Compositions and methods for purifying lentivirus particles

Abstract

This disclosure provides substances and methods related to the purification of viral vectors. In particular, this disclosure provides peptides, compositions, adsorbents and related methods that can remove process-related impurities (e.g., host cell proteins, nucleic acids and culture medium components) and product-related impurities (e.g., product fragments, product aggregates and inactive forms derived from product degradation by or related to other species in the cell culture recovery) from a sample during the production and purification of lentiviral particles.

Description

[Technical Field]

[0001] Cross-reference of related applications This application claims priority and interest under U.S. Provisional Patent Application No. 63 / 518,035, filed on 7 August 2023, which is incorporated herein by reference in its entirety for all purposes.

[0002] Reference to electronically submitted materials The entire computer-readable nucleotide / amino acid sequence listing, identified as a single 100,535-byte file named "NCSU_42204_601_SequenceListing.xml" filed concurrently with this specification and created on August 5, 2024, is incorporated herein by reference.

[0003] This disclosure provides substances and methods related to the purification of viral vectors. In particular, this disclosure provides peptide ligands, compositions, adsorbents and related methods that can remove process-related and product-related impurities from a sample during lentivirus production and purification. [Background technology]

[0004] With FDA approvals recently expanding to ex vivo cell therapies such as ABECMA® and CARVYKTI® for multiple myeloma, BREYANZI® for B-cell lymphoma, SKYSONA® for cerebral adrenoleukodystrophy, and ZYNTEGLO® for β-thalassemia, lentiviruses are attracting public attention as replication-deficient viral vectors for gene therapy and cell therapy. Lentiviral vectors (LVVs) contain 8.5-9kb single-stranded RNA within a capsid, encased in a coat 80-100nm in diameter. The envelope displays many copies of vesicular stomatitis virus glycoprotein G (VSV-G), which affects the virus's stability and cell targeting. This VSV-G protein interacts with the low-density lipoprotein receptor (LDL-R) on the target cell membrane, mediating viral infection through membrane fusion.

[0005] Despite its clinical relevance, LVV has not yet benefited from established biomanufacturing platforms. In particular, the downstream segment of bioprocessing remains a significant bottleneck in the industry, resulting in long waiting times and high costs for patients (up to $2.8 million per treatment). Traditionally, LVV is produced by transfecting adherent or suspension HEK293 cells with the target gene, along with packaging and transfer plasmids. After virus production, mammalian cells and debris are removed from the cell culture medium (CCF) by centrifugation and / or microfiltration. Following clarification, various purification processes are employed on membranes, nanofibers, monoliths, or resins, including tangential flow filtration (TFF), size exclusion chromatography (SEC), ultracentrifugation, PEG precipitation, and anion exchange chromatography (AEX). However, these unit operations suffer from long process times, low scalability, and limitations in LVV particle recovery and host cell protein (HCP) removal.

[0006] Industrial layouts for the production of adeno-associated viruses (AAVs), recently integrated with layouts inspired by established downstream process platforms for monoclonal antibodies, highlight the crucial role of affinity chromatography in the purification of viral vectors for cell therapy and gene therapy. A considerable number of affinity adsorbents are available for AAVs, including panselective and cellotype-specific resins. While these resins yield excellent product yields (50-60%) and purity, they require harsh elution conditions (pH approximately 2-2.5) which can lead to product degradation and aggregation, resulting in loss of transduction activity. Conversely, commercial processes for LVV purification are not well developed. Heparin has been applied as a pseudoaffinity ligand, with reported recovery rates of approximately 20-50% of HCP and HCP removal rates of 95%. However, since heparin is extracted from animal tissue, there are concerns about contamination, and therefore its use is not recommended in Good Manufacturing Processes (GMP) processes. Furthermore, heparin-based adsorbents are not compatible with typical clean-in-place (CIP) procedures used in bio-manufacturing. This is because these adsorbents rapidly lose selectivity and more than 50% of their binding capacity after the initial caustic wash. Other affinity strategies described in the literature rely on the expression of affinity tags on the LVV envelope. Histidine-tagged LVV particles can be purified by immobilized metal affinity chromatography. Monoliths functionalized with sodium iminodiacetate and nickel showed an elution efficiency of 69%, but the binding capacity was moderate (6.7 × 10 LVV units transduction to cells per 1 mL of adsorbent (TU / mL)). 8However, using imidazole in the elution buffer inactivates the virus. In another approach, streptavidin-functionalized magnetic beads have been used to purify LVV particles presenting cTag8, yielding over 60% with 3-log reduction of host cell DNA and 2-log reduction of HCP, but the yield dropped to 20% when the ligand was transferred to a monolith. More recently, single-domain camel antibody (V H Using the library from H), we identified ligands that target the vesicular stomatitis virus G (VSV-G) protein. These ligands are now commercially available in chromatographic resin format (CaptureSelect® Lenti VSVG Affinity Matrix), but the number of virus particles (vp / mL) per 1 mL of resin is approximately 10 11 While it provides a certain binding capacity and good product purity, it can denature the product and requires elution conditions (0.8M arginine) that cannot withstand the limitations of the caustic (10-25mM NaOH aqueous solution) CIP process. [Overview of the project]

[0007] Embodiments of the present disclosure include peptides for purifying lentiviruses from a sample. In some embodiments, the peptide comprises at least one cationic amino acid and / or at least one anionic amino acid, at least one aromatic amino acid, and one or more aliphatic amino acids.

[0008] In some embodiments, the peptide is 5 to 20 amino acids long. In some embodiments, the peptide is at least 8 amino acids long. In some embodiments, the peptide is 8 amino acids long. In some embodiments, the peptide is 12 to 20 amino acids long.

[0009] In some embodiments, the peptide comprises at least one cationic amino acid and at least one anionic amino acid. In some embodiments, the cationic amino acid is adjacent to the anionic amino acid.

[0010] In some embodiments, each of the at least one cationic amino acids is independently selected from histidine, lysine, and arginine. In some embodiments, the at least one cationic amino acid is lysine.

[0011] In some embodiments, each of the at least one anionic amino acids is independently selected from aspartic acid and glutamic acid. In some embodiments, the at least one anionic amino acid is glutamic acid.

[0012] In some embodiments, each of the at least one aromatic amino acids is independently selected from histidine, phenylalanine, tyrosine, and tryptophan. In some embodiments, the at least one aromatic amino acid is phenylalanine. In some embodiments, the at least one aromatic amino acid is histidine.

[0013] In some embodiments, the one or more aliphatic amino acids are selected from alanine, glycine, isoleucine, leucine, proline, and valine.

[0014] In some embodiments, the peptide does not contain asparagine, glutamine, and / or tryptophan.

[0015] In some embodiments, the peptide is a cyclic peptide. In some embodiments, the cyclic peptide is cyclized by a disulfide bond between two cysteine ​​residues.

[0016] In some embodiments, the peptide comprises an amino acid sequence having at least 80% sequence identity with one of SEQ ID NOs: 1 to 86. In some embodiments, the peptide comprises an amino acid sequence having one, two, three, four, or five substitutions compared to one of SEQ ID NOs: 1 to 86. In some embodiments, the peptide comprises an amino acid sequence described in any of SEQ ID NOs: 1 to 86.

[0017] In some embodiments, the peptide further comprises a linker. In some embodiments, the linker is bonded to the C-terminus of the peptide and is a glycine-rich linker.

[0018] In some embodiments, the lentivirus is a vesicular stomatitis virus glycoprotein (VSV-G) pseudotype lentivirus.

[0019] In some embodiments, the peptide binds to the VSV-G envelope protein. In some embodiments, the peptide targets the binding site of the low-density lipoprotein receptor LDL-R on VSV-G.

[0020] Embodiments of this disclosure include compositions for purifying lentiviruses from a sample, which also include compositions comprising at least one peptide as disclosed herein.

[0021] In some embodiments, at least one peptide is bound to a solid support. In some embodiments, the solid support includes non-porous or porous particles, membranes, plastic surfaces, fibers, woven or nonwoven fiber mats, hydrogels, microplates, monoliths, and / or microfluidic devices. In some embodiments, the solid support includes polymethacrylate, polyolefin, polyester, polystyrene, polysaccharides, polyvinyl ether, iron oxide, silica, titania, and / or zirconia.

[0022] Further embodiments of this disclosure provide adsorbents comprising at least one peptide or composition as disclosed herein. In some embodiments, the adsorbent has a binding capacity of at least 10 LVV units (TU / mL) transduction to cells per 1 mL of adsorbent. 8 In some embodiments, the adsorbent has a binding capacity of at least 10 LVV units (TU / mL) transduction to cells per 1 mL of adsorbent. 9 That is the case.

[0023] In some embodiments, the adsorbent contains at least 10 LVV units (TU / mL·min) transduction to cells per 1 mL of adsorbent and per 1 minute of purification time. 8 This results in increased productivity. In some embodiments, the adsorbent contains at least 10 LVV units (TU / mL·min) transduction to cells per 1 mL of adsorbent and per 1 minute of purification time. 9 It brings productivity.

[0024] Additional embodiments of the present disclosure include methods for purifying lentiviruses from a sample. In some embodiments, the method includes contacting a sample containing the lentivirus with at least one peptide, composition, or adsorbent, as disclosed herein, wherein the at least one peptide binds to the lentivirus and elutes the lentivirus from the at least one peptide.

[0025] In some embodiments, the sample is a biological fluid. In some embodiments, the biological fluid is a cell culture medium. In some embodiments, the biological fluid includes a supernatant and / or cell lysates. In some embodiments, the biological fluid is derived from a virus-producing cell line.

[0026] In some embodiments, the lentivirus is a vesicular stomatitis virus glycoprotein (VSV-G) pseudotype lentivirus.

[0027] In some embodiments, the elution is performed at a pH of approximately 6.0 to 8.0. In some embodiments, the method further includes a washing step before eluting the lentivirus.

[0028] In some embodiments, the method reduces host cell proteins by at least 100 times.

[0029] In some embodiments, the method yields at least 35% of the LVV units transductioned into the cells.

[0030] In some embodiments, the method ensures that at least 10 LVV units are transductioned to cells per milliliter of adsorbent used and per minute of purification time. 8 Productivity can be obtained.

[0031] Embodiments of the present disclosure also include lentiviruses purified using any of the methods described herein. In some embodiments, the lentiviruses have a transduction activity of at least 35%. [Brief explanation of the drawing]

[0032] [Figure 1]The goal was to identify candidate LV-binding peptide sequences. (A) A screening mix was initially prepared containing red fluorescently labeled HEK293 HCP (titer 0.3 mg / mL) and either green fluorescently labeled LVV (titer approximately 109 TU / mL) or green fluorescently labeled VSV-G protein (EVSV-G and FLVSV-G) (titer 0.2 mg / mL). (B) Aliquots of 100 μL of 8-mer peptide-ChemMatrix library beads, equilibrated in 20 mM PBS with 75 mM NaCl (pH 6.5), were incubated in 250 μL of the screening mix for 30 minutes at room temperature. (C) The beads were washed and individually sent to a microfluidic bead sorting device connected to a fluorescence microscope. In the device, (D) each bead exhibiting high green fluorescence emission and a green-to-red signal ratio was retained, and all other beads were discarded. (E) All retained beads were washed for 5 minutes with 20 mM citrate buffer containing 0.5 M MgCl2 (pH 6.0) and imaged again. (F) Beads that did not show loss of green fluorescence were discarded, while (G) beads that lost the fluorescence signal were collected in a 96-well plate. (H) After washing with 0.1 M glycine (pH 2.5), water, and 30% acetonitrile (v / v) in water, the collected beads were individually analyzed by Edman degradation using a PPSQ-33A protein sequencer (Shimadzu, Kyoto, Japan) to identify the supported peptide sequences. Finally, (I) the libraries were screened against complete LVV particles, EVSV-G ectodomains, and FLVSV-G FLs, and the identified sequences were grouped using Weblogo in homology plots. [Figure 2]The values ​​were obtained by purifying LVV from HEK293 CCF (LVV titer approximately 1010 vp / mL (equivalent to approximately 108 TU / mL), HCP titer approximately 0.3 mg / mL) using GKEAAFAA-Poros (SEQ ID NO: 3) resin. These values ​​represent the LVV recovery rate (LVV units transduction into cells) and HEK293 HCP removal (orange histogram) measured in HT1080 cells by p24 ELISA (total particles), qPCR (total number of encapsulated transgenes), and transduction assay. The loading process was performed at RT 1 minute. Yield values ​​were measured by p24 ELISA (green histogram), RT-qPCR (blue histogram), and transduction activity on HT1080 cells (dark blue-green histogram). [Figure 3] This complex is formed by VSV-G and the peptides (A)C-cyclo[GSRQFVADSDRD]C-GSG (red, SEQ ID NO: 87), (B)C-cyclo[GSRSFVGDSDRD]C-GSG (reddish-purple, SEQ ID NO: 88), (C)FEKISAAE-GSG (green, SEQ ID NO: 89), (D)FEKISNAE-GSG (blue, SEQ ID NO: 90), (E)GKEAAFAA-GSG (dark blue-green, SEQ ID NO: 91), (F)SRQFVCDSDRD-GSG (yellow, SEQ ID NO: 93), (G)SRTFVCDSDRD-GSG (orange, SEQ ID NO: 95), and (H)SVFRIGLSD-GSG (mustard yellow, SEQ ID NO: 97). VSV-G is shown in gray, LDL-R in a grain-like image, amino acids that interact on VSV-G are shown in dark blue, and amino acids that interact on LDL-R are shown in red. The presumed binding sites identified by draggability testing of the solvent-contactable surface of VSV-G are shown as grain-like pale blue, pale green, and pale brown surfaces. [Figure 4A]This is a breakthrough curve obtained by loading HEK293 CCCF (LVV titer approximately 1010 vp / mL (equivalent to approximately 108 TU / mL), HCP titer approximately 0.3 mg / mL) onto FEKISNAE (SEQ ID NO: 1)-Poros resin with a residence time of either 1 or 2 minutes. The LVV titer in the eluate was measured by real-time qPCR (vg / mL) and transduction assay (TU / mL). [Figure 4B] This is a breakthrough curve obtained by loading HEK293 CCCF (LVV titer approximately 1010 vp / mL (equivalent to approximately 108 TU / mL), HCP titer approximately 0.3 mg / mL) onto GKEAAFAA (SEQ ID NO: 3)-Poros resin with a residence time of either 1 or 2 minutes. The LVV titer in the eluate was measured by real-time qPCR (vg / mL) and transduction assay (TU / mL). [Figure 4C] This is a breakthrough curve obtained by loading HEK293 CCCF (LVV titer approximately 1010 vp / mL (equivalent to approximately 108 TU / mL), HCP titer approximately 0.3 mg / mL) onto SRAFVGDADRD (SEQ ID NO: 4)-Poros resin with a residence time of either 1 or 2 minutes. The LVV titer in the eluate was measured by real-time qPCR (vg / mL) and transduction assay (TU / mL). [Figure 4D] This is a breakthrough curve obtained by loading HEK293 CCCF (LVV titer approximately 1010 vp / mL (equivalent to approximately 108 TU / mL), HCP titer approximately 0.3 mg / mL) into SFVRIGLSD (SEQ ID NO: 5)-Poros resin with a residence time of either 1 minute or 2 minutes. The LVV titer in the eluate was measured by real-time qPCR (vg / mL) and transduction assay (TU / mL). [Figure 5A] The caustic stability test of GKEAAFAA (SEQ ID NO: 3)-Poros resin was performed as a continuous cycle, with LVV purification from HEK293 CCF, along with intermediate CIP, using 0.5M NaOH (15CV with RT1 min, followed by a static contact time of 30 minutes). [Figure 5B]The caustic stability test of FEKISAAE (SEQ ID NO: 2)-Poros resin was performed as a continuous cycle, with LVV purification from HEK293 CCF, along with intermediate CIP, using 0.5M NaOH (15 CV with RT1 minute, followed by a static contact time of 30 minutes). [Figure 5C] The caustic stability test of SRAFVGDADRD (SEQ ID NO: 4)-Poros resin was performed as a continuous cycle, with LVV purification from HEK293 CCF, along with intermediate CIP, using 0.5M NaOH (15 CV with RT1 minute, followed by a static contact time of 30 minutes). [Figure 5D] The caustic stability test of SFVRIGLSD (SEQ ID NO: 5)-Poros resin was performed as a continuous cycle of LVV purification from HEK293 CCF, along with intermediate CIP, using 0.5M NaOH (15CV with RT1 min followed by a static contact time of 30 minutes). [Figure 6] This describes the relative infectivity of LVV samples in chromatography buffers. LVV particles were incubated for 30 minutes at 25°C with an initial titer of 5.0 107 TU / mL in citrate-based, phosphate-based, and histidine-based buffers with varying pH levels and added salts (various concentrations of NaCl or MgCl2). The relative infectivity titer was calculated as the ratio of the transduction activity of the LVV sample in HT1080 cells to the transduction activity of the LVV particles in DMEM medium. [Figure 7] This is a calibration curve for determining the surface density of primary amine groups on modified Poros resin. Data points were obtained by analyzing aqueous solutions of ethanolamine at various concentrations using the Kaiser (ninhydrin) test. [Figure 8A]This is a chromatogram obtained by purifying LVV from clarified HEK293 cell culture recovery. LVV particles were bound and eluted using a peptide-based adsorbent, GKEAAFAA (SEQ ID NO: 3)-Poros resin, from HEK293 CCCF (LVV titer approximately 1010 vp / mL (equivalent to approximately 108 TU / mL), HCP titer approximately 0.3 mg / mL). Equilibration and washing steps were performed using 50 mM PIPES buffer with 100 mM NaCl (pH 7.4) (RT: 1 min). Elution was performed using 50 mM PIPES buffer with 0.65 M NaCl (pH 7.4) (RT: 1 min). [Figure 8B] This is a chromatogram obtained by purifying LVV from the recovered clarified HEK293 cell culture. LVV particles were bound and eluted using a peptide-based adsorbent, EHFEHWSE (SEQ ID NO: 12)-Poros resin, from HEK293 CCCF (LVV titer approximately 1010 vp / mL (equivalent to approximately 108 TU / mL), HCP titer approximately 0.3 mg / mL). Equilibration and washing steps were performed using 50 mM PIPES buffer with 100 mM NaCl (pH 7.4) (RT: 1 min). Elution was performed using 50 mM PIPES buffer with 0.65 M NaCl (pH 7.4) (RT: 1 min). [Figure 8C] This is a chromatogram obtained by purifying LVV from clarified HEK293 cell culture recovery. LVV particles were bound and eluted using a peptide-based adsorbent, EWKAAFIW (SEQ ID NO: 8)-Poros resin, from HEK293 CCCF (LVV titer approximately 1010 vp / mL (equivalent to approximately 108 TU / mL), HCP titer approximately 0.3 mg / mL). Equilibration and washing steps were performed using 50 mM PIPES buffer with 100 mM NaCl (pH 7.4) (RT: 1 min). Elution was performed using 50 mM PIPES buffer with 0.65 M NaCl (pH 7.4) (RT: 1 min). [Figure 8D]This is a chromatogram obtained by purifying LVV from clarified HEK293 cell culture recovery. LVV particles were bound and eluted using the peptide-based adsorbent FEKISNAE (SEQ ID NO: 1)-Poros resin from HEK293 CCCF (LVV titer approximately 1010 vp / mL (equivalent to approximately 108 TU / mL), HCP titer approximately 0.3 mg / mL). Equilibration and washing steps were performed using 50 mM PIPES buffer with 100 mM NaCl (pH 7.4) (RT: 1 min). Elution was performed using 50 mM PIPES buffer with 0.65 M NaCl (pH 7.4) (RT: 1 min). [Figure 8E] This is a chromatogram obtained by purifying LVV from clarified HEK293 cell culture recovery. LVV particles were bound and eluted using a peptide-based adsorbent, SKSAAEHE (SEQ ID NO: 6)-Poros resin, from HEK293 CCCF (LVV titer approximately 1010 vp / mL (equivalent to approximately 108 TU / mL), HCP titer approximately 0.3 mg / mL). Equilibration and washing steps were performed using 50 mM PIPES buffer with 100 mM NaCl (pH 7.4) (RT: 1 min). Elution was performed using 50 mM PIPES buffer with 0.65 M NaCl (pH 7.4) (RT: 1 min). [Figure 9] This complex is formed by VSV-G and the (A)CR2 domain (PDB ID: 5OYL) and (B)CR3 domain (PDB ID: 5OY9) of LDL-R. VSV-G is shown as light blue, and LDL-R as a grain-like image. Interacting amino acids on VSV-G are shown as dark blue, and interacting amino acids on LDL-R are shown as red. [Figure 10]This is an AKTA chromatogram obtained by loading a clarified HEK293 cell culture medium containing Cas9 / GFP-LVV (LVV titer: 2.02.106 TU / mL, HCP titer: 0.05 mg / mL) onto GKEAAFAA (SEQ ID NO: 3)-Poros® resin. The resin was packed into a 1 mL column, equilibrated with 100 mM NaCl (pH 7.4) in 50 mM PIPES buffer, loaded with 150 mL of culture medium with a 1 min RT, washed, and then eluted with 0.650 M NaCl (pH 7.4) in 50 mM PIPES with a 1 min RT. The UV absorbance of the eluted product was monitored at 280 nm. [Figure 11] This is a caustic stability test of GKEAAFAA-Poros® (SEQ ID NO: 3) resin, performed as a continuous cycle of LVV purification from HEK293 CCF with intermediate CIP using 0.5M NaOH (15 CV at RT-1 minute followed by a static contact time of 15 minutes). (A) Amount of virus and HCP removed from the resin. (B) Process yield determined by transduction assay and RT-qPCR in HT1080 cells. [Figure 12] The titer of LVV units (TU / mL) transduction into cells is represented by the titer (histogram) measured by a transduction assay and the HCP titer (red dots) measured by ELISA obtained by transient transfection of HEK293 cells cultured in three cell culture media, namely Peak Expression, BalanCD, and LV-Max, using GOI plasmids (A) pALD-LentiEGFP-K and (B) dCAS9-VP64-GFP. [Figure 13]This study compares the transfection of HEK293F cells cultured in BalanCD medium with three transfection reagents—TransIT, PEIpro, and LV-Max—to produce LVV by transient transfection with the plasmids pALD-LentiEGFP-K, pALD-Rev-K, pALD-VSV-GK, and pALD-GagPol-K. Cells were transfected with TransIT and LV-Max once they reached a density of 4·10⁶ cells / mL (viability over 98%), while cells were transfected with PEIpro once they reached a density of 2.5·10⁶ cells / mL (viability over 98%). The plasmid:transfection reagent ratio (m / v) and total plasmid concentration per 1 mL of cell culture medium were as follows: (A) 1:3 and 1 μg / mL, (B) 1:1.85 and 1.6 μg / mL, (C) 1:3 and 3 μg / mL, (D) 1:3 and 1.6 μg / mL, (E) 1:3 and 2.5 μg / mL, (F) 1:1 and 2.5 μg / mL, and (G) 1:2.5 and 2.5 μg / mL. [Figure 14] This is a confocal image of GKEAAFAA (SEQ ID NO: 3)-Poros® beads extracted from the tip of a column loaded with pure LVV particles labeled with the green fluorescent dye Syto13. [Modes for carrying out the invention]

