Delivery system for gene editing tools

Abstract

The present invention relates to a protein-ribonucleic acid complex comprising (I) at least one multi-functional protein or a set of at least two proteins and (II) a nucleic acid editor complex, a polynucleotide encoding the multi-functional protein (I) and its uses in the delivery of nucleic acids and gene editing tools. The present invention also relates to a nucleic acid delivery system, a plurality of polynucleotides, pharmaceutical compositions, a cell or a cell line, and kits. The present invention further relates to medical uses of the nucleic acid delivery system or the cell or cell line. Also provided are uses of a cell or a cell line in therapeutic and diagnostic applications, as well as in tissue engineering. Finally, the present invention further relates to an in vitro or in vivo method of editing a nucleic acid molecule in a target cell, and a method for nucleic acid delivery.

Description

[0001] Delivery system for gene editing tools

[0002] The present invention relates to a protein-ribonucleic acid complex comprising (I) at least one multifunctional protein or a set of at least two proteins and (II) a nucleic acid editor complex, a polynucleotide encoding the multi-functional protein (I) and its uses in the delivery of nucleic acids and gene editing tools. The present invention also relates to a nucleic acid delivery system, a plurality of polynucleotides, pharmaceutical compositions, a cell or a cell line, and kits. The present invention further relates to medical uses of the nucleic acid delivery system or the cell or cell line. Also provided are uses of a cell or a cell line in therapeutic and diagnostic applications, as well as in tissue engineering. Finally, the present invention further relates to an in vitro or in vivo method of editing a nucleic acid molecule in a target cell, and a method for nucleic acid delivery.

[0003] BACKGROUND OF THE INVENTION

[0004] The fast-paced developments of various gene editing technologies in the last decade have revolutionized the field of molecular biology, enabling the precise manipulation of virtually any gene with single-nucleotide precision. Especially the recent development of base editors (BEs) (Komor et al., 2016) and prime editors (PEs) (Anzalone et al., 2019), mediating single-nucleotide conversions, deletion / insertions (Anzalone et al., 2022; Choi et al., 2022), or even whole gene insertions (Yamall et al., 2022) all without the necessity of highly mutagenic double-stranded DNA breaks (DSBs), represent a milestone towards the ultimate goal of curing genetic disorders. However, efficient and safe delivery of the respective editing complexes remains a challenge.

[0005] Viral vector-based systems have emerged as a promising method for delivering genome editing machineries. However, despite the advances in this field, there remains a need for improved efficiency and specificity to maximize the potential of these delivery systems. While adeno-associated viruses (AAVs) are considered safe and efficient for in vivo gene delivery, they also face limitations that hinder their broad application in gene editing. These limitations include constrained cargo capacity, potential immunogenicity, and the risk of insertional mutagenesis. Importantly, while their long persistence is usually regarded as a benefit, for the same reason, AAV-based delivery is presumably not ideally suited for certain prime editing strategies. Third-generation prime editing system PE3(b) uses a nicking guide RNA (ngRNA) in addition to the primary prime editing guide RNA (pegRNA) to stimulate the intended edit being incorporated. The long persistence of AAVs, when used for PE3(b) strategies, could lead to long-term continuous nicking events even after the desired edit has been incorporated. Consequently, the risk of unintended mutations at the nick site is considerably higher, and the permanent nick might lead the DNA repair machinery to occupy that site, potentially repressing the transcription of the gene intended for repair.

[0006] Lentiviral vectors have been employed in the generation of chimeric antigen receptor (CAR) T cells, modified to target and eliminate cancer cells, ultimately resulting in FDA-approved CAR-T cell therapies. However, using lentiviral vectors to deliver gene editing complexes raises several severe safety concerns. Lentiviruses integrate their genetic payload into the host genome, providing stable and long-term transgene expression, which is problematic when delivering gene editing machineries, as the persistent expression may increase the risk of undesired on- and off-target effects. The random integration of lentiviral vectors into the host genome also raises safety concerns since their integration can disrupt essential endogenous genes or regulatory elements, potentially leading to oncogenesis. In addition, both AAV-based and lentiviral vector systems pose the risk of high immunogenicity due to their persistent expression of the editing machinery. Consequently, repaired cells are highly likely to be rejected by the host’s immune system.

[0007] Lipid nanoparticles (LNPs) on the other hand have gained attention as non-viral delivery vehicles for nucleic acids and are often favored over AAVs due to their packaging capacity and range of therapeutic applications. However, they often exhibit limited cell-type specificity, potential cytotoxicity, and may require complex formulation optimization for efficient delivery.

[0008] In contrast, virus-like particle (VLP) systems offer several advantages, including the ability to package and deliver genome editing complexes without the risk of insertional mutagenesis or the activation of host immune responses due to viral components or persistent transgene expression. VLPs can be easily pseudotyped, enabling the modulation of cell or tissue tropism by utilizing the diversity of known viruses with different tropisms. This feature allows for the reengineering of glycoproteins to target specific cell types or tissues, further enhancing the versatility and applicability of VLP systems.

[0009] SUMMARY OF THE INVENTION

[0010] According to the present invention this object is solved by a polynucleotide encoding the multifunctional fusion protein of the present invention.

[0011] According to the present invention this object is solved by a complex of at least one multi-functional fusion protein of the present invention with a specific mRNA, sgRNA or pegRNA; or with a geneediting ribonucleoprotein (RNP) complex comprising a specific mRNA, sgRNA or pegRNA.

[0012] According to the present invention this object is solved by using the multi-functional fusion protein of the present invention and / or the polynucleotide encoding it for: generating a nucleic acid delivery system, preferably a virus-like particle (VLP), delivering nucleic acids, preferably gene-editing nucleic acids, epigenome modulator nucleic acids, transcriptome-editing nucleic acids, and / or epitranscriptome modulator nucleic acids.

[0013] According to the present invention this object is solved by a nucleic acid delivery system, said nucleic acid delivery system comprising:

[0014] (A) at least one multi-functional fusion protein of the present invention;

[0015] (B) a gene-editing ribonucleoprotein (RNP) complex, which comprises

[0016] (Bl) an mRNA, a non-coding RNA (ncRNA), or a guide RNA, such as sgRNA or pegRNA, which is tagged by a nucleic acid binding domain; and

[0017] (B2) gene-editing polypeptide(s) or protein(s),

[0018] (C) a viral envelope protein or engineered variants thereof,

[0019] (D) a targeting domain, such as a glycoprotein or engineered variants thereof,

[0020] (E) a moiety mediating endosomal escape, such as a glycoprotein or engineered variants thereof,

[0021] (F) optionally, a reporter gene product such as a luciferase or a fluorescent protein, and

[0022] (G) optionally, a nucleic acid-stabilizing protein, preferably Csy4.

[0023] In one embodiment, the gene-editing ribonucleoprotein (RNP) complex (B) is not a protein fusion to gag. Instead, an indirect transport mechanism occurs via the guide RNA listed in (Bl).

[0024] According to the present invention this object is solved by a plurality of polynucleotides, comprising

[0025] (1) a first polynucleotide comprising a nucleic acid sequence encoding a multi-functional fusion protein as defined herein;

[0026] (2) a second polynucleotide comprising a nucleic acid sequence encoding a gene editor protein;

[0027] (3) a third polynucleotide comprising a nucleic acid sequence encoding a guide RNA, such as a sgRNA or pegRNA, which is optionally tagged by a nucleic acid binding domain; and

[0028] (4) a fourth polynucleotide comprising a nucleic acid sequence encoding an envelope protein.

[0029] Contrary to prior art approaches, the present invention uses in one embodiment a nucleic acid binding domain (NAB) as a covalent part of an analogous “RNA exporter protein” that is not the genome editing (GE) agent to bind a second polypeptide-polynucleotide complex non-covalently. The second polypeptide-polynucleotide complex comprises at least one further NAB that is a GE protein or a fusion thereof to other effector domain(s), such as a reverse transcriptase (RT) domain, and its respective guiding polynucleotide, that is, a guide RNA (gRNA) or a prime editing guide RNA (pegRNA). In one embodiment, the second complex is a genome editing complex, and can also be referred to as a ribonucleoprotein (RNP) complex.

[0030] Thus, in one embodiment of the present invention the system or the protein-ribonucleic acid complex comprises at least two different NABs, wherein the first NAB is an RNA-binding protein (RBP) and the second NAB is a GE protein in complex with a gRNA or pegRNA. In one embodiment of the present invention the system or the protein-ribonucleic acid complex comprises at least two different NABs, wherein the first NAB is an RBP that is an RNA-aptamer binding protein (ABP) that is also a covalent part of the “RNA exporter protein” that non-covalently binds a second NAB that is a GE protein in complex with a gRNA or pegRNA. Contrary, prior art polynucleotide complexes encode one RNA-binding domain that can be a genome-editing protein as part of a covalent fusion, e.g. of an “RNA exporter protein”.

[0031] In summary, the present invention describes a packaging system for GE-RNPs using a first polypeptide chimera comprising membrane-bending domains (MBDs) fused towards NCS motifs and NABs that preferentially are RBPs. In one embodiment, this polypeptide chimera binds to a second complex that is a GE-RNP complex, wherein the guide RNA component of the second complex comprises at least an RNA aptamer that is a stem-loop (e.g., a hairpin) that is bound by a respective ABP encoded on the first polypeptide chimera.

[0032] In one embodiment the RBP is not a genome -editing protein.

[0033] According to the present invention this object is solved by one or more vectors comprising the plurality of polynucleotides of the present invention, wherein the one or more vectors are preferably three vectors.

[0034] According to the present invention this object is solved by a pharmaceutical composition comprising: the polynucleotide of the present invention which encodes the fusion protein of the present invention, the nucleic acid delivery system of the present invention, the plurality of polynucleotides of the present invention or the one or more vectors of the present invention; optionally, pharmaceutically acceptable excipients and / or carrier.

[0035] According to the present invention this object is solved by a cell or cell line comprising the fusion protein of the present invention, the polynucleotide of the present invention, the nucleic acid delivery system of the present invention, the plurality of polynucleotides of the present invention or the one or more vectors of the present invention. According to the present invention this object is solved by a kit comprising: the polynucleotide of the present invention which encodes the fusion protein of the present invention, or the nucleic acid delivery system of the present invention, or the plurality of polynucleotides of the present invention, or the one or more vectors of the present invention, or the cell or cell line of the present invention.

[0036] According to the present invention this object is solved by a nucleic acid delivery system, preferably a virus-like particle, produced by means of transfection, lipofection, electroporation, photoporation, sonoporation, particle bombardment, microinjection, magnetofection, transduction, such as viral transduction, cell-penetrating peptides, calcium-phosphate, nanoparticles, combinations thereof, or otherwise inserting the plurality of polynucleotides of the present invention or the one or more vectors of the present invention into a cell and expressing the components of the nucleic acid delivery system from the plurality of polynucleotides or the one or more vectors of the present invention in the cell, thereby allowing the nucleic acid delivery system, preferably the virus-like particle, to assemble, such as spontaneously assemble, in the cell.

[0037] According to the present invention this object is solved by providing the nucleic acid delivery system of the present invention for use in medicine.

[0038] According to the present invention this object is solved by providing the nucleic acid delivery system of the present invention for use in the treatment of genetic diseases, preferably monogenetic diseases and / or polygenetic diseases.

[0039] According to the present invention this object is solved by an in vitro or in vivo method of editing a nucleic acid molecule in a target cell by nuclease mediated editing, base editing or prime editing, homology directed repair, non-homologous end-joining directed insertion, or microhomology-mediated end-joining directed insertion, such as by co-delivery of an integrase-deficient (D64V) lentivirus (IDLV), wherein, preferably, the IDLV is packaged using a psi-NC interaction, a PP7-PCP, MS2-MCP, and / or a C4-Csy4 interaction, comprising: contacting the target cell with a nucleic acid delivery system of the present invention, or with the pharmaceutical composition of the present invention, thereby installing one or more modifications to the nucleic acid at a target site. According to the present invention this object is solved by a method for nucleic acid delivery, preferably gene-editing nucleic acids. Said method comprises the steps of:

[0040] (I) providing a multi-functional fusion protein of the present invention to a cell, wherein said cell produces in its nucleus a gene-editing ribonucleoprotein (RNP) complex, which comprises an mRNA or a guide RNA, such as a sgRNA or pegRNA, wherein said geneediting ribonucleoprotein (RNP) complex is as defined herein;

[0041] (II) allowing the multi-functional fusion protein to locate to the nucleus of said cell;

[0042] (III) forming a complex of the multi-functional fusion protein with the gene-editing ribonucleoprotein (RNP) complex in the nucleus;

[0043] (IV) allowing the complex of (III) to be exported to the cytosol of the cell;

[0044] (V) budding of virus-like particles comprising the gene-editing ribonucleoprotein (RNP) complex;

[0045] (VI) binding to a target cell surface, preferentially via a specific receptor;

[0046] (VII) internalization into a target cell, preferentially via endocytosis;

[0047] (VIII) transport into the cytosol, e.g. via endosomal escape;

[0048] (IX) dissociation of the RNP (and optionally bound ‘nucleic acid editing complex) from the nucleic acid binding domain;

[0049] (X) optionally transport into the nucleus for genome editing or transactivation, and

[0050] (X) optionally binding of the guide -RNA-editing complex to an RNA target sequence.

[0051] DESCRIPTION OF THE PREFERRED EMBODIMENTS OF THE INVENTION

[0052] Before the present invention is described in more detail below, it is to be understood that this invention is not limited to the particular methodology, protocols and reagents described herein as these may vary. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of the present invention which will be limited only by the appended claims. Unless defined otherwise, all technical and scientific terms used herein have the same meanings as commonly understood by one of ordinary skill in the art. For the purpose of the present invention, all references cited herein are incorporated by reference in their entireties.

[0053] Concentrations, amounts, and other numerical data may be expressed or presented herein in a range format. It is to be understood that such a range format is used merely for convenience and brevity and thus should be interpreted flexibly to include not only the numerical values explicitly recited as the limits of the range, but also to include all the individual numerical values or sub-ranges encompassed within that range as if each numerical value and sub-range is explicitly recited. As an illustration, a numerical range of " 1 to 20" should be interpreted to include not only the explicitly recited values of 1 to 20, but also include individual values and sub-ranges within the indicated range. Thus, included in this numerical range are individual values such as 1, 2, 3, 4, 5 .... 17, 18, 19, 20 and sub-ranges such as from 2 to 10, 8 to 15, etc. This same principle applies to ranges reciting only one numerical value, such as "higher than 100". Furthermore, such an interpretation should apply regardless of the breadth of the range or the characteristics being described.

[0054] Multi-functional fusion protein, its encoding nucleic acid and its uses

[0055] As discussed above, the present invention provides a multi-functional fusion protein, such as a protein- ribonucleic acid complex comprising:

[0056] (I) at least one multi-functional protein or a set of at least two proteins, wherein the at least one multi-functional protein or the set of at least two proteins comprise the following protein domains:

[0057] (a) at least one plasma membrane-interacting domain;

[0058] (b) at least one oligomerization domain;

[0059] (c) a membrane-bending domain;

[0060] (d) at least one shuttling domain comprised of at least one nuclear export signal (NES) and nuclear localization signal (NLS);

[0061] (e) a nucleic acid-binding (NAB) domain, capable of binding a single strand or double-strand RNA or DNA motif; and

[0062] (f) a budding domain, which comprises a monopartite or multipartite motif that enables ESCRT-dependent or ESCRT-independent budding;

[0063] (g) optionally, a reverse transcriptase domain;

[0064] (h) optionally, a maturation factor;

[0065] (i) optionally, a reporter-signal generating domain;

[0066] ( / ) optionally one nucleic acid-stabilizing domain and

[0067] (II) a nucleic acid editor complex, which comprises an mRNA, a guide RNA or a ncRNA, which is bound via the nucleic acid-binding domain (e) of (I), preferably as ribonucleoprotein complex, wherein only the RNA species comprises the nucleic acid motif, bound by NAB domain (e) of (I), wherein the RNA is preferentially a gRNA, and wherein the RNA species comprises optionally an autonomous 3 ’-stabilization motif and / or a motif that is protected by an additional heterologously expressed nucleic acid-stabilizing domain ( / ) of (I).

[0068] In one embodiment, the nucleic acid-binding (NAB) domain (e) of (I) is capable of binding a specific single stranded or double -stranded RNA or DNA motif.

[0069] The multi-functional protein of the present invention is designed to:

[0070] (1) shuttle between the nucleus and cytosol, (2) bind to specific nucleic acids in the cell,

[0071] (3) bud off the plasma membrane of a packaging / sender cell along with specific nucleic acids and co-packaged proteinaceous components, and

[0072] (4) internalize into a recipient / target cell, delivering the specific nucleic acids, or co-delivering specific nucleic acid and proteinaceous components for cellular processing, such as gene editing.

[0073] - Plasma membrane-interacting domain (a)

[0074] The multi-functional fusion protein of the present invention comprises at least one plasma membraneinteracting domain (a).

[0075] Said plasma membrane-interacting domain (a) is preferably a lipid group. A preferred lipid group is / are C12 to C18 fatty acid(s), preferably myristoyl (C14), palmitoyl (C16), and / or isoprenoid(s), preferably C-famesyl. In one embodiment, the myristoyl moiety is N-terminal or near the N-terminus. In one embodiment, the famesyl moiety is C-terminal or near the C-terminus. In one embodiment, the palmitoylation is an S-palmitoylation. In one embodiment, palmitoylation is an O- or N- palmitoylation.

[0076] In one embodiment, the plasma membrane-interacting domain (a) is preferably a polypeptide domain capable of binding membrane proteins or membrane components, which are, for example, fatty acid chains, phospholipids, glycolipids, phosphorglycerids, sphingolipids, sterols. In one embodiment, said polypeptide domain is a pleckstrin homology (PH) domain, such as phospholipase C-51 pleckstrin homology domain5 (PH).

[0077] Further such polypeptide domains are C2 domain, FYVE domain, phox homology (PX) domains, PHD finger domain, PROPPINs (P-propellers that bind polyphosphoinositides) domain, epsin N- terminal homology (ENTH) domain, BAR (Bin-Amphiphysin-Rvs) domains, Tubby domain. The skilled artisan knows further plasma membrane -interacting domains.

[0078] - Oligomerization domain (b)

[0079] The multi-functional fusion protein of the present invention comprises at least one oligomerization domain (b).

[0080] The at least one oligomerization domain (b) preferably enables homo- or hetero-multimerization of one or multiple polypeptide species.

[0081] The multi-functional protein of the present invention preferably forms dimers.

[0082] The at least one oligomerization domain (b) enables homo-and / or hetero multimerizing, such as dimer, trimer, tetramer, etc. Preferably, the oligomerization domain (b) comprises or contains coiled coils or PDZ domains. In one embodiment, the oligomerization domain (b) comprises or contains de novo generated homo / hetero multimerization domains composed of coil-only, sheet-only, or mixed-type proteins.

[0083] In one embodiment, the multi-functional fusion protein of the present invention is referred to as “ENVLPE” (Engineered Nucleocytosolic Vehicles for Loading of Programmable Editors).

[0084] In one embodiment, the multi-functional fusion protein of the present invention comprises at least two oligomerization domains (b). In one embodiment, the multi-functional fusion protein of the present invention is an ENVLPE+, which is a prime editing system that has been improved by introducing homo-oligomeric coiled-coils. In one embodiment, the multi-functional fusion protein of the present invention comprises at least two oligomerization domains (b), and is referred to as ENVLPE+.

[0085] - Membrane-bending domain (c)

[0086] The multi-functional fusion protein of the present invention comprises a membrane-bending or a curvature-forming domain (c).

[0087] The membrane-bending domain (c) preferably induces plasma membrane bending and formation of nanospheres with a diameter, preferably in the range from about 50 nm to about 1 pm.

[0088] In one embodiment, the membrane-bending domain (c) is a structural polyprotein, such as the HIV-1 gag polyprotein, and / or modified / engineered / truncated variants thereof.

[0089] - Nuclear-cytosolic shuttling signal (d) / shuttling domain (d)

[0090] The multi-functional fusion protein of the present invention comprises at least one nuclear export signal (NES), and nuclear localization signal (NLS).

[0091] In a preferred embodiment, the at least one shuttling domain (d) is a nuclear-cytosolic shuttling domain. According to this invention, the terms “shuttling domain”, “nuclear-cytosolic shuttling domain” and “nuclear-cytosolic shuttling signal” can be used interchangeably.

[0092] In one embodiment, the at least one shuttling domain (d) facilitates the shuttling between the nucleus and the cytosol. Importantly, without NCS motifs, the MBD-ABP is localized exclusively at the plasma membrane in the cytoplasmic environment, while the GE-RNP, which comprises at least a nuclear localization signal, is spatially separated in the nucleus. Thus, a system not comprising an NCS motif cannot function efficiently and merely relies on the short period during mitosis, where the spatial separation of the nuclear and cytoplasmic environment disappears during the breakdown of the nuclear envelope.

[0093] The present inventors have thus solved the above problem by tethering NCS sequences comprising at least one nuclear localization signal (NLS) and at least one nuclear export signal (NES). As a consequence, the MBD-ABP fusion travels from the cytoplasmic environment using its NLS; there, it binds the GE-RNP via the grafted aptamer handle in the RNA component of the GE-RNP and forms a non-covalent MBD-ABP-GE-RNP complex. Due to the NES, the MBD-ABP-GE-RNP complex is exported back to the cytoplasmic compartment, where the MBD binds to the plasma membrane for membrane-bending and budding.

[0094] The NES facilitates the export of the multi-functional protein towards the cytosol when the multifunctional protein is localized in the nucleus.

[0095] In one embodiment, the at least one shuttling domain (e.g., the nucleocytoplasmic shuttling (NCS) domain according to (I)(e)), i.e., the at least one nuclear export signal (NES) and the at least one nuclear localization signal (NLS), are present on the same polypeptide chain, already responsible for a) membrane anchoring, b) budding from the plasma membrane, and c) oligomerization. Thus, the tripartite functions a)-c) are combined in a single molecule, which can be referred to as a membranebudding domain, MBD.

[0096] Thus, unlike prime editing systems of the prior art, the invention employs both nuclear localization signals (NLS) and nuclear export signals (NES) on a single polypeptide combined with an aptamerbinding protein (ABP) to facilitate non-covalent packaging of the GE-RNP. The at least one shuttling domain (according to (I)(d)) enables efficient trafficking between nuclear and cytoplasmic compartments, ensuring the effective loading of nuclear GE-RNPs onto membrane-budding domains (MBDs) for delivery.

[0097] In one embodiment, the multi-functional fusion protein of the present invention comprises at least one exogenous and / or heterologous shuttling domain, such as at least one exogenous and / or heterologous nuclear-cytosolic shuttling domain. In one embodiment, the multi-functional fusion protein of the present invention comprises at least one additional (i.e. non-inherent) shuttling domain, such as at least one additional (i.e. non-inherent) nuclear-cytosolic shuttling domain. In one embodiment, the multifunctional fusion protein of the present invention comprises at least one shuttling domain comprising at least one optimized NLS and / or at least one optimized NES. In one embodiment, the multifunctional fusion protein of the present invention comprises at least one additional (i.e. non-inherent) NLS and / or at least one additional (i.e. non-inherent) NES.

[0098] In one embodiment, the NES (d) is the NES of HIV-1 (NESHIV-I), preferably wherein the NES comprises an amino acid sequence according to SEQ ID NO: 65. In one embodiment, the NES (d) is the NES of DBRl.

[0099] The NLS is a mono- or bipartite NLS and facilitates the import of the multi-functional fusion protein towards the nucleus when the multi-functional protein is localized in the cytosol.

[0100] In one embodiment, the NLS (d) of (I) is the NLS of c-Myc protein (NLSC-Myc), NLS of SV40 (NLSSV4O) or a synthetic NLS, such as attenuated variants of NLSC-Myc and NLSsv4o, or other mono- or bipartite synthetic NLS, preferably wherein the NLS comprises an amino acid sequence according to any one of SEQ ID NO: 48 to 54.

[0101] The invention introduces a particle system harnessing a virus-like particle (VLP)-based platform as an exemplary platform for delivering genome editing ribonucleoproteins (GE-RNPs) that addresses critical limitations in existing systems. By combining nucleocytoplasmic shuttling (NCS) motifs, aptamer-based guide RNA recruitment, and optimized stabilization mechanisms for prime editors, this platform surprisingly provides an innovative, efficient, and versatile solution for genome editing applications.

[0102] The inventors surprisingly found that the combination of an NES together with an NLS in one protein, namely the multi-functional fusion protein (I) of the present invention, enables the protein’s specific shuttling functionality where it shuttles between the nucleus and the cytosol and enables thereby the binding of its RNA / RNP cargo in the nucleus and the transport towards the cell membrane. The inventors are the first ones to show that by introducing the shuttling NLS / NES gag in the context of an RNP-editor loading to VLPs, the editing efficacy of the resulting RNP-VLPs can strongly be increased.

[0103] In a preferred embodiment, the NES and NLS together form a shuttling motif.

[0104] For example, NLS-NES-NLSFLAG, wherein NLSFLAG refers to a synthetic NLS, preferably NESc-wyc-NESniv-i -NLSFLAG.

[0105] - Nucleic acid-binding domain (e) The multi-functional fusion protein of the present invention comprises at least one nucleic acidbinding domain (e).

[0106] Said nucleic acid-binding domain (e) is capable of binding single strand or double-strand RNA or DNA, namely through specific interactions, including recognition of linear motifs, specific secondary structures, tertiary structures, or combinations thereof.

[0107] Preferably, nucleic acid-binding domain (e) is capable of binding specific RNA molecules, including mRNA, guide RNA (such as sgRNA, pegRNA), and DNA, which can be tagged with specific RNA aptamer motifs.

[0108] In one embodiment, the nucleic acid-binding domain (e) of (I) is a C4-aptamer binding protein (Csy4- enzyme) domain, preferably a Csy4 domain comprising an amino acid sequence according to SEQ ID NO: 64. In one embodiment, the nucleic acid-binding domain (e) of (I) is an RNA-binding PP7 coat protein (PCP) domain, optionally single chain tandem PCP, preferably a PCP domain comprising an amino acid sequence according to SEQ ID NO: 77, 78, or 79. In one embodiment, the nucleic acidbinding domain (e) of (I) is an MS2 coat protein (MCP) domain, preferably a MCP domain comprising an amino acid sequence according to SEQ ID NO: 58, or a single chain tandem MCP. In one embodiment, the nucleic acid-binding domain (e) of (I) is a boxB domain or the N Protein of bacteriophage P22.

[0109] In one embodiment, the nucleic acid-binding domain (e) of (I) is a Com-com, which is a Com protein that binds to the structure of a com RNA hairpin, as described in Zalatan etal. (2014). In one embodiment, the nucleic acid-binding domain (e) of (I) a psi-binding nucleocapsid (NC) domain of HIV-1 Gag. The NC domain of HIV-1 gag can be used to package psi-tagged RNAs. All aptamer binding protein domains can be arranged as single chain tandem protein.

[0110] - Budding domain (f)

[0111] The multi-functional fusion protein of the present invention comprises a budding domain (g).

[0112] Said budding domain preferably comprises a monopartite or multipartite motif that enables ESCRT- dependent or ESCRT-independent budding.

[0113] In one embodiment, the budding domain () of (I) is a HIV-1 p6 domain, preferably a p6 domain comprising an amino acid sequence according to SEQ ID NO: 75, an equine infectious anemia virus (EIAV) p9 domain, a murine leukaemia virus (MLV) pl2 domain, a rouse sarcoma virus (RSV) or human T-cell leukemia virus type 1 (HTLV-1) pl9 domain, aretroviral L domain, a vesicular stromatitis virs (VSV) M protein, an influenza virus M2 protein, an alphavirus F13 protein, or an arenavirus Z protein. - optionally, a reverse transcriptase domain (g)

[0114] The multi-functional fusion protein of the present invention optionally comprises a reverse transcriptase domain (g).

[0115] In one embodiment, the RT domain (g) of (I) is pol of HIV- 1 or a mutant thereof, wherein the enzymatic function of the integrase domain is inactivated by mutating at least one of the amino acids of pol of HIV-1 at position D64 (such as pol D64V), DI 16, E152, and / or F185 (i.e. the key amino acids of pol of HIV- 1).

[0116] - optionally, a maturation factor (h)

[0117] The multi-functional fusion protein of the present invention optionally comprises a maturation factor (h).

[0118] Importantly, the inventors found a way to modularly package the RNP via the guide RNA and maturation of gag occurs much more naturally compared to prior work where the Cas component is directly fused to the gag as a fusion protein and its release relies on an additional protease cleavage site. Packaging via the guide RNA and maturation of gag according to this invention is thus much simpler and more modular since any RNP, and any Cas protein that uses a guide RNA (e.g., Cas9, Casl2, Casl3) can be packaged without the need to optimize the protein fusion, such as the linker and the proteolytic site, compared to prior work where the Cas component is directly fused to gag as fusion protein. Also in the case where coiled-coiled adapters are used instead of a direct fusion to gag, the guide-RN A mediated packaging is superior to the Cas-protein-mediated packaging of the editor RNP complex because the integrity of the RNP com ponent with the least stability and shortest half-life time (i.e. the guide RN A component) is used as the handle, thereby ensuring that fully functional RNP editor complexes are loaded instead of Cas proteins without functional guideRNA.

[0119] Importantly, the inventors further discover that maturation of gag is dispensable, as shown using miniGag constructs. The inventors further discovered that the dissociation of the RNP-payload occurs naturally due to the non-covalent packaging. Moreover, the dissociation of the RNP-payload is not dependent on proteolysis.

[0120] In one embodiment, the maturation factor (h) of (I) is an enzyme capable of cleaving the multifunctional fusion protein of the present invention into multiple functional domains, preferably when it is interacting with the membrane.

[0121] In one embodiment, the maturation factor (h) is the HIV-1 protease. In one embodiment, at least two maturation sites upstream and downstream of the NAB domain (e) are present.

[0122] Preferably, the multi-functional fusion protein of the present invention does not contain a proteolytic activity. In one embodiment, the release of the ribonucleoprotein complex does not rely on proteolytic maturation of the fusion protein. In one embodiment, the multi-functional fusion protein of the present invention optionally comprises proteolytic activity to enhance infectivity.

[0123] Importantly, the inventors found a mechanism that dissociates Gag maturation from RNP-packaging, thus making the functionality (4), i.e. simultaneous co-packaging and delivery of protein components modular gene editors without the need for further modification (e.g., fusion or proteolytic activation) of the gene editor proteins highly beneficial over prior art work.

[0124] - optionally, a reporter-signal generating domain (i)

[0125] The multi-functional fusion protein of the present invention optionally comprises a reporter-signal generating domain (i).

[0126] In one embodiment, the reporter-signal generating domain (i) of (I) is fluorescence, bioluminescence, photoabsorption or photoacoustic detection, detection by Nuclear Magnetic Resonance, Positron emission tomography, as well as read out via biochemical methods, RNA or DNA detection methods such as RT-qPCR, qPCR, or next-generation sequencing.

[0127] - optionally a nucleic acid-stabilizing domain (j)

[0128] The multi-functional fusion protein or a set of at least two proteins of the present invention optionally comprise a nucleic acid-stabilizing domain (j) to specifically bind to an aptamer motif on a pegRNA whereby the pegRNA is stabilized and protected.

