Targeting MTHFD2 disrupts mitochondrial redox homeostasis and restores venetoclax sensitivity in acute myeloid leukemia

This study establishes that targeting the mitochondrial one-carbon metabolism enzyme MTHFD2 disrupts redox homeostasis and nucleotide synthesis in acute myeloid leukemia, thereby suppressing disease progression and restoring sensitivity to venetoclax in both treatment-naive and resistant models.

The Gist

The Big Picture: Finding a Weak Spot in the Enemy's Armor

Imagine Acute Myeloid Leukemia (AML) as a very aggressive, fast-growing army of bad cells. For a long time, doctors have had a powerful weapon called Venetoclax to fight this army. However, the enemy is smart; they often build shields that make Venetoclax useless, leading to drug resistance.

This paper is like a team of detectives (scientists) who found a hidden, critical weakness in the leukemia army's supply lines. They discovered that these cancer cells rely on a specific machine called MTHFD2 to survive. If you break this machine, the cancer starves and its internal "fire alarms" go off, killing the cells.

Here is how they figured it out, broken down into simple steps:

1. The "One-Carbon" Factory

Think of a cell as a busy city. To build new buildings (DNA) and keep the lights on (energy), the city needs a constant supply of bricks and electricity.

• The Bricks: The cancer cells need to make "bricks" called nucleotides to copy their DNA and multiply rapidly.
• The Electricity: They also need to manage oxidative stress (think of this as internal rust or fire). If there is too much "rust," the cell explodes.

The enzyme MTHFD2 is the manager of a special factory inside the cell's power plant (the mitochondria). It does two crucial jobs:

1. It helps manufacture the "bricks" (nucleotides) needed for the cancer to grow.
2. It produces a special "fire extinguisher" (NADPH) that keeps the internal rust (ROS) under control.

2. The "Bad Guy" vs. The "Good Guy"

The researchers wanted to see what happens if they shut down this MTHFD2 factory.

• The Cancer Cells: When they turned off MTHFD2 in leukemia cells, the factory stopped. The cancer cells ran out of bricks (couldn't copy DNA) and their internal fires started raging because they lost their fire extinguishers. The cancer cells died.
• The Healthy Cells: When they turned off MTHFD2 in healthy blood stem cells, nothing bad happened. The healthy cells didn't seem to need this factory as much as the cancer cells did.
  • Analogy: Imagine a luxury sports car (cancer) and a reliable family sedan (healthy cell). The sports car needs high-octane, specialized fuel to run at top speed. If you cut off that fuel, the sports car stops dead. The family sedan, however, can switch to regular gas and keep driving just fine.

3. The "Double-Whammy" Strategy

The scientists found that the cancer cells were so dependent on MTHFD2 that they couldn't survive without it. But they also found a way to use this weakness to fix the problem of drug resistance.

• The Problem: Some cancer cells have learned to ignore Venetoclax (the current drug).
• The Solution: The researchers tested a new drug called DS18561882, which acts like a wrench thrown into the MTHFD2 factory.
• The Result:
  1. Alone: DS18561882 killed the cancer cells.
  2. With Venetoclax: When they used DS18561882 alongside Venetoclax, the cancer cells were crushed even faster.
  3. The Resistant Cells: Even the cancer cells that had built a shield against Venetoclax were terrified of the MTHFD2 wrench. The new drug made them sensitive to Venetoclax again!

4. Why This Matters

This is a game-changer for two reasons:

1. Precision: It targets the cancer's specific metabolic weakness without hurting the healthy blood cells (the "family sedan" keeps running).
2. Reviving Old Drugs: It offers a way to save patients who have stopped responding to Venetoclax. By breaking the MTHFD2 factory, the researchers essentially "disarmed" the cancer's shield, allowing the old drug to work again.

The Bottom Line

The scientists discovered that leukemia cells are like a house built on a shaky foundation (MTHFD2). While the house looks strong from the outside, if you pull out that one specific beam, the whole thing collapses.

By using a new drug (DS18561882) to pull that beam, they can stop the cancer from growing and, more importantly, help patients who were previously out of options to respond to treatment again. It's a new key to unlock a door that was previously stuck shut.

Technical Summary

1. Problem Statement

Acute Myeloid Leukemia (AML) remains a lethal malignancy with high rates of relapse and resistance to standard therapies, particularly to the BCL-2 inhibitor Venetoclax. While one-carbon metabolism is known to be dysregulated in cancer, the specific role of the mitochondrial enzyme MTHFD2 in AML pathogenesis and therapeutic resistance was incompletely understood.

• Knowledge Gap: Previous studies suggested MTHFD2 inhibition was effective, but the primary inhibitor used (TH9619) was found to be unable to penetrate mitochondria, instead inhibiting nuclear MTHFD2 and cytosolic MTHFD1. Consequently, the specific contribution of mitochondrial MTHFD2 to AML survival, redox balance, and drug resistance remained unresolved.
• Clinical Need: There is an urgent need to identify metabolic vulnerabilities that can sensitize AML cells to Venetoclax, especially in patients with resistant subtypes (e.g., FLT3, TP53, or spliceosome mutations).

