More Matt Kaeberlein - This Molecule Could Beat Rapamycin in Longevity

#1

Hmmmm… something new.

This Molecule Could Beat Rapamycin in Longevity

1 Like
#2

CGPT5.1 Video Summary:

A. Executive Summary

Matt Kaeberlein and Ben Blue discuss Ora Biomedical’s strategy to reboot longevity drug discovery using high-throughput in vivo screening in C. elegans and to redirect those findings toward stress- and radiation-resilience indications.

Ora’s core platform is the “WormBot”: automated robotics plus AI image analysis to run standardized lifespan experiments at scale. Each unit can screen ~100 molecules per month with proper replication, and Ora currently has <10,000 interventions in its database (single agents, dose curves, combinations). Despite this relatively small coverage of chemical space, they report multiple small molecules that extend worm lifespan more than rapamycin, including a PI3K/mTOR inhibitor now used as their positive control and single-agent hits with up to ~60% median lifespan extension. They emphasize the field’s over-focus on a few pathways (mTOR, autophagy, etc.) and under-investment in unbiased discovery.

The “Million Molecule Challenge” is their proposed scale-up: quantify lifespan effects of 1,000,000 interventions in worms. They argue that for roughly $5M over ~3 years, this would generate the foundational dataset needed for serious AI-guided longevity discovery and likely identify hundreds of better-than-current hits. They highlight unexpected findings (e.g., ivermectin as a dramatic lifespan shortener) and frequent non-intuitive drug–drug interactions.

For translation, Ora is prioritizing stress-resilience applications: UV and radiation protection (for radiotherapy patients, air/space crews, and future astronauts), resistance to organophosphate-like neurotoxins, and improved wound healing, moving hits from worms to human cell assays and then mouse models. They frame these as realistic clinical wedges that align with defense, space, and dermatology markets, while maintaining a longer-term ambition of genuine longevity therapeutics.

B. Bullet Summary (12–20 bullets)

  • Ora Biomedical is a spin-out from Matt Kaeberlein’s lab focused on in vivo high-throughput longevity drug discovery using C. elegans.
  • Their “WormBot” platform automates worm lifespan experiments with imaging and AI, enabling ~100 fully replicated compounds per bot per month.
  • Current database contains <10,000 interventions (single agents, doses, combinations) but is already larger and more standardized than most of the historical field.
  • They report multiple small molecules with greater lifespan extension in worms than rapamycin, including a PI3K/mTOR inhibitor now used as their internal positive control.
  • Some single molecules allegedly extend worm lifespan by up to ~60% at lower doses than rapamycin.
  • Early combinatorial screens show frequent antagonistic interactions and occasional strong synergies, undermining naive stacking of “good” drugs.
  • An FDA-approved library screen rediscovered ivermectin as a near-lethal “hit,” illustrating that lifespan screening will surface both pro- and anti-survival chemotypes.
  • Pathway-level analysis shows wide effect distributions within pathways (e.g., mTOR/PI3K agents can markedly help or harm lifespan depending on compound and dose).
  • Autophagy pathway inhibitors tend to shorten lifespan, which is consistent with existing biology and validates the platform’s face plausibility.
  • The “Million Molecule Challenge” proposes lifespan testing 1,000,000 interventions in worms; they argue it could be done for ~US$5M and would transform the field.
  • They claim NIH and similar funders have largely ignored such discovery-first proposals in favor of mechanistic, hallmark-driven projects.
  • Ora is now targeting radiation and toxin resilience as near-term indications: starting with UV-induced DNA damage in worms, then human cells, then mouse models.
  • Pilot work suggests some longevity hits also increase resistance to UV and organophosphate-like neurotoxic stress in worms.
  • They position these resilience drugs for radiotherapy support, military/CBRN contexts, occupational exposures, and eventually astronaut protection.
  • Wound-healing models reveal that some longevity drugs enhance tissue repair while others (e.g., high-dose mTOR inhibitors) can impair it.
  • They foresee highly personalized “resilience stacks,” where different individuals receive different combinations based on specific stress exposures and biology.
  • Ora collaborates with AI companies (e.g., Ryzome Biosciences) to use their dataset for dose-response prediction and more efficient future screening.
  • The team explicitly prioritizes large effect sizes in worms (≥40–60% lifespan extension) as worth serious preclinical follow-up, dismissing marginal 10–15% gains.
  • They argue that discovery science in longevity has been severely neglected and that large-scale unbiased animal screens are the missing infrastructure.
  • Open-access components (Million Molecule Challenge leaderboard, crowd-sponsored experiments) are used for community engagement and education.

