Intracellular Phosphoflow Cytometry of Acute Myeloid Leukemia Patient-Derived Xenotransplants

Here, a phosphoflow cytometry-based method is described to analyze signaling downstream of the mTORC1, JAK/STAT5, and MAPK pathways in acute human myeloid leukemia cells xenografted into mice and obtained from bone marrow aspirates. p-STAT5, p-4EBP1, p-RPS6, and p-ERK1/2 levels are simultaneously measured using a next-generation spectral flow cytometer with high sensitivity.

To adapt and resist approved treatments, acute myeloid leukemia (AML) cells activate specific molecular pathways that lead to changes in gene expression, protein levels and activity. In this protocol, an approach is reported to explore targets phosphorylated downstream of oncogenic signaling in AML: p-STAT5 (Tyr694), p-4EBP1 (Thr37/46), p-RPS6 (Ser240/244), and p-ERK1/2 (Thr202/Tyr204). This method enables the assessment of how these pathways—major regulators of stemness maintenance, immune evasion, protein synthesis, and adaptation to oxidative and metabolic stress—are modulated by one or more tested compounds in bone marrow cells harvested from live mice by aspiration before and after the treatment phase. This minimally invasive method preserves cell integrity and reduces stress compared to bone-crushing techniques, which can induce damage and potentially affect experimental outcomes. To optimize intracellular antibody staining for flow cytometric analysis, a protocol was developed using paraformaldehyde fixation and methanol permeabilization. This approach ensures high staining precision and minimizes background noise, enabling reliable detection of intracellular signaling markers. One of the main advantages of this protocol is the development of a multiparametric antibody panel, allowing for simultaneous assessment of the four pathways within the same sample. Using a next-generation spectral flow cytometer with high sensitivity, dynamic shifts in pathway activation were observed depending on treatment conditions compared to pretreatment baseline levels in the same mice. This methodology enables precise in vivo analysis of signaling pathway modulation in patient-derived xenograft bone marrow samples without requiring euthanasia of the animals, providing valuable insight into the adaptive mechanisms of AML cells, and can guide the evaluation of therapeutic strategies aimed at targeting these pathways to overcome resistance.

Acute myeloid leukemia (AML) is an aggressive hematologic malignancy characterized by the accumulation of immature myeloid progenitor cells in the bone marrow and peripheral blood. This disrupts normal hematopoiesis, leading to life-threatening cytopenia and systemic complications. Although advances in chemotherapeutic regimens, targeted therapies, and hematopoietic stem cell transplantation have improved outcomes for some patients, overall, 5-year survival rates remain around 30%, with worse prognosis in older patients or those with adverse genetic profiles¹. A significant challenge in the management of AML is the frequent emergence of drug resistance, which contributes to relapse and treatment failure². This underscores the importance of gaining a deeper understanding of the molecular and cellular mechanisms driving the progression and therapeutic resistance of AML.

To address this challenge, a new method is presented for bone marrow aspirations coupled with multiplex intracellular phosphoflow cytometry, offering a powerful tool to investigate intracellular signaling pathways in AML patient-derived xenograft (PDX) models (Figure 1A,B). Although bone marrow aspiration in PDX models has been previously described³, this protocol has been optimized to preserve leukemic cells for phosphoflow analysis. The overall goal of this method is to provide a minimally invasive procedure that is longitudinally applicable during leukemia progression and therapeutic response with single-cell resolution. By allowing for repeated sampling from the same animal, this technique offers a more accurate representation of disease evolution under treatment. The rationale behind the development of this technique lies in the need for a high-resolution, dynamic assessment of intracellular signaling in AML cells. Traditional methods, such as Western blotting, require large cell numbers, lack single-cell resolution, and multiplexing⁴. In contrast, phosphoflow cytometry preserves cellular heterogeneity and enables the detection of multiple phosphorylated signaling proteins in distinct leukemic subpopulations⁵, offering key insights into pathway activation in response to AML treatments.

