Sandbis virus-like particle trVLP-ASV with transcriptional and replication capabilities and application of Sandbis virus-like particle trVLP-ASV

Abstract

The invention relates to the technical field of bioengineering, in particular to a virus-like particle trVLP-ASV which has transcription and replication capabilities, is based on a Sindbis virus and displays a receptor binding domain of SARS-CoV-2 Spike protein and application. The genome of the trVLP-ASV provided by the invention does not contain an RNA replicase nsP4 gene, but inversely expresses the nsP4 gene in a BHK-21 cell, so that the possibility of biological hazards is greatly reduced; the trVLP-ASV which mainly depends on ACE2 as a receptor is formed by carrying out engineering transformation on a structural protein E2 of a Sindbis virus. The method can be used as a platform for analyzing the virus immune escape process and identifying adaptive mutation promoting virus cross-species transmission, has the advantages of safety, rapidness, flexibility and high efficiency, and has important significance on monitoring of continuous evolution of viruses and development of clinical coping strategies.

Description

Technical Field

[0001] The present invention relates to the field of bioengineering technology, and in particular to a Sindbis virus-like particle trVLP-ASV with transcription and replication capabilities and its applications. Background Art

[0002] Mutations in the SARS-CoV-2 genome are mainly located in the spike protein (Spike), which is crucial for determining the host range and immune escape ability of the virus. Therefore, characterizing and understanding the impact of Spike mutations on virus host range and immune escape is crucial for developing effective treatment and prevention strategies. Although multiple methods have been developed to analyze the interaction between spike variants and the ACE2 receptor, these methods are often limited by biosafety issues, lack of flexibility, or technical complexity.

[0003] Utilizing yeast or mammalian cell surface display technology is a commonly used method. Multiple laboratories have constructed deep saturation mutation libraries of the receptor binding domain (RBD) or full-length Spike through this method. These mutation libraries have been used to study how Spike mutations affect receptor binding and neutralizing antibody escape, usually evaluated by fluorescence or magnetic sorting. Such studies have greatly deepened our understanding of potential virus escape mutations and promoted the development of antibody therapeutic drugs. However, these surface display-based methods are limited to analyzing protein-protein interactions and lack the complete virus infection process and antibody-mediated virus neutralization. In addition, for yeast display, the glycoproteins expressed on the surface of yeast cells are different from those on mammalian cells, which may limit the effectiveness of this method in simulating Spike-receptor binding.

[0004] In contrast, virus-based systems provide a more realistic model of virus invasion and neutralization. Although the use of live SARS-CoV-2 virus enables the study of antibody escape and Spike mutations, its biosafety requirements and the engineering challenges posed by its large genome have prompted the development of Spike-based pseudovirus or recombinant virus systems. For example, lentivirus-based pseudoviruses provide a safe and efficient method for evaluating how specific mutations affect antibody neutralization and virus infection. However, this method cannot fully mimic the evolution of Spike because it relies on a predefined mutant library that is both time-consuming and laborious to construct. In contrast, Spike-based recombinant vesicular stomatitis virus (rVSV) supports virus evolution studies. Replacing the VSV-G protein with the Spike of SARS-CoV-2 can generate replication-competent recombinant VSV, which is widely used to identify SARS-CoV-2 escape mutations and analyze the virus's evolutionary trajectory under antibody pressure. However, since VSV is a negative-strand RNA virus, its packaging technology is extremely challenging and time-consuming, and creating a mutant library artificially in this system is even more complex. In addition, although VSV is classified as a biosafety level 2 (BSL-2) virus, the recombinant VSV system may still pose potential safety risks. Therefore, it is necessary to develop safer systems that are easier and faster to package, especially for packaging virus particles containing the Spike protein or large mutant libraries designed to mimic the evolution of the Spike protein.

[0005] In view of this, the present invention is specifically proposed. Summary of the Invention

[0006] To solve the above technical problems, the present invention provides a Sindbis virus-like particle trVLP-ASV with transcriptional and replicative capabilities and its applications.

[0007] Specifically, the technical solution of the present invention is as follows: In a first aspect, the present invention provides a Sindbis virus-like particle trVLP-ASV with transcriptional and replicative capabilities, whose genome does not contain the RNA replicase nsP4 gene; and whose E2 protein contains the receptor-binding domain of the Spike protein of SARS-CoV-2.

[0008] In the present invention, the Sindbis virus genome lacks the replicase nsP4 gene. The preferred way to make the genome not contain the nsP4 gene is to delete the base sequence of nsP4 at positions 5763-7358 of the genome.

[0009] In the present invention, the receptor-binding domain of the structural protein E2 itself (P171-C220 of E2) is deleted, and the receptor domain of Spike is inserted. The ways to make the E2 protein contain the receptor-binding domain of the Spike protein are selected from: inserting the receptor-binding domain of Spike of SARS-CoV-2 and its variants (amino acids N331-T531 of Spike) or inserting the receptor-binding domain of SARS-CoV-1 (amino acids R306-F527 of Spike), at least one of which.

[0010] The trVLP-ASV described in the present invention is constructed based on the Sindbis virus, and it can only replicate and amplify in host cells that trans - express nsP4.

[0011] The trVLP-ASV described in the present invention lacks the receptor-binding domain of the Sindbis virus itself, and at the same time inserts the receptor-binding domain of the Spike of SARS-CoV-2 or SARS-CoV-1 virus, so it uses ACE2 as the main receptor.

[0012] The trVLP-ASV described in the present invention, as a virus particle, its invasion of host cells can be inhibited by antibodies against the receptor-binding domain of SARS-CoV-2.

[0013] In the second aspect, the present invention provides a method for constructing the Sindbis virus-like particle trVLP-ASV with transcription and replication ability. The receptor-binding domain of Spike and its mutants or mutant libraries can be inserted into the E2 region of the Sindbis virus genome by means such as gibson assembly.

[0014] In the third aspect, the present invention provides a method for preparing the Sindbis virus-like particle trVLP-ASV with transcription and replication ability. After transfecting the plasmid into host cells expressing ACE2 and nsP4, the RNA transcription of trVLP-ASV is driven by the CMV promoter, and then the relevant proteins are translated and assembled into virus-like particles.

[0015] In the fourth aspect, the present invention provides the application of the Sindbis virus-like particle trVLP-ASV with transcription and replication ability in the construction of a SARS-CoV-2 immune escape model.

[0016] In the fifth aspect, the present invention provides the application of the Sindbis virus-like particle trVLP-ASV with transcription and replication ability in the study of the host range of SARS-CoV-2.

[0017] Beneficial effects: The present invention provides a Sindbis virus-like particle trVLP-ASV with transcription and replication capabilities and its applications. The genome of the trVLP-ASV provided by the present invention does not contain the nsP4 gene of the Sindbis virus, thereby restricting the replication of trVLP-ASV to specific host cells and reducing biological hazards. By engineering the structural protein E2 of the Sindbis virus and then inserting the receptor-binding domain (RBD) of the Spike protein, the E2 protein contains the receptor-binding domain of the Spike protein, thereby forming a trVLP-ASV that mainly relies on ACE2 as the receptor. The trVLP-ASV provided by the present invention can utilize the error-prone characteristics of the nsP4 replicase to construct a mutant library, or a plasmid mutant library of the receptor-binding domain can be constructed in vitro and then transfected into host cells to be packaged into a trVLP-ASV mutant library. The mutant library constructed by trVLP-ASV can be used as a platform for analyzing the immune escape process of SARS-CoV-2 and identifying key adaptive mutations for cross-species transmission, with the advantages of safety, rapidity, flexibility, and high efficiency, and is of great significance for the monitoring of the continuous evolution of SARS-CoV-2 and the development of countermeasures. In addition, by replacing the receptor-binding domain, there is also the potential to expand to other viruses for similar analyses, such as SARS-CoV-1. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will describe the drawings required for the examples or the prior art description.

[0019] Figure 1 The construction process of a virus-like particle (trVLP) system capable of transcription and replication. Figure 1 In the figure: Figure A shows the construction process from ESV1 to trVLP-ESV1 and the expected phenotypes; Figure B shows the EGFP fluorescence microscopy images and the corresponding bright-field images after 12 hours of infection under different conditions; Figures C and D show the flow cytometry maps corresponding to the experiment in Figure B and the quantitative data of the cell population (n = 3 biological replicates); Figure E shows the virus titers of ESV1 and trVLP-ESV1 measured in BHK-21 cells expressing nsP4 (24 hours after infection, MOI = 0.1). Note that the host cells used here overexpress human EGFR (n = 3 biological replicates); Figure F shows the long-term passage experiment of trVLP-ESV1 and the changes in the level of the EGFP reporter gene in the corresponding host cells.