2. Methodology

The authors employed a multi-faceted approach combining genetic ablation, stable isotope tracing, pharmacological inhibition, and patient-derived xenograft models:

• Genetic Ablation: Used CRISPR/Cas9 to delete MTHFD2 in human AML cell lines (MOLM14, NOMO1, OCI-AML2/3) and a genetically engineered mouse model (GEMM) with Mx1-Cre driven deletion of Mthfd2 in an MLL-AF9 leukemia model.
• Metabolic Tracing: Utilized 2,3,3-²H-Serine and U-¹³C-Serine isotopologue tracing coupled with HILIC-MS to distinguish between mitochondrial and cytosolic one-carbon flux, specifically quantifying contributions to dTTP, purine synthesis (IMP, GMP, AMP), NADH, and NADPH production.
• Pharmacological Inhibition: Tested DS18561882, a small-molecule inhibitor capable of penetrating mitochondria, and compared its effects to TH9619.
• Redox Analysis: Measured mitochondrial Reactive Oxygen Species (ROS) using MitoSOX and assessed the rescue of cell viability via overexpression of antioxidant enzymes (SOD2/Catalase) and exogenous purine supplementation.
• Clinical Cohort Testing: Screened 60 primary patient-derived AML (PD-AML) samples and healthy donor bone marrow cells to assess drug sensitivity and synergy with Venetoclax, Azacitidine (AZA), Cytarabine (Ara-C), and Doxorubicin (DOXO).
• Resistance Models: Generated Venetoclax-resistant AML cell lines (MOLM14-R) to test the efficacy of MTHFD2 inhibition in refractory disease.

3. Key Contributions

• Validation of Mitochondrial MTHFD2: The study definitively establishes that mitochondrial MTHFD2 (not just nuclear/cytosolic forms) is a critical dependency for AML survival.
• Mechanistic Insight: It elucidates a dual mechanism of action: MTHFD2 supports AML by driving de novo purine synthesis and maintaining mitochondrial redox homeostasis (via NADPH production).
• Therapeutic Synergy: The paper demonstrates that targeting mitochondrial MTHFD2 with DS18561882 synergizes with Venetoclax, effectively restoring sensitivity in Venetoclax-resistant AML models.
• Safety Profile: Confirms that MTHFD2 inhibition is selective for leukemic cells, sparing healthy hematopoietic stem and progenitor cells (HSPCs).

4. Key Results

A. MTHFD2 is Essential for Leukemia but Dispensable for Normal Hematopoiesis

• Genetic Deletion: CRISPR-mediated deletion of MTHFD2 significantly reduced AML cell proliferation, induced cell cycle arrest, and increased apoptosis in vitro. In vivo, Mthfd2 deletion in the MLL-AF9 mouse model significantly delayed leukemia onset and reduced white blood cell counts.
• Normal HSPCs: Conversely, deletion of Mthfd2 in healthy murine hematopoietic cells did not affect steady-state blood counts, lineage composition, or the long-term reconstitution potential of HSPCs in competitive transplantation assays.

B. Metabolic Mechanisms: Purine Synthesis and Redox Balance

• Purine Dependency: Isotopic tracing revealed that AML cells rely heavily on the mitochondrial arm of one-carbon metabolism for de novo purine synthesis (IMP, GMP, AMP). Deletion of MTHFD2 caused a significant drop in purine levels, which could be partially rescued by exogenous purine supplementation.
• Redox Homeostasis: MTHFD2 deletion led to a marked decrease in mitochondrial NADPH production, resulting in elevated mitochondrial superoxide (ROS).
• Rescue Experiments: The anti-leukemic effects of MTHFD2 inhibition were almost completely reversed when cells were treated with both exogenous purines and mitochondrial ROS neutralization (SOD2/Catalase overexpression), confirming that the dual loss of nucleotides and redox balance drives cell death.

C. Pharmacological Inhibition with DS18561882

• Mitochondrial Penetration: Unlike TH9619, DS18561882 successfully blocked mitochondrial one-carbon flux (reducing M+1 labeling in dTTP and purines) and increased mitochondrial ROS, phenocopying genetic deletion.
• Selectivity: DS18561882 reduced viability in a panel of human and mouse AML cell lines and primary PD-AML samples in a dose-dependent manner but had minimal effect on healthy donor bone marrow cells.
• Genotype Correlation: Samples with IDH2, KIT, RUNX1, or ASXL1 mutations showed higher sensitivity to DS18561882, while KRAS-mutant samples were less responsive.

D. Synergy with Venetoclax and Reversal of Resistance

• Synergy: DS18561882 showed strong synergistic cooperation with Venetoclax (but not Ara-C or DOXO) in reducing PD-AML viability.
• Venetoclax Resistance: In samples classified as "Venetoclax-Low Responders" (often harboring FLT3, TP53, or spliceosome mutations), the combination of Venetoclax and DS18561882 significantly reduced viability compared to Venetoclax alone.
• Resistant Cell Lines: Venetoclax-resistant MOLM14 cells (MOLM14-R), which overexpress MCL-1, exhibited extreme sensitivity to DS18561882 (IC₅₀ dropped from ~2.0 µM to ~0.003 µM).
• Mechanism of Synergy: The combination treatment resulted in a supra-additive accumulation of mitochondrial ROS, exceeding the levels seen with either agent alone, leading to catastrophic oxidative stress and cell death.

5. Significance

• Therapeutic Strategy: This study provides a strong preclinical rationale for targeting mitochondrial MTHFD2 in AML, specifically as a strategy to overcome Venetoclax resistance.
• Precision Medicine: The identification of specific genetic mutations (e.g., IDH2, ASXL1) that predict sensitivity to MTHFD2 inhibition could guide patient stratification.
• Safety: The finding that MTHFD2 is dispensable for healthy HSPC function suggests a wide therapeutic window, minimizing the risk of hematopoietic toxicity associated with current chemotherapy regimens.
• Future Directions: The results support the development of clinical trials combining MTHFD2 inhibitors (like DS18561882) with Venetoclax and/or Azacitidine for high-risk or resistant AML patients.

In conclusion, the paper redefines MTHFD2 as a critical mitochondrial metabolic vulnerability in AML, linking nucleotide synthesis and redox homeostasis to therapeutic resistance, and offers a promising combinatorial approach to improve AML outcomes.