An engineered fusion protein and its mediated active budding vesicle system and use in repairing ex vivo marginal donor organs

Abstract

The application provides an engineered fusion protein and an active budding vesicle system mediated by the engineered fusion protein and application in repairing an ex vivo marginal donor organ, and belongs to the technical field of synthetic biology, regenerative medicine and organ transplantation. The engineered fusion protein provided by the application comprises a functional fragment of a human ARRDC1 protein, a flexible connecting peptide and an RNA binding functional domain. The application utilizes the engineered fusion protein and the active budding vesicle system mediated by the engineered fusion protein to deliver nucleic acid molecules, small molecule compositions and polypeptides in an ex vivo normal temperature perfusion, and can realize deep repair of various marginal donor organs such as fatty liver and old donor pig liver within 4-8 hours, and the performance is that steatosis reduction is greater than 60%, fibrosis reduction is greater than 50%, and epigenetic age reversal is 58%-77%, and the system is easy to clean, and is high in safety.

Description

Technical Field

[0001] This invention relates to the fields of synthetic biology, regenerative medicine and organ transplantation, and in particular to an engineered fusion protein and its mediated active budding vesicle system and its application in repairing ex vivo marginal donor organs. Background Technology

[0002] Organ shortage is a global medical crisis. To expand the donor pool, clinicians are forced to widely use "marginal donor organs," such as those from fatty liver disease, elderly donor kidneys, organs donated after cardiac death (DCD), and damaged hearts or lungs. These organs exhibit reversible functional impairment or aging characteristics, posing an extremely high risk of primary non-function or delayed functional recovery after direct transplantation. Ex vivo normothermic mechanical perfusion (NMP) provides a controlled time window (typically 4-12 hours) for assessing and repairing such organs before transplantation. However, within the limited time of NMP, achieving safe, efficient, and deep active organ repair faces two major delivery bottlenecks: First, there is a trade-off between vector efficiency and safety: Currently, viral vectors and lipid nanoparticles remain the main choices for delivering gene therapy and therapeutic RNA. However, these methods have significant limitations, which manifest in multiple dimensions, including in vivo distribution, payload capacity, immunogenicity, dose-limiting toxicity, and manufacturing processes. Viral vectors (such as AAV) exhibit delayed expression and are difficult to clear; synthetic vectors (such as LNP) are highly inflammatory, have uneven penetration, and tend to accumulate only in the liver, making extrahepatic delivery difficult. In endogenous delivery systems, the ARRDC1-mediated ARMMs system suffers from low packaging efficiency and contains non-endogenous viral components such as NiV, posing safety risks (Targeted Intracellular Delivery via Precision Programming of ARRDC1-Mediated Microvesicles, J Extracell Vesicles, 2025, PMID: 41392542). Meanwhile, the endogenous PEG10-mediated SEND system (Mammalian retrovirus-like protein PEG10 packages its own mRNA and can be pseudotyped for mRNA delivery, Science, 2021, PMID: 34413232) lacks active vesicle budding, resulting in insufficient yield. A new system that integrates the advantages of both systems while overcoming their disadvantages is urgently needed. Furthermore, both systems currently lack data on organ-level delivery of nucleic acids and proteins.

[0003] Second, there is a lack of universal platforms and patent circumvention strategies: Existing technologies are mostly fragmented solutions targeting single organs or single pathologies, lacking a universal platform that can adapt to the anatomical and physiological characteristics of different organs and achieve efficient nucleic acid delivery and controllable expression. At the same time, current mainstream repair strategies rely excessively on nucleic acid delivery, making them susceptible to restrictions from related core patents. Therefore, there is an urgent need in this field for an innovative delivery system that can overcome the aforementioned efficiency bottlenecks and achieve deep and safe multi-organ repair through multi-modal delivery (including but not limited to nucleic acids) within the NMP window, thereby enabling donor transplant repair, organ enhancement, and future in-situ repair and functional enhancement of the corresponding organs.

[0004] Based on this, the present invention is proposed. Summary of the Invention

[0005] The purpose of this invention is to provide a novel industrialized and engineered fusion protein and its mediated vesicle carrier, which solves the inherent defects of existing endogenous delivery systems through structural fusion design.

[0006] Another objective of this invention is to provide a standardized application method for this vector in the repair of various marginal donor organs, so as to achieve rapid recovery of organ function and significant rejuvenation at the epigenetic level.

[0007] Furthermore, this invention aims to provide nucleic acid-free repair strategies, including chemical small molecule cocktail therapy and direct protein transduction technology, to circumvent existing patent barriers and improve treatment safety.

[0008] To achieve the above-mentioned objectives, the present invention provides the following technical solution: This invention provides an engineered fusion protein comprising a functional fragment of a human ARRDC1 protein, a flexible linker peptide, and an RNA-binding domain; The amino acid sequence of the functional fragment of the human ARRDC1 protein is shown in SEQ ID NO: 1; The flexible linker peptide is (GGGGS)n, where n is an integer from 2 to 4; The amino acid sequence of the RNA-binding functional domain is shown in SEQ ID NO: 3.

[0009] The present invention also provides a fusion gene encoding the engineered fusion protein, the nucleotide sequence of which is shown in SEQ ID NO: 6.

[0010] The present invention also provides an active budding vesicle system mediated by the engineered fusion protein or the fusion gene, the active budding vesicle system comprising: vesicles formed by the engineered fusion protein or the fusion gene; a membrane fusion protein VSV-G located on the vesicle membrane; and drug molecules encapsulated within the vesicles.

[0011] Preferably, the drug molecule includes nucleic acid molecules, small chemical molecule compositions, or polypeptides; Preferably, the nucleic acid molecule includes mRNA, self-replicating mRNA, or RNA interference molecule; the 5' or 3' untranslated region of the nucleic acid molecule contains a packaging signal sequence.

[0012] More preferably, the mRNA is an mRNA encoding EZH2 and / or E2F3, and / or an RNA interference molecule targeting STAT3 and / or ZFX; the molar ratio of mRNA EZH2, mRNA E2F3, siRNA STAT3, and siRNA ZFX is (1-3):(1-3):1:1.

[0013] More preferably, the self-replicating mRNA encodes Oct4, Sox2, Klf4, Glis1, and Lin28A; the molar ratio of Oct4, Sox2, Klf4, Glis1, and Lin28 is (2-4):1:1:1:1.

[0014] Preferably, the chemical small molecule composition comprises an epigenetic regulator and a key signaling pathway inhibitor; the epigenetic regulator comprises a histone deacetylase (HDAC) inhibitor, a GSK-3β inhibitor, a TGF-β signaling pathway inhibitor, and a lysine-specific demethylase 1 (LSD1) inhibitor; the key signaling pathway inhibitor is a KAT acetyltransferase inhibitor module, comprising an inhibitor of histone acetyltransferase KAT3A / B and an inhibitor of histone acetyltransferase KAT6A.

[0015] Preferably, the polypeptide is a polypeptide of a regeneration factor or a reprogramming factor, and the polypeptide is fused with a cell-penetrating peptide.

[0016] Preferably, the packaging signal sequence is the 5'UTR of human PEG10 mRNA, with a nucleotide sequence as shown in SEQ ID NO:9, or a variant thereof having at least 80% sequence identity and capable of enriching the nucleic acid by ≥10-fold when co-expressed with the engineered fusion protein.

[0017] Preferably, the small molecule chemical composition includes valproic acid, CHIR99021, RepSox, transphenylcyclopropylamine, C646, and WM-1119.

[0018] Preferably, the engineered fusion protein is cloned into an expression vector and co-transfected into cells with an expression plasmid carrying the membrane fusion protein VSV-G and a reporter plasmid carrying the nucleic acid molecule to obtain the actively budding vesicles; or the expression vector cloning the engineered fusion protein is co-transfected into cells with an expression plasmid carrying the membrane fusion protein VSV-G and an empty reporter plasmid vector, and simultaneously co-incubated with a small chemical molecule composition or peptide to obtain the actively budding vesicles.

[0019] The present invention also provides the application of the engineered fusion protein, the fusion gene, or the active budding vesicle system described herein in the repair of ex vivo marginal donor organs.

[0020] The present invention also provides a method for repairing ex vivo marginal donor organs using the aforementioned active budding vesicle system, comprising the following steps: (1) Connect the ex vivo marginal donor organ to an ambient temperature mechanical perfusion system that matches its type; (2) Introducing an effective dose of an active budding vesicle system into a room-temperature mechanical perfusion system; (3) Maintain perfusion for 4-8 hours; before transplantation, replace and clean with fresh perfusion fluid at least 3 times the capacity of the perfusion system to obtain the repaired ex vivo organ.

[0021] Preferably, the ex vivo marginal donor organ is selected from the liver, kidney, heart, or lung.

[0022] Compared with the prior art, the present invention has the following beneficial effects: This invention provides an engineered fusion protein and its mediated active budding vesicle system, serving as a drug delivery platform. The delivery process involves engineering production cells to express the ARRDC1 gene, mediating the formation of budding vesicles. Simultaneously, the target payload (drug molecule) is directly loaded into the vesicles using PEG10, and can be collected and purified from the budding vesicles of the production cells. The target payload is efficiently and actively loaded into the small vesicles connected to ARRDC1 by the PEG10 endogenous protein, thus avoiding the use of Nipah virus or HIV viral elements to load proteins and RNA, as is common in the classic ARRDC1 pathway.

[0023] This delivery platform does not simply splice ARRDC1 and PEG10, but rather achieves increased vesicle yield and mRNA encapsulation efficiency through precise domain fusion and linker design. This dual, non-linear improvement in both yield and efficiency constitutes a synergistic effect. Furthermore, the fusion protein can be synthesized in one step, enabling industrial-scale production.

