Copackaging of multiple adeno-associated viral vectors in a single production step

Abstract

Limiting factors in large preclinical and clinical studies utilizing adeno-associated virus (AAV) for gene therapy are focused on the restrictive packaging capacity, the overall yields, and the versatility of the production methods for single AAV vector production. Furthermore, applications where multiple vectors are needed to provide long expression cassettes, whether because of long cDNA sequences or the need of different regulatory elements, require that each vector be packaged and characterized separately, directly affecting labor and cost associated with such manufacturing strategies. To overcome these limitations, we propose a novel method of vector production that allows for the packaging of multiple expression cassettes in a single transfection step. Here we combined two expression cassettes in predetermined ratios before transfection and empirically demonstrate that the output vector recapitulates the predicted ratios. Titration by quantitative polymerase chain reaction of AAV vector genome copies using shared or unique genetic elements allowed for delineation of the individual vector contribution to the total preparation that showed the predicted differential packaging outcomes. By copackaging green fluorescent protein (GFP) and mCherry constructs, we demonstrate that both vector genome and infectious titers reiterated the ratios utilized to produce the constructs by transfection. Copackaged therapeutic constructs that only differ in transcriptional elements produced a heterogeneous vector population of both constructs in the predefined ratios. This study shows feasibility and reproducibility of a method that allows for two constructs, differing in either transgene or transcription elements, to be efficiently copackaged and characterized simultaneously, reducing cost of manufacturing and release testing.

FIG. 1.

Differential packaging of expression cassettes combined before transfection. (A) The vectors contained different transgenes that were used for delineation. DNA was amplified from each preparation and ran on 2% agarose gel. Green fluorescent protein (GFP) band is 171 bp and mCherry is 191 bp. Lanes 1–3 contain DNA amplified from vectors in crude lysate (postbenzonase treatment); lanes 4–6 contain DNA amplified from vectors purified via iodixanol gradient originating from the crude lysates in lanes 1–3. Vector constructs were copackaged in AAV9 at GFP-to-mCherry ratios of 1:9, 1:1, and 9:1, respectively. Gel is one representative image of three separate copackaging experiments. (B) Quantitative PCR using transgene-specific primers on iodixanol-purified GFP or mCherry vectors was performed to determine the respective contribution of each individual vector in the total preparation when copackaged at 1:9, 1:1, and 9:1 ratios, respectively. Each vector yield is expressed as a percentage of total vector genome, with 100% obtained from the summation of the titers determined using transgene-specific primers for either GFP or mCherry. Data represent the average of three separate experiments for each ratio.

FIG. 2.

Total vector genomes resulting from copackaging. DNA was extracted from postbenzonase-treated crude lysates (Crude) or after iodixanol purification (Purified) of GFP and mCherry vectors copackaged at 1:9, 1:1, or 9:1 ratios, respectively. Total vector genome titer was determined either directly, using a common CMV enhancer (Single Primer Set), or from the summation of each vector titer using transgene-specific primers (Double Primer Set). Data represent the average total titer from crude (final volume of 3 mL) or purified samples (final volume of 0.2 mL) at each ratio assayed in triplicate.

FIG. 3.

In vitro characterization of copackaged reporter vectors. The percent contribution of either AAV9-GFP or AAV9-mCherry to the total infectious titer (GFP+mCherry) was determined via single-cell fluorescence assay in C12 cells. Data represent the average of two separate experiments for each ratio.

FIG. 4.

Copackaging of therapeutic constructs. (A) AAV9-LSP-coGAA and AAV9-DES-coGAA were copackaged and purified at 1:9 (lanes 1, 4, and 7), 1:1 (lanes 2, 5, and 8), and 9:1 ratios (lanes 3, 6, and 9), respectively. AAV9-LSP-coGAA band is 288 bp and AAV9-DES-coGAA is 453 bp. DNA was amplified from each preparation and ran on 1.5% agarose gel. (B) DNA extracted from copackaged AAV9-LSP-coGAA and AAV9-DES-coGAA was subjected to quantitative PCR to determine the respective contribution of each individual vector in the total preparation when copackaged at 1:9, 1:1, and 9:1 ratios, respectively. Each vector yield is expressed as a percentage of total vector genome, with 100% obtained from the summation of the titers determined using promoter-specific primers for either LSP or DES. Data represent the average of three separate experiments for each ratio. LSP, liver-specific promoter; DES, desmin promoter; coGAA, human codon-optimized acid α-glucosidase.