Production of Adeno-Associated Virus Vectors in Cell Stacks for Preclinical Studies in Large Animal Models

Summary

Here we provide a detailed procedure for large-scale production of research-grade AAV vectors using adherent HEK 293 cells grown in cell stacks and affinity chromatography purification. This protocol consistently yields >1 x 10¹³ vector genomes/mL, providing vector quantities appropriate for large animal studies.

Abstract

Adeno-associated virus (AAV) vectors are among the most clinically advanced gene therapy vectors, with three AAV gene therapies approved for humans. Clinical advancement of novel applications for AAV involves transitioning from small animal models, such as mice, to larger animal models, including dogs, sheep, and nonhuman primates. One of the limitations of administering AAV to larger animals is the requirement for large quantities of high-titer virus. While suspension cell culture is a scalable method for AAV vector production, few research labs have the equipment (e.g., bioreactors) or know how to produce AAV in this manner. Moreover, AAV titers are often significantly lower when produced in suspension HEK 293 cells as compared to adherent HEK293 cells. Described here is a method for producing large quantities of high-titer AAV using cell stacks. A detailed protocol for titering AAV as well as methods for validating vector purity are also described. Finally, representative results of AAV-mediated transgene expression in a sheep model are presented. This optimized protocol for large-scale production of AAV vectors in adherent cells will enable molecular biology laboratories to advance the testing of their novel AAV therapies in larger animal models.

Introduction

Gene therapy utilizing adeno-associated virus (AAV) vectors has made huge strides over the past three decades^1,2. Demonstrated improvements in a diverse range of genetic diseases, including congenital blindness, hemophilia, and diseases of the musculoskeletal and central nervous system, have brought AAV gene therapy to the forefront of clinical research^3,4. In 2012, the European Medicines Agency (EMA) approved Glybera, an AAV1 vector expressing lipoprotein lipase (LPL) for the treatment of LPL deficiency, making it the first marketing authorization for a gene therapy treatment in either Europe or the United States⁵. Since then, two additional AAV gene therapies, Luxturna⁶ and Zolgensma⁷, have received FDA approval, and the market is expected to expand quickly over the next 5 years with as many as 10-20 gene therapies expected by 2025⁸. Available clinical data indicate that AAV gene therapy is a safe, well-tolerated, and efficacious modality making it one of the most promising viral vectors, with above 244 clinical trials involving AAV registered with ClinicalTrials.gov. The increasing interest in clinical applications involving AAV vectors requires robust and scalable production methods to facilitate the evaluation of AAV therapies in large animal models, as this is a critical step in the translational pipeline⁹.

For AAV vector production, the two main requirements are the AAV genome and the capsid. The genome of wild-type (wt)-AAV is single-stranded DNA that is approximately 4.7 kb in length¹⁰. The wt-AAV genome comprises inverted terminal repeats (ITRs) found at both ends of the genome, which are important for packaging, and the rep and cap genes¹¹. The rep and cap genes, necessary for genome replication, assembly of the viral capsid, and encapsulation of the genome into the viral capsid, are removed from the viral genome and provided in trans for AAV vector production¹². The removal of these genes from the viral genome provides room for therapeutic transgenes and all the necessary regulatory elements, including the promoter and polyA signal. The ITRs remain in the vector genome to ensure proper genome replication and viral encapsulation^13,14. To improve the kinetics of transgene expression, AAV vector genomes can be engineered to be self-complementary, which mitigates the need for conversion from single-stranded to double-stranded DNA conversion during AAV genome replication, but reduces the coding capacity to ~2.4 kb¹⁵.

Beyond AAV genome design, capsid serotype selection determines the tissue and cell tropism of the AAV vector in vivo². In addition to tissue tropism, different AAV serotypes have been shown to display different gene expression kinetics¹⁶. For example, Zincarelli et al.¹⁷ classified different AAV serotypes into low expression serotypes (AAV2, 3, 4, 5), moderate expression serotypes (AAV1, 6, 8), and high expression serotypes (AAV7 and 9). They also categorized AAV serotypes into slow-onset expression (AAV2, 3, 4, 5) or rapid-onset expression (AAV1, 6, 7, 8, and 9). These divergent tropisms and gene expression kinetics are due to amino acid variations in the capsid proteins, capsid protein formations, and interactions with host cell receptors/co-receptors¹⁸. Some AAV capsids have additional beneficial characteristics such as the ability to cross the blood-brain barrier following intravascular administration (AAV9) or reside in long-living muscle cells for durable transgene expression (AAV6, 6.2FF, 8, and 9)^19,20.

This paper aims to detail a cost-effective method for producing high-purity, high-titer, research-grade AAV vectors for use in preclinical large animal models. Production of AAV using this protocol is achieved using dual-plasmid transfection into adherent human embryonic kidney (HEK)293 cells grown in cell stacks. Furthermore, the study describes a protocol for heparin sulfate affinity chromatography purification, which can be used for AAV serotypes that contain heparin-binding domains, including AAV2, 3, 6, 6.2FF, 13, and DJ^21,22.

A number of packaging systems are available for the production of AAV vectors. Among these, the use of a two-plasmid co-transfection systems, in which the Rep and Cap genes and Ad helper genes (E1A, E1B55K, E2A, E4orf6, and VA RNA) are contained within one plasmid (pHelper), has some practical advantages over the common three-plasmid (triple) transfection method, including reduced cost for plasmid production^23,24. The AAV genome plasmid containing the transgene expression cassette (pTransgene), must be flanked by ITRs, and must not exceed ~4.7 kb in length. Vector titer and purity can be affected by the transgene due to potential cytotoxic effects during transfection. Assessment of vector purity is described herein. Vectors produced using this method, which yield a 1 x 10¹³ vg/mL for each, were evaluated in mice, hamsters, and ovine animal models.

