Abstract

Background

   A tyrosine kinase inhibitor (TKI) such as Imatinib (IM) is the
   preferred treatment for Chronic Myeloid Leukemia (CML). However, the
   emergence of IM resistance presents a significant challenge to disease
   management. A characteristic of cancer cells, including IM-resistant
   CMLs, are characterized by heightened uptake of glucose and aberrant
   glycolysis in the cytosol, which is known as the Warburg effect. In
   addition to its potential to modulate the Warburg effect, Chiglitazar
   (Chi), a compound that regulates glucose metabolism, has also been
   investigated for its implication in cancer treatment. This suggests
   that combining Chi with IM may be a therapeutic strategy for overcoming
   IM resistance in CML.

Methods

   Sensitive and IM-resistance CML cells were treated with Chi in vitro,
   followed by detecting of extracellular acidification rate (ECAR) using
   a Seahorse XF Analyzer. CML cell proliferation, cell cycle
   distribution, and apoptosis were tested by CCK-8 assay and flow
   cytometry. RNA sequencing was utilized to investigate potential
   transcriptional changes induced by Chi usage. In vivo studies were
   conducted on immunodeficient mice implanted with CML cells and given
   Chi and/or IM later. Tumor growth was monitored, as well as tumor
   burden and survival rates between groups.

Results

   Our metabonomic, transcriptomic, and molecular biology studies
   demonstrated that Chi, in part, diminished the Warburg effect by
   reducing glucose and lactate production in imatinib-resistant CML cells
   through the PPARγ/mTOR/PKM2 pathway. This modulation of glucose
   metabolism resulted in reduced cell proliferation and enhanced
   sensitivity to IM in imatinib-resistant CML cells in vitro. Rescue
   assay by introducing shPPARγ or mTOR activator verified the underlying
   regulatory pathway. Also, the combination of Chi and IM synergistically
   increased the sensitivity of IM in vivo and prolonged the survival of
   imatinib-resistance CML transplanted mice.

Conclusions

   Our results demonstrated the potential of Chi to overcome IM resistance
   in vitro and in vivo. By inhibiting the Warburg effect through the
   PPARγ/mTOR/PKM2 pathway, Chi resensitizes CML cells towards imatinib
   treatment. Combining IM with Chi is an alternative therapeutic option
   for CML management, especially for IM-resistant CML patients.

Supplementary Information

   The online version contains supplementary material available at
   10.1186/s40164-024-00589-1.

   Keywords: Chronic myeloid leukemia, Glucose metabolism, Imatinib
   resistance, Chiglitazar, PPARγ

Background

   CML is an aggressive myeloproliferative disorder caused by the
   translocation of chromosomes 9 and 22 within hematopoietic stem cells
   [[47]1]. This event elicits the expression of the aberrantly activated
   BCR::ABL protein, which governs the dysfunction of key intracellular
   signaling pathways/proteins essential for the survival and
   proliferation of hematopoietic progenitor cells [[48]2, [49]3]. Despite
   the transformative impact of tyrosine kinase inhibitors (TKIs) such as
   imatinib (IM), which target BCR::ABL fusion oncoproteins harbored in a
   significant proportion of patients (25% initially and up to 90% in
   TKI-resistant cases) [[50]4–[51]6]. Numerous investigations have
   revealed cellular processes linked to drug resistance, including the
   activation of key signaling pathways, cell cycle disruptions, abnormal
   epigenetic events, aberrant stem cell niche microenvironments, and
   metabolic shifts [[52]7–[53]9]. The clinical use of TKIs like imatinib,
   nilotinib, and dasatinib has substantially extended the overall
   survival of CML patients [[54]10–[55]13]. However, long time TKI
   therapy is ultimately associated with drug toxicity, acquired
   resistance, and significant economic burden [[56]14, [57]15].
   Furthermore, about half of the patients experience rapid relapse upon
   treatment cessation [[58]16]. Thus, it is evident that new therapeutic
   strategies based on the biological characteristics of TKI-resistant CML
   cells are urgently needed.

   As evidenced by stable isotope-assisted metabolomics, primitive CML
   cells rely heavily on oxidative metabolism [[59]8, [60]17]. Metabolic
   reprogramming offers a promising therapeutic alternative for CML
   treatment. Recent studies have demonstrated that the peroxisome
   proliferator-activated receptor (PPAR), specifically its α, β/δ, and γ
   subtypes, exerts a profound influence on the Warburg effect and
   metabolic state of tumor cells [[61]18, [62]19]. This influence is
   manifested through various mechanisms, including the direct regulation
   of pivotal metabolic enzymes such as phosphofructokinase (PFK) and
   citric acid cycle enzymes. PPAR also elicits its effects through
   inhibiting mTOR signaling pathway and suppressing HIF-1α and Wnt
   expression activity [[63]20–[64]22]. In addition, PPAR ligands also
   possess anticancer properties against various cancers such as colon,
   pancreatic, breast, prostate, and leukemia [[65]23, [66]24]. It has
   recently been discovered that thiazolidinediones (TZDs), a class of
   antidiabetic drugs, are undergoing clinical trials for the treatment of
   hematopoietic malignancies [[67]24, [68]25]. It is important to note
   that TZDs’ side effects have limited their clinical applications in
   cancer treatment [[69]26]. Our previous study demonstrated that Chi
   exerted its antileukemic effects by affecting stem cell survival
   through the PPARα/HIF1α/PGK1 pathway [[70]27]. As of yet, little is
   known about Chi’s metabolic regulation and potential antileukemic
   effects on IM-resistant CML cells.

