Analysis of the diabetic arterial transcriptome to define novel biomarkers of macrovascular disease

By analyzing arterial transcriptome data and validating findings in plasma proteomics, this study identified four novel protein biomarkers (ACP5, LEFTY2, LILRA5, and PSME2) that are differentially expressed in people with diabetes and improve the accuracy of cardiovascular disease risk prediction models.

The Gist

The "Broken Plumbing" Mystery: Why Diabetes Attacks the Arteries

Imagine your body is a massive, high-tech city. To keep everything running, the city relies on a complex network of water pipes (your arteries) to deliver essential supplies to every building (your cells).

Now, imagine a city where a specific type of resident—let’s call them "The Sugar Gang" (Diabetes)—is moving in. This gang doesn't just sit around; they cause chaos. They leave sticky residue in the pipes, causing them to clog, harden, and eventually burst. This is what doctors call Atherosclerotic Cardiovascular Disease (ASCVD).

For a long time, doctors have known that people with diabetes have much "rustier" and "clogged" pipes than others, but they haven't quite known exactly what the Sugar Gang is doing to the pipe walls at a microscopic level.

This research paper is like a team of forensic investigators who used super-powered microscopes to look at the "blueprints" of the pipes to find out exactly how the damage is happening.

---

1. The Investigation: Reading the "Instruction Manuals"

The researchers didn't just look at the pipes; they looked at the transcriptome.

The Analogy: Think of every cell in your artery as a tiny factory. These factories have a massive library of instruction manuals (DNA). When a factory is working normally, it reads certain manuals to keep the pipe smooth. But when diabetes hits, the factories start reading the wrong manuals—manuals for "How to build inflammation" or "How to create calcium buildup."

The scientists compared the "instruction manuals" being read in the arteries of healthy people versus people with diabetes. They looked at two different parts of the city: the big main water lines (the aorta) and the smaller neighborhood pipes (the tibial artery).

2. The Discovery: The "Inflammation Alarm"

They found that in both types of pipes, the "factories" were reading manuals related to inflammation and immune responses.

It’s as if the pipes, sensing the "Sugar Gang," started calling for an emergency response team (immune cells). However, instead of fixing the problem, this constant emergency call actually caused more damage, making the pipes thicker and more brittle.

3. The "Smoking Gun": Four New Clues

The researchers wanted to see if they could find "clues" floating in the city's water supply (the blood) that would tell them how much damage was happening in the pipes without having to actually cut the pipes open.

They identified four specific proteins (tiny chemical messengers) that act like "Warning Flares" in the blood:

1. ACP5
2. LILRA5
3. PSME2
4. LEFTY2

When these four "flares" are high in your blood, it’s a strong signal that the "Sugar Gang" is actively damaging your arterial pipes.

4. Why This Matters: A Better Weather Forecast

Currently, doctors use "risk models" (like a weather forecast) to predict if a patient is likely to have a heart attack or stroke. These models look at things like age, blood pressure, and cholesterol.

The researchers tested their four new "Warning Flares" by adding them to the existing weather forecast.

The Result: The forecast became much more accurate. It’s the difference between saying, "It might rain today" and saying, "It’s going to rain, and there’s a 90% chance of a thunderstorm because we can see the dark clouds approaching." By adding these proteins, doctors can better identify who is at the highest risk, potentially saving lives through earlier intervention.

---

Summary in a Nutshell

• The Problem: Diabetes makes your arteries clog and harden much faster.
• The Science: Researchers looked at the genetic "instruction manuals" in the arteries and found they were stuck in "emergency/inflammation mode."
• The Breakthrough: They found four specific proteins in the blood that act as early warning signals for this damage.
• The Goal: Using these signals, doctors can create better "early warning systems" to predict and prevent heart attacks and strokes in people with diabetes.