[0024] The RNA-binding domain (PEG10 NC domain) in the fusion protein provided by this invention has a consistent structural necessity across different payload delivery modes. When the delivery payload is mRNA, self-replicating mRNA, or RNA interference molecules, the CCHC-type zinc finger motif of the NC domain directly recognizes and binds to the PEG10 UTR packaging signal sequence at the 5' end of the target nucleic acid, driving the active enrichment of the nucleic acid payload within the vesicle; this function is indispensable. When the delivery payload is a small molecule composition or functional peptide, although small molecule and protein payloads do not rely on the NC domain for sequence-specific recognition, the NC domain, as an indivisible structural element of the fusion protein, indirectly ensures vesicle yield, particle size uniformity, and membrane structural integrity by maintaining the correct spatial folding of the ARRDC1 budding domain and its directional anchoring on the inner surface of the vesicle membrane. Removal of the linker peptide or disruption of the spatial integrity of the NC domain both lead to a significant decrease in vesicle yield and drug loading efficiency, demonstrating that this domain is essential for the overall function of the fusion protein in all three delivery modes, rather than serving only the single purpose of nucleic acid recognition. Therefore, the three-component structure design of the fusion protein provided by this invention has universal applicability to the efficient delivery of three payload types: nucleic acids, small molecules, and proteins.

[0025] This invention utilizes engineered fusion proteins and their mediated active budding vesicle system to deliver nucleic acid molecules, small molecule compositions, and peptides during in vitro, room-temperature perfusion. It can achieve deep repair of various marginal donor organs, such as fatty liver and liver from aged donor pigs, within 4-8 hours, demonstrating a reduction of >60% in steatosis, >50% in fibrosis, and a 58%-77% reversal of epigenetic age. Furthermore, the system is easily cleared and highly safe. In addition, this invention provides a non-nucleic acid repair strategy based on direct transduction of chemical small molecules and proteins, effectively circumventing existing patent barriers. This invention provides an efficient, safe, and universal platform technology for liver repair and multi-organ rejuvenation.

[0026] More importantly, the "self-reconstruction" technology path pioneered by this invention demonstrates profound application prospects and imagination. Based on the ex vivo perfusion repair platform of this invention, "active maintenance and renovation" of diseased or aging organs can be achieved in the future. For example, for patients with early-stage renal insufficiency, the kidney can be temporarily removed through minimally invasive surgery and placed in an integrated ex vivo ambient temperature mechanical perfusion system. Using the engineered vesicle system provided by this invention, specific regeneration factors, reprogrammed small molecule combinations, and nucleic acid regulatory cocktails such as shRNA or CRISPRi can be delivered to complete the "one-click repair and rejuvenation" of organ cells within hours. Subsequently, the functionally reconstructed kidney is re-implanted into the patient. This process uses the patient's own organ entirely, completely avoiding the immune rejection reaction in allogeneic transplantation, thus making it possible for human organs to achieve "regular maintenance and functional regeneration" like precision equipment, providing a revolutionary solution for the treatment of chronic organ degenerative diseases. Detailed Implementation

[0027] The technical solutions provided by the present invention will be described in detail below with reference to the embodiments, but they should not be construed as limiting the scope of protection of the present invention.

[0028] Example 1: Construction of engineered fusion proteins and verification of synergistic effects of fusion proteins

[0029] The fusion gene ARRDC1-PEG10NC encoding the engineered fusion protein was synthesized and cloned into the pcDNA3.1 vector to obtain the plasmid pAR-PEG10NC.

[0030] I. Fusion Gene Synthesis

[0031] 1. Design of fusion protein coding sequences

[0032] The fusion protein structure is: ARRDC1(1-420aa)-(GGGGS)3-PEG10NC (280-325aa).

[0033] ARRDC1 (1-420 aa) is the membrane budding domain; it is the functional fragment of amino acids 1-420 in the human ARRDC1 protein (NCBI Reference Sequence: NP_689498.1, full length 433 aa). The amino acid sequence is shown in SEQ ID NO: 1, which contains the PSAP motif and the PPXY motif (X = glutamate E / serine S, PPEY / PPSY). MGRVQLFEISLSHGRVVYSPGEPLAGTVRVRLGAPLPFRAIRVTCIGSCGVSNKANDTAWVVEEGYFNSSLSLADKGSLPAGEHSFPFQFLLPATAPTSFEGPFGK IVHQVRAAIHTPRFSKDHKCSLVFYILSPLNLNSIPDIEQPNVASATKKFSYKLVKTGSVVLTASTDLRGYVVGQALQLHADVENQSGKDTSPVVASLLQKVSYKA KRWIHDVRTIAEVEGAGVKAWRRAQWHEQILVPALPQSALPGCSLIHIDYYLQVSLKAPEATVTLPVFIGNIAVNHAPVSPRPGLGLPPGAPPLVVPSAPPQEEAEAEAAAGGPHFLDPVFLSTKSHSQRQPLLATLSSVPGAPEPCPQDGSPASHPLHPPLCISTGATVPYFAEGSGGPVPTTSTLILPPEYSSWGYPYEAPPSYEQ (SEQ ID NO: 1).

[0034] (GGGGS)3 is a flexible linker peptide with an amino acid sequence of (GGGGS)n. In this embodiment, n=3, i.e., GGGGSGGGGSGGGGS (SEQ ID NO: 2).

[0035] PEG10 (280-325aa) is an RNA-binding domain: it is a functional fragment of the human PEG10 protein nucleocapsid (NC) domain, with the amino acid sequence shown in SEQ ID NO: 3. It contains a CCHC-type zinc finger motif (CLYCGTGGHYADNC, 14aa, SEQ ID NO: 4) located at positions 295-308, which is essential for mRNA packaging function.

[0036] RLTQEEKERRRKLNLCLYCGTGGHYADNCPAKASKSSPAGKLPGPA (SEQ ID NO: 3).

[0037] Based on the above fusion protein structure and sequence, a fusion gene was synthesized. The sources of each coding sequence are as follows: The amino acid sequence of ARRDC1 (1-420aa) is encoded by NM_152285.4. Its encoding nucleotide sequence, after codon optimization for the human HEK293 cell expression system, is a component of the fusion gene synthesis fragment shown in SEQ ID NO:6 (corresponding to nucleotides 22-1278 of SEQ ID NO:6, encoding amino acids 2-420 of SEQ ID NO:1; the ATG of the first methionine is provided by the start codon at positions 19-21 of SEQ ID NO:6).

[0038] The coding sequence of the (GGGGS)3 flexible linker peptide, after codon optimization, is: GGTGGAGGCGGTTCAGGCGGAGGTGGCTCTGGCGGTGGCGGATCG (SEQ ID NO: 5).

[0039] PEG10 NC (280-325aa): Based on human PEG10 mRNA (NM_001172438.2), the nucleotide sequence encoding amino acids 280-325 (positions 838-975, 138bp) was taken.

[0040] The fusion gene (ARRDC1 functional fragment + linker peptide + PEG10 NC functional fragment) is 1443 bp in length and was synthesized by a gene synthesis company. The final synthesized fragment, including the two end attachments (protective bases, HindIII / XhoI restriction sites, Kozak sequence, start codon ATG, and stop codon TGA), is 1476 bp in length.

[0041] Codon optimization was performed concurrently during synthesis (specifically for the human HEK293 cell expression system), with a Kozak sequence (GCCACCATG) introduced at the 5' end and a stop codon (TGA) introduced at the 3' end. Restriction enzyme sites were designed at both ends of the fusion gene: 5' end: HindIII recognition sequence (AAGCTT); 3' end: XhoI recognition sequence (CTCGAG). During gene synthesis, 6 protective bases were added outside each restriction site, resulting in the final synthesized fragment structure shown in SEQ ID NO: 6.

[0042] 2. Cloning vector

[0043] Target vector: pcDNA3.1(+) (Invitrogen, catalog number V79020).

[0044] II. Construction of plasmid pAR-PEG10NC

[0045] 1. Required reagents and their catalog numbers: Table 1 Molecular cloning reagents

[0046] Table 2 PCR Validation Reagents

[0047] Table 3 Culture media and antibiotics

[0048] 2. Detailed cloning steps

[0049] Step 1: Double enzyme digestion

[0050] Simultaneously, the synthesized fusion gene fragment and pcDNA3.1(+) vector were double-digested. The mixture was then incubated in a 37°C water bath for 2 hours.

[0051] Table 4. Enzyme digestion system for fusion gene (50 μL)

[0052] Table 5. Enzyme digestion system of pcDNA3.1(+) vector (50 μL)

[0053] Step 2: Rubber cutting and recycling

[0054] The enzyme digestion products were separated by 1% agarose gel electrophoresis. The target bands (fusion gene approximately 1443 bp; linearized vector approximately 5400 bp) were cut off with a clean scalpel and recovered using the QIAquick Gel Extraction Kit according to the instructions. The final step was elution with 30 μL of EB buffer preheated to 65°C and quantified by NanoDrop.

[0055] Step 3: Connect

[0056] Table 6. Connecting system (20 μL)

[0057] Incubate overnight in a 16°C water bath (15 hours).

[0058] Step 4: Conversion

[0059] Add 10 μL of the ligation product to 50 μL of DH5α competent cells and proceed as follows: ice bath for 30 minutes; heat shock at 42°C for 90 seconds; ice bath for 2 minutes; add 900 μL of antibiotic-free LB medium; revive at 37°C with shaking at 200 rpm for 1 hour; spread 150 μL onto LB solid medium containing 100 μg / mL ampicillin; incubate overnight at 37°C with inverted incubation for 14 hours.

