Ethical approval
For primary CML patient-derived samples, cells were obtained from peripheral blood or leukapheresis product. All samples were from patients that were in chronic-phase at the time of diagnosis. All patients gave written informed consent in agreement with the Declaration of Helsinki and with the approval of the National Health Service (NHS) Greater Glasgow and Clyde Institutional Review Board. Ethical approval was granted to the research tissue bank (REC 15/WS/0077) and for using surplus human tissue in research (REC 10/S0704/60). These results are not sex specific. While more male patient samples were used (17) than female (8) this was due to the amount of cells available for each patient in biobank. The median age at diagnosis of patients was 47-years old, ranging from 24 to 70 years old.
Animal experiments were performed in accordance with Home Office regulations and under approved project licence PPL PP2518370 and personal licence PIL IE2DD924E.
Primary samples
Primary CML samples were sourced from CML patients either from 50 mL peripheral blood or leukapheresis product. Patients were in chronic phase CML at the time of diagnosis and gave informed consent in accordance with the Declaration of Helsinki and approval of the National Health Service (NHS) Greater Glasgow and CLyde Institutional Review Board. The CD34+ cells were isolated using the CliniMACS (Miltenyi Biotec) and purity verified to be >95% by flow cytometry (Apoptosis and CD34 analysis). Non-leukemic cells were isolated from femoral head material using human CD34 MicroBeads (Miltenyi Biotec), according to the manufacturer’s instructions. Purity was verified by flow cytometry to be >90%.
Statistical analyses
No statistical methods were used to determine sample size. For experiments, a minimum of four samples was used to give adequate power. The investigators were not blinded to sample or treatment during experiments. Non-parametric data was analysed using Kruskal–Wallis test.
Patient information
Please see Supplementary Table 1 for available patient information.
Reagents
Imatinib mesylate was purchased from LC Laboratories (I-5508). A stock solution of 10 mM was prepared in sterile Milli-Q water and stored at 4 °C. MSDC-0160 (Apex Biotechnology: B3702) UK5099 (Merck: PZ0160-5MG) and Omacetaxine mepesuccinate (ChemGenex Pharmaceuticals) were made up in DMSO. Stock concentrations were 50 mM for MSDC-0160, and 10 mM for UK5099 and Omacetaxine.
Cell culture
Primary CML samples were thawed and recovered overnight in Plasmax, a physiological cell culture medium37. This medium was supplemented with nutrients and growth factors as described in Supplementary Table 2, then filter sterilised through a 0.2 µM filter (Fisher Scientific: 10509821). For cell line experiments, the medium was RPMI with 10% dialysed FBS and 1% penicillin/streptomycin. For tracing experiments 11 mM 13C6 glucose was added to glucose-free RPMI. Cells were counted and viability calculated using a CASY counter with the following parameters to gate live cells: E-cur 7.5-22.5, N-cur 5.25-20.5. Doubling time over 48 h was calculated by dividing the cell density into four times the seeding density (4 × 4 × 105cells/mL = 16). For stromal co-culture experiment, irradiated M2-10B4 and S1/S1 mouse cell lines that are genetically engineered to express human cytokines were seeded (8 × 104 cells each in 1000 μL) in DMEM supplemented with 10% FBS and hydrocortisone onto collagen-coated plates. The following day, medium was removed and 2 × 105 primary CML resuspended in 1000 μl of Plasmax (13-UC-glucose), seeded on top of the feeder cell layer for 24-h culture in absence or presence of 50 μM MSDC-1060. Non-adherent cells were then harvested, and samples prepared for LC-MS. All cell lines were obtained form DSMZ and regularly tested to ensure they were free of contamination such as mycoplasma.
RNA extraction
RNA was extracted from 200,000 CML CD34+ cells using an RNA easy mini kit (Qiagen) according to the manufacturer’s instructions.
