EP1141249A1

EP1141249A1 - Method for separating viral particles

Info

Publication number: EP1141249A1
Application number: EP99964739A
Authority: EP
Inventors: Francis Blanche; Anne Barbot; Béatrice Cameron
Original assignee: Aventis Pharma SA
Current assignee: Centelion SAS
Priority date: 1998-12-31
Filing date: 1999-12-30
Publication date: 2001-10-10
Also published as: WO2000040702A1; US20020037565A1; AU779267B2; KR100762033B1; CN1332796A; JP2003523169A; NZ512504A; US20030170710A1; KR20020044106A; IL143381A; AU3049300A; BR9916654A; CN1195057C; HUP0104794A3; US6537793B2; CA2356373A1; IL143381A0; NO20013271D0; PL349074A1; HUP0104794A2

Abstract

The invention concerns a novel method for purifying and quantifying viral particles. More particularly, the invention concerns a method for purifying and quantifying adenovirus by ion-exchange chromatography. The invention also concerns a method for identifying different adenovirus serotypes.

Description

METHOD OF SEPARATING VIRAL PARTICLES

The invention relates to a new method for the purification and quantification of viral particles. More particularly the invention relates to a method of purification and quantification of adenovirus by ion exchange chromatography. The invention also relates to a method for identifying the various adenovirus serotypes.

Gene therapy is currently undergoing remarkable development and various clinical studies in humans have been ongoing since the first trials in 1990. Among the commonly used methods for transferring genes, viral vectors ^s' are particularly promising, and among - adenoviruses occupy a prominent place.

The development of adenovirus vectors in gene therapy requires access to two types of technology which are currently limiting for the production of viral stocks: the first is to have a rapid method, with high sensitivity and very selective for the quantification of the viral particles in the samples coming from the stages of construction and amplification of the virus considered, this point is particularly important for the optimization of the production process of the viral stocks: the second is to have a purification process reliable, reproducible, simple and easily extrapolated on an industrial scale for the purification of viral particles.

The production of clinical batches of adenovirus remains a long process due to the number of transfection and amplification steps whose productivity is not optimized. Recombinant adenoviruses are usually produced by introducing ^viral DNA into an encapsidation line. followed by ^a mechanical or chemical lysis of the cells after about two or three days of culture (the kinetics of the adenoviral cycle being 24 to 36 hours). According to another variant, the culture is continued for a longer period (8 to 12 days), and the viruses are harvested directly in the supernatant, after spontaneous release by a phenomenon of autolysis of the packaging cells (WO98 / 00524) .

Generally between 2 and 7 amplification cycles are necessary to build up viral stocks. An important limitation to the optimization of the viral stocks production process lies in the particle titration methods viral. Indeed, biological methods are relatively sensitive and precise methods but particularly long to implement (approximately 4 to 15 days depending on the dosage used ie activity of the transgene (tdu) or obtaining range (pfu)). Faster analytical methods have been developed but these do not have a sufficient degree of precision and sensitivity when the titrations of viral particles must be carried out, without prior purification, in lysates, crude cell extracts or culture supernatants . This is the reason why the successive amplification cycles are carried out with multiplicities of infection (MOI) estimated approximately. As a result, the amplification steps are not very reproducible, sometimes even longer and / or more numerous than would be necessary with an optimized process. The rapid and precise determination of the titles of adenovirus solutions would make it possible to adjust the multiplicity of infection of each step so as to optimize the entire production process of adenovirus stocks.

The quantification method for viral particles must satisfy several conditions. First, it must be sensitive enough to ensure the determination of viral particles in diluted preparations or with a low titer (typically <1 x 10 ° viral particles per ml (w / ml)) without resorting to an enrichment step prior. The dosage of viral particles must be carried out directly in lysates or crude preparations, without ^it being necessary to perform a step of purification or pretreatment. In addition, this method must ensure high selectivity to overcome possible interference with the many compounds present in the lysates or crude cell extracts, the proportions of which can vary depending on the culture conditions.

A quantitative analytical method based on anion exchange chromatography has been described in the literature (Huygue et al., Human Gene Ther. 6: 1403-1416, 1995; PW Shabram et al., Human Gene Ther. 8: 453-465. 1997). This method, which has a detection limit of the order of 1 x 10 ^s pv / ml, is applicable to the titration of purified viral preparations. However, the sensitivity of this method decreases when the analysis is carried out on lysates or crude cell extracts. The detection limit is estimated at 2 to 5 x 10 ° pvml in such samples and this method does not make it possible to quantify the adenoviral particles in very diluted and unpurified preparations such as lysates of infected cells during the transfection steps and ^{amplification} of the virus for which the Adenoviral titer is typically of the order of 1 x 10 ^s pv / ml to 1 x 10 ° pv / ml. In addition, this method also does not make it possible to quantify the adenoviral particles from preparations obtained in certain production media devoid of animal proteins. Indeed, such media contain, at the end of culture, compounds of the type, sugars, amino acids, vitamins, or phenol red, etc. among which some ^of them may interfere with adenovirus particles during the quantification of viruses and lead to overestimate largely the title of the preparation. Finally, the chromatographic method reported by Shabram et al. requires a sample pre-treatment with nuclease broad spectrum of ^activity (Benzonase) to remove nucleic acids that interfere with the detection and measurement of particles.

Regarding the preparative separation methods of adenoviruses, chromatography has been used for many years for the purification of adenoviral particles [Haruna, I., Yaosi, H., Kono. R. and Watanabe, I. Virology (1961) 13. 264-267; Klemperer, H. G. and Pereira, H. G. Virology (1959) 9, 536-545; Philipson, L., Virology- (1960) 10, 459-465]. More recently, methods describing the large-scale purification of recombinant adenoviruses have been described (international patent applications WO96 / 27677, WO97 / 08298, WO98 / 00524, WO98 / 22588).

Application WO98 / 00524 describes in particular a purification method using the strong anion exchange resin Source 150 ^and which makes it possible to obtain, in a single chromatographic step. adenovirus preparations whose purity is at least equivalent to that obtained from preparations purified by ultracentrifugation in cesium chloride gradient. This degree of purity is very high and reaches the standards required for clinical studies in humans (WHO Expert Committee on Biological Standardization, Forty-ninth Report. WHO Technical Report Series, WHO Geneva, in press).