[0060] Step 5: Colony PCR screening

[0061] Single clones were picked and cultured in LB broth containing ampicillin for 3 hours. 1 μL of the bacterial culture was then used directly as a PCR template. Verification primers: F (within the CMV promoter): CGCAAATGGGCGGTAGGCGTG (SEQ ID NO: 7); R (within the PEG10 NC segment): TCCTGCCCTTGACCCTGCGTT (SEQ ID NO: 8). The expected amplified fragment size was approximately 1550 bp (positive clone).

[0062] Table 7 PCR Procedure

[0063] Step 6: Sequencing Verification

[0064] After overnight amplification of positive clones, plasmids were extracted using the QIAprep Miniprep Kit and sent for Sanger sequencing (primers were the same as those used for colony PCR, with the addition of T7 promoter primers to cover the full length).

[0065] Sequencing confirmed the following: (1) The full-length sequence of the fusion gene was correct and there were no base mutations; (2) The reading frame was continuous and correct; (3) The HindIII and XhoI restriction sites were intact; (4) The Kozak sequence (GCCACCATG) was present; (5) The stop codon (TGA) was present.

[0066] Step 7: Large-scale plasmid extraction

[0067] Logically sequenced clones were inoculated into 250 mL of LB liquid medium containing 100 μg / mL ampicillin and incubated overnight at 37°C with shaking at 200 rpm. Plasmids were extracted using the QIAfilter Plasmid Maxi Kit, dissolved in TE buffer, and quantified using NanoDrop (A260 / A280 should be between 1.8 and 2.0). The plasmids were then aliquoted and stored at -20°C.

[0068] The plasmid pAR-PEG10NC was finally obtained and used for subsequent HEK293FT cell transfection experiments.

[0069] III. Verification of the synergistic effect of fusion proteins

[0070] The fusion protein-mediated active budding vesicle system requires co-expression of plasmid pAR-PEG10NC with VSV-G and reporter mRNA with PEG10UTR.

[0071] Four experimental groups were set up to prepare vesicles under the same conditions: Group A (ARRDC1-Tat), Group B (full-length PEG10), Group C (the present invention, ARRDC1-PEG10NC), and Group D (linked peptide without linkages, GGGGS). n ,ARRDC1-direct fusion-PEG10NC.

[0072] 1. Co-expression strategy

[0073] A three-plasmid transient co-transfection system was used to simultaneously transfect HEK293FT cells with the following three plasmids to simulate the vesicle production environment: Table 8 Co-expression strategies

[0074] Among them, plasmid ②: VSV-G plasmid reference: Addgene #8454 (pMD2.G, Zhang lab).

[0075] Plasmid ③ was constructed as follows: 1) Reporter mRNA selection The reporter mRNA used was Cre recombinase mRNA, consistent with the original validation system of the Segel 2021 Science SEND system (functional readout was performed using the loxP-GFP reporter cell line, and the proportion of GFP-positive cells after Cre-mediated recombination was used as a quantitative indicator of delivery efficiency).

[0076] Report the complete structure of the plasmid: HsPEG10 5'UTR (343nt) → Cre coding sequence (1029bp) → first 500bp of HsPEG10 3'UTR → BGH polyA signal.

[0077] Reference for Cre recombinase coding sequence: Addgene #13775 (codon-optimized human version).

[0078] 2) Human PEG10 5'UTR sequence

[0079] The human PEG10 5'UTR sequence is derived from nucleotides 1-343 (343 bp) of NCBI Reference Sequence NM_001172438.2, representing the untranslated region upstream of the PEG10 mRNA transcription start site and the start codon ATG. This sequence serves as an mRNA packaging signal, specifically recognized by the PEG10 NC domain (SEQ ID NO:3, containing the CCHC-type zinc finger motif CLYCGTGGHYADNC, SEQ ID NO:4) in the engineered fusion protein of this invention, thereby driving the active enrichment of target mRNA carrying this 5'UTR within ARRDC1-mediated active budding vesicles. This invention fuses this sequence to the 5' end of a reporter mRNA (such as Cre recombinase mRNA) as a functional packaging signal for delivery mediated by the fusion protein shown in SEQ ID NO:6. Its complete nucleotide sequence is as follows (5' → 3'): acacgcgcttcaacttcggttggtgtgtgtcgaagaaacctgactgcgccctgaggagaacagcggagaaggtccaccgagcctggcgaaaggtccgctgagcgggctgtcgtccggagccactccgggctgcggagcacccagtggagaccgcgcctggctcaggtgtgg gaccccatccttcctgtcttcgcagaggagtcctcgcgtggtgagtatgcgaaataagcgggttttgaaaacaaaaaaagaaggagtggaagaggggggccaggatccaggcctccatccccacagaagtgaagctacagctgggaggtctcctcccacccccaaccgtcacc (SEQ IDNO:9).

[0080] 3) Human PEG10 3'UTR 500bp

[0081] The human PEG10 3'UTR originates from NM_001172438.2 (positions 2606-3105), representing nucleotides 1-500 after the stop codon in the coding region. It works in conjunction with the 5'UTR as an mRNA packaging signal. Segel 2021 confirmed that the proximal 500 bp is the smallest functional unit. (Note: The full length of the PEG10 3'UTR is approximately 4 kb. The original Segel 2021 study demonstrated through systematic truncation experiments that efficient functional mRNA transfer can be achieved with only the proximal 500 bp. Therefore, plasmid ③ in this invention uses the proximal 500 bp as the 3' packaging signal to simplify construction and preserve complete function.)

[0082] 4) Plasmid ③ Complete element map

[0083] Table 9. Graph of complete elements of plasmid ③

[0084] 5) Functional verification method

[0085] Reporting cell line: HEK293FT-loxP-GFP cells.

[0086] Readout method: After vesicles incubate target cells for 72 hours, the percentage of GFP-positive cells is detected by flow cytometry. After successful CremRNA delivery and translation, Cre recombinase removes the termination sequence between loxP sites, activating GFP expression. The GFP positivity rate represents the functional mRNA delivery efficiency.

[0087] 2. Transfection procedure

[0088] Table 10 Transfection Procedure

[0089] 3. Plasmid ① replacement schemes for each control group

[0090] The following groups (A, B, C, D) all used the same vesicle preparation process, with the only variable being the different constructs of plasmid ①.

[0091] Table 11 Alternatives for Plasmid ①

[0092] Note: Group A pARRDC1-Tat was constructed using existing technology of ARRMs system ARRDC1-Tat as described in Wang et al., Nature Communications, 2018, 9:960 (PMID: 29511190). This study, based on the original direct fusion structure of ARRDC1-Tat, inserted a (GGGGS)2 flexible linker peptide between ARRDC1 and Tat to maintain topological comparability with Group C (this invention). This modification was only used for structural alignment of the control group and is not within the scope of protection of this invention. Tat is derived from the HIV-1 viral protein and is a non-endogenous element, posing a safety risk—this is the core motivation for this invention to replace Tat with endogenous PEG10 NC. Vector backbone: pcDNA3.1(+); CMV promoter driven; AmpR screening.

[0093] The structure of the ARRDC1-Tat fusion protein is: ARRDC1(1-433aa)-(GGGGS)2-HIV-1 Tat(1-86aa). Vector backbone: pcDNA3.1(+), CMV promoter driven; selection marker: ampicillin resistance (AmpR).

[0094] 4. Vesicle collection and purification (general procedure)

[0095] Collect the culture supernatant and centrifuge at 300×g for 10 minutes to remove dead cells. Transfer the supernatant and centrifuge at 2,000×g for 20 minutes to remove cell debris. Filter through a 0.45μm sterile filter to remove large particulate impurities. Centrifuge at 100,000×g (SW41 Ti rotor) at 4°C for 90 minutes. Discard the supernatant, resuspend the precipitate in PBS, and centrifuge again at 100,000×g at 4°C for 60 minutes (washing step). Resuspend the final precipitate in 200μL PBS; this is the vesicle component, which can be aliquoted and stored at -80°C or used directly for detection.

[0096] 5. Determine the key performance characteristics of the four delivery platforms.

[0097] Key performance tests covered the following four indicators: (1) vesicle yield, (2) mRNA encapsulation efficiency, (3) number of guide RNA gene editor complexes (base editor ABE8e / sgRNA RNPs) per vesicle, and (4) number of CRISPR / guide RNA complexes (SpCas9 / sgRNA RNPs) per vesicle. Group A (ARRDC1-Tat existing technology control), Group B (full-length PEG10 SEND existing technology control), and Group C (ARRDC1-PEG10NC of this invention) were tested for all four indicators to evaluate the delivery performance advantage of Group C over existing ARMMs / SEND systems (results detailed in Table 14). Group D (peptide-free reverse control) focused on testing the first two basic functional indicators—given that Group D was significantly lower than Group C in terms of vesicle yield and mRNA encapsulation efficiency (see Table 15), its payload capacity was no longer comparable at the structural level, therefore, RNP loading was not further quantified.

[0098] (1) Vesicle production

[0099] 1) Testing instruments

[0100] Nanoparticle Tracking Analysis (NTA), Instrument: NanoSight NS300 (Malvern Panalytical).

[0101] 2) Experimental Procedure

[0102] Vesicle samples were serially diluted with PBS (1:100 to 1:10,000) to achieve a particle count of 20-100 particles / frame in the NTA detection field of view. Five 60-second video segments were acquired for each sample, with the camera sensitivity set to detect particles >50 nm in diameter. The software automatically tracked the Brownian motion trajectory of each particle, calculating the hydrodynamic diameter and particle concentration. The output was the number of particles per milliliter of sample (particles / mL), calculated as the mean ± standard deviation of the five video segments. Three biological replicates were set up for each group (n=3, i.e., three independent transfections and vesicle preparations), and the mean ± SD was reported.

