Engineered subtilisin protease degrades active KRAS in cancer cells, leading to differential cell targeting

An engineered subtilisin protease (RASp) specifically targets and degrades active KRAS G12C by cleaving its switch II region, thereby disrupting downstream MEK-ERK signaling and inducing selective cell death in KRAS-dependent cancer cells while sparing normal cells.

The Gist

The Big Idea: A "Smart Bomb" for Cancer Cells

Imagine the human body as a bustling city. In this city, there are traffic lights called RAS proteins. Normally, these lights work perfectly: they turn green to let cells grow and divide when needed, and they turn red to stop when things are done.

However, in many cancers (like pancreatic cancer), a mutation breaks these traffic lights. They get stuck on Green forever. This causes cells to grow out of control, leading to a traffic jam that becomes a tumor.

For decades, scientists have tried to fix these broken lights with "brakes" (drugs that try to stop the signal). But the lights are so sticky and complex that it's been incredibly hard to build a brake that works without stopping the whole city.

This paper introduces a new strategy: Instead of trying to fix the broken light, why not just remove the broken light entirely?

The researchers created a "molecular scissors" (an engineered enzyme) that acts like a smart bomb. It doesn't just cut anything; it specifically hunts down the broken, "stuck-on" traffic lights and shreds them, leaving the healthy ones alone.

---

How the "Smart Scissors" Work

The scientists took a natural enzyme (a protein cutter) from a bacterium and re-engineered it. They gave it a very specific job: Find the "Switch II" region of the RAS protein.

Here is the clever part:

• Healthy RAS: The "Switch II" region is folded up tight, like a secret compartment. The scissors can't see it.
• Cancer RAS: Because the cancer RAS is stuck in the "ON" position, this secret compartment pops open. It's like a door swinging wide open.

The scissors are designed to only fit through that open door. So, they ignore the healthy, folded-up RAS and only attack the dangerous, open, cancer-causing RAS.

The Two Types of Scissors

The team built two versions of these scissors to see which one worked best:

1. The "Nitrite" Scissors (RASpN): These work using a chemical (nitrite) that is naturally found in high amounts inside cancer cells. It's like a key that fits perfectly into a lock already present in the enemy's base.
2. The "Imidazole" Scissors (RASpI): These require a different chemical (imidazole) that isn't naturally common in cells. You'd have to bring it in from the outside.

The Result: The "Nitrite" scissors were much faster and more powerful. They were like a speedster compared to the other version.

The Experiment: Testing the Scissors

The researchers tested this in two different scenarios:

1. The "Fake City" (Lab Cells):
They put the scissors into normal human cells (HEK 293T).

• What happened: The scissors cut the RAS protein, but the cells didn't die.
• Why? These normal cells have backup plans. Even if you remove one traffic light, they can still function. They are resilient.

2. The "Cancer City" (Pancreatic Cancer Cells):
They put the scissors into cancer cells (MIA PaCa-2) that rely entirely on the broken RAS light to survive.

• What happened: Within 24 hours, the cancer cells were wiped out.
• Why? These cancer cells are like a house built on a single, shaky pillar. When the scissors cut that one pillar (the active RAS), the whole house collapses. The healthy cells were fine, but the cancer cells died because they had no other way to survive.

The "Goldilocks" Discovery (Concentration Matters)

There was a fascinating twist in the lab tests.

• When there were too many scissors: They were so aggressive they cut everything, even the healthy, folded-up RAS.
• When there were just a few scissors: They became very picky. They waited for the "open door" (the active cancer RAS) to appear before cutting.

This is great news for therapy. In a real patient, you wouldn't flood the body with scissors. You'd use a small, controlled amount. This ensures the scissors only hunt the cancer cells and leave the healthy ones completely alone.

The Bottom Line

This paper proves that we can build a biological tool that:

1. Identifies cancer cells by their "open" broken switches.
2. Destroys the specific protein driving the cancer.
3. Spares healthy cells that don't need that specific protein to survive.

It's a shift from trying to "turn off" a broken switch to simply snipping the wire that powers the cancer, leaving the rest of the body's electrical grid untouched. This opens the door for a new kind of cancer treatment that is highly targeted and potentially much less toxic than current chemotherapy.

Technical Summary

1. Problem Statement

• The Challenge of RAS: The RAS family of small GTPases (KRAS, NRAS, HRAS) is the most frequently mutated gene family in human cancers. Mutations lock RAS in an active, GTP-bound state, driving uncontrolled cell proliferation via the MAPK and mTOR signaling pathways.
• Therapeutic Limitations: Historically considered "undruggable," current FDA-approved KRAS inhibitors (e.g., for G12C) are not specific to the active, GTP-bound state and can impact healthy RAS signaling. Furthermore, they do not degrade the protein, potentially allowing for resistance mechanisms.
• The Opportunity: The "Switch I" and "Switch II" regions of RAS undergo an order-to-disorder transition upon binding GTP. This conformational change exposes cryptic cleavage sites (specifically the QEEYSAM sequence in Switch II) that are hidden in the inactive, GDP-bound state. This suggests a potential strategy for selectively degrading active RAS using engineered proteases.

