RU2588387C2 - Improved methods of purifying vectors based on recombinant aav - Google Patents

Abstract

FIELD: biotechnologies.

SUBSTANCE: invention relates to biotechnology and virology. Described is method of purifying vectors based on recombinant adenoassociated virus (rAAV). Vectors based on recombinant AVV according to invention deprived of technological impurities, including components of product, such as cell nucleic acids, cell proteins, helper virus and components.

EFFECT: method involves bringing stream of initial materials in contact with medium for chromatography on apatite in presence of polyethylene glycol (PEG) and elution of rAAV particles using elution buffer containing less than 3 % (wt/Vol) PEG.

34 cl, 7 dwg, 6 tbl, 13 ex

Description

CROSS REFERENCE TO RELATED APPLICATIONS

This application claims the priority of provisional application for the grant of US patent No. 61/187601, filed June 16, 2009, incorporated herein by reference in full.

FIELD OF THE INVENTION

The present invention generally relates to the field of purification of recombinant adeno-associated virus (rAAV) -based vectors that can be used for gene transfer and, specifically, for gene therapy or vaccination. More specifically, it relates to methods for purifying recombinant rAAV-based vectors that essentially do not contain components associated with the production method, such as cellular nucleic acids, cellular proteins, helper virus, and medium components.

BACKGROUND OF THE INVENTION

Adeno-associated viruses (AAV) are characterized by unique traits that make them attractive as vectors for gene therapy and gene vaccines. Infection of AAV cells in culture does not lead to cytopathology, and the naturally occurring infection of humans and other animals is silent, asymptomatic, and is not involved in the etiology of any human disease. In addition, AAV infects a variety of cell types, including many mammalian cells, which allows targeted action on many different tissues in vivo . AAV infects slowly dividing and non-dividing cells and can persist, in fact, throughout the life of these cells as a transcriptionally active nuclear episoma (extrachromosomal element). Integrated copies of the rAAV vector in organs such as the liver or muscle are extremely rare. Effective long-term gene transfer to a number of cell types has been reported, including eye cells, central nervous system, and muscle. See, for example, X. Xiao et al., J. Virol. 70 (11): 8098-8108 (1996); RR Ali et al., Hum. Mol. Genet. 5 (5): 591-94 (1996). Current clinical studies are mainly focused on the use of vectors based on rAAV serotype 2, but several reports demonstrate that other serotypes of AAV, including rAAV-1, rAAV-4, rAAV-5 and rAAV-8, are characterized by unique bio-distribution in vivo , which makes them attractive viral serotypes from the point of view of testing in clinical trials.

Adeno-associated virus (AAV) is a replication deficient parvovirus whose genome from single-stranded DNA is approximately 4.7 thousand n. length, and includes an inverted terminal repeat (ITR) length of 145 nucleotides. The nucleotide sequence of AAV serotype 2 (AAV2) is presented in Srivastava et al., J. Virol., 45: 555-564 (1983), as amended by Ruffing et al., J. Gen. Virol., 75: 3385-3392 (1994). The cis-acting sequences directing viral DNA replication (rep), encapsulation / packaging and chromosome integration of the host cell are contained within the ITR. The three AAV promoters p5, p19 and p40 (identified by their relative position on the map) drive the expression of two internal open AAV reading frames encoding the rep and cap genes. Two rep promoters (p5 and p19), coupled to differential splicing of a single AAV intron in nucleotides 2107 and 2227, produce four rep proteins (rep78, rep68, rep52 and rep40) from the rep gene. Rep proteins have multiple enzymatic properties that ultimately are responsible for the replication of the viral genome. The cap gene is expressed from the p40 promoter, and it encodes three capsid proteins VP1, VP2, and VP3. Alternative splicing and non-consensus translation start sites are responsible for the production of three related capsid proteins. The only consensus polyadenylation site is located at position 95 of the AAV genome map. AAV life cycle and genetics are reviewed in Muzyczka, Current Topics in Microbiology and Immunology, 158: 97-129 (1992).

AAV particles contain a protein capsid, which includes three capsid proteins VP1, VP2, and VP3, inside which the gene of a linear single-stranded DNA is ~ 4.6 kb. Individual particles serve as a package of only one molecular chain of DNA, but it can be a plus or minus chain. Particles containing any of the chains are infectious, and replication occurs by converting the parent single chain to duplex and subsequent amplification, as a result of which single daughter chains are displaced and packaged in capsids. Duplex or single-stranded AAV genomes (sometimes referred to as “proviral DNA” or “provirus”) can be inserted into bacterial plasmids or phagemids and transfected into adenovirus-infected cells. See Carter, Handbook of Parvoviruses, Vol. I, pp. 169-228 (1989), and Berns, Virology, pp. 1743-1764, Raven Press, (1990) for a general review of AAV.

The production of an rAAV-based vector basically requires four common elements: 1) a permissive host cell for replication; 2) helper virus functions, which may be provided by suitable helper viruses, such as adenovirus or herpes virus, or, alternatively, plasmid constructs containing minimal adenoviral helper functions; 3) trans-packaging design rep-cap; and 4) a suitable environment for the product.

Recombinant AAV particles can be obtained from packaging cell lysates. See, for example, Chirico and Trempe (1998) J. Virol. Methods 76: 31-41. However, the cell lysate contains various cellular components, such as host cell DNA, host cell proteins, medium components and helper virus or helper virus plasmid DNA, which must be separated from the rAAV-based vector before it is suitable for use in vivo . Recent successes of rAAV production include the use of processes with a suspension of non-adhesive cells in mechanically mixed bioreactors and production conditions under which rAAV-based vectors are released into the medium or supernatant, with a decrease in the concentration of host cell components present in the product material, but this medium nevertheless, it contains appreciable amounts of impurities formed during the process. See US patent No. 6566118 and PCT WO 99/11764. Thus, rAAV particles can be collected from the medium and / or cell lysate and further purified.

Methods involving density gradient centrifugation used to clean rAAV-based vectors and, in particular, rAAV-2, are not scalable. Recent reports of rAAV-2-based vectors describe purification methods that use ion exchange chromatography, including reverse ion exchange chromatography (including cationic and anionic chromatography). See, for example, US Patent No. 6,566,118 and PCT WO 99/11764 for methods of using a combination of reverse ion exchange chromatography to purify vectors based on recombinant adeno-associated virus from a culture supernatant and / or cell lysate. Further improvements in the preparation of a concentrated rAAV solution include the use of deoxycholate treatment of the cell lysate, separation in the iodixinol gradient before affinity chromatography, resulting in a high titer of rAAV2 (Clark et al., Hum. Mol. Genet. 10 (6): 1031-39 ( 1999); Zolotukhin et al., Gene Therapy 6 (6): 973-985 (1999)). O'Riordan et al. (O'Riordan et al., J. Gene Med. 2: 444-454 (2000); US Pat. No. 7015026) also report a scalable method for chromatographic purification of vectors based on a recombinant adeno-associated virus and, in particular, as shown in the examples , rAAV-2 based vectors using ion exchange chromatography, hydroxyapatite chromatography, Cellufine sulfate affinity chromatography, and zinc metal chelate chromatography.

Recent findings indicate that rAAV capsid serotypes, such as rAAV-1, 4, 5, and 8, weakly bind to anionic resins, in the form of a purified stock virus solution or in the presence of impurities obtained during production, such as cell DNA host, host cell proteins, serum albumin, components of the medium and components of the helper virus. Thus, purification of these capsid serotype data typically involves anion exchange chromatography in combination with other purification methods, such as density gradient centrifugation of iodoxynol. See, for example, Zolotukhin et al., Methods 28 (2): 158-167 (2002) and Kaludov et al., Hum. Gene Therapy 13: 1235-1243 (2002); and US Patent Publication No. 2004/0110266 A1. However, these methods cannot easily scale to industrial scale processes.

Thus, in the development of recombinant AAV-based vectors, such as those used in gene therapy and gene vaccines, there is a need for methods for purifying rAAV-based vectors from components generated during production, including helper virus, as well as helper virus proteins, cellular proteins, host cell DNA, and environmental components present in the rAAV production stock solution. Such methods should be effectively used on a scale that is suitable for the practical application of gene therapy methods. In addition, there is a need to develop methods for purifying rAAV vectors that scale to produce highly purified commercial high titer stock solutions that can be used for gene therapy and gene vaccines involving rAAV. More specifically, there is a need to develop methods for purifying rAAV-based vectors that weakly bind to chromatographic resins and, in particular, to anionic resins.

The descriptions of all publications, patent applications and patents cited in this specification are incorporated herein by reference, as if for each individual publication, patent application, or patent it was specifically and separately indicated that they are incorporated by reference. In particular, all publications cited herein are expressly incorporated herein by reference for the purpose of describing and disclosing compositions and methods that may be used according to the invention. Although the invention provided herein is described in some detail by way of illustration and example for the purpose of clarity of understanding, those skilled in the art will readily understand in light of the presentation of the present invention that certain changes and modifications may be made with respect to it without departing from the spirit and scope of the attached claims

SUMMARY OF THE INVENTION

The invention relates to methods for separating a population of particles of recombinant adeno-associated virus (rAAV) of any capsid serotype from impurities generated during production by trapping rAAV particles on apatite chromatography medium in the presence of polyethylene glycol (PEG). The methods of the invention comprise a preceding treatment (such as, for example, centrifugation, Benzonase® treatment, anion exchange filtration and / or tangential flow filtering), as well as a subsequent treatment (such as, for example, heat inactivation, filtration, hydrophobic interaction chromatography, gel filtration and / or anion exchange chromatography). The preceding and subsequent methods can be used separately or in various combinations.

The invention relates to methods for separating a population of particles of recombinant adeno-associated virus (rAAV) from impurities formed during the process in a feed stream, comprising the steps of: (a) bringing the feed stream containing rAAV particles into contact with apatite chromatography medium in the presence of polyethylene glycol ( PEG), where rAAV particles bind to apatite chromatography medium; and (b) eluting the rAAV particles bound to the apatite chromatography medium using an elution buffer containing less than 3% (w / v) PEG. In certain embodiments, the apatite chromatography medium is ceramic hydroxyapatite (CHT) or ceramic fluorapatite (CFT). In certain embodiments, the rAAV particles bound to the apatite chromatography medium are eluted with an elution buffer containing less than 3% (w / v) PEG. In certain embodiments, rAAV particles bound to the apatite chromatography medium are eluted with an elution buffer in the absence of PEG.

In some embodiments, the specific binding of apatite chromatography medium is from 10 ⁶ to 10 ¹⁶ DNase resistant particles (DRP) per milliliter. In some embodiments, the specific binding of apatite chromatography medium is from 10 ⁸ to 10 ¹⁶ DNase resistant particles (DRP) per milliliter. In some embodiments, the specific binding of apatite chromatography medium is from 10 ¹⁰ to 10 ¹⁶ DNase resistant particles (DRP) per milliliter. In some embodiments, the specific binding of apatite chromatography medium is between 10 ¹² and 10 ¹⁶ DNase-resistant particles (DRP) per milliliter. In some embodiments, the specific binding of apatite chromatography medium is from 10 ¹⁴ to 10 ¹⁶ DNase resistant particles (DRP) per milliliter.

In some embodiments, the method further comprises, prior to the apatite chromatography step, an anion exchange filtration step, where rAAV particles are in the anion exchange filtrate. In some embodiments, the method further comprises concentrating the rAAV particles from the anion exchange filtrate by tangential flow filtration prior to the apatite chromatography step. In some embodiments, the method further comprises the step of linking the rAAV particles in the feed stream eluted from the apatite chromatography medium to the anionic chromatography medium. In some embodiments, the method further comprises a heat inactivation step to inactivate the helper virus. In some embodiments, the method further comprises the step of binding the rAAV particles in the feed stream to a chromatography medium based on hydrophobic interactions after chromatography on apatite.

In some embodiments, a feed stream containing rAAV particles is contacted with apatite chromatography medium in the presence of polyethylene glycol (PEG) and a base buffer. In some embodiments, the implementation of the main buffer is characterized by a pH between 7.2 and 10, a pH between 7.4 and 10, a pH between 7.6 and 10, a pH between 7.8 and 10, a pH between 8.0 and 10.0, pH between 8.2 and 10.0, pH between 8.4 and 10.0, pH between 8.6 and 10.0, pH between 8.8 and 10, pH between 9.0 and 10.0, pH between 9 , 2 and 10, a pH between 9.4 and 10.0, a pH between 9.6 and 10.0, or a pH between 9.8 and 10.0. In some embodiments, the implementation of the main buffer is characterized by a pH value approximately equal to any of 7.2, 7.6, 7.8, 8.0, 8.2, 8.4, 8.6, 8.8, 9.0, 9.2, 9.4, 9.6, 9.8, and 10.0. Any base buffer known in the art can be used. In some embodiments, the implementation of the main buffer contains borate. In some embodiments, the implementation of the main buffer is a borate.

In some embodiments, a feed stream containing rAAV particles is contacted with apatite chromatography medium in the presence of polyethylene glycol (PEG). For example, from about 3% (w / v) to about 10% (w / v) PEG can be used. In some embodiments, a feed stream containing rAAV particles is contacted with apatite chromatography medium in the presence of about 3% (w / v), about 3.5% (w / v), about 4% ( wt./about.), approximately 4.5% (wt./about.), approximately 5% (wt./about.), approximately 5.5% (wt./about.), approximately 6% (wt. / vol.), approximately 6.5% (w / v), approximately 7% (w / v), approximately 7.5% (w / v), approximately 8% (w / v .), approximately 8.5% (w / v), approximately 9% (w / v), approximately 9.5% (ma ss / vol.), or about 10% (w / v) PEG.

In some embodiments, the PEG is characterized by an average molecular weight between about 5000 (PEG5000) grams per mole and about 15000 (PEG15000) grams per mole, for example, about 5000 grams per mole (PEG5000), about 6000 (PEG6000) grams per mole, about 7000 (PEG7000) grams per mole, approximately 8000 (PEG8000) grams per mole, approximately 9000 (PEG9000) grams per mole, approximately 10,000 (PEG10000) grams per mole, approximately 11000 (PEG 11000) grams per mole, approximately 12000 (PEG12000) grams per mole, approximately 13000 (PEG 13000) grams n and a mole, about 14000 (PEG14000) grams per mole and about 15000 (PEG15000) grams per mole. In certain embodiments, PEG is characterized by an average molecular weight of approximately 5000 (PEG5000) grams per mole. In certain embodiments, PEG has an average molecular weight of approximately 6,000 (PEG6,000) grams per mole. In certain embodiments, PEG is characterized by an average molecular weight of approximately 8000 (PEG8000) grams per mole. In certain embodiments, PEG is characterized by an average molecular weight of approximately 10,000 (PEG 10,000) grams per mole. In certain embodiments, PEG is characterized by an average molecular weight of approximately 15,000 (PEG15,000) grams per mole.