[0103] 3) Calculation formula

[0104] NTA software automatically outputs particle concentration; the original calculation principle is as follows: Vesicle yield (particles / mL) = NTA traced particle number / detection volume (mL) × dilution factor.

[0105] The final report value is: Reported value = average particle concentration (particles / mL) × resuspension volume (mL) / original culture supernatant volume (mL).

[0106] (2) mRNA encapsulation efficiency

[0107] 1) Detection principle

[0108] Quantification was performed using an RNase protection assay combined with digital PCR (dPCR) and quantitative RT-PCR (qRT-PCR). Nuclease treatment degrades free mRNA outside vesicles, while the vesicle membrane protects the internal mRNA from degradation. The encapsulation efficiency was calculated by comparing the amount of mRNA before and after treatment.

[0109] 2) Required reagents

[0110] Table 12 Reagents, Product Numbers, and Uses

[0111] 3) Experimental Procedure

[0112] (a) Sample grouping and processing

[0113] Take the same batch of vesicle samples and process them in three tubes: Table 13 Different treatments

[0114] (b) RNA extraction and reverse transcription

[0115] Total RNA was extracted from each tube according to the TRIzol LS manufacturer's instructions. An equal volume of isopropanol was added to precipitate the RNA, followed by washing with 70% ethanol and dissolution in RNase-free water. The RNA concentration in each tube was quantified using NanoDrop, and 2 μL of the precipitate was simultaneously run on a Bioanalyzer (Agilent 2100) to detect RNA integrity (RIN value). An equal volume of RNA (100 ng) was taken from each tube and reverse transcribed using SuperScript IV VILOMaster Mix according to the manufacturer's instructions (50℃ for 10 minutes, then 85℃ for 5 minutes to terminate).

[0116] (c) qRT-PCR quantification

[0117] Each tube of cDNA was diluted 1:5 and used as a qPCR template. Cre mRNA quantification primers / probes: targeting the Cre coding region (avoiding the UTR to prevent detection of plasmid DNA), using the TaqMan dual-labeled probe system. Three technical replicates were set for each sample. A standard curve was plotted using in vitro transcripts of Cre mRNA with known copy numbers, and Ct values ​​were converted to mRNA copy number / μL. Simultaneously, the housekeeping gene (GAPDH) was used to detect RNA extraction efficiency in each tube, and internal control calibration was performed (only used for tube 1; absolute quantification was the primary method for tubes 2 and 3).

[0118] (d) Encapsulation ratio calculation formula

[0119] mRNA encapsulation rate (%) = [mRNA copy number in vesicles (tube 2) / total mRNA copy number (tube 1)] × 100%.

[0120] Note: The detection value of tube 3 should be less than 2% of that of tube 1; otherwise, it indicates that the RNase digestion is incomplete and the experiment needs to be repeated.

[0121] (3) Number of guide RNA gene editor complexes contained in each vesicle

[0122] 1) Definition of the detection object

[0123] "CRISPR / guide RNA complex" refers to the ribonucleoprotein complex (RNP) formed by the pre-assembly of SpCas9 protein and single-stranded guide RNA (sgRNA). The target of detection is the average number of RNPs loaded per vesicle.

[0124] 2) Detection strategy: Two-step method (protein quantification + particle quantification)

[0125] The average number of RNP complexes carried by each vesicle was calculated by combining absolute protein quantification (SpCas9 ELISA) and absolute vesicle particle quantification (NTA).

[0126] 3) Sample preparation

[0127] Based on the three-plasmid co-transfection system, a fourth plasmid, pU6-sgRNA-Kras (expressing a single-stranded guide RNA targeting MmKras), was additionally co-transfected. This allowed the sgRNA to spontaneously assemble with the SpCas9 protein, generated from the translation of overexpressed SpCas9 mRNA, into an RNP within the production cell, which was then packaged into vesicles. (Note: SpCas9 mRNA is expressed with the HsPEG10 UTR packaging signal at the 5' end, i.e., Hs.cargo(SpCas9), ensuring efficient loading into vesicles.)

[0128] 4) SpCas9 protein quantification (intravesicles)

[0129] (a) Vesicle lysis

[0130] Take a vesicle sample with a quantified particle count (pre-quantified by NTA), add RIPA lysis buffer (containing protease inhibitor), lyse on ice for 30 minutes, and sonicate (20% power, 3 × 5 sec pulses). Centrifuge at 13,000 × g for 15 minutes, and use the supernatant for protein quantification.

[0131] (b) SpCas9 ELISA quantification

[0132] A commercially available SpCas9 ELISA kit (Takara Bio #632663, detection range 0.1-20 ng / mL) was used. A standard curve was plotted using recombinant SpCas9 protein at a known concentration (purified Addgene #62988). The vesicle lysis supernatant was diluted to the linear range for detection to obtain the SpCas9 mass (ng). The molecular weight of SpCas9 is approximately 160 kDa (160,000 Da).

[0133] SpCas9 number = SpCas9 mass (g) / molecular weight (g / mol) × Avogadro's constant (6.022 × 10²³).

[0134] (c) Calculation of the number of RNPs per vesicle

[0135] RNP count per vesicle = Total number of SpCas9 molecules / Total number of vesicle particles determined by NTA; Note: This calculation assumes that each RNP contains one SpCas9 molecule, and the sgRNA quantity in the RNP is verified by Northern Blot or ddPCR to be approximately 1:1 in stoichiometry with SpCas9.

[0136] (4) Number of CRISPR / guide RNA complexes contained in each vesicle

[0137] 1) Definition of the detection object

[0138] "Guide RNA gene editor complex" refers to a complex formed by the pre-assembly of a base editor (BE) protein and sgRNA. The base editor used in this patent is ABE8e (adenine base editor version 8e, fused with nCas9-adenosine deaminase), with a molecular weight of approximately 166 kDa.

[0139] 2) Detection Strategy

[0140] Similar to the CRISPR / RNP detection strategy, a base editor-specific ELISA (or Western Blot quantification) combined with NTA is used for joint calculation.

[0141] 3) Sample preparation

[0142] pCargo(ABE8e)-PEG10UTR (containing the ABE8e coding sequence and the HsPEG10 UTR packaging signal at the 5' end) and pU6-sgRNA were co-transfected into production cells, and vesicles were collected after 48 hours. ABE8e reference plasmid: Addgene #112095 (David Liu lab).

[0143] 4) ABE8e protein quantification (Western Blot absolute quantification method)

[0144] (a) SDS-PAGE and Western Blot

[0145] Vesicle samples (with known particle count determined by NTA) were added to 4× Loading Buffer, denatured at 95°C for 10 min, and separated by 12% SDS-PAGE electrophoresis. PVDF membranes were transferred (100V, 90 min) and blocked with 5% skim milk TBST for 1 h. Primary antibody: anti-Cas9 antibody (recognizing the nCas9 domain, Abcam ab204448, 1:2000), incubated overnight at 4°C. Secondary antibody: HRP-labeled goat anti-rabbit IgG (1:10000), incubated at room temperature for 1 h. Chemiluminescence imaging was performed using ECL, and images were acquired using a ChemiDoc imaging system.

[0146] (b) Absolute quantitative standard curve

[0147] Simultaneously, known amounts of recombinant nCas9 protein (0.1 ng, 0.5 ng, 1 ng, 5 ng, 10 ng) were loaded into the gel as standards. The grayscale values ​​of the bands were quantified using ImageJ or ImageLab software, and a standard curve (within the linear range) was plotted. The absolute mass of ABE8e protein in the vesicle lysis buffer was calculated based on the standard curve.

[0148] (c) Calculation of the number of base editor complexes per vesicle

[0149] Number of BE complexes per vesicle = Total number of ABE8e molecules / Total number of vesicle particles determined by NTA; Number of ABE8e molecules = Mass of ABE8e (g) / 166,000 (g / mol) × 6.022 × 10²³.

[0150] 6. Statistical Analysis Methods

[0151] Statistical software: GraphPad Prism 9.0 or R 4.2.0. All p-values ​​were Bonferroni corrected. All experiments were set up with n=3 biological replicates (3 independent transfections and vesicle preparations), and results are expressed as mean ± standard deviation (Mean ± SD).

[0152] Table 14 Comparison of Key Data for Delivery Platforms in Groups A, B, and C

[0153] It is known that the vesicle yield and mRNA encapsulation efficiency of the present invention (ARRDC1-PEG10NC) are significantly improved compared with existing delivery platforms, and each vesicle can load up to 61 base editor complexes with guide gRNA or 82 class II CRISPR / gRNA complexes, which is about 4 times that of group A (ARRDC1-Tat) and 10 times that of group B (full-length PEG10).

[0154] Table 15 Comparison of key data for delivery platforms in groups C and D

[0155] Compared with group C (this invention), group D showed a 47.6% decrease in vesicle production and a 54.2% decrease in mRNA encapsulation efficiency, both of which were significantly lower than those in group C (p < 0.001). Transmission electron microscopy revealed a significant decrease in vesicle morphological uniformity in group D, with an increased proportion of abnormal aggregates (approximately 23%), suggesting that the direct fusion of the linkerless peptides resulted in steric hindrance between the ARRDC1 budding domain and the PEG10 NC domain, leading to functional interference between the two.

[0156] Given the significant decrease in vesicle yield and mRNA encapsulation efficiency, as well as impaired vesicle morphology uniformity in group D, this study concludes that group D vesicles are no longer practically feasible for further loading SpCas9 / sgRNA RNPs or ABE8e base editor complexes. Even if RNP loading quantification were attempted, the loading per vesicle would inevitably be significantly lower than 50% of the corresponding value in group C (due to the combined limitation of decreased vesicle yield and encapsulation efficiency), and would not provide additional information for demonstrating the necessity of the linker peptides in this invention. Based on the above judgment, this study did not perform the latter two RNP loading quantification tests on group D, and instead focused detection resources on groups A, B, and C (i.e., a horizontal comparison between prior art controls and this invention, see Table 14 for details).

