DUAL BIFUNCTIONAL VECTORS FOR AAV PRODUCTION

Abstract

The present invention relates novel combinations of nucleic acid constructs for the production of recombinant parvoviral gene therapy vectors. In particular the invention relates a combination preferably no more than two construct, the first construct expressing both the parvoviral Cap and Rep proteins, and the second construct at least comprising the transgene flanked ITRs and optionally again comprising an expression cassette forthe Cap proteins. The nucleic acid constructs are preferably baculoviral vectors for the production of rAAV in insect cells.

Claims

1. A cell comprising one or more nucleic acid constructs, comprising: (i) a first expression cassette comprising a first promoter operably linked to a nucleotide sequence encoding an mRNA, translation of which in the cell produces at least one of parvoviral Rep 78 and 68 proteins; ii) a second expression cassette comprising a second promoter operably linked to a nucleotide sequence encoding an mRNA, translation of which in the cell produces at least one of parvoviral Rep 52 and 40 proteins; (iii) a third expression cassette comprising a third promoter operably linked to a nucleotide sequence encoding parvoviral VP1, VP2, and VP3 capsid proteins; and, (iv) a nucleotide sequence comprising a transgene that is flanked by at least one parvoviral inverted terminal repeat sequence, wherein, at least one of the first and second expression cassette are present on a first nucleic acid construct with the third expression cassette, and upon transfection of the cell with the one or more nucleic acid constructs, the first promoter is active before the second and third promoters.

2. The cell according to claim 1, wherein the nucleotide sequence comprising the transgene flanked by the parvoviral inverted terminal repeat sequence is present on a second nucleic acid construct.

3. The cell according to claim 2, wherein the second nucleic acid construct further comprises a fourth expression cassette comprising a fourth promoter operably linked to a nucleotide sequence encoding parvoviral VP1, VP2, and VP3 capsid proteins, wherein the first promoter is active before the second, third and fourth promoters, wherein optionally, the third and fourth promoters are identical, and wherein optionally, the parvoviral VP1, VP2, and VP3 capsid proteins encoded by the nucleotide sequences in the third and fourth expression cassettes are identical.

4. The cell according to claim 3, wherein the at least one of parvoviral Rep 78 and 68 proteins and the at least one of parvoviral Rep 52 and 40 proteins comprise a common amino acid sequence comprising the amino acid sequence from the second amino acid to the most C-terminal amino acid of the at least one of parvoviral Rep 52 and 40 proteins, wherein the common amino acid sequences of the at least one of parvoviral Rep 78 and 68 proteins and the at least one of parvoviral Rep 52 and 40 proteins are at least 90% identical, and wherein the nucleotide sequence encoding the common amino acid sequence of the at least one of parvoviral Rep 78 and 68 proteins and the nucleotide sequence encoding the common amino acid sequences of the at least one of parvoviral Rep 52 and 40 proteins are less than 90% identical.

5. The cell according to claim 4, wherein the common amino acid sequences of the at least one of parvoviral Rep 78 and 68 proteins and the at least one of parvoviral Rep 52 and 40 proteins are at least 99% identical.

6. The cell according to claim 4, wherein the nucleotide sequence encoding the common amino acid sequence of the at least one of parvoviral Rep 78 and 68 proteins has an improved codon usage bias for the cell as compared to the nucleotide sequence encoding the common amino acid sequences of the at least one of parvoviral Rep 52 and 40, or wherein the nucleotide sequence encoding the common amino acid sequence of the at least one of parvoviral Rep 52 and 40 proteins has an improved codon usage bias for the cell as compared to the nucleotide sequence encoding the common amino acid sequences of the at least one of parvoviral Rep 78 and 68 proteins.

7. The cell according to claim 6, wherein the difference in codon adaptation index between the nucleotide sequences coding for the common amino acid sequences in the at least one of parvoviral Rep 78 and 68 proteins and the at least one of parvoviral Rep 52 and 40 proteins is at least 0.2.

8. The cell according to claim 1, wherein the first promoter is a constitutive promoter.

9. The cell according to claim 1, wherein at least one of the second, third and fourth promoters is an inducible promoter.

10. The cell according to claim 9, wherein the inducible promoter is a viral promoter that is induced at least 24 hours after transfection or infection of the cell with the virus.

11. The cell according to claim 1, wherein at least one of the first and second nucleic acid construct is stably integrated in the genome of the cell.

12. The cell according to claim 1, wherein the cell is an insect cell, and wherein at least one the first and second nucleic acid construct is an insect cell-compatible vector.

13. The cell according to claim 12, wherein the insect cell-compatible vector is a baculoviral vector.

14. The cell according to claim 13, wherein: (a) the first promoter is selected from a deltaEI promoter and an EI promoter; and, (b) the second, third and fourth promoters are selected from a polH promoter and a p10 promoter.

15. 3The cell according to claim 13, wherein at least one expression cassette comprises at least one ecdysone responsive element and/or at least one baculovirus enhancer element selected from the group consisting of hr1, hr2, hr2.09, hr3, hr4, hr4b and hr5.

16. The cell according to claim 1, wherein the nucleotide sequence encoding an mRNA, translation of which in the cell produces only at least one of parvoviral Rep 78 and 68 proteins, comprises an intact parvoviral p19 promoter.

17. The cell according to claim 1, wherein the at least one of parvoviral Rep 78 and 68 proteins, the at least one of parvoviral Rep 52 and 40 proteins, the parvoviral VP1, VP2, and VP3 capsid proteins and the at least one parvoviral inverted terminal repeat sequence are from an adeno associated virus (AAV).

18. The cell according to claim 4, wherein the first nucleic acid construct is DuoBac CapRep6 (SEQ ID NO. 10) and the second nucleic acid construct is DuoBac CapTrans1 (SEQ ID NO. 12), and the first and second constructs are optionally present in a 3 : 1 molar ratio.

19. A method for producing a recombinant parvoviral virion in a cell, comprising: (a) culturing a cell according to claim 1 under conditions such that recombinant parvoviral virion is produced; and, (b) recovering the recombinant parvoviral virion.

