Fixative eXchange (FX)-seq: Scalable Single-nucleus RNA Sequencing Analysis of PFA-fixed or FFPE Tissue

The paper introduces Fixative-eXchange (FX)-seq, a scalable single-nucleus RNA sequencing method that utilizes organocatalysts and platinum-based crosslinking to overcome low reverse transcription yields, thereby enabling high-quality transcriptomic analysis of heavily PFA-fixed and FFPE clinical tissue samples.

The Gist

Imagine you have a massive, dusty library of books that tells the story of human health and disease. These books are the cells in your body. For decades, scientists have been able to read the "fresh" books from living patients, but the vast majority of medical history is locked away in FFPE archives—tissues that have been preserved in formaldehyde and paraffin wax (like a time capsule) for years or even decades.

The problem? Trying to read these old, preserved books is like trying to read a page where the ink has been glued to the paper. The preservation chemicals (formaldehyde) act like super-strong glue, cross-linking the genetic instructions (RNA) so tightly that the reading machines (enzymes) can't get through. Previous attempts to unstick them were like using a sledgehammer: they either broke the pages (degraded the RNA) or didn't work well enough to read the whole story.

Enter FX-seq (Fixative eXchange sequencing), a new method developed by the team at Yonsei University. Think of FX-seq as a gentle, high-tech restoration workshop for these old books.

Here is how it works, broken down into simple steps:

1. The "Glue" Problem

When tissue is preserved, formaldehyde creates "methylol adducts." Imagine these as sticky tape stuck all over the letters of your DNA. When scientists try to copy the text (a process called Reverse Transcription), the enzymes get stuck on the tape and stop working.

2. The "Solvent" Solution (The Organocatalyst)

The researchers found a special chemical "solvent" (an organocatalyst) that acts like a magic eraser. Instead of using harsh heat that might burn the book, this solvent gently dissolves the sticky tape under mild conditions, freeing up the letters so they can be read again.

3. The "Safety Net" (The Platinum Crosslinker)

There was a catch: when they tried to dissolve the tape, some of the pages started to fall out of the book (RNA leakage). To fix this, they added a second step. They used a special "safety net" made of platinum molecules.

• The Analogy: Imagine the RNA is a loose thread in a sweater. When you wash it (remove the glue), the thread might unravel and fall out. The platinum crosslinker acts like a tiny, invisible safety pin that holds the thread in place without changing the pattern of the sweater. It secures the RNA so nothing is lost during the cleaning process.

4. The "Bodyguard" (PVSA)

During the process, there's a risk of "gremlins" called RNases (enzymes that eat RNA) attacking the delicate pages. Usually, scientists use expensive protein-based bodyguards to stop them. But FX-seq uses PVSA, a cheap, heat-stable chemical that acts like a shield. It blocks the gremlins without sticking to the pages itself, ensuring the RNA stays pristine.

Why This is a Big Deal

Before FX-seq, scientists could only study fresh tissue, which is hard to get and doesn't last. With FX-seq, they can now unlock the entire global archive of medical history.

• The "Time Travel" Effect: They can take a tissue sample from a cancer patient diagnosed 10 years ago, turn it into a single-cell map, and see exactly how the tumor evolved.
• The "Detective" Work: In the paper, they used this on a patient with a tumor that had stopped responding to medication (imatinib). By reading the "glued" books from the old tissue, they discovered two hidden paths the cancer cells took to become resistant. One path involved changing the "spelling" of the genes (methylation), and the other involved waking up "ghost genes" (germline genes) that usually sleep in the body. This gives doctors new clues on how to treat resistant cancers.
• The "Map" Maker: They can even take a thin slice of tissue that a pathologist has already stained with ink (H&E stain) to look at under a microscope, scrape off the cells, and turn them into a 3D map of the tumor's neighborhood, showing exactly which cells are friends and which are enemies.

The Bottom Line

FX-seq is like giving scientists a universal translator for the world's largest medical library. It allows them to read the stories hidden in old, preserved tissues with the same clarity as fresh ones. This means we can finally learn from decades of past medical cases to cure diseases today, without needing to perform new surgeries just to get a "fresh" sample. It turns the "dead" archives of the past into the "living" data of the future.

Technical Summary

1. The Problem

Single-nucleus RNA sequencing (snRNA-seq) is a powerful tool for understanding cellular heterogeneity, but its application to clinical archives is severely limited.

• The Barrier: Formalin-fixed, paraffin-embedded (FFPE) tissues and heavily paraformaldehyde (PFA)-fixed samples constitute billions of archived clinical specimens. However, formalin fixation creates crosslinks (methylol adducts) between nucleic acids and proteins, and modifies nucleobases.
• Consequences: These modifications inhibit Watson-Crick base pairing and reverse transcription (RT), leading to:
  • Low cDNA synthesis yields.
  • Premature termination of RT enzymes.
  • RNA leakage from nuclei during processing.
  • Poor data quality (low Unique Molecular Identifiers [UMIs] and gene counts) compared to fresh samples.
• Limitations of Existing Methods: Previous attempts (snPATHO-seq, snFFPE-seq, snRandom-seq) suffer from transcriptome bias (probe panels), inefficiency (random primers causing rRNA dominance), or limited scalability and reliance on harsh heat/protease treatments that degrade RNA.