[0020] Figure 2 The process of engineering trVLP to use ACE2 as the main receptor. Figure 2In Chinese: Figure A shows the schematic diagram of the engineering design of trVLP. ASV represents the Sindbis virus targeting ACE2, and Ref_RBD represents the receptor-binding domain (RBD) sequence from the SARS-CoV-2 strain Wuhan-Hu-1 (NC_045512). The second round of engineering modification adopted the mutations reported in the literature; Figures B-C are typical microscopic imaging images showing the intracellular fluorescence expression of trVLP-ASV1-Ref_RBD (B) and trVLP-ASV2-Ref_RBD (C) at 8 hours and 24 hours after infection, with an MOI of approximately 2. The data are from 3 biological replicate experiments, scale bar: 100 μm; Figure D is flow cytometry data showing the proportion of trVLP-infected cells corresponding to the conditions in Figures B and C (n = 3 biological replicates); Figure E is the specificity index of trVLP-ASV, which represents the ratio of the infection efficiency of trVLP in host cells expressing human ACE2 to that in host cells lacking hACE2 (n = 3 biological replicates); Figure F is the titer determination of trVLP in the culture supernatant 24 hours after infection, using different versions of trVLP (MOI approximately 2, n = 3 biological replicates); Figure G is a comparison of the infection rates of trVLP containing different SARS-CoV-2 variant Spike RBDs in host cells expressing or not expressing human ACE2, with data collected 24 hours after infection (MOI = 0.5, n = 3 biological replicates); Figure H is a comparison of the infection rates of trVLP containing SARS-CoV-1 Spike RBD in host cells expressing or not expressing human ACE2; data were collected at 8 hours and 24 hours after infection (MOI = 2, n = 3 biological replicates).

[0021] Figure 3 The study used trVLP-ASV to simulate SARS-CoV-2 antibody neutralization. Figure 3In the figure: Panel A is a schematic diagram of the antibody neutralization experiment based on trVLP; Panel B is a microscopic image of mCherry fluorescence (trVLP-labeled gene), showing the neutralization effects of different antibodies on two types of trVLP (trVLP-ASV2-Ref_RBD and trVLP-ASV1-JN.1_RBD). The images were collected 8 hours after infection, with an MOI of approximately 0.5. Scale bar: 100 μm. Due to the low titer of trVLP-ASV2-JN.1_RBD, trVLP-ASV1-JN.1_RBD was used in the neutralization experiment; Panel C is the antibody neutralization assessment based on flow cytometry. After treating the cells with different gradient concentrations of antibodies, the mCherry expression level was detected by flow cytometry to measure the trVLP infection, and the IC50 value was calculated through the infection curve. The data are from two biological replicate experiments; Panel D uses the VSV-ΔG pseudovirus system to evaluate the antibody neutralization activity. Through the VSV-ΔG pseudovirus system, the neutralization ability of multiple antibodies against the Spike proteins of SARS-CoV-2 strains (D614G mutation) and JN.1 strains was evaluated, and the IC50 value was calculated. The data are from two biological replicate experiments.

[0022] Figure 4 This is a study on simulating antibody escape by leveraging the evolutionary ability of trVLP-ASV2. Figure 4 In the figure: Panel A is the flow chart of the trVLP evolution experiment for simulating antibody escape; Panel B is an example of the mutation spectrum of the evolved RNA. Mutations with a mutation enrichment ratio exceeding 5% are considered significantly enriched and are highlighted in the figure; Panel C is the summary of the mutation enrichment data in the trVLP evolution experiments driven by multiple antibodies. The data are from two biological replicate experiments; Panels D - E show the mutation positions and verification results. The left panel shows the mutation positions in the RBD region under the antibody selection pressure of REGN10987 or LY-CoV1404. The structural information is from PDB: 6XDG and PDB: 7MMO respectively. The right panel evaluates the impact of these mutations on antibody escape using the VSV-ΔG pseudovirus system. The data are based on two biological replicate experiments; Panel F shows the main enriched mutations in the DMS library after SA55 antibody treatment. The table lists the mutation data and enrichment fold in the trVLP-ASV1-JN.1_RBD mutation library after treatment with 10 μg / ml SA55 antibody. The results are the average of two biological replicate experiments; Panel G is the display and verification of the SA55 antibody-enriched mutations. The left panel shows the mutation positions enriched in the trVLP-ASV1-JN.1_RBD library after SA55 antibody treatment. The structural information is from PDB: 7Y0W and PDB: 9IU1. The right panel evaluates the impact of two mutations on antibody escape. The experiment was conducted using the VSV-ΔG pseudovirus system, and each experimental condition included two biological replicates.

[0023] Figure 5 Study on trVLP-ASV simulating the binding of SARS-CoV-2 spike protein to different ACE2 homologs. Figure 5 In it: Panel A is a schematic diagram of the experimental design, showing the protocol of using trVLP to transduce host cells expressing ACE2 homologs from 9 different species to evaluate the ACE2-mediated cross-species infection ability; Panel B are the key ACE2 residues determined by previous studies for SARS-CoV-2 binding. The table summarizes the key residues within the binding interface of SARS-CoV-2 RBD to human ACE2 (within 4 Å) and analyzes their conservation in nine species; Panel C shows the infection efficiency of cells expressing different ACE2 homologs to trVLP variants. The bar graph compares the infection efficiencies of three trVLP variants - trVLP-ASV2-Ref_RBD, trVLP-ASV2-XBB.1.5_RBD, and trVLP-ASV2-JN.1_RBD - in host cells expressing ACE2 of different species. The MOI of trVLP was estimated to be ~0.4 based on cells expressing human ACE2, n = 3 biological replicates; Panel D is the correlation analysis of the infection efficiency between trVLP and SARS-CoV-2 virus. This figure evaluates the relationship between the infection efficiency of trVLP-ASV2-Ref_RBD mediated by ACE2 of different species and live SARS-CoV-2 virus. The data of trVLP-ASV2-Ref_RBD are the average of three biological replicates (from Panel C); Panel E shows the binding strength of ACE2 homologs to Ref_RBD and XBB.1.5_RBD, evaluating the binding strength of ACE2 from different species to two RBD variants (Ref_RBD and XBB.1.5_RBD). The RBD concentration is 10 μg / ml, and the results are the average of three biological replicates.

[0024] Figure 6 To identify functional mutations related to the cross-species transmission of SARS-CoV-2 through continuous trVLP adaptive evolution. Figure 6In: Panel A is a schematic diagram of the experimental design and workflow for identifying key mutations that promote viral cross-species transmission; Panel B is a table summarizing the mutations enriched after 5 rounds of adaptive evolution of trVLP-ASV2-Ref_RBD in host cells expressing 9 different species of ACE2; Panel C is a nucleotide-level analysis of the mutation enrichment, showing the mutation enrichment profiles of trVLP-ASV2-Ref_RBD after adaptive evolution in host cells overexpressing human or murine ACE2. The figure shows representative biological replicate results (a total of two replicates); Panel D compares the binding abilities of Ref_RBD and its two variants to host cells expressing different species of ACE2. The results are based on three biological replicates, and the RBD concentration is 10 μg / ml.

[0025] Figure 7 is Figure 1 supplementary data of. Figure 7 In: Panel A reviews several high-throughput methods for studying the effects of SARS-CoV-2 Spike mutations on virus entry and antibody escape. This review provides a comprehensive comparison of available tools; Panel B is flow cytometry data from a serial passage experiment. The supernatant from each round of passage was used to transduce WT BHK-21 cells expressing EGFR and BHK-21 cells co-expressing nsP4 and EGFR. Flow analysis was performed 24 hours after infection, showing representative data from three biological replicates.

[0026] Figure 8 is Figure 2 supplementary data of. Figure 8 In: Panels A and B are flow cytometry results comparing the infection of trVLP-ASV1-Ref_RBD in different host cells. Infection was performed at an MOI of approximately 2, and the data were obtained 8 hours after infection, n = 3 biological replicates; Panel C is flow cytometry data comparing the infection efficiency of trVLP displaying the Spike protein of SARS-CoV-2 variants in host cells with or without human ACE2; Panel D is flow cytometry data comparing the infection efficiency of trVLP displaying the Spike RBD of SARS-CoV-1 in host cells with or without human ACE2.

