Mitochondrially Transcribed dsRNA Mediates Manganese-induced Neuroinflammation

This study reveals that excess manganese induces neuroinflammation by disrupting mitochondrial transcriptome processing, leading to the accumulation of double-stranded RNA that activates cytosolic MDA5 sensors and triggers type I interferon responses in both human organoids and mouse models of hypermanganesemia.

The Gist

The Big Picture: A Toxic Metal and a "False Alarm"

Imagine your brain is a bustling, high-tech city. Inside every building (cell) in this city, there is a power plant called the mitochondria. These power plants run on fuel to keep the city lights on and the traffic moving.

Manganese is a metal that is actually essential for the city to function—it's like a specific type of oil needed for the engines. But, just like oil, if you spill too much of it, it causes a massive mess. When there is too much manganese in the brain (a condition called hypermanganesemia), it causes a disease that looks like Parkinson's, leading to tremors, stiffness, and confusion.

For a long time, scientists knew manganese was toxic, but they didn't know exactly how it broke the brain. This paper reveals a surprising new mechanism: Manganese tricks the brain's immune system into thinking it's under viral attack.

---

The Story: How the "False Alarm" Happens

Here is the step-by-step breakdown of what the researchers discovered:

1. The Glitch in the Power Plant

Inside the mitochondria, there is a complex assembly line that reads blueprints (RNA) to build parts for the power plant.

• The Problem: When too much manganese is present, it jams this assembly line. It's like pouring sand into a printer; the machine starts spitting out crumpled, half-finished pages.
• The Result: These crumpled pages are double-stranded RNA (dsRNA). In a healthy cell, these are quickly shredded and thrown away. But with manganese, they pile up.

2. The "Leak"

Usually, the power plant is a sealed room. But because of the manganese-induced mess, these crumpled RNA pages start leaking out of the mitochondria and into the main hallway of the cell (the cytoplasm).

3. The Security Guard Gets Confused

The cell has security guards (proteins like MDA5) whose job is to patrol the hallway. Their only job is to look for dsRNA.

• The Logic: In nature, cells only have dsRNA when a virus (like the flu) has invaded. So, the security guard's rule is simple: "If I see dsRNA, it's a virus! Sound the alarm!"
• The Mistake: The guard sees the manganese-induced RNA leak and thinks, "Virus! Virus!" even though there is no virus.

4. The False Alarm (Inflammation)

The guard sounds the Type I Interferon alarm. This is the cell's "Code Red" defense system.

• The cell starts pumping out cytokines (chemical messengers that call for backup).
• This causes neuroinflammation: the brain's immune cells (astrocytes and microglia) get angry, swell up, and start attacking the neighborhood.
• The Tragedy: The brain isn't fighting a virus; it's fighting itself. This chronic inflammation damages neurons, leading to the movement problems and cognitive issues seen in manganese poisoning.

---

The Evidence: From Test Tubes to Mouse Brains

The researchers didn't just guess; they tested this in three different "cities":

1. The Test Tube City (Cell Lines): They put manganese in human cells. They saw the RNA pile up, the security guards get confused, and the inflammatory alarm go off.
2. The Mini-Brain City (Organoids): They grew tiny, 3D human brains in a dish (cerebral organoids). When they added manganese, the "adult" cells in the mini-brain (specifically astrocytes, which are like the brain's support staff) started screaming with inflammation. Interestingly, the younger cells didn't react as strongly, suggesting the damage happens as the brain matures.
3. The Real City (Mouse Models): They looked at mice that have a genetic defect preventing them from getting rid of manganese (similar to a human genetic disease). These mice had high manganese levels, and sure enough, their brains were full of the RNA leaks and the inflammatory alarm was blaring.

Why This Matters

This discovery changes how we view manganese poisoning. It's not just that the metal poisons the cells directly; it's that the metal hijacks the immune system.

• The Analogy: It's like a smoke detector that goes off because someone burned toast (manganese), not because the house is on fire (virus). But because the alarm keeps screaming, the fire department (immune system) arrives, sprays water everywhere, and accidentally floods the house, causing more damage than the burnt toast ever would have.

The Takeaway

This paper suggests that for people suffering from manganese toxicity (whether from mining accidents, welding fumes, or genetic diseases), the treatment might not just be about removing the metal. We might also need to calm down the false alarm.