RNA-seq
Libraries were generated with the TruSeq Stranded mRNA LT Kit (Illumina) and ran on the Illumina Next-Seq 500 using the Illumina High-Output 75 cycles kit (2 × 36 cycles, paired end reads, single index). FastQ files were prepared with bcl2fastq (v. 2.20.0.422, Illumina). QC, alignment, and parsing of files into count matrices was performed in command line, with subsequent differential gene expression (DGE) analysis performed in R (1.2.1335). Adapter trimming was performed using Scythe (version 0.981), and Sickle (version 0.940), was used to trim bases with quality scores of <20. Prior to and after this pre-processing fastqc (version 0.11.2) was run to ascertain sequence quality, alongside the efficacy of the pre-processing steps. Trimmed reads were indexed and aligned using Hisat2 (version 2.1.0). Hisat2 indexes (GRCh38 genome_tran) were obtained from the John Hopkins Center for Computational Biology, 2020. Samtools (version 0.1.19044428 cd) view was used to convert the resulting.sam to.bam files, whilst samtools sort was used to sort the.bam files. Assembly was achieved through the use of stringtie (John Hopkins Center for Computational Biology, 2020), with output.gtf files converted to count matrices using the python script prepDE.py (stringtie version 1.3.3b.Linux_x86_64). Reads were assembled using an annotated reference human genome (GRCh38.p13), obtained from GENCODE (GENECODE, 2020). DESeq2 (version 1.26.0) was used to generate results sets from the gene and transcript count matrices. G genes with read counts too low to allow for the calculation of p and adjusted p-values (padj: Benjamini-Hochberg) were removed from the data sets leaving gene and transcript counts of sizes 16,069 and 45,218 respectively. Microarray analysis was carried out in R studio (version 1.1.4).
Microarray analysis
Data were analysed using Limma (version 3.34.9).
GSEA analysis
GSEA (version 4.1) was conducted on pre-ranked lists (ranked by pi score calculated by multiplying LOG fold change by -LOG (corrected p-value)).
Palmitate conjugation for stable isotope tracing experiments
For conjugation of palmitate, a 20 mM stock was made in 100% EtOH. This was incubated on a shaking heater, 60 oC until the solution clarified. The palmitate was then added to 10% BSA (ultra-fatty acid free in EBSS, Roche, 03117057001) to give a palmitate:BSA ratio of 1:3 BSA. This solution was left for 15 min in water bath (37 oC), and then added to final concentration in complete medium.
LC-MS sample preparation
Cells were counted, and viability calculated using a CASY counter with the following parameters: E-cur 7.5-22.5, N-cur 5.25-20.5. Timepoints were chosen to ensure equal viability within patient-derived samples for between experimental arms. Cell number was multiplied by peak volume (size) and this number was used to calculate volume of solvent to extract in. Cells were pelleted by centrifugation (400 × g, 10 min, room temperature) at which medium was removed for YSI analysis. Cells were then washed twice with ice cold PBS (Calcium and Magnesium free: Thermo Fisher Scientific) with pulse centrifugation (12,000 × g, 15 s, 4 oC) used between and after washes. Cells were then extracted in LC-grade ACN:MeOH:H2O solvent (−20 oC, 5:3:2) by disrupting the pellet by pipetting followed by a 5 second vortex. Finally, debris was pelleted by centrifugation (16,000 × g, 10 min, 4 oC), supernatant transferred to LC-MS glass vials that were stored at −80 oC until analysis.