   This study aimed to explore the potential of glucose metabolism
   modulation as a strategy to resensitize CML cells to IM therapy by
   employing Chi to influence glycolysis. Our findings demonstrated the
   synergistic effect of Chi and IM in inhibiting the CML cell
   proliferation. We showed that Chi treatments significantly reduce
   extracellular acidification rates (ECAR) and cellular ATP production in
   CML cells. As a result of these energy metabolic modulations, CML cells
   were induced by combined therapy to undergo apoptosis. Our study
   suggests the potential of combing Chi and IM to overcome IM resistance
   in CML cells.

Methods

Reagents and cell culture

   IM was purchased from MedChemExpress LLC (Shanghai, China) and
   dissolved in dimethyl sulfoxide (DMSO) (Invitrogen Corp., Waltham, MA,
   USA) to yield a 10 mM stock solution. Chi, obtained from Chipscreen
   Biosciences Co., Ltd. (Shenzhen, Guangdong, China), was dissolved in
   DMSO to obtain a 100 mM stock solution. MHY1458 was purchased from
   Selleck Inc. (Shanghai, China) dissolved in DMSO to obtain a 10 mM
   stock solution. Subsequently, it was diluted to specified
   concentrations in culture media for other experiments. The CML cell
   lines K562 and Ba/F3 (BCR::ABL-T315I) were kindly provided by Professor
   Liu XL and Professor Jiang Q (Peking University People’s Hospital) in
   Beijing, China. K562 and Ba/F3 cells were cultured in RPMI-1640
   (Hyclone, Thermo Fisher Scientific, Logan, USA) supplemented with 10%
   fetal bovine serum (Gibco, Life Technologies, NY, USA), 100 units/mL
   penicillin, and 100 μg/mL streptomycin at 37 °C in a 5% CO[2]
   incubator. K562 IM-resistant cell line (K562R) was cultured in imatinib
   at gradually increased concentrations and cultured in a supplemented
   medium containing 1 μM imatinib mesylate to acquire drug resistance. IM
   and simvastatin (SIM) were obtained from MedChemExpress (Shanghai,
   China) and prepared accordingly.

Cell proliferation assay

   The cell-killing effects of IM and Chi were evaluated using the Cell
   Counting Kit-8 (Shandong Sparkjade Biotechnology Co., Ltd, China).
   Cells were seeded in 96-well plates at a density of 4 × 10^4 cells per
   well and were treated with designated doses of IM or Chi, either alone
   or in combination, for 48 h. Following the addition of the CCK-8
   reagent, absorbance was measured at 450 nm using a microplate reader
   (ELx800; BioTek Instruments Inc, Winooski, VT). All experiments were
   performed in triplicate and repeated three times. The half-maximal
   inhibitory concentration (IC50) was calculated using SPSS 20.0 (SPSS
   Inc, Chicago, IL, USA). The combination index (CI) was computed using
   the Chou-Talalay method with Calcusyn software, which was used to
   quantify therapeutic synergy based on cell viability data (CI < 1
   indicates synergy).

Cell apoptosis and cell cycle assay

   For apoptosis evaluation, cells were subjected to treatment with either
   Chi or IM, independently or in combination, for 48 h. Subsequently,
   they were stained following the manufacturer’s guidelines using
   Annexin-V-FITC/PI (eBioscience, San Diego, CA, USA). The proportion of
   Annexin-V positive (apoptotic) cells was ascertained through flow
   cytometry (FACSCalibur, BD Biosciences) and data were processed using
   FlowJo software. In parallel, cell cycle analysis was performed on
   similarly treated cells, which were fixed in ethanol, stained with
   PI/RNase, and examined via flow cytometry. The resulting cell cycle
   distribution was scrutinized using flow cytometry and FlowJo software.

Gene knockdown and expression assay

   We designed the shPPARγ sequence (GGAGAACAATCAGATTGAAGC) and cloned
   into lentiviral vector ordered from Genechem (Shanghai, China). The
   shPPARγ plsmid was transfected into K562R cells using lipofectamine
   2000 transfectiong kit (ivitrogen) 24 h after Chi treatment following
   the protocol. For apoptosis assay under PPARγ knockdown condition, we
   transfected the plasmid and cultured the K562R cell with the drug
   containing medium for 48 h at the same day and then processed the cells
   to FACS assay.

RNA sequencing and data analysis

   Total RNA samples from different conditions were subjected to RNA
   sequencing (RNA-seq) on Illumina Hiseq 2500 platform. Released
   sequencing data were trimmed and quality filtered using cutadapt
   ([71]http://cutadapt.readthedocs.io, version 1.2.0). Alignment was
   performed using Hisat2 with human (hg38) genome as a reference
   [[72]28]. Quantified transcripts were generated by Feature Count
   program [[73]29]. They then subjected to differential expression
   analysis using Deseq2 program [[74]30]. The Venn diagram was generated
   using the ggVenn program. GO analysis was done using Database for
   Annotation, Visualization and Integrated Discovery (DAVID) [[75]31,
   [76]32]. GSEA analysis was carried out to verify the correlation
   between differential expressed genes and biological states.

Glucose and lactic acid measurement

   For glucose uptake analysis, 2 × 10^6 K562 or K562R cells were seeded
   into a 24-well plate and treated with imatinib and/or siglitin sodium
   for 24 h. A glucose uptake assay kit (A154-1-1, Nanjing Jiancheng
   Bioengineering Institute, China) was used following the manufacturer’s
   instructions. The glucose levels in the culture medium were compared
   before and after the 24-h treatment, and absorbance at 488 nm was
   measured and normalized to the cell number. This comparison allowed for
   the assessment of glucose uptake among the different groups. For lactic
   acid analysis, 2 × 10^6 K562 or K562R cells were seeded into a 24-well
   plate and incubated at 37 °C for 24 h. Following treatment with
   imatinib and/or siglitin sodium for 24 h, the cells were lysed to
   collect intracellular lactic acid. A lactate assay kit (BC2235,
   Solarbio, China) was used per the manufacturer’s instructions.
   Absorbance at 450 nm was recorded.

