RAS-mutant clones drive extramedullary acute myeloid leukemia

This study identifies RAS-mutant clones as a key molecular driver of extramedullary acute myeloid leukemia (eAML), demonstrating that these mutations promote leukemic tissue infiltration and extramedullary growth through a JAML-PI3K/AKT signaling axis, thereby revealing AKT inhibition as a potential therapeutic strategy.

The Gist

The Big Picture: The "Hiding" Cancer

Imagine Acute Myeloid Leukemia (AML) as a bad weed growing in a garden (your bone marrow). Usually, doctors check the soil to see how bad the weeds are and decide how to treat them.

But sometimes, these weeds don't just stay in the soil. They send runners out to grow in the flower beds, the fence posts, and the walls of the garden. This is called Extramedullary AML (eAML). It's like the cancer is hiding in the walls of the house, making it very hard to kill with standard treatments that only target the soil.

This paper asks a big question: Why does the cancer decide to hide in the walls? And how can we stop it?

---

1. The Culprit: The "Rebel Engine" (RAS Mutations)

The researchers looked at 85 samples of these "wall-growing" cancers. They found that in 41% of the cases, the cancer cells had a specific broken part in their engine called the RAS gene.

• The Analogy: Think of a normal cell as a car with a working speedometer. It knows when to stop and when to go. A cell with a RAS mutation is like a car with a stuck gas pedal. It's constantly revving, but more importantly, it has a special "off-road" mode that makes it really good at driving over rough terrain (tissue) and breaking through fences (barriers).

The study found that these "stuck gas pedal" cars were much more common in the wall-growing cancers (eAML) than in the soil-growing cancers (bone marrow). In fact, the cancer seemed to prefer to grow these mutations when it moved to the walls.

2. The Superpower: Breaking Through Walls

The researchers wanted to prove that this broken engine actually caused the cancer to invade new areas. They took normal cancer cells and used a molecular "scissors" (CRISPR) to install the stuck gas pedal (the RAS mutation) into them.

• The Experiment: They put these modified cells in a maze.
  • Normal cells: Stuck in the starting room.
  • RAS-mutant cells: Zoomed through the maze, broke through the walls, and invaded the next room.

They tested this in chicken embryos and mice, and the result was the same: The cells with the broken engine were much better at invading new territory and growing into solid tumors outside the bone marrow.

3. The Secret Weapon: The "Crowbar" (JAML)

So, how does the RAS mutation give the cancer this superpower? The researchers looked at the genetic "instruction manual" of the cells and found a key player: a protein called JAML.

• The Analogy: If the RAS mutation is the engine, JAML is the crowbar.
  • When the engine (RAS) is stuck, it tells the cell to build more crowbars (JAML).
  • These crowbars help the cancer cells pry open the tight seals between your body's cells, allowing them to slip into the walls and hide.

The study showed that if you remove the crowbars (silence the JAML gene), the cancer cells lose their ability to invade, even if they still have the stuck gas pedal.

4. The Chain Reaction: The "Power Line" (PI3K/AKT)

The researchers discovered that the crowbar (JAML) doesn't just pry open doors; it also flips a switch on the wall that turns on a power line called PI3K/AKT.

• The Analogy:
  1. RAS (Stuck Gas) turns on the JAML (Crowbar) factory.
  2. The Crowbar flips the PI3K/AKT (Power Line) switch.
  3. The Power Line sends electricity to the cell, telling it: "Go! Move! Break through!"

5. The Solution: Cutting the Power

Here is the most exciting part for patients. The researchers found that if they used a drug to cut the power line (an AKT inhibitor), the cancer cells stopped moving.

• The Takeaway: Even though we can't easily remove the "stuck gas pedal" (RAS mutation) right now, we can cut the power line (block AKT) that the cancer uses to invade your body.
• Real-world hope: Drugs that block this power line (like MK-2206 or Capivasertib) already exist and are being tested in other cancers. This study suggests they could be a "silver bullet" for stopping AML from hiding in the walls.

Summary

1. The Problem: AML often hides in body tissues outside the bone marrow, where standard chemo can't reach it.
2. The Cause: A specific genetic mutation (RAS) is very common in these hiding spots. It acts like a stuck gas pedal.
3. The Mechanism: This mutation forces the cancer to build a "crowbar" (JAML) that pries open tissue barriers and turns on a "power line" (AKT) that drives invasion.
4. The Cure: We might be able to stop this invasion by using existing drugs that cut the power line (AKT inhibitors), preventing the cancer from hiding in the walls and causing relapse.

In short: The cancer has a secret weapon to break into your tissues. The researchers found the weapon, figured out how it works, and identified a switch to turn it off.

Technical Summary

1. Problem Statement

Extramedullary acute myeloid leukemia (eAML), also known as myeloid sarcoma, is a manifestation where myeloid blasts invade non-hematopoietic tissues. While it occurs in approximately 9–20% of AML cases, its molecular drivers remain poorly defined. Current knowledge relies heavily on bone marrow (BM) profiling, which may not reflect the molecular landscape of extramedullary sites. Furthermore, eAML often serves as a reservoir for systemic relapse and is frequently resistant to standard therapies. The specific mechanisms driving the infiltration of leukemic blasts into extramedullary tissues and the potential therapeutic vulnerabilities of these clones are largely unknown.