20. The method according to claim 19, wherein the cell is an insect cell and/or the parvoviral virion is an AAV virion.

21. The method according to claim 19, wherein recovery of the recombinant parvoviral virion comprises at least one of affinity-purification of the virion using an immobilised anti-parvoviral antibody, and filtration over a filter having a nominal pore size of 30 - 70 nm.

22. A nucleic acid construct, comprising: (i) a first expression cassette comprising a first promoter operably linked to a nucleotide sequence encoding an mRNA, translation of which in the cell produces at least one of parvoviral Rep 78 and 68 proteins; ii) a second expression cassette comprising a second promoter operably linked to a nucleotide sequence encoding an mRNA, translation of which in the cell produces at least one of parvoviral Rep 52 and 40 proteins; and (iii) a third expression cassette comprising a third promoter operably linked to a nucleotide sequence encoding parvoviral VP1, VP2, and VP3 capsid proteins.

23. A nucleic acid construct comprising a nucleotide sequence comprising a transgene that is flanked by at least one parvoviral inverted terminal repeat sequence.

Description

DESCRIPTION OF THE FIGURES

[0146] FIG. 1: In a TripleBac AAV production three baculoviruses comprising Rep, Cap and Transgene cassettes are co-infected in expresSF+ insect cells. In contrast, in the DuoBac process the Cap and Rep cassettes are combined on one baculovirus genome and co-infected into expresSF+ insect cells with a separate baculovirus containing a transgene cassette. In the DuoDuoBac production process the Cap-Rep and Cap-Trans expression cassettes are combined on two baculoviruses and co-infected in expresSF+ cells.

[0147] FIG. 2: Schematic overview of the expression cassettes and orientations of the Cap-Rep and Cap-Trans DuoBac baculovirus constructs used in the examples as well as the used single expression cassette baculoviruses.

[0148] FIG. 3: Viral titers as measured in the CLB of BacCap2 or BacCap3 DuoBac AAV productions. The productions were performed at a volumetric ratio of 5% Cap-Rep baculovirus stock and 1% transgene stock. High titers were obtained with construct DuoBac CapRep2, 3, 4 and 7 whereas low titers were obtained from DuoBac CapRep1 and 6.

[0149] FIG. 4: Total/full ratio of wtAAV5 and AAV2/5 DuoBac productions. Low total/full ratio’s (<2) are observed in AAVs produced from all DuoBac constructs. These total full ratios are significantly lower than normally observed in TripleBac AAV productions (>5 total/full, Table 2).

[0150] FIG. 5: SDS Page gel run with purified AAV material made with DuoBac CapRep 1-5. Construct DuoBac CapRep6 was not included because of low yield. DuoBac CapRep3 and DuoBac CapRep7 display correct capsid stoichiometry of 1:1:10, while DuoBac CapRep2, 4 and 5 display suboptimal capsid stoichiometry (low VP1 for DuoBac CapRep 2, 4, 5 or very high VP1 in case of DuoBac CapRep1).

[0151] FIG. 6: Gc/ip of AAVs produced with DuoBac constructs DuoBac CapRep1-6. Infectivity of produced AAVs mirrors the VP123 capsid stoichiometry of the DuoBac constructs. Here low VP1 results in low infectivity (high gc/ip) for DuoBac CapRep2, 4 and 5, while high or normal VP1 results in high infectivity (low gc/ip) for DuoBac CapRep3 and 1.

[0152] FIG. 7: SDS Page gel run with purified AAV material made with DuoBac and TripleBac production processes. The ideal capsid VP1, 2, 3 protein stoichiometry of 1:1:10 for AAV was maintained after switching to the DuoBac process (lanes 1-2, 11, 13 vs Lanes 5 - 10, 12, 14).

[0153] FIG. 8: Comparison of the total/full ratio between the DuoBac and TripleBac AAV productions.

[0154] FIG. 9: SDS Page gel run with purified DuoDuoBac and TripleBac produced AAVs. When comparing AAV made with DuoDuoBac and TripleBac process, a similar VP123 stoichiometry of 1:1:10 was observed.

[0155] FIG. 10: Formaldehyde gel run with genomic AAV DNA obtained from AAVs produced with a DuoDuoBac or TripleBac production process. AAVs produced with different Rep:Cap ratio’s using DuoDuoBac have similar genomic DNA packaged into the AAV particle. The DuoDuoBac AAV fragments match the DNA fragments found after a TripleBac production. The main band was 2.4 kb long and represents a single copy of the transgene.

EXAMPLES

[0156] In the examples presented the inventors aim to examine the effects of using double expression cassettes (e.g. Bac.Cap-Rep with Bac.Cap-Trans or Bac.Cap-Rep with Bac.Trans) on product quality and vector yield. In Example 1 the inventors characterize the effect of the molecular optimization of double Rep-Cap cassettes on wtAAV5 and AAV2/5 yield and product quality. In example 2, the inventors produce wtAAV5 with an optimized wtAAV5 Cap-Rep and transgene baculovirus (DuoBac) and compare it against wtAAV5 produced with a triple infection. In example 3 the inventors extrapolate the DuoBac yields to larger production scale versus the Triple Bac system. Lastly, in example 4 the inventors examine the effect of using various combinations of Cap-Trans and Cap-Rep double baculoviruses (DuoDuoBac) on the quality and vector yields and compare these to triple infection wtAAV5 productions.

Methods and Materials

Expression Cassettes

[0157] In brief, Cap-Rep DuoBac constructs (DuoBac CapRep 1 - 7) comprise a combination of a Cap cassettes (wtAAV5 or AAV2/5) under control of a Polyhedrin (PoIH) or P10 promoter and a Rep cassette. Here the Rep cassette is of split design with Rep52 and Rep78 controlled by a PoIH and dIE1 promoter, respectively. DuoBac CapTrans1 combines a wtAAV5 Cap cassette under control of the PoIH promoter with a BacTrans4 transgene cassette. Single expression cassette constructs were needed as well, both for DuoBac and TripleBac AAV productions. These constructs were always kept the same and are BacCap1 or BacCap2, (wtAAV5) and BacRep1, split-.Rep cassette. FIG. 2 summarizes the orientations used in the cassette designs, while Tables 1A and 1B summarize the different promoter/start codon combinations used per construct.