[0129] In one embodiment, the set of at least two proteins of the present invention are 2 to 10 proteins, such as 2 to 5 proteins.

[0130] In one embodiment, the nucleic acid-stabilizing domain (j) is the Csy4 / Cas6f protein in combination with the C4 aptamer on the 3’ of the pegRNA. The Csy4 / C4 combination can be used for RNP packaging, while simultaneously processing and stabilizing the pegRNA to enhance prime editing (PE).

[0131] For further details, see also Figures 10 and 11. The inventors surprisingly found that Csy4 / C4 is not equivalent to other aptamer systems. So far, Csy4 has been used only biotechnologically for its processing activity to process fusion arrays of gRNAs and crRNAs to single components. The inventors are the first ones to show that Csy4 can be used for VLP -mediated export as well as protection of pegRNAs at the same time since pegRNA stability is a major bottleneck. Importantly, pegRNAs have to remain stable for a very long time, from producer cell line to VLPs, in the recipient cells all across the cytosol into the nucleus until it reaches its target.

[0132] Unlike systems of the prior art, the present invention includes a dedicated 3 ’-protection system based on Csy4 / C4 that represents a significantly better 3 ’-protection for prime editing systems compared to prime editing systems of the prior art (e.g., evopreQl) as shown by the inventors in direct comparisons.

[0133] In one embodiment, Csy4 is not used to prevent pegRNA circularization. Instead of implementing Csy4 to prevent pegRNA circularization, Csy4 is used in one embodiment of this invention to stabilize its inherently degradation-prone 3 ’ extension further.

[0134] Alternatively, instead of the Csy4 / C4 system, the Rotavirus N-terminal fragment (aa 1-162) of the NSP3 protein comprising an acid sequence according to SEQ ID NO: 323 could be used for protecting an engineered pegRNA that ends with a 3’-gacc. To exactly terminate at gacc, DNA coding for the pegRNA needs to have downstream of the gacc motif a 3’-tRNA or 3’-tRNA mimicking motif or a ribozyme motif, such as the genomic or anti-genomic Hepatitis D virus (HDV) ribozyme, encoded. The tRNA is processed by various enzymes, such as RNAse P, which cleaves upstream of the tRNA and thus, when placed downstream of the gacc motif, can be used to generate exact 3 ’-ends. The genomic or anti -genomic HDV ribozyme is used with the same intention to generate exact 3 ’-ends when placed downstream of gacc. In contrast to tRNA, ribozymes are autocatalytically and thus do not require any host processing enzymes.

[0135] In one embodiment, the multi-functional fusion protein of the present invention is based on groupspecific antigen (gag), such as HIV-1 gag, or on gag-pol polyprotein.

[0136] Group-specific antigen, or gag, is the polyprotein that contains the core structural proteins of a retrovirus. It makes up all the structural units of viral conformation and provides supportive framework for mature virion. All orthoretroviral gag proteins are processed by the protease (PR or pro) into MA / matrix), CA (capsid), NC (nucleocapsid) parts, and sometimes more. If Gag fails to cleave into its subunits, virion fails to mature and remains uninfective. Gag-Pol protein mediates, with Gag polyprotein, the essential events in virion assembly, including binding the plasma membrane, making the protein-protein interactions necessary to create spherical particles, recruiting the viral Env proteins, and packaging the genomic RNA via direct interactions with the RNA packaging sequence (Psi). Gag-Pol polyprotein may regulate its own translation, by the binding genomic RNA in the 5'-UTR. At low concentration, the polyprotein would promote translation, whereas at high concentration, the polyprotein would encapsidate genomic RNA and then shut off translation.

[0137] In preferred embodiments, the multi-functional fusion protein of the present invention based on (group-specific antigen) gag and / or gag-pol, which preferably comprises any one of the following structures:

[0138] Gag A-myr-[MA]-[CA]-p2-[NC]-pl-[p6]

[0139] Gag-pol D64V A-myr-[MA]-[CA]-p2-[NC]-p 1 -[p6*]-PR-RT-INTD64v

[0140] GagL2i s A-myr-[MAL2is]-[CA]-p2-[NC]-pl-[p6]

[0141] Gag AZF+L21 S" PCP A-myr-[MAL2is]-[CA]-p2-[NCzFimutAZF2-PCP]-pl-[p6]

[0142] GagAZF+L21 S / NLS-NES- PCP A-myr-[MAL21s]-[CA]-p2-[NCzFlmutAZF2+NLS-NES-PCP]-pl-[p6] wherein

[0143] A-myr refers to an N-terminal myristoyl;

[0144] MA refers to the matrix protein of Gag, such as pl7 matrix protein of HIV- 1;

[0145] MALZIS refers to the matrix protein containing the mutation L21S;

[0146] CA refers to the capsid protein of Gag, such as p24 capsid protein of HIV- 1; pl refers to spacer peptide 1 of Gag;

[0147] NC refers to the nucleocapsid protein of Gag comprising zink finger motif 1 (ZF1) and zinc finger motif 2 (ZF2) such as p7 capsid protein of HIV- 1, p2 refers to refers to spacer peptide 2 of Gag; p6 refers to the 6kDa polypeptide at the C-terminus of Gag; p6* refers to the cryptic p6 peptide, which is expressed after ribosomal frame shifting;

[0148] PR refers to the protease of Gag;

[0149] RT refers to the reverse transcriptase of Gag;

[0150] INT refers to the integrase of Gag;

[0151] PCP refers to PP7 coat protein;

[0152] NCAZF-PCP refers to a nucleocapsid protein wherein ZF1 was inactivated by mutation of 3 cysteines to serine and wherein ZF2 was replaced by PCP; NCAZF+NLS-NES-PCP refers to a nucleocapsid protein wherein ZF1 was inactivated by mutation of 3 cysteines to serine and wherein ZF2 was replaced by NLS-NES-PCP;

[0153] NLS-NES refers to s shuttling motif containing both a nuclear localization signal and a nuclear export signal, preferably NLS-NES-NLSFLAG, wherein NLSFLAG refers to a synthetic NLS such as NLSc-Myc-NESniv-1 -NLSFLAG;

[0154] For further details, see also Figures 1, 2A, 3A.

[0155] In one embodiment, there is no direct fusion between gag and the effector protein.

[0156] In one embodiment, the multi-functional fusion protein of the present invention is based on truncated gag and / or on truncated gag-pol polyprotein.

[0157] In preferred embodiments, the multi-functional fusion protein of the present invention based on truncated gag, non-gag variants, and / or gag-pol is truncated in MA and CA and the NC domain is replaced by a shuttling motif and PCP, more preferably comprising any one of the following structures: miniGagAZF+L2is / NLs-NEs-PCP '-myr-| MA .i 2-i i4|-|CA 'i33-277|-p2-CC-NLS-NES-PCP-|p61

[0158] PHminiGagAZF+L2is / NLs-NEs-PCP PH-[MAAI2-I I4]-[CAAI33-277]-P2-CC-NLS-NES-PCP-[P6] wherein

[0159] A-myr refers to an N-terminal myristoyl;

[0160] PH refers to a pleckstrin homology domain;

[0161] MA I2-II4 refers to the matrix protein of Gag wherein amino acid residues 12 to 114 are deleted;

[0162] CAAI33-277 refers to the capsid protein of Gag wherein amino acid residues 133 to 277 are deleted; p2 refers to refers to spacer peptide 2 of Gag;

[0163] CC refers to a dimerization domain, preferably the general control transcription factor GCN4 NLS-NES refers to s shuttling motif containing both a nuclear localization signal and a nuclear export signal, preferably NLS-NES-NLSFLAG, wherein NLSFLAG refers to a synthetic NLS such as NLSc-Myc-NESniv-1 -NLSFLAG;

[0164] PCP refers to PP7 coat protein; p6 refers to the 6kDa polypeptide at the C-terminus of Gag; For further details, see also Figure 7A.

[0165] In preferred embodiments, the multi-functional fusion protein (I) of the present invention is based on group-specific antigen (gag), more preferably HIV-1 gag, and / or a gag-pol polyprotein, or a mini-Gag variant, and comprises an amino acid sequence selected from any one of SEQ ID NOs: 1 to 81. In preferred embodiments, the multi-functional protein (I) comprises an amino acid sequence selected from any one of SEQ ID NOs: 1, 2, 13, 19, 24, 25, 26, 27, 60, and 61.

[0166] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 60, and the following structure: N-myr-[MA]-[CA]-p2-[NC]-p 1 -[p6] .

[0167] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 26 or 27, and the following structure: A-myr-[MA]-[CA]-p2-[NC]-pl-[p6*]-PR-RT-INTD64v.

[0168] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 61, and the following structure:

[0169] A-myr-[MAL2is]-[CA]-p2-[NC]-pl-[p6].

[0170] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 1, and the following structure:

[0171] V-myr-[MAL2iS]-[CA]-p2-[NCzFimutAZF2-PCP]-pl-[p6].

[0172] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 2, and the following structure:

[0173] V-myr-[MAL21s]-[CA]-p2-[NCzFlmutAZF2+NLS-NES-PCP]-pl-[p6].

[0174] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 13, and the following structure:

[0175] A-myr-[MAAi2-ii4]-[CAAi33-277]-p2-CC-NLS-NES-PCP-[p6].

[0176] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 19, and the following structure:

[0177] PH-[MAAI2-II4]-[CAAI33-277]-P2-CC-NLS-NES-PCP-[P6]. In one embodiment, the multi-functional fusion protein of the present invention is based on an ENVLPE variant, such as a variant referred to as “ENVLPE+”.

[0178] In preferred embodiments, the multi-functional fusion protein (I) of the present invention is based on ENVLPE+, and comprises an amino acid sequence selected from any one of SEQ ID NOs: 393 to 396.

[0179] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 393, and the following structure:

[0180] Gag A-myr-[MA]-[CA]-p2-[NC]-p 1 -[p6] .

[0181] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 394, and the following structure:

[0182] Gag-pol D64v A-myr-[MA]-[CA]-p2-[NC]-p 1 -[p6*]-PR-RT-INTD64v

[0183] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 395, and the following structure:

[0184] Gag-L21S-(ZFl-C>D,AZF2)-GCN4-HiBiT-NLS-NES-PCP.

[0185] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 396, and the following structure: PH-Gag(A12-114_A133-277_ANC-pl)-GCN4-HiBiT-NLS-NES-NLS-PCP-p6. In one embodiment, the multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 396 can be referred to as miniENVLPE.

[0186] In one embodiment, the multi-functional fusion protein of the present invention is based on Vesicular stomatitis virus (VSV) glycoprotein.

[0187] In preferred embodiments, the multi-functional fusion protein (I) of the present invention is based on Vesicular stomatitis virus (VSV) glycoprotein, and comprises the amino acid sequence according to SEQ ID NO: 406.

[0188] In one embodiment, the multi-functional fusion protein of the present invention is based on a phage capsid protein, such as an engineered version of a PP7 coat protein (PCP) and / or P22 bacteriophage. In one embodiment, the capsid function of PP7 / PCP is partially reactivated, whereby PCP / PP7 is not only used as an aptamer-binding protein, but also partially fulfills the requirement of the “oligomerization” domain.

[0189] The inventors used the PP7 aptamer system as an execution example for an ABP-aptamer system, where the RNA aptamer is the PP7 aptamer (PP7), and the ABP is an engineered version of a PP7 coat protein (PCP). Thus, the inventors included additional data that only uses the p6 motif of HIV-1 Gag for budding stimulation and demonstrated successful functionality with alternative non-Gag budding modules. This independence enhances modularity and applicability across diverse contexts.

[0190] In preferred embodiments, the multi-functional fusion protein (I) of the present invention is based on a phage capsid protein, and comprises an amino acid sequence selected from any one of SEQ ID NOs: 453 to 491. In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 453, and the following structure:

[0191] P22-gp5-XTEN-GCN4-trimer-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 453 is GA1835.

[0192] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 454, and the following structure: P22-gp5-XTEN-GCN4-trimer-HIV-NES-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 454 is GA2130.

[0193] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 455, and the following structure: P22-gp5-HIV-NES-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 455 is GA2131.

[0194] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 456, and the following structure: P22-gp5-XTEN-GCN4-trimer-NLS-NES-NLS-PCP-p6-HiBit-PH-PLC -delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 456 is GA2132.

[0195] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 457, and the following structure:

[0196] P22-gp5-NLS-NES-NLS-PCP-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 457 is GA2133.

[0197] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 458, and the following structure:

[0198] P22-gp5-NLS-NES-NLS-tdPCP-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 458 is GA2156. In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 459, and the following structure:

[0199] P22-gp5-XTEN-GCN4-trimer-NLS-NES-NLS-PCP-p6-HiBit-PH-PLC -delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 459 is GA2166.

[0200] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 460, and the following structure: P22-gp5-PCP-p6-HiBit-PH-PLC -delta. An internal reference number that has been used for a multifunctional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 460 is GA2206.

[0201] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 461, and the following structure:

[0202] P22-gp5-GCN4-trimer-p6-HiBit-PH-PLC -delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 461 is L1484.

[0203] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 462, and the following structure: P22-gp5-XTEN-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 462 is L1485.

[0204] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 463, and the following structure: P22-gp5-XTEN-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 463 is L1486.

[0205] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 464, and the following structure: P22-gp5-XTEN-GCN4-trimer-PCP-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 464 is L1583.1. In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 465, and the following structure:

[0206] P22-gp5-NLS-NES-NLS-(EAAAR)x4-PCP-(EAAAR)x4-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 465 is GA2198.

[0207] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 466, and the following structure:

[0208] P22-gp5-NLS-NES-NLS-(EAAAR)x2-PCP-(EAAAR)x4-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 466 is L1577.

[0209] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 467, and the following structure:

[0210] P22-gp5-NLS-NES-NLS-(EAAAR)xl-PCP-(EAAAR)x4-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 467 is L1578.

[0211] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 468, and the following structure:

[0212] P22-gp5-NLS-NES-NLS-(EAAAR)x4-PCP-(EAAAR)x2-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 468 is L1579.

[0213] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 469, and the following structure:

[0214] P22-gp5-NLS-NES-NLS-(EAAAR)x4-PCP-(EAAAR)xl-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 469 is L1580.

[0215] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 470, and the following structure:

[0216] P22-gp5-NLS-NES-NLS-(EAAAR)xl-PCP-(EAAAR)xl-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 470 is L 1581. In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 471, and the following structure: P22-gp5-NLS-NES-NLS-(EAAAR)x2-PCP-(EAAAR)x2-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 471 is L1582.

[0217] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 472, and the following structure: mGL-XTEN-P22-gp8(141-303). An internal reference number that has been used for a multifunctional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 472 is GA1837.

[0218] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 473, and the following structure: NLS-NES-NLS-PCP-XTEN-P22-gp8( 141-303). An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 473 is GA2128.

[0219] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 474, and the following structure: NLS-NES-NLS-tdPCP-XTEN-P22-gp8(141-303). An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 474 is GA2129.

[0220] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 475, and the following structure: HIV-NES-XTEN-P22-gp8(141-303). An internal reference number that has been used for a multifunctional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 475 is GA2134.

[0221] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 476, and the following structure: tdPCP-XTEN-GCN4-trimer-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC -delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 476 is GA2189. In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 477, and the following structure: tdPCP-active-XTEN-GCN4-trimer-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC -delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 477 is GA2190.

[0222] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 478, and the following structure: tdPCP_GCN4-trimer-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 478 is L1563.

[0223] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 479, and the following structure: tdPCP-XTEN-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 479 is L1564.

[0224] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 480, and the following structure: tdPCP-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 480 is L1565.

[0225] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 481, and the following structure: tdPCP-active-GCN4-trimer-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC -delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 481 is L1566.

[0226] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 482, and the following structure: tdPCP-active-XTEN-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 482 is L1567. In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 483, and the following structure: tdPCP-active-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 483 is L1568.

[0227] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 484, and the following structure: PCP-XTEN-GCN4-trimer-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 484 is L1569.

[0228] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 485, and the following structure: PCP_GCN4-trimer-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 485 is L1570.

[0229] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 486, and the following structure: PCP-XTEN-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 486 is L1571.

[0230] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 487, and the following structure: PCP-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 487 is L1572.

[0231] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 488, and the following structure: PCP-active-XTEN-GCN4-trimer-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 488 is L1573. In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 489, and the following structure:

[0232] PCP-active-GCN4-trimer-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 489 is L1574.

[0233] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 490, and the following structure:

[0234] PCP-active-XTEN-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 490 is L1575.

[0235] In a preferred embodiment, the multi-functional fusion protein of the present invention comprises an amino acid sequence according to SEQ ID No: 491, and the following structure:

[0236] PCP-active-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC -delta. An internal reference number that has been used for a multi-functional fusion protein of the present invention comprising an amino acid sequence according to SEQ ID No: 491 is L1576.

[0237] The multi-functional fusion protein of the present invention exhibits a highly modular functionality which is achieved by one single protein, such as based on Gag and / or based on a phage capsid protein, such as an engineered version of a PP7 coat protein (PCP), and / or based on P22 bacteriophage, which comprises all of the following 8 functionalities:

[0238] (1) budding off of the membrane with encapsulated cargo,

[0239] (2) shuttling between nucleus and cytosol via optimized combined NLS / NES motifs to pick up nuclear cargo and transport to the budding site,

[0240] (3) binding and transport of specific RNA molecules, including mRNA, sgRNA, pegRNA, and optionally RNA for reverse transcription, tagged with specific RNA aptamer motifs,

[0241] (4) simultaneous co-packaging and delivery of protein components, such as modular gene editors, e.g., Cas9-fusions to base editors or reverse transcriptase without the need for further modification (e.g., fusion or proteolytic activation) of the gene editor proteins,

[0242] (5) cellular uptake into the target cell via modular glycoproteins,

[0243] (6) delivery of the RNP into the nucleus of the target cell,

[0244] (7) optionally maturation of VLP for cellular uptake and delivery of RNAs, or RNPs or HDR- donor DNA,

[0245] (8) optional reverse transcriptase (RT) functionality for synthesizing DNA donors for HDR- mediated insertions, wherein the HDR donor-templates can be bound as RNA via one of the nucleic acid-binding domains, preferably NC or PCP, before reverse transcription in the VLP or the target cell. For further details, see also Figure 12.

[0246] In one embodiment, the HDR donor-templates can be bound as RNA via one of the nucleic acid-binding domains, preferably NC or PCP before reverse transcription in the VLP or the target cell.

[0247] As discussed above, the present invention provides a polynucleotide encoding the multi-functional fusion protein of the present invention.

[0248] The polynucleotide encoding the multi-functional fusion protein of the present invention preferably comprises or consists of a nucleotide sequence selected from any one of SEQ ID NOs: 82 to 159, 393 to 396, 406, or 453 to 491.

[0249] As discussed above, the present invention provides the use of the multi-functional fusion protein of the present invention and / or the polynucleotide encoding it for generating a nucleic acid delivery system, preferably a virus-like particle (VLP).

[0250] As discussed above, the present invention provides the use of the multi-functional fusion protein of the present invention and / or the polynucleotide encoding it for delivering nucleic acids, preferably geneediting nucleic acids, epigenome modulator nucleic acids, transcriptome -editing nucleic acids, and / or epitranscriptome modulator nucleic acids.

[0251] Complex of the fusion protein with RNPs

[0252] As discussed above, the present invention provides a complex of at least one multi-functional fusion protein of the present invention with a gene-editing ribonucleoprotein (RNP) complex comprising an mRNA or a guide RNA.

[0253] A “guide RNA” is an RNA that recognizes a region of interest of RNA or DNA by binding to it. It forms stable complexes with a genome editing protein. These gene-editing ribonucleoprotein complexes are used to introduce deletions, insertions or otherwise alterations of the targeted RNA or DNA. They occur naturally, serving important functions, but can also be designed to be used for targeted editing, such as with CRISPR-Cas9, CRISPR-Casl2, or on RNA targets via Casl3.

[0254] As discussed above, the gRNA is a short synthetic RNA composed of a scaffold sequence necessary for Cas-binding and a nucleotide spacer, which has a length of about 20 nucleotides, that defines the genomic target to be modified. Thus, the genomic target of the Cas protein is defined by the target sequence present in the gRNA.

[0255] Preferably, the guide RNA is a sgRNA, pegRNA, or Cas 13.

[0256] In type II CRISPR / Cas system, single guide RNA (sgRNA) directs the target specific regions. Single guide RNAs result from the combination of two RNA molecules, the trans-activating RNA (tracrRNA) is responsible for Cas9 endonuclease activity while the CRISPR-RNA (crRNA) binds specifically to the DNA target region. Therefore, tracrRNA and crRNA are two key components and are joined at the tetraloop which results in the formation of sgRNA. The Cas9 endonuclease binds to the stem loop structure of the sgRNAThe spacer region usually consists of 18-20 base pairs and identifies the specific complementary target region, which is cleaved by Cas9 after its binding.

[0257] A “gRNA” as used herein refers to a single stranded RNA with nucleotide spacer and a scaffold, which contains a nucleotide nucleic acid binding domain (preferably PP7 or the Csy4 motif “C4”) within an exposed loop or at the 3 ’ end of the RNA.

[0258] A prime editing guide RNA (“pegRNA”) as used herein refers to a prime editing gRNA with extended 3’ end including a primer binding sequence (PBS) and a template for reverse transcription containing the desired RNA sequence. It directs the prime editor protein to the target locus and also encodes the desired edit.

[0259] Structured RNA motifs, such as tEvopreQl, added to the 3' terminus of pegRNAs can enhance their stability and prevent degradation of the 3' extension.

[0260] Ribonucleoproteins (RNP) include a family of proteins present in the nucleus and cytoplasm that bind RNA to guide to a target nucleic acid and introduce a double strand break (DSB)

[0261] The protein(s) that is / are part of the RNP comprise(s) gene-editing polypeptide(s) or protein(s).

[0262] The “nucleic acid editor ribonucleoprotein complex”, as used herein, including proteins from the CRISPR-Cas family, TALEs, pumilios, and de novo proteins, such as de novo DNA / RNA binding proteins, preferably a Cas protein such as Cas9 or Cas 13 bound to the multifunctional protein via its aptamer-tagged gRNA component such that binding during packaging and unbinding in the target cell is independent of the modifications of the protein component, e.g. with base editors, reverse transcriptase, epigenome or epitranscriptome editors.

[0263] In a preferred embodiment, the gene-editing polypeptide(s) or protein(s) is / are CRISPR / Cas9 effectors. According to the present invention, any nucleotide sequence coding for a specific amino acid sequence shall be included when referring to the specific amino acid sequence.

[0264] In one embodiment, the CRISPR / Cas9 effectors comprise: a CRISPR-associated endonuclease (Cas protein or Cas9 nuclease), a reverse transcriptase domain, such as the MLV RT, Base editor domains, such as TadA, and / or

[0265] Transactivating / repressing domains, such as VPR or KRAB.

[0266] In a preferred embodiment, the gene-editing polypeptide(s) or protein(s) comprise / s: a. double-strand break generation activity, b. nicking activity, or c. nuclease-dead binding activity towards target nucleic acids, and d. optionally a fusion with a nuclear localization signal (NLS) and / or nuclear export signal (NES), and e. optionally, a fusion towards an effector domain towards an RNA- or DNA-dependent DNA polymerase activity, a base editor including but not limited to adenine and cytidine deaminases, optionally additionally fused to one or more copies of uracil glycosylase inhibitors (UGI), a transcriptional activator or repressor, an epigenetic modifier including DNA- and histone- (de)methylases or (de)acetylases, generic DNA-nicking domains, including but not limited to FokI, and recruiters thereof.

[0267] In an embodiment, where the complex comprises a complex of at least one multi-functional fusion protein of the present invention with a gene-editing ribonucleoprotein (RNP) complex comprising an mRNA, the mRNA preferably comprises an aptamer-tag.

[0268] Nucleic acid delivery system

[0269] As discussed above, the present invention provides a nucleic acid delivery system.

[0270] The nucleic acid delivery system of the present invention comprises:

[0271] (A) at least one multi-functional fusion protein of the present invention;

[0272] (B) a gene-editing ribonucleoprotein (RNP) complex, which comprises

[0273] (Bl) an mRNA or a guide RNA, such as a sgRNA or pegRNA, and

[0274] (B2) gene-editing polypeptide(s) or protein(s),

[0275] (C) a fusogenic (glyco)protein or protein complex comprising at least one multi-functional protein or a set of at least two proteins, wherein the multifunctional domain or the set of at least two proteins comprise a receptor-targeting and a fusogenic domain to fuse the VLP to the target cell membrane, (D) a targeting domain, such as a glycoprotein or engineered variants thereof,

[0276] (E) a moiety mediating endosomal escape, such as a glycoprotein or engineered variants thereof,

[0277] (F) optionally, a reporter-signal generating gene product, such as a luciferase or a fluorescent protein, and

[0278] (G) optionally, a nucleic acid-stabilizing protein, preferably Csy4.

[0279] The nucleic acid delivery system of the present invention is preferably a virus-like particle (VLP).

[0280] In one embodiment, the fusogenic (glyco)protein or protein complex according to (C) is an envelope protein, or an engineered variant thereof.

[0281] -Multi-functional fusion protein(s) (A)

[0282] The nucleic acid delivery system of the present invention comprises at least one multi-functional fusion protein of the present invention.

[0283] In one embodiment, the nucleic acid-binding (NAB) domain is fused to Vicugna pacos (Alpaca) / Synthetic CANTDcbl, which binds to the capsid of HIV-1 Gag. The NAB-CANTDcbl fusion comprises the shuttling motif and is co-expressed with Gag-Pol.

[0284] In a preferred embodiment, the nucleic acid delivery system comprises two multi-functional fusion proteins of the present invention, which preferably comprises an amino acid sequence selected from any one of SEQ ID NOs: 26, 27, 56, 57 and 66.

[0285] In one embodiment, the nucleic acid editor complex (II) comprises a guide RNA, preferably a sgRNA or a pegRNA, which comprises the nucleic acid motif, bound by NAB domain (e) of (I) and an apoprotein of the nucleic acid editor complex, wherein the guide RNA is preferentially selected from a nucleic acid sequence according to any one of SEQ ID NOs: 160 to 169, while the apo-protein is preferentially selected from the SEQ ID NOs: 28 to 35.

[0286] In one embodiment, the nucleic acid editor complex (II) comprises a guide RNA, preferably a sgRNA or a pegRNA, which comprises a nucleic acid sequence according to any one of SEQ ID NOs: 325 to 392. The nucleic acid sequence according to any one of SEQ ID NOs: 325 to 392 refers to the region of the DNA that is translated into a peg / sgRNA, where the thymine (T) then becomes a uracil (U). Thus, the sequence according to any one of SEQ ID NOs: 325 to 392 refers to the respective nucleic acid sequence that is translated into a peg / sgRNA. In one embodiment, the nucleic acid editor complex (II) is a gene-editing ribonucleoprotein (RNP) complex, which comprises gene-editing polypeptide(s) or protein(s), preferably

[0287] CRISPR / Cas9 effectors, more preferably a CRISPR-associated endonuclease (Cas9 nuclease or Casl3),

[0288] TALEs, pumilio proteins, de novo DNA / RNA binding proteins, or mutants thereof, and wherein the gene-editing polypeptide(s) or protein(s) preferably further comprise a nuclear localization signal (NLS), preferably NLS of c-Myc protein (NLSc-Myc), NLS of SV40 (NLSsv4o) or a synthetic NLS, such as attenuated variants of NLSC-Myc and NLSsv4o, or other mono- or bipartite synthetic NLS, more preferably wherein the NLS comprises an amino acid sequence according to any one of SEQ ID NOs: 48 to 54, or a nuclear export signal (NES), preferably the NES of HIV- 1 (NESHIV-I) or DBR1, more preferably a NES comprising an amino acid sequence according to SEQ ID NO: 65, or combinations thereof.

[0289] - mRNA or ribonucleoprotein (RNP) complex (B)

[0290] The nucleic acid delivery system of the present invention comprises a gene-editing ribonucleoprotein (RNP) complex, which comprises

[0291] (Bl) an mRNA or a guide RNA, such as a sgRNA or pegRNA, and

[0292] (B2) gene-editing polypeptide(s) or protein(s),

[0293] - Nucleic acid (Bl)

[0294] In one embodiment, the nucleic acid is an mRNA.

[0295] In one embodiment, the nucleic acid is a guide RNA (gRNA), such as a sgRNA or a pegRNA, as discussed above.

[0296] The nucleotide sequence of the nucleic acid depends, for instance, on the intended base editing in target loci such as e.g. B2M.

[0297] As discussed above, the gRNA is a short synthetic RNA composed of a scaffold sequence necessary for Cas-binding and a nucleotide spacer or flexible guide region, which has a length of about 15 to 30 nucleotides, preferably of about 20 nucleotides, that defines the genomic target to be modified. Thus, the genomic target of the Cas protein is defined by the target sequence present in the gRNA.

[0298] In one embodiment, the nucleotide spacer or flexible guide region is a CRISPR spacer.

[0299] In one embodiment, the gRNA is a pegRNA.

[0300] Said specific nucleic acid comprises an RNA-motif which can be bound by the nucleic acid-binding domain (f) of the multi-functional fusion protein(s) (A). Said RNA-motif is preferably an aptamer.

[0301] Preferred examples are

[0302] PP7 coat protein aptamer,

[0303] MS2 coat protein aptamer, com operator (which is the structure of a com RNA hairpin),

[0304] BoxB aptamer system,

[0305] P22,

[0306] Csy4 aptamer, and

[0307] HIV-lpsi.

[0308] The RNA-motif is preferably the same as the RNA-motif of the gene editing ribonucleoprotein (RNP).

[0309] - Gene editing component(s) (B2)

[0310] The nucleic acid delivery system of the present invention comprises a gene-editing components, in particular gene-editing polypeptide(s) or protein(s) (B2).

[0311] As discussed above, the gene-editing polypeptide(s) or protein(s) is / are preferably CRISPR / Cas9 effectors.

[0312] In one embodiment, the CRISPR / Cas9 effectors comprise: a CRISPR-associated endonuclease (Cas protein or Cas9 nuclease), a reverse transcriptase domain, such as MLV-RT, a base editor domain, such as TadA, and / or an epigenetic modifier domain, such as VPR.

[0313] In a preferred embodiment, the gene-editing polypeptide(s) or protein(s) comprise / s: a. double-strand break generation activity, b. nicking activity, or c. nuclease-dead binding activity towards target nucleic acids, and d. optionally a fusion with a nuclear localization signal (NLS) and / or nuclear export signal (NES), and e. optionally, a fusion towards an effector domain towards an RNA- or DNA-dependent DNA polymerase activity, a base editor including but not limited to adenine and cytidine deaminases, optionally additionally fused to one or more copies of uracil glycosylase inhibitors (UGI), a transcriptional activator or repressor, an epigenetic modifier including DNA- and histone- (de)methylases or (de)acetylases, generic DNA-nicking domains including but not limited to FokI, and recruiters thereof.

[0314] The gene-editing component preferably contains an RNA-motif which can be bound by the nucleic acid-binding domain (f) of the multi-functional fusion protein(s) (A). Said RNA-motif is preferably an aptamer.