[0033] Lentiviral vectors have rapidly become an essential tool for generating life-saving cell therapies. However, their manufacturing technologies are in their infancy and can only provide limited production volumes, so the application of these therapies is limited to a small subset of patients living in developed countries. While access to healthcare depends on many factors, introducing biomanufacturing technologies that are productive, robust, and scalable at affordable prices is crucial for delivering advanced therapies to a wider patient population worldwide. Disclosed herein is a collection of peptide ligands for purifying LVV by affinity chromatography. By incorporating criteria for peptide sequence affinity, selectivity, and stability under various user-defined conditions, ligands with unique combinations of high binding capacity, impurity removal, yield of LVV units transduced into cells, and longevity were identified.

[0034] As further described herein, these criteria were applied to the experimental identification of suitable peptides and the in silico discovery of pseudotyped VSV-G-targeting peptides. Among the sequences identified by library screening, for GKEAAFAA (SEQ ID NO: 3), the binding capacity was 5·10 9 TU per mL of resin, the yield of LVV units transduced into cells was 60 - 70%, the HCP was reduced by more than 200-fold, and stability against caustic washing was also observed. Similarly, among the in silico-designed sequences, for the alkali-stable SRAFVGDADRD (SEQ ID NO: 4) and SFVRIGLSD (SEQ ID NO: 5), approximately 6.24·10 9Binding capacity of TU / mL, yield of 38–45%, and HCP removal exceeding 200-fold were observed. These ligands can be produced as short peptides at an affordable price and on a scalable basis. Recent studies have shown that when produced on a scale of 10 kg or more, the cost of an 8-mer peptide can be as low as $60 per gram. Considering that approximately 25 grams of the peptide are needed to functionalize one liter of resin, the cost of resin functionalized with this peptide would range from $7.5 to $9,000 per liter, providing a competitive alternative to affinity resins that rely on protein ligands.

[0035] In addition, experiments were conducted to optimize the parameters that affect the purification performance of these peptide ligands, including the material composition and morphology of the chromatography matrix, as well as the density and presentation of the ligands. Therefore, the design of affinity adsorbents was optimized by using the lead peptides GKEAAFAA (SEQ ID NO: 3), FEKISNAE (SEQ ID NO: 1), and SRAFVGDADRD (SEQ ID NO: 4) in combination with matrix assemblies. Specifically, various resin materials (i.e., polystyrene-divinylbenzene, polymethyl methacrylate, polyvinyl ether, and agarose), particle sizes (45-90 μm), pore sizes (40-1000 nm), and functional densities (0.02-0.1 mmol of peptide per 1 mL of resin) were supplied. In addition, various materials (cellulose and Natrix® grafted with polyethyleneimine (PEI), and inert polymer webs impregnated with porous polyacrylamide hydrogel) were tested for membranes, pore size (approximately 0.3 μm), and functional density (approximately 0.1 mmol of peptide per 1 mL of membrane). The resulting adsorbents differed significantly in terms of binding capacity, LVV yield, and LVV purity. Resins and membranes that functioned well, and were functionalized with peptides, exhibited high capacity (in these resins, the transduction of LVV units (TU / mL) per 1 mL of adsorbent was 5-10 9 In that membrane, 7.10 8 ), and productivity (2.9·10 in that resin) 9TU / mL·min, 4.5-10 in that membrane 8 It features a TU / mL·min ratio, and its eluate allows for a substantial reduction (110-170 times) of LVV units transduction to cells and impurities. As further described herein, GKEAAFAA (SEQ ID NO: 3)-Poros® resin was incorporated into the LVV purification process in four steps: (i) clarification and nuclease treatment, (ii) affinity-based capture in binding and elution modes, (iii) polishing in flow-through mode, and (iv) formulation by limit filtration / dialysis filtration and sterile filtration. These processes yielded a 33% yield, a residual HCP level of 200 ng per mL, and 1.25-10 active LVV particles per hour and per liter of resin. 14 This production capacity has become possible. The results presented herein indicate that this technology can be used for large-scale LVV manufacturing.

[0036] 1.Definition Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art. In case of any conflict, including definitions, this document shall prevail. Preferred methods and materials are described below, but similar or equivalent methods and materials to those described herein may also be used in the implementation or testing of this disclosure. The phrase "in one embodiment" may, but not necessarily, refer to the same embodiment as used herein. Furthermore, the phrase "in another embodiment" may, but not necessarily, refer to a different embodiment as used herein. Thus, various embodiments of the present invention may be easily combined without departing from the scope or spirit of the invention, as described below. All publications, patent applications, patents and other references referenced herein are incorporated by reference in their entirety. The materials, methods and examples disclosed herein are illustrative and not intended to limit the scope.

[0037] The terms “comprise(s),” “include(s),” “having,” “has,” “can,” and “contain(s),” and their variations, when used herein, are intended to be open-ended transitional phrases, terms, or words that do not preclude the possibility of further actions or structures. The singular forms “a,” “and,” and “the” include multiple referents unless otherwise clearly indicated in the context. This disclosure also contemplates other embodiments “comprising,” “consisting of,” and “essentially consisting of,” the embodiments or elements described herein, whether expressly described or not.

[0038] Where numerical ranges are specified herein, intermediate values ​​of the same precision are explicitly assumed. For example, for the range 6–9, the digits 7 and 8 are intended in addition to 6 and 9, and for the range 6.0–7.0, the digits 6.0, 6.1, 6.2, 6.3, 6.4, 6.5, 6.6, 6.7, 6.8, 6.9, and 7.0 are explicitly intended.

[0039] As used herein, "correlated with ~" means "compared to ~".

[0040] As used herein, "peptide" and "polypeptide" generally refer to polymer compounds of two or more amino acids linked via a main chain by a peptide amide bond (--C(O)NH--), unless otherwise specified. The term "peptide" typically refers to a short amino acid polymer (e.g., a chain with fewer than 25 amino acids), while the term "polypeptide" typically refers to a longer amino acid polymer (e.g., a chain with more than 25 amino acids).

[0041] As used herein, “sequence identity” generally refers to the degree to which two polymer sequences (e.g., peptides, polypeptides, nucleic acids, etc.) have the same continuous composition of monomer subunits. “Percent sequence identity” (or “percent sequence similarity”) is calculated by (1) comparing two sequences that are optimally aligned across a comparison window (e.g., the length of the longer sequence, the length of the shorter sequence, or a specified window), (2) determining the number of positions containing identical (or similar) monomers (e.g., the number of times the same amino acid occurs in both sequences, or the number of times similar amino acids occur in both sequences) to obtain the number of matched positions, (3) dividing the number of matched positions by the total number of positions in the comparison window (e.g., the length of the longer sequence, the length of the shorter sequence, or a specified window), and (4) multiplying the result by 100 to obtain the percentage sequence identity or percentage sequence similarity. For example, if both peptide A and peptide B are 20 amino acid long and have all identical amino acids except for one position, then peptide A and peptide B have 95% sequence identity. If non-identical amino acids share the same biophysical characteristics (for example, both are acidic), then peptides A and B will have 100% sequence similarity. As another example, if peptide C is 20 amino acids long and peptide D is 15 amino acids long, and 14 of the 15 amino acids in peptide D are identical to a portion of peptide C, then peptides C and D have 70% sequence identity, but peptide D has 93.3% sequence identity with respect to the optimal comparison window of peptide C. For the purposes of calculating “percent sequence identity” (or “percent sequence similarity”), any gap in aligned sequences is treated as a mismatch at that position.

[0042] The term "sequence similarity" refers to the degree to which two polymer sequences (e.g., peptides, polypeptides, nucleic acids, etc.) have similar polymer sequences. For example, similar amino acids are amino acids that share the same biophysical characteristics and belong to the same family, e.g., acidic amino acids (e.g., aspartic acid (Asp, D), glutamic acid (Glu, E)), basic amino acids (e.g., lysine (Lys, K), arginine (Arg, R), histidine (His, H)), nonpolar amino acids (e.g., alanine (Ala, A), valine (Val, V), leucine (Leu, L), isoleucine (Ile, I), proline (Pro, P), phenylalanine (Phe, F), methionine (Met, M), tryptophan (Trp, W)), polar uncharged amino acids (e.g., They can be grouped into glycine (Gly, G), asparagine (Asn, N), glutamine (Gln, Q), cysteine ​​(Cys, C), serine (Ser, S), threonine (Thr, T), tyrosine (Tyr, Y)), aliphatic amino acids (e.g., alanine (Ala, A), valine (Val, V), leucine (Leu, L), isoleucine (Ile, I), glycine (Gly, G), and possibly methionine (Met, M)), and aromatic amino acids (histidine (H or His), phenylalanine (F or Phe), tyrosine (Y or Tyr), and tryptophan (W or Trp)).

[0043] As used herein, the terms “purified” or “to purify” refer to the removal of components (e.g., contaminants) from a sample. For example, a virus is purified by removing process-related and product-related contaminants that cause contamination. Process-related contaminants include proteins and nucleic acids produced by the host expressing the virus, and may also include viruses other than the target virus. Product-related contaminants include fragments of the target virus, and may also include transgenes that are not encapsulated or are misencapsulated, and may also include aggregates of the target virus as well as other proteins and nucleic acids. Removal of these contaminants increases the percentage of active virus in the sample.

[0044] As used herein, the term “sample” refers to any composition or mixture containing the target biological substance. The sample may originate from a biological source or other source. Biological sources include eukaryotic and prokaryotic sources, such as bacterial, fungal, plant and animal cells, tissues and organs. The sample may also include diluents, buffers, surfactants, hybrids, and fragments known to be mixed with the target biological substance. The sample may be “partially purified” (e.g., by performing one or more purification steps, such as filtration) or may be obtained directly from host cells or organisms producing the target molecule (e.g., the sample may include recovered cell culture medium).

[0045] As used herein, the terms “target” or “targeted biological substance” generally refer to any of the target proteins, peptides, polypeptides, nucleic acids, ribonucleoprotein complexes, nucleic acid constructs, supramolecular constructs, viruses, viral constructs, virus-like particles, cells, organelles, small molecules, and combinations thereof that may be present in a sample (e.g., a biological fluid) containing one or more process-related impurities and / or product-related impurities. In some embodiments, the target or targeted biological substance is a viral vector (e.g., lentivirus, adeno-associated virus) or virus-like particles.

[0046] As used herein, the term “host cell” refers to any cell line used to produce LVV.

[0047] As used herein, the term “encapsulated transgene” refers to any RNA sequence that encodes one or more products (e.g., proteins).

[0048] As used herein, the term “LVV units transduction into cells” refers to the number of active lentiviral vectors in a sample required to transduction into a cell, that is, to cause the cell to express one or more proteins encoded by a transgene encapsulated in a lentivirus.

[0049] As used herein, the term “host cell protein” or “HCP” refers to any protein produced or encoded by a host cell that is irrelevant to the intended product. HCPs are generally undesirable in the final drug substance.

[0050] As used herein, the term "host cell DNA" refers to any nucleic acid produced by the host cell that is unrelated to the intended product. Host cell DNA is generally undesirable as the final drug substance.

[0051] As used herein, the terms “plasmid DNA” or “pDNA” refer to any nucleic acid used to transfect host cells in order to produce LVV. pDNA is generally undesirable as the final drug substance.

[0052] As used herein, “mixture” comprises the target biological substance (preferably purified) and one or more contaminants or impurities. In some embodiments, the mixture is produced from host cells or host organisms expressing the target biological substance (either naturally or recombinantly). Examples of such mixtures include cell culture media, cell lysates, and clarified bulk (e.g., clarified cell culture supernatant).

[0053] 2. Peptides and compositions for purifying lentiviruses a. peptide Current platforms for lentivirus purification present several challenges: (i) concerns about product contamination and inactivation due to suboptimal process conditions, (ii) insufficient yield, and (iii) high costs and poor reusability. More recently, affinity ligands offering high binding capacity and product purity have been developed; however, these ligands require conditions that can denature the product and limit its reusability. In searching for robust affinity ligands for lentivirus purification, we developed a selection of specific peptides that selectively bind to lentiviruses, are elutable under near-physiological conditions, and can be reused multiple times without losing binding strength and selectivity, using experimental and in silico methods.

[0054] The peptides of this disclosure may be used for the purification of any lentivirus. Lentiviruses are a subfamily of enveloped retroviruses. This family includes, for example, human immunodeficiency viruses (e.g., HIV-1 and HIV-2), simian immunodeficiency virus (SIV), bovine immunodeficiency virus (BIV), feline immunodeficiency virus (FIV), canine arthritis encephalitis virus (CAEV), and equine infectious anemia virus (EIAV). Lentiviruses have a diploid RNA genome that is integrated into the host chromosome as proviral DNA for genome replication. Lentiviral infection is persistent due to their ability to integrate their genome into the host chromosome and to evade the host's immune response. Unlike other retroviruses, lentiviruses have the ability to efficiently infect and transduce non-proliferating cells (e.g., terminally differentiated cells).

[0055] Undesirable properties of lentiviruses can be removed or modified by recombination, thereby utilizing their beneficial characteristics for therapeutic purposes or as delivery vehicles. Thus, engineered lentiviruses can be produced that safely self-inactivate in a replication-deficient manner, while still retaining their beneficial ability to integrate into their host chromosomes, along with their beneficial ability to transduce into non-dividing cells for stable expression. In some embodiments, the peptides of this disclosure may be used to purify engineered or recombinant lentiviruses.

[0056] The engineered or recombinant lentivirus may be a pseudotyped lentivirus. When used herein in the context of viral vectors, the term pseudotyped refers to a viral vector whose cell-type specificity is modified by incorporating the envelope protein of an exogenous virus. This approach is well known in the art and is described, for example, in Bischof et al. (Methods Mol Biol. 2010;614:53-68). This approach can be used to alter host tropism and / or increase or decrease viral stability.

[0057] In some embodiments, the peptides described herein target the vesicular stomatitis virus glycoprotein (VSV-G) pseudotyped lentivirus. Lentiviral vectors are often pseudotyped with the vesicular stomatitis virus envelope glycoprotein (VSV-G). VSV-G pseudotyped lentiviral vectors often possess excellent mechanical stability, which allows for the spinocuration and production of high-titer vector stocks. The vesicular stomatitis virus (hereinafter abbreviated as VSV), also known as vesicular stomatitis virus Indiana strain (VSIV), is a virus of group V (e.g., negative sense ssRNA) viruses of the Rhabdoviridae family. The virus, its host, and the diseases induced by infection with the virus are well known in the art. VSV-G is a transmembrane glycoprotein known to efficiently bind to immature, non-infectious, envelope-deficient retrovirus-like particles produced by packaging cells that produce infectious pseudotyped viruses in vitro under cell-free conditions. The use of VSV-G to pseudotype lentiviruses is described, for example, in Burns et al. (Proc Natl Acad Sci USA. 1993; 90(17): 8033-8037).

[0058] The peptides of this disclosure may target any binding site found on their lentivirus, for example, a VSV-G pseudotype lentivirus. In some embodiments, the peptides bind to the VSV-G envelope protein. In some embodiments, the peptides target the binding site of the low-density lipoprotein receptor LDL-R on VSV-G.

[0059] In accordance with these embodiments, the disclosure provides peptides comprising at least one cationic amino acid and / or at least one anionic amino acid, at least one aromatic amino acid, and one or more aliphatic amino acids. In some embodiments, the peptide comprises at least one cationic amino acid and at least one negatively charged amino acid.

[0060] The at least one cationic amino acid and at least one anionic amino acid may be arranged along the length of the peptide in any orientation. In some embodiments, the cationic amino acid is adjacent to the anionic amino acid. For example, in some embodiments, the cationic amino acid is immediately C-terminal to the anionic amino acid. Alternatively, or in addition to this, the cationic amino acid is immediately N-terminal to the anionic amino acid. In some embodiments, the cationic amino acid is one or more amino acids away from the anionic amino acid.

[0061] The at least one cationic amino acid may be any amino acid exhibiting a naturally occurring or synthetic positively charged residue. In some embodiments, each of the at least one cationic amino acid is independently selected from histidine, lysine, and arginine. In some embodiments, the at least one cationic amino acid is lysine.

[0062] The at least one anionic amino acid may be any amino acid exhibiting a naturally occurring or synthetic negatively charged residue. In some embodiments, each of the at least one anionic amino acid is independently selected from aspartic acid and glutamic acid. In certain embodiments, the at least one anionic amino acid is glutamic acid.

