⁺ T cells while increasing CD8⁺ T cell population. These results suggest that dual-targeted inhibition by JMC14 may effectively modulate both tumor-intrinsic pathways and the TME, thereby offering a distinct advantage over single-targeted inhibitors. Nonetheless, additional validation with diverse tumor models is essential to identify tumor types likely responsive to dual inhibition of PI3Kδ and CSF1R.

In summary, JMC14 represents a first-in-class dual PI3Kδ/CSF1R inhibitor to our knowledge, which possesses robust antitumor efficacy in both hematologic malignancies and solid tumors. JMC14 may serve as a valuable tool to dissect the complex interplay between the PI3Kδ and CSF1R pathway. These findings underscore JMC14 not only as a promising therapeutic candidate but also as a unique lead for the rational design of next-generation PI3Kδ/CSF1R inhibitors with improved efficacy and safety profiles.

Author contributions

XZ, JD, CHY, and LHM conceptualized and designed the study. XZ, YW, PS, and HYL conducted the experiments and analyzed the data. CHY and CT were responsible for compound design and synthesis. LX, PS, and XZ carried out the in vivo experiments. LJT performed CSF1R kinase assay. XZ and LHM prepared the draft of the manuscript. JD, CHY, and LHM supervised the study.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41401-025-01575-x.