Table 1: Composition of required solutions. Necessary information, including percentages and volumes, of components needed for various solutions throughout the protocol. Please click here to download this Table.

Protocol

1. Double plasmid transfection of HEK293 cells in cell stacks

1. Thaw a cryo-vial of HEK293 cells in a bead bath set at 37 °C.
   NOTE: Pre-warm complete DMEM to 37 °C while cells are thawing to ensure the cold temperature does not shock cells when plating. Ensure the cells have a low passage number, ideally less than 20, to ensure optimal growth and transfection efficiency. Ensure that the cells are certified to be mycoplasma-free.
2. Transfer the contents of the cryo-vial dropwise into a 15 mL conical tube containing 10 mL of pre-warmed complete DMEM and centrifuge the cells at 500 x g for 5 min.
3. Aspirate the media, and then resuspend the HEK293 cells in 20 mL of pre-warmed complete DMEM. Seed the cells in a 15 cm plate and incubate at 37 °C, with 5% CO_2.
4. Split the cells from one 15 cm plate into three for seeding in the cell culture chamber.
   1. Once cells are 80% confluent, aspirate the media and gently wash the plate with 3 mL of PBS to not disrupt the monolayer. Then, aspirate PBS and add 3 mL of trypsin.
   2. Incubate for 2 min at 37 °C until the cells lift from the plate, and then neutralize trypsin by adding 7 mL of complete DMEM to the plate.
   3. Collect all the media and cells into a 15 mL tube and pellet the cells by centrifuging at 500 x g for 5 min.
   4. Aspirate the supernatant from the 15 mL tube and resuspend the cell pellet in 3 mL of complete DMEM. Add 1 mL to each 15 cm plate containing 20 mL of complete DMEM; gently rock the plates to distribute the cells evenly, and incubate at 37 °C, with 5% CO₂.
5. Once the cells are 80% confluent, repeat steps 1.4.1 and 1.4.2. Collect the supernatant in 50 mL conical tubes, and gently invert the tube to ensure the cells are homogeneous.
6. Determine the cell density by mixing 10 µL of the samples of cells with 10 µL of trypan blue and adding the mixture to a cell counting slide for analysis in the cell counter.
7. Mix 1 L of pre-warmed complete DMEM with the needed cell suspension to seed the cell culture chamber (surface area of 6360 cm²) with 1 x 10⁴ cells/cm². Pour the cell mixture into the cell culture chamber and gently rotate to evenly distribute the cells across each monolayer (Figure 1) and incubate at 37 °C, with 5% CO_2.
8. In addition to the cell culture chamber, plate a 15 cm plate with 1 x 10⁴ cells/cm² as a reference for confluency.
9. Following ~65-h incubation, check the reference plate for confluency-ideally ~80%-90% confluent.
   ​NOTE: Pre-warm complete DMEM for adding to the cell culture chamber at 37 °C.

Figure 1: Maneuvering of cell stack for cell seeding and transfection. For seeding cell stack, start by removing one of the vent caps and pouring in 1 L of pre-warmed complete DMEM with needed quantity of HEK293 cells (A). Evenly distribute cells and media by tightening both vent caps and bring all media to the corner of the cell stack with one of the vent caps and place it in that corner (B), place the cell stack on its side (C), and then turn the cell stack 90° (D) so that the vent ports are up (E). Gently lower the cell stack to its normal horizontal position and ensure all the chambers of the cell stack are completely covered in media (F). When transfecting, unscrew both vent caps and slowly pour out old media into a waste sterile waste container for even flow not to disturb the monolayer of the cells (G). Please click here to view a larger version of this figure.

10. Prepare polyethyleneimine (PEI)/DNA mixture at a concentration ratio of 3:1 (w/w).
    1. Prepare the DNA mixture in a 50 mL conical tube by adding 475 µg of pTrangene and 1425 µg of pHelper to 40 mL of reduced-serum medium to create a 3:1 ratio of pHelper:pTrangene.
       NOTE: The PEI/DNA mixture calculator can be found using Table 2.
    2. Add 5.7 mL of PEI (1 g/L) to the reduced-serum medium and the DNA mixture dropwise. Then, vortex briefly and incubate for 10 min at room temperature.
       NOTE: As PEI/DNA incubates at room temperature, it will become slightly cloudy.
11. After 8 min of PEI/DNA incubation, remove the media from the cell culture chamber.
    ​NOTE: Ensure to loosen both orange caps to maintain a smooth flow of media to avoid dislodging of cells.
12. Add PEI/DNA to 1 L of pre-warmed complete DMEM, and slowly pour the mixture into the cell culture chamber port. Distribute the liquid evenly to all the rows (Figure 1) and incubate for 72 h at 37 °C, with 5% CO₂.

2 Harvesting AAV and chemical lysis of the transfected HEK293 cells

1. Shake the cell culture chamber vigorously to dislodge the cells until the media appears cloudy from dislodged cells and pour into four 500 mL centrifuge tubes.
2. Centrifuge the tubes at 18,000 x g for 30 min at 4 °C to pellet the cells. Pour the clarified supernatant into 1 L polyethylene terephthalate copolyester (PETG) bottle.
   NOTE: If one does not have access to a high-speed centrifuge, centrifuge at 12,000 x g for 40 min. The pelleted cells may not be solid at this speed and will slide as pouring out supernatant.
3. Resuspend the cell pellets in 500 mL centrifuge tubes with 50 mL of lysis buffer and incubate for 60 min at 37 °C.
4. Centrifuge the tubes at 18,000 x g for 30 min, and then transfer the supernatant into the same 1 L PETG bottle. Discard the pelleted cell debris.
   ​NOTE: Purify the clarified supernatant immediately and store at 4 °C for up to 72 h. For longer-term storage, store at -80 °C. Do not store at -20 °C.