LC-MS
The LC system composed of a ZIC-pHILIC column (SeQuant, 150 × 2.1 mm, 5 µm, Merck KGaA) with a ZIC-pHILIC guard column (SeQuant, 20 × 2.1 mm) with an UltiMate 3000 HPLC system (Thermo Fisher Scientific). The aqueous mobile-phase solvent was 20 mM ammonium carbonate-0.1% ammonium hydroxide solution and with acetonitrile being used for organic mobile phase. A linear biphasic LC gradient was conducted from 80% organic to 80% aqueous for 15 min for a total run time of 22 min. The column temperature was maintained at 45 °C flow rate set to 200 µL/min. The MS used in this study was a qExactive Plus Orbitrap Mass Spectrometer (Thermo Fisher Scientific) operating in polarity switching mode. The MS set up was calibrated using a custom CALMIX in both ionization modes before analysis and a tune file targeted towards the lower m/z range was used. Full scan (MS1) data was acquired in both ionization modes in profile mode at 70,000 resolution (at m/z range 75–1000), an automatic gain control (AGC) target of 1 × 106 (max fill time of 250 ms), with spray voltages +4.5 kV (capillary +50 V, tube: +70 kV, skimmer: +20 V) and −3.5 kV (capillary −50 V, tube: −70 kV, skimmer: −20 V) and s-lens RF level of 50 for the front optics. The capillary temperature 375 °C, probe temperature 50 °C, sheath gas flow rate 25, auxiliary gas flow rate 15a.u. 15 and sweep gas flow rate of 1a.u. The mass accuracy achieved for all metabolites was below 5ppm. Data acquisition was achieved with Thermo Xcalibur 4.3.73.11 software.
LC-MS analysis
The peak areas of different metabolites were determined using Tracefinder 4.1 software (Thermo Fisher Scientific). Metabolites were identified by accurate mass of the singly charged ion and by known retention times on the pHILIC column. A commercially available standard compound mix (Merck: MSMLS-1EA) had been analysed previously on our LC-MS system to determine accurate ion masses and retention times. The 13 C labelling was determined by quantifying peak areas for the accurate mass of all isotopologues of each metabolite.
Steady state analysis
Two independent experiments were performed, data normalised to first sample in each experiment, processed through Autoplotter (version 2.3D)66, samples as experiments, and normalised data as replicates. These data were processed on Metabonalyst using the Rlog transformation and mean-centring to ensure data followed a normal distribution. For MSDC-0160 treated samples a batch correction (to non-drug control) was used prior to analysis using Metabonalyst. Volcano plots were generated, and an FDR adjusted t-test threshold 0.1 was utilised. Note both fold changes and p-values are log transformed. Top-25 scoring metabolites are included in heatmap.
Analysis of stable isotope tracing experiments
Biological and technical replicates were processed through Autoplotter (version 2.3D)66 and natural abundance was corrected for. Fraction of enrichment was calculated as m + 2 and above. For CML samples treated with or without imatinib, multiple paired T-tests were conducted on Log10 transformed values in Graphpad Prism. The correction for multiple comparisons was the two-stage step-up method of Benjamini, Krieger and Yekutieli with an FDR (Q) value of 10%. For volcano plots Log10 FC vs Log10 (q value) are plotted with q values > 5% in red and >10% in blue. For normal vs CML comparisons, the process was the same with the exception that unpaired t-tests were used. For PC activity in normal vs CML, this is denoted by fraction m + 3. For imatinib vs untreated we look at relative activity to PDH (m + 3/m + 2 ratio) and conducted multiple t-tests with correction for multiple comparisons being the two-stage step-up method of Benjamini, Krieger and Yekutieli with an FDR (Q) value of 10%. To examine contribution to carbon pool we corrected contribution of each isotopologue by the number of carbons it contributes to a given metabolite, i.e., for glutamine contribution from m + 5 is greater than m + 3. Note that for cell line experiments standard deviation is shown.
Bioanalyser
A YSI 2900 biochemistry analyser was used as per the manufacturer’s instructions to quantify glucose, lactate, glutamine, and glutamate in the media. The concentration of each metabolite was normalised to cell number and the rate of uptake or secretion per hour was calculated relative to cell free medium. The exchange rate per 48 h [secretion (+) or consumption (−)] for a specific metabolite (x) was obtained according to the following equation:
$$frac{Delta {{{{{rm{metabolite}}}}}}}{({cell},{number},{day},0+{cell},{number},{day},2)/2}$$
where Δmetabolite = ((x) mmol spent medium − (x) mmol cell-free medium).