[0315] Preferred examples are:

[0316] PP7 coat protein aptamer,

[0317] MS2 coat protein aptamer, com operator (which is the structure of a com RNA hairpin),

[0318] BoxB aptamer system,

[0319] P22,

[0320] Csy4 aptamer.

[0321] The RNA-motif is preferably the same as the RNA motif of the specific nucleic acid (Bl).

[0322] - Viral envelope protein (C)

[0323] The nucleic acid delivery system of the present invention comprises a viral envelope protein or viral envelope glycoprotein (C).

[0324] In one embodiment, the nucleic acid delivery system of the present invention comprises engineered variants of said viral envelope protein.

[0325] In one embodiment, the viral envelope protein is Vesicular stomatitis virus G (VSV G) protein.

[0326] In one embodiment, the viral envelope protein is a variant of VSV-G with a mutated target recognition domain and an additional fusion to a receptor-binding moiety, such as nanobodies or anticalines.

[0327] - Targeting domain

[0328] The nucleic acid delivery system of the present invention comprises a targeting domain (D).

[0329] In one embodiment, the targeting domain (D) is a glycoprotein.

[0330] In one embodiment, the targeting domain (D) is an engineered variant of said glycoprotein.

[0331] Examples for a glycoprotein are VSV-G, MMLV-Env, FMLV-Env, TP MV H protein, SARS-CoV2 spike protein, Nipah virus G protein.

[0332] - Moiety mediating endosomal escape (E) The nucleic acid delivery system of the present invention comprises a moiety mediating endosomal escape (E).

[0333] In one embodiment, the moiety (E) is a glycoprotein.

[0334] In one embodiment, the moiety (E) is an engineered variant of said glycoprotein.

[0335] Examples for a glycoprotein are VSV-G, MMLV-Env, FMLV-Env, TP MV F protein, SARS-CoV2 spike protein, Nipah virus G protein.

[0336] - Reporter gene product (F) nucleic acid delivery system of the present invention optionally comprises a reporter gene product (F).

[0337] In one embodiment, the moiety (F) is a luciferase or a fluorescent protein.

[0338] The inventors disclose the development of a novel virus-like particle (VLP) system that offers significant advantages over existing delivery methods.

[0339] The inventor’s objective was to reduce viral components, particularly the proteins encoded on the Pol frame (protease, integrase, reverse transcriptase), to minimize the potential immunogenic load. It is thus an object of the present invention to provide a VLP system with a reduced immunogenic load.

[0340] In recent years, several VLP technologies with both lentiviral and retroviral origin have been developed for the delivery of genome editing systems. However, the VLP system disclosed herein has been optimized to achieve superior efficiency and effectiveness compared to existing systems. The VLP system of this invention is preferably based on a multi-functional fusion protein, preferably an engineered Gag protein fusion with an aptamer-binding protein, specifically the PP7 coat protein (PCP), which can bind to a PP7 RNA aptamer or the Csy4 protein binding to the C4 aptamer. To ensure efficient packaging of PP7-tagged sgRNAs or pegRNAs, the multi-functional fusion protein (such as the Gag-PCP fusion protein) also contains a nuclear localization signal (NLS) and a nuclear export signal (NES). This dual-signal strategy enables the multi-functional fusion protein to shuttle between the nucleus and the cytoplasm, ultimately ensuring efficient packaging and delivery of the genome editing complexes in the form of ribonucleoproteins (RNPs).

[0341] Many recent VLP designs based on the retro- / lentiviral gag polyprotein rely on the N-terminal fusion of a gene editing effector to the C-terminus of gag, utilizing an additional protease cleavage site for RNP release (Banskota etal., 2022, Hamilton et al., 2021, Mangeot et al. 2019, Haldrup et al. 2023), which can be transported into the nucleus of the receiver cell. While this is a very natural design for VLP -based protein delivery which even tackles the bottlenecks of cargo release and nuclear trafficking, it does not prevent the delivery of unloaded proteins which could reduce its effectiveness. Also, switching to other cargo proteins can impact packaging efficiency and might require reoptimization.

[0342] In contrast, the system of this invention disclosed herein directly captures the (pe)gRNA via the aptamer-binding protein making use of the inherent instability of (pe)gRNAs in the absence of Cas9, ensuring that only functional and active RNP complexes are packaged into the VLPs. With the added benefit of pseudotyping and tropism modulation, this innovative approach has the potential to greatly enhance the application of genome editing technologies in various fields of research and therapeutic development.

[0343] In one embodiment, the RNA species comprises an autonomous 3 ’-stabilization motif, preferably truncated evopreQl comprising a ribonucleic acid sequence according to SEQ ID NO: 171, and / or a motif comprising a ribonucleic acid sequence according to SEQ ID NO: 216 that is protected by an additional heterologously expressed nucleic acid-stabilizing domain ( / ) of (I), preferably C4 / Csy4 comprising an amino acid sequence according to SEQ ID NO: 64, the N-terminal domain of rotavirus NSP3 comprising an acid sequence according to SEQ ID NO: 323, wherein the N-terminal domain of rotavirus NSP3 optionally protects 3’-gacc RNA ends, or combinations thereof.

[0344] Polynucleotides and vectors

[0345] As discussed above, the present invention provides a plurality of nucleotides.

[0346] Said plurality of polynucleotides comprises:

[0347] (1) a first polynucleotide comprising a nucleic acid sequence encoding a multi-functional fusion protein of the present invention;

[0348] (2) a second polynucleotide comprising a nucleic acid sequence encoding a gene editor protein;

[0349] (3) a third polynucleotide comprising a nucleic acid sequence encoding a guide RNA, such as a sgRNA or pegRNA, which is bound via the nucleic acid-binding domain (e), preferably as ribonucleoprotein complex; and

[0350] (4) a fourth polynucleotide comprising a nucleic acid sequence encoding a viral envelope protein.

[0351] As discussed above, the present invention provides one or more vectors comprising the plurality of polynucleotides of the present invention.

[0352] In one embodiment, the polynucleotide encoding the multi-functional protein (I) comprises a nucleotide sequence selected from any one of SEQ ID NOs: 82 to 106. In one embodiment, the polynucleotide encoding the multi-functional protein (I) comprises a nucleotide sequence selected from any one of the SEQ ID NOs. as depicted below:

[0353] In one embodiment, the present invention relates to three vectors or plasmids.

[0354] As an example, see Figure 9.

[0355] A further aspect of this invention relates to a set of polynucleotides encoding the set of at least two proteins (I), preferably comprising the nucleotide sequence of SEQ ID NO: 107 (psPAXD64V (Gag-Pol) and a nucleotide sequence selected from any one of SEQ ID NOs. 82 to 106.

[0356] Pharmaceutical compositions

[0357] As discussed above, the present invention provides a pharmaceutical composition.

[0358] A pharmaceutical composition of the present invention comprises: the polynucleotide of the present invention which encodes the fusion protein of the present invention, the nucleic acid delivery system of the present invention, or the plurality of polynucleotides of the present invention; optionally, pharmaceutically acceptable excipients and / or carrier.

[0359] Producer cells, cell lines and kits

[0360] As discussed above, the present invention provides a cell or cell line. Said cell or cell line comprises the multi-functional fusion protein of the present invention, the polynucleotide of the present invention, the nucleic acid delivery system of the present invention, the plurality of polynucleotides of the present invention.

[0361] The cell or cell line is preferably a mammalian cell or cell line, more preferably a human cell or cell line. The cell or cell line preferably has high transfectability, including but not limited to HEK293T cells and other human immortalized cells / cell lines.

[0362] The cell or cell line of the present invention is preferably a producer cell or cell line, which produces the nucleic acid delivery system of the present invention, preferably a VLP.

[0363] In one embodiment, the VLP will be in the supernatant of such a producer cell / cell line.

[0364] In one embodiment, the cell or cell line can be co-treated with exogenous compounds, such as by means of lipofection, electroporation, photoporation, sonoporation, particle bombardment, microinjection, magnetofection, viral transduction, cell-penetrating peptides, calcium-phosphate, nanoparticles, or combinations thereof, wherein the exogenous compounds are preferably selected from synthetic nanoparticles, fluorophores, small pharmaceutical compounds, positron emission tomography (PET) tracers, peptides, proteins, lipids, co-factors, small biomolecules, and combinations thereof.

[0365] In one embodiment, the exogenous compounds are compounds are not produced by the packaging cell line. In one embodiment, the exogenous compounds are imported for co-packaging into the VLPs. In one embodiment, the packaging cell line (to a lesser degree a cell line implanted in vivo) can be cotreated with exogenous compounds even if this is toxic to the packaging cell line as long as the VLP- product is as desired. In one embodiment, the system is a ‘self-referential system’.

[0366] A further aspect of this invention relates to the use of a cell or cell line according to this invention for:

[0367] - ex vivo generating a therapeutic cell line, a diagnostic cell line, or a cell line for tissue engineering,

[0368] - generating a VLP-distributing cell line, which preferably can be implanted into a combination of cells, a tissue and / or organism to produce VLPs therein.

[0369] Importantly, the cell or cell line according to this invention can be used for generating cells, such as cells that can communicate with other cells to, e.g., support diagnostic and / or therapeutic applications.

[0370] In one embodiment, the cell or cell line according to this invention can be used for tissue engineering and / or for preclinical applications. As discussed above, the present invention provides kits.

[0371] A kit of the present invention comprises: the polynucleotide of the present invention which encodes the fusion protein of the present invention, or the nucleic acid delivery system of the present invention, or the plurality of polynucleotides of the present invention, or the one or more vectors of the present invention, or the cell or cell line of the present invention.

[0372] As discussed above, the present invention provides a nucleic acid delivery system, preferably a viruslike particle, produced by transfecting, transducing, electroporating, lipofecting, photoporating, sonoporating, particle bombardment, microinjecting, magnetofecting, or otherwise inserting the plurality of polynucleotides of the present invention or the one or more vectors of the present invention into a cell and expressing the components of the nucleic acid delivery system from the plurality of polynucleotides or the one or more vectors of the present invention in the cell, thereby allowing the nucleic acid delivery system, preferably the virus-like particle, to spontaneously assemble in the cell.

[0373] Medical uses

[0374] As discussed above, the present invention provides the nucleic acid delivery system of the present invention for use in medicine.

[0375] As discussed above, the present invention provides the nucleic acid delivery system of the present invention for use in the treatment of genetic diseases.

[0376] In one embodiment, the genetic disease is a monogenetic disease, such as cystic fibrosis, sickle cell anemia or Huntington's disease.

[0377] In one embodiment, the genetic disease is a polygenetic disease, such as diabetes, coronary artery disease or Alzheimer's disease.

[0378] Base editing methods

[0379] As discussed above, the present invention provides an in vitro or in vivo method of editing a nucleic acid molecule in a target cell.

[0380] In one embodiment, the in vitro or in vivo method of editing a nucleic acid molecule in a target cell occurs by nuclease mediated editing, base editing or prime editing, homology directed repair, non- homologous end-joining directed insertion, or microhomology-mediated end-joining directed insertion, such as by co-delivery of an integrase-deficient (D64V) lentivirus (IDLV), wherein, preferably, the IDLV is packaged using a psi-NC interaction, a PP7-PCP, MS2-MCP, and / or a C4-Csy4 interaction. In one embodiment, the in vitro or in vivo method of editing a nucleic acid molecule in a target cell occurs by nuclease mediated editing, base editing or prime editing, or homology directed repair, such as by codelivery of an integrase-deficient (D64V) lentivirus (IDLV), wherein, preferably, the IDLV is packaged using a psi-NC interaction, a PP7-PCP, MS2-MCP, and / or a C4-Csy4 interaction, and the method comprises the step of delivering Cas9.

[0381] Thus, a combination of VLP (which delivers, e.g., the Cas9 nuclease) and IDLV (which delivers, e.g., the template for homologous recombination) is used for an in vitro or in vivo method of editing a nucleic acid molecule in a target cell.

[0382] Said method comprises: contacting the target cell with a nucleic acid delivery system of the present invention, or with the pharmaceutical composition of the present invention, thereby installing one or more modifications to the nucleic acid at a target site.

[0383] Preferably, the target cell is a mammalian cell, more preferably a human cell.

[0384] In one embodiment, the target cell is in a subject, preferably a mammal, more preferably a human.

[0385] Preferably, the one or more modifications to the nucleic acid are associated with reducing, relieving or preventing symptoms of a disease or disorder.

[0386] In one embodiment, the combination of VLP (which delivers, e.g., the Cas9 nuclease) and IDLV (which delivers the template for recombination) comprises a homology directed repair (HDR)-based template, a non-homologous end-joining (NHEJ) directed insertion-based template, a REPLACE-template or a microhomology-mediated end-joining (MMEJ) directed insertion-template, such as PITCh.

[0387] Non-homologous end-joining directed insertion (NHEJ)-based templates have been described in Suzuki et al (2016), which shall be incorporated herein by reference.

[0388] REPLACE-based templates have been described in Danner et al. (2021), which shall be incorporated herein by reference.

[0389] Microhomology-mediated end-joining (MMEJ) directed insertion-template have been described in Nakade et al. (2014), which shall be incorporated herein by reference. Method for nucleic acid delivery

[0390] As discussed above, the present invention provides a method for nucleic acid delivery, preferably geneediting nucleic acids.

[0391] Said method comprises the steps of:

[0392] (I) providing a multi-functional fusion protein of the present invention to a cell, wherein said cell produces in its nucleus a gene-editing ribonucleoprotein (RNP) complex, which comprises a genome editor protein and a guide RNA, such as an aptamer-tagged sgRNA or pegRNA, wherein said gene-editing ribonucleoprotein (RNP) complex is as defined herein,

[0393] (II) allowing the multi-functional fusion protein to locate to the nucleus of said cell;

[0394] (III) forming a complex of the multi-functional fusion protein with the gene-editing ribonucleoprotein (RNP) complex in the nucleus;

[0395] (IV) allowing the complex of (III) to be exported to the cytosol of the cell;

[0396] (V) budding of virus-like particles comprising the gene-editing ribonucleoprotein (RNP) complex;

[0397] (VI) binding of the VLPs to a target cell surface, preferably via a specific receptor,

[0398] (VII) internalization of the VLPs into the target cell, preferably via endocytosis,

[0399] (VIII) transport of the internalized complex from (VII) into the cytosol, such as via endosomal escape,

[0400] (IX) dissociation of the RNP (with bound nucleic acid) or a nucleic acid from the nucleic acid binding domain of (I),

[0401] (X) optionally, transport of the RNP into the nucleus for genome editing or transactivation,

[0402] (XI) optionally, binding of the guide-RNA-editing complex to an RNA target sequence.

[0403] Preferred embodiments

[0404] - Abstract

[0405] Recent gene editing technologies, such as prime editors (PEs) and base editors (BEs), have shown great potential for the cure of genetic diseases. However, efficient and safe delivery of these genome editing complexes still poses a major challenge. After careful evaluation of existing delivery systems, the inventors developed a system for the transgene-free delivery of functional ribonucleoprotein (RNP) complexes that address current bottlenecks of existing systems such as efficient (pe)gRNA packaging. The highly optimized system comprises a chimeric fusion of Gag and the PP7 coat protein (PCP), which binds to a PP7 aptamer grafted into the (pe)gRNA scaffold, and thus allows for modular packaging of any RNP configuration. With the delivery system of the present invention, the inventors demonstrate the delivery of functional RNP complexes for prime editing and other CRISPR effectors in various cell types, including cortical neurons, T cells, and induced pluripotent stem cells. Finally, the inventors utilize a minimal-Gag variant with deletions in MA, CA, and NC, which shows comparable efficacies and is not reliant on the co-expression of HIV-1 wild-type gag-pol while offering improved efficacy and safety for delivery of genome editing complexes and thereby providing a valuable resource for laboratory implementation with additional potential to advance the somatic therapeutic application of gene editing technologies.

[0406] - Results

[0407] HIV-l-based VLPs for mRNA delivery.

[0408] Initial results with the eVLP system (Banskota et al.. 2022) showed highly effective delivery of base editors, but achieved modest targeting efficacy with an analogously designed eVLP architecture-based prime editor (PE). Motivated by this observation, the inventors sought to systematically decipher the bottlenecks of such a delivery system using the extensively studied HIV-l-based VLP system.

[0409] According to various studies (Wang et al., 2019; Ma etal., 2016; Hendel et al., 2015), the half-lives of sgRNAs are magnitudes shorter compared to sgRNAs embedded in a ribonucleoprotein (RNP) complex, leading to the conclusion that especially pegRNAs, which carry a 3 ’-extended region encoding the desired edit and a primer binding site (PBS), are particularly prone to exonucleolytic degradation. The inventors reasoned that having the recruitment mechanism depend on the least stable component, that is the (pe)gRNA, of the complex would increase the likelihood of packaging functional RNPs exclusively.

[0410] The inventors based their system on the Gag protein of the Human Immunodeficiency Virus type 1 (HIV-1), which is a 55 kDa precursor responsible for assembly, budding, and maturation with the Gag-Pol protein possessing additional enzymatic functions, including reverse transcriptase and integrase activities, crucial for the virus lifecycle.

[0411] The inventors modified the first zinc finger motif within the NC domain by mutating all cysteine residues to serine to prevent unspecific RNA binding and replaced the second zinc finger motif by a non-oligomerizing mutant of the aptamer-binding domain PP7 coat protein (PCP) to bind and recruit RNA molecules tagged with the PP7 aptamer (Figure IB). In addition, the inventors introduced L21S in the MA domain of Gag which has been shown previously to enhance the budding efficacy (Sherer et al., 2009)). The inventors utilized this architecture previously to secrete barcoded introns out of a cell for non-invasive transcript quantification (Truong et al., 2022) a concept that has recently gained strong interest (Homs et al., 2023).

[0412] VLPs containing the mRNA information can be delivered to other cells when equipped with a suitable surface glycoprotein such as VSV-G (Lu et al., 2019). The inventors thus co-expressed VSV-G and mGreenLantem (mGL) with a PP7 aptamer in its 3’UTR to determine essential elements required for efficient cargo packaging, budding, or target cell entry (Figure 2A). Since the proteolytic processing of VLPs is assumed to be as critical for VLP transduction efficacy as it is tightly regulated, the most efficient VLP delivery systems rely on mosaic setups with the protease activity provided by lentiviral Gag-pol plasmids like psPAX2 with an inactivated integrase domain (D64V). Consistent with these findings, the inventors also observed that the HIV-1 protease was essential for delivery of PP7-tagged mRNA coding for mGL (mGLi i ?). as the HIV-1 protease inhibitor darunavir significantly decreased delivery efficiency when supplied during production (Figure 2B).

[0413] The inventors analyzed the impact of the L21S mutation, which is most beneficial when present in only half of the particle-forming on the Gag that has the PCP grafted into the ZF2 of the NC domain (Figure 2C); this combination is superior, especially compared to directly grafting PCP into the NC domain of Gag-Pol (Figure 2D). This was not surprising, since modifications of the NC domain, more specifically, the zinc finger motifs ZF1 and ZF2, often result in detrimental effects on the assembly of particles (Grigorov et al., 2007, Boutant et al., 2020). Thus, introduction of L21S likely compensates for the allosteric switch mediated naturally by the NC domain after binding to its genomic RNA. The inventors next set out to assess which fraction of green fluorescence of the transduced cells originated from mRNA delivery rather than protein delivery. Using the CS 1 aptamer as a negative control for PCP-specific loading of mGL mRNA, the inventors found that still, almost all cells showed green fluorescence, even though the fluorescence intensity was significantly reduced. Even co-transfection of microRNA (miR) targeting mGL did not completely abolish fluorescence, indicating that a substantial amount of mGL protein was co-transduced into the target cells (Figure 2E).

[0414] To provide a calibration for the performance of the mosaic Gag-Pol: Gag -PCP system against recently published experiments with de novo protein cages, the inventors compared EPN-24-PCP nanocages with a C-terminal fusion of PCP or single-chain tandem-PCP and found its fluorescence intensity to be roughly on the same level as the non-chimeric Gag-PCP without Gag-Pol, but 8-fold lower than the co-expression of Gag-PCP and the integrase-deficient Gag-Pol (Figure 2F).

[0415] Nuclear-cytosolic shuttling of Gag enables efficient delivery of various genome editors. While Cas9 delivery encoded as mRNA that needs to be translated in the recipient cell is straightforward, delivering (pe)gRNAs is more challenging. Because (pe)gRNAs do not naturally reside in the cytosol, the inventors reasoned that the VLP system benefits from an active relocalization of (pe)gRNAs to the cytosol. The inventors consequently introduced a nuclear localization sequence (NLS) and a nuclear export sequence (NES) to Gag-PCP in order to shuttle (pe)gRNAs out of the nucleus (Figure IB and 3A). For (pe)gRNA packaging, the inventors utilized a previous design from the inventors where the PP7 aptamer was grafted into the tetraloop (junction between tracrRNA and crRNA) of the SpCas9 gRNA / pegRNA scaffold (Figure 3A).

[0416] To test the VLP system for the delivery of functional prime editors, the inventors used a dedicated fluorescent reporter system in which mutations in the chromophore of mGreenLantem (G65S and Y66H) are corrected to shift the fluorescence profile from blue back to green (Figure 3B). Firstly, the inventors determined which individual component of the prime editing complex (pegRNA + prime editor) was limiting. To this end, the inventors provided one component via plasmid transfection while delivering the other component as PP7-tagged RNA via the VLP system of this invention. The results indicate that the pegRNA was delivered very effectively, while the delivery of prime editor (PE) mRNA was limiting, which lead to the assumption that packaging of the long mRNA encoding the prime editor might be the bottleneck (Figure 3C).

[0417] Immunofluorescence imaging revealed that the steady-state localization of the engineered shuttling Gag was in the cytosol as long as an NES motif was present (Figure 18A-C). The inventors used leptomycin B (LMB), a specific nuclear export inhibitor, to demonstrate the efficacy of the inventor’s ENVLPE system in harnessing nucleocytosolic shuttling as a process for efficient packaging. The inventors tracked the localization of FLAG-tagged Gag-PCP chimeras containing the shuttling motif, which initially had a strong steady-state presence in the cytosolic compartment; however, upon the addition of LMB, the signal accumulated in the nucleus within just one hour, indicating a rapid shuttling process. Thus, nuclear accumulation of Gag-PCP was evident as early as one hour after the addition of leptomycin B. This observation suggests that Gag-PCP is continuously shuttled between the nucleus and cytosol (Figure 18C). In addition, the inventors have included multiple other nucleocytosolic shuttling controls and variants (Figure 18A-C).

[0418] Table 1: Nucleocytosolic shuttling controls and variants used in Figure 18

[0419] SEQ

[0420] Description ID Amino Acid Sequence

[0421] NO:

[0422] MGARASVLSGGELDRWEKIRSRPGGKKKYKLKHIVWASRELERFAVN

[0423] PGLLETSEGCRQILGQLQPSLQTGSEELRSLYNTVATLYCVHQRIEIKDT

[0424] KEALDKIEEEQNKSKKKAQQAAADTGHSNQVSQNYPIVQNIQGQMVH

[0425] QAISPRTLNAWVKVVEEKAFSPEVIPMFSALSEGATPQDLNTMLNTVG

[0426] GHQAAMQMLKETINEEAAEWDRVHPVHAGPIAPGQMREPRGSDIAGT

[0427] Gag-L21S- TSTLQEQIGWMTNNPPIPVGEIYKRWIILGLNKIVRMYSPTSILDIRQGP

[0428] PCP KEPFRDYVDRFYKTLRAEQASQEVKNWMTETLLVQNANPDCKTILKA comprising 502 LGPAATLEEMMTACQGVGGPGHKARVLAEAMSQVTNSATIMMQRGN

[0429] NLS-NES, FRNQRKIVKSFNSGKEGHTARNSRAPRKKGGSGGVSGWRLFKKISGGS full sequence GPAAKRVKLDSGLQLPPLERLTLGAAPAAKKKKLDYKDDDDKGGASS

[0430] KTIVLSVGEATRTLTEIQSTADRQIFEEKVGPLVGRLRLTASLRQNGAK

[0431] TAYRVNLKLDQADVVDSGLPKVRYTQVWSHDVTIVANSTEASRKSLY

[0432] DLTKSLVATSQVEDLVVNLVPLGRGGSGGTERQANFLGKIWPSYKGRP

[0433] GNFLQSRPEPTAPPEESFRSGVETTTPPQKQEPIDKELYPLTSLRSLFGN

[0434] DPSSQ

[0435] MGARASVLSGGELDRWEKIRSRPGGKKKYKLKHIVWASRELERFAVN

[0436] PGLLETSEGCRQILGQLQPSLQTGSEELRSLYNTVATLYCVHQRIEIKDT

[0437] Gag-L21S-

[0438] KEALDKIEEEQNKSKKKAQQAAADTGHSNQVSQNYPIVQNIQGQMVH

[0439] PCP

[0440] QAISPRTLNAWVKVVEEKAFSPEVIPMFSALSEGATPQDLNTMLNTVG comprising

[0441] 503 GHQAAMQMLKETINEEAAEWDRVHPVHAGPIAPGQMREPRGSDIAGT

[0442] ANLS-

[0443] TSTLQEQIGWMTNNPPIPVGEIYKRWIILGLNKIVRMYSPTSILDIRQGP

[0444] ANES, full

[0445] KEPFRDYVDRFYKTLRAEQASQEVKNWMTETLLVQNANPDCKTILKA sequence

[0446] LGPAATLEEMMTACQGVGGPGHKARVLAEAMSQVTNSATIMMQRGN

[0447] FRNQRKIVKSFNSGKEGHTARNSRAPRKKGGSGGVSGWRLFKKISGGA SSKTIVLSVGEATRTLTEIQSTADRQIFEEKVGPLVGRLRLTASLRQNGA KTAYRVNLKLDQADVVDSGLPKVRYTQVWSHDVTIVANSTEASRKSL YDLTKSLVATSQVEDLVVNLVPLGRGGSGGTERQANFLGKIWPSYKG RPGNFLQSRPEPTAPPEESFRSGVETTTPPQKQEPIDKELYPLTSLRSLFG NDPSSQ MGARASVLSGGELDRWEKIRSRPGGKKKYKLKHIVWASRELERFAVN PGLLETSEGCRQILGQLQPSLQTGSEELRSLYNTVATLYCVHQRIEIKDT KEALDKIEEEQNKSKKKAQQAAADTGHSNQVSQNYPIVQNIQGQMVH QAISPRTLNAWVKVVEEKAFSPEVIPMFSALSEGATPQDLNTMLNTVG GHQAAMQMLKETINEEAAEWDRVHPVHAGPIAPGQMREPRGSDIAGT

[0448] Gag-L21S- TSTLQEQIGWMTNNPPIPVGEIYKRWIILGLNKIVRMYSPTSILDIRQGP

[0449] PCP KEPFRDYVDRFYKTLRAEQASQEVKNWMTETLLVQNANPDCKTILKA comprising 504 LGPAATLEEMMTACQGVGGPGHKARVLAEAMSQVTNSATIMMQRGN

[0450] NLS-ANES, FRNQRKIVKSFNSGKEGHTARNSRAPRKKGGSGGVSGWRLFKKISGGS foil sequence GPAAKRVKLDSGGGSGGGSGGGSGAAPAAKKKKLDYKDDDDKGGAS SKTIVLSVGEATRTLTEIQSTADRQIFEEKVGPLVGRLRLTASLRQNGA KTAYRVNLKLDQADVVDSGLPKVRYTQVWSHDVTIVANSTEASRKSL YDLTKSLVATSQVEDLVVNLVPLGRGGSGGTERQANFLGKIWPSYKG RPGNFLQSRPEPTAPPEESFRSGVETTTPPQKQEPIDKELYPLTSLRSLFG NDPSSQ MGARASVLSGGELDRWEKIRSRPGGKKKYKLKHIVWASRELERFAVN PGLLETSEGCRQILGQLQPSLQTGSEELRSLYNTVATLYCVHQRIEIKDT KEALDKIEEEQNKSKKKAQQAAADTGHSNQVSQNYPIVQNIQGQMVH QAISPRTLNAWVKVVEEKAFSPEVIPMFSALSEGATPQDLNTMLNTVG GHQAAMQMLKETINEEAAEWDRVHPVHAGPIAPGQMREPRGSDIAGT

[0451] Gag-L21S- TSTLQEQIGWMTNNPPIPVGEIYKRWIILGLNKIVRMYSPTSILDIRQGP

[0452] PCP KEPFRDYVDRFYKTLRAEQASQEVKNWMTETLLVQNANPDCKTILKA comprising 505 LGPAATLEEMMTACQGVGGPGHKARVLAEAMSQVTNSATIMMQRGN

[0453] ANLS-NES, FRNQRKIVKSFNSGKEGHTARNSRAPRKKGGSGGVSGWRLFKKISGGS foil sequence GPAAGSVGLDSGLQLPPLERLTLGAAPAAGSGGLDYKDDDDKGGASS KTIVLSVGEATRTLTEIQSTADRQIFEEKVGPLVGRLRLTASLRQNGAK TAYRVNLKLDQADVVDSGLPKVRYTQVWSHDVTIVANSTEASRKSLY DLTKSLVATSQVEDLVVNLVPLGRGGSGGTERQANFLGKIWPSYKGRP GNFLQSRPEPTAPPEESFRSGVETTTPPQKQEPIDKELYPLTSLRSLFGN DPSSQ MGARASVLSGGELDRWEKIRSRPGGKKKYKLKHIVWASRELERFAVN PGLLETSEGCRQILGQLQPSLQTGSEELRSLYNTVATLYCVHQRIEIKDT KEALDKIEEEQNKSKKKAQQAAADTGHSNQVSQNYPIVQNIQGQMVH QAISPRTLNAWVKVVEEKAFSPEVIPMFSALSEGATPQDLNTMLNTVG GHQAAMQMLKETINEEAAEWDRVHPVHAGPIAPGQMREPRGSDIAGT

[0454] Gag-L21S- TSTLQEQIGWMTNNPPIPVGEIYKRWIILGLNKIVRMYSPTSILDIRQGP

[0455] PCP KEPFRDYVDRFYKTLRAEQASQEVKNWMTETLLVQNANPDCKTILKA comprising

[0456] 506 LGPAATLEEMMTACQGVGGPGHKARVLAEAMSQVTNSATIMMQRGN superNLS- FRNQRKIVKSFNSGKEGHTARNSRAPRKKGGSGGVSGWRLFKKISGGS

[0457] NES, foil GPAAKRVKLDGSPPKKKRKVEDSGLQLPPLERLTLGAAPAAKKKKLD sequence YKDDDDKGGASSKTIVLSVGEATRTLTEIQSTADRQIFEEKVGPLVGRL RLTASLRQNGAKTAYRVNLKLDQADVVDSGLPKVRYTQVWSHDVTI VANSTEASRKSLYDLTKSLVATSQVEDLVVNLVPLGRGGSGGTERQA NFLGKIWPSYKGRPGNFLQSRPEPTAPPEESFRSGVETTTPPQKQEPIDK ELYPLTSLRSLFGNDPSSQ

[0458] Gag-L21S-

[0459] PCP comprising

[0460] 507 GSGPAAKRVKLDSGLQLPPLERLTLGAAPAAKKKKLDYKDDDDKGG NLS-NES, partial sequence Gag-L21S-

[0461] PCP comprising

[0462] ANLS- 508 GSGPAAGSVGLDSGGGSGGGSGGGSGAAPAAGSGGLDYKDDDDKGG

[0463] ANES, partial sequence

[0464] Gag-L21 S-

[0465] PCP comprising GSGPAAKRVKLDSGGGSGGGSGGGSGAAPAAKKKKLDYKDDDDKGG

[0466] NLS-ANES, partial sequence

[0467] Gag-L21S-

[0468] PCP comprising

[0469] GSGPAAGSVGLDSGLQLPPLERLTLGAAPAAGSGGLDYKDDDDKGG

[0470] ANLS-NES, partial sequence

[0471] Gag-L21S-

[0472] PCP comprising , , , GSGPAAKRVKLDGSPPKKKRKVEDSEGSGLQLPPLERLTLGAAPAAKK superNLS- KKLDYKDDDDKGG

[0473] NES, partial sequence

[0474] Figure 18 clearly demonstrates that the inventor’s Gag-PCP chimera is indeed a nucleocytosolic shuttling protein. From Figure 18B it is evident that both the NLS and NES motifs are required for an optimal ENVLPE performance and that the VLP system of this invention is the best-performing variant. Accordingly, both shuttling motifs (NLS and NES) are required for an ideal ENVLPE performance. Figure 18C shows that 1 h after adding a nuclear export inhibitor, leptomycin B (LMB), FLAG-tagged Gag-PCP chimeras started accumulating in the nucleus. The immunofluorescence and inhibitor studies used in Figure 18 substantiate the shuttling mechanism's impact on efficient packaging and delivery and proves that Gag-PCP is actively shuttling between the nuclear and cytosolic compartments.

