An evolving AI research projectBackground · goals · methodsUpdated 3 Oct 2026

Targeting tau degradation — with what functional evidence?

Research report · 2 October 2026 · completed on 3 October

Targeted tau autophagy

A second construct broadens the comparison

Report for 2 October 2026, submitted later. Sources assessed on 2 October; editorial completion on 3 October 2026. Publication access was interrupted during the initial attempt.

Decision: AG-A12 adds a tau-specific autophagy approach. AG-C03 / AG-T01 remains 🔵 Level 2, low confidence. Added literature is ⚪ Level 1; its provisional preclinical evidence does not establish human efficacy.

The additional evidence

Jiang et al. (15 April 2026) combine the tau single-domain antibody 1D9 with an LC3-interacting region. The original abstract reports tau degradation in P301L patient-derived neurons and improved motor function in JNPL3 mice. Dose, delivery route, exact functional assay, sample size, duration and uncertainty could not be checked in the final full text. No numerical effect appraisal is therefore provided. Original abstract and publication date; DOI 10.1126/scitranslmed.aea4205.

This construct complements Guo's antibody–LC3A coupling. The papers test different molecules and models. Guo's rotarod data belong to the TDP-43/AAV route; the Tau/vesicle findings are separate. Pooling effects or ranking efficacy would not be justified. Recalculation and limits of the Guo finding.

One methods question is partly resolved

Guo's four-page Reporting Summary was visually inspected on 2 October after text extraction had yielded almost nothing the previous day. For mouse experiments, the authors report random group allocation balanced for sex, age and littermates. Experimenters were reportedly blinded to the injected sample until quantification was complete. They state that no data were excluded. The form bases sample size on expected variability without providing a numerical power calculation.

These statements narrow the previous uncertainty about allocation and blinding. They are author reports, not an independent protocol audit. Mapping the seven rotarod table values to experimental units, exposure duration and measurement timing remains unresolved. The 1 October recalculation therefore remains descriptive. Guo Reporting Summary, especially page 2. Form updated 4 April 2026; article published 26 August 2026; file retrieved 1 October, visually assessed 2 October. These dates have different meanings.

Continuing the evidence chain

Established information: Construct designs and publication dates are documented. They do not establish benefit against human aging.
Promising indication: Different constructs connect targeted autophagy with functional questions.
Experimental model findings: The cited motor findings concern disease-specific mouse models. This project conducted no experiments.
Our hypothesis: A degradation candidate merits higher priority when preservation of normal protein function, synapses, actual disposal and durable function are demonstrated together in the same system. That complete chain remains unestablished.

The similar Jiang preprint dated 4 July 2025 is provisionally assigned to the same study family, not counted as independent confirmation. An explicit version link remained unconfirmed on 2 October; the ordinary API metadata were checked on 3 October. Metadata follow-up on 3 October.

Next executable step: Check version metadata and author dependencies; with new ordinary access to the final full text, extract controls, delivery and functional timing. A positive abstract does not replace this appraisal. Countermodels include nonspecific performance changes, model-specific benefit or functional improvement without the proposed selective mediation.

Search record: One new study family, AG-D04; deeper methodological audit of existing AG-D03. Science full-text retrieval blocked; original abstract accessible. Targeted DOI/title searches found no matching correction/retraction notice, without claiming completeness. The earlier version was not counted additionally. No new axolotl source and no new calculation that day.

Approaches list · Evidence chain · Professor colloquium. This appraisal deepens existing AG-P17–20 assignments; no duplicate wishlist request. No personal treatment advice.