Subcellular transcriptome sequencing with single cell APEX-seq identifies regulators of cell-cell interactions

This paper introduces single-cell APEX-seq, a proximity labeling-based method that maps subcellular transcriptomes at single-cell resolution to reveal interaction-dependent cell states and identify regulators like CTSW that enhance CAR T cell persistence and tumor killing, which are often missed by conventional scRNA-seq.

The Gist

Imagine you are trying to understand a bustling city. For years, scientists have been able to take a census of every person in that city (Single-Cell RNA Sequencing, or scRNA-seq). They can tell you who lives there, what their job is, and how many people are in each neighborhood.

But there's a problem: This census only counts people standing in the middle of the room. It misses everyone who is at the front door, the delivery trucks outside, or the secret messages being passed through the windows.

In a cell, the "front door" is the Endoplasmic Reticulum (ER). This is where the cell builds the proteins it needs to talk to other cells (like sending letters or waving hello). If you only look at the "middle of the room" (the total cell contents), you miss the most important conversations happening at the border.

This paper introduces a new, super-powered tool called scAPEX-seq that lets scientists peek specifically at that "front door" of thousands of individual cells at once.

Here is the story of how they did it and what they found, explained with some simple analogies.

1. The Problem: The "Blind Spot" in Cell Communication

Cells talk to each other using proteins on their surface. Think of these proteins as handshakes, flags, or megaphones.

• Old Method (scRNA-seq): Imagine trying to guess what a person is shouting by listening to the noise inside their house. You might hear some shouting, but you miss the specific words being yelled at the neighbors because the walls dampen the sound.
• The Issue: Many of these "shouts" (proteins) are made in a specific factory inside the cell (the ER). Standard methods often miss them because they are rare or get lost in the noise of the rest of the cell's contents.

2. The Solution: The "Magnetic Tag" (scAPEX-seq)

The researchers invented a way to tag only the messages being prepared at the "front door" (the ER) so they can be found later.

• The Magic Paint: They use a tiny, invisible "paint" (a chemical probe) that only sticks to RNA molecules that are currently sitting at the ER.
• The Magnet: Once the RNA is painted, they use a magnetic bead to grab only those painted molecules.
• The Upgrade (APEX-seq 2): The old version of this paint was sticky and hard to use (like trying to catch a fish with a net made of cheese). The new version, APEX-seq 2, uses a better "paint" (called Phenol-Azide) that gets into cells easily and sticks better. It's like switching from a cheese net to a high-tech, super-strong magnet. This allowed them to catch way more messages with fewer cells.

3. The Experiment: Watching Cells Fight and Talk

The team used this new tool to watch two different "battles" between cells.

Battle A: The Tumor vs. The Macrophage (The Police and the Criminal)

They mixed cancer cells with immune cells (macrophages).

• What happened: When the immune cells got close to the cancer, they started changing their behavior.
• The Discovery: The old method (looking at the whole cell) saw a blurry picture. The new method (scAPEX-seq) saw clear, distinct groups. It revealed that the immune cells were sending out specific "danger signals" and "handshakes" that the old method completely missed. It was like realizing the police weren't just standing there; they were actively negotiating with the criminals in ways no one knew about.

Battle B: The CAR T-Cell vs. The Tumor (The Super-Soldier)

This was the big one. They looked at CAR T-cells, which are genetically engineered immune cells designed to hunt cancer.

• The Mystery: Sometimes these super-soldiers work great, but other times they get tired (exhausted) and give up after a few days. Scientists didn't know exactly why or when they decided to quit.
• The Discovery: By looking at the "front door" messages, they found two secret groups of CAR T-cells:
  1. The "Burnout" Group: These cells were already tired and giving up.
  2. The "Super-Soldier" Group: A hidden group that looked different. They were sending out messages that kept them strong, young, and ready to fight for a long time.

4. The Big Breakthrough: Finding the "Stamina Gene" (CTSW)

The researchers found a specific gene in the "Super-Soldier" group called CTSW.