[0063] In some embodiments, the peptide further comprises aromatic amino acids. In some embodiments, each of the at least one aromatic amino acid is independently selected from histidine, phenylalanine, tyrosine, and tryptophan. In certain embodiments, the at least one aromatic amino acid is phenylalanine. In certain embodiments, the at least one aromatic amino acid is histidine.

[0064] The peptide further comprises one or more aliphatic amino acids. These one or more aliphatic amino acids may be arranged in any orientation along the peptide. For example, one or more aliphatic amino acids may separate their aromatic amino acids from their cationic or anionic amino acids, or from both cationic and anionic amino acids. In some embodiments, these one or more aliphatic amino acids may be adjacent to each other, resulting in a series of two or more aliphatic amino acids along the length of the peptide. In some embodiments, these one or more aliphatic amino acids are selected from alanine, glycine, isoleucine, leucine, proline, and valine.

[0065] In some embodiments, the peptide does not contain asparagine, glutamine, tryptophan, or any combination thereof.

[0066] In some embodiments, the peptide has an amino acid sequence that is at least 80% sequence-identical to one of SEQ ID NOs: 1 to 86. In some embodiments, the peptide has an amino acid sequence that is at least 85% sequence-identical to one of SEQ ID NOs: 1 to 86. In some embodiments, the peptide has an amino acid sequence that is at least 90% sequence-identical to one of SEQ ID NOs: 1 to 86. In some embodiments, the peptide has an amino acid sequence that is at least 95% sequence-identical to one of SEQ ID NOs: 1 to 86. In some embodiments, the peptide has an amino acid sequence that contains one of SEQ ID NOs: 1 to 86.

[0067] In some embodiments, the peptide has an amino acid sequence that is at least 80% sequence-identical to one of SEQ ID NOs: 1 to 12. In some embodiments, the peptide has an amino acid sequence that is at least 85% sequence-identical to one of SEQ ID NOs: 1 to 12. In some embodiments, the peptide has an amino acid sequence that is at least 90% sequence-identical to one of SEQ ID NOs: 1 to 12. In some embodiments, the peptide has an amino acid sequence that is at least 95% sequence-identical to one of SEQ ID NOs: 1 to 12. In some embodiments, the peptide has an amino acid sequence that contains one of SEQ ID NOs: 1 to 12.

[0068] In some embodiments, the peptide has an amino acid sequence that is at least 80% sequence-identical to one of SEQ ID NOs: 1 to 5. In some embodiments, the peptide has an amino acid sequence that is at least 85% sequence-identical to one of SEQ ID NOs: 1 to 5. In some embodiments, the peptide has an amino acid sequence that is at least 90% sequence-identical to one of SEQ ID NOs: 1 to 5. In some embodiments, the peptide has an amino acid sequence that is at least 95% sequence-identical to one of SEQ ID NOs: 1 to 5. In some embodiments, the peptide has an amino acid sequence that contains one of SEQ ID NOs: 1 to 5.

[0069] In some embodiments, the peptide has an amino acid sequence that is at least 80% sequence-identical to at least one of SEQ ID NOs: 1, 3, and 4. In some embodiments, the peptide has an amino acid sequence that is at least 85% sequence-identical to at least one of SEQ ID NOs: 1, 3, and 4. In some embodiments, the peptide has an amino acid sequence that is at least 90% sequence-identical to at least one of SEQ ID NOs: 1, 3, and 4. In some embodiments, the peptide has an amino acid sequence that is at least 95% sequence-identical to at least one of SEQ ID NOs: 1, 3, and 4. In some embodiments, the peptide has an amino acid sequence that is at least one of SEQ ID NOs: 1, 3, and 4.

[0070] In some embodiments, the peptide has an amino acid sequence having one, two, three, four, or five substitutions compared to at least one of SEQ ID NOs: 1-86. In some embodiments, the peptide has an amino acid sequence having one, two, three, four, or five substitutions compared to at least one of SEQ ID NOs: 1-12. In some embodiments, the peptide has amino acids having one, two, three, four, or five substitutions compared to at least one of SEQ ID NOs: 1-5. In some embodiments, the peptide has amino acids having one, two, three, four, or five substitutions compared to at least one of SEQ ID NOs: 1, 3, and 4.

[0071] Amino acid "substitution" or "replacement" refers to replacing one amino acid at a given position or residue within a polypeptide sequence with another amino acid at the same position or residue. Amino acid substitutions or replacements can be conserved, semi-conservative, or non-conservative. The expression "conservative amino acid substitution" or "conservative mutation" refers to replacing one amino acid with another amino acid that shares a common property (e.g., aromaticity, charge, polarity, etc.). A functional way to define common properties among individual amino acids is to analyze the normalized frequency of amino acid changes among corresponding proteins of the same species (Schulz and Schirmer, Principles of Protein Structure, Springer-Verlag, New York (1979)). According to such analysis, groups of amino acids may be defined when amino acids within a group preferentially exchange with each other, i.e., those groups have the most similar impact on the overall protein structure (Schulz and Schirmer, cited above). Examples of conserved amino acid substitutions include substitutions of amino acids with common properties, such as arginine to lysine and lysine to arginine substitutions that can maintain a positive charge, aspartic acid to glutamic acid and glutamic acid to aspartic acid substitutions that can maintain a negative charge, threonine to serine substitutions that can maintain a free -OH group, and asparagine to glutamine substitutions that can maintain a free -NH2 group. "Semi-conservative mutations" include amino acid substitutions of amino acids within the same broad group (e.g., aliphatic) but not within the same subgroup (e.g., polar or nonpolar). Examples include asparagine to aspartic acid substitution or lysine to asparagine substitution. "Non-conservative mutations" involve amino acid substitutions between various groups, such as tryptophan to lysine substitution or serine to phenylalanine substitution.

[0072] The peptides of this disclosure may be of any length that provides specificity to the target lentivirus. In some embodiments, the peptide is 5 to 20 amino acids long. In some embodiments, the peptide is 5 to 10 amino acids long. In some embodiments, the peptide is 5 to 15 amino acids long. In some embodiments, the peptide is 8 to 10 amino acids long. In some embodiments, the peptide is 8 to 15 amino acids long. In some embodiments, the peptide is 8 to 20 amino acids long. In some embodiments, the peptide is 10 to 15 amino acids long. In some embodiments, the peptide is 10 to 20 amino acids long. In some embodiments, the peptide is 12 to 20 amino acids long. The peptide may be 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 amino acids long. In certain embodiments, the peptide has a length of 8 to 14 amino acids. In certain embodiments, the peptide has a length of 8 amino acids.

[0073] In some embodiments, the peptide is a cyclic peptide. The peptide can be cyclized by any method available to those skilled in the art. For example, the N-terminus and C-terminus can be condensed to form a peptide bond by known procedures. The peptide of the present invention can also be cyclized by attaching functional groups present in the side chains of the amino acids within the peptide. Examples of functional groups that can form covalent bonds include -COOH and -OH, -COOH and -NH2, -COOH and -SH, and -SH and -SH. Examples of amino acid pairs that can be used to cyclize the peptide include Asp and Lys, Glu and Lys, Asp and Ser, Glu and Ser, Asp and Thr, Glu and Thr, Asp and Cys, Glu and Cys, and Cys and Cys. Other examples of amino acid residues that can form covalent bonds with each other include cysteine-like amino acids, such as Cys, hCys, β-methyl-Cys, and penicillamine (Pen), and non-proteinogenic alpha-amino acids having a valine structure substituted at the β position with a sulfanyl group (which can form disulfide crosslinks with each other). Preferred cysteine-like amino acid residues include Cys and Pen. In some embodiments, the peptide is cyclized by a disulfide bond between two cysteine ​​residues.

[0074] The group used to cyclize a peptide does not need to be an amino acid in the peptide itself. Examples of functional groups that can form a covalent bond with the amino terminus of a peptide include carboxylic acids and esters. Examples of functional groups that can form a covalent bond with the carboxyl terminus of a peptide include -OH, -SH, -NH2, and -NHR (wherein R is (C1-C6)alkyl, (C1-C6)alkenyl, and (C1-C6)alkynyl).

[0075] Preferably, the reaction conditions used to cyclize the peptide are mild enough not to degrade or otherwise damage the peptide. Suitable groups for protecting various functional groups as needed are well known in the art, as are various reaction schemes for preparing such protected molecules (e.g., Greene & Wuts, 1991, 2nd ed., John Wiley & Sons, NY).

[0076] Any of the peptides and related sequences of the present disclosure may be included in a cyclized peptide. Therefore, the peptide sequences of the present disclosure may include additional amino acids or other groups to facilitate cyclization, as described above. In some embodiments, the peptide is cyclized and includes an amino acid sequence that has at least 80% sequence identity with one of SEQ ID NOs: 13-16.

[0077] In some embodiments, the peptides obtained from this disclosure have a dissociation constant (K) with respect to a lentivirus (e.g., its VSV-G protein). D ) is about 10 -3 The value is less than or equal to M. In some embodiments, the peptide obtained from this disclosure has a dissociation constant (K) with respect to a lentivirus (e.g., its VSV-G protein). D ) is about 10 -4 The value is less than or equal to M. In some embodiments, the peptide obtained from this disclosure has a dissociation constant (K) with respect to a lentivirus (e.g., its VSV-G protein). D ) is about 10 -5 The value is less than or equal to M. In some embodiments, the peptide obtained from this disclosure has a dissociation constant (K) with respect to a lentivirus (e.g., its VSV-G protein). D ) is about 10 -6 It is M or less.

[0078] As those skilled in the art will recognize based on this disclosure, the peptides provided herein can be conjugated to a linker. In some embodiments, the linker can facilitate the presentation of the peptide on a solid support, thereby enabling better capture of the target lentivirus. In other embodiments, the peptides provided herein are not conjugated to a linker but can still bind to the target lentivirus and can be purified from cell culture by other means. In some embodiments, one or more of the peptides include a linker at the C-terminus of the peptide. Given that preferred linkers generally have sequences that result in flexible peptides, the linked peptide may have substantially any amino acid sequence. Low molecular weight amino acids, such as glycine, alanine, and serine, are commonly used in the production of flexible peptides. A variety of different linkers are commercially available and considered suitable for use, including, but are not limited to, glycine-serine polymers, glycine-alanine polymers, and alanine-serine polymers. In a given embodiment, the linker is [Gly-Ser-Gly] n This includes, in certain embodiments, the linker includes GSG and GGG.

[0079] b. Compositions and adsorbents Also described herein are compositions and adsorbents comprising one or more peptides of the Disclosure. In some embodiments, each peptide of the composition or adsorbent is conjugated to a support. The support may include, but is not limited to, particles, beads, plastic surfaces, resins, fibers, and / or membranes. In some embodiments, the solid support includes non-porous or porous particles, membranes, plastic surfaces, fibers, woven or nonwoven fiber mats, hydrogels, microplates, monoliths, and / or microfluidic devices. In some embodiments, the solid support includes fine particles and / or nanoparticles. In some embodiments, the solid support includes hydrogels. In some embodiments, the solid support includes membranes.

[0080] In some embodiments, the solid support comprises polymethacrylate, polyolefin, polyester, polystyrene, polysaccharide, polyvinyl ether, iron oxide, silica, titania, agarose, and / or zirconia. Each support may be made from any suitable material, including but not limited to synthetic or natural polymers, metals, and metal oxides. Some supports may be magnetic, for example, magnetic beads, fine particles, and / or nanoparticles. Suitable synthetic polymers include, but are not limited to, polymethacrylate, polysulfone, polyethersulfone, polyvinyl ether, and polyethylene glycol. Suitable natural polymers include, but are not limited to, cellulose, agarose, and chitosan. Suitable metal oxides include, but are not limited to, iron oxide, silica, titania, and zirconia.

[0081] In certain embodiments, the solid support comprises particles or beads. In some embodiments, the particles or beads have a diameter of about 50 μm, about 60 μm, about 70 μm, about 80 μm, about 90 μm, or about 100 μm. In certain embodiments, the solid support comprises porous particles. In some embodiments, the porous particles contain pores with a pore size of at least about 50 nm, at least about 60 nm, at least about 70 nm, at least about 80 nm, at least about 90 nm, or at least about 100 nm or more.

[0082] The ligand density on the solid support may vary. In some embodiments, the composition or adsorbent has a high ligand density (e.g., greater than 0.05 mmol / mL). In some embodiments, the composition or adsorbent has a relatively low ligand density (e.g., less than 0.05 mmol / mL).

[0083] In some embodiments, the composition or adsorbent comprises a single-type support made from a single-type support material, and all of the peptides are conjugated to the support formed from that single-type support material. In these embodiments, the composition or adsorbent may comprise one or more different types of peptides, each conjugated to a single-type support made from that single-type support material. In other embodiments, the composition or adsorbent comprises multiple types of supports. Each type of support may be made from the same type of support material or different types of support materials. In these embodiments, the composition or adsorbent may comprise one or more different types of peptides, each conjugated to a different type of support, as further described herein. In yet another embodiment, the peptides of the composition can be conjugated to a soluble compound, such as a stimulus-responsive polymer chain, to remove lentiviruses by affinity precipitation.

[0084] In some embodiments, the adsorbents obtained from this disclosure have a binding capacity to lentiviruses of at least about 10 LVV units (TU / mL) transduction to cells per 1 mL of adsorbent. 7 In some embodiments, the adsorbents obtained from this disclosure have a binding capacity to lentiviruses of at least about 10 LVV units (TU / mL) transduction to cells per 1 mL of adsorbent. 8 In some embodiments, the adsorbents obtained from this disclosure have a binding capacity to lentiviruses of at least about 10 LVV units (TU / mL) transduction to cells per 1 mL of adsorbent. 9 In some embodiments, the adsorbents obtained from this disclosure have a binding capacity to lentiviruses of at least about 10 LVV units (TU / mL) transduction to cells per 1 mL of adsorbent. 10In a given embodiment, the adsorbent obtained from this disclosure has a binding capacity to lentiviruses, with approximately 10 LVV units (TU / mL) transduction to cells per 1 mL of adsorbent. 9 ~10 10 That is the case.

[0085] In some embodiments, the adsorbent allows for at least 10 LVV units to be transductioned to cells per milliliter of adsorbent used and per minute of purification time. 7 A productivity of is obtained. In some embodiments, the adsorbent allows for at least 10 LVV units to be transductioned to cells per milliliter of adsorbent used and per minute of purification time. 8 A productivity of is obtained. In some embodiments, the adsorbent allows for at least 10 LVV units to be transductioned to cells per milliliter of adsorbent used and per minute of purification time. 9 Productivity can be obtained.

[0086] In some embodiments, the adsorbent reduces host cell proteins by at least 100 times. In some embodiments, the adsorbent reduces host cell proteins by at least 150 times. In some embodiments, the adsorbent reduces host cell proteins by at least 200 times.

[0087] 3.How to use As further described herein, the Disclosure also provides methods for purifying lentiviruses from a sample. In some embodiments, the sample is a biological fluid. Accordingly, the Disclosure provides methods for purifying lentiviruses from one or more product-related and / or process-related impurities or contaminants.

[0088] In some embodiments, the method comprises contacting a sample containing a target lentivirus with at least one peptide, composition, or adsorbent as described herein, wherein at least one peptide ligand of the Disclosure binds to the target lentivirus. According to these embodiments, the method comprises purifying the target lentivirus by eluting it from the at least one peptide. The method of the Disclosure may further comprise washing the composition or adsorbent to remove one or more product-related and / or process-related impurities or contaminants from the target lentivirus bound to the peptide ligand. In some embodiments, the method may be carried out under any binding conditions suitable for use with the peptide, composition, or adsorbent, including both static and dynamic binding conditions.

[0089] As further described herein, the peptides of this disclosure offer advantages over currently available compositions and methods for purifying lentiviruses. For example, currently available ligands used for lentivirus purification employ elution conditions that damage the eluted lentivirus product. In contrast, the peptide ligands of this disclosure release the bound target lentivirus under milder conditions, thereby not damaging its product.

[0090] The binding affinity of the peptide, composition, and / or adsorbent to the lentivirus can be altered by changing the properties and concentration of one or more product-related and / or process-related impurities or contaminants, the properties and concentration of the host cell protein, the composition, concentration, and pH of the mixture, and / or the loading conditions and residence time of the contact and washing steps, compared to one or more product-related and / or process-related impurities or contaminants. Any of these variable factors can be changed to suitable variable factors according to the methods of this disclosure, resulting in an increase or decrease in binding affinity as required in this disclosure.

[0091] In some embodiments, the contact step may include a low pH buffer of pH 5-9. In some embodiments, the contact step may include a low pH buffer of pH 5-8. In some embodiments, the contact step may include a low pH buffer of pH 5-7. In some embodiments, the contact step may include a low pH buffer of pH 6-9. In some embodiments, the contact step may include a low pH buffer of pH 6-8. In some embodiments, the contact step may include a low pH buffer of pH 6-7. In some embodiments, the contact step may include a low pH buffer of pH 7-9. In some embodiments, the contact step may include a low pH buffer of pH 7-8.

[0092] In some embodiments, the elution is carried out at a pH of about 5.0 to about 8.0. In some embodiments, the elution is carried out at a pH of about 5.0 to about 7.5. In some embodiments, the elution is carried out at a pH of about 5.0 to about 7.0. In some embodiments, the elution is carried out at a pH of about 5.0 to about 6.5. In some embodiments, the elution is carried out at a pH of about 5.0 to about 6.0. In some embodiments, the elution is carried out at a pH of about 5.0 to about 5.5. In some embodiments, the elution is carried out at a pH of about 5.5 to about 8.0. In some embodiments, the elution is carried out at a pH of about 6.0 to about 8.0. In some embodiments, the elution is carried out at a pH of about 6.5 to about 8.0. In some embodiments, the elution is carried out at a pH of about 7.0 to about 8.0. In some embodiments, the elution is carried out at a pH of about 7.5 to about 8.0. In some embodiments, the elution is carried out at a pH of about 6.0 to about 7.0. In some embodiments, the elution is carried out at a pH of about 5.5 to about 7.5.

[0093] In some embodiments, the yield of the transgene encapsulated in the lentiviral vector by the method of the present disclosure is at least 30%. In some embodiments, the yield of the encapsulated transgene by the method of the present disclosure is at least 35%. In some embodiments, the yield of the encapsulated transgene by the method of the present disclosure is at least 40%. In some embodiments, the yield of the encapsulated transgene by the method of the present disclosure is at least 45%. In some embodiments, the yield of the encapsulated transgene by the method of the present disclosure is at least 50%. In some embodiments, the yield of the encapsulated transgene by the method of the present disclosure is at least 55%. In some embodiments, the yield of the encapsulated transgene by the method of the present disclosure is at least 60%. In some embodiments, the yield of the encapsulated transgene by the method of the present disclosure is at least 70%. In some embodiments, the yield of the encapsulated transgene by the method of the present disclosure is at least 80%. In some embodiments, the yield of the encapsulated transgene by the method of the present disclosure is at least 90%.

[0094] In some embodiments, the method of the present disclosure yields at least 35% of LVV units transduction into cells. In some embodiments, the method of the present disclosure yields at least 40% of LVV units transduction into cells. In some embodiments, the method of the present disclosure yields at least 45% of LVV units transduction into cells. In some embodiments, the method of the present disclosure yields at least 50% of LVV units transduction into cells. In some embodiments, the method of the present disclosure yields at least 55% of LVV units transduction into cells. In some embodiments, the method of the present disclosure yields at least 60% of LVV units transduction into cells. In some embodiments, the method of the present disclosure yields at least 70% of LVV units transduction into cells. In some embodiments, the method of the present disclosure yields at least 80% of LVV units transduction into cells. In some embodiments, the method of the present disclosure yields at least 90% of LVV units transduction into cells.

