Dissecting epigenome dynamics in human immune cells upon viral and chemical exposure by multimodal single-cell profiling

This study presents a multimodal single-cell epigenome atlas of human immune cells exposed to various viral and chemical stressors, revealing distinct and coordinated remodeling of chromatin accessibility and DNA methylation landscapes that define specific cell states, such as T cell exhaustion and inflammatory responses.

The Gist

Imagine your immune system as a massive, bustling city. The cells are the citizens, and their DNA is the city's master blueprint. Usually, this blueprint is locked away in a secure vault (the nucleus), and only specific parts are opened up for the workers to read and use. This "opening up" of the blueprint is called chromatin accessibility.

When a virus or a chemical invades, the city needs to react instantly. It doesn't just change the workers' behavior; it actually rewrites parts of the blueprint, deciding which instructions to unlock and which to lock down permanently.

This paper is like a high-tech, real-time surveillance report of that city during four different crises: HIV, the Flu, SARS-CoV-2 (COVID-19), and exposure to pesticides. The researchers didn't just look at the workers (RNA); they looked at the blueprints themselves (DNA) to see exactly how the city's rules were being rewritten.

Here is the breakdown of their findings, translated into everyday language:

1. The Big Picture: A City Under Siege

The researchers took blood samples from nearly 300,000 individual immune cells from people exposed to these four threats. They used a super-powerful microscope (single-cell sequencing) to look at the "openness" of the DNA in every single cell.

Think of it like checking the doors of every house in the city.

• Open doors = The cell is ready to read that instruction (e.g., "Fight the virus!").
• Locked doors = The cell is ignoring that instruction.

They found that while every crisis changes the city, COVID-19 caused the most dramatic and chaotic remodeling of the blueprints, especially in the "police force" cells (monocytes).

2. The HIV Story: The Exhausted Soldiers

In people with long-term HIV, the researchers found a specific group of soldiers (CD8+ T-cells) that had become "exhausted."

• The Analogy: Imagine a soldier who has been fighting a war for years without rest. They are still there, but they are tired, confused, and their orders are mixed up.
• The Finding: The blueprint for these exhausted soldiers showed that the "stop" signs were being removed, and "sleep" instructions were being unlocked. Specifically, a family of master switches called FOXP was flipping the wrong switches, turning off the soldiers' ability to fight effectively. It's like the city's command center accidentally locked the weapons room and opened the nap room.

3. The COVID-19 Story: The Confused Police

This was the most surprising part. In severe COVID-19, the "police" cells (CD14+ monocytes) didn't just get stronger; they got weird.

• The Analogy: Imagine the police force suddenly deciding to stop patrolling the streets (ignoring inflammation) and instead start acting like a secret police force that suppresses the population (immune suppression).
• The Finding:
  • The "NF-kB" Switch: This is the "Siren" that usually screams "FIRE! FIGHT!" In severe COVID, the blueprint for this siren was locked shut. The cells stopped screaming.
  • The "AP-1" Switch: This is the "Stress Response" switch. In severe COVID, this switch was wide open.
  • The Result: The cells stopped trying to clear the virus and started acting like "Myeloid-Derived Suppressor Cells" (MDSCs). They essentially put on a blindfold and stopped fighting, which is why the body can't clear the virus in severe cases. It's a case of "immune paralysis."

4. The Flu and Pesticides: The Quiet Neighbors

Interestingly, the Flu and the pesticide exposure didn't cause the same massive blueprint rewriting as HIV or COVID.

• The Analogy: If COVID is a hurricane tearing roofs off houses, the Flu and pesticides are more like a heavy rainstorm. The city gets wet and a little messy, but the fundamental structure of the buildings (the DNA blueprints) stays mostly the same. The changes were subtle and didn't rewrite the city's core rules.

5. The Double-Check: The Two-Layered Map

The researchers didn't just look at the "open doors" (accessibility); they also looked at the "sticky notes" on the blueprint (DNA methylation).

