TW202511280A - Methods of preparing surface modified viral capsids - Google Patents

Abstract

The present disclosure relates to compositions of surface modified viral capsids having high physical titers. The present disclosure also relates to a method of preparing surface modified viral capsids using an efficient process of (a) reacting (i) a viral capsid comprising a plurality of surface available primary amines and (ii) a capsid-reactive linker comprising a terminal tetrafluorophenyl (TFP) ester and a first member of a crosslinker reactive pair to provide a composition comprising surface functionalized viral capsid; and (b) reacting the composition comprising surface functionalized viral capsid with a functionalized ligand comprising a second member of the crosslinker reactive pair.

Description

Method for preparing surface-modified virus shell

The introduction of molecules carrying genetic information into cells is a useful tool in modern medicine and basic research. Preferred methods include the use of gene delivery vehicles derived from viruses including adenovirus, retrovirus, lentivirus, vaccinia virus, and adeno-associated virus. Among these viruses, recombinant adeno-associated virus (AAV) has become a preferred virus for in vivo gene therapy due to its lack of pathogenicity, inability to replicate, and stable expression. More than 100 clinical trials are underway using AAV-based envelopes (also called vectors or virions), and the FDA recently approved two AAV gene therapy products, voretigene neparvovec-rzyl (LUXTURNA) for the treatment of genetic retinal diseases and onasemnogene abeparvovec-xioi (ZOLGENSMA) for the treatment of spinal muscular atrophy.

However, most clinical trials using AAV as a transgene delivery vehicle have demonstrated its key limitations: (i) its reduced therapeutic index (i.e., high doses of the encapsidate are often required to achieve therapeutic efficacy); (ii) its broad biodistribution; and (iii) its poor potency in the pre-existence of neutralizing antibodies.

One limitation of AAV is actually its tropism, which results in transgene expression in tissues other than those in which it is desired. It is also well understood that both host and vector-related immune challenges need to be overcome for long-term gene transfer.

Most gene therapy applications to date have used serotype 2 (AAV2). Its ubiquity is partly explained by the transduction of many postmitotic cells in vivo in mammals (e.g., muscle cells, hepatocytes, or neurons). This serotype is also used to transfer genes into muscle and liver in clinical trials for hemophilia B, and into the retina to treat Leber Congenital Amaurosis.

However, there are still complications and limitations to the use of this vector. First, since the transduction efficiency is generally low in vivo, high doses of the vector are usually required, resulting in increased toxicity. One of the most important complications is due to the fact that 50-90% of the human population is seropositive for AAV2 and has produced neutralizing antibodies (NAbs) against AAV2 that impair gene delivery. The discovery of natural AAV isolates (1 to 12) in human and animal species and the genetic modification of the capsids of these AAV serotypes using molecular tools have produced promising results in preclinical animal models and Phase I/II clinical trials, which has promoted exciting clinical translation in the near future. However, its therapeutic index is still low, which means that high concentrations of the vector still need to be administered and are accompanied by adverse effects. At the same time, clinical batch manufacturing of AAV capsids has greatly improved in the past decade and large-scale manufacturing methods are now available for gene therapy to enter the pre-industrial pharmaceutical stage. Nevertheless, if high doses of capsids are required for Phase III and commercialization, current methods will not be able to support the demand.

Since these strategies for demonstrating the potential of novel AAV serotypes (and related genetic variants) are not considered satisfactory due to the inability to achieve precise tropism that would enable selective transduction of target cell types, further efforts are currently being made to increase the tropism of AAV-derived capsids. In fact, to counteract the lack of specificity of AAV-derived capsids, extremely large amounts of AAV-derived capsids need to be administered to reach therapeutic thresholds, which is undesirable due to safety concerns and manufacturing limitations.

Various attempts have been made for this purpose, such as the genetic introduction of peptide epitopes with targeting specificities onto the viral surface. Other strategies use linker molecules with two specificities (e.g., bispecific antibodies), one for the viral coat and the other for the receptor; the introduction of an adapter domain (Z domain of protein A-biotin) for non-covalent attachment of protein ligands.

For example, in document WO00/002654, the alteration of tropism is mainly to prevent AAV from binding to the viral receptors of the original target cells. In a specific embodiment, this document also mentions an increased affinity for the target cells. Still in this document, the antibody fragment is linked to the capsid. According to an alternative embodiment, the other end of the antibody can be coupled to a ligand to improve the affinity for the target (preparation of a "double-chain antibody").

Combinations of biological and chemical conjugation on AAV capsids have also been proposed in the past to improve the selectivity of AAV-derived capsids for target tissues.

For example, WO2005/106046 proposes a method for combining genetic modification and chemical modification of the shell. The chemical modification as the second stage of the process relies on the presence of residual cysteine, the shell of which enriches the genetic pathway in the first step. Thus, AAV particles can be modified with ligands, polymers, gold nanoparticles, fluorescent molecules, magnetic or biochemically active substances. However, the coupling is carried out via disulfide, thioester and/or thioether bonds and NCS bonds.

The article by E. D. Horowitz et al., “Glycated AAV Vectors: Chemical Redirection of Viral Tissue Tropism” Bioconjugate Chemistry, 2011, 22, 529-532, addresses the issue of cell tropism and other technical issues. Specifically, the generation of non-natural amino acid side chains via capsid glycosylation is used as an orthogonal strategy to engineer AAV capsids to display novel tissue tropisms for gene therapy applications.

In WO2015/062516, non-natural amino acids (e.g., amino acids including an azido group) are genetically engineered into the capsid prior to the coupling step by click chemistry to alter the AAV capsid and its tropism for target cells.

WO 2022/101363 describes a method for preparing surface-modified viral exosomes. However, the use of the described NHS esters for surface functionalization requires substantial purification, which greatly reduces the potency of the resulting composition.

Therefore, there is a need to find methods to modify AAV-derived capsids by chemical coupling to increase their ability to target specific organs or tissues, especially by in vivo gene delivery.

It would also be desirable to modify the AAV-derived capsid at a step that does not require altering the AAV amino acid capsid sequence.

Additionally, there is a need for new surface-modified AAV-derived capsids with improved virus-mediated gene transfer to specific cell types.

More generally, new methods are needed to chemically couple ligands of any nature to AAV-derived capsid surfaces (i.e., various chemical moieties), for example, to improve "specific activity" and/or "therapeutic index" and thereby allow the therapeutic dose to be reduced.

The present invention provides novel recombinant adeno-associated virus (rAAV) vector particles, methods for their production, and their therapeutic and/or diagnostic uses.

In one aspect of the invention, a composition comprising a surface-modified viral exosome having a titer of at least 1.0E+10 vg/ml is provided, wherein the surface-modified viral exosome comprises a ligand covalently coupled to the viral exosome via a linker.

In some embodiments, the surface-modified viral exosome has Formula V: Formula V wherein: is a viral coat including a nucleic acid payload, as appropriate; ^SP1 and ^SP2 are a bond or a spacer, respectively; Q is a cross-linking moiety; and L is a ligand.

In another aspect of the present invention, a method for preparing a composition comprising a surface-modified viral exosome of Formula V is provided, comprising the following steps: (a) combining (i) a viral exosome comprising a plurality of surface-available primary amines with (ii) an exosome-reactive linker comprising a tetrafluorophenyl (TFP) ester and a first member (CRP1) of a crosslinker-reactive pair, thereby providing a composition comprising a surface-functionalized viral exosome; and (b) combining (i) a functionalized ligand comprising a second member (CRP2) of a crosslinker-reactive pair with (ii) a composition comprising a surface-functionalized viral exosome; thereby providing a surface-modified viral exosome.

The preferred features of each aspect of the present invention are the same as those of each of the other aspects after appropriate modification. The referenced documents mentioned herein are incorporated herein to the maximum extent permitted by law. Although the content of the present invention and its advantages have been described in detail, it should be understood that various changes, substitutions and modifications can be made herein without departing from the spirit and scope of the present invention as defined by the scope of the attached application.

4.1. Definition

As used herein, the term "rAAV" refers to a recombinant virion comprising a recombinant nucleic acid construct encapsidated within an AAV capsid.

The terms "AAV," "adeno-associated virus," "AAV virus," "AAV virion," "AAV virion," "AAV particle," "adeno-associated virus vector," and "AAV vector" are used herein synonymously with rAAV.

Recombinant nucleic acid constructs (synonymously "recombinant viral genomes") include a polynucleotide cargo (synonymously "cargo") positioned between the inverted terminal repeat sequences of AAV. The cargo can be an expressible polynucleotide or DNA construct that provides a template for homology-directed repair. In various embodiments, the polynucleotides can be expressed to encode a protein (e.g., a transgene-encoded therapeutic protein) or to encode a miRNA, siRNA, or guide RNA for gene editing or RNA editing machinery (e.g., CRISPR, ADAR, and ADAT).

The term "tropism" as used herein refers to the preferential infection and/or transduction of certain cells or tissues by a viral capsid. In a preferred embodiment, to modify the tropism of an AAV capsid, the capsid is endowed with certain characteristics that it does not naturally possess (e.g., a certain affinity for a receptor located on the surface of a target cell).

As used herein, viral titers are expressed in scientific notation using a simplified form that can be more reliably reproduced by digital text recognition, such as 1.0E+3 vg/m. Those skilled in the art will understand that 1,000 vg/m, 1.0 × 10^3 vg/m, and 1.0E+3 vg/m are all equivalent representations of the same quantity.

As used herein, the term "MOI" or "multiplicity of infection" represents the ratio of the number of viral particles to the number of host cells in a given infection medium. A value of MOI = 1 means that for a single phage particle there is, on average, a single host cell. 4.2. Other Interpretation Provisions

Although methods and materials similar or equivalent to those described herein can be used in the practice or testing of the subject methods and compositions, suitable methods and materials are described below. In addition, the materials, methods, and examples are illustrative only and are not intended to be limiting. All publications, patent applications, patents, and other references mentioned herein are incorporated herein by reference in their entirety.

It is important to note that, unless the context clearly dictates otherwise, the singular forms "a," "an," and the like as used herein and in the appended claims include plural referents. Thus, for example, reference to "an antibody or antigen-binding fragment" includes a plurality of such antibodies and antigen-binding fragments and reference to "the recombinant adeno-associated virus" includes reference to one or more recombinant adeno-associated viruses and their equivalents known to those skilled in the art, etc. It should be further noted that claims may be designed to exclude any optional elements. Therefore, this statement is intended to serve as a pre-emptive basis for incorporating statements of claim elements using exclusive terms such as "solely," "only," or using a "negative" limitation.

It should be understood that certain features of the present invention that are described in the context of separate embodiments for clarity may also be provided in combination in a single embodiment. Conversely, various features of the present invention that are described in the context of a single embodiment for simplicity may also be provided individually or in any suitable subcombination. All combinations of embodiments of the present invention are expressly contemplated herein and disclosed herein as if each and every combination were individually and expressly disclosed. In addition, all subcombinations of the various embodiments and elements thereof are also expressly contemplated herein and disclosed herein as if each and every combination were individually and expressly disclosed. .

The publications discussed herein are provided solely for their disclosure prior to the filing date of the present application. The dates of publications provided may be different from the actual publication dates which may need to be independently confirmed.

When providing a range of values, it should be understood that the endpoints of the range are included. In addition, the present invention covers every intermediate value between the upper and lower limits of that range (unless the context clearly requires otherwise, it is accurate to one tenth of the unit of the lower limit) and any other stated or intermediate values in that stated range. The upper and lower limits of those smaller ranges may be independently included in those smaller ranges, and are also included in the present invention in addition to any limit explicitly excluded in the stated range. If the stated range includes one or both of the limits, the present invention also includes ranges excluding either or both of those included limits.

Ranges recited herein should be understood as shorthand for all values within the range, including the recited endpoint. For example, a range of 1 to 50 should be understood to include any number, combination of numbers, or sub-range consisting of the following groups: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, and 50, and including sub-ranges such as 11 to 48 or 39 to 41. 4.3. Compositions

In one aspect, the present invention provides a composition comprising a surface-modified viral exosome having a titer of at least 1.0E+10 vg/ml, wherein the surface-modified viral exosome comprises a ligand covalently coupled to the viral exosome via a linker.

The surface-modified viral coat confers improved transduction efficiency, improved cell type selectivity, or both improved transduction efficiency and improved cell type selectivity to the recombinant virions when compared to recombinant virions comprising a viral coat having the same primary amino acid sequence but not modified to include a targeting ligand, of which the coat is a part.

Due to the efficiency of production of surface-modified viral exosomes described herein, the compositions of the present invention have relatively high physical titers compared to compositions prepared by other methods. In some embodiments, the composition comprising surface-modified viral exosomes has a physical titer of at least 1E+10 vg/mL, such as at least 1.5E+11 vg/ml, at least 2.0E+11 vg/ml, or at least 2.5E+11 vg/ml, as determined by droplet digital PCR (ddPCR). In some embodiments, the composition has a physical titer of at least 1.0E+12 vg/ml, at least 1.5E+12 vg/ml, at least 2.0E+12 vg/ml, or at least 2.5E+12 vg/ml. In some embodiments, the composition has a physical titer of at least 1.0E+13 vg/ml, at least 1.5E+13 vg/ml, at least 2.0E+13 vg/ml, or at least 2.5E+13 vg/ml.

In some embodiments, the composition comprising the surface-modified viral exosome has a physical titer of 1.0E+10 vg/ml to 5.0E+13 vg/ml, e.g., 1.0E+11 vg/ml to 5.0E+13 vg/ml, 1.0E+12 vg/ml to 5.0E+13 vg/ml, 1.0E+13 vg/ml to 5.0E+13 vg/ml, 2.0E+10 vg/ml to 5.0E+13 vg/ml, 2.0E+11 vg/ml to 5.0E+13 vg/ml, 2.0E+12 vg/ml to 5.0E+13 vg/ml, 2.0E+13 vg/ml to 5.0E+13 vg/ml, 3.0E+10 vg/ml to 5.0E+13 vg/ml, vg/ml, 3.0E+11 vg/ml to 5.0E+13 vg/ml, 3.0E+12 vg/ml to 5.0E+13 vg/ml, 3.0E+13 vg/ml to 5.0E+13 vg/ml, 4.0E+10 vg/ml to 5.0E+13 vg/ml, 4.0E+11 vg/ml to 5.0E+13 vg/ml, 4.0E+12 vg/ml to 5.0E+13 vg/ml, 4.0E+13 vg/ml to 5.0E+13 vg/ml, 5.0E+10 vg/ml to 5.0E+13 vg/ml, 5.0E+11 vg/ml to 5.0E+13 vg/ml or 5.0E+12 vg/ml to 5.0E+13 vg/ml.

In some embodiments, the composition comprising the surface-modified viral exosome has a physical titer of 1.5E+11 vg/ml to 3.0E+12 vg/ml, 2.0E+11 vg/ml to 3.0E+12 vg/ml, or 2.5E+11 vg/ml to 3.0E+12 vg/ml.

The composition of the present invention comprises a surface-modified viral capsid of Formula V: (V) In which: is a viral shell optionally comprising a nucleic acid payload; each of ^SP1 and ^SP2 is independently selected from a bond or a spacer; Q is a cross-linking moiety; and L is a ligand.

The variable "LCR" refers to the ligand to shell ratio and corresponds to the amount of ligand coupled to the shell surface. In some embodiments, LCR is an integer from 10 to 500, such as 10 to 400, 10 to 300, 10 to 200, 10 to 100, 50 to 500, 50 to 400, 50 to 300, 50 to 200, 50 to 100, 100 to 500, 100 to 400, 100 to 300, 100 to 200, 200 to 500, 200 to 400, 200 to 300, 300 to 500, 300 to 400, or 400 to 500. In some embodiments, LCR is an integer between 100 and 200, such as 100-190, 100-180, 100-170, 100-160, or 100-150.