[0475] Genome editors are delivered as pre-assembled ribonucleoproteins (RNPs).

[0476] Since any cargo mRNA is already translated in the VLP-producing cell, the inventors wanted to ascertain if the respective components were co-delivered as RNA molecules or as pre-assembled ribonucleoproteins (RNP). Others have demonstrated previously that aptamer-engrafted gRNAs can facilitate the recruitment of RNPs into VLPs (Lyu et al., 2019). To this end, the inventors tested if the separate packaging of each component before co-transduction would abrogate the editing efficacy by preventing any RNP assembly in VLP-producing cells. The inventors observed a substantial reduction in editing events, which led the inventors to believe that the efficacy the inventors initially observed originated from RNP delivery. Interestingly, the PP7 tag had little to no influence on efficacy during separated packaging of the iPE mRNA, suggesting that the residual editing activity originates from unspecific packaging of either the mRNA or the iPE protein. The inventors proceeded to co-package the PP7-tagged pegRNA with the PP7-tagged iPE mRNA, achieving just above 50% editing in a HEK293T PE reporter cell line. The inventors validated the importance of all key components of the delivery system, particularly that of the NLS-NES module for (pe)gRNA shuttling, which resulted in a 6.4-fold increase in editing.

[0477] To determine any proportion of potential iPE mRNA delivery, the inventors transfected the reporter cell line with a microRNA (miR) targeting the 3 ’ UTR of the iPE mRNA, prior to VLP delivery As the editing efficacy in the recipient cells was not reduced at all in the presence of miRs, the inventors concluded that editing events in the reporter cell line were exclusively mediated by RNP delivery and not by de novo translation of iPE mRNA and subsequent assembly with the co-delivered pegRNA. The same miR was pre-validated in an previous experiment, where it significantly reduced the translation of VLP -delivered mGreenLantem mRNA.

[0478] With the aim of developing a one-component packaging system, the inventors grafted the PP7-coat protein into the NC domain of Gag-pol and were able to deliver functional prime editing RNPs into target cells to mediate prime editing events, albeit with 3 -fold reduced efficiency, when compared to the mosaic set-up using Gag-pol and Gag-PCP.

[0479] The inventors have thus established and validated a model for prime editor packaging and delivery improved by key modifications, where a PP7-tagged (pe)gRNA initially forms an RNP with a nucleus- localized genome editor protein to then be shuttled out of the nucleus by a modified VLP-forming gag unit. This system prevents the potential packaging of ‘empty’ prime editor proteins, while also providing higher stability for the gRNA prior to nuclear export.

[0480] Optimization of the RNP delivery system.

[0481] After establishing the system for delivery for genome editors, the inventors set out to optimize packaging stoichiometries as well as the Gag-PCP construct. Intriguingly, observations from mRNA delivery did not seamlessly translate to RNP delivery. For example, while the unsuccessful attempts to entirely omit Gag-pol2o64v indicated that certain factors of gag-pol are crucial for efficient mRNA delivery, the cost of omitting Gag-polo64v was much smaller when delivering RNPs (Figures 2F and 4A). The inventors more closely investigated the components provided by Gag-polo64v and found small individual effects for each component. Intriguingly, the negative impact of protease inhibition or knockout on RNP delivery was quite small compared to mRNA delivery but also occurs with posttransduction treatment (Figures 2B and 4B).

[0482] The inventors lastly optimized the amount of the VSV-G for pseudotyping (Figure 4C) as well as the stoichiometry of all delivery components (Figures 4D and 4E). The inventors specifically found that a slight excess (2: 1) of cargo over VLP gives optimal results. Regarding the proportion of each budding component, Gag-polo64v gave optimal results when supplied in slight deficit in relation to the inventors' engineered Gag. As a ratio of 1: 1 was also close to optimal, the inventors decided to utilize this ratio to construct a combined plasmid where both components are expressed at identical strength (Gag-polo64v / GagNLSNEs AZF2+L2is-PCP). The optimal ratio for RNP formation appeared to be a 3 -fold excess of pegRNA over the prime editor, which the inventors assumed to be transferrable to other Cas9 effector strategies as well.

[0483] Application of the modular delivery system for delivery of base editors and other Cas9 effectors. Since packaging occurs via the handle of the (pe)gRNA, the inventors naturally sought to generalize to the transduction of other Cas9 effectors, which is useable without any modifications. The inventors initially tested the delivery of active Cas9 and adenine base editors (ABEs) using a fluorescent reporter system, which can detect mutagenic end joining (mutEJ) events (InDeis) as well as certain adenine base editing (BE) events. Specifically, when the stop codon is deleted, either via mutEJ or BE, one of the green fluorescence protein eUnaG in one of the three downstream frames will be activated (depending on InDei size) (Figure 5A). Upon RNP delivery to this stable reporter cell line, the inventors observed an InDei rate of above 60% in HEK293T cells and 12% in hiPSCs (Figure 5B).

[0484] The inventors confirmed that as with prime editing, all components are essential to mediate any level of editing at all, and that the additional introduction of the shuttling motif increases the editing efficacy by up to 5-fold (Figure 5B).

[0485] The inventors next wanted to apply the delivery system of the present invention for base editing of a reference target site in the B2M locus, comparing the efficacy between various types of cell lines (Figure 5C). The inventors found that the present delivery system was able to facilitate the corresponding base edit to a high proportion in all tested cell lines, reaching an efficiency of 90% in cortical neurons and, at the lower end, still above 55% in hiPSCs. We, therefore, demonstrate the capability of mediating gene editing even in cell lines with potential clinical relevance, which are typically challenging to genetically modify using conventional techniques such as lipofection.

[0486] To also utilize the system of this invention for CRISPR activation, the inventors delivered RNPs consisting of dCas9-VPR and a gRNA targeting the MAPT transcription start (= MAPT promoter) site to a reporter cell line, in which MAPT expression can be followed by firefly luciferase (FLuc) signal. The inventors observed a marked increase in bioluminescence signal after 48 hours (Figure 5D).

[0487] Comparison of the delivery system of the present invention to other VLP delivery systems.

[0488] The inventors next sought to compare the present invention to a recently published and widely recognized MMLV-based delivery system from Banskota et al. (2022). They introduced “eVLPs”, which build on previous iterations of gag -mediated protein delivery (Mangeot et al., 2019; Hamilton et al., 2021) to achieve the transfer of Cas9 or base editors into target cells via direct fusion of a Cas9 effector to an engineered MLV-gag protein. Consequently, the VLP design and cargo loading process is quite different from eVLPs in a couple of critical aspects (see Figures 6A and 6B). As mentioned before, in the present invention the RNP complex is recruited non-covalently via an aptamer handle on the (pe)gRNA (Figure 6A). In contrast, with eVLPs the RNP cargo is covalently linked to Gag, and the release in the target cell is facilitated via additional cleavage sites for the MMLV protease (Figure 6B). The inventors believe that this key differentiator provides an advantage in terms of efficacy but also for modularity and convenience:

[0489] (1) (pe)gRNA stability likely benefits from assembly into RNPs in the nucleus already, that is prior to re-localization to the cytosol via a shuttling component.

[0490] (2) having the “loading handle” located on the (pe)gRNA circumvents packaging of “empty” Cas9 effector proteins while the packaging of (pe)gRNAs in complex with an RNP is favored naturally due to the relative short half-life of unbound gRNAs (Wang et al.. 2019; Ma et al., 2016; Hendel etal., 2015).

[0491] (3) Thirdly, cargo release is independent of any proteolytic cleavage, omitting one kinetic bottleneck entirely. Being independent of the protease activity from Gag-pol also offers the chance of developing and further modifying non-mosaic VLPs. In terms of convenience and modularity, the utilization of Gag-fusion constructs requires modifications to the Gag polyprotein itself, which could affect VLP assembly and may require re-optimization for new cargos.

[0492] In contrast, the delivery system of the present invention can be easily repurposed for any (pe)gRNA- guided strategy without changes to the cargo or packaging constructs beyond a designated (pe)gRNA scaffold containing the PP7 aptamer (as demonstrated in Figure 6A).

[0493] The inventors initially compared the present invention to eVLPs regarding the delivery of ABEs to HEK293 cells carrying the eTLR reporter, which reported an approximate 3 -fold increase in editing for the present invention over the BE-eVLP system (Figure 6C). For a more detailed comparison, the inventors next set up titration experiments to compare both systems for ABE8e and Cas9 delivery to target the B2M locus in HEK293T and hiPSC, quantifying base edits / InDels viaNGS (Figures 6D and 6E). While in HEK293T InDei generation saturates at very low VLP amounts for both systems with comparable performances (Figure 6E), the performance of the present invention in base editing of B2M is superior over eVLPs by achieving similar editing efficacy with roughly a magnitude lower volume (Figure 6D). eVLP base editing does not reach saturation within the tested range, it is therefore possible that it matches the efficacy of the present invention if much higher VLP doses are used. Surprisingly, iPSC editing remains at a moderate level regardless of the delivery system or edit type, only marginally (ABE) or not at all responding (active Cas9) to changes in the applied VLP volume (Figures 6D and 6E).

[0494] The inventors also attempted a comparison between both systems for PE delivery, even though Banskota et al. (2022) did not yet report the delivery of prime editors. As there was no eVLP available for prime editing as of yet, the inventors used the eVLP plasmids provided via Addgene to construct an eVLP-like derivate for PE by exchanging the Cas9 effector domain for the iPE domain within the Gag-Cas9 fusion protein. When tested on the PE reporter from Figure 3A, the inventors observed that it could also mediate prime edits, though not to the extent of the present invention (Figure 6F).

[0495] Owing to the complicated cargo release mechanism in eVLPs, the inventors wondered if non-covalent cargo release of the present invention, could still be improved further. The inventors reasoned that if cargo release still poses a bottleneck, the transport of RNPs to the nucleus as a PCP-RNP complex would be equally viable. Assuming that in the chimeric assembly setup the gag-PCP polyprotein is also proteolytically processed just like in eVLPs, a design where the NES is repositioned further downstream to the C-terminus (so that the PCP domain only contains NLS motifs after processing) would utilize both proteolytic cleavage and non-covalent recruitment for cargo release. However, this design did not improve editing efficacy, indicating that cargo release from the non-covalent interaction between PCP and the PP7 aptamer is already sufficient in the recipient cell and does not require an additional proteolytic release (Figure 6G). Nevertheless, other ABP / aptamer pairs with much higher affinity, efficiency of cargo release might benefit from the protease-mediated release.

[0496] Due to the exposed nature of the 3 ’-extension of a pegRNA, comprising the reverse transcribed template (RTT) and the primer binding site (PBS), Nelson et al. developed multiple 3’ protection motifs using nature-derived RNA pseudoknots (e.g., (t)evopreQl and tmpknot, Nelson et al., 2022). Albeit the 3’- stability of pegRNAs is extended by such motifs, the inventors think that during its voyage across the producer cell’s nucleus and cytosol to the budding location of the VLPs at the plasma membrane, and vice versa back in the recipient cell from the endosomal membrane across to cytosol to its destination site in the nucleus would benefit from design considerations that only package RNPs with intact 3 ’-end, i.e., full-length RTT and PBS. In the current pegRNA design, RNPs with non-fiinctional 3 ’-truncated pegRNAs could still be packaged and delivered. To prevent this, the inventors investigated moving the PP7 aptamer from the tetraloop position within the gRNA scaffold to the 3’ end of the pegRNA downstream of the PBS but upstream of the (t)evopreQl motif, effectively restricting packaging of only pegRNAs with an intact 3’ RTT and PBS region. Indeed, this resulted in a further increase in editing efficacy (Figure 6F), which now also constitutes the final design for PE delivery of this invention. The inventors did not translate this design to gRNA-based strategies, i.e., nuclease- and base-editing-based genome editing techniques, which natively do not have 3 ’-extensions. 3 ’-extensions are usually accompanied by a major decrease in Cas9 activity, while tetraloop modifications, as well as insertions into the first or second stem -loop of the tracrRNA (2ndor 3rdstem-loop of the gRNA, when counting the tetraloop as the 1ststem -loop and the nexus not as stem -loop), are generally well tolerated with only a minor decrease in activity for some targets. Since pegRNAs already have 3 ’-extensions, insertion of an additional aptamer is also not expected to be detrimental to the activity of prime editing RNPs. Consequently, stem-loop modifications, especially the tetraloop and the stem-loops of the tracrRNA, are better suited for gRNAs to prevent unnecessary loss in Cas9 activity. So far, the inventors inserted the nuclear export and import sequences (NLS-NES) within the NC domain together with the aptamer-binding protein (ABP). As a consequence, maturation of the VLPs via the maturase / protease domain leads to the release of a modified NC domain comprising the ABP and the NLS-NES, where the NES might lead to a less efficient import of the genome editing RNP complex if the ABP-NLS-NES containing NC domain is still tethered via the ABP-aptamer pair, e.g., PCP-PP7 interaction, to the RNP. Thus, the inventors moved the NES portion towards the C-terminus after the budding domain (p6), while the NLS and the ABP were still grafted into the NC domain. When comparing the NC-grafted original and the C-terminal NES version of the VLP system of this invention, no additional benefits could be observed (Figure 6G), indicating that the non-covalent interaction of PCP-PP7 that is sufficient to export the genome editor from the nucleus and package it into VLPs due to the high local concentrations of cargo and packaging components, while in the recipient cells.

[0497] As mentioned before, in the present invention the RNP complex is recruited non-covalently via an aptamer handle on the (pe)gRNA (Figure 6A). Importantly, the naked RNA component of a GE-RNP, that is, the gRNA, pegRNA, or variant thereof, is extremely unstable without the proteinogenic component due to ubiquitous endo- and exonucleases in the cellular environment while vice versa the proteinogenic component is relatively stable in its naked protein-only form, the inventors deduct that prior arts’ systems cannot exclude that non-productive proteinogenic components of a GE agent without the gRNA component are packaged. The present inventors solved this problem by packaging it via the least-stable component of a GE-RNP, the gRNA or pegRNA. To achieve this, an aptamer handle is grafted into the gRNA or pegRNA.

[0498] Importantly, the packaging mechanism via non-covalent binding of the aptamer-tagged guide RNA prioritizes the least stable component of GE-RNPs - the guide RNA - as the anchoring point for packaging. This approach avoids inefficiencies associated with covalent fusions towards the proteinogenic component of a GE-RNP of the prior art, enabling precise, selective, and efficient loading of fully functional GE-RNPs.

[0499] The inventors further recognized that due to the covalent fusions in the prior arts, the release of the GE- RNP from the MBD is also limiting in the recipient cells and must rely on proteolytic processing of the MBD-GE-RNP fusion complex to enable the nuclear trafficking of the released GE-RNP to the recipient cell’s nucleus for genome editing. In contrast, the system of the present invention allows the release of the GE-RNP from the MBD in the recipient cells due to the non-covalent nature of how the GE-RNP is tethered to the MBD, making proteolytic release mechanisms via site-specific proteases non-obligatory. The inventors have shown that even in the absence of a maturation factor, that is, an active protease, the efficacy of the delivery system stayed nearly the same. Thus, considering the guide RNAs’ limited stability and the reduced affinity between modified guide RNAs towards the respective Cas proteins, the inventors found that the most efficient technical solution to ensure the loaded CRISPR Cas components contain fully functional CRISPR Cas RNPs is to use the guide RNA as loading handle.

[0500] Since the Cas protein for mammalian genome editing applications is usually tagged with an NLS and the guide RNA is expressed from an RNA polymerase III promoter that resides mainly in the nuclear compartment, the inventors conclude that to enable efficient loading of nuclear RNPs into MBD-ABP chimeras that are usually extranuclear, the MBD-ABP chimera has to shuttle between the nuclear and the cytoplasmic compartment. Hence, the inventors proposed that the fusion of the MDP-ABP chimera towards at least one shuttling domain that comprises at least one nuclear localization signal and at least one nuclear export signal may pose a suitable solution for efficient CRISPR Cas RNP loading into membrane-budding domains fused to aptamer-binding proteins. This solution was then carefully experimentally validated against suitable controls, most notably against controls lacking nucleocytoplasmic shuttling motifs (ANLS-NES), validating the inventors’ original hypothesis (see, e.g., Figure 3E; Figure 5B; Figure 15B-D; Figure 18).

[0501] Of note, the inventors’ solution provides a much more modular and sustainable solution compared to Gag-Cas direction fusions since the inventor’s system enables a quick adaption to potential new genome editing systems, where only an aptamer handle has to be introduced to the guide RNA components, whereas the budding module remains unmodified.

[0502] In contrast, systems described in the prior art based on Gag-Cas fusion require tedious optimization when the Cas component is exchanged. The modular design supports rapid adaptation to new genome editing tools by introducing aptamer handles into guide RNAs without modifying the core delivery components. This is a stark contrast to Gag-Cas fusion systems of the prior art, which require extensive re-optimization for each new genome editor. Thus, the flexibility and scalability of the system of the present invention is highly advantageous, e.g., it is an advantage of the present invention that the VLP system does not require tedious optimization when the Cas component is exchanged.

[0503] In one embodiment, the VLP system of this invention does not comprise a covalent fusion of Cas9 effectors to a budding module. Importantly, ENVLPE-mediated packaging, i.e. the VLP system of this invention, achieves a higher functional RNP loading capacity (1: 1 ratio of pegRNA:PE) compared to v3 and v3b PE-eVLP systems (Figure 29A-B).

[0504] Further engineering of a homomeric VLP assembly setup. As described in the previous section, the non-covalent cargo recruitment makes the present invention less dependent on HIV-1 protease activity. This provides the opportunity to omit Gag-polo64v from VLP production entirely, and the inventors concordantly observed that the omission of Gag-polo64v leads to a substantial yet tolerable decrease in efficacy of RNP delivery. Importantly, anon-chimeric VLP setup allows for further engineering of the gag polyprotein, which the inventors pursued by implementing minimal Gag variants (Accola et al., 2000) with deletions in MA, CA and NC. The inventors tested the PCP fusion of “midi -Gag” (AMA12-114, ANC) and mini-Gag (AMA12-114, ACA133-277, ANC) (Figure 7A) and observed that while midi-Gag-PCP did not lead to satisfying editing efficacy in the targeted cells, miniGag-PCP seems to be a promising option (Figure ID, Figure 7B). In addition, the inventors tested an alternative mode of membrane association and fused the phospholipase C-51 pleckstrin homology domain (PH) to the N-terminus of various gag variants (Figures 7 and 8).

[0505] The inventors observed increased editing, especially with the PH-miniGag variant (Figure 7B). The inventors also hypothesized that the localization equilibrium of the shuttling minimal Gag might have shifted based on its smaller size easing up nuclear import. The inventors, therefore, tested weakened NLS variants and combinations within the NLS-NES-NLS shuttling motif, but none of them led to a significant increase in editing over the initial mini-Gag version.

[0506] Dual-function domains for ultra-high-affinity cargo loading and stabilization.

[0507] Previous studies showed that nucleic acid binding domains can be either fused to the C-terminus of Gag, inserted into the NC domain or replacing portions of the NC domain, therefore the inventors set out to find the best possible cargo loading capability with minor impact on Gag’s budding capability Indeed, when the inventors compared their choice to replace ZF2 in the NC domain to other possible variants (PCP vs. tandem copies of PCP (tdPCP) and NC ZF2 grafting vs. C-terminal fusion), the inventors’ initial design was by far the best variant (PCPZF2 > tdPCPc-terminus > tdPCPzF2 > PCPc-terminus) for the delivery of PE VLPs as mosaic chimera VLPs mixture WT Gag / Gag-pol via psPAX2o64v (Figure 10C). The choice to replace ZF2 in NC was based on the prior knowledge that PCP forms a homodimer and its C-terminal fusion and dimerization might interfere with the budding process by sterically blocking accessibility of the ESCRT-I (via TSG101) and ESCRT-III (via ALIX) to the p6 domain located at the C-terminus of Gag. On the other hand, tdPCP is quite bulky and probably hinders Gag -Gag homo- multimerization due to its spatial requirements, while its tandem nature (tdPCP does not form dimers with other tdPCPs) makes it a more suitable choice for a C-terminal fusion in contrast to the non-tandem counterpart.

[0508] In order to improve the editing efficacy per VLP, RNP loading efficiency or half-life have to be improved even further, either by a much higher affinity ABP-aptamer handle or by increasing the effective steady-state concentration of functional non-degraded RNPs. Although protected by a 3’ (t)evopreQl motif, the 3 ’-end is probably still the weakest spot of a PE RNP complex. The inventors made use of a CRISPR protein Csy4 (alias Cas6f) from Pseudomonas aeruginosa that recognizes a highly conserved 16 nt RNA motif (Csy4 motif C4), wherein the nucleotides 2-16 form a hairpin with a 5 bp stem and a 5 nt loop (SEQ ID NO: 215, 216). When embedded in a stretch of RNA, C4 is bound by Csy4 with an exceptional affinity of 50 pM, which is 20 times higher than the PP7-PCP (Kd = 1 nM) interaction. More importantly, Csy4 also precisely cleaves 3 ’-proximal of the C4 hairpin without leaving any unpaired nucleotides as 3 ’-overhang. After processing the hairpin, Csy4 will remain bound to the hairpin with a non-altered affinity. The inventors hypothesized that these characteristics make Csy4 an ideal tool to protect the 3 ’-end of pegRNAs and / or to replace the PCP-PP7 interaction for loading RNPs into the VLPs, or optimally both. When Csy4 is co-expressed in the VLP production cells and the 3’- end of the pegRNA comprises a C4 motif downstream of the PP7, RTT, and PBS but either upstream, downstream of even without the (t)evopreQl (QI) pseudoknot motif, editing efficacy in the recipient cells was dramatically improved compared to controls where C4 is omitted but Csy4 is co-expressed, or vice versa, or to the reference VLPs. For all tested conditions, packaging of the pegRNAs into the VLPs was based on the PP7 aptamer inserted either in the scaffold or at the 3 ’ end upstream of Q 1 and / or C4 (Figure 10A, Figure 10B). Here, the effect of the improved efficacy and efficiency of the PE VLPs can be solely traced back to the 3 ’-protection effect of Csy4 first cleaving off the vulnerable ssRNA / polyU stretch (from Pol III termination) and shielding the C4 hairpin 3 ’-end by clamping it with 50 pM affinity.

[0509] To combine the additional layer of 3 ’-protection with potentially better cargo loading due to the exceptionally high Csy4 / C4 affinity, the inventors replaced the grafted PCP domain in the ZF2 region of the NC domain with Csy4 (Figure 1C). The inventors also created an alternative version, where Csy4 is fused after p6 instead, since in contrast to PCP, Csy4 does not dimerize, and thus, interference with the budding process is less likely (Figure 1C). When again applied on the inventors’ ‘blue mGL’ reporter system, it outperformed the reference approach using PCP-PP7, independent if Csy4 is either grafted into the NC domain or fused to the C-terminus of Gag (3’-Ql-C4 > 3’-C4-Ql > 3’-C4, Figure 10C). Of note, these results are somewhat surprising since direct transfection of Csy4 into the reporter cell lines together with iPE and the respective C4-modified pegRNA did not yield any improvement in editing performance indicating Csy4’s unique property only beneficial for scenarios where gRNA stability and availability is the major bottleneck. (Figure 11A, Figure 1 IB).

[0510] Combining VLPs + ID LVs for H DR-mediated genome editing.

[0511] Besides small nucleotides changes which can be mediated by base- and prime-editors, delivery of an HDR-donor template in combination with targeted Cas9 double strand break can be used to introduce large edits or gene insertions. To enable HDR, Cas9 nuclease and an aptamer-tagged gRNA was delivered as RNPs to HEK293T via the VLP system of the present invention targeting the ‘blue mGL’ reporter system (Figure 12, left). Additionally, an integrase-defective lentivirus encoding the portions of the mGreenLantem chromophore flanked by homology arms to provide a DNA template (after reverse transcription of the delivered RNA transfer vector) for HDR-based repair (Figure 12, right). The VLP as well as the IDLV were pseudotyped with the ecotropic envelope protein of the Moloney murine leukemia virus (MMLV- Env A607-622) targeting the murine Slc7al receptor (alias mCATl) that is expressed on the inventors’ reporter cell line. To enhance HDR, cells were treated with 0.5 pM AZD7648 during transduction to suppress NHEJ, a repair pathway otherwise competing with HDR.

[0512] Short summary.

[0513] Virus-like particles (VLPs) have emerged as promising delivery systems for genome editing in vivo due to modular tropism and their intrinsic ability to deliver different cargo types, such as ribonucleoproteins (RNPs), without necessarily introducing a transgene. Herein, the inventors present a novel VLP design derived from the HIV-1 Gag polyprotein, which can efficiently deliver different RNPs to target cells. The inventors demonstrate that the Gag polyprotein (Gag-PCP) specifically recruits (pe)gRNA-loaded RNPs to be packaged as cargo via the interaction between PCP and the PP7 aptamer, which was grafted into to the (pe)gRNA scaffold (but can also be added to the 3 ’ end of pegRNAs specifically). Furthermore, the inventors show that exporting the (pe)gRNA from the nucleus of the producer cell is a critical bottleneck that can be alleviated via shuttling of the Gag protein, specifically by introducing a nuclear localization (NLS) and nuclear export sequence (NES) into the Gag-PCP fusion protein. The delivery system was optimized using the inventors' HEK293T reporter cell lines initially and then successfully applied for the delivery of functional RNPs to primary cell lines, human induced pluripotent stem cells (hiPSCs), and “hard-to-transfect” cell lines such as hiPSC-derived cortical neurons. Simultaneously, the inventors comprehensively demonstrate the modularity and versatility of the inventors' system by delivering various Cas9 effector RNPs to mediate double-strand breaks (DSB), base editing (BE), prime editing (PE) or transactivation.

[0514] The cargo recruitment mechanism of the present invention differs quite significantly in comparison to recent Gag-cargo fusion approaches, providing advantages for efficacy and modularity:

[0515] (1) The RNP complex is bound and packaged via the PP7 aptamer within the (pe)gRNA, enabling straightforward packaging of virtually any Cas9 effector protein without modification in any way.

[0516] (2) Binding to the pre-assembled RNP complex with a nuclear-cytosolic shuttling Gag presumably increases the stability of (pe)gRNAs in the nucleus and prevents packaging of empty Cas9 effector proteins at the same time. (3) Additionally, the non-covalent cargo recruitment facilitates release in the target cell without further processing, thereby granting independence from any protease activity. Omitting other apparently non-essential enzymatic components of Gag-pol, enables the formation of homomeric, Gag-pol-free VLPs and importantly also allows for the implementation and engineering of structurally altered VLPs such as a minimal-Gag variant.

[0517] (4) The present invention includes an additional stabilization of the gRNA, especially pegRNAs, by either co-expressing Csy4 / Cas6f together with tagging the (pe)gRNA with a C4 motif inserted into a suitable position close to the 3 ’-end. Alternatively, Csy4 / Cas6f can be grafted into the budding module to fulfill two functions: ABP / aptamer pair for loading of the RNPs and protection of the RNPs at the same time.

[0518] In summary, the inventors' novel VLP system of the present invention offers a more efficient, effective, and modular method for the delivery of genome editing complexes as ribonucleoproteins. The optimized design, which utilizes an aptamer-binding protein and dual localization signals, ensures the successful packaging of active RNPs, providing a significant advantage over existing systems.

[0519] The option to omit gag-pol from VLP assemblies poses a key innovation compared to other delivery VLP technologies.

[0520] The following examples and drawings illustrate the present invention without, however, limiting the same thereto.

[0521] BRIEF DESCRIPTION OF THE DRAWINGS

[0522] Figure 1 Design of a nucleocytosolic shuttling RNP delivery vehicle.

[0523] (A) Graphical depiction of the Gag / Gag-pol polypeptide, a lentiviral packaging vector, which codes for Gag (matrix (MA), capsid (CA), nucleocapsid (NC) and p6 domain). The alternative frame codes for the enzymatic components including the functional domains of claim 1, which are highlighted with (a)- (k). (a) plasma membrane-interacting domain, (b) oligomerization domain, (c) membrane-bending domain, (d) shuttling domain, (e) nucleic acid-binding (NAB) domain, (f) budding domain, (g) reverse transcriptase domain, (h) maturation factor, (i) reporter-signal generating domain, (j) nucleic acidstabilizing domain

[0524] (B) Schematic illustration of the HIV-1 derived shuttling Gag-PCP, whereas PCP serves as an example of a nucleic acid binding (NAB) domain. The first zinc finger motif was inactivated by C>S mutation, while the second zinc finger was replaced by the PP7 coat protein (PCP). The additional nuclear localization and nuclear export sequences mediate Gag shuttling between nucleus and cytosol for efficient packaging of the nuclear localized (pe)gRNA. For (co)-expression of Gag / Gag-Polo64v, a mutant Gag-Polo64v was used to inactivate the integrase function (encoded on the Pol frame) and to preclude random integration.

[0525] (C) Schematic illustration of the shuttling Gag Csy4 fusion protein, where the Csy4 domain was inserted in the NC domain of HIV-1 Gag (top), or fused C-terminally (bottom).

[0526] (D) Illustration of the shuttling -mini -Gag -PCP fusion protein (top) and with the additional N-terminal Pleckstrin (PH) domain (bottom).

[0527] Figure 2 Delivery of PP7-tagged mRNA.

[0528] (A) Illustration of mRNA packaging using non-shuttling Gag -PCP. The PP7 aptamer is encoded in the 3’UTR of the cargo mRNA. The VLPs were pseudotyped with VSV-G.

[0529] (B) Effect of the protease inhibitor Darunavir on mGreenLantem mRNA delivery. Darunavir was added during production (prod.) or during transduction (TD) or both (prod. + TD).