[0095] In some embodiments, the method disclosed herein produces at least 10 LVV units transduction to cells per milliliter of adsorbent used and per minute of purification time. 7 This results in a productivity of at least 5 × 10 LVV units transduction to cells per milliliter of adsorbent and per minute of purification time by the method disclosed herein. 7 This results in a productivity of at least 10 LVV units transduction to cells per milliliter of adsorbent used and per minute of purification time. In some embodiments, the method of this disclosure yields at least 10 LVV units transduction to cells per milliliter of adsorbent used. 8 This results in a productivity of at least 5 × 10 LVV units transduction to cells per milliliter of adsorbent used and per minute of purification time. In some embodiments, the method disclosed herein yields at least 5 × 10 LVV units transduction to cells per milliliter of adsorbent used and per minute of purification time. 8This results in a productivity of at least 10 LVV units transduction to cells per milliliter of adsorbent used and per minute of purification time. In some embodiments, the method of this disclosure yields at least 10 LVV units transduction to cells per milliliter of adsorbent used. 9 This results in a productivity of at least 5 × 10 LVV units transduction to cells per milliliter of adsorbent used and per minute of purification time. In some embodiments, the method disclosed herein yields at least 5 × 10 LVV units transduction to cells per milliliter of adsorbent used and per minute of purification time. 9 This will become a source of productivity.

[0096] In some embodiments, the method of the Disclosure reduces host cell proteins by at least 80-fold (for example, compared to a method that does not use the peptide ligand of the Disclosure) when purifying a lentivirus. In some embodiments, the method of the Disclosure reduces host cell proteins by at least 100-fold. In some embodiments, the method of the Disclosure reduces host cell proteins by at least 150-fold. In some embodiments, the method of the Disclosure reduces host cell proteins by at least 200-fold. In some embodiments, the method of the Disclosure reduces host cell proteins by at least 250-fold.

[0097] Embodiments of the present disclosure also include lentiviruses purified using any of the methods described herein. In some embodiments, the lentivirus has a transduction activity of at least 35%. In some embodiments, the lentivirus has a transduction activity of at least 40%. In some embodiments, the lentivirus has a transduction activity of at least 45%. In some embodiments, the lentivirus has a transduction activity of at least 50%.

[0098] In some embodiments, the biological fluid is a cell culture medium. In some embodiments, the cell culture medium includes a supernatant and / or cell lysates. In some embodiments, the cell culture medium is derived from a mammalian cell culture medium. In some embodiments, the cell culture medium is derived from HEK cells. In some embodiments, the HEK cells are selected from the group consisting of HEK293S cells, HEK293T cells, HEK293F cells, HEK293FT cells, HEK293FTM cells, HEK293SG cells, HEK293SGGD cells, HEK293H cells, HEK293E cells, HEK293MSR cells, and HEK293A cells, or their derivatives or variants. In some embodiments, the cell culture medium is derived from yeast cells. In some embodiments, the cell culture medium is derived from a virus-producing cell line. Examples of virus-producing cell lines include, but are not limited to, MDCK-S, MDCK-A, Vero cells, LLC-MK2D, PER.C6, EB66, and AGE1.CR cells, HT1080 cells, and HeLa cells, or their derivatives or variants.

[0099] 4. Materials and Methods Materials: Dimethyl sulfoxide (DMSO), thioanisole, anisole, ethane-1,2-dithiol (EDT), polyblen, citric acid, hydrochloric acid (HCl), magnesium chloride hexahydrate (MgCl2·6H2O), phosphate-buffered saline (PBS) at pH 7.4, and the Kaiser test kit were obtained from MilliporeSigma (St. Louis, MO). N,N'-dimethylformamide (DMF), dichloromethane (DCM), virus-producing cells, LV-MAX production medium, LV-MAX transfection kit, LV-MAX Lentiviral Packaging Mix, Opti-MEM Reduced Serum Medium, Vivid Colors™ pLenti6.3 / V5-GW / EmGFP Expression Control Vector, Stbl3™ Chemically Competent E. coli, TrypLE™ expressing enzyme, fetal bovine serum (FBS), 5,5'-dithio-bis-(2-nitrobenzoic acid), PureLink™ HiPure Plasmid Maxiprep Kit, NHS-AlexaFluor 488 (AF488), NHS-AlexaFluor 594 (NHSAF594), Dulbecco's Phosphate Buffered Saline (DPBS), TaqMan™ Fast Virus 1-Step Multiplex Master Mix, custom-made probes and primers from TaqMan®, Purelink Viral RNA / DNA Kit, Turbo DNAse, Proteinase K, CaptureSelect® Lenti VSVG Affinity Matrix, Poros® 50 HE Heparin Affinity Resin, Poros® 50 OH resin, and high-glucose DMEM supplemented with GlutaMAX® and pyruvate were obtained from ThermoFisher Scientific (Waltham, MA).Fmoc / tBu-protected amino acids, hexafluorophosphate azabenzotriazole tetramethyluronium (HATU), piperidine, diisopropylethylamine (DIPEA), N-methyl-2-pyrrolidone (NMP), and trifluoroacetic acid (TFA) were purchased from ChemImpex (Wood Dale, Illinois). Cell culture flasks T-75 and T-25, 96-well culture plates, DNAse / RNAse-free water, and ampicillin were obtained from VWR (Radnor, PA). Shaking flasks and Plasmid bacterial growth medium were also used. + The yeast extract, peptone, and granular agar were obtained from Thomson (Oceanside, CA). The HT1080 cell line was received from American Type Culture Collection (AATC) (Manassas, VA). Tris(hydroxymethyl)aminomethane hydrochloride (Tris-HCl), monosodium phosphate dihydrate, sodium citrate dihydrate, sodium hydroxide (NaOH), and sodium chloride (NaCl) were supplied by Fisher Chemical (Hampton, NH). The resin, Aminomethyl ChemMatrix (particle diameter: 75-150 μm, loading amount: 0.6 mmol per gram of resin), was obtained from PCAS Biomatrix, Inc. (Saint-Jean-sur-Richelieu, Quebec, Canada). The HIV1 p24 ELISA kit was purchased from Abcam (Waltham, MA). Its ectodomain and full-length VSV-G ( E VSV-G and FL VSV-G was provided by Merck (Darmstadt, Germany).

[0100] The plasmids pALD-LentiEGFP-K, pALD-Rev-K, pALD-VSV-GK, and pALD-GagPol-K were purchased from Aldevron (Fargo, ND), and dCAS9-VP64_GFP was a gift from Feng Zhang (Addgene plasmid #61422, http: / / n2t.net / addgene:61422). 20TransIT-VirusGEN® Transfection Reagent for LVV production was purchased from Mirus (Madison, WI). Virus-producing cells derived from HEK293F, LV-MAX production medium, LV-MAX transfection kit, TrypLE® expression enzyme, fetal bovine serum (FBS), PureLink® HiPure Plasmid Maxiprep Kit, dye Syto13, iodine-activated resins SulfoLink and UltraLink iodoacetyl resin, Purelink Viral RNA / DNA Kit, 0.5M Bond-Breaker TCEP Solution, POROS® 50 OH Hydroxyl Activated Resin, and high-glucose DMEM supplemented with GlutaMAX® and pyruvate were obtained from ThermoFisher Scientific (Waltham, MA). Trifluoroacetic acid (TFA), N,N'-succinimidyl carbonate (DSC), 4-dimethylaminopyridine (DMAP), Fmoc / tBu-protected amino acids, piperidine, diisopropylethylamine (DIPEA), N-methyl-2-pyrrolidone (NMP), and hexafluorophosphate azabenzotriazole tetramethyluronium (HATU) were purchased from Chem-Impex (Wood Dale, Illinois). Cell culture flasks T-75 and T-25, 96-well culture plates, DNAse / RNAse-free water, 1,4-piperazinediethanesulfonic acid (PIPES) sesquisodium salt, BalanCD HEK293 medium, isopropanol, and ampicillin were supplied from VWR (Radnor, PA). The HT1080 cell line was purchased from the American Type Culture Collection (AATC) (Manassas, VA).N,N'-dimethylformamide (DMF), dichloromethane (DCM), sodium hydroxide, sodium chloride, sodium bicarbonate, a 0.45 μm polyethersulfone (PES) vacuum filter, iodoacetyl chloride (IAC), acetonitrile, isopropanol, yeast extract, peptone, and granular agar were obtained from Fisher Chemical (Hampton, NH). The HIV1 p24 ELISA Kit was purchased from Abcam (Waltham, MA). The HEK293 HCP ELISA kit was obtained from Cygnus (Southpoint, NC). ToyoPearl® Amino-750F and ToyoPearl® AF-Amino-650M were obtained from Tosoh Bioscience (Tokyo, Japan). The epoxy-activated Eshmuno resin and the Natrix® membranes functionalized with GKEAAFAAC, SRAFVGDADRDC, and FEKISNAEC were supplied by Merck Life Sciences KGaA (Darmstadt, Germany). The pre-packed HiTrap Capto Core 700 column, Peak Expression culture medium, and regenerated cellulose membranes with a nominal pore size of 1 μm and a diameter of 50 mm were purchased from Cytiva (Marlborough, MA). The transfection reagent PEIpro was purchased from Polyplus (Illkirch, France). A 0.2 μm syringe filter Nalgene made of polyethersulfone (PES), a centrifugal filter Amiconultra (100 kDa), ethane-1,2-dithiol, 2-mercaptoethanol, branched polyethyleneimine (molecular weight approximately 25,000 g / mol) (PEI), and benzonase were obtained from MilliporeSigma (Burlington, MA). Lyophilized peptides FEKISNAEC and GKEAAFAAC, as well as iodoacetyl-activated agarose resin, were obtained from GenScript (Piscataway, NJ).

[0101] Production of lentiviral particles (LV). LVVs were produced using the LV-MAX system (ThermoFisher Scientific (Waltham, MA)) according to the manufacturer's protocol. Plasmid pLenti6.3 / V5-GW / EmGFP was introduced into Stbl3 Chemically Competent E. coli cells and selected on LB agar plates supplemented with 100 μg / mL ampicillin. Selected colonies were then cultured in Plasmid + The cells were grown in medium, the plasmid was extracted, and purified using the PureLink® HiPure Plasmid Maxiprep Kit. Suspension HEK293F cells were grown in LV-MAX medium and subcultured until a final cell density of approximately 5.5-10⁴ was reached. 6 Live cells / mL were achieved. 12 hours before transfection, the cells were densified to a density of 3.5-10. 6 Adjust to cells / mL, culture overnight, and final concentration 4.7-10 6 The solution was diluted to cells / mL. In a 125 mL flask, 25.5 mL of cell culture suspension was combined with 1.5 mL of LV-MAX supplement. In a 5 mL vial, 1.5 mL of OptiMEM I was mixed with 45 μg of LV-MAX Lentiviral Packaging Mix (a predetermined mixture of plasmids pLP1, pLP2, and pLP / VSVG) and 30 μg of Vivid Colors® pLenti6.3 / V5-GW / EmGFP Expression Control Vector plasmid. This mixture was slowly added to 1.5 mL of OptiMEM I and 180 μL of transfection reagent and incubated at room temperature for 10 minutes. This mixture was slowly added to virus-producing cells and placed in a 37°C, 8% CO2 incubator with gentle shaking at 125 rpm. After 6 hours, 1.2 mL of LV-MAX enhancer was added to the cell suspension. The LVV particles were collected after 48 hours by centrifugation at 1300g for 15 minutes, and then filtered using a 0.45 μm surfactant-free cellulose acetate (SFCA) filter (ThermoFisher Scientific, Waltham, MA). All LVV samples were immediately stored at -80°C until further use.

[0102] Stability test of buffer solutions. Cell culture fluid (CCF) clarified using a 7kDa Zeba Micro Spin Desalting Column (ThermoFisher Scientific, Waltham, MA), containing LVV particles, was buffered with (i) 20mM citrate buffer or 20mM histidine slow solution containing 75mM NaCl at pH 6.0, 6.5, or 7.0; (ii) 20mM PBS containing 75mM NaCl at pH 6.2, 6.5, or 7.0; (iii) 20mM citrate (pH 6.0) with 0.1M, 0.25M, or 0.5M MgCl2 added; or (iv) DMEM medium. The samples were incubated at room temperature for 30 minutes and then serially diluted in DMEM medium supplemented with 8 μg / mL of polyblen. The LVV titer in the sample was determined by the transduction assay described below, and its infectivity titer was expressed by comparing it with that of an LVV sample in DMEM medium.

[0103] LVV particles, E VSV-G and FL Fluorescent labeling of VSV-G and HEK293 HCP. NHS-ester dyes Alexafluor594 (NHS-AF594, red) and NHS-Alexafluor488 (NHS-AF488, green) were first dissolved in DMSO at a concentration of 10 mg / mL. LVV particles were purified by centrifugation and resuspended in PBS at pH 7.4 according to the procedure described by Jiang et al. (Sci Rep 5, 13875 (2015)). E VSV-G and FL VSV-G and its HEK293F cell culture medium were buffered with PBS at pH 7.4 using a Zeba Spin Desalting Column (7 kDa molecular weight cutoff (MWCO)) (ThermoFisher Scientific, Waltham, MA). 100 μL of LVV particles (approximately 10 11 vp / mL, approx. 10 9Aliquots of TU / mL or VSV-G protein (0.2 mg / mL) were mixed with 3 μL of dye NHS-AF488 and incubated in the dark at room temperature for 1 hour with gentle shaking. The same procedure was used to label HEK293F HCP (approximately 0.3 mg / mL) with NHS-AF594. Unreacted dye was removed using a Zeba Dye removal column (ThermoFisher Scientific, Waltham, MA), and the samples were stored at 4°C.

[0104] Peptide library preparation and screening. Following the split-couple and recombination method using Aminomethyl ChemMatrix resin, a library of 8-mer linear peptides in the format X1X2X3X4X5X6X7X8G was constructed using the protected amino acids Fmoc-Ala-OH, Fmoc-Asn(Trt)-OH, Fmoc-Glu-(OtBu)-OH, Fmoc-Gly-OH, Fmoc-His(Trt)-OH, Fmoc-IleOH, Fmoc-Lys(Boc)-OH, Fmoc-Phe-OH, Fmoc-Ser(tBu)-OH, and Fmoc-Trp(Boc)-OH, via the chemical properties of Fmoc / tBu. Library synthesis was automated using the peptide synthesizer Syro I (Biotage, Uppsala, Sweden). In short, in a 5 mL reaction vial, aliquots of resin were combined with 3 equivalents (eq.) of a 0.5 M protected amino acid in DMF, 3 eq. of HATU in 0.5 M in DMF, and 6 eq. of DIPEA in 0.5 M in DIPEA. Coupling was performed at 45°C for 20 minutes, followed by washing with DMF. After each reaction step, a Kaiser test was performed to confirm the completion of amino acid coupling. Fmoc protection between two consecutive residues was removed using 20% ​​(v / v) piperidine in DMF at room temperature. Once chain extension was complete, the peptide library was deprotected by acid hydrolysis using a TFA:thioanisole:EDT:anisole (90:5:3:2) cocktail for 2 hours at room temperature. Subsequently, the deprotected library was washed and stored in dry DMF.

[0105] Equilibrate 20 μL of peptide library beads aliquots in 20 mM phosphate buffer with 75 mM NaCl at pH 6.5, and add AF594-labeled HEK293T HCP (approx. 0.3 mg / mL) and AF488-labeled LVV (approx. 10 mg / mL). 11 vp / mL, approx. 10 9 TU / mL) or AF488-labeled VSV-G protein ( E VSV-G or FLThe beads were combined with a 200 μL screening mix containing either VSV-G (0.2 mg / mL). After 30 minutes at room temperature, the beads were collected by centrifugation at 5000 g and resuspended in 200 mL of 20 mM phosphate buffer with 75 mM NaCl at pH 6.5. The library beads were screened using a microfluidic device (Center Valley, PA) developed in previous research and introduced into the fluorescence microscope Olympus IX81 (Chu, W. et al. J Chromatogr A 1635, 461632 (2021), Barozzi, A., et al., Int J Mol Sci. 21, 3769 (2020), Day, K. et al. Bioconjug Chem 30, 3057-3068 (2019), Kilgore, R. et al. J Chromatogr A 1687, 463701 (2023), Prodromou, R. et al. Advanced Functional Materials 31, (2021), Prodromou, R. et al. Advanced Functional Materials 31, (2021), and Chu, W. et al. J Chromatogr A 1679,463363(2022). Individual beads were imaged, and green fluorescence (AF488) and red fluorescence (AF594) values ​​were recorded. The beads were then washed for 5 minutes with 20 mM citrate buffer containing 0.5 M MgCl2, imaged again, and new green and red fluorescence values ​​were recorded. All values ​​of fluorescence, emission ratio, and emission decrease were determined in real time by image analysis using a custom MATLAB code (MathWorks, Natick, MA). Beads showing a high ratio of green fluorescence to red-to-green emission before washing, and (ii) a decrease of more than 75% in green fluorescence after washing were isolated, while all other beads were discarded. The selected beads were finally analyzed via Edman degradation after sequencing their peptide ligand candidates using a PPSQ-33A protein sequencer (Shimadzu, Kyoto, Japan).

[0106] Evaluation of in silico-designed VSV-G binding peptides and the interaction between VSV-G and the peptides. The crystal structure of the complex formed by vesicular stomatitis virus glycoprotein G (VSV-G) and low-density lipoprotein receptors (PDB ID: 5OY9 and 5OYL) was analyzed to identify paired residues and estimate their contribution to the binding energy. Nine design sequences were used, namely four disulfide-cyclic sequences (C-cyclo[GSRQFVADSDRD]C-GSG (SEQ ID NO: 87), C-cyclo[GSRSFVGDSDRD]C-GSG (SEQ ID NO: 88), C-cyclo[GSRAFVADADRD]C-GSG (SEQ ID NO: 98), C-cyclo[GSRAFVGDAD]C-GSG (SEQ ID NO: 99)), and five linear sequences (SRQFVCGDSDRD-GSG (SEQ ID NO: 100), SRSFCDSDRD-GSG (SEQ ID NO: 101), SRAFVGDADRD-GSG (SEQ ID NO: 102), AFVGDADRD-GSG (SEQ ID NO: 103), and SFVRIGLSD-GSG (SEQ ID NO: 104) were constructed in Avogadro along with experimentally identified sequences, namely FEKISNAE-GSG (SEQ ID NO: 90), FEKISAAE-GSG (SEQ ID NO: 89), FEKISTAE-GSG (SEQ ID NO: 105), GKEAAFAA-GSG (SEQ ID NO: 91), and SKSAAEHE-GSG (SEQ ID NO: 106), and then modeled in GROMOS using the GROMOS 54A7 force field. In short, each peptide sequence was (i) placed in a simulation box containing 2,000 TIP3P water molecules with periodic boundaries, (ii) equilibrated in a maximum gradient descent step of 10,000 molecules, (iii) heated to 300K in an NVT ensemble at 250 ps using a 1 fs time step, and (iv) equilibrated to 1 atm by a 500 ps NPT simulation using a 2 fs time step.Subsequently, the NPT ensemble was used to perform synthesis by applying a Nose-Hoover thermostat (300K) and a Parinello-Rahman barostat (1atm), respectively. The equations of motion were integrated using a leap-frog algorithm with 2 fs steps, covalent bonds were constrained using the LINCS algorithm, and the Lennard-Jones interaction and short-range electrostatic interaction were calculated using cutoff values ​​of 0.8 nm and 1.2 nm. The Ewald method of the particle mesh was used for long-range electrostatic interactions, and the lists of bonded and unbonded interactions (cutoff 1.2 nm) were updated every 2 fs and 6 fs. The structure of VSV-G was prepared using Protein Prep Wizard (PPW, Schrodinger, New York, NY) by adding lost atoms and explicit hydrogen, removing salt ions and small ligands, and optimizing the hydrogen bond network. Two ionization states of VSV-G (one at pH 6.0 and one at pH 7.4) were obtained, and the structure was minimized using PROPKA. Next, the structure was analyzed using SiteMap to identify the presumed peptide binding site on VSV-G. Specifically, this site had high S scores (greater than 0.8) and D scores (greater than 0.9). Using the docking software HADDOCK (High Ambiguity Driven Protein-Protein Docking) v.2.4, the candidate peptide ligand was docked in silico to VSV-G at pH 6.0 and pH 7.4. VSV-G residues present within the predetermined binding site and peptide residues X1X2[…]X. n The residue was marked as "active," and all surrounding residues were marked as "inactive." Using the dMM-PBSA score, VSV-G:peptide complex assemblies with a Cα RMSD greater than 7.5 Å were ranked, and the top complexes were refined by 200 ns MD simulations to determine the binding free energy (ΔG). B ) was estimated.