TABLE-US-00001 Cap-Rep DuoBac promoter/start codon combinations per construct construct (promoter-start codon) Rep52 (promoter-start codon) Rep78 promoter-Vp1 start-codon) Cap DuoBac CapRep1 PoIH-ATG-Rep52 dIE1-ATG-Rep78 PoIH-CTG-wtAAV5 DuoBac CapRep2 PoIH-ATG-Rep52 dIE1-ATG-Rep78 P10-CTG-wtAAV5 DuoBac CapRep3 PoIH-ATG-Rep52 dIE1-ATG-Rep78 PoIH-ACG-AAV2/5 DuoBac CapRep4 PoIH-ATG-Rep52 dIE1-ATG-Rep78 P10-ACG-AAV2/5 DuoBac CapRep5 PoIH-ACG-ShortRep - P10-ACG-AAV2/5 DuoBac CapRep6 PoIH-ATG-Rep52 dIE1-ATG-Rep78 PoIH-ACG-wtAAV5 DuoBac CapRep7 PoIH-ATG-Rep52 dIE1-ATG-Rep78 P10-DoubleATG-wtAAV5 BacRep1 PoIH-ATG-Rep52 dIE1-ATG-Rep78 - BacCap1 - - PoIH-CTG-wtAAV5 BacCap2 - - PoIH-ACG-wtAAV5

TABLE-US-00002 Cap-Trans DuoBac transgene and Cap promoter/start codon combination Construct Transgene (promoter-VP1 startcodon) Cap DuoBac CapTrans 1 BacTrans 4 PoIH-ACG-wtAAV5

Cell Culture and Baculovirus Amplification

[0158] ExpresSF+ insect cells were maintained in SF-900II SFM medium (Gibco) in shaker flasks at 28° C. at 135 RPM. Fresh baculovirus was generated for the productions of each example. Here ExpresSF+ cells were inoculated with frozen baculovirus stocks at a concentration of 3 ul stock /ml insect cells. 72 hours after the start of infection fresh baculovirus was harvested by centrifuging the cells at 1900 xg for 15 minutes and storing the cell supernatant.

Production and Purification of AAV

[0159] AAV material was generated by volumetrically co-infecting expresSF+ insect cells with various combinations of freshly amplified recombinant baculoviruses comprising double expression cassettes (Cap-Rep and Cap-Trans) or single expression cassettes (Cap, Rep, Trans) or a combination of double expression (Cap-Rep) and single (Trans) expression cassettes. The exact ratios are described in the examples. Following a 72 hour incubation at 28° C., cells were lysed in lysis buffer (1.5 M NaCl, 0.5 M Tris-HCl, 1 mM MgCl.sub.2, 1% Triton x-100, pH=8.5) for 1 hour. Next, genomic DNA was digested with benzonase (Merck) at 37° C. for 1 hour after which cell debris was pelleted at 1900 xg for 15 minutes (crude lysate samples). Supernatant was stored at 4° C. until the start of purification. AAV was then purified from crude lysed bulk (CLB) by batch binding with AVB Sepharose (GE healthcare). In brief, AVB sepharose resin was washed in 0.2 M HPO.sub.4 pH=7.5 buffer, after which clarified crude lysate was added to the resin and incubated 2 hours at room temperature (RT) in an incubator shaking at 85 rpm. Resin was washed again in 0.2 M HPO.sub.4 pH=7.5 buffer. Next, bound virus was eluted from the resin with the addition of 0.2 M Glycine pH=2.5. The pH of the eluted virus was immediately neutralized by the addition of 0.5 M Tris-HCl pH=8.5 and stored at -20° C. until further use.

Titration by Q-PCR and Total/Full Ratio Measurement by A260/A280 or HPLC

[0160] Viral titers of the crude lysates and purified AAV batches were determined by Q-PCR. Q-PCRs were run with primers specific for the promotor region of the transgene. Q-PCRs were run on an Applied Biosystems 7500 fast Q-PCR systems. Total/full ratios of purified AAV batches were measured by UV/Vis spectrophotometry. 1 ul of 10% SDS was mixed with 100 ul of purified AAV and incubated at 75° C. for 10 minutes. Following heat treatment, the absorbance at 260 and 280 nm was measured on a Nanodrop. Using the calculation described by Sommer et al. 2003 the total/full ratio of the AAV material was calculated. Alternatively, total particles were measured by HPLC. Here purified AAV material is loaded onto a size exclusion column. Total particles are determined via integrating the area under the curve of the capsid peak. Total/full ratio is subsequently calculated by dividing the total particles with the virus titer measured by the Q-PCR.

Total Protein Gels of Purified AAV Batches

[0161] Purified AAV batches were diluted in 4x Laemmli Sample Buffer (Biorad) supplemented with 10% β-mercaptoethanol (Bio-Rad), heated for 5 minutes at 95° C. and loaded on a 4-20% Mini-PROTEAN® TGX Stain-Free gel (Biorad). After 35 minutes of electrophoresis at 200 Volt in TGS buffer (Biorad) the gel stain was developed by exposing the gel for 5 minutes under UV light and visualizing the bands on a Chemidoc touch imager (Biorad).

Infectivity Assay in HelaRC32

[0162] The number of genome copies required for a single infectious particle (gc/ip) was determined with a limiting dilution based infectious titer assay. In brief, HelaRC32 (ATCC) cell that stably express AAV-derived Rep and Cap proteins were transduced with a series of AAV dilutions in replicates of 10 and infected with or without WT adenovirus 5 (wtAd5) at a wtAd5:HeLaRC32 MOI of 50. Plates were incubated for 48 h at 37° C. and wells were assessed for the presence or absence of vector genome DNA by means of Q-PCR using a vector genome-specific primer probe set. The number of infectious particles per seeded vector genome was calculated according to the Spearman-Kärber method [5].

