Improved AAV capsid production in insect cells

Abstract

To provide an adeno-associated virus vector (AAV) that can be very efficiently produced in insect cells.SOLUTION: The invention provides a nucleic acid construct comprising expression control sequences for expression in an insect cell of a nucleotide sequence comprising an open reading frame, the open reading frame sequence encodes: i) AAV capsid proteins VP1, VP2 and VP3; and ii) an ATG translation initiation codon for VP1. The nucleotide sequence also comprises, upstream of the open reading frame, an alternative start codon which is out of frame with the open reading frame.SELECTED DRAWING: None

Description

[Technical Field]

[0001] This invention relates to the production of adeno-associated viruses in insect cells and to adeno-associated viruses that provide enhanced infectivity. This invention also relates to means and methods involving an adeno-associated virus vector library. [Background technology]

[0002] Adeno-associated virus (AAV) can be considered one of the most promising viral vectors for human gene therapy. AAV has the ability to efficiently infect both dividing and non-dividing human cells, the AAV viral genome is integrated into a single chromosomal region in the host cell's genome, and most importantly, even though AAV is present in many humans, it has never been associated with any disease. Considering these advantages, recombinant adeno-associated virus (rAAV) is being evaluated in gene therapy clinical trials for hemophilia B, malignant melanoma, cystic fibrosis, and other diseases. Numerous clinical trials and approvals of gene therapy drugs in Europe, such as Alipogene tiparvovec (Glybera®, uniQure), suggest that AAV has the potential to become a main stay in clinical practice.

[0003] Generally, there are two main types of production systems for recombinant AAV. On the one hand, there are conventional production systems using mammalian cell types (293 cells, COS cells, HeLa cells, KB cells, etc.), and on the other hand, there are production systems using insect cells.

[0004] The mammalian production system suffers from several drawbacks, including the limited number of rAAV particles produced per cell (10 4 (Order of magnitude of individual particles) (Clark, 2002, Kidney Int.) 61 (Note 1): This is reviewed in sections 9-15) and may involve complex large-scale manufacturing. Regarding clinical research, 10 15Sometimes, more rAAV particles than a single individual particle are required. To produce this number of rAAV particles, 5,000 175 cm³ particles are needed. 2 Approximately 10 cells, which are equivalent to the cells in a flask. 11 It is thought that transfection and culture using individual cultured human 293 cells will be required, which may take up to 10 11 This means transfecting 293 individual cells. Therefore, large-scale production of rAAV using mammalian cell culture systems to obtain material for clinical trials has already proven problematic, and large-scale commercial production may not even be feasible. Furthermore, vectors for clinical use produced in mammalian cell cultures are always at risk of being contaminated with undesirable, potentially pathogenic material present in mammalian host cells.

[0005] To overcome these problems in mammalian production systems, an AAV production system using insect cells has been developed (Urabe et al., 2002, Hum. Gene Ther.). 13(1935-1943; U.S. Patent Application Publication No. 20030148506; and U.S. Patent Application Publication No. 20040197895). The wild-type AAV capsid derived from wild-type virus consists of approximately 60 capsid proteins, i.e., VP1, VP2, and VP3 in a stoichiometric ratio of approximately 1:1:10. Without being constrained by theory, stoichiometry appears to be important for achieving the superior potency of recombinant AAV, i.e., superior transduction. In wild-type virus, i.e., in mammalian cells, achieving a stoichiometric ratio of approximately 1:1:10 for the three AAV capsid proteins (VP1, VP2, and VP3) depends on a combination of alternating use of two splice acceptor sites and the less-than-optimal use of the ACG start codon for VP2. However, with regard to AAV production in insect cells, the expression strategy needed modification because it is not reproduced in insect cells as it occurs in mammalian cells. To achieve improved capsid protein production in insect cells, Urabe et al. (2002, cited above) used a construct transcribed into a single polycistronic messenger capable of expressing all three VP proteins without splicing, in which the initial translation start codon was replaced by an ACG codon. International Publication No. 2007 / 046703 discloses further improvement in the infectivity of rAAV vectors produced by baculoviruses by further optimizing the ratio of AAV capsid proteins in insect cells.

[0006] Urabe et al. (J. Virol., 2006, 80(4):1874~1885) reported that AAV5 particles produced in a baculovirus system using ACG as the start codon for the VP1 capsid protein had insufficient transduction efficiency or potency, and that, in contrast to AAV2 which has VP1 expressed from the ACG start codon, mutating the +4 position in the AAV5 VP1 coding sequence to a G residue did not improve infectivity. Urabe et al. constructed a chimeric AAV2 / 5 VP1 protein in which the N-terminal portion of at least 49 amino acids of AAV5 VP1 was replaced with the corresponding portion of AAV2 VP1, and this chimeric protein improved the transduction properties of virions.

[0007] In a further approach, AAV capsid protein expression was improved by inserting one or more amino acid residues between the suboptimal (other than ATG) translation initiation codon and the codon encoding the amino acid residue corresponding to the amino acid residue at position 2 of the wild-type capsid amino acid sequence in the AAV capsid coding sequence (Lubelski et al., International Publication No. 2015137802).

[0008] Despite improvements in insect cell-based production of AAV capsids for the manufacture of AAV gene therapy vectors for use in medical treatment, there remains a need to further improve AAV capsid production and to provide new methods for selecting improved AAV capsid constructs for expression in insect cells. [Overview of the project] [Problems that the invention aims to solve]

[0009] The inventors have surprisingly found that an AAV capsid, an expression construct encoding transcripts for VP1, VP2, and VP3 proteins from overlapping reading frames, can be produced very efficiently in insect cells from an expression construct in which VP1 is translated from an AUG start codon. Prior art constructs containing an ATG start codon do not yield the VP1:VP2:VP3 ratio of approximately 1:1:10 observed in wild-type AAV, and therefore, without theoretical constraints, do not produce potent AAV. The expression construct identified in the present invention enables the efficient production in insect cells of a substantial volume of very potent AAV gene therapy vector for use in medical treatment. Such vectors are at least comparable in potency and volume to AAV gene therapy vectors produced from alternative start codons such as CTG or GTG (see Figure 4).

[0010] Therefore, the construct of the present invention contains an additional out-of-frame start codon 5' from the VP1 ATG start codon, which appears to reduce translation initiation at the VP1 start codon, and enables translation of sufficient amounts of VP1, VP2, and VP3. Without being constrained by theory, such constructs may enable the expression of VP1, VP2, and VP3 amino acid sequences so that they are found in wild-type viruses.

[0011] As shown in the examples, such constructs were identified by using a library of AAV capsid expression constructs against insect cells. Constructs were selected based on the primary requirement of highly efficient production of AAV capsids in insect cells and the secondary requirement of high infectivity in selected target cells. Thus, we also provide a highly efficient selection method that provides AAV capsid expression constructs having improved properties, such as improved production and / or improved infectivity.

[0012] Therefore, in a first embodiment of the present invention, a nucleic acid construct comprising an expression control sequence for the expression of a nucleotide sequence comprising an open reading frame in insect cells, wherein the open reading frame sequence is i) Adeno-associated virus (AAV) capsid proteins VP1, VP2, and VP3; and ii) AUG translation start codon for VP1 A nucleic acid construct is provided in which the nucleotide sequence codes for a start codon that is offset from the open reading frame upstream of the open reading frame. In other words, the alternative start codon is preferably located 3N+1 or 3N+2 nucleotides upstream of the start codon.

[0013] In another embodiment, the present invention provides a method for providing a nucleic acid construct encoding a parvovirus capsid protein for production in insect cells, wherein the nucleic acid construct has one or more improved properties, and the method is a) A step of providing a plurality of nucleic acid constructs, each construct comprising a nucleotide sequence encoding a parvovirus capsid protein operatively linked to an expression regulatory sequence, and at least one parvovirus inverted terminal repeat (ITR) sequence adjacent to the nucleotide sequence encoding the parvovirus capsid protein operatively linked to the expression regulatory sequence; b) A step of transferring the plurality of nucleic acid constructs into insect cells capable of expressing parvovirus Rep protein; c) Exposing the insect cells to conditions that enable the expression of parvovirus capsid protein and parvovirus rep protein, thereby allowing the nucleic acid construct to be packaged in a parvovirus capsid to provide parvovirus virions; d) A step of recovering parvovirus virions from the insect cells and / or insect cell supernatant; e) A step of bringing the parvovirus virion into contact with the target cell to enable infection of the target cell; f) Step of recovering the nucleic acid construct from the target cells. This provides a method that includes [something]. [Brief explanation of the drawing]

[0014] [Figure 1]This is a schematic representation of the library preparation and selection process. (a) First, a DNA library is provided. In this particular example, the library of expression constructs has a variety of start codons (XXX) for AAV5 VP1 and random nucleotides at selected positions (N) (SEQ ID NO: 71), and examples of such constructs are listed (1 is SEQ ID NO: 1; 2 is SEQ ID NO: 63; n is SEQ ID NO: 65); (b) The DNA library is transferred into a vector construct having an expression cassette with a promoter (P) for the expression of the AAV5 VP1, VP2, and VP3 capsid proteins (Cap(VP123)), the expression cassette being flanked by two AAV inverted terminal repeats (ITRs) that enable capsidation in the AAV capsid. Expression cassettes for Rep52 and Rep78 are also provided; (c) The Cap and Rep constructs are then transferred into insect cells, in this example, Sf9 cells. (d) Transfer may be via a baculovirus vector, which allows for control of the multiplicity of infection; therefore, in insect cells, the expressed Rep52 and Rep78 proteins replicate and capsidize the AAV vector genome containing the capsid expression cassette. As stated, when a baculovirus vector is used, the multiplicity of infection can be well controlled, and this multiplicity of infection is preferably kept below 1 for the Cap construct, so that there is an average of only one library member per Sf9 cell and cross-packaging is avoided. Only Cap expression cassettes that effectively produce capsids are thought to capsidize the vector genome; (e) The capsid containing the vector genome is then tested for infectivity, i.e., efficient transfer of the vector genome into target cells. Vector particles with a vector genome may be non-infectious, for example, while vector particles with a vector genome and a VP1:VP2:VP3 ratio of approximately 1:1:10 are highly infectious. In this example, the HeLaRC32 cell line, which is capable of replicating the AAV vector genome, is used. The vector genome can then be identified from the target cells.For example, a vector genome sequence or a portion thereof containing the various sequences shown in (a) can be determined. Alternatively, an identifier sequence can be determined to identify library members of (a) that have successfully infected target cells. Thus, in combination, steps (c) and (e) make it possible to select a capsid expression construct that enables efficient production in insect cells and produces infectious virions in target cells. A selected expression construct that dominates a population may be a particularly suitable candidate. The selected candidate expression construct (without adjacent ITRs) can then be used, for example, in a baculovirus vector or inserted into a cell line to produce an AAV gene therapy vector. [Figure 2A] These plots show the percentage (y-axis) of library members (x-axis) possessing a particular start codon at each stage of the selection process; A) This plot shows the distribution of start codons in the constructed plasmid DNA library with expression cassettes adjacent by AAV ITR. The prevalence rates vary between approximately 4% and 8%. [Figure 2B] B) This plot shows the distribution of start codons in the constructed baculovirus library with an inserted expression cassette flanked by an AAV ITR. Note that the prevalence varied between approximately 4% and 9%, and the distribution profile of this library is very similar to that of the plasmid library. [Figure 2C] C) This plot shows the distribution of start codons contained in the AAV library prepared from the baculovirus library in Figure 2B). Note that the distribution of start codons is very similar to that of the baculovirus library, ranging from approximately 4% to 9%, with one exception: the ATG start codon, which has a prevalence far below 0.5%. [Figure 2D]D) This plot shows the distribution of start codons contained in the AAV vector genome in cells infected with the AAV library of Fig. 2C). Note that the distribution of start codons is very different here. CTG and GTG are the most prevalent start codons with a prevalence of about 50%. The ATG start codon, which was not well represented in the AAV library, has a prevalence of about 8% here, while the remaining start codons had a prevalence of about 1% - 3%. [Figure 2E] E) This plot integrates each of the plots in Fig. 2A) - D). Note the drop for the ATG codon with respect to the AAV library, and the peaks for CTG, GTG, and ATG in the cell library. [Figure 3] The selected sequences (ATG1, ATG2, etc.) were then cloned into a baculovirus vector for the expression of the AAV5 capsid. Clones of the baculovirus vector were then analyzed by SDS - PAGE to assess the VP1:VP2:VP3 ratio. CTG1 did not produce good clones, while CTG2 and GTG2 did produce and exhibited the stoichiometry shown previously (Lubelski et al., WO 2015137802). The TAG clone resulted in a low titer, and the TGA clone did not appear to produce VP1. It was surprising that ATG1 and ATG2 produced good clones with a stoichiometry similar to CTG2. [Figure 4]This is a graph of AAV efficacy assays. The relative efficacy of various AAV vectors containing the SEAP reporter gene under the control of the CMV promoter was tested in Huh7 cells (A) and HeLa cells (B). Cells were infected with various infection multiplicities (AAV(gc) with 106, 105, and 104 genomic copies per cell), and SEAP reporter gene expression was determined. The ATG1 construct yielded the most potent vector, while ATG2, CTG2, and GTG2 had similar profiles, while TGA was significantly weaker, likely because TGA contained little to no VP1 protein. The GTG1 AAV vector yielded a low titer of gc / ml and therefore did not allow infection at an MOI of 106. [Figure 5]This is a schematic diagram of the ATG sequence context for efficient AAV capsid protein expression. A) The upper boxes, from left to right, show the codons in the open reading frame for VP1. The box with the VP1 start codon contains "ATG". The lower boxes are offset from the open reading frame for VP1. Upstream of the ATG codon is an alternative start codon (start), and downstream of that is a stop codon (stop). B) The sequence of a dominant sequence selected from the library is shown (SEQ ID NO: 1). In the out-of-frame duplicate reading frame (OOF), the ATG start codon is found upstream of the in-frame reading frame for Cap. The OOF has a TGA downstream stop codon in the sequence derived from the wild-type AAV5 sequence, and that OOF, when translated from the OOF start codon, would result in the 6-amino acid short peptide MHHGK (SEQ ID NO: 72); C) The sequence from a further out-of-frame duplicate reading frame from another upstream start codon is shown. The upper scenario has an out-of-frame CTG start codon (SEQ ID NO: 2) along with a stop codon further downstream of the sequence derived from the AAV5 sequence (see, in particular, SEQ ID NO: 70). This scenario would result in the translation of a larger protein sequence of approximately 158 amino acids, terminated by a TAG stop codon. The lower scenario also has a CTG start codon along with a stop codon in the mutated sequence, immediately downstream of the start codon, and this scenario would result in the short 4-amino acid peptide MEIW (SEQ ID NO: 73) when translated from OOF (SEQ ID NO: 9);D) Schematic diagrams of the expression of VP1, VP2, and VP3 capsid proteins from constructs depicted in Figures 5A-C. The DNA contains an expression cassette with a promoter (P) and an open reading frame (Cap(VP123)) for the capsid proteins. Transcription initiation is indicated by an arrow. Transcription yields mRNA, from which the OOF protein may be translated first, followed by the VP1, VP2, and VP3 capsid proteins. The OOF sequence overlaps with the VP1 translation start. [Figure 6A] Schematic diagrams of various vector vehicle configurations for preparing AAV libraries. Figure 6A: A configuration used in the examples is shown where an expression cassette (gray box) expressing the AAV capsid protein is contained within the baculovirus genome between the AAV ITRs. The AAV produced from that configuration contains a vector genome having an ITR adjacent to the expression cassette. [Figure 6B] Figure 6B: A configuration is shown where a vector vehicle (e.g., baculovirus) contains an expression cassette for a parvovirus capsid protein and a sequence identifier (ID) is placed between the vector genome ITR sequences. The AAV produced from that configuration contains a vector genome having an ITR adjacent to the sequence identifier. Since the Cap sequence and the ID are linked, the ID and the Cap sequence are associated in one genome, for example, by identifying the sequence identifier by sequencing, so that the corresponding cap sequence can be determined because the baculovirus vector is constructed by means such that the combination of the identifier sequence and the Cap expression sequence is known a priori. [Figure 6C] Figure 6C: The construct can also include a reporter gene in the parvovirus vector genome.