In some embodiments, a feed stream comprising rAAV particles is contacted with apatite chromatography medium in the presence of from about 3% (w / v) to about 10% (w / v) PEG6000. In some embodiments, a feed stream comprising rAAV particles is contacted with apatite chromatography medium in the presence of about 3% (w / v) PEG6000. In some embodiments, a feed stream comprising rAAV particles is contacted with apatite chromatography medium in the presence of approximately 4% (w / v) PEG6000. In some embodiments, a feed stream comprising rAAV particles is contacted with apatite chromatography medium in the presence of approximately 5% (w / v) PEG6000. In some embodiments, a feed stream containing rAAV particles is contacted with apatite chromatography medium in the presence of about 6% (w / v) PEG6000. In some embodiments, a feed stream comprising rAAV particles is contacted with apatite chromatography medium in the presence of about 7% (w / v) PEG6000. In some embodiments, a feed stream comprising rAAV particles is contacted with apatite chromatography medium in the presence of approximately 8% (w / v) PEG6000. In some embodiments, a feed stream comprising rAAV particles is contacted with apatite chromatography medium in the presence of approximately 9% (w / v) PEG6000. In some embodiments, a feed stream comprising rAAV particles is contacted with apatite chromatography medium in the presence of approximately 10% (w / v) PEG6000.

In some embodiments, a feed stream comprising rAAV particles is contacted with apatite chromatography medium in the presence of between about 3% (w / v) and about 10% (w / v) PEG8000. In some embodiments, a feed stream comprising rAAV particles is contacted with apatite chromatography medium in the presence of approximately 3% (w / v) PEG8000. In some embodiments, a feed stream comprising rAAV particles is contacted with apatite chromatography medium in the presence of approximately 4% (w / v) PEG8000. In some embodiments, a feed stream comprising rAAV particles is contacted with apatite chromatography medium in the presence of approximately 5% (w / v) PEG8000. In some embodiments, a feed stream comprising rAAV particles is contacted with apatite chromatography medium in the presence of about 6% (w / v) PEG8000. In some embodiments, a feed stream containing rAAV particles is contacted with apatite chromatography medium in the presence of approximately 7% (w / v) PEG8000. In some embodiments, a feed stream comprising rAAV particles is contacted with apatite chromatography medium in the presence of approximately 8% (w / v) PEG8000. In some embodiments, a feed stream comprising rAAV particles is contacted with apatite chromatography medium in the presence of approximately 9% (w / v) PEG8000. In some embodiments, a feed stream comprising rAAV particles is contacted with apatite chromatography medium in the presence of approximately 10% (w / v) PEG8000.

In some embodiments, a feed stream comprising rAAV particles is contacted with apatite chromatography medium in the presence of between about 3% (w / v) and about 10% (w / v) PEG10000. In some embodiments, a feed stream comprising rAAV particles is contacted with apatite chromatography medium in the presence of about 3% (w / v) PEG10000. In some embodiments, a feed stream containing rAAV particles is contacted with apatite chromatography medium in the presence of about 4% (w / v) PEG 10000. In some embodiments, a feed stream containing rAAV particles is contacted with apatite chromatography medium in the presence of approximately 5% (w / v) PEG 10000. In some embodiments, a feed stream containing rAAV particles is brought into contact with apatite chromatography medium in the presence of approximately only 6% (w / v) PEG 10000. In some embodiments, a feed stream containing rAAV particles is contacted with apatite chromatography medium in the presence of about 7% (w / v) PEG 10000. In some in embodiments, a feed stream containing rAAV particles is contacted with apatite chromatography medium in the presence of approximately 8% (w / v) PEG 10000. In some embodiments, a feed stream containing rAAV particles is contacted with the medium chromatography on apatite in p in the presence of about 9% (w / v) PEG 10000. In some embodiments, a feed stream containing rAAV particles is contacted with apatite chromatography medium in the presence of about 10% (w / v) PEG 10000.

In some embodiments, a feed stream comprising rAAV particles is contacted with apatite chromatography medium in the presence of between about 3% (w / v) and about 10% (w / v) PEG15000. In some embodiments, a feed stream comprising rAAV particles is contacted with apatite chromatography medium in the presence of approximately 3% (w / v) PEG15000. In some embodiments, a feed stream containing rAAV particles is contacted with apatite chromatography medium in the presence of about 4% (w / v) PEG 15000. In some embodiments, a feed stream containing rAAV particles is contacted with apatite chromatography medium in the presence of approximately 5% (w / v) PEG 15000. In some embodiments, a feed stream containing rAAV particles is brought into contact with apatite chromatography medium in the presence of approximately only 6% (w / v) PEG 15000. In some embodiments, a feed stream containing rAAV particles is contacted with apatite chromatography medium in the presence of about 7% (w / v) PEG 15000. In some in embodiments, a feed stream containing rAAV particles is contacted with apatite chromatography medium in the presence of approximately 8% (w / v) PEG 15000. In some embodiments, a feed stream containing rAAV particles is contacted with the medium chromatography on apatite in p in the presence of approximately 9% (w / v) PEG 15000. In some embodiments, a feed stream comprising rAAV particles is contacted with apatite chromatography medium in the presence of approximately 10% (w / v) PEG15000.

In some embodiments, a feed stream containing rAAV particles is contacted with apatite chromatography medium in a buffer containing about 20 mM borate with a pH of 9.0 and about 5% PEG (such as PEG6000). In some embodiments, a feed stream is mixed in-line with an equal volume of buffer containing about 40 mM borate at pH 9.0 and about 10% PEG to give a final concentration of about 20 mM borate at pH 9.0 and about 5% PEG.

In some embodiments, apatite chromatography medium with rAAV particles bound to this medium is washed to remove impurities from the production process before eluting rAAV particles. In some embodiments, apatite chromatography medium is washed one or more times with wash buffer containing decreasing PEG concentrations to remove impurities from the production process. In some embodiments, the apatite chromatography medium is washed one or more times with wash buffer containing between about 3% (w / v) and about 10% (w / v) PEG. In some embodiments, the implementation of the wash buffer contains approximately any concentration of 10% (w / v), 9.5% (w / v), 9% (w / v), 8.5% (w. / vol.), 8% (w / v), 7.5% (w / v), 7% (w / v), 6.5% (w / v), 6 % (w / v), 5.5% (w / v), 5% (w / v), 4.5% (w / v), 4% (w / v) .), 3.5% (w / v), and 3% (w / v) PEG. In some embodiments, the apatite chromatography medium is washed one or more times with wash buffer containing 7.5% (w / v) PEG6000. In some embodiments, the apatite chromatography medium is washed one or more times with wash buffer containing 7.5% (w / v) PEG8000. In some embodiments, the apatite chromatography medium is washed one or more times with wash buffer containing 7.5% (w / v) PEG10000. In some embodiments, the apatite chromatography medium is washed one or more times with wash buffer containing 7.5% (w / v) PEG15000. In some embodiments, the apatite chromatography medium is washed one or more times with wash buffer containing 5% (w / v) PEG6000. In some embodiments, the apatite chromatography medium is washed one or more times with wash buffer containing approximately 5% (w / v) PEG8000. In some embodiments, the apatite chromatography medium is washed one or more times with a wash buffer containing approximately 5% (w / v) PEG 10000. In some embodiments, the apatite chromatography medium is washed one or more times with a wash buffer containing approximately 5% (mass./about.) PEG15000. In some embodiments, the apatite chromatography medium is washed one or more times with wash buffer containing less than about 3% (w / v) PEG6000. In some embodiments, the apatite chromatography medium is washed one or more times with wash buffer containing less than about 3% (w / v) PEG8000. In some embodiments, the apatite chromatography medium is washed one or more times with a wash buffer containing less than about 3% (w / v) PEG 10000. In some embodiments, the apatite chromatography medium is washed one or more times with a wash buffer containing less than than approximately 3% (w / v) PEG15000. In some embodiments, the apatite chromatography medium is washed one or more times with PEG-free wash buffer.

In some embodiments, the wash buffer comprises buffers known in the art. In some embodiments, the wash buffer comprises a buffer selected from the group consisting of borate, N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES), and Tris-HCl. In some embodiments, the wash buffer contains or is borate. In some embodiments, the wash buffer contains or is HEPES. In some embodiments, the wash buffer contains or is Tris-HCl. In some embodiments, a wash buffer exists at a basic pH. In some embodiments, the wash buffer has a pH between pH 7.0 and pH 10.0, between pH 7.2 and pH 10.0, between pH 7.4 and pH 10.0, between pH 7.6 and pH 10, 0, between pH 7.8 and pH 10.0, between pH 8.0 and pH 10.0, between pH 8.2 and pH 10.0, between pH 8.4 and pH 10.0, between pH 8, 6 and pH 10.0, between pH 8.8 and pH 10.0, between pH 9.0 and pH 10.0, between pH 9.2 and pH 10.0, between pH 9.4 and pH 10.0 between pH 9.6 and pH 10.0, or between pH 9.8 and pH 10.0. In some embodiments, the wash buffer has a pH of 7.0, 7.2, 7.4, 7.6, 7.8, 8.0, 8.2, 8.4, 8.6, 8.8, 9.0, 9.2, 9.4, 9.6, 9.8, or 10.0. In some embodiments, the wash buffer contains or is borate at a pH between 8.0 and 10.0. In some embodiments, the wash buffer contains or is borate at pH 8.0. In some embodiments, the wash buffer contains or is borate at pH 9.0. In some embodiments, the wash buffer contains or is borate at pH 10.0. In some embodiments, the wash buffer contains or is HEPES at a pH between 7.0 and 10.0. In some embodiments, the wash buffer contains or is at HEPES pH 7.0. In some embodiments, the wash buffer contains or is HEPES at pH 8.0. In some embodiments, the wash buffer contains or is at HEPES pH 9.0. In some embodiments, the wash buffer contains or is HEPES at pH 10.0. In some embodiments, the wash buffer contains or is Tris-HCl at a pH between 7.0 and 10.0. In some embodiments, the wash buffer contains or is Tris-HCl at pH 7.0. In some embodiments, the wash buffer contains or is Tris-HCl at pH 8.0. In some embodiments, the wash buffer contains or is Tris-HCl at pH 9.0. In some embodiments, the wash buffer contains or is Tris-HCl at pH 10.0. In some embodiments, the wash buffer further comprises between 100 and 500 mM phosphate. In some embodiments, the wash buffer further comprises between 50 and 250 mM NaCl.

In some embodiments, the washing step comprises first washing with a wash buffer containing about 30 mM borate at a pH of about 9.0 and about 7.5% PEG; a second wash with wash buffer containing about 150 potassium phosphate, about 20 mM borate at a pH of about 9.0, and about 5% PEG; a third wash with wash buffer containing about 20 mM borate at a pH of about 9.0 and about 5% PEG; and a fourth wash with wash buffer containing about 20 mM HEPES at a pH of about 7.0 and 150 mM NaCl.

In some embodiments, the implementation of rAAV particles associated with the chromatography medium on apatite, elute with elution buffer containing low concentrations of PEG or not containing PEG. In some embodiments, the elution buffer contains less than about 3% (w / v) PEG, less than about 2% (w / v) PEG, or less than about 1% (w / v) PEG . In some embodiments, the elution buffer contains about 2.5% (w / v), about 2% (w / v), about 1.5% (w / v), about 1% (w / w) ./vol.) or approximately 0.5% (w / v) PEG, or does not contain PEG. In some embodiments, the elution buffer contains less than about 3% (w / v) PEG6000. In some embodiments, the elution buffer contains less than about 3% (w / v) PEG8000. In some embodiments, an elution buffer of less than about 3% (w / v) PEG10000. In some embodiments, the elution buffer contains less than about 3% (w / v) PEG15000. In some embodiments, the rAAV particles bound to the apatite chromatography medium are eluted with an elution buffer in the absence of PEG. In some embodiments, the elution buffer comprises a buffer selected from the group consisting of borate, N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES), and Tris-HCl. In some embodiments, the elution buffer contains or is borate. In some embodiments, the elution buffer contains or is HEPES. In some embodiments, the elution buffer contains or is Tris-HCl. In some embodiments, an elution buffer exists at a neutral pH. In some embodiments, the elution buffer contains or is at HEPES at neutral pH. In some embodiments, the elution buffer contains or is Tris-HCl at neutral pH. In some embodiments, the elution buffer further comprises less than 100 mM phosphate. In some embodiments, the elution buffer further comprises less than 50 mM phosphate. In some embodiments, the elution buffer further comprises between 50 and 250 mM NaCl. In some embodiments, the rAAV particles bound to the apatite chromatography medium are eluted with an elution buffer containing about 50 mM potassium phosphate, about 20 mM HEPES at a pH of about 7.0, and about 150 mM NaCl.

In some embodiments, a method for separating rAAV particles from impurities obtained during a production process in a feed stream comprises the steps of: (a) bringing the feed stream containing rAAV particles into contact with apatite chromatography medium in the presence of about 5% (w / v) .) PEG in the main buffer at a pH of approximately 9.0, where rAAV particles bind to the apatite chromatography medium; (b) washing the apatite chromatography medium with a first washing buffer containing approximately 30 mM borate at a pH of approximately 9.0 and approximately 7.5% PEG; (c) washing the chromatography medium on apatite with a second washing buffer containing approximately 150 potassium phosphate, approximately 20 mM borate at a pH of approximately 9.0 and approximately 5% PEG; (d) washing the apatite chromatography medium with a third washing buffer containing approximately 20 mM borate at a pH of approximately 9.0 and approximately 5% PEG; (e) washing the apatite chromatography medium with a fourth washing buffer containing approximately 20 mM HEPES at a pH of approximately 7.0 and 150 mM NaCl; and (f) eluting the rAAV particles bound to the apatite chromatography medium with an elution buffer containing about 50 mM potassium phosphate, about 20 mM HEPES at a pH of about 7.0, and about 150 mM NaCl.