In some embodiments, the composition comprises free ligand (in any form not coupled to the surface of the coat, including functionalized ligand and functionalized ligand coupled to unbound TFP linker) less than bound ligand (in the form of a surface-modified viral coat). The amount of free ligand in the composition can be measured by gold standard analysis of AAV, such as analytical ultracentrifugation (AUC) and mass spectrometry. In some embodiments, the composition comprises 20% or less free ligand, such as 15% or less, 10% or less, 5% or less, 4% or less, 3% or less, 2% or less, or 1% or less. 4.3.1. Bifunctional Linkers

The bifunctional linker of the present invention comprises two components: i) TFP ester and ii) a member or ligand of a crosslinker reactive pair. Optionally, the bifunctional linker further comprises a spacer. In certain embodiments, the bifunctional linker comprises a TFP ester, an optional spacer, and a member of a crosslinker reactive pair. In certain embodiments, the bifunctional linker comprises a TFP ester, a spacer, and a member of a crosslinker reactive pair.

The crosslinker reactive moiety of the bifunctional ligand is preferably selected to have mutual reactivity and selective reactivity with the crosslinker reactive moiety of the functionalized ligand of the present invention.

In some embodiments, the shell reactive linker has the following structure: Wherein ^SP1 is a bond or spacer known in the art; and a is an integer from 1 to 5, 2 to 4, such as 1, 2, 3, 4 or 5. In some embodiments, a is 4. 4.3.1.1 Spacer

In some of these embodiments, the spacer includes 10 to 30 non-hydrogen atoms. In other embodiments, the spacer includes ₁₀ to 30 divalent groups selected from -CH2- (methylene), -O- (ether), -C(=O)- (carbonyl), and -N(R)- (diamine, where R is H or _C1-3 alkyl (e.g., methyl)). In some of these embodiments, the spacer includes 5-40 methylene groups, _1-20 ether groups, 1-5 amines, and/or 1-4 carbonyl groups. In some of these embodiments, the spacer includes 10-15 methylene groups, 4-6 ether groups, 1-2 amines, and/or 3-4 carbonyl groups.

In some embodiments, the shell reactive linker has the following structure: wherein: X is a —CH ₂ — or —O— group, b is an integer from 2 to 20, n is an integer from 2 to 20, m is an integer from 2 to 10; and a is an integer from 1 to 5, 2 to 4, for example, 1, 2, 3, 4 or 5.

In some embodiments, the shell reactive linker has the following structure: wherein: n is an integer from 2 to 20, m is an integer from 2 to 10; and a is an integer from 1 to 5, 2 to 4, for example, 1, 2, 3, 4 or 5.

In some embodiments, the shell reactive linker has the following structure: wherein: n is an integer from 2 to 20; m is an integer from 2 to 10; and a is an integer from 1 to 5, 2 to 4, for example, 1, 2, 3, 4 or 5.

In some embodiments, the shell reactive linker has the following structure:

In some embodiments, n is an integer from 2 to 15.

In some embodiments, m is selected from 2, 3, and 4.

In some embodiments, n is 4 to 12 and m is 2, 3 or 4.

In some embodiments, n is 12 and m is 3 and the shell reactive linker is called DBCO-PEG ₁₂ -TFP and has the following structure: .

In some embodiments, n is 12 and m is 4 and the shell reactive linker is called DBCO-PEG ₄ -TFP and has the following structure: 4.3.2. Reactivity of crosslinking agents

In some embodiments, CRP1 includes a reactive moiety selected from the following: an azide; an alkyne; a 1,4-triazole; a 1,3-nitrone; a cyclooctyne or a derivative thereof, such as dibenzylazacyclooctyne (referred to herein as DBCO) or a derivative thereof; a triazine; a tetrazine; a strained dienophile; an aryl or alkyl phosphine; an isocyanide; a benzylguanine group, a benzylcytosine group, or a chloroalkyl group.

In some embodiments, CRP1 includes a cyclooctyne reactive moiety. In some embodiments, CRP1 is selected from OCT, MOFO, DIFO, DIMAC, COMBO, DIBO, DIBAC (DBCO), BARAC, BCN, and TMTH, such as those shown below:

In some embodiments, CRP ¹ comprises DBCO. In some embodiments, CRP ¹ has the following formula: Wherein m is 1 to 20. In some embodiments, m is 1 to 15. In some embodiments, m is 1 to 10. In some embodiments, m is 4. In some embodiments, the wave key indicates connection to SP ¹ . 4.3.3. Connector

The surface-modified viral capsid of the present invention includes a linker formed by coupling a bifunctional linker to a viral capsid. In some embodiments, the linker has the following formula: wherein ^SP1 and ^SP2 are independently bonds or spacers; and Q comprises a cross-linking moiety.

In some embodiments, the acyl group of the linker can be bound to the surface of the viral coat using a primary amine, thereby forming an amide bond. In certain embodiments, ^SP2 is bound to a ligand. 4.3.3.1 Spacer

The linker may further include one or more spacer moieties. The spacer moiety is not particularly limited and may be any spacer known in the art, including but not limited to one or more divalent groups, such as _-CH2- (methylene), -O- (ether), -C(=O)- (carbonyl) and -N(R)- (amine, wherein R is H or _C1-3 alkyl).

In certain of these embodiments, the spacer (SP ¹ , SP ² or both SP ¹ and SP ² ) comprises 10 to 30 non-hydrogen atoms. In other embodiments, the spacer comprises 10 to 30 divalent groups selected from -CH ₂ - (methylene), -O- (ether), -C(=O)- (carbonyl) and -N(R)- (diamine, wherein R is H or C _1-3 alkyl (e.g., methyl)). In certain of these embodiments, the spacer comprises 5-40 methylene groups, 1-20 ether groups, 1-5 amines and/or 1-4 carbonyl groups. In certain of these embodiments, the spacer comprises 10-15 methylene groups, 4-6 ether groups, 1-2 amines and/or 3-4 carbonyl groups.

In some embodiments, SP ¹ , SP ² , or both SP ¹ and SP ² include one or more PEGs (i.e., -(-(O-CH ₂ -CH ₂ ) _n )- or -([PEG] _n )-) and SP ² is -YC(O)-, -YC(O)O-, -Y-NHC(O)-, -Y-NHC(S)-, or -YC(O), wherein Y is a bond or one or more PEGs.

In certain embodiments, ^SP1 comprises -([PEG] _n )-, wherein n is 1 to 100, such as 1 to 50, 1 to 25, 1 to 15, 1 to 10, 1 to 5, 4 to 20, 4 to 15, 4 to 12, or 4 to 10. In some embodiments, n is selected from 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, or 15.

In some embodiments, SP ¹ , SP ² , or both SP ¹ and SP ² include 2 to 20 PEGs, e.g., 2 to 15, 2 to 10, 2 to 5, 5 to 20, 5 to 15, 5 to 10, 10 to 20, 10 to 15, or 15 to 20. In some embodiments, SP ¹ includes 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, or 15 PEGs.

In some embodiments, SP ² comprises -([PEG]n)-C(O)-, wherein n is 1 to 100, e.g., 1 to 25, 1 to 20, 1 to 15, 1 to 10, 2 to 25, 2 to 20, 2 to 15, 2 to 10, 4 to 15, or 4 to 10. 4.3.4. Crosslinking Moiety - Q

In the present embodiments, "crosslinking moiety" and "Q" are used interchangeably and refer to the reaction product of the crosslinker moiety as described above.

In some embodiments, the crosslinking moiety comprises a product selected from the following reactions: CuAAC reaction, SPAAC reaction, SPANC reaction, IEEDD reaction, Staudinger ligation, [4+1] cycloaddition reaction. In some of these embodiments, the reaction is selected from: SPAAC, SPANC and IEEDD reaction.

In some embodiments, the crosslinking moiety comprises a 7- or 8-membered carbon ring containing 0-3 heteroatoms selected from O or N. In some of these embodiments, the crosslinking moiety comprises or .

In one aspect of the invention, the surface-modified viral exosome comprises a moiety (Q) formed by a reaction between reactive moieties of a cross-linking agent as described herein.

In some embodiments, Q comprises the product of a CuAAC reaction. In some embodiments, Q comprises the product of a SPAAC reaction. In some embodiments, Q is the product of a SPANC reaction. In some embodiments, Q comprises the product of an IEEDD reaction. In some embodiments, Q comprises the product of a Staudinger ligation reaction. In some embodiments, Q comprises the product of a [4+1] cycloaddition reaction. In some embodiments, Q comprises the product of a tension-promoted reaction, such as SPAAC, SPANC, and IEEDD.

In some embodiments, Q includes a cyclic group. In some embodiments, Q includes a bicyclic group. In some embodiments, Q includes a tricyclic group. In some embodiments, Q includes a 5-8 membered carbon ring containing 0 to 3 heteroatoms selected from O, S or N. In some embodiments, Q includes an 8-membered ring containing 0 to 1 heteroatoms selected from O and N. In some embodiments, Q includes a 5-membered ring containing 0 to 3 heteroatoms selected from O and N. In some embodiments, Q is a triazole ring. In some embodiments, Q includes a 6-membered ring containing 0-3 heteroatoms selected from O and N. In some embodiments, Q includes a 6-membered ring containing 2 N heteroatoms.

In some embodiments where Q comprises a cyclic group, Q is as follows, wherein Z is a 7- or 8-membered carbon ring containing 0-3 heteroatoms selected from O or N.

In some embodiments, Q includes the structure shown below: .

In some embodiments, Q comprises the reaction product of a cyclooctyne reactive moiety and an azide reactive moiety.

In some embodiments, Q comprises one of the following structures: and .

In some embodiments, Q comprises one of the following structures: and , wherein m is 1 to 10, such as 2 to 8, 3 to 7 or 4 to 6. 4.3.5. Viral shell

In the embodiments of the present invention, the type of viral exosome is not particularly limited. In some embodiments, the viral exosome is selected from non-enveloped viruses, such as adenovirus or adeno-associated virus. In some embodiments, the viral exosome is a protein exosome of an enveloped virus (e.g., retrovirus, lentivirus, herpes simplex virus, and bacilli). Embodiments include non-natural exosomes and include biological or chemical changes or alterations of natural exosome proteins.

In some embodiments, the viral coat is an adeno-associated virus or a recombinant adeno-associated virus (rAAV or AAV are used interchangeably herein). Such AAV particles are capable of transducing a wide range of post-mitotic cells in mammals in vivo, such as (but not limited to) muscle cells, hepatocytes, and neurons.

In some embodiments, AAV comprises VP1, VP2 and/or VP3 capsid proteins of a natural AAV serotype. In some embodiments, AAV comprises one or more of non-natural VP1, VP2 and/or VP3 capsid proteins. In some such embodiments, the non-natural VP1, VP2 or VP3 capsid proteins differ from the natural capsid in primary amino acid sequence. In some embodiments, in addition to or in addition to changes in primary amino acid sequence, the non-natural capsid comprises biological or chemical alterations or changes in the natural AAV capsid proteins.

In some embodiments, the AAV capsid is composed of three overlapping capsid proteins (VP1, VP2, VP3) containing a unique VP1 N-terminus, a VP1/VP2 common portion, and a VP1, VP2, and VP3 common portion.

In certain embodiments, one or more of the coat proteins include natural amino acid residues that correspond to a primary sequence of a wild-type coat protein. In alternative embodiments, a primary sequence of one or more of the coat proteins includes amino acid residues that are engineered into a wild-type coat protein sequence. In certain of these embodiments, the engineered amino acids include one or more amine groups present on the coat surface and involve surface functionalization of one or more of the coat proteins. In certain embodiments, the natural or engineered amine group that is surface functionalized using the methods of the invention is lysine.

In various embodiments, the capsid protein is one of the native AAV serotypes AAV1, AAV2, AAV3B, AAV5, AAV6, AAV8 or AAV9. In various embodiments, the capsid protein is selected from the capsid proteins disclosed in PCT/US2014/060163, USP9695220, PCT/US2016/044819, PCT/US2018/032166, PCT/US2019/031851 and PCT/US2019/047546, the entire contents of which are incorporated herein by reference.

In some embodiments, the exosome is AAV2. In some embodiments, the exosome is AAV5.

Likewise, adeno-associated viruses may be selected from synthetic serotypes generated by non-natural methods such as, but not limited to, capsid mutation, peptide insertion or deletion in the capsid sequence, capsid shuffling or recombinant engineering of various serotypes.

AAV for use in the present invention is produced by any method known in the art (without limitation). For example, AAV can be produced by several methods including transient transfection of HEK293 cells, stable cell lines infected with Ad or HSV, mammalian cells infected with Ad or HSV (expressing rep-cap and transgene), or insect cells infected with a bacilliform virus capsid (expressing rep-cap and transgene). AAV produced by any of these methods can be used to produce the surface functionalized and surface modified viral capsids described herein. In certain embodiments, AAV is produced by transient transfection of HEK293 cells using the calcium phosphate-HeBS method using the following two plasmids: pHelper PDP2-KANA encoding AAV Rep2-Cap2 and adenoviral helper genes (E2A, VA RNA, and E4) and pVector ss-CAG-eGFP.

In some embodiments, the AAV of the present invention includes one or more sequences optionally from a source foreign to a virus.

According to certain embodiments, the AAV comprises one or more wild-type coat proteins from a natural serotype.

According to another specific embodiment, the AAV comprises a genetically modified capsid protein. In certain embodiments, the genetically modified capsid protein is a natural serotype engineered to include one or more genetic modifications (mutations, insertions or deletions). In an alternative embodiment, the AAV capsid is composed of one or more of the synthetic capsid proteins. In a specific embodiment, the AAV capsid is engineered to modify the natural tropism, for example to reduce heparin binding.

In the context of the present invention, synthetic capsids include any combination of capsid proteins from natural, genetically modified, and artificially generated serotypes (e.g., random mutagenesis, sequence shuffling, computer design, etc.) that are capable of assembling and generating new AAVs that are not known to be natural.

Currently, more than 100 AAV serotypes have been identified, which differ in the ability of the capsid protein to bind specific cell surface receptors that can transduce different cell types. AAV2 was the first serotype selected to be cloned into bacterioplasms and has since been used as a control for the identification of other serotypes. Twelve serotypes (AAV1, AAV2, AAV3, AAV4, AAV5, AAV6, AAV7, AAV8, AAV9, AAV10, AAV11, and AAV12) have been fully tested for their ability to transduce specific cell types and to distinguish capsid protein motifs that bind specific cell surface receptors for cell attachment. In the context of the present invention, an AAV exosome selected from AAV1, AAV2, AAV3, AAV4, AAV5, AAV6, AAV7, AAV8, AAV9, AAV10, AAV11 and AAV12 is preferred. However, it will be appreciated that any other AAV exosome may be used.

In one embodiment, the adeno-associated virus (AAV) particle is selected from AAV1, AAV2, AAV3, AAV4, AAV5, AAV6, AAV7, AAV8, AAV9, AAV10, AAV11 and AAV12. The most commonly used gene transfer systems to date are derivatives of viruses, such as adeno-associated virus type 2 (AAV2), AAV9 and AAV8. In specific embodiments, the AAV capsid is selected from AAV2 and AAV9, wherein the capsid protein is further modified to reduce or modify the natural tropism, such as to reduce heparin binding, as appropriate. In some embodiments, the AAV particle is selected from AAV2, AAV5 and AAV9. In some embodiments, the AAV particle is AAV2. In some embodiments, the AAV particle is AAV9. In some embodiments, the AAV particle is AAV5.

In certain embodiments, the binding site of the AAV capsid capable of binding to heparan sulfate proteoglycans has been removed.

In certain embodiments, the removal of heparin binding is achieved by replacing arginine 585 or arginine 588 of VP1 and/or at least one of the similar arginines in VP2 or VP3 with a different amino acid (e.g., alanine). In some embodiments, at least one of arginine 448 and arginine 451 in VP2 or 383 and 386 in VP3 is altered.