Central to the biology of AML are signaling pathways that regulate cellular proliferation, survival, and metabolic adaptation, including the Mitogen-Activated Protein Kinase (MAPK), Mechanistic Target of Rapamycin Complex 1 (mTORC1), and Janus Kinase/Signal Transducer and Activator of Transcription 5 (JAK/STAT5) signaling pathways. Beyond their roles in leukemic cell proliferation and survival, these pathways are also critically involved in key oncogenic processes such as stemness maintenance, immune evasion, and adaptation to oxidative and metabolic stress⁴. In addition to promoting cell growth and survival, the cross-talk among these pathways orchestrates critical processes such as transcription, translation, and cellular metabolism, enabling AML cells to sustain their growth and resist apoptotic signals, even in the face of therapeutic interventions^6,7 (Figure 1A).

The MAPK pathway, which includes key effectors such as p-ERK1/2 (Extracellular signal-regulated kinase 1/2), plays a crucial role in the integration of extracellular signals, such as growth factors and cytokines, to regulate cell proliferation and survival. ERK1/2 activation occurs through the RAS (Rat Sarcoma)-RAF (Rapidly Accelerated Fibrosarcoma)-MEK (MAPK/ERK Kinase)-ERK cascade, where RAS-GTP recruits RAF, leading to sequential phosphorylation of MEK1/2 and then ERK1/2 at Thr202/Tyr204. Once phosphorylated, ERK1/2 dimerizes and translocates to the nucleus, where it phosphorylates transcription factors such as MYC (Myelocytomatosis viral oncogene homolog), ELK1 (ETS Like-1 Protein), and AP-1 (Activator Protein-1), promoting cell proliferation, differentiation block, and survival⁸. In AML, mutations in FLT3 (fms-like tyrosine kinase 3), RAS, or KIT frequently result in constitutive ERK activation^9,10 (Figure 1A).

Mutations in FLT3, RAS, or KIT also cause the upregulation of mTORC1 signaling, which enables AML growth and therapeutic resistance by supporting oncogenic processes such as metabolic rewiring, modulation of protein synthesis, ribosome biogenesis, and autophagy¹¹. Through the regulation of mRNA translation, mTORC1 facilitates the production of oncogenic proteins and other essential factors for the progression of AML. A key group of mTORC1 substrates includes 4E-BPs (Eukaryotic initiation factor 4E-binding proteins). In their hypophosphorylated state, 4E-BPs bind to eIF4E (Eukaryotic initiation factor 4E), inhibiting cap-dependent translation. Phosphorylation of 4E-BP1 at Thr37/46 by mTORC1 causes the release of eIF4E, enabling the initiation of translation for key oncogenic mRNAs, such as MYC, CCND1 (Cyclin D1), and MCL-1 (Myeloid cell leukemia 1), thereby promoting leukemic proliferation and survival^8,12. Additionally, RPS6 phosphorylation at Ser240/244, mediated by S6K1 (Ribosomal protein S6 kinase beta-1) downstream of mTORC1, enhances ribosome biogenesis, and mRNA translation, increasing the synthesis of proteins necessary for metabolic adaptation, stress resistance, and rapid proliferation^8,13. Notably, mTORC1 activity is tightly linked to metabolic adaptation, a critical survival strategy employed by AML cells under therapeutic stress^13,14,15 (Figure 1A).

The JAK/STAT5 pathway is another crucial signaling axis in AML, particularly in cases with mutations affecting cytokine receptors or signaling mediators such as JAK2, FLT3, and CALR (calreticulin)^16,17. STAT5 is activated in response to cytokine signaling through receptors such as FLT3 and JAK2. Upon ligand binding, associated Janus kinases (JAKs) phosphorylate STAT5 at Tyr694. Phosphorylated STAT5 dimerizes and translocates to the nucleus, where it binds to specific DNA sequences to regulate the transcription of genes involved in cell survival, proliferation, and differentiation⁸. In AML, constitutive activation of STAT5, often due to mutations in FLT3 or JAK2, leads to persistent expression of genes that promote leukemogenesis¹⁸ (Figure 1A).