[0530] (C) Impact of the inclusion of the L21 S mutation in the MA domain of Gag when introduced in Gag- PCP or Gag-pol.

[0531] (D) Comparison of mGreenLantem mRNA delivery via the Gag-pol-PCP or the mosaic set-up using Gag-pol and Gag -PCP.

[0532] (E) Specificity of Gag -PCP towards PP7-tagged aptamers and the effect of microRNA (miR) transfection (depicted by TF) targeting the 3’ UTR of mGreenLantem.

[0533] (F) Comparison of mRNA delivery efficiency based on the EPN-24-PCP nanocage or Gag -PCP.

[0534] Figure 3 Proof of concept of the nucleocytosolic shuttling RNP delivery vehicle, using fluorescence reporter cell lines to quantify prime editing (PE) events.

[0535] (A) Depiction of packaging and cargo components of the delivery system of the present invention. The cargo RNP, consisting of a Cas9 effector of choice and a corresponding (pe)gRNA. Recmitment of the RNP is mediated via a PP7 aptamer grafted into the (pe)gRNA scaffold, which strongly interacts with the PCP domain in Gag. The cargo RNP complex can already form in the nucleus, stabilizing both components, efficient cargo packaging requires ribonucleoprotein (RNP) translocation from the nucleus into the cytosol, which is enabled by the NLS-NES motif in Gag. After cargo packaging and budding, proteolytic VLP maturation occurs. Notably, cargo release is independent of this process as the RNP is not covalently linked to Gag.

[0536] (B) Prime editing was monitored in a reporter cell line with stably integrated mGreenLantem harbouring two mutations in the chromophore (G65S, Y66H), which shift its excitation range from green to blue. Successful PE reverses the mutations back to wild type, resulting in green fluorescence.

[0537] (C) To determine the limiting component of the prime editor complex, either the mRNA encoding the prime editor protein or the pegRNA were transfected by standard lipofection (TF) while the complementary component was transduced via VLP. (D) The different PE efficiencies were obtained for separate packaging of the PP7-tagged pegRNA and iPE mRNA (VLP1 and VLP2) as well as co-packaging of both components into one VLP (VLP1 only).

[0538] (E) The importance of each component of the delivery system was assessed for PE delivery using the PE reporter. The control conditions demonstrate the importance of the shuttling motif and the PP7:PCP interaction for delivery efficiency.

[0539] (E) Prime editing efficiencies of the nucleocytosolic shuttling Gag-PCP. Quantification of prime editing events in the ‘blue’ mGreenLantem reporter cell line (HEK293T) after delivery of iPE mRNA and PP7- aptamer tagged pegRNAs . N / C indicates the position of the reverse transcriptase (RT) within the iPE complex.

[0540] (F) To determine if the prime editor is delivered as mRNA or rather as RNP, the PE reporter cell line was transfected with a miR targeting the prime editor mRNA. Both, pegRNA and mRNA, were tagged with a PP7 aptamer.

[0541] (G) PE delivery efficiency of the Gag-pol and Gag-PCP mosaic set-up in comparison to a direct insertion of PCP into Gag-poloew.

[0542] Figure 4 Optimization and characterization of the delivery system.

[0543] (A) Analysis of the impact on PE delivery when the L21S mutation as well as the PCP domain is included in one or both packaging components.

[0544] (B) Inactivation of each component encoded in the 2nd generation packaging vector psPAX2 and the impact of the protease inhibitor Darunavir.

[0545] (C) Optimization of the amount of VSV-G glycoprotein for pseudotyping.

[0546] (D) Optimization of the stoichiometry of Gag-Pol and Gag-PCP (packaging plasmids) as well as the ratio of packaging plasmids to cargo plasmids.

[0547] (E) Titration of the cargo plasmids encoding the pegRNA and nCas9:RT.

[0548] Figure 5 Delivery of alternative RNP effectors for introduction of InDeis, base editing (BE) or transactivation of an endogenous target locus.

[0549] (A) The enhanced traffic light reporter (eTLR) was used to quantify DSB events. By targeting the stop codon region, all potential InDei events lead to a read-through activating eUnaG in any of the three downstream ORFs. The reporter can also be used to detect a successful base conversion within that same stop codon, leading to continued translation of the downstream eUnaG reporter.

[0550] (B) Delivery of Cas9 and PP7-tagged gRNA as RNPs. Quantification of InDeis in HEK293 reporter cell line via flow cytometry.

[0551] (C) Delivery of ABE8CNLS and PP7-tagged gRNA as RNPs. Quantification of base conversion within the TLR reporter in cortical neurons via flow cytometry.

[0552] (D) Quantification of base conversion upon delivery of an ABE8CNLS RNP into various cell lines. The gRNA targets the splice donor (SD) of the B2M locus leading to a functional KO of the B2M locus. (E) Transactivation of the MAPT gene expression by delivery of a dCas9-miniVPRNLs RNP into a HEK293T cell line the inventors described previously (Truong et al., 2021). MAPT expression of the reporter cell line can be followed by luciferase expression.

[0553] Figure 6 Comparison of the delivery system of the present invention to eVLPs.

[0554] (A) Packaging modalities: Cas9 binds and stabilizes the (pe)gRNA in the nucleus; nuclear cytosolic shuttling of Gag-PCP enables binding of the fully assembled RNP via PCP:PP7 interaction. The system allows for modular packaging of essential any RNP complex if the (pe)gRNA is tagged with the PP7 aptamer.

[0555] (B) eVLP consists of a chimeric fusion of Gag and a genome editing effector protein of choice. The fusion protein shuttles between nucleus and cytosol to package the unprotected (pe)gRNA. In contrast to the present invention, packaging of the Cas protein without (pe)gRNA is possible, the genome editor is released after budding upon proteolytic cleavage of the protease encoded in Gag-pol.

[0556] (C) Comparison of base editing efficacy regarding editing the eTLR reporter.

[0557] (D) Titration curves of base conversion of the B2M splice donor after delivery of the present invention and eVLPs to HEK293T or iPSC. Quantification of base editing via NGS.

[0558] (E) Quantification of InDeis in the 2A7 locus after Cas9-RNP delivery into HEK293T and iPSC.

[0559] (F) Comparison of RNPs for prime editing with a derivate of the delivery system of the present invention and eVLP system from the Liu group (eVLP-like), where the inventors exchanged the Cas9 effector domain for the iPE domain within the Gag-Cas9 fusion protein. The eVLP-like iPE was codelivered with a pegRNA to correct the mutations of the ‘blue’ mGreenLantem reporter, vl refers to the co-delivery of iPE and the pegRNA from previous experiments, with the PP7 aptamer grafted into the scaffold. In v2, the inventors translocated the PP7 aptamer to the 3 ’ of the pegRNA preventing packaging of 3’ truncated pegRNAs.

[0560] (G) Illustration of repositioning the NES motif from the shuttling domain to the C-terminus (top). Comparison of both NES positions regarding delivery of prime editors to the ‘blue’ mGreenLantem reporter cell lines (bottom).

[0561] Figure 7 Engineering of a minimal Gag variant for delivery of functional RNPs.

[0562] (A) Schematic of WT HIV-1 Gag (top) and minimal Gag (bottom) which is truncated in MA and CA. The NC domain is completely replaced by the shuttling motif and PCP. To improve particle formation at the membrane, a Pleckstrin homology domain (PH) was fused to the N-terminus of Gag.

[0563] (B) Set up of a miniGag variant (darker green) with deletions in MA, CA and NC in combination with the phospholipase C-51 pleckstrin homology domain5 (PH) for membrane association and a GCN4 coiled coil for oligomerization, as well as alternative nuclear-cytosolic shuttling motifs. Quantification of the prime editor complex delivery to reverse the G65S / Y66H mutation of the mGreenLantem reporter via flow cytometry. (C) Comparison of shuttling PH-miniGag against Gag-PCP- and eVLP- delivery regarding base editing efficiency of the TLR reporter locus in HEK293.

[0564] Figure 8 Optimization of the shuttling motif for minimal Gag variants.

[0565] (A) Schematic depiction of shuttling PH-miniGag-PCP carrying multiple deletions (AMA12-114, ACAI 33-277, ANC).

[0566] (B) Combination of several NLS-NES shuttling motifs with miniGag-PCP.

[0567] Figure 9 Workflow ofVLP production process.

[0568] (A) Co-transfection of HEK293T with packaging plasmid, cargo plasmid and pseudotyping plasmid. The cargo plasmid encodes for the PP7 aptamer-tagged (pe)gRNA and a suitable genome editing effector protein. 48-72 h post-transfection of the producer cells, the supernatant (SN) containing the VLPs, is collected and sterile-filtered to remove cell debris and left-over transfection reagents. Subsequently, the filtered SN is concentrated by using a molecular weight cutoff filter (100 kDa).

[0569] (B) For pseudo-transduction of the target cells, the supernatant is applied to the cells without exceeding 20% of the total volume of the receiver cell’s original growth medium.

[0570] (C) For optimizing the VLP system, cells were analyzed 72 h post-pseudo-transduction via targeted amplicon sequencing of the respective targets or FACS analysis if fluorescence reporter cell lines were used.

[0571] Figure 10 Implementation of the C4 aptamer and Csy4 enzyme to optimize the delivery system.

[0572] (A) Depiction of packaging and cargo components of the delivery system additionally utilizing the Csy4 ABP-aptamer system either only as protection for the pegRNA 3 ’ end (left) or for both protection and recruitment of the pegRNA simultaneously (right), replacing PCP-PP7.

[0573] To increase pegRNA stability, a Csy4-aptamer (C4) was introduced at different positions of PP7- aptamer modified pegRNAs in combination with tEvopreQl (QI) and co-expressed with the Csy4 / Cas6f protein in the VLP producer cell line. Prime editing efficacy was assessed using the blue- shifted mGreenLantem reporter system and quantifying the percentage of green cells and median fluorescence intensity separately or combined.

[0574] (B) The omission of Csy4 protein expression in the VLP producer cell line was assessed using the best-performing C4 aptamer pegRNA variant from (A)Prime editing efficacy was assessed using the blue-shifted mGreenLantem reporter system and quantifying the percentage of green cells and median fluorescence intensity separately (left) or combined (right).

[0575] (C) To switch from PP7-mediated recruitment to Csy4-mediated recruitment, PCP was exchanged for the Csy4 protein, which was grafted into the NC domain (as PCP before) or placed at the C-terminus of the Gag-PCP fusion protein. For the pegRNA, the PP7 aptamer was correspondingly omitted and the C4 aptamer was placed at different positions in the 3’ end. Prime editing efficacy was referenced against multiple PCP-PP7 configurations, again using the blue-shifted mGreenLantem reporter system and quantifying the percentage of green cells and median fluorescence intensity separately (left) or combined (right).

[0576] Figure 11 Transfection ofHEK293T to test pegRNAs comprising the C4 aptamer as stabilization motifs for prime editing.

[0577] (A) To assess the effect of the C4 aptamer at the 3’ of pegRNAs in transient transfection settings, prime editing efficiencies were obtained in absence and presence of the Csy4 protein.

[0578] (B) Quantification of prime editing efficiencies for pegRNAs containing the PP7 and C4 aptamer with varying position at the 3 ’-end. The expression of Csy4 and iPE is coupled via a ribosomal skipping mechanism (2A).

[0579] Figure 12 Delivery of a DNA donor template for homology-directed repair of the ‘blue ’ mGreenLantem reporter.

[0580] The ‘blue’ mGreenLantem (G65S, Y66H) reporter cell line shows blue fluorescence unless the chromophore mutations are reversed, e.g., via prime editing or after a targeted double-strand break at the chromophore site followed by homologous recombination (HR or HDR) mediated repair using a donor DNA template coding for the reversed chromophore mutations or delivery (left).

[0581] To enable HDR, Cas9 and a PP7-tagged gRNA targeting the reporter locus was delivered to HEK293T via Gag-PCP VLPs. Additionally, a psi-tagged RNA encoding the mGreenLantem chromophore flanked by homology arms and long terminal repeats (LTRs) was packaged and reverse transcribed via an integrase-deficient (D64V) lentivirus (IDLV) to provide a DNA template for HDR (right). The VLP as well as the IDLV were pseudotyped with the ecotropic envelope protein of the murine leukemia vims (MLV-Env A607-622) targeting the murine Slc7al (mCATl) receptor. To enhance HDR, cells were treated with 0.5 pM AZD7648 prior to transduction.

[0582] Figure 13 Implementation of Csy4 for improved prime editing (PE) guide RNA (pegRNA) stability.

[0583] (A) Schematic of Csy4-protected PE RNP complex formation for packaging via ENVLPE. Csy4 cleaves off the 3’ overhanging nucleotides at the pegRNA’s C4 aptamer and remains bound, thereby providing additional 3 ’-stabilization against exonucleolytic attack. The PE-RNP complex is then exported by the shuttling ENVLPE and combined with the budding module.

[0584] (B) Comparison of ENVLPE VLPs packaged with different 3 ’-stabilized PE RNP complexes comprising the indicated 3 ’-modifications and covalent or non-covalent Csy4 tethers across a range of VLP doses. PE-mediated reversion of the ‘blue’ mGL reporter in HEK293T cells is quantified via FACS and displayed as a percentage of green cells (lines, left y-axis) or the percentage multiplied by the median fluorescence intensity (MFI) of green cells (dashed lines, right y-axis). PP7, PP7 aptamer bound by PP7 coat protein (PCP); QI, evopreQl pseudoknot; C4, Csy4 aptamer hairpin bound and processed by Csy4; elements are described from 5’ to 3’ as they appear at the 3 ’-end of the respective pegRNA. Points represent mean ± SD (n = 3 biological replicates). Log2(VreiSN): relative transduction volumes in 2x dilution steps. Editing rates of the lowest VLP titer were analyzed via Bonferroni MCT after one-way AN OVA (**** > < 0.0001; **p < 0.01).

[0585] (C) Comparison of different VLP systems packaged with PE RNPs with different 3 ’-protection strategies applied to ‘blue’ mGL HEK293T cells across a range of VLP doses. La indicates the fusion of the N-terminal fragment (aa 1-194) of La (SSB) used in PE7. Solid and dashed lines represent the metrics as defined in (B). Points represent n = 3 individual replicates (orange) and n = 2 (green and blue). Log2(VreiSN): relative transduction volumes in 2x dilution steps.

[0586] Figure 14 Engineering of a minimal Gag variant for delivery of functional RNPs.

[0587] (A) Schematic representation of engineered minimal Gag versions with a majority of MA and CA deleted as indicated. The shuttling motif and PCP replace the complete NC domain which also ablates the function of Gag to homo-oligomerize. Homo-oligomerization is restored by the introduction of GCN4 coiled coils (CCGCNQ. A phospholipase C-51 pleckstrin homology domain (PH) was fused to the N-terminus to replace the native N-myristoylation site. The resulting particles are formed from homomeric assemblies.

[0588] (B) Schematic representation of full-length Gag-PCP as used in ENVLPE for reference.

[0589] (C) Quantification of PE events in the ‘blue’ mGL HEK293T reporter cell line via flow cytometry 72 h after transduction. Respective ENVLPE variants, shuttling domain variants, and membrane anchors (MY: HIV-1 N-myristoylation; PH: phospholipase C-51 pleckstrin homology domain) are indicated. Bars represent mean± SD (n = 3 biological replicates).

[0590] (D) Comparison of the original ENVLPE (left) to variants without Gag / Gag-Polo64v, and the addition of PH to full-length Gag-PCP as in (A, top) to miniENVLPE versions with different engineered nucleocytosolic shuttling motifs. PE was quantified in the ‘blue’ mGL reporter HEK293T cell line via flow cytometry 72 h after transduction with VLPs. Bars represent mean ± SD (n = 3 biological replicates).

[0591] (E) Validation of the Csy4-C4 3 ’-protection module in the untethered or direct recruitment mode (see Figure 22) in the context of miniENVLPE. Experiments were conducted in the ‘blue’ mGL HEK293T cell line. Bars represent mean± SD (n = 3 biological replicates).

[0592] (F) The influence of proteolytic activity on PE efficacy in the ‘blue’ mGL HEK293T reporter cell line in the context of the ENVLPE / Gag-PolD64v mosaic setup and the homomeric miniENVLPE. Bars represent mean ± SD (n = 3 biological replicates).

[0593] Selected results of Bonferroni multiple comparisons test (MCT) after one-way ANOVA analysis are shown for (C) and (E) (****p < 0.0001). Figure 15 Optimized ENVLPE+benchmarked for PE and BE at several endogenous sites.

[0594] (A) Depiction of the improved module with an additional GCN4 domain (ENVLPE+) to assist Gag oligomerization (left). Editing performance of ENVLPE+compared to ENVLPE (right). Experiments were performed on the ‘blue’ mGL HEK293T line. Bars represent mean ± SD (n = 3 biological replicates). Results from apaired / -test are indicated (**p < 0.01).

[0595] (B) Evaluation of ENVLPE, miniENVLPE, ENVLPE+, and v3b PE-eVLP delivery on several endogenous loci in HEK293T cells (HEK3 and RUNX1) and in Neuro-2a cells (Dnmtl) using either the ‘PE2’ (default) or ‘PE3’ (indicated on top) strategy. Editing efficacy was analyzed via targeted amplicon sequencing, and precision was calculated by the ratio of intended edits divided by the total number of all non-WT reads x 100%. Bars represent mean ± SD (n = 3 biological replicates, except in HEK3 PECsy4 ENVLPE (CAG) and PECsy4 ENVLPE+(CMV) where n = 2). HEK3 and Dnmtl data were analyzed using 2-way ANOVA with Bonferroni MCT 0.0001). The conditions of RUNX1 editing were analyzed by an unpaired two-tailed / -test (****p < 0.0001).

[0596] (C) Additional assessment of on- and off-target editing in HEK293T of the respective VLP systems in HEK293T cells compared to plasmid-transfection of PE components targeting the respective targets. The same non-targeting control (NTC) of HEK4 on-target editing is shown for both edit types. Bars represent mean ± SD (n = 3 biological replicates). Selected results of Bonferroni MCT after one-way ANOVA are indicated (**** < 0.0001; ***p < 0.001; ns p > 0.05).

[0597] (D) Evaluation of VLP -mediated BE in the HEK293 traffic light reporter (TLR) cell line and other endogenous sites and cell lines. Editing efficacy was analyzed via targeted amplicon sequencing; bars represent mean ± SD (n = 3 biological replicates).

[0598] Figure 16 Using ENVLPE+to generate hypoimmunogenic T cells ex vivo.

[0599] (A) Schematic illustration of TCR / CD3 and MHCI knockout by targeting B2M and TRBC1 / 2 in primary T lymphocytes.

[0600] (B) Ex vivo delivery of adenine base editor RNPs via CAGENVLPE+to primary T lymphocytes. Quantification of B2M knock-out (KO) and the effect on MHCI surface expression were analyzed 72 h after VLP delivery via NGS or flow cytometry (FC) after MCHI immunostaining, respectively. Log2(VreiSN): relative transduction volumes in 2x dilution steps.

[0601] (C) Ex vivo delivery of adenine base editor RNPs via CAGENVLPE+to primary T lymphocytes. Quantification of TRBC1 / 2 KO and the effect on TCR surface expression were analyzed 72 h after VLP delivery via NGS or FC after CD3 immunostaining, respectively. Log2(VreiSN): relative transduction volumes in 2x dilution steps.

[0602] (D) Analysis of functional MHCI / CD3 double KO in primary T lymphocytes by BE ENVLPE+vs. v4 BE-eVLPs. KO efficiency was quantified via FC (functional) and NGS 72 h after delivery of the indicated VLP systems. Bars represent mean ± SD (n = 3). Titers of the respective VLP systems were measured via ELISA (5.43 x IQ11v4 BE-eVLP particles per 50,000 cells; 1.34 x 1010ENVLPE+ particles per 50,000 cells; Figure 31). For dual targeting, 46 pl ofVLPs containing either B2M- targeting sgRNAs or ' / 'AW / 2-targcting sgRNAs were used for the eVLP or ENVLPE system, respectively.

[0603] Figure 17 Subretinal injection of ENVLPE particles for mRNA delivery into Cre-reporter mice and for PE RNP delivery to mouse models for inherited retinal degeneration.

[0604] (A) Depiction of subretinal injection ofVLPs into mouse retina. CHO, choroid; RPE, retinal pigment epithelium; PR, photoreceptor.

[0605] (B) Genetic construct of the reporter mouse indicating delivery of mRNA-encoded Cre activity with a red-green switch of fluorescent proteins expressed in RPE. Example 3D rendering of an imaging volume of the posterior portion of the intact eye acquired by 2-photon imaging after subretinal injection of 1.5 pl sample (containing 3.63 x 108ENVLPE particles). Green fluorescence was observed in 17 out of 18 eyes; additional images are shown in Figure 30.

[0606] (C) Schematic diagram of 4 bp deletion in the splice donor region of exon 4 of Mfrp in rd6 mice, showing the sequences of SEQ ID NOs: 492 to 495, 516 and 517.

[0607] (D) Immunoblot analysis of MFRP protein levels in rd6 mouse eyes, three weeks after prime editing. (E, F) Analysis of gDNA-editing efficacy and indel rate. Bars represent median with 95% CI (n = 12 biological replicates, except for ‘untreated’ where n = 3). Selected statistical comparisons using the Kolmogorov-Smimov test are shown (**** < 0.0001).

[0608] (G) Schematic diagram of the nonsense-mutation in exon 3 of Rpe65 in rd 12 mice, showing the sequences of SEQ ID NOs: 496 to 501, 518 and 519.

[0609] (H) Immunoblot analysis of RPE65 protein levels in rdl 2 mice after prime editing.

[0610] (I, J) Analysis of the corresponding editing efficacy and error rate, including indels. Bars represent median with 95% CI (n = 4 (untreated); n = 7 (eVLP); n = 6 (eVLP 13 x excess); n = 10 (ENVLPE ) biological replicates). Selected statistical comparisons using the Kolmogorov- Smimov test are shown (**p < 0.01).

[0611] (K) Scotopic electroretinogram (ERG) A -wave amplitudes upon light stimulus of -0.3 log[cd s m2| flash three weeks after injection of the different VLP systems or controls. Bars represent median with 95% CI (w = 2 (rd!2),' n = 9 (PBS); n = 56 (v3 PE-eVLP light orange); n = 8 (v3 PE-eVLP dark orange), n = 40 (ENVLPE+); n = 10 (C57BL / 6) biological replicates). Mice were dark-adapted overnight prior to analysis. Selected statistical comparisons using the Kolmogorov-Smimov test are shown (**** < 0.0001; ns p > 0.05).

[0612] (L) Representative ERG traces from (K).

[0613] (M) Schematic diagram of the visual cycle. Extracted whole-eye retinoids were analyzed by HPLC to quantify 1 1 -c / .s-rctinal as well as all- / ram-rctin l esters. The mobile phase for elution consisted of 99.4% hexanes and 0.6% ethyl acetate. Retinoid levels were determined using a standard curve. Results on the right show median and 95% CI for n = 4 control samples and n = 5 ENVLPE+and eVLP samples.

[0614] In D-M, injection volumes were adjusted to equalize the injected VLP titers from v3 PE-eVLP and ENVLPE+particles per eye. VLP titers were quantified via ELISA (Figure 31).

[0615] Figure 18 Impact of nucleocytosolic shuttling motifs on VLP-mediated packaging of CRISPR RNPs.

[0616] (A) Amino acid sequences of respective combinations of nuclear localization (NLS) and nuclear export sequences (NES) variants used in (B) and (C), showing the sequences of SEQ ID NOs: 507 to 511.

[0617] (B) Evaluation of different combinations of NLS or NES on the delivery of prime editors into the mGL HEK293T reporter line.

[0618] (C) ANLS-NES ENVLPE performed less effectively in a traffic light reporter system in base editing mode. In this system, base editing removes a premature stop codon, enabling the translation of a green fluorescent protein.

[0619] (D) Immunofluorescence staining of FLAG-tagged Gag variants carrying nuclear localization or export sequences depicted in (A); 48 h post-transfection into HEK293T cells. Addition of 20 nM Leptomycin B (LMB) blocks XPO1 / CRM1 -mediated nuclear export of HIV-1 REV NES, resulting in nuclear accumulation of a nucleocytosolic shuttling protein (indicated with a white arrow).

[0620] Figure 19 Optimization and characterization of the ENVLPE system for PE RNP delivery in HEK293T ‘blue ’ mGL cells.

[0621] (A) To determine the limiting component of the prime editor complex, either the mRNA encoding the prime editor protein or the pegRNA was transfected (TF), while the complementary component was transduced (VLP). For the condition where only iPE was delivered via VLP, the iPE mRNA was tagged with a PP7 aptamer to enable recruitment (VLP*).

[0622] (B) While being transduced with ENVLPE carrying PE RNPs targeting the ‘blue’ mGreenLantem reporter, recipient cells were additionally transfected with a microRNA targeting the 3 ’UTR of the iPE mRNA to investigate if any proportion of PE efficacy originated from mRNA delivery instead. miR functionality was independently verified in Figure S1E.

[0623] (C) Analysis of the impact on PE delivery when the L21S mutation, as well as the PCP domain, is included in one or both packaging components.

[0624] (D) Inactivation of indicated domains encoded in the 2ndgeneration IDLV packaging vector PSPAX2D64V and the impact of the protease inhibitor darunavir.

[0625] (E) Co-optimization of mass stoichiometry between Gag / Gag-Polo64v and Gag-PCP (packaging plasmids) and the ratio between packaging and cargo plasmids. The pegRNA:iPE ratio was set to 3: 1 in all conditions. (F) Co-optimization of mass stoichiometry between pegRNA and iPE (cargo plasmids), and the ratio between packaging and cargo plasmids. Gag / Gag-Polo64v and Gag-PCP (packaging plasmids) were expressed 1: 1 for all conditions.

[0626] (G) Effect of inserting the nucleocytosolic shuttling motif and PCP directly into the NC domain of Gag-PolD64v. Bars represent mean± SD (n = 3 biological replicates). Results of an unpaired / -test are shown (**** < 0.0001).

[0627] Figure 20 ENVLPE-mediated Cas9 delivery in reporter cortical neurons and transactivation of MAPT in HEK293T cells.

[0628] (A) Delivery of Cas9 and PP7-tagged gRNA as RNPs targeting the enhanced traffic-light reporter (eTLRv2); eTLRv2 signals all possible indel events via read-through activation of mGreenLantem in any of the three downstream ORFs. Double-strand breaks (DSBs) were quantified in hiPSC-derived cortical neurons stably expressing eTLRv2 via flow cytometry. Bars represent mean± SD (n = 3 biological replicates).

[0629] (B) Relative light units (RLU) of an EXSISERS MAPT HEK293T reporter cell line after transduction with dCas9-miniVPR ENVLPE to induce MAPT via CRISPRa. Bars represent mean ± SD (n = 3 biological replicates). Results of an unpaired / -test are shown

[0630] (**** < 0.0001).

[0631] Figure 21 Impact of Csy4 / C4 protection in transient transfection settings, validation of PCP- grafting position, and assessment of the versatility of the final pegRNA 3 ’ configuration.

[0632] (A) PE efficacies in ‘blue’ mGL reporter HEK293T line transfected with plasmids (z.e., not transduced with ENVLPE) coding for iPE and different pegRNAs with the indicated 3 ’-modified pegRNAs in the presence or absence of Csy4.

[0633] (B) PE efficacies of ENVLPE variants with different grafting positions of PCP (dimeric) and tandem dimers of PCP (monomeric), quantified by flow cytometry of the ’blue’ mGreenLantem reporter HEK293T cell line. NC indicates the replacement of the nucleocapsid ZF2 domain with (td)PCP, while C denotes the C-terminal fusion of (td)PCP downstream of Gag after p6. Bars represent mean ± SD (n = 3 biological replicates). Selected results are shown of Bonferroni multiple comparisons test (MCT) after one-way ANOVA analysis (****p < 0.0001).

[0634] (C) Checking the COM aptamer as an alternative to PCP under non-saturating conditions (8 x dilution). Experiments were performed on the ‘blue’ mGL HEK293T line. The position of the COM aptamer in the pegRNA is denoted as sc: scaffold, 3’: 3’ end. Bars represent mean± SD (n = 3 biological replicates).

[0635] Figure 22 Utilization of Csy4 for simultaneous high-affinity recruitment. (A, B) For the ENVLPEcsy4 strategy, Csy4 was inserted into Gag NC analogously to ENVLPE in Figure 1A, with the only difference being that an additional C-terminal fusion of Csy4 was tested. In this instance, the Csy4-C4 system can simultaneously be utilized for RNP recruitment and pegRNA stabilization by fusing Csy4 to shuttling Gag, thereby replacing the PCP-PP7 interaction entirely.

[0636] (C) Editing efficacy in the ‘blue’ mGL HEK293T line. The red curve represents ENVLPEcsy4 (C-term. fusion), which employs the C4-Csy4 interaction for RNP packaging and pegRNA stabilization, in comparison to ENVLPE in combination with untethered Csy4 for pegRNA protection (blue curve, duplicated from Figure 13B). Log2(VreiSN): relative transduction volumes in 2x dilution steps.

[0637] (D) Delivery of RNPs for PE via direct fusion of Csy4 to Gag (ENVLPEcsy4) to test different orientations of the C4 and QI motifs on the pegRNA 3’ end. ‘NC’ denotes Csy4 grafting into the NC domain, ‘C’ denotes fusion to the C terminus of Gag. Experiments were performed on the ‘blue’ mGL HEK293T line. Bars represent mean ± SD (n = 3 biological replicates). Data were analyzed using 2- way AN OVA with Bonferroni MCT (**** < 0.0001).

[0638] Figure 23 Evaluation of La / SSB-mediated 3 ’-protection in the context of ENVLPE and context specificity of Csy4 / C4 PE RNP protection system.

[0639] (A) ENVLPE packaged with PE RNPs comprising the final 3 ’-configuration ‘PP7-C4-Q1’ was compared with and without (un)tethered Csy4 expression under non-saturated conditions (8x dilution) for prime editing efficacy on the ‘blue’ mGL HEK293T line. Selected results of Bonferroni MCT after one-way ANOVA are indicated (****p < 0.0001; **p < 0.01; *p < 0.05; ns p > 0.05).

[0640] (B) Comparison of ENVLPE -mediated delivery of Csy4 / C4-protected prime editors vs. La / SSB- protected prime editors (PE7) across a range of VLP doses on the ‘blue’ mGL HEK293T line. The PE7 protection system is based on the fusion of an N-terminal fragment (aa 1-194) from La / SSB towards the C-terminus of the prime editor to bind and shield the polyU stretch of the pegRNA (with or without evopreQl) from 3’-exonucleolytic attack. Log2(VreiSN): relative transduction volumes in 2x dilution steps. Points represent n = 3 biological replicates (except ‘PE7 PP7-Q1’, where n = 2). Selected results of Bonferroni MCT after one-way ANOVA are indicated for the 8x dilution (log2(VrdSN) = -3, **** < 0.0001).

[0641] Figure 24 Impact of VLP s on cell viability and cytotoxicity.