• The Analogy: Think of Accessibility as whether a door is open or closed. Think of Methylation as a sticky note on the door saying "Do Not Enter" or "Open for Business."
• The Finding: Usually, if a door is open, the sticky note says "Open." If it's closed, the note says "Do Not Enter."
• The Twist: In the formation of immune memory (how your body remembers a virus years later), these two layers work together perfectly. When a T-cell becomes a "Memory Cell" (a veteran soldier), the blueprint opens up for "memory instructions" (like BATF and AP-1), and the sticky notes are removed. It's a coordinated dance where the door opens and the warning sign is taken down simultaneously.

The Takeaway

This paper gives us a "Google Earth" view of how our immune system's internal software gets hacked or reprogrammed by different enemies.

• HIV tricks the soldiers into giving up (exhaustion).
• Severe COVID tricks the police into standing down and suppressing the body's own defenses (paralysis).
• Flu and Pesticides cause a ruckus but don't rewrite the core operating system.

By understanding exactly which doors are being locked and which sticky notes are being placed, scientists can hope to design drugs that force the doors back open or remove the wrong sticky notes, helping the immune system fight back effectively in the future.

Technical Summary

1. Problem Statement

Environmental and pathogen exposures (viruses like HIV-1, SARS-CoV-2, Influenza, and chemicals like organophosphates) induce profound remodeling of the gene-regulatory landscape in human immune cells. While transcriptomic changes (gene expression) in response to these exposures are well-documented, the underlying gene regulatory programs and epigenetic mechanisms (chromatin accessibility and DNA methylation) driving these responses remain poorly characterized at high resolution. Existing studies often lack:

• Comprehensive integration of multiple exposure types (viral vs. chemical).
• Multimodal single-cell profiling (combining chromatin accessibility and DNA methylation).
• Longitudinal insights into disease progression (e.g., HIV stages, COVID-19 severity).

2. Methodology

The authors employed a multimodal single-cell epigenomics approach to map the regulatory landscape of human immune cells across diverse exposure cohorts.

• Cohorts and Samples:

  • 92 PBMC samples from 4 distinct cohorts:
    1. SARS-CoV-2: Healthy controls, and mild, moderate, and severe COVID-19 cases.
    2. HIV-1: Longitudinal cohort (pre-infection, acute, chronic) with antiretroviral therapy (ART) intervention.
    3. Influenza A: Pre-challenge and post-challenge (3, 6, 28 days).
    4. Organophosphates (OP): Grouped by urinary TCPY levels (low, medium, high exposure).
  • Total Cells: 271,299 high-quality single nuclei retained after QC.
  • Regulatory Elements: 319,420 reproducible peaks identified as candidate cis-regulatory elements (cCREs).

• Sequencing and Assays:

  • snATAC-seq: Performed using 10x Genomics Chromium to profile chromatin accessibility.
  • snmC-seq2: Single-nucleus DNA methylation profiling for a subset of 39 samples (HIV and OP cohorts) to enable multimodal integration.
  • Integration: Canonical Correlation Analysis (CCA) was used to integrate snATAC-seq and snmC-seq data, transferring cell type annotations between modalities.

• Computational Analysis:

  • Clustering & Annotation: Graph-based clustering (Harmony for batch correction) identified 22 clusters mapped to 12 broad immune cell types (e.g., CD14+ monocytes, T cells, B cells, NK cells).
  • TF Activity Inference: Used ChromVAR for accessibility-based TF activity and a custom tool (methylTFR) to infer TF activity from DNA methylation footprints (calculating deviation scores relative to genomic background).
  • Differential Analysis: Pseudo-bulk differential accessibility (using ChrAccR) and differential methylation (using RnBeads) were performed to identify Differentially Accessible Regions (DARs) and Differentially Methylated Regions (DMRs).
  • Cross-Modality Correlation: Analyzed the relationship between chromatin accessibility and DNA methylation at TF binding sites (TFBS).

3. Key Contributions

1. First Multimodal Atlas: Creation of a comprehensive single-nucleus chromatin accessibility atlas of human immune cells across viral and chemical exposures, integrated with matched DNA methylation data.
2. Multimodal TF Activity Framework: Development and application of a method to infer transcription factor (TF) activity from DNA methylation footprints, allowing for a direct comparison with accessibility-based activity.
3. Exposure-Specific Signatures: Identification of distinct epigenetic signatures for HIV-induced exhaustion, SARS-CoV-2 severity, and chemical exposure, highlighting both shared stress responses and unique regulatory programs.