However, when the virus titer preparations purified is low (eg in the case of an adenovirus with low productivity, or when the cleaning should be done from a stock obtained at ^a stage ^of early amplification) , or even when the virus production medium leads to the presence of compounds co-eluted with the adenovirus (as for example in the case of media lacking calf serum), the limited performances of the chromatographic techniques described above do not allow either of quantify, nor purify adenoviral particles in a single step from such starting material.

The problem therefore arises of being able to have a method for titrating viral particles from crude preparations which is at the same time rapid, sensitive and highly selective. The problem also arises of having a reliable, reproducible purification method making it possible to obtain, from these same crude preparations, and preferably in a single step, viral preparations of pharmaceutical quality.

It has now been found, and it is that which is the subject of the present invention, that certain chromatography supports have suφrenantly quite exceptional properties for the separation of viral particles and in particular of adenoviruses. These properties allow the titration and / or purification of viral particles from crude preparations, without prior treatment, with very high sensitivity and selectivity. The use of these supports also provides, unexpectedly, a simple and rapid analysis method for separating and identifying by chromatography, adenoviruses of different serotypes or adenoviruses modified at the fiber or hexon level.

The present invention relates to a method of separating viral particles from a biological medium characterized in that ^it comprises at least one chromatography step carried out on a support comprising a matrix, groups of ion exchangers, said groups being grafted onto said matrix by means of a flexible arm.

The matrix can be chosen from agarose. dextran, ^acrylamide, silica, poly [styrene-divinylbenzene], alone or in admixture. Preferably the matrix consists of agarose. more preferably, it is about 6% crosslinked agarose.

The supports consisting of crosslinked agarose beads onto which flexible functionalized ion exchange arms are grafted have been developed for the preparative and industrial chromatography of biomolecules. These materials have especially been developed for the capture step ^(that is, the initial step of the purification process) biomolecules from crude mixtures simply released, ^that is to say, cleared of suspended solids constituents . Their performances have been optimized in the sense of a very high capacity for fixing the solutes on the support, a very low back pressure at high linear flow rate of liquid, a low cost as well as a very high chemical resistance cleaning agents used for regeneration.

Advantageously, the flexible arm is hydrophilic in nature and consists of a polymer of synthetic or natural origin. Among the polymers of synthetic origin, mention may be made of polymers consisting of monomers of polyvinyl alcohols, of polyacrylamides, of polymethacrylamides or of polyvinyl ethers. As a polymer of natural origin, mention may in particular be made of polymers of polysaccharide nature chosen from starch, cellulose, dextran and agarose. Preferably the degree of polymerization of the flexible arm is approximately 30 monomeric units, more preferably, the flexible arm is a dextran with an average molecular weight of approximately 5000 Da.

Preferably, the flexible arm is functionalized by grafting a group capable of interacting with an anionic molecule. Most generally, the group consists of an amine which can be ternary or quaternary. In the context of the present invention, it is particularly advantageous to use a strong anion exchanger. Thus, preferably used according to the invention a chromatography support as indicated above, functionalized with quaternary amines.

As a particularly preferred support for the implementation of the invention, mention may be made of Q Sepharose® XL (Amersham Pharmacia Biotech). The use of this support is mentioned in one of the examples of application WO98 / 39467. Purified adenoviruses are modified by treatment with polyethylene glycol (PEG). After reaction, the modified adenoviruses, the unmodified adenoviruses and the PEG are separated by passage through a column of Q Sepharose® XL. It is therefore a simple separation between starting products and end products of a chemical reaction. Those skilled in the art could not assume that this column could be successfully used for the separation of adenovirus from a complex biological medium containing various contaminating species (host DNA. RNAs, proteins, lipids, lipoproteins, endotoxins ...). such as a cell lysate ^of packaging. Nor did it appear on reading this document that the Q Sepharose® XL can be used for preparatory purposes because it is known that the majority of supports lose their effectiveness when large quantities of products are injected.

Other strong anion exchange supports having similar characteristics of matrix composition, particle size distribution, porosity, chemical nature of the flexible arm, graft density can be used for the preparative or analytical separation of the particles. adenovirals. Advantageously, the matrix consists of agarose crosslinked at 6%, it is grafted with flexible arms made of dextran and functionalized with strong anion exchange groups. The support has a particle size preferably between 40 and 200 μm approximately; The term "approximately" referring to the grain size signifies that the value to be taken into consideration is within a range of between +/- 20% relative to the value expressed. Preferably, this difference is between +/- 10% and more preferably it is between +/- 5% relative to the value expressed.

More particularly preferably, the particle size is between

45 and 165 μm and centered on 90 μm.

Also advantageously, the matrix has a dispersion such that 95% of the particles have a diameter between 0.1 and 10 times the average diameter of the particles, and preferably between 0.3 and 3 times the average diameter of the particles.

The Q Sepharose® XL, used in the examples which follow, non-exhaustively illustrates the performance of the supports which can be used in the context of the invention.