[0157] The above results indicate that the flexible linker peptide (GGGGS)3 is not a simple structural spacer, but a key design element enabling the synergistic function of the ARRDC1 budding domain and the PEG10 NC domain. Its mechanism of action is as follows: (a) it provides a spatial spacing of approximately 15-20 Å between the two domains, allowing them to fold independently; (b) it provides conformational flexibility, ensuring that the recognition between the CCHC-type zinc finger motif (SEQ ID NO:4) of the NC domain and the 5'UTR packaging signal is not spatially constrained by the ARRDC1 host structure; and (c) it maintains directional anchoring on the inner surface of the vesicle membrane, ensuring vesicle yield and size uniformity. In summary, the flexible linker peptide (GGGGS)... n (n=2~4) is one of the necessary conditions for the present invention to achieve unexpected technical effects, and it is also a key innovative element that distinguishes the present invention from the direct fusion design of unlinked peptides (Group D control).

[0158] Example 2: Repair of ex vivo marginal donor livers – Combined intervention with EZH2 / E2F3 / STAT3 / ZFX

[0159] I. Selection and Modeling Methods for Edge Donor Models

[0160] Rat fatty liver model (marginal donors of steatosis): Male SD rats, 8 weeks old, weighing 200-220 g (purchased from Vital River Pharmaceuticals, Beijing), were induced by a high-fat diet (HFD), with 60% kcal derived from fat (Research Diets D12492), and fed continuously for 12 weeks. Successful model establishment criteria: hepatic steatosis area >30% (confirmed by Oil Red O staining); serum ALT >80 U / L; liver / body weight ratio >5%. The rats were anesthetized by intraperitoneal injection of sodium pentobarbital (50 mg / kg), and after irrigation with cold UW solution (University of Wisconsin solution) via portal vein catheterization, the liver was harvested in situ.

[0161] II. AR-PEG vesicle preparation process (completed in vitro before in vitro injection)

[0162] Table 16 Production Cell Culture

[0163] Table 17 Three-plasmid co-transfection (in vitro, performed in culture dishes)

[0164] Transfection reagent: PEI MAX (Polysciences 24765-1), PEI:DNA = 3:1 (w / w); 6 hours after transfection, change the medium to serum-free DMEM (to reduce the interference of serum proteins on vesicle purification); collect the culture supernatant at 48 hours and 72 hours after transfection, and process them together.

[0165] Vesicle collection and purification: Centrifuge at 300×g for 10 minutes to remove dead cells and large debris. Centrifuge at 2,000×g for 20 minutes to remove cell debris. Filter through a 0.45 μm sterile membrane to remove large particulate impurities. Centrifuge at 100,000×g (SW41 Ti rotor) at 4°C for 90 minutes, discarding the supernatant. Resuspend the precipitate in PBS and centrifuge again at 100,000×g at 4°C for 60 minutes (wash). Resuspend the final precipitate in perfusion buffer (Williams E medium) and quantify the particle concentration using NTA. Adjust to the target concentration based on the NTA results, and aliquot for immediate use (prepare fresh and store at 4°C for no more than 24 hours).

[0166] Table 18 Vesicle Quality Control and Release Standards

[0167] The above plasmid ③ contains a nucleic acid combination: - mRNA EZH2 - mRNA E2F3 - siRNA STAT3 - siRNA ZFX in a molar ratio of 2:2:1:1, all packaged together in the same vesicle. Nucleic acid information and construction methods are as follows: Table 19 Sequence information of EZH2 and E2F3 nucleic acid payloads in plasmid ③

[0168] Table 20 Sequence information of STAT3 and ZFX nucleic acid payloads in plasmid ③

[0169] Table 21 Design of complete elements for plasmid ③ (pCargo-four factor)

[0170] Note: The mRNA component (EZH2+E2F3) and the siRNA component (siSTAT3+siZFX) are expressed in the same plasmid. The mRNA is actively loaded via PEG10 UTR recognition, while the siRNA is passively encapsulated into the vesicle lumen in the production cell via hairpin transcription. The molar ratio of 2:2:1:1 was achieved through optimization of promoter strength and copy number for each expression cassette.

[0171] III. Ex vivo room temperature mechanical perfusion (NMP)

[0172] NMP parameters: Dual-cycle perfusion, 37℃ (controlled by constant temperature water bath), hepatic artery pressure 5-10 mmHg (controlled by peristaltic pump), portal vein pressure 3-5 mmHg (assisted by gravity perfusion), 95% O2 / 5% CO2 mixed gas bubble oxygenation. Perfusion solution: Williams E medium + 20% bovine serum albumin + insulin (0.1 U / mL) + dexamethasone (2 μg / mL). 200 mL / cycle, continuous monitoring with an ultrasonic flow meter, hepatic artery flow maintained at 4 mL / min.

[0173] The excised liver was connected to the NMP system, perfusion was initiated, and stabilized for 30 minutes (confirming the start of bile secretion, normal perfusion fluid color, and no significant leakage). After stabilization, AR-PEG vesicle suspension was slowly injected via a portal vein cannula: dose 5 × 10¹² particles / liver, injection volume 10 mL, injection rate approximately 1 mL / min. Perfusion continued after injection, with a total perfusion time of 6 hours (starting from the injection start time). Perfusion fluid samples (2 mL) were collected every 2 hours during perfusion to measure functional indicators such as ALT, LDH, glucose, and lactate, as well as to continuously collect and measure bile. After 6 hours of perfusion, the liver was replaced and rinsed with at least 3 times the volume of fresh, pre-cooled perfusion fluid (3 times, 200 mL each time). Liver biopsies were taken (100 mg from each lobe), flash-frozen in liquid nitrogen, and used for subsequent molecular biological and histological analysis. The results are shown in the table below; the detection methods for each data point are referenced in Table 46.

[0174] Table 22 Core data on liver repair (based on combined intervention of transcription factors)

[0175] Example 3 Repair of ex vivo aged donor pig liver—based on the OSKGL combinatorial model of self-replicating mRNA

[0176] I. Ex vivo model of aged donor pig liver

[0177] Liver from elderly donor pigs (aged > 5 years, corresponding to human > 70 years) was immediately connected to an ex vivo perfusion system after acquisition.

[0178] II. AR-PEG Vesicle Preparation Procedure (Based on Self-Replicating mRNA OSKGL)

[0179] OSKGL consists of five factors: Oct4 (POU5F1), Sox2, Klf4, Glis1, and Lin28 (Lin28A). Oct4 is a core factor for maintaining pluripotency, activating target genes such as Sox2 and Nanog to initiate epigenetic reprogramming. Sox2 and Oct4 synergistically bind enhancers to maintain pluripotency and promote chromatin opening. Klf4 is a zinc finger transcription factor that inhibits the p53 / p21 pathway and synergistically activates the pluripotency gene network with Oct4 / Sox2. Glis1 is a key cofactor for enhancing reprogramming efficiency, activating pluripotency genes while inhibiting aging pathways, significantly improving OSK reprogramming efficiency. Lin28 (Lin28A) is an RNA-binding protein that inhibits let-7 miRNA maturation, upregulates pluripotency-related mRNAs, enhances reprogramming efficiency, and improves mitochondrial function.

[0180] All five genes used porcine (Sus scrofa) sequences to match the porcine liver model. Human reference numbers were also provided for homologous design. The self-replicating mRNA backbone used either Venezuelan equine encephalitis virus (VEEV) or Sindbis virus alphavirus replication subsystems.

[0181] Table 23 Five-Factor Gene Sequence Information

[0182] Synthesis instructions: In this embodiment, all five reprogramming factors underwent codon optimization using porcine (Sus scrofa) CDS as templates (for porcine hepatocyte expression systems) to match downstream porcine liver ex vivo perfusion models. Each factor was processed as follows: Oct4 (POU5F1): Take the pig-derived NM_001113060.1 CDS (1083 bp, encoding 360 aa), optimize the codons, clone it into the VEEV (Venezuelan equine encephalitis virus) or Sindbis (Sindbis virus) alphavirus replicon backbone, and link the 5′ end to the HsPEG10 5′UTR packaging signal (SEQ ID NO:9); Sox2: The homology between porcine (NM_001123197.1) and human (NM_003106.4) SOX2 proteins is >97%. The codon-optimized version of the human NM_003106.4 CDS can be directly used. It is also cloned into the VEEV / Sindbis replicon backbone, and the signal is packaged with HsPEG10 5′UTR at the 5′ end. Klf4: Take the pig source NM_001031782.1 CDS (1440 bp, encoding 479 aa), optimize the codons, and clone it into the VEEV / Sindbis replicon backbone; Glis1: A porcine XM_021096612.1 CDS (1866 bp, encoding 621 aa) was used, codon optimized, and cloned into the VEEV / Sindbis replicon backbone. Although Glis1 was added to the OSKGL combination at the same molar ratio as OSK (molar ratio 1), its synergistic effect on reprogramming aged cells was particularly crucial. It can activate multiple pro-reprogramming pathways such as N-Myc, Mycl1, and Lin28, without introducing c-Myc-related tumorigenic risk. Lin28A: Take the pig-derived NM_001123133.1 CDS (630 bp, encoding 209 aa), optimize the codons, and clone it into the VEEV / Sindbis replicon backbone.