2. Methodology

The study utilized a multi-tiered approach combining in vitro biochemistry, cell lysate assays, and live-cell models to validate an engineered protease.

• Protease Engineering: The authors utilized RASProtease (RASp), an engineered variant of Bacillus subtilis subtilisin designed to target the Switch II region of RAS. Two variants were tested:
  • RASpI: Requires imidazole as a cofactor.
  • RASpN: Requires nitrite as a cofactor (endogenous to cancer cells).
  • RASp:* An inactive control with a catalytic site mutation (S221A).
• In Vitro Kinetics:
  • Fluorescent Peptide Assays: Used a QEEYSAM-AMC peptide substrate to measure cleavage rates ($k_{obs}$) via stopped-flow spectroscopy.
  • Protein Degradation: Incubated purified WT and G12C mutant KRAS (in both GDP and GMPPNP/GTP-bound states) with RASp variants. Reactions were analyzed via SDS-PAGE and NMR spectroscopy to quantify degradation kinetics and substrate preference.
• Cellular Models:
  • HEK 293T: Used as a control model expressing wild-type (WT) KRAS. Transfected with eGFP-KRAS to visualize degradation.
  • MIA PaCa-2 (MP2): A pancreatic cancer cell line harboring homozygous KRAS G12C. This model is known to be KRAS-dependent for survival.
• Delivery Systems:
  • Transient Transfection: Used for initial expression studies in 293T and MP2 cells.
  • Lentiviral Transduction: Used to generate stable cell lines expressing RASpN, allowing for long-term induction via doxycycline.
• Readouts: Western blotting (for KRAS, MEK, ERK, AKT levels), Immunofluorescence microscopy (for cell viability and FLAG-tagged RASp expression), and cell counting.

3. Key Contributions

• Identification of RASpN as the Superior Variant: The study demonstrated that the nitrite-dependent variant (RASpN) is approximately 100-fold more active ($k_{obs} \approx 1.57 s^{-1}$) than the imidazole-dependent variant (RASpI, $k_{obs} \approx 0.012 s^{-1}$).
• Discovery of Concentration-Dependent Selectivity: A critical finding was that RASpN exhibits differential substrate preference based on concentration:
  • At stoichiometric (high) concentrations, RASpN cleaves both GDP-bound (inactive) and GTP-bound (active) KRAS, with a slight preference for the inactive form.
  • At limiting (low, therapeutic-relevant) concentrations (1:10,000 ratio), the preference switches to the active (GTP-bound) KRAS. This is attributed to the inherent conformational dynamics of Switch II, which are more accessible in the active state when the protease is scarce.
• Validation in Cancer-Relevant Models: The study successfully moved beyond peptide substrates to demonstrate degradation of endogenous KRAS G12C in a relevant pancreatic cancer cell line (MIA PaCa-2).

4. Key Results

• Biochemical Degradation:
  • RASpN rapidly degraded both WT and G12C KRAS in vitro.
  • NMR analysis confirmed that at low protease concentrations, GTP-bound KRAS is degraded significantly faster ($k_{obs} \approx 0.077 hr^{-1}$) than GDP-bound KRAS ($k_{obs} \approx 0.025 hr^{-1}$).
• Downstream Signaling Inhibition:
  • In MIA PaCa-2 cells, induction of RASpN led to the near-total depletion of KRAS.
  • This depletion resulted in a strong reduction of MEK and phosphorylated-MEK (pMEK), and a near-total ablation of AKT (mTOR pathway). ERK and pERK levels were also modestly depleted.
• Differential Cell Killing (Therapeutic Window):
  • MIA PaCa-2 (KRAS G12C): Induction of RASpN caused extensive cell death, with nearly 100% of cells dying within 24 hours.
  • HEK 293T (WT KRAS): Under the same conditions, >50% of cells remained viable after 24 hours.
  • Conclusion: The engineered protease selectively kills cells dependent on oncogenic, active RAS while sparing cells with wild-type RAS cycling between active and inactive states.

5. Significance

• Proof of Concept for "Active-Only" Degradation: This work validates the hypothesis that engineered proteases can exploit the conformational dynamics of RAS to selectively degrade the oncogenic, active form, a feat difficult for small molecules to achieve.
• Therapeutic Potential: The ability to induce rapid cell death in KRAS-dependent cancer cells (MIA PaCa-2) while sparing normal cells (293T) suggests a high therapeutic index. This approach offers a potential "pan-RAS" therapeutic strategy that could overcome resistance to current G12C-specific inhibitors.
• Novel Regulatory Mechanism: The use of endogenous nitrite as a cofactor (RASpN) allows for activation within the tumor microenvironment without requiring exogenous drug delivery for cofactor activation, simplifying the therapeutic delivery profile.
• Platform Technology: The study demonstrates a framework for re-engineering proteases to target other "undruggable" proteins by leveraging order-to-disorder transitions in their active states.

In summary, this paper presents a robust, biochemically validated strategy using an engineered subtilisin (RASpN) to selectively degrade active KRAS, effectively shutting down oncogenic signaling and inducing apoptosis in cancer cells while sparing normal cells.