[0015] definition As used herein, the term "operatively linked" refers to the linkage of polynucleotide (or polypeptide) elements in a functional relationship. A nucleic acid is "operatively linked" when it is placed in a functional relationship with another nucleic acid sequence. For example, a transcriptional regulatory sequence is operatively linked to a coding sequence if it affects the transcription of the coding sequence. By operatively linked is meant that the linked DNA sequences are typically adjacent and, where two protein-coding regions are to be joined, are adjacent and in the reading frame.

[0016] A “regulatory sequence” refers to a nucleic acid sequence that regulates the expression of a nucleotide sequence to which it is operationally linked. When a regulatory sequence controls and regulates the transcription and / or translation of a nucleotide sequence, it is “operationally linked” to the nucleotide sequence. Therefore, regulatory sequences may include promoters, enhancers, internal ribosome entry sites (IRESs), transcription terminators, start codons preceding protein-coding genes, splicing signals for introns, and stop codons. The term “regulatory sequence” is intended to include at least sequences designed so that their presence affects expression, and may also include additional beneficial components. For example, leader sequences and fusion partner sequences are regulatory sequences. The term may also include nucleic acid sequence designs such that undesirable latent start codons in-frame and out-of-frame are removed from the sequence. The term may also include nucleic acid sequence designs such that undesirable latent splice sites are removed. The aforementioned terms include sequences referred to as poly-A sequences, poly-A tails, i.e., sequences that direct the addition of a series of adenine residues at the 3' end of mRNA, or polyadenylation sequences (pA). Expression regulatory sequences may also be designed to enhance mRNA stability. Expression regulatory sequences that affect transcriptional and translational stability, such as promoters, and sequences that affect translation, such as Kozak sequences, are known in insect cells. Expression regulatory sequences may be of a nature that modulates the nucleotide sequence to which the expression regulatory sequence is operatively linked, so that a lower or higher expression level is achieved.

[0017] As used herein, the terms “promoter” or “transcriptional regulatory sequence” refer to a nucleic acid fragment that functions to control the transcription of one or more coding sequences, is located upstream of the transcription start site of the coding sequence in the direction of transcription, and is structurally identified by the presence of any other DNA sequence including, but not limited to, a binding site for DNA-dependent RNA polymerase, a transcription start site, and transcription factor binding sites, repressor and activator protein binding sites, as well as any other sequence of nucleotides known to those skilled in the art to act directly or indirectly to regulate the amount of transcription from the promoter. A “constitutive” promoter is a promoter that is active in most tissues under most physiological and developmental conditions. An “inducible” promoter is a promoter that is physiologically or developmentally regulated, for example, by the application of a chemical inducer. A “tissue-specific” promoter is one that is active only in a specific type of tissue or cell.

[0018] The terms “substantially identical,” “substantially identical,” “essentially similar,” or “essentially similar” mean that two peptides or two nucleotide sequences, when optimally aligned by a program such as GAP or BESTFIT using default parameters, share at least a certain percentage of sequence identity as defined elsewhere herein. GAP aligns two sequences over their entire lengths using the Needleman and Wunsch global alignment algorithm to maximize the number of matches and minimize the number of gaps. Generally, GAP default parameters with a gap creation penalty of 50 (nucleotides) / 8 (protein) and a gap elongation penalty of 3 (nucleotides) / 2 (protein) are used. For nucleotides, the default score matrix used is nwsgapdna, and for proteins, the default score matrix is ​​Blosum62 (Henikoff & Henikoff, 1992, PNAS 89, 915-919). When an RNA sequence is said to be essentially similar to a DNA sequence or to have a certain degree of sequence identity, it is clear that thymine (T) in the DNA sequence is considered equivalent to uracil (U) in the RNA sequence. Sequence alignment and scores for sequence identity percentages can be determined using computer programs such as the GCG Wisconsin Package, version 10.3, available from Accelrys Inc., 9685 Scranton Road, San Diego, CA 92121-3752 USA, or the open-source software Emboss for Windows (current version 2.7.1-07). Alternatively, similarity or identity percentages can be determined by searching databases such as FASTA and BLAST.

[0019] The nucleotide sequences encoding the parvovirus Rep protein or Cap protein of the present invention can also be defined by their ability to hybridize with each other under moderate, preferably stringent, hybridization conditions. Stringent hybridization conditions are defined herein as conditions that enable nucleic acid sequences of at least about 25, preferably about 50, 75 or 100, and most preferably about 200 or more nucleotides to hybridize at about 65°C in a solution containing about 1 M of salt, preferably 6 × SSC, or any other solution having comparable ionic strength, and to wash at 65°C in a solution containing about 0.1 M of salt or less, preferably 0.2 × SSC, or any other solution having comparable ionic strength. Hybridization is preferably carried out overnight, i.e., for at least 10 hours, and washing is preferably carried out for at least 1 hour with at least two changes of the washing solution. These conditions are generally thought to enable specific hybridization of sequences with approximately 90% or more sequence identity.

[0020] Moderate conditions are defined herein as conditions that enable a nucleic acid sequence of at least about 50 nucleotides, preferably about 200 or more nucleotides, to hybridize at a temperature of about 45°C in a solution containing about 1 M of salt, preferably 6 × SSC, or any other solution having comparable ionic strength, and to wash at room temperature in a solution containing about 1 M of salt, preferably 6 × SSC, or any other solution having comparable ionic strength. The hybridization is preferably carried out overnight, i.e., for at least 10 hours, and the washing is preferably carried out for at least 1 hour with at least two changes of the washing solution. These conditions are generally considered to enable specific hybridization of sequences having up to 50% sequence identity. Those skilled in the art will be able to modify these hybridization conditions to specifically identify sequences with varying degrees of identity between 50% and 90%. [Modes for carrying out the invention]

[0021] This invention relates to the use of animal parvoviruses, particularly dependent viruses such as infectious human or monkey AAV, and their components (e.g., animal parvovirus genomes) for use as vectors for the introduction and / or expression of nucleic acids in mammalian cells. In particular, this invention relates to the enhancement of the infectivity of such parvovirus vectors when produced in insect cells.

[0022] Viruses of the Parvoviridae family are small DNA animal viruses. Parvoviridae can be divided into two subfamilies: Parvovirinae, which infects vertebrates, and Densovirinae, which infects insects. Members of the Parvovirinae subfamily are referred to herein as parvoviruses and include the genus Dependvirus. As can be inferred from the name of the genus Dependvirus, members of Dependvirus are unique in that they usually require co-infection with a helper virus, such as an adenovirus or herpesvirus, for productive infection in cell culture. The Dependvirus genus includes AAV, which typically infects humans (e.g., serotypes 1, 2, 3A, 3B, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13) or primates (e.g., serotypes 1 and 4), and closely related viruses that infect other warm-blooded animals (e.g., cattle, dog, horse, and sheep adeno-associated viruses). Further information on parvoviruses and other members of the Parvoviridae genus is provided in Chapter 69 of Kenneth I. Berns, "Parvoviridae: The Viruses and Their Replication," in Fields Virology (Vol. 3, 1996). For convenience, the present invention is further illustrated and described herein by reference to AAV. However, it is understood that the present invention is not limited to AAV and can be equally applied to other parvoviruses.

[0023] The genomic organization of all known AAV serotypes is very similar. The AAV genome is a linear single-stranded DNA molecule less than approximately 5,000 nucleotides (nt) in length. Inverted terminal repeats (ITRs) flank specific coding nucleotide sequences for unstructured replication (Rep) and structural (VP) proteins. VP proteins (VP1, -2, and -3) form a capsid. The terminal 145nt are self-complementary and organized to form energetically stable intramolecular double helixes that form a T-shaped hairpin. These hairpin structures function as starting points for viral DNA replication and act as primers for cellular DNA polymerase complexes. After wtAAV infection in mammalian cells, Rep genes (i.e., Rep78 and Rep52) are expressed from the P5 promoter and P19 promoter, respectively, and both Rep proteins have a function in viral genome replication. The splicing event in Rep ORFs actually results in the expression of four Rep proteins (i.e., Rep78, Rep68, Rep52, and Rep40). However, in mammalian cells, it has been shown that unspliced ​​mRNA encoding Rep78 and Rep52 proteins is sufficient for AAV vector production. In insect cells as well, Rep78 and Rep52 proteins are sufficient for AAV vector production. Three capsid proteins, VP1, VP2, and VP3, are expressed from a single VP reading frame from the p40 promoter. wtAAV infection in mammalian cells depends on a combination of alternating use of two splice acceptor sites and suboptimal utilization of the ACG start codon for VP2 for capsid protein production.

[0024] In insect cells, expression of transcripts containing an AAV open reading frame encoding the VP1 (with an AUG start codon), VP2, and VP3 proteins, i.e., mRNA, does not normally produce sufficient amounts of VP1, VP2, and VP3 capsid proteins to yield potent AAV in a ratio of approximately 1:1:10. Potency is defined herein as the ability of an AAV vector to transfer its vector genome into target cells and enable efficient expression of the transgene. We have now surprisingly found that AAV capsids can be produced very efficiently in insect cells from expression constructs encoding transcripts for the VP1, VP2, and VP3 proteins, where VP1 is translated from an AUG start codon.

[0025] The expression constructs identified in this invention enable the efficient production in insect cells of substantial quantities of highly potent AAV gene therapy vectors for use in medical treatment. Such vectors are at least as potent, if not superior, to AAV gene therapy vectors produced from alternative start codons such as CTG or GTG (Figure 4). With respect to nucleic acid sequences, nucleic acid sequences are understood to be enumerated as DNA sequences listing A, T, C, and G, or as RNA sequences listing A, U, C, and G. Expression constructs may usually refer to DNA sequences, while expressed nucleotide sequences are understood to refer to RNA sequences, i.e., mRNA transcribed or expressed from expression constructs.

[0026] The construct of the present invention encodes an additional out-of-frame start codon 5' from the VP1 start codon, which appears to result in a reduction of translation initiation at the VP1 start codon, while allowing for further translation of sufficient amounts of both VP2 and VP3. Without being constrained by theory, such an out-of-frame 5' start codon would result in interference with transcription initiation at the VP1 AUG start codon, allowing for leaky pseudo-ribosome scanning similar to that occurring in wild-type AAV. Without being constrained by theory, the synthesis of short peptides from these alternative start codons (e.g., translation termination of an out-of-frame reading frame before the VP2 coding sequence) could allow the ribosome to continue scanning downstream of the VP1 AUG start codon or restart the ribosome, enabling translation of VP2 and VP3 from the same transcript.

[0027] Such constructs provide at least comparable, if not superior, production of AAV capsids in insect cells compared to those produced in the prior art. When used to create expression constructs for insect cells, such constructs are advantageous because they may allow the VP1, VP2, and VP3 nucleotide sequences to remain unchanged, as they are found in wild-type viruses. Constructs according to the present invention may allow the amino acid sequences of the VP1, VP2, and VP3 capsid proteins to be substantially identical to, or identical to, those of, the capsid proteins found in wild-type viruses. Therefore, this expression strategy is generally applicable to any parvovirus vector construct or AAV vector construct and may not require any further adjustment of the 5' sequence or the AAV capsid open reading frame sequence.

[0028] Therefore, in a first embodiment of the present invention, a nucleic acid construct comprising an expression control sequence for the expression of a nucleotide sequence comprising an open reading frame in insect cells, wherein the open reading frame sequence is i) Adeno-associated virus (AAV) capsid proteins VP1, VP2, and VP3; and ii) ATG translation start codon for VP1 A nucleic acid construct is provided in which the nucleotide sequence codes for a nucleotide sequence that includes an alternative start codon upstream of the open reading frame, which is offset from the open reading frame.