3 AAV Vector purification using heparin affinity chromatography

1. Remove the crude lysate from -80 °C and leave at 4 °C overnight to thaw. Once thawed, use a 0.22 µM filter to filter the crude lysate.
2. To passivate the centrifugal concentrator, add 4 mL of filter pre-treatment buffer to a centrifugal concentrator for each heparin sepharose column being used. Passivate the centrifugal concentrator at room temperature for 2-8 h. Set up passivation immediately before the purification steps.
3. Set up the tubing and the pump (Figure 2).
   1. Place the tubing in a peristaltic pump and run 20 mL of 1 M NaOH. Next, run 50 mL of molecular grade water, and then run 50 mL of basal DMEM.
   2. Attach 5 mL heparin sepharose column to tubing and run 25 mL of basal DMEM to remove the preservative.
4. Run 0.2 µM of the filtered crude lysate through the column at a flow rate of 1-2 drops/s.

Figure 2: Set up for peristaltic pump for AAV purification. Run the tubing from the crude lysate, through the peristaltic pump, and into the heparin matrix column. Please click here to view a larger version of this figure.

NOTE: Ensure not to introduce bubbles or allow the column to run dry, as this will compromise the column and prevent the elution of AAV. Discard the column if it runs dry and use a new column for the remainder of crude lysate.

5. Load all of the crude lysate onto the heparin column and use the following solutions to wash the column.
   1. Wash using 50 mL of 1x Hank's Balanced Salt Solutions (HBSS) without Mg²⁺ and Ca²⁺.
   2. Wash using 15 mL of 0.5 % N-Lauroylsarcosine in HBSS without Mg²⁺ and Ca²⁺.
   3. Wash using 50 mL of HBSS without Mg²⁺ and Ca²⁺.
   4. Wash using 50 mL of HBSS with Mg²⁺ and Ca²⁺
   5. Wash using 50 mL of 200 mM NaCl/HBSS with Mg2+ and Ca²⁺.
   6. Elute 5 x 5 mL (25 mL total) with 300 mM of NaCl/HBSS with Mg²⁺ and Ca²⁺ and label the elutions as E1-E5 (each elution is of 5 mL).
6. Concentrating the virus using a centrifugal concentrator
   1. Spin the centrifugal concentrator containing the pre-treatment buffer at 900 x g for 2 min. Discard the flow-through.
   2. Wash the centrifugal concentrator filter with 4 mL of HBSS with Mg²⁺ and Ca²⁺ and centrifuge at 1000 x g for 2 min; discard the flow-through.
   3. Add the elution E2 to the centrifugal concentrator. Spin at 1000 x g for 5 min and discard the flow-through.
   4. Finish adding E2 and then add E3 to the centrifugal concentrator and spin at 1000 x g for 5 min until the concentrated virus is approximately 1 mL.
      NOTE: Avoid centrifuging the vector such that the volume is below the level of the filter. Do not concentrate E1, E4, or E5 in the centrifugal concentrator, as they contain very little to no vector and contain contaminants.
   5. Remove the concentrated virus from the centrifugal concentrator using a p200 filtered tip and place it into a sterile 1.5 mL centrifuge tube.
   6. Rinse the centrifugal concentrator with 200 µL of HBSS with Mg²⁺ and Ca²⁺ to dislodge any remaining AAV from the filter. Pipette up and down vigorously multiple times (for ~30 s) to dislodge any virus adhered to the membrane and place in the 1.5 mL centrifuge tube with the remainder of the virus. Mix the tube well.
   7. Aliquot 5 µL for DNA extractions and store the purified vector at -80 °C.
7. Wash the column using 25 mL of 2 M NaCl. Further, use 25 mL of 0.1% Triton X-100, pre-heated to 37 °C to wash the column. Next, wash the column using 50 mL of sterile dH₂O, and then wash using 25 mL of 20% ethanol.
8. Ensure the column membrane is fully saturated in 20% ethanol, as this is the storage solution. Seal the column with plugs provided and store at 4 °C.
9. Store the tubing in 1 M NaOH.
   ​NOTE: If cleaned properly, heparin sepharose columns can be re-used up to five times.

4 AAV genomic DNA extraction

1. Prepare the reaction mix mentioned in Table 3 in a PCR tube for DNase treatment.

Component	Volume
Purified AAV vector	5 μL
10x DNase Buffer	2 μL
DNase	1 μL
ddH2O	12 μL
Final Volume	20 μL

Table 3: DNase treatment master mix formula. Recommended components and volumes required for DNase treatment of AAV viral vectors during DNA extraction.

2. Vortex the PCR tube to mix and pulse the PCR tube to spin down the contents.
3. Using a thermocycler, incubate at 37 °C for 20 min followed by 75 °C for 15 min to heat inactivate the DNase.
4. Add 5 µL of Proteinase K.
5. Using a thermocycler, incubate at 50 °C for 60 min and then at 95 °C for 30 min to heat inactivate Proteinase K.
6. Use a DNA clean-up kit to remove potential contaminants.
   NOTE: This step was performed using a commercially available blood and tissue clean up kit (Table of Materials).
   1. Add 200 µL of AL buffer (Blood and tissue clean up kit, Table of Materials) to the PCR tube containing the DNase/Proteinase K treated vector.
   2. Vortex the PCR tube and incubate at 56 °C for 10 min in a thermocycler.
   3. Pipette the liquid from the PCR tube into a sterile spin column sitting in a collection tube.
   4. Add 200 µL of 100% ethanol to the column and mix thoroughly by vortexing.
   5. Centrifuge at 6,000 x g for 1 min and discard the flow through.
   6. Add 500 µL of buffer AW1 (Blood and tissue clean up kit, Table of Materials) to the spin column.
   7. Centrifuge at 6,000 x g for 1 min and discard the flow through.
   8. Add 500 µL of buffer AW2 (Blood and tissue clean up kit, Table of Materials) to the spin column.
   9. Centrifuge at 15,000 x g for 3 min and discard the flow through.
   10. Place the spin column into a sterile 1.5 mL centrifuge tube and add 200 µL of buffer AE (Blood and tissue clean up kit, Table of Materials) directly to the spin column membrane.
   11. Incubate at room temperature for 1 min.
   12. Centrifuge at 6,000 x gfor 1 min to elute the DNA.
   13. Store the DNA at -20 °C.

