Multiple System Atrophy is associated with brain somatic mutations in clonal haematopoiesis genes

This study reveals that brains of patients with Multiple System Atrophy (MSA) exhibit a significant enrichment of somatic mutations in clonal haematopoiesis genes, particularly those affecting myeloid and lymphoid lineages, which are often found across multiple brain regions and may indicate a disease-associated proliferative process potentially involving peripheral infiltration.

The Gist

The Big Picture: A Mystery in the Brain's "Library"

Imagine the human brain as a massive, ancient library. Every cell in the brain is a book containing the instructions (DNA) for how that cell should work. Usually, all the books are identical copies of the same master blueprint.

Multiple System Atrophy (MSA) is a devastating disease where parts of this library start to crumble. The books get damaged, the shelves collapse, and the "librarians" (cells) stop working. We know the disease involves a sticky protein called alpha-synuclein clumping up like spilled glue, but we didn't know why the books were getting damaged in the first place.

Scientists have long suspected that while our "birth certificates" (inherited genes) might not be the problem, the books might get typos as we get older. These are called somatic mutations.

The Detective Work: Looking for "Typo" Clones

In this study, researchers from London and Cambridge acted like high-tech detectives. They wanted to find out if specific "typos" in the brain's DNA were causing MSA.

They focused on two main suspects:

1. The Glue Maker (SNCA): The gene that makes the sticky alpha-synuclein protein.
2. The Security Guards (Tumor Suppressor Genes): Genes that usually stop cells from growing out of control. When these get broken, cells can start acting like a rogue gang.

The Problem: Finding a typo in a library of billions of books is hard. Most sequencing machines are like a noisy crowd; they can't hear a single whisper (a rare mutation) over the background noise.

The Solution: The team used a super-advanced technique called Duplex Sequencing.

• The Analogy: Imagine you are trying to read a very faint message written on a piece of paper. Instead of just reading it once, you ask two people to read it independently. If they both say the exact same thing, you know it's real. If one says "cat" and the other says "bat," you know one of them made a mistake. This technique allowed the scientists to find typos that were incredibly rare, appearing in only a tiny fraction of cells.

The Discovery: The "Rogue Gang" in the Brain

The researchers looked at brain tissue from 20 people with MSA and 9 healthy people. They checked three different rooms in the library: the Cerebellum (balance), the Putamen (movement), and the Cortex (thinking).

Here is what they found:

1. The "Glue" Gene was Clean
They checked the SNCA gene (the glue maker) for typos. Result: Nothing. No new typos were found. This suggests the problem isn't that the glue-making instructions got mutated; it's likely that the instructions are being followed too literally, or the glue is just accumulating for other reasons.

2. The "Security Guards" Were Broken (Clonal Haematopoiesis)
This is the big discovery. They found that the "Security Guard" genes (specifically genes involved in Clonal Haematopoiesis) were broken in many MSA brains.

• The Analogy: Imagine a security guard in a building who gets a promotion to "Team Leader" because they have a special badge (a mutation). Suddenly, this guard and their clones take over a whole floor of the building, pushing out the normal guards. In healthy people, this happens rarely in the brain. In MSA patients, it happened much more often.

3. The "Rogue Gang" Was Everywhere
In healthy brains, if a rogue guard appeared, they usually stayed in one small room. But in MSA brains, these rogue clones were found in all three rooms (Cerebellum, Putamen, and Cortex) of the same person.

• The Analogy: It's as if the rogue guards didn't just start in one room; they either traveled from the outside (the blood) into the brain, or they started early and spread everywhere like a wildfire.

4. The "Bad Guys" Were Stronger
The researchers noticed that the mutations that actually changed the protein (the "bad" mutations) were found in higher numbers than the harmless ones.

• The Analogy: It's like a weed garden. The weeds that are good at surviving (the mutations that help the cell grow) are taking over the garden faster than the normal flowers. This suggests these rogue cells are actively multiplying and spreading within the brain.

Why Does This Matter? The "Trojan Horse" Theory

The most exciting part of the paper is the theory of where these rogue cells are coming from.