[0642] HEK293T cells and induced pluripotent stem cells (iPSCs) were transduced with the indicated VLP systems. 48 hours post-TD, cell viability and cytotoxicity assays were performed. Puromycin 2 pg mF1was used as a positive control for cytotoxicity. Selected results of Bonferroni MCT after one-way ANOVA analysis are indicated (****p < 0.0001; *p < 0.05; ns p > 0.05).

[0643] Figure 25 Evaluation of alternative Csy4-independent 3 ’ stabilization strategies. (A) Depiction of the secondary structure of the pegRNA variant carrying PP7-C4-Q 1 (left) used throughout the manuscript, showing the sequence of SEQ ID NO: 512; and a variant (right) where truncated evopreQl (tevopreQl) is followed by an HDV ribozyme to remove the 3 ’-terminal polyU stretch, a product of RNA polymerase III promoter termination, showing the sequence of SEQ ID NO: 513.

[0644] (B) Comparison of the different alternative strategies for 3’-protection from 3’-exonucleolytic attack. dCsy4 indicates a processing-defective variant of Csy4 (H29A) that retains its binding capability for its C4 motif. vl-v4 indicate different minor modifications in the exact placement between C4 or QI to the HDV ribozyme with varying nucleotide linkers. Experiments were performed on the ‘blue’ mGL HEK293T line. Data were analyzed via one-way ANOVA with Bonferroni MCT (****p < 0.0001; **p < 0.01).

[0645] (C) Comparison of a variety of pseudoknot structures with and without a downstream HDV ribozyme as alternative 3’-protection motifs. Experiments were performed on the ‘blue’ mGL HEK293T line. Data were analyzed via one-way ANOVA with Bonferroni MCT (****p < 0.0001).

[0646] Figure 26 Effect of Csy4 protection on ENVLPE-mediated Cas9 delivery or base editing.

[0647] (A) Illustration of sgRNA structures, either with insertion of the PP7 aptamer into the scaffold in combination with the C4 aptamer and downstream tevopreQl at the 3 ’-end, or all the components at the 3’-end (showing the sequences of SEQ ID NOs: 514 and 515).

[0648] (B, C) Investigation of the effect of the C4 / Csy4 protection strategy on ENVLPE-mediated Cas9 and base editor delivery using the gRNAs scaffolds depicted in (A). Csy4 was either co-expressed with the Cas9 effector via a 2A peptide (‘P2A’), co-transfected as a separate plasmid (‘Co-TF’), or fused to ENVLPE (‘recruitment’). For (B), indels were quantified via flow cytometry analysis of the TLR reporter cell line. For (C), on-target base edits and bystander edits of the endogenous B2M locus in HEK293T cells were quantified vzaNGS. Results were analyzed via one-way ANOVA; selected results from Bonferroni MCT are indicated by asterisks (****p < 0.0001; ns p > 0.05). Sc, scaffold- PP7; 3’, 3’-PP7.

[0649] Figure 27 Investigation of the ENVLPE size distribution and release mechanism.

[0650] (A) Analysis of the impact of Gag / Gag-Polo64v co-expression on PE and PEcsy4 efficacy for ENVLPE. Experiments were performed on the ‘blue’ mGL HEK293T line.

[0651] (B) Assessment of VLP size distributions comparison to HIV-1 Gag via cryo-electron microscopy. The thick dotted line represents the median, while the thin dotted lines constitute the quantiles. Results were analyzed by the Kruskal-Wallis test with Dunn’s MCT (****p < 0.0001; ns p > 0.05).

[0652] Representative cryo-EM micrographs of the respective species are shown on the right; HIV-1 denotes particles made of HIV- 1 Gag / Gag-Polo64v expressed from psPAX2o64v. Scale bar = 50 nm. (C) Investigating the protease dependence of ENVLPEcsy4 by functional inactivation (PrD25A or omitting Gag-Pol (A) under non-saturating conditions (8x dilution). Experiments were performed on the ‘blue’ mGL HEK293T line.

[0653] (D) Benefit of CCGCN4 on chimeric and homomeric ENVLPEcsy4 assemblies assessed under nonsaturating conditions (8x dilution). Experiments were performed on the ‘blue’ mGL HEK293T line. For (A), (C), and (D), selected results are shown from one-way ANOVA with Bonferroni MCT (**** < 0.0001; ***p < 0.001; **p < 0.01).

[0654] Figure 28 Assembly and RNP recruitment mechanisms in ENVLPE and eVLP variants.

[0655] The Cas9 effector protein binds the stabilized gRNA-Cas-effector ribonucleoprotein (RNP) complex in the nucleus; nucleocytosolic shuttling (NLS-NES) of Gag-PCP enables binding of the fully assembled RNP via PCP-PP7 interaction. The system enables modular packaging of various RNP complexes and only requires modification of the (pe)gRNA with a PP7 aptamer. Csy4 cleaves and protects the pegRNA in the RNP complex. v4 eVLPs consist of an MMLV-Gag fusion to a genome-editing effector protein of choice. The fusion protein shuttles between the nucleus and cytosol to package the unprotected gRNA. In contrast to ENVLPE+, packaging of the Cas a o-enzyme (without gRNA) is possible; the genome editor is released after budding only upon proteolytic processing of the Gag polyprotein by the protease encoded in MMLV Gag / Gag-Pol.

[0656] Figure 29 Characterization of RNP loading into different VLP systems, increasing ENVLPE titers by switching from CAG to CMV promoter-driven plasmids, and the benefit ofPEcsy4 forPE- eVLP -mediated RNP delivery.

[0657] (A) Quantification of Cas9 molecules per VLP matched with pegRNA molecules per VLP for ENVLPE and v3(b) PE-eVLPs. VLPs produced in 15 cm dishes were purified by ultracentrifiigation and resuspended in 1 / 1000thof the original volume prior to analysis.

[0658] (B) Corresponding Cas9:pegRNA packaging ratios.

[0659] (C) Optimization of ENVLPE+plasmid ratios at non-saturating conditions (8x dilution) after switching ENVLPE+component expression to CMV promoter-driven plasmids. The chosen ratio for the benchmarking experiments in Figure 15B-D is indicated in dark green.

[0660] (D) To test the compatibility of the Csy4 / C4 protection system with v3 / v3b PE-eVLPs, Csy4 was supplied as a co-expressing plasmid from a separate plasmid for the PE-eVLP systems. Experiments were performed on the ‘blue’ mGL HEK293T line. Transduction was performed at 2x dilution. Selected results are shown of Bonferroni MCT after one-way ANOVA analysis (****p < 0.0001; ***p < 0.001; **p < 0.01).

[0661] Figure 30 Delivery of CreNLs mRNA to activate a red / green-reporter locus. (A) Indicated volumes of Crc i s mRNA-loaded ENVLPE were injected subretinally into the eyes of mT / mG reporter mice (2.53 x 108particles pE1in PBS). Two-photon excitation imaging-based 3D reconstruction of the posterior portion of intact mouse eyes are shown, 1 week after subretinal injection. Green fluorescent cells indicate successful reporter activation by Cre in the red / green- reporter locus.

[0662] (B) Quantification of Crc-,i s mRNA delivery by ENVLPE vs. ENVLPE+to a HEK293T reporter cell line carrying a green / red-reporter locus (inverse signal logic compared to (A)). Log2(VreiSN): relative transduction volumes in 2x dilution series.

[0663] Figure 31 Quantification ofVLP titers using anti-p24 (HIV-1) or anti-p30 (MMLV) ELISA.

[0664] (A) Quantification of ENVLPE+and v3 PE-eVLP titers after ultracentrifugation prior to subretinal injections. The VLP amount per volume was calculated based on the assumption that 2500 HIV-1 Gag monomers (or 1800 MMLV Gag monomers) form the capsid.

[0665] (B) Assessment of VLP titers for ex vivo base editing of T lymphocytes.

[0666] Figure 32 Correlation of in vivo editing of Rpe65 in rd!2 mice with corresponding h-waves.

[0667] Correlation plot of Rpe65 prime editing efficacy and functional restoration quantified by 6-wave amplitudes in response to a light stimulus after ENVLPE+-mediated iPEcsy4 delivery; points correspond to the subset of matching pairs of data points also shown in Figure 171 and Figure 17K.

[0668] Figure 33 (A) Schematic illustration of the PPP1R12C (AAVS1) locus. The insertion cassette encodes the ‘blue’ mGL reporter as well as a Puromycin N-acetyltransferase.

[0669] (B) Analysis of copy number variation (CNV) in the context of the iPSC genotype.

[0670] (C) Example immunostainings of iPSC cell markers (POU5F1, NANOG, and SOX2) in the ‘blue’ mGL cell line after CRISPR / Cas9-mediated knock-in. Scale bar = 200 pm.

[0671] (D) Immunostaining to show the ability of the generated cell line to differentiate into all three lineages. Scale bar = 50 pm.

[0672] (E) Immunostaining of the cell line after differentiation into cortical neurons. Scale bar = 50 pm.

[0673] Figure 34 (A) Illustration illustration of the PPP1R12C (AAVS1) locus for insertion of the enhanced traffic light reporter (eTLR)v2. A puromycin '-accty 1 transferase cassette mediates resistance to antibiotic selection.

[0674] (B) Analysis of copy number variation (CNV) in the context of the iPSC genotype.

[0675] (C) Example immunostainings of pluripotency markers (POU5F1, NANOG, and SOX2) in the ‘eTLRv2 iPSC, after CRISPR / Cas9-mediated knock-in. Scale bar = 200 pm.

[0676] (D) Immunostaining to show the ability of the generated cell line to differentiate into all three lineages. Scale bar 50 = pm. (E) Immunostaining of the cell line after differentiation into cortical neurons. Scale bar = 50 pm.

[0677] Figure 35

[0678] (A) Left: Schematic depiction of a generic VLP composed of a membrane lipid bilayer decorated with fusogen / glycoproteins (dotted area) to enable the fusion of the vesicle with a recipient cell that expresses a suitable membrane protein that is recognized by the fusogen / gly coprotein. The hatched area indicates the potential region that can be harnessed to load a desired payload, such as a geneediting RNP. The black ring indicates the homo-oligomerization domain of an engineered polypeptide chain that forms the central core domain of such a membranous vesicle. The homo-oligomerization may be optionally nucleated / scaffolded by an auxiliary scaffolding protein.

[0679] Right: Schematic depiction of the domain organization of such a polypeptide to organize such a membranous enclosed vesicle to package custom payloads; glycoproteins / fusogens are not displayed but are always part of such a vesicle that is designed to deliver a custom payload to a recipient cell. Besides the homo-oligomerization domain, as depicted on the left, other domains are required to achieve budding and packaging of gene-editing RNPs. The respective domains / fimctions are indicated.

[0680] (B) Oligomerized protein spheres can also bud off the membrane as multi-sphere assemblies that form together one single enclosed membranous vesicle.

[0681] (C) One design example is a fusion protein comprising the P22 phage gp5, XTEN linker, GCN4 mutant that preferentially forms trimers over dimers (GCN4*), a nucleocytosolic shuttling motif (NCS). PP7 coat protein (PCP), HIV-1 p6, HiBiT (that only has a reporter function for LgBiT complementation assays that has no relevance for budding or cargo loading), and a membrane-tether domain PLC-H (pleckstrin homology domain).

[0682] (D) The effect of the different designs on the delivery efficacy for prime editing RNPs quantified by FACS of the ‘blue’ mGL reporter HEK293T cell line. Bars represent mean ± SD (n = 3 biological replicates). P22 gp8 indicates the scaffolding protein that is or is not fused to PCP-NLS-NES, whereas NLS-NES indicates the nucleocytosolic shuttling motif (NCS). P22 gp8 may help nucleating gp5 homo-oligomerization and thus its assembly. XTEN-GFN4 indicates the presence (+) or absence (-) of XTEN-GCN4* in the P22 gp5 fusion proteins depicted in (C). AVSV-G indicates the condition where no glycoproteins were transfected into the producer cells. All other conditions were always cotransfected with plasmids encoding VS V-G glycoprotein to enable vesicle fusion with the recipient cells.

[0683] (E) The effect of the different designs on the delivery efficacy for prime editing RNPs quantified by FACS of the ‘blue’ mGL reporter HEK293T cell line. Bars represent mean ± SD (n = 3 biological replicates). Gp5 and gp8 were fused or not fused to NES, NCS-PCP motifs; XTEN-GFN4 indicates the presence (+) or absence (-) of XTEN-GCN4* in the P22 gp5 fusion proteins. All conditions were always co-transfected with plasmids encoding VSV-G glycoprotein to enable vesicle fusion with the recipient cells. (F) The effect of the different designs on the delivery efficacy for prime editing RNPs quantified by FACS of the ‘blue’ mGL reporter HEK293T cell line. Bars represent mean ± SD (n = 3 biological replicates). XTEN-GFN4 indicates the presence (+) or absence (-) of XTEN-GCN4* in the P22 gp5 fusion proteins. NCS (+) or (-) indicates the presence or absence of a nucleocytosolic shuttling motif. All conditions were always co-transfected with plasmids encoding VSV-G glycoprotein to enable vesicle fusion with the recipient cells.

[0684] (G) The effect of the different designs on the delivery efficacy for prime editing RNPs quantified by FACS of the ‘blue’ mGL reporter HEK293T cell line. Bars represent mean ± SD (n = 3 biological replicates). XTEN-GFN4 indicates the presence (+) or absence (-) of XTEN-GCN4* in the P22 gp5 fusion proteins. PCP and tdPCP indicate that either a single or tandem (td) copy of PCP was fused to gp5. AVSV-G indicates the condition where no glycoproteins were transfected into the producer cells. All other conditions were always co-transfected with plasmids encoding VSV-G glycoprotein to enable vesicle fusion with the recipient cells.

[0685] Figure 36 Representative cryo-EM micrographs of the budded membrane-enveloped vesicles from producer cells transfected with plasmids expressing the indicated proteins.

[0686] Figure 37

[0687] (A) Left: Schematic depiction of a generic VLP composed of a membrane lipid bilayer decorated with fusogen / glycoproteins (dotted area) to enable the fusion of the vesicle with a recipient cell that expresses a suitable membrane protein that is recognized by the fusogen / gly coprotein. The hatched area indicates the potential region that can be harnessed to load a desired payload, such as a geneediting RNP. The black ring indicates the homo-oligomerization domain of an engineered polypeptide chain that forms the central core domain of such a membranous vesicle. The homo-oligomerization may be optionally nucleated / scaffolded by an auxiliary scaffolding protein.

[0688] Right: Schematic depiction of the domain organization of such a polypeptide to organize such a membranous enclosed vesicle to package custom payloads; glycoproteins / fusogens are not displayed but are always part of such a vesicle that is designed to deliver a custom payload to a recipient cell. Besides the homo-oligomerization domain, as depicted on the left, other domains are required to achieve budding and packaging of gene-editing RNPs. The respective domains / functions are indicated. In contrast to Figure 35, alpha-helical rigid linkers of varying lengths are used.

[0689] (B) One design example is a fusion protein comprising the P22 phage gp5, multiple EAAAR repeats that act as alpha-helical rigid linkers, and a nucleocytosolic shuttling motif (NCS). PP7 coat protein (PCP), HIV-1 p6, HiBiT (that only has a reporter function for LgBiT complementation assays that has no relevance for budding or cargo loading), and a membrane-tether domain PLC-H (pleckstrin homology domain). (C) The effect of the different designs on the delivery efficacy for prime editing RNPs quantified by FACS of the mGL reporter HEK293T cell line. Bars represent mean ± SD (n = 3 biological replicates). XTEN-GCN4 indicate the reference condition where a flexible XTEN linker was used together with GCN4* . - indicate the absence of any linkers. 4L4R indicates that 4 repeats of EAAARx4 are inserted left of the PCP, and 4 repeats of EAAARx4 are inserted right of the PCP. All other nomenclatures follow the same logic.

[0690] Figure 38

[0691] (A) Left: Schematic depiction of a generic VLP composed of a membrane lipid bilayer decorated with fusogen / glycoproteins (dotted area) to enable the fusion of the vesicle with a recipient cell that expresses a suitable membrane protein that is recognized by the fiisogen / gly coprotein. The hatched area indicates the potential region that can be harnessed to load a desired payload, such as a geneediting RNP. The black ring indicates the homo-oligomerization domain of an engineered polypeptide chain that forms the central core domain of such a membranous vesicle. Note that, in contrast to Figure 35 and Figure 37, the intended packaging area is within the oligomerization core.

[0692] Right: Schematic depiction of the domain organization of such a polypeptide to organize such a membranous enclosed vesicle to package custom payloads; glycoproteins / fusogens are not displayed but are always part of such a vesicle that is designed to deliver a custom payload to a recipient cell. Besides the homo-oligomerization domain, as depicted on the left, other domains are required to achieve budding and packaging of gene-editing RNPs. The respective domains / fimctions are indicated.

[0693] (B) One design example is a fusion protein comprising the PCP and variants thereof, XTEN linker, GCN4 mutant that preferentially forms trimers over dimers (GCN4*), a nucleocytosolic shuttling motif (NCS). PP7 coat protein (PCP), HIV-1 p6, HiBiT (that only has a reporter function for LgBiT complementation assays that has no relevance for budding or cargo loading), and a membranetether domain PLC-H (pleckstrin homology domain). PCP fulfills here at least two roles: it serves as an aptamer-binding protein and also as a homo-oligomerization domain since PCP is derived from the PP7 phage coat protein that is able to form capsids on its own.

[0694] (C) The effect of the different designs on the delivery efficacy for prime editing RNPs quantified by FACS of the ‘blue’ mGL reporter HEK293T cell line. Bars represent mean ± SD (n = 3 biological replicates). Indicated presence or absence of various motifs, as depicted in (B), are indicated. PCP and tdPCP indicate that either a single or tandem (td) copy of PCP was fused to gp5. FG loop indicates the presence or absence of the “FG loop (Tars et al., 2000, doi: 10.1006 / viro.2000.0373)” in tdPCP / PCP. WT PCP contains are “FG loop” that mediates dimer-dimer interaction; when using PCP only as an aptamer-binding protein, the “FG loop” is usually inactivated by deletion or mutation to prevent PCP dimers binding to other PCP dimers and thus multimerization. Mixed indicates that half of the plasmids transfected contain FG-active loops, while the other half does not. AVSV-G indicates the condition where no glycoproteins were transfected into the producer cells. All other conditions were always co-transfected with plasmids encoding VSV-G glycoprotein to enable vesicle fusion with the recipient cells.

[0695] EXAMPLES

[0696] EXAMPLE 1 Materials and Methods

[0697] 1. Molecular Cloning and DNA analysis

[0698] 1.1 PCR

[0699] Single-stranded primer deoxyribonucleotides (Integrated DNA Technologies (IDT)) were resolubilized 100 pM) in nuclease-free water. PCR reaction with plasmid and genomic DNA templates was performed with Platinum SuperFi II PCR Master Mix (Thermo Fisher Scientific) according to manufacturer’s protocol. PCR reactions were purified by DNA agarose gel electrophoresis and subsequent DNA extraction using Monarch® DNA Gel Extraction Kit (New England Biolabs (NEB)).

[0700] 1.2 DNA digestion with restriction endonucleases

[0701] Samples were digested with NEB restriction enzymes according to the manufacturer’s protocol in a total volume of 40 pl with 1-3 pg of plasmid DNA. Afterwards, DNA fragments were purified by agarose gel-electrophoresis and subsequent purification using Monarch® DNA Gel Extraction Kit (NEB).

[0702] 1.3 Ligation and Gibson assembly

[0703] Concentrations of agarose-gel purified DNA fragments were determined by a spectrophotometer (NanoDrop 1000, Thermo Fisher Scientific). Ligations were carried out with 50-100 ng backbone- DNA (DNA fragment containing the ori) in 20 pl volume, with molar 1: 1-3 backbone: insert ratios, using T4 DNA ligase (Quick Ligation™ Kit, NEB) at room temperature for 5-10 min. Gibson assemblies were performed with 75 ng backbone DNA in a 15 pl reaction volume and molar 1: 1-5 backbone: insert ratios, using NEBuilder® HiFi DNA Assembly Master Mix (2x) (NEB) for 20-60 min at 50 °C.

[0704] 1.4 DNA agarose gel-electrophoresis

[0705] 1% (m / m) agarose (Agarose Standard, Carl Roth) gels were prepared in lx TAE-buffer and 1: 10.000 SYBR Safe stain (Thermo Fisher Scientific). Gel electrophoreses were carried out for 20-40 min at 100 V. For size determination, 1 kb Plus DNA Ladder (NEB) was used. DNA samples were mixed prior to loading with Gel Loading Dye (Purple, 6x) (NEB).

[0706] 1.5 Bacterial strains (E. coli) for molecular cloning Chemically competent Stable cells (NEB® Stable) were used for transformation of circular plasmid DNA. For plasmid amplification, carbenicillin (Carl Roth) was used as a selection agent at a final concentration of 100 pg / ml. All bacterial cells were incubated in Lysogeny Broth (LB) medium or on LB agar plates including the respective antibiotics.

[0707] 1.6 Bacterial transformation with plasmid DNA

[0708] Transformation was performed by mixing 1-5 pl of ligation or Gibson reaction with 50 pl thawed, chemically competent cells, which were incubated on ice for 30 min before heat shocking at 42 °C for 30 s. Afterwards, cells were incubated on ice for 5 min and finally mixed with 450 pl SOC-medium (NEB). Transformed cells were then plated on agar plates containing an appropriate type and concentration of antibiotics according to the supplier’s information. Plates were incubated overnight at 37 °C.

[0709] 1.7 Plasmid DNA purification and Sanger sequencing

[0710] E. coll colonies with correct potential constructs were inoculated from agar plates in 2 ml LB medium at 37 °C with the respective antibiotics and incubated for at least 6 h or overnight. Plasmid DNA was extracted with Monarch® Plasmid Miniprep Kit (NEB) according to the manufacturer’s protocol and sent out for Sanger sequencing (GENEWIZ, Azenta Life Sciences). Sanger sequencing validated clones were inoculated in 100 ml LB medium overnight at 37 °C containing the respective antibiotic selection agent. Plasmid DNA was extracted using the Plasmid Maxi Kit (QIAGEN).

[0711] 1.8 Genomic DNA isolation

[0712] 72 hours after transfection or transduction in 96-well format, genomic DNA was isolated with the Quick-DNA 96 Kit (Zymo Research) according to the manufacturer’s protocol with an elution volume of 30 pl.

[0713] 1.9 Amplicon PCR and purification

[0714] PCR was performed as described above using ~50 ng of gDNA and appropriate primers for each target. Amplicon lengths were designed to approach 250 bp for sequencing. PCR purification was performed using the DNA Clean & Concentrator-5 Kit (Zymo Research) according to the manufacturer’s protocol with an elution volume of 30 pl

[0715] 1.10 Amplicon sequencing and analysis

[0716] Following an initial PCR on the genomic DNA, a second outer PCR using barcoded primers was performed. PCR products of each experiment were purified as described above, normalized, and pooled. The mixture was gel-purified, normalized to 20 ng / pl, and submitted for Amp-EZ sequencing (GENEWIZ, Azenta Life Sciences). The resulting fastq files containing paired reads were analyzed with Geneious via barcode separation and CRISPR editing analysis within the entire range covered by reads or at least the full sequence area between the genomic primer binding sites.

[0717] 2. Mammalian cell culture

[0718] 2.1 Cell lines and cultivation

[0719] Cells were cultured at 37 °C, 5% CO2, and H2O saturated atmosphere.

[0720] HEK293T (ECACC: 12022001, Sigma-Aldrich) and HEK293 eTLR (originate from HEK293, ECACC 85120602) cells were maintained in advanced DMEM (Gibco™, Thermo Fisher Scientific) supplemented with 10% FBS (Gibco™, Thermo Fisher Scientific), GlutaMAX (Gibco™, Thermo Fisher Scientific), 100 pg / ml Penicillin-Streptomycin (Gibco™, Thermo Fisher Scientific), 10 pg / ml Piperacillin (Sigma- Aldrich) and 10 pg / ml Ciprofloxacin (Sigma- Aldrich). Cells were passaged at 90% confluence by removing the medium, washing with DPBS (Gibco™, Thermo Fisher Scientific) and detaching the cells with Accutase® solution (Gibco™, Thermo Fisher Scientific). Cells were then incubated for 5-10 min at room temperature until a visible detachment of the cells was observed. Accutase™ was subsequently inactivated by addition of pre-warmed DMEM including 10% FBS and all supplements. Cells were then transferred into a new flask at an appropriate density or counted and plated on 96-well or 6-well format for plasmid transfection or VLP transduction.

[0721] Jurkat E6.1 cells (ECACC 88042803) were maintained in RPMI media 1640 (Gibco™, Thermo Fisher Scientific) with GlutaMAX™ (Gibco™, Thermo Fisher Scientific) and 100 pg / ml Penicillin- Streptomycin (Gibco™, Thermo Fisher Scientific). For transduction, they were seeded to 96-well format at 25,000 cells / well. The proper density was achieved by centrifugation 200 ref for 2 min and subsequent dilution RPMI 1640 to the desired density.

[0722] 2.2 hiPSC cultivation hiPSCs were invariantly incubated in Essential 8 Flex medium (Gibco, Thermo Fisher Scientific) at 37 °C under 5% CO2 saturation on (v / v) Vitronectin-coated (A31804, Thermo Fisher Scientific) plates. At 70% confluency, the cells were subcultivated with StemMACS Passaging Solution XF (Miltenyi Biotec) by incubating the cells for 6 min at room temperature. Subsequently, StemMACS Passaging Solution XF was aspirated, and cells were collected by adding 1 ml Essential 8 Flex medium to the cells. Collected cells were transferred to a flask containing 2 ml Essential 8 Flex medium and seeded on 96-, 48-, 24- or 6-well plates at the appropriate cell densities.

[0723] 2.3 hiPSC-derived cortical neurons hiPSCs were plated on multiwell plates coated with 1% (v / v) Geltrex (A1413302, Gibco, Thermo Fisher Scientific) 24 h before neuronal initiation, maintained in Essential 8 medium (A2858501, Gibco, Thermo Fisher Scientific). For neuronal initiation, smNPC maintenance medium — a 1: 1 composition of N2 and B27 medium supplemented with 100 ng pF1FGF8b (130-095-740, Miltenyi Biotec), 100 nM LDN193189 (130-103-925, Miltenyi Biotec) and 10 pM SB431542 (S1067, Selleck Chemicals) — was added to the cells. N2 medium was composed of DMEM / F12 + GlutaMax (10565018, Gibco, Thermo Fisher Scientific), l x N2 supplement (17502048, Gibco, Thermo Fisher Scientific), 5 pg mF1insulin (I9278-5ML, Sigma-Aldrich), 100 pM MEM non-essential amino acids (11140035, Gibco, Thermo Fisher Scientific), 50 pM p-mercaptoethanol (M6250, Sigma- Aldrich) and 0.5x (50 U mF1) penicillin-streptomycin (15140122, Gibco, Thermo Fisher Scientific). B27 medium was composed of neurobasal medium (21103049, Gibco, Thermo Fisher Scientific), 1 x B27 supplement (11530536, Gibco, Thermo Fisher Scientific), 1 x GlutaMax Supplement (35050038, Life Technologies) and 0.5 x (50 U mF1) penicillin-streptomycin. On day 14, the neuroepithelial sheet was replated onto a six-well plate, coated with 15 pg mF1poly-L-omithine hydrobromide and 10 pg mF1laminin. On day 21, medium was aspirated, and cells were washed once with Dulbecco’s phosphate-buffered saline (DPBS) (14190169, Gibco, Thermo Fisher Scientific). Cells were detached using Accutase (6964, Sigma-Aldrich) under incubation for 10 min at 37 °C. The Accutase reaction was stopped by transferring the cells into a 15 ml Falcon containing 6 ml DMEM / F-12, GlutaMax. Cells were sedimented by centrifugation at 200 r.c.f. for 5 min, and cell pellets were resuspended in smNPC maintenance medium supplemented with 100 pM Y-27632 dihydrochloride (ALX-270-333- M005, Enzo Life Sciences) and counted using a Neubauer improved, precision disposable plastic hemocytometer C-chip (PK361, Carl Roth). Appropriate cell densities were seeded and cells were cultured in smNPC maintenance medium supplemented with 100 pM Y-27632 dihydrochloride for 24 h at 37 °C under invariant 5% CO2 saturation. The next day, the medium was aspirated and neuronal differentiation was initiated by adding neuronal induction medium - a 1 : 1 composition of N2 and B27 medium, supplemented with 200 pM ascorbic acid (10389701, Th. Geyer) and 20 ng mF1brain-derived neurotrophic factor (130-093-811, Miltenyi Biotec). In the case of detachment of the neuronal network, cells were reseeded onto new plates coated with 15 pg mF1poly- L-omithine hydrobromide (P3655, Sigma-Aldrich) and 10 pg mF1laminin (L2020-1MG, Sigma- Aldrich). Neuronal networks were dissociated into single cells using Accutase. Cells were then carefully washed once with DPBS, and incubated with Accutase for 15 min at 37 °C under invariant 5% CO2 saturation. The Accutase reaction was stopped by transferring the cell solution into a 15 ml Falcon tube, containing 3 ml neuronal induction medium supplemented with 100 pM Y-27632 dihydrochloride. Cells were released from Y-27632 dihydrochloride by a medium change 24 h after reseeding.

[0724] 2.4 Plasmid transfection for VLP production

[0725] HEK293T cells were seeded on a 6-well plate in 3 ml advanced DMEM with a density of 200,000 cells / ml. One day post seeding, the cells were transfected with the packaging plasmid (333 ng), pseudotyping plasmid (200 ng) and the cargo plasmids. The first cargo plasmid encodes for the PP7 aptamer-tagged (pe)gRNA (500 ng) as well as a suitable genome editing effector protein (167 ng). The lipofection was carried out according to manufacturer’s protocol (jetOPTIMUS, Polyplus). DNA amounts were kept constant in all transient experiments to yield reproducible complex formation and comparable results. In 6-well plates, a total amount of 1.2 pg of plasmid DNA was used per well. After 48-72 h cultivation of the producer cells, the supernatant (SN) was collected.

[0726] 2.5 Isolation and transduction of VLPs

[0727] The supernatant (SN), containing the VLPs, was sterile -fdtered to remove cell debris and remaining lipofection reagents. Subsequently, the filtered SN was concentrated to approximately 1 / 30 of the original volume using a molecular weight cutoff filter (100 kDa). For pseudo-transduction (TD), the respective reporter cell line was seeded on a 96-well plate (25,000 cells in 200 pl advanced DMEM). 6h post seeding, the supernatant of one 6-well production was applied to a triplicate of receiver cells on the 96-well plate without exceeding 20% of the total volume of the receiver cell’s original growth medium. For optimizing the VLP system, cells were analyzed 72 h post TD when delivering an RNP complex and 24h post TD when delivering mRNA.

[0728] Storage capabilities of the VLPs vary depending on pseudotyping. VSV-G pseudotyped VLPs can be stored at 4 °C for at least three days without significant loss of infectivity. For other pseudotyped VLPs the inventors recommend to shock-freeze the VLPs using liquid nitrogen.