5 Titration of AAV vector genomes using quantitative polymerase chain reaction and a Simian Virus 40 (SV40) probe

NOTE: Perform all qPCR work in a PCR hood using filtered pipette tips to avoid external DNA contamination. If the AAV genome does not encode an SV40 polyA sequence, use a probe against the ITR described elsewhere²⁵. Ensure the plasmid DNA selected as standard contains SV40 polyA sequence.

1. Stock standard preparation
   1. Dilute the stock plasmid DNA standard (pTransgene plasmid containing SV40 polyA sequence) to a final concentration of 10 µg/µL and store at -20 °C in 6 µL aliquots.
   2. Determine the copy number present in plasmid DNA standard using the following online calculator²⁶.
      NOTE: Use a plasmid DNA used for the standard produced by a commercial vendor to ensure quality and correct concentration. Prepare a large quantity of standard (e.g., 10 mL) to conduct bridging studies when transitioning to a newly prepared standard.
2. Prepare the following reagent mix mentioned in Table 4 for both the samples and the standard in a 1.5 mL centrifuge tube.
   NOTE: Prepare sufficient overage of master mix. See Table 5 for primer/probe sequences.

Component	Volume
Universal qPCR master mix (2X)	10 μL
Molecular grade water	4.5 μL
40x SV40 polyA primer/probe	0.5 μL
Final Volume	15 μL

Table 4: qPCR master mix for AAV titration. Recommended components and volumes required for qPCR of DNA extracted from AAV viral vectors.

Component	Sequence
Forward primer	5’-AGCAATAGCATCACAAATTTCACAA-3’
Reverse primer	5’-CCAGACATGATAAGATACATTGATGAGTT-3’
Probe	/56-FAM/AGCATTTTT/Zen/TTCACTGCATTCTAGTTGTGGTTTGTC/3IABkFQ

Table 5: Primer sequences against the SV40 polyA DNA sequence. Sequences of the primers and probe used for qPCR titration, which bind to specific areas of AAV viral vectors that contain the SV40 polyA sequence.

3. Pipette the master mix up and down to mix.
4. Set up the dilution plate.
   1. Use a clear 96-well plate to prepare standard and sample dilutions, add 45 µL of molecular grade water to each well in every other column starting with column 1 (columns 1, 3, 5, 7, 9, and 11).
   2. Add 5 µL of the standard to well A1 and pipette to mix.
   3. Use a new filtered pipette tip to create a 1/10 dilution from well A1 to B1.
   4. Continue a series of 10-fold dilutions down the column until reaching G1.
   5. Do not add to H1, as this will act as a negative control.
   6. Apply the first sample (S1) by adding it to well A3, forming a 1/10 dilution. Pipette this mixture and transfer 5 µLto well B3. Discard the pipette tip after this transfer.
   7. With a new pipette tip, mix the solution in well B3 and form a 1/100 dilution. Transfer 5 µLof this mixture to well C3 and discard the tip after the transfer.
   8. With a new pipette tip, pipette up and down the solution in well C3 to make a 1/1000 dilution. Discard the tip.
   9. Continue diluting samples without adding any samples to columns 2, 4, 6, 8, 10, or 12.
   10. Once all samples are diluted, mix the contents in wells of column 1, and then transfer 20 µL to column 2.
   11. Repeat this for columns 3, 5, 7, 9, and 11 to create replicates of each standard and sample dilution. Refer to Figure 3 for plate layout.
       ​NOTE: When following plate set-up in Figure 3, samples diluted in rows G and H will only have 1/10 and 1/100 dilutions.

Figure 3: Plate layout for qPCR AAV titration. Blue indicates the placement of the serial dilution of the standard; green indicates the placement of the negative control; purple indicates the placement of the dilution of the samples. Each standard, negative, or sample is added in replicate. An example for the concentration of the standard has been added to show the dilution series of the standard, and placement of sample dilutions have been added to their respective wells. Please click here to view a larger version of this figure.

5. Titration by SV40 polyA detection-based qPCR
   1. Add 15 µL of qPCR master mix to each well of a white-semi skirted 96-well qPCR plate.
   2. Transfer 5 µL of each sample from the clear 96-well plate to the white-semi skirted 96 well qPCR plate.
   3. Use a multichannel pipette to ensure adequate mixing of the qPCR master mix and sample.
   4. Seal the plate with a sealing film and centrifuge the qPCR plate at 1500 x g for 30 s.
   5. Run the qPCR reaction on plate-based real-time PCR amplification and detection instrument, using the suggested conditions in Table 6.

Section	Cycles	Time	Temperature	Description
Pre-incubation	1x	5 min	95 °C	DNA denaturation.
Amplification	38x	15 s	95 °C	Amplification of DNA. Settings can be modified if using alternative primers with different annealing temperatures.
		60 s	60 °C
Cooling	1x	60 s	40 °C	Plate Cooling. End of run.

Table 6: Thermocycler protocol for hydrolysis probe-based qPCR titration. Recommended thermocycler protocol for use of probe-based qPCR titration of DNA extracted purified AAV vectors.

NOTE: For qPCR AAV titration worksheet see Table 7.