Metabolic profiling of CML cells

   Glycolysis levels in K562 or K562R cells were gauged using lactate
   production (BC2235, Solarbio, China) and glucose uptake assays
   (A154-1-1, Nanjing Jiancheng Bioengineering Institute, China),
   following the respective assay kits’ instructions. Concurrently,
   real-time analysis of cellular metabolism was conducted on XFe96
   extracellular flux analyzer (Seahorse Bioscience, North Billerica, MA,
   USA). Cell plates were pre-coated with poly-lysine (Pernoside, Wuhan,
   China) to facilitate wall attachment of K562 and K562R cells. 1 h
   before the assay, changed the culture medium to pre-warmed hippocampal
   basal medium supplemented with 1 mM pyruvate, 2 mM glutamine, and 10 mM
   glucose. Cells then were incubated in a CO[2]-free incubator at 37 °C
   for 1 h. Glucose (10 mM), oligomycin (2 μM), and 2-Deoxy-d-glucose
   (2-DG, 100 mM) were injected into ports A, B, and C, respectively, for
   ECAR analysis. Results were normalized based on the cell numbers.

Western blot analysis

   Whole cell lysates (30 μg/lane) of each sample went through 10% or
   12.5% SDS-PAGE gel and subsequently transferred to a PVDF membrane
   (Millipore Corp., Burlington, MA, USA). After blocking non-specific
   binding, PVDF membranes were incubated at 4 °C overnight with primary
   antibodies (PARP, rabbit monoclonal, 1:1000; Caspase3, rabbit
   monoclonal, 1:1000; GAPDH, rabbit monoclonal, 1:1000; STAT5, rabbit
   monoclonal, 1:1000; P-STAT5, rabbit monoclonal, 1:1000; PPARγ, rabbit
   monoclonal, 1:1000; MTOR, rabbit monoclonal, 1:1000; PKM2, mouse
   monoclonal, 1:1000; Cell Signaling Technology, Inc., Danvers, MA, USA).
   PVDF membranes were then incubated with horseradish
   peroxidase-conjugated secondary antibodies for 1 h at room temperature.
   Immunoreactive bands were detected using ECL buffer (Aase Bio,
   Shanghai, China).

Chronic myeloid leukemia cell derived xenograft model

   NOD-Prkdc−/−IL2rg−/− (NOD/SCID) mice (6 weeks old) obtained from the
   Animal Care Institute of Xiamen University were raised in a
   pathogen-free environment. To establish a Cell Derived Xenograft (CDX)
   model, mice received 1.5 Gy sublethal irradiation followed by tail vein
   injection of 1 × 10^6 K562R cells and were randomly divided into 4
   groups (n = 14). On day 7, peripheral venous blood samples were taken,
   and all groups of mice were euthanized for further analysis of human
   CD45 expression in the spleen (Spleen tissue human CD45 > 1%, 7 days
   post-transplantation). 0 mice were administered oral gavage carrier
   (methylcellulose/tween-80), imatinib (50 mg/kg/day), chiglitazar
   (15 mg/kg/day), or their combination for 12 days to determine drug
   efficacy. The percentage of human CD45 + CML cell infiltration in bone
   marrow and spleen leukocytes was measured to determine the drug
   response in all experimental groups of mice. For Ki67
   immunohistochemistry (IHC) and hematoxylin and eosin (H&E) staining,
   tissue sections were deparaffinized, rehydrated, and antigen-retrieved.
   Ki67 primary antibody (TA800648, ZSGB-Bio) was applied overnight at
   4 °C. After washing, a secondary antibody and 3,3′-Diaminobenzidine
   (DAB) chromogen were used for signal detection. Counterstaining was
   done with hematoxylin. Animals were used in accordance with a protocol
   approved by the Laboratory Animal Center of Xiamen University.

Statistical analysis

   Data were presented as mean ± standard deviation (SD) from three
   independent experiments and analyzed using GraphPad Prism 6 software
   (GraphPad Software, San Diego, CA, USA). Statistical comparisons
   between two groups were conducted using Student’s t-tests (unpaired,
   two-tailed) or one-way analysis of variance (ANOVA) followed by Tukey’s
   post-hoc tests. P < 0.05 was considered statistically significant. All
   data generated and analyzed during this study are available upon
   request from the corresponding author.