[0027] Figure 9 is Figure 3 and Figure 4 supplementary data of. Figure 9In the figure: Panel A is a schematic diagram showing the genomic regions used to measure mutation frequencies in trVLP; Panel B shows the change in mutation frequency in the trVLP mCherry region over time. The left bar graph shows the mutation frequencies, and the right graph shows the results of linear fitting. The data are from three biological replicates; Panel C shows representative mCherry fluorescence and bright-field snapshots demonstrating the effect of antibody supernatant on neutralizing trVLP. The images were taken 8 - 10 hours after trVLP infection. The data are from three biological replicates, showing representative images. The scale bar represents 100 µm.

[0028] Figure 10 is Figure 4 supplementary data of. Figure 10 In the figure, Panel A shows the nucleotide-level analysis of mutations in the RBD region of trVLP-ASV2-Ref_RBD and trVLP-ASV2-JN.1_RBD after treatment with neutralizing antibodies. The data are from two biological replicates. Panel B is a heatmap showing the enrichment of RBD mutants in the trVLP-ASV1-JN.1_RBD library after treatment with the SA55 antibody. The heatmap values represent the average of enrichment fractions from two biological replicates.

[0029] Figure 11 is Figure 5 supplementary data of. Figure 11 In the figure, Panel A shows the phylogenetic analysis of ACE2 orthologous genes from nine species. Panel B is an mCherry fluorescence and bright-field image showing trVLP infection in mouse ACE2-expressing host cells at an MOI of 0.4. The images were taken 8 hours after infection. The scale bar marked in the figure is 100 µm.

[0030] Figure 12 is Figure 6 supplementary data of. Figure 12 Shows the mutation analysis of trVLP-ASV2-Ref_RBD after adaptive evolution in host cells expressing ACE2 orthologous genes from nine different species. The data are from the results of two biological replicate experiments. Detailed implementation

[0031] The continuous evolution of SARS-CoV-2 poses a major challenge to global public health. Developing a safe, rapid, and versatile method for evaluating the functions of Spike protein mutations in immune escape and cross-species transmission is of great significance. To this end, we constructed a virus-like particle (trVLP-ASV) system based on Sindbis virus with transcriptional and replicative capabilities, and engineered it to carry the receptor-binding domain (RBD) of SARS-CoV-2. The trVLP-ASV system can only replicate and spread in BHK-21 cells that simultaneously express the viral RNA replicase nsP4 and human ACE2, providing a controllable and safe model for studying the mechanism of SARS-CoV-2 invasion into cells. By leveraging the evolutionary ability of trVLP-ASV mediated by error-prone RNA replication and the ability of trVLP-ASV to rapidly construct mutant libraries, we screened out RBD mutants that can escape antibody inhibition or enhance ACE2 adaptability, and these ACE2 proteins are derived from a series of mammals. The research results show that trVLP-ASV is a safe, rapid, and flexible platform, which is of great value for analyzing the immune escape mechanism of SARS-CoV-2 and identifying key adaptive mutations, and these mutations are crucial for monitoring virus evolution and formulating prevention and control strategies.

[0032] In this study, based on the ESV1 virus derived from the Sindbis virus (a positive-strand RNA virus), we constructed a safe and versatile virus-like particle (trVLP-ASV) system with transcriptional and replicative capabilities. It has a high mutation rate and is easily packaged by recombinant DNA. To mimic the protein-protein interactions driving the evolution of SARS-CoV-2 through Spike mutations, our system introduced two key innovations. First, the RNA replicase (nsP4) was removed from the RNA genome of ESV1 and stably expressed in the host genome, thereby restricting the replication of trVLP-ASV to specific host cells and reducing the biological hazard. Second, we engineered the E2 protein by deleting the receptor-binding domain of the Sindbis virus itself, inserting the RBD of the Spike protein, and introducing a series of point mutations to further reduce its ACE2-independent invasion, thus forming trVLP-ASV that mainly depends on ACE2 as the receptor. By exploiting the error-prone replication property of RNA and the ability of trVLP-ASV to easily package mutant libraries, we identified a series of antibody escape mutations in Spike and studied how these mutations mediate the adaptation of trVLP-ASV to ACE2 receptors from nine mammalian species. Overall, our system effectively mimics virus evolution, similar to rVSV, while avoiding the complex packaging process, the difficulty of generating mutant libraries, the long production timeline, and potential biosafety issues. Therefore, it provides a powerful tool for predicting the evolution and cross-species transmission of SARS-CoV-2, and it may be applicable to the study of other virus systems beyond SARS-CoV-2, such as SARS-CoV-1.

[0033] The continuous transmission and evolution of SARS-CoV-2 highlight the importance of understanding its evolutionary dynamics. Based on an ESV1 virus system derived from the Sindbis virus, the present invention obtained an ACE2-targeted trVLP-ASV system through engineering to mimic the evolutionary process of the Spike RBD of SARS-CoV-2. By implementing multiple engineering innovations, our system provides a safe, simple, and versatile alternative to traditional live viruses, recombinant viruses, or pseudoviruses.

[0034] By deleting nsP4 replicase from the RNA genome and expressing it trans - complementarily in host cells, the trVLP - ASV system minimizes biosafety risks by confining replication to specific cells. Since the level of RNA replicase required for virus spread is lower than that of structural proteins (e.g., SARS - CoV - 2 N protein), integrating the replicase into the host genome may have a relatively minor impact on virus titers. Additionally, modification of the Sindbis structural proteins enables its native RBD to be replaced by the Spike RBD of SARS - CoV - 2, thus targeting the ACE2 receptor with high specificity. Further rational engineering reduces the affinity of the E2 protein for its native receptor, significantly enhancing the specificity of trVLP - ASV for host cells expressing ACE2.

[0035] This specificity ensures accurate modeling of SARS - CoV - 2 receptor interactions, paving the way for virus evolution studies. Moreover, the trVLP - ASV system exploits the short positive - strand RNA genome of the Sindbis virus and its high mutation rate to efficiently generate mutant libraries, enabling the study of Spike mutant functions. Using this system, we identified escape mutants for 10 antibodies, most of which have been validated or are consistent with previous studies, highlighting the reliability and robustness of the system. Furthermore, it can map Spike adaptation on different ACE2 receptors, which lays the foundation for its potential value in studying SARS - CoV - 2 host range expansion. This ability makes the system highly useful for understanding virus evolution and adaptation.

[0036] Compared with existing platforms, including yeast display, mammalian cell display, and pseudovirus - based models, the trVLP - ASV method circumvents several key limitations. However, there is still room for improvement and expansion. For example, while our system is specifically targeted at Spike - ACE2 interactions, further engineering of the Sindbis - virus - based trVLP - ASV backbone to display other proteins can significantly expand its application scope. Additionally, a remaining challenge is the residual binding of the E2 protein to its native receptor, resulting in background infection. Addressing this interference is an important direction for future research and may be achieved by optimizing virus glycoprotein modularity. Similar design principles have been explored in coronavirus receptor design studies, and applying such a modular approach to SARS - CoV - 2 Spike glycoprotein may enhance our ability to predict virus evolution and engineer virus vectors for therapy. Given the relatively modular and genetically tractable nature of the Sindbis virus structural proteins, our work provides a starting point for exploring glycoprotein design principles in this system. Moreover, although our modifications reduce potential biosafety risks, implementing additional measures such as a segmented - protein strategy can further enhance safety.

[0037] Overall, the trVLP-ASV system fills the gap between traditional pseudovirus models and live virus systems, providing a safer and more efficient platform for studying the evolution of SARS-CoV-2. It can identify antibody escape mutations and explore host range adaptation in a controlled environment. In addition, the design of trVLP-ASV may provide valuable insights for the development of oncolytic viruses, vaccines, and other antiviral tools, demonstrating its broad potential in the fields of virology and therapeutic innovation.

[0038] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0039] At the endpoints and any values disclosed in this specification, the exact range or value is not limited to that precise range or value, and these ranges or values should be understood to include values approaching these ranges or values. For numerical ranges, between the endpoint values of each range, between the endpoint values of each range and individual point values, and between individual point values, they can be combined with each other to obtain one or more new numerical ranges, and these numerical ranges should be regarded as specifically disclosed herein.

[0040] In the description of this specification, the description referring to terms such as "one embodiment", "some embodiments", "specific implementation manners", or "some specific implementation manners" means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the embodiments of the present invention. In this specification, the schematic descriptions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0041] In the embodiments provided in this specification, for those without specific technical or conditions indicated, they shall be carried out according to the technologies or conditions described in the literature in this field or according to the product specifications. For the reagents or instruments without the manufacturer indicated, they are all conventional products that can be obtained through regular channels.