[0029] The expression of nucleotide sequences according to the present invention is understood to be related to the expressed mRNA. Therefore, the alternative start codon is an object contained within the mRNA, i.e., the alternative start codon is contained within the sequence 5' from the open reading frame encoding the capsid protein, and the alternative start codon is located 3' from the transcription start site of the nucleic acid construct. Hence, the alternative reading frame is located 5' from the VP1 AUG codon contained within the expressed mRNA. The open reading frame according to the present invention is understood to be a single open reading frame, i.e., the sequences encoding the capsid proteins VP1, VP2, and VP3 are understood to be duplicated. In other words, the VP2 and VP3 proteins are encoded by the same sequence as the VP1 sequence. Such an open reading frame may be a contiguous open reading frame, but may not be contiguous, and may contain, for example, intron sequences. The open reading frame in which VP1, VP2, and VP3 are translated is a single, adjacent open reading frame, and it is preferable that no further transcripts are transcribed in insects from transcripts in which capsid proteins can be translated (for example, one transcript encoding VP1, another encoding VP2, and yet another transcript encoding VP3).

[0030] The out-of-frame start codon is preferably selected from the group consisting of CUG, ACG, AUG, UUG, CUC, and CUU. The surrogate start codon is more preferably selected from AUG or CUG. The surrogate start codon is most preferably AUG. As shown in the Examples section, sequences having the most common AUG start codon mostly contained out-of-frame start codons. Upstream out-of-frame start codons are mostly relatively strong codons such as UUG, CUG, GUG, AUG, and ACG. Weaker start codons such as CUC and CUU were also observed. The most common and most preferred out-of-frame surrogate start codon is AUG.

[0031] An alternative start codon can be the start of an alternative open reading frame. Therefore, an alternative start codon is understood to contain a codon that allows the ribosome to initiate translation. Sometimes, if the start codon is near the 5' capped end of the mRNA, for example, such a sequence may not be allowed to function as a start codon. Because the gene code is a triplet encoding amino acids, a nucleic acid sequence is understood to be able to be translated into three different amino acid sequences, depending on where translation begins and ends. An out-of-frame alternative start codon is upstream of the VP1 AUG start codon, and the gene code following the alternative start codon is preferably such that a translation termination occurs that prevents the ribosome from initiating or initiating translation from the VP1 AUG start codon. Similarly, without theoretical constraint, an out-of-frame alternative start codon upstream of the VP1 AUG start codon allows for the initiation of translation from the mRNA. The alternative open reading frame preferably terminates downstream of the VP1 AUG start codon. For example, if an A follows immediately after the VP1 AUG start codon, the UGA triplet in the AUGA sequence encodes a termination codon. Therefore, it is preferable that an alternate open reading frame beginning with an upstream alternate start codon contains the VP1 AUG start codon.

[0032] Therefore, in a further embodiment according to the present invention, a nucleic acid construct for expression in insect cells of a nucleotide sequence comprising an open reading frame is provided, wherein the open reading frame sequence encodes adeno-associated virus (AAV) capsid proteins VP1, VP2, and VP3, and the AUG translation start codon for VP1, and the nucleotide sequence comprises an alternate open reading frame beginning with an alternate start codon, the alternate open reading frame encompassing the AUG translation start codon for VP1.

[0033] An alternative open reading frame may begin at least 100, 90, 80, 70, 60, 50, 40, 30, 20, or 10 nucleotides 5' from the VP1 AUG start codon and terminate thereafter. An alternative open reading frame may begin at least 5' from the VP1 AUG start codon and terminate thereafter at least 500, 400, 300, 200, 150, 100, 90, 80, 70, 60, 50, 40, 30, 20, or 10 nucleotides. An alternative open reading frame may begin at least 50 nucleotides 5' from the VP1 AUG start codon and terminate thereafter at least 500 nucleotides. An alternative open reading frame may begin at least 40 nucleotides 5' from the VP1 AUG start codon and terminate thereafter at least 200 nucleotides. An alternative open reading frame may begin at most 30 nucleotides 5' from the VP1 AUG start codon and terminate at most 50 nucleotides thereafter. An alternative open reading frame may begin at most 10 nucleotides from the VP1 AUG start codon and terminate at most 20 nucleotides thereafter. In one alternative embodiment, the alternative open reading frame terminates before the VP3 start codon, preferably before the VP2 start codon. For example, such an alternative open reading frame shown in the example may begin 4 nucleotides upstream and terminate at 14 nucleotides thereafter, or begin 8 nucleotides upstream and terminate at 4 or more nucleotides thereafter.

[0034] Such alternative open reading frames may preferably be contained within the DNA sequences encoding adeno-associated virus (AAV) capsid proteins VP1, VP2, and VP3, including the sequence upstream of the VP1 ATG start codon sequence, which is the sequence encoded by nucleotides 105-155 of the DNA sequence of Sequence ID No. 70. The sequence upstream of the ATG start codon is transcribed into RNA. Such alternative open reading frames may also be contained within the DNA sequences encoding adeno-associated virus (AAV) capsid proteins VP1, VP2, and VP3, including the sequence upstream of the VP1 ATG start codon sequence, which is the sequence encoded by nucleotides 1-155 of the DNA sequence of Sequence ID No. 70. The upstream sequence encodes the polyhedrin promoter and 5' leader sequence upstream of the ATG VP1 start codon (105-155).

[0035] Therefore, the alternative open reading frames of the present invention described above are preferably translated into peptides in insect cells. In one embodiment, the peptide has a length of at least 4 amino acids, at least 5 amino acids, or at least 6 amino acids. In one embodiment, the translated amino acid sequence includes or consists of SEQ ID NO: 72 or SEQ ID NO: 73. In another embodiment, the peptide has a length of at most 200, 150, 100, 50, 40, 30, 20, or 10 amino acids. In a further embodiment, a nucleic acid construct according to the present invention encoding an alternative open reading frame is translated into a peptide having a length of 2 to 200 amino acids, 2 to 100, 2 to 50, or preferably 2 to 10 amino acids. Thus, a nucleic acid construct according to the present invention described herein, comprising an alternative open reading frame following an alternative start codon, encodes a peptide. The peptide length may depend on the sequence following the VP1 start codon, i.e., the VP1 encoding sequence, which may be derived from, for example, a naturally occurring AAV sequence, or from a synthetic or artificial AAV capsid sequence (e.g., a codon-optimized or enhanced variant). Therefore, the length depends on the location of the stop codon (TGA, TAA, TAG) in the out-of-frame reading frame, starting from an alternative start codon upstream of the VP1 ATG start codon. The sequence downstream of the start codon may be mutated to introduce an out-of-frame upstream start codon and an in-frame stop codon. Thus, the peptide length can be intentionally selected. This allows for the introduction of an out-of-frame stop codon that does not introduce a change in the amino acid sequence with respect to the VP1 coding sequence, in other words, a silent mutation in the VP1 reading frame. The introduced out-of-reading-frame stop codon may be introduced by one, two, or three point mutations in three consecutive nucleic acids in the reading frame.Triplet sequences (i.e., TGA, TAA, or TAG) can also be inserted into the VP1 coding sequence, resulting in a one-amino acid insertion relative to the length of the coding sequence and an additional amino acid change to the VP1 coding sequence (i.e., one triplet in the VP1 coding sequence can be changed to two triplets by the insertion of an out-of-frame stop codon).

[0036] In another embodiment, a nucleic acid construct comprising an expression regulatory sequence for the expression of a nucleotide sequence including an open reading frame in insect cells, wherein the open reading frame sequence is i) Adeno-associated virus (AAV) capsid proteins VP1, VP2, and VP3; and ii) AUG translation start codon for VP1 A nucleic acid construct is provided, wherein the nucleotide sequence codes for VP1 AUG and contains nucleotides 1-8 of a nucleotide sequence selected from the group consisting of SEQ ID NOs. 32-62 refer to RNA sequences, and therefore the nucleic acid construct is understood to have a corresponding DNA sequence that codes for an RNA sequence, such as those listed in SEQ ID NOs. 32-62 refer to RNA sequences, and thus the nucleic acid construct is understood to have a corresponding DNA sequence that codes for an RNA sequence, such as those listed in SEQ ID NOs. The nucleotide sequence preferably contains a G nucleotide immediately downstream of VP1 AUG. The nucleic acid construct according to the present invention includes a sequence selected from the group consisting of SEQ ID NOs. 1-31 that codes for the VP1 start codon, and more preferably the VP1 start codon corresponds to positions 9-11 of SEQ ID NOs. 32 sequences are most preferably derived from SEQ ID NOs. 1 and 32, i.e., nucleotides 1-8 of SEQ ID NOs. 1 and 32 are preferred, and it is preferred that they have a G directly adjacent to VP1 ATG, and most preferably that either codes for the entire sequence of SEQ ID NOs. 1.

[0037] In a further embodiment, a nucleic acid construct according to the present invention is provided, wherein the second codon of the open reading frame of VP1 encodes an amino acid residue selected from the group consisting of alanine, glycine, valine, aspartic acid, and glutamic acid. This second amino acid residue may be derived from a codon inserted between the start codon and, for example, a second codon derived from a wild-type AAV VP1 sequence, or the second codon of the VP1 nucleotide sequence may be a mutated codon (for example, by mutating the nucleic acid immediately following the VP1 ATG codon to G). It is most preferable that the second codon of VP1 encodes valine. It is more preferable that the second codon is selected from the group consisting of GUA, GUC, GUU, and GUG, and it is preferable that the second codon is GUA. The open reading frame optionally includes, after the second codon, one or more codons encoding further additional amino acid residues, e.g., 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 additional amino acids, but preferably fewer than 60, 50, 40, 35, 30, 25, 20, 19, 18, 17, 16, 15, or 14 additional amino acid residues. As will be easily understood, the codons encoding the additional amino acid residues should be in-frame with the open reading frame of the capsid protein.

[0038] Accordingly, in one embodiment, an AAV vector is provided containing a VP1 capsid protein having valine at position 2 of VP1, for example, by modification of position 2 of the wild-type VP1 capsid protein sequence, or by insertion of a valine codon between positions 1 and 2 of the wild-type VP1 capsid protein sequence, or because naturally occurring or already selected VP1 capsid proteins contain valine at position 2. Such capsids, preferably produced in insect cells, may be particularly useful in the medical treatments described herein.

[0039] In one embodiment, when an open reading frame is compared to a wild-type capsid protein, the open reading frame encoding the capsid protein further includes codons encoding one or more amino acid residues inserted between the ATG translation start codon of VP1 and the codon encoding the amino acid residue immediately adjacent at its 3' end in the corresponding wild-type capsid protein. For example, the open reading frame includes codons for 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 additional amino acid residues compared to the corresponding wild-type capsid protein. Preferably, the open reading frame includes codons for 60, 50, 40, 35, 30, 25, 20, 19, 18, 17, 16, 15, or fewer than 14 additional amino acid residues compared to the corresponding wild-type capsid protein. As should be easily understood, codons encoding additional amino acid residues should be in-frame with the open reading frame of the capsid protein. Of these codons encoding additional amino acid residues compared to the corresponding wild-type capsid protein, the first codon, i.e., the codon immediately adjacent at its 3' end to the suboptimal translation initiation codon, encodes an amino acid residue selected from the group consisting of alanine, glycine, valine, aspartic acid, and glutamic acid. Therefore, if there is only one additional codon between the translation initiation codon and the codon encoding the amino acid residue corresponding to residue 2 in the wild-type sequence, that additional codon encodes an amino acid residue selected from the group consisting of alanine, glycine, valine, aspartic acid, and glutamic acid. If there are more than one additional codons between the translation initiation codon and the codon encoding amino acid residue 2 in the wild-type sequence, the codon immediately following the translation initiation codon encodes an amino acid residue selected from the group consisting of alanine, glycine, valine, aspartic acid, and glutamic acid. The additional amino acid residue immediately following the suboptimal translation initiation codon (i.e., at the 3' end of the translation initiation codon) is preferably valine.In other words, in a preferred embodiment of the present invention, the codon immediately following the suboptimal translation initiation codon encodes valine.

[0040] The sequence encoding the AAV capsid protein in step a) may be a naturally occurring capsid sequence, such as those of AAV1-AAV13, whose nucleotide and amino acid sequences are listed as SEQ ID NOs. 13-38 in International Publication No. 2015137802 by Lubelski et al., which is incorporated herein by reference in its entirety. Thus, a nucleic acid construct according to the present invention may include the entire open reading frame for the AAV capsid protein disclosed by International Publication No. 2015137802 by Lubelski et al. Alternatively, the sequence may be artificial, for example, a hybrid form, or codon-optimized, for example, by AcmNPv or the codon usage of the fall armyworm (Spodoptera frugiperda). For example, the capsid sequence may consist of the VP2 and VP3 sequences of AAV1, while the remaining VP1 sequence is from AAV5. Preferred capsid proteins are AAV5, preferably provided in SEQ ID NO: 22, or AAV8, preferably provided in SEQ ID NO: 28, as listed in International Publication No. 2015137802 by Lubelski et al. Therefore, in preferred embodiments, the AAV capsid protein is the AAV serotype 5 or AAV serotype 8 capsid protein modified according to the present invention. More preferably, the AAV capsid protein is the AAV serotype 5 capsid protein modified according to the present invention. It is understood that the exact molecular weight of the capsid protein and the exact position of the translation start codon may differ among various parvoviruses. However, those skilled in the art will know how to identify the corresponding position in the nucleotide sequence derived from parvoviruses other than AAV-5. Alternatively, the sequence encoding the AAV capsid protein may be an artificial sequence, for example, as a result of directed evolution experiments. This directed evolutionary method experiment may include the creation of capsid libraries by DNA shuffling, error-prone PCR, bioinformatics rational design, and site-saturated mutagenesis. The resulting capsids are based on existing serotypes but contain various amino acid or nucleotide changes that enhance the characteristics of such capsids.The resulting capsid may be a “shuffled capsid”—a combination of various parts of an existing serotype—or it may contain entirely novel alterations, i.e., the addition, deletion, or substitution of one or more amino acids or nucleotides, organized into groups or extending throughout the entire length of the gene or protein. See, for example, Schaffer and Maheshri; Proceedings of the 26th Annual International Conference of the IEEE EMBS San Francisco, CA, USA; September 1-5, 2004, pp. 3520-3523; Asuri et al. (2012) Molecular Therapy 20(2):329-3389; Lisowski et al. (2014) Nature 506(7488):382-386.