Also provided herein are methods for separating a population of recombinant adeno-associated virus (rAAV) particles from impurities obtained during the production of impurities in a feed stream, comprising the steps of: (a) bringing the feed stream containing rAAV particles into contact with a hydrophobic interaction chromatography medium (HIC) in a high salt concentration buffer, where rAAV particles and impurities from the production process bind to the HIC medium; and (b) eluting the rAAV particles bound to the HIC medium with a medium salt buffer. In some embodiments, the HIC is selected from the group consisting of Tosoh Butyl 650M resin, Tosoh SuperButyl 650C, Tosoh Phenyl 650C, EMD Fractogel Phenyl and Tosoh Has (butyl). In some embodiments, the high salt concentration buffer contains or is 0.5 M and 2.0 M citrate (e.g., sodium citrate). In some embodiments, the high salt concentration buffer contains approximately any concentration of citrate of 0.5 M, 0.75 M, 1.0 M, 1.25 M, 1.5 M, 1.75 M, and 2.0 M. In some embodiments, the implementation of the buffer with an average concentration of salts contains less than 0.5 M citrate (for example, sodium citrate) or is a buffer with such a concentration. In some embodiments, the implementation of the buffer with an average concentration of salts contains from 0.5 M to about 0.3 M citrate. In some embodiments, the medium salt concentration buffer contains approximately any citrate concentration of 0.45 M, 0.4 M, 0.35 M, 0.3 M, and 0.25 M. In some embodiments, the high salt concentration buffer is additionally contains from 1 to 100 mm phosphate. In some embodiments, the implementation of the buffer with an average concentration of salts further comprises between 1 and 100 mm phosphate. In some embodiments, the medium salt concentration buffer elutes the rAAV particles without eluting the rAAV particles with empty capsids, partially denatured capsids, less infectious capsids and / or partially empty capsids.

In any of the embodiments described herein, rAAV particles have an AAV capsid serotype selected from the group consisting of AAV-1, AAV-2, AAV-3, AAV-4, AAV-5, AAV-6, AAV- 7, AAV-8, AAV-9, AAV-10, AAV-11, AAV-12, AAV-13, AAV-14, AAV-15 and AAV-16. In some embodiments, the rAAV particles have an AAV capsid serotype selected from the group consisting of AAV-1, AAV-4, AAV-5, and AAV-8. In some embodiments, the rAAV particles have the AAV-1 capsid serotype. In some embodiments, the rAAV particles have an AAV-4 capsid serotype. In some embodiments, the rAAV particles have an AAV-5 capsid serotype. In some embodiments, the rAAV particles have an AAV-8 capsid serotype. In some embodiments, rAAV particles comprise an AAV capsid protein from an AAV serotype selected from the group consisting of AAV-1, AAV-2, AAV-3, AAV-4, AAV-5, AAV-6, AAV-7, AAV- 8, AAV-9, AAV-10, AAV-11, AAV-12, AAV-13, AAV-14, AAV-15 and AAV-16. In some embodiments, the rAAV particles have an AAV capsid serotype, which is a weakly binding anionic substance. In some embodiments, an AAV capsid serotype, which is a weakly binding anionic substance, is selected from the group consisting of AAV-1, AAV-4, AAV-5, and AAV-8. In some embodiments, a composition comprising rAAV particles further comprises impurities obtained from production in culture. In some embodiments, the impurities obtained from the production in the culture include damaged rAAV particles, impurities from the host cell, impurities from the helper virus and / or impurities from the cell culture. In some embodiments, the implementation of the impurities from the host cell include the DNA of the host cell, plasmid or protein of the host cell. In some embodiments, contaminants from a helper virus include adenoviral particles, adenoviral DNA, or adenoviral proteins. In some embodiments, impurities from the cell culture include medium components, serum albumin, or other whey proteins. In some embodiments, impurities from the cell culture include medium components. In some embodiments, cell culture impurities do not include serum albumin or other whey proteins.

It should be understood that one, some or all of the properties of the various embodiments described herein can be combined with the formation of other embodiments of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 presents the results of the cleavage of Benzonase® clarified supernatant harvest culture products rAAV. The results demonstrate that no high molecular weight DNA is present after Benzonase® cleavage.

Figure 2 shows a typical spectrophotometric trace of a typical resin screened for binding affinity with rAAV as described in Example 4. Optical density (AU) and conductivity (mS / cm) are indicated.

Figure 3 shows a breakthrough analysis in the presence and absence of PEG. Two model cultures of rAAV products were used to evaluate the capacity of apatite resin (CFT type I). Top panel: breakthrough during loading of serum-containing or serum-free feed streams in the presence or absence of approximately 5% (w / v) PEG6000 in the feed. The loading volumes refer to the initial feed stream before 1: 1 dilution in the process, and they are normalized per ml of resin volume. Bottom panel: loading volumes (ml) at which a 1% breakthrough was observed, and recovery in the eluted fraction. The TFF charges used in the experiment were at a concentration of approximately 10 ¹⁶ DRP / ml for rAAV vectors. In the presence of approximately 5% (w / v) PEG6000, 150 ml of TFF collection was loaded onto 1.2 ml of CFT resin without breakthrough, which was defined as the presence of> 1% rAAV in column washout, corresponding to a loading of 1.8 · 10 ¹⁴ general DRP rAAV.

Figure 4 shows a typical CHT I chromatogram. Together, UV A ₂₈₀ optical density measurements (AU, optical density unit) and conductivity (mS / cm) for 3 mm Amersham Skid are shown. The brackets mark the main segments of the program described in Example 7. “NaOH” marks the stage of column decontamination.

Figure 5 shows the relative purity of rAAV-based vectors eluted from apatite resins. Panel A shows the vector distribution between flow / slip (FT), washing with a high phosphate concentration / 5% (w / v) PEG6000 (PO ₄ ), washing with phosphate and PEG6000 (WII / WIII) and the eluate. None of the differences between runs is significant within the accuracy of the analytical technique, and the lack of material balance is typical. Panel B shows the corresponding tracks with Sypro stained orange SDS-PAGE, which contained fractions eluted from the apatite column. Each sample was loaded on 2 · 10 ¹¹ DRP / track; the apparent difference in migration between the tracks is a salt-related artifact due to the need to concentrate the eluate with CFT by evaporation to a volume that corresponds to application to the gel. The only predominant bands appeared to correspond to AAV capsid proteins. Panel C shows the relevant Ad5 western blotting paths, with the order of the paths being changed for clarity, which demonstrates the comparable purity of Ad5 proteins.

Figure 6 presents the assessment of purification during the process by SDS-PAGE. Obtained in the process samples from a representative collection of product culture were separated in denaturing / reducing 10% polyacrylamide gel and stained with Sypro orange. All samples after collection were loaded in an amount of 1 · 10 ¹⁰ DRP / lane. Two previously obtained samples before the TFF concentration stage (initial clarification stage and AEX slip) can only be loaded in an amount of 1 · 10 ⁹ DRP / lane due to volume limitations of the gel. Beta-galactosidase (B-Gal) was loaded at 50 ng / lane to evaluate gel sensitivity and uniformity. Three capsid proteins AAV1 (VP1, 2 and 3) are indicated.

The figure 7 presents the stepwise output of the cleaning rAAV, as described in examples 1-12. The total number of DRP present in the supernatant before collection was determined as 100%. The output at each stage is the total number of DRP obtained relative to the total number of DRP processed at each stage. The total yield of the whole process was approximately 28%. D4 sup: product culture; AEX FT: anion exchange column slip (Mustang® Q); captured: by chromatography on apatite; heating: inactivation by heating or destruction by heating; HIC: hydrophobic interaction chromatography; SEC: gel filtration; AEX: anion exchange.

DETAILED DESCRIPTION

The aim of the present invention is the provision of methods for separating a population of particles of recombinant adeno-associated virus (rAAV) of any serotype of AAV capsid from impurities associated with product culture, such as damaged rAAV particles, helper virus, helper virus proteins, plasmids, cellular proteins and DNA, medium components, whey proteins, etc. In addition, the methods of the present invention provide commercially scalable, orthogonal processes consistent with regulatory requirements for isolating a population of rAAV particles with harvests or crop streams of products with high rAAV titer. Populations of rAAV particles isolated by the methods of the present invention are substantially free of impurities, including impurities associated with the culture of production and / or technological impurities, such as damaged rAAV particles, helper virus, helper virus proteins, plasmids, cell proteins and DNA, medium components, whey proteins and glucans. The methods of the present invention are particularly suitable for rAAV-based vector serotypes that weakly bind to anions, such as, for example, rAAV-1, rAAV-4, rAAV-5 and rAAV-8. The invention further relates to a method for isolating a population of rAAV particles with a high titer substantially devoid of impurities, including impurities associated with the product culture and / or process impurities suitable for use in gene therapy applications, without the need for density gradient centrifugation.

Definitions

As used herein, the term “isolated” or “purified” refers to the preparation of rAAV particles free of at least some other components that may also be present where rAAV particles are in nature or where the particles are obtained from. Thus, for example, isolated rAAV particles can be obtained using a purification method to enrich them from the starting mixture, such as a culture lysate or product culture supernatant. Enrichment can be measured in various ways, for example, by assessing the proportion of DNase-resistant particles (DRPs) present in the solution, or by assessing infectivity, or it can be measured with respect to a second, potentially inhibiting substance present in the starting mixture, such as impurities, including impurities associated with the culture of the product, and / or technological impurities, such as helper virus, components of the environment, and the like, as defined below.

An rAAV preparation is noted as “substantially devoid” of helper virus if the ratio of AAV infectious particles to helper virus infectious particles is at least about 10 ² : 1; preferably at least about 10 ⁴ : 1, more preferably at least about 10 ⁶ : 1; even more preferably at least about 10 ⁸ : 1. The preparations are also preferably devoid of equivalent amounts of helper virus proteins (i.e., proteins that may result from the presence of the same concentration of helper virus if the impurities of the helper virus particles mentioned above are present in a degraded form). An admixture of viral and / or cellular protein can mainly be observed as the presence of Coomassie stained bands on SDS gels (for example, as the appearance of bands other than the corresponding capsid proteins AAV VP1, VP2 and VP3).

As used herein, the term “weakly binding anionic substance” or “low binding affinity anionic substance” interchangeably refers to an rAAV particle having a capsid serotype which in the presence of impurities (including impurities associated with the culture of the product and / or process impurities) does not bind to an affinity sufficient to allow separation of rAAV particles from other impurities associated with the rAAV production culture. Such capsid serotypes are known in the art and include, but are not limited to, AAV-1, AAV-5, AAV-8, and AAV-4. As described in the art, such weakly binding anionic substances are generally purified by methods that include at least one density gradient centrifugation step including iodixinol gradient centrifugation (commercially available under the trade name Optiprep®) or cesium chloride.

As used herein, the term “helper virus” or “contaminating helper virus” refers to a virus used in the production of copies of a helper virus-dependent virus vector, such as an adeno-associated virus, which is not capable of replicating on its own. Helper virus is used for joint infection of cells together with the viral vector and provides the proteins necessary for replication of the genome of the viral vector. The term encompasses intact viral particles, empty capsids, viral DNA, and the like. Helper viruses commonly used to produce rAAV particles include adenovirus, herpes simplex virus, cytomegalovirus, Epstein-Barr virus and vaccinia virus.

As used herein, the term “product culture” refers to a vessel containing components necessary for the production of rAAV-based vector particles. Product cultures, by way of non-limiting examples, include the following components: 1) a suitable host cell; 2) the function of the helper virus; 3) AAV rep and cap genes and products of these genes; 4) therapeutic transgene flanked by ITR AAV sequences; and 5) a suitable medium, medium components, and medium additives, including but not limited to whey, whey proteins, vitamins, essential and non-essential amino acids, and glucose, which are known to support rAAV production.

As used herein, the terms “impurities”, “impurities associated with a product culture”, “process impurities”, “impurities obtained from a product process”, “impurities” or “contaminants” that are used interchangeably herein, without limitation, include to compositions of media known in the art that provide the production of rAAV vectors; additives to media such as salts, calf serum, amino acid supplements, vitamin supplements, growth factors, serum albumin and other low molecular weight proteins present in media formulations known in the art; to permissive host cells, to proteins of the host cell or DNA of the host cell; helper viruses, helper virus proteins or helper virus DNA, such as wild-type adenovirus or herpes virus proteins; and other non-rAAV vectors or culture materials for rAAV-based vector products introduced during the purification process, for example, glucans or chromatographic buffers used in the purification of rAAV-based vectors from raw material streams.

As used herein, the term “collection of product culture” is defined as a solution containing rAAV-based vector particles produced by rAAV-based vector product cultures by means known in the art, including, but not limited to, transfection processes, production of a stable cell line, Ad- hybrid production systems or baculovirus production systems. In addition, the term “product culture collection” as used herein refers to material isolated from a product culture vessel and includes materials isolated by lysing rAAV producer cells by means known in the art and materials isolated from rAAV production cultures maintained under culturing conditions known in the art, with the release of rAAV particles released into the medium from intact cells. Product culture collection may contain some or all of the following components without limitation: rAAV-based vector particles, product culture components, such as medium components, host cell proteins, DNA host cell DNA, host cells, helper virus, helper virus proteins , Helper virus DNA, plasmid DNA, carrier virus DNA, serum, serum proteins and media additives.

As used herein, the term “feed stream” refers to a source of particles of an rAAV-based vector that is loaded onto, passed through, or applied to the chromatographic phase. The feed streams of the present invention include product culture fees and materials isolated in previous chromatographic steps of the invention, whether the material is present in the slip from the previous step, whether it was bound or eluted in the previous step, whether it was present in the dead volume in the previous step, or whether it was present in any fraction obtained during the purification of rAAV particles. Such feed streams may include one or more “impurities,” “impurities associated with the culture of the product,” “technological impurities,” “impurities obtained from the production process,” or “impurities” or “contaminants,” as defined herein .

The terms “capture,” “bonded,” “bonded,” or “bonded,” as used herein, interchangeably refer to bonding, adhesion, or adherence of a component of a feed stream to a chromatographic medium. The components may bind to the chromatographic medium by any force or chemistry known in the art, including but not limited to hydrophobic interactions, ionic interactions (including anionic and cationic), affinity, metal chelate interactions, and chelating. Components can bind to a chromatographic medium in more than one type of chemistry, for example, apatite chromatographic medium.

The terms “apatite resin”, “apatite chromatographic medium”, “apatite matrix” or “apatite medium”, as used interchangeably herein, refer to a chromatographic medium composed of mineral calcium phosphate, and include, but are not limited to, ceramic chromatographic media hydroxyapatite (CHT) and ceramic fluorapatite (CFT).

The terms “mixed mode” or “multimodal” refer to chromatographic media that are capable of binding to more than one type of chemistry. Mixed-mode chromatographic media include, but are not limited to, apatite chromatographic media that are capable of metal affinity binding via calcium atoms, hydrogen bonding via hydroxyl groups present on the core, repulsion of a positive charge and attraction of a negative charge by calcium atoms and repulsion of a negative charge and attraction positive charge through phosphate groups present on the medium.

A general reference to a “composition” or “compositions” includes the compositions of the invention and is applicable thereto.

As used herein, the singular nouns include plural references unless otherwise indicated. For example, the expression "viral particle" includes one or more viral particles.