[0042] Example 1 The method involved in this example is as follows: (1) Cell culture and construction of stable transfected cell lines.

[0043] In this study, the hamster kidney fibroblast cell line BHK-21 (iCell, C001), the human embryonic kidney cell line HEK293T (ATCC), and Expi293F cells (Thermo Fisher, A14527) were used. BHK-21 cells were cultured in Minimum Essential Medium (MEM) which contained EBSS and L-glutamine (HyClone, SH30024.01), and supplemented with 10% fetal bovine serum (ExCell, FSP500), 1% penicillin-streptomycin-glutamine (Gibco, 10378016), and 1% non-essential amino acid solution (Gibco, 11140050). HEK293T cells were cultured in DMEM medium which contained 10% fetal bovine serum (ExCell, FSP500) and 1% penicillin-streptomycin-glutamine (Gibco, 10378016). Expi293F cells were cultured in SMM293-TII expression medium (Sino Biological, M293TII) containing 1% penicillin-streptomycin-glutamine (Gibco, 10378016). BHK-21 and HEK293T cells were incubated at 37°C in a humidified environment with 5% carbon dioxide. Expi293F cells were cultured at 37°C on a shaker at 160 rpm with 5% carbon dioxide. All cell lines were routinely tested and confirmed to be free of mycoplasma contamination.

[0044] In this study, the Piggybac transposon system was used to generate stable cell lines that overexpressed human ACE2 or EGFR and carried the BFP reporter gene. To transfect the cells, 2 μg of the expression plasmid and 0.5 μg of the transposase plasmid were co-transfected into wild-type BHK-21 cells or BHK-21 cells expressing nsP4 replicase using Lipofectamine LTX reagent and PLUS reagent (Thermo Fisher Scientific, 15338100). One week after transfection, BFP-positive cells were sorted by flow cytometry and amplified for subsequent experiments. For other stable cell lines, a lentiviral transduction system was used. To generate stable cell lines expressing nsP4, 3 μg of the Lenti-pEF1α-iRFP-NLS-P2A-ΔnsP3-nsP4 plasmid, 2.66 μg of the pCMV-dR8.91 packaging plasmid, and 0.34 μg of the PMD2.G envelope plasmid were co-transfected into HEK293T cells at 80% confluence in a 6-well plate using PEI reagent. After 8 hours, the medium was replaced with fresh medium. The lentiviral supernatant was collected 48 hours after transfection, centrifuged at 500×g for 4 minutes, and filtered through a 0.45-μm filter. Subsequently, 500 μL of the filtered lentiviral supernatant was added to BHK-21 cells at 30% confluence in a 6-well plate. The medium was replaced 12 hours after infection. One week later, iRFP-positive cells were sorted by flow cytometry and amplified for further experiments. Meanwhile, we established stable cell lines expressing ACE2 homologous genes from nine species (human, cynomolgus macaque, marmoset, squirrel monkey, mouse, pig, koala, and ferret). The ACE2 plasmid encoding the ZsGreen1 reporter gene (kindly provided by Dr. Qiang Ding of Tsinghua University) was packaged into lentivirus by the same procedure. These lentiviral particles were used to transduce BHK-21 cells expressing nsP4. After sorting out ZsGreen1-positive cells by flow cytometry and culturing them, stable cell lines were established for subsequent experiments.

[0045] (2) trVLP-ASV packaging and titer quantification.

[0046] In this study, we used plasmid transfection to generate trVLP-ASV RNA through Pol II transcription under the control of the CMV promoter. These RNAs underwent a translation, transcription, and replication process similar to that of Sindbis virus RNA within the cells, ultimately generating trVLP-ASV particles. For standard trVLP-ASV production, 3 μg of the trVLP-ASV plasmid was transfected into BHK-21 cells expressing nsP4 (expressing EGFR or ACE2), which were seeded at a density of 50% in one well of a 6-well plate, using Lipofectamine LTX reagent plus PLUS reagent (Thermo Fisher Scientific, 15338100). After 8 to 12 hours, the medium was replaced with fresh medium. After another 36 hours, the supernatant was collected, centrifuged at 500×g for 5 minutes, filtered through a 0.45-μm membrane, aliquoted, and stored at −80°C for later use. To produce the trVLP-ASV deep mutational scanning library, the plasmid amount was increased to 15 μg and transfection was carried out in a 10-cm culture dish. In each replicate experiment, the supernatants from 4 culture dishes were mixed, filtered, aliquoted, and stored as described above.

[0047] To determine the trVLP-ASV titer, we adopted the serial dilution method combined with flow cytometry, taking advantage of the fluorescent protein reporter gene carried by trVLP-ASV. Before infection, BHK-21 cells expressing nsP4 (expressing EGFR or ACE2) were seeded in a 24-well plate at a density of 150,000 cells per well and incubated for 12 hours. Six serial dilutions of the trVLP-ASV supernatant (0 μL, 1 μL, 5 μL, 10 μL, 50 μL, and 100 μL) were added to each well, and finally 500 μL of medium was added to each well, with three technical replicates set for each dilution. Another three wells were set for cell counting to determine the initial cell number. One hour after the addition of the trVLP-ASV dilutions, the medium was replaced, and the cells were washed twice with DPBS. Eight hours after infection, the cells were collected, and the proportion of fluorescent cells was measured by flow cytometry. To minimize the impact of secondary trVLP-ASV infection on the measurement accuracy, we chose 8 hours after infection as the analysis time point. At this time, the fluorescent signal of the primary infection had been fully expressed, while the secondary infection event had not produced detectable fluorescent signals, ensuring an accurate assessment of trVLP-ASV infection.

[0048] (3) Molecular cloning.

[0049] The nucleotide sequence of ESV1 virus is based on the Girdwood strain of Sindbis virus (Addgene, accession numbers #127692 and #127695), with modifications including mutations, linkers, and EGF sequences. As described in previous studies, the sequence of the SARS-CoV-2 spike protein is from the NCBI database, and the antibody sequence information is from relevant literature. The ACE2 sequences of nine different species and their corresponding plasmids were generously provided by Dr. Qiang Ding of Tsinghua University. The detailed sequence information of the key plasmids used in this study is listed in Table S1. Plasmid construction was carried out using Gibson assembly. DNA sequences were synthesized by Tsingke, and primers were ordered from GENEWIZ. The constructed plasmids were cloned and cultured in Escherichia coli Trelief5α (Tsingke, TSC-C01) and confirmed by Sanger sequencing (RiboBio).

[0050] (4) Construction of the trVLP-ASV1-JN.1_RBD deep scanning mutant library.

[0051] We constructed a deep saturation mutant library covering the SA55 binding region (K481-T531) of JN.1 RBD. Since the packaging titer of trVLP-ASV2-JN.1_RBD was low, which hindered the packaging of the library, trVLP-ASV1 was selected as the packaging backbone. The target region was divided into two blocks, each containing 25 amino acids. Two BsmBI restriction sites were introduced into each block, and an intermediate vector was pre-constructed. Primer libraries encoding all 19 possible amino acid substitutions within each block were synthesized, which were surrounded by BsmBI recognition sequences (GenScript). The mutant DNA library was amplified by PCR and cloned into the intermediate vector using Golden Gate assembly technology. Then, the structural protein sequence in the intermediate vector was excised using NotI and AvrII enzymes and inserted into the trVLP-ASV1 backbone vector. The resulting plasmid library was amplified in Escherichia coli Trelief5α (Tsingke, TSC-C01) and verified by next-generation sequencing (Novogene).

[0052] (5) Expression and purification of recombinant antibodies and receptor binding domains.

[0053] The light and heavy chains of the antibody were cloned into the pcDNA3.1 expression vector by Gibson assembly. For each antibody, 25 μg of the light chain plasmid and 25 μg of the heavy chain plasmid were co-transfected into 50 million Expi 293F cells. The cells were cultured in 250 mL shake flasks at 37°C, 5% CO2, and 160 rpm for 96 hours. After incubation, the cell supernatant was collected by centrifugation at 1000 rpm for 3 minutes, filtered through a 0.45 μm membrane (Millex®-HP SLHP033R), and the monoclonal antibody was purified using protein A / G beads (Sangon Biotech, C600981-0001) according to the manufacturer's instructions. The purified antibody was concentrated and buffer exchanged into RNase-free water using an ultrafiltration tube (Merck UFC9030). Antibody purity was analyzed by SDS-PAGE.