Beyond their individual contributions, these pathways converge to regulate both transcription and translation, shaping the proteome of AML cells in ways that promote survival and resistance. In particular, mRNA translation is emerging as a key factor in AML pathophysiology, as it allows for the rapid production of oncogenic proteins and stress response factors that enable AML cells to adapt to environmental challenges and evade the effects of targeted therapies. Dysregulation of translation machinery, such as eukaryotic initiation factors (eIFs) or ribosomal proteins, has been implicated in therapeutic resistance and poor prognosis in AML¹⁹. A detailed investigation of the roles of the MAPK, mTORC1, and JAK/STAT5 pathways in transcriptional and translational regulation is essential to gain a comprehensive understanding of the molecular mechanisms underlying AML progression and resistance. Such insights are critical for identifying new biomarkers of treatment response and designing novel therapeutic strategies that target these pathways to overcome resistance. This article provides a protocol specifically designed to investigate these signaling networks in AML patient-derived xenograft (PDX) models.

One of the key advantages of this protocol is the integration of bone marrow aspiration with intracellular phosphoflow cytometry, allowing for a dynamic and minimally invasive assessment of signaling pathway activation in AML patient-derived xenograft (PDX) models. This is particularly valuable for monitoring the activation status of key pathways such as MAPK, mTORC1, and JAK/STAT5 in response to targeted therapies. The combination of these techniques enables the acquisition of a comprehensive and high-resolution understanding of AML biology, ultimately aiding in the development of more effective therapeutic strategies.

The following experiments were performed with approval from the McGill University Animal Care Committee (CIHR PJT-186019) and the Institutional Review Board of the Jewish General Hospital (11-047). In this protocol, male NOD-scid IL2Rg^null-3/GM/SF (NSGS) mice, aged 8 weeks to 6 months and weighing 20-30 g, were previously transplanted with human patient-derived AML cells. Details of the reagents and the equipment used in this study are listed in the Table of Materials.

1. Bone marrow aspiration

CAUTION: Safety precautions surrounding the use of needles should be followed.

1. Peri-operative analgesia: 1-4 h prior to the bone marrow aspiration, inject the mice subcutaneously (s.c.) with 0.1 mg/kg of buprenorphine slow release. This ensures analgesia for up to 72 h post-procedure.
2. Prepare the surgical area under the laminar flow hood with a chirurgical pad, hair trimmer, heating lamp, and all necessary material for bone marrow aspiration (25 G syringes, 1.5 mL microcentrifuge tubes containing 0.5 mL of PBS on ice, 70% isopropanol swabs). Secure the nose cone for isoflurane inhalation using adhesive tape (Figure 2A).
3. Prepare two 25 G syringes for each aspirate, one for bone puncture ("dry needle") and one whose hub has been flushed with PBS for aspiration ("flushed needle").
4. Anesthetize the mouse using isoflurane inhalation delivered by a certified vaporizer, according to the manufacturer's instructions and veterinary guidelines. Local procedures may vary. For reference, this protocol followed McGill University's standard operating procedures for mouse anesthesia²⁰.
5. Confirm anesthesia depth via lack of response to foot pinch and muscle relaxation. Apply veterinary ointment to the eyes to prevent dryness while the animal is under anesthesia. Continuously monitor respiratory rate during anesthesia.
6. Shave the entire leg where the aspiration will be performed using the hair trimmer. Disinfect the aspiration site (femur) with an antiseptic pad saturated with isopropyl alcohol (Figure 2B).
7. Expose the articular surface of the femur, where the puncture site is located, by bending the knee and immobilizing the leg using the non-dominant hand. Place the thumb on the tibia, the index finger on the femur, and the middle finger, stabilizing everything on the outer side of the two bones (Figure 2C).
   NOTE: Ensure very stable positioning of the leg using the non-dominant hand to prevent any movement of the leg during the aspiration that may lead to failure or injury. If the immobilization of the leg is not stable, reposition the hand and retry prior to any puncture.
8. While immobilizing with the non-dominant hand, disinfect the knee area again with isopropyl alcohol (Figure 2C) to move the remaining hairs away to improve the visualization of the bone structure.
9. Use the first dry 25 G syringe to create a hole in the femoral articular surface by positioning it in the middle of the femur joint and rotating it without forcing (Figure 2D,E). Ensure the needle of the syringe is completely aligned (0° deviation) with the longitudinal axis of the femur to enter the marrow correctly.
   NOTE: Maintain this alignment to ensure smooth entry without causing unnecessary damage to surrounding tissues. When the needle enters the middle of the femoral bone, a slight decrease in resistance indicates successful penetration into the marrow cavity (Figure 2F). Avoid applying excessive force on the needle at this step, as it may cause loss of correct alignment, fracture, or trauma. By using a rotatory drilling motion, the chances are greater that the needle will follow the path of least resistance and enter the femoral cavity correctly. A needle properly positioned inside the femur should stay vertical even when the hand holding it is removed.
10. Once in the marrow, gradually remove the syringe by rotating (Figure 2G) and insert the second flushed syringe into the hole made by the first needle (Figure 2H,I).
    NOTE: A second syringe is used because the first one is often blocked by bone during drilling.
11. Apply vacuum to the syringe inserted within the femur while gently rotating and gradually withdrawing the syringe. Approximately 10-20 µL of bone marrow should be visible within the hub of the syringe (Figure 2J).
    NOTE: If needed, gently re-advance the needle within the bone to optimize marrow extraction.
12. Transfer the bone marrow by flushing the contents of the syringe into the chilled microcentrifuge tube containing 500 µL of PBS (Figure 2K). Keep on ice and proceed to steps 2-3 as soon as possible.
13. Apply gentle pressure to the puncture site for 30 s with an isopropanol swab to stop any bleeding (Figure 2L).
14. Place the mouse in a clean recovery cage warmed using a warming lamp. Monitor breathing and mobility until the animal is awake and fully recovered before placing it in communal cages with the rest of the mice. Check for signs of bleeding.