[0107] Amination and peptide conjugation of Poros® 50 OH resin. Using a nitrogen stream, 10 mL of Poros® 50 OH resin was first dried, washed with DMF, and resuspended in 50 mL of CDI solution at 100 mg / mL in DMF. The sample was kept at room temperature with stirring. After 5 hours, the resin was thoroughly washed with DMF and dried under a nitrogen stream. Subsequently, the resin was mixed with 100 mL of 5% (v / v) ethylenediamine in DMF and incubated at 45°C at 100 rpm with shaking. After 12 hours, the resin was washed with DMF, then with DCM, dried under nitrogen, and stored at 4°C. The density of primary amine groups on the modified Poros 50 resin beads was determined by Kaiser assay. In short, 10 mg of resin was mixed with 1 mL of DMF, 0.1 mL of KCN in H2O / pyridine, and 0.1 mL of ninhydrin. The mixture was then placed in boiling water for 5 minutes, cooled to room temperature, and the supernatant was diluted 100-fold with DMF. The UV absorbance of the solution was measured at 425 nm using a UV-1800 spectrophotometer (Shimadzu, Kyoto, Japan), and a calibration curve was constructed using ethanolamine. On the aminated Poros resin, the predetermined peptide sequence was synthesized using an automated peptide synthesizer Alstra (Biotage, Uppsala, Sweden) according to the following procedure.

[0108] LVV was purified from HEK293 cell culture supernatant using peptide-Poros resin. The peptide-Poros resin prepared as described above, along with its controls CaptureSelect® Lenti VSVG and Poros® 50 HE Heparin Affinity Resin, were flow-packed into 1 mL Tricorn 5 / 50 columns (Cytiva, Marlborough, MA) and introduced into an AKTA Avant FPLC system (Cytiva, Marlborough, MA). The resin packing quality was evaluated by measuring the peak symmetry of the conductivity signal generated by pulse injection of 1 M NaCl aqueous solution (target value: 1~1.2). The resin was equilibrated with 10 CV of equilibration buffer (Table 1). Clarified HEK293 CCF (LVV titer approximately 0.5~2·10)8 LVV (with a TU / mL and HCP titer of approximately 0.3–0.1 mg / mL, including variations in LVV activity across different manufacturing batches) was loaded onto the resin with a residence time (RT) of either 1 minute or 3.5 minutes. After loading, the resin was washed with 20 CV of washing buffer, and the bound LVV was eluted with 9 CV of elution buffer (Table 1). After elution, the resin was regenerated using 10 CV of 0.1 M glycine containing 2 M NaCl at pH 2.0.

[0109] [Table 1]

[0110] Measurement of dynamic binding capacity. GKEAAFAA (SEQ ID NO: 3)-Poros resin, FEKISNAE (SEQ ID NO: 1)-Poros resin, FEKISAAE (SEQ ID NO: 2)-Poros resin, FEKISTAE (SEQ ID NO: 11)-Poros resin, SRAFVGDADRD (SEQ ID NO: 4)-Poros resin, and SFVRIGLSD (SEQ ID NO: 5)-Poros resin, along with their controls, CaptureSelect® Lenti VSVG and Poros® 50 HE Heparin Affinity Resin, were flow-packed into 1 mL Tricorn 5 / 50 columns (Cytiva, Marlborough, MA) and introduced into an AKTA Avant FPLC system (Cytiva, Marlborough, MA). After equilibration with 10CV of 50mM PIPES buffer in 100mM NaCl buffer at pH 7.4, the resin is clarified with HEK293 CCF (LVV titer approximately 0.5-2.10) until the LVV titer in the eluate reaches 70-80% of its corresponding feedstock titer. 8A TU / mL (HCP titer approximately 0.3-0.1 mg / mL) was continuously loaded with either a 1-minute or 2-minute RT. The eluate was partitioned into 3 mL fractions and analyzed as described below to determine the titer of the encapsulated transgene and the LVV units transduction into cells that were present in the eluate. The dynamic binding capacity at 10% breakthrough (DBC10%) was calculated as described in previous studies, and the void volume of the system was measured by pulsed acetone injection, which was used to adjust the DBC10% value.

[0111] Stability of peptide-Poros resins. GKEAAFAA (SEQ ID NO: 3)-Poros resin, FEKISNAE (SEQ ID NO: 1)-Poros resin, FEKISAAE (SEQ ID NO: 2)-Poros resin, FEKISTAE (SEQ ID NO: 11)-Poros resin, SRAFVGDADRD (SEQ ID NO: 4)-Poros resin, and SFVRIGLSD (SEQ ID NO: 5)-Poros resin, as well as their controls, CaptureSelect® Lenti VSVG and Poros® 50 HE Heparin Affinity Resin, were flow-packed into 1 mL Tricorn 5 / 50 columns (Cytiva, Marlborough, MA) and introduced into an AKTA Avant FPLC system (Cytiva, Marlborough, MA). After equilibrating 100 mM NaCl buffer with 10 CV of 50 mM PIPES buffer at pH 7.4, the resin is treated with clarified HEK293 CCF (LVV titer approximately 0.5-2·10). 8A solution (TU / mL, HCP titer approximately 0.3 mg / mL) was loaded at 30 CV with a 1-minute RT. The resin was washed with 20 CV of binding buffer, and the bound LVV was eluted at 1-minute RT with 4 CV of 50 mM PIPES buffer containing 650 mM NaCl buffer at pH 7.4. After elution, the resin was regenerated at pH 2.0 with 10 CV of 0.1 M glycine containing 2 M NaCl, followed by a cleaning-in-place (CIP) with 15 CV of 0.5 M NaOH at 1-minute RT, and then a 30-minute static incubation. Both the regeneration and CIP processes were performed at 1-minute RT. An additional cycle was repeated in which LVV was purified from the clarified HEK293 CCF by intermediate CIP. The chromatographic fraction was analyzed as described below to measure the yield and purity of LVV.

[0112] Analytical characteristics of chromatographic fractions. The titers of p24 protein and HEK293 HCP in chromatographic samples were measured by ELISA using kits from Abcam (Cambridge, MA) and Cygnus (Southport, NC), respectively, according to the manufacturer's instructions.

[0113] RT-qPCR chromatography samples were initially treated with TurboDNAse, followed by RNA isolation using the Purelink Viral RNA / DNA Kit (ThermoFisher Scientific, Waltham, MA). These samples were then combined with TaqMan Fast Virus, custom TaqMan probes, and the primers listed below, and analyzed using the CFX Duet Real-Time qPCR System (Bio Rad, Hercules, Ca). Plasmid pLenti6.3 / V5-GW / EmGFP was used as the standard. [Table 2]

[0114] The total amount of double-stranded DNA (dsDNA) was measured using the DNA Quantification Quant-iT (trademark) PicoGreen (trademark) dsDNA Assay Kit (ThermoFisher Scientific, Waltham, MA) according to the manufacturer's protocol.

[0115] HT1080 cells were seeded at a density of 7,000 cells / well in high-glucose DMEM medium supplemented with fluorescence flow cytometry (FFC) GlutaMAX®, pyruvate, and 10% (v / v) FBS. The plates were centrifuged at 900 g for 5 minutes and placed in an incubator at 37°C and 5% CO2. The chromatographic fraction containing LVV particles was serially diluted (10×) in DMEM medium (without FBS or antibiotics) supplemented with 8 μg / mL polyblen. After 4 hours, the consumed cell culture medium in the 96-well plate was replaced with 0.1 mL of the diluted sample and incubated for 12 hours. Subsequently, the sample was replaced with fresh DMEM medium supplemented with 10% (v / v) FBS. After 72 hours, the cells were isolated from the plate by incubation for 15 minutes at 37°C with 150 μL of a mixture consisting of TrypLE® Express Enzyme:DPBS (75:25 (v:v)). GFP expression was performed using a flow cytometer CytoFlex (Beckman Coulter, Brea, CA). + The cell fraction was quantified, and the number of transduction units (TU / mL) per 1 mL was calculated using Equation 1 (below).

number

[0116] Resin functionalization with peptide ligands. Following the method described in a previous study, the surface functional groups of Poros® resin and Eshmuno® resin were converted to primary amino groups. (Automated peptide synthesizer Initiator) + Using Alstra® (Biotage, Uppsala, Sweden), the sequences GKEAAFAA (SEQ ID NO: 3), GKEAAFAA-G (SEQ ID NO: 107), GKEAAFAA-GSG (SEQ ID NO: 91), GKEAAFAA-GSGSGSG (SEQ ID NO: 108), GKEAAFAA-GSGPGSG (SEQ ID NO: 109), GKEAAFAA-PEG3 (SEQ ID NO: 3), and FEKISNAE (SEQ ID NO: 1) were synthesized on a resin. Each amino acid was coupled by incubating a solution of 5 equivalents (eq.) of the protected amino acid and HATU in dry DMF, and a solution of 6 eq. of DIPEA in NMP (both at a concentration of 0.5 M), with the chromatography resin at 70°C for 20 minutes (achieved by microwave heating). After each coupling, Fmoc deprotection was achieved by using 20% ​​(v / v) piperidine in DMF for 30 minutes at room temperature. The peptide chain was deprotected for 2 hours at room temperature using a cleavage cocktail containing TFA, thioanisole, anisole, and EDT(90 / 5 / 3 / 2)(v / v). After deprotection, the resin was washed with DMF and DCM, dried in an N2 stream, and stored at 4°C. The bromide-activated WorkBead resin was initially reacted with 25%(v / v) NH4OH (1:1 ratio) at room temperature while gently mixing. 21Before the next step, after 16 hours, the amination resin was sequentially washed with 10 cubic centimeters of each solvent (water, ethanol, and acetonitrile). A mixture consisting of 0.717 mL of IAC, 1.112 mL of TEA, and 20 mL of ACN was added to 20 mL of the amination resin and reacted slowly with gentle mixing at room temperature, while protected from light. After 3 hours, the resin was washed with 10 cubic centimeters of ACN, acetone, and DMF. Unreacted primary amines in the resin were acetylated by reacting 10 mL of the resin with 6.6 mL of acetic anhydride, 10.2 mL of DIPEA, and 20 mL of NMP at room temperature for 3 hours. The completion of the reaction was converted by a negative result of the Kaiser assay. As described in previous studies, conjugation of cysteine-derivativeized peptides was performed on the iodoacetyl-functionalized resin. 22In short, the resin was rinsed with water, 50 mM Tris, 5 mM EDTA-Na, and 25 mM TCEP (pH 8.5) (coupling buffer). Peptides FEKISNAEC (SEQ ID NO: 110) and GKEAAFAAC (SEQ ID NO: 11) were solubilized in coupling buffer at a concentration of 10 mg / mL, and added to the precipitated resin at a ratio of 2 mL of peptide solution per 1 mL of resin. The conjugation reaction was allowed to proceed at room temperature for 2 hours by inversion mixing, and then quenched with 25 mM 2-mercaptoethanol in coupling buffer. The resin activation, peptide conjugation, and quenching steps were performed in the dark. The resin was thoroughly rinsed with 1 M sodium chloride and water, and finally stored in 20% (v / v) ethanol at 4°C. Conjugation of GKEAAFAAC (SEQ ID NO: 11) with iodoacetyl-functionalized Poros® resin was repeated by incubation with 21 mg, 52 mg, 104 mg, and 312 mg of peptide per 1 mL of resin. The residual amount of peptide in the coupling solution was measured by quantitative Ellman assay. Briefly, 100 μL of coupling solution was combined with 100 μL of 5,5'-dithio-bis-(2-nitrobenzoic acid stock solution) (Ellman reagent) at a concentration of 4 mg / mL in 0.1 M sodium phosphate buffer (pH 8.0) containing 1 mM EDTA. The sample was incubated at room temperature for 15 minutes and analyzed by UV spectrophotometry at a wavelength of 412 nm. Solutions of GKEAAFAAC (SEQ ID NO: 11) at concentrations of 0.2 mg / mL, 0.1 mg / mL, 0.05 mg / mL, and 0.025 mg / mL in the coupling buffer were used as standard substances. Finally, the ligand density was determined by mass balance, by dividing the difference between the initial and residual amounts of the peptide in the coupling solution by the volume of iodoacetyl-functionalized Poros® resin.

[0117] Amine activation of cellulose membrane. The regenerated cellulose membrane was first rinsed with DMF and air-dried. A solution consisting of 35 mL of DMF, 1.80 g of DMAP, and 2.24 g of DSC was added to the membrane at 100 cm². 2The total surface area of ​​the membrane was immersed at room temperature for 3 hours and mixed in a rocker. After the reaction, the membrane was rinsed with DMF, then with DMSO and IPA. The membrane was maintained in IPA and at 4°C until the next step. The membrane was air-dried and aminated by immersion at room temperature in a solution consisting of 10 g of PEI in 90 mL of MilliQ water and shaking in a rocker. After 2 hours, the membrane was vigorously rinsed with water, rinsed with ACN, and functionalized with IAC according to the same procedure described herein.

[0118] Production and recovery of LVV. Virus-producing cells (ThermoFisher Scientific, Waltham, MA) were cultured in BalanCD LV-MAX or Peak Expression medium at 8% CO2 and 37°C, and before transfection, at least 4 passages were used to obtain 3.5–5.5·10 LVV. 6 The cell density reached 1.5-10 cells / mL. When using the PEIpro transfection reagent, the cells were transfected to 1.5-10 6 Dilute to cells / mL 24 hours before transfection, and apply 2.5-10 immediately before transfection. 6 The cells were adjusted to the desired level per mL. The PEIpro and plasmid were then mixed in DMEM medium in a mass ratio of 1:3 (DNA:PEI) at 1 × 10⁶. 6 The cells were lysed with 1 μg of total DNA per cell (10% of the total cell culture volume), mixed, incubated at room temperature for 15 minutes, and then added to the cell suspension. When using the LV-Max system, LVV was produced according to the manufacturer's protocol. When using the Mirus transfection reagent, the plasmid was first diluted to a volume of complexing solution equal to 10% of the cell culture volume and to a plasmid amount of 1.6 μg / mL of cell culture suspension. After plasmid dilution, 3 μL of transfection reagent per 1 μg of plasmid was added to the same vial, mixed by inverting the tube 5 times, incubated at room temperature for 15 minutes, and then added to the HEK293F cell suspension. 6Cells were added at a concentration of cells / mL. For all three systems, 48 ​​hours after transfection, the cells were removed by centrifugation at 1300g for 15 minutes, and the supernatant was treated with 50 U / mL benzonase and 2 mM MgCl2 for 30 minutes at 37°C. Finally, clarification was performed by filtration using a 0.45 μm polyethersulfone (PES) vacuum filter. If not to be used immediately, all samples were stored at -80°C.

[0119] LVV purification using peptide-functionalized resin. The resin was flow-packed to a controllable Tricorn 5 / 50 column with a final volume of 1 mL and equilibrated with 10 column volumes (CV) of binding buffer (25 mM PIPES, 100 mM NaCl, pH 7.4). 10–35 mL of clarification material was loaded by downflow at a linear velocity of 305 cm / hour (equivalent to a residence time (RT) of 1 minute). After washing the resin with 20 CV of binding buffer, LVV elution was performed by upflow with 3 CV of 25 mM PIPES and 650 mM NaCl (pH 7.4), and 3 CV of 1 M NaCl (pH 7.4). A clean-in-place (CIP) wash was performed with 15 CV of 0.5 M NaOH (aqueous solution), followed by a 15-minute incubation. The resin was then washed one last time with 10 CV of equilibration buffer to restore neutral pH. All chromatography steps were performed at a flow rate of 1 mL / min (RT: 1 min) while continuously monitoring the conductivity, pH, and UV absorbance of the column eluate at 254 nm, 260 nm, and 280 nm. The affinity resin CaptureSelect® Lenti VSVG was operated according to the manufacturer's instructions.

[0120] Confocal imaging of GKEAAFAA-Poros™ after loading fluorescently tagged LVV particles. Lentiviruses were purified by sucrose gradient ultracentrifugation according to the method described by Jiang et al. The resulting LVV pellet was approximately 1.10 11 Suspended in 0.5 mL of 25 mM PIPES and 100 mM NaCl (pH 7.4) overnight at 4°C until a titer of vp / mL was achieved. Approximately 2 μL of 5 mM Syto13 solution in DMSO was added. 11Fluorescent labeling of LVV particles was performed by incubation at room temperature in the dark for 30 minutes. Excess dye was removed using Pierce Dye Removal Columns (ThermoFisher, Ma). The fluorescently labeled LVV was loaded onto GKEAAFAA (SEQ ID NO: 3)-Poros® resin according to the procedure described herein. After loading, beads extracted from the leading and trailing ends of the column were imaged using a Leica Stellaris Confocal Microscope (Wetzlar, Germany).

[0121] LVV purification using peptide-functionalized membranes. Natrix® membranes functionalized with peptides GKEAAFAAC (SEQ ID NO: 111), FEKISNAEC (SEQ ID NO: 112), and SRAFVGDADRDC (SEQ ID NO: 112) were punched out with a 22mm diameter disc, and two layers were placed in a 25mm Whatman filter holder (Cytiva, Marlborough, MA). The membranes were equilibrated with 25mM PIPES and 100mM NaCl (pH 7.4) at 50 membrane volumes (MV) at 10MV / min, and then 2-5mL of clarification material was loaded downflow at 3MV / mL. After washing the membranes with 50MV binding buffer, LVV elution was performed upflow with 25mM PIPES and 650mM NaCl (pH 7.4) at 50MV at 10MV / mL. Finally, the membrane was regenerated with 0.1 M glycine and 2 M NaCl (pH 2.0) at 50 MV, and CIP was performed with 0.5 M NaOH (aqueous solution) at 50 MV. Lentivirus purification was performed on a Mustang Q device (MV: 0.86 mL) according to the published instructions. Briefly, the membrane was first equilibrated with 10 mM histidine and 150 mM NaCl (pH 7.0) at 50 MV, loaded with 150 MV of clarification material, washed with 60 MV of binding buffer, and LVV elution was performed in three steps using 10 mM histidine (pH 7.0) and NaCl concentrations of 0.4 M, 1.0 M, and 1.5 M (each containing 20 MV at 10 MV / min). All chromatographic steps were performed while continuously monitoring the conductivity, pH, and UV absorbance of the column eluate at 254 nm, 260 nm, and 280 nm.

[0122] LVV polishing, buffer exchange, and sterile filtration. A 1 mL column packed with CaptoCore700 resin was equilibrated with 25 mM PIPES and 100 mM NaCl (pH 7.4) at 10 CV with a RT of 1 minute, and then filtered downflow with GKEAAFAA (SEQ ID NO: 3)-POROS (trademark) (LVV titer: 6.2.10). 9A 20 CV elution stream obtained from a column (vp / mL, HCP titer: 2.6 μg / mL) was loaded at RT 2 min. The resin was then cleaned upflow with 15 CV of 1 M NaOH in 30% (v / v) isopropanol:water, followed by static contact for 30 minutes. All chromatography steps were performed while continuously monitoring the conductivity, pH, and UV absorbance of the column eluate at 254 nm, 260 nm, and 280 nm. After polishing, the LVV was set to 1.2 using 25 mM PIPES, 10% sucrose, and 20 mM MgCl2 (pH 7.4) 10 diavolute. . 10 10 The samples were concentrated by centrifugation to a titer of vp / mL using an Amicon filter (MWCO: 100kDa) at 3500g for 30 minutes. Finally, the samples were filtered using a Nalgene PES 0.2μm syringe filter and either analyzed immediately or stored at -80°C.