[0183] The five genes were each prepared as five independent self-replicating mRNAs, which were then mixed and loaded into AR-PEG vesicles in a molar ratio of 3:1:1:1:1 (Oct4:Sox2:Klf4:Glis1:Lin28A). The preparation procedures for other AR-PEG vesicles were performed as described in Example 2.

[0184] III. Ex vivo room temperature mechanical perfusion (NMP)

[0185] NMP parameters: dual circulation perfusion, 37°C, hepatic artery pressure 8-12 mmHg, portal vein pressure 3-5 mmHg.

[0186] Intervention: After perfusion stabilization, AR-PEG vesicles loaded with a self-replicating mRNA combination were injected: self-replicating mRNA Oct4 - self-replicating mRNA Sox2 - self-replicating mRNA Klf4 - self-replicating mRNA Glis1 - self-replicating mRNA Lin28 (molar ratio 3:1:1:1:1). Dosage: 1 × 10¹³ vesicles / liver, administered over 6 hours. The results are shown in the table below. All indicators in the AR-PEG treatment group were significantly better than those in the model group. Data detection methods are detailed in Table 46.

[0187] Table 24 Core Data on Liver Repair from Elderly Porcine Donors

[0188] Example 4 Repair of ex vivo marginal donor kidneys—based on Sox9, Hnf4a

[0189] I. Ex vivo marginal donor kidney model

[0190] This embodiment uses an aged rat kidney model as the experimental group.

[0191] Table 25. Rat model of aged kidney

[0192] II. AR-PEG Vesicle Preparation Process (Based on Sox9, Hnf4a)

[0193] Table 26 Kidney repair factor gene sequences (Sox9, Hnf4a)

[0194] Synthesis instructions: Sox9: The complete CDS (1686 bp, encoding 561 aa) of rat NM_080403.1 was obtained, and the codons were optimized for rat renal tubular cells. The HsPEG10 5'UTR packaging signal (SEQ ID NO:9) was added to the 5' end, and the mRNA was transcribed in vitro.

[0195] Hnf4a: The complete CDS (1374 bp, encoding 457 aa, isoform 1) of rat NM_022180.2 was obtained, and the codons were optimized for rat renal tubular cells. The HsPEG10 5'UTR packaging signal (SEQ ID NO:9) was added to the 5' end, and the RNA was transcribed into mRNA in vitro.

[0196] Sox9 mRNA and Hnf4a mRNA were prepared independently (each carrying an HsPEG10 5'UTR packaging signal), mixed in an equimolar ratio (1:1), and co-loaded into AR-PEG vesicles. The remaining vector construction and vesicle preparation procedures were performed according to the original protocols of Examples 2 and 4.

[0197] Table 27 Plasmid ③ Design (Kidney Repair Version)

[0198] III. Ex vivo room temperature mechanical perfusion (NMP)

[0199] NMP parameters: single artery perfusion, 32°C (subnormal temperature), pressure 80-100 mmHg. Intervention: perfusion of AR-PEG vesicles for 6 hours. Results are shown in the table below. All indicators in the AR-PEG treatment group were significantly better than those in the model group. The detection methods for each indicator are shown in Table 46.

[0200] Table 28 Core Data on Liver Repair from Elderly Porcine Donors

[0201] Example 5 Repair of ex vivo marginal donor hearts—based on Gata4, Tbx5

[0202] I. Ex vivo marginal donor kidney model

[0203] Table 29. DCD Model of Cardiac Arrest

[0204] II. AR-PEG Vesicle Preparation Process (Based on Gata4, Tbx5)

[0205] Table 30. Gene sequences of cardiac repair factors

[0206] Table 31 Plasmid ③ Design (Heart Repair Version)

[0207] Other AR-PEG vesicle preparation procedures were performed according to Example 2.

[0208] III. Ex vivo room temperature mechanical perfusion (NMP)

[0209] NMP parameters: Langendorff perfusion (self-prepared, containing NaCl 118.5 mM, KCl 4.7 mM, CaCl2 1.4 mM, MgSO4 1.2 mM, NaHCO3 25 mM, KH2PO4 12 mM, glucose 11 mM), 37℃, perfusion pressure 60-80 mmHg (constant pressure perfusion), 95% O2 / 5% CO2 oxygenation. Intervention: 15 minutes after perfusion stabilization (T=0), AR-PEG vesicle suspension (5×10¹² particles / heart, injection volume = 2 mL, rate 0.5 mL / min) was injected via the coronary artery for 4 hours. The results are shown in the table below. The AR-PEG treatment group showed significantly better results than the model group. The detection methods for each indicator are shown in Table 46.

[0210] Table 32 Core Data on Cardiac Repair

[0211] Note: The baseline for left ventricular developmental pressure (LVDP) in the normal control group was 60±8 mmHg, while the baseline for the model group and the AR-PEG treatment group was 25±8 mmHg.

[0212] Example 6 Repair of ex vivo marginal donor lungs – based on Klf2

[0213] I. Ex vivo marginal donor lung model

[0214] Table 33 Acute pulmonary edema / injury model (LPS induction)

[0215] II. AR-PEG Vesicle Preparation Process (Based on Klf2)

[0216] Table 34 Lung repair factor Klf2 gene sequence information

[0217] Table 35 Plasmid ③ Design (Lung Repair Version)

[0218] Other AR-PEG vesicle preparation procedures were performed according to Example 2.

[0219] III. Ex vivo room temperature mechanical perfusion (NMP)

[0220] NMP parameters: EVLP perfusion system (XVIVO Perfusion System), perfusion fluid: Steen solution (XVIVOPERFUSION), pulmonary artery cannulation (perfusion fluid inlet), left atrial cannulation (perfusion fluid outlet), endotracheal intubation (ventilation). Temperature: 37℃, perfusion pressure: 15-20 mmHg, tidal volume: 7 mL / kg, respiratory rate (RR): 7 breaths / min, 95% O2 / 5% CO2. Intervention: After perfusion stabilization for 20 minutes (T=0), AR-PEG vesicles (5×10¹² particles / lung pair, injected via pulmonary artery at a rate of 1 mL / min) were administered for 5 hours. Results are shown in the table below. The AR-PEG treatment group showed significantly better results than the model group. The detection methods for various indicators are shown in Table 46.

[0221] Table 36 Core Data on Lung and Heart Repair

[0222] Note: The reference range for the oxygenation index PaO2 / FiO2 is 350-450 mmHg. The reference value for the lung W / D ratio in normal rats is 4.0-4.5 (without edema).

[0223] Example 7: Ex vivo liver repair – based on a small molecule chemical cocktail (VC6TF+KATi)

[0224] Chemical cocktail formulation: contains VPA (2 mM), CHIR99021 (3 μM), RepSox (0.5 μM), transphenylcyclopropane (10 μM), C646 (5 μM) and WM-1119 (1 μM).

[0225] Method A (direct perfusion): The above-mentioned small molecule composition is directly dissolved in liver NMP perfusion fluid.

[0226] Method B (vesicle delivery): The small molecule composition is co-incubated with cells that produce AR-PEG vesicles (refer to Example 2) to encapsulate the small molecule in the secreted vesicles. The drug-loaded vesicles are then collected, purified, and obtained.

[0227] NMP repair: Infuse the isolated fatty liver with drug-containing perfusion fluid or drug-loaded vesicles and perfuse in a double circulation manner at 37°C for 6 hours.

[0228] Results: Compared with the model group, both treatment methods significantly improved liver condition. The area of ​​steatosis decreased by >55%, the area of ​​fibrosis decreased by >48%, and the levels of inflammatory factors decreased significantly. Epigenetic age analysis showed a trend towards younger age of onset. The vesicle delivery group showed more uniform distribution and slightly higher efficiency in specific liver lobes.

[0229] Example 8: Ex vivo liver repair – based on direct protein transduction (OSKGL-TAT fusion protein)

[0230] I. Ex vivo marginal donor lung model

[0231] Liver of aged mice (24 months old).

[0232] II. AR-PEG Vesicle Preparation Process (Based on Klf2)

[0233] Protein loading: OSKGL (Oct4, Sox2, Klf4, Glis1, Lin28) fusion proteins (5 independent proteins, each fused with the TAT cell-penetrating peptide YGRKKRRQRRR, SEQ ID NO: 16) carrying the TAT cell-penetrating peptide were prepared by recombinant expression. The 5 independent proteins were quantified separately and then mixed in a molar ratio of 3:1:1:1:1 (Oct4:Sox2:Klf4:Glis1:Lin28), consistent with the molar ratio of the self-replicating mRNA in Example 3, for easy comparison of effects.

[0234] Table 37 Structural Design of Each TAT Fusion Protein

[0235] Expression and purification of recombinant OSKGL-TAT protein: 1. The CDS of each transcription factor (Oct4 / Sox2 / Klf4 / Glis1 / Lin28A, human sequence, codon optimized to E. coli preferred codon) were cloned into the pET-28a(+) multiple cloning site (NcoI / XhoI).

[0236] 2. Insert the GGGGS flexible linker peptide and the TAT sequence (TACGGCAAGAAGCGCCGCCAGCGCCGCCGG (SEQ ID NO: 17), encoding YGRKKRRQRRR) at the 3' end (before the XhoI site) of each CDS.

[0237] 3. The pET-28a(+) vector carries an N-terminal His6 tag (MGSSHHHHHHSSGLVPRGSH, SEQ ID NO: 18) for affinity purification.

[0238] 4. After sequencing to confirm the correct reading frame, transform BL21(DE3) competent cells and screen positive clones with 50 μg / mL kanamycin.

[0239] 5. Pick a single colony and inoculate it into 5 mL of LB / kanamycin medium, then incubate overnight at 37°C with shaking (seed culture).

[0240] 6. Inoculate the seed culture at a ratio of 1% into 500 mL of TB / kanamycin medium and incubate at 37°C and 200 rpm until OD600 = 0.7.