2. Live/dead and surface marker staining

NOTE: Keep cells on ice throughout the entire procedure. Prepare master mixes of antibody staining solutions (for step 2.4 and step 2.7) prior to step 2 and step 3, keeping them at 4 °C protected from light. Prepare a fresh 1.6% formaldehyde solution (step 3.1) and a 100% methanol solution keeping at -20 °C (step 3.6).

1. Count the viable cells using acridine orange/propidium iodide (AO/PI) with the cell counter²¹.
2. Transfer the appropriate volume for 1 million cells into a new 1.5 mL tube. If the volume is not sufficient, take the entire volume. Pool the remaining cells and aliquot into 3 replicate tubes (1 M each) for isotype control staining and 1 tube for unstained control.
3. Centrifuge cells at 500 x g for 5 min at 4°C. Aspirate gently and discard the supernatant using a vacuum aspiration system.
4. Add 200 µL per sample of fluorescent cell staining solution diluted 1/100 in PBS to label the dead cells. Gently resuspend the cells with the pipette. Do the same with the extra tubes for isotype control staining.
5. Incubate on ice for 10 min protected from light.
6. Centrifuge the cells at 500 x g for 5 min at 4 °C. Aspirate gently and discard the supernatant using a vacuum system.
7. Add 100 µL per sample of hCD45-BUV395 antibody diluted 1/100 in PBS containing 2% fetal bovine serum to label the human hematopoietic cells. Do the same with the extra tubes kept for isotype controls.
8. Incubate on ice for 15 min protected from light. Proceed immediately to step 3.

3. Fixation and permeabilization

1. Prepare a fresh 1.6% formaldehyde/PBS solution: For every sample, mix 29.5 µL of 37% formaldehyde stock solution + 970.5 µL of PBS (account for some extra volume as a precaution). Keep the 1.6% formaldehyde/PBS solution at room temperature.
   CAUTION: Formaldehyde should be handled in a chemical hood and disposed of in accordance with institutional safety regulations.
2. Centrifuge the cells at 500 x g for 5 min at 4 °C. Aspirate and discard the supernatant using a vacuum system.
3. Resuspend in 1 mL of 1.6% formaldehyde/PBS solution.
4. Incubate for 10 min at room temperature, protected from light.
5. Centrifuge the cells at 500 x g for 5 min at 4 °C. Aspirate and discard the supernatant using a vacuum system.
6. Add 1 mL of 100% methanol chilled at -20 °C directly into the tube.
7. Incubate samples at -20 °C for 30 min protected from light.
   NOTE: At this point, it is possible to store fixed and permeabilized cells at -80 °C for up to one month before analysis.

4. Phosphoflow staining

NOTE: The following antibody panel was validated using the flow cytometer equipped with the following lasers: 320 nm, 355 nm, 405 nm, 488 nm, 561 nm, 637 nm, and 808 nm. This panel should be re-validated if different cytometers are used with different laser and detector setups.

