(T1030-01-06) Detection and Quantification of AAV Genome: Effects of Sample Pretreatment and Cross Platform Comparison of ddPCR and qPCR

Purpose: Viral vectors are increasingly being used to treat or cure genetic diseases. Adeno-associated viruses (AAVs) have proved to be a useful tool in this endeavor due to their ease of manipulation and production, long-term efficacy, wide range, yet tunable tissue tropism, and excellent safety profile. Currently, there are three AAV-based therapies approved by the United States Food and Drug Administration, with the most recent, Hemgenix®, approved in December 2022 for the treatment of hemophilia B. Accurate detection and quantification of any viral vector is an integral part the discovery, manufacturing, and pre-clinical/clinical testing phases of product development. While there are many methods to quantify different biological characteristics of a vector, molecular methods for genome copy (GC) number quantification are widely used due to their high accuracy and ease of implementation. As a result, qPCR and ddPCR based assays are increasingly being used for regulated GC quantification. However, limited regulatory requirements for PCR-based methods raise questions on the appropriateness of the available platforms and guidance for assay validation. In addition, recommendations vary with respect to pre-PCR treatment of samples. DNase I treatment is commonly used to remove residual, free DNA for more accurate quantification of encapsidated genomes, which are more likely to reflect active vector numbers. Some protocols recommend the addition of a proteinase K treatment to assist with capsid denaturation, while other protocols rely on incubation at 95°C. Restriction enzyme treatment is recommended for ddPCR amplification of AAV to reduce the complex secondary structure of the single stranded DNA in order to reduce droplet rain and improve accuracy. Here, we examine the effects of sample pre-treatment steps on the quantification of GC of a commercially available AAV2 reference material via ddPCR and qPCR. Finally, we discuss the pros and cons of both platforms with respect to assay design, performance, and costs.

Methods: AAV2 reference material containing the eGFP transgene was obtained from a commercial source. Primers and hydrolysis probe targeting the eGFP transgene were obtained from IDT. The same primers and probes were used in both the qPCR and ddPCR assays. qPCR was performed on a QuantStudio 7 Pro (ThermoFisher) and ddPCR was performed on the QX200 AutoDG System (BioRad). Amplification conditions were optimized using a double-stranded synthetic DNA obtained from IDT for both qPCR and ddPCR. The lower limit of detection for both platforms was established utilizing serial dilutions of the synthetic DNA standard. qPCR and ddPCR were performed on serially diluted AAV2 reference material that had been pre-treated with various combinations of DNase I, Proteinase K, and/or heat. For ddPCR, the effect of the addition of the Hae III restriction enzyme was also assessed. For qPCR, GC numbers were determined by interpolation of the Cq values with a standard curve prepared from serial dilutions of the synthetic dsDNA. Since AAV2 contains a single-stranded genome, final GC copy numbers determined by qPCR were multiplied by two to adjust for the use of a double-stranded DNA standard. For ddPCR, GC numbers were determined via absolute quantification.

Results: After optimization, both qPCR and ddPCR platforms gave consistent and robust amplification. However, amplitude plots of ddPCR results indicated significant rain in the reactions which was reduced by the addition of Hae III restriction enzyme to the ddPCR master mix (Figure 1). As a result, GC number determined in ddPCR increased with Hae III treatment (8.95E+06 GC without vs. 1.37E+08 GC with). Similar results were obtained between both qPCR and ddPCR platforms (Table 1). However, AAV GC numbers varied with respect to pretreatment of the material. On both platforms, simple heating to open AAV capsids resulted in the highest GC numbers while DNase I treatment reduced GC numbers, consistent with removal of unencapsidated DNA. In all conditions, results from GC number determinations were higher than those reported by the manufacturer (15-77% higher, depending upon pretreatment conditions).

Conclusion: The results reported here demonstrate the similarities between qPCR and ddPCR platforms in the ability to quantify AAV2 GC numbers. However, these results also indicate that sample pre-treatment steps will affect the results of AAV2 GC quantification in both qPCR and ddPCR platforms. The choice of platform and sample pretreatment protocols will depend on the context of use of the assay. For example, applications which require a broad dynamic range of quantification may focus on the development of qPCR-based methodology while for applications that require higher sensitivity, ddPCR-based methods may be more appropriate. In addition, selection and characterization of reference materials to be used to assess assay performance should be considered. Method development must carefully evaluate these parameters for robust assay performance and accurate results within the context of use.