AI生命延续学日报 2026/10/17
Today’s Summary
Monte Rosa's NEK7 degrader hit a median 85% drop in hsCRP across 108 subjects in its Phase 1 trial; Tenvie's CNS-penetrant NLRP3 inhibitor dosed its first patient the same day — anti-inflammaging now has two early-stage clinical programs running in parallel.
Big pharma is buying instead of building, Prenuvo just opened 200K MRI scans to researchers, and senescent cell-derived extracellular vesicles can spread the aging signal on their own — together these signals suggest the field is shifting from mechanism discovery toward data infrastructure and clinical pipeline assembly.
Readers should verify the peer-review status of Monte Rosa's GFORCE-1 data and the post-hoc analysis limitations in the Longeveron preprint before drawing conclusions about the near-term clinical value of the NLRP3 pathway.🔥 Today’s Top 9 Highlights
1. AI Flags SRSF1 Gene as Extending Lifespan by 14% in Aged Mice
According to Google News aggregated reporting , a study reports that SRSF1 — an RNA splicing factor flagged by AI screening — is linked to a “rejuvenation” effect, extending lifespan in aged mice by 14%. This is a secondary source; the original paper, journal name, sample size, and intervention details are all unverified. What we can confirm: the work sits at the intersection of AI-assisted target discovery and aging biology, and an animal result doesn’t translate directly to humans. 🧬
2. Blocking the STING Pathway Protects Bone Density in Aged Mice 🦴
According to Lifespan.io’s coverage , a Chinese research team published in Advanced Science showing that macrophage-specific silencing of the Sting gene in 12–18-month-old mice significantly preserved bone density and structure while reducing bone marrow adipocyte numbers. The study also mapped out a signaling chain: bone marrow adipocytes secrete DPP4, which cooperates with TLR4 and mitochondrial DNA leakage to activate STING, which then drives CD38 to degrade extracellular NAD⁺ and impair mitochondrial function in neighboring bone progenitor cells. The drug combo (STING inhibitor H-151 + NMN) performed comparably to the established anti-osteoporosis drug alendronate. Caveats: mouse study, small sample, combo design makes it hard to isolate each drug’s contribution, no human data yet.
3. Monte Rosa’s NEK7 Degrader Posts Multi-Marker Inflammation Drops in Phase 1 🔬
According to Longevity.Technology’s coverage , Monte Rosa Therapeutics ran GFORCE-1 (Phase 1/pilot, 4-week dosing) in 108 obese, high-cardiovascular-risk subjects. The molecular glue degrader MRT-8102 drove a median ~80–90% drop in NEK7 protein levels; downstream markers followed — hsCRP down 85%, IL-6 down 54%, and plaque-associated signals calprotectin, S100A12, and SAA down 56%, 46%, and 51% respectively. Adverse event rates (33%) were close to placebo (30%), with no increase in infection risk. Data came from a company press release — no peer-reviewed publication yet. A 4-week biomarker win ≠ reduced cardiac events; the follow-on Phase 2b GFORCE-2 is expected to kick off in H1 2027.
4. Prenuvo Opens 200K Whole-Body MRI Scans to 50+ Research Institutions 📊
According to Longevity.Technology’s coverage , Prenuvo announced the Prenuvo Research Network on September 29th, adding over 200,000 whole-body MRI scans to a research biobank. A preliminary analysis of 78,000+ voluntary screeners found 7% had clinically significant brain abnormalities (aneurysms, infarcts, meningiomas). Partners include UCSF, City of Hope, Moffitt Cancer Center, and 50+ others. Key caveat: participants are self-selected paying customers — not a representative population. The causal link between imaging findings and improved survival outcomes hasn’t been established; each sub-study calls for prospective validation.
5. Longeveron’s MSC Therapy Preprint Hints at Slowing Alzheimer’s Neuroinflammation 🧠
According to Longevity.Technology’s coverage , Longeveron posted a post-hoc exploratory analysis on medRxiv (based on the published CLEAR MIND Phase 2a trial, n=49): laromestrocel (Lomecel-B) hit p < 0.05 on the free water MRI metric in the left hippocampus, with neuroinflammation reduction moving in sync with cognitive score (MoCA) improvement and a drop in plasma neurogranin. Caveats: preprint, not yet peer-reviewed; post-hoc design; small sample; correlation ≠ causation; the product doesn’t have FDA approval; the company is running an investor roadshow at the same time.
6. Kailera’s Obesity Drug Finishes Phase 3 Enrollment — Results Due 2028 💊
According to Longevity.Technology’s coverage , Kailera Therapeutics wrapped enrollment across three KaiNETIC Phase 3 trials (roughly 4,900 participants total), with a 10 mg top dose and a 76-week duration per trial. KaiNETIC-3 includes a direct head-to-head against open-label semaglutide 2.4 mg. Top-line data are expected mid-2028. No efficacy or safety data to analyze yet; whether a GLP-1/GIP dual agonist translates to longer healthspan is a question a 76-week trial simply can’t answer.