[0041] In a preferred embodiment of the present invention, the open reading frame encoding the VP3 capsid protein begins with a non-standard translation initiation codon selected from the group consisting of ACG, ATT, ATA, AGA, AGG, AAA, CTG, CTT, CTC, CTA, CGA, CGC, TTG, TAG, and GTG. The non-standard translation initiation codon is preferably selected from the group consisting of GTG, CTG, ACG, and TTG, and more preferably CTG.

[0042] The nucleotide sequences of the present invention for the expression of AAV capsid proteins preferably further comprise at least one modification of the nucleotide sequence encoding the AAV VP1 capsid protein, selected from G at nucleotide position 12, A at nucleotide position 21, and C at nucleotide position 24 of the VP1 open reading frame, wherein the nucleotide position corresponds to the nucleotide position of the wild-type nucleotide sequence. “Potential / possible false start site” or “Potential / possible false translation start codon” is understood herein to mean an in-frame ATG codon located in the coding sequence of the capsid protein(s). The elimination of possible false start sites for translation in the VP1 coding sequences of other serotypes will be well understood by those skilled in the art, as is the elimination of putative splice sites that may be recognized in insect cells. For example, a nucleotide modification at position 12 is not required for recombinant AAV5 because the T nucleotide does not produce a false ATG codon. Various modifications of the wild-type AAV sequence for proper expression in insect cells can be achieved by applying well-known genetic engineering techniques, such as those described in Sambrook and Russell (2001) "Molecular Cloning: A Laboratory Manual" (Part 3), Cold Spring Harbor Laboratory, Cold Spring Harbor Laboratory Press, New York. Various further modifications of the VP coding region that can increase the production of VP and virions, or have other desired effects such as altered directionality, or reduce the antigenicity of virions are known to those skilled in the art. These modifications are within the scope of the present invention.

[0043] The nucleotide sequence of the present invention encoding the AAV capsid protein is preferably operatively linked to an expression regulatory sequence for expression in insect cells. Therefore, in a second embodiment, the present invention relates to a nucleic acid construct comprising a nucleic acid molecule according to the present invention, wherein the nucleotide sequence of an open reading frame encoding the adeno-associated virus (AAV) capsid protein is operatively linked to an expression regulatory sequence for expression in insect cells. These expression regulatory sequences are thought to include at least a promoter that is active in insect cells. The present invention can be put into practice using techniques known to those skilled in the art for expressing foreign genes in insect host cells.Methodologies for molecular manipulation and expression of polypeptides in insect cells include, for example, Summers and Smith. 1986. A Manual of Methods for Baculovirus Vectors and Insect Culture Procedures, Texas Agricultural Experimental Station Bull. No. 7555, College Station, Tex.; Luckow. 1991. In Prokop et al., Cloning and Expression of Heterologous Genes in Insect Cells with Baculovirus Vectors' Recombinant DNA Technology and Applications, 97-152; King, LA and RD Possee, 1992, The baculovirus expression system, Chapman and Hall, United Kingdom; O'Reilly, DR, LK Miller, VALuckow, 1992, Baculovirus Expression Vectors: A Laboratory Manual, New York; WH Freeman and Richardson, CD, 1995, Baculovirus Expression Protocols, Methods in Molecular The information is described in Biology, Vol. 39; U.S. Patent No. 4,745,051; U.S. Patent Application Publication No. 2003148506; and International Publication No. 03 / 074714. Particularly suitable promoters for the transcription of the nucleotide sequences of the present invention encoding the AAV capsid protein are, for example, polyhedron promoters (polH), such polH promoters are provided in SEQ ID NO: 70 (or listed as SEQ ID NO: 53 and its abbreviated form SEQ ID NO: 54 in International Publication No. 2015137802 by Lubelski et al.).However, other promoters that are active in insect cells and can be selected according to the present invention, such as the polyhedrin (polH) promoter, the p10 promoter, the p35 promoter, the 4×Hsp27 EcRE+ minimum Hsp70 promoter, the delta E1 promoter, the E1 promoter, or the IE-1 promoter, and further promoters described in the above references are known in the art.

[0044] Nucleic acid constructs for the expression of AAV capsid protein in insect cells are preferably insect cell-compatible vectors. “Insect cell-compatible vector” or “vector” is understood to be a nucleic acid molecule capable of productive transformation or transfection of insects or insect cells. Exemplary biological vectors include plasmids, linear nucleic acid molecules, and recombinant viruses. Any vector may be employed, as long as it is insect cell-compatible. While vectors may be integrated into the insect cell genome, the presence of the vector in insect cells does not need to be permanent, and transient episomal vectors are also included. Vectors can be introduced by any known means, for example, by chemical treatment of cells, electroporation, or infection. In preferred embodiments, the vector is a baculovirus, a viral vector, or a plasmid. In more preferred embodiments, the vector is a baculovirus, i.e., the construct is a baculovirus vector. Baculovirus vectors and methods for using baculovirus vectors are described in the references cited for molecular manipulation of insect cells.

[0045] In a third aspect, the present invention relates to insect cells comprising the nucleic acid constructs of the present invention as defined above. Any insect cells that can enable AAV replication and be maintained under culture may be used according to the present invention. For example, the cell lines used may be those derived from the fall armyworm, Drosophila cell lines, or mosquito cell lines, such as Aedes albopictus cell lines. Preferred insect cells or cell lines are those derived from insect species susceptible to baculovirus infection, including, for example, expressSF+(registered trademark), Drosophila Schneider 2(S2) cells, Se301, SeIZD2109, SeUCR1, Sf9, Sf900+, Sf21, BTI-TN-5B1-4, MG-1, Tn368, HzAm1, Ha2302, Hz2E5, and High Five from Invitrogen.

[0046] A preferred insect cell according to the present invention further comprises: (a) a second nucleotide sequence comprising at least one AAV inverted terminal repeat (ITR) nucleotide sequence; (b) a third nucleotide sequence comprising a Rep52 or Rep40 coding sequence operatively ligated to an expression regulatory sequence for expression in insect cells; and (c) a fourth nucleotide sequence comprising a Rep78 or Rep68 coding sequence operatively ligated to an expression regulatory sequence for expression in insect cells.

[0047] In the context of this invention, “at least one AAV ITR nucleotide sequence” is understood to mean a palindromic sequence containing mostly complementary and symmetrically arranged sequences, also referred to as “A,” “B,” and “C” regions. The ITR functions as a replication origin, being a site that has a “cis” role in replication, i.e., a recognition site for a trans-acting replication protein (e.g., Rep78 or Rep68) that recognizes the palindrome and specific sequences within it. One exception to the symmetry of the ITR sequence is the “D” region of the ITR. The D region is unique (it does not have a complement within a single ITR). Nicking of single-stranded DNA occurs at the junction between the A and D regions. The junction is the region where new DNA synthesis begins. The D region is usually located on one side of the palindrome and provides directionality to the nucleic acid replication steps. AAV replicating in mammalian cells typically has two ITR sequences. However, it is possible to manipulate the ITR so that binding sites are on both strands of the A region and the D region is symmetrically located one on each side of the palindrome. With respect to a double-stranded circular DNA template (e.g., a plasmid), nucleic acid replication supported by Rep78 or Rep68 then proceeds in both directions, and a single ITR is sufficient for AAV replication of a circular vector. Therefore, one ITR nucleotide sequence may be used in the context of the present invention. However, it is preferable that two or another even number of regular ITRs are used. It is most preferable that two ITR sequences are used. Considering the safety of the viral vector, it is sometimes desirable to construct a viral vector that cannot reproduce further after initial introduction into cells. Such a safety mechanism to suppress undesirable vector reproduction in the recipient can be provided by using rAAV having a chimeric ITR as described in U.S. Patent Application Publication 2003148506. In a preferred embodiment, the nucleotide sequences encoding the parvovirus VP1, VP2, and VP3 capsid proteins include at least one in-frame insertion of a sequence encoding an immune-evading repeat, such as as described in International Publication 2009 / 154452. This insertion leads to the formation of so-called self-complementary or monomeric double-stranded parvovirus virions.In preferred embodiments, the sequences encoding parvovirus VP1, VP2, and VP3 capsid proteins include monomeric double-stranded or self-complementary genomes. For the preparation of monomeric double-stranded AAV vectors, the AAV Rep proteins and AAV capsid proteins are expressed in insect cells according to the present invention in the presence of a vector genome containing at least one AAV ITR, with increased Rep52 and / or Rep40 protein expression compared to Rep78 and / or Rep68 protein expression. Monomeric double-stranded AAV vectors can also be prepared by expressing the AAV Rep proteins and AAV Cap proteins in insect cells in the presence of a vector genome construct flanked by at least one AAV ITR, as described, for example, in International Publication No. 2011 / 122950, ​​with decreased nicking activity of Rep78 and / or Rep60 compared to helicase / capsidation activity of Rep52 and / or Rep40.

[0048] The number of vectors or nucleic acid constructs used is not limited in this invention. For example, one, two, three, four, five, six, or more vectors may be used to produce AAV in insect cells according to this invention. If six vectors are used, one vector may encode AAV VP1, another may encode AAV VP2, yet another may encode AAV VP3, yet another may encode Rep52 or Rep40, while Rep78 or Rep68 may be encoded by another vector, and the last vector may contain at least one AAV ITR. Additional vectors may be used to express, for example, Rep52 and Rep40, and Rep78 and Rep68. If fewer than six vectors are used, the vectors may contain at least one AAV ITR and various combinations of VP1, VP2, VP3, Rep52 / Rep40, and Rep78 / Rep68 coding sequences. It is preferable that two or three vectors are used, and more preferable that two vectors are used as described above. When two vectors are used, it is preferable that the insect cell includes (a) a first nucleic acid construct for the expression of the AAV capsid protein as defined above, wherein the construct further comprises third and fourth nucleotide sequences as defined in (b) and (c) above, the third nucleotide sequence comprising a Rep52 or Rep40 coding sequence operatively linked to at least one expression regulatory sequence for expression in the insect cell, and the fourth nucleotide sequence comprising a Rep78 or Rep68 coding sequence operatively linked to at least one expression regulatory sequence for expression in the insect cell; and (b) a second nucleic acid construct comprising a second nucleotide sequence as defined in (a) above, comprising at least one AAV ITR nucleotide sequence. When three vectors are used, it is preferable to use the same configuration as that used for two vectors, except that separate vectors are used for the expression of the capsid protein and the Rep52, Rep40, Rep78, and Rep68 proteins. The sequences in each vector can be in any order relative to each other.For example, if a vector contains an ITR and an ORF containing a nucleotide sequence encoding the VP capsid protein, the VP ORF may be located in the vector such that the VP ORF is replicated or not replicated if there is DNA replication between the ITR sequences. As another example, the ORF containing the Rep-coding sequence and / or the nucleotide sequence encoding the VP capsid protein may be in any order in the vector. The second, third, and further nucleic acid constructs are also preferably insect cell-compatible vectors, and are understood to be preferably baculovirus vectors as described above. Alternatively, in the insect cells of the present invention, one or more of the first nucleotide sequence, the second nucleotide sequence, the third nucleotide sequence, and the fourth nucleotide sequence, and any further nucleotide sequences of any choice may be stably incorporated into the genome of the insect cell. Those skilled in the art know how to stably introduce nucleotide sequences into an insect genome and how to identify cells having such nucleotide sequences in their genomes. Incorporation into the genome can be assisted, for example, by using a vector containing a nucleotide sequence highly homologous to a region of the insect genome. The use of specific sequences, such as transposons, is another method for introducing nucleotide sequences into the genome.

[0049] Therefore, in a preferred embodiment, an insect cell according to the present invention includes (a) a first nucleic acid construct according to the present invention, wherein the first nucleic acid construct further comprises the third and fourth nucleotide sequences defined above; and (b) a second nucleic acid construct comprising the second nucleotide sequence defined above, wherein the second nucleic acid construct is preferably an insect cell-compatible vector and more preferably a baculovirus vector.

[0050] In a preferred embodiment of the present invention, the second nucleotide sequence present in the insect cell of the present invention, i.e., the sequence comprising at least one AAV ITR, further comprises at least one nucleotide sequence encoding a gene product of interest (preferably for expression in mammalian cells), wherein the at least one nucleotide sequence encoding the gene product of interest is preferably incorporated into the AAV genome produced in the insect cell. The at least one nucleotide sequence encoding the gene product of interest is preferably a sequence for expression in mammalian cells. The second nucleotide sequence comprises two AAV ITR nucleotide sequences, and the at least one nucleotide sequence encoding the gene product of interest is preferably located between the two AAV ITR nucleotide sequences. The nucleotide sequence encoding the gene product of interest (for expression in mammalian cells) is preferably incorporated into the AAV genome produced in the insect cell if the nucleotide sequence is located between two regular ITRs or on either side of an ITR that has been manipulated to have two D regions. Therefore, in a preferred embodiment, the present invention provides an insect cell according to the present invention, wherein the second nucleotide sequence comprises two AAV ITR nucleotide sequences, and at least one nucleotide sequence encoding a gene product of interest is located between the two AAV ITR nucleotide sequences.