In a specific embodiment, the AAV of the present invention comprises at least one coat protein mutated from the wild type, for example, wherein the modified/mutated protein is selected from the wild type proteins VP1, VP2 and/or VP3. Alternatively, two of the proteins VP1, VP2 and/or VP3 in the coat are mutated, or all three proteins VP1, VP2 and VP3 in the coat are modified. In a specific embodiment, at least a portion (e.g., an amino acid) of at least one of the proteins to be modified in the coat is mutated (replaced, inserted or deleted). However, it is also possible to mutate multiple portions of the proteins VP1, VP2 and VP3, for example multiple amino acids (e.g., 2, 3, 4, 5, 6, 7, 8, 9, 10 or any other number of portions or amino acids) in the coat. In certain embodiments, at least one of arginine 484, 487, 585 and 588 and lysine 532 in VP1 and/or a similar arginine in VP2 or VP3 is replaced with a different amino acid (eg, alanine).

Primary amines are present at the N-terminus of each coat protein and in the side chain (ε) of lysine (Lys, K) amino acid residues in the coat protein sequence. These primary amines are surface available and thus can react with TFP esters (e.g. of the coat reactive linker (see below)) to provide amide bonds. 4.3.6. Ligands

The ligands for use in the present invention are not particularly limited, as long as the ligand is suitable for coupling as described herein. In some embodiments, the ligand is selected from protein ligands having homologs (e.g., receptors) located on the surface of mammalian cells. In some of these embodiments, the homologous proteins are involved in the transduction of surface-modified viral capsids.

In some embodiments, the ligand is a cell type specific ligand. In certain embodiments, the ligand is selected from a polypeptide, a protein, a monosaccharide or polysaccharide, a steroid hormone, an RGD motif peptide, a vitamin, a small molecule or a targeting peptide. Also contemplated are antibodies (e.g., single chains) or fragments thereof and nanobodies or DARPins (designed anchor protein repeat proteins) (genetically engineered antibody mimics that typically exhibit highly specific and high affinity target protein binding); enzymes, such as proteases, glycosidases, lipases, peptidases; immunoglobulins, such as CD47 (don't eat me signal); IgG proteases, such as IdeZ and IdeS; protein-based adjuvants and small molecule adjuvants for vaccination.

According to one embodiment, the cell type specific ligand is derived from a protein, such as transferrin, epidermal growth factor EGF, basic fibroblast growth factor bFGF.

According to one embodiment, the cell type specific ligand is derived from a monosaccharide or polysaccharide, such as galactose, N-acetylgalactosamine and mannose.

According to one embodiment, the cell type specific ligand is derived from a vitamin, such as folate.

According to one embodiment, the cell type specific ligand is derived from a small molecule comprising naproxen, ibuprofen or other known protein binding molecules.

In certain embodiments, the ligand is selected from a protein ligand, such as a growth factor or an interleukin; a toxin subunit, such as cholera toxin B subunit; a lectin, such as isolectin B4 or wheat germ agglutinin; an adhesion factor, such as lactadherin; an antibody or a single-chain variable fragment thereof, such as an anti-CD-34 antibody; more specifically, an E. coli recombinant scFv CD-34 antibody fragment; a peptide, such as deltamorphin opioid receptor ligand; and a gene editing nuclease, such as Cas9, DARPin (e.g., MP0112). In some embodiments, the lectin is selected from wheat germ agglutinin (WGA), isolectin B4 (IB4), Maackia amurensis lectin, Lens culinaris lectin, Wisteria floribunda lectin, and Pha-L. In some embodiments, the lectin is WGA. In some embodiments, the lectin is IB4.

In other embodiments, the ligand is an oligonucleotide, for example, such as the aptamers described in Kelly, L., Maier, KE, Yan, A et al., A comparative analysis of cell surface targeting aptamers. Nat Commun 12, 6275 (2021), which is incorporated herein by reference. 4.3.7. Load

In some embodiments, the nucleic acid cargo is encapsulated inside the surface-modified shell of the present invention. The nucleic acid cargo can be any type of nucleic acid molecule that can be usefully transduced into cells by rAAV.

In some embodiments, the cargo or cargo of the rAAV of the present invention may be an expressible polynucleotide. In some embodiments, the expressible polynucleotide encodes a protein (e.g., encoding a therapeutic protein). In some embodiments, the expressible polynucleotide encodes a transgene. In some embodiments, the expressible polynucleotide may be transcribed to provide a guide RNA, a transactivating CRISPR RNA (tracrRNA), a messenger RNA (mRNA), a micro RNA (miRNA), or a shRNA.

In some embodiments, the cargo provides a DNA homology construct for homology-directed repair.

In some embodiments, the nucleic acid molecule is a nucleic acid molecule encoding an intracellular antibody (e.g., neutralizing certain intracellular proteins), a nucleic acid molecule encoding a peptide toxin (e.g., blocking ion channels in pain pathways), a nucleic acid molecule encoding an optogenetic actuator (e.g., turning on or off neuronal activity using light), a nucleic acid molecule encoding a pharmacogenetic tool (e.g., turning on or off neuronal signaling using a chemical ligand with non-interfering pharmacological effects), a nucleic acid molecule encoding a CRISPR-based editor for precision gene editing, a nucleic acid molecule encoding a CRISPR-epigenetic tool for regulating gene expression, and/or a nucleic acid molecule encoding a suicide gene that induces cell death.

In some embodiments, the cargo encodes an immunogenic polypeptide (e.g., for use in vaccination). The nucleic acid may encode any immunogen of interest known in the art, including, but not limited to, immunogens from human immunodeficiency virus, influenza virus, gag protein, tumor antigens, cancer antigens, bacterial antigens, viral antigens, and the like. Alternatively, the immunogen may be present within (e.g., incorporated into) or tethered to (e.g., by covalent modification of) a viral coat.

An immunogenic polypeptide or immunogen may be any polypeptide suitable for protecting an individual from a disease, including but not limited to microbial, bacterial, protozoan, parasitic, fungal, and viral diseases. For example, the immunogen may be an orthomyxovirus immunogen (e.g., an influenza virus immunogen (e.g., an influenza virus hemagglutinin (HA) surface protein or an influenza virus nucleoprotein gene or an equine influenza virus immunogen)) or a lentivirus immunogen (e.g., an equine infectious anemia virus immunogen, a simian immunodeficiency virus (SIV) immunogen, or a human immunodeficiency virus (HIV) immunogen (e.g., HIV or SIV envelope GP160 protein, HIV or SIV matrix/capsid protein, and HIV or SIV gag, pol, and env gene products)). The immunogen may also be an orthomyxovirus immunogen (e.g., a Lassa fever virus immunogen, such as a Lassa fever virus nuclear coat protein gene and a Lassa fever envelope glycoprotein gene), a poxvirus immunogen (e.g., vaccinia (e.g., vaccinia L1 or L8 gene)), a flavivirus immunogen (e.g., a yellow fever virus immunogen or a Japanese encephalitis virus immunogen), a filovirus immunogen (e.g., an Ebola virus immunogen or a Marburg virus immunogen, such as NP and OP genes), a bunyavirus immunogen (e.g., RVFV, CCHF and SFS virus), or a coronavirus immunogen (e.g., an infectious human coronavirus immunogen, such as a human coronavirus envelope glycoprotein gene, or a porcine infectious gastroenteritis virus immunogen, or an avian infectious bronchitis virus immunogen, or a severe acute respiratory syndrome (SARS) immunogen (e.g., S [S1 or S2], M, E or N protein or immunogenic fragment thereof)). The immunogen may further be a poliovirus immunogen, a herpes immunogen (e.g., CMV, EBV, HSV immunogen), a mumps immunogen, a measles immunogen, a rubella immunogen, a diphtheria toxin or other diphtheria immunogen, a pertussis antigen, a hepatitis (e.g., hepatitis A, hepatitis B or hepatitis C) immunogen, or any other vaccine immunogen known in the art.

In some embodiments, the immunogen can be any tumor or cancer cell antigen. Optionally, the tumor or cancer antigen is expressed on the surface of a cancer cell. Exemplary cancer and tumor cell antigens are described in SA Rosenberg, (1999) Immunity 10:281). Illustrative cancer and tumor antigens include, but are not limited to, BRCA1 gene products, BRCA2 gene products, gp100, tyrosinase, GAGE-1/2, BAGE, RAGE, NY-ESO-1, CDK-4, β-catenin, MUM-1, caspase-8, KIAA0205, HPVE, SART-1, PRAME, p15, melanoma tumor antigen (Kawakami et al., (1994) Proc. Natl. Acad. Sci. USA 91:3515; Kawakami et al., (1994) J. Exp. Med., 180:347; Kawakami et al., (1994) Cancer Res. 54:3124), including MART-1 (Coulie et al., (1991) J. Exp. Med. 180:35), gp100 (Wick et al., (1988) J. Cutan. Pathol. 4:201) and MAGE antigens (MAGE-1, MAGE-2 and MAGE-3) (Van der Bruggen et al., (1991) Science, 254:1643), CEA, TRP-1; TRP-2; P-15 and tyrosinase (Brichard et al., (1993) J. Exp. Med. 178:489); HER-2/neu gene product (U.S. Patent No. 4,968,603); CA 125; HE4; LK26; FB5 (endosialin); TAG 72; AFP; CA19-9; NSE; DU-PAN-2; CA50; Span-1; CA72-4; HCG; STN (sialyl Tn antigen); c-erbB-2 protein; PSA; L-CanAg; estrogen receptor; milk fat globulin; p53 tumor suppressor protein (Levine, (1993) Ann. Rev. Biochem. 62:623); mucin antigen (International Patent Publication WO 90/05142); telomerase; nucleoplasmic protein; prostatic acid phosphatase; papillomavirus antigen; and antigens associated with the following cancers: melanoma, adenocarcinoma, thymoma, sarcoma, lung cancer, liver cancer, colorectal cancer, non-Hodgkin's lymphoma, Hodgkin's lymphoma), leukemia, uterine cancer, breast cancer, prostate cancer, ovarian cancer, cervical cancer, bladder cancer, kidney cancer, pancreatic cancer, brain cancer, kidney cancer, stomach cancer, esophageal cancer, head and neck cancer, and others (see, e.g., Rosenberg, (1996) Annu. Rev. Med. 47:481-91).

In some embodiments, the nucleic acid encodes any polypeptide that is desired to be produced in vitro, in vitro, or in vivo in cells. For example, a viral vector can be introduced into cultured cells and the expressed protein product isolated therefrom.

Those skilled in the art will appreciate that the nucleic acid cargo of interest can be operably associated with appropriate control sequences. For example, the nucleic acid cargo can be operably associated with expression control elements (e.g., transcription/translation control signals, replication origins, polyadenylation signals, internal ribosome entry sites (IRES), promoters, enhancers, and the like).

Those skilled in the art will also appreciate that a variety of promoter/enhancer elements may be used depending on the desired level and tissue-specific expression. Depending on the desired expression pattern, the promoter/enhancer may be constitutive or inducible. The promoter/enhancer may be natural or exogenous and may be a natural or synthetic sequence. If exogenous, it is expected that the transcription start region is not found in the wild-type host and needs to be introduced.

The promoter/enhancer element may be native to the target cell or individual to be treated and/or native to the nucleic acid cargo. The promoter/enhancer element is generally selected so that it can function in the target cell of interest. In representative embodiments, the promoter/enhancer element is a mammalian promoter/enhancer element. The promoter/enhancer element may be constitutive or inducible.

Inducible expression control elements are generally used in those applications where it is desirable to provide regulation of the expression of a nucleic acid payload. Inducible promoter/enhancer elements for gene delivery can be tissue-specific or tissue-preferential promoter/enhancer elements, and include muscle-specific or preferential (including heart, skeletal and/or smooth muscle), neural tissue-specific or preferential (including brain-specific), eye (including retina-specific and corneal-specific), liver-specific or preferential, bone marrow-specific or preferential, pancreas-specific or preferential, spleen-specific or preferential, and lung-specific or preferential promoter/enhancer elements. In one embodiment, a CNS cell-specific or CNS cell-preferential promoter is used. Examples of neuron-specific or preferential promoters include, but are not limited to, neuron-specific enolase, synaptophysin, and MeCP2. Examples of astrocyte-specific or preferential promoters include, but are not limited to, neurofibromic acid protein and S100β. Examples of ependymal cell-specific or preferential promoters include, but are not limited to, wdr16, Foxj1, and LRP2. Examples of microneuronal cell-specific or preferential promoters include, but are not limited to, F4/80, CX3CR1, and CD11b. Examples of oligodendrocyte-specific or -preferred promoters include, but are not limited to, myelin basic protein, cyclic nucleotide phosphodiesterase, proteolipid protein, Gtx, and Sox10. The use of CNS cell-specific or -preferred promoters can increase the specificity achieved by the AAV vector. Other inducible promoter/enhancer elements include hormone-inducible and metal-inducible elements. Exemplary inducible promoter/enhancer elements include, but are not limited to, Tet on/off elements, RU486-inducible promoters, corticosterone-inducible promoters, rapamycin-inducible promoters, and metallothionein promoters.

In embodiments where the nucleic acid is transcribed and then translated in the target cell, specific initiation signals are generally employed to efficiently translate the inserted protein coding sequence. Such exogenous translation control sequences, which may include the ATG initiation codon and adjacent sequences, may be of various origins (natural and synthetic).

In some embodiments, when the load includes a gene editing nuclease (e.g., Cas9), the load further includes a nucleic acid molecule, such as a gRNA and/or a specific DNA pre-inserted into the host genome. In some of these embodiments, the load includes a transgene known to be associated with a genetic disease.

Those skilled in the art are aware of other gene editing nucleases besides Cas9, such as Cpf1, TALEN, ZFN or nested endonucleases. In addition, the modification can be conveniently performed using a DNA-guided Argonaute interference system (DAIS). Ago proteins are heterologously expressed by a polynucleotide introduced into the cell in the presence of at least one exogenous oligonucleotide (DNA guide) that provides the Ago protein with cleavage specificity at a preselected locus. TALEN and Cas9 systems are described in WO 2013/176915 and WO 2014/191128, respectively. Zinc finger nucleases (ZFNs) were first described by Kim, YG; Cha, J.; Chandrasegaran, S. (“Hybrid restriction enzymes:zinc finger fusions to Fok I cleavage domain” (1996). Proc Natl Acad Sci USA 93 (3):1156-60). Cpfl is a Class 2 CRISPR Cas system described by Zhang et al. (Cpfl is a single RNA-guided Endonuclease of a Class 2 CRIPR-Cas System. (2015). Cell；163:759-771). The AGO gene family was first described in Guo S, Kemphues KJ. (Par-1, a gene required for establishing polarity in C. elegans embryos, encodes a putative Ser/Thr kinase that is asymmetrically distributed. (1995). Cell;81(4):611-20).

In some embodiments, the vector is an "empty" capsid particle (i.e., without the vector genome) comprising, consisting of, or consisting essentially of a chimeric AAV capsid protein of the invention. As described in U.S. Patent Application No. 5,863,541, the chimeric AAV capsid of the invention can be used as a "capsid vehicle." Molecules that can be covalently linked, bound to, or encapsulated by a viral capsid and transferred into a cell include DNA, RNA, lipids, carbohydrates, polypeptides, small organic molecules, or combinations thereof. In addition, a molecule can be attached (e.g., "tethered") to the outside of a viral capsid to transfer the molecule into a host target cell. In one embodiment of the invention, the molecule is covalently linked (i.e., coupled or chemically coupled) to a capsid protein. Methods of covalently linking molecules are known to those skilled in the art.

The exosomes of the present invention can also be used to generate antibodies against novel exosome structures. Alternatively, exogenous amino acid sequences can be inserted into viral exosomes to present antigens to cells, for example, for administration to an individual to generate an immune response against the exogenous amino acid sequences. 4.4. Methods of Preparing Surface Functionalized Exosome Compositions

In one aspect of the invention, a method of preparing a composition comprising a surface functionalized viral exosome as described above is provided. Compared to previous processes that rely on an exosome reactive linker containing NHS esters, the method described herein uses a relatively low amount of an exosome reactive linker (i.e., an exosome reactive linker containing TFP esters). The reduced amount of exosome reactive linker eliminates the need for a post-reaction purification step, thereby preventing viral loss and improving cell transduction compared to surface-modified viral exosomes prepared using an exosome reactive linker containing NHS esters.

