Hypothesis: CD47-functionalized liver-detargeted AAV capsids combined with TREE-driven transgenes enable safe, dose‑sparing multi‑tissue transduction for longevity interventions

Dr. Isaac Björk· Mar 21, 2026

Mechanism: CD47-functionalized AAV capsids evade Kupffer cell phagocytosis and complement activation, while TREE promoters target transgene expression to injured tissues. Readout: Readout: This leads to multi-tissue transduction, reduced TMA risk, and a significant 'LIFESPAN BOOST' in the longevity intervention platform.

Background

Recent advances show that engineered AAV capsids can achieve strong cardiomyocyte or muscle transduction while minimizing liver uptake [https://pubmed.ncbi.nlm.nih.gov/35810332/], [https://aavnergene.com/novel-aav-capsids/aav-capsid-tropisms/]. Ancestral reconstruction and directed evolution further reduce immunogenicity [https://pmc.ncbi.nlm.nih.gov/articles/PMC11296317/]. Nevertheless, neutralizing antibodies (NAbs) at low titers blunt transgene expression [https://pmc.ncbi.nlm.nih.gov/articles/PMC10014285/], and high systemic doses risk complement‑mediated thrombotic microangiopathy (TMA) [https://pmc.ncbi.nlm.nih.gov/articles/PMC10014285/]. Tissue regeneration enhancer elements (TREEs) offer damage‑responsive transcription but do not solve vector clearance or complement activation [https://doi.org/10.1101/2025.04.28.651096].

Hypothesis

Displaying the "self" signal CD47 on the surface of liver‑detargeted AAV capsids will inhibit Kupffer‑cell phagocytosis and downstream complement activation, thereby permitting lower vector doses that avoid TMA while preserving transduction of heart, skeletal muscle, and CNS. When paired with TREE‑driven cassettes, transgene expression will be restricted to injured or senescent cells within those tissues, creating a self‑limiting, multi‑tissue therapeutic platform for systemic longevity interventions.

Mechanistic Rationale

CD47–SIRPα interaction delivers an inhibitory phagocytic signal to macrophages; engineering AAV capsids to present CD47 peptides or fused CD47 domains should reduce opsonin‑dependent uptake by liver Kupffer cells, decreasing the source of complement‑activating immune complexes.
Reduced Kupffer‑cell uptake lowers circulating cytokine bursts (IL‑6, TNF‑α) that amplify the classical complement cascade, mitigating the risk of TMA observed at high doses.
Lower effective dose diminishes the absolute load of capsid available for NAb binding; combined with transient NAb adsorption using empty capsids, the vector can evade humoral immunity without immunosuppression.
Liver‑detargeting mutations (e.g., THGTPAD, NLPGSGD) already reduce hepatocyte transduction; adding CD47 further shifts biodistribution toward extracardiac muscles and CNS, as shown by ancestral variants with high muscle tropism [https://aavnergene.com/novel-aav-capsids/aav-capsid-tropisms/].
TREE promoters respond to DNA damage or senescence markers, ensuring transgene activity only where needed, which reduces ectopic expression and potential toxicity.

Predictions & Experimental Design

In vitro – HEK293T cells transfected with plasmids encoding AAV capsids bearing CD47 inserts will show reduced binding to isolated murine Kupffer cells in flow cytometry assays compared with wild‑type capsids.
In vivo (murine) – Intravenous administration of 5×10¹³ vg/kg of CD47‑functionalized, liver‑detargeted AAV9 carrying a TREE‑driven secreted reporter (e.g., GLuc) will yield:
- ≥70% reduction in liver transgene mRNA versus standard liver‑detargeted capsid.
- 2‑3‑fold increase in cardiac and skeletal muscle GLuc activity.
- Serum complement C3a and sC5b‑9 levels comparable to saline controls, whereas standard AAV at the same dose elevates these markers.
NAb challenge – Pre‑immunize mice with AAV9 capsid; co‑administer empty CD47‑capsid decoys (1:10 ratio) and assess transgene expression. Expect <30% reduction in expression versus >80% reduction with non‑decoy controls.
Safety – Monitor hematology, renal function, and histopathology for TMA signs over 28 days; hypothesis predicts no microthrombi or endothelial injury.

Potential Pitfalls

Overexpression of CD47 on capsids could interfere with proper assembly or packaging efficiency; titration of CD47 copy number will be required.
Species differences in SIRPα affinity may limit translatability; humanized CD47 peptides may be needed for preclinical models.
TREE leakiness in non‑injured tissues could still cause off‑target effects; incorporating microRNA detargeting motifs for ubiquitous tissues may refine specificity.

If validated, this approach would directly address the three major fronts highlighted in the field—tissue specificity, immunogenicity, and scalable multi‑tissue delivery—while providing a falsifiable framework: either CD47 decoration lowers complement activation and enables dose‑sparing multi‑tissue transduction, or it does not, in which case alternative immune‑evasive strategies must be pursued.

Background

Hypothesis

Mechanistic Rationale

CD47–SIRPα interaction delivers an inhibitory phagocytic signal to macrophages; engineering AAV capsids to present CD47 peptides or fused CD47 domains should reduce opsonin‑dependent uptake by liver Kupffer cells, decreasing the source of complement‑activating immune complexes.

Reduced Kupffer‑cell uptake lowers circulating cytokine bursts (IL‑6, TNF‑α) that amplify the classical complement cascade, mitigating the risk of TMA observed at high doses.

Lower effective dose diminishes the absolute load of capsid available for NAb binding; combined with transient NAb adsorption using empty capsids, the vector can evade humoral immunity without immunosuppression.

Liver‑detargeting mutations (e.g., THGTPAD, NLPGSGD) already reduce hepatocyte transduction; adding CD47 further shifts biodistribution toward extracardiac muscles and CNS, as shown by ancestral variants with high muscle tropism [https://aavnergene.com/novel-aav-capsids/aav-capsid-tropisms/].

TREE promoters respond to DNA damage or senescence markers, ensuring transgene activity only where needed, which reduces ectopic expression and potential toxicity.

Predictions & Experimental Design

In vitro – HEK293T cells transfected with plasmids encoding AAV capsids bearing CD47 inserts will show reduced binding to isolated murine Kupffer cells in flow cytometry assays compared with wild‑type capsids.

In vivo (murine) – Intravenous administration of 5×10¹³ vg/kg of CD47‑functionalized, liver‑detargeted AAV9 carrying a TREE‑driven secreted reporter (e.g., GLuc) will yield:

≥70% reduction in liver transgene mRNA versus standard liver‑detargeted capsid.
2‑3‑fold increase in cardiac and skeletal muscle GLuc activity.
Serum complement C3a and sC5b‑9 levels comparable to saline controls, whereas standard AAV at the same dose elevates these markers.

NAb challenge – Pre‑immunize mice with AAV9 capsid; co‑administer empty CD47‑capsid decoys (1:10 ratio) and assess transgene expression. Expect <30% reduction in expression versus >80% reduction with non‑decoy controls.

Safety – Monitor hematology, renal function, and histopathology for TMA signs over 28 days; hypothesis predicts no microthrombi or endothelial injury.

Potential Pitfalls

Overexpression of CD47 on capsids could interfere with proper assembly or packaging efficiency; titration of CD47 copy number will be required.

Species differences in SIRPα affinity may limit translatability; humanized CD47 peptides may be needed for preclinical models.

TREE leakiness in non‑injured tissues could still cause off‑target effects; incorporating microRNA detargeting motifs for ubiquitous tissues may refine specificity.