If we can teach the cell's security guards to ignore the manganese-induced RNA leaks, or if we can block the "alarm bell" (the interferon response), we might be able to stop the brain inflammation and protect patients from the devastating effects of this neurotoxin.

Technical Summary

1. Problem Statement

Manganese (Mn) is an essential trace element, but excess accumulation (hypermanganesemia) due to environmental exposure or genetic defects (e.g., mutations in SLC30A10 or SLC39A14) causes severe neurotoxicity known as manganism. This condition is characterized by astrogliosis, neuronal loss, and neuroinflammation, often mimicking Parkinson's disease and dystonia.

• The Gap: While the neuropathological features are well-documented, the molecular mechanisms linking mitochondrial dysfunction to the specific inflammatory cascade in manganese toxicity remain poorly understood.
• The Hypothesis: The authors hypothesized that manganese disrupts mitochondrial transcriptome processing, leading to the accumulation of double-stranded RNA (dsRNA) in mitochondria. This dsRNA escapes into the cytoplasm, activating innate immune sensors (specifically MDA5) and triggering a Type I Interferon (IFN) response, thereby driving neuroinflammation.

2. Methodology

The study employed a multi-modal approach combining cell biology, human organoid models, and in vivo mouse genetics:

• Cell Lines: Used HAP1 (wild-type and SLC30A10 knockout) and HeLa cells to model acute manganese exposure.
• Human Cerebral Organoids: Differentiated human iPSCs into 100-day-old unguided cerebral organoids to model complex brain tissue architecture and developmental stages (neuroprogenitors, neurons, astrocytes).
• In Vivo Model: Utilized Slc30a10⁻/⁻ mice (global knockout), a genetic model of chronic hypermanganesemia, comparing them to wild-type littermates at 8 weeks of age.
• Key Techniques:
  • MitoString (NanoString): Custom panels to quantify mitochondrial RNA processing intermediates (tRNA-mRNA junctions) and non-coding transcripts from both Heavy (H) and Light (L) strands.
  • Immunofluorescence: Used the J2 monoclonal antibody (specific for dsRNA >40bp) to detect dsRNA, co-stained with mitochondrial markers (GRSF1, Tom20) and cell type markers (GFAP, Nissl).
  • Reporter Assays: Luciferase reporters driven by the IFN-$\beta$ promoter to measure antiviral signaling.
  • Pharmacological & Genetic Interventions: Used inhibitors (IMT1 for POLRMT, Cyclosporine A for mPTP, V5 peptide for Bax-Bak, GSK8612 for TBK1, H151 for STING) and siRNA knockdowns (MDA5, RIG1) to dissect the signaling pathway.
  • Single-Cell RNA Sequencing (scRNAseq): Performed on 100-day organoids to identify cell-type-specific responses.
  • Cytokine Profiling: Luminex multiplex assays and ELISA to measure secreted inflammatory mediators (e.g., IL-8, CXCL10).
  • Metal Quantification: ICP-MS to measure Mn, Fe, Zn, and Ca levels.

3. Key Contributions

• Novel Mechanism: Identified mitochondrial dsRNA accumulation as the primary trigger for manganese-induced neuroinflammation, distinct from oxidative stress or direct cytotoxicity.
• Pathway Elucidation: Mapped the specific signaling axis: Mn $\rightarrow$ Mitochondrial Transcriptome Dysregulation $\rightarrow$ dsRNA Accumulation $\rightarrow$ Cytoplasmic Release $\rightarrow$ MDA5 Sensing $\rightarrow$ TBK1/IRF3 Activation $\rightarrow$ Type I IFN Response.
• Cell-Type Specificity: Demonstrated that astrocytes are the primary drivers of this inflammatory response in the brain, while neurons exhibit stress responses but lack the robust pro-inflammatory IFN signature.
• In Vivo Validation: Confirmed that this mechanism operates in a genetic model of human hypermanganesemia (Slc30a10⁻/⁻ mice), linking the in vitro findings to human disease pathology.