[0054] The codon-optimized SARS-CoV-2 RBD mutants were cloned into the pcDNA3.1 expression vector. Each vector contained a signal peptide at the N-terminus and a His-tag at the C-terminus. For each RBD variant, 50 μg of the plasmid was transfected into 50 million Expi293F cells using the PEI reagent. The cells were cultured in 250 mL shake flasks at 37°C, 5% CO2, and 160 rpm for 96 hours. Then, the cell supernatant was collected by centrifugation at 1000 rpm for 3 minutes, filtered through a 0.45 μm membrane (Millex®-HP SLHP033R), and the His-tagged RBD was purified using Ni-IDA resin (Sangon Biotech, C600811) according to the manufacturer's protocol. The purified RBD was then concentrated and buffer exchanged into RNase-free water using an ultrafiltration tube (Merck UFC9030). The purity of the RBD samples was verified by SDS-PAGE.

[0055] (6) Live cell imaging.

[0056] The BHK-21 host cells transduced with different trVLPs were imaged at the indicated time points using an EVOS® FL Auto Imaging System (Thermo Fisher Scientific, version 1.7). The system was equipped with an LPlanFL PH2 10x / 0.30 objective lens, and bright-field and fluorescence images (GFP or mCherry) were captured. Subsequently, the images were adjusted and cropped using ImageJ2 (https: / / imagej.net / software / imagej2 / ) for presentation.

[0057] (7) Flow cytometry.

[0058] In this study, flow cytometry was applied to two types of experiments. (1) Measuring the expression of fluorescent proteins (GFP or mCherry) by flow cytometry to evaluate trVLP-ASV infection. After treatment with trypsin, the cells were resuspended in flow buffer (DPBS + 5% FBS) to prepare a single-cell suspension. Then, the samples were analyzed using a BD LSRFortessa flow cytometer and BD FACSDiva 7.0 software. (2) Analyzing the binding of His-tagged RBD or its mutants to ACE2 expressed on the surface of BHK-21 cells by flow cytometry to evaluate the RBD-ACE2 interaction. Cells grown to approximately 80% confluence were trypsinized for 3 minutes and digestion was immediately stopped. The cells were collected by centrifugation and prepared as a single-cell suspension at a density of 10 7 cells / mL in the medium. A specific concentration of RBD or its mutants was added, and then gently rotated and incubated with an anti-His antibody (PE anti-6X His-tag® antibody, Abcam, ab237338, diluted 1:1000) at 4°C for 1 hour. Then the cells were centrifuged at 300×g for 3 minutes, washed once and resuspended in DPBS to prepare a single-cell suspension. These samples were analyzed using a similar method as above. The data were processed and analyzed using MATLAB scripts.

[0059] (8) trVLP-ASV RNA sequencing and mutation identification.

[0060] We performed trVLP-ASV RNA sequencing and mutation identification according to a protocol adapted from previous studies. Briefly, culture supernatants containing trVLP-ASV were collected, and RNA was extracted using the TRIzol LS kit (HaiGene, #B1901). Approximately 100 ng of RNA was reverse-transcribed into cDNA using the HiScript III cDNA Synthesis Kit (Vazyme, R312-02) and sequence-specific primers. Subsequently, using 5 µL of cDNA as a template, PCR amplification was performed using the KOD One PCR Mix in a 50-µL reaction volume. The PCR reaction was carried out for 28 cycles using primers designed for specific regions of the trVLP-ASV genome, such as the RBD or Cherry regions. The PCR products were sequenced using the FastNGS sequencing (NGS) service of Tsingke Biotechnology Company. Generally, sequencing coverage of several hundred times per nucleotide was sufficient to accurately identify mutations. The cleaned sequencing data was aligned to the reference genome using BWA (version 0.7.17-r1188). Mutations were identified using an in-house developed Python script. According to established empirical thresholds, mutations with a frequency exceeding 5% were considered significantly enriched and were prioritized for further analysis. The selected mutations were experimentally validated to confirm their functions.

[0061] (9)trVLP neutralization assay and identification of antibody escape mutations by trVLP evolution and DMS.

[0062] The trVLP-ASV neutralization assay was performed using a BHK-21 cell line stably expressing nsP4 and human ACE2. Twelve hours before infection, 3×10 5 cells per well were seeded in a 12-well plate to achieve a cell number of approximately 5×10 5 at the time of infection. trVLP-ASV with a multiplicity of infection of ~0.5 was added to 1 mL of medium together with gradient concentrations of antibody and incubated at 37 °C for 1 hour, then added to the cells. After incubation for 8 hours, the cells were collected and prepared into a single-cell suspension. The proportion of infected cells was determined by flow cytometry based on the mCherry fluorescence signal. The neutralization effect was evaluated by calculating the percentage of infected cells, and the half-maximal inhibitory concentration value was obtained accordingly. Due to the relatively low titer of trVLP-ASV2-JN.1_RBD, trVLP-ASV1-JN.1_RBD was used for the JN.1_RBD-related neutralization assay.

[0063] To simplify the identification of antibody escape mutants and based on the literature reporting that the antibody concentration in more than 90% of the culture supernatants expressing antibodies is higher than 1 μg / mL, we directly used the antibody supernatants for antibody-driven trVLP-ASV evolution experiments. Briefly, in one well of a 6-well plate, when HEK293T cells reached ~50% confluence, 3 μg of heavy chain and 3 μg of light chain plasmids were co-transfected into the cells using PEI transfection reagent. After 12 hours, the fresh medium was replaced. Four to five days later, the culture supernatant was collected, centrifuged at 500×g, filtered through a 0.45 μm filter, aliquoted, and stored at 4°C for up to 3 days for evolution experiments.

[0064] The evolution experiment for each antibody included two rounds of incubation processes ( Figure 4 Figure A in 5 . In the first round, 3×10 5 cells were seeded into one well of a 12-well plate 12 hours before infection. Subsequently, trVLP-ASVs (MOI = 0.5, approximately 50 - 100 μL of the supernatant of trVLP-ASV), 100 μL of the culture supernatant containing the antibody, and fresh medium were mixed to a final volume of 1.5 mL. The mixture was incubated at 37°C for one hour and then added to the cells. After 60 hours of incubation, the supernatant was collected for the second round of experiments. The cells were prepared according to the same protocol as the first round. Then, 100 μL of the first-round supernatant, 250 μL of the culture supernatant containing the antibody, and fresh medium were mixed to a total volume of 1.5 mL. The mixture was incubated at 37°C for 1 hour and then added to the cells. After another 60 hours of incubation, the supernatant was collected for RNA extraction, sequencing, and mutation analysis.

[0065] For the DMS library, 10 mL of medium containing 10 μg / mL SA55 antibody was co-incubated with 5 million trVLP-ASV viroid particles (~5000-fold coverage) for 1 hour and then added to the cells. After 60 hours, the supernatant was collected, RNA was extracted, reverse transcription was performed, and the deep saturation mutation region was amplified using specific primers, sequenced by NGS sequencing, and then the abundance of each mutant was counted, and the enrichment fold was calculated by dividing the number of mutants calculated from the RNA and plasmid libraries by the normalized library size.

[0066] (10) VSV pseudovirus preparation and neutralization experiment.

[0067] The preparation of SARS-CoV-2 pseudovirus followed the previously established experimental procedure. Briefly, 20 μg of the plasmid encoding the Spike protein mutant of SARS-CoV-2 was transfected into HEK293T cells in a 10 cm culture dish using PEI reagent. Meanwhile, 7.0×10 4Pseudotyped ΔG-luciferase (G*ΔG-luciferase) rVSV virus at TCID50 / mL was added to the medium. After 8 hours, the medium was removed, and the cells were washed three times with DPBS, then 10 mL of fresh medium was added. Subsequently, the cells were incubated for 48 hours. Then the culture supernatant was collected, centrifuged at 500×g for 3 minutes, concentrated using an ultrafiltration tube (Merck UFC9030), aliquoted, and stored at -80 °C for later use. Spike plasmids used for pseudovirus preparation included the strain D614G, mutants based on D614G (D614G / R408S, D614G / K444T, D614G / V455D, D614G / G446S, D614G / N450S, and D614G / N450T), Spike proteins from the JN.1 strain, and JN.1 Spike protein mutants (G504D and G504E). For the VSV-based neutralization assay, antibodies were serially diluted in 96-well plates and incubated with pseudovirus at 37 °C for 90 min. Subsequently, BHK-21 cells expressing ACE2 were added to the wells and incubated for 24 hours. After incubation, the supernatant was discarded, and luciferase luminescence was measured using a luciferase detection kit (YEASEN#11401ES76) according to the reference protocol provided by the manufacturer. Fluorescence was detected using a Varioskan LUX multimode microplate reader (Thermo Fisher). Each experiment was performed with at least two biological replicates.

