The AAV viral vector — recombinant adeno-associated virus (rAAV) — has become the dominant platform for in vivo gene delivery, underpinning multiple FDA-approved gene therapies and thousands of preclinical studies. Its combination of broad tissue tropism, minimal immunogenicity, and a well-characterized safety record makes rAAV the go-to choice for neuroscience, ophthalmology, liver-directed therapy, and an expanding range of disease targets.
What Is an AAV Viral Vector?
Adeno-associated virus is a small (~25 nm), non-enveloped, replication-defective member of the Parvoviridae family. Its genome is a single-stranded DNA (ssDNA) molecule of approximately 4.7 kb flanked by inverted terminal repeats (ITRs). In wild-type AAV, the coding regions between the ITRs are the rep (replication) and cap (capsid) genes. In the recombinant vector (rAAV), both rep and cap are removed and replaced with the therapeutic transgene cassette, leaving only the ITRs — which are required for genome packaging and second-strand synthesis.
This design has three important consequences:
- Cargo limit: the effective packaging capacity of rAAV is approximately 4.7 kb (ITRs included), constraining the size of the expression cassette.
- Non-integrating persistence: rAAV genomes persist predominantly as circular episomes in the nucleus of transduced cells rather than integrating into the host chromosome. This episomal state is stable in post-mitotic cells (neurons, cardiomyocytes, photoreceptors) and supports durable, long-term transgene expression without the insertional mutagenesis risk associated with integrating vectors.
- Low immunogenicity: the absence of viral coding sequences dramatically reduces innate immune sensing of rAAV-transduced cells, contributing to the vector's favorable clinical safety profile.
The clinical track record of rAAV now includes FDA-approved products such as Luxturna (voretigene neparvovec-rzyl, AAV2, Spark Therapeutics) for RPE65-mediated inherited retinal dystrophy and Zolgensma (onasemnogene abeparvovec-xioi, AAV9, Novartis) for spinal muscular atrophy type 1 — milestones that validated rAAV as a human gene therapy platform. The landmark phase I/II clinical trial by Nathwani et al., 2011 (N Engl J Med) demonstrated sustained factor IX expression in hemophilia B patients after a single peripheral-vein infusion of AAV8, confirming the feasibility of systemic rAAV gene transfer in humans.
AAV Serotype Selection Guide
The capsid protein determines which cell-surface receptors an AAV viral vector binds, and therefore which tissues it transduces most efficiently. Choosing the correct serotype is the first and most consequential design decision. The table below summarizes primary tropism, key model species, and notes for the ten serotypes most commonly used in research and therapy. Tropism data are drawn from the systematic in vivo comparison by Zincarelli et al., 2008 (Mol Ther) and the engineered capsid characterization by Chan et al., 2017 (Nat Neurosci).
| Serotype | Primary tissue tropism | Key species | Notes |
|---|---|---|---|
| AAV1 | Skeletal muscle, heart, CNS, lung | Mouse, NHP | Moderate-onset systemic expression; good muscle transduction after intramuscular injection |
| AAV2 | CNS, retina (RPE), liver, muscle | Mouse, Human | Most studied serotype; slow-onset expression; used in Luxturna (subretinal); robust in vitro transduction of HEK293 cells |
| AAV5 | CNS, lung, retina, pancreas | Mouse, NHP | Slow onset; efficient photoreceptor transduction; well-characterized in NHP retina studies |
| AAV6 | Skeletal muscle, heart, lung | Mouse | Rapid-onset expression; favored for cardiac and diaphragm transduction; hybrid of AAV1 and AAV2 capsids |
| AAV8 | Liver, skeletal muscle, CNS, retina | Mouse, NHP, Human | High liver transduction after IV delivery; used in Nathwani 2011 hemophilia B trial (scAAV2/8) |
| AAV9 | Liver, lung, heart, CNS (crosses BBB) | Mouse, NHP, Human | Highest systemic biodistribution of natural serotypes; crosses the blood-brain barrier in neonates; used in Zolgensma (SMA) |
| AAV-DJ | Liver, heart, kidney (in vitro) | Mouse, in vitro | Engineered hybrid of AAV2/8/9 capsids; exceptionally high in vitro transduction efficiency; rapid-onset expression in liver after IV |
| PHP.eB | CNS neurons and glia (pan-CNS) | Mouse (C57BL/6) | Engineered AAV9 variant; IV injection transduces ~69% of cortical neurons and ~55% of striatal neurons in adult C57BL/6 mice; does not cross BBB efficiently in NHP |
| PHP.S | Peripheral nervous system (DRG, enteric, cardiac neurons) | Mouse (C57BL/6) | Companion to PHP.eB; IV injection transduces ~82% of DRG neurons; useful for autonomic and sensory circuit studies |
| Retrograde-AAV (rAAV2-retro) | Retrograde axonal transport to projection neurons | Mouse, Rat | Engineered for efficient retrograde transduction; enables labeling or manipulation of presynaptic neurons from injection into the axon terminal field; key tool for circuit dissection (Tervo et al., 2016, Neuron) |
AAV Vector Design: Promoters and Transgene Considerations
After selecting a serotype, the next critical decision is the promoter. The promoter controls the cell types in which the transgene is expressed, the expression level, and the size of the remaining packaging space. The seminal comparative studies by Gray et al., 2011 (Hum Gene Ther) and Nathanson et al., 2009 (Front Neural Circuits) established the in vivo specificity and strength of many commonly used AAV promoters.