The method comprises the step (a): combining a viral exosome (I) comprising a plurality of surface-available primary amines (p) with an exosome-reactive linker (II) comprising a terminal tetrafluorophenyl (TFP) ester and a terminal first member (CRP1) of a crosslinker-reactive pair to provide a composition comprising a surface-functionalized viral exosome (III), as shown in the following reaction diagram 1: Scheme 1. Method for preparing surface functionalized virus capsids.

The amide bond is formed by the reaction of a primary amine of the primary sequence of the shell as shown above (for example, a surface available lysine residue) with a TFP ester-containing shell reactive linker (only one surface lysine amine is shown for clarity).

A plurality of surfaces can be reacted with a plurality of TFP esters of a shell reactive linker using a primary amine (p). In some embodiments, p is an integer from 10 to 500, such as 10 to 400, 10 to 300, 10 to 200, 10 to 100, 50 to 500, 50 to 400, 50 to 300, 50 to 200, 50 to 100, 100 to 500, 100 to 400, 100 to 300, 100 to 200, 200 to 500, 200 to 400, 200 to 300, 300 to 500, 300 to 400, or 400 to 500. In some embodiments, p is an integer from 100 to 200, for example, from 100 to 190, from 100 to 180, from 100 to 170, from 100 to 160, or from 100 to 150.

The variable "q" in the surface functionalized viral exosome (III) reflects the number of amides formed by the reaction of the surface available primary amine (p) of the viral exosome (I) with the TFP ester of the exosome reactive linker (II). In some embodiments, q is the same as p. In some embodiments, q is an integer from 10 to 500, such as 10 to 400, 10 to 300, 10 to 200, 10 to 100, 50 to 500, 50 to 400, 50 to 300, 50 to 200, 50 to 100, 100 to 500, 100 to 400, 100 to 300, 100 to 200, 200 to 500, 200 to 400, 200 to 300, 300 to 500, 300 to 400, or 400 to 500. In some embodiments, q is an integer from 100 to 200, such as 100 to 190, 100 to 180, 100 to 170, 100 to 160, or 100 to 150.

SP ¹ is a bond or one or more PEG (i.e., -(-(O-CH ₂ -CH ₂ ) _n )- or -([PEG] _n )-). In some embodiments, n is 0. In some embodiments, n is 1-100, e.g., 1-50, 1-25, 1-15, 1-10, 1-5, 4-20, 4-15, 4-12, or 4-10.

In some embodiments, the molar ratio of viral exosome (I) to exosome-reactive linker (II) in reaction (a) is 1:100 to 1:50,000, e.g., 1:100 to 1:33,000, 1:100 to 1:10,000, 1:100 to 1:5,000, 1:100 to 1:1,000, 1:100 to 1:500, 1:100 to 1:250. In some embodiments, the molar ratio of viral exosome (I) to exosome-reactive linker (II) is 1:500 to 1:10,000.

In some embodiments, the viral exosome (I) is purified prior to reacting with the exosome-reactive linker (II). It has been found that when the viral exosome is first purified, a lower molar ratio of viral exosome to exosome-reactive linker is required to prepare a composition with potent viral cell transduction. Suitable purification methods include, but are not limited to, density gradient centrifugation, chromatography (e.g., ion exchange chromatography or affinity chromatography), and combinations thereof. In some embodiments, the density gradient centrifugation is iodixanol density gradient ultracentrifugation.

In some embodiments, the viral capsid is used without purification. In some embodiments, chemical modification is performed directly on crude cell lysate (e.g., unpurified AAV). In some embodiments, cell lysate (e.g., AAV9, AAV2-VR4) is effectively modified with, for example, NHS and TFP esters of a capsid-reactive linker without purification. In some embodiments, AAV9 is effectively modified with a TFP ester of a capsid-reactive linker without purification. In some embodiments, AAV9 is effectively modified with an NHS ester of a capsid-reactive linker without purification. In some embodiments, AAV2-VR4 is effectively modified with a TFP ester of a capsid-reactive linker without purification. In some embodiments, AAV2-VR4 can be efficiently modified using NHS esters of the coat-reactive linker without purification. In some embodiments, a lower concentration of TFP ester compared to NHS is required.

In some embodiments, the viral coat is freeze-thawed prior to surface functionalization.

In some embodiments, the reaction is carried out in an aqueous medium. Suitable aqueous media include at least one buffer. In some embodiments, the buffer is selected from N-[2-hydroxyethyl]-hexahydropyrazine-N'-[2-ethanesulfonic acid] (HEPES), MOPS, MES, phosphates and bicarbonates. In some embodiments, the concentration of at least one buffer is different. In some embodiments, the concentration of at least one buffer is no more than about 0.1M, such as about 0.01M to about 0.1M or about 0.05M to about 0.1M. In some embodiments, the aqueous medium includes about 0.01M to about 0.1M sodium bicarbonate.

In some embodiments, the aqueous medium includes at least about 200 mM of at least one salt, such as at least about 250 mM, at least about 300 mM, at least about 350 mM, at least about 400 mM, at least about 450 mM, or at least about 500 mM. In some embodiments, the salt is selected from chloride salts, phosphate salts, sulfate salts, and citrate salts. In some embodiments, the salt is selected from sodium salts, potassium salts, calcium salts, and magnesium salts. In some embodiments, the aqueous medium includes at least about 200 mM sodium chloride.

In some embodiments, the ionic strength of the aqueous medium is at least about 150 mM, such as at least about 200 mM, at least about 250 mM, at least about 300 mM, at least about 350 mM, at least about 400 mM, at least about 450 mM, or at least about 500 mM.

In some embodiments, the aqueous medium further comprises at least one surfactant. In some embodiments, the surfactant is Pluronic® F68. In some embodiments, the aqueous reaction medium comprises about 0.001% to about 0.005% Pluronic® F68.

In some embodiments, the pH of the aqueous medium is about 6 to about 10, e.g., about 7 to about 10, about 8 to about 10, about 9 to about 10, about 6 to about 9, about 7 to about 9, about 8 to about 9, about 6 to about 8, about 7 to about 8, or about 6 to about 7. In some embodiments, the pH of the aqueous medium is about 8 to about 9.

In some embodiments, the temperature of the reaction is about 0° C. to about 50° C., such as about 10° C. to about 40° C. or about 20° C. to about 30° C. In some embodiments, the temperature of the reaction is room temperature (about 23° C.).

In some embodiments, the duration of the reaction is from about 5 minutes to about 24 hours, such as from about 30 minutes to about 24 hours or from about 1 hour to about 24 hours.

In some embodiments, after the reaction is complete, an excess quencher is added to consume unbound linkers. In some embodiments, the quencher comprises an amine-containing compound. In some embodiments, the quencher is selected from glycine and Tris buffer.

In some embodiments, no purification step is performed after forming the surface functionalized viral exocoel. Exemplary purification methods to be avoided include centrifugation, precipitation, and chromatography. It has been found that in some embodiments, the use of such purification steps results in the loss of viral particles.

In the examples where no purification was performed, the resulting composition contained some amount of starting materials and reagents in addition to the surface functionalized virus exosome product.

The method of the present invention further comprises step (b): reacting a composition comprising a surface functionalized viral capsid (III) comprising a first member (CRP1) of a crosslinker reactive pair with a functionalized ligand comprising a second member (CRP2) of a crosslinker reactive pair to provide a composition comprising a surface modified viral capsid (V).

The first crosslinker reactive pair reacts with the second crosslinker reactive pair to form the crosslinking moiety Q (discussed above), thereby coupling the viral coat and the ligand, as shown in the following reaction diagram 2: Reaction diagram 2. Coupling of viral capsid and ligand. In reaction diagram 2, is a functionalized ligand, ^SP2 is a bond or spacer and L is a ligand.

In some embodiments, the functionalized ligand is purchased from a commercial supplier. In other embodiments, the functionalized ligand is prepared by contacting a ligand comprising a first reactive moiety with a ligand-reactive linker comprising a second reactive moiety and a second member (CRP2) of a crosslinker reactive pair.

In some embodiments, the ligand-reactive linker is selected from TCO-PEGn-NHS; tetrazine-PEGn-NHS; azido-PEGn-NHS; phosphine-NHS; maleimide-PEGn-succinimidyl ester; DBCO-PEGn-TFP ester and DBCO-PEGn-NHS ester, wherein n is 1 to 100, preferably 4 to 10.

In some embodiments, the ligand-reactive linker is DBCO-PEGn-TFP ester, wherein n is 1 to 100, preferably 4 to 10. In some embodiments, the ligand-reactive linker is DBCO-PEGn-NHS ester, wherein n is 1 to 100, preferably 4 to 10. In some embodiments, the ligand-reactive linker is DBCO-PEG4-TFP. In some embodiments, the ligand-reactive linker is DBCO-PEG5-TFP. In some embodiments, the ligand-reactive linker is DBCO-PEG6-TFP. In some embodiments, the ligand-reactive linker is DBCO-PEG7-TFP.

In some embodiments, the functionalized ligands for use in the present invention are as described in WO2022101363, the entire contents of which are incorporated herein by reference.

In some embodiments, SP ² of the functionalized ligand is selected from -YC(O)-, -YC(O)O-, -Y-NHC(O)-, -Y-NHC(S)-, or -YC(O), wherein Y is a bond or one or more PEGs (i.e., -(−(O−CH ₂ −CH ₂ ) _n )- or -([PEG] _n )-). In some embodiments, SP ² is -([PEG] _n )-C(O)-, wherein n is 1 to 100, e.g., 1 to 25, 1 to 20, 1 to 15, 1 to 10, 2 to 25, 2 to 20, 2 to 15, 2 to 10, 4 to 15, or 4 to 10. In some embodiments, SP ² is -([PEG] _n )-C(O)-, wherein n is 1 to 100, preferably 1 to 10.

In some embodiments, the functionalized ligand has Formula IV: in: is a ligand.

In these embodiments, the surface (shown above) can be coupled to ^SP2 using a primary amine.

In some embodiments, CRP2 comprises a reactive moiety selected from the following: an azide; an alkyne; a 1,4-triazole; a 1,3-nitrone; a cyclooctyne or a derivative thereof, such as a dibenzylcyclooctyne or a derivative thereof; a triazine; a tetrazine; a strained dienophile; an aryl or alkyl phosphine; an isocyanide; a benzylguanine group, a benzylcytosine group, or a chloroalkyl group. In certain of these embodiments, CRP2 comprises an azide. In some embodiments, CRP2 consists of -N ₃ .

In some embodiments, the functionalized ligand is WGA-[PEG] _n -N ₃ (also referred to herein as WGA-[PEG] _n -azide), wherein n is 1 to 20, such as 1 to 10 or 4 to 10. In some embodiments, the functionalized ligand is WGA-PEG ₄ -azide. In some embodiments, the functionalized ligand is WGA-PEG ₅ -azide. In some embodiments, the functionalized ligand is WGA-PEG ₆ -azide. In some embodiments, the functionalized ligand is WGA-PEG ₇ -azide.

In some embodiments, the functionalized ligand is NGF-[PEG] _n -N ₃ (also referred to herein as NGF-[PEG] _n -azide), wherein n is 1 to 20, such as 1 to 10 or 4 to 10. In some embodiments, the functionalized ligand is NGF-[PEG] n -azide, wherein n is 1 to 20, preferably 4 to 10. In some embodiments, the functionalized ligand is NGF-PEG ₄ -azide. In some embodiments, the functionalized ligand is NGF-PEG ₅ -azide. In some embodiments, the functionalized ligand is NGF-PEG ₆ -azide. In some embodiments, the functionalized ligand is NGF-PEG ₇ -azide.

In some embodiments, the molar ratio of viral exosome (used in step (a)) to functionalized ligand (used in step (b)) is 1:100 to 1:50,000, e.g., 1:100 to 1:33,000, 1:100 to 1:10,000, 1:100 to 1:5,000, 1:100 to 1:1,000, 1:100 to 1:500, 1:100 to 1:250. In some embodiments, the molar ratio of viral exosome to functionalized ligand is 1:500 to 1:10,000.

In some embodiments, the molar ratio of viral exosome to exosome-reactive linker in step (a) is the same as the molar ratio of viral exosome to functionalized ligand in step (b), i.e., 1: 1. In some embodiments, the molar ratio of viral exosome to exosome-reactive linker in step (a) is different from the molar ratio of viral exosome to functionalized ligand in step (b).

In certain embodiments, the reaction is carried out in an aqueous medium. Suitable aqueous media include at least one buffer. In some embodiments, the buffer is selected from N-[2-hydroxyethyl]-hexahydropyrazine-N'-[2-ethanesulfonic acid] (HEPES), MOPS, MES, phosphates and bicarbonates. The concentration of at least one buffer may vary. In some embodiments, the concentration of at least one buffer is no more than about 0.1M, such as about 0.01M to about 0.1M or about 0.05M to about 0.1M. In some embodiments, the aqueous medium includes about 0.01M to about 0.1M sodium bicarbonate.

In some embodiments, the aqueous medium includes at least about 200 mM of at least one salt, such as at least about 250 mM, at least about 300 mM, at least about 350 mM, at least about 400 mM, at least about 450 mM, or at least about 500 mM. In some embodiments, the salt is selected from chloride salts, phosphate salts, sulfate salts, and citrate salts. In some embodiments, the salt is selected from sodium salts, potassium salts, calcium salts, and magnesium salts. In some embodiments, the aqueous medium includes at least about 200 mM sodium chloride.

In some embodiments, the ionic strength of the aqueous medium is at least about 150 mM, such as at least about 200 mM, at least about 250 mM, at least about 300 mM, at least about 350 mM, at least about 400 mM, at least about 450 mM, or at least about 500 mM.

In some embodiments, the aqueous medium further comprises at least one surfactant. In some embodiments, the surfactant is Pluronic® F68. In some embodiments, the aqueous reaction medium comprises about 0.001% to about 0.005% Pluronic® F68.

In some embodiments, the pH of the aqueous medium is about 6 to about 10, e.g., about 7 to about 10, about 8 to about 10, about 9 to about 10, about 6 to about 9, about 7 to about 9, about 8 to about 9, about 6 to about 8, about 7 to about 8, or about 6 to about 7. In some embodiments, the pH of the aqueous medium is about 8 to about 9.

In some embodiments, the reaction temperature is about 0°C to about 50°C, such as about 10°C to about 40°C or about 20°C to about 30°C. In some embodiments, the reaction temperature is about 0°C to about 50°C, such as about 10°C to about 40°C or about 20°C to about 30°C (about 23°C).

In some embodiments, the duration of the reaction is from about 5 minutes to about 24 hours, such as from about 30 minutes to about 24 hours or from about 1 hour to about 24 hours.

In some embodiments, the surface-modified viral exosomes are purified. Exemplary purification methods include centrifugation, precipitation, and chromatography. In some embodiments, the purification method is centrifugation ultrafiltration. In some embodiments, the purification method is affinity chromatography.

In some embodiments, the surface-modified viral coat is body-modified.

In some embodiments, the surface-modified viral exosomes are stable for at least 3 weeks (e.g., at least 4 weeks, at least 5 weeks, at least 6 weeks, at least 7 weeks, at least 8 weeks, at least 9 weeks, at least 10 weeks, at least 11 weeks, at least 12 weeks, at least 13 weeks, at least 14 weeks, or at least 15 weeks) at low temperatures comprising -150°C (e.g., -140°C, -130°C, -120°C, -110°C, -100°C, -90°C, -80°C, -70°C, -60°C, -50°C, -40°C, -30°C, -20°C, or -10°C).