[0068] (11) Identifying mutations that promote cross-species transmission through trVLP evolution.

[0069] We used serial passage to study the adaptive evolution of trVLP-ASV in BHK-21 cells (expressing nsP4) that were engineered to display ACE2 proteins from nine species. Twelve hours before each experiment, 3×10 5 cells were seeded in one well of a 12-well plate. In the first round of the experiment, trVLP-ASV was added at an MOI of approximately 0.5, and fresh medium was added to adjust the total volume to 1.5 mL. After 3 days of incubation, the supernatant was collected for the next round of passage. For high-affinity RBD-ACE2 combinations (e.g., Ref_RBD with human ACE2), 100 μL of the supernatant was transferred to a fresh well, and 1.4 mL of fresh medium was added. For low-affinity RBD-ACE2 combinations (e.g., Ref_RBD with koala ACE2), 300 μL of the supernatant was used, and the total volume was adjusted to 1.5 mL with fresh medium. Each trVLP-ASV-ACE2 combination was passaged for five rounds, with each round lasting 3 days. After the fifth round, the supernatant was collected for RNA extraction, sequencing, and mutation analysis.

[0070] (12) Quantitative and statistical analysis.

[0071] For the VSV pseudovirus neutralization assays, such as the trVLP-ASV neutralization assay, trVLP-ASV evolution assay, and DMS evolution assay, two biological replicates were performed. The half-maximal inhibitory concentration values were calculated using the MATLAB nlinfit function with a four-parameter logistic model. To evaluate the significant differences between samples, a two-tailed two-sample t-test was performed using the MATLAB ttest2 function.

[0072] The results obtained in this example according to the above method are as follows: (1) Construct a virus-like particle (trVLP-ASV) system with transcriptional and replicative capabilities.

[0073] We aimed to construct a safe virus platform that could rapidly and easily package virus-like particles using recombinant DNA. The platform should also incorporate directed evolution and deep mutational scanning to analyze the effects of mutations in the Spike protein on its function. To this end, we chose to modify Sindbis virus, a positive-strand RNA virus belonging to the genus Alphavirus and commonly regarded as a biosafety level 2 (BSL-2) model virus. Sindbis virus has a relatively short genome (~11.7 kb), making it highly suitable for recombinant DNA-based packaging. Inspired by earlier studies that developed epidermal growth factor receptor (EGFR)-specific virus 1 (ESV1) by modifying the E2 protein of Sindbis virus, we sought to redirect the modified Sindbis virus ESV1 to utilize ACE2 as a receptor.

[0074] We first focused on enhancing the safety of the system ( Figure 1 Panel A in). Although Sindbis virus belongs to the BSL-2 class of pathogens, its unpredictable evolutionary properties still require further enhancement of safety. Our previous studies have demonstrated that the Sindbis RNA genome lacking nsP4 can still replicate and transcribe in cells expressing nsP4. Therefore, we hypothesized that ESV1 would also amplify in cells providing nsP4 after deletion of nsP4. To test this, we constructed two virus RNA genomes derived from ESV1: one similar to ESV1 but carrying a GFP reporter gene, and the other with nsP4 deleted ( Figure 1 Panel A in). These RNAs were expressed from DNA vectors driven by the CMV promoter.

[0075] Since the nsP4 expression vector integrated into the host genome shares partial sequences with its viral RNA genome (including the ΔnsP3 region, which retains the first amino acid of nsP4 after protease cleavage, and the nsP4 C-terminal region overlapping with the subgenomic promoter), we codon-optimized the shared sequences of the nsP4 vector. This significantly reduced sequence similarity, thereby minimizing the risk of RNA recombination. Subsequently, we transfected these ESV1-derived vectors into BHK-21 cells stably expressing codon-optimized nsP4. Two days later, we collected the supernatant and used it to infect BHK-21 cells expressing or not expressing nsP4. As expected, the ESV1-derived RNA genome lacking nsP4 successfully replicated in BHK-21 cells expressing nsP4, and the generated particles only infected BHK-21 cells expressing nsP4. In contrast, the particles generated from the complete ESV1 RNA genome could infect BHK-21 cells regardless of whether nsP4 was expressed ( Figure 1 in Panels B, C, and D of

[0076] Since the nsP4-deficient virus-like particles can only maintain transcription and replication in specific host cells, we named them trVLP-ESV1 according to the literature convention. Subsequently, we measured the titers of ESV1 (i.e., containing nsP4) and trVLP-ESV1 (i.e., lacking nsP4) and found that the viral titers of both were comparable ( Figure 1 in Panel E of Figure 1 ), indicating that trVLP-ESV1 can infect specific host cells as effectively as wild-type virus particles. Next, we evaluated the stability of trVLP-ESV1 and its potential for viral recombination (with the nsP4 gene carried in the host) by serially passaging trVLP-ESV1. Even after five passages (10 days), these trVLPs still maintained high infectivity only in host cells expressing nsP4 ( Figure 7 in Panel B of

[0077] This finding is consistent with the phenotype of the lack of recombination events, where RNA recombination between viral RNA and nsP4 allows infection of cells lacking nsP4 expression.

[0078] Therefore, by creating trVLP-ESV1-like virus particles lacking the nsP4 gene, which can only replicate and amplify in cells expressing nsP4, we improved the safety of the engineered Sindbis virus vector platform. In addition, these particles maintained stable infectivity even after multiple passages.

[0079] Next, we set out to modify the E2 protein of trVLP-ESV1 so that it can use ACE2 as the receptor to enter host cells. Since the receptor-binding domain (RBD) of the Spike protein of SARS-CoV-2 can not only bind to ACE2 but also serves as the target of approximately 90% of neutralizing antibodies, we chose to display the RBD of Spike on the E2 protein. Compared with EGF in trVLP-ESV1 (EGF consists of only 53 amino acids and can be relatively easily displayed on E2), the Spike RBD is approximately 200 amino acids long, which may be more challenging to integrate. To reduce steric hindrance and minimize the binding ability of E2 to its native receptor, we deleted the EGF and part of the receptor-binding region of E2 itself, and then inserted the Spike RBD through two linker sequences ( Figure 2 Panel A in

[0080] ). The RBD sequence is derived from the Wuhan strain of SARS-CoV-2. We named this modified construct trVLP-ASV1-Ref_RBD, where ASV represents the ACE2-targeting Sindbis Virus.

[0080] To characterize this modified trVLP-ASV, we transfected the plasmid expressing trVLP-ASV1-Ref_RBD driven by CMV into BHK-21 cells stably expressing nsP4 and human ACE2. The supernatant was collected two days later and then used to infect BHK-21 cells stably expressing nsP4 (i.e., host cells), with or without ACE2. As expected, we observed that the infection rate of host cells expressing ACE2 was significantly higher than that of cells not expressing ACE2 at 8 hours post-infection (primary infection) ( Figure 2 Panel B in Figure 8 Panels A and B in ). However, at 24 hours post-infection (primary infection plus secondary infection after trVLP-ASV amplification), we noticed a significant amplification of trVLP-ASV in cells without ACE2 ( Figure 2 Panels B and D in ). We speculated that this low specificity might be due to the entry of trVLP-ASV mediated by the remaining binding ability of the Sindbis structural protein to its own receptor.

[0081] (3) The interaction between the Sindbis structural protein and the remaining part of its own receptor can be eliminated by additional mutations in the structural protein.

[0082] To verify this, we introduced multiple mutations into the E3 and E2 structural proteins, which were reported to disrupt their binding to their own receptors. We named this modified version trVLP-ASV2-Ref_RBD ( Figure 2Figure A). Infection experiments showed that this new version (designated V2) exhibited lower ACE2-independent entry and higher host specificity ( Figure 2 Figures C, D, and E). Similar conclusions were also drawn by measuring the trVLP-ASV titers ( Figure 2 Figure F). Specifically, for V1, the amplified trVLP-ASV titer was approximately 8.6-fold higher in cells expressing ACE2 than in cells expressing ACE2-negative, while the fold change for V2 increased to 24.6-fold with only a slight decrease in the overall titer.

