High resolution interaction surface mapping by PRISMA reveals novel ARID1A interactions

This study utilizes the PRISMA technique combined with quantitative mass spectrometry to map high-resolution interaction surfaces of the cancer-associated protein ARID1A, successfully recapitulating known bindings while identifying novel interactors like TOX4, SIN3A, CDK2, and CCNA2, and revealing a new regulatory mechanism involving CDK2 phosphorylation that impacts cell proliferation.

The Gist

Imagine your cell's DNA as a massive, dusty library containing the instructions for building and running a human body. To read these instructions, the cell needs a special team of workers called the SWI/SNF complex. Think of this team as a group of librarians who can push heavy bookshelves (chromatin) aside to let the readers (RNA polymerase) access the books (genes).

One specific librarian, named ARID1A, is the team's "glue" or "scaffold." Without ARID1A, the whole team falls apart, and the library becomes a chaotic mess. This is why mutations in ARID1A are so common in cancer; the library shuts down, and the cell starts building things incorrectly.

However, ARID1A is a strange character. It's a giant protein, but half of it is like a floppy, shapeless noodle (called an "intrinsically disordered region"). Scientists have long known that this noodly part is crucial for holding hands with other proteins, but because it's so floppy and the handshakes are often weak or fleeting, it's been incredibly hard to figure out exactly who ARID1A holds hands with and where on its noodly body those hands connect.

The New Tool: The "Velcro Wall"

In this paper, the researchers used a clever new trick called PRISMA (Protein Interaction Screen on a peptide Matrix).

Imagine you want to find out which specific part of a long, floppy noodle grabs onto a specific type of Velcro. Instead of trying to grab the whole noodle at once, you cut the noodle into hundreds of tiny, overlapping strips and stick them all onto a giant wall. Then, you spray the wall with a mist of the "Velcro" proteins floating in the cell.

Wherever the proteins stick to a strip, you know exactly which tiny piece of the noodle they like. This is what PRISMA does: it maps the entire length of ARID1A, strip by strip, to see exactly which amino acids are the "sticky spots."

What They Found

Using this "Velcro wall" method, the researchers discovered several exciting things that traditional methods missed:

1. Re-confirmed the Knowns: They saw the team members (other parts of the SWI/SNF complex) sticking to the known "glue" spots on ARID1A, proving their method works.
2. The "Handshake" with SIN3A: They found that ARID1A directly shakes hands with a protein called SIN3A (a gene silencer) at a specific, tiny spot. This explains how ARID1A helps turn off certain genes.
3. New Friends: TOX4 and CDK2: They discovered two new friends ARID1A holds hands with:
   • TOX4: A protein involved in editing the library's catalog. Interestingly, ARID1A seems to hold hands with TOX4 only when ARID1A is "tagged" with a special sticker called ubiquitin (which usually marks proteins for recycling).
   • CDK2/Cyclin A: These are the cell's "clockwork" proteins that control the cell cycle (when a cell divides). ARID1A holds hands with them in a way that suggests they are coordinating the cell's division schedule.

The Big Discovery: The "Off Switch" at Position 363

The most exciting part of the story involves a specific spot on ARID1A called Serine 363.

Think of Serine 363 as a tiny dimmer switch on a lamp.

• When the switch is ON (Phosphorylated): The cell is ready to divide. ARID1A is busy organizing the "microtubules" (the scaffolding that pulls chromosomes apart during cell division).
• When the switch is OFF (Mutated): The researchers created a version of ARID1A where this switch was broken (it couldn't be turned on).

The Result: The cells with the broken switch couldn't divide properly. They stopped growing, and the "microtubule scaffolding" fell apart. It was like trying to build a house without a crane; the construction crew (the cell) just sat there confused.

Why This Matters

This paper is a game-changer for two reasons:

1. The Method: It proves that the "Velcro wall" (PRISMA) is a superpower for finding weak, fleeting connections that traditional methods miss. It's like finding a whisper in a noisy room by listening to the echo off a specific wall.
2. The Biology: It shows us that ARID1A isn't just a static glue; it's a dynamic manager that gets "switched on and off" by chemical signals (phosphorylation) to control how cells divide. If we can understand how to fix or manipulate this switch, we might be able to develop new treatments for cancers where ARID1A is broken.

In short: The researchers took a giant, floppy protein, sliced it into tiny pieces, and mapped out exactly who it hugs and when. They found that a tiny chemical switch on ARID1A is critical for keeping the cell's construction crew working, and they built a new tool that will help scientists find similar secrets in other proteins.

Technical Summary

1. Problem Statement

ARID1A is a defining scaffold subunit of the cBAF (canonical BAF) chromatin remodeling complex, which regulates gene expression, DNA repair, and cell fate. It is the most frequently mutated subunit in SWI/SNF complexes in human cancers (~6% of all cancer mutations) and is linked to neurodegenerative diseases.

Despite its importance, the molecular basis of many ARID1A interactions remains obscure.

• Structural Challenge: ARID1A is a large protein (2,285 amino acids) where the N-terminal half consists largely of Intrinsically Disordered Regions (IDRs). These regions contain Short Linear Motifs (SLiMs) and Post-Translational Modification (PTM) sites crucial for interactions but are notoriously difficult to study using traditional methods.
• Methodological Limitation: Traditional Affinity Purification Mass Spectrometry (AP-MS) often fails to detect weak, transient, or low-abundance interactions mediated by IDRs and SLiMs because these complexes are often lost during washing steps or require specific cellular contexts/PTMs not present in standard lysates.
• Knowledge Gap: While hundreds of ARID1A interactors are reported, their precise binding interfaces (amino acid residues and motifs) are largely unmapped, hindering the understanding of regulatory mechanisms.