• The Analogy: Think of CTSW as a fitness coach inside the cell.
• What it does: When the researchers turned up the volume on this coach (overexpression), the CAR T-cells became stronger. They didn't get tired as easily, they multiplied faster, and they kept killing cancer cells for weeks instead of days.
• The Result: When they turned the coach off (knockout), the cells gave up quickly.
• Real World Proof: They checked data from real cancer patients and found that people with high levels of CTSW in their immune cells survived longer and responded better to immunotherapy.

5. Why This Matters

This paper is like upgrading from a black-and-white photo to a 4K video with sound.

• Before: We could see the cells, but we missed the most important conversations happening at the edges.
• Now: With scAPEX-seq, we can see exactly how cells prepare their "weapons" and "signals" to talk to neighbors.
• The Future: This helps scientists design better cancer treatments. Instead of just guessing which drugs might work, they can now engineer immune cells (like the CAR T-cells) to be tougher, smarter, and longer-lasting by tweaking the specific genes (like CTSW) that control their stamina.

In a nutshell: The scientists built a better microscope that doesn't just look at the whole cell, but zooms in on the "front door" where the action happens. This helped them find a secret "stamina switch" that could make cancer-fighting immune cells much more effective.

Technical Summary

1. Problem Statement

Single-cell RNA sequencing (scRNA-seq) has revolutionized the understanding of cellular heterogeneity and tissue architecture. However, it has a critical limitation: it measures total cellular RNA abundance, ignoring subcellular RNA localization.

• The Gap: RNA localization is crucial for function (e.g., splicing, translation, protein secretion). Transcripts encoding cell surface receptors and secreted proteins (ligands) are often localized to the Endoplasmic Reticulum (ER) membrane. In conventional scRNA-seq, these transcripts are often low-abundance and easily missed ("dropout"), making it difficult to infer Cell-Cell Interactions (CCIs).
• Existing Limitations: Current subcellular methods rely on imaging (MERFISH, seqFISH), which are limited by throughput, spatial resolution (~1 µm), and the need for pre-selected probe sets. Bulk subcellular fractionation cannot be applied to single cells. There is a need for an unbiased, high-throughput, sequencing-based method to profile subcellular transcriptomes at single-cell resolution.

2. Methodology: Development of scAPEX-seq

The authors developed single-cell APEX-seq (scAPEX-seq), a proximity labeling-based method integrated with droplet microfluidics.

A. Optimization of Bulk APEX-seq (APEX-seq2)

Before moving to single cells, the authors improved the original bulk APEX-seq protocol to increase sensitivity and RNA recovery:

• Probe Replacement: Replaced the original Biotin-Phenol (BP) probe with Phenol-Azide (PA). PA has better cell permeability and allows for copper-free Click chemistry (using DBCO-biotin), which is less damaging to RNA.
• Workflow Simplification: Eliminated RNA elution and purification steps. Instead, RNA-cDNA hybrids were amplified directly on streptavidin beads.
• Results: These changes improved RNA recovery by 10-fold (from ~0.1% to ~1%) and increased signal-to-noise ratios, allowing analysis with 1/10th the input material (4 µg total RNA).

B. Integration with Single-Cell Workflow (scAPEX-seq)

To adapt this for single cells using the 10x Genomics platform, the authors solved the challenge of enriching biotinylated transcripts without mixing cell barcodes:

1. Labeling: Cells expressing APEX2 targeted to a specific compartment (e.g., ER membrane) are labeled with PA and H₂O₂ for 1 minute.
2. Barcoding (Pre-Enrichment): Cells are encapsulated in Gel Beads-in-Emulsion (GEMs) for a single cycle of reverse transcription (RT). This attaches cell-specific barcodes to full-length RNA-cDNA hybrids before any PCR amplification.
3. Pooling & Enrichment: Emulsions are broken, and the RNA-cDNA hybrids are pooled. The azide groups are converted to biotin via Click chemistry, followed by streptavidin bead enrichment.
4. Dual Sequencing:
   • scAPEX-seq: The enriched fraction (biotinylated, subcellular-specific transcripts) is amplified and sequenced.
   • Supernatant-seq: The unenriched supernatant (representing the whole transcriptome minus the enriched fraction) is sequenced as a control.