Formaldehyde Gel Electrophoresis With Genomic AAV DNA

[0163] Genomic AAV DNA was isolated from purified AAV batches with the PCR purification Nucleospin kit (Machery Nagel). Prior to the electrophoresis run 500 ng of AAV genomic DNA was denatured for 10 minutes at 95° C. in formaldehyde loading buffer (1 ml 20x MOPS, 3.6 ml 37% Formaldehyde, 2 ml 5 mg/ml Orange G in 67% sucrose, to 10 ml with MQ) and immediately put on ice. Next, samples were run on a 1% agarose gel made in 1x MOPS (40 mM MOPS, 10 mM NaAc, 1mM EDTA, pH=8.0) supplemented with 6.6% formaldehyde. Samples were then run for 2 hours at 100 volts in 1x MOPS supplemented with 6.6% formaldehyde running buffer. After the run, DNA was stained with SYBR Gold (Thermofisher) and bands were visualized on a Chemidoc touch imager (Biorad).

Design of Experiments (DoE) Methodology

[0164] To study the effects of upstream bioprocess variance on the total:full ratios of the DuoBac and TripleBac systems two studies were subjected to Design of experiments (DoE) methodology and analysis. The two studies were performed using slightly different methods, however in both cases experimental variance was introduced in shaker flasks and AAV purification was performed using comparable methods. In addition, for both studies, two types of analysis were performed on purified samples for each experimental condition: qPCR was used to determine the vector genome copy number (gc), while SEC-HPLC was used to determine the total amount of particles regardless of content. These two metrics were subsequently used to calculate the total:full ratio, representing the proportion of total AAV capsids relative to full capsids containing a genome copy. The differences between both studies are described in the two subsequent sections.

DoE DuoBac System: Design Space and Experimental Platform

[0165] By means of a Central Composite Design (CCD), experimental variance was introduced during DuoBac-mediated transduction of Sf+ cells as listed in Table_2. This yielded a total of 17 experimental conditions (“production cultures”) with three replicate mid-points.

TABLE-US-00003 Design space for the DuoBac transduction system Factor Low Mid High BacTrans5 (% vol.) 0.33 1 3 DuoBac CapRep3 (% vol.) 0.33 1 3 VCD at TOI (x10.sup.6 VC/mL) 1 1.45 1.9

[0166] Amplified baculovirus and seed cells were generated in 10 L wave bags (Flexsafe, Sartorius) using rocking motion bioreactors (BioWave PU-Biostat, Sartorius). The media used throughout this study was Sf900 II media (ThermoFisher). The settings for all incubations were as follows T=28° C.; agitation at 25 rpm and 8 ° angle; DO=50%; and an airflow rate of 0.2 L/min. One dedicated bioreactor was used for amplification of cells at a working volume of 5 L and an initial VCD at 1.2 x 10.sup.6 VC/mL (reactor A). 18.5 hours after inoculating reactor A, two bioreactors were inoculated at a concentration of 0.8 x 10.sup.6 VC/mL and a working volume of 5.25 L (reactors B and C). 15.75 mL baculovirus Working Seed Virus (WSV) was added to reactors B and C 18 hours after cell inoculation for separate amplification of baculoviruses BacTrans5 and DuoBac CapRep3. After an additional 48 hours of incubation all reactors were harvested. The resulting materials (cells and baculovirus) were used to prepare AAV production cultures.

[0167] For production cultures, a fresh media-exchange step was implemented prior to transduction to control VCD at TOI and media composition. This media exchange involved gentle centrifugation of each seed culture at 300 g, discarding the supernatant and resuspending cells in fresh media to achieve a target VCD at TOI. Production culture composition was done as specified in Table 2.

[0168] After 70 hours, the transduction was terminated by consecutive steps of lysis (addition of 10% v/vof a 10x lysis buffer, incubation for 60 minutes at 37° C. and 135 rpm), benzonase treatment (addition of 10 units Benzonase per mL, incubation for 60 minutes at 37° C. and 135 rpm), clarification (centrifugation for 15 minutes at 4100 g at RT) and filtration (filtration through a 0.22 .Math.m bottle top filter under a vacuum). The filtrates were incubated at RT for 12 hours for adventitious viral inactivation. Remaining filtrates were purified using a batch binding affinity chromatography protocol which involved (1) preparation of AVB Sepharose HP resin in 0.2M phosphate buffer pH 7.5 (1:1 volumetric ratio); (2) addition and incubation of 250 .Math.L resin suspension to 40 mL of filtrate for 4 hours at 40 rpm; (3) centrifugation of resin at 4100 g for 5 minutes; (4) washing pellets with 0.2 M phosphate buffer pH 7.5; (5) extracting the pellet using 500 .Math.L 0.5 M Glycine/HCl pH 2.5 during an incubation of 4 minutes; (6) centrifuging the used pellet using a benchtop centrifuge; (7) neutralizing the supernatant using 200 .Math.L Tris/HCl pH8.5 buffer; and (8) filtering the neutralized eluate with a 0.22 .Math.m PVDF syringe filter. The purified materials were used for qPCR and SEC-HPLC analysis to determine total:full ratios.

Results

Example 1: Characterization of wtAAV5 and AAV2/5 Cap-Rep DuoBac Constructs

[0169] AAV production in insect cells is commonly performed by co-infecting three baculoviruses comprising Rep, Cap and Trans cassettes. To improve the statistical chance that all three components are present in the cell at the same time the Cap and Rep expression cassettes were moved to a single baculovirus (FIG. 1). To investigate if the quality and quantity of wtAAV5 and AAV2/5 produced in a double infection setting may be improved, the inventors swapped the single Rep expression cassette for a split Rep expression cassette and optimized the promoter/VP1 start codon combination of Cap. Introduction of the split Rep cassette can give better control over the timing and expression strength of Rep52 and Rep78. Furthermore, optimization of the VP123 ratio of the capsids is essential for generating infective AAV.