A reference to an “approximately” set number or parameter in this document includes (and describes) embodiments that are directed to a given number or parameter, as such. For example, a description with reference to “approximately X” includes a description of “X”.

It should be understood that the aspects and embodiments of the invention described herein include aspects and options consisting of, or essentially consisting of, aspects and embodiments of.

RAAV Vectors Production

Numerous methods for producing rAAV-based vectors are known in the art, including transfection, production of a stable cell line, and infectious hybrid virus production systems, which include adenovirus-AAV hybrids, herpes-AAV hybrids, and baculovirus-AAV hybrids. All rAAV production cultures for the production of viral particles rAAV viral particles require: 1) suitable host cells, including, for example, cell lines of human origin, such as HeLa, A549 or 293 cells, or cell lines derived from insects, such as SF -9, in the case of baculovirus production systems; 2) a suitable helper virus function provided by wild-type adenovirus or mutant adenovirus (such as temperature-sensitive adenovirus), herpes virus, baculovirus, or plasmid construct that provides helper functions; 3) AAV rep and cap genes and products of these genes; 4) a transgene (such as a therapeutic transgene) flanked by ITR AAV sequences; and 5) a suitable medium and medium components to support rAAV production. Suitable media known in the art can be used to produce rAAV-based vectors. These media, by way of non-limiting examples, include those manufactured by Hyclone Laboratories and JRH, including Eagle's Modified Medium (MEM), Dulbecco's Modified Eagle's Medium (DMEM), specially formulated formulations, such as those described in US Pat. No. 6,566,118, and Sf-900 II SFM described in US patent No. 6723551, and each of the patents is incorporated herein by reference in full, in particular, with respect to specially made compositions of the environment for use in the production of vectors based on recombinant AAV.

Suitable culture media for rAAV production of the present invention can be supplemented with serum or recombinant proteins of serum origin to a concentration of 0.5-20% (vol./about. Or mass./about.). Alternatively, as is known in the art, rAAV-based vectors can be obtained under serum-free conditions, which can also be referred to as media without animal products. One of skill in the art can understand that commercial or custom-made media formulated to support the production of rAAV-based vectors can also be supplemented with one or more cell culture components known in the art, including but not limited to glucose, vitamins, amino acids and / or growth factors, to increase the titer of rAAV in product cultures.

RAAV production cultures can be grown under certain conditions (over a wide temperature range, for various periods of time, etc.) suitable for the particular host cell used. As is known in the art, rAAV production cultures include attachment-dependent cultures that can be cultured in suitable vessels for attachment-dependent cells, such as, for example, rotating vials, hollow fiber filters, microcarriers, and in fermenters with a carrier layer or with a fluidized bed layer. RAAV-based vector production cultures may also include suspension-adapted host cells, such as HeLa, 293 and SF-9 cells, which can be cultured in various ways, for example, in roller flasks, in mechanically mixed bioreactors and in disposable systems such like a wave bag system.

The rAAV-based vector particles of the invention can be harvested from rAAV production cultures by lysing host cells in a production culture or by harvesting a consumed medium from the production culture, while ensuring that cells are cultured under conditions known in the art to cause the release of rAAV particles into the medium from intact cells, as more fully described in US Pat. No. 6,566,118. Suitable cell lysis methods are also known in the art and include, for example, multiple freeze / thaw cycles, ultra treatment wook, microfluidization and processing chemicals, such as detergents and / or protease.

RAAV-based vector cleaning

When harvesting the culture, rAAV products of the present invention may contain one or more of the following components: (1) host cell proteins; (2) DNA of the host cell; (3) plasmid DNA; (4) helper virus; (5) helper virus proteins; (6) DNA helper virus DNA; and (7) medium components, including, for example, whey proteins, amino acids, transferrins and other low molecular weight proteins. In addition, rAAV product cultures further include rAAV particles containing an AAV capsid serotype selected from the group consisting of AAV-1, AAV-2, AAV-3, AAV-4, AAV-5, AAV-6, AAV- 7, AAV-8, AAV-9, AAV-10, AAV-11, AAV-12, AAV-13, AAV-14, AAV-15 and AAV-16. In some embodiments, the rAAV particles have an AAV capsid serotype selected from the group consisting of AAV-1, AAV-4, AAV-5, and AAV-8.

In some embodiments, the rAAV product culture collection is clarified to remove debris of the host cells. In some embodiments, product culture is clarified by filtration through a series of porous filters, including, for example, a D0HC Millipore Millistak + HC Pod filter, an A1HC Millipore Millistak + HC Pod filter, and a 0.2 μm Opticap XL 10 Millipore Express filter with a hydrophilic membrane and pores SHC Hydrophilic Membrane. Clarification can also be achieved by several other standard methods known in the art, such as centrifuging or filtering through a 0.2 μm or larger cellulose acetate filter known in the art.

In some embodiments, harvesting the rAAV product culture is further treated with Benzonase® to cleave any high molecular weight DNA present in the product culture. In some embodiments, the Benzonase® cleavage is carried out under standard conditions known in the art, including, for example, a final concentration of Benzonase® of 1-2.5 units / ml at a temperature in the range of room temperature to 37 ° C for a period of 30 minutes to a few hours.

RAAV particles can be isolated or purified using one or more of the following purification steps: anion exchange filtering, tangential flow filtering (TFF) to concentrate rAAV particles, capture of rAAV particles by apatite chromatography, heat inactivation of helper virus, capture of rAAV by chromatography based hydrophobic interactions; buffer change by gel filtration (SEC); nanofiltration and rAAV capture by anion exchange chromatography. These stages can be used separately, in various combinations or in a different order. In some embodiments, the method includes all steps in the order described below.

Anion exchange filtration

In some embodiments, a clarified and Benzonase®-treated product culture collection is optionally subjected to anion exchange filtration under conditions in which the rAAV vector is present in the breakthrough and the contaminating helper virus is retained on the charged filter. With an ionic harvesting force of rAAV product, rAAV particles can be separated from a helper virus, such as an adenovirus, by passing through an anion filter, such as a Mustang® Q filter (Pall Corp., East Hills, NY). One of skill in the art can determine the size and number of filters needed to achieve the optimal procedure for lowering the level of adenovirus (LRV) and adenoviral proteins present in clarified, processed Benzonase® and passed through anion filter culture products. In some embodiments, the LRV is at least one order and more than ten orders of magnitude. In a preferred embodiment, the LRV is at least two or more than eight orders of magnitude. In a more preferred embodiment, the LRV is at least six orders of magnitude.

Tangential Flow Filtration Concentration (TFF)

In some embodiments, the anion filtration slip of the clarified Benzonase®-treated feed stream is concentrated by tangential flow filtration ("TFF") before apatite chromatography is applied to the medium. Large-scale concentration of viruses using TFF ultrafiltration is described by R. Paul et al., Human Gene Therapy, 4: 609-615 (1993). Concentration of the feed stream by TFF provides a technically controlled volume flow of the feed materials subjected to the chromatography steps of the present invention and allows the use of more reasonable column sizes without the need for long periods of recirculation. In some embodiments, the rAAV feed stream is concentrated at least two times and at least ten times prior to concentration. In some embodiments, the feed stream is concentrated at least ten times, and at least twelve times before concentration. In some embodiments, the feed stream is concentrated at least twelve times and before concentration at least fifty times. One skilled in the art will also understand that TFF can also be used at any stage of the cleaning process where it is required to change buffers before performing the next stage of the cleaning process.

Capture rAAV by chromatography on apatite in the presence of polyethylene glycol (PEG)

FDA-approved methods for the purification of proteins and other biological products suitable for use in clinical trials in humans and pharmaceuticals are based on orthogonal processes on a commercial scale. It is believed that a multi-stage cleaning scheme includes an orthogonal process if it uses separation mechanisms that are different from each other, despite the fact that each stage is an axis of Cartesian space. For example, a two-step process using anion exchange chromatography and hydrophobic interaction chromatography (HIC) will be considered orthogonal. The processes for removing impurities, such as impurities associated with a product culture, or process impurities from harvesting a product culture or from a feed stream described herein are orthogonal processes including the steps of capturing and skipping the final product (i.e., rAAV-based vectors) on a number of chromatographic media. RAAV-based vectors (specifically rAAV-2), as has been demonstrated in the art, bind to anionic resins. RAAV-based vectors, such as rAAV-1, -5, and -8, have been shown to bind much less strongly than rAAV-2 to anion exchange media in the presence of production components such as serum albumin, helper virus components, components the production environment and the DNA of the host cell, which leads to a less efficient purification scheme with lower quality.

Previous purification strategies described in the art for low affinity binding anionic substances such as AAV-1 included the iodixinol gradient step, which reduced the relative concentration of impurities, such as product impurities and process impurities, to achieve stronger binding rAAV-based vectors with anion exchangers. The iodixinol gradient steps are difficult to scale to the commercial scale processes described herein.

The inventors of the present application have discovered that rAAV-based vector particles can be separated from impurities, such as impurities associated with product culture, and process impurities, by entrainment and elution from apatite resins. Thus, in addition to capturing the product from the untreated feed stream, the apatite column removes a number of process impurities, including host and adenovirus proteins, glucans and serum proteins, and also provides additional purification factors for the helper virus (such as the Ad5 helper virus) .

Apatite resins include chromatographic media containing calcium phosphate minerals, including, but not limited to, ceramic hydroxyapatite (CHT) and ceramic fluoroapatite (CFT). Chromatography media on apatite is also referred to as mixed-mode or multimodal media because apatite has functional groups that provide more than one kind of binding chemistry. Without intending to relate this to theory, apatite media provide an opportunity for calcium-based metal affinity bonding, hydrogen bonding, repulsion of a positive charge, attraction of a positive charge, repulsion of a negative charge and attraction of a negative charge through a variety of different chemical groups including hydroxyl residues present on the core , positively charged calcium atoms and negatively charged phosphate radicals present on the resin. Each type of binding chemistry contributes to the mixed binding mode exactly as it participates in single-mode chromatography. However, unlike single-mode chromatography, different types of chemistry of binding and elution are not independent and can work in opposite ways. For example, an increase in ionic strength may control hydrophobic binding. (T. Kawasaki, M. Niikura, and Y. Kobayashi, J. Chrom. 515: 125-148 (1990) and PS Gagnon, P. Ng, J. Zhen, C. Aberin, and J. He, BioProcess Int ' l 4: 50-60 (2006)). Specifically, CHT and CFT are spherical macroporous forms of hydroxyapatite (Ca ₅ (PO ₄ ) ₃ OH) ₂ , sintered at high temperature to convert the mineral from crystalline to ceramic. This forms a chromatography medium with a macroporous structure providing a large surface area, limited mass transfer resistance, high strength and base resistance. Sintering at different temperatures and for different times leads to the formation of various physical structures of type I and II, which are chemically identical, but provide different capacities for different classes of molecules. CFT differs from CHT in that it is a fluoroapatite and hydroxyapatite composite obtained by chemically replacing hydroxyl groups with fluorine groups to increase stability under acidic conditions. Resins CFT and CHT are commercially available (e.g., from Bio-Rad Laboratories, Inc.).

The authors of the present invention unexpectedly found that the presence of polyethylene glycol (PEG) in the loading buffer significantly increases the capacity and reproducibility (by reducing the variation in the yield of rAAV particles in the breakthrough) of the binding of rAAV-based vector particles to apatite resins. Without regard to theory, one of the features of rAAV vectors, which distinguishes them from most technological impurities, is the large physical particle size. This size difference is used in the capture and washing steps by incorporating polyethylene glycol (PEG) in the chromatographic binding and washing buffers to preferably increase the separation coefficients of the larger molecules in the bound state based on the adoption of an energetically advantageous form by hydrated shells. While the use of PEG in the purification of vectors based on viruses and bacteriophages is described in this field, in contrast to the present invention, it was used mainly as a precipitating agent for the physical aggregation and removal of viral particles from the solution. Since PEG in this area is known to promote aggregation and precipitation of viral particles, and rAAV in this area is described to aggregate at ionic strengths below 200 mM (Wright et al., Molecular Therapy 12: 171-178 (2005 )), the effect of PEG on the binding of rAAV-based vector to apatite resins was unpredictable. PEG is known in the art for facilitating the binding of immunoglobulin molecules to ion exchange resins, as described, for example, in Gagnon, J. Chromtogr. 743A: 51-55 (1996), and with charged hydrophobic resins with a mixed mode, as described, for example, in Gagnon et al., 22 ^nd International IBC Conference on Antibody Production and Development, March 4-6, 2009.

The inventors of the present application determined, based on experimentation with PEG6000 in the concentration range of 3-10% (w / v) in the feed stream, that the relative concentration of about 5% (w / v) PEG6000 is optimal. One skilled in the art will recognize that other types and molecular weights of PEG may be used, including, but not limited to, PEG8000, PEG 10000, and PEG 15000, and that the relative concentration of PEG in the final concentration of the rAAV-based vector solution can be empirically determined so that, with suitable the concentration of PEG rAAV-based vector particles in solution are sent to bind to an apatite resin, but do not form aggregates or are not physically precipitated.

In some embodiments, the rAAV-based vector particles are separated from the culture-related impurities by trapping on an apatite resin in the presence of PEG and eluting the bound rAAV particle from the apatite resin in phosphate buffer. In preferred embodiments, the rAAV-based vector particles are separated from product culture impurities by trapping on an apatite resin in the presence of PEG and eluting the bound rAAV particle from the apatite resin in the buffer in the absence of PEG. In some embodiments, capsid-containing rAAV particles that are weakly binding anionic substances are separated from the product culture from product culture impurities by trapping on an apatite resin in the presence of PEG, and the bound rAAV particles are eluted from the apatite resin in the buffer in the absence of Peg. In more preferred embodiments, rAAV particles containing serotype 1 capsids (rAAV-1 serotype) are separated from product culture impurities by capture on apatite resin in the presence of PEG, and particles containing rAAV-1 serotype capsid bound to the resin, elute in buffer in the absence of PEG. In some embodiments, rAAV vector particles are separated from product culture impurities by a method comprising loading a feed stream in a phosphate-free loading buffer but in the presence of PEG, and eluting the bound rAAV with an apatite resin in an elution buffer containing phosphate and not containing PEG.

While apatite chromatography in the presence of PEG represents an effective capture or binding strategy for the purification of rAAV-based vectors, many process impurities also linger on the apatite resin at pH 7.0. Without being bound by theory, the proteins present in the feed stream at a basic pH (pH above 7.0) are more likely to have a net negative charge and are repelled by negatively charged phosphate binding sites present on the apatite resin, thereby reducing the total cation exchange the binding ability of the chromatographic resin. However, due to the multimodal nature of apatite resins, bonding by attraction with a positive charge and metal affinity properties will all take place early.

