Hypothesis: Long, hypomethylated cfDNA fragments drive inflamm‑aging via TLR9 activation

Hypothesis

Aging shifts cfDNA toward longer, hypomethylated fragments that act as potent TLR9 ligands, creating a feed‑forward loop that amplifies inflamm‑aging.

Mechanistic reasoning

1. Methylation governs nuclease access – Low CpG methylation increases DNASE1L3 cutting efficiency, altering fragment size and end motifs [5]. In older plasma, global hypomethylation at ~2 000 age‑related DMCs [1] therefore favors generation of longer cfDNA pieces.
2. Chromatin remodeling enlarges protected fragments – Nucleosome repeat length (NRL) rises with age, moving signals from heterochromatin to euchromatin [4]. Euchromatin is less methylated and more accessible, so the nuclease‑protected DNA surrounding these nucleosomes retains hypomethylated CpGs.
3. Size‑selected fragments retain immunostimulatory motifs – cfDNA fragments >175 bp, enriched in elderly samples [3], are sufficient to contain multiple CpG dinucleotides. Hypomethylated CpGs within these fragments are optimal TLR9 agonists, whereas methylated CpGs blunt signaling.
4. TLR9 signaling fuels inflamm‑aging – Plasma cfDNA activates TLR9 on monocytes and macrophages, driving NF‑κB–dependent transcription of MCP‑1, IL‑6 and other inflamm‑aging mediators [2]. The observed correlation between hypomethylation and elevated MCP‑1 may thus reflect direct TLR9 stimulation rather than a passive biomarker.

Testable predictions

• Prediction 1: Size‑fractionated cfDNA from individuals >65 yr will show a higher proportion of hypomethylated CpGs in the >175 bp fraction compared with <175 bp fractions, whereas young adults (<30 yr) will display the opposite pattern.
• Prediction 2: Synthetic cfDNA oligonucleotides mimicking the long, hypomethylated profile (>175 bp, ≤20 % methylation) will trigger significantly greater NF‑κB reporter activity in human PBMCs via TLR9 than methylated or short fragments; blockade with TLR9 antagonist or DNase1L3 inhibition will abolish this effect.
• Prediction 3: In aged mice, genetic or pharmacological reduction of DNASE1L3 activity will decrease long cfDNA fragments, lower serum MCP‑1/IL‑6 levels, and improve frailty scores without altering total cfDNA concentration.
• Prediction 4: Administration of a TLR9 antagonist to old mice will attenuate inflamm‑aging biomarkers (MCP‑1, IL‑6, tissue macrophages) even when long cfDNA levels remain high, uncoupling fragment abundance from downstream inflammation.

Experimental approach

1. Human plasma stratification – Collect plasma from age‑stratified cohorts (n≥50 per group). Use size‑exclusion chromatography to isolate <175 bp and >175 bp fractions. Perform EM‑seq on each fraction to quantify methylation at the 2 000 age‑related DMCs [1].
2. In vitro immunomodulation – Isolate human monocytes, expose them to equal ng amounts of each fraction, measure NF‑κB luciferase phosphorylation and cytokine secretion. Include TLR9‑KO or inhibitor controls.
3. Mouse intervention – Employ DNASE1L3 heterozygous knock‑down or small‑molecule inhibitor in 20‑month‑old mice. Serial blood draws assess cfDNA size distribution (Bioanalyzer) and methylation (targeted bisulfite sequencing). Quantify plasma cytokines and perform frailty indexing.
4. Pharmacologic blockade – Treat a parallel set of aged mice with the TLR9 antagonist ODN 2088; compare inflamm‑aging readouts to DNASE1L3‑reduced group.

Falsifiability

If long cfDNA fractions do not exhibit greater hypomethylation than short fractions, or if synthetic long hypomethylated cfDNA fails to activate TLR9-dependent pathways, the core mechanistic link is refuted. Similarly, if DNASE1L3 reduction does not shift fragment size or lower inflamm‑aging markers, the causal role of nuclease accessibility in generating immunostimulatory cfDNA is unsupported.

This hypothesis integrates methylation, fragmentomics, and innate immunity into a single, testable model of how cfDNA contributes to the inflamm‑aging phenotype.