A method for separating or removing empty AAV capsids from intact AAV capsids

Abstract

A method for separating or removing empty AAV capsids from intact AAV capsids in an aqueous mixture comprising empty and intact AAV capsids, wherein the mixture is contacted with a solid-phase surface bearing a primary amino group in a first alkaline environment, whereby (i) the intact AAV capsids bind to the solid-phase surface while the empty AAV capsids are at least partially unbound to the solid-phase surface, or (ii) both the intact and empty AAV capsids bind to the solid-phase surface, and subsequently the empty AAV capsids are at least partially eluted by a second alkaline environment having a pH higher than the pH of the first alkaline environment, provided that the second alkaline environment does not elute the intact AAV capsids from the solid-phase surface.

Description

[0001] The present invention relates to a method for separating or removing empty AAV capsids from intact AAV capsids using solid-phase extraction.

[0002] Preparations of recombinant adeno-associated virus (AAV) typically consist of two capsid subpopulations. The desired capsids are appropriately filled with their intended therapeutic DNA payload. These are referred to as intact capsids. The impure preparation reagents also contain capsids that are not filled with their intended DNA load. These are referred to as empty capsids. Some preparations may contain 90% or more empty capsids, while others contain less than 50%. Current clinical guidelines recommend reducing the content of empty AAV capsids to less than 10%. This can be achieved by ultracentrifugation, but the equipment and process are difficult to scale up and are prone to failure due to minor operator errors. Many consider chromatography to be superior to centrifugation.

[0003] Chromatographic methods for separating empty AAV capsids from intact AAV capsids are known [1-3]. Using strong anion exchangers and eluting with a salt gradient at alkaline pH conditions has been shown to sufficiently reduce empty AAV capsids for many AAV serotypes. However, the results are sometimes inadequate in one or more respects. First, the degree of separation is often limited. There are known cases of achieving baseline resolution, meaning that the peaks corresponding to empty AAV capsids and intact AAV capsids are completely separated, but these are exceptions. The individual elution peaks typically overlap to some extent, sometimes substantially, sometimes almost completely. This makes it necessary to sacrifice intact AAV capsids that overlap with empty AAV capsids in order to eliminate sufficient empty capsids. The recovery of intact capsids is thus impaired.

[0004] A related compromise also involves the limited separation achieved by strong anion exchangers: it burdens the industrial-scale use of the technique. Many users prefer step gradient elution in large-scale chromatography techniques because they are simpler and can save money on equipment, buffer preparation, fraction collection, and analysis. Unfortunately, step gradients are affected by reproducibility issues because each individual process variable must be invariant between processing batches. Among other variables, the composition of the buffer in terms of pH and conductivity must be exactly the same, the composition of the chromatographic medium must be exactly the same, the processing temperature must be the same, etc. These parameters are not invariant in actual practice, and the results will inevitably vary with the degree of recovery of intact AAV capsids and / or the elimination of empty AAV capsids.

[0005] Linear gradient elution supports better reproducibility. Due to variations in materials or conditions, the elution curve may vary slightly, but the relative relationship between the peaks of the eluted full and empty capsids is not affected by these variations and remains resolved. However, manufacturing-scale equipment that supports linear gradient elution is more expensive, requires more training, more buffer, more fraction analysis, and more time than step gradients.

[0006] The third limitation is related to the extreme chemical conditions for empty-full separation on strong anion exchangers. The required pH is typically 9.0 or higher, sometimes up to 10.0 or higher [1-3]. Generally, the higher the pH, the better the separation. Unfortunately, the chemical stress generated by pH values above 9.0 reduces the stability of many biologics, and the higher the pH above 9, the higher the level of chemical stress. Chromatograms of highly purified AAV capsids typically show peaks in addition to the empty and full AAV capsids, which are interpreted as representing partially dissociated capsids and free capsid proteins. DNA released from previously intact full AAV capsids due to damage may also be evident.

[0007] The term strong anion exchanger refers to chromatographic media with quaternary amine ligands. The adjective "strong" specifically refers to their ability to maintain their full charge over a wide pH range, e.g., from approximately pH 2.0 to pH 12.0. Common trade names generally include Q, QA, QAE, referring to quaternary, quaternary amine, and quaternary aminoethyl, respectively. Other names for strong anion exchangers include TEAE (triethylaminoethyl) or TMAE (trimethylaminoethyl), which correspond to the same degree of amine derivatization but different naming conventions. TEAE and TMAE still describe quaternary amine ligands.

[0008] Examples of less derivatized amino chromatographic ligands are well known by names such as DEAE (diethylaminoethyl). DEAE represents a tertiary amine ligand. Tertiary amine ligands are examples of so-called weak anion exchangers because they maintain their full charge over a relatively limited pH range. DEAE ligands start to lose charge at pH values as low as 7.5. They lose most of their charge at pH 9.0 and essentially all of their charge at pH 12. The low charge at pH 9.0 translates to low capacity for AAV, thus reducing productivity. It also indicates that weaker anion exchange ligands, such as secondary or primary amine ligands, may produce more adverse results compared to quaternary amino ligands. In this regard, it is notable that weak exchangers like DEAE are rarely discussed for separating empty and full AAV capsids.

[0009] Anion exchangers are most commonly used with a salt gradient elution at a fixed pH, such as a salt gradient generated by increasing the sodium chloride concentration. They are sometimes eluted with a decreasing pH gradient. The charge on a strong exchanger remains constant over a range of approximately 2 to 12, but the charge on a protein changes. The electronegativity of a protein weakens as the pH decreases, while the electropositivity of the protein strengthens. The decreasing electronegativity means a decreased attraction to the positively charged anion-exchange surface. The increasing electropositivity translates to an increased repulsion from the positively charged anion-exchange surface. These two phenomena work together to effect elution with decreasing pH. Exposing a strong anion exchanger to an increasing pH gradient results in the bound components binding more tightly.

[0010] Cation exchangers are widely used for AAV purification, but most commonly use a salt gradient elution, resulting in the empty and full AAV capsids eluting together in one peak. They can also be eluted with a pH gradient, but since cation exchangers are negatively charged while anion exchangers are positively charged, gradient elution of cation exchangers requires an increasing pH value. Eluting a cation exchanger with an increasing pH gradient enables a degree of separation between the empty and full AAV capsids, but the degree of separation is not as great as that provided by salt elution of a strong anion exchanger.

[0011] Overview

[0012] The present invention is based on the unexpected finding that separation of full AAV capsids from empty AAV capsids can be achieved using a solid phase modified with a primary amine group. The method of the present invention is capable of overcoming several practical limitations associated with known methods for removing empty AAV capsids, which use strong (quaternary amine) anion exchangers that are eluted with an increasing salt gradient at a fixed alkaline pH [1-3]. The main enabling feature of the method of the present invention is the unexpected finding that, despite the absence of excess salt, most of the empty AAV capsids can be selectively removed by increasing the pH within an alkaline range. This is contrary to the behavior of empty AAV capsids in known anion-exchange methods, in which the binding of the capsids is strong at alkaline pH values in the absence of excess salt [1-3] and becomes stronger at higher pH values. The term "excess salt" herein refers to salt that is in excess of the buffering components used to control the pH of the operating solution, such as sodium chloride (NaCl) or other salts that may be added to Tris buffer or Bis-Tris-propane buffer or other buffers to effect elution of bound components.