These specific mutations are usually found in blood cells (specifically white blood cells) as people age. This is called Clonal Haematopoiesis.

• The Theory: The scientists suspect that these mutated blood cells are acting like Trojan Horses. They travel from the bloodstream, sneak into the brain, and turn into "microglia" (the brain's immune cells). Once inside, they start causing inflammation and damage, contributing to the disease.

Think of it like this: The brain is a fortress. The disease isn't just an internal collapse; it's an invasion from the outside. The "invaders" are mutated blood cells that have learned to survive and multiply, and they are making the brain's environment toxic.

The Takeaway

• What they found: People with MSA have a higher number of "rogue" mutated cells in their brains compared to healthy people. These mutations are the same ones usually found in blood cells.
• What it means: The disease might be driven by a battle between the brain's immune system and these invading rogue cells.
• The Future: This opens a new door for treatment. Instead of just trying to fix the brain, doctors might be able to treat the blood. If we can stop these rogue blood cells from entering the brain or make them less aggressive, we might be able to slow down or stop MSA.

In short: The brain isn't just breaking down on its own; it's being invaded by a mutated army from the blood, and this paper is the first map showing us exactly where that army is hiding.

Technical Summary

1. Problem Statement

Multiple System Atrophy (MSA) is a sporadic, rapidly progressive neurodegenerative disorder characterized by central nervous system (CNS) alpha-synuclein ($\alpha$-syn) inclusions, specifically glial cytoplasmic inclusions (GCIs). While the disease is largely sporadic (heritability <7%), recent studies have identified a germline association with the KCTD7 gene and somatic copy number variations (CNVs) in the SNCA gene. However, the role of somatic single nucleotide variants (SNVs) and small insertions/deletions (indels) in MSA pathogenesis remains unexplored.

The authors hypothesized that:

1. Somatic mutations in SNCA or KCTD7 could drive the unique filament structures seen in MSA.
2. Somatic mutations in clonal haematopoiesis (CH) genes (tumour suppressors frequently mutated in Alzheimer's disease and blood cancers) might be enriched in MSA brains, potentially indicating an infiltration of mutant peripheral immune cells (e.g., microglia) or a localized proliferative process within the brain.

2. Methodology

The study employed a highly sensitive, targeted deep-sequencing approach to detect low-frequency somatic mutations in post-mortem brain tissue.

• Cohort:
  • MSA Cases: 20 patients (10 Olivopontocerebellar Atrophy [OPCA] subtype, 10 Striatonigral Degeneration [SND] subtype).
  • Controls: 9 unaffected individuals.
  • Tissue: DNA extracted from three brain regions: Cerebellum, Putamen, and Cingulate Cortex.
• Target Gene Panel:
  • SNCA (including coding exons, promoters, enhancers, and alternative transcripts).
  • KCTD7.
  • 10 Tumour Suppressor (TS) genes frequently mutated in Alzheimer's microglia: ASXL1, ATRX, CBL, DNMT3A, KMT2D, TET2, TP53, BCR, MLH1, STAT3.
  • Note: 7 of these TS genes are known drivers of Clonal Haematopoiesis (CH).
• Sequencing Technology:
  • Platform: Illumina NovaSeq X (Paired-end 150 bp).
  • Library Prep: Agilent SureSelect XT HS2 with Unique Molecular Identifiers (UMIs) to tag original DNA molecules.
  • Duplex Sequencing: The protocol generated consensus sequences from both DNA strands (Duplex) and single strands (Hybrid) to distinguish true somatic variants from PCR/sequencing errors.
  • Coverage: Mean depths ranged from 963× (Duplex) to 7,249× (Hybrid-MS1).
• Bioinformatics Pipeline:
  • Error Correction: UMI-based read collapsing at three stringency levels (Full Duplex, Hybrid-MS2, Hybrid-MS1).
  • Variant Calling: GATK Mutect2 for SNVs; specific filtering for indels using Duplex data.
  • Validation: Artificial mosaic spike-in mixtures (down to 0.125% Variant Allele Fraction) were used to benchmark sensitivity (84.1% at 0.25% VAF).
  • Filtering: Custom Python pipeline to remove systematic false positives; manual inspection via IGV for all candidate variants.
• Statistical Analysis: Non-parametric tests (Mann-Whitney U, Fisher's exact) and linear regression models adjusted for age.