D. Claims & Evidence Table

Claim (video) Evidence presented in the discussion Your assessment
Rapamycin is still the best-in-class small-molecule longevity drug in terms of effect size and reproducibility. Kaeberlein states that ~20 years after first rapa lifespan studies, no one has clearly surpassed it in mammals; this aligns with ITP and other rodent data showing robust lifespan extension with rapamycin across strains and sexes. Strong for mammals: multiple independent studies show rapamycin extends lifespan and healthspan in mice; no other small molecule has a stronger mammalian dataset yet. See e.g. Harrison et al. 2009, Bitto et al. 2016.
Ora has discovered mTOR/PI3K inhibitors that extend C. eleganslifespan more than rapamycin. Internal data; they say one such compound is now their standard positive control and is being submitted to the NIA ITP/CITP. No structures or peer-reviewed lifespan curves are presented. Speculative–Moderate: plausible in worms—many mTOR/PI3K chemotypes exist—but currently supported only by company claims and community presentations, not peer-reviewed publications.
They have single small molecules achieving up to ~60% lifespan extension in worms at doses below those used for rapamycin. Described verbally; no quantitative curves or stats shown in the transcript. Speculative: large effects in worms are biologically believable (e.g., daf-2, strong CR), but magnitude, reproducibility, and generality remain unverified externally.
A million-molecule lifespan screen in worms could be completed in ~3 years for about US$5M. Back-of-the-envelope scaling from current capacity (9 wormbots, ~1,000 interventions/month, <10k done so far), assuming linear scale-up in hardware, space, and staff. Weak–Speculative: directionally plausible (worm work is cheap), but ignores real-world overhead, QC, software engineering, and failure modes. No detailed costed plan provided.
Most AI-driven longevity discovery efforts are hamstrung by lack of a large, high-quality, in vivo dataset. They note that almost all current AI approaches train on sparse, noisy in vitro/pathway data, whereas Ora’s dataset is dense in vivo phenotyping. Empirically, DrugAge and similar resources are small and heterogeneous. Moderate: broadly true that AI is limited by training data; also true that high-throughput in vivo datasets are rare. But “most AI efforts are hamstrung” is qualitative and somewhat overstated. See DrugAge, LongevityMap.
Longevity interventions often increase resistance to multiple forms of stress, including radiation and toxins. They reference classical work in worms/flies where long-lived mutants show increased stress resistance; their own early data show some hits protect worms from UV and organophosphate-like stress. Moderate–Strong for the general principle: long-lived mutants (daf-2, age-1, etc.) often display enhanced stress resistance (Lithgow & Walker 2002). Weak for their specific molecules until peer-reviewed.
Combinations of individually beneficial longevity drugs often antagonize each other instead of adding benefits. They cite their metformin+FDA library screen and anecdotal reporting from Brian Kennedy’s mouse work, where “good+good” frequently yields neutral or negative outcomes. No numeric rates given. Moderate: consistent with general pharmacology (polypharmacy interactions) and some published examples in aging (e.g., certain rapa+metformin regimes), but systematic data are limited.
Autophagy inhibitors tend to shorten lifespan in their screen. They state that pathway-grouped analysis shows autophagy-pathway inhibitors as net negative for lifespan, which matches established biology. Moderate: directionally consistent with autophagy’s known pro-longevity role (Levine & Kroemer 2008), but Ora’s specific data are unpublished.
Longevity biotech has over-focused on a handful of pathways and molecules, neglecting unbiased discovery. They point to hallmarks-of-aging–driven funding, repeated focus on mTOR, NAD, senolytics, etc., and NIH reluctance to fund exploratory screens. Moderate: qualitatively accurate—funding skew is real—but not systematically quantified. NIH does fund some unbiased screens; scale is indeed small relative to mechanism-driven projects.
Radiation-resilience indications (radiotherapy support, military, space) are a tractable wedge for aging-pathway drugs. They argue these areas have clear acute endpoints and existing defense/space interest; some longevity pathways (DNA repair, oxidative stress, senescence) are mechanistically relevant. Speculative but rational: conceptually sound, and radioprotectors exist (e.g., amifostine in radiotherapy link), but no clinical data yet for Ora’s hits.

E. Actionable Insights (5–10 items)

These are meta-actionable for a serious longevity practitioner, not prompts to self-administer untested drugs.

