Viral Vectors 101: AAV Variables That Matter

Using AAV vectors in complex biological systems can be tricky at best, and downright infuriating at worst. While it is tempting to just dive right in and start injecting your virus, a successful AAV experiment starts with validation and optimization. Although there are different ways to ultimately achieve success, we find it is usually faster and easier to do the validation first. 

AAV-mediated gene expression can differ extensively depending on the experimental system used, so you’ll need to know what expression pattern you need, and the factors in your system that can affect expression, before starting the selection process. Factors to think about are (1) your experimental conditions and measurements; (2) the cells you’re targeting; (3) how much expression is needed to read your experimental output; and (4) the consequences to your experiment if there’s off-target expression.  

Once you know that, you’re ready to start selecting, testing, and optimizing three key variables of an AAV vector: serotype, promoter, and titer.

Figure 1: Flow chart depicting a) steps for selecting serotypes and promoters to test in your AAV system and b) steps for testing and optimizing your serotypes, titer, and promoter. Note that several steps in testing are effected by multiple AAV variables. 

Serotypes

Serotype refers to the different protein compositions of AAV capsids, which determine how a virus enters the cell. Different serotypes utilize a number of different receptors, co-receptors, and other mechanisms to gain entry into the cell (Dudek et al., 2018; Korneyenkov & Zamyatnin, 2021), meaning what works great in one system or cell type can completely fail in a different system (Issa et al., 2023). This can be particularly burdensome in the brain where many different cell types, or even different phenotypes of the same cell type, are intermingled in the same brain nucleus: a serotype that works great in a GABA neuron, for instance, may not transduce the adjacent norepinephrine neuron at all. There may be more than one serotype that will efficiently transduce your cells. In fact, we usually recommend testing more than one. 

Most AAV serotypes have broad tropism (the ability of a virus to enter a particular cell type), but there are differences in tropism between animal species or even strains of the same species. What works well in a mouse may not work as well in a rat or non-human primate. Likewise, in vitro efficacy doesn’t always translate to in vivo efficacy. You’ll therefore need to test potential AAV vectors in your specific experimental system, rather than relying on other systems as proxies. 

Finally, the size and receptor interactions of a serotype can affect spread from the injection site  (Kanaan et al., 2017), axonal transport  (Murlidharan et al., 2014), and postsynaptic transmission (Oh et al., 2020). For example, AAV1 and AAV9 tend to spread farther from the injection site than AAV2 (Watakabe et al., 2015) (Figure 1). You’ll need to consider if such properties could help, or harm, your experimental goals when designing and testing your AAV. 

Figure 2: Comparison of AAV serotypes 1, 2, 5, 6, 8, and 9 in two different brain regions.  Image modified from Aschauer et al., 2013, Figure 1. Used under Creative Commons Attribution License.

Within the nervous system, remember that some serotypes can be transported retrogradely and/or anterogradely along the axon. Others, like AAV1 and AAV9, can also transduce postsynaptic cells when used at very high titers (Zingg et al., 2017), which can lead to cells in far away brain regions expressing your gene of interest. This can cause issues in tracing studies, where it could potentially label cells not directly connected to the injection site, or when using chemogenetic tools, where modification of cellular activity outside of your experiment site or circuit can occur.

Promoters

The promoter controls how the gene of interest will be expressed; it can be ubiquitous, cell-type selective, or inducible. 

Ubiquitous promoters 

These often-strong promoters offer the most flexibility for expression in a large number of cell types and are useful if you're going to be working in several different systems or want to transduce a large number of different cells. Promoters such as chicken-beta-actin or CAG can drive strong early gene expression, but can become methylated over time, resulting in decreased expression (Damdindorj et al., 2014; Nieuwenhuis et al., 2021). It is important to note that this expression pattern can be affected by the cell and/or tissue type, so you’ll need to test for this in your system if you’re measuring expression over a period of weeks (in vitro) or months (in vivo). 

Selective promoters

Selective promoters, such as synapsin, glial fibrillary associated protein (GFAP), or tyrosine-hydroxylase (TH), can limit which cells express your gene. Off-target expression can still occur (Fig. 3) - for example, some neurons can transiently express TH - so you’ll need to check for this in your system and then test each vector before starting your experiments (Rolland et al., 2016; Su et al., 2004; Taschenberger et al., 2017).

Figure 3: Comparison of two astrocyte selective promoters demonstrating off-target expression in both oligodendrocytes (Olig2) and neurons (NeuN). Image from Griffin et al., 2019, Figure 4. Used under Creative Commons Attribution 4.0 International License.

Inducible promoters 

This type of promoter allows you to decide when to turn on, or turn off, gene expression by administering an inducing agent like doxycycline (Favre et al., 2002). They are useful when you need to control the timing of expression in your experiment. Some inducible constructs can exhibit leaky expression in the absence of the inducing drug, so be sure to test for this when validating your viral vector.