The Q Sepharose® XL has a size distribution of beads ranging from 45 to 165 μm, centered on 90 μm. These characteristics of size and distribution of the beads make this support a preparative type chromatographic exchanger. The theory as well as the chromatographic practice indicate that such a support has very modest performances for the separation of compounds exhibiting similar chromatographic behaviors as for the interaction of ion exchange. Similarly, such a support generates wide chromatographic peaks poorly resolved, in particular due to the large size and the very wide distribution of the beads which constitute it. These expected chromatographic characteristics are verified for biomolecules in general such as proteins, which are eluted in the form of large poorly separated peaks (see Data File Pharmacia Biotech N ° 18-1 123-82). On the other hand, and completely unexpectedly, the adenovirus particles are eluted from this type of support in the form of an extremely fine peak. very symmetrical. Compared to proteins, such as albumin, the efficiency of a column filled with Q Sepharose® XL, measured by the Height Equivalent to a Theoretical Tray (HEPT) or the number of theoretical trays per unit of column length (N / m). is 50 to 100 times higher for adenovirus (N / m: 35,000) than for proteins like bovine serum albumin (N / m: 600). See for example FIG. 1. Thus, when it is used under optimized chromatographic conditions, this type of gel and in particular Q Sepharose® XL gel, gives a chromatographic peak for the adenovirus of a fineness unmatched by the supports. generally recommended for the separation of biomolecules. Among the supports recommended for the separation of biomolecules, mention may be made of supports whose base matrix is of the divinylbenzene polystyrene type (such as, for example, the resins Source 15Q and Source 30Q, or the resins of the type Poros HQ, Poros DE2, or Poros D ). Mention may also be made of supports whose base matrix is of the ethylene glycol methacrylate copolymer type, such as, for example, Toyopearl DEAE, QAE resins. and Super Q, or the Fractogel TMAE, TMAE HiCap, DMAE, or DEAE type resins, the ion exchange functional groups of which are located on linear polymer chains of polyacrylamide type grafted onto the matrix.

The effectiveness of the supports used in the context of the present invention for the separation of adenovirus particles leads to a very high sensitivity of detection of the particles. When these supports are used in analytical chromatographic columns, the unexpected chromatographic behavior of the viral particles makes it possible to quantify the adenovirus with a detection limit much lower than the detection limit of the methods described previously. This detection limit is at least ten times lower, which makes it possible to reach a detection limit of the order of 1 × 10 ^s pv / ml in crude cell lysate type preparations and a detection limit of the order of 1 x 10 'pv / ml for purified viral preparations.

This type of support also ^ensures a very high selectivity with respect to contaminants present in the samples to be analyzed such as proteins and nucleic acids. The proteins are in the form of very broad peaks which elute well before the viral peak. Nucleic acids are eluted from the column with a salt concentration much higher than the concentration necessary for the elution of the virus. This characteristic, very different from that obtained with the chromatographic method previously described (Huygue et al., Human Gene Ther. 6: 1403-1416, 1995) makes it possible to overcome the interference of this type of compound with the viral peak. Finally, even when the preparations to be analyzed contain species co-eluted with the viral particles, the very specific form of the viral peak easily makes it possible to identify it and to carry out its quantification.

Thus, the supports used in the context of the invention make it possible to identify and quantify very easily and with great precision the peak of the adenovirus when it is analyzed from preparations containing a large quantity of proteins and d 'nucleic acids. The quantitative analysis of the particles as well as the purification is also possible from preparations obtained with very varied viral production media, or media devoid of constituents of animal origin, such as albumin, whether it is d bovine, human or another origin (Figure 2). It is also important to note that the method described in the present invention is applicable to the analysis of samples containing nucleic acids without prior treatment with a nuclease without affecting either the sensitivity or the selectivity of the method.

In this regard, another subject of the invention relates to the use of this type of support and of Q Sepharose® XL in particular, for the preparative separation or the purification of viral particles, in particular of adenovirus, from media. biological. Such a separation process can optionally comprise a prior stage of chromatography on another support such as those used in the process which is the subject of application WO98 / 00524, and in particular the Source 15Q resin. Such a preliminary step may prove to be advantageous in particular cases, for example if an excessive quantity of contaminants is present in the biological medium.

Another object of the invention relates to the use of this type of support and of Q Sepharose® XL in particular, for the quantitative analysis or the titration of viral particles, in particular of adenovirus, from biological media.

The biological medium from which is made the purification or titration of virus may be a cell supernatant ^of producing encapsidation of the virus or a cell lysate of encapsidation. or a prepurified solution of said virus. When the preparative separation or the purification of viral particles is carried out from a cell supernatant ^of producing encapsidation or a lysate, it may be advantageous to perform a preliminary step of ultrafiltration, preferably this step is carried out by ultrafiltration tangential on membrane having a cutoff threshold between 300 and 500 kDa.

The purification process according to the invention makes it possible to obtain viral preparations of high quality in terms of purity with high yields of particles (of the order of 75 to 80%) in one step, from a diluted stock. or / and very rich in contaminants and this, under production conditions fully compatible with industrial requirements and with the regulations concerning the production of therapeutic molecules.

Another object of the invention relates to a method for quantifying adenovirus, characterized in that the viral particles are separated by chromatography on a support of Q Sepharose® XL type and the amount of adenovirus is determined by measuring the absorbance of the chromatography fractions. The method of the invention allows easier and more precise monitoring of the production kinetics, directly on homogeneous samples of supernatant, without pretreatment, which allows better reproducibility and better control of the production processes of stocks of viral particles. .

The subject of the invention is also the use of a Sepharose® XL type Q chromatography support for the identification of different adenovirus serotypes. Indeed, and suφrenantly, it has been observed that this type of support makes it possible to separate and identify in a simple and rapid manner a large variety of adenoviruses of different serotypes directly from a sample of viral production medium. by determining the retention time and the absorbance ratio at 260 nm and 280 nm from the chromatographic peak.

Regarding the implementation of chromatography media in the scope of the present invention, the separation of viral particles for analytical or preparative purposes may be carried out by applying to the chromatography column a ^gradient salt elution or according to a mode isocratic, ie at constant salt concentration. For the preparative methods, the chromatographic support can be used in a chromatography column of conventional type or in a column suitable for high performance chromatography systems, using for example the Q Sepharose® XL support, or even in a so-called system. with "fluidized bed" or expanded, using for example the Streamline® Q XL support. The size of the chromatographic column is determined according to the amount of virus present in the starting material.

The viral preparation to be purified can be applied to the support in a buffer whose conductivity is such that the virus is not retained on the support while the nucleic acids are fixed. Advantageously, the conductivity is adjusted to 45 mS / cm. This particular mode of implementation then makes it possible to separate, by simple filtration through the support Q Sepharose® XL, the virus from the nucleic acids originating from the host cell contaminating the viral preparation.

The assay and purification and characterization methods of the different serotypes described in the present invention can be applied to different types of virus, and of adenovirus in particular, whether they are wild virus or recombinant viruses carrying a transgene of interest.

