Construction method of multi-epitope subunit vaccine of norovirus

Abstract

The invention relates to a construction method of a multi-epitope subunit vaccine of norovirus. The method comprises the following steps: determining a standardized construction principle and a design parameter system of the multi-epitope subunit vaccine by designing four different multi-epitope construction strategies; the method solves the ubiquitous problems of single construction mode, strong immunological enhancement dependence, insufficient antigen spectrum coverage and lack of a high-throughput screening system in the existing multi-epitope vaccine design, significantly shortens the vaccine design period, reduces the experiment cost, and improves the design accuracy. The multi-epitope subunit vaccine of the norovirus is obtained through the construction method, the amino acid sequence of the multi-epitope subunit vaccine is shown as any one of SEQ ID NO.3-5, and the strong immunogenicity and broad-spectrum protection potential of the multi-epitope subunit vaccine are verified through animal experiments.

Description

Technical Field

[0001] This invention relates to the field of vaccine construction technology, and in particular to a method for constructing a multi-epitope subunit vaccine for norovirus. Background Technology

[0002] Norovirus is one of the leading pathogens causing acute gastroenteritis worldwide. Currently, several norovirus vaccine candidates have entered clinical trials, but most employ design technologies such as virus-like particle vaccines, adenovirus vector vaccines, subunit recombinant vaccines, and DNA or mRNA vaccines. Due to the frequent mutations in the norovirus genome, its diverse genotypes (10 known gene groups, 49 genotypes), and the lack of effective in vitro culture systems, traditional vaccine development faces significant challenges.

[0003] Multi-epitope subunit vaccines, as a next-generation vaccine development strategy, are gradually gaining attention. These vaccines integrate multiple T-cell and B-cell epitope fragments to mimic the immune response mechanism of natural antigens, offering advantages such as flexible design, structural stability, high safety, simple manufacturing processes, and the ability to be customized for different populations. However, existing multi-epitope vaccine designs generally suffer from the following technical bottlenecks: Limited construction methods: Multi-epitope sequences are typically constructed in a linear tandem configuration, lacking a systematic design for the connection with adjuvants; the impact of different connection sequences on spatial structure and immunogenicity has not been fully assessed; Strong dependence on immune enhancement: Traditional peptide vaccines have weak immunogenicity and must rely on exogenous adjuvants for enhancement; the mechanism by which the selection and arrangement of adjuvants affect vaccine stability remains unclear; Insufficient antigen coverage: Most studies predict epitopes based on only a small number of strains, without systematically evaluating their HLA diversity coverage and conservation in the population; Lack of high-throughput screening systems: Existing designs are mostly based on single-model construction, lacking systematic comparison and optimization of different construction strategies. Summary of the Invention

[0004] Objective of the Invention: The technical problem to be solved by the present invention is to address the shortcomings of existing technologies by providing a method for constructing a multi-epitope subunit vaccine for norovirus. Specifically, the present invention provides a method for constructing a multi-epitope subunit vaccine for norovirus based on multiple construction strategies, sufficient antigen spectrum coverage, and a high-throughput screening system.

[0005] To address the aforementioned technical problems, this invention discloses a method for constructing a multi-epitope subunit vaccine against norovirus. The specific technical solution is as follows: In a first aspect, the present invention provides a norovirus multi-epitope subunit vaccine, the amino acid sequence of which is shown in any one of SEQ ID NO.3 to 5.

[0006] In a second aspect, the present invention provides a gene encoding a multi-epitope subunit vaccine for norovirus as described in the first aspect.

[0007] The nucleotide sequence of the gene is shown in any one of SEQ ID NO. 6 to 8.

[0008] Thirdly, the present invention provides the application of the norovirus multi-epitope subunit vaccine described in the first aspect in the preparation of drugs for the prevention of norovirus.

[0009] Fourthly, the present invention provides a method for constructing a multi-epitope subunit vaccine against norovirus, comprising the following steps: Step 1: Obtain and screen proteomic data of norovirus gene groups; In some embodiments of the present invention, the norovirus gene groups include norovirus GI, GII and GIV gene groups, and the proteomic data for screening norovirus gene groups are obtained through homology screening, transmembrane helix number prediction, antigenicity prediction and sensitization prediction. Step 2: Immune cell epitope prediction; Step 3: Multi-epitope subunit vaccines were constructed using the full epitope splicing construction strategy, the optimal epitope screening construction strategy, the epitope duplication enhancement construction strategy, and the heat shock protein adjuvant substitution construction strategy, respectively. Step 4: Screen the multi-epitope subunit vaccine constructed in Step 3 through property prediction, vaccine immunization simulation, structure prediction, molecular docking analysis, molecular dynamics analysis, and immunological evaluation. In some embodiments of the present invention, the property prediction includes antigenicity, sensitization, and physicochemical property prediction; the structure prediction includes secondary and tertiary structure prediction and verification and refinement of structural conformation rationality; the molecular docking analysis includes molecular docking and interaction analysis with TLR3 (PDB ID: 7C76).