Results

IM-resistant cells showed a increased glucose uptake and lactate production

   We first sought to evaluate the cellular responses of IM-resistant CML
   to IM treatment. Therefore, human and mice IM-resistant CML cell lines
   K562R (drug induced) and Baf3-T315I (T315I mutation) were exposed to
   48 h IM treatment in this study [[77]33]. Cell viability assays were
   performed to assess the proliferative abilities of cell lines when
   treated with IM. The IC50 values of IM for K562, K562R, and Baf3 cells
   were 0.48 μM, 8.01 μM, and 24.37 μM (Fig. [78]1a). Then we assessed
   cell line apoptosis using Annexin V-FITC/PI staining. The FACS analysis
   revealed significantly lower apoptosis percentage in K562R
   (Fig. [79]1b, Additional file [80]1: Fig. S1a–c). We chose 5 μM as IM’s
   concentration to conduct further tests which is between the IC50 values
   of IM for K562 and K562R (Fig. [81]1a). To explore the underlying
   transcriptional mechanism of IM resistance, RNA sequencing was
   performed on K562 and K562R cell lines. Differential expression
   analysis revealed 967 downregulated and 636 upregulated genes
   (Fig. [82]1c, Additional file [83]2: Table S1). Further, KEGG and GO
   enrichment analyses of differential expressed genes (DEGs) highlighted
   significant alterations in pathways associated with glycolysis and
   oxidative phosphorylation (Fig. [84]1d, Additional file [85]3: Table
   S2). These findings reinforce previous research suggesting that
   CML-resistant cells rely on enhanced oxidative metabolism for survival,
   albeit without a fully elucidated mechanism [[86]34]. So, we assessed
   the glucose uptake and lactate production in K562 and K562R cells by
   measuring the glucose and lactate levels in the culture medium.
   Interestingly, K562R cells displayed enhanced glucose uptake and
   lactate production, indicating a metabolic shift towards glycolysis, a
   phenomenon known as the Warburg effect (Fig. [87]1e). To test the
   Warburg effect in IM-resistant cells, we detected the Oxygen
   Consumption Rate (OCR) and Extracellular Acidification Rate (ECAR)
   level. IM-resistant cells showed higher ECAR levels and lower OCR
   levels than IM-sensitive cells (Fig. [88]1f, Additional file [89]1:
   Fig. S1d). Our results confirmed that IM-resistant cells’ metabolic
   program depends more on glycolysis.

Fig. 1.

   [90]Fig. 1
   [91]Open in a new tab

   IM-resistant cells showed a increased glucose uptake and lactate
   production. a Cell viability percentage of K562, K562R and Baf3-T315I
   cells after 48 h of imatinib treatment. b Flow cytometry apoptosis
   assay of K562 and K562R cells following imatinib treatment. c Heatmap
   showing differentially expressed genes in K562 and K562R cells
   (q < 0.01 and |log fold change (FC)| > 2). d KEGG and GO enrichment
   analysis of DEGs (q < 0.01 and |log fold change (FC)| > 2). e Glucose
   uptake and lactate production in K562 and K562R cells after IM
   treatment for 24 h. f For the glycolysis analysis, cells were firstly
   seeded in plate for 6 h, and then cells were sequentially treated with
   glucose (10 mM), oligomycin (2 μM), 2-DG (100 mM). Data are presented
   as mean ± SD. NS < 0.1234, *P < 0.03, **P < 0.0021, ***P < 0.0002,
   ****P < 0.0001

Chiglitazar diminished the glucose uptake in IM-resistant cells

   Chiglitazar, a common antidiabetic drug, has previously shown its
   therapeutic effect in the treatment of acute myeloid leukemia [[92]27].
   Chi reducing glucose metabolism in leukemic stem-like cells inspired us
   to investigate its potential application in the treatment of CML. We
   first used the CCK-8 assay to evaluate the cell-killing ability of Chi
   in IM-sensitive cell lines (K562) and resistant cell lines (K562R and
   Baf3-T315I). Chi exhibited dose-dependent cytotoxicity in K562R cells,
   and there was no significant difference between sensitive and resistant
   cell lines with IC50 values of 21.34 μM (K562), 23.29 μM (K562R), and
   25.33 μM (Baf3-T315I), respectively (Fig. [93]2a). Next, we focus on
   evaluating the apoptosis and cell cycle progression of K562R cell with
   Chi treatment. Chi did not significantly induce apoptosis in K562R
   cells even after 48 h of treatment, with the apoptosis rate still below
   30% (Fig. [94]2b, c). Cell cycle distribution assay was conducted after
   treated with varying concentrations of Chi (0, 10, 20 μM) for 24 h and
   found that higher dose Chi treatment (20 μM) but not lower dose (10 μM)
   caused cell cycle arrest at the S phase which is consistant with our
   apoptotic assay (Fig. [95]2b, c). In addition, we investigated the
   influence of Chi on glucose uptake and lactate production in K562 and
   K562R cells. The results showed a significant decrease in glucose
   uptake and lactate production in each cell type after Chi treatment
   (Fig. [96]2d). We confirmed that IM-resistant cells had a higher level
   of glucose uptake and lactate production (Fig. [97]1e). Further, Chi
   significantly reduced resistant cells’ glucose uptake and lactate
   production to a level comparable to that of sensitive cells, indicating
   that Chi significantly inhibited the Warburg effect in resistant cells
   (Fig. [98]2d). Finally, we evaluated whether Chi would affect the
   expression of apoptosis-related proteins PARP and Caspase3, and the
   results showed a limited effect of Chi on their expression, which was
   consistent with our flow cytometry results (Fig. [99]2b, e). We also
   evaluated the impact of Chi on the expression of its target proteins:
   PPARα/γ/δ, leukemia stemness-maintaining STAT5, and cell cycle-related
   protein CyclinD1. The results showed that Chi could significantly
   inhibit the levels of P-STAT5 and reverse the downregulation of PPAR
   family proteins and the upregulation of CyclinD1 in resistant cell
   lines (Fig. [100]2f). Taken together, our results suggested that Chi
   significantly reduced the Warburg effect in IM-resistant CML cells.

Fig. 2.