1. Prepare an antibody master mix solution with the following dilutions in a commercially available staining buffer (Table 1), calculating a volume of 50 µL for every sample. Keep at 4 °C protected from light.
2. For washes (steps 4.8, 4.10), keep PBS on ice (2 mL for every sample).
3. Prepare a second master mix solution for the isotype control technical replicates in the staining buffer (Table 1), calculating a volume of 50 µL for every sample. Keep at 4 °C protected from light. A minimum of two isotype control samples is recommended for technical reproducibility.
4. Centrifuge the cells from step 3.7 at 500 x g for 5 min at 4 °C. Aspirate and discard the supernatant using a vacuum system.
   NOTE: Following fixation and permeabilization, the cells may appear more translucent.
5. Add 50 µL of antibody mix solution (step 4.1) or isotype control mix solution (step 4.2). Gently resuspend the cells with a pipette.
6. Incubate all the samples overnight at 4 °C protected from light.
7. Centrifuge cells at 500 x g for 5 min at 4 °C. Aspirate and discard the supernatant using a vacuum system.
8. Wash with 1000 µL of PBS by resuspending gently with the pipette.
9. Centrifuge cells at 500 x g for 5 min at 4 °C. Aspirate and discard the supernatant using a vacuum system.
10. Perform a second wash with 1000 µL of PBS by resuspending gently with the pipette.
11. Centrifuge cells at 500 x g for 5 min at 4 °C. Aspirate and discard the supernatant using a vacuum system.
12. Resuspend samples in 200 µL of PBS. Transfer to FACS tubes or plates for analysis using flow cytometry (Figure 3).
13. Prepare single-color compensation controls with beads by incubating 0.1 µL of every antibody with compensation beads, according to manufacturer's instructions (see Table of Materials).

The experimental design aims to monitor the phosphorylation state of selected proteins of interest at different timepoints, for example, 1 day pretreatment initiation and 15 days post-treatment initiation. This baseline serves as a critical control to verify the veracity of subsequent changes in signaling pathway activation after treatment. After fifteen days of treatment, a second comparison is performed to evaluate pathway modulation in treated mice versus vehicle-treated controls (Figure 4). This two-step approach allows for the distinction of treatment-induced effects from baseline variability.

For this study, an AML PDX mouse model was utilized. The mice were transplanted with patient-derived AML cells characterized by the following mutations: DNMT3^AMR882H with a variant allele fraction (VAF) of 48 %, NPM1^W288fs VAF 36%, FLT3^ITD VAF 10%, and IDH2^R140Q VAF 44%. After transplantation, three weeks were allowed for tumor growth before initiating treatment. The FLT3-mutant-transplanted PDX mice were treated for fifteen days with one of three regimens: placebo, gilteritinib (15 mg/kg daily), or a combination of venetoclax (50 mg/kg daily), 5-azacitidine (2.5 mg/kg daily), and cedarizudine (3 mg/kg daily) (Figure 4A).

With respect to the results obtained, it is noteworthy to mention that treatment with a venetoclax-based regimen resulted in significantly enhanced STAT5 and RPS6 phosphorylation, suggesting that these pathways contribute to therapy resistance (Figure 4E). In contrast, gilteritinib, a drug targeting FLT3, did not reduce the phosphorylation status of mutant FLT3 effectors. It is possible that in the AML specimen tested, the presence of additional secondary mutations other than FLT3-ITD (e.g., IDH2) contributes to activation of these pathways or that the drug treatment over two weeks results in selective elimination of AML cells with decreased signaling. Since all the phosphoproteins analyzed are gated from the same cell population, it is possible to compare phosphoprotein activation. In this experiment, the strongest correlation in phosphoprotein staining appears to be between p-STAT5 and p-RPS6 (Supplementary Figure 1A), which are both upregulated in this experimental group after treatment (Figure 4E). There seems to be a lesser relationship between p-STAT5 and p-ERK (Supplementary Figure 1B). Importantly, cell size (assessed using FSC-A) is not a major contributor to overall p-STAT5 positivity, highlighting that staining positivity is not merely a product of cell size (Supplementary Figure 1C). This experimental framework enables a comprehensive evaluation of the effects of these treatments on signaling pathway modulation.