[0170] Constructs DuoBac CapRep1-7 (Table 1A and FIG. 1) were designed to optimize the expression of wtAAV5 and AAV2/5 Cap and balance them with Rep expressed from a split Rep cassette. To assess the impact of these changes on the AAV vector yields and quality, DuoBac productions were performed with a therapeutically relevant transgene (BacTrans4). AAVs were produced in expresSF+ insect cells (50 ml) with 5% freshly amplified Cap-Rep baculovirus and 1% freshly amplified transgene baculovirus. Following the production, viruses were purified and several assays were performed on the resulting AAV material. Virus titers (by Q-PCR) were determined on the crude lysates. Total/full ratio’s (by HPLC/Q-PCR) and capsid stoichiometry (by SDS-page gel) were determined on purified AAVs. The number of genome copies required for 1 infectious particle (gc/IP) was determined with an infectivity assay in HelaRC32 cells.

[0171] FIG. 3 summarises the viral titers measured in crude lysates of the wtAAV5 and AAV2/5 DuoBac productions. High viral yields (>1e11 gc/ml) were obtained with constructs DuoBac CapRep2, 5 and 7, while relatively low yields were observed with constructs DuoBac CapRep1 and 6. The total/full ratio of the purified virus batches was determined by dividing the total particles/ml (as determined by HPLC) by the genome copies/ml (as determined by Q-PCR). In general, low total/full ratio’s (<2.0) were observed with all DuoBac constructs (FIG. 4). This observation contrasts significantly with the total/full ratio normally observed in TripleBac AAV productions which normally falls above 5 (see example 2). Capsid stoichiometry of the purified AAVs was determined by SDS-Page gel electrophoresis (FIG. 5, capsid stoichiometry of DuoBac CapRep6 could not be determined due to low virus yield). Capsid stoichiometry was significantly impacted depending on which DuoBac construct was used. DuoBac CapRep3 and 7 display correct capsid stoichiometry of 1:1:10, while DuoBac CapRep2, 4 and 5 display suboptimal capsid stoichiometry (low VP1 for DuoBac CapRep2, 4 and 5 or very high VP1 in the case of DuoBac CapRep1). The effect that these changes could have on AAV infectivity was determined by a limiting dilution infectivity assay in HelaRC32 (FIG. 6). AAV infectivity results mirrored the capsid stoichiometry results. Here DuoBac CapRep1, 3 and 6 showed high infectivity (low gc/ip), due to normal or high VP1 in the capsid. While DuoBac CapRep2, 4 and 5 (high gc/ip) showed reduced infectivity due to a low amount of VP1 in the Capsid. Table 3 summarizes the data from these experiments.

TABLE-US-00004 Summary of the quality parameters of AAV produced with DuoBac constructs DuoBac CapRep1-7 construct (promoter-startcodon) Rep52 (promoter-startcodon) Rep78 (promoter-Vp1 startcodon) Cap gc/ml in CLB Total-full ratio VP123 ratio gc/ip DuoBac CapRep1 PoIH-ATG-Rep52 dIE1-ATG-Rep52 PoIH-CTG-wtAAV5 5.57 E+10 n/d High VP1 129 DuoBac CapRep2 PoIH-ATG-Rep52 dIE1-ATG-Rep52 P10-CTG-wtAAV5 1.76 E+13 5,0 low VP1 20825 DuoBac CapRep3 PoIH-ATG-Rep52 dIE1-ATG-Rep52 PoIH-ACG-AAV2/5 8.45 E+12 0,3 normal 60 DuoBac CapRep4 PoIH-ATG-Rep52 dIE1-ATG-Rep52 P10 -ACG-AAV2/5 2.64 E+12 0,4 low VP1 3591 DuoBac CapRep5 PoIH-ACG-ShortRep - P10 -ACG-AAV2/5 2.37 E+11 1,3 low VP1 9651 DuoBac CapRep6 PoIH-ATG-Rep52 dIE1-ATG-Rep52 PoIH-ACG-wtAAV5 8.8 E+10 2 n/d 52,8 DuoBac CapRep7 PoIH-ATG-Rep52 dIE1-ATG-Rep52 P10 -doubleATG-wtAAV5 5.3 E+11 1,5 normal n/d

[0172] From these results it appears that promoter competition has a significant impact on the virus titers for wtAAV5 DuoBac constructs (PoIH Rep + PoIH Cap= low titer for wtAAV5, DuoBac CapRep1 and 6), but less for AAV2/5 (PolH Rep+ PolH Cap = high titer for AAV2/5, DuoBac CapRep3). Introducing a P10 promoter before the wtAAV5 cassette improves the titer (DuoBac CapRep2), but results in a suboptimal VP123 stoichiometry. Introducing a stronger start codon in front of VP1 (double ATG) rescues VP123 stoichiometry and produces high titers (DuoBac CapRep7). This shows that balancing the promoter type and initiation strength for Cap VP1 is essential for generating high titers with correct AAV capsid stoichiometry. Furthermore, process complexity is reduced by combining Rep and Cap on the same baculovirus. This combination of AAV genes also led to clear improvements to the total/full ratio. How DuoBac AAV production compares to TripleBac AAV production will be examined in example 2.

Example 2: Comparison of AAV5 DuoBac (Bac.Cap-Rep and Bac.Transgene) and Triple Bac (Bac.Cap, Bac.Rep Bac.Transgene) AAV Productions

[0173] The previous example showed that by combining the Cap and Rep cassette on the same baculovirus and molecularly optimizing the Cap cassette we were able to produce an improved AAV product. This example compares AAV produced by a DuoBac and TripleBac process. To compare the two production systems DuoBac (DuoBac CapRep 7: Cap wtAAV5-Rep) productions were compared to TripleBac AAV productions (BacCap1 wtAAV5, BacRep1) with respect to vector yields and quality. Both a reporter and two therapeutically relevant transgenes were used in the AAV productions (BacTrans 1, 3 and 4). To perform AAV productions, expresSF+ insect cells (50 ml or 2.5 L) were inoculated with multiple volumetric ratios of freshly amplified baculovirus stocks. Inoculation volumes ranged between 1 to 5% of the culture volume. Following production, viruses were purified and several assays were performed on the material. Virus titers (in gc/ml by Q-PCR) were determined on crude lysates and purified AAVs. Total/full ratio’s (by A260/A280) and VP123 ratio (by SDS-page gel) were determined on purified AAV material.