In some embodiments, the surface-modified viral exosomes are stable at -140°C. In some embodiments, the surface-modified viral exosomes are stable at -130°C. In some embodiments, the surface-modified viral exosomes are stable at -120°C. In some embodiments, the surface-modified viral exosomes are stable at -110°C. In some embodiments, the surface-modified viral exosomes are stable at -100°C. In some embodiments, the surface-modified viral exosomes are stable at -90°C. In some embodiments, the surface-modified viral exosomes are stable at -80°C. In some embodiments, the surface-modified viral exosomes are stable at -70°C. In some embodiments, the surface-modified viral exosomes are stable at -60°C. In some embodiments, the surface-modified viral exosomes are stable at -50°C. In some embodiments, the surface-modified viral exosomes are stable at -40°C. In some embodiments, the surface-modified viral exosomes are stable at -30°C. In some embodiments, the surface-modified viral exosomes are stable at -20°C. In some embodiments, the surface-modified viral exosomes are stable at -10°C.

In some embodiments, the surface-modified viral shells are stable for at least 4 weeks. In some embodiments, the surface-modified viral shells are stable for at least 5 weeks. In some embodiments, the surface-modified viral shells are stable for at least 6 weeks. In some embodiments, the surface-modified viral shells are stable for at least 7 weeks. In some embodiments, the surface-modified viral shells are stable for at least 8 weeks. In some embodiments, the surface-modified viral shells are stable for at least 9 weeks. In some embodiments, the surface-modified viral shells are stable for at least 10 weeks. In some embodiments, the surface-modified viral shells are stable for at least 11 weeks. In some embodiments, the surface-modified viral shells are stable for at least 12 weeks. In some embodiments, the surface-modified viral capsid is stable for at least 13 weeks. In some embodiments, the surface-modified viral capsid is stable for at least 14 weeks. In some embodiments, the surface-modified viral capsid is stable for at least 15 weeks. 4.5. Examples Example 1. Significant loss of physical potency of AAV2 after NHS modification

In previous experiments performed using NHS esters as the reactive portion of the shell, it was determined that two separate cleaning steps were required to produce ligand-functionalized shell compositions that exhibited robust transduction. First, cleaning was required after the shell was functionalized with a linker (e.g., NHS-PEGn _- DBCO) to remove excess reactive linker, since the linker would effectively quench the functionalized ligand during the cross-linking step and result in contamination of the final product with excess ligand cross-linked to the NHS- _PEGn -DBCO linker, thereby competing with the AAV shell for transduction.

When transitioning to in vivo experiments using NGF-modified vectors, physical AAV titers were measured for the first time before and after modification and a surprising 90% loss of AAV particles was observed during the modification process ( Figure 1 ). This result was surprising because despite the 90% loss, a significant increase in transduction efficiency was observed when coupled to a ligand such as WGA.

Commercially available wild-type AAV2 with EGFP loading as a reporter gene for cell transduction was used, which had a viral titer of 2.7E+12 Vg/ml in PBS/Pluronic 0.001%/NaCl 200mM. This AAV2 was modified with NHS-PEG ₄ -DBCO (2.6 nMol) in 20ul of PBS/Pluronic 0.001%/NaCl 200mM at room temperature for 3 hours, and then reacted with 55 pMol of WGA-PEG ₄ -azolide at room temperature for 1 hour followed by overnight at 4°C. The surface-modified viruses were either (1) added directly to PC12 cells or (2) centrifuged one to three times in an Amicon 100Kda centrifugal unit, each at 10000 g for 1 minute.

As shown in Figure 1 , modification with NHS-PEG ₄ -DBCO resulted in a 90% reduction in AAV2 titer as measured by ddPCR. To further explore this result, WGA-conjugated AAV2 was applied to PC12 cells and transduction efficiency was monitored before and after reaction cleanup. From these experiments it was apparent that virus loss occurred after a single centrifugation step using a 100Kda centrifugal filter unit. Further centrifugation did not significantly reduce the yield. Other cleanup procedures, including passivation of the filter unit, precipitation of AAV2, and size-screening analysis, were evaluated for transduction efficiency in HEK293 cells and ddPCR with physical titer as a readout for virus loss.

In other experiments, wild-type AAV2 was applied directly to HEK293 cells or one or more of the following clean-up procedures were performed: 1. Centrifugation at 10000g for 1 minute on 100Kda Amicon centrifuge units. 2. Passivation of 100Kda Amicon centrifuge units by pretreatment with 3E+9 Vg of wild-type AAV2 carrying the tdTomato reporter gene for 60 minutes followed by centrifugation of modified AAV2 at 10000g for 1 minute. 3. Precipitation of modified AAV2 by incubation in 8% PEG8000 for 1 hour at 4°C with slow agitation, followed by complete precipitation at 4°C without agitation for 3 hours. The sample was then centrifuged at 2818g for 30 min at 4°C, the supernatant was subsequently removed and the pellet was resuspended in PBS/Pluronic 0.001%/NaCl 200mM. 4. The modified AAV2 was sieved by centrifugation at 1000g for 2 minutes on a column filled with agarose 4B resin. The sample was then concentrated to 24ul using a vacuum concentrator at 45°C.

Separately, AAV2 (0.0045 pMol) was modified with the coat-reactive linker TFP-PEG ₄ -DBCO (45 pMol) in PBS/Pluronic 0.001%/NaCl 200 mM at room temperature overnight. The reaction was then quenched with 50 mM glycine to remove residual linker.

Samples were analyzed by adding them to AAV2-permissive HEK293 cells and monitoring transduction efficiency using Bio-Rad S3e cell sorter imaging and FACS, and by quantifying physical titers by performing ddPCR using primers targeting the ITR region.

As can be seen in Figures 2a and 2b , all purification methods resulted in a substantial reduction in AAV2 recovery, with centrifugal filtration units (untreated and passivated) performing the worst, PEG precipitation slightly improving yield, and agarose 4B size screening further slightly increasing recovery. In contrast, modification of AAV2 with the TFP linker did not require a purification step and did not result in a reduction in the yield of modified virus, as demonstrated by transduction efficiency in HEK293 cells and physical titer measured by ddPCR. Example 2. Reaction conditions for use with TFP ester coat reactive groups.

Commercially available wild-type AAV2 with EGFP as a reporter gene for cell transduction was used. The virus titer was 2.7E+12 Vg/ml in PBS/Pluronic 0.001%/NaCl 200 mM.

0.0045 pMol of AAV (1 ul) was reacted with 448 pMol of NHS-PEG ₄ -DBCO (virus: linker ratio of 1:100,000) in 22 ul of PBS/Pluronic 0.001%/NaCl 200 mM at room temperature for 3 hours. After that, 55 pMol of WGA-PEG ₄ -azide (virus: linker ratio of 1:12,000) was incubated at room temperature for 2 hours and then kept at 4°C overnight. The virus was added to the cells the next day.

0.0045 pMol of AAV (virus:conjugate 1:30,000) was incubated with a reduced amount of linker (i.e., 148 pMol of TFP-PEG ₄ -DBCO) in a final volume of 22 ul and overnight at room temperature. Glycine (50 mM pH 6.5) was added the next day to quench any unbound, active TFP-PEG ₄ -DBCO. This was followed by an overnight incubation at room temperature with 148 pMol of WGA-PEG ₄ -azide (virus:conjugate ratio 1:30,000). The next day, virus was added to the cells. Various reaction parameters were varied, including (a) the reaction buffer (i.e., sodium bicarbonate or different formulations of PBS/Pluronic 0.001%/NaCl 200 mM) and (b) the use of quenching reagents. For each protocol tested, appropriate controls (unmodified virus) were processed and used to normalize the data reported in cells. A summary of the protocols evaluated is provided in the table below. Table 1. Summary of protocols Reaction Reaction buffer Response time Quenching with glycine 1 0.1M sodium bicarbonate in water, pH 8.5 2 days yes 2 0.1 M sodium bicarbonate, pH 8.5 in PBS/Pluronic 0.001%/NaCl 200 mM 2 days yes 3 0.1M sodium bicarbonate in water, pH 8.5 2 days no 4 PBS/Pluronic 0.001%/NaCl 200mM 2 days yes 5 0.2M sodium bicarbonate in water, pH 8.5 2 days yes Using NHS control PBS/Pluronic 0.001%/NaCl 200mM 1 day no In vitro administration to PC12 cells

PC12 cells were maintained at 37°C in DMEM/F12 medium containing 5% horse serum, 5% fetal bovine serum, and 100 U of penicillin/streptomycin. PC12 cells were incubated with WGA-modified AAV2 prepared with either an NHS linker or a TFP linker. The medium was then replaced, and the cells were maintained at 37°C and imaged using a Zeiss AxioObserver A1 microscope 4 days after infection. After 7 days of infection, cells were harvested and analyzed using Bio-Rad's S3e cell sorter. Results

As is evident from the imaging analysis ( FIG3 ), reactions 2, 3, and 4 are the most efficient. Without wishing to be bound by theory, it is believed that PBS/Pluronic 0.001%/NaCl 200 mM prevents viral aggregation and that higher concentrations of sodium bicarbonate are detrimental to the conjugation of the TFP linker to the virus.

For cytometry analysis, the mean fluorescence intensity (MFI) of EGFP was used as an indicator of how much virus could enter each cell ( Figure 4a ) and EGFP+ cells were used to define the percentage of cells transduced by AAV ( Figure 4b ). MFI and percentage of EGFP+ cells were normalized by control. Among the protocols using TFP that resulted in good cell transduction, the conditions associated with reaction 2 were the best. Based on the comparison of reactions 2-4, it was concluded that glycine and 0.1 M sodium bicarbonate improved the modification of the virus by the TFP linker, thereby resulting in good cell transduction.

Additionally, similar amounts of WGA ligand-virus entered cells with either TFP-reaction 2 or NHS. However, the percentage of transduced cells was higher with TFP-reaction 2 compared to NHS. This data, and the fact that three times less TFP-PEG ₄ -DBCO linker was used compared to NHS-PEG ₄ -DBCO, suggests that the TFP linker chemistry may improve the chemical functionalization of AAV and may be performed at lower molar ratios, thereby avoiding the need for reaction cleanup. Example 3. Comparison of TFP- and NHS- linkers in vitro AAV2 and AAV5

The two AAV serotypes used were purchased commercially. AAV2 had a titer of 2.7E+12 vg/ml and contained a cargo of EGFP reporter gene under the CAG promoter, while AAV5 had a titer of 1E+13 vg/ml and contained an EGFP reporter gene under the CMV promoter. NHS- mediated chemical functionalization and WGA coupling of AAV2 and AAV5

For NHS-mediated chemical functionalization of AAV2, 1 µl (0.0045 pMol) of AAV2 was reacted with DBCO-PEG 4 -NHS at various virus:linker molar ratios (see table below) in a reaction volume of 20 µl PBS (+0.001% Pluronic and ₂₀₀ mM NaCl) for 3 hours at room temperature with shaking. Table 2 Virus : conjugate molar ratio Virus Connector 1 3,300 1 10,000 1 33,000 1 100,000 1 330,000

WGA (0.1 nMol) was dissolved in PBS and reacted with 20 molar equivalents of azide-PEG ₄ -NHS for 3 hours at room temperature with shaking. Unreacted azide groups were removed using 10 kDa MWCO centrifugal filtration. NHS-mediated DBCO-modified AAV2 was further incubated with 50 pMol WGA-PEG ₄ -azide for 2 hours at room temperature with shaking and then kept at 4°C overnight.

NHS-mediated chemical functionalization of 1 µl (0.0166 pMol) AAV5 was performed as described above for AAV2, but using different virus:linker ratios as indicated in the table below. Table 3 Virus : conjugate molar ratio Virus Connector 1 150 1 300 1 1,000 1 3000 1 10,000 1 30,000 1 100,000 TFP- mediated chemical functionalization and WGA coupling to AAV2 and AAV5

For TFP-mediated chemical modification, 1 µl of AAV2 (0.0045 pMol) was reacted with DBCO-PEG ₄ -TFP and WGA-PEG ₄ -azide at different molar virus:linker:ligand ratios (see table below). Table 4 Molar ratio of virus : conjugate : ligand Virus Connector Ligand 1 100 100 1 500 500 1 1,000 1,000 1 3,300 3,300 1 10,000 10,000 1 33,000 33,000

The volume of DBCO-PEG ₄ -NHS conjugate was minimized (1 µl) by preparing dilutions from 20 mM stock. Further 11 µl of PBS (+0.001% Pluronic and 200 mM NaCl) and 0.1 M sodium bicarbonate buffer (pH 8.3) were added and the reaction was kept shaking overnight at room temperature. The reaction was stopped by adding 50 mM glycine. WGA (0.1 nMol) was dissolved in PBS and reacted with 20 molar equivalents of azide-PEG ₄ -NHS at room temperature for 3 hours with shaking. Unreacted azide groups were removed using 10 kDa MWCO centrifugal filtration. WGA-PEG ₄ -azide was added at the indicated molar ratios (see table above) and the reaction was kept shaking at room temperature overnight.

TFP-mediated chemical modification of 1 µl (0.0166 pMol) of AAV5 was performed as described above for AAV2, but using different virus:linker:ligand ratios as indicated in the table below. Table 5 Molar ratio of virus : conjugate : ligand Virus Connector Ligand 1 500 500 1 1,000 1,000 1 3,000 3,000 1 10,000 10,000 1 30,000 10,000 1 100,000 10,000 In vitro administration to PC12 cells

PC12 cells were maintained in DMEM/F12 medium (+ 10% horse serum, 5% fetal bovine serum, 15 mM HEPES, 2.5 mM Glutamax and 100 U penicillin/streptomycin) and incubated at 37°C in a humidified atmosphere of 5% _CO2 . Modified AAV was added to the cells and incubated overnight. The medium was replaced and the cells were maintained at 37°C for 5 days. Data acquisition

Transduction efficiency was determined by imaging PC12 cells with a Nikon A1R confocal microscope. To quantify efficiency, cells were collected and prepared for flow cytometry. Flow cytometry data were acquired on an S3e cell sorter from Bio-Rad and analyzed using Flow Jo. The number of viral genome copies was quantified using the Bio-Rad Droplet Digital PCR System. Results

AAV2 modified with DBCO-PEG ₄ -NHS at a virus:linker ratio of 1:100,000 and with 2.5uM WGA-PEG ₄ -azide showed the highest transduction in PC12 cells compared to the other virus:linker ratios tested. In contrast, AAV2 modified with DBCO-PEG ₄ -TFP and WGA-PEG ₄ -azide showed the highest transduction at a virus:linker:ligand ratio of 1:10,000:10,000. TFP modification requires less linker to provide the same transduction efficiency as using NHS chemistry. When using another AAV serotype (AAV5), it was observed that TFP-mediated modification required a lower amount of virus:conjugate (1:1000 molar ratio) to achieve optimal transduction efficiency compared to NHS-mediated modification (1:30,000 virus:conjugate molar ratio). Example 4. Comparison of TFP- and NHS- conjugates in vivo

Through a series of in vitro experimental implementations using TFP and NHS linkers, it was demonstrated that TFP chemistry improves the modification of adeno-associated virus (AAV) with ligands of interest, prevents viral loss and is applicable to more than one serotype. Here, the goal was to validate the superior performance of TFP on NHS-modified AAV2 in vivo in mice. For the ligand, a mutant neuronal growth factor ligand (NGF ^R121W ) was chosen that targets a population of noxious peptidergic neurons expressing the TrkA/p75 receptor complex receptor. It was further found that AAV2 could be effectively modified using TFP on a larger scale. The modified AAV2 was injected subcutaneously into the plantar surface of the hind paw and the transgene expression was analyzed in the skin, dorsal root ganglion (DRG) and spinal cord. Methods Chemical functionalization and coupling of NGF to AAV via NHS and TFP chemicals

Wild-type commercial AAV2 with an EGFP reporter gene was used for chemical functionalization. Its viral titer in PBS/Pluronic 0.001%/NaCl 200mM was 2.7E+12Vg/ml.