3. Key Contributions

• First Application of Duplex Sequencing in MSA: This is the first study to apply duplex sequencing with UMIs to search for small somatic mutations in MSA brain tissue, offering significantly higher specificity than previous UMI-only studies.
• Discovery of CH Enrichment: The study provides the first evidence that somatic mutations in clonal haematopoiesis genes are significantly enriched in MSA brains compared to controls.
• Subtype-Specific Patterns: The research delineates distinct mutational patterns between the OPCA and SND subtypes of MSA, linking mutation burden and distribution to specific pathological regions.
• Differentiation from Blood Contamination: Through regional VAF analysis, the authors argue that the mutations are likely intrinsic to the brain tissue (via infiltration or local proliferation) rather than simple blood contamination artifacts.

4. Key Results

• SNCA and KCTD7: No coding somatic SNVs or indels were found in SNCA. While a few KCTD7 variants were detected (including one coding deletion), the sample size was too small to draw conclusions, and no clear excess was observed.
• Clonal Haematopoiesis (CH) Enrichment:
  • Prevalence: MSA brains showed a higher burden of CH mutations (median 1 vs. 0 in controls; $p=0.054$). 75% of MSA brains harbored at least one CH mutation vs. 44% of controls.
  • Multi-Regional Distribution: CH mutations were found in multiple brain regions in 12/20 MSA cases (60%) compared to 1/9 controls (11%; $p=0.020$). Notably, 11 MSA cases had mutations in all three sampled regions, whereas 0 controls did ($p=0.005$).
  • Specific Genes: Mutations were found in DNMT3A and TET2 (myeloid CH) and a recurrent mutation in KMT2D (lymphoid CH) across three MSA cases.
• Variant Allele Fraction (VAF) and Selection:
  • Non-synonymous (protein-altering) CH mutations had higher VAFs than synonymous mutations ($p=0.076$), suggesting a potential proliferative advantage.
  • Regional VAF Gradients: In SND cases, the putamen (most affected region) showed significantly higher VAFs for CH mutations compared to the cerebellum ($p=0.013$). OPCA cases showed no such gradient, suggesting different mechanisms of accumulation.
  • Cortex: Interestingly, the cingulate cortex often harbored the highest VAFs for CH mutations, potentially reflecting cell-type composition differences.
• Indels and Disease Duration: A significant correlation was found between indel burden and disease duration specifically in the SND putamen ($\rho=0.78, p=0.0075$), suggesting disease-associated DNA damage.

5. Significance and Implications

• Pathogenic Mechanism: The findings suggest that MSA pathogenesis may involve a disease-associated proliferative process extending beyond peripheral haematopoiesis. The presence of CH mutations (specifically in DNMT3A, TET2, and KMT2D) in the brain implies that mutant monocytes or lymphocytes may infiltrate the CNS, adopt a microglial phenotype, and drive neuroinflammation.
• Neuroinflammation Link: This supports the hypothesis that peripheral immune cells with somatic mutations contribute to the neuroinflammatory environment in MSA, potentially exacerbating $\alpha$-syn aggregation and demyelination.
• Therapeutic Opportunities: If confirmed, the peripheral origin of these mutations offers new therapeutic avenues. Strategies could include preventing the infiltration of pathogenic monocytes or modulating their activity, distinct from targeting CNS-resident pathology.
• Biomarker Potential: Peripheral CH burden could serve as an accessible biomarker for MSA if replicated in larger cohorts.
• Limitations: The study is limited by the small sample size, the lack of matched peripheral blood for direct comparison, and the inability to definitively identify the specific cell type (microglia vs. infiltrating monocytes) harboring the mutations without single-cell sequencing.

In conclusion, this study shifts the focus from purely neuronal/oligodendroglial somatic mutations to a potential haematopoietic contribution in MSA, revealing a complex interplay between somatic mutations, clonal expansion, and neurodegeneration.