Colocalization and discordance between plasma and brain protein quantitative trait loci

This study reveals significant discordance between plasma and brain protein quantitative trait loci (pQTLs), demonstrating that while circulating pQTLs effectively capture systemic and immune pathways, tissue-specific data is crucial for accurately interpreting brain-related protein associations and prioritizing therapeutic targets.

The Gist

Imagine your body is a massive, bustling city with different neighborhoods, each running its own unique operations. Two of the most important neighborhoods in this study are the Blood District (plasma) and the Brain District (specifically the dorsal prefrontal cortex).

Scientists are trying to understand how our DNA acts as the "blueprint" for the proteins (the workers) that keep this city running. They look for specific spots in the DNA that control how many of these workers are made; these spots are called pQTLs.

Here is what the researchers found when they compared the blueprints for these workers in the Blood District versus the Brain District:

1. The "Same Blueprint, Different Neighborhoods" Problem
Usually, scientists like to measure proteins in the blood because it's easy to get a sample, kind of like checking the city's main news feed. They assume that what's happening in the Blood District reflects what's happening everywhere else. However, this study found that this assumption isn't always true.

When they compared the DNA blueprints for the same proteins in both neighborhoods, they found that only about 80% of the time did the blueprints match up perfectly. In the other 20% of cases, the DNA instructions in the blood didn't match the instructions in the brain at all.

2. The "Opposite Directions" Twist
Even when the blueprints did match (the 80% that colocalized), there was a funny twist. In about 20% of those matching cases, the DNA instructions actually told the proteins to do the exact opposite thing.

• Analogy: Imagine a DNA switch that says, "Turn up the volume" in the Blood District, but the exact same switch says, "Turn down the volume" in the Brain District. If you only looked at the blood, you'd think the protein was getting louder, but in the brain, it was getting quieter.

3. Who Lives Where?
The researchers looked at which proteins live in which neighborhoods:

• The Commuters: Proteins that had matching blueprints in both blood and brain tended to be the "commuters"—they are highly active in the immune system and general body tissues.
• The Locals: The proteins that didn't match up were more likely to be "locals" who only hang out in the brain and don't travel to the blood.

4. The "Neuroticism" Test Drive
To see why this matters, the scientists ran a simulation using a trait called neuroticism (a tendency to feel anxious or worried). They used the DNA blueprints from the blood to predict how proteins affect neuroticism, and then compared that to using blueprints from the brain.

• They found 13 proteins linked to neuroticism.
• Shockingly, for 6 of those proteins, the blood blueprints suggested they would increase neuroticism, while the brain blueprints suggested they would decrease it.

The Bottom Line
The study concludes that checking the blood is a great way to understand proteins that circulate through the whole body and the immune system. However, if you are interested in proteins that stay mostly in the brain, relying only on blood samples can give you a misleading picture. To get the full story, you need to look at the specific "neighborhood" where the protein actually lives.

Technical Summary

Technical Summary: Colocalization and Discordance Between Plasma and Brain Protein Quantitative Trait Loci

Problem Statement
Protein quantitative trait loci (pQTLs) are instrumental for interpreting complex trait genetics, identifying biomarkers, and pinpointing therapeutic targets. However, a critical limitation exists in current research: while circulating proteins are frequently utilized in pQTL studies due to the accessibility of blood-based measurements, their regulatory mechanisms may not accurately reflect those in disease-relevant tissues, such as the brain. This discrepancy raises concerns regarding the validity of using plasma-derived genetic instruments for understanding central nervous system pathologies.

Methodology
To address this, the authors conducted a comprehensive assessment of the colocalization and discordance between plasma and dorsal prefrontal cortex (DPC) cis-pQTLs. The study integrated data from four large-scale pQTL studies to evaluate the genetic overlap between these two tissue contexts. The analytical framework included:

• Colocalization Analysis: Determining the extent to which plasma and brain cis-pQTLs share causal variants.
• Directional Concordance: Investigating whether colocalized loci exhibit consistent or opposing directions of genetic effects.
• Tissue-Specific Profiling: Utilizing data from the Genotype-Tissue Expression (GTEx) project to characterize gene expression profiles and correlate them with pQTL colocalization status.
• Mendelian Randomization (MR): Conducting MR analyses using neuroticism as an illustrative complex trait to compare effect estimates derived from plasma-based instruments versus those derived from brain-based instruments.

Key Results
The analysis yielded several critical findings regarding the relationship between plasma and brain proteomics:

• Limited Colocalization: Across the proteins examined, at most 80% of the cis-pQTLs demonstrated evidence of colocalization between plasma and the dorsal prefrontal cortex.
• Directional Discordance: Among the loci that did colocalize, approximately 20% exhibited opposite directions of genetic effects, suggesting that a shared genetic variant can regulate protein levels in opposing ways depending on the tissue context.
• Expression Patterns: Proteins with colocalized cis-pQTLs were significantly more likely to exhibit high gene expression in systemic tissues and immune cells. Conversely, proteins lacking colocalization were more likely to have high expression specifically in brain tissues.
• Impact on Trait Association: In the MR analysis of neuroticism, 13 proteins were identified as significantly associated with the trait. Notably, six of these proteins displayed opposite effect directions when instruments were derived from plasma compared to the dorsal prefrontal cortex.

Significance and Claims
The paper posits that while circulating pQTLs remain informative for proteins involved in systemic and immune pathways, they may be insufficient or misleading for proteins with localized brain expression. The authors claim that incorporating tissue-specific data is essential for proteins that do not colocalize well between blood and brain. Furthermore, the study highlights that considering multiple tissue contexts can refine the interpretation of protein-trait associations and improve the prioritization of candidate therapeutic targets, particularly when genetic effects differ in direction or magnitude across tissues. The findings underscore the necessity of validating tissue relevance when translating pQTL findings into biomarkers or drug targets for neurological conditions.