| Promoter | Expression pattern | Ideal use case |
|---|---|---|
| CMV (cytomegalovirus) | Ubiquitous; high initial expression that diminishes over time in vivo | In vitro overexpression; short-term in vivo studies; large-scale antigen production |
| CAG (CMV enhancer + chicken beta-actin) | Ubiquitous; strong and durable in vivo | Strong long-term transgene expression across all cell types; reporter lines |
| EF1α (elongation factor 1-alpha) | Ubiquitous; strong and stable | High-level constitutive expression; 1,179 bp full-length or 493 bp short form (nEF1α) for larger payloads |
| CBA / CBh (chicken beta-actin / hybrid) | Ubiquitous; robust long-term expression including motor neurons | Preferred over CMV for sustained CNS/PNS expression; CBh (793 bp) packs more payload than full CBA (1,678 bp) |
| Synapsin I (hSyn) | Pan-neuronal; mature neurons only | Standard choice for neuron-specific transgene expression; 485 bp — leaves good packaging headroom |
| CaMKII (calcium/calmodulin-dependent protein kinase II) | Excitatory neurons (neocortex, hippocampus, striatum) | Selective manipulation of pyramidal and excitatory neurons; 1,293 bp limits remaining cargo space |
| GFAP (glial fibrillary acidic protein) | Astrocytes | Astrocyte-targeted transgene delivery; short GFAP104 (845 bp) available for larger transgenes |
| TBG (thyroxine-binding globulin) | Hepatocyte-specific | Liver-directed gene therapy; ~410 bp — compact and highly liver-specific for metabolic disease models |
| MeCP2 (methyl-CpG-binding protein 2, truncated) | Pan-neuronal (low-moderate level) | Neuronal expression with a 229 bp footprint; enables packaging of large transgenes (>4 kb coding sequences) into rAAV |
| U6 (RNA polymerase III) | Ubiquitous (drives short non-coding RNA) | Expression of shRNA, sgRNA (CRISPR), or microRNA; not for protein-coding transgenes |
Beyond promoter choice, three additional design factors govern rAAV vector performance:
- ITR integrity: both 5′ and 3′ ITRs must be intact for efficient genome packaging. Damaged or truncated ITRs reduce vector yield and genome stability.
- Cargo limit: the total insert size (promoter + transgene + poly-A signal + any additional regulatory elements) must fit within approximately 4.7 kb. Oversized genomes are packaged at severely reduced efficiency and may be truncated.
- Codon optimization: for human transgenes expressed in rodent models (or vice versa), codon optimization for the host species significantly increases protein output, sometimes by an order of magnitude, without changing the amino acid sequence.
How AAV Vectors Are Produced
Recombinant AAV viral vector production requires three genetic components: (1) the vector plasmid carrying the ITR-flanked transgene cassette, (2) the helper plasmid supplying rep and cap genes, and (3) adenoviral helper functions. These are delivered by one of three production platforms:
- Triple-transfection in HEK293 cells: the most widely used research-scale method. Three plasmids — vector, rep/cap, and adenoviral helper — are co-transfected into HEK293 cells. Scalable from 15 cm plates to multi-layer cell factories. Typical yields are 1013–1014 vector genomes (vg) per batch from a T150-scale production.
- Baculovirus-insect cell (Sf9) system: preferred for large-scale GMP manufacturing. Two or three baculovirus constructs are used to infect Sf9 cells, enabling production in bioreactor suspension culture. Yields can exceed 1015 vg per liter of culture.
- Helper-free plasmid methods: all required sequences are combined into two plasmids (or a single self-contained plasmid), eliminating the need for a separate adenoviral helper plasmid. Reduces the risk of replication-competent AAV (rcAAV) contamination.
After production, crude lysate is purified by iodixanol density gradient ultracentrifugation (a rapid, scalable method that separates full from empty capsids by buoyant density) or by affinity chromatography (using AVB Sepharose or POROS CaptureSelect resins specific for individual serotypes). Final quality control includes measurement of AAV titer (typically by droplet digital PCR or qPCR for genome copies, and by ELISA for capsid particles) and purity assessment by SDS-PAGE and silver staining to confirm the correct VP1/VP2/VP3 capsid protein ratio.
