Aging/Research reports/2026-10-10
TIMP2: Three studies—how independent is the evidence?
Project status: 🔵 Level 2 · AG-C03 / AG-T01. Our synthesis still requires testing. Scientific evidence: preclinical literature; today’s task audits source dependencies. No new efficacy finding or independent experimental confirmation by this project.
The central finding
The TIMP2 literature contains different experiments. Their independence cannot be inferred from publication counts. Today we document specific connections that were previously unresolved. All three remain registry studies; this does not increase the established independent confirmation of the project mechanism.
Three connections across the sources
| Source pair | Documented connection |
|---|---|
| AG-R05 (2017) ↔ AG-R01 (2026) | Shared author Joseph Castellano; AG-R01 refers to the earlier treatment protocol. R05 R01 |
| AG-R05 (2017) ↔ AG-R08 (2023) | AG-R08 acknowledges Tony Wyss-Coray, an AG-R05 author. No shared author name was found in the inspected complete lists. R05 R08 |
| AG-R08 (2023) ↔ AG-R01 (2026) | AG-R01 acknowledges Eva Czirr, an AG-R08 author, for advice on microglial isolation/culture. No shared author name in the inspected lists. R08 R01 |
Record interests and materials separately: AG-R05 discloses Alkahest interests, AG-R08 Alkahest/Grifols employment, and AG-R01 a TIMP2 patent licensed to Alkahest. AG-R01 specifies recombinant mouse TIMP2 from R&D Systems; AG-R08 describes human TIMP2 constructs, including variants, produced by Proteos/Grifols. An identical protein batch is not established. R05: interests R08: methods/interests R01: methods/interests
How the conclusion changes
Our inference: We assess independence on four separate levels: research team, protocol, material and data. A shared author or technical advice establishes a connection; it does not establish reuse of the same animals or measurements. Conversely, a different author list alone does not prove fully independent reproduction. Commercial connections are disclosed and considered in appraisal; they do not replace scrutiny of the results.
The concrete decision is to retain three different testing contexts while not counting them as three confirmed independent reproductions of AG-T01. Without suitable experimental units, comparators and uncertainties, no pooled effect is calculated. An arbitrary numerical “independence discount” would be equally unjustified.
Mechanism and countermodel remain open
The previously reported functional comparison in AG-R05 is retained. The AG-R08 statistical limitation and unresolved degradation/export questions also remain. Direct neuronal or matrix effects can still account for part of the functional findings. Relationships between papers do not resolve these biological alternatives.
Established knowledge for this audit: the specifically documented bibliographic and methodological connections. Promising indications: earlier TIMP2 functional findings with their respective limitations. Experimental results: external mouse/cell findings, with no newly analysed measurements today. Hypothesis: selective harmful-cargo degradation with preserved synapses and improved function; AG-T01 stays at Level 2 with low mechanistic confidence. Persistent benefit after withdrawal, human efficacy and adequate safety are not established.
Next executable step
AG-R01 cites GEO GSE297916/GSE297917 as generated data. Data availability Next compare the public series metadata with the described groups and experimental units: which samples are independent, which measurements come from the same animals, and can treatment and genotype be assessed separately? The series themselves were not audited today. A reliable metadata comparison must precede any decision on a bounded analysis. Missing sample linkage must be recorded as non-identifiability.
Source status and teaching update
- AG-R05: Castellano et al., Nature; first published 19 April 2017, DOI 10.1038/nature22067.
- AG-R08: Britton et al., eNeuro; publisher date 15 June 2023, differing PMC heading 12 June 2023; first-online date still unresolved; DOI 10.1523/ENEURO.0031-23.2023.
- AG-R01: Hemmer, Philippi et al., Nature Communications; first published 12 August 2026, DOI 10.1038/s41467-026-74906-z.
- Retrieved/audited: 10 October 2026. Targeted original sections and indexed original text; no complete repeat of effect appraisal or raw-data analysis. Correction/retraction search found no attributable new notice; not comprehensive. Source delta 0: still 28 distinct studies. New adjacent search leads not yet included; search log.
- Evidence chain · Approaches registry · Professor research colloquium. AG-P17/18/19 cover dependencies, intervention ledgers and countermodels. No duplicate supplement; axolotl sample-linkage gaps remain open.
Continue exploring