Technical Summary

Technical Summary: Analysis of the Diabetic Arterial Transcriptome to Define Novel Biomarkers of Macrovascular Disease

1. The Problem

Diabetes mellitus (DM) approximately doubles the risk of atherosclerotic cardiovascular disease (ASCVD), yet the precise molecular mechanisms by which DM alters the human arterial wall remain poorly understood. While preclinical models suggest roles for oxidative stress and inflammation, there is a lack of unbiased, large-scale human transcriptomic data characterizing the diabetic artery. Furthermore, existing clinical risk prediction tools (such as SCORE2) do not fully capture the unique biological risk profile of individuals with DM, highlighting a need for novel, biologically informed biomarkers.

2. Methodology

The researchers employed a multi-omic, cross-cohort approach to bridge the gap between tissue-level gene expression and systemic plasma biomarkers:

• Transcriptomic Discovery (GTEx Cohort): The study analyzed RNA-sequencing data from the Genotype-Tissue Expression (GTEx) project. They identified differentially expressed genes (DEGs) associated with DM in two distinct anatomical sites: the thoracic aorta (n=427) and the tibial artery (n=645). The analysis controlled for multiple covariates, including age, sex, race, BMI, hypertension, and technical factors like RNA integrity (RIN) and ischaemic time.
• Proteomic Validation (UK Biobank): To translate tissue findings into clinically accessible markers, the authors cross-referenced the DEGs with plasma proteomic data from the UK Biobank (UKB) Olink Explore 3072 panel (n=52,996). They sought proteins that were both concordantly differentially abundant in the plasma of DM patients and associated with ASCVD events.
• Risk Model Augmentation: The study evaluated the incremental predictive value of the identified protein biomarkers by adding them to the established SCORE2 (for non-diabetics) and SCORE2-Diabetes (for diabetics) models. Performance was assessed using Area Under the Curve (AUC), Net Reclassification Improvement (NRI), and calibration plots.

3. Key Contributions

• Large-scale Arterial Mapping: This represents one of the largest analyses of the human arterial transcriptome in the context of diabetes, providing a high-resolution view of how DM affects different vascular beds.
• Identification of a Shared Molecular Signature: The study identified a core set of 22 DEGs common to both the aorta and the tibial artery, suggesting a systemic rather than purely site-specific diabetic vascular pathology.
• Translation of Tissue Hits to Plasma Biomarkers: The study successfully identified four specific proteins (ACP5, LEFTY2, LILRA5, and PSME2) that bridge the gap between arterial gene expression and measurable systemic inflammation/risk.

4. Results

• Transcriptomic Findings: In the thoracic aorta, 619 DEGs were identified; in the tibial artery, 356 DEGs were identified. Enrichment analysis revealed that both sites were heavily influenced by inflammatory and immune-related biological processes (e.g., dendritic cell differentiation and cytokine response). Notably, many upregulated genes were associated with macrophage expression.
• Proteomic Validation: Of the shared DEGs, four proteins were validated in the UKB plasma proteomic dataset. These proteins showed concordant differential abundance in DM patients and were significantly associated with cardiovascular risk factors (e.g., HbA1c, BMI, PASI) and incident ASCVD events (MI, stroke, and cardiovascular mortality).
• Improved Risk Prediction: The addition of these four proteins to the SCORE2/SCORE2-Diabetes models resulted in a statistically significant increase in discriminative performance (AUC). The models also showed improved population-level classification and calibration, demonstrating that these proteins provide incremental information not captured by traditional clinical risk factors.

5. Significance

This study provides a molecular roadmap of diabetic macrovascular disease, shifting the focus from general risk factors to specific, biologically driven pathways—primarily involving myeloid/macrophage-driven inflammation.

By demonstrating that tissue-derived transcriptomic hits can be successfully validated as plasma proteins that improve clinical risk models, the authors provide a framework for future biomarker discovery. While the study does not establish causality, the findings suggest that these proteins could serve as novel targets for both early risk stratification and potential therapeutic intervention in the management of diabetic cardiovascular complications.