3. Key Contributions

• Technical Innovation: The first method to map subcellular transcriptomes (specifically ER-associated) at single-cell resolution with high throughput.
• Methodological Refinement: The creation of "APEX-seq2" significantly improved RNA recovery and sensitivity, making single-cell applications feasible.
• Biological Insight: Demonstrated that subcellular RNA organization reveals cell states and interaction regulators that are invisible to conventional scRNA-seq.

4. Key Results

A. Validation of scAPEX-seq Specificity and Sensitivity

• Applied to co-cultures of tumor cells (MC38) and macrophages (RAW264.7).
• Specificity: scAPEX-seq showed strong enrichment (Log2FC > 0.5) for secretory/ER-localized genes compared to the supernatant fraction.
• Sensitivity: Detected 70% of true-positive ER transcripts.
• Resolution of Cell States: While conventional scRNA-seq and Supernatant-seq failed to distinguish between co-cultured (CC) and non-co-cultured (NCC) cells, scAPEX-seq resolved distinct interaction-dependent cell states. It identified hundreds of differentially expressed genes (DEGs) and unique ligand-receptor interactions (e.g., Ccl2-Ccr2, PD-L1-PD1) missed by standard methods.

B. Short-Term CAR T Cell Activation (2-hour Coculture)

• Experimental Setup: HER2+ tumor cells co-cultured with anti-HER2 CAR T cells for 2 hours.
• Discovery: scAPEX-seq resolved four distinct CAR T cell subclusters (Major Effector, Major Memory, Secondary Effector-like, Secondary Memory-like), whereas Supernatant-seq only saw two broad clusters.
• Key Finding: The "Secondary Memory" cluster showed unique upregulation of interferon signaling and survival markers. Crucially, scAPEX-seq detected early upregulation of NT5E (CD73) and its localization to the ER, indicating rapid engagement of immunosuppressive adenosine pathways.
• Validation: CRISPR knockout of NT5E in CAR T cells reduced exhaustion markers (PD1, TIM3) and improved tumor killing in repeated stimulation assays.

C. Long-Term CAR T Cell Persistence (21-day Coculture)

• Experimental Setup: CAR T cells subjected to 7 cycles of repeated antigen stimulation.
• Discovery: scAPEX-seq identified a unique "Late Effector" cluster that was absent in Supernatant-seq. This cluster exhibited high cytotoxicity (IFNG, GZMB) and low exhaustion markers (TOX, LAG3).
• Key Finding: This persistent population was characterized by the upregulation of CTSW (Cathepsin W).
• Mechanism & Validation:
  • CTSW overexpression promoted a stem-like memory phenotype (CD62L+CD45RA+) and sustained proliferation.
  • CTSW overexpression led to reduced CD25 expression, preventing premature exhaustion.
  • CTSW knockout resulted in rapid tumor escape and T cell failure.
  • Clinical data analysis (CIDE database) confirmed that high CTSW expression correlates with improved survival in melanoma and lung cancer patients treated with immunotherapy.

5. Significance

• Overcoming scRNA-seq Limitations: scAPEX-seq bypasses the "dropout" problem for low-abundance, surface, and secreted proteins by enriching for their transcripts at the ER membrane. It captures changes in RNA localization and accessibility, not just abundance.
• Decoding Cell-Cell Interactions: By focusing on the "secretome" at the single-cell level, the method provides a more accurate map of ligand-receptor interactions, revealing regulatory networks that drive immune cell fate (e.g., exhaustion vs. persistence).
• Therapeutic Impact: The identification of CTSW as a regulator of CAR T cell persistence offers a new target for engineering more durable and effective immunotherapies for solid tumors.
• Scalability: The method is compatible with standard droplet-based workflows, making it a scalable tool for studying complex tissue microenvironments, developmental biology, and disease mechanisms where RNA localization is key.