[0123] Analysis of chromatographic samples. HT1080 cells were cultured in DMEM supplemented with 10% FBS at 5% CO2 and 37°C until 80–90% confluence for the transduction assay. Cells were released from the culture flask using trypsin, counted using a hemocytometer and trypan blue for cell viability, and seeded at 7,000 cells / mL in a 96-well plate. The plate was centrifuged at 900g for 5 minutes and incubated for 4 hours. At the start of the transduction assay, the culture medium in the plate was replaced with an equal volume of sample prepared by serially diluting (10x) the LVV-containing fraction in DMEM medium supplemented with 8 μg / mL polyblen. After 12 hours, the consumed medium was replaced with fresh DMEM medium supplemented with 10% (v / v) FBS, and the cells were incubated for 60 hours. Using a flow cytometer CytoFLEX (Beckman, Brea, CA), the fraction of cells expressing GFP was measured, and the value of transduction units (TU) per 1 mL was calculated using equation 2 (below). %GFP + We considered the LVV transduction concentration only for dilutions containing 1% to 25% cells.

number

[0124] HEK293 HCP ELISA and p24 ELISA. HIV ELISA (Abcam, Waltham, MA) and HEK293 HCP ELISA (Cygnus, Southpoint, NC) kits were used according to the manufacturer's instructions to measure the titers of p24 protein and HEK293 HCP, respectively. From the HCP titer values, the reduction in HCP (RV) and logarithmic reduction (LRV) were derived using the following equations.

number

[0125] Real-time quantitative PCR (RT-qPCR) was performed as described in previous studies. Briefly, samples treated with DNAse were purified using the Purelink Viral RNA / DNA Kit (ThermoFisher Scientific, Waltham, MA), and the encapsulated RNA was isolated. Subsequently, these samples were combined with TaqMan Fast Virus, a custom TaqMan probe, and the primers listed herein, and analyzed using the CFX Duet Real-Time qPCR System (Bio Rad, Hercules, CA). Plasmid pALD-LentiEGFP-K was used as the standard. [Examples]

[0126] The attached examples are provided as illustrations of a partial scope and specific embodiments of the present disclosure and are not intended to limit the scope of the present disclosure.

[0127] Example 1 Selection of chromatographic buffers for library screening Lentiviral particles are highly sensitive to the physicochemical properties of aqueous environments, and their transduction activity can be irreversibly damaged by small fluctuations in pH, salt concentration, temperature, and osmotic pressure. This limits the acceptable range of chromatographic buffers suitable for LVV purification in the context of bioprocesses. Particularly restrictive is the limitation on elution pH, which is limited to a range of 6–8 and cannot be utilized as in the affinity purification of antibodies and AAVs. Therefore, we explored various formulations of binding and elution buffers compatible with LVV, initially focusing on citrate-based, phosphate-based, and histidine-based solutions with varying ionic strengths and pH levels. 10 8 Preliminary stability tests (Figure 6) performed by incubating TU / mL LVV particles in various buffers for 30 minutes showed that (i) no significant loss of infectivity was observed in citrate and phosphate buffers with 75 mM NaCl (pH 6.0 and 7.0), and (ii) a significant decrease in infectivity (over 40%) was observed in 20 mM histidine buffer with 75 mM NaCl (pH 6.0). To prepare the elution buffers, instead of lowering the pH, sodium chloride and magnesium chloride were used to increase their ionic strength. Magnesium chloride in sodium citrate buffer did not affect LVV activity at concentrations in the range of 100–500 mM at pH 6.0 (Figure 6). Based on these results, 20 mM phosphate buffer with 75 mM NaCl (pH 6.5) and 20 mM citrate buffer with 500 mM MgCl2 (pH 6.0) were selected as equilibration and elution buffers, respectively, for library screening.

[0128] Identification of peptide ligand candidates that target LV LVV vectors used in ex vivo applications (including all FDA-approved drugs ABECMA, CARVYKTI, BREYANZI, SKYSONA, and ZYNTEGLO) are designed using pseudotyping. This approach uses human CD4 + This involves replacing the wild-type envelope glycoprotein gp120, which underlies the tropism of HIV virus against T cells, with a heterologous glycoprotein. Most LVV pseudotyping to date has utilized vesicular stomatitis virus G glycoprotein (VSV-G), which, by targeting ubiquitous cell membrane phospholipids, confers high stability and a high ability to transduce to a wide variety of cell types to its vector particles. Other proteins used in LVV pseudotyping, namely feline endogenous virus (RD114) envelope glycoprotein, measles virus hemagglutinin and fusion glycoprotein, gibbon leukemia virus envelope protein, rabies virus glycoprotein, and Molony's mouse leukemia virus 4070A-envelope protein (amphotropic), have not yielded comparable efficacy. Therefore, VSV-G-pseudotyped LVV is expected to be used in the design of ex vivo cell therapies in the foreseeable future.

[0129] Under this assumption, two model targets were selected for ligand selection: one VSV-G protein and a mature VSV-G pseudotype LVV particle. The VSV-G protein has three domains, namely the ectodomain ( E It includes a VSV-G (presented on the viral surface), a transmembrane domain (which fixes the protein to the lipid layer of the viral envelope), and a cytoplasmic domain. In principle, E Only VSV-G can be targeted by ligands immobilized on its surface, but the transmembrane domain or full-length VSV-G in the lipid layer of the tertiary structure of its ectodomain ( FLRegarding any role of the intercalation of VSV-G), information is unavailable. Therefore, to avoid the ligand selection being biased towards a model target that does not represent its product, E both VSV-G and FL VSV-G were employed in library screening.

[0130] The selection of peptide ligand candidates was first carried out by screening a solid-phase peptide library using a device for ligand development. The technology relies on orthogonal fluorescent labeling to ensure the selection of ligands that have strong and selective binding and can also release their targets when exposed to mild elution conditions. For this purpose, a device for imaging and sorting microfluidic beads was designed, introduced into a fluorescence microscope, and routinely used to implement a protocol for peptide ligand discovery (Figure 1). That is, (i) a solid-phase peptide library prepared on hydrophilic and translucent porous particles as a one-bead-one-peptide library was incubated with its target labeled with a green fluorescent dye, numerous impurities contained therein, and the whole proteome of HEK293 cells, and collectively labeled with a red fluorescent dye (Note: To mimic industrial LVV recoveries, the screening mix was adjusted to an HCP titer of approximately 0.3 mg / mL, and an LVV titer of approximately 10 9 TU / mL or a VSV-G titer of 0.2 mg / mL ( E VSV-G or FL(ii) The beads were prepared using either VSV-G) (here, Alexafluor488 was used as the green dye and Alexafluor594 as the red dye because they have a low tendency to alter the structure and behavior of the labeled protein). (ii) After incubation, the library beads were thoroughly washed and individually transferred to the microfluidic device, where they were imaged and the image indicators were analyzed in real time. (iii) Beads showing high green-only fluorescence (meaning selective and strong target affinity) were retained in the imaging chamber while all other beads were drained. (iv) The beads were then exposed to a flow of 20 mM citrate buffer with a predetermined elution buffer, 0.5 M MgCl2 (pH 6.0), and imaged again. (v) Beads showing strong loss of green fluorescence (meaning the ability to release their product under elution conditions defined by the operator) were selected while all other beads were discarded. Steps (i) through (v) are automated by custom Matlab code, which enables library screening at a rate of 350 beads per hour. Finally, (vi) the sequences held by the given beads are identified by Edman decomposition. E VSV-G, FL Table 2 reports a list of peptides identified for VSV-G and complete LVV particles, while their sequence homology plots are reported in Figure 1I. The identified sequences are ostensibly amphiphilic, each containing at least one aromatic amino acid (Phe or Trp) and multiple aliphatic amino acids (Ala, Ile). The abundance of Glu is rather surprising, appearing at least once in 72% of the identified sequences and twice in 30%. Due to the negative charge on the LVV surface, the presence of Glu is unexpected, as it is rooted in the phospholipid bilayer forming the envelope, making anion exchange chromatography the primary separation method. Simultaneously, the hydrophilic and anionic characteristics of the candidate ligands reduce the risk of binding HCPs, thus requiring peptides rich in cationic and hydrophobic amino acids.

[0131] Example 2 Evaluation of Candidate Peptide Ligands under Dynamic Conditions Most commercially available chromatography resins feature pore diameters in the range of 20 - 100 nm. These adsorbents are suitable for the purification of protein-based biopharmaceuticals but not ideal for large virus vectors such as LVV and baculovirus, whose diameter can reach 100 nm. Therefore, in the bioseparation community, chromatography substrates with large pore sizes, such as membranes and monoliths, are expected to become mainstream in the downstream processing of virus vectors. In this study, established chromatography adsorbents were used to evaluate a given peptide, avoiding uncertainties related to peptide surface density and presentation. Notably, Poros™ 50 OH resin, which features pores with a maximum diameter of 1000 nm, is well-suited for the purification of LVV particles.

[0132] By converting its hydroxyl groups to primary amino groups, Poros™ 50 OH beads were modified to reach a functional density of 172 μmol per gram of resin (Figure 7), and these beads were utilized for the conjugation of the sequences identified by library screening (Table 2). With a recommended residence time (RT) of 3.5 minutes on the Poros™ resin and using a given binding buffer (20 mM phosphate buffer with 75 mM NaCl (pH 6.5)) and elution buffer (20 mM citrate buffer with 0.5 M MgCl2 (pH 6.0)), the obtained peptide-Poros resin was evaluated for the binding and elution of LVV under dynamic conditions by loading clarified HEK293 CCF (LVV titer of approximately 10 10 vp / mL (equivalent to approximately 10 8 TU / mL) and an HCP titer of approximately 0.3 mg / mL, note: some variation in the titer of total LVV units and LVV units that transduce cells was observed across different production batches).

[0133] The yield and purity values ​​of LVV listed in Table 3 indicate SIEINSSE (SEQ ID NO: 10), GEFENINW (SEQ ID NO: 7), EWKAAFIW (SEQ ID NO: 8), SKSAAEHE (SEQ ID NO: 6), GKEAAFAA (SEQ ID NO: 3), SNEIEIAN (SEQ ID NO: 9), and FEKISNAE (SEQ ID NO: 1) as promising ligand candidates. Specifically, these sequences reduce HEK293 HCP by 10–70 times, reduce cellular DNA by up to 70%, and reduce the yield of encapsulated transgenes to a range of 30–50%. Considering that LVV particles are susceptible to changes in buffer composition, conductivity, and pH (which often lead to significant loss of infectivity), the transduction activity of purified LVV was evaluated on HT1080 human fibrosarcoma cells as an additional indicator to guide the selection of candidate ligands. The LVV recovery rates obtained from specific sequences (shown in Table 4 along with other purification parameters) indicate that FEKISNAE (SEQ ID NO: 1) and GKEAAFAA (SEQ ID NO: 3) perform comparably to the affinity adsorbent Poros® 50 HE Heparin and CaptureSelect® Lenti VSVG affinity resin, resulting in a yield of 38%–41% of LVV units transduction into cells. However, these values ​​are considered quite low when assembled in the context of achieving affordable LVV production.

[0134] In exploring ways to improve the LVV yield, we first investigated the effect of residence time during the loading process. When loading at a flow rate of approximately 0.3 mL / min (RT 3.5 min), the amount of clarified HEK293 CCF loaded into the column was approximately 3.10 per 1 mL of resin. 11 30 mL calculated based on the load reference value of VP, and approximately 10 of its raw materials 10Based on the LVV titer results in vp / mL, the total loading time was 1.75 hours. Combining the recovery and clarification steps, chromatographic washing and elution steps, and incubation period with purified LVV and HT1080 cells, the total process time was approximately 3 hours. Comparing this time to the half-life of VSV-G pseudotype LVV at room temperature (estimated at 35 hours), it was suggested that the recovery rate of LVV transduction into cells could be negatively affected by the long processing time. To mitigate this inconvenience, additional testing of specific peptides was performed by reducing the residence time of all chromatographic steps from 3.5 minutes to 1 minute, thereby reducing the processing time from 3 hours to approximately 50 minutes. Based on the recovery rates of LVV particles and transgenes, sequences GEFENINW (SEQ ID NO: 7) and SNEIEIAN (SEQ ID NO: 9) were selected. Based on the recovery rates of LVV units transduction into cells, FEKISNAE (SEQ ID NO: 1), EWKAAFIW (SEQ ID NO: 8), SKSAAEHE (SEQ ID NO: 6), and GKEAAFAA (SEQ ID NO: 3) were selected. The results shown in Table 4 demonstrate that shortening the load residence time is beneficial to the performance of all resins. In particular, the yields of FEKISNAE (SEQ ID NO: 1), GEFENINW (SEQ ID NO: 7), and GKEAAFAA (SEQ ID NO: 3) increased by 1.7 to 3 times, while their HCP removal increased from 1.4 to 1.6 to 1.8 to 2.7 LRV. Notably, the yield and purity of production obtained with Poros™ 50 HE Heparin also doubled, demonstrating that the need for shorter load times is not tied to the specific chemical composition of the ligand. It should also be noted that the performance of the peptide-based adsorbent is equivalent to that of the control affinity resin in terms of recovered lentiviral particles, transgenes, and purity. At the same time, further process optimization is needed to improve the economics of LVV production, given the yield value well below 50%.

[0135] [Table 3-1] [Table 3-2]

[0136] [Table 4-1] [Table 4-2]

[0137] [Table 5]

[0138] Example 3 By adjusting the composition of the chromatography buffer, LVV purification can be optimized. The increased yield and purity of LVV, achieved solely by shortening the residence time in the loading process, suggested that further adjustments to the chromatography process could lead to even greater increases in yield and purity. Therefore, the performance of FEKISNAE (SEQ ID NO: 1)-Poros resin, GEFENINW (SEQ ID NO: 7)-Poros resin, and GKEAAFAA (SEQ ID NO: 3)-Poros resin was improved by optimizing the composition, concentration, and pH of the chromatography buffer. We first explored adding arginine to the washing buffer and MgCl2 to the elution buffer, and then evaluated buffers with varying basic compositions and conductivity while maintaining a constant RT1 minute in the loading process.

[0139] As shown in Table 5, adding 50 mM arginine to the washing buffer (20 mM phosphate buffer with 75 mM NaCl (pH 6.5)) slightly increased the HCP LRV obtained from FEKISNAE (SEQ ID NO: 1)-Poros resin from 1.82 to 2.01 (a 102-fold decrease and corresponding to a residual HCP titer of less than 3 μg / mL), suggesting a potential strategy for improving HCP removal. However, when the MgCl2 concentration in the base elution buffer (20 mM citrate buffer (pH 6.0)) increased to 0.5 M to 1 M, the LRV recovery rate decreased, so that buffer was discarded.

[0140] To explore additional buffer systems with different basic compositions, Tris was first used to prepare new equilibration, washing, and elution buffers. Notably, the new washing buffer (50 mM Tris buffer with 130 mM NaCl (pH 8.0)) increased the HCP LRV with FEKISNAE (SEQ ID NO: 1)-Poros resin to 2.39 (a 246-fold decrease and equivalent to a residual HCP titer of 1.2 μg / mL), and the HCP LRV with GKEAAFAA (SEQ ID NO: 3)-Poros resin increased to 2.05 (a 112-fold decrease, equivalent to 2.6 μg / mL). In addition, the new elution buffer (50 mM Tris and 1 M NaCl (pH 8.0)) increased the yield of transductionable LVV units to cells, with LVV units obtained by FEKISNAE (SEQ ID NO: 1)-Poros increasing to 35% and LVV units obtained by GKEAAFAA (SEQ ID NO: 3)-Poros increasing to 38%. Under the same conditions, the GEFENINW (SEQ ID NO: 7)-Poros resin yielded excellent purification, but the yield was insufficient. This inferior performance, coupled with the presence of asparagine (N) and tryptophan (W) residues in this peptide (which are likely to be easily degraded by deamidation and oxidation to aspartic acid), led to the abandonment of this candidate ligand.

[0141] [Table 6]

[0142] A new set of wash buffer, elution buffer, and 50 mM HEPES / PIPES buffer (pH 7.4) with 100 mM NaCl and 0.65 M NaCl added, respectively, was then evaluated, starting with GKEAAFAA (SEQ ID NO: 3) as the highest-performing ligand. As summarized in Figure 2, both the yield of transduction into cells for LVV units and HCP removal was significantly improved, reaching 51% and LRV 2.26 (i.e., a 182-fold reduction, with a residual potency of approximately 1.6 μg / mL) respectively when using the PIPES-based buffer. Therefore, four or more candidate sequences, including FEKISNAE (SEQ ID NO: 1) and its candidate ligands SKSAAEHE (SEQ ID NO: 6), EWKAAFIW (SEQ ID NO: 8), and EHFEHWSE (SEQ ID NO: 12), selected from Table 3 based on their sequence similarity to GKEAAFAA (SEQ ID NO: 3) and FEKISNAE (SEQ ID NO: 1), were re-evaluated using the PIPES-based buffer. The resulting chromatograms are shown in Figure 8, and the corresponding values ​​for LVV yield and purity are shown in Table 6. In summary, these results strongly suggest the use of piperazine ethanesulfonate-based buffers, and more broadly, demonstrate the impact of chromatographic treatment on the transduction activity of purified LVV particles, while no decrease was observed in incubation of LVV particles in either HEPES buffer or PIPES buffer, and significant loss of infectivity was achieved by chromatographic treatment (i.e., when tested in HT1080 cells with the same ratio of encapsulated transgene to cell, the transduction activity of purified LVV particles was approximately half that of the pre-chromatographic counterpart).

[0143] [Table 7]

[0144] Example 4 In silico discovery and experimental evaluation of VSV-G binding peptides The results presented in the previous section demonstrate that peptide sequences identified by screening peptide libraries against VSV-G consistently performed better than sequences selected against complete LVV particles. Sequences targeting ligand sites on VSV-G are presented as a promising route to discovering peptide ligands for LVV purification. In this regard, insights into the interaction were gained from the published crystal structures of the complex formed by VSV-G and the cysteine-rich domains of the low-density lipoprotein receptor (LDL-R CR2 and LDL-R CR3), which are cell surface receptors that play a crucial role in LVV cell entry. Two cationic residues on VSV-G, His8 and Lys47, and residues targeting the anionic residues Asp69 and Asp73 on CR2, and Asp108 and Asp112 on CR3, contribute significantly to its LDL-R binding energy. In addition, the LDL-R CR2 and LDL-R CR3 domains are characterized by a compact tertiary structure firmly maintained by three disulfide bonds, a structure similar to that of scaffolds (i.e., Notchin, Avimer, and bicyclic peptides) used to discover small protein affinity ligands. Finally, analysis of pairwise interactions between active residues on LDL-R CR2 and LDL-R CR3 is reported in Figure 9 and Table 7, showing that the VSV-G binding site consists of residues in a continuous segment. In connection with this work, these data suggest that (i) small peptide mimes of LDL-R CR2 and LDL-R CR3 designed in silico to target the same CR binding site of VSV-G, and (ii) the highest-performing ligands identified by library screening, SKSAAEHE (SEQ ID NO: 6), GKEAAFAA (SEQ ID NO: 3), and FEKISNAE (SEQ ID NO: 1) (all containing at least one glutamic acid residue), can indeed target the LDL-R CR binding site of VSV-G.