2. Methodology: FX-seq

The authors developed Fixative eXchange (FX)-seq, a scalable snRNA-seq protocol designed to recover high-quality transcriptomes from heavily fixed tissues. The method integrates chemical optimization with combinatorial barcoding.

Key Technical Innovations:

1. Organocatalytic Decrosslinking:
   • Instead of harsh heat, FX-seq employs a specific organocatalyst (Cat. 2) to remove PFA adducts under mild conditions (55°C).
   • Cat. 2 was screened and found to have superior reaction kinetics for removing formaldehyde adducts compared to previously reported catalysts, significantly improving RT yield.
2. Regiospecific Pt(II) Crosslinking (RNA Retention):
   • To prevent RNA leakage during the decrosslinking and isolation steps, the protocol introduces an additional crosslinking step after PFA removal but before RT.
   • A novel Pt(II)-based crosslinker (coupling cisplatin derivatives with a PEG linker) is used. It specifically targets Guanine-N7 residues.
   • This crosslinking is regioselective and does not interfere with Watson-Crick base pairing, thereby stabilizing RNA within the nucleus without hindering enzymatic processes.
3. Chemical RNase Inhibition (PVSA):
   • To combat RNA degradation caused by endogenous RNases during the extensive handling and heating steps, the authors replaced expensive protein-based RNase inhibitors with Polyvinyl Sulfonic Acid (PVSA).
   • PVSA acts as a non-reactive competitor that blocks RNase active sites. It is thermostable, cost-effective, and does not leave adducts on RNA that could interfere with downstream enzymes.
4. Scalable Combinatorial Barcoding:
   • The protocol utilizes an updated sci-RNA-seq3 workflow for combinatorial cell barcoding, allowing for the processing of hundreds of thousands of nuclei in a single experiment.

3. Key Contributions

• Chemical Optimization: The discovery and validation of the organocatalyst (Cat. 2) and the Pt(II) crosslinker as a dual strategy to simultaneously reverse fixation damage and prevent RNA loss.
• Protocol Standardization: A robust, scalable workflow that successfully isolates nuclei from heavily fixed tissues (PFA and FFPE) without the need for fresh tissue.
• Spatial Integration: A method to link snRNA-seq data with pathological annotations by physically separating tumor and non-tumor regions from H&E-stained FFPE sections prior to sequencing.

4. Results

The authors validated FX-seq across multiple biological contexts:

• Mouse Brain (PFA-fixed):
  • FX-seq recovered significantly more UMIs and gene counts compared to untreated or Tris-buffered controls.
  • It enabled accurate classification of distinct neuronal subtypes (e.g., distinguishing Medium Spiny Neurons from inhibitory interneurons) that were indistinguishable in control samples.
  • Performance was comparable to snRNA-seq of fresh, lightly fixed nuclei, despite the heavy fixation.
• Mouse Brain (FFPE and Sections):
  • Successfully applied to FFPE blocks and thin sections (4 µm and 10 µm), including H&E-stained sections.
  • While sectioning reduced total UMIs (due to physical cutting), the transcriptomic profiles remained consistent with block-level data, allowing for co-clustering of cell types across different sample preparations.
• Human Clinical Specimens:
  • Gastrointestinal Stromal Tumor (GIST): Analyzed ~172,000 nuclei from an imatinib-resistant GIST FFPE block. FX-seq revealed rare immune populations and distinct tumor sub-clusters. Trajectory analysis identified two paths of transcriptional shift leading to resistance, involving DNA methylation changes and germline-specific gene expression.
  • Colorectal Cancer (CRC): Applied to archived FFPE slides with pathologist annotations. The method successfully separated tumor vs. non-tumor regions, identified epithelial progenitor vs. tumor cells, and inferred Copy Number Variations (CNVs) consistent with known CRC subtypes (iCMS2). It also resolved heterogeneity in Cancer-Associated Fibroblasts (CAFs).

5. Significance

• Unlocking Clinical Archives: FX-seq transforms the billions of existing FFPE archives into a resource for single-cell transcriptomics, enabling retrospective studies of disease progression, drug resistance, and biomarker discovery without requiring fresh tissue collection.
• Precision Medicine: By integrating pathological diagnosis (H&E staining) with single-nucleus transcriptomics, FX-seq allows for the molecular characterization of specific tumor regions, facilitating better subtype classification and therapeutic planning.
• Animal Studies: The method allows for PFA-perfusion of animals, fixing the transcriptome in situ. This eliminates the stress-induced gene expression artifacts common in fresh tissue dissociation and enables the study of tissues with high endogenous RNase activity.
• Scalability: The ability to process hundreds of thousands of nuclei from fixed samples makes large-scale cohort studies feasible, bridging the gap between bulk histology and single-cell genomics.

In conclusion, FX-seq represents a paradigm shift in handling fixed tissues, overcoming the historical limitations of RNA degradation and low yield through a sophisticated combination of organocatalysis, regioselective crosslinking, and chemical inhibition, thereby opening new avenues for precision medicine and basic research using archived clinical samples.