2. Methodology

The authors employed a high-resolution peptide interactomics strategy called PRISMA (PRotein Interaction Screen on a peptide MAtrix) combined with quantitative mass spectrometry.

• PRISMA Assay Design:
  • A tiled peptide array was created covering the entire human ARID1A sequence (2,285 aa) using 20-mer peptides with a 10-amino acid overlap, resulting in 228 peptide spots.
  • Screening: The array was incubated with nuclear extracts from RPE1 cells (treated with nuclease to eliminate DNA-mediated interactions).
  • Detection: Bound proteins were identified and quantified using shotgun mass spectrometry (LC-MS/MS).
• Data Filtering:
  • Proteins identified by at least two peptides were quantified.
  • A strict filter was applied to remove "promiscuous" background binders (signals below the 75th percentile) and to require consecutive binding across at least two overlapping peptides to confirm specific interaction sites.
• Validation & Functional Characterization:
  • Structural Validation: PRISMA data was compared against cryo-EM structures (6LTJ) and Cross-linking Mass Spectrometry (XL-MS) data.
  • Biochemical Validation: Co-immunoprecipitation (Co-IP), Far-Western blotting on peptide arrays, and in vitro pull-down assays using recombinant proteins.
  • Functional Analysis:
    • Generation of ARID1A knockout (KO) RPE1 cell lines reconstituted with Wild Type (WT), Phospho-dead (S363A), or Phospho-mimetic (S363D) ARID1A.
    • Omics: RNA-seq and quantitative proteomics (TMT-MS) to assess transcriptional and protein-level changes.
    • Phenotyping: Live-cell imaging (Incucyte) to monitor cell proliferation and mitotic defects.

3. Key Contributions & Results

A. High-Resolution Interaction Map

The study generated a comprehensive map of ARID1A interaction surfaces, identifying 1,454 high-confidence interactors.

• Validation of Known Interactions: The assay successfully recapitulated binding of BAF complex subunits (e.g., SMARCA4, SMARCB1) to the conserved C-terminal ASD1/ASD2 domains and the LXXLL motifs. It also confirmed known SLiM-mediated interactions, such as YAP1 binding to a PPXY motif in the N-terminal region.
• Novel Interactors Identified:
  1. SIN3A (Transcriptional Repressor): PRISMA identified a direct interaction between ARID1A and SIN3A via a predicted Sin3-interacting domain (SID) SLiM. This was validated by showing ARID1A is required for SIN3A chromatin recruitment and that the ARID1A C-terminus directly binds the PAH domain of SIN3A.
  2. TOX4 (PP1 Regulatory Subunit): A novel interactor involved in transcriptional elongation. TOX4 was found to bind ARID1A peptides containing kinase consensus sites. Crucially, TOX4 was shown to bind poly-ubiquitinated forms of ARID1A, explaining why this interaction was missed in previous AP-MS studies.
  3. CDK2 and Cyclin A2: The screen identified weak but specific binding of CDK2/Cyclin A2 to disordered regions of ARID1A (specifically residues 1391–1430 and 531–600), overlapping with a CDK2 consensus phosphorylation site (Ser363).

B. Discovery of a Novel Regulatory Mechanism: Ser363 Phosphorylation

The study focused on Serine 363 (Ser363), a CDK2 consensus site identified in the PRISMA data.

• Cell Cycle Regulation: Phosphorylation of Ser363 is cell-cycle dependent, peaking in G1/S and decreasing in G2/M.
• Mutant Analysis:
  • S363A (Phospho-dead): Cells expressing this mutant showed a mild proliferation defect.
  • S363D (Phospho-mimetic): The protein was unstable and degraded rapidly, preventing functional analysis.
• Transcriptional & Proteomic Impact: RNA-seq and proteomics of S363A cells revealed a specific downregulation of genes and proteins involved in microtubule organization, mitotic spindle assembly, and kinesin complexes. This effect was observed at both the mRNA and protein levels, a rare correlation for SWI/SNF subunits.
• Phenotypic Consequence: Live-cell imaging demonstrated that cells expressing the S363A mutant exhibited severe mitotic defects and failed to undergo cell division, confirming that Ser363 phosphorylation is essential for normal mitosis.

4. Significance

• Methodological Advancement: The study demonstrates that PRISMA is superior to traditional AP-MS for mapping interactions involving intrinsically disordered regions and weak/SLiM-mediated interactions. It provides a resource for the community to explore sequence-to-function relationships at high resolution.
• Mechanistic Insight: It uncovers a direct, phosphorylation-dependent regulatory axis for ARID1A. The dynamic phosphorylation of Ser363 by CDK2/Cyclin A2 acts as a switch to regulate the transcription of microtubule-related genes, linking chromatin remodeling directly to cell division machinery.
• Therapeutic Implications:
  • The identification of TOX4 as an interactor with ubiquitinated ARID1A suggests new avenues for targeting ARID1A-deficient cancers (e.g., synthetic lethality with PLK1 inhibition, which is known to be lethal in ARID1A loss).
  • The discovery of the SIN3A-ARID1A direct link clarifies how ARID1A recruits repressive complexes to chromatin.
  • Understanding the specific role of ARID1A in microtubule regulation via Ser363 could inform strategies to target cell cycle defects in ARID1A-mutated tumors.

In summary, this paper transforms the understanding of ARID1A from a static scaffold to a dynamically regulated hub, utilizing peptide array technology to reveal previously invisible interaction networks and a critical phosphorylation-dependent mechanism governing cell division.