[0145] Therefore, we designed in silico ensembles of candidate ligands whose sequences and structures mimic the LDL-R CR2 and LDL-R CR3 domains. Specifically, we designed four disulfide-cyclic sequences: C-cyclo[GSRQFVADSDRD]C-GSG (SEQ ID NO: 87), C-cyclo[GSRSFVGDSDRD]C-GSG (SEQ ID NO: 88), C-cyclo[GSRAFVADADRD]C-GSG (SEQ ID NO: 98), and C-cyclo[GSRAFVGDAD]C-GSG (SEQ ID NO: 99), and five linear sequences: SRQFVCGDSDRD-GSG (SEQ ID NO: 100), SRSFCDSDRD-GSG (SEQ ID NO: 101), SRAFVGDADRD-GSG (SEQ ID NO: 102), AFVGDADRD-GSG (SEQ ID NO: 103), and SFVRIGLSD-GSG (SEQ ID NO: 104). The sequence homology of the designed peptide and its cognitive CR2 and CR3 domains, as well as the small root mean square deviation (RMSD), provide confidence in the LDL-R mimicry behavior of the proposed sequence. The eight designed peptides, SKSAAEHE (SEQ ID NO: 6), GKEAAFAA (SEQ ID NO: 3), FEKISNAE (SEQ ID NO: 1), and their variants FEKISAAE (SEQ ID NO: 2) and FEKISTAE (SEQ ID NO: 11), were docked in silico to the VSV-G (PDB IDs: 5OY9 and 5OYL) crystal structure at various aqueous environments representing different buffers used during the purification process, namely, ionic strengths representing the binding buffer, pH 150 mM and 7.4, and 0.7 M and pH 7.4 or 1 M and pH 6, representing two alternative elution buffers (i.e., 50 mM PIPES buffer with 0.65 M NaCl (pH 7.4) and 20 mM citrate with 0.5 M MgCl2 (pH 6.0)). In peptide docking, we focused on the putative binding sites identified on the solvent-contacting surface of the protein, assuming that it is "ligand-like," meaning that its physicochemical and topological properties make it truly compatible with binding to biomolecular ligands.Another important constraint imposed during docking is the -GSG tripeptide attached to the C-terminus, which does not interact with the target VSV-G. This forces the -GSG tripeptide to orient outward from the binding pose, thus mimicking the orientation constraint imposed on the peptide by its conjugation on the surface of the chromatographic resin. In previous studies, this constraint has yielded excellent accuracy in estimating the target binding energy. The resulting VSV-G:peptide complex (selected based on its raster size and initial scoring using X-score) is given a reliable value for its binding free energy (ΔG). b To obtain the selected complex, a 250 ns MD simulation was performed under explicit solvent conditions representing the binding buffer and elution buffer. The selected complex is shown in Figure 3, with the corresponding dissociation constant (K D,in silico These are listed in Table 8.

[0146] [Table 8-1] [Table 8-2]

[0147] [Table 9]

[0148] The molecular docking results support the design criteria for the mimic sequences. Specifically, (i) with the sole exception of SKSAAEHE (SEQ ID NO: 6), the pairwise binding interactions of all peptides forming the complex reproduce the interactions of the VSV-G:LDL-R complex, and (ii) the binding strength of the VSV-G:peptide complex in its binding environment matches the binding strength of the VSV-G:LDL-R CR2 precursor and VSV-G:CR3 precursor (9.3-9.7 kcal / mol, K). D Approximately 5·10 -8 ~10 -7 It is moderately lower than M (5.9~8.7 kcal / mol, KD Approximately 5·10 -7 ~5.10 -6 (iii) All peptides except M) and (iii)SRTFVCDSDRD (SEQ ID NO: 94) showed equivalent affinity to the second binding site (indicated by the green surface in Figure 3). The ability of peptide ligands to target multiple binding sites on the target surface with true affinity suggests the formation of a multi-site interaction network between the virus and the peptide-functionalized surface. The LVV coat presented approximately 216 VSV-G proteins, with VSV-Gs arranged at 12.5–31.3 Å apart from each other (suggesting an LVV radius of 40–50 nm). In addition, the peptide density on the resin (approximately 30 μmol per gram) and the specific surface area of ​​the resin (approximately 30 m²) were also suggested. 2 Based on / g), the peptides are presented at approximately 18 Å from each other. This suggests the formation of 8–10 VSV-G:peptide interactions per bound LVV particle. The cooperation of multiple affinity interactions results in a strong avidity-like binding that promotes efficient and selective LVV capture, as indicated by the binding capacity and product purity values ​​shown below.

[0149] Furthermore, as shown in Table 8, the dissociation constant (K) of the VSV-G:peptide complex D The ionic strength of the environment increased 540–750 times when the ionic strength increased from 150 mM to 1.3 M (representing 50 mM PIPES elution buffer containing 0.65 M MgCl2), and increased 1,550–1,900 times when the pH decreased to 7.4–6.0 (representing 20 mM citrate elution buffer). This suggests that the adsorbed virus can be effectively released under conditions that protect its transduction activity, as confirmed by the product yield values. Analysis of its molecular simulation orbitals showed that VSV-G:peptide dissociation is strongly influenced by (i) the loss of Coulombic interaction between the anionic residue in the peptide ligand and its cationic counterpart on VSV-G, mainly Lys47, Arg342, and Arg354 (contributing approximately 34–41% of the binding energy at pH 7.4).

[0150] The second major contributor to the free energy of binding, the hydrogen bond and polar interaction network, for example, the network formed by Gln10, Ser179, Asn180, Ser183, Thr350-352, and Glu353 (contributing 31-39% of the binding energy), also disappeared upon the addition of MgCl2 (a known chaotropic substance that destabilizes the interaction between electrostatic and hydrogen bonds). Increased ionic strength and decreased pH also slightly contract the solvent-contact pocket of VSV-G, and this rearrangement significantly reduces the structural complementarity of the putative pocket to the peptide ligand, promoting its detachment from the coat protein. The energies of both the VSV-G:LDL-R CR2 and VSV-G:LDL-R CR3 complexes also decreased when the simulation environment was switched to elution conditions. However, under the control elution conditions (ΔG b The residual intensity at approximately 7.2–8.2 kcal / mol is higher than the intensity observed in the VSV-G:peptide complex, suggesting that product elution from the protein ligand is even more difficult, which explains why stronger denaturation conditions are required for lentiviral elution from antibody-based ligands (e.g., 0.8 M arginine is recommended for elution from CaptureSelect® Lenti VSVG affinity resin).

[0151] Based on the predicted VSV-G affinity and binding loss at pH 7.4 under the applied elution conditions, peptides C-cyclo[GSRAFVGDAD]C (SEQ ID NO: 16), SRQFVCGDSDRD (SEQ ID NO: 17), SRAFVGDADRD (SEQ ID NO: 4), and SFVRIGLSD (SEQ ID NO: 5) were conjugated onto Poros resin, and LVV was evaluated by purifying it from the clarified HEK293 cell culture recovery using an optimized PIPES-based buffer system. The results summarized in Table 9 confirm the criteria adopted in the in silico peptide design. Specifically, the cyclic peptide that yielded the highest HCP removal value was registered in this test (i.e., a residual HCP titer of approximately 0.34 μg / mL (corresponding to an 871-fold reduction)), but rather resulted in an insufficient amount of product. Conversely, SRAFVGDADRD (SEQ ID NO: 4) and SFVRIGLSD (SEQ ID NO: 5) allowed for a yield of LVV units transduction into cells that was equivalent to that obtained with FEKISNAE (SEQ ID NO: 1) and GKEAAFAA (SEQ ID NO: 3), while maintaining a more than 100-fold reduction in HCP and approximately 68-fold reduction in DNA. Therefore, they were selected for further characterization.

[0152] [Table 10]

[0153] These results support the development (in silico or in vitro) of VSV-G targeting peptides as affinity ligands for the purification of lentiviruses from recombinant materials. Thanks to their moderate affinity and ability to form multiple interactions resulting in product capture facilitated by strong avidity, VSV-G targeting peptides can outperform antibody-based ligands in product yield under non-denaturing conditions, while being comparable to antibody-based ligands in terms of binding capacity and selectivity. Furthermore, the use of chemically stable amino acids and the loss of tertiary structural properties in short peptides in constructing resin-bound libraries or in silico ensembles of LDL-R mimetic compounds encourages the selection of ligands that are likely to be more robust than protein binders. The latter aspect is particularly relevant in the manufacture of biopharmaceuticals, as it affects the number of uses that affinity resins can withstand. This number represents a crucial determinant of process operating costs and, ultimately, the price of the drug to the patient.

[0154] Example 5 Dynamic binding capacity and alkali stability of peptide-Poros resins Based on the results in Tables 6 and 9, the dynamic binding capacity (DBC) of the adsorbents FEKISNAE (SEQ ID NO: 1)-Poros resin, GKEAAFAA (SEQ ID NO: 3)-Poros resin, SRAFVGDADRD (SEQ ID NO: 4)-Poros resin, and SFVRIGLSD (SEQ ID NO: 5)-Poros resin was determined. 10% ) and stability were measured. Conventional literature (by loading a pure virus solution into DBC) 10% Unlike measuring the value of (which does not represent the actual process flow), the titer is approximately 10 8Breakthrough experiments were performed by loading the clarified bioreactor recovery containing TU / mL LVV particles. Measurements were taken at two residence times: 2 minutes and 1 minute. 2 minutes was recommended for CaptureSelect® Lenti VSVG affinity resin and adopted for comparability in this work, while 1 minute was adopted to reduce the process time of LVV particles and transduce LVV units into cells, achieving higher yield units. The results are recorded in Figure 4 and summarized in Table 10 (DBC). 10% From the corresponding values, all peptide-based adsorbents, DBC 10% However, the values ​​are shown to be high, comparable to or exceeding those of commercially available affinity resins. Specifically, GKEAAFAA (SEQ ID NO: 3)-Poros resin and SFVRIGLSD (SEQ ID NO: 5)-Poros resin showed values ​​of 1.91·10 at RT1 min. 10 vg / mL and 3.99·10 10 Prominent DBC in vg / mL 10% The FEKISNAE (SEQ ID NO: 1)-Poros resin and SRAFVGDADRD (SEQ ID NO: 4)-Poros resin are characterized by DBC 10% Value 5.84·10 9 vg / mL and 6.89·10 9 The value was slightly low at vg / mL (RT1 min), but still quite high. For reference, the DBC of CaptureSelect™ Lenti VSVG resin. 10% At RT2 minutes, the ratio was 9.73.10 10 The concentration is vp / mL, and Poros (trademark) 50 HE Heparin Affinity Resin is 1.0·10 at RT0.5 min. 8The value is TU / mL. The ability of peptide-functionalized adsorbents to capture an equivalent amount of LVV particles while reducing process time by 50% may stem from the "flexible" biorecognition mechanism of the peptide ligand. As suggested by docking tests (which returned several high-probability binding poses at each sequence (Figure 3)), the interaction with VSV-G formed by the peptide ligand appears to be less orientation-dependent than that formed by the protein. This may lead to the peptide-functionalized surface facilitating rapid binding at multiple sites on LVV particles, manifesting as faster absorption kinetics, resulting in equal binding capacity at shorter residence times or higher capacity at longer residence times.

[0155] The ratio of LVV titers in the eluate (C) did not reach the corresponding value at plateau loading (C0). Segura et al. and Moreira et al. also reported a C / C0 plateau of approximately 0.8 when measuring the LVV binding capacity of heparin-functionalized resins (see Biotechnol.Bioeng.,2005,90:391-404 and Int.J.Mol.Sci.2023,24,3354, respectively). To evaluate the role of LVV loss in tubing at plateau values ​​of LVV titer, HEK293 CCCF was loaded into an FPLC system without a column, and the transduction assay of the eluted fraction was performed immediately after it was dispensed into the fraction collector. As expected, analysis of the eluted material revealed a 5-10% loss of LVV transduction activity, which can be attributed to shear, non-ideal temperature, or adsorption to the inner wall of the chromatography apparatus (Note: The FPLC system is constructed with inert tubing).

[0156] [Table 11]

[0157] Along with binding capacity, another critical parameter in downstream bioprocesses is resin stability to clean-in-place (CIP) washing. Caustic treatment with concentrated sodium hydroxide aqueous solution (0.1-0.5 M) (established in antibody production) is now moving on to the production of viral vectors for in vivo and ex vivo gene therapy. Commercial resins for AAV purification, POROS CaptureSelect AAVX and AVIPure affinity resins, were designed to withstand multiple cycles of reuse in intermediate caustic washing. However, at present, these ligands have not yet reached the chemical stability of the latest generation of protein A for mAb purification, which, through decades of engineering, can withstand many cycles of washing with 0.5 M NaOH. Similarly, affinity technologies for LVV purification are still in their early stages, and newly introduced ligands have not yet reached the molecular engineering pathways that lead to high chemical stability; therefore, the recommended CIP conditions for Poros® 50 HE Heparin and CaptureSelect® Lenti VSVG resins are limited to 25 mM NaOH.

[0158] The instability of protein-based ligands is often linked to deamidation of asparagine / glutamine (N / Q) residues and loss of tertiary structure due to exposure to high pH, ​​as observed with native protein A. Conversely, three of the four specified peptides, namely GKEAAFAA (SEQ ID NO: 3), SRAFVGDADRD (SEQ ID NO: 4), and SFVRIGLSD (SEQ ID NO: 5), do not contain either N or Q, feature only an α-helix secondary structure, and can be recovered immediately upon incubation at neutral pH. On the other hand, FEKISNAE (SEQ ID NO: 1) is expected to be converted to FEKISDAE upon alkaline washing by deamidation of N to aspartic acid (D). Indeed, upon exposure to flowing 50 mM NaOH, the FEKISNAE (SEQ ID NO: 1)-Poros resin lost approximately 50% of its binding capacity, and after 30 minutes of static contact with 0.1 M NaOH, it showed no measurable binding of any LVV particles. Therefore, alkali-stable variants FEKISAAE (SEQ ID NO: 2) and FEKISTAE (SEQ ID NO: 11) were designed in silico to have VSV-G binding and elution activity equivalent to that of the cognitive sequence (Table 8).

[0159] The adsorbents GKEAAFAA (SEQ ID NO: 3)-Poros resin, FEKISAAE (SEQ ID NO: 2)-Poros resin, SRAFVGDADRD (SEQ ID NO: 4)-Poros resin, and SFVRIGLSD (SEQ ID NO: 5)-Poros resin were subjected to LVV purification from HEK293CCF in 5 consecutive cycles, along with intermediate CIP in 0.5M NaOH. The lifetime tests shown in Figure 5 support the criteria adopted in peptide design. Specifically, GKEAAFAA (SEQ ID NO: 3) maintained its binding capacity and selectivity, and 10 9It consistently adsorbed TU / mL+, and the average yield was over 40% (over 50% depending on qPCR) of LVV units transduction to cells, and HEK293 HCP was reduced 130-300 times (Figure 5A). FEKISAAE (SEQ ID NO: 2) showed purification performance comparable to its cognitive FEKISNAE (SEQ ID NO: 1), but with significantly higher stability. The adsorbent maintained its volume (approximately 5-10) throughout the subsequent 5 purification cycles. 8 The TU / mL was maintained, and a more than 200-fold reduction in HCP was achieved, along with a consistent product yield of over 50% (Figure 5B). Conversely, with FEKISTAE (SEQ ID NO: 11)-Poros resin, column-bound LVV significantly decreased within an increasing number of CIP cycles (data not shown), and the slight acidity of the threonine residue may indicate a negative charge that reduces the binding capacity of this peptide. Finally, SRAFVGDADRD (SEQ ID NO: 4) and SFVRIGLSD (SEQ ID NO: 5) showed high LVV binding (approximately 10%). 9 The elution yield (TU / mL) and elution yield (over 38%), as well as impurity removal over continuous cycles, were maintained.

[0160] Example 6 Lentiviral purification performance of chromatography resins functionalized with affinity peptide ligands targeting VSV-G In the following series of examples, experiments were conducted to characterize peptides by evaluating their performance on various matrices, focusing on binding capacity, recovery rate of LVV units transduction into cells, and removal of host cell proteins (HCPs). To this end, various material compositions, resins with particle and pore diameters, and functional densities were tested. Experiments were also conducted to test membranes with varying fiber morphology, porosity, and ligand distribution.

[0161] Material composition, bead diameter, pore size distribution, specific surface area, and ligand density are critical design parameters that determine the binding capacity and selectivity of a chromatography substrate, i.e., the throughput and quality of its purified product. Most published literature in this field focuses on the isolation of therapeutic proteins, particularly monoclonal antibodies, using affinity (e.g., protein A, protein G, and protein L) mixed modes and ion-exchange ligands. Nevertheless, the impact of the parameters listed above on productivity and quality is almost certainly more pronounced in viral vectors, given their larger size and complexity and lower stability. At the same time, this dependence is quite difficult to predict and must be confirmed experimentally. Several studies have investigated these phenomena in the context of adeno-associated virus vector (AAV) purification, focusing on either affinity adsorbents for product capture or anion exchange for product polishing. On the other hand, similar studies on lentiviral vectors (LVVs) are relatively insufficient, despite their increasing relevance in cell and gene therapies. This is partly due to the difficulty in handling and analyzing LVVs, and partly due to the recent introduction of affinity ligands targeting pseudotyped LVVs with VSV-G protein. To explore bridging this knowledge gap, this study focused on LVV purification using chromatographic substrates of varying composition and morphology, starting with resins made from polystyrene (PS, Poros), poly(methyl methacrylate) (PMMA, ToyoPearl®), poly(vinyl ether) (PVE, Eshmuno), polyacrylamide / azalactone (Ultralink), or cross-linked agarose (WorkBeads). The pore sizes of these matrices ranged widely (50–10000 nm), while the bead diameters ranged from 45–75 μm.

[0162] When using resins from Eshmuno, ToyoPearl (registered trademark), and Poros (trademark), particularly high DBC values ​​are achieved. 10% (1.9-5.0·10 per 1 mL of resin)9 A TU was obtained. This is likely due to its higher functionalization, which results in a higher ligand density (0.05–0.2 mmol of peptide per 1 mL of resin, mmol / mL). In contrast, the ligand densities of the agarose-based and polyacrylamide / azalactone ace resins were more moderate (0.02–0.05 mmol / mL), and their volumes were consistently 10 9 The values ​​were below TU / mL (Table 11). The recovery rate of bound LVV particles appeared to correlate with pore diameter; agarose resins such as Poros, ToyoPearl® 750, Eshmuno 800-50, and GenScript featured larger pores that returned 60-70% of the supplied LVV, while other resins showed lower product yields.

[0163] Based on the recovery rate, we derived a critical process-related parameter (TU / mL·min) defined as the number of LVV units transduced from 1 mL of resin per minute, representing productivity. The values ​​differed significantly, by only 2.4 to 10. 8 2.9-10 TU / mL·min of GKEAAFAA (SEQ ID NO: 3)-Ultralink resin 9 The GKEAAFAA (SEQ ID NO: 3)-Poros resin was in the range of TU / mL·min (41 times higher than the commercially available adsorbent CaptureSelect® Lenti VSVG Affinity Resin).

[0164] Finally, all resins except ToyoPearl® 750 and Ultralink showed a reduction of more than 100 times in HCP. These results indicate that LVV purification performance (primarily promoted by the peptide GKEAAFAA (SEQ ID NO: 3)) is significantly influenced by the morphological characteristics of its base matrix. Based on these results, the GKEAAFAA (SEQ ID NO: 3)-Poros® resin (with a binding capacity approximately twice as high as commercially available standard VSV-G Capture Select) was selected for further characterization.