Borate buffers are widely used in the art as primary buffer systems because of their preferred manufacturing characteristics, including but not limited to ease of preparation, optimal solubility, excellent buffer capacity, and low cost. Thus, borate buffers as a model buffer system were tested for trapping rAAV on apatite resins. It will be apparent to one of skill in the art that other major buffer systems can be evaluated to determine if they reduce the level of binding of process impurities to apatite resins in the presence of PEG. Other core buffers can be tested and applied to rAAV capture. In some embodiments, the apatite loading buffer without PEG contains a borate buffer. In preferred embodiments, the borate buffer is formulated at a pH of from about 8.0 to about pH 9.9. In a preferred embodiment, the borate buffer is formulated at a pH of approximately 9.0. In some embodiment, the borate buffer is in a concentration of from about 5 mM to about 500 mM. In more preferred embodiments, the borate buffer is formulated as approximately 20 mM borate, pH 9.0. In some embodiments, a 20 mM borate buffer at pH 9.0 specifically reduces the uptake of low molecular weight process impurities onto the apatite resin.

In some embodiments, the feed stream is loaded onto an apatite resin in a phosphate-containing buffer in the presence of PEG by mixing the feed stream with a phosphate buffer containing PEG at a concentration twice that of the final. In some embodiments, the pH of the phosphate buffer is between pH 6.5 and pH 7.0. In some embodiments, the PEG is PEG6000. In some embodiments, the concentration of PEG6000 in the loading buffer is between about 3% (w / v) and about 10% (w / v). In more preferred embodiments, the concentration of PEG6000 in the loading buffer is about 5% (w / v). In some embodiments, the concentration of phosphate in the apatite resin loading buffer is between 5 mM and 500 mM.

In some embodiments, the binding capacity of the apatite resin in the presence of PEG is increased relative to the binding ability of the apatite resin in the absence of PEG. In some embodiments, the binding ability of the apatite resin to particles of the rAAV-based vector in the feed stream in the presence of PEG is increased from about half an order to about ten orders of magnitude relative to the binding ability of the apatite resin in the absence of PEG. In preferred embodiments, the binding capacity of the apatite resin with respect to rAAV-based vector particles in the feed stream in the presence of PEG is increased eight-fold. In some embodiments, the binding ability of the apatite resin to rAAV-based vector particles in the feed stream in the presence of PEG is at least about 10 ⁶ particles of rAAV per ml of resin to about 10 ¹⁶ particles per ml of resin (e.g., approximately any the amount of 10 ⁶ , 10 ⁷ , 10 ⁸ , 10 ⁹ , 10 ¹⁰ , 10 ¹¹ , 10 ¹² , 10 ¹³ , 10 ¹⁴ , 10 ¹⁵ , 10 ¹⁶ particles per ml of resin). In some embodiments, the binding capacity of the apatite resin in the presence of PEG is approximately 10 ¹⁴ particles per ml of resin.

While this unexpected binding capacity of approximately 10 ¹² -10 ¹⁴ DRP rAAV-1 per ml of apatite resin in the presence of PEG provides a highly efficient, economical scaling of commercial purification of rAAV-1, one skilled in the art will recognize that the binding capacity is the maximum number or the number rAAV-1, which will bind per ml of resin, and is not intended to operational limit the scope of the invention. In fact, the inventors understand that rAAV-1-based vector harvest cultures that contain less than 10 ¹⁴ -10 ¹⁶ DRP rAAV-1 / ml can be purified by the present invention.

In some embodiments, the implementation of the rAAV particles associated with the apatite resin, washed before eluting the rAAV particles from the resin. In some embodiments, the apatite chromatography medium is washed one or more times with a wash buffer containing decreasing concentrations of PEG to remove process impurities. In some embodiments, apatite chromatography medium is washed one or more times with wash buffer containing between about 3% (w / v) and about 10% (w / v) PEG. In some embodiments, the implementation of the washing buffer contains approximately any concentration of 10% (w / v), 9.5% (w / v), 9% (w / v), 8.5% (w / w) ./vol.), 8% (w / v), 7.5% (w / v), 7% (w / v), 6.5% (w / v), 6% (w / v), 5.5% (w / v), 5% (w / v), 4.5% (w / v), 4% (w / v) vol.), 3.5% (w / v) and 3% (w / v) PEG. In some embodiments, the apatite medium is washed with a wash buffer containing PEG at a concentration higher than the concentration of PEG used to allow the binding of rAAV particles to the apatite medium. In some embodiments, the apatite medium is further washed with decreasing concentrations of PEG. In some embodiments, the wash buffer comprises buffers known in the art. In some embodiment, the wash buffer comprises a buffer selected from the group consisting of borate, N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES), and Tris-HCl. In some embodiments, the wash buffer has a basic pH. In some embodiments, the wash buffer has a pH between pH 7.0 and pH 10.0, between pH 7.2 and pH 10.0, between pH 7.4 and pH 10.0, between pH 7.6 between pH 10 , 0, between pH 7.8 and pH 10.0, between pH 8.0 and pH 10.0, between pH 8.2 and pH 10.0, between pH 8.4 and pH 10.0, between pH 8 6 and pH 10.0, between pH 8.8 and pH 10.0, between pH 9.0 and pH 10.0, between pH 9.2 and pH 10.0, between pH 9.4 and pH 10, 0, between pH 9.6 and pH 10.0 or between pH 9.8 and pH 10.0. In some embodiments, the implementation of the wash buffer is characterized by a pH of 7.0, 7.2, 7.4, 7.6, 7.8, 8.0, 8.2, 8.4, 8.6, 8.8 , 9.0, 9.2, 9.4, 9.6, 9.8, or 10.0. In some embodiments, the wash buffer further comprises between 100 and 500 mM phosphate. In some embodiments, the wash buffer further comprises between 50 and 250 mM NaCl.

In some embodiments, rAAV-based vectors isolated from a feed stream by capture onto an apatite resin in the presence of PEG are eluted into a buffer with low PEG concentrations. In some embodiments, low PEG concentrations are between about 2.9% (w / v) and about 0.1% (w / v) PEG. In some embodiments, rAAV-based vectors isolated from the feed stream by capture onto an apatite resin in the presence of PEG are eluted into the buffer without PEG. In preferred embodiments, rAAV-based vectors isolated from the feed stream by capture onto an apatite resin in the presence of PEG are eluted into a phosphate-free buffer containing no PEG.

In some embodiments, rAAV-based vectors isolated from a feed stream by capture onto an apatite resin in the presence of PEG are eluted into a phosphate-containing buffer. In some embodiments, rAAV-based vectors isolated from the feed stream by capture onto an apatite resin in the presence of PEG are eluted into a buffer containing phosphate at a concentration of between about 0.1 mM and about 500 mM (e.g., between about 1 mM and about 250 mM, between about 10 mM and about 100 mM). In preferred embodiments, rAAV-based vectors, vectors isolated from the feed stream by capture onto an apatite resin in the presence of PEG, are eluted in 50 mM phosphate buffer.

The inventors of the present application have found that rAAV-based vectors present in the feed stream can be isolated by capture on an apatite resin in the presence of PEG. However, if helper viruses used in the culture of production (such as adenovirus) are present in the apatite resin feed stream, they are captured by the apatite resin in the presence of PEG. The rAAV-based vector particles captured by the apatite resin in the presence of PEG can easily be separated from the adenovirus by their elution profile in phosphate buffers. The rAAV-based vector particles bound to the apatite resin in the presence of PEG are eluted in the absence of PEG into buffers containing a small amount of phosphate, such as 0 mM; at this time, adenoviral helper virus particles are retained on apatite resins at phosphate concentrations used to elute the rAAV-based vector particles. It was experimentally found that rAAV-based vectors elute as a single acute peak in a low phosphate concentration of 50 mM in the absence of PEG, while helper viruses such as adenovirus, if present, are retained on the resin. In internal control experiments, when 8 ⁹ DNase-resistant particles (DRP) of infectious adenovirus were added to the rAAV feed streams and chromatographed on apatite resins in the presence of PEG, rAAV-based vectors were captured on an apatite resin and eluted in 50 mM phosphate buffer without PEG, while approximately 4 orders of adenoviral proteins were retained on apatite resin. Thus, in some embodiments, rAAV-based vectors present in the feed stream are separated from contaminating helper viruses by capture on an apatite resin in the presence of PEG and eluting into phosphate buffers without PEG. In some embodiments, phosphate buffers are formulated at concentrations at which the contaminating helper virus remains bound to the apatite resin. In some embodiments, two to eight orders of magnitude adenovirus is retained per ml of apatite resin. In some embodiments, rAAV-based vectors present in the feed stream are isolated by eluting from an apatite resin in 0-500 mM phosphate buffer (e.g., 0-400 mM, 0-300 mM, 0-200 mM, 0-100 mM , 0-50 mM) under conditions in which the helper virus remains bound to the apatite resin.

Production systems known in the art for the production of rAAV-based vectors may include production media containing serum in the range of 0.5-20% (v / v), or may be completely devoid of serum. In addition, the purification schemes described in this field can include one or more concentration steps, which can lead to an increase in the level of whey proteins and other serum components in the feed stream applied to the apatite resin. For example, the culture supernatant of the product described herein, which is formulated with 1% (v / v) serum, was concentrated approximately twenty times at the TFF stage, so that the feed stream loaded onto the apatite resin contained such a large amount of impurity whey protein, as 20%, compared with the concentrate flow of raw materials from a culture of products made without serum. The inventors of the present application tested the apatite uptake methods provided herein with concentrates of a feed stream from product cultures made up in the presence or absence of serum. It was found that the presence of serum proteins in the feed stream does not affect the functionality of the apatite chromatography stage.

Inactivation of a helper virus by heating (Destruction by heating)

If infectious adenovirus is used as a source of helper virus in product cultures for rAAV production, an optional heat inactivation step (heat kill) may be inserted to inactivate any remaining adenoviral particles that may be present in the feed stream. The heat inactivation step takes advantage of one of the main differences between AAV and adenovirus: adenovirus particles are inactivated at temperatures of about 54-56 ° C, while the virus particles AAV and rAAV are stable and are not affected by these temperatures. In the present invention, the inventors have set up the heat inactivation step so that it is adapted to a larger scale organization of the process, for example, 250 L product cultures carried out as described herein. In particular, the apatite eluate was inactivated by heating in a 5 L sterile disposable biotechnological container on a swing platform with temperature control set to a temperature of up to 53 ° C at a swing speed of 40 rpm, with a mixing angle of 12 ° (20 L Wave water bath) . The apatite eluate was incubated on a platform until a temperature of 52 ° C was reached, and then kept at that temperature for an additional 10 minutes. MgCl _{2 was} added to the apatite eluate at a final concentration of 2 mM to stabilize the rAAV vector during heating. One of skill in the art can understand that the scale, final heating temperature, and heating time can be tested empirically to find the optimal conditions for inactivation of adenoviral particles while maintaining the infectivity and integrity of rAAV particles. The heat inactivation step may be skipped when the rAAV particles are purified from product cultures that use plasmid constructs to provide helper function.

Hydrophobic Interaction Chromatography

Hydrophobic Interaction Chromatography (HIC) is a method for separating biomolecules based on differences in hydrophobicity of their surface. Thus, the HIC is considered to be orthogonal to the other stages of purification in the AAV production process. Chromatographic media for HICs contain hydrophobic ligands, such as linear chain hydrocarbons (e.g. propyl (C3), butyl (C4), hexyl (C6), or octyl (C8)) or aromatic hydrocarbons (e.g. phenyl). In pure water, the hydrophobic effect is too weak for the functional interaction between the ligand and the proteins, or between the proteins themselves. However, lyotropic salts enhance hydrophobic interactions, and the addition of salt directs the proteins to be captured by HIC. For this reason, HIC resins typically load at a high salt concentration and elute at a low salt concentration. As one skilled in the art understands, ammonium sulfate [(NH ₄ ) ₂ SO ₄ ] is the most widely used salt for controlling protein uptake by HIC chromatography because of the high lyotropic coefficient of ammonium and sulfate ions in the Hoffmeister series and the high solubility of the salt. In the present invention, rAAV particles present in the feed stream were loaded onto the HIC resin by in-line mixing in a ratio of 75:25 (volume: volume) 2 M ammonium sulfate + 50 mM Tris bis buffer (pH 7.0): feed stream materials, respectively. In-line mixing of the feed stream with the loading buffer prevents the risk of any precipitation of the rAAV-based vector with the high concentration of ammonium sulfate present in the buffer. As one skilled in the art understands, the concentration of salt (ammonium sulfate) can be controlled to achieve the optimal concentration for rAAV binding. Thus, in some embodiments, the concentration of ammonium sulfate is between 1 M and 3 M. In some preferred embodiments, the concentration of ammonium sulfate in the loading buffer is 2 mM. As one skilled in the art understands, in-line mixing of ammonium sulfate and a feed stream is carried out for convenience and to control a flow device, but it is easy to mix a feed stream with a suitable concentration of a loading buffer by any means known in the art, and then load the flow solution starting materials + buffers for loading onto HIC chromatographic medium.

Co-solvents can also affect hydrophobic interactions. For example, ethylene or propylene glycol can weaken the interaction between the protein and the immobilized ligand, and thus can be used to improve elution profiles. Thus, the HIC column was washed with a buffer mixture of 75:25 (volume: volume) 2 M ammonium sulfate + 50 mm Tris bis-buffer (pH 7.0): 50 mm bis-Tris (pH 7.0) + 10% propylene glycol (volume: volume) (EMD Biosciences) and rAAV was eluted with a low salt buffer with propylene glycol (800 mM ammonium sulfate + 50 mM Tris bis-buffer (pH 7.0) + 4% propylene glycol). Under these elution conditions, any remaining helper viruses and proteins present in the feed stream loaded onto the column remain bound to the column. Propylene glycol in this example was added to the buffers to obtain a sharper peak elution profile compared to a wide elution profile of the buffer without propylene glycol, but the latter is not required for the process.

Non-limiting examples of suitable hydrophobic resins include Tosoh Butyl 650M, Tosoh SuperButyl 650C, Tosoh Phenyl 650C and EMD Fractogel Phenyl (Tosoh Bioscience LLC, PA).

Waste from methods for producing rAAV require strict disinfection before disposal, for at least two reasons: (1) the product contains a viral vector; and (2) live adenovirus type 5 (Ad5) is commonly used in product cultures as a helper virus for rAAV production. Liquid wastes from chromatographic operations, as a rule, are first disinfected with bleach at the place of use, and then subjected to further disinfection by incubation at high pH before neutralization and disposal.