[0083] Having achieved highly specific ACE2-receptor-mediated infection with V2, we next investigated the generality of this system for other SARS-CoV-2 variant RBDs, including XBB.1.5, HK.3, and JN.1, as well as the RBD of SARS-CoV-1. We replaced Ref_RBD in V2 with these four different RBDs and measured the infection rate 24 hours post-infection ( Figure 2 Figures G, H; Figure 8 Figures C and D). Notably, all the resulting VLPs exhibited high infection specificity, indicating that each RBD variant on E2 could effectively mediate ACE2-dependent trVLP-ASV host entry.

[0084] After successfully achieving highly specific ACE2-mediated trVLP-ASV infection, we further investigated the applicability of this system for other SARS-CoV-2 variant RBDs, including XBB.1.5, HK.3, and JN.1, as well as the RBD of SARS-CoV-1. We replaced Ref_RBD in V2 with these four different RBDs and measured the infection rate 24 hours post-infection ( Figure 2 Figures G and H; Figure 8 Figures C and D). Notably, all the generated VLPs showed high infection specificity, indicating that each RBD variant on E2 could effectively mediate ACE2-dependent host entry of trVLPs.

[0085] In summary, we successfully developed a new trVLP-ASV platform - trVLP-ASV2-Ref_RBD, which displays the spike RBD of SARS-CoV-2 on the E2 protein to mediate ACE2-dependent host invasion. This platform has a high degree of infection specificity and can be widely applied to a variety of SARS-CoV-2 RBD variants and SARS-CoV-1 RBD.

[0086] (4) Using trVLP-ASV2 to simulate the immune escape of SARS-CoV-2 neutralizing antibodies.

[0087] After determining that Spike RBD-ACE2 interaction mediated the majority of cell entry of trVLP-ASV2, we next investigated how neutralizing antibodies affected this process and whether our trVLP-ASV system could be used to mimic the antibody escape process of SARS-CoV-2.

[0088] To answer these questions, we first evaluated the neutralization effects of three antibodies - LY-CoV1404, REGN10987, and SA55 - on trVLP-ASV based on Ref_RBD or the RBD of the most recent JN.1 strain ( Figure 3 Panel A in [[ ]]). LY-CoV1404 and REGN10987 have received Emergency Use Authorization from the US Food and Drug Administration (FDA) for use in COVID-19 treatment regimens, and SA55 was approved for clinical trials in China during the epidemic. Interestingly, we found that LY-CoV1404 could effectively neutralize trVLP-ASV based on Ref_RBD, but its neutralization ability against trVLP-ASV based on JN.1_RBD ( Figure 3 Panels B and C in [[ ]]) was reduced by approximately eight-fold. Previous studies have shown that LY-CoV1404 targets RBD residues V445 and G446, which are affected by immune escape through the V445P / G446S mutations identified in the XBB variant. However, JN.1_RBD carries V445H and G446S mutations, which may explain the reduced neutralization efficiency of LY-CoV1404 against trVLP-ASV based on JN.1_RBD. As for REGN10987, it showed potent neutralization ability against trVLP based on Ref_RBD but completely lost its neutralization against trVLP based on JN.1_RBD ( Figure 3 Panels B and C in [[ ]]), which is consistent with previous reports - probably attributed to the mutations of V445H and G446S. In contrast, as shown in previous studies, SA55 showed a high degree of neutralization ability against trVLP displaying either Ref_RBD or JN.1_RBD ( Figure 3 Panels B and C in [[ ]]).

[0089] To validate the neutralization results based on trVLP-ASV, we used the VSV pseudovirus assay to evaluate the neutralization potency of these antibodies. We found that the VSV pseudovirus assay reproduced the results of the trVLP-ASV-based assay, and a consistent trend was observed when comparing the original strain and the JN.1 strain ( Figure 3 Panel D in [[ ]]). Collectively, these results highlight the applicability of our trVLP-ASV system for mimicking antibody-mediated inhibition of virus entry.

[0090] Next, we sought to evaluate whether the system could mimic antibody escape driven by viral mutations. As a prerequisite for evolutionary studies, we first estimated the mutation rate of our trVLP-ASV system. To avoid the influence of purifying selection on the non-structural protein region and structural protein region of the RNA genome, we measured the mutation rate of the mCherry reporter gene ( Figure 9 Panel A in Figure 9 ), and its mutation rate was approximately 0.021 bp / kb / d (

[0091] Panel B in Figure 9 ). Although this rate is lower than the mutation rate of Sindbis virus reported previously, it is still sufficient to generate a large mutant library at high trVLP-ASV titers.

[0092] Furthermore, previous studies have shown that in approximately 90% of cases, the antibody concentration in the cell supernatant expressing the antibody can exceed 1 μg / mL. Taking advantage of this, we tested the neutralizing ability of the unpurified supernatant to simplify the experimental procedure and reduce the labor intensity of antibody purification. We selected 10 antibodies against six different epitope regions on the RBD: LY-CoV 016 (Group A), BRII-196 (Group A), AZD8895 (Group B), REGN10933 (Group B), DXP-593 (Group C), LY-CoV555 (Group C), REGN10987 (Group D), LY-CoV1404 (Group D), VIR-7831 (Group E), and SA55 (Group F). Neutralization assays using the antibody-containing supernatant showed that all tested antibodies could strongly neutralize trVLP-ASV2-Ref_RBD ( Figure 9 Panel C in

[0092] ), further highlighting the strong potency of these antibodies even in unpurified form. Figure 4 Therefore, we conducted antibody-driven trVLP-ASV evolution experiments using these antibody supernatants, following a standardized protocol ( Figure 4 Panel A in ). In this experiment, we tested all 10 antibodies against trVLP-ASV2-Ref_RBD. We also evaluated the potency of SA55 against trVLP-ASV2-JN.1_RBD, as it is one of the few antibodies that still retains neutralizing activity against the recent JN.1 strain. After two rounds of directed evolution, we collected the supernatant, extracted RNA from trVLP-ASV, reverse-transcribed the RNA, and specifically amplified the RBD region for next-generation sequencing and bioinformatics analysis. A mutation was considered enriched if it exceeded 5% of the total bases at a given locus. According to this definition, nine out of ten antibodies were found to have unique enriched mutations that did not appear in the control group (i.e., trVLP-ASV RNA incubated with supernatant lacking the corresponding antibody) (Figure 4 Figures B and C in Figure 10 Figure A in

[0093] Many of these enriched mutations coincide with potential escape mutations inferred previously based on yeast display, including N460K and N487K (LY-CoV016), L455S (BRII-196), F486S (AZD 8895 and REGN10933), V483A and E484K (DXP-593), Q493R and Q493K (LY-CoV555), K444T, V445D and G446S (REGN10987), and K444T and V445D (LY-CoV1404). The G504D mutation identified in SA55-treated trVLP-ASV2-Ref_RBD was also supported by other studies. These findings suggest that these mutations may enable trVLP-ASV2-Ref_RBD to escape immune neutralization. Notably, the antibody VIR-7831 effectively blocked the entry of trVLP-ASV but did not enrich for effective mutations, possibly because its epitope is close to the N-terminus of the RBD and far from the ACE2 binding motif.

[0094] To further validate these observations, we selected two antibodies with a relatively large number of enriched mutations, REGN10987 (6 mutations) and LY-CoV1404 (3 mutations), for validation. We introduced the mutations enriched by these antibodies into the Spike D614G mutant based on the VSV-ΔG system to construct pseudoviruses and evaluated the impact of these mutations on antibody escape. For REGN10987, five out of six mutations (three of which were previously inferred by yeast display) showed significant antibody escape ( Figure 4 Figure D in Figure 4 Figure E in

[0095] For SA55, we performed evolutionary experiments using trVLP-ASV2-JN.1_RBD and identified two escape mutations, G502D and G504D. To go beyond the mutational scope provided by directed evolution, we constructed a deep saturation mutagenesis library covering residues K481–T531 in JN.1_RBD, which contain the binding interface for SA55. After a single round of SA55 antibody treatment, we found numerous enriched mutations concentrated in the Y501–H505 region (Figure 4F and G; Figure 10B), overlapping with the major binding interface of SA55 and JN.1_RBD. Among them, G502D and G504D were also detected by the previously mentioned directed evolution analysis, while G504D and G504E were the most prominent mutations. Notably, Y501D, G502E, V503E, G504D, G504E, and H505E were also observed in yeast display or rVSV-based evolutionary experiments. Some mutations, such as G504E, require changes in two consecutive bases, highlighting the advantage of deep saturation mutagenesis in identifying such variants. Validation using VSV-based pseudoviruses confirmed that both G504D and G504E completely escaped SA55 neutralization (Figure 4G). These results highlight the powerful ability of our trVLP system in simulating and predicting antibody escape mutations.

