The tumour suppressor RBM5 activates the helicase DHX15 to regulate splicing

The tumour suppressor RBM5 enforces apoptotic alternative splicing by acting as a dual-action gatekeeper that physically stalls spliceosome progression while simultaneously activating the DHX15 helicase to unwind the branch helix.

The Gist

The Big Picture: The Cell's "Quality Control" Manager

Imagine your body's DNA as a massive, chaotic library of instruction manuals. To build a protein (the worker that does the actual job in your cells), the cell needs to copy a specific page from the manual. But these pages are messy: they have important sentences (exons) mixed with long, useless paragraphs of gibberish (introns).

Splicing is the process of cutting out the gibberish and gluing the good sentences together. The machine that does this is called the spliceosome. It's like a highly sophisticated, automated pair of scissors and glue.

However, sometimes the machine makes mistakes. It might glue the wrong sentences together, creating a broken protein. In cancer, these mistakes happen way too often, leading to cells that grow out of control.

This paper introduces a specific "Quality Control Manager" named RBM5. Its job is to stop the spliceosome from making bad cuts, acting as a tumor suppressor (a guard against cancer). The researchers finally figured out exactly how this manager works by taking a high-resolution 3D photo (using Cryo-EM) of the manager standing right next to the scissors.

---

The Cast of Characters

1. The Spliceosome (The Assembly Line): A giant, complex machine that builds proteins.
2. RBM5 (The Supervisor): A tumor suppressor protein. If RBM5 is missing or broken, the assembly line runs wild, creating dangerous proteins that cause cancer.
3. DHX15 (The Demolition Crew): A helicase enzyme. Think of it as a specialized tool that can unwind tangled wires or pull things apart.
4. SF3B1 (The Gatekeeper): A part of the spliceosome that holds the "gibberish" (intron) in place while it's being checked.
5. U2SURP (The Glue/Connector): A helper protein that ties everyone together.

---

The Discovery: How RBM5 Stops the Machine

The researchers discovered that RBM5 doesn't just stand around watching; it actively jams the machine and then calls in the demolition crew. They found two main ways it does this:

1. The "Traffic Jam" (Physical Blocking)

Imagine the spliceosome is a train moving along a track. To get to the next station (the next step of making a protein), the train needs to pick up a new carriage (called the tri-snRNP).

RBM5 attaches itself to the side of the train (specifically to the SF3B1 gatekeeper). It sticks out a part of its body (a "helix-loop-helix" shape) that physically blocks the new carriage from hooking up.

• The Result: The train stops. The assembly line is frozen. This prevents the machine from finishing the job on "bad" instructions that shouldn't be used.

2. The "Demolition Call" (Activating the Helicase)

While the train is stopped, RBM5 has another trick. It has a special handle (called a G-patch) that grabs the DHX15 demolition crew.

• The Action: RBM5 doesn't just hold DHX15; it activates it. It positions DHX15 right at the exit of the train, ready to grab the "gibberish" (the intron) and rip it out.
• The Result: If the instructions were bad, DHX15 pulls the whole thing apart, destroying the bad assembly before it can become a dangerous protein.

The Helper: There's a third character, U2SURP, which acts like a piece of duct tape or a connector cable. It helps hold RBM5, the gatekeeper (SF3B1), and the demolition crew (DHX15) all in the right spot so they can work together efficiently.

---

Why This Matters for Cancer

The paper shows that in many cancer genomes, the "hands" of this system are broken.

• Mutations: Cancer cells often have mutations in the genes for RBM5, SF3B1, or DHX15.
• The Consequence: These mutations are like cutting the brake lines on the train or breaking the handle on the demolition crew.
  • If the brake (RBM5) is broken, the train keeps moving even when it shouldn't.
  • If the demolition crew (DHX15) can't be called, the bad proteins get built.
  • If the connector (U2SURP) is broken, the team falls apart.

When this system fails, the cell starts producing the wrong versions of proteins. These broken proteins tell the cell to divide uncontrollably, leading to tumors.

The Takeaway

Think of this discovery as finally seeing the blueprints of a factory's safety inspector. We now know that RBM5 is a dual-action supervisor:

1. It physically blocks the machine from moving forward if the instructions look suspicious.
2. It summons a demolition crew (DHX15) to tear the bad instructions apart.

By understanding exactly how these parts fit together, scientists can now look for drugs that might fix broken versions of these proteins or target the specific spots where cancer cells try to bypass this safety check. It's a crucial step in understanding how to stop cancer at the very moment it tries to rewrite the cell's instruction manual.

Technical Summary

1. Problem Statement

Pre-mRNA splicing is a critical step in gene expression, frequently dysregulated in cancer. The RNA-binding protein RBM5 is a known tumor suppressor that regulates alternative splicing to control apoptosis and cell-cycle genes. However, its precise molecular mechanism within the spliceosome has remained elusive.