[0174] Table 4 summarizes the 50 ml production results, while Table 5 summarizes the 2.5 L production results. Both at 50 ml and 2.5 L scale, DuoBac productions outperform TripleBac productions in both virus yields and total/full ratio. Depending on the inoculation volumes or transgenes used forthe production, titers (in gc/ml) in the CLB improved by 4 to 10-fold with DuoBac CapRep 7 as compared to the equivalent TripleBac production. Total genome copies purified from the productions were increased with a similar factor. Interestingly total/full ratios were also improved with the DuoBac process. Here, the used transgene seems to influence the amount this parameter improves, but the total/full ratio was consistently improved in the DuoBac productions (approximately 2-8 fold depending on the transgene cassette used for production). Expression of VP123 capsid proteins was identical between the DuoBac and TripleBac AAV productions (FIG. 7), maintaining the ideal stoichiometry of 1:1:10.

[0175] Reducing process complexity by combining the Cap and Rep expression cassettes on the same baculovirus resulted in clear improvements in yield and total/full ratio (FIG. 8), whilst maintaining the ideal VP protein stoichiometry of AAV. Although not investigated here, it is likely that process robustness (batch to batch variation) can be improved with a DuoBac process because of the reduction from three to two variables.

TABLE-US-00005 Production results for the 50 ml DuoBac to TripleBac comparison Volumetric Ratio’s 50 ml productions gc/ml crude lysate Total gc Total/Full Ratio DuoBac CapRep7 : BacTrans4 5:1 5.30 E+11 2.65 E+13 1,5 BacCap1 : BacRep1 : BacTrans4 1:1:1 6.10 E+10 3.05 E+12 2,4 BacCap1 : BacRep1 : BacTrans4 1:5:1 2.70 E+10 1.35 E+12 1,4 BacCap1 : BacRep1:BacTrans4 5:5:1 1.5 e11 7.50 E+12 2,1

TABLE-US-00006 Production results for the 2.5 Liter DuoBac to TripleBac comparison Volumetric Ratio’s 2.5 L productions gc/ml crude lysate Total gc from 2.5 L Total/Full Ratio BacCap1 : BacRep1 : BacTrans1 1:1:1 8.90 E+10 2.25 E+14 6,7 DuoBac CapRep7 : BacTrans1 1:1 1.10 E+12 2.80 E+15 1,5 BacCap1 : BacRep1 : BacTrans3 1:1:1 2.40 E+11 5.90 E+14 16,3 DuoBac CapRep7 : BacTrans3 1:1 6.80 E+12 1.70 E+16 1,8

Example 3: Comparison of DuoDuoBac (Bac.Cap-Rep and Bac.Cap-Trans) to TripleBac AAV (Bac.Cap, Bac.Rep Bac.Transqene)

[0176] Previous studies showed that the Cap:Rep baculovirus inoculation ratio of a TripleBac AAV production had a direct impact on the total/full ratio and titer yield of an AAV production. Here increased Rep baculovirus inoculation resulted in a reduction in Capsid production and total/full ratio. In contrast, an increased Cap baculovirus inoculation ratio increased the total/full ratio and yield. By introducing a Cap cassette on both the Rep and Transgene baculoviruses, thereby creating a double DuoBac process or DuoDuoBac process (FIG. 1), we have more freedom controlling the Cap:Rep ratio in the cell during an AAV production. Also it would allow us to explore Cap:Rep production ratios (especially high Cap ratios) that are impossible to achieve in a TripleBac AAV process (due to too high inoculation volume which inhibits AAV production).

[0177] In this Example we aim to investigate the impact of changing the Cap:Rep ratios during insect cell infection on AAV quality and yield, this was achieved by varying the DuoBac CapTrans1 to DuoBac CapRep6 inoculation ratio. The DuoDuoBac AAV production was compared to TripleBac AAV productions. AAV productions were performed in expresSF+ insect cells at a 50 ml scale. Inoculation volumes ranged between 1 to 5% of the culture volume for each baculovirus. Following production, viruses were purified with AVB sepharose. Virus titers (gc/ml as determined by Q-PCR) were measured in the crude lysates and purified AAVs. Total/full ratio’s (by A260/A280) and capsid composition (by SDS-page gel) were determined on purified AAVs. In addition, the genomic DNA packaged into the AAV particle was also investigated by formaldehyde gel electrophoresis.

[0178] Table 6 summarises the result of the DuoDuoBac and TripleBac AAV productions. For the DuoDuoBac productions it lists the used inoculation conditions as well as what the equivalent inoculation conditions would be needed to achieve a similar ratio with a TripleBac AAV production. In all of the of the DuoDuoBac AAV productions tested, the vector yields in the crude lysate fell between 7 e+11 to 1.4 e+12 gc/ml, as compared to 6-7e+11 for the tested TripleBac productions, meaning a 2-fold titer increase is observed for the best DuoDuoBac condition. The total/full ratio of all DuoDuoBac productions was reduced as compared to TripleBac productions. When comparing DuoDuoBac productions, a lower total/full ratio was generally observed when more Rep was present, whilst a higher total/full ratio was linked to an increase in Cap. The best condition tested was the 1:3 DuoBac CapTrans1 to DuoBac CapRep6 co-infection, which resulted in an average titer in the CLB of 1.2 e+12 gc/ml with a total/full ratio of ~1.5. Compared to its closest TripleBac equivalent (5:5:1 ratio), the titer was improved by 2-fold (1.2 e+12 vs 6 e+11), while the total/full ratio was improved approximately 4-fold (1.5 vs 6). When comparing the expression of capsid proteins VP-1, -2 and -3 between DuoDuoBac and TripleBac productions, a similar stoichiometry of 1:1:10 was observed for all conditions tested (FIG. 9). This indicated that introducing the Cap cassette on the Rep and Transgene baculoviruses did not alter the optimal ratio, maintaining it at 1:1:10. Also, the genomic DNA packaged into the AAV particle was similar between DuoDuoBac and TripleBac productions (FIG. 10). Genomic AAV DNA isolated from both productions resulted in an identical banding pattern on a formaldehyde gel. The main band was 2.4 kb long and represents a single copy of the BacTrans4 transgene.