   [101]Fig. 2
   [102]Open in a new tab

   Chiglitazar diminishes the Warburg effect in imatinib-resistant K562R
   cells. a Cell viability percentage of K562R cells after 48 h of
   Chiglitazar treatment. b Induction of apoptosis by 48 h of treatment
   with Chiglitazar, and cell cycle distribution upon 24 h Chiglitazar
   treatment in K562R cells. c Typical representative diagrams of
   apoptosis and cell cycle. d Glucose uptake and lactate production in
   K562R cells treated with Chiglitazar for 48 h. e Western blotting
   analysis of PARP and Caspase-3 before and after Chi treatment for 48 h.
   f Western blotting analysis of PPAR protein family, STAT5, and CyclinD1
   expression levels after Chi (20 μM) treatment for 48 h. GAPDH was used
   as a loading control. Data are presented as mean ± SD. *NS < 0.1234,
   *P < 0.03, **P < 0.0021, ***P < 0.0002, ****P < 0.0001

Chiglitazar resensitized imatinib-resistant CML cells

   To evaluate the therapeutic potential of Chi in CML, we conducted a
   CCK8 assay to evaluate the cell viability with different concentrations
   of Chi and IM. As previously shown, the cytotoxic effects of either Chi
   or IM alone on IM-resistant cells were limited (Figs. [103]1, [104]2).
   The combination of different concentrations of Chi and IM at a fixed
   ratio (4:1) significantly decreased cell viability in both sensitive
   and IM-resistant cell lines (Fig. [105]3a). The combination index (CI)
   in both sensitive and IM-resistant cell lines was less than 0.85
   (Fig. [106]3b). The sensitive cell line already displayed significant
   apoptosis upon IM treatment, which was significantly enhanced by the
   addition of Chi (Fig. [107]3c). Higher concentrations of IM had limited
   apoptosis effects in drug-resistant and T315I mutation cell lines, but
   were greatly enhanced by co-treatment with Chi (Fig. [108]3c, d). As
   shown in Fig. [109]3e, the combination of Chi and IM not only affected
   cell apoptosis but also the cell cycle, with cell cycle arrest observed
   in S phase (Fig. [110]3e). Glucose, lactate, and pyruvate generation
   were significantly decreased in cells treated with combined Chi and IM
   (Fig. [111]4a, Additional file [112]1: Fig. S1e). Western blot revealed
   upregulation in the expression of metabolism-associated gene PPARγ and
   downregulation of the expression of P-STAT5 (Fig. [113]4b). Thus,
   combining Chi and IM significantly induced the apoptosis of
   IM-resistant cell lines.

Fig. 3.

   [114]Fig. 3
   [115]Open in a new tab

   Chiglitazar resensitized Imatinib-resistant CML cells. a Cell viability
   percentage of Chi-only (20 μM), IM-only (5 μM), and their combination
   for 48 h in K562, K562R, and Baf3-T315I cells. b Combination index (CI)
   of Chi and IM in the indicated two cell line cells was analyzed by
   CalcuSyn software. CI > 1 indicates antagonist effect; CI = 1 indicates
   additive effect; CI < 1 indicates synergistic effect. c Apoptosis rate
   in three indicated cell lines treated with Chi-only (20 μM), IM-only
   (5 μM), and their combination were analyzed using flow cytometry. d The
   representative apoptosis diagrams after 48 h of combined treatment with
   Chi (20 μM) and IM (5 μM). e The representative cell cycle distribution
   diagrams after 24 h of combined treatment with Chi (20 μM) and IM
   (5 μM). Data are presented as mean ± SD. NS < 0.1234, *P < 0.03,
   **P < 0.0021, ***P < 0.0002, ****P < 0.0001

Fig. 4.

   [116]Fig. 4
   [117]Open in a new tab

   Chiglitazar reprogrom the metabolism of Imatinib-resistant CML cells. a
   Glucose uptake, lactate production, and pyruvate production in K562 and
   K562R cells after Chi-only (20 μM), IM-only (5 μM), and their
   combination for 24 h. b Western blotting analysis of
   full-length/cleaved PARP, Caspase-3, STAT5, PPAR protein family, and
   CyclinD1 expression levels after Chi-only (20 μM), IM-only (5 μM), and
   their combination treatment for 48 h in K562 and K562R cells. GAPDH was
   used as a loading control. Data are presented as mean ± SD.
   NS < 0.1234, *P < 0.03, **P < 0.0021, ***P < 0.0002, ****P < 0.0001