Figure 1: Visualization of the target signaling pathways and the strategy for their analysis. (A) Overview of the signaling pathways involving p-STAT5 (Tyr694), p-4EBP1 (Thr37/46), p-RPS6 (Ser240/244), and p-ERK1/2 (Thr202/Tyr204). In the p-STAT5 pathway, JAK undergoes autophosphorylation, increasing its catalytic activity, then phosphorylates specific tyrosine residues, creating docking sites for STAT proteins. Phosphorylated STAT5 (Tyr694) homodimerizes and translocates into the nucleus, where it binds to specific STAT-responsive elements in the promoter regions of target genes. In the RAS-RAF-MEK-ERK cascade, autophosphorylation of the FLT3 receptor or other receptor tyrosine kinases (RTKs) upon ligand binding leads to the exchange of GDP for GTP on RAS, thereby activating RAS. Activated RAS-GTP recruits RAF to the plasma membrane, where RAF undergoes a conformational change and becomes active. RAF then phosphorylates and activates MEK1/2, a dual-specificity kinase, which in turn phosphorylates ERK at Thr202/Tyr204 residues. Activated ERK1/2 translocates into the nucleus, where it phosphorylates specific transcription factors. ERK1/2 is also able, through intermediate signaling pathways, to increase mTORC1 activity and phosphorylate eIF4E at specific sites to increase translation activity. Activated by nutrients and growth factors through FLT3 and other RTKs, mTORC1 phosphorylates downstream targets like S6K1 and 4E-BP1. S6K1, in turn, phosphorylates RPS6, which enhances ribosome function and activates translation. Hyperphosphorylation of 4E-BP1 releases eIF4E, allowing the assembly of the translation initiation complex eIF4F and activation of cap-dependent translation. All of these pathways are highly involved in AML, contributing to acquired adaptations by increasing cell proliferation, metabolism, and resistance to apoptotic signals. (B) presents the main steps of the bone marrow aspiration and intracellular phosphoflow staining coupled protocol. Please click here to view a larger version of this figure.

Figure 2: Visualization of critical steps during the bone marrow aspiration procedure. The surgical area and necessary materials are set up under a laminar flow hood. After anesthesia by isoflurane inhalation (A), the femoral articular surface is prepared for aspiration (B,C). The femoral bone is punctured by a first syringe aligned to the middle of the femoral bone (D). It is critical to avoid applying strong pressure to the needle and instead perform a rotatory drilling motion; this is usually sufficient for the needle to perforate the surface of the bone and follow the path of the femoral cavity. Next, the bone marrow is aspirated with a second syringe through the puncture site (I,J). After transferring the bone marrow aspirate to a microcentrifuge tube (K), gentle pressure is applied to stop the bleeding in the area (L), and the mouse is placed into a warm and clean cage for recuperation. Please click here to view a larger version of this figure.

Figure 3: Workflow for the gating strategy analysis, including isotype controls. (A) represents gating strategies for live cells, exclusion of doublets, and hCD45-positive AML cells. (B) shows signals for each phosphoprotein in brown and isotype controls in black (two biological replicates). Please click here to view a larger version of this figure.

Figure 4: Workflow for the detection and analysis of intracellular phosphoproteins after immuno-stating of viable bone marrow AML cells. (A) NSG-SGM3 mice engrafted with patient-donated AML cells were treated with vehicle (control) or venetoclax, 5-azacytidine, and cedazuridine (vene/aza/CDZ) or gilteritinib. After 15 days, Bone Marrow (BM) aspirates were processed. (B) represents the gating strategy for the analysis of viable AML cells (e.g., hCD45/Zombie-NIR positive) 15 days after treatment. (C,D) represent the normalization of the signal for each of the phosphoantibodies as indicated, in 10,000 AML cells isolated from mice treated with venetoclax/5-azacytidine/cedazuridine or vehicle (B) or gilteritinib or vehicle (C). The traces represent the signals in AML cells isolated from the control mouse treated with vehicle (light blue), vene/aza/CDZ (red) or gilteritinib (purple). (E) The histogram represents the mean value changes in the staining intensities for each phosphoprotein tested on live human AML CD45⁺ cells from three mice and SD. Error bars represent standard deviation. The asterisks indicate significant differences in staining intensities as assessed by ANOVA with p values below 0.05 (*) and 0.01 (**). Please click here to view a larger version of this figure.