[0051] Gene products of interest, including ITRs, are typically 5,000 nucleotides (nt) or less in length. In another embodiment, oversized DNA, i.e., DNA longer than 5,000 nt, can be expressed in vitro or in vivo using the AAV vectors described herein. Oversized DNA is understood herein as DNA exceeding the maximum AAV packaging limit of 5 kbp. Therefore, it is also possible to create AAV vectors capable of producing recombinant proteins encoded by genomes typically larger than 5.0 kb. For example, the inventors created an rAAV5 vector containing a partially unidirectionally packaged fragment of hFVIII in insect cells. The total size of the vector genome, containing at least 5.6 kb, was packaged into two populations of FVIII fragment-containing AAV5 particles. These variant AAV5-FVIII vectors were shown to promote the expression and secretion of active FVIII. This propagation was confirmed in vitro, where an AAV vector containing the gene product of interest encoding factor VIII, following infection of Huh7 cells, resulted in the production of active FVIII protein. Similarly, tail vein delivery of rAAV.FVIII in mice resulted in the production of active FVIII protein. Molecular analysis of the capsidated product clearly showed that the 5.6 kbp FVIII expression cassette was not completely capsidated in the AAV particle. Without being constrained by any theory, the inventors hypothesize that the + and - DNA strands of the capsidated molecule exhibited a missing 5' end. This hypothesis is consistent with previously reported unidirectional (3' end-starting) packaging mechanisms that operate according to the "head-full principle" with a 4.7–4.9 kbp limit (see, for example, Wu et al.

[2010] Molecular Therapy 18(1):80–86; Dong et al.

[2010] Molecular Therapy 18(1):87–92; Kapranov et al.

[2012] Human Gene Therapy 23:46–55; and especially Lai et al.

[2010] Molecular Therapy 18(1):75–79).Although only approximately 5 kb of the entire 5.6 kb vector genome was capsidized, the vector was potent and led to the expression of active FVIII. We demonstrated that the correct template for FVIII production assembled in target cells based on the partial complementarity of the + and - DNA strands, followed by the synthesis of the second strand.

[0052] Therefore, the second nucleotide sequence defined herein may include a nucleotide sequence encoding at least one “gene product of interest” for expression in mammalian cells, positioned so that the nucleotide sequence is incorporated into the AAV genome replicated in insect cells. Any nucleotide sequence may be incorporated into mammalian cells transfected with AAV produced according to the present invention for subsequent expression, provided that the construct remains within the packaging capacity of the AAV virion. The nucleotide sequence may encode, for example, a protein, or it may express an RNAi agent, i.e., an RNA molecule capable of RNA interference, such as shRNA (short hairpin RNA) or siRNA (short interfering RNA). “siRNA” means a small interfering RNA, which is a short, double-stranded RNA that is not toxic in mammalian cells (Elbashir et al., 2001, Nature). 411 :494~98;Caplen et al., 2001, Proc. Natl. Acad. Sci. USA 98 :9742~47). In a preferred embodiment, the second nucleotide sequence may comprise two nucleotide sequences, each encoding a gene product of interest for expression in a mammalian cell. Each of the two nucleotide sequences encoding the product of interest is positioned to be incorporated into the rAAV genome replicated in an insect cell.

[0053] Products of interest for expression in mammalian cells may be therapeutic gene products. Therapeutic gene products may be polypeptides, RNA molecules (siRNAs), or other gene products that, when expressed in target cells, provide desired therapeutic effects, such as the removal of undesirable activity, the removal of infected cells, or the complementation of gene defects that cause a deficiency in enzyme activity. Examples of therapeutic polypeptide gene products include CFTR, factor IX, lipoprotein lipase (LPL, preferably LPL S447X; see International Publication No. 01 / 00220), apolipoprotein A1, uridine diphosphate glucuronosyltransferase (UGT), retinitis pigmentosa GTPase regulator interacting protein (RP-GRIP), cytokines or interleukins such as IL-10, dystrophin, PBGD, NaGLU, Treg167, Treg289, EPO, IGF, IFN, GDNF, FOXP3, factor VIII, VEGF, AGXT, and insulin. Alternatively, or in addition as a second gene product, the second nucleotide sequence defined herein may include a nucleotide sequence encoding a polypeptide that acts as a marker protein for assessing cell transformation and expression. Suitable marker proteins for this purpose include, for example, the fluorescent protein GFP, and selectable marker genes HSV thymidine kinase (for selection in HAT medium), bacterial hygromycin B phosphotransferase (for selection against hygromycin B), Tn5 aminoglycoside phosphotransferase (for selection against G418), and dihydrofolate reductase (DHFR) (for selection against methotrexate), CD20, and low-affinity nerve growth factor genes. Sources for obtaining these marker genes and methods for using them are provided in Sambrook and Russell (2001) "Molecular Cloning: A Laboratory Manual" (Part 3), Cold Spring Harbor Laboratory, Cold Spring Harbor Laboratory Press, New York.Furthermore, the second nucleotide sequence defined above in this specification may, if deemed necessary, include a nucleotide sequence encoding a polypeptide that can act as a fail-safe mechanism to heal a subject from cells transduced with the rAAV of the present invention. Such nucleotide sequences, often referred to as suicide genes, encode proteins that can convert a prodrug into a toxic substance capable of killing transgenic cells expressing the protein. Suitable examples of such suicide genes include, for example, the Escherichia coli (E. coli) cytosine deaminase gene, or one of the thymidine kinase genes derived from herpes simplex virus, cytomegalovirus, and varicella-zoster virus, in which case ganciclovir can be used as a prodrug to kill transgenic cells in a subject (e.g., Clair et al., 1987, Antimicrob. Agents Chemother.). 31 (See pages 844-849).

[0054] In another embodiment, the gene product of interest may be an AAV protein, particularly a Rep protein such as Rep78 or Rep68, or a functional fragment thereof. The nucleotide sequences encoding Rep78 and / or Rep68, when present in the rAAV genome of the present invention and expressed in mammalian cells transduced with the rAAV of the present invention, enable the integration of rAAV into the genome of the transduced mammalian cells. Expression of Rep78 and / or Rep68 in mammalian cells transduced or infected with rAAV may provide advantages to certain uses of rAAV by enabling the long-term or persistent expression of any other gene product of interest introduced into the cell by rAAV.

[0055] In the rAAV vector of the present invention, it is preferable that at least one nucleotide sequence (or more) encoding a gene product of interest for expression in mammalian cells is operatively ligated to at least one mammalian cell-compatible expression regulatory sequence, such as a promoter. Many such promoters are known in the art (see Sambrook and Russell, 2001, cited above). Constitutive promoters that are widely expressed in many cell types, such as the CMV promoter, may be used. However, promoters that are inducible, tissue-specific, cell-type-specific, or cell cycle-specific would be more preferable. For example, for liver-specific expression, the promoter may be selected from the α1-antitrypsin promoter, thyroid hormone-binding globulin promoter, albumin promoter, LPS (thyroxine-binding globin) promoter, HCR-ApoCII hybrid promoter, HCR-hAAT hybrid promoter and apolipoprotein E promoter, LP1, HLP, minimal TTR promoter, FVIII promoter, hyperon enhancer, and ealb-hAAT. Other examples include the E2F promoter for tumor-selective expression, particularly for neuronal cell tumors (Parr et al., 1997, Nat. Med. 3:1145~9), or the IL-2 promoter for use in mononuclear blood cells (Hagenbaugh et al., 1997, J Exp Med; 185:2101~10).

[0056] AAV can infect many mammalian cells; see, for example, Tratschin et al., Mol. Cell Biol., 5(11):3251~3260 (1985), and Grimm et al., Hum. Gene Ther., 10(15):2445~2450 (1999). However, AAV transduction of human synovial fibroblasts is significantly more efficient than in similar mouse cells; see, Jennings et al., Arthritis Res, 3:1 (2001). The cellular tropism of AAV differs depending on the serotype; see, for example, Davidson et al., Proc. Natl. Acad. Sci. USA, 97(7):3428~3432 (2000) (which discusses the differences between AAV2, AAV4, and AAV5 in terms of targeting and transduction efficiency of mammalian CNS cells).

[0057] As stated, the AAV sequences that can be used in the present invention for AAV production in insect cells may be derived from the genome of any AAV serotype. Generally, AAV serotypes have genomic sequences that are quite homologous at the amino acid and nucleic acid levels, provide the same set of gene functions, produce virions that are essentially physically and functionally equivalent, and replicate and associate by virtually the same mechanisms. For an overview of the genomic sequences and genomic similarities of various AAV serotypes, see, for example, GenBank accession numbers U89790, J01901, AF043303, AF085716, Chlorini et al. (1997, J.Vir.71:6823~33), Srivastava et al. (1983, J.Vir.45:555~64), Chlorini et al. (1999, J.Vir.73:1309~1319), Rutledge et al. (1998, J.Vir.72:309~319), and Wu et al. (2000, J.Vir.74:8635~47). Human or adeno-associated virus (AAV) serotypes are preferred sources of AAV nucleotide sequences for use in the context of the present invention, with AAV serotypes that typically infect humans (e.g., serotypes 1, 2, 3A, 3B, 4, 5, 6, 7, 8, 9, 10, 11, 12, and 13) or primates (e.g., serotypes 1 and 4) being more preferred.

[0058] For use in the context of the present invention, the AAV ITR sequences are preferably derived from AAV1, AAV2, AAV5, and / or AAV4. Similarly, the Rep52, Rep40, Rep78, and / or Rep68 coding sequences are preferably derived from AAV1, AAV2, and / or AAV4. The sequences encoding the VP1, VP2, and VP3 capsid proteins for use in the context of the present invention may be selected from any of the 42 known serotypes, more preferably from AAV1, AAV2, AAV3, AAV4, AAV5, AAV6, AAV7, AAV8, or AAV9, or from newly developed AAV-like particles obtained, for example, by capsid shuffling techniques and AAV capsid libraries. In a preferred embodiment, the sequences encoding the VP1, VP2, and VP3 capsid proteins are derived from AAV5 or AAV8, and more preferably from AAV5.

[0059] AAV Rep and ITR sequences are particularly conserved among most serotypes. Rep78 proteins from various AAV serotypes are, for example, more than 89% identical, and overall nucleotide sequence identity at the genomic level among AAV2, AAV3A, AAV3B, and AAV6 is around 82% (Bantel-Schaal et al., 1999, J. Virol., 73(2):939~947). Furthermore, it is known that Rep sequences and ITRs from many AAV serotypes efficiently cross-complement (i.e., functionally substitute) corresponding sequences from other serotypes in the production of AAV particles in mammalian cells. U.S. Patent Application Publication 2003148506 reports that AAV Rep and ITR sequences efficiently cross-complement other AAV Rep and ITR sequences in insect cells as well.

[0060] The AAV VP protein is known to determine the cellular tropism of AAV virions. The VP protein coding sequence is considerably less conserved among various AAV serotypes than the Rep protein and gene. The ability of the Rep and ITR sequences to cross-complement corresponding sequences of other serotypes allows for the production of pseudotype AAV particles containing the capsid protein of one serotype (e.g., AAV3) and the Rep and / or ITR sequences of another AAV serotype (e.g., AAV2). Such pseudotype AAV particles are part of the present invention.

[0061] As stated, modified “AAV” sequences may also be used in the context of the present invention, for example, in the production of rAAV vectors in insect cells. Such modified sequences include, for example, sequences having at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, or more than that proportion of nucleotide and / or amino acid sequence identity with AAV1, AAV2, AAV3, AAV4, AAV5, AAV6, AAV7, AAV8, or AAV9 ITR, Rep, or VP (e.g., sequences having about 75–99% nucleotide sequence identity), and may be used in place of wild-type AAV ITR, Rep, or VP sequences.

[0062] Although similar in many respects to other AAV serotypes, AAV5 differs from other known human and monkey AAV serotypes more than other known human and monkey AAV serotypes. Considering this, the production of AAV5 may differ from the production of other serotypes in insect cells. When producing rAAV5 using the method of the present invention, one or more vectors preferably contain a nucleotide sequence comprising AAV5 ITRs, where one nucleotide sequence comprises AAV5 Rep52 and / or Rep40 coding sequences, and another nucleotide sequence comprises AAV5 Rep78 and / or Rep68 coding sequences. Such ITR and Rep sequences can be modified as desired to obtain efficient production of rAAV5 or pseudotyped rAAV5 vectors in insect cells. For example, the start codon of the Rep sequence can be modified.

[0063] In a preferred embodiment, the first nucleotide sequence, the second nucleotide sequence, the third nucleotide sequence, and optionally the fourth nucleotide sequence are stably incorporated into the genome of an insect cell.

[0064] A preferred AAV according to the present invention is a virion whose genome contains at least one nucleotide sequence encoding the gene product of interest, preferably not a native AAV nucleotide sequence, and the AAV virion comprises a VP1 capsid protein having methionine at amino acid position 1 and valine at position 2. For example, an AAV virion obtainable from the insect cells defined above is more preferred in the method defined below herein.

[0065] The advantage of the AAV virions of the present invention is their enhanced infectivity. Without being constrained by any theory, infectivity appears to increase with increasing amounts of VP1 protein in the capsid in relation to the amounts of VP2 and / or VP3 in the capsid, in combination with valine at position 2 of VP1. The infectivity of AAV virions is understood herein to mean the efficiency of transduction of the transgene contained in the virion, as can be inferred from the transgene expression rate and the amount or activity of the product expressed from the transgene.

[0066] The AAV virion of the present invention preferably comprises a gene product of interest that encodes a polypeptide gene product selected from the group consisting of CFTR, factor IX, lipoprotein lipase (LPL, preferably LPL S447X; see International Publication No. 01 / 00220), apolipoprotein A1, uridine diphosphate glucuronosyltransferase (UGT), retinitis pigmentosa GTPase regulator interacting protein (RP-GRIP), cytokines or interleukins such as IL-10, dystrophin, PBGD, NaGLU, Treg167, Treg289, EPO, IGF, IFN, GDNF, FOXP3, factor VIII, VEGF, AGXT, and insulin. The gene product of interest more preferably encodes factor IX or factor VIII protein.

[0067] In another embodiment, the present invention thus relates to a method for producing AAV in insect cells. The method preferably comprises the steps of (a) culturing insect cells as defined above herein under conditions such that AAV is produced; and optionally (b) recovering the AAV. Growth conditions for insect cells under culture, and the production of heterologous products in insect cells under culture, are well known in the art and are described, for example, in references cited for molecular manipulation of insect cells.