Chiglitazar resensitized K562R cells through PPARγ/mTOR/PKM2 pathway

   The metabolic reprogramming of cancer cells is characterized by the
   Warburg effect, which is defined by increased glucose uptake and
   lactate production. Studies have indicated that prolonged IM usage
   correlates with increased glycolysis, potentially linked to IM
   resistance [[118]35, [119]36]. Elevated glucose uptake and lactate
   production in the resistant cell line suggested an aberrant metabolic
   profile within the glycolytic pathway (Fig. [120]4b). Our analysis
   revealed that CML-resistant cells exhibit higher ECAR levels when
   compared to sensitive cells (Fig. [121]1f). Combing Chi and IM
   treatment significantly reduced the IM-resiatant cells’ ECAR level
   (Fig. [122]5a). Also, IM-resistant cells’ OCR levels were restored to
   those similar to IM sensitive cells when treated with Chi and IM
   combined regimen (Fig. [123]5a, Additional file [124]1: Fig. S1d). To
   elucidate how Chi restores IM sensitivity in CML cells, we conducted
   transcriptome sequencing studies. Interestingly, RNA-Seq showed that
   Chi-only treatment didn’t change K562R’s transcriptome profile a lot,
   but combining IM with Chi promoted the transcriptional change induced
   by IM treatment (Fig. [125]5b). Our RNA-Seq result verified the
   synergistic effect of Chi and IM in IM-resistant cells (Fig. [126]3).
   Then we combined three groups of DEGs and listed out 431 DEGs
   represented the genes that changed by adding Chi into traditional IM
   treatment in IM-resistant cells (Fig. [127]5c, Additional file [128]4:
   Table S3). KEGG enrichment analysis was employed to uncover the
   underlying biological process, signaling pathways, and related
   diseases. Glycolysis, cell cycle, apoptosis, mTOR pathway, PPAR
   pathway, and CML disease were highly enriched (Fig. [129]5d, Additional
   file [130]5: Table S4). GSEA analysis of apoptosis and mTOR pathway
   also confirmed that they were positively correlated to combined
   treatment over IM-only treatment (Fig. [131]5e). Given the roles of
   AMPK, HIF1, and mTOR signaling pathways in metabolism, we investigated
   whether Chi and IM-induced metabolic alterations were associated with
   changes in these proteins. Our results demonstrated a decrease in mTOR
   phosphorylation coupled with an increase in AMPK phosphorylation in
   K562R cells following Chi and IM combined treatment (Fig. [132]6a).
   Tumor cells maintain PKM2 activation to sustain glycolytic levels,
   thereby evading drug-induced cell death. Examination of Glut1, HKII,
   PKM2, and PFKP expression levels in CML cells treated with combined Chi
   and IM revealed a downregulation of Glut1 and a substantial decrease in
   pyruvate kinase PKM2 expression in both cell types following the
   combined strategy (Fig. [133]6b). Then we performed rescue assay to
   verify the PPARγ/mTOR/PKM2 pathway’s regulatory function in IM
   resistance mechanism. We used shPPARγ to transient knockdown PPARγ in
   IM-resistant cells after Chi treatment for 24 h (Fig. [134]6c). Our
   qPCR and immunoblot results confirmed the downregulation of PPARγ
   (Fig. [135]6d, Additional file [136]1: Fig. S1f). Also, PPARγ knockdown
   lead to the upregulation of Glut1 and PKM2 (Fig. [137]6d). Then we
   tested the apoptosis ratio of IM-resistant cells treated with shPPARγ.
   Downregulation of PPARγ significantly redued the apoptosis-induction
   effect when IM-resistant cells treated with IM and Chi combined
   treatment (Fig. [138]6e). Also, we introduced mTOR activator MHY1458
   (MHY) to test whether MHY can mimic the apoptosis-induction and
   metabolic-reprogram effect of Chi treatment. We chose 1 and 5 μM (lower
   than 72 h IC50: 20.7 μM) to avoid the cytotoxic effect of MHY
   (Additional file [139]1: Fig. S1g). mTOR activation was showed by
   immunoblot test after MHY treatment (Additional file [140]1: Fig. S1h).
   The apoptosis assay verified that activation of mTOR by MHY could
   miminc the Chi’s apoptosis-induction in IM-resistant cells
   (Fig. [141]6f, Additional file [142]1: Fig. S1i). After MHY treatment,
   the pyruvate and lactate production were significant reduced which is
   similar to Chi treatment (Fig. [143]6g). Through a combination of
   metabonomic, transcriptomic, and molecular biology studies, Chi was
   verified to resensitize IM-resistant CML cells via the PPARγ/mTOR/PKM2
   pathway.

Fig. 5.

   [144]Fig. 5
   [145]Open in a new tab

   RNA-Seq identifies a potential pathway by which Chiglitazar
   resensitized K562R cells. a ECAR and OCR levels in different conditions
   using the Seahorse Bioscience Extra Cellular Flux Analyzer. For the
   glycolysis analysis, cells were firstly treated with Chi (20 μM), IM
   (2.5 μM for K562 cells and 5 μM for K562R cells) and their combination
   for 6 h, and then cells were sequentially treated with glucose (10 mM),
   oligomycin (2 μM), 2-DG (100 mM). b Heatmap showing DEGs in K562R cells
   treated with Chi (20 μM), IM (5 μM) and their combination for 48 h
   (q < 0.05 and |log fold change (FC)| > 0.5). c Venn diagram of
   different DEG groups (q < 0.05 and |log fold change (FC)| > 0.5). d
   KEGG pathway enrichment analysis of 431 overlapping DEGs. e GSEA
   analysis of mTOR pathway and apoptosis comparing combination therapy
   over IM-only treatment. Data are presented as mean ± SD. NS < 0.1234,
   *P < 0.03, **P < 0.0021, ***P < 0.0002, ****P < 0.0001

Fig. 6.

   [146]Fig. 6
   [147]Open in a new tab

   Chiglitazar resensitized K562R cells through PPARγ/mTOR/PKM2 pathway. a
   Western blotting analysis of mTOR, p-mTOR, AMPK, p-AMPK, HIF1, and
   Wnt5a expression levels after Chi-only (20 μM), IM-only (5 μM), and
   their combination treatment for 48 h in K562R cells. b Western blotting
   analysis of p-BCR/ABL, BCR/ABL, P-CRKL, Glut-1, HK2, PKM2 and PFKP
   expression levels after Chi-only (20 μM), IM-only (5 μM), and their
   combination treatment for 48 h in K562 and K562R cells. GAPDH was used
   as a loading control. c Schematic diagram of shPPARγ treatment in K562R
   cells. d The log2-fold changes in gene expression are shown. e
   Induction of apoptosis in K562R cells by 48 h of treatment with
   Chiglitazar, and shPPARγ upon 24 h treatment. f Induction of apoptosis
   by 48 h of treatment with MHY and Chiglitazar in K562R cells. g
   Pyruvate and lactate production after MHY and Chiglitazar treatment for
   48 h. Data are presented as mean ± SD. NS < 0.1234, *P < 0.03,
   **P < 0.0021, ***P < 0.0002, ****P < 0.0001