[0013] In general aspects, the present invention relates to a method for separating or removing intact AAV from empty AAV capsids in an aqueous mixture containing intact and empty AAV capsids, wherein the aqueous mixture containing intact and empty AAV capsids is contacted with a solid phase having a primary amino group.

[0014] In another aspect, the present invention relates to a method for separating or removing empty AAV capsids from intact AAV capsids in an aqueous mixture containing empty and intact AAV capsids, wherein the mixture is contacted with the surface of a solid phase having a primary amino group in a first basic environment, whereby

[0015] (i) intact AAV capsids bind to the solid phase surface while empty AAV capsids are at least partially unbound to the solid phase surface, or

[0016] (ii) both intact and empty AAV capsids bind to the solid phase surface, and subsequently the empty AAV capsids are at least partially eluted by a second basic environment having a pH higher than that of the first basic environment, provided that the second basic environment does not elute intact AAV capsids from the solid phase surface.

[0017] In one embodiment of the present invention, the pH of the first basic environment can be pH > 7, specifically pH > 7 to pH 8.

[0018] In another embodiment of the present invention, the pH of the second basic environment can be greater than the pH of the first environment. In yet another embodiment of the present invention, after at least partial elution of the empty AAV capsids, the solid phase surface can be contacted with a third basic environment having a pH greater than that of the second environment.

[0019] In yet another embodiment of the present invention, after at least partial elution of the empty AAV capsids, the solid phase surface can be contacted with a fourth basic environment having a pH less than that of the third or second basic environment and having a higher salt concentration compared to the salt concentration of the first, second, or third basic environment.

[0020] According to the present invention, the pH of the second basic environment can especially be in the range of pH 8.0 to pH 9.0, pH 8.1 to pH 8.9, pH 8.2 to pH 8.8, pH 8.3 to 8.7, or pH 8.4 - 8.6.

[0021] According to the present invention, the pH of the third basic environment can especially be in the range of pH 8.5 to pH 10.5, or pH 8.5 to 10.0, or pH 8.5 to 9.5.

[0022] The overlap of the pH range values of the first, second, or third pH value is virtual. It is easy for those skilled in the art to know that if the pH value of the first alkaline environment has been selected as pH 8 (the range span from pH > 7 to pH 8), then the corresponding pH value of the second alkaline environment should be selected from the range of pH 8.0 to pH 9.0, which means that the pH value of the second alkaline environment cannot be the lowest value of the pH range, i.e., pH 8, but a higher value, such as 8.1 or 8.2, etc. The same consideration is correct regarding the virtual overlap of the pH ranges of the second and third alkaline environments.

[0023] In another embodiment of the method of the present invention, the salt concentration is up to 1 M, corresponding to the concentration of alkali metal salts such as NaCl, specifically 1 mM to 1,000 mM, or 2.5 mM to 250 mM, such as 2.5 mM, or 5 mM, or 12.5 mM, 25 mM, or 50 mM, or 100 mM, or 150 mM, or 200 mM, or 250 mM.

[0024] In a specific embodiment of the present invention, the alkaline environment may contain a magnesium salt with a magnesium ion concentration in the range of 1.0 mM to 5.0 mM, particularly 1.5 mM to 3.0 mM, or 2.0 mM to 2.5 mM.

[0025] An advantageous aspect of the present invention is the fact that the full and empty AAV capsids can be of any serotype, particularly selected from the group consisting of: natural and recombinant serotypes, chimeric (hybrid) types, and combinations thereof.

[0026] The method of the present invention can be applied in any device commonly used in the art. For example, a solid-phase surface with a primary amine group can be arranged in a chromatographic device. Generally, the solid-phase surface with a primary amine can be a monolithic column, a packed particle column, a packed nanofiber column, a membrane adsorber, or a hydrogel.

[0027] The subject matter of the present invention also lies in the use of a solid-phase extraction material having a solid-phase surface with a primary amine group for separating or removing empty AAV capsids from an aqueous mixture containing empty AAV capsids and full AAV capsids.

[0028] Before, after, or both before and after the method of the present invention, there are additional processing steps to reduce the empty capsid content to a greater extent than can be achieved by any one of the processing steps alone. Currently known additional processing steps for reducing the amount of empty capsids include ion exchange chromatography using a quaternary ammonium ion exchanger and density gradient ultracentrifugation. In one embodiment, the combination of the present invention with ion exchange chromatography on a quaternary anion exchanger can produce a fraction of intact capsids with a sufficiently low concentration of empty capsids such that no additional processing to remove the empty capsids is required. The two steps can be performed in any order. In another embodiment, density gradient ultracentrifugation can be performed after the method of the present invention. In another embodiment, density gradient ultracentrifugation can be performed after the combination of the present invention with ion exchange chromatography on a quaternary ammonium anion exchanger.

[0029] No specific mechanism or mechanism by which the present invention works has been developed; it is only known that it is fundamentally different from known cation exchange or anion exchange chromatography and that it achieves an excellent separation of empty AAV capsids from intact AAV capsids compared to the salt elution criteria on strong anion exchangers. The superiority of the separation refers to any combination of the following three attributes: the separation can be carried out under mild chemical conditions that are favorable for maintaining the stability of the intact capsids, and / or the degree of separation between empty and intact AAV capsids is increased, and / or the degree of separation is sufficient to make step gradient elution a viable method.

[0030] The method can be used for analytical or preparative applications. The method is applicable to all AAV serotypes. The specific chromatographic conditions and degree of separation between empty and intact AAV capsids may vary between serotypes and between different recombinant constructs within a serotype. This is a well-known characteristic of empty-full separation carried out by known ion exchange chromatography methods and simply reflects the natural variation in capsid surface chemistry between different AAV serotypes. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 Shows a comparison of the results of separating empty AAV capsids from intact AAV capsids by salt elution from a quaternary amino solid phase and a primary amine solid phase.

[0032] Figure 2 Shows a comparison of the results of separating empty AAV capsids from intact AAV capsids by salt elution from a quaternary amino solid phase and elution from a primary amine solid phase by an increasing pH gradient.

[0033] Figure 3Shows the equilibration and loading results of a primary amine column under conditions where both empty AAV capsids and full AAV capsids bind. The empty AAV capsids are removed by a second wash step, and the full AAV capsids are eluted by an increasing pH gradient.

[0034] Figure 4 Shows the equilibration and loading results of a primary amine column under conditions where empty AAV capsids do not bind. Subsequently, the full AAV capsids are eluted in a step gradient with increasing salt concentration while decreasing the pH value.

[0035] Figures 5 to 10 Shows the results of a series of experiments. Samples containing empty and full AAV8 capsids are applied to a solid phase with primary amine groups at pH 8.0. In each experiment, the loaded column is washed with a buffer of higher pH and then eluted with an increasing pH gradient, ending at pH 9.5.

[0036] Figure 11 : The relative recovery of empty AAV capsids and full AAV capsids from the gradient is shown graphically.