[0068] The method preferably further includes a step of affinity purification of AAV using an anti-AAV antibody, preferably an immobilized antibody. The anti-AAV antibody is preferably a monoclonal antibody. Particularly suitable antibodies are single-chain camelid antibodies or fragments thereof, obtainable, for example, from camels or llamas (e.g., Muyldermans, 2001, Biotechnol). 74(See pp. 277-302). The antibody for AAV affinity purification is preferably an antibody that specifically binds to an epitope of the AAV capsid protein, and the epitope is preferably an epitope present on the capsid proteins of one or more types of AAV serotypes. For example, the antibody may be produced or selected based on specific binding to the AAV2 capsid, but at the same time, the antibody may also specifically bind to the AAV1, AAV3, and AAV5 capsids.

[0069] In another embodiment of the present invention, a method for providing a nucleic acid construct encoding a parvovirus capsid protein, wherein the nucleic acid construct has one or more improved properties, and the method is a) A step of providing a plurality of nucleic acid constructs, each construct comprising a nucleotide sequence encoding a parvovirus capsid protein operatively linked to an expression regulatory sequence, and at least one parvovirus inverted terminal repeat (ITR) sequence adjacent to the nucleotide sequence encoding the parvovirus capsid protein operatively linked to the expression regulatory sequence; b) A step of transferring the plurality of nucleic acid constructs into insect cells capable of expressing parvovirus Rep protein; c) Exposing the insect cells to conditions that enable the expression of parvovirus capsid protein and parvovirus rep protein, thereby allowing the nucleic acid construct to be packaged in a parvovirus capsid to provide parvovirus virions; d) A step of recovering parvovirus virions from the insect cells and / or insect cell supernatant; e) A step of bringing the parvovirus virion into contact with the target cell to enable infection of the target cell; f) Step of recovering or identifying the nucleic acid construct from the target cells. A method is provided that includes this.

[0070] As shown in the Examples section and described above, this method is particularly useful for first selecting nucleic acid constructs that are highly functional in insect cells, in the sense that the construct can produce a capsid containing a considerable amount of vector genome, and it is also possible to create a construct contained in a capsid, which is highly effective in transferring the construct DNA into target cells and subsequently expressing the construct DNA.

[0071] With respect to multiple nucleic acid constructs, it is understood that this means constructs that vary with respect to the expression regulatory sequence and / or the nucleic acid sequence encoding the amino acid sequence of the capsid protein and / or the amino acid sequence of the capsid protein and / or the ITR sequence(s). Therefore, any modification in nucleic acid constructs may be attempted. With respect to any improvement in properties, the properties may be in relation to a reference sequence, e.g., a wild-type sequence or a prior art nucleic acid construct for the production of AAV capsids in insect cells. Any properties that may require improvement may be attempted, relating to sequences that may vary in multiple nucleic acid constructs. Such properties may include, but are not limited to, improved potency, increased yield, and improved target cell selectivity.

[0072] Creating molecular diversity or inducing mutagenesis is the first step in the method of the present invention. Multiple nucleic acids encode mutant sequences (i.e., a library of mutant nucleic acids) by introducing random point mutations into a reference sequence to be improved, for example, by error-prone (EP) PCR. As stated, random mutations may be contained in non-coding sequences and / or coding sequences. The frequency of mutations that can be introduced can be varied by varying the amount of template and PCR cycles, and the mutagenic primers used. When the reference is made to multiple sequences, this reference is understood to involve 100 or more, preferably 1,000 or more, 10,000 or more, 100,000 or more, or 1,000,000 or more different sequences, depending on the mutations to be introduced into the multiple nucleic acid constructs. The terms “library” and “multiple” are understood to have the same meaning herein, for example, in the sense that the terms refer to a number of different sequences that may be related, i.e., that may have substantial sequence identity. Each member of a library, i.e., each distinct sequence, may be represented more than once in the library. For example, if a library contains 1,000 unique sequences, the library may contain a total of 1,000,000 sequences. This means that, on average, 1,000 copies of each library member exist in the library.

[0073] Mutagenesis may be carried out in any manner known to those skilled in the art. For example, such mutagenesis may be random, or it may be directed (i.e., targeting a specific sequence / structure within a nucleic acid construct). Random mutagenesis can be performed to achieve a low mutation rate and provide sequences encoding a Cap protein having, for example, 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 or more amino acid changes (compared to the starting sequence on which mutagenesis is performed).

[0074] Techniques that can be used to perform random mutagenesis include E. coli XL1red, UV irradiation, chemical methods (e.g., deamination, alkylation, or base analog mutagens), or PCR methods (e.g., DNA shuffling, site-directed random mutagenesis, or error-prone PCR).

[0075] Error-prone PCR is a modification of a standard PCR protocol designed to alter and enhance the innate error rate of a polymerase. Taq polymerase can be used because of its naturally high error rate, which has an error bias towards AT-to-GC conversion. However, it is also possible to use alternative forms of polymerase where the bias allows for an increase in mutation type deformation (i.e., more GC-to-AT conversion).

[0076] Error-prone PCR reactions typically contain higher concentrations of MgCl2 compared to basic PCR reactions to stabilize non-complementary pairs. Adding MgCl2 can also increase the error rate. Other means of modifying the mutation rate include varying the nucleotide ratios in the reaction or including nucleotide analogs such as 8-oxo-GTP or dITP. The mutation rate can also be modified by changing the number of effective doublings by increasing / decreasing the number of cycles, or by changing the initial template concentration.

[0077] In any case, regardless of the method by which mutations are introduced, the resulting multiple sequences are then cloned into nucleic acid constructs to obtain multiple nucleic acid constructs. Each nucleic acid construct contains one or more parvoviruses or AAV ITRs (typically flanked by two AAV ITRs) adjacent to a nucleotide sequence encoding a parvovirus capsid protein, operatively linked to an expression regulatory sequence. The nucleic acid constructs may also optionally contain reporter gene expression cassettes, such as a green fluorescent protein (GFP) expression cassette, between the ITRs, under the control of promoters such as CMV and baculovirus p10 promoters. The multiple constructs can then be introduced into a target vector, such as a baculovirus vector, to obtain a library of baculoviruses. This introduction can be easily achieved using common biomolecular techniques such as homologous recombination, and also by using commercially available systems such as Bac-to-Bac. Each baculovirus in the library contains a single nucleic acid construct, which has the intended sequence modification. The complexity of the library is preferably maintained when the baculovirus library is created (i.e., the amount of unique sequences in the baculovirus library remains approximately the same as in the nucleic acid library). Therefore, the nucleic acid construct defined in step a) of the above method is preferably contained in the baculovirus vector.

[0078] Subsequently, multiple constructs are transferred into insect cells. It is preferable to use multiple baculoviruses. This is because using baculoviruses allows for better control of the multiplicity of infection. Therefore, when using a baculovirus library, it is preferable to keep the multiplicity of infection below 1, below 0.5, and more preferably below 0.1. For example, using a moi of 0.5, the majority of insect cells are thought to have a single baculovirus per cell; however, a significant portion of these cells have two baculoviruses from the library per cell, and most cells are not infected. The number of baculoviruses per cell is influenced by the Poisson distribution. By lowering the moi, the number of cells with more than one baculovirus is further reduced. However, knowing the multiplicity of infection is not necessary according to the present invention. For example, as shown in the Examples section, dilution series of multiple baculoviruses (dilution serious) can also be used, and a dilution can be selected to provide optimal AAV vector library production (e.g., the highest titer and / or the least cross-packaging).

[0079] Insect cells provided with multiple constructs may also express the parvovirus Rep protein. For example, the Rep expression construct may be introduced into the cells using an additional baculovirus containing the Rep expression construct. It is preferable to use a relatively high degree of infection multiplicity so that Rep is not a limiting factor, i.e., so that when a cell is provided with one of the multiple constructs, it is highly likely that the cell will also have the Rep expression construct. Alternatively, a stable cell line containing the Rep expression construct can be used, which can constitutively express the Rep protein or inductively express Rep when one of the multiple constructs is introduced into the cell. In either case, insect cells capable of expressing the parvovirus Rep protein and provided with one (or more) of the multiple constructs according to the present invention are then subjected to conditions that enable the expression of the parvovirus capsid protein and the parvovirus rep protein, thereby allowing the nucleic acid construct to be packaged in the parvovirus capsid to provide a parvovirus virion. For example, when a baculovirus system is used, this process largely involves culturing the cells for a period of time. When a baculovirus vector system is used, it is preferable to select conditions that do not allow for the spread of baculovirus to the extent that many, if not many, cells contain some members of the construct library. It is preferable to select conditions such that the majority of cells containing constructs from the library contain a single construct and produce only the parvovirus capsid encoded by that construct, which also contains the single construct. If conditions are selected in which more than one type of construct will be contained in the insect cells, one of the constructs will produce infectious or potent AAV, and the less potent or non-infectious constructs will be cross-packaged, and this cross-packaging makes it difficult to determine which of all the packaged constructs can produce potent AAV. In other words, having low cross-packaging allows for a more stringent and effective selection.

[0080] Next, parvovirus virions are recovered from insect cells and / or insect cell supernatant. Numerous methods are available for the recovery of parvovirus virions, including those described in the Examples section. Conventional methods such as density (step) gradient centrifugation (iodixanol, CsCl) and / or tangential flow filtration may also be used. Such conventional methods may be useful, for example, when a modification that may be effective against affinity chromatography is introduced into the capsid sequence. Nevertheless, depending on which construct can be selected, it may also be intended to include a specific affinity chromatography step as one of the features. Thus, improvement of the specific affinity chromatography feature may also be one of the features that can be intended to be improved. Nevertheless, efficient production in insect cells, and infectivity or potency, remain features that need to remain and / or can be improved.

[0081] In another embodiment, a parvovirus virion library produced by the method described above is provided. In a further embodiment, a parvovirus library comprising diverse parvovirus vectors is provided, the parvovirus vector library comprising parvovirus vector capsids, each parvovirus capsid comprising a parvovirus vector genome comprising an expression cassette for the expression of a parvovirus capsid protein. A parvovirus vector library comprising diverse parvovirus vectors preferably comprises parvovirus vector capsids, each parvovirus capsid comprising a parvovirus vector genome comprising an expression cassette for the expression of a parvovirus capsid protein in insect cells. A parvovirus vector library comprising diverse parvovirus vectors preferably comprises parvovirus vector capsids, substantially each parvovirus capsid comprising a parvovirus vector genome comprising an expression cassette for the expression of a parvovirus capsid protein in insect cells, and the parvovirus vector genome is more preferably capsidized. Alternatively, as stated, the vector genome does not necessarily have to contain an expression cassette, but may also contain a sequence identifier that can identify the parvovirus amino acid sequence (and / or expression cassette encoding the parvovirus amino acid sequence) to which the vector genome is capsidized (see Figure 6). In other words, the library substantially contains a parvovirus capsid that can contain any sequence within the capsidized vector genome, insofar as the sequence contained in the vector genome can identify the corresponding parvovirus capsid (i.e., the amino acid sequence of the parvovirus capsid) to which the vector genome is contained.

[0082] The specific identifier sequences that can be intended (see Figure 6B) are preferably at least 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 nucleotides in length. The specific identifier sequences that can be intended may be at most 50, 60, 70, 80, or 90 nucleotides in length. Using identifier sequences of at least 15 nucleotides allows for approximately 10e9 unique possible combinations. Having longer sequence identifiers may allow for greater redundancy and more reliable identification. The sequence identifier may be a priori conjugated to a specific capsid sequence. Thus, in such a scenario, if the sequence identifier is sequenced or detected, the corresponding capsid expression cassette can be identified by referring to the table. Alternatively, the sequence identifier may be used by capturing and / or sequencing a vector vehicle genome, such as a baculovirus genome, so that the capsid expression cassette sequence or a portion thereof associated with the sequence identifier can be determined. Such analysis and / or sequencing may be performed later. Such means and methods for sequencing using high-throughput techniques to identify sequences from complex libraries are well known in the art.

[0083] Libraries produced in accordance with the present invention as described above may be provided as crude lysates or purified products. In particular, it is preferable that such libraries be produced from a viral vector containing a vector genome and encoding a parvovirus capsid protein. Preferred vectors used to produce libraries may be baculovirus vectors containing an expression cassette for a parvovirus capsid protein that is active in insect cells. Alternatively, any suitable alternative viral vector library can be readily conceived and cell lines can be planned, and systems based on adenoviruses, HSVs, lentiviral vectors, etc., may be used instead of baculoviruses, and the expression cassette for the capsid protein may be suitable for expression in mammalian cells such as HeLa cells, 293 cells, CHO cells, A549, 293T, COS, etc. (or may be selected for expression). Such alternative vector vehicles and suitable cell lines that can be conceived are well known in the art and are described, for example, in Gene and Cell Therapy—Therapeutic Mechanisms and Strategies, Vol. 4, edited by Nancy Smith Templeton, 2015, CRC Press. Thus, in alternative embodiments, the means and methods described herein can be readily used with mammalian cells combined with a suitable mammalian viral vehicle instead of using baculovirus vectors and insect cells. In any case, parvovirus vector libraries provided according to the present invention, such as AAV vector libraries, are produced using vector vehicles that allow control of the copy number per producing cell, and the quality of the vector library is significantly improved compared to plasmid-produced libraries that do not allow such control.

