Systems-level longitudinal immune profiling reveals individualized immunotypes and genetic associations

By integrating whole-genome sequencing with longitudinal multi-omics profiling in 101 healthy individuals over two years, this study reveals reproducible individualized immune patterns, identifies three distinct immunotypes with specific metabolic and inflammatory signatures, and uncovers genetic associations linking polygenic scores to memory B cell abundance and activity.

The Gist

Imagine your immune system not as a single, static shield, but as a bustling, 24/7 city. Every person has their own unique version of this city, with different populations of "citizens" (immune cells), different "news broadcasts" (gene activity), and different "smoke signals" sent out into the air (proteins in the blood).

This paper is like a two-year-long documentary that followed 101 healthy people to map out these unique immune cities. The researchers wanted to answer a big question: Is your immune system a chaotic mess that changes every day, or does it have a stable, personal "personality" that stays with you?

Here is the story of what they found, broken down into simple concepts:

1. The "Fingerprint" of Your Immune System

The researchers took blood samples from these 101 people six times over two years. They looked at three different layers of data:

• The Citizens: Counting the different types of immune cells (like T-cells, B-cells, and monocytes).
• The News: Reading the genes inside the cells to see what they were "thinking" or doing.
• The Smoke Signals: Measuring proteins floating in the blood that tell the story of inflammation or health.

The Discovery: Just like your face or your fingerprints, your immune system has a distinct "signature." Even though the number of cells might wiggle up and down a little, the pattern of your immune system remained surprisingly stable over two years. If you met your immune system in Year 1 and again in Year 2, you would recognize it immediately. It's not a random storm; it's a consistent climate.

2. The Three "Immune Personalities" (Immunotypes)

When the researchers looked at everyone's data together, they didn't see a smooth gradient. Instead, they found that the healthy population naturally sorted itself into three distinct "teams" or personalities:

• Team "The Defenders" (Cluster A):

  • Who they are: Mostly women.
  • The Vibe: These people have a lot of B-cells and CD4+ T-cells. Think of them as the "diplomats and archivists" of the immune city. They are great at remembering past infections and making antibodies.
  • Health Check: They generally had the best metabolic health (lower blood pressure, better cholesterol) and less inflammation.
  • The Secret: Their style is heavily influenced by their DNA. It's like they were born with a blueprint that says, "Be a B-cell specialist."

• Team "The Alarmists" (Cluster B):

  • Who they are: Mostly men.
  • The Vibe: These people are dominated by monocytes and dendritic cells. Think of them as the "firefighters and scouts." They are always on high alert, scanning for trouble.
  • Health Check: Their city is a bit noisier. They have higher levels of inflammation markers (like CRP and IL-6). It's not that they are sick, but their immune system is running a "low-grade fire drill" more often than the others.

• Team "The Hit Squad" (Cluster C):

  • Who they are: A mix of men and women.
  • The Vibe: These people are packed with CD8+ T-cells and NK cells. Think of them as the "special forces" or "assassins" who are ready to destroy infected cells immediately.
  • Health Check: They have a unique mix. They aren't as inflamed as Team B, but they have slightly worse cholesterol numbers than Team A. They are the "aggressive protectors."

3. The Genetic "Architect"

The researchers also looked at the people's DNA to see if genetics built these cities.

• The Finding: Genetics played a huge role in Team A (The Defenders). If you have certain genes, you are statistically likely to have more memory B-cells.
• The Twist: However, for the other teams, genetics wasn't the only boss. The environment, lifestyle, and how your body handles stress (like diet or sleep) seemed to shape the "Alarmist" and "Hit Squad" styles just as much as your DNA did.

4. Why This Matters (The "So What?")

Imagine you go to a doctor for a check-up. Today, doctors often say, "Your immune system is normal" or "It's a bit high." This paper suggests that "normal" is actually a spectrum of three different healthy styles.

• Precision Medicine: If you are "Team Alarmist," a doctor might know to look closer at your heart health or inflammation markers, even if you feel fine. If you are "Team Hit Squad," they might know you are primed to fight viruses aggressively.
• Vaccines and Treatments: Understanding which "team" you belong to could help doctors predict how well you will respond to a vaccine or a new drug. A vaccine might work differently for a "Diplomat" than for a "Special Forces" soldier.

The Big Takeaway

Your immune system isn't a random fluctuation of numbers. It is a stable, personalized ecosystem. You have a unique "immune personality" that is a mix of your DNA and your life experiences. By understanding which of the three main "teams" you belong to, we can move from a "one-size-fits-all" approach to medicine to a truly personalized one that respects your unique biological identity.

Technical Summary

1. Problem Statement

The human immune system exhibits substantial inter-individual variation, yet the mechanisms coordinating this variability across different molecular layers (cells, transcripts, proteins) and its persistence over time remain incompletely defined. While previous studies have identified that immune traits vary widely among healthy individuals, there is a lack of systems-level understanding regarding:

• How cellular composition, gene expression, and circulating proteins are coordinated to form stable, individualized "immunotypes."
• The extent to which these immunotypes are shaped by genetic architecture versus non-heritable factors (environment, physiology).
• How these personalized immune states relate to systemic metabolic and inflammatory health.

2. Methodology

The study employed a longitudinal, multi-omics integrative approach on the Swedish SciLifeLab SCAPIS Wellness Profiling (S3WP) cohort.