Oxygen consumption rate measurements (Seahorse analysis)
Seahorse (XF96) assay was conducted as previously described12. CMLCD34+ were treated for 24 h with 1 µM imatinib prior to analysis.
Lentivirus production
Lentiviruses for pLentiCRISPRv.2 were produced by the calcium phosphate method using pCMV-VSV-G (envelope plasmid: RRID: Addgene_8454) and psPAX2 (packaging constructs: RRID: Addgene_12260) vectors and human embryonic kidney (HEK) 293FT cells for transfection.
Multi-omic analysis
The joint pathway tool of Metabonalyst67 was used for analysis of paired data. Here we used the hypergeometric test for enrichment analysis, the topology measure was degree centrality, and the integration method was combined queries.
Gene ontology analysis
Panther35, (version 17.0) was used for mapping to different sets and Fisher’s exact test used to compare fraction of number of upregulated genes per pathway/total number of upregulated genes with same fraction of all expressed genes.
Survival analysis
A multivariable Cox proportional hazards model was fitted to TCGA data in R software. Two datasets were analysed, one including and the other excluding the FAB M3 subtype. To simplify the model, a backward stepwise model selection procedure was applied to the complex Cox survival model, which originally included age, sex, FLT3_ITD, protocol, transplant_type, PC, FAB, and cytogenetic_risk predictors (full model). The reduced model (reduced model) was obtained by retaining age, FLT3_ITD, protocol, transplant_type, PC, and cytogenetic_risk predictors, while dropping the interaction term between PC and FAB, from the original model.
The resulting models can be represented as:
Full model: proportional hazard ~ age + sex + FLT3_ITD + protocol + transplant_type + PC * (FAB + cytogenetic_risk)
Reduced model: proportional hazard ~ age + FLT3_ITD + protocol + transplant_type + PC * cytogenetic_risk.
Western blot analysis
Cells were lysed in RIPA buffer containing mini-Complete protease inhibitor cocktail and phosphatase inhibitors (both Roche). Total protein concentration was quantified using a Pierce BCA kit (Thermo Fisher Scientific: 23227). Equal amounts of protein (5–7.5 µg) were heated at 95 °C for 5 min and separated (120 V) in 4–12% gels (Novex) for SDS-PAGE. Proteins were transferred onto PDVF membranes (Thermo Fisher Scientific: 21882) then blocked in 2% BSA (in Tris-buffered saline, 0.01% Tween (TBS-T)) for one hour. Next, membranes were incubated overnight at 4 °C with the primary antibodies. The primary antibodies used were p-AMPK (Cell signalling, Cat #2531, diluted 1:1000), AMPK (Cell signalling, Cat #2532, diluted 1:1000), p-CRKL (Cell signalling, Cat #3181, diluted 1:500), CDK (Cell signalling, clone POH1, Cat #9116, diluted 1:1000), PC (Proteintech Cat# 16588-1-AP) and H3 (Active Motif, clone MABI 0301, Cat #39763 diluted 1:1400). The membranes were rinsed three times with TBS-T, then incubated with secondary HRP-linked antibodies; Anti-rabbit IgG HRP-linked Ab (Cell Signalling Cat#7074, diluted 1:10,000) and Anti-mouse IgG HRP-linked Ab (Cell Signalling Cat#7076, diluted 1:10,000), for 1 hour at room temperature. The SuperSignal West Femto Maxi detection system was used (Thermo Fisher Scientific: 34095) and imaging was carried out using a LI-COR Odyssey Fc gel-doc system. All uncropped and unprocessed scans are in Source Data File.
Apoptosis and CD34 analysis
Cells were stained with Annexin V (fluorescein isothiocyanate (FITC, BioLegend: Cat# 640906, 5 µL/test), 7-AAD (BD Bioscience: Cat# 559925, 5 µL/test) and CD34+ (APC, BD Bioscience, clone 581, Cat# 555824, 2 µL/test) in 50 µL Hanks’ Balanced Salt Solution (HBSS) for 20 min (room temperature in dark). CML cells were analysed by flow cytometry (BD FACSVerse, BD FACSuite version 1.0.6.5320) and data were analysed using FloJo (version 10).