Table 1: List of optimized dilutions for antibodies.

Supplementary Figure 1: Comparison of p-STAT5, p-RPS6, p-ERK1/2, and FSC-A in a representative AML PDX sample. This figure shows an example of the relationship between p-STAT5, p-ERK, and p-RPS6 in mouse "E9" at day 15 of treatment with venetoclax/5-azacitidine/cedazuridine (result from Figure 4C). In this sample, the strongest correlation in phosphoprotein staining appears to be between p-STAT5 and p-RPS6 (A), which are both upregulated in this experimental group after treatment (Figure 4E). There seems to be a lesser relationship between p-STAT5 and p-ERK (B). Importantly, cell size (assessed using FSC-A) is not a major contributor to overall p-STAT5 positivity, highlighting that staining positivity is not merely a product of cell size (C). Please click here to download this File.

Critical steps
The use of immunologically based techniques in the study of complex signaling pathways through specific detection of phosphoproteins requires that experimental variables are tightly controlled²², sample preparation is meticulous, and complementary techniques are employed for validation. Integrating these practices ensures the reproducibility, precision, and robustness of the data, ultimately contributing to more reliable biological conclusions. These efforts not only enhance the scientific rigor of the method but also broaden its applicability in complex studies of cell signaling and regulation. Proper anesthesia and gentle handling are critical to minimizing physiological stress, which can significantly alter experimental outcomes, particularly in studies focused on immune responses or signaling pathways. Ensuring appropriate analgesic protocols further aids in maintaining animal welfare while preventing pain-related stress that might interfere with the data. Maintaining aseptic conditions during the procedure is vital to prevent infections, which can lead to inflammatory responses and potentially confound experimental results. This involves sterilizing equipment, using clean gloves, and disinfecting the aspiration site thoroughly.

Phosphorylated proteins are highly susceptible to dephosphorylation by phosphatases during the manipulation of biological specimens^22,23. To preserve phosphorylation states during processing, immediate fixation is essential. Two different fixation/permeabilization methods were tested: a cytofix/cytoperm kit and the common formaldehyde/methanol method outlined in the protocol. Among these, the methanol-based method provided cleaner results with brighter fluorescence signals, making it the preferred choice. Following fixation, samples can be stored short-term at -20 °C or long-term at -80 °C. In experiments, fixed samples stored for up to three months at -80 °C yielded consistent results. For antibody staining, an overnight incubation can enhance accuracy, especially for subtle phosphorylation shifts, although this depends on the specific experiment and drug effects under investigation. A shorter 1 h incubation was also tested on ice, protected from light, which produced acceptable results in certain cases. To ensure an accurate assessment of background noise, isotype controls are recommended as a reference for nonspecific binding, as shown in Figure 3.

Modifications and troubleshooting of the technique
To develop technical proficiency, trainees in our group practice the needle insertion technique on euthanized mice, where proper placement can be confirmed by dissection. The live animal procedure is performed most efficiently by two operators, where one can focus on the aspirations while an assistant is dedicated to the tasks of animal monitoring and sample handling. In rare instances where it is not possible to obtain an aspiration following several attempts, it is advisable to abort the procedure since the risk of injury increases with every attempt. Post-procedure monitoring of the animals is equally important. Observing for signs of infection, pain, or stress ensures the well-being of the mice and the reliability of the study. Housing conditions should support recovery, with appropriate bedding and nutrition to minimize external stressors.

For flow cytometric analysis, processing all fixed and preserved samples from different bone marrow aspiration days simultaneously is recommended. Variations in sample rehydration and staining due to differences in handling days or operators can affect the fluorescence sensitivity detected by the cytometer. To ensure consistent results, optimize the cytometer voltage settings to maintain all stained samples within the detection range before proceeding. The Alexa Fluor 488-conjugated p-4EBP1 protein exhibits the highest brightness among the markers in this panel, making it a reliable reference for setting the maximal detection range. While the antibody panel presented in this protocol did not require compensation adjustments, this may vary depending on the spectral flow cytometer used.

