Human promoter analysis of the Programmed Axon Death genes NMNAT2 and SARM1

This study functionally characterizes the human promoters of the axon death genes NMNAT2 and SARM1, revealing that specific naturally occurring and rare single-nucleotide variants significantly alter their transcriptional activity in ways that may increase susceptibility to neurodegenerative diseases.

The Gist

Imagine your body's nerves are like a vast network of electrical cables stretching from your brain to your toes. These cables (axons) are essential for sending messages, but sometimes, they get damaged and need to be cut off to save the rest of the system. This is a controlled process called Programmed Axon Death.

Two main characters control this process:

1. NMNAT2 (The Guardian): A protein that acts like a shield, keeping the "cut-off" switch turned off.
2. SARM1 (The Executioner): A protein that acts as the switch. If it gets turned on, it destroys the nerve cable.

Normally, the Guardian (NMNAT2) keeps the Executioner (SARM1) locked in a cage. But if the Guardian's numbers drop too low, the Executioner breaks free, and the nerve dies. This mechanism is linked to diseases like ALS (Lou Gehrig's disease) and various neuropathies.

The Big Question:
Scientists knew that mutations in the coding parts of these genes (the blueprints for building the proteins) could cause disease. But they didn't know much about the promoters.

Think of a gene as a song. The "coding" part is the melody. The promoter is the volume knob and the on/off switch. If the volume knob is broken, the song might be too quiet (not enough Guardian) or too loud (too much Executioner), even if the melody itself is perfect.

This paper investigates the "volume knobs" (promoters) of the human NMNAT2 and SARM1 genes to see if tiny changes there could make people more vulnerable to nerve diseases.

The Main Discoveries

1. The Human "Volume Knob" is Different from the Mouse's

Scientists previously studied mice and found their NMNAT2 gene had two switches (called CREs) that responded to stress signals (cAMP).

• The Finding: When they looked at humans, they found we only have one working switch (CRE2). The second one is broken in humans.
• The Analogy: Imagine a mouse has a car with two gas pedals. If you press either one, the car speeds up. Humans, however, only have one gas pedal. If you break that one, the car won't move at all.

2. Tiny Typos Can Break the Shield

The researchers looked at the "volume knob" of the human Guardian gene (NMNAT2) and found tiny typos (single-letter changes in the DNA) that exist naturally in people.

• The Result: Some of these typos turned the volume knob down so much that the Guardian protein became 50% weaker.
• The Consequence: If you have one of these typos, your nerve cables are less protected. If you also face an injury or a toxic drug (like chemotherapy), your nerves might die much faster than someone without the typo.
• Real-World Case: They found one specific typo in a patient with ALS. This typo made the Guardian gene work at less than half its normal strength, suggesting it helped cause the disease.

3. The Executioner's Switch is Too Loud in Some People

They also looked at the Executioner gene (SARM1).

• The Finding: They found a common typo in the SARM1 promoter that acts like a stuck "High Volume" button. It makes the Executioner protein slightly more active than usual.
• The Twist: This typo is actually quite common in the general population (about 8% of people have it), and it wasn't found more often in ALS patients than in healthy people.
• The Takeaway: While this specific typo doesn't seem to cause ALS on its own, it shows that small changes in the "volume knob" can change how much of these dangerous proteins are made. This could be a hidden factor in why some people get nerve diseases and others don't.

Why This Matters

Think of your nervous system as a house with a fire alarm (SARM1) and a fire extinguisher (NMNAT2).

• If the fire extinguisher is broken (low NMNAT2), a small spark can burn the house down.
• If the fire alarm is too sensitive (high SARM1), it might go off for no reason.

This paper shows that we don't just need to worry about broken fire extinguishers (coding mutations). We also need to worry about broken volume knobs (promoter mutations).

The Bottom Line:
By mapping out these "volume knobs," scientists have found new ways that our genes can go wrong. This helps explain why some people are more fragile when their nerves are under attack. It also opens the door for new treatments: instead of just trying to fix the broken protein, doctors might one day be able to turn up the volume on the Guardian gene or turn down the volume on the Executioner gene to save our nerves.

Technical Summary

1. Problem Statement

Programmed Axon Death (PAxD) is a critical pathway in neurodegenerative diseases, regulated primarily by the balance between NMNAT2 (a neuroprotective enzyme converting NMN to NAD) and SARM1 (an NAD-degrading enzyme that executes axon death). While coding variants in these genes are known to cause rare, severe axonal disorders, the contribution of non-coding regulatory variants (specifically in promoter regions) to susceptibility in common neurodegenerative diseases (like ALS and peripheral neuropathies) remains poorly understood.

The study addresses two main gaps:

1. The lack of functional characterization of human NMNAT2 and SARM1 promoter regions, particularly regarding how they differ from mouse models.
2. The potential impact of naturally occurring single-nucleotide variants (SNVs) in these promoters on gene expression levels and subsequent axon vulnerability.

