Methylation Risk Score Identifies an Interferon-Driven SLE Subset Distinct from Polygenic Risk

This study demonstrates that a methylation risk score (MRS) identifies a distinct, interferon-driven subset of systemic lupus erythematosus (SLE) characterized by specific autoantibodies and HLA-DRB1*03:01, which is phenotypically and mechanistically separate from the nephritis-associated risk defined by polygenic genetic scores.

The Gist

Imagine your body's immune system is a highly sophisticated security team. In a healthy person, this team knows exactly who the friends are and who the enemies are. But in Systemic Lupus Erythematosus (SLE), the security team gets confused and starts attacking the body's own organs.

For a long time, scientists thought this confusion was mostly written in the "hard drive" of our DNA (our genetics). They created a Polygenic Risk Score (PRS), which is like a genetic credit score. If your score is high, it means you have a lot of "bad genes" that make you prone to Lupus, specifically predicting severe kidney problems.

However, this new study suggests there's a second, hidden layer to the story: the Methylation Risk Score (MRS). Think of DNA methylation not as the hard drive itself, but as the software settings or the dimmer switches on the lights. You can have the same hard drive (genes) as someone else, but if your dimmer switches are turned up too high, the lights (genes) shine too bright and cause chaos.

Here is what the researchers found, broken down simply:

1. Two Different Types of "Bad Luck"

The study discovered that Lupus isn't just one big mess; it's actually two different types of messes driven by different causes:

• The "Genetic Kidney" Type (High PRS):

  • The Analogy: This is like a house with a weak foundation. No matter what you do, the basement (kidneys) is likely to flood.
  • The Science: People with a high genetic risk score are most likely to develop nephritis (kidney inflammation) and have specific antibodies (anti-dsDNA). This is the "classic" severe genetic risk.

• The "Interferon Alarm" Type (High MRS):

  • The Analogy: This is like a house where the smoke detectors are set to "ultra-sensitive." Even a tiny piece of toast burning (a minor trigger) sets off the entire sprinkler system, flooding the whole house.
  • The Science: This group has a high Methylation Risk Score. Their "dimmer switches" are turned up on genes that control Interferon, a chemical signal that sounds the alarm for the immune system.
  • The Result: These patients have high levels of Interferon in their blood. They are more likely to get skin rashes (like discoid lupus), have low blood cell counts, and carry a wide variety of autoantibodies (like anti-SSA).

2. The "Switch" and the "Genetic Blueprint"

The most exciting part of this study is that these two scores do not correlate.

• You can have a high genetic risk (bad foundation) but a normal methylation score (calm alarms).
• You can have a normal genetic risk but a high methylation score (chaotic alarms).

It's like having a car with a great engine (good genes) but a stuck accelerator (bad methylation). The car will still race out of control, but for a different reason than a car with a broken engine.

3. The "HLA-DRB1" Connection

The researchers found a specific genetic tag called HLA-DRB1*03:01.

• If you have this tag, you are much more likely to have the "Interferon Alarm" type of Lupus (High MRS).
• This group tends to have skin rashes, specific antibodies, and a very active immune system.
• Interestingly, the study found that as patients get older, their "methylation alarm" (MRS) tends to quiet down, perhaps because the disease burns out or because medications help turn the dimmer switches back down.

Why Does This Matter?

Imagine a doctor trying to treat Lupus. In the past, they might have treated everyone the same way.

• The Old Way: "You have Lupus, here is a general immunosuppressant."
• The New Way (Based on this study): "Let's check your scores."
  • If your Genetic Score is high, we watch your kidneys closely.
  • If your Methylation Score is high, we know your immune system is stuck in "high alert" mode. We might target treatments specifically to calm down that Interferon alarm, rather than just suppressing the whole immune system.

The Bottom Line

This paper tells us that Lupus is a "chameleon." It wears different masks.

• One mask is Genetic, leading to kidney trouble.
• The other mask is Epigenetic (chemical switches), leading to skin rashes, high alarms, and a specific type of antibody storm.

By measuring the "chemical switches" (Methylation Risk Score), doctors can finally identify a specific group of patients who are driven by a runaway Interferon alarm, allowing for more precise, personalized treatment in the future. It's like realizing that while two houses are on fire, one is burning because of a short circuit (genes), and the other is burning because someone left the oven on (methylation/environment)—and they need different fire extinguishers.

Technical Summary

1. Problem Statement

Systemic Lupus Erythematosus (SLE) is a heterogeneous autoimmune disease driven by a complex interplay of genetic, environmental, and epigenetic factors. While Polygenic Risk Scores (PRS) based on non-HLA genetic variants have been developed to predict disease severity (particularly nephritis), they do not fully capture the clinical heterogeneity of SLE. Epigenetic modifications, specifically DNA methylation, are known to be dysregulated in SLE, often linked to interferon (IFN) signaling. However, there is a lack of standardized tools to quantify these epigenetic changes and determine if they define distinct clinical subphenotypes that differ from those predicted by genetic risk alone. The study aims to investigate whether a Methylation Risk Score (MRS) can identify specific clinical and immunological subsets of SLE and how these relate to existing genetic risk scores and HLA-DRB1*03:01 status.

2. Methodology

The study employed a discovery-replication cohort design involving whole blood DNA analysis.