Standard functionalization via NHS was performed by reacting 0.2025 pMol of AAV (45 µl) with 20250 pMol of NHS-PEG ₉ -BG (virus: linker molar ratio = 1:100,000) in a total volume of 900 µl PBS/Pluronic 0.001%/NaCl 200 mM for 3 hours at room temperature. Afterwards, the virus was concentrated using Amicon 100 KDa centrifugal filtration to remove excess unbound linker and 200 pMol of NGF-SNAP was added. The reaction was incubated at room temperature for 2 hours and then kept at 4°C overnight. The next day, the virus was concentrated again using Amicon 100 KDa filtration to remove excess unbound NGF-SNAP.

For the TFP reaction, 0.1125 pMol of AAV was incubated with TFP-PEG ₄ -DBCO at different virus: linker ratios reported in the table below. The reaction was performed in PBS (+0.001% Pluronic and 200 mM NaCl) and 0.1 M sodium bicarbonate buffer (pH 8.3) at room temperature overnight, quenched with 50 mM glycine, and then incubated with an equimolar ratio of NGF-PEG ₄ -azide with linker (see table below) for 2 hours at room temperature and then overnight at 4°C. Table 6 TFP ratio virus : conjugate / ligand pMol virus ( titer 2.7 vg/ml ; 25µl) pMol linker TFP-PEG ₄ -DBCO ( in 1µl ) pMol ligand NGF-PEG ₄ -azide 1:100 0.1125 11.25 11.25 1:500 0.1125 56.25 56.25 1:1,000 0.1125 112.5 112.5 1:3,000 0.1125 337.5 337.5 1:10,000 0.1125 1125 1125

use NHS and NGF Of AAV9 The modification of the body and TrkA/p75 Over-expression HEK293 Transduction assays in cells

AAV9 modifications were scaled up and transduction was tested in the HEK293 cell line expressing TrkA/p75 using NGF as a ligand.

For bulk chemical modification, wild-type commercial AAV9 with an EGFP reporter gene was used. Its viral titer in PBS/Pluronic 0.001%/NaCl 200 mM was 5.9E+12 Vg/ml.

For the NHS reaction, 2832 VG of AAV were incubated with NHS-PEG ₄ -DBCO at a virus:conjugate ratio of 175 000. The reaction was performed in PBS (+0.001% Pluronic and 200 mM NaCl) at room temperature overnight, washed three times with 15 ml Amicon falcon (100 KDa cutoff), aliquoted and frozen at -80°C. Subsequently, AAV9-DBCO was thawed and incubated with 3 μM NGF-PEG ₄ -azolide at room temperature for 2 hours and then at 4°C overnight. Viruses with a range of titers (i.e. different MOIs) were added to HEK293 cell lines overexpressing TrkA and p75 receptors. As a control, unmodified AAV9 was applied to the cells at the same MOI.

use TFP and Coupling NGF Of AAV2 Body modification , Testing freeze-thaw stability and analyzing transduction in vivo in mice

For bulk chemical modification, wild-type commercial AAV2 with an EGFP reporter gene was used. Its viral titer in PBS/Pluronic 0.001%/NaCl 200 mM was 2.7E+12 Vg/ml.

For TFP reactions, the following titers of AAV were incubated with TFP-PEG ₄ -DBCO at a virus:conjugate ratio of 1:20,000. Reactions were performed overnight at room temperature in PBS (+0.001% Pluronic and 200 mM NaCl) and 0.1 M sodium bicarbonate buffer (pH 8.3), quenched with 50 mM glycine, and then incubated with 27 μM NGF-PEG ₄ -azide overnight at room temperature. Fresh or frozen samples were then used as indicated in Table 7. Table 7 Initial volume of AAV Total Vg Volume of the first reaction step Volume of the second reaction step condition 45ul 1.22E+11 29.3ul 74.3 ul Fresh Frozen 1 ml 2.7E+12 650ul 1.65 ml Frozen 5 ml 1.35E+13 3.25 ml 8.25 ml Frozen 12 ml 3.24E+13 7.8 ml 19.8 ml Frozen ddPCR

AAV2 modified with NHS and TFP at different ratios were processed for ddPCR using primers targeting inverted terminal repeats ( ITRs ) present in the viral genome.

To verify that the modified virus was actually active, it was also administered in vivo to PC12 cells expressing TrkA/p75 endogenously. PC12 cells were maintained at 37°C in DMEM/F12 medium containing 5% horse serum, 5% fetal bovine serum and 100 U of penicillin/streptomycin. PC12 cells were cultured with NGF-modified AAV2 prepared with NHS and TFP at different ratios. The medium was then replaced and the cells were maintained at 37°C and imaged with a Zeiss AxioObserver A1 microscope 7 days after infection. Paw injection in mice

Three hours before virus injection, 60 units of hyaluronidase were injected to promote viral spread. 5E+10 VG of AAV2 modified with NHS or TFP at different ratios were injected into the paws of mice anesthetized with isoflurane. Injections were performed using an insulin syringe positioned in the center of the paw. The volume injected was 30 µl. Histological analysis

After 3 weeks of intravital injection, mice were sacrificed and lumbar dorsal root ganglia (DRG) were collected. DRG were fixed in 2% paraformaldehyde (PFA) overnight, then washed in PBS and cleaned with ScaleS solution at 4°C for 2 days. They were flat mounted and imaged using a Zeiss AxioObserver A1 microscope.

In other experiments, DRGs were washed in PBS and blocked with 2% donkey/0.3% Triton/PBS overnight at 4°C. They were then incubated with 1:200 goat anti-rat Trka antibody in 2% donkey/0.3% Triton/PBS for 72 hours at 4°C. Thereafter, DRGs were washed 3 times with 0.3% Triton/PBS at 10-20 minute intervals. DRGs were then incubated with 1:200 donkey anti-goat Ax594 antibody in 2% donkey/0.3% Triton/PBS for 72 hours at 4°C, washed as described above, cleared overnight at 4°C, flat mounted, and imaged.

The skin of the injection site was collected, fixed in 2% paraformaldehyde (PFA) overnight, then washed in PBS and incubated with sucrose 30%/PBS overnight at 4°C. It was then embedded in OCT and cut at a thickness of 30 µm in a cryostat. The sections were blocked with 2% donkey/0.3% Triton/PBS for 2 hours at room temperature. The skin was then incubated with 1:200 goat anti-rat Trka and chicken anti-EGFP antibodies in 2% donkey/0.3% Triton/PBS overnight at 4°C. The skin sections were then washed 3 times with 0.3% Triton/PBS at intervals of 10-20 minutes. Skins were then incubated with 1:200 donkey anti-goat Ax594 and donkey anti-chicken Ax488 antibodies in 2% donkey/0.3% Triton/PBS at 4°C overnight, washed, mounted and imaged as described above.

Spinal cords were dissected from the lumbar enlargement, fixed overnight in 2% paraformaldehyde (PFA), then washed in PBS and incubated with sucrose 30%/PBS overnight at 4°C. They were then embedded in OCT and cut at 30 µm thickness in a cryostat. The sections were blocked with 2% donkey/0.3% Triton/PBS at room temperature for 2 hours and then incubated with 1:200 goat anti-rat Trka in 2% donkey/0.3% Triton/PBS overnight at 4°C. Spinal cord sections were washed 3 times with 0.3% Triton/PBS at 10-20 min intervals and further incubated with 1:200 donkey anti-goat Ax594 and isolectin GS-IB ₄ Alexa Fluor™ 647 conjugate in 2% donkey/0.3% Triton/PBS overnight at 4°C. Prior to imaging, spinal cord slices were washed three times and then mounted as described above. Image Analysis

Tissues were imaged using a Nikon A1R confocal microscope and analyzed using ImageJ. Cell counts in DRGs were performed manually. Object-based colocalization was performed using the Colocalization Image Creator plugin (1). See Lunde, A., Glover, JC A versatile toolbox for semi-automatic cell-by-cell object-based colocalization analysis. Sci Rep 10, 19027 (2020). https://doi.org/10.1038/s41598-020-75835-7. Results

Initial experiments confirmed by ddPCR that TFP functionalization did not cause virus loss while NHS functionalization resulted in a 90% reduction in yield ( Figure 5 ). Consistently, when virus was added to PC12 expressing the Trka receptor, cell transduction was not enhanced compared to the control, probably due to virus loss of NHS, while all ratios of PC12 tested using TFP expressed EGFP.

In vivo transduction was tested by subcutaneous injection of the virus into the paw. Wild-type AAV2 was ineffective in transducing DRG neurons (perhaps because it does not transport retrogradely from the nerve terminals to the DRG in the skin), NHS-modified NGF-AAV caused transduction of some neurons, and the highest rate of TFP-functionalized NGF-AAV2 caused efficient transduction of neurons. Figure 6 illustrates the quantification of this data, which shows the number of positive DRG neurons per ganglion under each condition. NGF-AAV2 shells modified with higher rates of TFP were more effective than NHS-modified particles, and a 1:10,000 ratio was the most effective. In addition, the data show increased transduction efficiency in L3-L5 ganglia, which is consistent with the fact that nerves from these ganglia innervate the hind limbs and paws.

The selectivity of TFP-functionalized NGF-AAV targeting to injured neurons was further quantified by co-staining positive ganglia with antibodies against TrkA. As shown in Figure 7 , almost all NGF-AAV2-targeted neurons also stained positive for TrkA on L3, L4, and L5 ganglia, indicating that targeting to injured neurons was virtually accurate.

Skin sections from injection sites transduced by wild-type AAV2 and TFP-functionalized NGF-AAV2 were analyzed ( Figure 8 ). Nociceptor nerve terminals in the skin were identified by co-staining sections with TrkA antibody (red channel). In samples injected with wild-type virus, no transduction of nerves was detected. In sections from NGF-AAV2 injected mice, extensive overlap with TrkA-positive fibers was observed. In addition, eGFP-positive muscle fibers were observed in both conditions, indicating that the natural tropism of AAV2 was not de-targeted using this approach.

The sensory innervation of the spinal cord of mice injected with wild-type or TFP-functionalized NGF-AAV2 was explored. Fibers innervating the dorsal horn were eGFP-positive only in spinal cord sections collected from animals injected with NGF-AAV2 ( Figure 9 ). Co-staining was further performed to identify the surface topographic organization of the dorsal horn and understand which subset of neurons contained eGFP-positive fibers. Using the marker Trka, the first sheet where peptidergic nociceptive neurons terminated was visualized, and the marker Ib4 was used to identify the second sheet where non-peptidergic nociceptive neurons protruded. eGFP-positive fibers overlapped with Trka but not with Ib4, confirming that NGF-AAV targets peptidergic nociceptive neurons. Notably, fibers double positive for Trka and eGFP were primarily located in the medial region of the first lamina, an area that receives input from axons innervating the paw.

The scale-up of AAV9 modification was also explored and transduction was tested in a HEK293 cell line expressing TrkA/p75 using NGF as a ligand. Figure 10a shows images of TrkA/p75 HEK293 cells transduced using AAV9 or NGF-AAV9 at different MOIs. Figure 10b shows analysis of transduction efficiency. It was found that AAV9 can be modified more efficiently on a large scale using NHS. The MOI required to transduce TrkA/p75 HEK293 cells was substantially reduced after coupling with NGF.

The scale-up modification of AAV2 and the stability of the modified virus to freeze-thaw cycles were further explored and transduction was tested in mice in vivo. Figure 11 presents a comparison of the transduction efficiency of lumbar DRG isolated from mice injected subcutaneously with AAV2 modified with TFP-PEG ₄ -DBCO at different scales and used freshly or stored at -80°C. DRG were collected 3 weeks after AAV injection in vivo, mounted flat and imaged with confocal microscopy. It was found that AAV2 can be effectively modified on a larger scale using TFP. After coupling with NGF, the MOI required to transduce TrkA/p75 HEK293 cells was substantially reduced.

In summary, these data indicate that TFP chemistry allows for more efficient production of surface-modified AAV2 than NHS chemistry to avoid viral loss. The TFP protocol can be easily scaled up and is compatible with in vivo experiments. Importantly, modification of AAV with NGF via TFP chemistry clearly demonstrated the ability to target the virus to a population of interest, in this case peptidergic neurons expressing the NGF receptor, TrkA. Example 5. NHS- versus TFP - conjugated AAV

Chemical modification of AAV with DBCO-PEG _(n) -NHS linker and further conjugation to ligands results in substantial virus loss due to column-based wash steps during the modification process. To overcome the problem of virus loss, DBCO-PEG _(n) -TFP linker was used since no wash step is required during the modification process using this chemistry. Here, NHS- vs. TFP-mediated modification of AAV was compared by monitoring transduction efficiency and virus loss. AAV2, AAV5, or δ-HSPG AAV2 (with mutations in its HSPG binding site) was modified with NHS or TFP linkers, conjugated to the ligand wheat germ agglutinin (WGA), and tested for transduction efficiency in PC12 cells. For serotype AAV2, NHS- and TFP-based chemistries were tested with virus purified by two methods that yielded varying degrees of sample purity. Based on this data, the ratio of virus:linker:ligand conjugation was identified for each chemical modification and serotype.

method

NHS- and TFP-based chemistries were tested on different AAV serotypes. For one of the serotypes (AAV2), NHS- and TFP-chemistries were tested on virus purified by two different protocols that differed in the degree of purity of the final viral product - using a cesium chloride gradient or by affinity purification followed by isopycnic centrifugation via an iodixanol gradient (which is expected to increase the purity of the AAV vector). AAV5 and delta-HSPG AAV2 were purified using only affinity chromatography and an iodixanol gradient.

All AAV serotypes used were purchased commercially. One AAV2 purified using a cesium chloride gradient had a titer of 2.7E+12 vg/ml and contained a load as an EGFP reporter gene under the CAG promoter, while another AAV2 purified by affinity chromatography and an iodixanol gradient had a titer of 1E+13 vg/ml and contained an EGFP reporter gene. AAV5 contained EGFP with a CMV promoter and had a titer of 1E+13 vg/ml. Delta-HSPG AAV2 had EGFP with a CAG reporter gene and had a titer of 5.6E+12 vg/ml.

The following table summarizes the purification and characteristics of the AAV used. Table 8 . Serotype gradient Further purification steps Starter AAV2 Cesium chloride without CAG AAV2 Iodixanol Affinity analysis CMV AAV5 Iodixanol Affinity analysis CMV ΔAAV2 Iodixanol Affinity analysis CAG

NHS- Mediated chemical modification and WGA right AAV2 , AAV5 and δ AAV2 The coincidence

For NHS-mediated chemical modification of AAV2 purified by CsCl gradient, 1 µl (corresponding to 0.0045 pMol) of AAV2 was reacted with DBCO-PEG ₄ -NHS at different virus:conjugate molar ratios (see table below) in a reaction volume of 20 µl PBS (+0.001% Pluronic and 200 mM NaCl) for 3 hours at room temperature with shaking. Table 9 . Virus : conjugate molar ratio Virus Connector 1 3300 1 10000 1 33000 1 100000 1 330000

WGA (0.1 nMol) was dissolved in PBS and reacted with 20 molar equivalents of azide-PEG ₄ -NHS at room temperature with shaking for 3 hours. Unreacted azide groups were removed by centrifugal filtration using 10 kDa MWCO. NHS-mediated DBCO-modified AAV2 was further incubated with 50 pMol WGA-PEG ₄ -azide at room temperature with shaking for 2 hours and then kept at 4°C overnight.