Limitations
One of the major limitations of this technique is the number of cells extracted during bone marrow aspiration. To ensure reliable results, it is essential to stain enough cells to analyze a minimum of approximately 10,000 cells per sample. Analyzing too few cells can lead to significant variations between samples, compromising the accuracy and interpretability of the data. Ensuring an adequate number of cells is critical for robust and reproducible mechanistic studies.

For data analysis, the gating strategy illustrated in Figure 3A is applied. First, after excluding doublets, the cells of interest are identified using a human CD45 antibody coupled with BUV395. Next, the gating for live cells is done using Zombie NIR. Finally, the median fluorescence intensity (MFI) is calculated for each phosphoprotein (Figure 3B). All the cells analyzed for all phosphorylated proteins were gated using the same strategy, ensuring consistency and comparability across AML cell populations.

Another limitation of this method is the potential for background noise, which can arise from nonspecific antibody binding, insufficient washing, cell autofluorescence, or improper permeabilization. This may affect signal specificity and reduce the accuracy of phosphoflow cytometry analysis. For users who encounter these issues, an additional washing step is recommended to reduce background noise, and Fc-blocking reagents can be used to minimize nonspecific binding. Additionally, dead cell fluorescence can interfere with results, making the use of viability dyes essential. Importantly, for accurate analysis, the number of cells analyzed should not be fewer than 500 AML cells to ensure reliable and reproducible results. It is also important to ensure that a similar number of cells is used during staining to maintain consistency across samples. Despite these limitations, careful optimization of controls and gating strategies can enhance the reliability of the method.

Significance of the technique
The bone marrow aspiration technique has already been published³; however, an updated technique is included in this protocol with an explanation highlighting two advantages: improved immobilization and the use of a two-needle approach to enhance precision and consistency in bone marrow aspiration, preventing any issue with blockage of the needle with bone fragments. For longitudinal studies, femoral aspiration can be performed multiple times, providing a powerful method to track biological changes over time in individual animals. A minimum one-month recovery interval is recommended between aspirations on the same femur to allow sufficient bone healing and regeneration of bone marrow. When using the contralateral femur, a shorter interval of one week is sufficient to ensure the mouse has recovered from the effects of analgesia and anesthesia without introducing undue stress or physiological variability.

Altogether, this combined murine bone marrow aspiration and flow cytometry technique provides valuable insights into key signaling pathways, including JAK/p-STAT5, mTORC1/p-4EBP1, mTORC1/p-RPS6, and MEK/p-ERK1/2. These pathways play significant roles in treatment resistance in AML. This technique allows for repeated sampling from the same animal, enabling real-time monitoring of disease progression and therapy response at the single-cell level.

Phosphoflow cytometry captures phosphorylation events and preserves cellular heterogeneity^4,23,24. This makes it particularly valuable for understanding the dynamic cellular responses to therapy in specific leukemic subpopulations.

Future applications
This protocol provides a valuable tool for studying the underlying mechanisms of treatment adaptation and the emergence of resistance in AML, paving the way for deeper insights and potential therapeutic advancements. Future studies could integrate this approach with single-cell transcriptomics or proteomics to further dissect the molecular mechanisms driving resistance and disease progression²⁵. Additionally, applying this method to other hematologic malignancies or inflammatory conditions could broaden its impact in translational research.

By refining this protocol and addressing its limitations, this approach could become a standard method for studying phosphoprotein signaling in hematologic diseases, ultimately aiding in the development of more effective therapeutic strategies. In conclusion, this protocol provides a valuable tool for studying the underlying mechanisms of treatment adaptation and the emergence of resistance in AML, paving the way for deeper insights and potential therapeutic advancements.

The authors have nothing to disclose.

This work was supported by Transition Grants for LH and FEM from the Cole Foundation, and grants from the Leukemia and Lymphoma Society of Canada and the Canadian Institutes for Health Research (PJT-186019) to LH and FEM. FEM is an FRSQ Junior 2 Clinical Scientist and LH is an FRSQ Junior 2 Scientist. VG holds a doctoral fellowship from the Cole Foundation. Figure 1 was created with BioRender under a licensed agreement. Flow cytometry plots were generated using FlowJo software. Special thanks to Dr. Colin Crist and Victoria Richard for granting access to their animal surgery facility.