CRISPR-Cas9 mediated deletion
To target the human PC gene, guides were designed using the optimized tool https://www.genscript.com/gRNA-database.html. Two guides were chosen and ordered from Integrated DNA Technologies. These were annealed and cloned in Bsmb I–digested lentiCRISPRv.2-puro (RRID: Addgene_52961). After stable integration of lentiCRISPRv.2 using lentiviral transfection and 1-week selection using puromycin (2.5 µg/ml), guides were validated by performing Western blotting. Oligonucleotides from IDT are as follows:
g1 forward: CACCGCAGGCCGCGGCCGATGAGAT
g1 reverse: AAACATCTCATCGGCCGCGGCCTGC
g2 forward: CTGAAGTTCCGAACAGTCCA
g2 reverse: TGGACTGTTCGGAACTTCAG
g3 forward: CACCGACAGGTGTTCCCGTTGTCCC
g3 reverse: AAACGGGACAACGGGAACACCTGTC
Patient derived xenografts experiments
For in vivo engraftment, 1 million CML CD34+ cells, were transplanted via tail vein into sublethally irradiated (2.5 Gy) female NRG-W41 (Jackson Laboratory: RRID:IMSR_JAX:026014, NOD.Cg-Rag1tm1MomKitW-41JIl2rgtm1Wjl) mice aged 8-10 weeks. Female mice are used as this gives higher engraftment. For housing, light is 12 h light and 12 h darkness, temperature was 20–24 °C, humidity was 45–65% and feeding (LBS biotech: SDS Cat# SDS801730) and water were ad libitum. Eight weeks following transplant, drug treatment was started with both imatinib (100 mg/kg/day (50 mg/kg BID); oral gavage) and MSDC-0160 (30 mg/kg; oral gavage once daily) given for 4 weeks. At the end of treatment, bone marrow cells were collected. This was done by placing inverted cut leg bones into 0.5 mL Eppendorf tubes with holes at bottom. These in turn were placed within 1.5 mL Eppendorf tubes containing PBS, centrifuged (12,000 × g, 20 s). Cells were stained (300 µl/test) with anti-mouse (APC-Cy7 BD Biosciences, clone 30-F11, Cat# 557659, RRID: AB_396774, 1 µl), anti-human CD45 (FITC; BD Biosciences, clone HI30, Cat# 555482, RRID: AB_395874, 10 µl), anti-human CD34 (APC; BD Biosciences, clone 581, Cat# 555824, RRID: AB_398614, 2 µl) and anti-human CD38 (PerCP; BioLegend, clone HIT2, Cat# 303520, RRID: AB_893313, 2 µl) antibodies for 20 min in dark (room temperature) prior to flow cytometry analysis as described above. This model does not result in high disease level, hence for animal welfare, post-irradiation sickness or tolerability of treatment were the main factors monitored for.
Fluorescence-activated cell sorting (FACS)
For isolation of the CD34+CD38− cell population from normal or CML samples, CD34+ cells were stained with anti-human CD34 and anti-human CD38 as described above and sorted using a FACSAria Fusion Cell sorter (BD Biosciences, BD FACSDiva Software v8.0.1).
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
- SEO Powered Content & PR Distribution. Get Amplified Today.
- PlatoData.Network Vertical Generative Ai. Empower Yourself. Access Here.
- PlatoAiStream. Web3 Intelligence. Knowledge Amplified. Access Here.
- PlatoESG. Automotive / EVs, Carbon, CleanTech, Energy, Environment, Solar, Waste Management. Access Here.
- PlatoHealth. Biotech and Clinical Trials Intelligence. Access Here.
- ChartPrime. Elevate your Trading Game with ChartPrime. Access Here.
- BlockOffsets. Modernizing Environmental Offset Ownership. Access Here.
- Source: https://www.nature.com/articles/s41467-023-40222-z