In addition to the foregoing provisions, the present invention also includes other characteristics and advantages which will emerge from the following examples, given by way of illustration and not limitation.

LEGENDS OF FIGURES

Figure 1: elution profile of purified adenovirus and bovine albumin on Q Sepharose® XL.

Figure 2: elution profile on Q Sepharose® XL of a viral culture supernatant obtained on a medium devoid of serum.

Figure 3: elution profile on Q Sepharose® XL of a preparation of purified adenovirus (2 x 10 ^l0 pv injected).

Figure 4: elution profile on 0 Sepharose® Fast Flow of a preparation of purified adenovirus (2 x 10 ^l0 pv injected). Figure 5: Comparison of the elution profiles of a preparation of adenovirus purified on Q Sepharose® XL and Q Sepharose® Fast Flow.

Figure 6: elution profile on Q Sepharose® HP of a preparation of purified adenovirus (2 x 10 ^l0 pv injected).

MATERIAL AND METHODS

1 - Adenovirus and production of recombinant adenoviruses defective for replication.

Adenoviruses are linear double-stranded DNA viruses about 36 kilobases in size. Their genome includes in particular a repeated inverted sequence (ITR) at each end, an encapsidation sequence (Psi), early genes and late genes. The main early genes are contained in the El. E2, E3 and E4 regions. Among these, the genes contained in the El region in particular are necessary for viral propagation. The main late genes are contained in regions L1 to L5. The genome of the Ad5 adenovirus has been fully sequenced and is accessible on the database (see in particular Genebank M73260). Likewise, parts or even all of other adenoviral genomes (Ad2, Ad7, Ad 12, etc.) have also been sequenced.

For their use in gene therapy, different vectors derived from adenoviruses have been prepared, incorporating different therapeutic genes. In each of these constructions, the adenovirus was modified so as to render it incapable of replication in the infected cell. Thus, the constructions described in the prior art are adenoviruses deleted from the E1 region essential for viral replication, at the level of which heterologous DNA sequences are inserted (Levrero et al. Gene 101 (1991) 195; Gosh-Choudhury et al. Gene 50 (1986) 161). Furthermore, to improve the properties of the vector, it has been proposed to create other deletions or modifications in the genome of the adenovirus. Thus, a thermosensitive point mutation was introduced into the mutant ts125, making it possible to inactivate the 72 kDa DNA binding protein (DBP) (Van der Vliet et al., 1975). Other vectors include a deletion of another region essential for viral replication and / or spread, the E4 region. The E4 region is indeed involved in the regulation of the expression of late genes, in the stability of late nuclear RNA, in the extinction of the expression of cell proteins host and in the efficiency of viral DNA replication. Adenoviral vectors in which the E l and E4 regions are deleted therefore have very limited transcription background noise and expression of viral genes. Such vectors have been described by example in applications WO94 / 28152, WO95 / 02697, WO96 / 22378). In addition, vectors carrying a modification at the level of the IVa2 gene have also been described (WO96 / 10088).

The recombinant adenoviruses described in the literature are produced from different serotypes of adenoviruses. There are in fact different serotypes of adenoviruses, the structure and properties of which vary somewhat, but which have a comparable genetic organization. More particularly, the recombinant adenoviruses can be of human or animal origin. As regards adenoviruses of human origin, mention may preferably be made of those classified in group C, in particular adenoviruses of type 2 (Ad2), 5 (Ad5); in group B, adenovirus type 7 (Ad7); or in group A, adenovirus type 12 (Ad 12). Among the various adenoviruses of animal origin, mention may preferably be made of adenoviruses of canine origin, and in particular all the strains of adenoviruses CAV2 [manhattan or A26 / 61 strain (ATCC VR-800) for example]. Other adenoviruses of animal origin are cited in particular in application WO94 26914, which is incorporated herein by reference.

In a preferred embodiment of the invention, the recombinant adenovirus is a human adenovirus of group C. More preferably, it is an adenovirus Ad2 or Ad5.

Several methods have been described for the generation of recombinant adenoviruses (C. Chartier et al., J. Virol. 70: 4805-4810, 1996; WO96 / 25506; J. Crouzet et al., Proc. Natl. Acad. Sci USA 94: 1414-1419, 1997; TC. He et al., Proc. Natl. Acad. Sci USA 95: 2509-2514, 1998). These methods make it possible to construct in E. coli plasmids carrying the adenoviral genome of interest, these plasmids are then digested with a restriction enzyme to excise the adenoviral genome of the plasmid. The adenoviral genome is then transfected into a line ^{encapsidation} then amplified.

Recombinant adenoviruses are usually produced by introducing viral DNA into the packaging line, followed by lysis of the cells after approximately 2 or 3 days (the kinetics of the adenoviral cycle being 24 to 36 hours), depending on to another variant, the culture is continued until 8 to 12 days and the viral particles are released spontaneously into the culture medium by autolysis of ^packaging cells.

The viruses used in the context of the following examples are adenoviruses containing the lacZ marker gene from E.coli (AV, ₀ CMV.lacZ). These viruses are derived from the Ad5 serotype and have the following structure:

- A deletion in the E1 region covering, for example, nucleotides 386 (Hinfl site) to 3446 (Sau3a site).

- A lacZ gene expression cassette, under the control of the CMV promoter, inserted at the level of said deletion.

- A deletion of the E3 region.

The construction of these viruses has been described in the literature (WO94 / 25073,

WO95 / 14102, WO96 / 25506, J. Crouzet et al., Proc. Natl. Acad. Sci USA 94: 1414-

1419, 1997). It is understood that any other construction can be used in the method according to the invention, and in particular viruses carrying other heterologous genes and / or other deletions (E1 / E4 or E1 / E2 for example).

The techniques of cell transfection, amplification and titration of adenoviruses have been described previously (FL Graham et al., Molecular Biotechnology 3: 207-220, 1995; Crouzet et al., Proc. Natl. Acad. Sci USA 94 : 1414-1419, 1997; WO96 / 25506).

The tdu assay technique for β-galactosidase activity encoded by the lacZ gene contained in AV viruses, ₀ CMV.lacZ is carried out as described by P. Yeh et al. (J. Virol. 70: 559-565, 1996).