6. Data analysis to determine AAV genome copy numbers.
   1. Fill in the spreadsheet data cells (Table 7A) with the concentration values obtained from the qPCR run for both standard and sample dilutions.
   2. Use concentration values from Table 7A to produce a standard curve (Table 7B).
      NOTE: The standard curve will be shown as a natural logarithm (y = a ln(x) + b) along with R² efficiency. A standard curve must have an efficiency close to 100 % and R² close to 1.0 (≥0.99).
   3. Fill in the slope efficiency by filling in this online calculator²⁷.
      NOTE: An efficiency between 90%-110% is acceptable. If the efficiency of qPCR is outside of this range, re-run the qPCR.
   4. Use concentration values from Table 7A to average the dilutions of each sample and determine the standard deviation of each sample (Table 7C).
      NOTE: Exclude dilutions from samples that are more than one standard deviation away from the average of the sample dilutions.
   5. Using the mean concentration of each dilution, multiply by the dilution factor, and then divide by five to get the vector genomes (vg)/µL of each sample (Table 7C).
   6. Calculate the vg/mL of each sample by multiplying the mean of each sample's concentrations by 80,000 (Table 7C).
   7. Average the vg/mL of each dilution to produce the final vg/mL of each sample (Table 7C).
      ​NOTE: The user must divide the mean concentration of each dilution by a factor five to account for the 5 µL loaded into each well for the qPCR run to produce the concentration in vg/µL. The factor of 80,000 accounts for the transition from each samples' mean concentration value to vg/mL. First, the mean of each sample's concentration value must be multiplied by 2 to account for the single-stranded genomes, as the primer-probe set only quantifies positive-sense, single-stranded DNA (ssDNA), and the AAV genome exists in an approximate 1:1 ratio between positive and negative-sense ssDNA^25,28. The mean of each sample's concentration value must be multiplied x40 to account for the sample dilution from 5 µL of purified vector (section 4.1) to 200 µL of extracted DNA (section 4.6.12). Lastly, the mean of each sample's concentration value must be multiplied x1000 to convert from vg/µL to vg/mL.

6 Assessment of vector quality and purity

1. Quality Control - Western Blot
   1. Prepare a 12% SDS PAGE gel.
   2. Perform polyacrylamide gel electrophoresis.
      NOTE: Load 6 x 10¹⁰ vg of samples per well.
   3. Transfer the proteins to polyvinylidene difluoride (PVDF) membrane.
   4. Blocking PVDF membrane
      1. Remove the membrane from the transfer apparatus and rinse in 0.1% PBST to remove loose acrylamide.
      2. Place the membrane in blocking solution for at least 1 h at room temperature or overnight at 4 °C.
         NOTE: Blocking buffer can be further supplemented with 2% goat serum.
   5. Incubation with primary antibody
      1. Decant the blocking buffer and add the primary antibody, an anti-AAV mouse monoclonal antibody at a 1:200 dilution.
      2. Incubate overnight at 4 °C.
      3. Decant the primary antibody and wash five times with 0.1% PBST for 5 min at room temperature with agitation.
   6. Incubation with secondary antibody
      1. Decant the wash solution and add HRP conjugated secondary antibody, diluted at a 1:7500 in blocking buffer, and incubate for 1 h at room temperature with agitation.
      2. Decant the secondary antibody and wash five times with 0.1% PBST for 5 min at room temperature with agitation.
      3. Perform a final wash with PBS at room temperature with agitation.
   7. Detect the proteins using enhanced chemiluminescent (ECL) substrate.
   8. Image the gel to visualize the viral proteins (VP1, VP2, and VP3 subunits) (Figure 4).

Figure 4: Western blot showing AAV capsid proteins. Lane A; MW ladder, Lane B; AAV6.2FF-hIgG01, Lane C; AAV6.2FF-hIgG02, Lane D; AAV6.2FF-hIgG03, and Lane E; AAV6.2FF-hIgG04. 6 x 10¹⁰ vg of each AAV6.2FF-hIgG was loaded into their respective lanes. Please click here to view a larger version of this figure.

2. Purity Control - SDS PAGE and Coomassie Stain
   1. Prepare SDS-PAGE gel and samples as described from step 6.1.1 and 6.1.2.
   2. Fix the gel in fixing solution for 1 h or overnight with gentle agitation. Change the fixing solution once during the first hour.
   3. Stain the gel in staining solution for 2-4 h with gentle agitation.
   4. Destain the gel with a destaining solution. Replenish the destaining solution several times until the background of the gel is fully destained (4-24 h).
   5. Store the destained gel in a storage solution.
   6. Image the gel to visualize all proteins stained by Coomassie staining solution.

Figure 5: Coomassie-stained gel. Lane A; MW ladder, Lane B; AAV6.2FF-hIgG01, Lane C; AAV6.2FF-hIgG02, Lane D; AAV6.2FF-hIgG03, Lane E; AAV6.2FF-hIgG04, Lane F; AAV6.2FF-hIgG05, and Lane G; AAV6.2FF-hIgG06. 6 x 10¹⁰ vg of each AAV6.2FF-hIgG was loaded into their respective lanes. Please click here to view a larger version of this figure.

3. Alternative purity control assay - HEK293 host cell protein detection ELISA
   1. Perform the HEK293 host cell protein detection via ELISA as per the manufacturer's instructions.
      NOTE: Use dilutions of 5 x 10^-2 and 1 x 10^-3 for purified rAAV samples. Once TMB is added to the well, incubate away from light. Linear regression cannot be used to analyze the results.
   2. Perform a point-to-point analysis, cubic spline, or four-parameter logistic fit method to interpolate concentrations of unknowns and multiply by the dilution factor to determine original sample concentration.