2. Methodology

The authors employed a combination of in silico analysis, molecular cloning, and functional cell-based assays:

• Cell Models: Used SH-SY5Y (human neuroblastoma) and HEK293T (human embryonic kidney) cell lines to test promoter activity.
• Luciferase Reporter Assays:
  • Cloned various lengths of the 5' upstream promoter regions of human NMNAT2 (ranging from -170 bp to -2528 bp) and SARM1 (ranging from -286 bp to -2501 bp) into the pGL4.10 vector.
  • Transfected constructs into cells and measured firefly luciferase activity normalized to a NanoLuc control.
• Site-Directed Mutagenesis:
  • Generated mutations in predicted regulatory elements (e.g., cAMP response elements) and naturally occurring variants found in population databases.
• cAMP Stimulation: Treated cells with dibutyryl cAMP (dbcAMP) to test responsiveness of the NMNAT2 promoter to cAMP signaling.
• qRT-PCR: Validated endogenous NMNAT2 mRNA expression changes in SH-SY5Y cells following dbcAMP treatment.
• Variant Screening:
  • Queried gnomAD v4.1.0 for variants in the NMNAT2 CRE2 region.
  • Analyzed Project MinE (ALS database, 4,366 cases, 1,832 controls) for variants in both NMNAT2 and SARM1 promoter regions.
• Bioinformatics: Used Clustal Omega for sequence alignment, MEME Suite for de novo motif discovery, and STAMP for transcription factor binding site prediction.

3. Key Contributions & Results

A. Human NMNAT2 Promoter Architecture

• Minimal Promoter Region: The first 286 bp upstream of the ATG start codon drives maximal promoter activity in human cells. Extending the region further upstream (-286 to -2528 bp) results in repressive activity, but only in SH-SY5Y cells (not HEK293T), suggesting cell-type-specific repressors.
• cAMP Response: Unlike mice, which have two cAMP response elements (CREs), the human promoter relies on a single, evolutionarily conserved CRE2.
  • Mutating CRE2 reduced promoter activity by ~70% and abolished the response to dbcAMP.
  • Mutating the non-conserved CRE1 had no effect.
  • dbcAMP treatment increased endogenous NMNAT2 mRNA by ~1.5-fold, confirming cAMP regulation.
• Other Regulatory Elements: Computational analysis identified potential motifs (MEME1–3) in the -286 to -170 bp region, but functional mutagenesis showed these do not act as discrete, essential drivers of maximal activity, suggesting they may act synergistically or are less critical than CRE2.

B. Impact of Natural Variants on NMNAT2

• gnomAD Variants: Eight rare SNVs were found in the human CRE2 region. Functional assays revealed that four of these variants (GNO4, GNO5, GNO6, GNO7) significantly reduced promoter activity (some by >50%). Three of these lie directly within the core transcription factor binding site (TGACG).
• ALS Patient Variant (Project MinE): A specific ultra-rare variant, PMV6, identified in an ALS patient, was found to reduce NMNAT2 promoter activity by ~55% (from 11.0 to 4.9 relative units). This suggests a mechanism where reduced NMNAT2 expression could lower the threshold for PAxD activation in disease.

C. Human SARM1 Promoter Architecture

• Minimal Promoter Region: Maximal SARM1 promoter activity requires 1093 bp upstream of the ATG. Extending to -2501 bp did not further increase activity.
• Common Variant Effect: A relatively common variant (PMV13, allele frequency ~0.08) found in both controls and ALS patients was tested.
  • PMV13 caused a modest but significant increase (~40%) in promoter activity.
  • While not enriched in ALS cases, its high frequency suggests it could modify disease susceptibility in a broader population or in other neurodegenerative contexts.

4. Significance and Implications

• Refined Regulatory Mechanisms: The study establishes that human NMNAT2 regulation differs from mouse models (one vs. two CREs) and identifies a specific repressive region (-286 to -2528 bp) that is cell-type dependent.
• Expansion of Disease Genetics: It demonstrates that non-coding promoter variants can significantly alter the expression of PAxD genes.
  • Loss-of-function in Promoter: Variants like PMV6 in NMNAT2 mimic the effect of coding loss-of-function mutations, potentially contributing to ALS or neuropathy susceptibility by lowering the neuroprotective threshold.
  • Gain-of-function in Promoter: Variants like PMV13 in SARM1 could increase the "death signal," potentially acting as a risk modifier.
• Therapeutic Targets: Understanding that cAMP signaling upregulates NMNAT2 via CRE2 offers a potential therapeutic avenue (e.g., enhancing cAMP signaling) to boost neuroprotection.
• Future Association Studies: The identified variants provide a new catalogue of functional alleles for future genome-wide association studies (GWAS) in neurodegenerative diseases, moving beyond coding regions to include regulatory variants that influence gene dosage.

Conclusion

This paper provides the first functional characterization of human NMNAT2 and SARM1 promoters. It reveals that naturally occurring SNVs in these regions can drastically alter transcription levels in directions that favor axon degeneration, highlighting the importance of regulatory variants in the etiology and susceptibility of neurodegenerative diseases.