[0179] In summary, a DuoDuoBac process results in improved vector yields and total to full ratios using a wide range of Bac.Cap-Rep to Bac.Cap-Trans inoculation ratios as compared to TripleBac. Increased freedom to change the Cap:Rep ratio in the production cell during AAV production (due to the presence of two Cap expression cassettes and the reduction of the number of baculovirus seeds used for infection) allows for steering and optimisation of the total/full ratio of the produced AAVs. We observed that an increase in Rep resulted in slightly lower yields and total/full ratio, while an increase in Cap resulted in higher total/full ratio. DuoDuoBac productions minimize the variation in yield and total/full ratio as compared to TripleBac. In addition, a DuoDuoBac AAV production allows us explore Cap:Rep ratios that cannot be feasibly reached with a TripleBac process. This expanded manoeuvring room offered by the DuoDuoBac process can potentially allow for the development of more robust AAV productions processes.

TABLE-US-00007 Production results for the 50 ml DuoDuoBac to TripleBac comparison Shaker Flask# ratio DuoBac CapTrans 1: DuoBac CapRep6 equivalent Cap:Rep:Transgene gc/ml Crude Lysate Total Gc Total/full ratio 1 5:5 10:5:5 9.6 E+11 4.80 E+13 2,0 2 5:5 10:5:5 1.1 E+12 5.50 E+13 2,0 3 5:3 8:3:5 1.2 E+12 6.00 E+13 2,5 4 5:3 8:3:5 1.3 E+12 6.50 E+13 1,9 5 3:5 8:5:3 9.1 E+11 4.55 E+13 2,7 6 3:5 8:5:3 7.3 E+11 3.65 E+13 1,1 7 3:3 6:3:3 1.2 E+12 6.00 E+13 1,7 8 3:3 6:3:3 1.3 E+12 6.50 E+13 1,9 9 1:5 6:5:1 8.9 E+11 4.45 E+13 1,3 10 1:5 6:5:1 8.2 E+11 4.10 E+13 1,7 11 1:3 4:3:1 1.1 E+12 5.50 E+13 1,1 12 1:3 4:3:1 1.1 E+12 5.50 E+13 1,8 13 1:1 2:1:1 1.4 E+12 7.00 E+13 2,2 14 1:1 2:1:1 1.3 E+12 6.50 E+13 2,5 15 3:1 4:1:1 1.3 E+12 6.50 E+13 3,6 16 3:1 4:1:1 8.8 E+11 4.40 E+13 3,2 17 5:1 6:1:1 8.8 E+11 4.40 E+13 4,6 18 5:1 6:1:1 1.1 E+12 5.50 E+13 5,0 19 1:1:1 Triple infection (BacCap2: BacRepl: BacTrans 4) 1:1:1 6.3 E+11 3.15 E+13 7,6 20 1:1:1 Triple infection (BacCap2: BacRepl: BacTrans 4) 1:1:1 6.0 E+11 3.00 E+13 6,9 21 1:5:1 Triple infection (BacCap2: BacRepl: BacTrans 4) 1:5:1 5.9 E+11 2.95 E+13 2,9 22 1:5:1Triple infection (BacCap2: BacRepl: BacTrans 4) 1:5:1 5.9 E+11 2.95 E+13 3,2 23 5:5:1 Triple infection (BacCap2: BacRepl: BacTrans 4) 5:5:1 1.2 E+12 6.00 E+13 6,3 24 5:5:1 Triple infection (BacCap2: BacRepl: BacTrans 4) 5:5:1 6.6 E+11 3.30+13 3.30 E+13 6,1

Example 4: Comparison of DuoDuoBac (Bac.Cap-Rep and Bac.Cap-Trans) to DuoBac AAV (Bac.Cap, Bac.Rep Bac.Transqene)

4.1 Cell Culture and Baculovirus Amplification

[0180] ExpresSF+ insect cells were cultured in SF-900II SFM medium under conditions as described above. Fresh baculovirus inocula were generated as described above.

4.2 DOE Studies in 1L Shake Flasks

4.2.1 DOE deSign

[0181] A Central Composite Design (CCD) was used to investigate two factors (volumetric infection ratios of the two amplified baculoviruses in a range of 0.33-3%) and their interactions. Statistical analysis was performed using Design Expert 11 (Statease, Minneapolis MN) and JMP 15 (SAS Institute Inc., Cary, NC). Quadratic Response Surface Models were generated using a rotatable CCD (α=1.414) and three center points. Genome copy titers in filtered crude lysed bulk and total particle to genome copies (tp/gc) ratios were set as responses. Only statically significant model terms (p<0.1) were included in each model and were selected through stepwise regression whilst maintaining model hierarchy.