- Production of AV adenoviruses, ₀ CMV.lacZ

Harvesting of the virus from the cultures of producing lines was carried out either by a conventional process using a series of freeze-thaw cycles, or by chemical lysis in the presence of 1% Tween-20, or by continuing the culture until 'to autolysis according to the method described in WO98 / 00524.

The culture media can vary according to the transcomplementing lines used or according to the quantities used. These environments can be the MEM, DMEM ... whether or not supplemented with calf serum and contain different concentrations of inorganic salts, sugar, amino acids, vitamins, hepes or phenol red.

Transcomplementing cells of the E1 region such as 293 or PER.C6 cells are transfected at 60-80% confluence in a culture dish with viral DNA obtained by digestion of a plasmid carrying the adenoviral genome of interest. The incubation lasts from 8 to 15 days, the harvest time is judged by the observation under the microscope of the cells which round off, become more refractive and adhere more and more weakly to the culture support. The virus is then released from the nucleus by 3 to 6 successive thawing cycles (dry ice ethanol at -70 ° C, water bath at 37 ° C). The virus thus obtained is used to re-infect transcomplementing cells at a given Multiplicity of Infection (MOI) which can vary between 10 and 500 viral particles per cell, the amplified virus is obtained as previously by continuing the incubation from 40 to 72 h.

According to another variant described in application WO98 / 00524, the cells are not harvested 40 to 72 hours post-infection, but the incubation is prolonged between 8 to 12 days so as to obtain total lysis of the cells without the need for proceed to freeze-thaw cycles. The virus is then spontaneously released into the supernatant. The supernatant is then clarified by filtration on filters with decreasing porosity depth (10 μm / 1.0 μm / 0.8-0.2 μm). The clarified supernatant is then concentrated by tangential ultrafiltration on a Millipore spiral membrane having a cutoff threshold of 300 kDa. The concentration factor is of the order of 20 to 100 times. According to another variant, the clarified supernatant can be used as it is for the purification of adenoviral particles by chromatography on a column of Q Sepharose® XL.

- Analysis of adenovirus preparations.

The various analytical techniques used to determine the quality of the viral preparations obtained (SDS-PAGE, Western blot analysis, IE-HPLC on a Resource 15Q column, etc.) have been described previously (WO98 / 00524).

2 - Analytical methods using the Q Sepharose® XL chromatography support. The operating conditions for the detection, identification and quantification of ^adenovirus particles from a culture of infected packaging cells are performed as described below.

A chromatography column filled with approximately 1 ml of Q Sepharose® XL (45-165 μm; Amersham-Pharmacia Biotech) is prepared in a column of HR 5/5 type (Amersham-Pharmacia Biotech). This column is mounted on a Waters HPLC system type 626 fitted ^with a detection system UV / visible diode array 996 operating in the range ^of 200-300 nm absorbance. This anion exchange column is used for the separation and quantification of the viral particles.

Before each analysis, the column is equilibrated at 30 ° C in a 20 mM Tris / HCl buffer, pH 7.5 at a flow rate of 1.5 ml / min. The sample to be analyzed containing the viral particles is injected onto the column. To obtain maximum resolution, the quantity of particles injected must be less than or equal to 2x l0 ¹² particles / ml of support. The volume injected has no appreciable influence on the separation of the species, at least for an injected volume less than or equal to 50 ml per ml of gel. After injection, the column is rinsed with 5 volumes of the same buffer, and the fixed species are eluted with a linear gradient from 0 to 1 M NaCl in the 20 mM Tris / HCl buffer, pH 7.5 over 30 column volumes . At the end of the gradient, the column is washed with 2 volumes of 0.5 N sodium hydroxide column before rebalancing for the next analysis.

A standard curve at 260 nm is constructed with a preparation of adenovirus particles purified either by a CsCl gradient or by chromatography. This standard preparation was previously titrated into particles by its absorbance at 260 nm in a 0.1% SDS solution using the conversion factor of 1 × 10 ¹⁰ particles per unit of absorbance at 260 nm.

Under these conditions, the adenovirus is eluted at the retention time of approximately 18 min and has an absorbance ratio at 260 nm compared to 280 nm of 1.30 ± 0.02 (see FIG. 3). The suitability software of the Millennium Waters chromatographic signal acquisition and processing station automatically determines after each analysis the value of N m (calculated at mid-height) and the asymmetry (calculated at 10% of the height) of the peak. The value of N m on the peak adenoviral is typically 35,000 ± 3,000 and the peak asymmetry factor is 1.05 ± 0.05.

3 - Preparative method for the purification of adenovirus by anion exchange chromatography.

The adenovirus is purified from cultures of packaging cells 293 or

PER.C6 (WO97 / 00326). The virus is produced and harvested in supernatants after autolysis as described above. It is then filtered through a 0.45 μm membrane (HT Tuffryn or polysulfone) just before purification. In the absence of contrary indication, the purification protocol is identical to the protocol used for the analytical separation of viral particles described above, but with a gradient ^of different elution. Elution is carried out with a gradient of 0.25 to 1 M NaCl over 30 column volumes. The volume of the column is adapted to the quantities of virus to be purified by considering a capacity of 1 × 10 ¹² particles per ml of chromatographic support. Similarly, the linear flow rate of the eluents is fixed at 300 cm / h.

EXAMPLES

Example 1: comparison of Q type Sepharose® XL supports with Q Sepharose® Fast Flow support.

This example illustrates the specific properties of Sepharose® XL type Q media with respect to Sepharose® Fast Flow 0 support.

The two supports consist of balls of identical basic structure

(cross-linked agarose at 6%), with the same particle size distribution (45-165 μm). They differ in the presence of flexible dextran arms carrying the Q type exchanger groups for Q Sepharose® XL, while the same Q type groups are fixed directly on the agarose matrix in the case of Q Sepharose® FF.

