AAV Gene Delivery: A Beginner's Complete Guide

AAV gene delivery — using recombinant adeno-associated virus (rAAV) as a vector — has become the dominant approach for stable in vivo gene transfer in both basic neuroscience and clinical gene therapy. For researchers encountering AAV for the first time, understanding its virology, the practical differences between serotypes, titer nomenclature, and promoter choice is essential before ordering the first vector. This guide covers every concept you need to get started.

What Is AAV? Virology Basics

Adeno-associated virus belongs to the family Parvoviridae, genus Dependoparvovirus. The native virus is a small (~25 nm), non-enveloped icosahedral particle carrying a single-stranded DNA (ssDNA) genome of approximately 4.7 kb, flanked at both ends by inverted terminal repeats (ITRs). In its natural biology, AAV is a defective virus: it cannot complete a productive replication cycle without a co-infecting helper virus (adenovirus or herpesvirus). In the absence of a helper, wild-type AAV establishes latency — it can integrate into a specific site (AAVS1) on human chromosome 19 or persist episomally.

In recombinant AAV (rAAV) — the form used in research and therapy — the viral rep and cap coding genes are removed and replaced with the therapeutic or research transgene cassette (promoter + gene of interest + polyadenylation signal). Only the two ITRs are retained from the viral genome. Rep and Cap are supplied in trans during production and are absent from the final vector. The consequence is critical for safety: rAAV is predominantly episomal in transduced cells and does not carry any viral coding sequences.

Low-level integration at random genomic sites does occur (~0.1% of transduced cells under typical research conditions), but this rate is orders of magnitude lower than integrating vectors such as lentivirus or gamma-retrovirus, giving rAAV a favorable genotoxicity profile. It is therefore scientifically inaccurate to say "AAV never integrates" — the correct statement is that rAAV is predominantly non-integrating and persists as circular episomes in the nucleus (Naso et al., BioDrugs 2017; Wang et al., Signal Transduct Target Ther 2024).

Implication for experiment design: because rAAV persists episomally, expression is durable in post-mitotic cells (neurons, hepatocytes, adult cardiomyocytes, skeletal muscle) for months to years. In actively dividing cells, the episomal genome is diluted with each round of cell division, leading to progressive loss of expression. rAAV is therefore not the appropriate choice for long-term transgene expression in rapidly proliferating cell populations.

Thirteen natural AAV serotypes have been identified (AAV1–13), and they differ substantially in their outer capsid protein sequence. Capsid sequence determines receptor binding, cell-entry efficiency, and tissue tropism — which is why serotype selection is one of the most consequential decisions when designing an AAV experiment.

Key AAV Parameters Every Researcher Must Understand

(a) Serotype — Tissue Tropism

The AAV serotype determines which cell types the capsid can transduce efficiently. Even if a given serotype can enter several tissues, efficiency varies enormously. The table below summarizes the most commonly used serotypes and engineered variants:

For a detailed comparison of serotype performance across tissues and species, see our AAV serotype comparison guide.

(b) Titer — What the Numbers Mean

AAV preparations are quantified in two ways that are frequently confused:

• Vector genomes per mL (vg/mL) — counts physical viral particles containing the transgene genome, measured by qPCR against the ITR or the transgene sequence. This is the standard unit for research-grade AAV. Typical ready-to-use titers: 1×10¹² to 1×10¹³ vg/mL.
• Transducing units per mL (TU/mL) — measures biological activity (the number of vector particles that successfully transduce a defined cell type in a defined assay). TU/mL is always lower than vg/mL because not every physical particle produces a productive transduction event.

Do not interchange these units when comparing products or planning dose escalation. A vector labeled "1×10¹³ vg/mL" and one labeled "1×10¹³ TU/mL" represent very different quantities of biologically active vector. For an in-depth discussion of titer measurement methods and purity assessment, see our AAV titer and purity guide.

(c) Promoter — Which Cells Express the Transgene

The capsid serotype determines which cells are transduced (the vector enters); the promoter determines which of those transduced cells actually express the transgene. A mismatch between serotype tropism and promoter activity is a common source of failed experiments.

• CAG (CMV enhancer + chicken beta-actin) — ubiquitous, strong, relatively resistant to in vivo silencing. The default choice when broad, durable expression is needed.
• EF1α — ubiquitous, moderate strength; generally resistant to methylation silencing in vivo.
• CMV — ubiquitous and very strong in vitro, but frequently silenced in vivo (especially in liver and brain) due to CpG methylation. Useful for short-term in vitro work; less reliable for chronic in vivo experiments.
• hSyn (human synapsin) — pan-neuronal; restricts expression to neurons even when the capsid transduces non-neuronal cells. Widely used in neuroscience.
• CaMKIIα — excitatory (glutamatergic) neurons; useful for cell-type–specific CNS studies.
• TBG (thyroxine-binding globulin) — hepatocyte-specific; standard for liver-directed metabolic disease studies.
• GFAP — astrocytes; pairs well with AAV5 or AAV9 for glial-targeted CNS work.

