STING-STAT3-SOX18 Axis Drives EndMT and Epigenetic Reprogramming in SAVI Lung Fibrosis

This study reveals that gain-of-function STING1 mutations in SAVI patients drive pulmonary fibrosis by triggering a non-canonical cGAS-STING-STAT3 signaling axis that induces endothelial-to-mesenchymal transition through SLUG activation and epigenetic silencing of SOX18, independent of TGF-beta.

The Gist

The Big Picture: A Broken Alarm System in the Lungs

Imagine your body's immune system is like a high-tech security system for a building (your body). Its job is to detect intruders (viruses or bacteria) and sound the alarm to call the police (immune cells) to fight them off.

In a rare disease called SAVI (Stimulator of Interferon genes-Associated Vasculopathy with onset in Infancy), this security system has a broken sensor called STING. Instead of waiting for a real intruder, the sensor is stuck in the "ON" position. It keeps screaming "INTRUDER!" even when the building is empty. This causes constant, chaotic inflammation.

The big mystery this paper solves is: Why does this broken alarm cause the lungs to turn into scar tissue (fibrosis)?

The Discovery: The "Shape-Shifting" Cells

Usually, when lungs get injured, they try to heal. But in SAVI, the healing process goes wrong. The researchers discovered that the cells lining the tiny blood vessels in the lungs (Endothelial cells) are the ones getting confused.

Think of these blood vessel cells as peaceful gardeners. Their job is to tend to the "garden" (the lung tissue), keeping the soil (blood flow) healthy and the plants (air sacs) growing.

In SAVI patients, the broken STING alarm forces these gardeners to shape-shift. They stop being gardeners and turn into construction workers (mesenchymal cells).

• Gardeners build and maintain.
• Construction workers in this context are actually overzealous. They start laying down too much concrete and steel (scar tissue/collagen).
• The result? The soft, spongy lung turns into a hard, stiff block of concrete. This is fibrosis, and it makes it impossible to breathe.

The Mechanism: The "Switch" and the "Brake"

The researchers found exactly how this shape-shifting happens. It involves a chain reaction inside the cell:

1. The Broken Alarm (STING): The STING protein is hyperactive.
2. The Signal (STAT3): STING flips a switch inside the cell called STAT3. Usually, STAT3 helps with normal healing, but here it's stuck in the "GO" position.
3. The Construction Manager (SLUG): The overactive STAT3 turns on a "Construction Manager" protein called SLUG. SLUG tells the cell: "Stop gardening! Start building scars!"
4. The Missing Brake (SOX18): Every cell has a "Brake" pedal that keeps it in its original job. For blood vessel cells, this brake is a protein called SOX18.
   • In healthy people, SOX18 keeps the gardeners happy and working.
   • In SAVI, the broken STING alarm actually smashes the brake pedal. It silences SOX18. Without the brake, the cell has no choice but to become a scar-building construction worker.

The Analogy: Imagine driving a car (the cell). STING is the gas pedal stuck to the floor. SOX18 is the brake. In SAVI, the gas is floored, and someone has cut the brake lines. The car (the cell) speeds out of control and crashes into a wall (scar tissue).

Why Other Treatments Didn't Work (and What Will)

The paper tested existing drugs used for lung fibrosis (like Nintedanib and Pirfenidone).

• The Result: These drugs are like trying to fix a broken car by painting it. They try to slow down the construction workers, but they don't fix the fact that the gardeners have been turned into construction workers in the first place. In fact, one of the drugs made the problem worse!
• The Solution: The researchers tested a drug that specifically targets the broken alarm (STING inhibitor).
  • The Result: When they turned off the broken STING alarm, the "gas pedal" stopped being stuck. The brake (SOX18) was restored. The construction workers stopped building scars, and the cells started acting like gardeners again.

The Takeaway

This paper changes how we understand lung scarring in SAVI.

• Old View: We thought the problem was just too much inflammation or too many scar-builders.
• New View: The problem starts with the blood vessel cells losing their identity because a broken alarm (STING) broke their brakes (SOX18).