4. Key Results

A. HIV-1 Infection and T-Cell Exhaustion

• Progression: Observed a stepwise increase in exhausted CD8+ T cells (Tex) from pre-infection to acute and chronic stages, accompanied by CD4+ T cell depletion.
• Regulatory Mechanism: Exhausted T cells showed increased accessibility at FOXP family motifs (specifically FOXP2).
• Target Genes: The CTLA4 locus (a key exhaustion marker) exhibited increased accessibility enriched with FOXP motifs, suggesting FOXP family members drive CTLA4 expression in CD8+ Tex cells, similar to the role of FOXP3 in Tregs.

B. SARS-CoV-2 and CD14+ Monocyte Reprogramming

• Severity Correlation: Severe COVID-19 induced the most significant epigenetic remodeling, specifically in CD14+ classical monocytes.
• Regulatory Switch:
  • Hypo-accessibility: Decreased accessibility at NF-κB motifs (RELA, RELB) and SPI motifs.
  • Hyper-accessibility: Increased accessibility at AP-1 (FOS/JUN) and CREB motifs.
• Functional Consequence: This shift suggests a transition from acute inflammatory regulation (NF-κB) to stress-response and potential immune paralysis (AP-1).
• Gene Silencing: Decreased accessibility at the CD4 locus and HLA Class II genes (HLA-DRA, HLA-DPA1), indicating a shift toward a Myeloid-Derived Suppressor Cell (MDSC)-like state with impaired antigen presentation.
• Multimodal Concordance: In severe cases, DNA methylation changes were concordant with accessibility changes. For example, IRF1 motifs were hypermethylated (consistent with reduced expression), while AP-1 motifs showed hypomethylation despite open chromatin, suggesting complex regulatory constraints.

C. Multimodal Epigenetic Dynamics in T-Cell Memory

• Inverse Correlation: Generally, chromatin accessibility and DNA methylation showed an inverse relationship at TF binding sites (e.g., high accessibility = low methylation).
• Memory Formation: In the transition from naive to memory T cells:
  • BATF, AP-1, and ETS motifs showed increased accessibility and decreased methylation, indicating active regulatory roles in memory formation.
  • NF-κB factors were primarily regulated by methylation changes with minor accessibility shifts.
  • FOXP1 and IRF4 showed coordinated epigenetic changes across both modalities.
• Cell-Type Specificity: The anticorrelation between accessibility and methylation was strongest in B cells and naive T cells, but weaker in memory T cells, suggesting memory states involve more complex, layered regulation.

D. Chemical (Organophosphate) and Influenza Exposure

• Limited Remodeling: Compared to SARS-CoV-2 and HIV, OP and Influenza exposures induced fewer differentially accessible regions at the single-cell level.
• Shared Stress Response: A shared stress-response program driven by AP-1 motifs was observed in myeloid subsets across all exposure cohorts.
• Correlation: Influenza-induced chromatin changes showed a weak positive correlation with SARS-CoV-2 changes in monocytes, suggesting a shared innate immune signature restricted to specific cell types.

5. Significance

• Mechanistic Insight: The study moves beyond correlation to identify specific transcription factor networks (FOXP in HIV, NF-κB/AP-1 switch in COVID-19) that drive immune cell dysfunction or adaptation.
• Biomarker Potential: The identified epigenetic signatures (e.g., FOXP motif accessibility in CD8+ T cells, NF-κB silencing in monocytes) serve as potential biomarkers for disease severity (COVID-19) and therapeutic response (HIV ART).
• Multimodal Validation: By integrating DNA methylation, the study validates that chromatin accessibility changes are often accompanied by coordinated methylation dynamics, providing a more robust view of the "epigenetic code" governing immune responses.
• Resource: The authors provide a publicly available atlas (GEO: GSE306525) and open-source tools (methylTFR) for the community to dissect stimulus-specific transcriptional regulation.

In summary, this work establishes a framework for understanding how environmental and pathogenic stimuli rewire the immune epigenome, revealing that severe viral infections like COVID-19 and chronic infections like HIV drive distinct, cell-type-specific regulatory switches that can be decoded through multimodal single-cell profiling.