(d) Transgene Packaging Capacity

The ITR-to-ITR packaging limit is approximately 4.7 kb. Transgene cassettes (promoter + coding sequence + polyadenylation signal) must fit within this window. Large genes pose a real challenge: the dystrophin coding sequence is ~11 kb and requires a truncated microdystrophin strategy. For transgenes approaching the limit, dual-vector trans-splicing approaches can extend effective capacity but at the cost of reduced transduction efficiency. Plan your cassette size early in vector design.

A related design choice is the genome format. Standard single-stranded AAV (ssAAV) uses the full ~4.7 kb capacity but must be converted to double-stranded DNA by the host cell before expression begins. Self-complementary AAV (scAAV) packages a pre-annealed double-stranded genome that skips this rate-limiting step — giving faster, often stronger early expression — but at roughly half the usable payload:

In short, scAAV trades capacity for speed: choose it for small cassettes (e.g., short reporters, shRNA, or compact recombinases) where rapid, robust expression matters, and ssAAV when the transgene needs the full packaging window (McCarty, Mol Ther 2008).

(e) Route of Administration

Delivery route interacts strongly with serotype to determine final biodistribution:

• Intracranial / stereotaxic injection — local CNS transduction; all major neurotropic serotypes work; AAV2, AAV9, PHP.eB, AAV2/1 commonly used.
• Intravenous (systemic) — broad distribution; liver preferentially transduced with most serotypes; CNS transduction requires BBB-crossing serotypes (AAV9, PHP.eB in C57BL/6 mice).
• Intramuscular — local muscle transduction; AAV2/1, AAV8, AAV9 efficient.
• Subretinal / intravitreal — retina; AAV2 historically standard; AAV5, AAV8, and engineered variants also used.
• Intrathecal — spinal cord and motor neurons; AAV9 and AAVrh10 used clinically.
• Intranasal — lung and olfactory epithelium; AAV5 preferred.

Pre-Existing Immunity: A Key Consideration

Wild-type AAV is ubiquitous in the environment and infects humans subclinically. As a result, a substantial fraction of the general population carries pre-existing neutralizing antibodies (NAbs) against common AAV serotypes. Seroprevalence studies in US and European populations report NAb rates of approximately 30–80% for AAV2 and approximately 3–15% for AAV5, with other serotypes falling between these extremes — though estimates vary by assay, titer threshold, and geographic population (Mingozzi & High, Blood 2013).

Practical implications:

• In clinical trials: patient screening for anti-AAV NAbs is now standard practice. NAb-positive patients are typically excluded from trials involving systemic administration, because even low NAb titers can dramatically reduce vector transduction.
• In rodent models: laboratory mice are not naturally exposed to human AAV serotypes and do not have pre-existing AAV NAbs. This is generally not a concern in naïve mice.
• Re-dosing: a first dose of rAAV generates a robust anti-capsid antibody response. Re-administration of the same serotype is largely neutralized by these antibodies — one of the most significant unresolved challenges in AAV gene therapy for diseases requiring repeat dosing. Switching serotypes or engineering novel capsids are active research strategies to address this limitation.

AAV Production: How Research-Grade Vectors Are Made

Understanding production helps you evaluate vector quality. Three main systems are in use:

1. Triple plasmid transfection of HEK293T cells — the standard research-lab method. Three plasmids are co-transfected: (1) the transgene plasmid carrying ITR-flanked GOI; (2) a packaging plasmid supplying Rep2 and the chosen Cap; (3) a helper plasmid providing adenoviral helper functions (E2A, E4, VA RNA). Virus is harvested from cells and medium 48–72 h post-transfection.
2. Baculovirus / Sf9 insect cell system — scalable and used for large-batch manufacturing; gene expression is driven by baculovirus promoters.
3. HSV helper virus system — used in some clinical-scale manufacturing platforms.

After production, vectors are purified by iodixanol density gradient ultracentrifugation (most common in research) or CsCl gradient; HPLC affinity chromatography is used in clinical manufacturing. Key quality controls include: vg titer by qPCR, empty capsid ratio by analytical ultracentrifugation or transmission electron microscopy, protein purity by silver-stained SDS-PAGE, and (for clinical use) sterility and endotoxin testing.