[0037] Figure 12 : The effect of the presence or absence of magnesium ions on the elution of empty and full AAV capsids is illustrated by comparing chromatograms. SUMMARY OF THE INVENTION

[0038] The term "solid phase with primary amine" refers to a solid phase that predominantly or only has primary amine ligands on its surface, where the term primary amine ligand describes a nitrogen atom connected to each of two hydrogen atoms by a single covalent bond and also connected to a carbon atom by a single covalent bond. Secondary amines should not be present on the solid phase surface or be present in very small amounts. Tertiary amines and quaternary amines should not be present or be present only in minor amounts. Negatively charged residues should not be present. Uncharged hydrophobic or hydrogen bond residues may be present. The primary amine ligand can be directly covalently linked to the solid phase through the ligand carbon atom. Alternatively, the primary amine ligand can be indirectly linked to the solid phase by covalently connecting the carbon atom of the ligand to a so-called spacer arm that is covalently linked to the solid phase. The primary amine ligand can also be part of a polymer structure covalently linked to the solid phase. The solid phase can be one or more porous membranes, one or more fibers, one or more porous or non-porous particles, or a monolithic solid phase, including a monolithic column synthesized from a single polymer mixture or a monolithic column first synthesized as a macroscopic framework on which a polymer phase containing a secondary ligand is synthesized. Any of the above solid phase materials can be provided in a housing to facilitate chromatography. The housing of the chromatographic solid phase is commonly referred to as a chromatographic device and often as a chromatographic column.

[0039] Chromatographic solid phases with primary amines are known and commercially available. As an example, a chromatographic solid phase sold under the name Toyopearl NH2-750F (Tosoh Bioscience), where "NH2" refers to a primary amine [https: / / www.separations.eu.tosohbioscience.com / solutions / process-media-products / by-mo de / ion-ex-change / anion-exchange / toyopearl-nh2-750f]. Marketing materials indicate that the primary amine ligand is in the form of a polyamine, meaning it is a polymer covalently fixed to the solid phase by repeating primary amine subunits. Another example is sold under the name Sartobind STIC PA (Sartorius), where "PA" refers to a primary amine [www.sartorius.com / shop / ww / en / usd / sartobind- -pa / c / M_Sartobind_STIC_PA]. Marketing materials indicate that the primary amine ligand is in the form of a polyamine, specifically polyallylamine, covalently fixed to the solid phase by repeating primary amine subunits. All major chromatographic solid phase producers manufacture products with amine derivatives on the surface, indicating that the field has the knowledge and resources required to produce primary amine-containing solid phases on a routine or experimental basis.

[0040] Given that chromatographic solid phases containing primary amines may not be named in a way that clearly reveals their composition, it would be useful to have an analytical method to determine whether a given product has the appropriate characteristics to practice the present invention. A simple way to make this determination is to equilibrate the chromatographic solid phase to be identified with a buffer (e.g., 10 mM Tris, 10 mM bis-Tris propane, 15 mM NaCl, pH 7.2), apply a sample consisting of protein standards for isoelectric focusing, wash the solid phase with the equilibration buffer, and then apply a 50 column volume (CV) linear gradient starting from the equilibration buffer and ending at 10 mM Tris, 10 mM bis-Tris propane, 15 mM NaCl, pH 9.5, with 10 CV held in the end point buffer. Elution of the protein across the linear gradient indicates that the solid phase to be identified is a primary amine-containing solid phase. An experimental control consisting of a quaternary amine-containing solid phase can be run in parallel. The protein will not elute. Protein standards for isoelectric focusing are available from various suppliers including ThermoFisher, Bio-Rad, and Serva.

[0041] In some cases, according to the above-described identification tests, a solid phase with polyethyleneimine (PEI) may allow elution of proteins with basic isoelectric points, and they may allow a certain degree of separation between empty capsids and intact capsids. Linear PEI is an amine-containing polymer in which the termini are occupied only by primary amines and there are secondary amines on each repeating subunit of the polymer. Branched PEI polymers also carry tertiary amines at the branch points. It is known that PEI solid phases can effect separation of certain proteins with a decreasing pH gradient, but like quaternary amine solid phases, only increasing pH gradient elution is known.

[0042] The term "equilibrated" or "equilibration" refers to a chemical conditioning step performed on the solid phase and / or the sample to create a specific chemical environment. The solid phase is typically conditioned by exposing them to a buffer containing the desired pH, salts, and salt concentrations. The sample is typically conditioned by titrating the pH, sometimes by dilution to reduce the salt concentration, sometimes by buffer exchange techniques including chromatography, or by dialysis, or by tangential flow filtration diafiltration. In the present invention, equilibration conditions can be set such that the solid phase binds empty AAV capsids and intact capsids, or equilibration conditions can be set such that a subset of the empty capsids, which may include most or all of the empty capsids, do not bind to the solid phase during the loading process. All of these methods and the criteria for choosing one or the other have been well known in the art for decades.

[0043] The intact and empty AAV capsids in any given sample can belong to any serotype, including natural and recombinant serotypes, including chimeric (hybrid) and novel serotypes. It should be understood that capsids of different serotypes have different charge characteristics, including differences in charge levels between empty AAV capsids and intact capsids, and such differences may also be apparent among different recombinant constructs within a serotype. This is important because it means that the specific behavior of any given formulation may vary with primary amine solid phases, as they also vary with strong anion exchangers. In turn, this means that some degree of optimization may be required when applying the method to any AAV sample that has not been tried before, as is also the case with known methods of separating empty and intact AAV capsids with a salt gradient on strong anion exchangers. The sample can be in the form of a cell culture harvest, a cell lysate, a partially purified preparation such as an elution fraction from an affinity chromatography column, a partially purified fraction eluted from an ion exchange chromatography column, a partially purified fraction eluted from a hydrophobic interaction chromatography column, or a capsid mixture obtained from any other purification method or combination of methods used to process AAV.

[0044] The term "loading" refers to the process of contacting an equilibrated sample with an equilibrated primary amine solid phase. This is typically done using chromatography equipment by subjecting the sample to an external force (e.g., gravity or pumping) to pass it through the equipment.

[0045] The term "adsorption" refers to the process of binding a biological product to a chemically complementary surface. Complementarity in the present invention is meant to include electrostatic charge. The negative electrostatic charge on the surface of the AAV capsid mediates their adsorption to the surface of a solid phase that has been rendered positively charged by covalent immobilization of primary amines. The adsorption of a biological product to a solid phase for chromatography is most commonly referred to as "binding".

[0046] The term "selective adsorption" refers to the application of conditions that allow at least one substance to adsorb while preventing the adsorption of one or more other substances. In the present invention, the operating pH can be adjusted to an alkaline range to prevent the adsorption of unwanted empty AAV capsids while allowing the adsorption of the desired intact AAV capsids.

[0047] The term "desorption" refers to the process of releasing a biological product from its previously adsorbed chemically complementary surface. In the present invention, this desorption can be achieved in a variety of ways, including but not limited to reducing the electronegativity of the biological product (AAV), for example by lowering the pH; or by reducing the electropositivity of the solid phase, for example by increasing the pH; or by introducing a competing substance (such as a salt) to disrupt the electrostatic interaction; or any combination of these methods. The desorption of a biological product from a chromatographic medium is commonly referred to as elution.