Ammonium sulfate present in HIC buffers reacts with bleach and sodium hydroxide to release hazardous chlorine and ammonia gases, respectively. Thus, the initial consideration for optimizing the HIC chromatography step was the development of a suitable buffer system that can be safely disinfected by methods known in the art.

As one of skill in the art would recognize, high salt concentration buffers used in hydrophobic interaction chromatography can be further screened for viscosity, which can result in high backfiltration pressures that can limit flow rates or cause mixing problems, thereby increasing the risk of precipitation of the product due to crystallization of salt in buffers occurring at temperatures that can be used for storage or operation. Based on the data shown in Table 6 below and taking into account the factors described above, the inventors chose 1 M sodium citrate (pH 7.0) as the optimal buffer for binding rAAV vectors to HIC chromatographic medium, although citrate buffers can be used in hydrophobic interaction chromatography in concentrations ranging from 0.5 M to 2.0 M.

Buffer change by gel filtration (SEC)

Numerous methods for changing a buffer described herein are known in the art, including TFF and dialysis. The use of gel filtration has the additional advantage of providing further purification from proteins, the dimensions of which allow them to pass through the pores of the resins, and this method is relatively fast in terms of the time required for the exchange of buffers. Buffers were exchanged at this stage to ensure that the HIC eluate in the previous step was exchanged for a suitable buffer to bind rAAV in the final step of the process, anion exchange chromatography.

Additional agent (virus removal)

Optionally, a further step of removing trace impurities, such as additional viruses that may be present in the feed stream, may be included in the process, thus providing a commercially reasonable orthogonal process. Thus, in some embodiments, the process further includes a virus removal filter. Examples of such filters are known in the art and include Millipore Viresolve NFR (50 nm), Pall Ultipore VF (50 nm) and Asahi 70 nm.

Anion exchange chromatography

The step of anion exchange capture of an rAAV-based vector, chromatographed on apatite, was carried out as a final concentration step and brought to readiness. Suitable media for anion exchange chromatography are known in the art and, by way of non-limiting examples, include Unosphere Q (Biorad, Hercules, California) and N-charged amino or imino resins, such as, for example, POROS 50 PI, or any DEAE-based, TMAE, tertiary or quaternary amine or PEI resins known in the art (US Pat. No. 6,989,264; N. Brument et al., Mol. Therapy 6 (5): 678-686 (2002); G. Gao et al., Hum. Gene Therapy 11: 2079-2091 (2000)). One of skill in the art can understand that wash buffers with suitable ionic strength can be determined so that rAAV remains bound to the resin, while other process impurities, including but not limited to glucans that can enter the product by leakage from various filters used in the cleaning stages will be washed away. In some embodiments, the wash buffer is 60 mM NaCl, and the rAAV vector elutes from the column in 130 mM NaCl, so that any trace residues of process impurities, such as serum albumin or helper virus, are retained on the column.

EXAMPLES

Example 1. Collection of rAAV-1 from the medium for cultivation - clarification and cleavage of Benzonase®

The rAAV-1 production waste medium (supernatant) from 250 L of rAAV-1 viral production culture produced by any method known in the art and containing the rAAV-1-based vector was clarified to remove any cells contained in the supernatant. The supernatant was passed through a series of filters, connected in series, including: (1) Millipore Millistak + ® HC Pod filter, grade D0HC (Millipore Corp., Bedford, MA) (4 times); (2) Millipore Millistak + ® HC Pod filter, grade A1HC; and (3) a 0.2 μm Opticap® XL10 Millipore Express SHC hydrophilic membrane filter, at a speed of 5 liters per minute (LPM), the speed being reduced in stages to 4 LPM.

All filters were pre-washed in reverse osmosis / deionized ("RO / DI") water according to the manufacturer's specifications. The slip was collected in a Benzonase® biotechnology cleavage container. The final concentration of Benzonase® 2 units / ml (EM Industries, catalog number 1.01695,0002) was dissolved in rAAV-1 medium and added to the clarified viral supernatant to reach a final concentration of 2.5 units / ml. The supernatant with Benzonase® was incubated at ambient temperature with constant recirculation at 4 LPM to ensure DNA cleavage. Benzonase® cleavage data is shown in the figure. 1, which demonstrated that after cleavage of Benzonase® there was no high molecular weight DNA.

Example 2. The removal of impurities associated with the product through the exchange of anions

The clarified and cleaved Benzonase® rAAV-1 supernatant of Example 1 was passed through a series of two by twenty two inch Pall Mustang® Q filters (“MQ”) connected in series (Pall Corp., catalog number NP6MSTGQP1). Before loading the viral supernatant rAAV-1, the filters were sanitized with 15 L of 0.5 M NaOH at 0.5 LPM for an incubation time of 15 min, charged by rinsing in 15 L of TMEG + 2 M NaCl (TMEG: 0.05 M Tris-HCl, pH 7.5, 1 mM 2-mercaptoethanol, 1 mM Na ₂ EDTA, 10% (v / v) glycerol) at a rate of 6 LPM and equilibrated 15 vector production media at 6 LPM. Then the supernatant was pressed at a speed of approximately 6 LPM through a series of filters and collected in a biotechnological container. Under the ionic strength of the production medium, the MQ anion exchange filter, as was shown, in addition to other impurities, removed the helper virus and the remaining DNA from the rAAV-1 supernatant by binding the impurities to the charged membrane. However, with the ionic strength of the product culture, the rAAV-1 based vector present in the supernatant was washed through the anion exchange membrane. During the optimization of the process, it was experimentally determined that the use of a single MQ filter led to the penetration of impurities into the process stream, including the Ad5 helper virus. Thus, a second filter was added to the process as a series or tandem.

Example 3. Concentration of the supernatant of a vector based on rAAV-1

The entire supernatant of rAAV-1 based vector production processed in Examples 1 and 2 was concentrated approximately 20 times using tangential flow filtration (“TFF”) from an initial volume of approximately 250 L to a volume of approximately 12.5 L. Tangential flow polyethersulfone filter cartridges with 100 kDa molecular weight cut-off, C-screen and total surface area of 5 m ² (Millipore Pellicon® 2 Biomax, catalog number P2B100C05), rinsed with 50 l of RO / DI water, disinfected 15 l of 0.5 M NaOH during the aging stage lasting 15 minutes, rinsed again with 100 l of WIFI (purified water HyPure ™ WFI; HyClone, Logan, UT), rinsed with 15 l of TMEG + 2M NaCl, and finally balanced with 15 l of rAAV production medium -one. The supernatant was passed through a TFF cartridge at a flow rate of approximately 3 LPM with a recirculation rate of 16 LPM. The TFF material remaining on the filter (“concentrate”) was concentrated to approximately 10.5 L and transferred to a reservoir. The filter was rinsed with approximately 2 L of product medium. Then the washing and concentrate were combined to obtain a final volume of approximately 12.5 L.

Concentration of rAAV-1 to a volume of 12.5 L by TFF also led to the concentration in solution of the remaining impurities associated with the production of rAAV-1, approximately 20 times. Thus, after the TFF step, a production retention step was introduced during which the TFF concentrate was filtered through a 0.22 μM 4-inch Opticap filter membrane (Millipore Opticap® catalog No. KVSC04HB3). This additional filtration step allows the passage of rFF-1 containing TFF concentrate to the next process step without diafiltration and buffer exchange. The material after TFF can be stored at 2-8 ° C for any period of time, including at least 24 hours to 3 months or more, before further processing without loss of stability, as measured by the yield or infectivity of the vector, as assessed by analyzes known in this area. Alternatively, the TFF carried out as described herein can be used at any stage of the purification process to concentrate the rAAV-1 based vector or buffer exchange.

Example 4. Screening resins for binding of the vector based on rAAV-1 compared with the binding of impurities obtained in the production process

Commercial, FDA-approved methods for purifying proteins and other biological products are based on embedded commercial-scale orthogonal processes. Orthogonal processes are processes that include more than one step or process for removing process impurities, including the capture and slip stages of the final product, such as, for example, an rAAV-1 based vector. RAAV-based vectors (specifically rAAV-2), as has been demonstrated in the art, bind to anionic resins. RAAV-based vectors, such as rAAV-1, -5, and -8, have been shown to bind much less strongly than rAAV-2 to anion exchangers in the presence of production components such as serum albumin, helper virus components, environmental components and DNA and the host cell, which leads to the appearance of a less efficient and lower-quality purification scheme.

Previous purification strategies described in the art for low affinity binding anionic substances such as AAV-1 include the iodixinol gradient step, which reduces the relative concentration of product components to achieve stronger binding of the rAAV-based vector to anion exchangers. The iodixinol gradient steps are difficult to expand to a commercial scale, such as those described herein. Thus, to optimize the purification of the rAAV-based vector for low-affinity anionic binders, such as rAAV-1, without the need for ultracentrifugation and stepwise gradients, a number of resins were screened for their ability to bind rAAV-based vectors or exclude rAAV-based vectors from a slip compared with the ability to bind or skip commonly observed technological impurities, including host cell DNA, helper virus, serum albumin, serum proteins (if serum in is included in the production environment) and other low molecular weight proteins found in product cultures to develop a commercially scalable, orthogonal, and efficient rAAV purification method.

Resin screening was performed using a 1 ml column (carrier layer height 5 cm) with linear flow rates recommended by the supplier for each resin, with significant overload in capacity relative to the manufacturer's recommendations. Tracking the spectrophotometric signal of the optical density of the ultraviolet at 280 nanometers (A ₂₈₀ ) was recorded for binding and elution with each resin. Peaks were analyzed by suitable analysis for a vector based on rAAV-1 and typical process impurities. The data presented in figure 2, is a typical tracking spectrophotometric signal for a typical resin from the list. Table 2 lists some of the resins screened and the relative binding affinities for the rAAV-1 based resins and various process impurities.

table 2 Resin Screening for rAAV-1 Binding Compared to Binding of Process Impurities Resin Resin type rAAV-1 Helper virus Host cell DNA Serum albumin Whey proteins UnoS pH 5.5 Cation exchange + 0 - + + UnoS pH 7.0 Cation exchange - - - - - UnoQ pH 8.5 Anion exchange + ++ ++++ ++ ++ UnoQ pH 7.0 Anion exchange + +++ ++++ ++ ++ Fractogel EMD SO ₃ Cation exchange + ++ - 0 0 Cht Apatite + 0 ++ + + Cft Apatite + 0 ++ + + Hic phenyl Hydrophobic - - - 0 0 HIC Butyl Hydrophobic + ++ - 0 0 HIC Hexyl Hydrophobic + ++ - 0 0 HIC PPG Hydrophobic - - - 0 0 Source s Ion exchange + 0 0 ++ 0 Source q Ion exchange + 0 0 ++ 0 TMAE Ion exchange + 0 0 ++ 0 IMAC FeCl ₃ - 0 0 - + + Superdex 200 Gel filtration Zero Zero crowding out crowding out crowding out Hw55 Gel filtration Zero Zero crowding out crowding out crowding out Hw65 Gel filtration Zero Zero crowding out crowding out crowding out Hw75 Gel filtration Zero Zero crowding out crowding out crowding out Explanation: (-) = present in the slip (no binding); + = weakly bound (very easily eluted in the gradient); ++ to ++++ = stronger binding (further eluted along the gradient)

Example 5. The development of chromatography on apatite in the presence of polyethylene glycol ("PEG") to capture rAAV-1

Based on the results of the resin screening carried out in Example 4, an apatite resin or a ceramic apatite resin was selected as one of the rAAV-1 capture resins. Initial experiments were carried out using a CFT II resin, but for further purification, the resin was changed to CHT I, as discussed in detail below. The data showed that both chromatographic resins functioned equivalently. Experiments were conducted to further increase the binding ability of apatite resin in relation to rAAV-1 and to improve the ability of the resin to separate particles of rAAV-1 and other process impurities. Two key improvements in the function of apatite resins were further developed as described herein: 1) addition of PEG; and 2) development of conditions related to the load buffer.

Based on the variable slip of the apatite column due to capacity issues at commercially reasonable column sizes, PEG was mixed with the concentrate after TFF (see Example 3 above) before loading onto the apatite resin to increase the binding of the rAAV-1 based vector to other technological impurities, such as serum albumin, helper virus and other protein impurities, which competed with the rAAV-1-based vector for binding to the column in Example 4.

Chromatography on apatite in the presence of PEG is an effective capture or binding strategy for the purification of rAAV-1-based vectors, although many process impurities are also retained by the apatite resin at pH 7.0.

Experiments were conducted to determine whether the modification of the buffer conditions could improve the separation of rAAV-1 from other process impurities. Small scale experiments were performed using an AKTAexplorer FPLC system (GE Healthcare, Piscataway, NJ) equipped with 1.2 ml Tricorn 5 columns (GE Healthcare) packed with CFT resin with a carrier layer height of 6 cm, and the system was started at a flow rate equal to 150 cm / h. These columns were evaluated for the capture of a vector based on rAAV-1 relative to the binding of bovine serum albumin ("BSA"), a model low molecular weight impurity, in various buffer systems in the presence or absence of 5% PEG6000.

rAAV-1 or BSA was injected in small volumes (<5% of the total volume) onto a CFT column in the test buffer system, in the presence or absence of 5% (w / v) PEG6000. Small volumes were used to eliminate the need for buffer exchange in the samples. Products were eluted using a gradient of 500 mM PO ₄ . 50 mM 2- (N-morpholino) ethanesulfonic acid (“MES”) was used to buffer the system at pH 6.50, and 20 mM borate was used to buffer the system at pH 9.0.

The data presented in table 3 demonstrate that the binding of the vector based on rAAV-1 with apatite resin is essentially the same at pH 6.5 or pH 9.0 in the presence of 5% (w / v) PEG6000, whereas the binding of BSA, a model low molecular weight impurity, decreased significantly at pH 9.0 in the presence or absence of 5% (w / v) PEG6000, as indicated by spectrophotometric monitoring (data not shown). Further analysis of the reduced ability of BSA to bind to apatite resin under basic loading buffer conditions (ie, pH 9.0) by enzyme-linked immunosorbent assay ("ELISA") showed that the majority of BSA (~ 78%) under pH 9 buffer conditions, 0 was in slip, while an additional level of release or decrease in BSA binding could be achieved during subsequent washing steps (~ 19%), so that only ~ 0.1% of the loaded BSA remained on the column, actually bound to the apatite resin and eluting together with vector on about rAAV-1 again. The rAAV-1 particles were stable at pH 9.0, as indicated by the absence of a loss of infectivity or the absence of a decrease in the number of eluted DNase-resistant particles (“DRP”).