• Knowledge Gap: While RBM5 was known to bind U2 snRNP complexes and inhibit splicing at specific introns, it was unclear when and how it acts during spliceosome assembly, how it recruits the helicase DHX15 (yeast Prp43), and how this interaction leads to the stalling or disassembly of the spliceosome to enforce tumor-suppressive splicing programs.
• Limitation of Previous Studies: Most high-resolution spliceosome structures were derived from in vitro systems using synthetic RNAs, which may not reflect the heterogeneous, nascent RNA environment found in vivo on chromatin.

2. Methodology

The authors employed a multi-disciplinary approach combining in vivo biochemistry, functional assays, and high-resolution structural biology:

• In Vivo Spliceosome Capture: They isolated high-molecular-weight (HMW) chromatin-associated material from human HEK293 cells expressing FLAG-tagged RBM5. Spliceosomal complexes were affinity-purified using anti-FLAG beads, preserving complexes bound to nascent RNA.
• Cryo-Electron Microscopy (Cryo-EM): Single-particle cryo-EM was used to determine the structure of the RBM5-bound complex at 3.3 Å resolution. This allowed for de novo model building and the identification of flexible regions using AlphaFold predictions.
• Functional Assays:
  • RT-PCR Splicing Assays: Used to test the splicing regulation of specific exons (e.g., TRPT1, DPP7, NUMB) in cells expressing wild-type RBM5 versus a G-patch domain-deleted mutant.
  • Mutational Analysis: Mapping cancer-associated mutations from the COSMIC database onto the newly determined structural interfaces.
• Computational Modeling: AlphaFold 3 was utilized to predict structures of G-patch proteins and validate interactions with SF3B1.

3. Key Contributions and Results

A. Structural Architecture of the RBM5-Stalled Complex

The study reveals the structure of a precatalytic spliceosome (A complex) arrested by RBM5 immediately after U2 snRNP branchpoint recognition.

• Composition: The complex contains the U2 snRNP (SF3A and SF3B subcomplexes), the pre-mRNA branch site (BS), and the polypyrimidine tract (PPT). Crucially, it includes RBM5, the helicase DHX15, and the bridging factor U2SURP.
• RNA Recognition: Despite the heterogeneity of nascent introns, the structure shows a well-resolved U2–pre-mRNA duplex. The branchpoint adenosine is bulged, and the PPT is clamped within a cavity formed by SF3B1 HEAT repeats.

B. Dual-Action Mechanism of RBM5

RBM5 acts as a "gatekeeper" with two distinct functions:

1. Steric Blockade (The "Brake"):
   • RBM5 binds to the outer surface of the SF3B1 HEAT repeats via two interfaces: a helix-loop-helix (HLH) motif and a C-terminal helix.
   • This binding physically blocks the docking of the U4/U6.U5 tri-snRNP and the Prp8 protein.
   • Consequence: The spliceosome is prevented from progressing to the pre-B and Bact complexes, effectively stalling assembly.
2. Helicase Activation (The "Engine"):
   • RBM5's C-terminal G-patch domain binds and activates DHX15.
   • RBM5 positions DHX15 onto the pre-mRNA segment exiting the SF3B1 ring.
   • U2SURP acts as a scaffold, tethering RBM5, SF3B1, and DHX15 together, stabilizing this regulatory complex.

C. Mechanism of Splicing Repression

The authors propose a model where RBM5 arrests the spliceosome to allow DHX15 to function:

• Once stalled, the activated DHX15 translocates along the pre-mRNA.
• DHX15 disrupts the U2–branch site interaction and the PPT–SF3B1 contacts.
• This leads to the disassembly of the stalled complex, preventing the completion of splicing at suboptimal or regulated sites (exon skipping).
• Evidence: Deleting the G-patch domain of RBM5 reduced its ability to repress specific exons (e.g., TRPT1, DPP7), confirming that DHX15 activation is required for full repression.

D. Cancer Relevance

• Mutation Mapping: The study identified that numerous somatic mutations found in cancer genomes (COSMIC database) map precisely to the interfaces between RBM5, SF3B1, U2SURP, and DHX15.
• Implication: These mutations likely disrupt the formation of this regulatory checkpoint, leading to defective splicing of apoptotic isoforms and contributing to oncogenesis.

4. Significance

• Mechanistic Insight: This is the first high-resolution structure of an in vivo spliceosome stalled by a tumor suppressor. It resolves the long-standing question of how RBM5 controls splicing: by acting as a dual-action gatekeeper that physically blocks assembly progression while simultaneously recruiting an engine (DHX15) to dismantle the complex.
• Cancer Biology: It provides a structural basis for understanding how mutations in splicing factors (RBM5, SF3B1) drive cancer. It suggests that the loss of this "quality control" checkpoint allows aberrant splicing events to proceed, promoting cell survival and proliferation.
• Methodological Advance: The successful isolation and structural determination of a complex from chromatin-associated nascent RNA demonstrates a powerful approach to studying splicing regulation in its native physiological context, bridging the gap between in vitro biochemistry and in vivo function.

In summary, the paper defines RBM5 as a critical regulator that couples the physical stalling of the spliceosome with the activation of a helicase to enforce specific alternative splicing programs, a mechanism whose failure is directly linked to cancer pathogenesis.