[0048] The term "selective desorption" refers to a situation where an adsorbed substance is released from the solid phase surface by changing the conditions such that one or more other substances remain adsorbed, and then a different set of conditions causes a different substance to be released from the solid phase surface. The change in conditions can be carried out stepwise. The change in conditions can also be carried out in a continuous manner, resulting in the early desorption of weakly bound substances at the early part of the continuum and the later elution of strongly bound substances at the later part of the continuum, ideally separated from each other.

[0049] The term "washing" refers to the process of exposing a loaded column to a cleaning buffer for the purpose of displacing unbound substances from the solid phase. In this context, the term "rinsed" has the same meaning. In the most basic case, the wash buffer has the same formulation as the equilibration buffer. In more complex configurations, the wash buffer may have the additional role of displacing a portion of the weakly bound contaminants from the solid phase so that they can be removed before the desired product is eluted. Or there may be more than one wash step, where the first employs the same conditions as the equilibration buffer, but the second employs conditions that displace a portion of the weakly bound contaminants from the solid phase so that they are removed before elution.

[0050] The term "elute" or "eluted" refers to the process of changing the chemical environment in which the solid resides such that the interaction between the primary amine-based solid phase and the substance that remains bound after the loading and washing steps is dissociated.

[0051] After removing empty AAV capsids from the solid phase, the intact AAV capsids can be eluted by any of a variety of methods. As described above, the intact AAV capsids can be eluted by further increasing the pH value in the absence or presence of excess salt. After removing empty AAV capsids, the intact AAV capsids can be eluted by increasing the pH value in the presence of salt. After removing empty AAV capsids, the intact AAV capsids can be eluted by introducing excess salt while maintaining the pH constant. After removing empty AAV capsids, the intact AAV capsids can be eluted by introducing excess salt while decreasing the pH value.

[0052] The following general, non-limiting description of a series of basic method options illustrates how to perform variations of the method and provides a platform for a more detailed discussion of operating variables. It should be understood that the buffer conditions mentioned in these scenarios are intended to provide a general concept of how to implement the method, and the buffer formulation needs to be optimized to accommodate mixtures of capsids from different serotypes, or even different recombinant constructs of a single serotype.

[0053] In one embodiment, elution is performed using only an increasing pH gradient. A primary amine-based solid phase in the form of a chromatography device, such as a monolith, is equilibrated to a near-neutral pH value, such as 20 mM Tris, 20 mM bis-Tris propane, pH 7.5 ± 0.2. A sample containing a mixture of empty and intact AAV capsids is equilibrated by exchanging the buffer to 20 mM Tris, 20 mM bis-Tris propane, pH 7.5 ± 0.2. The sample is loaded onto the chromatography device. The chromatography device is then eluted with a linear pH gradient from the equilibration buffer to an end buffer of 20 mM Tris, 20 mM bis-Tris propane, pH 9.5 ± 0.2, over 20 column volumes, or over 50 column volumes, or over 100 column volumes, where the number of column volumes is used as a means to regulate the rate of pH change during the gradient, also known as the gradient slope. This method can be particularly useful for analytical purposes because all empty capsids and intact capsid populations are eluted within the gradient. It may also be valuable as a starting point for developing preparative use conditions. It can also be used for preparative applications, but it should be realized that this method may expose the intact AAV capsids to a range of pH values that may have an adverse effect on capsid stability.

[0054] In one embodiment, elution is carried out using only an increasing salt gradient on a primary amine solid phase in the form of a chromatography device, such as a monolith column, which is equilibrated to a weakly basic value, such as 20 mM Hepes pH 7.5 ± 0.2. A sample containing a mixture of empty and full AAV capsids is equilibrated to the same conditions by buffer exchange. The sample is loaded onto the chromatography device. The chromatography device is then eluted with a linear sodium chloride gradient from an equilibration buffer of 20 mM Hepes, pH 7.5 ± 0.2, to an end buffer of 20 mM Hepes, 200 mM NaCl, pH 7.5 ± 0.2, over 20 device volumes, or over 50 device volumes, or over 100 device volumes, where the number of device volumes is used as a means to adjust the rate of change of salt concentration during the gradient, i.e., the gradient slope. This method can also be used for analytical purposes because, as described above, during sample loading, all empty capsids and full capsids bind to the primary amine solid phase. The degree of separation between empty and full AAV capsids is generally not as good as elution using an increasing pH gradient, but is still better than fractionation using a salt gradient with a strong anion exchanger at a fixed pH.

[0055] In a closely related embodiment, elution is carried out using only an increasing salt gradient on a primary amine solid phase in the form of a chromatography device, such as a monolith column, equilibrated to a weakly basic pH value, such as 10 mM Tris, 10 mM bis-Tris propane, pH 8.7 ± 0.2. A sample containing a mixture of empty and full AAV capsids is equilibrated to the same conditions by buffer exchange and loaded onto the solid surface. The chromatography device is then eluted with a linear sodium chloride gradient from an equilibration buffer of 10 mM Tris, 10 mM bis-Tris propane, 100 mM NaCl, pH 8.7 ± 0.2, over 20 device volumes, or over 50 device volumes, or over 100 device volumes, where the number of device volumes is used as a means to adjust the rate of change of salt concentration during the gradient, i.e., the gradient slope. This method can also be used for analytical purposes because, as described above, during sample loading, all empty capsids and full capsids bind to the primary amine solid phase. The degree of separation between empty and full AAV capsids will be better than that with a salt gradient at a lower fixed pH and generally better than fractionation using a salt gradient with a strong anion exchanger at a fixed pH, but generally worse than elution using an increasing pH value gradient on a primary amine solid phase.

[0056] In an embodiment utilizing mixed pH-salt elution, a primary amine-based solid phase in the form of a chromatography device, such as a monolith column, is equilibrated to a near-neutral pH value, such as 20 mM Hepes, pH 7.0 ± 0.2. The sample containing a mixture of empty and full AAV capsids is equilibrated by exchanging the buffer to 20 mM Hepes, pH 7.0 ± 0.2. The sample is loaded onto the chromatography device. Under these conditions, both empty particles and full particles bind to the solid phase. The device is washed with the equilibration buffer to remove unbound material. The chemical environment is changed by exposing it to a buffer with a higher pH value (such as 20 mM Tris, pH 8.5 ± 0.2), thereby washing / eluting the empty AAV capsids from the device. The buffer is flowed through the chromatography device until the UV absorbance of the column effluent reaches the baseline. After removing the empty AAV capsids from the device, the chemical environment is changed again by exposing it to a buffer with a lower pH value but containing an excess of salt (such as 20 mM Hepes, 50 mM NaCl, pH 7.0). This method may be advantageous for preparative use due to its mild elution conditions for full capsids. It is also considered by some users to be an example of a favorable step gradient elution form.