[0010] In step two, the immune cell epitope prediction includes prediction of cytotoxic T lymphocyte (CTL) epitopes and helper T lymphocyte (HTL) epitopes based on human leukocyte antigen (HLA) restriction. The immune cell epitope prediction also includes prediction of linear B cell (LBL) epitopes. In some embodiments of the present invention, the human leukocyte antigen (HLA) restriction is selected from the following HLA genotypes: MHC-I class: HLA-A*02:01, HLA-A*24:02, HLA-A*11:01, HLA-C*04:01, HLA-C*03:04, and HLA-B*56:01. Preferably, CTL epitopes are predicted based on MHC-I class HLA genotype restriction. MHC-II class: HLA-DRB1*15:01, HLA-DRB1*03:01, HLA-DQA1*05:01, and HLA-DQA1*03:01. Preferably, HTL epitopes are predicted based on MHC-II class HLA genotype restriction; The predicted immune cell epitopes include 5 CTL epitopes, 2 HTL epitopes, and 2 LBL epitopes, which are as follows: Five CTL tabletops: TYPGEQILF, NIIDPWIMK, TLEPIFIPV, LMAGNAFTA, VLMAGNAFT; Two HTL tabletops: PVAGGAIAA and FTAGKVIFA; Two LBL epitopes: FFRSYIPLKGGFGNTA and LKGGFGNTAI.

[0011] In step three, the multi-epitope subunit vaccine includes an adjuvant, an immune cell epitope, and a histidine tag, all of which are linked by a linker peptide.

[0012] In the aforementioned full epitope splicing construction strategy, the adjuvants include TLR7 agonists, TLR3 agonists, and PADRE, and the immune cell epitopes are all the immune cell epitopes predicted in step two. In some embodiments of the present invention, the full epitope splicing construction strategy uses a linker peptide to linearly and sequentially tandemly connect TLR7 agonists, PADRE immune adjuvants, CTL epitope regions, HTL epitope regions, LBL epitope regions, TLR3 agonists, and histidine His tags from the N-terminus to the C-terminus. The epitopes in the CTL, HTL, and LBL epitope regions are arranged and combined in their respective regions, and the positions of TLR3 and TLR7 agonists are interchanged to form multiple candidate vaccine sequences.

[0013] In the optimal epitope screening and construction strategy, the adjuvants include TLR7 agonists, TLR3 agonists, and PADRE. The immune cell epitopes are selected from each type of immune cell epitope predicted in step two, and the epitope with the best overall performance is combined. In some embodiments of the present invention, in the optimal epitope screening and construction strategy, the optimal CTL epitope, optimal HTL epitope, and optimal LBL epitope with optimal conditions are retained in the CTL epitope region, HTL epitope region, and LBL epitope region, respectively. The positions of the optimal CTL epitope, optimal HTL epitope, and optimal LBL epitope are arbitrarily arranged and combined. By swapping the positions of TLR3 agonists and TLR7 agonists, multiple sets of candidate vaccine sequences are formed. The optimal conditions are: the best epitope is the one that combines high affinity (percentile rank less than 0.5), good hydrophilicity (the best among various types of epitopes, the higher the positive value, the higher the hydrophilicity), and population coverage (covering the above-mentioned HLA genotype-restricted individuals).

[0014] In the epitope repetition enhancement construction strategy, the adjuvants include TLR7 agonists, TLR3 agonists, and PADRE, and the immune cell epitope is selected by repeating the predicted immune cell epitope from step two six times consecutively. In some embodiments of the present invention, the epitope repetition enhancement construction strategy retains and repeats each of the optimal CTL epitope, optimal HTL epitope, and optimal LBL epitope six times consecutively to form multiple sets of candidate sequences.

