The Archive vs. The Engine

Most research into cognitive aging treats the problem as neurodegeneration—a simple loss of physical hardware. But if we look at it through an 'archival' lens, the aging brain might instead be suffering from a computational search cost. The hardware stays mostly intact, but the index of accumulated experience has grown so dense that the latency for new processing spikes. I suspect the molecular basis for this index isn't just in synaptic weights, but in the accumulation of 8-oxoguanine (8-oxoG) at transcriptionally active genomic loci.

8-oxoG as a High-Pass Filter

I'm viewing 8-oxoG as a biological 'mark' of past neuronal activity. In post-mitotic neurons, high-frequency transcriptional events generate local oxidative stress, which results in 8-oxoG lesions. These marks shouldn't be seen as a failure of repair; their persistence https://pmc.ncbi.nlm.nih.gov/articles/PMC3834072/ effectively creates a 'transcriptional log' of the neuron’s history.

The problem is that this log creates a physical bottleneck during Base Excision Repair (BER). As the 8-oxoG 'Trap' idea suggests, OGG1 binds to these lesions but often stalls because the OGG1/APE1 handover loses efficiency in older neurons. This stalling works like a high-pass filter. For a new transcriptional burst to happen at a locus the neuron has used before, the BER machinery has to navigate those pre-existing archival marks first. From this perspective, cognitive decline is just the macro-level result of OGG1-stalling latency at the promoter regions of plasticity-related genes.

The AP-Site Bottleneck

We've seen evidence that upregulating APE1 preserves synaptic function even when 8-oxoG levels remain high https://pmc.ncbi.nlm.nih.gov/articles/PMC4281843/. This suggests the neuron can handle the 'archive' (8-oxoG) itself, but it’s crippled by the 'access request'—the toxic abasic/AP site generated during the repair process. In an aged brain, the sheer volume of 8-oxoG marks means any attempt at new learning triggers a massive wave of AP site intermediates. The neuron likely slows its search query to keep these toxic intermediates from hitting the threshold that triggers apoptosis via MUTYH-mediated strand breaks https://www.jci.org/articles/view/65053/.

Selective Index Pruning

If decline is a maintenance cost rather than broken hardware, we should stop looking for universal neuroprotection and start looking at selective index pruning. If we can identify which 8-oxoG marks are just 'stale' associative data, we could use targeted BER enhancers—like locus-specific APE1 recruitment—to clear the search path without wiping the entire archival record. This goes against the 'more repair is better' dogma. We don't just need more repair; we need smarter repair that can tell the difference between active computational needs and historical logs.

Testing the Model

There are a few ways to put this to the test:

1. Transcriptional Latency Mapping: We can use optogenetically driven gene expression to measure the time-to-peak mRNA levels in young versus aged neurons. I’d expect aged neurons to show significant latency specifically at loci with high 8-oxoG density.
2. dOGG1 Intervention: We could introduce a catalytically 'dead' OGG1 (dOGG1) that occupies 8-oxoG sites without initiating repair. If the hypothesis is correct, dOGG1 should actually reduce latency by preventing the formation of those toxic AP-site bottlenecks. It might 'freeze' the archive, but it should improve current processing speed.
3. Locus-Specific Pruning: By using CRISPR-dCas9 fused to OGG1/APE1, we could selectively clear 8-oxoG from a specific memory circuit. If the circuit's plasticity returns without needing to lower the neuron's general oxidative load, we'll know we're on the right track.