[0084] Next, the parvovirus virions are recovered, or to put it another way, a parvovirus vector library is provided, which may be a crude lysate or a purified product, and the parvovirus virions are then brought into contact with selected target cells to infect the target cells with parvovirus. Appropriate target cells can be selected, which may be suitable target cells such as liver cells, kidney cells, or neurons, where gene therapies are being developed. Appropriate target cells may be any cell line, such as HeLa cells, HEK293 cells, or HuH cells, or they may be primary cells. Contact may include delivery to a suitable animal model, such as rats, mice, or monkeys, and may also include various delivery routes, such as intravenous or intramuscular injection, and the subsequent target cells may even be selected candidate organs in such an animal model. In any case, any cell type can be selected, and the parvovirus virions can be brought into contact with any cell type by any means, i.e., in vivo or in vitro, to enable infection, i.e., the transfer of the nucleic acid construct contained in the capsid virions into the cells. It is understood that cells can also be co-infected with adenoviruses that support the transduction process, for example, by inducing transduction. Co-infection can be beneficial when we want to select cells in which, for example, a reporter gene construct is contained within a nucleic acid construct, enabling not only efficient DNA transfer but also efficient intracellular trafficking to deliver the nucleic acid construct to the nucleus (see Figure 6). If the capsid sequence is mutated and / or the stoichiometry of VP1, VP2, and VP3 is altered without theoretical constraints, co-infection may lead to interference with internal trafficking. For example, a capsid lacking VP1 can infect cells but will not enter the nucleus. Capsids containing nucleic acid constructs subsequently remain in endosomes (than) and are targeted for proteasome protein degradation. Therefore, it can be interesting to include a selection step based on the purpose of the selection process, i.e., to achieve efficient delivery of the nucleic acid construct to the nucleus and enable expression from the nucleic acid construct.The selection step may be mediated, for example, through a reporter gene or any other gene of interest. The selection step may also be in HeLa RC32 cells or similar cells, in which virions are amplified to achieve efficient delivery of the vector genome.

[0085] Finally, if the cells are used to infect target cells, preferably enabling efficient transduction, the nucleic acid construct is recovered from the target cells. The nucleic acid construct may be recovered from the entire cell population. The nucleic acid construct may also be recovered from a subset of the cell population, for example, a subset that exhibits reporter transgene expression and is therefore effective in transducing target cells. The nucleic acid construct may be recovered from the entire cell population, but especially from nuclei derived from all cells. In this way, nucleic acid constructs (and, incidentally, capsids similarly encoded by the nucleic acid constructs) that are expected to be excellent in transducing target cells can be selected. The recovered nucleic acid constructs can then be subjected to sequencing to identify them. As stated, the nucleic acid constructs may contain identifier sequences that identify the constructs. For example, if a baculovirus vector system and insect cells are used to create a parvovirus vector library, and the parvovirus vector genome contains an expression cassette for the parvovirus capsid in which it is contained, it can be understood that the expression cassette or a part thereof can be considered an identifier sequence. When an expression cassette is introduced into mammalian cells, it cannot produce an AAV capsid unless the expression cassette has an insect cell promoter that is active in mammalian cells. In particular, a portion of the modified nucleic acid construct (or the corresponding identifier sequence) can be subjected to sequencing, for example, after a PCR reaction in which a small portion is amplified in a short time. The entire nucleic acid construct or the entire capsid coding sequence can also be sequenced. Sequencing is understood to include high-throughput sequencing or any other suitable sequencing method known in the art.

[0086] Of particular interest may be the identification of improved sequences. If conditions are chosen such that improved sequences are very restrictive, then all recovered nucleic acid constructs and their sequences are improved nucleic acid constructs. Therefore, the recovery of nucleic acid constructs includes the selection of improved nucleic acid constructs. Nevertheless, improved sequences derived from recovered nucleic acid constructs may be confirmed or identified by comparing the population of recovered sequences with, for example, the population of sequences contained in the initially constructed library. Recovered sequences that are very dominant in the recovered population when compared with the initial population indicate that they are the desired improved nucleic acid constructs. Therefore, in addition to the recovery of nucleic acid constructs, an additional step may include the identification of library-derived nucleic acid constructs that correspond to improved nucleic acid constructs. Such identification may include comparison with, for example, the initial library, a baculovirus library containing nucleic acid constructs, or a population of nucleic acid constructs contained in parvovirus capsids.

[0087] Once a nucleic acid construct with improved properties has been selected for its characteristics and is provided or identified, the next step is step g) to create a nucleic acid construct for the production of a gene therapy vector, which contains a nucleotide sequence encoding a parvovirus capsid protein, operationally linked to an expression regulatory sequence recovered in step f). The nucleic acid construct for the production of a gene therapy vector does not have an expression construct for the parvovirus capsid protein flanked by a parvovirus ITR sequence. Therefore, the nucleic acid construct for the production of a gene therapy vector preferably contains an expression construct for the parvovirus capsid protein and may optionally contain further parvovirus components, such as a gene therapy construct, i.e., a therapeutic gene flanked by a parvovirus ITR, and / or a Rep expression construct, and all constructs are constructed for compatibility with insect cell production. Therefore, the created nucleic acid construct is preferably contained in a baculovirus vector or insect cells. Since AAV virus vectors are suitable candidates for gene therapy, it is particularly preferable that the parvovirus capsid protein, parvovirus Rep protein, and / or ITR nucleotide sequence are derived from adeno-associated virus. The recovered nucleic acid construct used to produce the nucleic acid construct for gene therapy vector production is understood to be an actual physical nucleic acid obtained, for example, by excising the sequence of interest from the recovered nucleic acid construct. Alternatively, the sequence of interest, such as a parvovirus capsid expression cassette or a portion thereof, may be amplified by a PCR reaction and then used. The sequence of interest may also be determined and produced by a de novo, for example, a DNA synthesizer.

[0088] The entire selection process is for identifying improved constructs for the production of gene therapy vectors based on insect cells for use in medical treatment, and in further embodiments, a method for producing parvovirus vectors comprising steps a) to g) described above, wherein insect cells are... Nucleic acid constructs created for the production of gene therapy vectors; Nucleic acid constructs containing nucleotide sequences that include at least one inverted terminal repeat (ITR) nucleotide sequence; and Nucleic acid constructs encoding parvovirus Rep protein that can express parvovirus Rep protein in insect cells. A method is provided comprising: providing insect cells, culturing the insect cells under conditions such that a parvovirus vector is produced; and optionally (b) recovering the produced parvovirus vector. The parvovirus vector is preferably an AAV vector. Therefore, any of the methods described above for the production of AAV vectors having VP1, VP2, and VP3 expression constructs having an out-of-frame start codon before the VP1 ATG codon is also applicable to any identified improved constructs and produced nucleic acid constructs for the production of gene therapy vectors.

[0089] In this document and in the claims herein, the verb “comprise” and its conjugations are used in the non-restrictive sense of “comprise,” meaning that the items following the word are included, but not items not specifically mentioned. In addition, references to elements with the indefinite articles “a” or “an” do not exclude the possibility of more than one element being present, unless the context explicitly requires that there be one or exactly one element. Thus, the indefinite articles “a” or “an” usually mean “at least one.”

[0090] All patents and references cited herein are incorporated herein by reference as a whole.

[0091] The following examples are provided for illustrative purposes only and are not intended to limit the scope of the invention. [Examples]

[0092] 1. Introduction Expression of AAV capsids in the baculovirus expression system (BEVS) requires modification of the expression cassette to promote a single mRNA transcript that yields three viral capsid proteins in appropriate ratios. A study by Urabe et al. (2002; cited above) demonstrated that adapting the start codon, combined with the removal of an intron splicing site, resulted in the expression of all three VP proteins in insect cells. Further studies showed that CTG and GTG can be used as efficient start codons for AAV production in the BEVS system. Additionally, alanine at the second position, for example, introduced into the AAV5 capsid sequence, resulted in an AAV5 capsid with the native VP1-VP3 capsid protein ratio.

[0093] However, in the rational design process, a limited subset of constructs and combinations is possible due to the labor-intensive research required to produce recombinant baculoviruses. Therefore, we used a library approach designed to construct a set of alternative start codons (17 in total) in combination with randomized context sequences within the AAV5 capsid to determine if there is still room to select improvements in the quality and yield of AAV capsids derived from the BEVS system (see Figure 1 for an overview of the method). The results and methods described below are not limited to AAV5 but can be similarly applied to other serotypes and other parvoviruses and can be similarly used to select improvements in other characteristics of parvovirus gene therapy vectors.

[0094] Materials and methods Structure design and plasmid library The following alternative putative start codons across various eukaryotes and prokaryotes were found in the literature and utilized as possible start codons for AAV5 VP1 production: ATT, ATG, ATA, AGA, AGG, AAA, CTG, CTT, CTC, CTA, CGA, CGC, TTG, TGA, TAA, TAG, and GTG. The constructs had the following contextual design: NNN NNN NNN GNN NNN (SEQ ID NO: 71). NNN When indicating the insertion of any of the above start codons for VP1, on the one hand, N represents A, T, C, or G randomly with an equal distribution. The "G" in the first trimer following the start codon is fixed. The theoretical complexity of this library is 7.1×10 7 (4 11 ×17), i.e., calculated as the maximum number of unique sequences that can be produced. The start codon library was synthesized at GeneArt (ThermoFisher), and the complete sequence with the AAV5 coding capsid sequence and gene expression sequence was cloned into an ITR-containing plasmid such that the produced AAV capsid encapsidated the capsid-coding gene as a transgene within the capsid itself. The plasmid library was generated at GeneArt, and 100 single colonies from the library were subjected to Sanger sequencing to confirm the complexity and diversity within the library.

[0095] Baculovirus library Utilizing the capabilities of the BEVS system and screening new designs for their compatibility with the BEVS system by that capability, the inventors generated a recombinant baculovirus library from the above-supplied plasmid library. The theoretical diversity of the library is 7.1×10 7 . In a standard recombination protocol, the inventors used 1 μg of Bsu36I-digested BacAMT5 baculovirus backbone (7.34×10 9 molecules) together with 1 μg of donor plasmid (8.12×10 10A single plasmid molecule was used. The limiting factor is the baculovirus scaffold, which represents a theoretical library complexity of up to 10³ times in the case of 100% recombination efficiency. When the pooled P0 library is amplified in SF9 cells, the baculovirus is expected to amplify approximately 1000 times, resulting in a P1 library that represents a sufficiently complex library.

[0096] AAV library creation For the creation of the AAV library, SF9 cells were inoculated at a rate of 1 million cells per ml. The MOI (Multiple Infection Intake) was calculated as follows: MOI = 0.7 × Virus Volume × Titer / Cell Density × Cell Volume. The inventors determined that the P1 passage of the baculovirus library was approximately 2 × 10⁻⁶. 11 The titer was determined to be gc / ml. On average, the TCID50 value of baculovirus is estimated to be approximately 2 log values ​​lower than the genome copy titer. 2 × 10 9 This yields an approximate TCID50 value of / ml. 2 × 10 for P1 baculovirus libraries. 9 Using the calculated infectivity titer, the first AAV library (MOI of 0.5) was created. By inoculating 3L of insect cells at a rate of 1 million cells per ml, the inventors obtained a capsid / transgene with an MOI of 0.5. In other words, less than one infectious particle per cell. Since the capsid is also the transgene, the (hence, capsid) cassette is thought to be amplified approximately 1000 times per cell by the replicase. This double infection is also statistically more efficient with respect to the Poisson distribution compared to triple infection. Three further AAV libraries were created using approximate MOIs of 5, 25, and 50. The AAV library created with an MOI of 0.5 was found to perform best in the selection method.

[0097] Purification and quantification of AAV AAV library material was purified from 3 L CLB via a 5 ml AVB Sepharose column (affinity chromatography) in an Akta Explorer. DNA was isolated, and qPCR was performed on each fraction using primers that amplified the AAV vector genome sequence. From these fractions, the inventors pooled them for a modified TCID50 assay in HeLa RC32 cells to apply selective pressure to novel mutants in the library. See below. Three other AAV organisms (MOIs of 5, 25, and 50, respectively) were isolated in a similar manner. DNA was isolated from all isolated AAV libraries for next-generation sequencing (NGS).

[0098] Selection pressure on AAV libraries Using a modified TCID50 assay in HeLa RC32 cells (Tessier J et al., J. Virol. 75(1):375~383, 2001), we selected the AAV variant that exhibited the highest efficacy. HeLa RC32 cells contain AAV2 replicase and capsid genes incorporated into their genome. Upon transduction by AAV, the transgene is amplified by the replicase and packaged into an AAV capsid, which is also produced within HeLa cells. The advantage of this cell line is that, in principle, the replicase acts as an amplifier for any AAV DNA that enters the nucleus. By performing a limiting dilution series of AAV and infecting HeLa cells with it, we were able to selectively amplify only those AAVs that manage to successfully reach the nucleus. In other words, we selected AAV capsids and constructs containing / encoding VP1:VP2:VP3 in favorable ratios. The dilution series used to transduce HeLa cells were 6400 gc / cell, 3200 gc / cell, 1600 gc / cell, 800 gc / cell, 400 gc / cell, and 200 gc / cell.

[0099] Isolation of AAV DNA Two days after transduction, HeLa cells were lysed and subjected to DNA isolation to recover the AAV vector genome, in which the vector genome that reached the nucleus was amplified in the HeLa cells. Before submission to next-generation sequencing (NGS), endpoint PCR was performed on the isolated DNA using a universal primer set against a capsid library.

[0100] NGS sequencing of various libraries NGS sequencing was performed on isolated DNA from plasmid libraries, P1 passages from baculovirus libraries, products from AAV libraries, and DNA isolated from pooled dilutions for each AAV library transduction. The prepared DNA for each sequencing reaction was sent to BaseClear for amplification and barcoding.

