Preprint — not peer-reviewed (bioRxiv, January 2023). All claims are preliminary and have not undergone formal peer review.
A dual mTOR/NAD+ acting gerotherapy
TL;DR
Li et al. report a genome-wide CRISPRa screen in K562 cells that identifies SIRT3 — the mitochondrial NAD+-dependent deacetylase — as the top genetic intersection between the mTOR/insulin pathway and the NAD+ pathway. The proposed mechanism is shared regulation of fatty acid oxidation (FAO): mTORC1 targets LPIN1 (a phosphatidate phosphatase governing lipid flux) while SIRT3 deacetylates and activates mitochondrial FAO enzymes. The authors then characterize BIOIO-1001, a novel pioglitazone-derived compound (lacking the thiazolidinedione ring) that activates this SIRT3-centered intersection. BIOIO-1001 shows efficacy in mouse models of NASH and ALS (SOD1-G93A transgenic mice), with ALS lifespan extension of ~26% and paralysis delay of ~29%. The mTOR–NAD+ cross-talk is proposed as a therapeutic target for multiple age-related diseases simultaneously. No NMN or NR was studied; the mTOR–NAD+ cross-talk is inferred from shared metabolic targets, not demonstrated by direct molecular epistasis.
Design
Genome-wide CRISPRa screen
- Cell line: K562 CRISPRa-competent cells
- Library: Human Genome-wide CRISPRa-v2
- Compound tested: BIOIO-1001 at 50 µM (duplicate)
- Hit identification: Top-3 sgRNA average phenotype, p < 0.05 threshold
- Validated top hit: SIRT3
- LPIN1 status: Directionally consistent but sub-threshold (not a statistically significant hit)
NASH mouse models
- HFD model: Male C57BL/6J mice on high-fat diet (60% kcal fat; 10 weeks); then BIOIO-1001 or pioglitazone (30 mg/kg/day oral gavage) for 10 days; n=6/group (transcriptional profiling subset; metabolic experiment n not explicitly stated per group in text)
- HTF-C (NASH) model: Male C57BL/6J mice on high trans-fat (40% kcal)/fructose (20% kcal)/cholesterol (2% w/w) diet (16 weeks); then BIOIO-1001 or vehicle (oral gavage) for 3 weeks; n=6 (biochemistry); n=10–13 (NAS/fibrosis scoring)
- Ex vivo validation: Primary hepatocytes from SIRT3 wildtype and SIRT3-deficient mice
ALS models
- Sporadic ALS (human iPSC motor neurons): BIOIO-1001 at 1, 5, 10 µM; motor neuron survival assessed at day 31 via ISL1/DAPI staining
- Familial ALS (SOD1-G93A mice): SV129 background; treatment initiated at 9 weeks of age (paper notes typical lifespan ~18 weeks in this model); continued to endpoint; n not explicitly stated in figure legends or methods
Endpoints
- Liver triglycerides, ALT/AST, NAS/fibrosis scoring, gene expression (hepatic inflammation/fibrosis markers)
- ALS: survival curves, paralysis onset, plasma triglycerides
- HFD: glucose tolerance, insulin sensitivity, AKT phosphorylation (Ser473) in muscle, plasma triglycerides, NEFAs
Key results
CRISPRa screen
- SIRT3 scored as top validated genetic hit for BIOIO-1001; follow-up experiments confirmed SIRT3 dependence in ex vivo hepatocytes (CPT1b and UCP3 expression lost in SIRT3-KO cells)
- LPIN1 was directionally consistent in the screen but did not reach statistical significance — the mTOR/LPIN1 arm of the proposed cross-talk is not directly validated by the screen data
NASH/metabolic
- Liver triglycerides significantly reduced by BIOIO-1001 vs. vehicle (n=6; ANOVA with Tukey’s test; p < 0.05)
- ALT/AST substantially lower than vehicle controls
- BIOIO-1001 more protective than pioglitazone at the transcriptional level for inflammation and fibrosis markers
- HFD study: plasma triglycerides uniquely reduced by BIOIO-1001 (not pioglitazone); improved glucose tolerance, insulin sensitivity; enhanced AKT Ser473 phosphorylation in muscle
- No PPARÎł activation (unlike pioglitazone)
ALS (SOD1-G93A mice)
- Lifespan extension: ~26% (vs vehicle)
- Paralysis delay: ~29% (vs vehicle)
- Plasma triglycerides increased in ALS mice treated with BIOIO-1001 (context-dependent; authors note elevated plasma triglycerides correlate with better ALS prognosis in humans)
- Sporadic ALS iPSC motor neurons: protective at 1, 5, 10 µM
Lipid metabolomics (liver)
- Phosphatidic acid (LPIN1 substrate) increased; diacylglycerol (LPIN1 product) decreased — consistent with LPIN1 modulation but mechanism of how BIOIO-1001 affects LPIN1 not directly shown
BIOIO-1001 chemical biology
- In silico docking (DiffDock): highest-confidence pose binds SIRT3 outside the NAD-Ribose pocket (PDB: 4BN4; SIRT3 in complex with ADP-ribose, 1.3 Å resolution) — direct NAD+ competitive binding not predicted
- Does not activate PPARÎł, mitochondrial pyruvate carrier, or a panel of nuclear hormone receptors
Extrapolation to humans
| Dimension | Status | Notes |
|---|---|---|
| Pathway (SIRT3/mTOR/FAO) conserved in humans? | yes | Conserved enzymes; but SIRT3’s role in human aging is not fully established |
| Phenotype (NASH, ALS) conserved from mice? | partial | NASH pathology relevant; SOD1-G93A is familial ALS only (~2% of ALS cases) |
| Replicated in humans? | no | Preprint; no human arms; no NIA ITP equivalent; needs-human-replication |
Limitations
- Preprint — not peer-reviewed. All claims are preliminary.
- LPIN1 was not a significant CRISPRa hit — the mTOR arm of the proposed dual cross-talk is under-supported by the screen data.
- Mechanistic directionality not established. The paper does not show mTOR regulating NAD+ pools or NAD+/SIRT3 regulating mTOR directly — cross-talk is inferred from shared lipid metabolic targets.
- n per group small (n=6 for biochemistry; n=10–13 for histology). Two different genetic backgrounds used: C57BL/6J for NASH/HFD studies; SV129 for SOD1-G93A ALS survival studies.
- SOD1-G93A model represents familial ALS with a rare mutation — very limited translatability to sporadic ALS or general aging.
- n not reported for ALS survival curves — cannot evaluate statistical power.
- AKT phosphorylation observed in muscle is AKT Ser473 (mTORC2 substrate), not a direct mTORC1 output — does not confirm mTORC1 modulation by BIOIO-1001.
- Conflicts of interest declared: T.R.P. is founder of BIOIO (the company that owns BIOIO-1001 as an asset); T.M.M. declares consulting/advisory relationships with Ionis, Biogen, Cytokinetics, Disarm Therapeutics, BIOIO, UCB, Regeneron, and Denali. No patent-inventor conflict is disclosed in the preprint’s COI section.