[0057] In another embodiment employing mixed pH-salt elution, a primary amine-based solid phase in the form of a chromatography device, such as a monolith column, is equilibrated to an alkaline pH, such as 20 mM Tris, pH 8.5 ± 0.2. The sample containing a mixture of empty and full AAV capsids is equilibrated by exchanging the buffer to 20 mM Tris, pH 8.5 ± 0.2. The sample is loaded onto the chromatography device. Under these conditions, the full particles bind to the solid phase, but the empty particles cannot do so. After removing the empty AAV capsids from the device, the chemical environment is changed again by exposing it to a buffer with a lower pH value but containing a small amount of salt (such as 20 mM Hepes, 50 mM NaCl, pH 7.0). This method is procedurally simpler than before and is thus attractive for preparative use. However, it may reduce the binding ability of full AAV capsids because the equilibration conditions close to the threshold interfere with the binding of full capsids.

[0058] In embodiments where the intact AAV capsid is eluted with pH and salt in another manner, a primary amine-based solid phase in the form of a chromatography device, such as a monolith column, is equilibrated to a near-neutral pH value, such as 20 mM bis-Tris propane, pH 8.5 ± 0.2. A sample containing a mixture of empty and intact AAV capsids is equilibrated by exchanging the buffer to 20 mM bis-Tris propane, pH 8.5. The sample is loaded onto the chromatography device. Then the chromatography device is eluted with a linear salt gradient, from the equilibration buffer (pH 7.5), to an end buffer of 20 mM bis-Tris propane, 100 mM NaCl, pH 8.5, over 20 device volumes, or over 50 device volumes, over 100 device volumes, where the number of device volumes is used as a means to adjust the rate of change of the salt concentration during the gradient, i.e., the gradient slope. Preliminary data indicate that the salt gradient supports the separation of empty and intact AAV capsids, similar to the separation achieved by strong anion exchangers, and the conditions are milder compared to strong anion exchangers, but still inferior compared to pH gradients. In most cases, the higher the pH value of the salt gradient, the better the resolution between empty and intact capsids.

[0059] It will be apparent to those skilled in the art that one or more corresponding elements from any of the above basic forms can be applied in other forms. For example, an elution gradient can be applied while changing both the pH value and the salt concentration, with both parameters increasing, or one or the other parameter increasing while the other decreases; a single-step elution can be converted to a multi-step elution; a linear gradient elution can be converted to a step elution; a linear gradient elution can be applied after a step that significantly reduces the amount still bound but does not completely eliminate them; a linear gradient elution can be applied after loading under equilibration conditions that result in most of the empty AAV capsids not binding. These variations and many others can be achieved without departing from the true nature of the defined elements and methods.

[0060] In one embodiment, the primary amine solid phase and the sample are equilibrated to an operating pH in the range of pH 5.5 ± 0.2 to pH 8.5 ± 0.2, or pH 6.0 ± 0.2 to pH 8.5 ± 0.2, or pH 6.5 ± 0.2 to pH 8.5 ± 0.2, or pH 7.0 ± 0.2 to pH 8.5 ± 0.2, or pH 7.5 ± 0.2 to pH 8.5 ± 0.2. It should be understood that the exact conditions may vary depending on the different serotypes of AAV and may also vary depending on the different recombinant constructs within a serotype. For samples known to tolerate a higher pH range, the pH range can be extended to higher values. Generally, the higher the pH value within this range, the fewer empty AAV capsids will bind. Any AAV that has not been tried before requires a simple experiment to determine the pH value at which empty AAV capsids cannot bind while essentially all intact AAV capsids bind. The lower the pH value within this range, the stronger the intact AAV capsids will bind and the higher the binding capacity of the intact capsids. However, for experienced personnel, empty AAV capsids may compete with intact AAV capsids for binding space, and this competition may limit the binding capacity of the intact capsids. These are routine process optimization issues that are considered during the development of all industrial purification methods.

[0061] In some embodiments, if the binding of empty capsids is not halted at pH 8.5 ± 0.2 and higher pH values that may compromise capsid stability are to be particularly avoided, salts can be added to the equilibration buffer, such as 2 mM NaCl, or 5 mM NaCl, or 10 mM NaCl, or 20 mM NaCl, or concentrations in between or higher. Achieving the lowest salt concentration that halts the binding of empty capsids at a moderate pH is generally most advantageous. If it is desired to evaluate the use of higher concentrations to displace empty AAV capsids, such as 30 mM, or 40 mM, or 50 mM, or 60 mM, or 70 mM, or concentrations in between or higher, they can be carefully evaluated to avoid accidental displacement and loss of intact capsids. If needed, this approach can be considered to reduce empty capsid binding, particularly at acidic, neutral, or very mildly basic pH values.

[0062] Many methods for equilibrating the sample to the conditions for loading the sample onto the chromatography column are known to those skilled in the art. Any of these methods can be used without changing the essence of the method. Among these methods are dialysis at the laboratory scale, tangential flow filtration, and buffer exchange chromatography. In some cases, sufficient sample equilibration can be achieved by titrating the sample to the target pH value and diluting the sample with water or a low-salt or salt-free buffer if necessary.

[0063] Equilibration provides an opportunity to prevent binding of empty AAV capsids, but the conditions for doing so may weaken binding of intact AAV capsids, thereby reducing the solid phase's ability relative to intact capsids. Since a particular goal of the preparation of the separation is to accommodate the maximum practical loading of the desired intact capsids, washing provides a potentially less lossy method to minimize the content of empty capsids prior to elution. Given the most effective washing conditions, equilibration conditions can be selected that favor high-capacity binding of the desired intact capsids, which generally corresponds to a neutral or lower pH and minimal or no additional salt.

[0064] In one embodiment, a primary amine solid phase loaded with a mixture of empty AAV capsids and intact AAV capsids is washed with a buffer having the same composition as the buffer used for equilibration of the solid phase.

[0065] In another embodiment, a primary amine solid phase loaded with a mixture of empty AAV capsids and intact AAV capsids is washed with a buffer having a higher pH than the equilibration buffer. For example, a primary amine solid phase equilibrated and loaded at pH 7.5 may first be washed with a buffer having the same composition as the buffer used for equilibration of the solid phase and then washed with a buffer of higher pH, such as pH 8.0 ± 0.2, or pH 8.1 ± 0.2, or pH 8.2 ± 0.2, or pH 8.3 ± 0.2, or pH 8.4 ± 0.2, or pH 8.5 ± 0.2, or pH 8.6 ± 0.2, or pH 8.7 ± 0.2. It should be understood that the exact conditions may vary for different AAV serotypes and may also vary for different recombinant constructs within a serotype. If empty AAV capsids are not completely removed by increasing the pH value, salt can be added. For example, for a buffer at pH 8.5 ± 0.2, NaCl can be added in an amount corresponding to a concentration of 2 mM, or 5 mM, or 10 mM, or 20 mM or higher or in between.