[0165] [Table 12]

[0166] Example 7 The effect of target gene (GOI) size on LVV purification. Further characterization of the GKEAAFAA (SEQ ID NO: 3)-Poros® resin was performed by evaluating the effects of GOI sequence and size. While the size limitations of ssRNA cargo in LVV remain debated, GOIs up to 10 kilobases are considered limiting in terms of process feasibility and sufficient generation of transduction units. Therefore, the purification of LVV (approximately 9.5 kilobases) encapsulating a GOI encoding a CRISPR Cas9 nuclease-GFP fusion from clarified HEK293F cell culture medium (CCF) was evaluated. As discussed herein, increasing GOI size leads to lower viral titer and requires a larger loading volume (Figure 10). Affinity purification resulted in a 40% viral recovery rate, a 4.5-fold enrichment factor, a 65-fold reduction in HCP, and a residual titer of 4.14 μg / mL. Lower recovery rates can be attributed to loading time, which increases from 30 minutes for GFP-LVV to 2.5 hours for Cas9 / GFP-LVV. Recent studies have shown that longer contact times between LVV and its chromatographic matrix lead to morphological changes in the LVV, stronger binding at multiple points, and decreased viral recovery rates.

[0167] Example 8 The effect of spacer arms and ligand density on the performance of ligand GKEAAFAA (SEQ ID NO: 3) Introducing a spacer arm between the ligand and the chromatographic support has been used to facilitate product capture. Optimizing ligand presentation is particularly relevant in the context of viral vector purification due to their large size, low curvature of the virion surface, and low reliance on multi-point interactions to achieve sufficient binding strength and capacity. Therefore, we evaluated the insertion of a single G(gly) or tripeptide GSG(gly-ser-gly) linking the C-terminus of GKEAAFAA (SEQ ID NO: 3) to POROS® resin beads. The resulting adsorbent was then subjected to a clarified HEK293 cell culture with assumed binding capacity (DBC per 1 mL of resin). 10% Approximately 5·10 9 Loading was performed at up to 75% of the TU, and at that point, no virus was detected in either the flow-through or wash fraction. Somewhat unexpectedly, no significant differences in LVV recovery rates were recorded among the various adsorbents (Table 12).

[0168] [Table 13]

[0169] Ligand density can also determine the performance of chromatographic adsorbents; higher ligand densities result in greater binding capacity, but also decrease binding selectivity and product recovery. To evaluate the impact of ligand density on LVV binding capacity and productivity, four lots of GKEAAFAA (SEQ ID NO: 3)-Poros® resin were produced with ligand densities ranging from 16 to 69 μmol per mL of resin (Table 13). Binding capacity did not substantially change with ligand density; even a fourfold difference in ligand density resulted in only a 1.4-fold increase. While steric effects due to the large size of LVV particles are likely the primary factor determining capacity, ligand density has only a secondary effect on affinity resins with a given pore diameter. Conversely, ligand density determines binding strength through interactions at multiple sites (e.g., avidity effect), thus determining the product yield during elution. As ligand density increased, LVV recovery decreased from 80% to 37%, reducing productivity, and the number of units transductioned to cells per minute per mL of resin decreased to 5-10. 8 from 3.3·10 8 (LVV bonding was not recorded on the control OH-Poros® resin).

[0170] [Table 14]

[0171] Example 9 Testing of the lifespan and stability of GKEAAFAA-Poros® resin. Gene accessibility and cell therapy depend on lower manufacturing costs and increased sustainability. While the majority of manufacturing costs are currently associated with upstream materials (e.g., culture media, plasmids, and transfection reagents), the development of stable cell lines for viral vector expression is likely to shift the focus of cost management to downstream components. In this context, the lifespan of chromatographic adsorbents is a critical factor in reducing operational costs and consumable waste flows. Affinity adsorbents, particularly due to their high cost, are expected to be reused over multiple cycles, each involving a regeneration process with a strong denaturing solvent and a clean-in-place (CIP) process with caustic conditions. For most affinity resins currently on the market, the recommended CIP conditions for AAV and LVV purification are significantly milder (10 mM NaOH) than the conditions routinely applied to established affinity adsorbents such as protein A / G resins (0.5 M NaOH).

[0172] Therefore, using 0.5 M NaOH (15 CV flow followed by 30 minutes of static contact), the ability of GKEAAFAA (SEQ ID NO: 3)-Poros (trademark) to perform continuous cycles of LVV purification, along with intermediate CIP, was tested. Binding capacity, yield of encapsulated transgenes, LVV units transduction into cells, and HCP removal values ​​measured over 50 cycles are summarized in Figure 11, clearly demonstrating the stability of GKEAAFAA (SEQ ID NO: 3). LVV DBC 10% The values ​​and yields are 2.10 per 1 mL of resin. 9 While TU and 54% fluctuated, the decrease in HEK293 HCP was consistently more than 100-fold. These results are attributed to the amino acid sequence of the peptide ligand, which does not contain residues that tend to deamidate (such as Asn(N) or Gln(Q)) or residues that tend to oxidize (such as Trp(W), Cys(C), or Met(M)).

[0173] Example 10 Membrane as an alternative substrate to chromatography resins for LVV purification The high binding capacity characteristic of chromatographic resins is rooted in the large surface area of ​​their pores; however, the curved morphology and limited diameter of these pores restrict diffusion, reducing the transport of larger biological agents, such as viruses. LVV particle binding was confirmed by loading fluorescently labeled LVV onto GKEAAFAA(SEQ ID NO: 3)-Poros™ and evaluating the distribution of green fluorescence across the bead volume, confirming that LVV was confined to the bead surface (Figure 12). Membranes offer an excellent alternative to resins because their open porosity removes diffusion restrictions and allows for significantly higher flow rates at lower pressure drops. However, to date, while most commercially available affinity membranes are usable for protein purification, only pseudo-sulfated cellulose affinity membranes have been used to purify influenza A with a 68% recovery rate. The development of affinity membranes is particularly useful for LVV purification because it protects the transduction activity of purified virions by shortening their contact time with the chromatographic matrix.

[0174] Therefore, peptide ligands GKEAAFAA (SEQ ID NO: 3), FEKISNAE (SEQ ID NO: 1), and SRAFVGDADRD (SEQ ID NO: 4) were conjugated onto commercially available Natrix® membranes and commercially available activated cellulose membranes. The Natrix® membrane contained a continuous hydrogel matrix embedded in a polyamide fiber scaffold. The ligands were conjugated onto the hydrogel matrix and the epoxide groups presented in its graft layer to achieve a functional density (approximately 0.05 mmol per gram) equivalent to that of peptide-functionalized resins. Along with LVV recovery and HCP removal, its DBC was also evaluated. 10% The obtained adsorbents were evaluated by measuring their properties at residence times (RT) of 0.25 minutes and 0.5 minutes (Table 14).

[0175] The combination of a smaller surface area and shorter residence time resulted in approximately a 10-fold decrease in binding capacity compared to the lead resin, GKEAAFAA(SEQ ID NO: 3)-Poros(Trademark). However, it is important to note that comparable productivity was achieved with GKEAAFAA(SEQ ID NO: 3)-Natrix(Trademark) and GKEAAFAA(SEQ ID NO: 3)-cellulose membranes by making full use of short residence times throughout the chromatography process. The difference in productivity is mainly due to the lower recovery rates obtained with the affinity membranes (34.0±3.0% and 32±2%, respectively). This can be attributed to the "trapping" of virions in the polymer matrix surrounding or coating the fibers, which shortens the LVV desorption rate and ultimately leads to lower recovery rates with shorter residence times. Notably, the functionalization strategies did not yield the binding selectivity of the affinity membranes, and consistently high purified LVV was obtained. In particular, SRAFVGDADRD(SEQ ID NO: 4)-Natrix® significantly reduced HCP by approximately 500 times, achieving the highest reported value in LVV affinity purification. In contrast, Mustang Q (a powerful anion exchange membrane widely used for LVV capture) yielded a higher product recovery rate than the affinity membrane, but the purity was significantly lower (the reduction in HCP was one-fifth that of the affinity membrane).

[0176] [Table 15]

[0177] Example 11 Demonstration of a benchtop scale process for LVV purification The results obtained from LVV expression and affinity purification were combined in a downstream process of step X, which included (i) clarification by centrifugation and microfiltration, (ii) affinity-based capture using GKEAAFAA (SEQ ID NO: 3)-Poros® resin in binding and elution modes, (iii) polishing in flow-through mode, (iv) concentration by tangential flow filtration for enrichment, and finally (v) dialysis filtration and sterile filtration. HEK293 cell culture was first clarified by centrifugation and filtration using a 0.45 μm vacuum filter, and then treated with benzonase to remove residual plasmids and HEK293 host cell DNA (hcDNA). After filtration, a very slight decrease (approximately 2%) in the titer of LVV units transduction to cells was observed.

[0178] The clarified culture solution is then placed on GKEAAFAA (SEQ ID NO: 3)-Poros (trademark) resin and then the DBC 10% The column was loaded to 100%. Previous observations of viral vector purification had shown that significantly overloading the column resulted in higher yields. LVV elution resulted in an LVV yield of 68% and a concentration factor of 2.53 (productivity 1.25). . 10 14 TU·hr -1 ·L resin -1 (This corresponds to...), and along with this, HCP decreased 120-fold, with a residual level of 6.7 μg / mL and residual DNA becoming undetectable. Most notably, the ratio of total particles to infected particles (TP / IP) significantly decreased from 145 to 46 during the affinity capture process, which corresponds to a 3-fold concentration of LVV units transduction into cells (Table 15).

[0179] Following the affinity step, a polishing step was performed using CaptoCore700 in flow-through mode to remove residual impurities. This CaptoCore700 has been used in previous studies for the purification of large viral vectors (e.g., lentiviruses and adenoviruses). The CaptoCore beads contain an outer layer (shell) whose pores allow HCP and other biomolecular contaminants to reach the inside of the beads, where they are captured by a combination of ion exchange and hydrophobic interactions, while removing LVVs due to size differences. This polishing step yielded a high yield (95%) and a further 20-fold reduction in HCP, achieving a residual level of 0.2 μg / mL. The eluate was then concentrated, buffered, and sterile filtered. This process resulted in an overall yield of 33%, a 13,000-fold reduction in HCP content, and undetectable residual DNA by PicoGreen Assay.

[0180] [Table 16]

[0181] Example 12 Optimization of LVV expression in suspension HEK293 cell culture medium Lentiviruses are produced almost ubiquitously by transfection of HEK293F cells using four plasmids: a plasmid carrying the target gene (GOI), an envelope plasmid, and two packaging plasmids. LVV can be produced in both adherent and suspension HEK293F cells, but suspension production is preferred due to its ease of scalability, higher titer, and the absence of the need for FBS supplementation. Inspired by studies on the effects of cell culture medium composition on LVV and HCP titers, three cell culture media—Peak Expression, BalanCD HEK293, and LV-Max—were tested. The LVV titers measured by transduction assays and HCP titers measured by ELISA are shown in Figure 12. When small GOIs encoding green fluorescent protein (GFP, approximately 4.5 kb) were encapsulated in capsids, no significant differences in LVV titers were observed across various media. However, HCP levels in the Peak Expression medium were 1.6 times higher than those obtained in the BalanCD medium, and HCP levels in LV-max were 2.1 times higher (Figure 12A). This suggests the role of media composition in determining cellular metabolism and, consequently, HCP production.

[0182] The effect of gene payload size on LVV production was also investigated by comparing a GOI encoding GFP (4.5kb) with a GOI encoding CRISPR Cas9 (9.5kb) fused with GFP. The results, summarized in Figure 12B, showed a similar trend, with BalanCD exhibiting higher productivity and purity than Peak Expression and LV-Max. However, the functional titer of LVV with Cas9 encapsulated in the capsid was approximately 1 / 100th that of its GFP-loaded counterpart. These results are consistent with the conclusions of Kumar et al., who investigated the effect of GOI size on LVV titer and observed that functional titer decreased to a similar extent as the size of the transgene increased. 48The authors presented two hypotheses for titer reduction: (i) during the assembly of viral particles, larger GOIs become more difficult to encapsulate, and the rate at which proviral RNA is transported from the cell nucleus to the cytoplasm decreases with size.

[0183] Transient transfection (a method commonly used for virus production) is mediated by a transfection reagent that forms a complex with a plasmid and transports it to its host cell. Various transfection reagents are documented in the literature, including cationic polymers of calcium phosphate, such as polyethyleneimine (PEI), and lipids. To evaluate the role of transfection reagents in functional titer, three commercially available products, namely TransIT, PEIpro, and LV-Max, were tested. The results, summarized in Figure 13, show that TransIT and LV-Max yielded titers three times higher than PEIpro. While the exact compositions of these transfection reagents are not disclosed, PEIpro is reported to be a pure cationic polymer, while TransIT and LV-Max contain mixtures of cationic polymers and cationic lipids.

[0184] Improving the efficiency and cost of downstream processing of viral vectors is crucial for creating next-generation medicines that are available to all patients who need them. While significant progress has been made in recent years, particularly in the frontline of AAVs for in vivo gene therapy, a mature purification toolkit for LVVs still does not exist. Embodiments of this disclosure include the discovery of selective, robust, and extensible peptide ligands targeting pseudotype VSV-G LVVs, and proceed to the evaluation of bioprocess-related parameters, namely binding capacity, lifetime, productivity, and end-to-end downstream process compatibility. The results of this disclosure demonstrate that peptide-functionalized adsorbents (i) have the ability to capture LVVs carrying various genetic payloads with high binding capacity, and (ii) yield an eluate rich in LVV units for transduction into cells, with significantly reduced protein and nucleic acid contaminants. The combination of high purification performance, long lifetime, and low manufacturing cost compared to commercially available affinity adsorbents places peptide-functionalized resins at the forefront of LVV purification technology. In addition, the peptide-functionalized membranes were evaluated as being able to operate with one-tenth the residence time characteristic of resin-based chromatography, aligning with the goal of improving productivity.

[0185] 6. Array The sequences related to embodiments of this disclosure are shown in the following table.

[0186] [Table 17-1] [Table 17-2]

Claims

1. A peptide for purifying lentiviruses from a sample, At least one cationic amino acid and / or at least one anionic amino acid, At least one aromatic amino acid, One or more aliphatic amino acids, The peptide comprising the above.

2. The peptide according to claim 1, having a length of 5 to 20 amino acids.

3. The peptide according to claim 1 or claim 2, having a length of at least 8 amino acids.

4. A peptide according to any one of claims 1 to 3, having a length of 8 amino acids.

5. A peptide according to claim 1 or claim 2, having a length of 12 to 20 amino acids.

6. The peptide according to any one of claims 1 to 5, comprising at least one cationic amino acid and at least one negatively charged amino acid.

7. The peptide according to any one of claims 1 to 6, wherein a cationic amino acid is adjacent to an anionic amino acid.

8. The peptide according to any one of claims 1 to 7, wherein each of the at least one cationic amino acid is independently selected from histidine, lysine, and arginine.

9. The peptide according to any one of claims 1 to 8, wherein the at least one cationic amino acid is lysine.

10. The peptide according to any one of claims 1 to 9, wherein each of the at least one anionic amino acid is independently selected from aspartic acid and glutamic acid.

11. The peptide according to any one of claims 1 to 10, wherein the at least one anionic amino acid is glutamic acid.

12. The peptide according to any one of claims 1 to 11, wherein each of the at least one aromatic amino acid is independently selected from histidine, phenylalanine, tyrosine, and tryptophan.

13. The peptide according to any one of claims 1 to 12, wherein the at least one aromatic amino acid is phenylalanine.

14. The peptide according to any one of claims 1 to 12, wherein the at least one aromatic amino acid is histidine.

15. The peptide according to any one of claims 1 to 14, wherein the one or more aliphatic amino acids are selected from alanine, glycine, isoleucine, leucine, proline, and valine.

16. A peptide according to any one of claims 1 to 15, which does not contain asparagine, glutamine, and / or tryptophan.

17. The peptide according to any one of claims 1 to 16, which is a cyclic peptide.

18. The peptide according to any one of claims 1 to 17, wherein the cyclic peptide is cyclized by a disulfide bond between two cysteine ​​residues.

19. The peptide according to any one of claims 1 to 18, comprising an amino acid sequence having at least 80% sequence identity with one of sequence numbers 1 to 86.

20. A peptide according to any one of claims 1 to 19, comprising an amino acid sequence having one of sequence numbers 1 to 86.

21. A peptide according to any one of claims 1 to 20, comprising an amino acid sequence having SEQ ID NO: 1, 3, or 4.

22. The peptide according to any one of claims 1 to 18, comprising an amino acid sequence having one, two, three, four, or five substitutions compared to one of sequence numbers 1 to 86.

23. The peptide according to any one of claims 1 to 20, comprising an amino acid sequence having one, two, three, four, or five substitutions compared to one of sequence numbers 1, 3, and 4.

24. The peptide according to any one of claims 1 to 22, further comprising a linker.

25. The peptide according to claim 24, wherein the linker is bonded to the C-terminus of the peptide, and the linker is a glycine-rich linker.

26. The peptide according to any one of claims 1 to 25, wherein the lentivirus is a vesicular stomatitis virus glycoprotein (VSV-G) pseudotype lentivirus.

27. A peptide according to any one of claims 1 to 26, which binds to a VSV-G envelope protein.

28. The peptide according to claim 27, which targets the binding site of the low-density lipoprotein receptor LDL-R on VSV-G.

29. A composition for purifying lentiviruses from a sample, comprising at least one peptide as described in any one of claims 1 to 28.

30. The composition according to claim 29, wherein the at least one peptide is bound to a solid support.

31. The composition according to claim 30, wherein the solid support comprises non-porous particles or porous particles, a membrane, a plastic surface, fibers, a woven or nonwoven fiber mat, a hydrogel, a microplate, a monolith, and / or a microfluidic device.

32. The composition according to claim 30 or claim 31, wherein the solid support comprises polymethacrylate, polyolefin, polyester, polystyrene, polysaccharide, polyvinyl ether, iron oxide, silica, titania, agarose and / or zirconia.

33. An adsorbent comprising at least one peptide according to any one of claims 1 to 25 or a composition according to any one of claims 29 to 32.

34. Each 1 mL of adsorbent contains at least 10 LVV units that transduce to cells. 8 The adsorbent according to claim 33, wherein the binding capacity (TU / mL) is...

35. Each 1 mL of adsorbent contains at least 10 LVV units that transduce to cells. 9 The adsorbent according to claim 33 or claim 34, wherein the binding capacity (TU / mL) is...

36. For every 1 mL of adsorbent and every 1 minute of purification time, at least 10 LVV units are transductioned to cells. 8 An adsorbent according to any one of claims 33 to 35 that provides a productivity (TU / mL·min).

37. For every 1 mL of adsorbent and every 1 minute of purification time, at least 10 LVV units are transductioned to cells. 9 An adsorbent according to any one of claims 33 to 36 that provides a productivity (TU / mL·min).

38. A method for purifying lentiviruses from a sample, Contacting a sample containing the lentivirus with at least one peptide according to any one of claims 1 to 28, a composition according to any one of claims 29 to 32, or an adsorbent according to any one of claims 33 to 37, such that the at least one peptide binds to the lentivirus, and The lentivirus is eluted from the at least one peptide. The method comprising the above.

39. The method according to claim 38, wherein the sample is a biological fluid.

40. The method according to claim 39, wherein the biological fluid is a cell culture medium.

41. The method according to claim 39 or claim 40, wherein the biological fluid comprises a supernatant and / or cell lysate.

42. The method according to any one of claims 39 to 41, wherein the biological fluid is derived from a virus-producing cell line.

43. The method according to any one of claims 38 to 42, wherein the lentivirus is a vesicular stomatitis virus glycoprotein (VSV-G) pseudotype lentivirus.

44. The method according to any one of claims 38 to 43, wherein the elution is performed at a pH of approximately 6.0 to approximately 8.

0.

45. The method according to any one of claims 38 to 44, further comprising a washing step before eluting the lentivirus.

46. The method according to any one of claims 38 to 45, which reduces host cell proteins by at least 100 times.

47. The method according to any one of claims 38 to 46, wherein the yield of the LVV unit transduction to cells is at least 35%.

48. For each milliliter of adsorbent used, at least 10 LVV units should be transductioned to the cells per minute of purification time. 8 The method according to any one of claims 38 to 47, which increases productivity.

Images