  1. Treat mTOR inhibition as validated, but do not assume rapamycin is the ceiling. Current mammalian data still support rapamycin as the most robust small-molecule longevity lever, but Ora’s worm results make it almost certain that superior chemotypes exist. For now, focus on optimizing rapa (or rapalog) regimens within known safety constraints; view “better mTOR inhibitors” as a pipeline, not as DIY options.
  2. Stop assuming supplement megastacks are additive. Their combination data plus Kennedy’s experience strongly suggest frequent antagonism. For self-experimentation, emphasize single-mechanism probes with good biomarkers (e.g., mTOR, GLP-1, SGLT2, inflammation) rather than poorly rationalized 10–20-drug stacks.
  3. Anchor your mental model on “resilience,” not just “lifespan.” Think in terms of specific stresses you are likely to face (metabolic, vascular, neurotoxic, radiation, hypoxia) and match interventions to these domains, using measurable endpoints (e.g., CGM metrics, lipid peroxidation markers, DNA damage markers in clinical contexts).
  4. Keep effect sizes in perspective. A 10–15% lifespan gain in worms is probably not worth chasing into mammals. Prioritize interventions that show large, consistent effects in multiple models, then ask what fraction of that benefit is realistically transferable to humans.
  5. Be cautious extrapolating worm effect sizes to humans. A 60% lifespan extension in C. elegans is impressive but does not map linearly to mammals. Treat it as “this pathway is powerful,” not “this molecule will add decades to human life.”
  6. Design your own n=1 work to be falsifiable. Their emphasis on replication and standardized conditions highlights how weak many human self-experiments are. For any intervention you try, pre-register (privately) what you expect to change and by how much (biomarkers, performance metrics), and be willing to stop if the data are negative.
  7. Watch the “resilience” angle for near-term clinical products. If and when Ora or others publish human or well-controlled animal data on radiation/toxin/wound-healing resilience, those indications will likely be first-in-class clinical footholds for aging biology. Those drugs may later become candidates for broader longevity use.
  8. Use the Million Molecule Challenge as a barometer of field maturity. If the community cannot fund a ~US$5–10M unbiased in vivo screen while pouring orders of magnitude more into narrow mechanistic bets, that’s a signal about misallocation of capital. For your own investing or collaboration choices, weigh that carefully.
  9. For teaching or outreach, use open MMC data as real-world material. The open leaderboard and raw curves can be exploited for high-school/undergrad education or exploratory analyses of dose–response and pathway distributions, which can sharpen intuition about what “hit” data actually look like.
  10. Mentally separate “cool discovery” from “clinically deployable.” Everything in this video is discovery-phase. Until you see: (1) peer-reviewed worm/other-model replication, (2) rodent lifespan/healthspan and safety data, and (3) at least phase 1–2 human data for some indication, none of these specific molecules should influence your personal protocol.

H. Technical Deep-Dive (mechanisms, methods, and strategy)

Model organism:

C. elegans is used because of:

  • Short lifespan (~3 weeks),
  • Well-characterized genetics (e.g., daf-2/IGF-1 pathway, insulin/FOXO, TOR, autophagy, mitochondrial stress),
  • Established links between lifespan extension and increased stress resistance (oxidative, thermal, UV).

This allows large-N, fully automated lifespan curves that are impractical in mice.

Platform architecture:

  • Hardware: “WormBots” – robotic imaging systems that scan agar plates with worms at defined intervals.
  • Software: AI/vision pipeline to classify living vs dead worms and derive survival curves automatically.
  • Experiment design: Each molecule tested with multiple replicates per plate, often multiple doses; positive controls (rapamycin, now an in-house mTOR/PI3K inhibitor) and negative controls (vehicle only).

Pathway-level analysis:

  • Compounds are annotated by known or inferred primary targets (e.g., mTOR, PI3K, autophagy regulators, kinases).
  • Grouping by pathway shows broad distributions: some compounds in a pathway extend lifespan substantially, others shorten it, emphasizing off-target effects, dose dependence, and network complexity.
  • Autophagy inhibitors clustering as lifespan-shortening is consistent with autophagy’s role in proteostasis and organelle quality control.

Combinatorial strategy:

  • Initial combination screens fix one “backbone” drug (e.g., metformin) and iterate over a library of other agents (often FDA-approved).
  • Result patterns: synergy, neutrality, antagonism.
  • These data are used to train models (with partners like Ryzome) to predict dose windows and interaction patterns, with the goal of reducing false negatives due to suboptimal dosing.