7. Tenvie’s CNS-Penetrant NLRP3 Inhibitor Enters Phase 1 Human Trial 🧪
According to Longevity.Technology’s coverage , Tenvie Therapeutics dosed its first subject on October 13th, kicking off a randomized dose-escalation Phase 1 safety and PK study in healthy volunteers. TNV313 is positioned as a fully CNS-penetrant oral NLRP3 inhibitor targeting neurological, cardiometabolic, and ophthalmic diseases; preliminary safety and PK data are expected in H1 2027. Phase 1 only evaluates safety and target tissue distribution — you can’t read neuro-degenerative efficacy into it yet.
8. Big Pharma Is Buying, Not Building — Aging Is Now Part of the M&A Playbook 💰
According to Longevity.Technology’s analysis , big pharma runs 62% of all registered industry trials in core aging pathways — but nearly every specific 2025–2026 deal was an external acquisition: Lilly–Insilico at $2.75B, Roche/Chugai–Gero at up to $1B. A ~$100B patent cliff before 2028 is driving life sciences M&A to $240B in 2025 (up 81% YoY). These numbers are base-case assumptions from the report, not realized value. Worth noting: AbbVie’s 11-year Calico collaboration ended in November 2025 with zero approved products — internal aging research has its own sky-high failure rate.
9. Senescent Cell-Derived Extracellular Vesicles Spread Aging Signals Independently of Classic SASP 🔬
According to Lifespan.io’s coverage , a cell study published in Aging compared the secretomes of H-RASG12V-induced senescent cells versus p16-overexpressing cells: p16 cells barely secreted soluble SASP factors like IL-1β or IL-6, but pumped out more small extracellular vesicles (sEVs). Those sEVs drove recipient cells to express p16 and SA-β-gal (senescence markers) and induced DNA damage (γH2AX) — without triggering classic inflammatory cytokine secretion. The effect was validated using sEVs from elderly donors applied to infant fibroblasts. Caveats: in vitro only, specific sEV cargo unidentified, targeting sEVs therapeutically remains unexplored.
⚡ Product & Feature Updates
Digbi Health Folds HRT Into Its Multi-Omics Precision Health Platform 💡
According to Longevity.Technology’s coverage , Digbi Health has added hormone replacement therapy (HRT) to its existing precision biology platform, letting members get both GLP-1 and HRT prescriptions from the same clinical team. Personalized plans are tailored using genomic, gut microbiome, continuous glucose, and blood biomarker data. The company also expanded to all 50 US states via OpenLoop Health and MD Integrations, covering 300+ employers. Digbi cites published research showing $3,012 in annual per-member healthcare savings — but that data predates the HRT launch, so the HRT module itself has no standalone efficacy or cost data yet. Platform is B2B (employers and health plans), not direct-to-consumer.
🧪 Cutting-Edge Research
Bisphenol Mixtures Reprogram Epigenetic State in Human Breast Organoids ⚗️
According to GEN’s coverage , researchers from Tufts University and Boston University published in PNAS using 3D breast organoids from 13 healthy young women to test the effects of BPA, BPS, and BPF — individually and as a mixture. Solo exposures showed limited changes, but low-dose mixed exposure (matching concentrations actually detected in human blood and urine) disrupted normal breast development in most donors’ organoids, inducing extracellular matrix remodeling, epithelial plasticity, and DNA methylation reprogramming. The molecular change patterns closely overlapped with data from invasive lobular carcinoma patients. An AI-assisted high-content screening platform was used to quantify developmental disruption. Caveats: in vitro organoid model, small donor sample, can’t prove chemicals cause cancer, and whether these molecular changes persist in human tissue and lead to disease still needs prospective study.
◎ Industry Shifts & Personal Impact
Cell & Gene Therapy Scale-Up Faces Systemic Bottlenecks 🏭
According to GEN’s coverage of the BioLife Solutions e-book report , attendees at the 2026 Cell Summit (researchers, clinicians, manufacturers, CDMOs, etc.) converged on a shared view: the next core challenge for CAR-T and other cell & gene therapies is building a scalable “operating system” — manufacturing, reimbursement, logistics, and data governance all have to work in sync, and any single broken link leaves patients waiting. This industry assessment doesn’t include specific efficacy data; it’s a good read for anyone tracking regenerative medicine accessibility and production infrastructure. Commercial sustainability remains to be proven.
⌘ Top Open-Source Projects
kkontras/NeuroAtlas: EEG Foundation Model Cross-Task Benchmark Suite 🧠
NeuroAtlas is an EEG foundation model evaluation framework covering four task categories: sleep staging, brain age estimation, epilepsy detection, and brain-computer interfaces (BCI). The paper was accepted at NeurIPS 2026. Great fit for developers and researchers working on neural signal processing, brain age biomarkers, and cross-task transfer learning. Repo published 2026-10-16, currently at 18 stars; check the repo README for license details, dataset terms, and environment requirements — not independently verified.
JackieHanLab/PAOPAC: Proteome-Driven Organ Proxy Aging Clock ⏱️
PAOPAC (Proteome-Aware Organ Proxy Aging Clock) uses proteomics data to build biological age proxy clocks for individual organs — a multi-omics aging measurement tool. Good fit for researchers working on proteomics and aging biology or organ-specific biological age estimation. Repo published 2026-09-21, 2 stars; datasets used, license, and algorithmic details are all in the repo — none independently verified.
❓ Related Questions