NHS-mediated chemical modification of 1 µl (corresponding to 0.0166 pMol) AAV2 purified by affinity chromatography and iodixanol gradient was performed as described above for AAV2, but using different virus:conjugate ratios as indicated in the table below. Table 10 . Virus : conjugate molar ratio Virus Connector Ligand 1 500 500 1 1000 1000 1 3000 3000 1 10000 5000 1 30000 5000 1 100000 5000 1 300000 5000

NHS-mediated chemical modification of 1 µl (corresponding to 0.0166 pMol) AAV5 was performed as described above for AAV2, but using different virus:conjugate ratios as indicated in the table below. Table 11 . Virus : conjugate molar ratio Virus Connector 1 500 1 1000 1 3000 1 10000 1 30000 1 100000 1 300000

NHS-mediated chemical modification of 1 µl (corresponding to 0.0093 pMol) of δ AAV2 was performed as described above for AAV2, but using different virus:conjugate ratios as indicated in the table below. Table 12 . Virus : conjugate molar ratio Virus Connector Ligand 1 3000 3000 1 10000 6000 1 30000 6000 1 60000 6000

TFP- Mediated chemical modification and WGA right AAV2 , AAV5 and δ AAV2 The coincidence

For TFP-mediated chemical modification, 1 µl of AAV2 (for 0.0045 pMol) purified with CsCl2 was reacted with DBCO-PEG4-TFP and WGA-PEG ₄ -azide at different molar virus:linker:ligand ratios (see table below). Table 13 . Virus : conjugate : ligand molar ratio Virus Connector Ligand 1 100 100 1 500 500 1 1000 1000 1 3300 3300 1 10000 10000 1 33000 33000

Since TFP is more stable than NHS, the volume of DBCO-PEG ₄ -NHS conjugate was kept relatively small (1 µl) by preparing dilutions from 20 mM stock. Further 11 µl of PBS (+0.001% Pluronic and 200 mM NaCl) and 0.1 M sodium bicarbonate buffer (pH 8.3) was added and the reaction was kept shaking overnight at room temperature. The reaction was stopped by adding 50 mM glycine. WGA (0.1 nMol) was dissolved in PBS and reacted with 20 molar equivalents of azide-PEG ₄ -NHS at room temperature for 3 hours with shaking. Unreacted azide groups were removed using 10 kDa MWCO centrifugal filtration. WGA-PEG ₄ -azide was added at the indicated molar ratios (see table above) and the reaction was kept shaking at room temperature overnight.

TFP-mediated chemical modification of 1 µl of AAV2, AAV5, and delta AAV2 purified by affinity chromatography and iodixanol gradient was performed as described above for AAV2, but using different virus:linker:ligand ratios as indicated in the table below. In addition, the final volume of ligand for the three viruses (AAV2, AAV5, and delta AAV2) was too large for the highest ratios, so a maximum virus:ligand ratio of 1:5000, 1:10000, and 1:6000 was used, respectively. Table 14 (AAV2) . Molar ratio of virus : conjugate : ligand Virus Connector Ligand 1 500 500 1 1000 1000 1 3000 3000 1 10000 5000 1 30000 5000 1 100000 5000 Table 15 (AAV5) . Molar ratio of virus : conjugate : ligand Virus Connector Ligand 1 500 500 1 1000 1000 1 3000 3000 1 10000 10000 1 30000 10000 1 100000 10000 Table 16 (ΔAAV2) . Molar ratio of virus : conjugate : ligand Virus Connector Ligand 1 3000 3000 1 10000 6000 1 30000 6000 1 60000 6000

In vitro administration PC12 Cells

PC12 cells were maintained in DMEM/F12 medium (+ 10% horse serum, 5% fetal bovine serum, 15 mM HEPES, 2.5 mM Glutamax and 100 U penicillin/streptomycin) and incubated at 37°C in a humidified atmosphere of 5% _CO2 . Modified AAV was added to the cells and incubated overnight. The medium was replaced and the cells were maintained at 37°C for 5 days. Data collection

Transduction efficiency was determined by imaging PC12 cells with a Nikon A1R confocal microscope. To quantify efficiency, cells were collected and prepared for flow cytometry. Flow cytometry data were acquired on an S3e cell sorter from Bio-Rad and analyzed using Flow Jo. The number of viral genome copies was quantified using the Bio-Rad Droplet Digital PCR System. Results

Compared to the other virus:linker ratios tested, AAV2 modified with DBCO-PEG ₄ -NHS and with 2.5uM WGA-PEG ₄ -azide showed the highest transduction in PC12 cells at a virus:linker ratio of 1:100,000 ( FIG. 12 ). In contrast, AAV2 modified with DBCO-PEG ₄ -TFP and WGA-PEG ₄ -azide showed the highest transduction at a virus:linker:ligand ratio of 1:10,000:10,000 ( FIG. 13 ). This indicates that TFP modification allows us to use less linker and still achieve the same transduction efficiency as NHS chemistry.

As observed with affinity chromatography and iodixanol gradient purified AAV2, the amount of TFP linker molecules required to modify the virus and obtain robust cell transduction was further reduced when using higher purity virus. In fact, the optimal ratio of virus:TFP linker:ligand ratio was reduced from 1:10,000:10,000 to 1:500:500 when AAV2 was purified in this manner rather than only with a cesium chloride gradient ( Figures 13 and 15 ). Also in this case, NHS chemistry required higher amounts of linker (1:100,000) to achieve cell transduction via TFP display ( Figures 14 and 16 ) . When using other AAV serotypes (AAV5 and delta AAV2), it was observed that the virus:conjugate ratio could be reduced to 1:1,000 and 1:3,000, respectively, for TFP-mediated modification ( Figures 17 and 19 ) , while NHS-mediated modification ( Figures 16 and 18 ) required virus:conjugate ratios of 1:100,000 and 1:60,000, respectively, to achieve optimal transduction efficiency. In addition to the lack of virus loss during the modification process, another advantage of using the DBCO-PEG ₄ -TFP linker for improved AAV modification is demonstrated here. Example 6. Incorporation of a vector functionalization step in AAV DSP

As shown in reaction Figure 3 below, chemical modification of AAV during AAV purification rather than after purification (i.e., original protocol) was tested (i.e., crude virus extract modification). Figure 3. Schematic illustration of the vector functionalization steps incorporated into AAV DSP using two approaches .

Recombinant AAV9 or AAV2 with an insertion in the VR4 region and carrying eGFP or tdTomato, respectively, under the CAG promoter as cargo were generated in HEK293 cells as described previously. Cells were harvested 3 days after transfection, freeze-thawed 3 times and then buffer exchanged to PBS (+0.001% Pluronic and 200mM NaCl) using 100KDa centrifugal filtration. AAV9 was then modified with NHS-PEG ₄ -DBCO at various concentrations (3mM, 1mM, 0.3mM, 0.1mM, 30µM, 10µM, 1µM, 0.3µM or 0µM) at room temperature for 3 hours. The samples were again exchanged for buffer using 100 KDa centrifugal filtration and then incubated overnight with 2.5 uM WGA ₄ -azide. The samples were applied to PC12 cells and transduction efficiency was evaluated 5 days later.

For comparison using TFP esters, AAV2-VR4 was modified with TFP-PEG ₄ -DBCO at various concentrations (3mM, 1mM, 0.3mM, 0.1mM, 30µM, 10 µM, or 0µM) in PBS (+0.001% Pluronic and 200mM NaCl) and 0.1M sodium bicarbonate buffer (pH 8.3) overnight at room temperature. The reaction was then quenched with 50mM glycine, the buffer was exchanged using 100KDa centrifugal filtration, and then incubated with 2.5uM WGA ₄ -azide overnight. The samples were applied to PC12 cells. Results

Surprisingly, the modification works directly on crude cell lysate (i.e., unpurified AAV). Figures 20A and 20B present _{images of PC12 cells transduced with AAV9-WGA chemically modified with NHS-PEG 4} -DBCO and TFP-PEG ₄ -DBCO at the cell lysate stage, respectively. As can be seen from Figures 20A and 20B , NHS and TFP esters can be used to effectively modify AAV9 or AAV2-VR4 cell lysate without purification. Lower concentrations of TFP esters are required compared to NHS. 5. Equivalent Content and Incorporation by Reference

While the present invention has been particularly shown and described with reference to preferred embodiments and various alternative embodiments, it will be understood by those skilled in the relevant art that various changes in form and details may be made therein without departing from the spirit and scope of the invention.

For all purposes, the entire contents of all literature references, issued patents, and patent applications (including U.S. Provisional Application No. 63/514,909 filed on July 21, 2023) cited in the body of this specification are incorporated herein by reference.

The entire contents of WO 2020/225363 and WO 2022/101363 are incorporated herein by reference for all purposes.

When read in conjunction with the accompanying drawings, the foregoing invention content and the following embodiments of the present invention will be better understood. For the purpose of illustrating the present invention, the accompanying drawings illustrate some but not all alternative embodiments. However, it should be understood that the present invention is not limited to the exact configuration and tools shown. These drawings, which are incorporated into and constitute part of this specification, help to illustrate the principles of the present invention.

Figure 1 illustrates the significant loss of physical titer of AAV2 after NHS modification and centrifugation as measured by ddPCR.

Figures 2a-2b show the transduction efficiency of compositions prepared using different purification methods after modification of AAV2 with NHS, as evaluated by transduction efficiency in HEK293 cells ( Figure 2a ) and physical titer measured from ddPCR ( Figure 2b ). The figures also show the transduction efficiency of AAV2 modified with TFP ester measured by the same method.

Figure 3a illustrates the comparison of transduction efficiency of AAV2 and WGA-PEG ₄ -azide in PC12 cells as confirmed by the number of PC12 cells expressing the reporter gene EGFP after infection with WGA-conjugated AAV2 modified with TFP-PEG ₄ -DBCO + WGA-PEG ₄ -azide using the following different protocols: Reaction #1 performed in 0.1 M sodium bicarbonate, pH 8.5 in water, quenched with glycine; Reaction #2 performed in 0.1 M sodium bicarbonate, pH 8.5 in PBS/Pluronic 0.001%/NaCl 200mM, quenched with glycine. Cells were imaged 4 days after AAV-infection.

Figure 3b illustrates a comparison of transduction efficiency demonstrated by the number of PC12 cells expressing the reporter gene EGFP after infection with WGA-conjugated AAV2 modified with TFP-PEG ₄ -DBCO + WGA-PEG ₄ -azolide using the following different protocols: Reaction #3 performed in 0.1 M sodium bicarbonate, pH 8.5 in water, without glycine; Reaction #4 performed in PBS/Pluronic 0.001%/NaCl 200 mM, quenched with glycine. Cells were imaged 4 days after AAV-infection.

Figure 3c illustrates a comparison of transduction efficiency demonstrated by the number of PC12 cells expressing the reporter gene EGFP after infection with WGA-conjugated AAV2 modified with TFP-PEG ₄ -DBCO + WGA-PEG ₄ -azolide using the following different protocols: reaction #5 performed in 0.1 M sodium bicarbonate, pH 8.5 in water, quenched with glycine; and a control reaction using NHS performed in PBS/Pluronic 0.001%/NaCl 200 mM, without glycine. Cells were imaged 4 days after AAV-infection.

Figures 4a-4b show a comparison of transduction efficiency as demonstrated by the number of PC12 cells expressing the reporter gene EGFP after infection with WGA-conjugated AAV2 modified with TFP-PEG ₄ -DBCO + WGA-PEG ₄ -azide using different protocols (1-5) and NHS-PEG ₄ -DBCO + WGA-PEG ₄ -azide. Cells were analyzed by cytometry 7 days after AAV-infection. Figure 4a shows the mean fluorescence intensity (MFI) of EGFP+ cells normalized by appropriate controls (ΔMFI). Figure 4b shows the percentage of EGFP+ positive cells normalized by appropriate controls.

FIG. 5 demonstrates the reduction in viral titer of viruses prepared using different ratios of TFP-PEG ₄ -DBCO + NGF-PEG ₄ -azide, NHS-PEG ₉ -BG + NGF-SNAP, and a portion of wild-type AAV2 as measured by ddPCR.

Figure 6 shows the transduction efficiency of lumbar DRG isolated from mice injected subcutaneously with AAV2 modified with different ratios of TFP-PEG ₄ -DBCO + NGF-PEG ₄ -azolide, NHS-PEG ₄ -BG + NGF-SNAP, and wild-type AAV2 (control). DRG were collected 3 weeks after intravital injection of AAV, flat-mounted and imaged with confocal microscopy. The graph shows the number of EGFP+ positive cells.

Figure 7 shows the selectivity of TFP-functionalized NGF-AAV targeting injured neurons obtained by co-staining positive ganglia with antibodies against Trka. It shows quantification of transduced cells co-localized with the injured neuron marker Trka in lumbar DRGs 3, 4, and 5.

Figure 8 shows transduction of skin collected from the paws of mice injected subcutaneously with wild-type AAV2 and AAV2 modified with TFP-PEG ₄ -DBCO + NGF-PEG ₄ -azolide (virus:conjugate ratio = 1:10,000). Nerve tracts in the dermis were identified by TrkA staining (red) and the presence of virus was identified by the fluorescent protein EGFP (green).

Figure 9 shows transduction of spinal cords collected from mice injected subcutaneously with wild-type AAV2 and AAV2 modified with TFP-PEG ₄ -DBCO + NGF-PEG ₄ -azolide (virus:conjugate ratio = 1:10,000). Viral transduction was identified by the fluorescent protein EGFP (green), and the first and second lamellae of the dorsal horn were identified by Trka (red) and Ib4 (cyan), respectively.

FIG. 10a shows images of TrkA/p75 HEK293 cells transduced with AAV9 or NGF-AAV9 at different multiplicities of infection (MOIs).

Figure 10b shows the fluorescence intensity of AAV9 and NGF-AAV9; after NGF was coupled to AAV9, the MOI required to transduce TrkA/p75 HEK293 cells was substantially reduced.

Figure 11 presents a comparison of transduction efficiency of lumbar DRG isolated from mice injected subcutaneously with AAV2 modified with TFP-PEG ₄ -DBCO at different scales. The number on the image indicates the total amount of vector modified at one time. All animals were injected with the same dose of modified vector (5E+10 VG).

Figure 12 shows the transduction efficiency of PC12 cells by WGA-conjugated AAV2 modified with DBCO-PEG ₄ -NHS and 50 pMol WGA-PEG ₄ -azolide purified using a cesium chloride gradient and using different virus:linker ratios. PC12 cells were imaged 5 days after AAV infection.

Figure 13 shows the transduction efficiency of PC12 cells by WGA-conjugated AAV2 modified with DBCO-PEG ₄ -TFP and WGA-PEG ₄ -azolide purified using a cesium chloride gradient and using different virus:linker:ligand ratios. PC12 cells were imaged 5 days after AAV infection.

Figure 14 shows the transduction efficiency of PC12 cells by DBCO-PEG ₄ -NHS and WGA-PEG ₄ -azolide modified WGA-conjugated AAV2 using affinity chromatography and iodixanol gradient purification and using different virus:linker ratios. PC12 cells were imaged 5 days after AAV infection.

Figure 15 shows the transduction efficiency of PC12 cells by DBCO-PEG ₄ -TFP and WGA-PEG ₄ -azol modified WGA-conjugated AAV2 using affinity analysis and iodixanol gradient purification and using different virus:linker ratios. PC12 cells were imaged 5 days after AAV infection.

FIG. 16 shows the transduction efficiency of PC12 cells by WGA-conjugated AAV2 modified with DBCO-PEG ₄ -NHS and 50 pMol WGA-PEG ₄ -azide using different virus:linker ratios.

Figure 17 shows the transduction efficiency of PC12 cells by DBCO-PEG ₄ -TFP and WGA-PEG ₄ -azolide modified WGA-conjugated AAV2 using different virus:linker:ligand ratios. PC12 cells were imaged 5 days after AAV infection.

Figure 18 shows the transduction efficiency of PC12 cells by WGA-conjugated δ-HSPG AAV2 modified with DBCO-PEG ₄ -NHS and WGA-PEG ₄ -azide using different virus:linker:ligand ratios. PC12 cells were imaged 5 days after AAV infection.

Figure 19 shows the transduction efficiency of PC12 cells by WGA-conjugated δ-HSPG AAV2 modified with DBCO-PEG ₄ -TFP and WGA-PEG ₄ -azolide using different virus:linker:ligand ratios. PC12 cells were imaged 5 days after AAV infection.