A purified adenovirus preparation (2 x 10 ¹⁰ pv) is injected onto a column of Q Sepharose® XL (1 ml of support) and eluted with a NaCl gradient as defined in paragraph 2 of the Materials and Methods section. The elution profile is presented in FIG. 3. Under the same conditions, an identical analysis is carried out on a similar column filled with Q Sepharose® Fast Flow (FF) support. The elution profile is presented in FIG. 4. The comparison of the chromatographic performances is presented in the table below.

Table 1: comparison of the chromatographic performances with the Q Sepharose® XL and Q Sepharose® FF supports

As shown in Figures 3 and 4, the virus retention time is similar in the two cases (t = 18 min for Q Sepharose® XL and t = 20 min for Q Sepharose® FF), but the support Q Sepharose® XL has a much higher efficiency than the Q Sepharose® FF support.

Similarly, analysis of a crude cell extract containing 2 × 10 ⁹ viral adenovirus particles on the two supports (FIG. 5) shows that only the Q Sepharose® XL support exhibits an identified and quantifiable virus peak. On the other hand, the proteins present in the preparation are separated with identical efficiency for the two supports studied (FIG. 5).

These results indicate that the presence of the flexible arms carrying the exchanger groups is an essential element of this type of support. The presence of these flexible arms contributes significantly to the advantageous chromatographic performance of the Q Sepharose® XL for the separation of adenoviruses.

Example 2: comparison of Q-type Sepharose® XL supports with Q-type Sepharose® HP support.

This example illustrates the specific properties of Sepharose® XL type Q media with respect to Sepharose® HP 0 support.

The two supports consist of beads of identical basic structure (cross-linked agarose at 6%). The Q Sepharose® XL support has a size distribution of beads ranging from 45 to 165 μm centered on 90 μm. The particle size of the support Q Sepharose® HP is 34 ± 10 μm. The particle size of the Q Sepharose® HP support is finer and much less dispersed than that of the Q Sepharose® XL support. The Q Sepharose® HP support should therefore have chromatographic performances far superior to the performances of the Q Sepharose® XL.

A purified adenovirus preparation (2 x 10 ¹⁰ pv) is injected onto a column of Q Sepharose® XL (1 ml of support) and eluted with a NaCl gradient as defined in paragraph 2 of the Materials and Methods section. The elution profile is presented in FIG. 3. Under the same conditions, an identical analysis is carried out on a similar column filled with Q Sepharose® HP support. The elution profile obtained with support 0 Sepharose® HP is presented in Figure 6.

The results presented in FIG. 6 show that the Q Sepharose® HP support has an efficiency (N / m, 15,000) substantially lower than the Q Sepharose XL support. In addition, the viral peak has a sensitive lag on the Q Sepharose® HP support (asymmetry, 1.6), whereas it is rigorously symmetrical on the Q Sepharose® XL support.

Unexpectedly, despite its finer particle size and its less dispersed particle size distribution, the Q Sepharose® HP support does not achieve the performance of the Q Sepharose® XL support for the separation of viral particles. On the other hand, it has much higher performance for the separation of proteins, as indicated by the supplier (Amersham-Pharmacia Biotech) and as it is confirmed in our experimental conditions with bovine albumin separation experiments (table below). below).

Table 2: comparison of the chromatographic performances with Q Sepharose® XL and Q Sepharose® HP supports These results confirm that the presence of the flexible arms carrying the strong anion exchange groups is an essential element in the chromatographic performance of this type of support for the separation of the viral particles. These results also indicate that another important parameter to take into account when defining the choice of support is the particle size and in particular the size dispersion.

Example 3: comparison of Sepharose® XL type Q supports with Fractogel® TMAE (S) and Source 15Q supports.

This example illustrates the specific properties of Sepharose® XL type Q media with respect to the Fractogel® TMAE (S) support and Source 15Q support.

The three supports consist of balls of different structure and composition. Q Sepharose® XL consists of 6% cross-linked agarose. The Fractogel® TMAE support is a crosslinked polymethacrylate resin and the Source 15Q support consists of resin beads of the polystyrene-divinylbenzene type. The particle size of the Fractogel® TMAE (S) (20-40 μm) and Source 15Q (15 μm) supports is much lower and much less dispersed than that of the Q Sepharose® XL support (45-165 μm). Furthermore, the three supports present the strong exchanger groups. These are located on flexible arms fixed to the matrix in the case of Fractogel® TMAE (S) and Q Sepharose® XL, unlike Source 150 whose exchanger groups are grafted directly onto the matrix.

A preparation of purified adenovirus (2 x 10 ⁱ⁰ vp) is injected onto a column of Q Sepharose XL (1 ml support) and eluted with a NaCl gradient as defined in paragraph 2 of the section Materials and Methods. The elution profile obtained is presented in Figure 3. Under the same conditions, an identical analysis is carried out on a similar column filled with Source 15Q support and and another analysis is carried out on a column filled with Fractogel® TMAE (S) support.

Unexpectedly, despite its much finer grain size, the support

Fractogel® TMAE (S) does not reach the performance of the Q Sepharose® XL support for the separation of adenoviruses. On the other hand, it has much higher performance for the separation of proteins. Similarly, despite its very fine particle size and its almost monodispersed particle size distribution. Source 15Q support n ^'reaches not, either, Q Sepharose XL support performance for the separation of adenoviruses.

Table 3: comparison of chromatographic performance with Source 15Q, Q Sepharose® XL and Fractogel® TMAE (S) supports.

These results indicate that the presence of flexible arms carrying the exchanger groups is not solely responsible for the specific chromatographic performances of the Q Sepharose® XL support for the separation of adenoviruses. The chemical composition of these arms, their grafting density, the nature and the porosity of the matrix on which the flexible arms are grafted are all parameters which can influence the chromatographic performance of the support for the separation of viral particles.

Example 4: detection and quantification of wild adenovirus particles of various serotypes.

This example illustrates a method of detection and identification of wild adenovirus particles of various serotypes based on the use of the Sepharose® XL type Q chromatographic support.

The various wild adenoviruses were produced by infection of A549 cells cultured on DMEM medium and harvested after 3 cycles of freezing-thawing of the cells. The preparations were then filtered through an Acrodisc membrane (type HT Tuffryn 0.45 μm; GelmanSciences) before analysis.