[0096] (5) Simulating SARS-CoV-2 cross-species transmission mediated by RBD mutations.

[0097] After demonstrating the versatility of the trVLP-ASV system, we further explored whether this system could also be used to analyze the interactions between RBD and ACE2 homologs from different species, thereby providing potential insights into the cross-species transmission of SARS-CoV-2.

[0098] For this purpose, we overexpressed ACE2 proteins from nine different species in host cells and infected these cells with trVLP-ASV2-Ref_RBD, trVLP-ASV2-XBB.1.5_RBD, and trVLP-ASV2-JN.1_RBD ( Figure 5 Figure A and B; Figure 11Figure A). We measured the infection efficiency of each trVLP at 8 hours post-infection. Since previous studies have quantified the infection efficiency data of the original SARS-CoV-2 strain (carrying Ref_RBD) in cells expressing ACE2 proteins of these species, we further evaluated whether the infection efficiency of trVLP-ASV2-Ref_RBD was consistent with these data. Notably, we observed a strong linear correlation, indicating that trVLP-ASV can reliably reflect the performance of SARS-CoV-2 in terms of cross-species receptor-binding potential ( Figure 5 Figures C and D).

[0099] Next, we compared the infection efficiency of trVLP-ASV displaying three different RBDs in host cells expressing various ACE2 receptors. Previous studies have shown that although the original SARS-CoV-2 strain was unable to infect mice, later Omicron lineages acquired this ability. Based on this, we examined the infection efficiency of trVLPs displaying the RBDs of the original Ref_RBD or Omicron variants XBB.1.5 and JN.1 in mouse ACE2-overexpressing host cells. We found that trVLPs displaying Ref_RBD were unable to effectively infect these mouse ACE2-overexpressing host cells, while trVLPs displaying XBB.1.5_RBD and JN.1_RBD showed strong infectivity ( Figure 5 Figure C; Figure 11 Figure B). These results confirmed previous findings and highlighted the application value of the trVLP system in studying the host range of SARS-CoV-2.

[0100] In addition, we observed a significant increase in the infection efficiency of trVLP-ASV2-XBB.1.5_RBD in cells expressing ACE2 from cynomolgus monkeys, rhesus macaques, and squirrel monkeys. Mechanistically, these three ACE2 homologous proteins share mutations in the RBD-binding region, which may explain the stronger binding affinity of XBB.1.5 RBD for these ACE2 variants ( Figure 5 Figure B). Since previous studies have not examined these interactions, we verified our findings by measuring the binding affinity between host cells expressing these ACE2s and purified Spike RBD. As a positive control, cells expressing human ACE2 showed similar binding ability for Ref_RBD and XBB.1.5_RBD at 10 μg / ml. Then, we compared the binding ability of the two RBDs in cells expressing ACE2 from cynomolgus monkeys, rhesus macaques, and squirrel monkeys. As expected, we found that XBB.1.5_RBD showed significantly higher binding ability than Ref_RBD on these three ACE2s ( Figure 5(Figure E). Overall, these observations are consistent with the view that the trVLP system is a powerful platform for analyzing cross-species RBD-ACE2 interactions.

[0101] Given the success of these experiments, we next explored how to utilize the evolutionary capacity of trVLP-ASV to mimic RBD mutation-driven viral cross-species transmission. We designed a serial evolution experiment in which trVLP-ASV was serially passaged in cells expressing ACE2 from different species. We reasoned that in the case of poor binding of ACE2 orthologs to the RBD, cell invasion by trVLP-ASV would rely on the basal interaction between E2 and its own receptor, resulting in relatively slow amplification of viral particles. In this scenario, any RBD mutation that promotes ACE2 binding would allow the mutant to expand faster and eventually become enriched in the population, which could then be identified by sequencing ( Figure 6 Figure A).

[0102] To systematically mimic cross-species transmission, we performed five rounds of serial evolution experiments with trVLP-ASV2-Ref_RBD in host cells expressing ACE2 from nine species. Overall, no enriched mutations were observed in three species, while one to four enriched mutations (some with frequencies as high as 99%) were identified in the other six ACE2s ( Figure 6 Figures B and C; Figure 12 Figure A). Importantly, we found that the enrichment of mutations was closely correlated with the binding affinity of ACE2-Ref_RBD. Specifically, ACE2 from humans, mandrills, and pigs showed high affinity for Ref_RBD, while ACE2 from the other six species showed low or almost no affinity.

[0103] Next, we investigated the functions of the enriched mutations during evolution. Some of these mutations, such as Q493R and Q498H, have been previously reported to alter the host range of SARS-CoV-2, further confirming the ability of our trVLP-ASV system to capture important RBD mutations involved in cross-species transmission. To explore the mechanism of these enriched mutations, we purified the two most frequently enriched mutations in Ref_RBD (i.e., Q498H and Y449H) and tested their binding ability to ACE2. As a negative control, the Q498H mutation had no significant effect on the binding of Ref_RBD of human ACE2 at 10 μg / ml, which is consistent with the lack of enriched mutations in the evolution experiments using human ACE2. Similarly, koala ACE2 served as an additional negative control, as the Q498H mutation did not change its binding ability, which is consistent with the lack of Q498H enrichment during evolution. However, Q498H significantly enhanced the binding ability of Ref_RBD to ACE2 from five other species, including marmosets, capuchins, squirrel monkeys, mice, and ferrets ( Figure 6 It is worth noting that among these five ACE2s, Q498H was enriched in the evolution experiment ( Figure 6 In contrast, the Y449H mutation (previously identified in the C.1.2 variant and thought to be associated with antibody escape51) did not show a meaningful change in ACE2 affinity ( Figure 6 The presence of β-catenin in the mitochondria is not an artifact of mitochondrial expansion (D), which may reflect the limited dynamic range of the assay or alternative adaptive evolutionary mechanisms not driven by protein-protein interactions.

[0104] Thus, we have demonstrated the utility of trVLP-ASV in modeling cross-species viral transmission mediated by RBD mutations and identified enriched mutations associated with cross-species transmission. These findings highlight the potential of trVLP-ASV as a platform for studying the dynamics of viral transmission between different hosts.

[0105] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A Sindbis virus-like particle trVLP-ASV with transcription and replication capabilities, characterized in that, Its genome does not contain the RNA replicase nsP4 gene; and its E2 protein contains the receptor-binding domain of the Spike protein of SARS-CoV-2.

2. The Sindbis virus-like particle trVLP-ASV with transcription and replication capabilities according to claim 1, characterized in that, The method for making the genome not contain the nsP4 gene is: deleting the base sequence of nsP4 at positions 5763 - 7358 of the genome.

3. The Sindbis virus-like particle trVLP-ASV with transcription and replication capabilities according to claim 1, characterized in that, The method for making the E2 protein contain the receptor-binding domain of the Spike protein is: inserting the receptor-binding domain of Spike of SARS-CoV-2 and its variants; and / or inserting the receptor-binding domain of Spike of SARS-CoV-1.

4. The Sindbis virus-like particle trVLP-ASV with transcription and replication capabilities according to any one of claims 1-3, characterized in that, It replicates and amplifies in host cells that express nsP4 in trans.

5. The Sindbis virus-like particle trVLP-ASV with transcription and replication capabilities according to any one of claims 1-3, characterized in that, It uses ACE2 as the main receptor.

6. The Sindbis virus-like particle trVLP-ASV with transcription and replication capabilities according to any one of claims 1-3, characterized in that, Its invasion of host cells is inhibited by antibodies against the receptor-binding domain of SARS-CoV-2.

7. Method for constructing Sindbis virus-like particle trVLP-ASV with transcription and replication ability according to any one of claims 1-6, characterized in that, The receptor-binding domain of Spike and its mutants or mutant libraries are inserted into the E2 region of the Sindbis virus genome by means of gibson assembly.

8. A method for preparing the Sindbis virus-like particle trVLP-ASV with transcription and replication capabilities according to any one of claims 1-6, characterized in that, After transfecting plasmids into host cells expressing ACE2 and nsP4, the RNA transcription of trVLP-ASV is driven by the CMV promoter, and then the relevant proteins are translated and assembled into virus-like particles.

9. Use of the Sindbis virus-like particle trVLP-ASV with transcriptional and replicative ability according to any one of claims 1 - 6 in the construction of a SARS-CoV-2 immune escape model.

10. Use of the Sindbis virus-like particle trVLP-ASV with transcriptional and replicative ability according to any one of claims 1 - 6 in the study of the host range of SARS-CoV-2.

Images