FIG. 20a shows images of PC12 cells transduced with AAV9-WGA chemically modified with NHS-PEG ₄ -DBCO on crude cell extracts (lysates) using different concentrations of ligand during the modification reaction.

FIG. 20 b shows images of PC12 cells transduced with AAV9-WGA chemically modified with TFP-PEG ₄ -DBCO on crude cell extracts (lysates) using different concentrations of ligand during the modification reaction.

Claims (112)

A composition comprising a surface-modified viral capsid having a titer of at least 1.0E+10 vg/ml, the surface-modified viral capsid comprising: A ligand covalently coupled to the viral capsid via a linker. The composition of claim 1, wherein the linker has Formula VI: Wherein: ^SP1 and ^SP2 are independently bonds or spacers; and Q is a cross-linking moiety. The composition of claim 2, wherein the spacer comprises one or more divalent groups selected from: -CH ₂ -, -O-, -C(=O)-, and -N(R)-, wherein R is H or C _1-3 alkyl. The composition of any one of claims 1 to 3, wherein the surface-modified shell has Formula V: (V) In which: is the viral coat, optionally including nucleic acid cargo; SP ¹ and SP ² are independently a bond or a spacer; Q is a cross-linking moiety; and L is a ligand. The composition of any one of claims 1 to 4, wherein the composition has a surface-modified viral capsid physical titer of 1.0E+10 vg/ml to 5.0E+13 vg/ml. The composition of any one of claims 1 to 5, wherein the ligand to shell ratio (LCR) is 1 to 480, preferably 10 to 150, or 30 to 80. A composition as claimed in any one of claims 1 to 6, wherein the composition comprises less than 20% free ligand. The composition of any one of claims 1 to 7, wherein the viral capsid is an adeno-associated virus capsid (AAV), optionally selected from AAV1, AV2, AV3, AV4, AV5, AV6, AV7, AV8, AV9, AV10, AV11 and AAV12. The composition of claim 8, wherein the AAV is AAV2, AV5 or AAV9. The composition of claim 8 or 9, wherein the AAV is AAV2. The composition of claim 8 or 9, wherein the AAV is AAV5. The composition of claim 8 or 9, wherein the AAV is AAV9. The composition of any one of claims 1 to 12, wherein the shell comprises at least one protein selected from VP1, VP2 and VP3. The composition of any one of claims 1 to 13, wherein the capsid protein is a wild-type capsid protein. The composition of any one of claims 1 to 13, wherein the coat protein is a genetically modified coat protein. The composition of claim 15, wherein the genetic modification is disruption of a heparin sulfate proteoglycan binding site. The composition of claim 16, wherein at least one of arginine 585 of VP1, arginine 588 of VP1, arginine 488 of VP2, arginine 451 of VP2, arginine 383 of VP3, and arginine 386 of VP3 is replaced by a different amino acid. The composition of any one of claims 15 to 17, wherein the capsid is delta AAV2. The composition of any one of claims 1 to 17, wherein the ligand is selected from cell type specific ligands, polypeptides, proteins, monosaccharides, polysaccharides, steroid hormones, RGD motif peptides, vitamins, small molecules, antibodies, nanobodies, enzymes, and immunoglobulins. The composition of claim 19, wherein the ligand is a protein ligand, a toxin subunit, a lectin, an adhesion factor, an antibody or a single-chain variable fragment thereof, a peptide, and a gene-editing nuclease. The composition of claim 20, wherein the lectin is selected from wheat germ agglutinin (WGA), isolectin B4 (IB4), Maackia amurensis lectin, Lens culinaris lectin, Wisteria floribunda lectin, and Pha-L. The composition of claim 20, wherein the lectin is WGA. The composition of claim 20, wherein the ligand is a protein, preferably a growth factor or a cytokine. The composition of claim 23, wherein the protein ligand is neural growth factor (NGF). The composition of any one of claims 1 to 24, wherein the cross-linking moiety comprises at least one of an 8-membered ring and a triazole ring. The composition of claim 25, wherein the cross-linking moiety comprises both an 8-membered ring and a triazole ring. A composition as in any one of claims 1 to 26, wherein the cross-linking moiety is selected from the product of the following reactions: Cu(I)-catalyzed azido-alkyne cycloaddition (CuAAC) reaction, strain-promoted alkyne-azido-cycloaddition (SPAAC) reaction, strain-promoted alkyne-nitrone cycloaddition (SPANC) reaction, inverse electron demand Diels-Alder (IEEDD) reaction, Staudinger ligation, and [4+1] cycloaddition reaction. The composition of claim 27, wherein the reaction is a strain-promoted alkyne-azide cycloaddition (SPAAC) reaction. The composition of any one of claims 1 to 28, wherein the cross-linking moiety comprises: or . The composition of claim 29, wherein the cross-linking moiety comprises: or Where m is 1 to 10. The composition of claim 30, wherein the reaction is an inverse electron demand Diels-Alder (IEEDD) reaction. The composition of any one of claims 1 to 31, wherein the viral capsid comprises a nucleic acid cargo. The composition of claim 32, wherein the nucleic acid encodes a protein or an immunogenic polypeptide. A method of preparing a composition as claimed in any of the preceding claims, the method comprising the steps of: (a) combining: (i) a viral exosome comprising a plurality of surface-available primary amines; and (ii) an exosome-reactive linker comprising: a tetrafluorophenyl (TFP) ester and a first member (CRP1) of a crosslinker-reactive pair; thereby providing a composition comprising a surface-functionalized viral exosome; and (b) combining: (i) a functionalized ligand comprising a second member (CRP2) of a crosslinker-reactive pair; (ii) the composition comprising a surface-functionalized viral exosome; thereby providing a composition comprising the surface-modified viral exosome. The method of claim 34, wherein the composition comprising the surface functionalized viral exosome produced in step (a) is used in step (b) without purification or separation steps. The method of claim 34 or 35, wherein the combination of step (b) comprises adding the functionalized ligand directly to the composition comprising the surface functionalized viral capsid produced in step (a). The method of any one of claims 34 to 36, wherein the shell reactive linker is according to formula (II): . The method of any one of claims 34 to 37, wherein step (a) occurs according to the following reaction scheme A: Wherein: p is an integer from 10 to 500, and q is an integer from 10 to 500. The method of any one of claims 34 to 38, wherein the molar ratio of viral capsid (I) to capsid-reactive linker (II) is 1:100 to 1:50,000. The method of claim 39, wherein the molar ratio of viral capsid (I) to capsid-reactive linker (II) is 1:100 to 1:33,000. The method of claim 39, wherein the molar ratio of viral capsid (I) to capsid-reactive linker (II) is 1:100 to 1:10,000. The method of claim 39, wherein the molar ratio of viral capsid (I) to capsid-reactive linker (II) is 1:100 to 1:1,000. The method of any one of claims 34 to 42, further comprising purifying the viral exosome prior to combining step (a). The method of claim 43, wherein the viral exosome is purified by density gradient centrifugation, chromatography, or a combination thereof. The method of claim 44, wherein the density gradient centrifugation is iodixanol density gradient ultracentrifugation. The method of claim 44, wherein the analysis is affinity analysis. The method of any one of claims 34 to 46, wherein the viral exosome is used in step (a)(i) without prior purification. The method of any of claims 34 to 47, wherein SP ¹ is one or more PEG (ie, -(-(O-CH ₂ -CH ₂ ) _n )- or -([PEG] _n )-), wherein n is 2-20. The method of any one of claims 34 to 48, wherein CRP1 comprises a reactive moiety selected from the group consisting of: an azide; an alkyne; a 1,4-triazole; a 1,3-nitrone; a cyclooctyne or a derivative thereof, such as a dibenzylcyclooctyne or a derivative thereof; a triazine; a tetrazine; a strained dienophile; an aryl or alkyl phosphine; an isocyanide; a benzylguanine, a benzylcytosine, and a chloroalkane. The method of claim 49, wherein the CRP1 reactive moiety comprises cyclooctyne. The method of claim 50, wherein the cyclooctyne is selected from dibenzyl cyclooctyne analogs. The method of claim 51, wherein the dibenzyl cyclooctyne analog is selected from the group consisting of dibenzyl cyclooctyne (DIBO), dibenzoazacyclooctyne (DBCO), and biarylazacyclooctyne ketone (BARAC), and functional derivatives thereof. The method of any one of claims 34 to 52, wherein CRP1 comprises the following structure: Wherein m is an integer between 4 and 12, and the waveform key indicates connection to the shell reactive connector. The method of any one of claims 34 to 53, wherein the shell reactive linker has the formula: Wherein: n is an integer from 2 to 20, and m is an integer from 2 to 10. The method of claim 54, wherein the shell reactive linker has the formula: . The method of any one of claims 34 to 55, wherein the shell reactive linker is DBCO-PEG ₄ -TFP. The method of any one of claims 34 to 56, wherein step (a) occurs in an aqueous medium which optionally further comprises at least one buffer. The method of claim 57, wherein the at least one buffer is selected from HEPES, MOPS, MES, phosphate, and bicarbonate. The method of claim 57 or 58, wherein the concentration of at least one buffer does not exceed about 0.1 M. The method of claim 59, wherein the concentration of the at least one buffer is about 0.01 M to about 0.1 M. The method of any one of claims 57 to 60, wherein the aqueous medium further comprises at least about 200 mM of at least one salt. The method of claim 61, wherein the at least one salt is selected from chloride salts, phosphate salts, sulfate salts, citrate salts, sodium salts, potassium salts, calcium salts and magnesium salts. The method of claim 62, wherein the at least one salt is sodium chloride. The method of any one of claims 57 to 63, wherein the ionic strength of the aqueous medium is at least about 150 mM. The method of any one of claims 57 to 64, wherein the aqueous medium further comprises at least one surfactant. The method of claim 65, wherein the concentration of the at least one surfactant is about 0.001% to about 0.005%. The method of claim 65 or 67, wherein the at least one surfactant comprises Pluronic® F68. The method of any one of claims 57 to 68, wherein the pH of the aqueous medium is about 6 to about 10. The method of any one of claims 34 to 68, wherein step (a) occurs at a reaction temperature of about 0°C to about 50°C. The method of any one of claims 34 to 69, wherein step (a) occurs for a reaction duration of about 5 minutes to about 24 hours. The method of any one of claims 34 to 70, wherein step (a) further comprises the step of adding an excess of a quenching agent. The method of claim 71, wherein the quencher comprises an amine-containing compound. The method of claim 71 or 72, wherein the quencher is selected from glycine or Tris buffer. The method of any one of claims 34 to 73, wherein step (b) occurs according to the following reaction scheme B: in: is a functionalized ligand, wherein ^SP2 is a bond or a spacer and L is a ligand; and r is an integer from 10 to 500. The method of claim 74, wherein ^SP1 and ^SP2 are independently selected from one or more divalent groups including a bond, _-CH2- , -O-, -C(=O)- and -N(R)-, wherein R is H or _C1-3 alkyl. The method of claim 75, wherein SP ² is -([PEG] _n )-C(O)-, wherein n is 1 to 100, preferably 1 to 10. The method of any one of claims 34 to 76, wherein the functionalized ligand has the formula: wherein Y is a bond or one or more PEGs. A method as in any one of claims 34 to 77, wherein CRP2 comprises a reactive moiety selected from the group consisting of: an azide; an alkyne; a 1,4-triazole; a 1,3-nitrone; a cyclooctyne or a derivative thereof; a dibenzylcyclooctyne or a derivative thereof; a triazine; a tetrazine; a strained dienophile; an aryl or alkyl phosphine; an isocyanide; a benzylguanine group, a benzylcytosine group, or a chloroalkyl group. A method as in claim 78, wherein CRP2 comprises a stacked nitride. The method of any one of claims 34 to 79, wherein the functionalized ligand is WGA-[PEG] n-azide, wherein n is 1 to 20, preferably 4 to 10. The method of any one of claims 34 to 80, wherein the functionalized ligand is WGA-PEG ₄ -azide. The method of any one of claims 34 to 80, wherein the functionalized ligand is WGA-PEG ₅ -azide. The method of any one of claims 34 to 80, wherein the functionalized ligand is WGA-PEG ₆ -azide. The method of any one of claims 34 to 80, wherein the functionalized ligand is WGA-PEG ₇ -azide. The method of any one of claims 34 to 79, wherein the functionalized ligand is NGF-[PEG] n-azide, wherein n is 1 to 20, preferably 4 to 10. The method of any one of claims 34 to 79 and 85, wherein the functionalized ligand is NGF-PEG ₄ -azide. The method of any one of claims 34 to 79 and 85, wherein the functionalized ligand is NGF-PEG ₅ -azide. The method of any one of claims 34 to 79 and 85, wherein the functionalized ligand is NGF-PEG ₆ -azide. The method of any one of claims 34 to 79 and 85, wherein the functionalized ligand is NGF-PEG ₇ -azide. The method of any one of claims 34 to 89, wherein the molar ratio of the viral exosome (I) in step (a) to the functionalized ligand in step (b) is 1:100 to 1:50,000. The method of claim 90, wherein the molar ratio of the viral exosome (I) in step (a) to the functionalized ligand in step (b) is 1:100 to 1:33,000. The method of claim 90, wherein the molar ratio of the viral exosome (I) in step (a) to the functionalized ligand in step (b) is 1:100 to 1:10,000. The method of claim 90, wherein the molar ratio of the viral exosome (I) in step (a) to the functionalized ligand in step (b) is 1:100 to 1:1,000. The method of any one of claims 34 to 93, wherein the molar ratio of viral exosome to exosome-reactive linker in (a) is the same as the molar ratio of viral exosome to functionalized ligand in (b). The method of any one of claims 34 to 94, wherein (a) and (b) occur in the same reaction vessel. The method of any one of claims 34 to 95, wherein CRP1 and CRP2 are reacted in (b) to form Q. The method of claim 96, wherein (i) CRP1 is DBCO and CRP2 is a stacked nitride or (ii) CRP1 is a stacked nitride and CRP2 is DBCO. The method of claim 96 or 97, wherein Q comprises a triazole. The method of any one of claims 96 to 98, wherein Q is: or . The method of any one of claims 96 to 99, wherein Q is: or Where m is 1 to 10. The method of any one of claims 34 to 100, further comprising preparing the functionalized ligand by reacting a ligand comprising a first reactive moiety with a ligand-reactive linker comprising a second reactive moiety and CRP2. The method of claim 101, wherein the ligand comprises a _-NH2 first reactive moiety. A method as claimed in claim 101 or 102, wherein the ligand-reactive linker comprises a second reactive moiety selected from isothiocyanates, isocyanates, azidoyl groups, NHS esters, sulfonyl chlorides, aldehydes, glyoxals, epoxides, oxiranes, carbonates, aryl halides, imidoesters, carbodiimide anhydrides, benzoyl fluorides and TFP esters. A method as in any one of claims 98 to 103, wherein the ligand-reactive linker is selected from TCO-PEGn-NHS; tetrazine-PEGn-NHS; azido-PEGn-NHS; phosphine-NHS; maleimide-PEGn-succinimidyl ester; DBCO-PEGn-TFP ester, and DBCO-PEGn-NHS ester, wherein n is 1 to 100, preferably 4 to 10. The method of claim 104, wherein the ligand-reactive linker is DBCO-PEGn-TFP ester, wherein n is 1 to 100, preferably 4 to 10. The method of claim 105, wherein the ligand-reactive linker is DBCO-PEG4-TFP. The method of claim 105, wherein the ligand-reactive linker is DBCO-PEG5-TFP. The method of claim 105, wherein the ligand-reactive linker is DBCO-PEG6-TFP. The method of claim 105, wherein the ligand-reactive linker is DBCO-PEG7-TFP. The method of any one of claims 34 to 109, further comprising purifying the composition comprising the surface-modified viral exosome. The method of any one of claims 34 to 110, wherein the viral capsid comprises a nucleic acid cargo. The method of claim 111, wherein the nucleic acid encodes a protein or a polypeptide.