The various preparations are then analyzed by chromatography according to the protocol described in paragraph 2 of the Materials and Methods section. However, in ^the example presented, the analyzes were performed with a column having a slightly higher volume of 1 ml (≈ 1.35 ml), which explains that the time 21

Adenovirus 5 retention is greater (25.3 min) than the reference retention time (18 min) indicated in the Materials and Methods section.

Table 4: Chromatographic characterization of various wild adenoviruses

This example shows that the various adenoviruses not only have a variable retention time according to the serotypes considered but also an absorbance ratio 260 nm / 280 nm characteristic of the serotype in question. The combination of these two criteria which can be measured simultaneously during a single chromatographic analysis therefore constitutes a means of rapid and reliable identification of the serotype of the adenovirus present in the chromatographed preparation.

A correlation has been sought between the retention time of the virus on the column and the fiber and hexon characteristics of adenoviruses of type 7, 3, 4, 5 and 2, for which the sequences are published (1998 J, Virol. (1998) 72 p. 7909 and Arch, Virol (1997) 142 p. 1307). No correlation is revealed from the identity data of sequences of the head of the fiber, nor even from the differences in the number of repetitions of β sheets in the fiber stalk (see table below). By cons, and suφrenantly. a correlation has been demonstrated between the retention time of the virus and the sequence identity of the hexon (see table below). This correlation is in perfect agreement with the overall charge of the hexon at pH 7; this correlation shows that the increase in the charge of the hexon at pH 7 corresponds to an increase in the retention of the virus on the column. Nuances may exist depending on the charge at pH 7 of the exposed part L l of the hexon, as indicated by the data obtained with the type 3 virus.

Table 5: correlation between the retention time of the virus and the sequence identity of the hexon.

Example 5: detection and quantification of the recombinant adenovirus particles during the stages of production of a viral stock.

This example illustrates the use of Sepharose® XL type Q supports for the detection and quantification of the AV, ₀ CMV.lacZ recombinant adenovirus particles produced during the transfection and amplification steps with different packaging cell lines ( cells 293 or PER.C6).

This extremely rapid analysis method provides, in a few minutes, and from a culture supernatant, the title of the adenovirus solution resulting from each step. This fast and sensitive method makes it possible to optimize the amplification conditions of the next step which can therefore be carried out under conditions of determined and controlled MOI. This analysis method was tested to control the production of recombinant adenoviruses of type AV _{1 0} CMV.lacZ during the steps of transfection and amplification on 293 or PER.C6 cells.

The adenovirus AV _{1 0} CMV.lacZ was produced after transfection of 293 or PER.C6 cells with the plasmid pXL2822 digested with Pacl (Crouzet et al., Proc. Natl. Acad. Sci USA 94: 1414-1419, 1997) then infection of cells respectively 293 or PER.C6 at a determined multiplicity of infection (MOI). The initial transfection is carried out with 5 to 10 micrograms of viral DNA obtained by digestion of the plasmid.

During cell lysis, the virus was harvested by 3 freeze-thaw cycles of the cells. The preparations were then filtered through an Acrodisc membrane (of the HT Tuffryn type) of 0.45 μm before analysis.

The various preparations are then analyzed by chromatography according to the protocol described in paragraph 2 of the Materials and Methods section. The titer of the adenovirus solution is determined by reference to a standard curve produced under the conditions described in paragraph 2 of the Materials and Methods section.

The results are presented in the table below.

Table 6: Detection and quantification of particles ^of recombinant adenovirus AV, ₀ CMV.lacZ obtained during the production of viral stocks in two transcomplementing lines (293 or PER.C6).

The total viral particles obtained were assayed to determine the concentration of infectious particles (pfu) and particles having the activity of the transgene (tdu). The term pfu ("plaque forming unit") corresponds to the infectious power of an adenovirus solution, and is determined by infection of an appropriate cell culture, and measures, generally after 15 days, the number of plaques of infected cells. These assays are based on biological methods and the values obtained may appear different depending on the conditions used (J. Virol. (1996) 70 p. 7498). In fact, the formation ^of a beach in a line Transcomplementing does not necessarily describe the infectivity of the virus in other target cells (Biotechniques (1997) 22 p. 447). The results obtained are presented in the table below and the values obtained are in perfect agreement with the data in the literature (P. Yeh et al. J. Virol. (1996) 70 p. 559). Indeed, P. Yeh et al. describe AdRSVβGal viruses purified by cesium chloride gradient having tdu / pfu ratios of 0.49 to 0.68 which is very close to the results obtained below.

Table 7: relationship between the concentration of infectious particles (pfu) and particles with the activity of the transgene (tdu).

Example 6: detection and quantification of the recombinant adenovirus particles produced during the transfection and amplification steps on an IGRP2 cell.

This example illustrates the use of Sepharose® XL type Q supports for the detection and quantification of recombinant adenovirus particles. AV ₃₀ CMV.lacZ produced during the transfection and amplification steps on an IGRP2 cell.

The adenovirus AV ₃₀ CMV.lacZ was produced after transfection of IGRP2 packaging cells with the plasmid pXL3005 digested with Pacl (the plasmid pXL3005 derives from the plasmid pXL281 1 described in (Crouzet et al., Proc. Natl. Acad. Sci USA 94: 1414-1419, 1997) by exchange of the RSV promoter with the CMV promoter) then infection of IGRP2 cells (WO96 / 22378) at a determined multiplicity of infection (MOI). During cell lysis, the virus was harvested by 3 freeze-thaw cycles of the cells. The preparations were then filtered through an Acrodisc membrane (of the HT Tuffryn type) of 0.45 μm before analysis.

The various preparations are then analyzed by chromatography according to the protocol described in the Materials and Methods section. The results are presented in the table below.

Table 8: Detection and quantification of particles ^of recombinant adenovirus _{AV? 0} CMV.lacZ obtained during the production of viral stocks.

The total viral particles obtained were assayed to determine the concentration of particles having ^the transgene activity (tdu 3.2 x ^10s / ml). A ratio vp / tdu of 43 is comparable to the ratio vp / tdu 43 and 55 obtained in ^Example 5 and used to correlate the physical measurement 1, 39 x 10 '° vp / ml of viral particles in biological measuring 3.2 x 10 tdu / ml of transduction units. Example 7: purification of the virus by chromatography on Q Sepharose® XL resin.