The Hope: This suggests that for patients with SAVI (and potentially other inflammatory lung diseases), the best treatment isn't just trying to stop the scarring. It's to fix the alarm system (inhibit STING). If we stop the alarm from screaming, the cells will remember how to be gardeners, and the lungs might stop turning into concrete.

Technical Summary

1. Problem Statement

• Clinical Context: Stimulator of Interferon Genes (STING)-Associated Vasculopathy with onset in Infancy (SAVI) is a rare autoinflammatory disease caused by gain-of-function (GOF) mutations in the STING1 gene. A hallmark of SAVI is severe, early-onset interstitial lung disease (ILD) and pulmonary fibrosis, leading to respiratory failure and early mortality.
• Knowledge Gap: While SAVI is known to be driven by type I interferon (IFN) signaling, the specific cellular mechanisms driving the unique pattern of lung fibrosis in SAVI remain unclear. Unlike Idiopathic Pulmonary Fibrosis (IPF), which features fibroblastic foci, SAVI lung pathology is characterized by prominent vascular injury and microvascular remodeling.
• Hypothesis: The authors hypothesized that constitutive STING activation in endothelial cells drives a specific form of Endothelial-to-Mesenchymal Transition (EndMT) independent of canonical TGF-β signaling, leading to the loss of endothelial identity and subsequent fibrosis.

2. Methodology

The study employed a multi-modal approach combining clinical histopathology, patient-derived cellular models, and multi-omics profiling:

• Clinical Cohort & Histopathology:
  • Analyzed lung biopsies and CT scans from 13 genetically confirmed SAVI patients.
  • Used multiplex immunofluorescence (CODEX) and standard IHC to characterize cellular composition, specifically looking for endothelial markers (CD31, VE-cadherin) vs. mesenchymal markers (SMA/ACTA2).
  • Compared SAVI pathology with IPF and healthy controls.
• In Vitro Disease Models:
  • iPSC-Derived Endothelial Cells (iECs): Generated iPSCs from 4 SAVI patients (harboring STING1 GOF mutations) and 4 healthy controls (HC).
  • Isogenic Controls: Used CRISPR/Cas9 to correct the STING1 mutation in two SAVI patient lines to create isogenic controls (iso-SAVI), ensuring observed phenotypes were mutation-specific.
  • Primary Cells: Used Human Lung Microvascular Endothelial Cells (HLMECs) treated with the STING agonist 2'3'-cGAMP and/or IFN-β to mimic acute and chronic activation.
• Multi-Omics Profiling:
  • Bulk RNA-seq: To identify transcriptional signatures (EndMT, IFN response) across passages (P1 to P5).
  • ATAC-seq: To map chromatin accessibility changes and identify epigenetic remodeling.
  • ChIP-seq Integration: Integrated public datasets for BRD4, SOX18, ERG, and AP-1 to validate motif enrichment.
• Functional Assays:
  • Signaling Inhibition: Tested inhibitors for STING (IFM35883), TBK1, STAT3, and JAK1/2 (Baricitinib).
  • Genetic Manipulation: Used shRNA knockdown (STAT1, STAT3, STING) and cDNA overexpression (SOX18) in SAVI iECs.
  • Phenotypic Analysis: Assessed tube formation, Ac-LDL uptake, and flow cytometry for endothelial/mesenchymal markers.
• Therapeutic Testing: Evaluated FDA-approved antifibrotics (Pirfenidone, Nintedanib) against STING pathway inhibitors.

3. Key Contributions & Findings

A. Identification of EndMT as a Primary Driver

• Histopathology: SAVI lung biopsies showed extensive microvascular damage, perivascular fibrosis, and a loss of endothelial cells (CD31+) with a concomitant gain of SMA+ cells. Crucially, SAVI lungs lacked fibroblastic foci, distinguishing them from IPF.
• Cell Intrinsic Phenotype: SAVI-derived iECs spontaneously underwent EndMT upon serial passaging (P3–P5), losing endothelial markers (CD31, VE-cadherin) and gaining mesenchymal markers (SMA, SM22). This was rescued in isogenic-corrected lines, confirming the mutation as the driver.
• Fibroblast Independence: Primary SAVI fibroblasts did not spontaneously activate or show increased myofibroblast differentiation upon STING stimulation, suggesting fibroblasts are not the primary initiators of the disease.

