Body size dictates which anti-aging strategy evolution selects: mammals over 5-10 kg convergently repress telomerase as cancer defense, while smaller long-lived species keep it active.
New literature review: https://zenodo.org/records/21843358
Three example findings from this paper that I think are worth your time:
1. There's a body mass threshold that determines how species solve the longevity-cancer trade-off. Mammals exceeding 5-10 kg, like humans, elephants, whales convergently repress somatic telomerase and maintain shorter telomeres, using replicative senescence as an anti-cancer mechanism. Small long-lived species like bats and naked mole rats do the opposite: they keep telomerase active and rely instead on robust early-acting tumor suppression (Rb/p53) to prevent hyperplasia. The sharpest version of this: across 15 rodent species, telomerase activity coevolves with body mass, not lifespan. This isn't "telomeres matter for aging": it's that telomere strategy is body-size-dependent, and interventions need to account for which side of that threshold the target species sits on. (Tian et al. 2018; Seluanov et al. 2007)
2. Transposon silencing is a convergent longevity mechanism across three separate lineages and in one case it's causal, not correlative. The piRNA/PIWI pathway suppresses transposable elements, and independently evolved long-lived organisms converge on strengthening it:
- In C. elegans, downregulating active TE families extends lifespan, and ectopic activation of Piwi in somatic cells promotes longevity (Sturm et al. 2023)
- In 20-year-old Macrotermes natalensis termite queens, TE expression does not rise with age at all despite a sevenfold increase in overall gene expression because the piRNA pathway is upregulated with age (Post et al. 2022)
- In Drosophila, the adult fat body runs a somatic piRNA pathway; Piwi mutants show TE mobilization, elevated DNA damage, and reduced lipid stores (Jones et al. 2016)
Three species, three independent labs, and the C. elegans result is an intervention rather than an observation which is what makes this testable rather than merely suggestive.
3. The primary bottleneck for longevity therapeutics isn't finding targets, it's reaching them. The paper calls this a "delivery-before-discovery paradox": dual-AAV prime editing achieves 42% efficiency in mouse cortex and 46% in liver, but only 11% in cardiac tissue. We already have validated longevity targets we can't reach in the tissues where aging hits hardest. Heart delivery, not target identification, is the rate-limiting step and it directly constrains which mechanisms are therapeutically actionable at all.
This entire paper was generated autonomously by Thinktica, an AI platform that traces every claim to its source to avoid hallucinations. It came from a single prompt:
"What are the convergent genetic and epigenetic mechanisms shared across independently evolved long-lived species (>200 years), and which of these mechanisms have human orthologs that are druggable or editable with existing technologies (CRISPR, mRNA, small molecules)?"
84 references, 36 pages, produced in a 48-hour research run.
-----------
One limitation I want to state up front:
This version doesn't separately label hypothesis-derived reasoning from source-derived reasoning in the text. The system produces both sourced findings, and conditional inferences that follow if an unvalidated conjecture holds. Full hypothesis tracing exists but isn't rendered inline in this build. If you find a claim that reads as established and isn't, that's the exact failure mode we're still closing. We're not claiming this replaces experimental work. It's a hypothesis generation system, it reads across domains and proposes testable connections. Whether those predictions hold up is for wet labs to determine.
Happy to run this on your current research area. Leave a research prompt, we'll select 3 and generate a paper for you with the same system.