[0729] 2.6 Generation of stable cell lines carrying the fluorescent prime editing reporter via CRISPR / Cas9

[0730] Cell lines were created by cloning the reporter coding for the blue-shifted mGreenlantem2 mutant (G65S, Y66H) with a CAG-promoter and a bovine growth hormone (bGH) polyadenylation signal (pA) was cloned with homology arms for a safe harbour locus of choice. To create a HEK293T reporter cell line, cells were transfected with the CRISPR donor plasmid as well as a second plasmid encoding Cas9 and the corresponding gRNA. Cells were transfected according to the manufacturer's protocol (jetOPTIMUS, Polyplus) 24 hours post-seeding on a 6-well plate (600k cells in 3 ml per well) in the presence of 0.5 pM AZD7648 (UY-111783; MedChemExpress), a DNA-PCcs inhibitor, and a CAG-promoter-driven i53, a 53BP1 inhibitor, to inhibit NHEJ and shift the DNA repair towards HDR (see also Truong etal., 2022; Truong etal., 2023). The surviving polyclonal population was monoclonalized using limiting dilution in 96-well plates, selected clones were genotyped to determine homozygosity.

[0731] 2. 7 Immunoblot analysis

[0732] Cells were lysed with the appropriate volume of M-PER (Thermo Fisher Scientific), including protease inhibitors (Halt Protease Inhibitor Cocktail, Thermo Fisher Scientific), according to the manufacturer’s protocol. Cleared lysates were then equalized against the relative protein concentration determined using the NanoDrop 1000 (Thermo Fisher Scientific) and diluted with M-PER. Equalized lysates were prepared for SDS-gel-electrophoresis using XT sample buffer (Bio-Rad) and XT reducing agent (Bio-Rad) and denatured at 70 °C for 10 min or 95 °C for 5 min. Samples were loaded in 18-well 4-12% Criterion XT Bis-Tris Protein Gel, and electrophoresis was run at 200 V for 1 h in XT MOPS running buffer (Bio-Rad). Molecular mass markers (Amersham ECL Full-Range Rainbow Molecular Weight Markers, GE Healthcare Life Sciences, General Electric) were used in parallel to confirm the different protein species. Subsequently, an immunoblot was performed onto an Immobilon-P PVDF membrane (Merck) with a wet blotting system (Criterion Blotter, Bio-Rad) in ice- cold Towbin buffer (Bio-Rad) with 20% (v / v) methanol (Carl Roth) at 100 V for 30 minutes in an ice water reservoir or overnight (15 V at 4 °C). Next, the free valences on the PVDF membrane were blocked in blocking buffer containing 5% (m / v) skimmed milk (Carl Roth) in TBS-T (pH 7.6) with 0.1% (v / v) Tween-20 (Sigma- Aldrich) at room temperature for 1 h. Antibodies were diluted at the indicated dilution factors in blocking buffer and either incubated at room temperature for 2 h or overnight at 4 °C, followed by at least three washing steps (room temperature, 5 min, 60 r.p.m.). The HRP-conjugated secondary antibody (Abeam) was also diluted in blocking buffer (1:20,000) and subsequently washed again with TBS-T at least four times. HRP detection was performed using the SuperSignal West Femto Maximum Sensitivity Substrate or SuperSignal West Atto Ultimate Sensitivity Substrate (Thermo Fisher Scientific) on the Fusion FX7 / SL advanced imaging system using FusionCapt Advance SL4 v. 16.09b (Vilber Lourmat). The primary and secondary antibodies used were as follows: HIV-1 p55 Gag (full length) antibody (1:2,000, #ABIN2452023, antibodies- online GmbH), AC-15 mouse anti-beta-actin (HRP-coupled, 1: 100,000, ab6276, Abeam), goat antimouse IgG H&L (HRP-coupled, 1:20,000, ab97023, Abeam), goat anti-rat IgG H&L (HRP-coupled, 1:20,000, ab97057, Abeam) and goat anti-rabbit IgG H&L (HRP-coupled, 1:20,000, ab6721, Abeam). Quantification was performed using Image Lab (v.6. 1.0 build 7, Bio-Rad).

[0733] 2.8 FACS analysis.

[0734] FACS analysis was performed on the BD FACSaria II system (controlled with the BD FACSDiva Software (v.6.1.3, BD Biosciences)). Briefly, the main population of the cells was gated first according to their FSC-A and SSC-A. Secondly, single cells were gated using FSC-A and FSC-W. The final gate (green fluorescence) was used to determine the proportion of successfully edited / transduced cells.

[0735] 2.9 Biolumin escen ce qu antification

[0736] For bioluminescence bulk quantifications, cells were plated (and pseudo-transduced) in 96-well format. For bioluminescence detection of secreted NLuc, the supernatant was sampled (80 pL) at the indicated time points and detected using the Nano-Gio® Luciferase Assay System (Promega) on the Centro LB 960 (Berthold Technologies) plate reader with 0.5 s acquisition time. For dual-luciferase read-out using Nano-Gio® Dual-Luciferase® Reporter Assay System (Promega), NLuc and FLuc signals were read out on-plate 48 hours post-transfection. Signals were obtained with 0.5 s acquisition time 10 min after the addition of reagent 1 (ONE-Glo™ EX Luciferase) for FLuc and 10 min after addition of reagent 2 (NanoDLR™ Stop & Gio®) for NLuc. Reagent 2 includes an FLuc inhibitor.

[0737] 2.10 Statistics

[0738] Statistics were calculated with Prism 9 (Graphpad) as indicated.

[0739] 2.11 Protein domains

[0740]

[0741] 2.12 Proteins and protein chimeras

[0742] EXAMPLE 2: Further experiments

[0743] Delivery of Cas9 transactivators and base editors

[0744] The inventors wondered whether functional CRISPR transactivators such as dCas9-miniVPR RNPs could be delivered using ENVLPE to induce MAPT expression in a HEK293T reporter cell line, which reports MAPT expression as a firefly luciferase (FLuc) activity. The inventors observed a significant 2- fold increase in reporter signal 48 hours after CRISPRa-ENVLPE transduction, indicating successful gene activation (Figure 20B).

[0745] The inventors sought to extend ENVLPE to base editing (BE), a double-strand-break-free and therapeutically promising gene editing technology. The inventors first evaluated ENVLPE-mediated base editing in their HEK293-TLR reporter cell line, which can also report the stop codon removal via a>g editing as green fluorescence, and observed that 35% of the cells were successfully edited (Figure 5E). The inventors next tested BE ENVLPE on the 2A7 locus (Figure 5F,G). In multiple common cell types, where BE ENVLPE was able to facilitate base editing with high efficacy at the B2M locus in all of the cell lines tested, reaching 90% in iPSC-derived cortical neurons, and > 55% in hiPSCs (Figure 5G). The latter cells are typically challenging to access with conventional techniques such as lipofection or electroporation, which require tedious optimization.

[0746] 3’ protection of pegRNA with Csy4 increases editing efficacy

[0747] Compared to gRNAs, pegRNAs particularly have a short half-life because of their degradation-prone 3’ extension. Al -though their current design restricts the packaging of PE RNPs to those with intact 3’ ends that are protected by the evopreQ 1 pseudoknot, the inventors hypothesized that pegRNA stability may still limit the full potential of PE-ENVLPE. Accordingly, the inventors harnessed the CRISPR protein Csy4 (alias Cas6f) from Pseudomonas aeruginosa, which binds to a highly conserved 16 nt RNA motif (Csy4 motif, C4), wherein the nucleotides 2-16 form a hairpin with a 5 bp stem and a 5 nt loop. Csy4’s processing activity has been used extensively in various CRISPR / Cas9 systems, e.g., to process arrays of gRNAs / crRNAs into individual gRNA / crRNAs for multiplexed gene editing, or to enhance PE efficiency by preventing the pegRNA from folding back on itself due to the complementarity between the spacer and PBS region. Importantly, the C4 RNA aptamer is bound by Csy4 with an exceptionally high affinity (K ~50 pM) that is ~20-fold higher than the PCP-PP7 interaction (K ~1 nM). The C4 aptamer is precisely cleaved 3’ proximal to the hairpin without leaving any unpaired nucleotides as 3’-overhang. Csy4 remains bound with an unchanged affinity and thus may shield the 3 ’ end from a potential 3 ’ exonucleolytic attack.

[0748] Although C4-modified pegRNAs showed no additional benefits in transient transfection experiments in which pegRNA levels are naturally elevated (Figure 21A), the inventors hypothesized that Csy4- mediated 3 ’ protection of the pegRNA could prove advantageous in conditions where RNA stability is limiting, such as during RNP delivery (Figure 13A). When Csy4 is co-expressed in the VLP- production cells, recruitment of the RNP complex is still mediated by the PCP-PP7 system and the 3' end of the pegRNA of the inventors’ initial setup contains an additional C4 motif downstream of the essential elements (RTT, PBS, and PP7). The editing efficacy in the recipient cells was strongly improved when Csy4 was co-expressed in the producer cells, either as a direct fusion to iPE or as an untethered module via a 2A peptide (iPEcsy4) (Figure 13B), indicating that the increased efficacy is solely due to the 3' protective effect of Csy4. Notably, the median fluorescence intensity (MFI) was strongly elevated for cells edited in the presence of Csy4, indicating either multi-allelic editing, faster editing kinetics, or a combination of both. The inventors further observed that expressing Csy4 in an untethered fashion resulted in slightly higher % x MFI values than the fusion of Csy4 to iPE (Figure 13B). Interestingly, the inventors observed an almost complete loss of editing activity in the control condition without Csy4, suggesting that unprocessed 3'-terminal nucleotides downstream of the C4 aptamer are highly susceptible to degradation (Figure 13B).

[0749] Evaluation of pegRNA 3’ end motifs

[0750] The effectiveness of PCP grafting into Gag was evaluated at different grafting positions to confirm that the observed improvement in protective Csy4 efficacy was not due to compensation for potentially suboptimal PCP grafting (Figure 21B). Alternative aptamer recruitment systems, such as Com (Figure 21C) or Csy4 alone, as a dual-purpose protection and recruitment module (Figure 22), did not provide additional benefits.

[0751] Therefore, the inventors chose the 3' configuration 'PP7-C4-Q1' as the preferred mode for iPECsy4 because it can be used without Csy4 and thus provides the most flexibility (Figure 23A).

[0752] The inventors next compared iPEcsy4 with PE7, which utilizes the N-terminal fragment of the La protein fused to the prime editor to protect the 3 ’ ends of pegRNAs by binding to the 3 ’ polyU nucleotides, a scar left by RNA polymerase III termination. With ENVLPE-mediated delivery, iPEcsy4 clearly outperformed PE7 regarding editing of the mGL reporter line (Figure 13C). PE7-ENVLPE performance was increased by adding a tevopreQ 1 motif to the pegRNA 3 ’ end but still displayed lower editing rates compared to iPEcsy4 ENVLPE (Figure 23B). Thus, the Csy4 / C4 3'-protection module of one embodiment of this invention is clearly advantageous over evopreQl and the 3’- protection system PE7.

[0753] Evaluation of cell viability postPEcsy4 delivery and potential application of Csy4 for canonical gRNAs

[0754] To evaluate the potential toxicity of Csy4 due to its RNA processing activity, the inventors investigated the effect of PEcsy4 transductions via ENVLPE on cellular fitness in recipient HEK293T and iPS cells. However, the inventors could not find relevant differences in cell viability compared to the non-VLP control condition and no difference in cytotoxicity compared to other VLP-transduced conditions lacking Csy4 (Figure 24A,B), reinforcing the platform's safety profile. The inventors conclude that previous reports of Csy4’s cytotoxic effect arise from circumstances where Csy4 is excessively overexpressed in transient transfection settings.54Alternative strategies in which the protection and processing of the pegRNA 3’ end are separated into distinct processes did not prove to be beneficial (see Figure 25).

[0755] The inventors further tested whether Csy4-mediated 3 ’-stabilization of pegRNAs would also translate to conventional gRNA-based effectors that do not carry 3 ’-extensions. Likely due to the absence of 3’- vulnerable ends, gRNA-based nuclease and base-editing systems do not benefit from additional 3’- protection (Figure 26). In summary, Csy4 expressed via a 2A peptide with iPE (PEcsy4), combined with ‘PP7-C4-Q1’ pegRNAs, substantially improves prime editing efficacy in ENVLPE -mediated RNP delivery.

[0756] Engineering of minimal, homomeric shuttling vehicles

[0757] Current VLP -delivery tools based on cargo fusion utilize co-assembly with virally-derived Gag / Gag- Pol domains, as they are obligatorily dependent on the protease domain contained in the Pol frame for proteolytic cargo release. The inventors hypothesized that due to its aptamer recruitment and release mechanism, ENVLPE is not fundamentally dependent on protease-mediated release, even though it still utilizes mosaic assemblies with Gag / Gag— Pol domains composed of GagL2is,NLs-NEs— PCP, Gag, and Gag-Pol.

[0758] Omitting Gag / Gag-Polo64v in the ENVLPE setup only resulted in a modest decrease in efficacy (Figure 27A), indicating that Gag / Gag-Polo64v was not essential. In contrast, omitting Csy4 led to a strong diminution in RNP transfer efficacy, indicating that RNP stability is a major bottleneck in aptamer-based RNP delivery systems (Figure 27A).

[0759] The inventors identified the relatively mild impact of Gag / Gag-Polo64v omission as an opportunity to create a minimal VLP -like delivery system. The inventors thus attempted to replace as many HIV-1 elements as possible with functional equivalents of non-viral elements (Figure 14A,B). While preserving the nucleocytosolic shuttling motif and the aptamer-binding domain PCP, most of the MA / CA / NC regions were removed (AMA12-114, ACA133-277, ANC) except for the parts necessary to act as a spacer to induce membrane curvature during budding, and the budding domain p6 (Figure 14A). The now missing oligomerization role of Gag was compensated by GCN4 coiled coils, which form homomeric parallel dimers and trimers. Examining the functionality of this minimal design (miniENVLPE), the inventors again found that the nucleocytosolic shuttling component is a key element in forming an efficient RNP packaging and delivery system (Figure 14C). Replacing the N-myristylation signal (MY) with a pleckstrin homology domain from phospholipase C-51 (PH) that binds to phosphatidylinositol lipids, a component of the plasma membrane, further increased efficacy (Figure 14C). Remarkably, these miniature systems contained less than 13% of wild-type HIV-1 Gag sequences but were almost as active as full-length, homomeric GagL2is,NLs-NEs-PCP setups (ENVLPE AGag / Gag-PolD64V, Figure 14D).

[0760] Since the minimal system is considerably decreased in size which may shift its nucleocytosolic equilibrium, the inventors re-tested several NES-NLS combinations on the miniENVLPE system. Still, no substantial improvement over the original nucleocytosolic shuttling sequence was found (Figure 14D). Cryo-electron microscopy in the inventors’ experiment revealed that HIV-1 Gag VLPs had a median radius of ~55 nm, while ENVLPE and miniENVLPE had a smaller median radius of ~31.5 nm (Figure 27B).

[0761] Mechanistic insights yield optimized ENVLPE*

[0762] Further experiments on homomeric and minimal assemblies with different aptamers for recruitment provided additional evidence for the inventors’ key mechanistic insights on the ENVLPE system for PE: firstly, the PE efficacy improvement of PEcsy4 is also compatible with homomeric and minimal ENVLPE variants (Figure 14E, Figure 27A); secondly, cargo release is sufficient without proteolytic activity, but is restricted to aptamer-systems with intermediate affinities (nM range) since ultra-high affinity aptamer systems (pM range) limits cargo release in the recipient cells (Figure 14F, Figure 21C, Figure 27C); lastly, Gag / Gag-Pol is not essential for particle assembly, but is still generally beneficial (Figure 14F, Figure 27A, Figure 27C, D). Thus, The inventors addressed the high-affinity interactions of aptamer-coat protein pairs (e.g., PCP-PP7 and Csy4-C4) and clarified how these contribute to efficient packaging and release. Protease-deficient experiments validate the necessity of dynamic dissociation in recipient cells (Figures 14F and 27).

[0763] Since GCN4 coiled coils were successfully applied in miniENVLPE, the inventors also introduced them into ENVLPE (Figure 15A). Remarkably, this led to another significant increase in efficacy (Figure 15A), thus yielding a new GCN4-enhanced final configuration, which the inventor’s term ENVLPE . According to one embodiment of the present invention, the ENVLPE+variant and / or ENVLPE+system is characterized by an additional oligomerization domain.

[0764] According to one embodiment of the present invention, the ENVLPE+variant and / or ENVLPE+system is characterized by an additional oligomerization domain via coiled-coils.

[0765] According to one embodiment of the present invention, the ENVLPE+variant and / or ENVLPE+system is characterized by an enhanced efficacy in RNP delivery. In a further embodiment, the ENVLPE+variant and / or ENVLPE+system is characterized by an enhanced efficacy in RNP delivery compared to the initial ENVLPE system.

[0766] The inventors surprisingly found that an additional coiled-coil dimerization domain, as used in some embodiments of this invention (e.g. those referring to a ENVLPE+variant and / or ENVLPE+system) is highly advantageous.

[0767] In one embodiment, the ENVLPE+variant and / or ENVLPE+system comprises an additional coiled- coil dimerization domain.

[0768] Characterization of cargo loading efficiency

[0769] The inventors next combined ENVLPE+with PEcsy4 as the optimal configuration for PE and conducted comparisons with v3 and v3b PE-eVLPs. As they are based on distinct packaging mechanisms compared to ENVLPE+(Figure 28), the inventors chose to compare the amount of pegRNA and Cas9 effector protein per VLP. The inventors found that one ENVLPE+particle was loaded with an average of 78 Cas9 molecules and 77 pegRNAs (ratio 0.99). In contrast, similar Cas9 levels were observed for v3 and v3b PE-eVLPs but distributed in a higher number of VLPs per volume and with a Cas9:pegRNA ratio of 0.58 and 0.61 for v3 and v3b PE-eVLPs, respectively (Figure 29A, B). This result aligns with the inventors’ hypothesis that ENVLPE favors packaging of pegRNA:Cas9 complexes, while packaging of pegRNAs or Cas9 apoenzymes alone is improbable due to instability or absence of a recruitment handle. Conclusively, the inventors have confirmed in their quantitative assays that PE ENVLPE+ also has nearly 4-fold more functional PE RNP per VLP packaged than v3b PE-eVLP (Figure 29A-B), and a 1 : 1 stoichiometry for PE:pegRNA was measured while v3 / v3b PE-eVLP had an excess of pegRNA-ffee prime editors within the VLP fraction. This and the corresponding enhanced efficacy in PE mode are probably attributable to a combination of improved packaging efficacy via nucleocytosolic shuttling, in conjunction with aptamer-based packaging, and additional RNP protection via the Csy4 / C4 3'-protection system.

[0770] Benchmarking ENVLPE+against eVLPon endogenous targets

[0771] The inventors proceeded to benchmark the editing efficacy and precision of ENVLPE+in base- and prime-editing modes on established endogenous loci against the MMLV-based v4 BE-eVLP and v3b PE-eVLP systems (see Figure 28 for the mechanistic differences). Because eVLP production is driven by CMV promoters, which are considered to be one of the strongest promoters in HEK293T cells, the inventors also switched ENVLPE+to CMV promoter-driven expression and optimized the plasmid stoichiometries again (Figure 29C). The inventors wondered whether v3 or v3b PE-eVLP systems would benefit from Csy4 protection as well, and while both v3 and v3b PE-eVLP showed improvements with the additional 3 ’-protection to varying extents, ENVLPE+still achieved the highest prime-editing efficacy of all tested conditions (Figure 29D). The inventor’s work emphasizes how the guide-RNA-based packaging mechanism (e.g., PP7-PCP) of ENVLPE facilitates modular adaptation to emerging RNA-guided editors and effectors. The inventors clarified why they chose the RNA- aptamer-based packaging and how it avoids the limitations of protein fusions to CRIPSR-effectors for packaging. The inventors also showed that the dedicated 3 ’-protection system based on Csy4-C4 is compatible with other VLP systems, such as v3 and v3b PE-eVLPs (Figure 29D).

[0772] The inventors then performed a detailed comparison on the endogenous HEK3 locus (+1 t>a editing), including additional controls and comparisons to individual features of the ENVLPE system in HEK293T cells (NLS-NES, PEcsy4, miniENVLPE, GCN4 coiled coils, CMV promoter). Evaluation of the editing performance at endogenous loci again demonstrated the functional relevance of the features built into ENVLPE+and reaffirmed CMV-driven PEcsy4 ENVLPE+as the optimal configuration. This configuration also outperformed the state-of-the-art v3b PE-eVLP system in both prime editing modes, ‘PE2’ and in ‘PE3’ (Figure 15B). Thus, the PE ENVLPE+system demonstrates superior editing efficacy on various endogenous targets and cell lines compared to v3 / v3b PE-eVLP-based systems that rely on double recruitments of the pegRNA and the prime editor. This efficacy is likely attributable to the enhanced packaging efficiency of functional RNPs, as evidenced by the quantification of the pegRNA (RT-qPCR instead of Northern Blot) and the prime editing enzyme (ELISA), as requested (Figure 29A-B).

[0773] The inventors continued to benchmark the inventors’ system across three more loci and more editing types (t>a, g>c, t>g, +2 gctg>cacc), which showed that iPEcsy4 ENVLPE+displayed consistently higher editing than v3b PE-eVLP across multiple endogenous sites in HEK293T and in Neuro-2a cells (Figure 15B, 15C). No significant increase in off-targets compared to v3b PE-eVLP was found despite higher on-target efficacy (Figure 15C), reinforcing the platform's safety profile. On the other hand, plasmid transfection of prime-editing components led to elevated off-target rates (Figure 15C). Thus, the VLP system of the present invention is safer compared to plasmid transfection of primeediting components due to less off-target effects. As expected for PE, editing precision (proportion of correct edits within all edited reads) was consistently high across all systems, mostly above 90%. Importantly, the controls used in Figure 15B,D show that the absence of any element of the VLP system of the present invention negatively impacts performance (Figure 15B,D). The ANLS-NES control experiments in prime- and base editing modes clearly confirm that nucleocytosolic shuttling is also beneficial for base editing (BE). Moreover, the inventors furtjer demonstrated that ENVLPE+equipped with Csy4-protected prime editors (PEcsy4) exhibited a substantial enhancement in performance. Thus, the inventor’s innovative combination of nuclear-shuttling motifs with optimized aptamer-based recruitments and a 3 ’-protection system, is a carefully designed VLP system with superior performance.

[0774] The inventors extended ENVLPE+benchmarking against eVLPs to include base editing and found similar rates for editing of multiple endogenous sites as with v4 BE-eVLPs in HEK293T and iPSC cells, with miniENVLPE also mediating comparable editing efficacies (Figure 15D). Importantly, to demonstrate the conceptual innovation of utilizing nuclear-shuttling motifs combined with optimized aptamer-based recruitments and a dedicated 3 ’-protection system, the inventor’s experiments show that the absence of each of the elements is detrimental to the overall performance (Figures 15B and 15D). Thus, the combination of these carefully chosen elements has led to a well-designed and efficient VLP system with superior performance. The inventors found it interesting to study which components are absolutely necessary for an efficient RNP delivery module, providing a mechanistic understanding of the dependency on proteolytic maturation, aptamer-mediated cargo release, and particle assembly. It even leads to further enhancements of the inventor’s main technology (ENVLPE+). miniENVLPE also highlights the inventor’s efforts towards a budding module with minimal virus-derived protein parts, removing all non-essential viral sequences (87% of HIV- 1 -related sequences were removed in miniENVLPE) or replacing them with non-viral counterparts. The inventors have expanded the section on mechanistic studies and included further experiments showing that miniENVLPE can be applied successfully in a non-mosaic fashion to facilitate prime and base editing (Figure 15B,D).

[0775] Base editing with ENVLPI for ex vivo T-cell engineering

[0776] The inventors then sought to assess the therapeutic efficacy of their system in clinically relevant cell types with typical hypoimmunogenic edits (disruption of MHCI and TCR via double-strand break-free edits of B2M and TRBC1 / 2 locus in T lymphocytes, Figure 16), thereby allowing the assessment of the usefulness of ENVLPE for generating hypo-immunogenic T cells suitable for cell-based therapeutic interventions. To achieve sufficient titers, the inventors scaled up ENVLPE+particle production and transitioned to ultracentrifuge precipitation for enrichment.

[0777] The inventors then administered ENVLPE+to primary T lymphocytes, facilitating abase edit in splice donor sites of the B2M and TRBC1 / 2 loci to knock out MHCI and TCR, two common KO targets to generate hypoimmunogenic CAR-T cells (Figure 16A). ENVLPE+facilitated both base edits at high efficacy and in a dose-dependent fashion (Figure 16B, 16C).

[0778] For B2M modification, the inventors observed a direct correlation between editing efficacy and surface protein expression that is consistent with the requirement for biallelic modification to achieve complete knockout (Figure 16B). This relation was less clear for TRBC1 2. where the KO of CD3 more closely follows editing (Figure 16C). It is likely that this pattern is caused by only one TRBC locus being transcriptionally active at a time, resulting in a bias in both editing and expression towards the same allele. The inventors then proceeded to compare the efficacy of a simultaneous gene knockout with a single transduction of either ENVLPE+or v4 BE-eVLPs. The inventors’ findings revealed comparable performance for both systems, albeit at a remarkably 40-fold higher titer of v4 BE-eVLPs (5.43 x 1011v4 BE-eVLP particles and 1.34 x 1O10ENVLPE+particles per 50,000 cells; Figure 16D). Conclusively, the inventors achieved a 40-fold reduction in the number of VLPs required to attain equivalent half-maximum total edits in a multiplexed base editing configuration.

[0779] This outcome signifies a substantial enhancement in per-VLP performance.

[0780] Thus, the inventors have shown in a head-to-head comparison in ex vivo and in vivo applications that EVNLPE+has superior per-VLP editing efficacy. Particularly, the inventors hypoimmunogenized primary T cells by KO’ing the respective genes essential for MHCI and TCR formation. Remarkably, the inventors found that a 40 x higher VLP dose of v4 BE-eVLP was required to approximate the editing performance of ENVLPE+. Therefore, the present invention is a very promising path towards universal hypoimmunogenic T cells. A protein-ribonucleic acid complex, a polynucleotide encoding the multifunctional protein (I) of the protein-ribonucleic acid complex, and a nucleic acid delivery system according to one embodiment of the present invention is thus highly superior, since ~40x higher doses of v4 BE-VLPs were required to achieve similar editing efficacy in multiplexed ex vivo KO applications (TCR and MHCI disruption via inactivation of TRBC1 / 2 and B2M via base editing) in primary T lymphocytes (Figure 16D). Most strikingly, the inventors showed that ENVLPE+outperformed v3 PE- eVLPs in two mouse models for retinitis pigmentosa rd6 and rdl2.

[0781] Distribution of ENVLPE in vivo after subretinal injection

[0782] The inventors next set out to evaluate ENVLPE ’s editing performance in an in vivo setting and chose an ophthalmology context, since the eye has been used as a potent testbed for breakthroughs in precision genome editing. Multiple BE and PE strategies have been demonstrated in a diverse set of inherited retinal degeneration models, though virally delivered BE and PE limit the clinical applicability of the studies.

[0783] To first test ENVLPE’s transduction efficacy and local distribution of ENVLPE upon subretinal injection, the inventors applied VSV-G-pseudotyped ENVLPE loaded with Crc i s mRNAip? to mT / mG mice (Figure 17A), which report Cre-recombinase activity as a conversion from red-to-green fluorescence. The inventors injected 1.0, 1.5, and 2.0 pl of an ENVLPE solution (2.53 x 108particles pE1in PBS) and observed broad and homogeneous green fluorescence reporter activation across the retinal pigment epithelium (RPE) for 17 / 18 injected eyes after one week after injection (Figure 17B, 30A). More importantly, no gross change of the compact RPE layer's pentagonal and hexagonal cell morphology was evident, suggesting ENVLPE was well tolerated. Consistent with these results, ENVLPE also performed well as an mRNA-delivery tool in a titration experiment in a Cre-dependent green-to-red reporter HEK293 line (Figure 30B). The inventors demonstrated that ENVLPE+equipped with Csy4-protected prime editors (PEcsy4) exhibited a substantial enhancement in the functional restoration of light-dependent electric retinal activity in a rd! 2 retinitis pigmentosa mouse model (Figure 17). When the same number of VLPs of ENVLPE+and v3 PE-eVLP were injected, a 10-fold higher v3 PE-eVLP dose was found to be approximately equivalent to the restoration of PEcsy4 ENVLPE .

[0784] Prime editing of inherited retinal degeneration mouse models with PEcsy4 ENVLPEf

[0785] Encouraged by the results from mRNA delivery, the inventors sought to benchmark the performance of VLP -mediated PE in the inventors’ optimized configuration iPEcsy4 ENVLPE To this end, the inventors chose mouse models of retinitis pigmentosa (rd6) and of Leber congenital amaurosis (rd!2), which have recently been used to measure the editing performance of v3 PE-eVLPs after subretinal injection. The loss-of-fimction phenotype in the rd6 mouse model is caused by a 4 bp deletion in the splice donor region oiMfrp intron 4, abolishing protein expression (Figure 17C). To enable a direct head-to-head comparison with the recently published v3 PE-eVLP systems in vivo, the inventors employed the identical PE3b strategy for ENVLPE+(including identical spacers and 3 ’ extensions of the pegRNA, but adapted to the inventors’ scaffold and with 3 ’-modifications required for the Csy4 / C4 protection) (Figure 17C). The inventors then produced both VLP systems in parallel from 15 cm dishes with 12.5 x 106HEK293T cells / dish using original plasmids for v3 PE-eVLPs (from Addgene), concentrated the VLPs by ultracentrifuge precipitation using a sucrose cushion, and quantified the titers by ELISA against the capsid proteins in ENVLPE+(HIV-1 p24) or v3 PE-eVLP (MMLV p30), respectively (Figure 31). 5-7-weeks-old rd6 mice were injected with equal doses of PE3bcsy4 ENVLPE+or v3 PE3b-eVLP (1.6* 109particles pl1in 1 pl PBS). Three weeks post-injection, the inventors analyzed the protein expression of Mfrp in the eye cup via immunoblot and assessed the editing outcomes via targeted amplicon sequencing. Only ENVLPE+-treated eyes showed visibly restored MFRP protein expression (Figure 17D) mediated by the restoration of the splice donor site (median 8.6%, Figure 17E) with no detectable indels (Figure 17F), while samples from v3 PE3b- eVLP -treated mice exhibited MFRP levels beyond detection limit.