Chiglitazar resensitizes CML cells to imatinib in vivo

   To assess the clinical implications of our findings, we further
   evaluated the in vivo effects of Chi on IM-resistant CML using a K562R
   cell line-derived xenograft model (CDX) (Fig. [148]7a). Mice receiving
   different drug treatments showed no loss of weight throughout the
   experiment (Additional file [149]1: Fig. S1j). Upon euthanization of
   the mice, we noticed a significant reduction in spleen weight following
   combined treatment of IM and Chi compared to IM-only groups
   (Fig. [150]7b, c). Interestingly, while treated with Chi did not
   notably prolong the survival of CML cell-transplanted mice, combined
   treatment significantly extended the survival of IM-resistant mice
   (Fig. [151]7d). Flow cytometry analysis of human CD45+ cells in the
   bone marrow and spleen revealed a substantial reduction in human CML
   tumor blast following combined treatment with Chi and IM, compared to
   either treatment alone (Fig. [152]7e, f). Histological examination of
   the liver, spleen, kidney, and bone marrow did not reveal any
   significant toxicity in any of the treatment groups (Fig. [153]8).
   Furthermore, immunohistochemical analysis demonstrated a significant
   reduction in the proliferation marker Ki-67 in the combination
   treatment group compared to the other groups (Fig. [154]8). In summary,
   our in vivo data indicated that Chi, when combined with IM,
   significantly reduced the burden of IM-resistant CML cells and improved
   the IM-resistant mice models’ survival. These findings underscore the
   potential of this combination as a promising therapeutic strategy
   against IM-resistant CML.

Fig. 7.

   [155]Fig. 7
   [156]Open in a new tab

   Chiglitazar resensitizes CML cells to Imatinib in vivo. a Experimental
   design for CDX CML mice model. b Representative spleen images of CDX
   models after 12 days of gavage carrier, IM (50 mg/kg/day), Chi
   (15 mg/kg/day), or their combination. c Spleen weight after 12 days of
   gavage carrier, IM (50 mg/kg/day), Chi (15 mg/kg/day), or their
   combination. d Kaplan–Meyer survival analysis of CDX models treating
   with gavage carrier, IM (50 mg/kg/day), Chi (15 mg/kg/day), or their
   combination. e, f Flow cytometry analysis showing human CD45 + CML
   cells in the bone marrow and spleen after 12 days of gavage carrier, IM
   (50 mg/kg/day), Chi (15 mg/kg/day), or their combination. Data are
   presented as mean ± SD. NS < 0.1234, *P < 0.03, **P < 0.0021,
   ***P < 0.0002, ****P < 0.0001

Fig. 8.

   [157]Fig. 8
   [158]Open in a new tab

   HE staining in the liver, spleen, kidneys, and bone marrow and Ki-67
   staining in spleen across all treatment groups

Discussion

   TKIs have been widely used for treating CML patient that significantly
   improved the prognosis and life expectancy of CML patients [[159]7].
   However, the persistence of residual tumor cells necessitates continued
   medication [[160]25, [161]37]. Approximately 17% of patients display
   imatinib resistance within 5 years, highlighting the urgent need for
   innovative strategies to combat TKI resistance in CML treatment
   [[162]38–[163]41]. Previous studies have revealed metabolic shifts in
   CML resistance, including increased ROS production, imbalanced
   glutathione oxidation, and a heightened reliance on amino acid, fatty
   acid, and glutamine metabolism [[164]42–[165]46]. These studies
   provided a biological basis for metabolic intervention in IM-resistant
   CML management. Chiglitazar is a thiazolidinedione insulin sensitizer
   and has demonstrated potential in inhibiting leukemia stem cells
   [[166]27]. However, its therapeutic efficacy in CML cells remains
   unexplored. We demonstrate in this study that Chi-mediated reprogram of
   glucose metabolism is capable of resensitizing IM-resistant CML cells
   for tranditional IM therapy.