Table 3 Relative binding strength of rAAV-1 and BSA with apatite resin at pH 6.5 or pH 9.0, in the presence or absence of 5% (w / v) PEG6000 Binding Conditions BSA 5% (w / v) PEG6000 pH 6.50 ++ + pH 6.50 + 5% (w / v) PEG6000 ++ ++++ pH 9.0 + + pH 9.0 + 5% (w / v) PEG6000 + ++++ Explanation: + = weak binding; ++ = average binding; ++++ = strong binding

Example 6. The effect of serum in the culture of collecting rAAV-1 on the capture of rAAV-1 by chromatography on apatite in the presence of PEG

RAAV-1-based vector production cultures or rAAV-1-containing feed streams to be purified by the methods described herein may contain serum and whey proteins if product cultures were grown in a medium containing serum. While the production of an rAAV-1 based vector is described which uses very low serum concentrations (i.e., 1% or less) (see, for example, US Patent No. 6995006), concentrating the product culture or feed stream as described in example 3, can lead to the formation of a stream of source materials, effectively containing 20% serum and whey proteins as a result of 20-fold concentration of the collection of product culture. To evaluate the effect of serum components on the functionality of apatite chromatography, experiments were carried out on product cultures produced in the presence or absence of serum, in the presence or absence of PEG6000.

Two model cultures of rAAV-1 products were used to evaluate the functionality of apatite resin (CFT type I) by traditional breakthrough analysis in just four column loading experiments. In one experiment, the product culture contained serum, and in the second experiment, the product culture did not contain serum. Both feed streams were tested in the presence of 5% (w / v) PEG6000 or in the absence of PEG. The feed streams represented the collection process and were clarified culture supernatants that were passed through an anion exchange filter and concentrated 20 times by filtration in a tangential stream, as described herein. Type I CFT columns were loaded by mixing during 1: 1 chromatography a borate buffer feed stream at pH 9.0. The buffer contained 0% or 10% (w / v) PEG6000 (until a final concentration of 5% (w / v) PEG6000 was reached). Once the columns were loaded, the breakthrough was collected in a series of fractions that were analyzed for product by PCR DRP. Functionality was defined as the point at which the concentration of the product in the column effluent reached 1% of the concentration introduced into the column, after dilution was taken into account during chromatography. For column loads containing PEG, the rest of the chromatography process was started to evaluate the yield of the vector in the eluted fractions.

The data presented in FIG. 3 demonstrate that the addition of PEG6000 increases the ability of the apatite resin to bind rAAV-1 based vector, regardless of whether serum is present in the culture of the product. Without intending to relate the observed to theory, in the supernatant after TFF in the presence of PEG, rAAV-1 preferentially binds to an apatite resin compared to other technological impurities through anionic interaction with phosphate constituent groups, for which competition may occur due to the presence of phosphate in the elution buffer. In addition, the binding of rAAV-1 with an apatite resin further separates it from technological impurities due to interaction with metals, which can lead to competition in the presence of salt. The binding of rAAV-1 to the phosphate constituent groups is not primarily controlled by hydrophobicity, since capture and elution buffers are formulated to be mainly ionic in nature (i.e., the elution buffer contains 50 mM phosphate, compared to elution buffers containing 150 mM or more of phosphate, widely used in elution compositions for binding mainly due to hydrophobic interactions). Under such high salt conditions, low phosphate elution conditions, remaining helper virus, host cell DNA and other low molecular weight proteins contained in the product culture supernatant, if present, will remain on the resin. Surprisingly, these data demonstrated that, in the presence of PEG6000, rAAV-1 based vectors produced in serum-containing or serum-free media exhibit apatitic resin binding capacity of at least 1.2 × 10 ¹² DRP / ml (loading 1 ml) to more than 1.5 · 10 ¹⁴ DRP / ml resin (150 ml). In the absence of 5% (w / v) PEG6000, the binding capacity of the apatite resin for TFF collection was less than 2.4 · 10 ¹² DRP / ml for vectors produced in serum-containing medium and 7.2 · 10 ¹² DRP / ml for vectors produced in serum-free medium, while rAAV-1 based vectors were not obtained in the CFT eluate.

Example 7. Purification of rAAV-1 by Apatite Chromatography

A Type I CHT column was filled with 2 M NaCl and disinfected with 1 M NaOH. Before loading, 20 mM borate (pH 9) + 5% (w / v) PEG6000 was balanced with 6 column volumes (“CV”). The feed concentrate after TFF was loaded via 3 mm BioProcess Skid (GE Healthcare) onto a 923 ml CHT column (diameter 14 cm × support height 6 cm), prepared as described previously at a flow rate of 96 cm / h. The starting material concentrate after TFF was mixed during chromatography with an equal volume of 40 mM borate buffer (pH 9) + 10% (w / v) PEG6000 buffer to obtain a final concentration of 20 mM borate (pH 9) + 5% (mass. / rev.) PEG6000.

A series of 4 successive washes was performed to remove process impurities while retaining the rAAV-1 based vector on the column. Washing 1 (“displacement”) was carried out by mixing during chromatography 5 CV 50:50 (volume: volume) 20 mM borate (pH 9.0) + 5% (w / v) PEG6000: 40 mM borate (pH 9 , 0) + 10% (w / v) PEG6000 to displace the PEG6000 of all loaded solutions. It was further found that this step preferably increases the binding affinity of rAAV-1 based vectors. Rinse 2 was performed with 15 CV 150 mM potassium phosphate + 20 mM borate (pH 9) + 5% (w / v) PEG6000 to remove most of serum albumin and other low molecular weight protein impurities while retaining rAAV-1 on the column. Rinse 3 (“WII” in FIG. 5) was performed with 15 CV 20 mM borate (pH 9) + 5% (w / v) PEG6000 to remove any remaining phosphate, so that rAAV-1 remained bound to the column, while PEG6000 removed. Rinse 4 (“WIII” in FIG. 5) was carried out with 5 CV of 20 mM HEPES buffer (pH 7.0) +150 mM NaCl to remove PEG6000 and to adjust the salt concentration, thereby ensuring separation of rAAV-1 and any remaining helper virus or other technological impurities, such as protein impurities, which may remain bound to the column.

The rAAV-1-based vector was eluted from column 6 CV of 50 mM potassium phosphate + 20 mM HEPES (pH 7.0) +150 mM NaCl buffer. Figure 4 shows a typical spectrophotometric plot of the UV absorbance at 280 nm (A ₂₈₀ ) and conductivity for the chromatographic procedure on CHT I. Figure 5 shows the relative degree of purification of rAAV-based vectors eluted from apatite resin.

Glucan Removal by Apatite Chromatography

Glucans are carbohydrates similar to cellulose that seep into the process medium from a cellulose-based volumetric filter used to collect rAAV-1 particles from product cultures. Glucans at concentrations above ≈1 ng / ml may interfere with standard tests for Limulus amebocyte lysate (“LAL”) tests for contamination with bacterial endotoxins. As shown in Table 4 below, a Type I apatite column removed ~ 2.5 orders of glucan from the product culture. Under the buffer conditions described herein, a large majority of glucans were present in the breakthrough and did not bind to the column.

Table 4 Removing glucans in the process Processing stage Glucan concentration (ng / ml) Total amount per batch (ng) (scale 250 l) Lightening 10,4 2,600,000 Tff 136 1541016 Elution with CHT 1.9 4769 After HIC and SEC 0.2 662 The final eluate from the ion exchanger 0.1 71

Samples were analyzed for glucans using a LAL-based kinetic chromogenic analysis specific for glucans (Glucatell®, Cape Cod, MA).

Removal of helper virus Ad5 by chromatography on apatite

To confirm that chromatography on CHT removes Ad5 from the feed stream, a preliminary study with the introduction of an internal standard was performed using the feed stream from the final preceding process. The levels of added Ad5 were determined based on data from a phase I virus removal study with CFT II resins, and three different feed ratios of the feed stream were used: 6.6 ml; 13.5 ml; and 33 ml of feed stream after TFF per ml of CHT resin.

The data are presented in table 5, the removal of Ad5 by CHT was comparable to the removal of 4 LRV and showed approximately removal of the Ad5 virus by about 4 orders of magnitude, and, as it turned out, did not depend on the volume of the loaded feed stream within the estimated 5 orders of magnitude. The low yield of Ad5 is consistent with previous data indicating that under the buffer conditions used, Ad5 binds to apatite resin more strongly than rAAV-1.

Table 5 The total number of Ad5 infection units in the fractions of the column with CHT Fraction Ratio 6.6 ml 13.5 ml 33 ml Loading 9.0 · 10 ⁸ 1.310 ⁹ 9.0 · 10 ⁸ Leakage + Extrusion <2.5 · 10 ⁵ <3.6 · 10 ⁵ <6.910 ⁵ Flushing PO ₄ 2.9 · 10 ⁶ 2.7 · 10 ⁶ <2.6 · 10 ⁶ Flushing II and III 3.6 · 10 ⁶ 2.210 ⁶ 2,810 ⁶ Elution <6.910 ⁴ <6.910 ⁴ <3.5 · 10 ⁵ Decrease Order Value (LRV) > 4.1 > 4.3 > 3.4

A total of 8 · 10 ⁹ Ad5 infectious particles (i.e., the total number of particles with P: I ~ 10) were added to different volumes of the feed stream after TFF to run on 1.2 ml CHT columns. Column chromatography was performed at 100 cm / h, and fractions were collected for analysis of the DRP vector and Ad5 by analysis of the infectious titer. A standardized version of the Ad5 infectivity assay was used, since high concentrations in loading and elution samples with CHT are known to interfere with the cell-based assay. Ad5 removal was determined as the value of the reduction order (“LRV”), calculated as the logarithm of the total amount of loaded Ad5 divided by the total amount of Ad5 obtained in the elution fraction (value of the reduction order).

Example 8. Inactivation of the remaining helper virus by heating

The heat inactivation step was performed to inactivate and remove any remaining helper virus present in the eluate with CHT I. For small-scale experiments, the eluate CHT I was divided between two 1 L PETG (Nalgene®) bottles and MgCl _{2 was} added to a final concentration of 2 mM to increase the stability of the rAAV-1 based vector. The bottles were incubated in a water bath at 53.5 ° C with stirring until the temperature in the bottle reached approximately 52 ° C. The bottles were then cooled by transferring to an ambient temperature water bath, and stirred until the temperature reached 5 ° C above ambient temperature. The heat-treated mixture was filtered through a 0.22 μM Opticap® 4-inch filter membrane (Millipore Opticap®, catalog number KVSC04HB3). Alternatively, for experiments on a larger scale, the eluate with CHT I was inactivated by heating in a sterile disposable biotechnological container (5 L Custom Hyclone container, CX5-14 film) on a swing platform with temperature control, at a set temperature of 53 ° C, swing speed of 40 rpm min and a mixing angle of 12 ° (pan of a wave heater with a volume of 20 l). The CHT I eluate was incubated on a platform until a temperature of 52 ° C was reached, and then left for an additional 10 minutes. To stabilize rAAV-1, MgCl _{2 was} added during heating to a final concentration of 2 mM. After heating, the product was filtered through a 0.2 μm filter and incubated overnight at ambient temperature to minimize possible temperature effects with respect to the subsequently used hydrophobic interaction column.

Example 9. Capture of rAAV-1 by hydrophobic interaction chromatography ("HIC")

HIC is a method for separating biomolecules based on differences in hydrophobicity of their surface. As such, the HIC is considered to be orthogonal to the other stages of rAAV-1 purification in the process. HIC media contain hydrophobic ligands, such as linear chain hydrocarbons (e.g. propyl (C3), butyl (C4), hexyl (C6), or octyl (C8)) or aromatic hydrocarbons (e.g. phenyl). In pure water, the hydrophobic effect is too weak for functional interactions between the ligand and proteins, or between the proteins themselves. However, lyotropic salts enhance hydrophobic interactions, and the addition of such salts directs protein adsorption to the HIC medium. For this reason, HIC resins are typically charged at high salt concentrations and eluted from them at lower salt concentrations.

HIC Chromatography with Ammonium Sulfate Buffers

Briefly, a 170 ml butyl HIC column (6 cm diameter × 6 cm carrier layer height) (Toyopearl® Butyl 650M; Tosoh Biosciences, Montgomeryville, PA; catalog number 14702) was disinfected with several volumes of a 0.5 M NaOH column and balanced with a mixture of 75:25 (volume: volume) 2 M ammonium sulfate + 50 mm Bis Tris (pH 7.0): 50 mm Bis Tris (pH 7.0). The heat-treated rAAV-1 vector eluate from apatite was charged at a rate of 3.3 L / h while mixing during chromatography at 75:25 (volume: volume) 2 M ammonium sulfate + 50 mM Bis Tris (pH 7.0) : eluate rAAV-1 with apatite. Chromatographic mixing prevents the risk of any precipitation of the rAAV-1-based vector by the ammonium sulfate present in the buffer. The column was washed with one or more volumes of a 75: 25 mixture column (volume: volume) buffer 2 M ammonium sulfate + 50 mM Bis Tris (pH 7.0): buffer 50 mM Bis Tris (pH = 7.0) + 10% propylene glycol ( volume: volume) (EMD Biosciences). Propylene glycol in this example was added to buffers to narrow the elution profile compared to a wide elution profile without propylene glycol, although this is not necessary in this process. The rAAV-1-based vector was eluted from the column with 800 mM ammonium sulfate buffer + 50 mM Bis Tris (pH 7.0) + 4% propylene glycol. Under the elution conditions used, any remaining helper viruses and proteins present in the loading solution will remain bound to the column.

Many methods for producing rAAV-1 require strict disinfection before disposal due to the fact that the product is a viral vector and due to the use of live type 5 adenovirus (Ad5) as a helper virus in the production. Liquid wastes from chromatographic operations, as a rule, are first disinfected with bleach at the place of use, and then subjected to further disinfection by incubation at high pH before neutralization and disposal. Ammonium sulfate present in HIC buffers reacts with bleach and sodium hydroxide to release hazardous chlorine and ammonia gases, respectively. Thus, the initial consideration for optimizing the HIC chromatography step was the development of a suitable buffer system that can be safely disinfected by methods known in the art.

Screening for suitable buffers for binding rAAV-1 to HIC column

RAAV-1 based vectors were loaded onto the columns under a number of different buffer conditions, and the relative binding efficiency was determined by measuring the amount of rAAV-1 vector present in the slip fraction (Table 6). Evaluated buffers included high concentrations of lyotropic salts traditionally used with HIC chromatographic processes and several low pH buffers in which mixed mode interaction (HIC / cation exchange) could potentially occur. Both Tosoh Butyl 650M and EMD Phenyl resins linked vectors in some of the alternative buffers.