[0786] The inventors repeated their head-to-head comparison on a second mouse model, rdl2, which displays a more severe loss-of-fiinction phenotype caused by a nonsense mutation in exon 3 of Rpe65, which abolishes the canonical visual cycle leading to visual impairment (Figure 17G). Three weeks after subretinal injection of equal doses (1 x 109particles pE1in 1 pl PBS) of both preparations into the RPE of 5-7-weeks-old rdl2 mice, immunoblot analysis of RPE homogenates revealed that the restored RPE65 protein could only be detected in eyes treated with PE3bcsy4 ENVLPE+, whereas protein levels in eyes treated with v3 PE3b-eVLP remained below the detection limit (Figure 17H). Comparable results were found by analyzing Rpe65 exon 3 via targeted amplicon sequencing, which revealed substantial successful prime editing events with a median of 6.3% for PE3bcsy4 ENVLPE+ (Figure 171) without detectable indels (Figure 17J). In comparison, editing levels in the v3 PE3b- eVLP condition were below the detection threshold. The ERG response correlated with the observed editing results across all samples from ENVLPE+(Figure 32). rd! 2 mice display a virtually absent light-dependent retinal activity that can be quantified via electroretinography (ERG) (Figure 17K,L). Treatment with ENVLPE+evoked -5.5-fold higher b- wave signal (median 173.5 pV) compared to an equal dose of v3 PE-eVLPs (median 31.52 pV). A response of the comparable magnitude (median 102.2 pV) was evoked only after injecting a dose of v3 PE-eVLP that was approximately one order of magnitude higher (13.7x 109particles pl1) (Figure 17K,L). Conclusively, in in vivo experiments, when compared to v3 PE-eVLPs in rd6 and rd 12 mouse models for genetically inherited retinal degeneration, ENVLPE+ outperformed v3 PE-eVLPs in per VLP conditions (Figure 17C-M). It was evident that ENVLPE+ in prime editing mode exhibited superior performance in the functional restoration of rd6 and rd 12 mouse models for retinitis pigmentosa following subretinal injections of equimolar VLPs. Only when a 13-fold higher concentration of eVLP was injected did it approach a similar level of functional restoration as ENVLPE+ as measured via ERG (Figure 17K-L).

[0787] In rdl2 mice, the absence of RPE65 disrupts the visual cycle, leading to a deficiency of 11-czs'-retinal. High-performance liquid chromatography (HPLC) analysis of retinoids in PE3bcsy4 ENVLPE+-treated rd! 2 mouse eyes revealed significant restoration of I l -c / .s-rctinal after dark adaptation, reaching -20% of wild-type levels, suggesting a partially restored visual cycle (Figure 17M). Instead, 1 l -c / .s-rctinal levels in v3-PE3b-e VLP -treated samples were below the detection limit, though some restoration must have occurred due to the recordable A -wave signals (Figure 17K,L). Conclusively, the inventors surprisingly showed that mouse models for inherited retinal degeneration can be treated using the ENVLPE+system of the present invention. Moreover, treatment of the rd6 and rdl2 mouse models for genetically inherited retinal degeneration with ENVLPE+clearly outperformed the treatment with v3 PE-eVLPs.

[0788] Figures 35 to 38 show that the same results as shown above for Gag can be achieved in a Gagindependent manner by using phage capsid protein instead of HIV- 1 Gag. The function of homooligomerization is functionally replaced by P22 phage gp5 capsid protein and optionally modulated by homo-oligomerization domains, such as GCN4 and functional equivalent domains (Figure 35). To enable membrane tethering, phospholipase C-51 pleckstrin homology domain (PLC-H) was included in the fusion polypeptide together with other indicated domains enabling in cis and trans (e.g, fusogen), to enable budding, cargo loading of nuclear-localized RNPs, and fusion with the recipient cells. The supernatant of cells transfected with plasmids encoding such engineered polypeptides yield expectedly enveloped spherical protein cages resembling native P22 phages but with a lipid bilayer due to the fused membrane-tether (Figure 36). Instead of flexible linkers, such as XTEN, one can also use rigid linkers, such as repeats of EAAAR (or EAAAK as an alternative and any other alpha-helical polypeptide domains, Figure 37A and Figure 37B); designs containing multiple repeats of EAAAR instead of XTEN were also capable of delivering prime editing RNPs to cell lines carrying a fluorescence reporter system (Figure 37C). Alternatively, aptamer-binding proteins derived from bacteriophages, such as PP7 coat protein (PCP), can functionally homo-oligomerize on their own without additional domains that can be additionally modulated by partially reactivating the capsid function of PP7 / PCP (Figure 38A and Figure 38B), resulting in a novel functional RNP delivery module as shown by delivery prime editing RNPs capable of editing a fluorescence reporter (Figure 38C). Homo-oligomerization is generally enhanced by membrane-tethering, which serves as a scaffolding / nucleation site for enhancing homo-oligomerization by restricting theoretical 3D movements of monomers to 2D lateral movements, thus increasing the probability of monomermonomer interaction.

[0789] EXAMPLE 3 Transfection conditions for VLP production The features disclosed in the foregoing description, in the claims and / or in the accompanying drawings may, both separately and in any combination thereof, be material for realizing the invention in diverse forms thereof.

[0790] REFERENCES

[0791] Accola, M.A., Strack, B., and Gottlinger, H.G. (2000). Efficient Particle Production by Minimal Gag Constructs Which Retain the Carboxy-Terminal Domain of Human Immunodeficiency Virus Type 1 Capsid-p2 and a Late Assembly Domain. Journal of Virology 74, 5395-5402. 10.1128 / j vi.74. 12.5395- 5402.2000.

[0792] Anzalone, A.V., Randolph, P.B., Davis, J.R. et al. (2019). Search-and-replace genome editing without double-strand breaks or donor DNA. Nature 576, 149-157.

[0793] Anzalone, A.V., Gao, X.D., Podracky, C.J. etal. (2022). Programmable deletion, replacement, integration and inversion of large DNA sequences with twin prime editing. Nat Biotechnol 40, 731— 740.

[0794] Banskota, S., Raguram, A., Suh, S., Du, S.W., Davis, J.R., Choi, E.H., Wang, X., Nielsen, S.C., Newby, G.A., Randolph, P.B., et al. (2022). Engineered virus-like particles for efficient in vivo delivery of therapeutic proteins. Cell 185, 250-265.el6.

[0795] Boutant, E., Bonzi, J., Anton, H., Nasim, M. B., Cathagne, R., Real, E., & Bemacchi, S. (2020). The two zinc fingers in the nucleocapsid domain of the HIV-1 Gag precursor are equivalent for the interaction with the genomic RNA in the cytoplasm, but not for the recruitment of the complexes at the plasma membrane. bioRxiv, 2020-01.

[0796] Choi, J., Chen, W., Suiter, C.C. et al. (2022) Precise genomic deletions using paired prime editing. Nat Biotechnol 40, 218-226.

[0797] Danner, Eric et al. “A homology independent sequence replacement strategy in human cells using a CRISPR nuclease.” Open biologyvol. 11, 1 (2021): 200283. doi: 10.1098 / rsob.200283.

[0798] Grigorov, B., Decimo, D., Smagulova, F., Pechoux, C., Mougel, M., Muriaux, D., & Darlix, J. L. (2007). Intracellular HIV-1 Gag localization is impaired by mutations in the nucleocapsid zinc fingers. Retro virology, 4, 1-12. Hamilton, J. R., Tsuchida, C. A., Nguyen, D. N., Shy, B. R., McGarrigle, E. R., Espinoza, C. R. S., ... & Doudna, J. A. (2021). Targeted delivery of CRISPR-Cas9 and transgenes enables complex immune cell engineering. Cell reports, 35(9).

[0799] Hendel, A., Bak, R., Clark, J. etal. Chemically modified guide RNAs enhance CRISPR-Cas genome editing in human primary cells. Nat Biotechnol 33, 985-989 (2015).

[0800] Homs, F., Martinez, J.A., Fan, C., Haque, M., Linton, J.M., Tobin, V., Santat, L., Maggiolo, A.O., Bjorkman, P.J., Lois, C., et al. (2023). Engineering RNA export for measurement and manipulation of living cells. Cell 186, 3642-3658.e32.

[0801] Komor, A.C., Kim, Y.B., Packer, M.S., Zuris, J.A., and Liu, D.R. (2016). Programmable editing of a target base in genomic DNA without double-stranded DNA cleavage. Nature 533, 420-424.

[0802] Lu, B., Javidi-Parsijani, P., Makani, V., Mehraein-Ghomi, F., Sarhan, W.M., Sun, D., Yoo, K.W., Atala, Z.P., Lyu, P., and Atala, A. (2019). Delivering SaCas9 mRNA by lentivirus-like bionanoparticles for transient expression and efficient genome editing. Nucleic Acids Res. 47, e44.

[0803] Lyu, P., Javidi-Parsijani, P., Atala, A., and Lu, B. (2019). Delivering Cas9 / sgRNA ribonucleoprotein (RNP) by lentiviral capsid-based bionanoparticles for efficient “hit-and-run” genome editing. Nucleic Acids Res. 47, e99.

[0804] Ma H, Tu LC, Naseri A, Huisman M, Zhang S, Grunwald D, Pederson T. CRISPR-Cas9 nuclear dynamics and target recognition in living cells. J Cell Biol. 2016 Aug 29;214(5):529-37.

[0805] Mangeot, P.E., Risson, V., Fusil, F. et al. Genome editing in primary cells and in vivo using viral- derived Nanoblades loaded with Cas9-sgRNA ribonucleoproteins. Nat Commun 10, 45 (2019).

[0806] Nakade, Shota et al. “Microhomology-mediated end-joining -dependent integration of donor DNA in cells and animals using TALENs and CRISPR / Cas9.” Nature communications vol. 5 5560. 20 Nov. 2014, doi: 10.1038 / ncomms6560.

[0807] Nelson, J.W., Randolph, P.B., Shen, S.P. etal. Engineered pegRNAs improve prime editing efficiency. Nat Biotechnol 40, 402-410 (2022). Sherer, N.M., Swanson, C.M., Papaioannou, S., and Malim, M.H. (2009). Matrix mediates the functional link between human immunodeficiency virus type 1 RNA nuclear export elements and the assembly competency of Gag in murine cells. J. Virol. 83, 8525-8535.

[0808] Suzuki, Keiichiro et al. “In vivo genome editing via CRISPR / Cas9 mediated homology-independent targeted integration.” Nature vol. 540,7631 (2016): 144-149. doi: 10.1038 / nature20565.

[0809] Truong, D.-J.J., Armbrust, N., Geilenkeuser, J., Lederer, E.-M., Santl, T.H., Beyer, M., Ittermann, S., SteinmaBl, E., Dyka, M., Raffl, G., et al. (2022). Intron-encoded cistronic transcripts for minimally invasive monitoring of coding and non-coding RNAs. Nat. Cell Biol. 24, 1666-1676.

[0810] Truong, D.-J.J., Armbrust, N., Geilenkeuser, J., and Westmeyer, G.G. (2023). Generation of cell lines for the usage of Intron-encoded Scarless Programmable Extranuclear Cistronic Transcripts (INSPECT). Research Square. 10.21203 / rs.3.pex-2000 / vl.

[0811] Truong, DJ.J., Phlairaham, T., EBwein, B. et al. Non-invasive and high-throughput interrogation of exon-specific isoform expression. Nat Cell Biol 23, 652-663 (2021).

[0812] Wang H, Nakamura M, Abbott TR, Zhao D, Luo K, Yu C, Nguyen CM, Lo A, Daley TP, La Russa M, Liu Y, Qi LS. CRISPR-mediated live imaging of genome editing and transcription. Science. 2019 Sep 20;365(6459): 1301-1305.

[0813] Yamall MTN, loannidi El, Schmitt-Uhns C, Krajeski RN, Lim J, Villiger L, Zhou W, Jiang K, Gamshyants SK, Roberts N, Zhang L, Vakulskas CA, Walker JA 2nd, Kadina AP, Zepeda AE, Holden K, Ma H, Xie J, Gao G, Foquet L, Bial G, Donnelly SK, Miyata Y, Radiloff DR, Henderson JM, Ujita A, Abudayyeh OO, Gootenberg JS. (2023) Drag-and-drop genome insertion of large sequences without double-strand DNA cleavage using CRISPR-directed integrases. Nat Biotechnol. 2023 Apr;41(4):500-512. doi: 10.1038 / s41587-022-01527-4.

[0814] Zalatan JG, Lee ME, Almeida R, Gilbert LA, Whitehead EH, La Russa M, Tsai JC, Weissman JS, Dueber JE, Qi LS, Lim WA. Engineering complex synthetic transcriptional programs with CRISPR RNA scaffolds. Cell. 2015 Jan 15; 160(l-2):339-50. doi: 10.1016 / j.cell.2014. 11.052.

Claims

Claims1. A protein-ribonucleic acid complex comprising(I) at least one multi-functional protein or a set of at least two proteins, wherein the at least one multi-functional protein or the set of at least two proteins comprise the following protein domains:(a) at least one plasma membrane-interacting domain;(b) at least one oligomerization domain;(c) a membrane-bending domain;(d) at least one shuttling domain comprised of at least one nuclear export signal (NES), and nuclear localization signal (NLS);(e) a nucleic acid-binding (NAB) domain, capable of binding a single strand or double-strand RNA or DNA motif; and(f) a budding domain, which comprises a monopartite or multipartite motif that enables ESCRT-dependent or ESCRT-independent budding;(g) optionally, a reverse transcriptase domain; h) optionally, a maturation factor;(i) optionally, a reporter-signal generating domain;( / ) optionally one nucleic acid-stabilizing domain; and(II) a nucleic acid editor complex, which comprises an mRNA, a guide RNA or a ncRNA, which is bound via the nucleic acid-binding domain (e) of (I), preferably as ribonucleoprotein complex, wherein only the RNA species comprises a nucleic acid motif, bound by the NAB domain (e) of (I), wherein the RNA species is preferentially a gRNA, and wherein the RNA species comprises optionally an autonomous 3 ’-stabilization motif and / or a motif that is protected by an additional heterologously expressed nucleic acid-stabilizing domain ( / ) of (I).

2. The complex of claim 1, wherein the plasma membrane -interacting domain (a) of (I) is: a lipid group, such as C12 to C18 fatty acid(s) (preferably myristoyl (Cl 4), palmitoyl (Cl 6)) or isoprenoid(s) (preferably famesyl); or a polypeptide domain capable of binding membrane proteins or membrane components,such as pleckstrin homology (PH) domain, C2 domain, FYVE domain, phox homology (PX) domains, PHD finger domain, PROPPINs domain, epsin N- terminal homology (ENTH) domain, BAR (Bin-Amphiphysin-Rvs) domains, Tubby domain, and / or wherein the oligomerization domain (b) of (I) is a homo- and / or hetero-multimerizing domain, such as coiled coils, PDZ domains, de novo generated homo / hetero-multimerization domains composed of coil-only, sheet-only, or mixed-type proteins, and / or wherein the membrane-bending domain (c) of (I) induces plasma membrane bending and formation of nanospheres, preferably with a diameter in the range from about 50 nm to about 1 pm, and / or wherein the NES (d) of (I) is preferably the NES of HIV- 1 (NESHIV-I) or DBR1, more preferably a NES comprising an amino acid sequence according to SEQ ID NO: 65, and / or wherein the NLS (d) of (I) is preferably NLS of c-Myc protein (NLSC-Myc), NLS of SV40 (NLSSV4O) or a synthetic NLS, such as attenuated variants of NLSC-Myc and NLSsv4o, or other mono- or bipartite synthetic NLS, more preferably wherein the NLS comprises an amino acid sequence according to any one of SEQ ID NOs: 48 to 54, and / or wherein the nucleic acid-binding domain (e) of (I) is a C4-aptamer binding protein (Csy4- enzyme) domain, preferably a Csy4 domain comprising an amino acid sequence according to SEQ ID NO: 64, an RNA-binding PP7 coat protein (PCP) domain, preferably a PCP domain comprising an amino acid sequence according to SEQ ID NO: 77, 78, or 79, an MS2 coat protein (MCP) domain, preferably a MCP domain comprising an amino acid sequence according to SEQ ID NO: 58, a boxB domain, the N Protein of bacteriophage P22, a Com -com domain, or a psi-binding nucleocapsid (NC) domain of HIV- 1 Gag, optionally wherein all domains can be arranged as a single chain tandem protein, and / or wherein the budding domain (f) of (I) is a p6 domain, preferably a p6 domain comprising an amino acid sequence according to SEQ ID NO: 75, a p9 domain, a pl2 domain, a pl9 domain, an L domain, an M protein, an M2 protein, an F13 protein, or a Z protein, and / or wherein the complex comprises a reverse transcriptase domain (g) of (I), and an integrase domain, which is inactivated by mutating at least one of the amino acids at position D64 (such as pol D64V), DI 16, E152, and / or F185 of a pol protein of HIV-1.

3. The complex of claim 1 or 2, wherein (I) is a multi-functional protein, preferably based on group-specific antigen (gag), and / or on a phage capsid protein, such as an engineered version of a PP7 coat protein (PCP), and / or on P22 bacteriophage, more preferably HIV-1 gag, and / or a gag -pol polyprotein, or a mini-Gag variant, which preferably comprises any one of the following structures:(i) Gag: A-myr-[MA]-[CA]-p2-[NC]-pl-[p6](11) Gag-pol D64v: A-myr-[MA]-[CA]-p2-[NC]-pl-[p6*]-PR-RT-INTD64v(111) GagL2i s: A-myr-[MAL2is]-[CA]-p2-[NC]-pl-[p6](IV) Gag AZF+L21 S" PCP: A-myr-[MAL2is]-[CA]-p2-[NCzFimutAZF2-PCP]-pl-[p6](v) GagAZF+L21 S / NLS-NES-PCP: A-myr-[MAL21 s]-[CA]-p2-[NCzFlmutAZF2+NLS-NES-PCP]-p l -[p6](vi) miniGagAZF+L2i S / NLS-NES-PCP: X-myr-[MAAi2-ii4]-[CAAi33-277]-p2-CC-NLS-NES-PCP-[p6](vii) PHminiGagAZF+L2i S / NLS-NES-PCP: PH-[MAAI2-II4]-[CAAI33-277]-P2-CC-NLS-NES-PCP- [p6],(viii) Gag-L21S-(ZFl-C>D,AZF2)-GCN4-HiBiT-NLS-NES-PCP,(ix) PH-Gag(A12-114_A133-277_ANC-pl)-GCN4-HiBiT-NLS-NES-NLS-PCP-p6,(x) P22-gp5-XTEN-GCN4-trimer-p6-HiBit-PH-PLC-delta,(xi) P22-gp5-XTEN-GCN4-trimer-HIV-NES-p6-HiBit-PH-PLC-delta,(xii) P22-gp5-HIV-NES-p6-HiBit-PH-PLC-delta,(xiii) P22-gp5-XTEN-GCN4-trimer-NLS-NES-NLS-PCP-p6-HiBit-PH-PLC -delta,(xiv) P22-gp5-NLS-NES-NLS-PCP-p6-HiBit-PH-PLC-delta,(xv) P22-gp5-NLS-NES-NLS-tdPCP-p6-HiBit-PH-PLC-delta,(xvi) P22-gp5-XTEN-GCN4-trimer-NLS-NES-NLS-PCP-p6-HiBit-PH-PLC -delta,(xvii) P22-gp5-PCP-p6-HiBit-PH-PLC -delta,(xviii) P22-gp5-GCN4-trimer-p6-HiBit-PH-PLC -delta,(xix) P22-gp5-p6-HiBit-PH-PLC-delta,(xx) P22-gp5-XTEN-p6-HiBit-PH-PLC-delta,(xxi) P22-gp5-XTEN-GCN4-trimer-PCP-p6-HiBit-PH-PLC-delta,(xxii) P22-gp5-NLS-NES-NLS-(EAAAR)x4-PCP-(EAAAR)x4-p6-HiBit-PH-PLC-delta, (xxiii) P22-gp5-NLS-NES-NLS-(EAAAR)x2-PCP-(EAAAR)x4-p6-HiBit-PH-PLC-delta,(xxiv) P22-gp5-NLS-NES-NLS-(EAAAR)xl-PCP-(EAAAR)x4-p6-HiBit-PH-PLC-delta,(xxv) P22-gp5-NLS-NES-NLS-(EAAAR)x4-PCP-(EAAAR)x2-p6-HiBit-PH-PLC-delta,(xxvi) P22-gp5-NLS-NES-NLS-(EAAAR)x4-PCP-(EAAAR)xl-p6-HiBit-PH-PLC-delta, (xxvii) P22-gp5-NLS-NES-NLS-(EAAAR)xl-PCP-(EAAAR)xl-p6-HiBit-PH-PLC-delta, (xxviii) P22-gp5-NLS-NES-NLS-(EAAAR)x2-PCP-(EAAAR)x2-p6-HiBit-PH-PLC-delta,(xxix) mGL-XTEN-P22-gp8(141-303),(xxx) NLS-NES-NLS-PCP-XTEN-P22-gp8( 141-303),(xxxi) NLS-NES-NLS-tdPCP-XTEN-P22-gp8(141-303),(xxxii) HIV-NES-XTEN-P22-gp8(141-303),(xxxiii) tdPCP-XTEN-GCN4-trimer-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-delta,(xxxiv) tdPCP-active-XTEN-GCN4-trimer-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC -delta,(xxxv) tdPCP_GCN4-trimer-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-delta,(xxxvi) tdPCP-XTEN-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-delta, (xxxvii)tdPCP-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-delta,(xxxviii) tdPCP-active-GCN4-trimer-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC -delta, (xxxix) tdPCP-active-XTEN-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-delta, (xl) tdPCP-active-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-delta,(xli) PCP-XTEN-GCN4-trimcr-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-dclta.(xlii) PCP_GCN4-trimer-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-delta, (xliii) PCP-XTEN-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-delta, (xliv) PCP-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-delta,(xlv) PCP-active-XTEN-GCN4-trimer-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-delta, (xlvi) PCP-active-GCN4-trimer-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-delta, (xlvii) PCP-active-XTEN-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC-delta, (xlviii) PCP-active-NLS-NES-NLS-FLAG-p6-HiBit-PH-PLC -delta, wherein the multi-functional protein more preferably comprises an amino acid sequence selected from any one of SEQ ID NOs: 1, 2, 13, 19, 24, 25, 26, 27, 60, 61, 393 to 396, and 453 to 491.

4. The complex of any one of claims 1 to 3, wherein (I) is a set of at least two proteins, which preferably comprises an amino acid sequence selected from any one of SEQ ID NOs: 26, 27, 56, 57 and 66.

5. The complex of any one of claims 1 to 4, wherein the nucleic acid editor complex (II) comprises a guide RNA, preferably a sgRNA or a pegRNA, which comprises the nucleic acid motif, bound by NAB domain (e) of (I) and an apo-protein of the nucleic acid editor complex, wherein the guide RNA is preferentially selected from a nucleic acid sequence according to any one of SEQ ID NOs: 160 to 169, or selected from a nucleic acid sequence according to any one of SEQ ID NOs: 325 to 392, and the apo-protein is preferentially selected from an amino acid sequences according to any one of the SEQ ID NOs: 28 to 35, and / or wherein the nucleic acid editor complex (II) is a gene-editing ribonucleoprotein (RNP) complex, which comprises gene-editing polypeptide(s) or protein(s), preferably CRISPR / Cas9 effectors, more preferably a CRISPR-associated endonuclease (Cas9 nuclease or Casl3),TALEs, pumilio proteins, de novo DNA / RNA binding proteins, or mutants thereof, and wherein the gene-editing polypeptide(s) or protein(s) preferably further comprise a nuclear localization signal (NLS), preferably NLS of c-Myc protein (NLSC-Myc), NLS of SV40 (NLSsv4o) or a synthetic NLS, such as attenuated variants of NLSC-Myc and NLSsv4o, or other mono- or bipartitesynthetic NLS, more preferably wherein the NLS comprises an amino acid sequence according to any one of SEQ ID NOs: 48 to 54, or a nuclear export signal (NES), preferably the NES of HIV- 1 (NESHIV-I) or DBR1, more preferably a NES comprising an amino acid sequence according to SEQ ID NO: 65, or combinations thereof.

6. The complex of any one of claims 1 to 5, wherein the RNA species comprises an autonomous 3’- stabilization motif, preferably truncated evopreQ 1 comprising a ribonucleic acid sequence according to SEQ ID NO: 171, and / or a motif comprising a ribonucleic acid sequence according to SEQ ID NO: 216 that is protected by an additional heterologously expressed nucleic acid-stabilizing domain ( / ) of (I), preferably C4 / Csy4 comprising an amino acid sequence according to SEQ ID NO: 64, the N- terminal domain of rotavirus NSP3 comprising an amino acid sequence according to SEQ ID NO: 323, wherein the N-terminal domain of rotavirus NSP3 optionally protects 3’-gacc RNA ends, or combinations thereof.

7. A polynucleotide encoding the multi-functional protein (I) as defined in any one of claims 1 to 6, preferably comprising a nucleotide sequence selected from any one of SEQ ID NOs: 82 to 106, or 407 to 452, or a set of polynucleotides encoding the set of at least two proteins (I) as defined in any one of claims 1 to 6, preferably comprising the nucleotide sequence of SEQ ID NO: 107 (psPAXD64V) and a nucleotide sequence selected from any one of SEQ ID NOs: 82 to 106, or 407 to 452.

8. Use of a polynucleotide or a set of polynucleotides of claim 7 for- generating a nucleic acid delivery system, preferably a virus-like particle (VLP),- delivering nucleic acids, preferably gene-editing ribonucleoproteins, epigenome modulator nucleic acids, transcriptome-editing nucleic acids, and / or epitranscriptome modulator nucleic acids.

9. A nucleic acid delivery system, comprising(A) at least one multi-functional protein or a set of at least two proteins as defined in any one of claims 1 to 6;(B) a gene-editing ribonucleoprotein (RNP) complex as defined in claim 1 or 5;(C) a fusogenic (glyco)protein (envelope protein) or protein complex comprising at least one multi-functional protein or a set of at least two proteins, wherein the multifunctional domain or the set of at least two proteins comprise a receptor-targeting and a fusogenic domain to fuse the VLP to the target cell membrane;(D) a targeting domain, such as a glycoprotein or engineered variants thereof;(E) a moiety mediating endosomal escape, such as a glycoprotein or engineered variants thereof;(F) optionally, a reporter-signal generating gene product, such as a luciferase or a fluorescent protein, and(G) optionally, a nucleic acid-stabilizing protein, preferably Csy4, further optionally wherein the nucleic acid delivery system is a virus-like particle (VLP).

10. A plurality of polynucleotides, comprising(1) a first polynucleotide or a set of polynucleotides of claim 7;(2) a second polynucleotide comprising a nucleic acid sequence encoding a gene editor protein;(3) a third polynucleotide comprising a nucleic acid sequence encoding a guide RNA, such as a sgRNA or pegRNA, which is optionally tagged by a nucleic acid binding domain; and(4) a fourth polynucleotide comprising a nucleic acid sequence encoding an envelope protein, or one or more vectors comprising the plurality of polynucleotides, wherein the one or more vectors are preferably three vectors.

11. A pharmaceutical composition comprising: the polynucleotide or the set of polynucleotides of claim 7, the nucleic acid delivery system of claim 9, or the plurality of polynucleotides or the one or more vectors of claim 10; optionally, pharmaceutically acceptable excipients and / or carriers.

12. A cell or a cell line comprising: the multi-functional protein (I) as defined in any one of claims 1 to 6 or the set of at least two proteins (I) as defined in any one of claims 1 to 6, or the protein-ribonucleic acid complex of any one of claims 1 to 6; or the polynucleotide or the set of polynucleotides of claim 7; or the nucleic acid delivery system, preferably the VLP, of claim 9; or the plurality of polynucleotides or the one or more vectors of claim 10; wherein the cell or the cell line is preferably a mammalian cell or cell line, more preferably a human cell or cell line, optionally wherein the cell or cell line can be co-treated with exogenous compounds, such as by means of lipofection, electroporation, photoporation, sonoporation, particle bombardment,microinjection, magnetofection, viral transduction, cell-penetrating peptides, calcium-phosphate, nanoparticles, or combinations thereof, wherein the exogenous compounds are preferably selected from synthetic nanoparticles, fluorophores, small pharmaceutical compounds, positron emission tomography (PET) tracers, peptides, proteins, lipids, co-factors, small biomolecules, and combinations thereof.

13. A kit comprising: the polynucleotide of or the set of polynucleotides of claim 7, and / or the nucleic acid delivery system of claim 9, and / or the plurality of polynucleotides of claim 10, and / or one or more vectors of claim 10, and / or the pharmaceutical composition of claim 11, and / or the cell or cell line of claim 12.

14. A nucleic acid delivery system, preferably a virus-like particle, produced by transfecting, transducing, electroporating, lipofecting, photoporating, sonoporating, particle bombardment, microinjecting, magnetofecting, or otherwise inserting the plurality of polynucleotides or the one or more vectors of claim 10 into a cell and expressing the components of the nucleic acid delivery system from the plurality of polynucleotides or the one or more vectors in the cell, thereby allowing the nucleic acid delivery system, preferably the virus-like particle, to assemble in the cell.

15. The nucleic acid delivery system, preferably a virus-like particle, of claim 9 or 14, or the cell or cell line of claim 12 for use in medicine.

16. The nucleic acid delivery system, preferably a virus-like particle, of claim 9 or 14, or the cell or cell line of claim 12 for use in the treatment of genetic diseases, preferably monogenetic diseases and / or polygenetic diseases.

17. Use of a cell or cell line of claim 12 for- ex vivo generating a therapeutic cell line, a diagnostic cell line, or a cell line for tissue engineering,- tissue engineering,- preclinical applications, and / or- generating a VLP-distributing cell line, which preferably can be implanted into a combination of cells, a tissue and / or organism to produce VLPs therein.

18. An in vitro or in vivo method of editing a nucleic acid molecule in a target cell by nuclease mediated editing, base editing or prime editing, homology directed repair, non-homologous end-joining directed insertion, or microhomology-mediated end-joining directed insertion, such as by co-delivery of an integrase-deficient (D64V) lentivirus (IDLV), wherein, preferably, the IDLV is packaged using a psi- NC interaction, a PP7-PCP, MS2-MCP, and / or a C4-Csy4 interaction, comprising: contacting the target cell with a nucleic acid delivery system of claim 9 or 14, or with the pharmaceutical composition of claim 11, thereby installing one or more modifications to the nucleic acid at a target site; optionally wherein the target cell is a mammalian cell, preferably a human cell; or wherein the target cell is in a subject, preferably a mammal, more preferably a human; further optionally wherein the one or more modifications to the nucleic acid are associated with reducing, relieving, or preventing symptoms of a disease or disorder.

19. A method for nucleic acid delivery, preferably gene-editing nucleic acids, comprising the steps of:(I) providing the multi-functional protein (I) as defined in any one of claims 1 to 6 or the set of at least two proteins (I) as defined in any one of claims 1 to 6 to a cell, wherein said cell produces in its nucleus a gene-editing ribonucleoprotein (RNP) complex as defined in claim 1 or 5,(II) allowing the multi-functional protein or the set of at least two proteins to locate to the nucleus of said cell;(III) forming a complex of the multi-functional protein or the set of at least two proteins with the gene -editing ribonucleoprotein (RNP) complex in the nucleus;(IV) allowing the complex of (III) to be exported to the cytosol of the cell;(V) budding of virus-like particles (VLPs) comprising the gene-editing ribonucleoprotein (RNP) complex;(VI) binding of the VLPs to a target cell surface, preferably via a specific receptor;(VII) internalization of the VLPs into the target cell, preferably via endocytosis;(VIII) transport of the internalized complex from (VII) into the cytosol, such as via endosomal escape;(IX) dissociation of the RNP (with bound nucleic acid) or a nucleic acid from the nucleic acid binding domain of (I);(X) optionally, transport of the RNP into the nucleus for genome editing or transactivation, and(XI) optionally, binding of the guide-RNA-editing complex to an RNA target sequence.

Images