B. Discovery of a Non-Canonical STING-STAT3-SLUG Axis

• IRF3 Independence: Unlike canonical STING signaling which activates IRF3 and Type I IFNs, SAVI iECs showed IRF3-independent activation.
• STAT3 Dominance: While STAT1 (canonical) signaling was transient, STAT3 phosphorylation (pSTAT3) was constitutively elevated and nuclear in SAVI iECs.
• Mechanism: STING activation induces TBK1-mediated phosphorylation of STAT3. This leads to the upregulation of SLUG (SNAI2), a transcription factor that drives the mesenchymal program.
• Specificity: Knockdown of STAT3 (but not STAT1) partially preserved endothelial markers, confirming STAT3 as the pivotal driver of EndMT in this context.

C. Epigenetic Reprogramming and SOX18 Repression

• Chromatin Remodeling: ATAC-seq revealed massive chromatin closing (loss of accessibility) in SAVI iECs, particularly at regions enriched for endothelial maintenance motifs (AP-1, ETS, SOX).
• BRD4 Redistribution: Acute STING activation caused BRD4 (a chromatin reader) to redistribute from homeostatic endothelial enhancers to inflammatory loci.
• SOX18 Collapse: The most significant finding was the progressive repression of SOX18, a master regulator of endothelial identity.
  • SOX18 binding sites lost accessibility in SAVI iECs.
  • SOX18 protein levels dropped significantly.
  • Overexpression of SOX18 partially rescued the endothelial phenotype and upregulated downstream targets like ERG.
• Mechanism: The study proposes a model where acute STING activation triggers BRD4 redistribution and HDAC9 induction, leading to the epigenetic silencing of SOX18-dependent enhancers, locking cells into a mesenchymal state.

D. Therapeutic Implications

• Failure of Standard Antifibrotics: FDA-approved drugs Pirfenidone and Nintedanib failed to prevent EndMT in SAVI iECs; Nintedanib even accelerated the loss of endothelial markers.
• Success of Pathway Inhibition:
  • STING Inhibitors (IFM35883): Almost completely prevented spontaneous EndMT and preserved endothelial markers.
  • TBK1 and STAT3 Inhibitors: Also significantly attenuated the transition.
  • JAK Inhibitors (Baricitinib): Had limited efficacy, suggesting that blocking downstream IFN signaling is insufficient to stop the STING-STAT3-EndMT axis.

4. Significance

• Mechanistic Insight: The paper defines a novel, TGF-β-independent pathway for pulmonary fibrosis driven by endothelial dysfunction. It shifts the paradigm from "fibroblast-centric" (as in IPF) to "endothelial-centric" for SAVI-associated fibrosis.
• Molecular Axis: It elucidates the STING–STAT3–SLUG–SOX18 axis, revealing how chronic innate immune activation can epigenetically erase cell identity.
• Therapeutic Strategy: The findings suggest that current antifibrotic treatments are ineffective for SAVI because they do not target the root cause (STING-driven EndMT). Instead, direct STING or TBK1 inhibition represents a promising therapeutic strategy to preserve endothelial integrity and prevent fibrosis in SAVI and potentially other inflammatory lung diseases.
• Biomarker Potential: The specific loss of SOX18 and the presence of the EndMT signature could serve as biomarkers for disease progression and therapeutic response.

Conclusion

This study demonstrates that gain-of-function STING mutations drive pulmonary fibrosis in SAVI through a unique mechanism: the induction of Endothelial-to-Mesenchymal Transition (EndMT) via a non-canonical STING-STAT3-SLUG axis. This process is mediated by the epigenetic silencing of the endothelial master regulator SOX18. The research highlights that targeting the upstream STING pathway, rather than downstream fibrotic mediators, is essential for treating this specific form of inflammatory lung disease.