• Cohort: 101 clinically healthy individuals (ages 50–66, 51% female) followed for two years across six visits.
• Data Modalities:
  • Mass Cytometry (CyTOF): Profiling of 53 immune cell subsets across 18 major lineages in Peripheral Blood Mononuclear Cells (PBMCs).
  • Transcriptomics (RNA-seq): PBMC gene expression profiling (16,985–18,100 expressed genes).
  • Proteomics (Olink): Quantification of 794 plasma proteins covering immune, inflammatory, and metabolic pathways.
  • Genomics: Whole-genome sequencing (WGS) generating ~6.7 million high-confidence SNPs per individual.
• Analytical Framework:
  • Stability Analysis: Used t-SNE, Euclidean distance, and Intraclass Correlation Coefficients (ICC) via linear mixed-effects models to distinguish intra-individual stability from inter-individual variability.
  • Network Construction: Applied isometric log-ratio (ILR) transformations to handle compositional data, followed by linear regression to map associations between immune cell frequencies and gene expression.
  • Clustering: Used K-Nearest Neighbor (KNN) and Louvain algorithms to identify distinct immunotypes based on the integrated network.
  • Genetic Analysis: Conducted Genome-Wide Association Studies (GWAS) for cell frequencies (ctQTLs) and gene expression (eQTLs), and calculated Polygenic Scores (PGS) for memory B cells.

3. Key Contributions

• Longitudinal Validation: Demonstrated that individual immune signatures (particularly transcriptomic and proteomic) are highly stable over a two-year period, with intra-individual distances significantly lower than inter-individual distances.
• Multi-Omic Integration: Successfully linked cellular frequencies to transcriptional programs and circulating plasma proteins, revealing that molecular features (transcripts/proteins) often exhibit stronger individual segregation than cellular frequencies alone.
• Immunotype Stratification: Defined three distinct, stable immunotypes in healthy adults, characterized by coordinated cellular, transcriptional, and proteomic modules.
• Genetic Architecture: Identified that while genetics significantly influence specific memory lymphocyte populations (B and T cells), the broader immunotype stratification is driven more by physiological and environmental factors than by gross genetic differences.

4. Key Results

A. Individual Immune Signatures and Stability

• Molecular vs. Cellular Stability: PBMC transcriptomics and plasma proteomics showed strong individual-level segregation, whereas immune cell frequencies showed moderate clustering.
• High-ICC Features: Adaptive immune cells (memory CD4+, CD8+, and B cells) and specific genes (e.g., HLA family, ERAP2, SIGLEC14, cytotoxic effectors like GZMH) exhibited the highest inter-individual variability (ICC > 0.8).
• Sex Differences: Sex was a major driver of transcriptional variation (23% of genes), with females showing upregulation of X-chromosome escape genes and B-cell receptor pathways, while males showed higher expression of innate receptors (NKG2, TLRs). Age had no significant impact within the narrow 50–65 year range.

B. Three Distinct Immunotypes

Clustering analysis revealed three stable groups with distinct biological and clinical profiles:

1. Cluster A (Adaptive Lymphoid):
   • Composition: Enriched in B cells and CD4+ T cells; higher CD4:CD8 ratio.
   • Demographics: Predominantly female (73%).
   • Clinical Profile: Lowest cardiometabolic and inflammatory risk (high HDL/ApoA1, low CRP, LDL, triglycerides).
   • Genetics: Strongly associated with genetic predisposition for memory B cells.
2. Cluster B (Myeloid-Inflammation):
   • Composition: Enriched in monocytes and dendritic cells.
   • Demographics: Predominantly male (70%).
   • Clinical Profile: High inflammatory signature (elevated CRP, IL-6, GGT, Troponin T) and pro-inflammatory plasma proteins (OSM, CXCL10).
3. Cluster C (Cytotoxic):
   • Composition: Dominated by CD8+ T cells and NK cells.
   • Demographics: Balanced sex distribution.
   • Clinical Profile: Low inflammation but higher metabolic risk (lower HDL, higher LDL) compared to Cluster A. Characterized by lymphoid-associated receptors (FCRL5/6).

C. Genetic Associations

• ctQTLs: Identified 39 independent cell-type quantitative trait loci (ctQTLs) affecting 15 immune subtypes.
• Specificity: Genetic associations were almost exclusively linked to adaptive immunity (49% memory B cells, 36% memory CD8+ T cells). No significant ctQTLs were found for naïve cells.
• Polygenic Score (PGS): A PGS for memory B cell frequency correlated significantly with both cell abundance ($\rho=0.44$) and the transcriptional activity of the B-cell module.
• Genetic vs. Environmental: While genetics explained a significant portion of variance in memory B and T cells, the overall stratification into the three immunotypes was not strongly driven by genetic differences between clusters, suggesting environmental and physiological factors play a larger role in defining the broad immunotype.

5. Significance

• Precision Immunology: This study provides a framework for defining "normal" immune heterogeneity. It suggests that healthy individuals can be stratified into biologically meaningful groups that reflect their systemic physiological state (metabolic health, inflammation levels).
• Biomarker Discovery: The identification of stable, individual-specific immune signatures offers potential biomarkers for predicting susceptibility to infections, vaccine responsiveness, and immune-targeted therapies.
• Systems Biology: By integrating four omics layers over time, the study moves beyond static snapshots to reveal how genetic, cellular, and molecular layers interact to maintain a personalized immune "set point."
• Clinical Implications: The distinct immunotypes (Adaptive, Myeloid, Cytotoxic) highlight different axes of immune variation that may require tailored therapeutic or preventive strategies, particularly in the context of aging and cardiometabolic disease.

Limitations Noted: The cohort was limited to middle-aged Scandinavian adults (limiting generalizability), the sample size (N=101) restricted GWAS power, and the study was descriptive (healthy individuals only), preventing causal inference regarding disease outcomes.