Representative Results

Translation from small rodent models to larger animal models and eventual clinical application presents a significant challenge due to the large amount of AAV required to transduce larger animals and achieve therapeutic effects. To compare transduction efficiency of the rationally designed AAV6.2FF capsid, previously demonstrated a 101-fold increase in transduction efficiency in murine muscle cells compared to AAV6³, mice, hamsters, and lambs were all administered AAV6.2FF expressing a human monoclonal antibody (hIgG). The AAV6.2FF-hIgG-expressing vector contained a CASI promoter⁴, a Woodchuck hepatitis virus post-transcriptional regulatory element (WPRE), and a SV40 polyA sequence. Six-week-old female BALB/c (n = 4) mice were intramuscularly (IM) administered 1 x 10¹¹ vg of AAV6.2FF-hIgG in 40 µL, and blood was collected on days 0, 7, 14, 21, and 28 post AAV administration for hIgG monitoring. Four-week-old Syrian hamsters (two females and two males) were IM administered 1 x 10¹² vg of AAV6.2FF-hIgG in 40 µL, and blood was collected weekly to monitor for hIgG expression. Lastly, 10-day-old male Dorset lambs (n = 3) were IM administered 1 x10¹³ vg /kg of AAV6.2FF-hIgG in two to three 1 mL injections in the rump. Weekly blood collection was completed by jugular bleeds and hIgG was monitored weekly for 28 days. All animal experiments were approved by the Institutional Animal Care Committee of the University of Guelph.

Mice IM administered 5 x 10¹² vg/kg of AAV6.2FF-hIgG expressed between 171-237 µg/mL of hIgG in the serum by day 28 post-administration (Figure 6). Hamsters IM administered 2 x 10¹³ vg/kg of AAV6.2FF-hIgG expressed much higher levels than the mice, with serum hIgG levels of 495-650 µg/mL by day 28 post-administration. Lastly, sheep IM administered 1 x 10¹³ vg/kg expressed serum hIgG levels of 21-46 µg/mL by day 28 post-administration. Across all animal models, the vector did not appear to impede any health indices, such as weight, proving the safety and quality of the produced vectors. It is well known that AAV-mediated monoclonal antibody (mAb) expression varies considerably depending on the species and the mAb. To the best of the authors' knowledge, there are no reports of AAV-mAb expression in ovine species thus, there is no benchmark for expected expression levels. It is noteworthy that the sheep in this study doubled in weight over the 28-day post-injection period and that the animals in Figure 6 were all transduced with different AAV-mAbs and thus cannot be compared.

Figure 6: Intramuscular delivery of AAV6.2FF-hIgG leads to sustained serum hIgG expression in mice, hamsters, and sheep. Female BALB/c mice (n = 4) were IM administered 5 x 10¹² vg/kg of AAV6.2FF-hIgG, Syrian hamsters (2 male, 2 female), were IM administered 1 x 10¹³ vg/kg of AAV6.2FF-hIgG, and 10-day-old male Dorset lambs (n = 3) were IM administered 1 x 10¹³ vg/kg. Serum was monitored for hIgG expression over a period of 28 days. Data are represented as the mean ± standard deviation. Please click here to view a larger version of this figure.

Transfection of Cell Culture Chamber
Protocol
1. Allow DNA, OptiMEM and PEI to warm to room temperature prior to transfection
2. Input the number of cell stacks to be transfected (no overage required)
3. Ensure the DNA concentrations are correct as this will change the volumes required
Number of Cell Culture Chambers		1
Concentration of transgene plasmid		1	mg/mL
Concentration of pDGM6.2FF plasmid		1	mg/mL
		Amount Per cell culture chambers		Transfection Mastermix
	OptiMEM	48.1	mL	48.1	mL
1:3 ratio pTransgene:pHelper	pTransgene	475	µg	475	µL
	pHelper	1425	µg	1425	µL
4. Add required volumes of OptiMEM and plasmid DNA to a 50 mL conical tube
5. Invert 10 times to mix
3:1 PEI:DNA ratio	PEI Max	5.7	mL	5.7	mL
6. Add the required amount of PEI to the 50 mL tube containing the OptiMEM and plasmid DNA
7. Immeditely close the 50 mL tube and vortex and invert 3–5 times to mix.
8. Set a timer for 10 min for the PEI complexes to incubate
9. Move cell stacks to be transfected into the BSC
10. After 10 min, pour the trasnfection mix into one orange port.
11. Gently mix liquid throughout the cell stack
12.  Return the transfected cells stack to the incubator ensuring equal volume on each layer
13. Harvest cell culture chamber 72 h later

Table 2: Transfection calculator for cell culture chamber. Interactive worksheet to determine correct concentration of pTransgene, pHelper, and PEI for transfection of cell culture chamber.

Table 7: AAV Titration Calculator. Interactive worksheet to determine final concentration of AAV samples from qPCR, expressed as vg/mL. Please click here to download this Table.

Discussion

The production of recombinant AAV (rAAV) vectors described in this paper uses common materials, reagents, and equipment found in majority of the molecular biology research labs and facilities. This paper allows for high-quality in vitro and in vivo grade rAAV to be produced by the reader. Above all, this protocol for rAAV production, compared to more tedious protocols involving cesium chloride purification, is efficient and avoids the use of ultracentrifugation. Once HEK293 cells have been transfected, purified AAV is ready to use within 5 working days.

During the rAAV production and purification process, many steps can influence the purity and final titer of the vector. The first and most critical step is the health of the HEK293 cells, which has a direct effect on the vector titer. The use of HEK293 cells as opposed to other cell types or systems, such as HEK293T cells, is advantageous in that HEK293 cells have a greater ability to adhere to the plastic coating of plates and cells creating robust cell networks for continual cell growth. It is recommended to monitor cells for mycoplasma contamination and to avoid passaging HEK293 cells past ~40 passages or once their doubling time appears to be slowing down as this will lead to lower AAV yields. Furthermore, confirming cells are approximately 80% confluent prior to transfection ensures that cells are not too sparse or overgrown. With regard to the transfection components, the use of high-quality plasmid DNA (e.g., commercially prepared plasmid DNA) is highly recommended to ensure DNA remains in solution and interacts properly with chemical transfection delivery components, such as PEI. Lastly, the transfection reagent has a significant impact on rAAV yields. Here, PEI is used as the transfection agent. PEI is a stable cationic polymer that delivers exogenous DNA to the nucleus of cells through production of plasmid-polymer complexes, which are then taken up by cells and trafficked to the nucleus through host-cell processes²⁹. PEI-based transfections are quick and easy in comparison to other methods of DNA delivery, including calcium phosphate or lipofectamine. However, the ratio of PEI:DNA must be optimized.