[0241] 7. Add IPTG to a final concentration of 0.3 mM, lower the culture temperature to 16°C, and continue to culture with shaking for 17 hours (overnight induction of expression).

[0242] 8.4,000×g, centrifuged at 4℃ for 20 minutes to collect the bacterial cells, discarding the supernatant to obtain the bacterial cells.

[0243] 9. The bacterial cells were resuspended in lysis buffer (50 mM NaH2PO4, 300 mM NaCl, 10 mM imidazole, pH 8.0, containing 1 mM PMSF and a protease inhibitor Cocktail), 4 mL / g wet cells.

[0244] 10. Sonicate disruption (300W power, 3 seconds disruption / 5 seconds interval, 50 cycles, ice bath) until the lysate is clear.

[0245] Centrifuge at 20,000×g at 4℃ for 30 minutes and collect the supernatant (soluble protein fraction).

[0246] 12. Pass the lysate supernatant through a Ni-NTA resin column (5 mL, Qiagen) pre-equilibrated with equilibration buffer (50 mM NaH2PO4, 300 mM NaCl, 10 mM imidazole, pH 8.0) at a flow rate of 1 mL / min at 4°C.

[0247] 13. Wash with 20 column volumes of wash buffer (50 mM NaH2PO4, 300 mM NaCl, 20 mM imidazole, pH 8.0) to remove non-specific binding proteins.

[0248] 14. Elute the target protein with elution buffer (50 mM NaH2PO4, 300 mM NaCl, 250 mM imidazole, pH 8.0) and collect the peak elution fraction (detected by UV at 280 nm).

[0249] 15. Verify purification purity using SDS-PAGE (target >90%), and detect by Coomassie brilliant blue staining or silver staining.

[0250] 16. Add the eluted protein to a dialysis bag (molecular weight cutoff 10 kDa), and dialyze against PBS (pH 7.4) at 4°C. Change the solution 3 times (500 mL each time, 4 hours apart).

[0251] 17. After dialysis, the protein concentration was measured using the BCA protein quantification kit (Thermo Fisher).

[0252] 18. Concentrate each protein to 1-5 mg / mL (using an Amicon Ultra 10 kDa ultrafiltration tube, centrifuge at 3,000×g), aliquot into 100 μL / tube, and store at -80°C (avoid repeated freeze-thaw cycles).

[0253] 19. After quantifying the BCA of each protein, convert the molar concentration (moles = mass / molecular weight, Oct4-TAT molecular weight 44kDa, Sox2-TAT molecular weight 38kDa, Klf4-TAT molecular weight 56kDa, Glis1-TAT molecular weight 72kDa, Lin28-TAT molecular weight 26kDa) and mix them according to the molar ratio Oct4:Sox2:Klf4:Glis1:Lin28=3:1:1:1:1.

[0254] Example: To prepare a mixture with a total protein concentration of 100 μg / mL (total volume 1 mL), use: Oct4 40.8 μg + Sox2 12.5 μg + Klf4 18.3 μg + Glis1 23.6 μg + Lin28 8.5 μg = 103.7 μg (adjust each component to 100 μg). Store at 4°C after mixing and use within 24 hours.

[0255] Incubation steps for protein-loaded vesicles: Step 1: Preparation of standard AR-PEG vesicles (same as Examples 1-6, without small molecules) 1. HEK293FT cells were co-transfected with three plasmids (pAR-PEG10NC / VSV-G / pCargo-empty vector, i.e., blank vesicles without the target mRNA) according to the standard protocol. The vesicles were collected and purified by standard ultracentrifugation.

[0256] 2. After quantitative purification of TA, the vesicle concentration was adjusted to 1×10⁻⁶. 13 Granules / mL, store at 4℃, use on the same day.

[0257] 3. Mix the five OSKGL-TAT fusion proteins in a molar ratio of 3:1:1:1:1 to prepare a mixed protein solution (total protein concentration 100 μg / mL, dissolved in PBS pH 7.4).

[0258] 4. Mix the protein mixture and the vesicle suspension at a volume ratio of 1:1, i.e.: protein mixture: 500 μL (containing 50 μg total protein), AR-PEG vesicle suspension: 500 μL (containing 5 × 10⁻⁶ g total protein). 12 Particles), total volume: 1 mL, final protein concentration after mixing: 50 μg / mL, final vesicle concentration: 5 × 10⁻⁶. 12 Particles / mL.

[0259] 5. Transfer 1 mL of the mixture into a 1.5 mL Eppendorf tube and incubate at 37°C for 4 hours using a rotary mixer (10 rpm).

[0260] 6. During incubation, TAT transmembrane peptides bind to vesicle membrane phospholipids through electrostatic interactions, driving protein insertion or transmembrane penetration, thus achieving physical binding between the protein and vesicles.

[0261] 7. After incubation, centrifuge at 100,000×g at 4℃ for 60 minutes to precipitate the drug-loaded vesicles containing protein. Discard the supernatant (containing free protein).

[0262] 8. Resuspend the precipitate in 1 mL PBS, centrifuge again at 100,000×g for 30 minutes, and wash. Discard the supernatant.

[0263] 9. The final precipitate was resuspended in 100 μL of perfusion medium (Williams E medium, same as in Example 2) and NTA was quantified.

[0264] 10. Take 5 μL of the lysate and perform SDS-PAGE+Western Blot. Use anti-His antibody to confirm that each TAT protein has been successfully loaded into the vesicle.

[0265] III. Ex vivo room temperature mechanical perfusion (NMP)

[0266] Table 38. Vesicle Addition Amount and NMP Dosing Regimen

[0267] Results: Immunofluorescence confirmed successful delivery of OSKGL-TAT protein into the hepatocyte nucleus. After treatment, liver ATP levels increased by >90%, aging-associated secretory phenotype (SASP) markers significantly decreased, and epigenetic clock analysis showed an age reversal of approximately 40-50%.

[0268] Table 39 SASP Detection Indicators

[0269] Example 9 Security and Cleanup Verification

[0270] All organs were treated with the AR-PEG system (regardless of whether they were loaded with nucleic acids, small molecules, or proteins) and then replaced with three times the volume of perfusion fluid before transplantation. After washing, three types of safety indicators were tested: (1) residual VSV-G protein in the organ tissue; (2) residual active ingredients in the small molecule treatment group; and (3) the level of the inflammatory factor IL-6 in the perfusion fluid (compared with the AAV control group).

[0271] Table 40 Experimental Group Settings

[0272] The table now includes the specific drug loading and operating procedures for the newly added parallel control AAV group: AAV control group setup: This was performed in parallel with AR-PEG Example 2 under identical NMP conditions, with the only variable being the delivery vector changed from AR-PEG vesicles to AAV9 viral vector. The aim was to demonstrate the low immunogenicity advantage of the AR-PEG system by comparing IL-6 levels in the perfusion fluid.

[0273] Table 41 Specific Drug Loading in the AAV Control Group

[0274] Note: AAV was custom-packaged by Vigene Biosciences, and each construct was separately packaged and purified to ≥1×10⁻⁶. 13 vg / mL.

[0275] Dosing parameters set: EZH2:E2F3:shSTAT3:shZFX = 2:2:1:1 (corresponding to the molar ratio in Example 2), according to 1×10 12 Vg / liver administration (equivalent to 5 × 10 in the AR-PEG group) 12 Particles / liver), injected into the NMP perfusion circuit via portal vein cannulation at a rate of 1 mL / min, volume ≤5 mL, perfused in a double circulation at 37°C for a total of 6 hours (exactly the same as in Example 2). Displacement cleaning solution (applicable to all embodiments): After 1.6 hours of perfusion, vesicle circulation was stopped, and fresh pre-warmed (37°C) perfusion solution (Williams E + 2% BSA + insulin + dexamethasone, without any drugs) was used for replacement and cleaning.

[0276] 2. First cleaning: Inject 3 times the perfusion volume of fresh perfusion fluid (rat liver: 600 mL; aged pig liver: 3 × 2000 mL = 6 L), and perfuse continuously for 30 minutes at normal perfusion pressure. Discard the perfusion fluid (>90% of free components removed).

[0277] 3. Second rinse: Inject 3 times the volume of fresh perfusion solution, perfuse for 20 minutes, and discard (cumulative removal of >99% of free components).

[0278] 4. Third rinse: Inject 3 times the volume of fresh perfusion solution, perfuse for 10 minutes, and discard (cumulative removal of >99.9% of free components).

[0279] 5. After cleaning, take an organ biopsy for residual detection (50-100 mg per leaf), and take a sample of the last cleaning solution (5 mL) for IL-6 detection. The organ is then placed in cryopreservation to await transplantation or testing.

[0280] VSV-G protein residue detection (applicable to all organs in all embodiments)

[0281] Detection method: ELISA

[0282] 1. Organ biopsy (50 mg / leaf, 3 leaves in total): Add 500 μL RIPA lysis buffer (containing protease inhibitor), homogenize and lyse, centrifuge at 14,000×g for 15 minutes, collect the supernatant, and quantify the total protein to 1 mg / mL using BCA.

[0283] 2. ELISA procedure: Using anti-VSV-G monoclonal antibody (Sigma V5507) as the capture antibody, HRP-labeled secondary antibody (goat anti-mouse, 1:5000) was used for detection, TMB color development was performed, and the plate was read at 450 nm.

[0284] 3. Standard curve: Purified VSV-G protein (Kerafast EVU001) was serially diluted from 0.01 to 10 ng / mL (matrix matched: prepared in negative control liver lysate).

[0285] VSV-G residue (ng / g) = ELISA concentration (ng / mL) × lysis buffer volume (mL) / tissue weight (g).

