AAV is a non-enveloped small single-stranded DNA virus, its genome is 4.7 kb long, belongs to the genera of the parvoviridae and has a life cycle of two phases: latent infection and lytic phase [1]. Interestingly, AAV is non-pathogenic in humans [2]. The AAV reproductive and lytic phases in mammalian cells require a helper virus, such as adenovirus, herpesvirus, or vaccinia virus [1]. After viral entry, human AAV integrates its genome into q arm of chromosome 19, between q13.3 and qter, which is also termed AAVS1 [3]. The AAV genome integration into the AAVS1 is safe and does not cause insertional mutagenesis events that may lead to carcinogenesis [2, 4], although such optimism is under scrutiny [5, 6].

[enlarge]

Figure 1. A schematic representation of wild-type AAV and recombinant AAV-derived vector system.

AAV can also infect non-dividing cells and has a wide range of cell tropism [2]. The AAV genomic organization is relatively simple. Two inverted terminal repeats (ITRs) are situated at the 5’ and 3’end of the viral genome, which contains only two genes: Rep and Cap (Fig. 1). The ITRs contain the viral promoter and the signals that are responsible of the integration of the viral genome into the cellular chromosome [2]. Rep gene encodes four structural proteins termed Rep 78, Rep 68, Rep 52 and Rep 40, whereas Cap encodes three structural protein termed VP1, VP2 and VP3 [2]. Rep 68 and Rep 78 mediate the AAV genome integration into the safe AAVS1 site of the cellular chromosome [2]. Ogden PJ et al systematically evaluated the fitness landscape of mutant AAV2 Cap gene; AAV2 forms the first U.S. FDA-approved gene therapy [7]. Other experiments also used AAV2 [8]. More AAV-based gene therapies have been approved by FDA [9]. Chan KY et al optimized AAVs for specific gene delivery to central (AAV-PHP.eB) and peripheral (AAV-PHP.S) nervous systems [10]. Y Lu et al introduced OCT4, SOX2 and KLF4 as a polycistron in an AAV9 vector to regenerate retinal ganglion cells in mice [11]. Rauch JN et al infected mice to deliver shRNA against LRP1 with AAV-PHP.eB [12]. J Hordeaux et al integrated an microRNA 183 target sequence in AAV vectors to reduce the expression of transgenes in dorsal root ganglion, and thus reduce the dorsal root ganglion toxicity in AAV-based gene therapies [13].

The interest in AAV-mediated gene transfer is motivated by two major factors: the non-pathogenicity of the wild-type virus in humans and its ability to infect non-dividing cells [1, 2]. The engineering of AAV-derived vector systems only requires the removal of the Rep and Cap genes from the shuttle vector, which contains the two ITRs, an internal exogenous enhancer/promoter that drives the expression of the transgene and a poly A tail (Fig. 1). Rep and Cap genes are placed in a separate plasmid, which has to be co-transfected with the shuttle vector into packaging cells [2]. Following the two-plasmid co-transfection into packaging cells, an adenovirus infection is required to induce AAV lytic phase, which is induced by the adenoviral early E1 and E4 genes [2].

More recent protocols do not require the use of a helper virus to generate AAV-derived vector particles. These protocols utilize a plasmid encoding for the necessary helper virus factors, which is co-transfected into packaging cells, along with the shuttle vector and the plasmid that containing the Rep and Cap genes [15]. The possibility of using a three-plasmid co-transfection without the need of a helper virus has improved the clinical grade applications of AAV-derived vectors.

AAV vectors are commonly used in experiments involving brains or neurons. S Espinoza et al designed and injected into mouse brains an AAV9 vector expressing SINEUP RNA for GDNF expression [16]. Zhang X et al injected AAV2/9-CAG-DIO-taCasp3-TEVp, AAV2/9-hsyn-DIO-GCaMP6m, AAV9-EF1α-DIO-hM3D(Gq)-mCherry, AAV9-EF1α-DIO-hM4D(Gi)-mCherry, AAV9-EF1α-DIO-eYFP and AAV2/9-EF1α-DIO-ChR2-mCherry into the central nucleus of the amygdala and the paraventricular nucleus in mice to study the neuronal control of humoral immune responses in spleen [17]. Moya IM et al injected AAV8.TBG.PI.Cre.rBG from UPenn Core into mice to express Cre recombinase in hepatocytes [18]. Siciliano CA et al injected AAV5-CaMKIIα-eNpHR3.0-eYFP or AAV5-DIO-ChR2-eYFP, and their controls AAV5-CaMKIIα-eYFP and AAV5-DIO-eYFP into mouse medial prefrontal cortex to conduct optogenetic experiments [19]. Patzke C et al generated conditional synapsin-1 knockout human ES H1 cells with adeno-associated viruses through cre-recombination [20]. Marshel JH et al transduced neurons with an adeno-associated viral vector carrying enhanced YFP and optogenetic gene ChRmine along with the trafficking sequence, ER export signal, and the CaMKIIalpha promoter [21]. Szőnyi A et al injected into mouse brains AAV2/1-EF1a-DIO-GCaMP6f, AAV2/5-EF1αDIO-eYFP, AAV2/5-EF1α-DIO-mCherry, AAV2/5-CAG-FLEX-ArchT-GFP, AAV2/5-EF1α-DIO-hChR2(H134R)-eYFP for anterograde tracing and optogenetic experiments to study the role of brainstem nucleus incertus GABAergic cells in contextual memory formation [22]. Marvin JS et al transfected HEK293 cells with the adeno-associated virus (AAV) plasmid coding for the glutamate sensor iGluSnFR, helper plasmids encoding rep and cap genes (pRV1 and pH21), and adenoviral helper pFΔ6 from Stratagene using the calcium phosphate transfection method, and purified the virus particles from cell lysate by HiTrap heparin HP columns from GE Healthcare [23]. Zott B et al injected these kinds of GluSnFr viral constructs into mouse hippocampal CA1 regions to study the effect of beta-amyloid peptides on glutamate levels in brain neurons in vivo [24]. AAV2/5 serotype is thought to infect astrocytes preferentially in mouse brains [25]. M Aubert et al evaluated the infectiveness of four serotypes: AAV1, AAV8, AAV-PHP.S, and AAV-Rh10 among neurons of the peripheral nerve system [26].