Producing high-quality, high-titer AAV in-house typically takes 4–8 weeks from plasmid to purified, titered vector. Pre-manufactured, QC-tested ready-to-use vectors from BioHippo's AAV catalog (in partnership with Vector Biolabs) eliminate this bottleneck entirely.

Ready-to-Use AAV Vectors from BioHippo

BioHippo offers an extensive catalog of pre-titered, QC-tested adeno-associated virus vectors spanning all major serotypes, promoters, and common transgenes — from fluorescent reporters (GFP, mRFP) and recombinases (Cre, Flp) to shRNA controls and custom payloads. Browse ready-to-use vectors by serotype:

• AAV2/1 series — broad CNS and muscle tropism; CMV, CAG, hSyn, and U6 promoter options; available as single-stranded (ss) or self-complementary (sc) AAV
• AAV2/2 series — canonical benchmark serotype; moderate neuronal tropism; full reporter and recombinase lineup
• AAV5 series — CNS, astrocyte, and lung tropism; low human seroprevalence
• AAV8 series — liver-directed delivery; CMV promoter options
• AAV9 series — systemic CNS and muscle targeting; available with CAG, CMV promoters
• AAV-PHP.eB series — exceptional brain-wide coverage in C57BL/6 mice after IV delivery; CMV, CAG, hSyn promoters available

All BioHippo AAV vectors are manufactured under strict QC, titered by qPCR, and shipped at ≥1×10¹³ vg/mL ready for direct in vivo use. Browse the full AAV collection or request a quote for custom serotype or transgene combinations.

For researchers using AAV to express genetically encoded sensors, see our related post on GRAB sensors for real-time neurotransmitter detection.

Frequently Asked Questions About AAV

What is AAV and how does it differ from other viral vectors?

AAV (adeno-associated virus) is a small, non-enveloped, ssDNA virus of the family Parvoviridae repurposed as a gene delivery vector. Compared with other vectors: lentivirus integrates stably into the host genome (useful for dividing cells, but carries insertional mutagenesis risk); adenovirus delivers transgenes episomally but provokes strong innate and adaptive immune responses; rAAV occupies a middle ground — predominantly episomal (so minimal integration risk), low immunogenicity, and capable of durable expression in post-mitotic tissues. It cannot replicate without a helper virus and carries no viral coding genes in recombinant form.

How do I choose the right AAV serotype?

Start with your target tissue. For mouse CNS with IV delivery, AAV9 or PHP.eB (C57BL/6 only) are the top choices. For liver, use AAV8. For retina, AAV2 or AAV5. For muscle, AAV2/1 or AAV9. Then consider species compatibility (PHP.eB does not work in primates or non-C57BL/6 mice), pre-existing seroprevalence (if working with non-human primates or patient-derived cells, AAV5 has the lowest seroprevalence), and whether you need IV or local delivery. Consult our serotype comparison post for a detailed side-by-side analysis.

What titer of AAV should I use for my experiment?

Titer requirements vary widely by serotype, route, tissue, and transgene. As a starting point: in vivo stereotaxic CNS injection typically uses 1×10¹² to 2×10¹³ vg/mL at volumes of 0.3–1 µL per site. IV injection in mice commonly uses a dose of 1×10¹¹ to 5×10¹¹ total vg per animal. In vitro transduction of cultured neurons typically requires a multiplicity of infection (MOI) of 10⁴ to 10⁵ vg/cell. Always pilot a dose range. For a full discussion of how to calculate and validate working titers, see our AAV titer guide.

Is AAV safe to work with in the laboratory?

Recombinant AAV vectors carrying a non-pathogenic transgene are classified as BSL-1 to BSL-2 in most institutional biosafety guidelines — but the exact classification depends on the transgene insert (an oncogene or a toxin gene changes the risk profile) and your institution's biosafety committee (IBC) determination. Consult your IBC before starting any AAV work. Wild-type AAV has no known pathogenicity in humans, and rAAV lacks viral coding sequences. Standard laboratory PPE and biological waste disposal practices apply.

What are the limitations of AAV gene delivery?

Five limitations every researcher should understand: (1) Packaging capacity — the ~4.7 kb limit excludes many large genes without truncation or dual-vector strategies; (2) Episomal dilution in dividing cells — expression declines as cells divide, limiting use in proliferating populations; (3) Pre-existing immunity — NAbs in human subjects (and non-human primates) can block transduction; (4) Re-dosing — a primary AAV dose generates anti-capsid antibodies that largely neutralize a second dose of the same serotype; (5) Onset of expression — single-stranded rAAV requires second-strand synthesis by the host cell, adding days to weeks before peak expression; scAAV bypasses this but at the cost of halved packaging capacity (~2.4 kb).