Stress-resilience assays:

  • Radiation: UV exposure protocols induce DNA damage and mortality; worms are pre-treated with candidate compounds to assess shifts in survival curves and DNA damage markers.
  • Neurotoxins: Low-dose organophosphate-like exposures impair function or survival; candidates are screened for protection.
  • In vitro / ex vivo: Hits move into human cell lines and ex vivo skin explants to quantify DNA damage, repair pathway activation (e.g., γ-H2AX foci, p53 signaling), and wound closure dynamics.

Translational framing:

  • Rather than selling “anti-aging” to regulators, they pursue indications with clear mechanistic links and standard endpoints (e.g., radioprotection in radiotherapy, CBRN resilience, occupational toxin exposures, dermatologic repair/anti-photoaging).
  • Longevity is treated as an emergent outcome of improving organismal resilience to multiple stressors.

I. Fact-Check of Important Claims

  1. “Rapamycin is still the best small-molecule longevity drug we have.”
  1. “A million-molecule worm lifespan screen could be done for ~$5M.”
  • Evidence: No published cost model; this is a rough claim from the founders.
  • Check: There is no independent analysis. Comparable industrial HTS campaigns (in vitro, not in vivo) often cost far more when fully costed (automation, QC, computational infrastructure) (Paul et al., Nat Rev Drug Disc 2010). Treat as optimistic.
  1. “Long-lived mutants/interventions show broad stress resistance (radiation, toxins, etc.).”
  • Evidence: Classical gerontology literature shows daf-2, age-1, and similar mutants have enhanced resistance to heat, oxidative stress, and some toxins.
  • Check: Generally correct. See Lithgow & Walker, Nat Rev Mol Cell Biol 2002, Murakami & Johnson 1996. Direct evidence for ionizing radiation specifically is less abundant but conceptually aligned.
  1. “Most AI longevity discovery efforts lack robust in vivo training data.”
  • Evidence: Current public resources (DrugAge, GenAge, LongevityMap) have limited, heterogeneous data, primarily curated from small, manually run experiments; industry AI platforms often rely on omics, pathways, or cell-based screens.
  • Check: Directionally accurate. See DrugAge and overviews of AI in longevity like Zhavoronkov 2022. No large, standardized million-scale in vivo dataset exists.
  1. “Topical rapamycin is in clinical development for skin aging/wrinkles.”
  • Evidence: Kaeberlein references Chris Hladczuk’s Hayflick effort.
  • Check: There are clinical trials and papers on topical rapamycin for aging skin and conditions like facial angiofibromas: see Chung et al. 2019, and a current trial targeting wrinkles/skin aging (e.g., NCT05523306).
  1. “Combination longevity interventions frequently antagonize each other.”
  • Evidence: Video references internal Ora data and Brian Kennedy’s mouse experiences.
  • Check: Published literature on combination therapies in aging is limited but supports non-additivity in some cases. For example, some rapamycin+metformin regimens show complex, non-additive outcomes in mice (Strong et al. 2016), and polypharmacy risks in older adults are well documented (Maher et al. 2014). Claim is plausible but not quantitatively established for longevity drugs.

Overall, the transcript describes a conceptually coherent discovery platform and strategy. The biggest gap is peer-reviewed, mammalian-level evidence for Ora’s specific hits. Until those data exist, treat their molecules as early-stage candidates, not usable tools.

2 Likes
Your feedback requested on new approach to longevity research summaries
#3

And for those who saw the ivermectin reverence, here is some clarifying information:

Ora screened ivermectin as part of an FDA-approved library and found it caused >90% lifespan reduction, killing C. elegans within minutes. This was discovered only after unblinding but made perfect biological sense, since ivermectin is an anti-helminthic designed to kill worms. The team jokes about relief that it didn’t extend lifespan during the politically charged COVID ivermectin debates. The result validated the screening system by demonstrating its ability to detect dramatic negative hits. They argue that such findings illuminate unexploited opportunities such as discovering novel anti-parasitic combinations or agricultural worm-control agents. Overall, ivermectin is used as a case study showing how large-scale longevity screening will uncover both pro- and anti-survival molecules.

2 Likes
#4

I was hoping they would go more into detail about everything they tested, which they really didn’t do much of.

2 Likes
#5

If they manage to find something that works well for space or military usage they could get a lot of money from contracts which can be funnelled back into longevity research and development.

Also the molecule they’re talking about is Omipalisib if I am not mistaken. More info here: Crowdfunding Project Looks for a “Better Rapamycin” - #68 by AustraliaLongevity

2 Likes