In neuroscience applications, rAAV vectors are routinely used to express genetically encoded fluorescent reporters, optogenetic tools (channelrhodopsins, halorhodopsins), chemogenetic actuators (DREADDs), and GRAB sensors for real-time imaging of neurotransmitter dynamics in live animals.
BioHippo AAV Products and Custom Production Services
BioHippo offers a catalog of pre-made rAAV vectors and custom AAV production services to support a wide range of research applications. Browse the full selection — including vectors spanning the serotypes and promoters described above — in the BioHippo AAV collection.
Pre-made AAVs are available as ready-to-use, quality-controlled viral preparations spanning major serotypes (AAV1, 2, 5, 6, 8, 9, DJ, PHP.eB, PHP.S, Retrograde) and a broad range of tissue-specific promoters. Each lot is characterized for titer and purity before shipment.
Custom AAV production services are available for researchers requiring non-catalog serotypes, novel transgene cassettes, engineered capsids, or large-quantity manufacturing. Submit your vector design and production requirements to receive a project-specific quote.
Frequently Asked Questions About AAV Viral Vectors
How does an AAV viral vector get into cells?
An AAV viral vector enters cells through a multi-step process: the capsid binds serotype-specific cell-surface receptors and co-receptors (for example, AAV2 uses heparan sulfate proteoglycans as its primary receptor), triggering clathrin-mediated or macropinocytic endocytosis. Inside the endosome, acidification induces capsid conformational changes that facilitate endosomal escape. The vector particle then traffics along microtubules to the nucleus, where the ssDNA genome is imported through the nuclear pore. Once in the nucleus, the ssDNA is converted to double-stranded DNA (dsDNA) by host cell DNA polymerase and then circularizes to form a stable episome, where the transgene is expressed by host transcriptional machinery.
What is AAV gene therapy?
AAV gene therapy is the delivery of a therapeutic transgene into target cells using a recombinant AAV viral vector. The transgene — typically a corrective copy of a mutated gene or a gene encoding a therapeutic protein — is packaged between the ITRs of the rAAV genome. After transduction, the transgene persists as a non-integrating episome in the nucleus and is expressed long-term, particularly in post-mitotic cells such as neurons, cardiomyocytes, and photoreceptors. FDA-approved examples include Luxturna (AAV2) for retinal dystrophy and Zolgensma (AAV9) for spinal muscular atrophy.
Why is AAV preferred for gene therapy?
rAAV is preferred for gene therapy because it efficiently transduces both dividing and non-dividing cells, elicits a minimal innate immune response compared with adenoviral vectors, and has an established clinical safety record spanning two decades of trials. The removal of all viral coding sequences (rep and cap) from the recombinant vector means transduced cells are not recognized and cleared by cytotoxic T lymphocytes via viral antigen presentation — a critical advantage for long-term expression. Multiple FDA-approved products demonstrate the platform's therapeutic utility: Luxturna (AAV2, Spark Therapeutics) for RPE65-associated retinal dystrophy and Zolgensma (AAV9, Novartis) for SMA type 1, with numerous additional programs in late-stage clinical trials for hemophilia, Duchenne muscular dystrophy, and neurological diseases.
How many AAV serotypes are there?
Over 100 naturally occurring AAV serotypes and variants have been identified from human and non-human primate tissues. Research and clinical applications predominantly use AAV1, 2, 5, 6, 8, and 9 (characterized in detail by Zincarelli et al., 2008), along with engineered variants that extend the tropism space: AAV-DJ (enhanced in vitro transduction), PHP.eB and PHP.S (pan-CNS and peripheral nervous system transduction after IV injection in mice; Chan et al., 2017), rAAV2-retro (retrograde circuit tracing; Tervo et al., 2016), and Anc80L65 (ancestral reconstruction with broad tropism including inner ear hair cells).
Can rAAV integrate into the host genome?
Wild-type AAV2 integrates site-specifically into the AAVS1 locus on human chromosome 19q13 through the action of the Rep78 and Rep68 proteins, which recognize AAVS1 sequences. However, recombinant AAV (rAAV) lacks the rep gene entirely; without Rep proteins, site-specific integration cannot occur. Consequently, rAAV genomes persist predominantly as circular, double-stranded episomes in the nucleus. Random integration events do occur at a low frequency (estimated at 0.1–1% of transduced cells in some studies), but the rate is far lower than for integrating vectors such as lentivirus or gammaretrovirus. The episomal state is stable in post-mitotic cells for years but is diluted out in rapidly dividing cell populations, which limits rAAV utility for applications requiring long-term expression in proliferating tissues.