The use of cell stacks opposed to traditional adherent plates provides a more convenient and consistent method for the production of rAAV. Cell stacks involve less technical manipulation during transfection, mitigating the possible dislodgement of cells from the adherent monolayer, allowing for stronger cell networks, and better production of rAAV. Post transfection, the collection of supernatant and lysis of cells is efficient and consistent. To harvest rAAV from cells, physical cell lysis methods such as freeze-thawing are inefficient and inconsistent as there is no way to ensure that all the cells are lysed. Here, a chemical lysis procedure is described. The use of chemical lysis buffer ensures that all the cells are exposed to the lysis agent at a specific concentration, as well as additives such as protease inhibitor to prevent capsid degradation. This method more consistently lyses cells allowing for a more efficient harvest of rAAV. Furthermore, the pelleting and removal of cellular debris eliminates possible contaminants from the crude lysate, which can increase the time and cost of filtering lysate prior to purification.

Though rAAV might be present in high amounts in the crude lysate, the act of capturing and cleaning the rAAV can differ among various purification methods, such as cesium chloride gradients. Here, the use of heparin sepharose affinity chromatography purification provides a rapid and easy method for vector purification that results in ultrapure virus and does not require gradient ultracentrifugation. However, not all AAV serotypes contain heparin-binding domains, so for those serotypes this purification method would not be appropriate. For example, while AAV2 and AAV6 bind heparin sulfate, AAV4 and AAV5 do not³⁰. Though this method is not universal, it is efficient and easy for those capsids that do bind heparin sulfate. Unlike ultra-centrifugation gradients such as iodixanol, all rAAV particles are bound to the membrane of the column until they are eluted using high salt concentration washes, avoiding issues such as mixing of gradients and skill needed to recover fractions from the gradient. Elution through high salt concentration washes allows for controlled and precise elution of the virus into fractions that can be further concentrated. Only concentrating certain fractions of the eluted virus further removes potential contaminants while recovering >97% of the eluted virus. One limitation of heparin sepharose affinity chromatography purification is that it does not distinguish between empty and full particles. Additional analytical ultracentrifugation steps would need to be incorporated in order to remove empty capsids³¹.

qPCR is a cost-effective and rapid method for determining the number of vector genomes in an AAV vector prep. Though qPCR is a sensitive quantification method, it does not provide any information about the number of infectious particles in the prep. Moreover, the sensitivity of this assay can result in variability as technical errors during pipetting can lead to inter-assay variability. Thus, for any experiment that involves comparing different AAV vectors, it is critical that the vectors be titered on the same qPCR plate. Despite these limitations, qPCR is the most accurate method developed for titering AAV and is currently the most widely used and accepted method for quantification of AAV vectors³². The use of a primer/probe that binds to the SV40 polyA of a rAAV genome is based on the fact that this sequence is conserved across many AAV genome plasmids engineered in the laboratory. Beyond choosing a conserved sequence among the reader's rAAV genome plasmids, qPCR probes can be designed for all parts of the rAAV genome, including, but not limited to, promoters, transgenes, or post-transcriptional factors.

Assessing the purity and quality of the AAV product is an important step in the production process. Western blotting can be used to detect the VP1, VP2, and VP3 structural proteins that make up the capsid of AAV, as well as any altered forms of VP. VP1, VP2, and VP3 are typically present at a ratio of 1:1:10, but this can vary from 1:1:5 to 1:1:20 depending on the serotype. Therefore, it is critical to determine the ratio for each system empirically, particularly because the N-terminal region of VP1 has been shown to be important for infectivity and transduction³³. SDS-PAGE coupled with Coomassie staining is a rather straightforward method for detecting VP1, VP2, and VP3, as well as host cell protein contaminants. The use of fluorescent protein stains such as SYPRO Ruby offer higher sensitivity detection of host cell protein contaminants than Coomassie; however, not all laboratories have access to the imaging equipment required to visualize the gel. Finally, commercial ELISAs can be used to quantify host cell protein contaminants in AAV vectors produced in HEK293 cells.

This protocol provides an in-depth overview of producing high titer, high purity rAAV for serotypes that bind to heparin. In the representative results section, the use of novel AAV6.2FF expressing hIgG in a variety of animal models shows the safety and efficiency of this rAAV in vivo. Though the AAV6.2FF vector was administered IM, these high-quality, high purity viruses can be administered via a variety of different routes in vivo, as possible inflammatory contaminants, such as host-cell protein, have been eliminated through our purification processes.