[0066] In related embodiments, the method of washing with conditions different from the sample equilibrium pH conditions can be performed in an alternative manner that simplifies the overall method. The sample and the chromatography column can be equilibrated to different conditions. Specifically, the sample can be equilibrated to a pH and / or salt concentration that supports the highest capacity binding of intact capsids, while the column can be equilibrated to conditions that can substantially remove empty AAV capsids prior to the step of eluting the intact capsids. For example, the sample can be equilibrated with 50 mM Hepes, pH 7.5 ± 0.2, while the chromatography column is equilibrated with 50 mM Tris, pH 8.5 ± 0.2. When the sample is loaded onto the chromatography column, it has the effect of at least partially re-equilibrating the column to accommodate the conditions of the sample. At the end of the sample loading, the equilibrium / washing buffer at the higher pH condition will immediately start to displace the empty AAV capsids bound during the loading process, thus eliminating the need to include a second washing step. Buffer preparation and the overall method will be simplified.

[0067] In one embodiment, the intact AAV capsids are eluted only by increasing the pH. Depending on the serotype of the AAV capsid, this may include a single-step elution, or a multi-step gradient elution, or a linear gradient elution to pH 8.6 ± 0.2, or 8.7 ± 0.2, or 8.8 ± 0.2, or 8.9 ± 0.2, or 9.0 ± 0.2, or 9.1 ± 0.2, or 9.2 ± 0.2, or 9.3 ± 0.2, or 9.4 ± 0.2, or 9.5 ± 0.2, or lower, intermediate or higher values.

[0068] It is recognized that pH values approaching and exceeding 9.0 ± 0.2 increase the risk of damaging intact AAV capsids, and thus the pH of the elution can be moderated by adding salt. For example, a stepwise or gradient elution up to pH 9.5 ± 0.2 can be replaced by a stepwise or gradient elution up to pH 8.5 ± 0.2 in the presence of 2 mM NaCl, or 5 mM NaCl, or 10 mM NaCl, or 15 mM NaCl, or 25 mM NaCl, or lower, intermediate or higher values. Alternatively, by adding salt, the pH stepwise or gradient can be performed at pH 8.0 ± 0.2, or at pH 7.5 ± 0.2, or at pH 7.0 ± 0.2, or at pH 6.5 ± 0.2, or at pH 6.0 ± 0.2, or at pH 5.5 ± 0.2, or at lower, intermediate or higher values. It will be apparent to those skilled in the art that the more salt is added, the lower the pH. If desired, the amount of salt added at any pH value can be increased to any value within the range of 2.5 mM to 250 mM, such as 2.5 mM, or 5 mM, or 12.5 mM, 25 mM, or 50 mM, or 100 mM, or 150 mM, or 200 mM, or 250 mM, or lower, intermediate or higher values.

[0069] Any of the above embodiments can be carried out in the presence of divalent metal cations such as magnesium and / or calcium, wherein the concentration of either is from 0.1 mM to 10 mM, or 0.5 mM to 5 mM, or 1.0 mM to 2.75 mM, or 1.5 mM to 2.5 mM, or 1.75 mM to 2.25 mM, or 2.0 mM ± 0.1 mM, or higher values, such as up to 5 mM.

[0070] In one embodiment, from an analytical perspective, the inclusion of divalent metal cations such as magnesium and / or calcium may be advantageous because the metal ions help maintain the integrity of empty AAV capsids and tend to result in empty capsids eluting as a single peak. This makes it easier to measure the total amount of empty AAV capsids in a given sample as well as the relative amounts of empty and full AAV capsids.

[0071] In some embodiments, the presence of divalent metal cations such as magnesium and / or calcium may also be advantageous for preparation of separation because, in addition to often optimizing their separation from empty capsids, they stabilize the full capsids.

[0072] The presence of magnesium is generally expected to yield more desirable results in terms of AAV purity and / or recovery. However, other embodiments may include the absence or lack of divalent metal ions such as magnesium and / or calcium. This may be especially the case where the full AAV capsids exhibit sufficient stability under the selected separation conditions; and, the overall distribution of contaminants is such that fractionation in the absence of divalent metal cations can isolate the full capsid fraction with fewer contaminating empty AAV capsids compared to separation in the presence of such divalent cations.

[0073] To stabilize full capsids, and / or increase capsid solubility, and / or inhibit non-specific interactions between the capsids and the solid phase that may reduce separation efficiency or lower the recovery of full capsids, and / or simply reduce conductivity during separation, any of the foregoing embodiments can be carried out in the presence of arginine or histidine.

[0074] In some embodiments, arginine or histidine can directly replace NaCl or other salts for gradient separation. The water solubility limit of arginine is approximately 600 mM, and the solubility limit of histidine is approximately 200 mM. Arginine is more convenient to use due to its higher solubility, but histidine can be effectively used in many cases even though its solubility is more limited.

[0075] In embodiments where arginine or histidine concentrations sufficient to elute the capsid cannot be achieved, both can be combined with salts, such as combinations including arginine plus sodium chloride or histidine plus sodium chloride. In some such embodiments, a baseline level of arginine can be added to the equilibration, wash, and / or elution buffer, or only to the wash and elution buffers, or only to the elution buffer. Such a baseline level may be 25 mM or 50 mM, or an intermediate concentration or higher within its solubility range.

[0076] In another embodiment, arginine or histidine can be used across a rising pH gradient at a baseline level, such as at a concentration of 5 mM or 10 mM, or an intermediate concentration or higher within its solubility range.

[0077] Any of the above embodiments can be carried out in the presence of low molecular weight zwitterions such as glycine or alanine or betaine. Such additives can have the effect of improving capsid solubility and stability without increasing conductivity. Such zwitterions increase polarity but do not contribute to conductivity. They are also preferentially excluded from the protein surface, which gives them a stabilizing effect. Such additives may be particularly suitable for use in combination with a pH gradient to overcome the desolvation effects of low conductivity, where they may have an effect on improving peak sharpness and / or capsid stability.

[0078] In some embodiments, betaine can be used in the range of pH 5.0 to pH 10.0, while alanine and glycine will be restricted to the range of 5.0 to about 8.5. At pH values above 8.5, the nitrogen atoms of glycine and alanine begin to lose their positive charge and the molecules become net electronegative. This causes them to contribute to the conductivity of the sample and also suspends other positive effects of their zwitterionic form. The nitrogen atom on betaine is a quaternary amine that can maintain its charge up to about pH 12, thus maintaining its zwitterionic form throughout the range.

[0079] In one embodiment where the operating pH is 8.5 or lower, the concentration of glycine or alanine can be up to 2 M or higher, or 0.5 M to 2.5 M, or 1.0 M to 2.0 M, or 1.25 M to 1.75 M, or 1.4 M to 1.6 M, or about 1.5 M, or different ranges.

[0080] In one embodiment, when the pH is less than 10, the concentration of betaine can be up to 2 M or higher, or 0.5 M to 2.5 M, or 1.0 M to 2.0 M, or 1.25 M to 1.75 M, or 1.4 M to 1.6 M, or about 1.5 M, or different ranges.

[0081] Any of the above embodiments can be carried out in the presence of sugars, such as sugars intended to stabilize the intact capsid. Examples include but are not limited to sucrose, sorbitol, xylose, mannitol, trehalose, or other nonionic sugars in a concentration range of 0.5% to 25% or higher.

[0082] Any of the above embodiments can be carried out in the presence of glycerol, such as glycerol used to stabilize the intact capsid, for example, at a concentration of 0.5% to 25% or higher.