4. Key Results

A. Manganese Induces Mitochondrial dsRNA Accumulation

• Transcriptome Analysis: Mn exposure (and SLC30A10 deficiency) caused a specific accumulation of non-coding RNA sequences from the mitochondrial Light (L) strand, which are normally rapidly degraded. This indicates a failure in mitochondrial RNA processing.
• dsRNA Detection: Immunofluorescence confirmed increased dsRNA co-localizing with mitochondrial RNA granules (GRSF1) in Mn-treated cells. This signal was abolished by inhibiting mitochondrial transcription (POLRMT) or degrading dsRNA with RNase III.
• Specificity: The accumulation was specific to Mn; levels of Fe, Zn, and Ca remained unchanged.

B. Activation of the Type I Interferon Pathway

• Sensor Identification: In HeLa cells, Mn induced IFN-$\beta$ promoter activity. Knockdown experiments revealed that MDA5 (Melanoma Differentiation-Associated protein 5), but not RIG-I, is the critical cytosolic sensor for Mn-induced mitochondrial dsRNA.
• Signaling Cascade: The pathway requires dsRNA export from mitochondria (blocked by Cyclosporine A or V5 peptide) and activation of TBK1.
• Exclusion of mtDNA: Inhibition of the cGAS-STING pathway (H151) did not block Mn-induced IFN signaling, ruling out mitochondrial DNA as the trigger.

C. Inflammatory Cytokine Production

• In Vitro: Mn treatment induced the secretion of pro-inflammatory cytokines, specifically IL-8 (CXCL8) and CXCL10, in HeLa cells and human organoids.
• Independence from Cytotoxicity: Inhibiting the dsRNA pathway (e.g., with TBK1 inhibitors) reduced cytokine production without affecting cell survival, proving that inflammation is a distinct signaling event from cell death.

D. Human Cerebral Organoids and Cell-Type Specificity

• Developmental Timing: Mn-induced mitochondrial defects and inflammation were only observed in 100-day-old organoids (mature astrocytes), not in younger (30 or 60-day) organoids.
• scRNAseq Findings:
  • Astrocytes: A specific subpopulation of astrocytes (Cluster 2) drove the inflammatory response, upregulating IFN-stimulated genes (ISGs), CXCL8, and CXCL10. These cells showed reduced expression of RNA processing enzymes (HSD17B10, REXO2).
  • Neurons: Neurons upregulated stress response genes (ATF3, ATF5) but did not upregulate pro-inflammatory or IFN genes, suggesting a protective mechanism (possibly via ADAR editing or SOCS1).
• Immunofluorescence: Confirmed that ~36% of GFAP+ astrocytes in Mn-treated organoids accumulated mitochondrial dsRNA.

E. In Vivo Validation (Slc30a10⁻/⁻ Mice)

• dsRNA in Brain: Slc30a10⁻/⁻ mice showed increased mitochondrial dsRNA in the globus pallidus, co-localizing with both GFAP+ astrocytes and Nissl+ neurons.
• Transcriptomic & Proteomic Changes:
  • Brain cortex showed upregulation of ISGs, complement component C3, and astrocyte reactivity markers.
  • Cytokine profiling (Luminex) revealed a complex, sex-dependent inflammatory signature, with males showing stronger upregulation of cytokines (e.g., IL-12b, IL-6) than females.
  • Discriminant Partial Least Squares (D-PLSR) analysis confirmed that the cytokine profile strongly distinguished mutant from wild-type animals.

5. Significance

• Paradigm Shift: This study redefines manganese neurotoxicity not just as a metabolic failure or oxidative stress event, but as an autoimmune-like response triggered by endogenous mitochondrial dsRNA.
• Therapeutic Implications: The findings suggest that targeting the dsRNA-MDA5-TBK1 axis (e.g., with TBK1 inhibitors) could mitigate neuroinflammation and neurodegeneration in manganese toxicity without necessarily needing to chelate the metal immediately.
• Broader Relevance: The mechanism (mitochondrial dsRNA triggering IFN responses) may be a common pathway in other neurodegenerative diseases caused by environmental toxins or genetic defects affecting mitochondrial RNA processing.
• Astrocyte-Centric Pathology: It highlights the critical role of astrocytes as the primary initiators of neuroinflammation in manganese exposure, challenging previous views that focused heavily on microglia or direct neuronal toxicity.

In conclusion, the paper establishes a direct causal link between manganese-induced mitochondrial RNA processing defects, dsRNA accumulation, and Type I interferon-mediated neuroinflammation, providing a molecular basis for the pathogenesis of manganism.