[0015] In the heat shock protein adjuvant replacement construction strategy, the adjuvant includes amino acid fragments 407-426 of HSP65 and HSP70, and PADRE. The immune cell epitope is selected by repeating the predicted immune cell epitope from step two six times consecutively. In some embodiments of the present invention, in the heat shock protein adjuvant replacement construction strategy, based on the epitope repetition enhancement construction strategy, one end of the TLR7 agonist is replaced with heat shock protein HSP65, and the other end of the TLR3 agonist is replaced with amino acid fragments 407-426 of HSP70. The replaced HSP70 amino acid fragments 407-426 are repeated twice consecutively to form multiple candidate vaccine sequences. The amino acid sequence of HSP65 is shown in SEQ ID NO.1, and the amino acid sequence of the HSP70 407-426 amino acid fragment is shown in SEQ ID NO.2.

[0016] The linker peptides include any one or more combinations of EAAAK, RVRR, AAY, GGPPG, and KK.

[0017] The TLR7 agonist has a Unipol of Q5U7J2, the TLR3 agonist has a Unipol of Q74EI3, and the amino acid sequence of PADRE is AKFVAAWTLKAAA.

[0018] Fifthly, the present invention provides a multi-epitope subunit vaccine for norovirus constructed using the construction method described in the fourth aspect.

[0019] Beneficial effects: 1. This invention significantly improves vaccine construction efficiency and immunogenicity through systematic optimization of multiple construction strategies. By designing four different multi-epitope construction strategies, the effects of epitope arrangement order, repetitive sequence pattern, adjuvant type, and connection method on vaccine physicochemical properties and immunogenicity were systematically compared. Among these, repetitive epitope design and diverse adjuvant combinations significantly enhanced antigenicity and immune response strength. Antibody levels induced in vivo by vaccine candidates obtained through immune response prediction were more than twice that induced by traditional linear construction. This invention thus establishes standardized construction principles and design parameter systems for multi-epitope subunit vaccines. 2. Animal experiments validated the strong immunogenicity and broad-spectrum protective potential of the vaccine: This invention successfully expressed and purified three representative vaccine candidates (NV5, NV1, and NV4) using the E. coli system. Mouse immunization experiments showed that NV5 induced the highest level of antigen-specific antibody titers; NV1 significantly promoted IFN-γ secretion, exhibiting excellent cellular immune activation capacity; and NV4 produced a balanced IgM / IgG immune profile, possessing rapid response and long-lasting immunity characteristics. All three induced cross-recognition antibodies against real GII.4 virus-like particles, demonstrating their potential broad-spectrum protective effect. This result is the first time that computationally designed multi-epitope vaccines can achieve synergistic humoral and cellular immune responses in animal models. 3. Significantly Enhanced Vaccine Stability and Immune Initiation Potential: Through molecular docking and 150 ns molecular dynamics simulation analysis, the stability of the binding interface between the vaccine and the TLR3 receptor was optimized, and its conformational stability under physiological conditions was verified. A stable complex structure helps enhance innate immune recognition and promotes dendritic cell activation, thereby achieving more efficient immune initiation. This invention thus achieves a leap from the sequence level to the structural level in vaccine molecular design. 4. Constructing a transferable AI-assisted vaccine design paradigm. By integrating sequence screening, epitope prediction, structure optimization, and immune simulation, a closed-loop validation process from data-driven to empirical immune testing was achieved. This method can significantly shorten the vaccine design cycle, reduce experimental costs, and improve design accuracy, providing a generalized technical framework for AI-driven vaccine development. Attached Figure Description

[0020] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments, and the advantages of the present invention in the above and / or other aspects will become clearer.

[0021] Figure 1 The diagram shows a multi-epitope vaccine, where a represents strategy one, b represents strategy two, c represents strategy three, and d represents strategy four.

[0022] Figure 2 The three-level structure of vaccines NV6 and NV1 is comprehensively validated, where a represents NV6 and b represents NV1.

[0023] Figure 3 The three-level structure of the vaccine NV4 and NV5 is comprehensively validated, where a represents NV4 and b represents NV5.

[0024] Figure 4 The image shows a side view of molecular docking, with TLR3 shown in red; where a represents NV6 and TLR3, b represents NV1 and TLR3, c represents NV4 and TLR3, and d represents NV5 and TLR3.

[0025] Figure 5 The figures show the RMSD curves of the docking complex between the vaccine candidate and TLR3 obtained from molecular dynamics simulations, where a represents NV6-TLR3, b represents NV1-TLR3, c represents NV4-TLR3, and d represents NV5-TLR3.

[0026] Figure 6 Characterization of vaccines NV5, NV4, and NV1 is shown in Figure a. a schematic diagram of the three candidate vaccines; b. SDS-PAGE analysis of purified proteins, with lane M representing the marker, lane 1 representing NV5, lane 2 representing NV1, and lane 3 representing NV4; c. Schematic diagram of mouse immunization; d. Antibody responses in mice after immunization with PBS or the three vaccines. Each symbol represents a mouse, and the straight line represents the GMT group. Statistical significance: ns, P>0.05; *, P<0.05; **, P<0.01; ***, P<0.001.