2. Methodology

The study employed a multi-faceted approach combining clinical genomics, functional in vitro/in vivo assays, and transcriptomic analysis:

• Clinical Cohort & Sequencing: Targeted Next-Generation Sequencing (NGS) was performed on 85 eAML biopsies (FFPE and frozen) from four centers (Graz, Innsbruck, Kiel, Dresden). Mutations were analyzed in 49 myeloid neoplasm-associated genes.
• Comparative Analysis: eAML mutation frequencies were compared against three large BM-only AML cohorts (LB-MUG, TCGA-LAML, Beat-AML; total n > 1,300). Paired eAML and BM samples from 25 patients were analyzed to assess clonal evolution using normalized Variant Allele Frequencies (nVAF).
• Isogenic Cell Line Models: To establish causality, the authors used CRISPR/Cas9 and rAAV6 to introduce the most common mutation found in the cohort (NRAS G12D) into RAS-wildtype (WT) human leukemia cell lines (K562 and HEL). This created isogenic pairs (NRAS G12D/WT vs. NRAS WT/WT) with endogenous expression levels.
• Functional Assays:
  • In vitro: Transwell migration and invasion assays.
  • In vivo (Chicken): Chorioallantoic membrane (CAM) assays to assess tissue infiltration and tumor formation.
  • In vivo (Mouse): Subcutaneous injection of cells into NRG nude mice to model extramedullary growth.
• Transcriptomics & Mechanism: RNA-Seq was performed on murine tumors to identify differentially expressed genes. Candidate genes were validated via qPCR and Western blot. The study utilized the Beat-AML cohort to correlate gene expression with leukocyte migration signatures and PI3K/AKT pathway activity.
• Therapeutic Intervention: Functional validation involved siRNA-mediated knockdown of the target gene (JAML) and pharmacological inhibition of the AKT pathway (using MK-2206).

3. Key Contributions

• Molecular Landscape of eAML: This study provides one of the largest genomic analyses of eAML to date, identifying RAS or RAS-modifying gene mutations (RASMUT) in 41% of cases.
• Clonal Expansion: Demonstrated that RASMUT clones are not only enriched in eAML compared to BM but also undergo clonal expansion or de-novo emergence specifically at the extramedullary site.
• Mechanistic Discovery: Identified Junctional Adhesion Molecule-Like (JAML) as the critical downstream effector linking RAS mutations to tissue infiltration.
• Signaling Axis: Elucidated a RASMUT-JAML-PI3K/AKT axis, showing that RAS mutations drive eAML growth primarily through PI3K/AKT activation (rather than MAPK/ERK) mediated by JAML upregulation.
• Therapeutic Strategy: Validated that pharmacological inhibition of AKT reverses the migratory and invasive phenotype of RASMUT cells, proposing AKT inhibitors as a targeted therapy for RASMUT-driven eAML.

4. Key Results

A. Genomic Enrichment and Clonal Dynamics

• Frequency: RASMUT (NRAS, KRAS, PTPN11, CBL, NF1) were found in 41% of eAML cases, significantly higher than the 26% observed in non-eAML AML cohorts (p=0.016).
• Clonal Evolution: In paired samples, normalized VAFs of RASMUT were significantly higher in eAML tissue than in matched BM (70% vs. 16%, p=0.003). In 72% of paired cases, the RASMUT clone expanded at the extramedullary site, while in 16% it appeared de-novo.
• Mutation Hotspots: NRAS codon 12 (40%) and KRAS codon 13 (46%) were the most frequent mutation sites.

B. Functional Role of RASMUT

• Migration/Invasion: Isogenic NRAS G12D cells (K562 and HEL) and murine Kras G12D HSPCs showed significantly increased migration and invasion compared to WT controls.
• In Vivo Growth: In CAM assays and subcutaneous mouse models, NRAS G12D cells formed significantly larger tumors and infiltrated tissues more aggressively than WT cells.
• Proliferation: Notably, the increased tumor burden was not due to increased proliferation (Ki-67 staining was unchanged), suggesting RASMUT enhances tissue infiltration and survival rather than cell cycle speed.

C. Mechanism: The JAML-PI3K/AKT Axis

• Transcriptomics: RNA-Seq of NRAS G12D tumors revealed 258 differentially expressed genes. Among the 65 upregulated genes, JAML was identified as the top candidate linked to migration.
• Correlation: In the Beat-AML cohort, JAML expression strongly correlated with a "leukocyte migration" gene signature (Spearman ρ=0.71) and a "PI3K/AKT activation" signature (Spearman ρ=0.43).
• Causality:
  • JAML Knockdown: siRNA-mediated silencing of JAML in NRAS G12D cells significantly reduced migration.
  • Pathway Activation: NRAS G12D cells showed elevated p-AKT (but not p-ERK). JAML knockdown reduced p-AKT levels.
  • Pharmacological Blockade: Treatment with the AKT inhibitor MK-2206 abolished the enhanced migration of NRAS G12D cells, confirming the dependency on this axis.

5. Significance

• Clinical Implications: The findings suggest that relying solely on bone marrow biopsies for molecular profiling in AML patients may miss critical extramedullary drivers (RASMUT). This could lead to inadequate treatment and relapse.
• Therapeutic Vulnerability: The study identifies AKT signaling as a specific therapeutic vulnerability in RASMUT-associated eAML. Since AKT inhibitors (e.g., capivasertib, MK-2206) are already in clinical use or trials, this offers a rapid translational pathway for treating refractory eAML.
• Biological Insight: The research clarifies that RAS mutations in AML drive disease progression not just through proliferation, but by reprogramming cells for tissue infiltration via the JAML-PI3K/AKT axis. This distinguishes the biology of eAML from systemic AML and highlights the need for site-specific therapeutic strategies.
• Future Directions: While JAML is a key driver, its dual role in immune regulation (promoting tumor growth but also potentially aiding CD8+ T cell function) necessitates careful development of targeted delivery methods to avoid immune suppression.

In conclusion, this paper delineates the molecular landscape of eAML, proving that RAS mutations are a primary driver of extramedullary infiltration via a JAML-mediated PI3K/AKT pathway, offering a new rationale for targeted AKT inhibition in this high-risk AML subset.