[0083] Any of the above embodiments can be carried out in the presence of organic additives to minimize non-specific interactions between the capsid and the primary amine-based solid phase surface. Such additives include surfactants in a concentration range of 0.01% to 1.0%, including nonionic surfactants. Such additives also include reagents in a concentration range of 1% to 10%, such as ethylene glycol or propylene glycol.

[0084] Those skilled in the art will recognize that the methods of the present invention can be implemented on ligands other than primary amine groups, which either completely lack more highly derivatized amines such as secondary amines, tertiary amines, or quaternary amines, or contain a very low proportion. Similarly, it can be recognized that the results are enhanced by including hydrophobic residues or residues that mediate more increased hydrogen bonding between the capsid and the solid phase.

[0085] All references cited herein are incorporated herein by reference to the extent that their incorporated content is not inconsistent with the explicit teachings herein.

[0086] The present invention is further illustrated by the following non-limiting examples. Examples

[0087] Example 1

[0088] The separation of empty AAV capsids from intact AAV capsids was compared by salt elution on a quaternary amino solid phase and a primary amine solid phase.

[0089] A 100 μL monolithic column with a primary amine surface was equilibrated with 20 mM bis-Tris propane, 2 mM magnesium chloride, pH 8.8. A cation exchange-purified AAV 8 sample was applied to the column and washed with the equilibration buffer. The monolithic column was then eluted with a 50 bed volume linear gradient with an end point of 20 mM bis-Tris propane, 2 mM magnesium chloride, 100 mM NaCl, pH 8.8. The same basic method was carried out on a quaternary amine (strong anion exchange monomer, CIMac QA), except that the NaCl concentration in the gradient end buffer was 200 mM. Neither column could completely separate the empty and intact AAV capsids, but the primary amine solid phase gave better results ( Figure 1 ).

[0090] Example 2

[0091] Comparison of the separation of empty / infectious virus capsids by elution with increasing pH of a primary amine solid phase and a strong (quaternary amine) anion exchanger using a salt gradient.

[0092] A 100 μl monolith column with a primary amine surface was equilibrated with 10 mM bis-Tris propane, 10 mM Tris, 2 mM magnesium chloride, pH 8.0. A cation exchange purified AAV8 sample was applied to the column and washed with the equilibration buffer. The column was then eluted with a 100 bed volume linear gradient with an end point of 10 mM bis-Tris propane, 10 mM Tris, 2 mM magnesium chloride, pH 9.5. The separation of empty and infectious AAV capsids was compared to that of a strong anion exchange (quaternary amine) monolith column eluted with a salt gradient. The strong anion exchange monolith column was equilibrated with 20 mM bis-Tris propane, 2 mM magnesium chloride, pH 9.0. A cation exchange purified AAV8 sample was applied to the column and washed with the equilibration buffer (the same sample used for the primary amine column experiment). The column was then eluted with a 50 column volume linear gradient with an end point of 20 mM bis-Tris propane, 2 mM magnesium chloride, 200 mM sodium chloride, pH 9.0. The results are shown in Figure 2 Figure []. As shown, in the absence of excess salt, elution of the primary amine column with an increasing pH gradient was significantly superior to the separation achieved with a strong anion exchanger using a salt gradient. In addition to demonstrating excellent separation of empty and infectious capsids, the elution order of other contaminants also demonstrated a completely different selectivity of the primary amine column compared to the strong anion exchanger. Contaminants that eluted before the empty AAV capsids on the strong anion exchanger eluted after the infectious AAV capsids on the primary amine column and were better separated.

[0093] Example 3

[0094] Removal of bound empty AAV capsids by a wash step prior to pH gradient elution of infectious capsids.

[0095] A primary amine monolith column was equilibrated with 10 mM bis-Tris propane, 10 mM Tris, 2 mM magnesium chloride, pH 8.0. A cation exchange purified AAV8 sample was applied to the column and washed with the equilibration buffer (the same sample used for Example 2). Empty AAV capsids were then removed with a wash step of 10 mM bis-Tris propane, 10 mM Tris, 2 mM magnesium chloride, pH 8.7. The column was then returned to its original equilibration conditions and eluted with a 40 bed volume linear gradient with an end point of 10 mM bis-Tris propane, 10 mM Tris, 2 mM magnesium chloride, pH 9.5. The results are shown in Figure 3As shown. This experiment demonstrated that primary amine columns bind empty AAV capsids and intact AAV capsids at a weakly basic pH (8.0), and then support selective removal of empty AAV capsids with an increasing pH gradient (up to pH 9.5), all in the absence of excess salt.

[0096] Example 4

[0097] Equilibrate under conditions where empty AAV capsids do not bind, and then elute intact AAV capsids in one step while increasing the salt concentration and decreasing the pH value.

[0098] Equilibrate the primary amine monolith with 10 mM bis-Tris propane, 10 mM Tris, 2 mM magnesium chloride, pH 8.5. Apply the cation exchange-purified AAV 8 sample to the column and wash with the equilibration buffer (the same as the samples used in Examples 2 and 3). Under these conditions, empty AAV capsids mostly fail to bind and are almost completely eliminated during a long wash. Then elute the intact AAV capsids from the column in steps of 10 mM bis-Tris propane, 10 mM Tris, 2 mM magnesium chloride, 50 mM NaCl, pH 7.5. The results are as Figure 4 shown. This experiment specifically documented the ability of the primary amine column to separate empty AAV capsids from intact AAV capsids using a highly simplified step gradient method to accommodate the simplest chromatographic instruments.

[0099] Example 5

[0100] The following examples illustrate the method of the present invention, starting at pH 8 and using different methods to separate empty AAV capsids by increasing the pH.

[0101] It is understood that capsids of different serotypes respond differently to the same conditions and that conditions need to be adjusted individually for each different product. This series of figures illustrates how easy it is to screen capsids from any serotype to obtain the most effective conditions.

[0102] In a series of experiments, samples containing empty and intact AAV8 capsids were applied to a solid phase with a primary amine group at pH 8.0 (20 mM Tris, 20 mM bis-Tris propane). In each experiment, the loaded column was washed with a buffer of higher pH (20 mM Tris, 20 mM bis-Tris propane), and then eluted with an increasing pH gradient, ending at pH 9.5 (20 mM Tris, 20 mM bis-Tris propane). The original chromatograms are shown as ( Figures 5 - 10 ). The relative recovery of empty AAV capsids and intact AAV capsids from the gradient is as Figure 11 shown.

[0103] Figure 5.Both empty and full AAV capsids bind at pH 8.0. During the entire washing process, both remain bound at pH 8.3. The full capsids, as indicated by their UV absorbance ratio of 0.68, start eluting immediately after application of the pH elution gradient. There is a distinct but partial separation from the full capsids that elute later, showing a distinct wavelength ratio of 1.32.

[0104] Figure 6 .Both empty and full AAV capsids bind at pH 8.0. Some empty AAV capsids are removed by washing at pH 8.4. As the pH elution gradient is applied, the remaining empty AAV capsids start eluting immediately. The full AAV capsids remain bound and elute within the pH gradient.