• Cohorts:
  • Discovery Cohort: 311 SLE patients (fulfilling ≥4 ACR-82 criteria) and 400 healthy controls from Uppsala and Linköping University Hospitals.
  • Replication Cohort: 175 SLE patients and 187 controls from Karolinska University Hospital.
  • Combined Analysis: 486 patients and 587 controls for comparative analysis between MRS and PRS.
• Epigenetic Profiling:
  • DNA methylation was assessed using the Illumina HumanMethylation450K (HM450K) BeadChip array.
  • MRS Construction: Based on 17 independent, top differentially methylated CpG sites (DMCs) identified in previous work (Imgenberg-Kreuz et al., 2018) with an effect size ($\Delta\beta$) $\ge$ 0.1.
  • Calculation: Z-scores were calculated for each CpG site relative to control means/SDs. For hypomethylated sites (16 of 17), the Z-score was multiplied by -1. The final MRS was the sum of these standardized scores.
• Genetic Profiling:
  • Genotyping was performed using the Illumina 200K Immunochip.
  • PRS Construction: A non-HLA Polygenic Risk Score comprising 57 SLE-associated SNVs (based on Reid et al.).
  • HLA Typing: Genotyping for the HLA-DRB103:01* tag SNP (rs1269852).
• Clinical & Immunological Data:
  • Clinical variables included ACR-82 criteria, disease activity (SLEDAI-2K or SLAM), and medication use.
  • Autoantibodies (anti-SSA, SSB, RNP, Sm, dsDNA) and serum IFN-$\alpha$2 levels (measured via Simoa) were assessed.
• Statistical Analysis:
  • Multivariable logistic regression (adjusted for age/sex) for binary outcomes.
  • Linear regression and Kruskal-Wallis tests for continuous variables.
  • Comparison of MRS and PRS associations with clinical subphenotypes.

3. Key Contributions

• Development of an SLE Methylation Risk Score (MRS): A standardized, quantitative score based on 17 specific CpG sites that reflects the cumulative epigenetic burden in SLE.
• Phenotypic Stratification: The study delineates a specific SLE subset characterized by high MRS, which is distinct from the subset defined by high PRS.
• Integration of Epigenetics and Genetics: The research demonstrates that epigenetic risk (MRS) and genetic risk (PRS) operate largely independently in SLE pathogenesis, capturing different biological pathways.
• Link to HLA-DRB1*03:01: The study establishes a novel association between the MRS and the HLA-DRB103:01* risk allele, suggesting a specific epigenetic mechanism linked to this genetic haplotype.

4. Key Results

A. MRS and Interferon Signaling

• IFN-$\alpha$ Correlation: Higher MRS was strongly associated with elevated serum IFN-$\alpha$2 levels ($p = 1.04 \times 10^{-14}$).
• Gene Annotation: 16 of the 17 CpG sites in the MRS are annotated to Interferon-regulated genes (e.g., MX1, PARP9, IFI44L), confirming the score captures an "Interferon Signature."

B. Clinical Associations of MRS

• Positive Associations: High MRS correlated with:
  • Discoid lupus rash.
  • Hematologic disorders.
  • Immunologic disorders (hypocomplementemia).
  • High disease activity (SLEDAI/SLAM) in the discovery cohort.
  • Positivity for multiple autoantibodies, specifically anti-SSA, anti-SSB, anti-RNP, and anti-Sm.
  • HLA-DRB103:01* carrier status.
• Treatment: Higher MRS was associated with prednisolone use, while lower MRS was linked to antimalarial use.

C. Comparison with Polygenic Risk Score (PRS)

• Distinct Phenotypes:
  • High PRS was associated with nephritis and anti-dsDNA positivity, but not with discoid rash or anti-SSA/SSB.
  • High MRS was associated with discoid rash, hematologic disorder, and anti-SSA/SSB, but not significantly with nephritis.
• Lack of Correlation: There was no significant correlation between MRS and PRS ($p=0.35$, $R^2=0.05$), indicating they measure independent risk factors.
• Overlap: Only one gene (IRF7) was represented in both the MRS CpG sites and the PRS SNPs.

D. HLA-DRB1*03:01 Interaction

• Patients carrying HLA-DRB103:01* had significantly higher MRS levels.
• The association between MRS and clinical features (like anti-SSA) was observed in both carriers and non-carriers, but the slope of the relationship between PRS and MRS differed significantly based on HLA-DRB103:01* status.

5. Significance and Conclusion

This study provides compelling evidence that SLE is not a monolithic disease but comprises distinct pathogenic pathways that can be stratified using integrated genetic and epigenetic data.

• Interferon-Driven Subset: The MRS identifies a specific "Interferon-high" subset of SLE patients characterized by HLA-DRB103:01* positivity, skin manifestations (discoid/malar rash), hematologic issues, and a broad autoantibody profile (SSA/SSB/RNP/Sm).
• Complementary Biomarkers: The MRS and PRS are complementary rather than redundant. While PRS predicts renal severity (nephritis), MRS predicts interferon-driven systemic and cutaneous manifestations.
• Mechanistic Insight: The findings suggest a feedback loop where HLA-DRB103:01* and autoantibodies drive IFN production, which in turn induces DNA hypomethylation in IFN-regulated genes, further amplifying the disease phenotype.
• Clinical Utility: The MRS offers a potential tool for patient stratification, enabling more personalized treatment approaches by identifying patients who may benefit from interferon-targeted therapies versus those with high genetic risk for renal involvement.

Limitations: The study notes that the regression models were not adjusted for cell type proportions (a potential confounder in methylation studies) and the cohort was predominantly Caucasian, necessitating validation in diverse ethnic populations. Additionally, the longitudinal stability of the MRS over time requires further investigation.