[0101] result AAV libraries were created by 0.5 MOI infection. After AAV library production, the libraries were used to infect HeLaRC32 cells. Plasmid libraries, baculovirus libraries, AAV libraries, and infected HeLaRC32 cells were processed and analyzed for next-generation gene sequencing to determine their complexity. Unique sequences were identified at each step, the copy number of each unique sequence was determined, the total number of sequences was determined, and the relative percentage to each start codon was determined and plotted (see Figures 2A-E). The resulting baculovirus library represented approximately 74% of the complexity of the plasmid library. There were no significant observations regarding the prevalence of start codons between the plasmid and baculovirus libraries (see Figures 2A and 2B), which is presumably due to the lack of selective pressure applied to the start codons. However, when ATG was used as the start codon, this sequence was found least frequently in the AAV capsid (see, for example, Figure 2C). Here, ATG represents less than 0.5% of the total library. This low percentage was expected, as strong start codons for VP1 mostly produce VP1 protein with little to no VP2 and VP3 protein production, of which VP3 is generally essential for capsid production. Regarding the remainder, there were no significant observations made regarding the percentage of codon usage among plasmid libraries, baculovirus libraries, and AAV capsid libraries, as all observations were within the normal range of variation (ranging from approximately 4-5% to approximately 8-9%). Finally, the AAV library generally represented approximately 96% of the complexity of the baculovirus library, suggesting a comprehensive shift in complexity in AAV production from baculovirus libraries. Finally, when HeLa RC32 cells were infected with the AAV library in limiting dilution series, we found that CTG and GTG are the two most abundant start codons for the production of potent AAV virus capsid particles in the baculovirus expression system.CTG and GTG together constituted nearly 50% of the entire sequence that successfully transduced and infected cells (i.e., transferred the vector genomic DNA into the nucleus, enabling amplification by HeLaRC32 cells). It was notable (though not shown) that only the codon immediately following the start codon was limited to G, while codons following the start codon predominantly encoded alanine, supporting the possibility that the Ala-encoding trimer may have a preference as a second codon for VP1 expression in insect cells, due to its amino acid sequence and / or DNA / RNA sequence. The sequences recovered from the cells suggest a recovery of only 5% of the AAV library complexity, which is remarkable. This low recovery indicates significant selective pressure.

[0102] It is interesting that ATG as a start codon is the third most frequently expressed start codon in isolated DNA from Hela RC32 cells, representing approximately 8% of the complete library. This proportion contrasts sharply with its representation in the AAV library, which is only 0.5%. The top 30 sequences with VP1 start codons are listed in Table 1 below, with the most prevalent (SEQ ID NO: 1), listed at the top, accounting for the vast majority of the population. Each sequence itself, when used as a substitution sequence in the context of a VP1 start codon, enables the efficient production of AAV capsids. While each sequence itself may possess certain unique properties that enable the efficient production of AAV capsids, additional fundamental characteristics can be identified from the sequences listed below, and these sequences may describe some general principles that influence the efficient production of potent AAV from ATG start codons (see, in particular, Figure 5). The basic characteristics may include, but are not limited to, an out-of-frame start codon preceding the VP1 start codon and / or a GT sequence immediately following the ATG codon, preferably resulting in valine at position 2 of the VP1 capsid. For the majority of the 30 clones, upstream out-of-frame start codons (ATG, CTG, ACG, TTG, and GTG) that could act as translation initiation sites were observed. Such out-of-frame start codons, when translated, are expected to result in a short peptide with a stop codon following the VP1 start codon. Out-of-frame CTT or CTC non-standard start codons may also be identified. Although CTT and CTC are not considered strong non-standard start codons, the inventors have observed that various capsids specifically containing these two start codons can be isolated from HeLa cells. Without being constrained by theory, this suggests that an out-of-frame start codon preceding the VP1 ATG can act as a decoy translation initiation context for the ribosome, thereby interfering with VP1 translation and enabling the leaky pseudo-ribosome scanning that can be observed with respect to wild-type AAV.More specifically, the synthesis of (short) peptides from these alternative start codons can allow the ribosome to continue scanning the mRNA transcript or to restart scanning. This delayed, leaky initiation can enable the translation of VP2 and VP3 from a single polycistronic mRNA transcript. Furthermore, this delayed, leaky initiation can almost certainly be similar to what happens when CTG, GTG, TTG, and ACG are introduced as non-standard start codons (European Patent No. 1,945,779 B1 granted; U.S. Patent No. 8,163,543 granted; Urabe et al., 2002; cited above), thereby preventing the ribosome from regularly initiating translation with the non-standard VP1 start codon and enabling sufficient initiation of VP2 and VP3 translation from their respective start codons in a single mRNA transcript.

[0103] [Table 1]

[0104] To confirm that the library selection process produced useful novel clones, two representative start codon constructs for ATG, CTG, and GTG, and one representative construct for TAG and TGA, were selected for recombination into stable baculovirus clones (Table 1). The viral capsid subunit ratio and potency were determined using these constructs. Furthermore, the inventors desired to confirm that the construct with the ATG start codon produced high yield and potent AAV.

[0105] [Table 2]

[0106] Unique start codon sequence contexts (the VP1 start codon is underlined) were selected and cloned as replacements in the AAV5 expression construct sequence (sequences 70 and 74; sequence 74 corresponds to nt148-167). Sequence number 31 was a dominant clone selected and identified from the MOI5 library. Several clones were generated for each candidate, and VP capsid expression was analyzed (Figure 3). Start codons with relative contexts had varying degrees of success in generating AAV capsids with good stoichiometry. Note that in most cases, three clones were tested for each construct to determine the stability of the baculovirus clone. In this regard, ATG1 had one stable producer (the second lane for ATG1 in Figure 3). For ATG2, there were an abundance of stable producers, all with good stoichiometry. The CTG1 construct could not be produced, while CTG2 produced a capsid stoichiometry similar to that described in International Publication No. 2015 / 137802 of the international patent application (data not shown). Similarly, GTG2 also exhibited good stoichiometry, but TAG (stop codon) produced only a very small amount, and TGA (stop codon) resulted in the production of a capsid without VP1. Therefore, it is surprising that the inventors were able to confirm that they could create an efficient AAV capsid construct, namely AAV5, in which ATG is used as a start codon that exhibits good stoichiometry.

[0107] Stable clones were selected for each start codon construct, and AAVs containing the SEAP reporter gene were produced using these stable clones under the control of the CMV promoter. All AAV constructs yielded titers (gc / ml) in a similar range. After titration, the inventors transduced both Huh7 and HeLa cells at three different MOIs, and SEAP activity was determined after 48 hours (Figures 5A and 5B). The two constructs with the ATG start codon produced capsids with similar or slightly improved potency compared to CTG and GTG, while the capsid lacking VP1(TGA) did not exhibit discernible SEAP activity above the expected background. Supporting evidence that valine may enhance potency is provided by the fact that the dominant intrinsic clone identified in Table 1 encodes valine at position 2. These results are consistent with observations from Figure 3, and these capsids exhibited a VP1:VP2:VP3 stoichiometry very similar to that of the CTG and GTG constructs.

[0108] Further embodiments are as follows: [Embodiment 1] A nucleic acid construct comprising an expression regulatory sequence for the expression of a nucleotide sequence containing an open reading frame in insect cells, wherein the open reading frame sequence is i) Adeno-associated virus (AAV) capsid proteins VP1, VP2, and VP3, and ii) ATG translation start codon for VP1 A nucleic acid construct that codes for a nucleotide sequence including an alternative start codon upstream of the open reading frame, which is misaligned with the open reading frame. [Embodiment 2] The nucleic acid construct according to Embodiment 1, wherein the alternative start codon is selected from the group consisting of CTG, ATG, ACG, TTG, GTG, CTC, and CTT. [Embodiment 3] The nucleic acid construct according to Embodiment 1 or Embodiment 2, wherein the nucleotide sequence includes an alternative open reading frame beginning with the alternative start codon, which includes the ATG translation start codon for the VP1. [Embodiment 4] The nucleic acid construct according to Embodiment 3, wherein the alternative open reading frame following the alternative start codon encodes a peptide of up to 20 amino acids. [Embodiment 5] A nucleic acid construct according to any one of Embodiments 1 to 4, wherein the nucleotide sequence adjacent to the open reading frame and containing the alternative start codon is nucleotide residues 1 to 8 of SEQ ID NO: 1. [Embodiment 6] The nucleic acid construct according to Embodiment 5, wherein the open reading frame containing the ATG translation initiation codon for VP1 has the nucleotide sequence ATGGTAAGCTTT of SEQ ID NO: 1, and the residues at positions 9-11 are the ATG translation initiation codon for VP1. [Embodiment 7] A nucleic acid construct according to any one of embodiments 1 to 6, wherein the second codon of the open reading frame encodes an amino acid residue selected from the group consisting of alanine, glycine, valine, aspartic acid, and glutamic acid. [Embodiment 8] The nucleic acid construct according to any one of Embodiments 1 to 7, wherein the AAV capsid protein is an AAV serotype capsid protein. [Embodiment 9] A nucleic acid construct according to any one of Embodiments 1 to 8, comprising a promoter selected from the group consisting of a polyhedrin promoter, a p10 promoter, a 4×Hsp27 EcRE+ minimum Hsp70 promoter, a delta E1 promoter, and an E1 promoter. [Embodiment 10] A nucleic acid construct according to any one of embodiments 1 to 9, which is a baculovirus vector. [Embodiment 11] An insect cell comprising a nucleic acid construct according to any one of Embodiments 1 to 10. [Embodiment 12] (a) A second nucleotide sequence containing at least one AAV inverted terminal repeat (ITR) nucleotide sequence, (b) A third nucleotide sequence comprising a Rep78 or Rep68 coding sequence operatively ligated to an expression regulatory sequence for expression in insect cells, (c) Optionally, a fourth nucleotide sequence containing a Rep52 or Rep40 coding sequence operatively linked to an expression regulatory sequence for expression in insect cells. The insect cell according to Embodiment 11, further comprising: [Embodiment 13] A method for producing AAV in insect cells, comprising (a) culturing the insect cells described in Embodiment 11 or Embodiment 12 under conditions that produce AAV, and optionally (b) recovering the AAV. [Embodiment 14] A method for providing a nucleic acid construct encoding a parvovirus capsid protein, wherein the nucleic acid construct has one or more improved properties, and the method is a) A step of preparing a plurality of nucleic acid constructs, wherein each construct comprises a nucleotide sequence encoding a parvovirus capsid protein operatively linked to an expression control sequence, and at least one parvovirus inverted end repeat (ITR) sequence adjacent to the nucleotide sequence encoding the parvovirus capsid protein operatively linked to the expression control sequence, b) A step of transferring the plurality of nucleic acid constructs into insect cells capable of expressing parvovirus Rep protein, c) A step in which the insect cells are subjected to conditions that enable the expression of parvovirus capsid protein and parvovirus rep protein, thereby enabling the nucleic acid construct to be packaged in a parvovirus capsid to provide parvovirus virions, d) A step of recovering parvovirus virions from the insect cells and / or insect cell supernatant, e) A step of bringing the parvovirus virion into contact with the target cell to enable infection of the target cell, f) Step of recovering the nucleic acid construct from the target cells. Methods that include... [Embodiment 15] The method according to Embodiment 14, wherein the nucleic acid construct defined in step a) is contained in the baculovirus vector. [Embodiment 16] The method of Embodiment 14 or Embodiment 15, further comprising step g) creating a nucleic acid construct for the production of a gene therapy vector, comprising a nucleotide sequence encoding a parvovirus capsid protein operatively linked to an expression control sequence, which is recovered in step f).

Claims

1. A method for producing an adeno-associated virus (AAV) encoding at least one gene product of interest, The method includes (A) the step of culturing insect cells, and (B) the step of recovering the AAV. The aforementioned insect cells, (a) A nucleic acid construct comprising an expression control sequence for the expression of a nucleotide sequence, The nucleotide sequence is, i) Open reading frames (ORFs) encoding AAV capsid protein VP1, AAV capsid protein VP2, and AAV capsid protein VP3; ii) The ATG translation start codon for the AAV capsid protein VP1, followed immediately by an inserted or mutated second codon encoding an amino acid selected from the group consisting of alanine, glycine, valine, aspartic acid, and glutamic acid; and iii) An alternative ORF located upstream of the ORF and having a reading frame offset from the ORF, which begins up to 8 nucleotides upstream of the ATG translation start codon for the AAV capsid protein VP1 and begins with an alternative start codon selected from the group consisting of CTG, ATG, ACG, TTG, GTG, CTC, and CTT; nucleic acid constructs including, (b) A second nucleotide sequence comprising at least one inverted terminal repeat sequence and encoding the at least one gene product of interest. Methods that include...

2. The method according to claim 1, wherein the alternative ORF encodes a peptide of up to 20 amino acids.

3. The method according to claim 1 or 2, wherein the nucleotide sequence adjacent to the ORF and containing the alternative start codon is nucleotide residues 1 to 8 of SEQ ID NO:

1.

4. The method according to claim 3, wherein the ORF containing the ATG translation initiation codon for VP1 has the nucleotide sequence ATGGTAAGCTT of SEQ ID NO: 1, and the residues at positions 9-11 are the ATG translation initiation codon for VP1.

5. The method according to any one of claims 1 to 4, wherein the AAV capsid protein is the capsid protein of AAV serotype 5.

6. The method according to any one of claims 1 to 5, wherein the nucleic acid construct comprises a promoter selected from the group consisting of a polyhedrin promoter, a p10 promoter, a 4×Hsp27 EcRE+minimumHsp70 promoter, a delta E1 promoter, and an E1 promoter.

7. The method according to any one of claims 1 to 6, wherein the nucleic acid construct is a baculovirus vector.

8. The insect cell is (c) A third nucleotide sequence comprising a Rep78 or Rep68 coding sequence operatively linked to an expression regulatory sequence for expression in insect cells, and (d) A fourth nucleotide sequence comprising a Rep52 or Rep40 coding sequence operatively linked to an expression regulatory sequence for expression in insect cells. The method according to any one of claims 1 to 7, further comprising:

9. The method according to any one of claims 1 to 8, wherein the at least one gene product of interest can be expressed in mammalian cells.

10. The method according to claim 9, wherein the at least one gene product of interest is an RNA molecule or polypeptide capable of RNA interference.

11. The method according to claim 10, wherein the gene product provides a therapeutic effect.

12. The method according to claim 11, wherein the therapeutic effect is a change in the activity of CFTR, factor IX, lipoprotein lipase (LPL), apolipoprotein A1, uridine diphosphate glucuronosyltransferase (UGT), retinitis pigmentosa GTPase regulator interacting protein (RP-GRIP), cytokines, interleukins, dystrophin, PBGD, NaGLU, Treg167, Treg289, EPO, IGF, IFN, GDNF, FOXP3, factor VIII, VEGF, AGXT, or insulin.

Images