[0105] Figure 7 :Both empty and full AAV capsids bind at pH 8.0. Most empty AAV capsids are removed by washing at pH 8.5. As the pH elution gradient is applied, the remaining empty AAV capsids start eluting immediately. The full AAV capsids remain bound and elute within the pH gradient.

[0106] Figure 8 :Both empty and full AAV capsids bind at pH 8.0. All empty AAV capsids are removed by washing at pH 8.6. The full AAV capsids remain bound and elute within the pH gradient.

[0107] Figure 9 :Both empty and full AAV capsids bind at pH 8.0. All empty AAV capsids are removed by washing at pH 8.7, but some full AAV capsids elute during the washing process. The remaining full AAV capsids remain bound and elute within the pH gradient.

[0108] Figure 10 :Both empty and full AAV capsids bind at pH 8.0. All empty AAV capsids are removed by washing at pH 8.8, but most of the full AAV capsids are also removed. The remaining full AAV capsids elute within the pH gradient.

[0109] Figure 11 :This figure illustrates the relative amounts of full and empty AAV capsids eluted in a gradient after washing steps at different pH values. The shaded vertical bars indicate the washing pH values that support the most effective elimination of empty AAV capsids and the best recovery of full capsids.

[0110] Example 6

[0111] Alter selectivity by the presence or absence of magnesium ions.

[0112] Figure 12The elution curves between two chromatographic experiments on a primary amine-containing solid phase were compared, where one curve was generated with 2 mM magnesium ions in two gradient buffers, while the other curve was generated without magnesium ions. Additionally, the buffers used for the two experiments were the same. The equilibration buffer was 10 mM Tris, 10 mM bis-Tris propane, 15 mM NaCl, pH 7.0 (with or without 2 mM magnesium chloride). The gradient endpoint was 10 mM Tris, 10 mM bis-Tris propane, 15 mM NaCl, pH 9.5, (with or without 2 mM magnesium chloride). In the presence of magnesium, the intact capsids eluted at approximately pH 8.77 (peak center). In the absence of magnesium, the intact capsids eluted at approximately pH 9.23 (peak center), approximately half a pH unit higher than in the presence of magnesium ions. In the presence of magnesium ions, the separation between the empty capsids and the intact capsids was also more distinct, and the recovery of the intact capsids was increased by approximately 50%.

[0113] References

[0114] All references cited herein are incorporated herein by reference to the extent that such incorporation is not inconsistent with the express teachings herein.

[0115] [1] M. Lock, M. Alvira, J. Wilson, Analysis of the particle content of recombinant adeno-associated virus serotype 8 vectors by ion-exchange chromatography, Human Gene Therapy Methods: Part B 23 (2012) 56 - 64.

[0116] [2] M. Lock, M. Alvira, Scalable purification method for AAV9, US Patent Application US20190002842A1, priority US201562266357P, WO2017160360A9

[0117] [3] X. Fu, W.C. Chen, C. Argento, P. Clarner, V. Bhatt, R. Dickerson, G. Bou-Assaf, M. Bahshaveshi, X. Lu, S. Bergelson, J. Pieracci, Quantitative analysis strategies for empty AAV capsids to support process development in adeno-associated virus, Human Gene Therapy. 30 (2019) 144–152.

Claims

1. A method for separating or removing empty AAV capsids from intact AAV capsids in an aqueous mixture containing empty and intact AAV capsids, wherein the mixture is contacted with a solid phase surface bearing a primary amino group in a first alkaline environment, whereby (i) intact AAV capsids bind to the solid phase surface while empty AAV capsids are at least partially unbound to the solid phase surface, and the bound empty AAV capsids are at least partially eluted by a second alkaline environment having a pH higher than the pH of the first alkaline environment, provided that the second alkaline environment does not elute intact AAV capsids from the solid phase surface; or (ii) both intact and empty AAV capsids bind to the solid phase surface, and subsequently the empty AAV capsids are at least partially eluted by a second alkaline environment having a pH higher than the pH of the first alkaline environment, provided that the second alkaline environment does not elute intact AAV capsids from the solid phase surface; wherein the pH of the second alkaline environment is in the range of pH 8.0 to pH 9.

0.

2. The method according to claim 1, wherein the pH of the first alkaline environment is pH 7 to pH 8.

3. The method according to claim 1, wherein after at least partially eluting the empty AAV capsids with the second alkaline environment, the solid phase surface is contacted with a third alkaline environment having a pH greater than the pH of the second alkaline environment.

4. The method according to claim 3, wherein the solid phase surface is contacted with a fourth alkaline environment having a pH less than the pH of the third or second alkaline environment but having a higher salt concentration compared to the salt concentration of the first, second, or third alkaline environment.

5. The method according to claim 1, wherein the pH of the second alkaline environment is in the range of pH 8.1 to pH 8.

9.

6. The method according to claim 1, wherein the pH of the second alkaline environment is in the range of pH 8.2 to pH 8.

8.

7. The method according to claim 1, wherein the pH of the second alkaline environment is in the range of pH 8.3 to pH 8.

7.

8. The method according to claim 1, wherein the pH of the second alkaline environment is in the range of pH 8.4 to pH 8.

6.

9. The method according to claim 3, wherein the pH of the third alkaline environment is in the range of pH 8.5 to pH 10.

5.

10. The method according to claim 3, wherein the pH of the third alkaline environment is in the range of pH 8.5 to 10.

0.

11. The method according to claim 3, wherein the pH value of the third alkaline environment is in the range of pH 8.5 to 9.

5.

12. The method according to claim 4, wherein the salt concentration contained in the first, second, third, and / or fourth alkaline environment corresponds to the concentration of an alkali metal salt of at most 1 M.

13. The method according to claim 4, wherein the salt concentration contained in the first, second, third, and / or fourth alkaline environment corresponds to the concentration of an alkali metal salt of 1 mM to 1,000 mM.

14. The method according to claim 4, wherein the salt concentration contained in the first, second, third, and / or fourth alkaline environment corresponds to the concentration of an alkali metal salt of 2.5 mM to 250 mM.

15. The method according to any one of claims 12-14, wherein the alkali metal salt is NaCl.

16. The method according to any one of claims 1-4, wherein the alkaline environment contains magnesium ions in a concentration range of 1.0 mM to 5.0 mM.

17. The method according to any one of claims 1-4, wherein the alkaline environment contains magnesium ions in a concentration range of 1.5 mM to 3.0 mM.

18. The method according to any one of claims 1-4, wherein the alkaline environment contains magnesium ions in a concentration range of 2.0 mM to 2.5 mM.

19. The method according to claim 1, wherein the full and empty AAV capsids have any serotype.

20. The method according to claim 19, wherein the full and empty AAV capsids have a serotype selected from the group consisting of natural and recombinant serotypes, chimeric types, mixed types, or combinations thereof.

21. The method according to claim 1, wherein the solid phase surface with a primary amino group is disposed in a chromatography device.

22. The method according to claim 1, wherein the solid phase surface with a primary amino group is a monolithic column, a packed particle column, a packed nanofiber column, a membrane adsorber, or a hydrogel.

Images