The starting material consists either of the culture lysate obtained by freeze-thaw cell ^for producing packaging of the virus or the supernatant obtained after lysis of the cells spontaneously.

In the experiment reported in this example, 153 ml of a culture autolysate of PER.C6 cells infected with the AV virus, ₀ CMV.LacZ containing 4.9 × 10 ^ι particles were injected into a column of 5.8 ml of Q Sepharose® XL. Equilibration of the column and elution of the virus were carried out at a flow rate of 300 cm / h with a gradient of 0.25 to 1 M NaCl over 30 column volumes as described for the analytical separation of the virus in the paragraph 2 of the Materials and Methods section. The viral peak (7.1 ml) was collected and then analyzed by various techniques (EI-HPLC, SDS-PAGE) as described below. The collected fraction is analyzed by high performance liquid chromatography (HPLC) on a Resource Q column (1 ml) in a following chromatographic system: 10 μl of the fraction purified by chromatography as described above are injected onto a Resource Q15 column ( 1 ml of gel; Pharmacia) equilibrated in 100 mM Tris / HCl buffer pH 8.0 containing 0.5 mM MgC12, (buffer B). After rinsing with 5 ml of buffer B, the adsorbed species are eluted with a linear gradient of 30 ml of NaCl (0 to 1 M) in buffer B at a flow rate of 1 ml / min. The eluted species are detected at 260 nm. After ^the purification step on a column of Q Sepharose XL. the fraction collected has a purity> 99% in viral particles (UV detection at 260 nm). The purification yield in viral particles is 82%.

This analysis by HPLC further shows that the residual bovine serum albumin present in the starting lysate is completely eliminated during the preparative chromatography.

The electrophoresis analysis of the purified adenoviral fraction by chromatography is carried out on polyacrylamide gel (4-20%) under denaturing conditions (SDS). The protein bands are then revealed with silver nitrate. This analysis shows that the adenoviral preparation obtained by chromatography has a level of purity at least equal to that of the preparation obtained conventionally by ultracentrifugation and that it does not have any additional protein band which would indicate contamination of the preparation with non-adenoviral proteins.

The adenoviral preparation obtained by chromatography has an absorbance ratio A 260 nm / 280 nm equal to 1.30 ± 0.05. This value, which is identical to that obtained for the best preparations obtained by ultracentrifugation. indicates that the preparation is free of contaminating proteins or contaminating nucleic acids.

The titration of the virus clearly shows the presence of infectious viral particles with a very satisfactory pv / pfu ratio (see table below) and the purified viral particles effectively have the expected infectious activity.

Table 9: purification of AV adenovirus, ₀ CMV.LacZ on Q Sepharose® XL support.

The method described in this example therefore makes it possible to purify the adenoviral particles without affecting their infectious power, directly from a lysate of packaging cells, without prior treatment (ultrafiltration for example or treatment with a nuclease) of the material to be purified.

Claims

1. A method of separating viral particles from a biological medium characterized in that ^it comprises at least one chromatography step carried out on a support comprising a matrix, groups of ion exchangers, said groups being grafted on said matrix by means of a flexible arm.

2. Method according to claim 1, characterized in that the matrix is chosen from agarose. dextran, acrylamide, silica, poly [styrene-divinylbenzene], alone or as a mixture.

3. Method according to claim 2, characterized in that the matrix consists of crosslinked agarose and preferably crosslinked agarose at 6%.

4. Method according to one of claims 1 to 3, characterized in that the matrix has a particle size between 40 and 200 microns approximately.

5. Method according to claim 4, characterized in that the matrix has a particle size between 45 and 165 microns and centered on 90 microns.

6. Method according to claim 4 or 5, characterized in that the matrix has a dispersion such that 95% of the particles have a diameter between 0.1 and 10 times the average diameter of the particles, and preferably between 0.3 and 3 times the diameter medium particles.

7. Method according to claim 1. characterized in that the flexible arm is a hydrophilic arm consisting of a polymer of synthetic or natural origin.

8. Method according to claim 7, characterized in that the flexible arm is a polymer of synthetic origin chosen from polyvinyl alcohols, polyacrylamides. polymethacrylamides or polyvinyl ethers.

9. Method according to claim 7, characterized in that the flexible arm is a polymer of natural origin of polysaccharide nature chosen from starch, cellulose, dextran and agarose.

10. Method according to claim 8 or 9, characterized in that the degree of polymerization of the flexible arm is approximately 30 monomeric units

1 1. Process according to claim 10, characterized in that the flexible arm is a dextran with an average molecular weight of approximately 5000 Da.

12. Method according to claim 1, characterized in that the ion exchange group is a strong anion exchange group.

13. Method according to claim 12, characterized in that the strong anion exchanger group is a quaternary amine.

14. Method according to one of claims 7 to 13 or according to claim 5, characterized in that the chromatography is carried out on a support of type Q Sepharose® XL.

15. The method of claim 1, characterized in that the biological medium is a supernatant of packaging cells producing said virus.

16. Method according to claim 1, characterized in that the biological medium is a lysate of packaging cells producing said virus.

17. The method of claim 1, characterized in that the biological medium is a prepurified solution of said virus.

18. The method of claim 1, characterized in that it comprises a prior step of ultrafiltration.

19. The method of claim 18 characterized in that the ultrafiltration is a tangential ultrafiltration on the membrane having a cutoff threshold between 300 and 500 kDa.

20. Use of Sepharose® XL type Q chromatography support for analytical and / or preparative separation of viral particles.

21. Use according to claim 20, characterized in that the viral particles are adenoviruses.

22. Use of Sepharose® XL type Q chromatography support for the identification of different adenovirus serotypes.

23. Use of Sepharose® XL type Q chromatography support for adenovirus titration.

24. Adenovirus quantification method characterized in that the viral particles are separated by chromatography on a support of type Q Sepharose® XL and the amount of adenovirus is measured by absorbance of the chromatography fractions.