Materials

Name	Company	Catalog Number	Comments
0.22 μm filter	Millipore Sigma	S2GPU05RE
0.25% Trypsin	Fisher Scientific	SM2001C
1-Butanol	Thermo Fisher Scientific	A399-4	CAUTION. Use under a laminar flow hood. Wear gloves
10 chamber cellstack	Corning	3271
1L PETG bottle	Thermo Fisher Scientific	2019-1000
30% Acrylamide/Bis Solution	Bio-Rad	1610158
96-well skirted plate	FroggaBio	FS-96
Adhesive plate seals	Thermo Fisher Scientific	08-408-240
Ammonium persulfate (APS)	Bio-Rad	161-0700	CAUTION. Use under a laminar flow hood. Wear gloves
Blood and Tissue Clean up Kit	Qiagen	69506	Use for DNA clean up in section 4.6 of protocol
Bromophenol blue	Fisher Scientific	B392-5	CAUTION. Use under a laminar flow hood. Wear gloves
Cell Culture Dishes	Greiner bio-one	7000232	15 cm plates
Culture Conical Tube	Thermo Fisher Scientific	339650	15 mL conical tube
Culture Conical Tube	Fisher Scientific	14955240	50 mL conical tube
Dulbecco's Modified Eagle Medium (DMEM) with 1000 mg/L D-glucose, L-glutamine	Cytiva Life Sciences	SH30022.01
ECL Western Blotting Substrate	Thermo Fisher Scientific	32209
Ethanol	Greenfield	P016EA95	Dilute ethyl alcohol(95% vol) to 20% for section 3.7.4 and 70% for section 6.1.1.1
Fetal Bovine Serum (FBS)	Thermo Fisher Scientific	SH30396.03
Glacial acetic acid	Fisher Scientific	A38-500	CAUTION. Use under a laminar flow hood. Wear gloves
Glycerol	Fisher Scientific	BP229-1
Glycine	Fisher Scientific	BP381-500
HBSS with Mg2+ and Ca2+	Thermo Fisher Scientific	SH302268.02
HBSS without Mg2+ and Ca2+	Thermo Fisher Scientific	SH30588.02
HEK293 cells	American Tissue Culture Collection	CRL-1573	Upon receipt, thaw the cells and culture as described in manufacturer’s protocol. Once cells have been minimally passaged and are growing well, freeze a subfraction for future in aliquots and store in liquid nitrogen. Always use cells below passage number 30. Once cultured cells have been passaged more than 30 times, it is recommended to restart a culture from the stored aliquots
HEK293 host cell protein ELISA kit	Cygnus Technologies	F650S	Follow manufacturer’s instructions
Heparin sulfate column	Cytiva Life Sciences	17040703
Kimwipe	Thermo Fisher Scientific	KC34120
L-glutamine (200 mM)	Thermo Fisher Scientific	SH30034.02
Large Volume Centrifuge Tube Support Cushion	Corning	CLS431124	Support cushion must be used with large volume centrifuge tubes uless the centrifuge rotor has the approriate V-bottom cushions
Large Volume Centrifuge Tubes	Corning	CLS431123-6EA	500 mL centrifuge tubes
MgCl2	Thermo Fisher Scientific	7791-18-6
Microcentrifuge tube	Fisher Scientific	05-408-129	1.5 mL microcentrifuge tube, sterilize prior to use
Molecular Grade Water	Cytiva Life Sciences	SH30538.03
N-Lauroylsarcosine sodium salt	Sigma Aldrich	L5125	CAUTION. Wear gloves
NaCl	Thermo Fisher Scientific	BP35810
Optimem, reduced serum medium	Thermo Fisher Scientific	31985070
Pasteur pipets	Fisher Scientific	13-678-20D	Sterilize prior to use
PBS (10x)	Thermo Fisher Scientific	70011044	Dilute to 1x for use on cells
Penicillin-Streptomycin Solution	Cytiva Life Sciences	SV30010
pHelper plasmid	De novo design or obtained from plasmid repository	NA
Pipet basin	Thermo Fisher Scientific	13-681-502	Purchase sterile pipet basins
Polyethylene glycol tert-octylphenyl ether (Triton X-100)	Thermo Fisher Scientific	9002-93-1	CAUTION. Wear gloves
Polyethylenimine (PEI)	Polyscience	24765-1	Follow manufacturer’s instructions to produce a 1L solution. 0.22μm filter and store at 4°C
Polypropylene semi-skirted PCR Plate	FroggaBio	WS-96
Polysorbate 20 (Tween 20)	Thermo Fisher Scientific	BP337-100	CAUTION. Wear gloves
polyvinylidene difluoride (PVDF) membrane	Cytiva Life Sciences	10600023	Use forceps to manipulate. Wear gloves.
Primary antibody	Progen	65158
Protein Ladder	FroggaBio	PM008-0500
Proteinase K	Thermo Fisher Scientific	AM2546
pTrangene plasmid	De novo design or obtained from plasmid repository	NA	Must contain SV40 polyA in genome to be compatible with AAV titration in section 5.0
Pump tubing	Cole-Parmer	RK-96440-14	Optimize length of tubing and containment of virus in fractions E1-E5
RQ1 Dnase 10 Reaction Buffer	Promega	M6101	Use at 10x concentration in protocol from section 4.0
RQ1 Rnase-free Dnase	Promega	M6101
Sample dilutent	Cygnus Technologies	I700	Must be purchased separately for use with HEK293 host cell protein ELISA kit
Secondary antibody, HRP	Thermo Fisher Scientific	G-21040
Skim milk powder	Oxoid	LP0033B
Sodium dodecyl sulfate (SDS)	Thermo Fisher Scientific	28312	CAUTION. Use under a laminar flow hood. Wear gloves
Sodium hydroxide (NaOH)	Thermo Fisher Scientific	SS266-4
SV40 polyA primer probe	IDT		Use sequence in Table X for quote from IDT for synthesis
Tetramethylethylenediamine (TEMED)	Thermo Fisher Scientific	15524010	CAUTION. Use under a laminar flow hood. Wear gloves
Tris Base	Fisher Scientific	BP152-5
Trypan blue	Bio-Rad	1450021
Ultra-Filter	Millipore Sigma	UFC810024	Ultra-4 Centrifugal 10K device must be used, as it has a 10000 molecular weight cutoff
Universal Nuclease for cell lysis	Thermo Fisher Scientific	88702
Universal qPCR master mix	NEB	M3003L
Whatman Paper	Millipore Sigma	WHA1001325
β-mercaptoethanol	Fisher Scientific	21985023	CAUTION. Use under a laminar flow hood. Wear gloves
CAUTION: Refer to the Materials Table for guidelines on the use of dangerous chemicals.