High salt concentration buffers used in hydrophobic interaction chromatography should then be screened for viscosity, which can result in high back pressure, which can limit flow rates or cause mixing problems, thereby increasing the risk of product precipitation due to salt crystallization in buffers occurring at temperatures that can be used for storage or operation. Based on the data shown in Table 6 below and after considering the factors described above, 1 M sodium citrate, pH 7.0, was chosen as the optimal buffer for binding vectors based on rAAV-1 to HIC chromatographic resin.

Table 6 AAV1 binding screening for alternative buffers Tested Buffer System Butyl 650M (% in slip) EMD-Phenyl (% in slip) 1.1 M sodium sulfate (pH 7) 0% 0% 1 M sodium citrate (pH 7) 0% 0% 1.3 M potassium phosphate (pH 7) 3% 3% 2.9 M NaCl, 50 mM sodium citrate (pH 4) 0% 28% 1 M glycine, 50 mM sodium citrate (pH 4) four% one% 50 mM potassium phosphate (pH 4.5) 3% NT 50 mM sodium citrate (pH 4) four% NT 2.9 M NaCl, 50 mM potassium phosphate (pH 4.5) 17% NT

The experiments were carried out at ambient temperature on Tricorn 5/50 columns (carrier layer height 6 cm, column volume 1.2 ml) using purified rAAV-1. The columns were balanced with the listed buffers, and ≈2 · 10 ¹¹ DRP or rAAV-1 was loaded onto the column. rAAV-1 was eluted in a linear gradient with a volume of 20 CV from 145 mM Bis Tris (pH 7.0), 10% (v / v) propylene glycol. The slip was collected and analyzed by the DRP test for the rAAV-1 fraction applied to a column that was washed and not bound.

Further characterization was carried out by investigating the removal of process impurities on a HIC column in various buffers that demonstrated good binding of rAAV-1 in a previous experiment. The data in Table 7 below for linking model impurities demonstrate that adenovirus and DNA, if present in the feed stream, are effectively separated in the HIC chromatography step.

Table 7 Relative binding of rAAV-1 compared to model process impurities in various HIC buffers Buffer system rAAV-1 Ad5 DNA 0.1 M sodium sulfate + Bis-Tris buffer, pH 7.0 + ++ - 0.8 M sodium sulfate bis Tris buffer pH 7.0 + ++ 0 1 M sodium citrate, pH 7.0 + ++ 0 2.9 M NaCl (50 mM sodium citrate for buffer at pH 4.0) - - 0 Explanation: “0” = no binding, material present in the slip; “-” = very weak binder; “+” = Strong binder; “++” = stronger binding agent

The experiments were carried out at ambient temperature on Tricorn 5/50 columns (carrier layer height 6 cm, column volume 1.2 ml). The columns were balanced with the listed buffers, and these samples were loaded onto the column. Each sample was eluted in a linear gradient of 20 CV. Samples were collected and analyzed for related material loaded.

HIC Chromatography with Sodium Citrate Buffers

The heat-treated rAAV-1 vector eluate from apatite was then loaded onto a HIC butyl column to further reduce the level of remaining process impurities and as a concentration and desalination step. A butyl column (6 cm diameter × 8.9 cm height of carrier layer) for a 373 ml HIC (Tosoh Biosciences Toyopearl® Butyl 650M, catalog number 14702) was decontaminated with several different volumes of a 0.5 M NaOH column and 5 CV mixture 75 was balanced: 75: 25 (volume: volume) 1 M citrate + 20 mm sodium phosphate: 20 mm sodium phosphate. The heat-treated rAAV-1 vector eluate with CHT I was loaded at a speed of 106 cm / h while mixing during chromatography at 75:25 (volume: volume) 1 M citrate + 20 mM sodium phosphate: CHT I eluate. Mixing by chromatography prevents the risk of any precipitation of the rAAV-1-based vector by the ammonium sulfate present in the buffer. The column was washed with 5 CV of a mixture of 75:25 buffers (volume: volume) 1 M citrate + 20 mm sodium phosphate: 20 mm sodium phosphate. The rAAV-1-based vector was eluted from a column with 6 CV 0.35 M citrate + 20 mM sodium phosphate. The column was then washed with 3.5 CV of 20 mM sodium phosphate buffer. In this wash with a low salt concentration (20 mM sodium), the fraction of rAAV-1-based vector particles elutes, which differs in the hydrophobicity of its elution profile from rAAV-1-based vector particles, which elute in a buffer with a higher salt concentration. In fact, if a fraction eluted with a low salt concentration was isolated and re-applied onto a HIC column under the described conditions, this population of the vector also eluted only in the low salt fraction, indicating that this fraction was not the result of a break in the column capacity. An infectivity analysis suggests that this rAAV-1 fraction is probably a population containing empty capsids, partially denatured capsids, less infectious capsid material, and partially full capsids. Thus, this observation may lead to improvements in the separation of rAAV-1 particles, which are less infectious and are thus less required as the material of the product. Under the elution conditions used, any helper viruses and proteins present in the loading solution will remain bound to the column and thus must be present in the low salt band.

Example 10. The exchange of buffer by gel filtration ("SEC")

The exchange of the buffer by gel filtration provides additional removal of proteins that are able to pass through the pores in resins in size, and is relatively fast in the time required for the exchange of buffers. The exchange of the buffer carried out at this stage should ensure that the HIC eluate from the previous stage has changed the buffer to anion exchange chromatography suitable for rAAV-1 binding in the final stage of the process. 3.2 L (diameter 14 cm × carrier layer height 21 cm) Amersham Superdex® prep resins 200 grade (Amersham / GE Healthcare, Piscataway, NJ; catalog number 17-1043-04) were packaged and prepared by disinfecting 2 M NaCl + 1 M NaOH, and 2.8 CV of 20 mM NaCl + 20 mM Tris (pH 8.0) was balanced. The HIC eluate was subdivided for SEC treatment in three consecutive cycles of approximately 400 ml each, with a loading of not more than 12.5% of the SEC column volume for each cycle. Product peaks (contained in void volumes) from three SEC cycles were collected in one biotechnological container. The eluate after HIC was applied to the column at a flow rate of 49 cm / h. The preceding solvent was displaced from the column and washed with 1.4 CV 20 mM NaCl + 20 mM Tris (pH 8.0), and the rAAV-1 based vector present in the eluate after HIC was in the void volume of the column. After collecting the void volume, as described, the second and third fractions were loaded, and collected from the same column sequentially as previously described for the first fraction.

Example 11. Additional agent (removal of viruses)

As an optional process for removing additional viruses that may be present as trace impurities and thus contribute to a commercially reasonable orthogonal process, a virus removal filter is introduced into the process. Examples of such filters are known in the art and include Millipore Viresolve® NFR (50 nm), Pall Ultipore® VF (50 nm) and Asahi 70 nm. Millipore Viresolve® NFR virus filter (Millipore 4 "Virosolve NFR filter, catalog number KZRV 04T C3) was prepared according to the manufacturer's instructions, rinsed with 20 mM NaCl + 20 mM Tris (pH 8.0) and the eluate after SEC was filtered through a membrane. Filter rinsed with several volumes of 20 mM NaCl + 20 mM Tris-HCl (pH 8.0) and this solution was combined with the filtered eluate after SEC.

Example 12. Anion exchange chromatography

The second stage of anion exchange vector capture based on rAAV-1 vector was performed as the final stage of concentration and and finalization on Unosphere® Q resin (Biorad, Hercules, CA). A 373 ml column (diameter 8.9 cm × support height 6 cm) was disinfected with several volumes of a 0.5 M NaOH column and equilibrated with 7 CV 20 mM NaCl + 20 mM Tris buffer (pH 8.0). A SEC void volume fraction or an optionally virus-filtered eluate was charged at a speed of 309 cm / h. The column was washed with 10 CV 60 mM NaCl. The ionic strength of the wash solution was chosen to keep rAAV-1 bound to the resin, washing away other process impurities, such as glucans, which could be leaked from various filters used in the cleaning steps. RAAV-1 based vector from Column 6 CV 130 mM NaCl. The ionic strength of the saline eluate with 130 mM NaCl flushes rAAV-1 off the column, while any remaining trace impurities, such as serum albumin or helper virus, remain bound.

Figure 6 compares the degree of purification after various stages of the process by SDS-PAGE. Process-related samples from a representative collection of product culture were separated on a denaturing / reducing 10% polyacrylamide gel and stained with Sypro orange. All samples after collection were loaded at 1 · 10 ¹⁰ DRP / lane. Two samples from the early stages obtained prior to the concentration step by TFF (initial clarification and anion exchange (“AEX”) slip steps) could only be loaded at 1 · 10 ⁹ DRP / lane due to volume limitations of the gel. Beta-galactosidase (B-Gal) was loaded at 50 ng / lane to assess the sensitivity and uniformity of gel staining. Three capsid proteins AAV1 (VP1, 2 and 3) are indicated.

Example 13. The percentage output of rAAV during cleaning

The data shown in FIG. 7 represents the percentage yield of rAAV infectious particles after each stage of the process according to a purification scheme from a representative culture of rAAV-1-based vector production. Percentage yield was calculated based on the total number of DRP vectors based on rAAV-1 obtained at each stage of the process divided by the total number of DRPs loaded or subjected to this stage of purification. The data demonstrate that at each stage of the cleaning process, a yield of approximately 60% or higher was achieved. In numerous experiments, the yield interval from each stage of the process was at least 60% to 90%. It is especially noteworthy that the yield interval at the capture stage (i.e., at the apatite chromatography stage) in individual experiments varied from 57% to more than 90%. In addition, the output interval of the HIC stage varied from 60% to 80%.

Although the above invention has been described in some detail by way of illustration and example for purposes of clarity of understanding, those skilled in the art will appreciate that some minor changes and modifications will be applied. Thus, the description and examples should not be intended to limit the scope of the invention.

Claims (34)

1. A method of isolating a population of particles of a recombinant adeno-associated virus (rAAV) from technological impurities in a feed stream, comprising the steps of:
(a) bringing the feed stream containing rAAV particles into contact with apatite chromatography medium in the presence of polyethylene glycol (PEG), where rAAV particles bind to the apatite chromatography medium; and
(b) eluting the rAAV particles bound to the apatite chromatography medium using an elution buffer containing less than 3% (w / v) PEG. 2. The method of claim 1, wherein the apatite chromatography medium is ceramic hydroxyapatite (CHT). 3. The method of claim 1, wherein the apatite chromatography medium is ceramic fluorapatite (CFT). 4. The method of claim 1, wherein the specific binding of the apatite chromatography medium to rAAV particles is characterized by 10 ¹⁴ to 10 ¹⁶ DNase-resistant particles per milliliter (DRP / ml). 5. The method of claim 1, further comprising the step of binding the rAAV particles in the feed stream eluted from the apatite chromatography medium to the anion chromatography medium. 6. The method according to claim 1, wherein the feed stream containing rAAV particles in step (a) is contacted with apatite chromatography medium in the presence of polyethylene glycol (PEG) and a base buffer. 7. The method according to claim 6, in which the main buffer is characterized by a pH between 7.6 and 10. 8. The method according to p. 6, in which the main buffer is characterized by a pH between 8.0 and 10.0. 9. The method according to claim 6, in which the main buffer is characterized by a pH between 9.0 and 10.0. 10. The method according to claim 6, in which the main buffer contains borate. 11. The method of claim 1, wherein the PEG has an average molecular weight of between about 5000 (PEG5000) grams per mole and about 15000 (PEG15000) grams per mole. 12. The method of claim 11, wherein the PEG has an average molecular weight of approximately 6,000 (PEG6000) grams per mole. 13. The method according to claim 1, wherein the feed stream containing rAAV particles in step (a) is contacted with apatite chromatography medium in the presence of between about 3% (w / v) and about 10% (w / w) ./ob.) PEG. 14. The method of claim 13, wherein the feed stream is contacted with apatite chromatography medium in the presence of approximately 5% (w / v) PEG6000. 15. The method of claim 13, wherein the feed stream is contacted with apatite chromatography medium in the presence of approximately 10% (w / v) PEG6000. 16. The method according to claim 1, further comprising the step of washing the chromatography medium on apatite with a washing buffer after the feed stream is brought into contact with the chromatography medium on apatite, but before elution of the rAAV particles from the chromatography medium on apatite. 17. The method of claim 16, wherein the apatite chromatography medium is washed one or more times with wash buffer containing approximately 7.5% (w / v) PEG and / or wash buffer containing approximately 5% ( mass./about.) PEG. 18. The method of claim 17, wherein the apatite chromatography medium is further washed with a wash buffer containing less than about 3% (w / v) PEG and / or a wash buffer not containing PEG. 19. The method of claim 16, wherein the wash buffer comprises a buffer selected from the group consisting of borate, N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid (HEPES), and Tris-HCl. 20. The method according to p. 16, in which the buffer for washing is characterized by a basic pH. 21. The method according to p. 20, in which the buffer for washing contains borate at a pH between approximately 8.0 and approximately 10.0. 22. The method according to p. 21, in which the buffer for washing contains borate at a pH of approximately equal to 8.0. 23. The method according to p. 21, in which the buffer for washing contains borate at a pH of approximately 9.0. 24. The method of claim 21, wherein the wash buffer contains borate at a pH of approximately 10.0. 25. The method according to p. 20, in which the buffer for washing additionally contains from 100 to 500 mm phosphate. 26. The method according to p. 20, in which the buffer for washing additionally contains from 50 to 250 mm NaCl. 27. The method of claim 1, wherein the rAAV particles bound to the apatite chromatography medium are eluted with an elution buffer containing low concentrations of PEG or not containing PEG. 28. The method of claim 27, wherein the elution buffer comprises a buffer selected from the group consisting of borate, N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid (HEPES), and Tris-HCl, at a neutral pH. 29. The method of claim 27, wherein the elution buffer contains less than about 3% (w / v) PEG6000. 30. The method of claim 29, wherein the elution buffer further comprises less than 100 mM phosphate. 31. The method of claim 30, wherein the elution buffer further comprises 50 mM phosphate. 32. The method according to p. 31, in which the buffer for elution additionally contains from 50 to 250 mm NaCl. 33. The method of claim 1, wherein the rAAV particles comprise an AAV capsid protein from an AAV capsid serotype selected from the group consisting of AAV-1, AAV-2, AAV-3, AAV-4, AAV-5, AAV-6 , AAV-7, AAV-8, AAV-9, AAV-10, AAV-11, AAV-12, AAV-13, AAV-14, AAV-15 and AAV-16. 34. The method of claim 33, wherein the rAAV particles comprise an AAV capsid protein from an AAV capsid serotype selected from the group consisting of AAV-1, AAV-4, AAV-5, and AAV-8.

Images