It is of interest to note that multiplicity of Infection/MOI or AAV titer differs up to an order when determined by two different appraches: ddPCR and qPCR [27].

The advantages and drawbacks of AAV-mediated gene transfer are reported in Table 1. Two main shortcomings of AAV-derived vectors are related to the limited capacity to accommodate transgenes, which is in the range of 4 kb, and insertional mutagenesis [2]. Dual AAVs with split inteins can remedy the limited capacity issue [14]. Unfortunately, AAV-derived vectors integrate their genomes randomly into cellular chromosomes, in contrast to wild-type AAV [4]. Some studies reported that AAV-mediated gene transfer induced hepatocellular carcinomas and angiosarcomas in mice [4]. Worryingly, a number of studies showed that the random integration of AAV-derived vectors occurs preferentially at actively transcribed genes of the target cell genome, which, in turn, may lead to critical DNA damage and/or silencing of tumor suppressor genes [4].

A list of companies and organization that provide the components for the production of AAV-derived vectors is reported in Table 2. Some of them also publish standardized protocols such as Salk GT3 Core [37].

The quantification of AAV-derived vector particles can help control and optimize the efficiency of AAV-mediated gene transfer into mammalian cells [15]. Several systems are used to assess the titers of AAV vector stocks, such as dot-blot hybridization [39], Southern blot [40], UV-based spectrophotometry [41], enzyme-linked immunosorbent assay (ELISA) [42], PicoGreen-based fluorimetry [43], quantitative real-time polymerase chain reaction (qPCR) [44] and droplet digital PCR (ddPCR) [45]. UV-based spectrophotometry measures the UV absorbance of denatured AAV-derived vector particles in solution [41]. The fluorescent dye PicoGreen stains the encapsidated AAV transfer vectors genome [43]. Droplet digital PCR (ddPCR) does not require a standard curve, as it can quantify directly and with high precision specific types of DNA copies [45]. Dynamic light scattering can assess the size of AAV-based vector particles and determine if some particles have formed aggregates. Next-generation sequencing (NGS) can examine the possible mis-incorporation of genetic elements inside AAV-derived vector particles [46, 47]. A number of assays are commercially available and are listed in Table 3.

Some companies provide purified AAV-derived vector particles encoding green fluorescent protein (GFP), which can be utilized to assess the efficiency of AAV-mediated gene transfer in a variety of mammalian cell types (Table 3). GFP is encoded in different serotypes of AAV-based vectors, in order to compare their gene transduction efficiency in particular cell types.

Studies are currently underway for the full characterization of in vivo transduction properties of various subtypes of AAV-based vector systems. This may require the application of spatial transcriptomics, which is an emerging technology that allows for the readout of thousands of mRNA sequences in tissue sections, without altering the overall context of the three-dimensional structure of the analyzed tissues [48-61]. Indeed, spatial transcriptomics can also be utilized for the three-dimensional analyses of viral vectors-encoded transgenes among various types of transduced cells and/or tissues, with a particular emphasis on the central nervous system [62-65].

A recent report has improved and characterized a promising ultrasensitive system, which was based on sequential fluorescence in situ hybridization (USeqFISH), to determine the three-dimensional transcriptomic profiling of endogenous and viral vector-encoded mRNA with a short barcode [66]. The study was conducted in the following animal models: mouse, marmoset and rhesus macaque. Single doses of AAV-derived vector particles were injected into the central nervous systems of the animals. A pool of six serotypes AAV serotypes was used in the experiments. The AAV-derived vectors pool included a new variant, termed AAV-PHP.AX, which has shown a particular efficiency in transducing neurons and astrocytes in the in vivo system [66-68]. The AAV-encoded mRNAs are typically abundant in transduced cell populations and may vary from 10 to 100 transcripts per cell [69]. The copy numbers of AAV-encoded transcripts are comparable to the expression levels of endogenous cellular genes. Usually, the spatial transcriptomics modalities involve dozens of probes per mRNA target to ensure an adequate specificity and signal intensity, which may require a rather long barcode ranging from 0.5 to 1 kilobase. The length of the barcode may pose an obstacle for the detection of AAV-encoded mRNA, due to the limited packaging capacity of AAV vector particles, which is lower than 4.7 kb. With the intent to reduce the length of the probes for the specific detection of the AAV-encoded mRNA barcode, USeqFISH combines signal amplification with sequential labeling to produce brighter and more sensitive mRNA-associated signals than conventional amplification systems for spatial transcriptomic analysis. The signal amplification can be based either on rolling-circle amplification (RCA) [70], or hybridization chain reaction (HCR) [71]. The increment of the signal sensitivity results in the use of just four, or even less, labelled oligonucleotide probes for the detection of each target gene, along with cell type-specific markers, in different regions of the mammalian brain [59, 60, 72]. Signal amplification systems combined with sequential labeling with shorter mRNA barcodes may allow for the simultaneous readout of approximately 50 genes in brain tissue sections.