[0027] Figure 7 Immunogenicity of candidate vaccines NV5, NV4, and NV1 in mice: a) Transmission electron microscopy image of purified GII.4-VLP; b) GII.4-specific antibody titers in serum collected on days 21, 35, and 49 after the first immunization. The endpoint titer was defined as the reciprocal of the highest dilution with absorbance > blank value. c) GII.4-specific IFN-γ response measured by ELISpot; d) Specific numerical value of GII.4-specific IFN-γ. Statistical significance: ns, P > 0.05; *, P < 0.05; **, P < 0.01. Detailed Implementation

[0028] To make the above-mentioned objectives, features, and advantages of this application more apparent and understandable, the specific implementation methods of this application are described in detail below with reference to embodiments. It is to be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of this application.

[0029] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0030] This application discloses a method for constructing a multi-epitope subunit vaccine against norovirus. Reverse vaccinology and subtractive proteomics methods are used to screen candidate proteins for vaccine design.

[0031] Example 1: Optimal Antigen Protein Screening 1. Complete acquisition of norovirus protein sequences Proteomic data for the GI, GII, and GIV gene groups of noroviruses that infect humans were retrieved from NCBI (National Center for Biotechnology Information) (https: / / www.ncbi.nlm.nih.gov / ). The results from NCBI show that each strain of norovirus GI, GII, and GIV contains three types of proteins: non-structural polyproteins, capsid protein VP1, and capsid protein VP2. Therefore, a total of nine protein sequences were retrieved: YP_009700102.2 nonstructural polyprotein [Norovirus GI], YP_009700103.1 VP1 [Norovirus GI], YP_009700104.1 VP2 [Norovirus GI]; YP_009701445.1 nonstructural polyprotein [Norovirus GI] GII], YP_009701446.1 VP1 [Norovirus GII], YP_009701447.1 VP2 [Norovirus GII]; YP_009725315.1 nonstructural polyprotein [Norovirus GIV], YP_009725316.1 VP1 [Norovirus GIV], YP_009725317.1 VP2 [Norovirus GIV].

[0032] The CD-HIT (Cluster Database at High Identity with Tolerance) (https: / / github.com / weizhongli / cdhit / releases) server was used to remove homologous proteins. A threshold of 60% was set, and paralogous proteins with sequence similarity greater than 60% were removed, that is, one redundant protein was removed, leaving eight protein sequences.

[0033] Homologous proteins were identified from humans (taxid: 9606) and three probiotics (including *Lactobacillus rhamnosus* taxid: 47715, *Lactobacillus casei* taxid: 1582, and *Lactobacillus johnsonii* taxid: 33959) using the BLASTp tool from the NCBI database. The screening criteria were: sequence identity <30%, alignment score >100, and E-value <0.005. The results showed that the viral protein sequence had no homology with the host protein, *Lactobacillus casei*, or *Lactobacillus johnsonii*. However, when compared with the complete protein set of *Lactobacillus lactis*, two homologous protein sequences, GI and GIV polyproteins, were found and removed from the candidate protein sequence list (Table 1), leaving six candidate proteins.

[0034] Table 1. Proteins identified as homologous to the human genome (alignment score > 100, sequence identity < 30%)

[0035] 2. Prediction of the number of transmembrane spirals The transmembrane helical number and location information of candidate proteins were predicted using online servers TMHMM (https: / / services.healthtech.dtu.dk / services / DeepTMHMM-1.0 / ) and HMMTOP (http: / / www.enzim.hu / hmmtop / index.php). Proteins with ≥2 transmembrane helices were eliminated. The remaining six candidate protein sequences were predicted to have 0 transmembrane helices by TMHMM and HMMTOP, meeting the criteria of having 0 or 1 transmembrane helices, which facilitates protein expression and purification.

[0036] 3. Antigenicity and sensitization prediction Protein antigenicity was predicted using the VaxiJen 2.0 online server (https: / / www.ddg-pharmfac.net / vaxijen / VaxiJen / VaxiJen.html), retaining proteins with an antigenicity score >0.5. Simultaneously, protein sensitization was predicted using the AllerTop 2.0 online server (https: / / www.ddg-pharmfac.net / allertop_test / ), eliminating sensitizing proteins. The results, as shown in Table 2, excluded GII VP2 and GIV VP2 due to their allergenicity. Among the remaining four candidate proteins, GII VP1, with the highest antigenicity score of 0.55, was selected as the optimal antigenic protein.