   In this study, we used transcriptomic sequencing to uncover important
   glucose metabolic differences between IM-resistant and sensitive cells,
   highlighting the discernible metabolic disparities. Metabolomic data
   also confirms the change in glucose metabolism. We then discovered that
   Chi could reduce glucose uptake and lactate production in sensitive and
   IM-resistant CML cells but didn’t induce CML cell apoptosis. It is
   consistent with the finding that CML cells with the BCR::ABL fusion
   gene tend to shift their energy metabolism toward active glycolysis.
   Our results also confirmed that IM-resistant CML cells, dependent on
   glycolysis, are more sensitive to glycolytic inhibition or glucose
   depletion (Fig. [167]2). TKI-resistant cells are reliant on glycolysis,
   particularly PKM2, which is a stress-associated rate-limiting enzyme
   [[168]47, [169]48]. The combination of Chi and IM therapy inhibits CML
   cells effectively regardless of the presence or absence of BCR::ABL
   mutations T315I. Chi appears to exert cellular effects on CML cells
   primarily by inducing metabolic shifts, altering cell cycles, and
   contrasting the apoptosis typically induced by conventional
   chemotherapeutic agents [[170]49, [171]50]. PPARγ is critical to
   glucose homeostasis and the adipogenic pathway, as it influences the
   activity of a network of genes involved in cell differentiation,
   proliferation, and apoptosis [[172]51, [173]52]. Although PPARγ has
   been suggested as a tumor suppressor, there is conflicting evidence
   regarding its role in tumorigenesis, including cancers related to CML
   [[174]25, [175]34, [176]53, [177]54]. Pyruvate kinase (PK) is a
   glycolysis regulator that orchestrates critical signaling pathways that
   promote cancer cell metabolism, proliferation, and migration
   [[178]55–[179]57]. Our study reveals that PPARγ activation reduces the
   phosphorylation of STAT5 but not its expression, which is essential for
   leukemia cell survival. Targeting PKM2 and STAT5 presents a potential
   drug target for eradicating IM-resistant CML, as supported by our
   findings of Chi and IM combination downregulating STAT5 via PKM2
   (Fig. [180]9). Interestingly, our previous study proved that Chi
   increased PPARα expression, resulting in the inhibition of glucose
   metabolism and apoptosis of AML cells. This is different from the main
   regulatory mechanism in IM-resistant CML cells. Consequently, it is
   possible that the effect of Chi on IM-resistant CML cells observed in
   our study may also be mediated by multiple mechanisms. The mechanism of
   Chi to partially inhibit the Warburg Effect could be explained by its
   direct activation of PPARγ. Metabolic shift is only part of the Warburg
   Effect, it would be interesting to test the enzymes involved in
   pyruvate metabolism prioritize like: efficient ATP production via
   mitochondrial oxidative phosphorylation. The mechanism by which Chi
   inhibits the mitochondrial oxidative phosphorylation seems mainly due
   to the limitation of metabolic substrates like pyruvate. In the future,
   it would be interesting to measure the metabolic substrate changes that
   are caused by the combinationof Chi and IM in vitro and in vivo using
   [U-13C]-glucose.

Fig. 9.

   [181]Fig. 9
   [182]Open in a new tab

   Schematic diagram of the function and mechanisms of Chiglitazar in CML
   cells. Chiglitazar activates PPARγ and therefore reduces the PKM2
   expression. This molecule inhibits the lactate production of
   IM-resistant CML cells by inhibiting the mTOR signalling pathway.
   Chiglitazar causes mitochondrial dysfunction to activate the caspase
   pathway and then induces apoptosis in CML cells

   In recent years, extensive research has been conducted regarding
   metabolism and imatinib drug resistance due to its significance in
   tumor growth, survival, and chemotherapy resistance [[183]44,
   [184]58–[185]60]. Despite insights from various perspectives,
   clinically applicable strategies to combat resistance remain elusive.
   Pioglitazone, a thiazolidinedione combined with imatinib, was effective
   in eliminating CML stem cells by inhibiting STAT5 through PPARγ
   activation [[186]25]. However, its usage is limited by adverse effects
   such as heart failure and bladder cancer [[187]26, [188]61]. This study
   demonstrates that Chiglitazar and imatinib synergistically inhibit
   CML-resistant cells’ glycolysis, which reduces the viability of
   IM-resistant cells and enhances their susceptibility to IM. Our
   findings present a promising therapeutic approach wherein the
   inhibition of the glycolytic pathway, coupled with TKI therapy, may
   overcome drug resistance. This warrants future clinical studies for an
   affordable intervention in TKI-resistant patients.

Conclusions

   Our study demonstrated the synergistic effect of Chi and IM on
   IM-sensitive and IM-resistant CML cells by activating the PPARγ and
   mTOR pathway, weakening the PKM2 and STAT5. This change thus causing
   mitochondrial dysfunction to activate the caspase pathway.
   Additionally, we demonstrated that Chi could improve the susceptibility
   of CML cells to IM. These experiments provide a mechanism basis for Chi
   as a solution for TKI-resistant patients.

Supplementary Information

   [189]40164_2024_589_MOESM1_ESM.pptx^ (2.2MB, pptx)

   Additional file 1: Figure S1. (a–c) Typical flow cytometry diagrams of
   apoptosis level in K562, K562R, and Baf3-T315I cells treated with
   Chi-only, IM-only, and their combination for 48 h. (d) OCR level of
   K562 and K562R cells. (e) Glucose uptake and lactate production in K562
   and K562R cells after Chi-only (20 μM), IM-only (5 μM), and their
   combination for 24 h. (f) Western blotting analysis of PPARγ protein
   after shPPARγ transfection. (g) Cell viability percentage of MHY1458
   for 24, 48, and 72 h in K562R cells. (h) Western blotting analysis of
   mTOR expression and phosphorylation levels after MHY1458 treatment. (i)
   Typical flow cytometry diagrams of apoptosis level in K562R cells
   treated with IM-only, and MHY combination for 24, 48, 72 h. (j) Weight
   changes of CDX models after drug treatment with Chi-only, IM-only, and
   their combination for 16 days. Data are presented as mean ± SD.
   NS < 0.1234, *P < 0.03, **P < 0.0021, ***P < 0.0002, ****P < 0.0001.
   [190]40164_2024_589_MOESM2_ESM.xls^ (368KB, xls)

   Additional file 2: Table S1. Differential expressed gene list comparing
   K562R over K562 cells.
   [191]40164_2024_589_MOESM3_ESM.xls^ (31KB, xls)

   Additional file 3: Tables S2. Selected enriched biological process and
   pathways generated by DEGs of table S1.
   [192]40164_2024_589_MOESM4_ESM.xls^ (101KB, xls)

   Additional file 4: Table S3. 431 DEGs generated from venn diagram.
   [193]40164_2024_589_MOESM5_ESM.xls^ (88KB, xls)

   Additional file 5: Table S4. KEGG enrichment analysis of 431 DEGs in
   table S3.

Acknowledgements