TWNK Substrate Context Separates RNA:DNA Exchange from Replication Claims

**Background:** TWNK encodes the mitochondrial replicative helicase Twinkle, also known historically as C10orf2 or PEO1. Recent public primary studies span RNA:DNA binding and annealing, DNA strand exchange, real-time DNA unwinding and rewinding, and mouse D-loop replication endpoints. Treating those endpoints as interchangeable can overstate what purified substrate assays show. **Objective:** This working paper asks whether public TWNK/Twinkle primary studies separate RNA:DNA binding and annealing/strand-exchange evidence from canonical DNA-unwinding, real-time pause/rewinding, and in vivo D-loop replication endpoints strongly enough to prevent substrate-context overreach. **Methods:** Four archived full-text primary records and one current single-molecule comparator were source-coded. Endpoint classes were RNA:DNA binding/annealing/exchange, DNA strand exchange/unwinding, real-time loading/unwinding/pause/rewinding, mouse D-loop/complete mtDNA replication, and smFRET DNA-unwinding context. A portable pilot counted endpoint vocabulary in the four archived primary texts. Quantitative summaries report source-study denominators, repository-file counts and vocabulary counts only; no pooled biological effect, sampling confidence interval or clinical estimate was performed. **Results:** Five public primary records were included. The pilot over four archived texts counted 398 RNA terms, 527 DNA-unwinding terms, 153 annealing terms, 173 strand-exchange terms, 73 replication-synthesis terms and 109 single-molecule terms. The Plaza 2025 repository resolved to three public bitstreams: two analysis notebooks and one source-data workbook. The endpoint matrix remained separable across RNA:DNA biochemical activity, DNA exchange/unwinding, optical-tweezers kinetics and mouse replication endpoints. Missing same-condition RNA:DNA exchange plus cellular mtDNA replication/deletion readouts remained unavailable rather than inferred. **Conclusions:** Public TWNK evidence supports a substrate-context claim ceiling, not a validated new repair or treatment mechanism. RNA:DNA biochemical findings can motivate follow-on replication-fork and compartment tests, but they should not be cited as direct evidence of cellular repair, mtDNA-deletion prevention, tissue rescue or patient-relevant benefit without paired functional and molecular readouts under the same TWNK perturbation. AI tools assisted literature-search triage, source coding, manuscript drafting, figure generation, formatting and QA preparation. Andrew Friedl remains solely responsible for the content.