[0286] Table 42. VSV-G Residue in Various Organs (Examples 2-6)

[0287] Detection of Small Molecule Active Ingredient Residues—LC-MS / MS Method (Specific to Example 7)

[0288] Table 43 Loading of small molecules (VC6TF+KATi, six types) and release criteria (Example 7)

[0289] LC-MS / MS detection process: 1. Take liver biopsy (100 mg) after three washes, grind with liquid nitrogen, add 500 μL acetonitrile:water (80:20, containing internal standard d5-VPA 100 ng / mL + d4-TCP 50 ng / mL), vortex and sonicate for extraction, centrifuge at 15,000×g, purify the supernatant with a C18 SPE column, dry with nitrogen, reconstitute with 100 μL mobile phase A, filter through a 0.22 μm membrane, and inject.

[0290] 2. Instruments: Waters ACQUITY UPLC + Xevo TQ-S triple quadrupole mass spectrometer; column: HSS T3 (1.8 μm, 2.1×100 mm); flow rate: 0.3 mL / min; gradient: 5%-95% acetonitrile (0.1% formic acid), 8 min; ESI positive / negative dual mode; MRM detection.

[0291] 3. Small molecule MRM ion pairs: Table 44 Small molecule MRM ion pairs

[0292] Tissue residue (ng / g) = LC-MS quantitative concentration (ng / mL) × extraction volume (mL) × dilution factor / tissue weight (g).

[0293] Results: After three replacement washes, the residual amounts of the six small molecules in the liver tissue were all below the release criteria (1 / 10 of their respective IC50), and were determined to be 'below the pharmacologically effective concentration', making them safe for transplantation.

[0294] Detection of the inflammatory factor IL-6 in the perfusion fluid (AR-PEG group vs. AAV control group)

[0295] At the end of perfusion (T=6h) and after three washes (T=6h+wash), 5mL of perfusion fluid was collected from the perfusion circuit (samples were collected simultaneously from each AR-PEG group and the AAV control group). The samples were centrifuged at 2,000×g for 10 minutes, and the supernatant was collected, stored at -80°C, and analyzed in batches. Rat IL-6 ELISA (R&D Systems DY506): The undiluted perfusion fluid was directly detected (no BCA normalization required).

[0296] Table 45 Results of IL-6 detection in perfusion fluid (inflammatory factor)

[0297] Note: Statistical methods: IL-6 intergroup comparison: one-way ANOVA + Tukey post-hoc test (n=5-6 / group), mean ± standard deviation reported, p<0.001 is considered significant.

[0298] Conclusion: The IL-6 levels in the perfusion fluid of all AR-PEG treatment groups (regardless of whether they were loaded with nucleic acids, small molecules, or proteins) were significantly lower than those in the AAV control group (p<0.001), with a reduction of approximately 85-90%. This demonstrates that the AR-PEG engineered vesicle system, as a delivery platform for endogenous protein sources, possesses excellent low immunogenicity characteristics, representing a core safety advantage compared to AAV viral vectors.

[0299] Based on the above three safety indicators, the AR-PEG engineered vesicle system of this invention can efficiently remove carrier components (VSV-G clearance rate >99.5%) after three replacement washes following perfusion of ex vivo organs. For small molecule delivery examples, it can effectively remove active ingredients to pharmacologically ineffective concentrations. Furthermore, the immunogenicity of the system is significantly lower than that of AAV viral vectors, providing a safe and reliable technical guarantee for pre-transplant repair of clinical organs.

[0300] Table 46 Detection methods for each indicator in Examples 2-8

[0301] As can be seen from the above embodiments, the present invention has the following advantages: 1. Profound Multi-Organ Repair and Rejuvenation Capabilities: Based on this efficient platform, this invention can induce profound repair in multiple marginal organs within a short (4-8 hours) in vitro perfusion window. For models such as fatty liver, aged kidney, ischemic heart, and damaged lung, the proposed solution can not only fully restore organ-specific functions but also achieve a 57% to 77% deep epigenetic age reversal, marking a breakthrough from functional improvement to cellular remodeling.

[0302] 2. Highly effective reversal of hepatic steatosis and fibrosis: By delivering a combination of transcription factors that have been systematically screened and validated (such as mRNA EZH2, mRNA E2F3, RNAi STAT3, and RNAi ZFX), this invention can simultaneously and significantly reverse steatosis (reduction >60%) and fibrosis (reduction >50%) in fatty liver in a single perfusion, and improve glucose metabolism. This is a comprehensive therapeutic effect that is difficult to achieve with existing single intervention methods.

[0303] 3. Deep rejuvenation and repair of aged donor pig liver: By delivering the self-replicating mRNA OSKGL combination (Oct4:Sox2:Klf4:Glis1:Lin28 = 3:1:1:1:1), this invention can significantly reverse the aging phenotype of aged pig liver within 6 hours, achieving an epigenetic age reversal of >60% and restoring functional indicators to youthful levels, providing an efficient solution for clinically salvaging elderly marginal donor livers.

[0304] 4. Excellent clinical translational safety: The system components are derived from endogenous proteins with low immunogenicity. Its unique vesicular properties allow for efficient removal via standard perfusion and washing procedures (residual carrier protein <1 ng / g tissue). The combination of transcription factors used has been documented in the literature as having a low tumorigenic risk, providing a safety guarantee for in vitro clinical applications.

[0305] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. An engineered fusion protein, characterized in that, The fusion protein includes a functional fragment of the human ARRDC1 protein, a flexible linker peptide, and an RNA-binding domain. The amino acid sequence of the functional fragment of the human ARRDC1 protein is shown in SEQ ID NO: 1; The flexible linker peptide is (GGGGS)n, where n is an integer from 2 to 4; The amino acid sequence of the RNA-binding functional domain is shown in SEQ ID NO:

3.

2. A fusion gene encoding the engineered fusion protein of claim 1, characterized in that, The nucleotide sequence of the fusion gene is shown in SEQ ID NO:

6.

3. An engineered fusion protein of claim 1 or a fusion gene-mediated active budding vesicle system of claim 2, characterized in that, The active budding vesicle system includes: vesicles mediated by the engineered fusion protein or the fusion gene; a membrane fusion protein VSV-G located on the vesicle membrane; and drug molecules encapsulated within the vesicle.

4. The active budding vesicle system as described in claim 3, characterized in that, The drug molecules include nucleic acid molecules, small chemical molecule compositions, or polypeptides; Preferably, the nucleic acid molecule includes mRNA, self-replicating mRNA, or RNA interference molecule; the 5' or 3' untranslated region of the nucleic acid molecule contains a packaging signal sequence; More preferably, the mRNA is an mRNA encoding EZH2 and / or E2F3, and / or an RNA interference molecule targeting STAT3 and / or ZFX; the molar ratio of mRNA EZH2, mRNA E2F3, siRNA STAT3, and siRNA ZFX is (1-3):(1-3):1:1; More preferably, the self-replicating mRNA encodes Oct4, Sox2, Klf4, Glis1, and Lin28A; the molar ratio of Oct4, Sox2, Klf4, Glis1, and Lin28 is (2-4):1:1:1:1; Preferably, the chemical small molecule composition comprises an epigenetic regulator and a key signaling pathway inhibitor; the epigenetic regulator comprises a histone deacetylase inhibitor, a GSK-3β inhibitor, a TGF-β signaling pathway inhibitor, and a lysine-specific demethylase 1 inhibitor; the key signaling pathway inhibitor is a KAT acetyltransferase inhibitor module, comprising an inhibitor of histone acetyltransferase KAT3A / B and an inhibitor of histone acetyltransferase KAT6A; Preferably, the polypeptide is a polypeptide of a regeneration factor or a reprogramming factor, and the polypeptide is fused with a cell-penetrating peptide.

5. The active budding vesicle system as described in claim 4, characterized in that, The packaging signal sequence is the 5'UTR of human PEG10 mRNA, with a nucleotide sequence as shown in SEQ ID NO:9, or a variant thereof having at least 80% sequence identity and capable of enriching the nucleic acid by ≥10-fold when co-expressed with the engineered fusion protein.

6. The active budding vesicle system as described in claim 4, characterized in that, The chemical small molecule composition includes valproic acid, CHIR99021, RepSox, transphenylcyclopropylamine, C646, and WM-1119.

7. The active budding vesicle system according to any one of claims 4 to 6, characterized in that, The engineered fusion protein is cloned into an expression vector and co-transfected into cells with an expression plasmid carrying the membrane fusion protein VSV-G and a reporter plasmid carrying the nucleic acid molecule to obtain the actively budding vesicles; or the expression vector cloning the engineered fusion protein is co-transfected into cells with an expression plasmid carrying the membrane fusion protein VSV-G and an empty reporter plasmid vector, and simultaneously co-incubated with a small chemical molecule composition or peptide to obtain the actively budding vesicles.

8. The application of the engineered fusion protein of claim 1, the fusion gene of claim 2, or the active budding vesicle system of any one of claims 3 to 7 in the repair of ex vivo marginal donor organs.

9. A method for repairing ex vivo marginal donor organs using the active budding vesicle system according to any one of claims 3 to 7, characterized in that, Includes the following steps: (1) Connect the ex vivo marginal donor organ to an ambient temperature mechanical perfusion system that matches its type; (2) Introducing an effective dose of an active budding vesicle system into a room-temperature mechanical perfusion system; (3) Maintain perfusion for 4-8 hours; before transplantation, replace and clean with fresh perfusion fluid at least 3 times the capacity of the perfusion system to obtain the repaired ex vivo organ.

10. The method for repairing ex vivo marginal donor organs as described in claim 9, characterized in that, The ex vivo marginal donor organs are selected from liver, kidney, heart or lung.

Images