4.2.2 Production and Purification of AAV

[0182] Amplified baculovirus and seed cells (preculture) were generated in 1L shake flasks at 28° C. at 135 rpm. The media used throughout this study was SF900 II media (ThermoFisher). Based on the VCD of the preculture, a calculated volume of culture is added to each 1 L shake flask to achieve the target seeding cell density of 1.3 x 10.sup.6 VC/mL in a final working volume of 400 mL. Additional SF900 II medium was added to each shake flask to bring the culture volume to 400 mL, as needed. Cell expansion in 1 L shake flasks was performed at 28° C. and 135 rpm. 15-21 hours after inoculation, a pool of amplified baculovirus inocula was added at a volumetric infection ratio according to DOE design. After infection, temperature set-point was increased to 30° C. and the cultures were continued for 68-76 hours at 135 rpm. After that, the cultures were harvested by adding 10% (v/v) of 10x lysis buffer (Lonza). 30 minutes after starting lysis, temperature setpoint was increased to 37° C. When the temperature set-point was reached, benzonase was added (9 unites/mL), after which the culture was incubated for additional 60 minutes. Clarification of crude lysed bulk was performed by centrifugation for 15 minutes at 4100 g and room temperature (20- 25° C.) followed by filtration through a 0.2 .Math.m membrane filter. Filtered bulks were then purified using AVB Sepharose HP resin from Cytiva. The product was eluted using 0.2 M glycine/HCI pH 2.4 buffer and subsequently neutralized using 60 mM Tris pH 8.5. Purified samples were subsequently analyzed by qPCR (to determine the vector genome copy number, GC concentration in crude lysate) and SEC-HPLC (to determine the total amount of total AAV particles). The results in Table 7 show that the DuoDuoBac system achieve higher vectors yields than a comparable DuoBac system over a wide range of infection ratios of the two baculoviruses.

TABLE-US-00008 The effects of different volumetric infection ratios of the two baculoviruses on AAV vector yields and total to full ratios for DuoBac and DuoDuoBac produced in 1 L shake flasks BacIDs Infection ratio range (%) GC (E+11 gc/mL) TP/GC ratio DuoBac BacCapRep6 + BacTrans4 0.33-3.00 0.1-2.7 2.0 - 3.2 DuoDuoBac BacCapTrans1 + BacCapRep6 0.33-3.00 1.7 - 4.3 3.4 - 25.3

4.3 Productions in 2 L Stirred Tank Bioreactors

4.3.1 Production and Purification of AAV

[0183] Amplified baculovirus and seed cells (preculture) were generated in 1 L shake flasks at 28° C. at 135 rpm. The media used throughout this study was SF900 II media (ThermoFisher). For each combination of baculoviruses, rAAV production was performed in duplicate, using two 2 L stirred tank reactors (STR, The UniVessel® SU, Satorious). Based on the VCD of the preculture, a calculated volume of culture is added to the 2 L STR to achieve the target seeding cell density of 0.5 x 10.sup.6 VC/mL in a final working volume of 2 L. Additional SF900 II medium was added to the 2 L STR to bring the culture volume to 2 L, as needed. Cell expansion in 2 L STR was performed at 28° C. Dissolved oxygen (DO) was maintained at 30% with a continuous fixed air flow through overlay at 0.2 L/min and oxygen addition through sparger at a flow of 0-150 ccm using a stirring speed of 100-300 rpm. 43-48 hours after inoculation, a pool of amplified baculovirus inocula was added at a volumetric infection ratio indicated in Table 8. After infection, temperature set-point was increased to 30° C. and the cultures were continued using the settings described above.

[0184] The cultures were harvested 68-76 hours post-infection by adding 10% (v/v) of 10x lysis buffer (Lonza). 30 minutes after starting lysis, temperature setpoint was increased to 37° C. When the temperature set-point was reached, benzonase was added (9 unites/mL), after which the culture was incubated for additional 60 minutes. Clarification of crude lysed bulk was performed by centrifugation for 15 minutes at 4100 g and room temperature (20-25° C.) followed by filtration through a 0.2 .Math.m membrane filter. Filtered bulks were then purified using a column packed with AVB Sepharose HP resin from Cytiva. The product was eluted using 0.2M glycine/HCI 2 M urea pH 2.4 buffer and subsequently neutralized using 60 mM Tris 2 M urea pH 8.5. Neutralized eluate was then loaded onto 5 mL Mustang Q membrane (Pall). Product elution was performed using 60 mM Tris 150 mM NaCl 2 M urea pH 8.5 buffer, followed by a nanofiltration using Planova 35N filter (0.01 m.sup.2). Finally, product was diafiltered against phosphate buffered saline (Merck) containing 5% sucrose and concentrated to a desired volume.

[0185] Purified samples were subsequently analyzed by qPCR (to determine the vector genome copy number, GC concentration in crude lysate), SEC-HPLC (to determine the total amount of total AAV particles), FIX potency assay and infectivity assay in HelaRC32. Table 8 shows that the DuoDuoBac system (BacCapTrans1 + BacCapRep6) outperforms the comparable DuoBac system (BacCapRep6 + BacTrans4) at least in terms of vector yield, potency and infectivity.

TABLE-US-00009 A comparison of various properties of AAV vectors produced with DuoBac or DuoDuoBac in 2 L tank bioreactors BacIDs Infection ratios (%) GC (E+11 gc/mL) TP/GC ratio Potency (RU) Infectivity (gc/ip) BacCapRep6 + BacTrans4 1 : 0.33 0.7 - 0.8 4.0-4.3 0.9 12.5 BacCapTrans1 + BacCapRep6 0.33 : 0.33 2.6 - 2.7 8.9 - 9.3 1.1 28.1

LITERATURE REFERENCES

[0186] 1. Chaabihi, H., et al., Competition between baculovirus polyhedrin and p10 gene expression during infection of insect cells. J Virol, 1993. 67(5): p. 2664-71. [0187] 2. Hill-Perkins, M.S. and R.D. Possee, A baculovirus expression vector derived from the basic protein promoter of Autographa californica nuclear polyhedrosis virus. J Gen Virol, 1990. 71 ( Pt 4): p. 971-6. [0188] 3. Pullen, S.S. and P.D. Friesen, Early transcription of the ie-1 transregulator gene of Autographa californica nuclear polyhedrosis virus is regulated by DNA sequences within its 5′ noncoding leader region. J Virol, 1995. 69(1): p. 156-65. [0189] 4. Bosma, B., et al., Optimization of viral protein ratios for production of rAAV serotype 5 in the baculovirus system. Gene Ther, 2018. 25(6): p. 415-424. [0190] 5. Grieger, J.C., S. Snowdy, and R.J. Samulski, Separate basic region motifs within the adeno-associated virus capsid proteins are essential for infectivity and assembly. J Virol, 2006. 80(11): p. 5199-210.