[0037] Table 2. Antigenicity, sensitization, and transmembrane helix prediction of candidate proteins.

[0038] Example 2: Immune Cell Epitope Prediction 1. Determination of Human Leukocyte Antigen (HLA) Restriction Human leukocyte antigen (HLA) restriction was performed using the Allele Net Frequency Database (http: / / allelefrequencies.net / hla.asp), and the following HLA genotypes were selected: MHC-I class: HLA-A*02:01, HLA-A*24:02, HLA-A*11:01, HLA-C*04:01, HLA-C*03:04 and HLA-B*56:01; MHC-II class: HLA-DRB1*15:01, HLA-DRB1*03:01, HLA-DQA1*05:01 and HLA-DQA1*03:01.

[0039] 2. Prediction of cytotoxic T lymphocyte (CTL) epitopes Machine learning-based Artificial Neural Network (ANN) algorithms were used with the servers IEDB-NetMHCPan4.1EL (http: / / tools.iedb.org / mhci / ) and NetCTL (https: / / services.healthtech.dtu.dk / services / NetCTL-1.2 / ). The Consensus algorithm was used with the tool IEDB-Consensus (http: / / tools.iedb.org / mhci / ). Automated benchmarking was performed using a method trained on eluted data from the Immune Epitope Database (IEDB), with NetMHCPan 4.1 EL demonstrating state-of-the-art performance. A Motif Matrix (MM) algorithm was used with the server SYFPEITHI (http: / / www.syfpeithi.de / bin / MHCServer.dll / EpitopePrediction.htm). An additive model based on Quantitative Structure-Activity Relationship (QSAR) was used with the server MHCPred (https: / / www.ddg-pharmfac.net / mhcpred / MHCPred / ).

[0040] When predicting CTL epitopes, the optimal antigen protein sequence and MHC-I class HLA genotype restrictions are input. The algorithm outputs suitable epitopes and their binding affinities. Nine CTL epitopes were predicted, as shown in Table 3. The method also explored the percentile ranking of the predicted CTL epitopes, where lower values ​​indicate better binding affinity to MHC molecules.

[0041] Table 3. Prediction and assessment of cytotoxic T lymphocyte epitopes (CTLs)

[0042] 3. Helper T lymphocyte (HTL) epitope prediction When predicting HTL epitopes, the optimal antigen protein sequence and MHC class II HLA genotype restrictions are input. Based on the algorithm described in CTL epitope prediction, suitable HTL epitopes and their binding affinity are output. Seven HTL epitopes were predicted, as shown in Table 4. The method also explored the percentile ranking of the predicted HTL epitopes, where lower values ​​indicate better binding affinity to MHC molecules.

[0043] Table 4. Prediction and assessment of helper T lymphocyte epitopes (HTLs)

[0044] 4. Linear B-cell (LBL) epitope prediction The following algorithms are used for predicting LBL epitopes: ABCpred (https: / / webs.iiitd.edu.in / raghava / abcpred / ), based on Artificial Neural Networks (ANN); SVMtrip (http: / / sysbio.unl.edu / SVMTriP / ), based on Support Vector Machines (SVM); and BepiPred-2.0 (https: / / services.healthtech.dtu.dk / services / BepiPred-2.0 / ), based on Random Forests (or Decision Trees). All of these tools fall under the category of Machine Learning (ML) methods. When predicting LBL epitopes, the optimal antigen protein sequence was input, and the above algorithms output suitable epitopes. The results predicted nine linear LBL epitopes, as shown in Table 5.

[0045] Table 5. Prediction and assessment of linear B-cell epitopes (LBLs)

[0046] 5. Analysis of population coverage and various properties of predicted epitopes. Epitopes were analyzed for sensitization and toxicity on AllerTop2.0 and ToxinPred (http: / / crdd.osdd.net / raghava / toxinpred / ), and their hydrophobicity and hydrophobicity were also analyzed, retaining epitopes with good hydrophilicity. The initially screened epitopes met the criteria of being non-allergenic, non-toxic, highly antigenic, high percentile, and capable of binding to restriction HLA supertypes, totaling 5 CTL epitopes (C1-C5), 2 HTL epitopes (H1, H2), and 2 LBL epitopes (B1, B2). The results are summarized in Table 6.

[0047] Table 6 Summary of CTL, HTL, and LBL epitopes used in constructing vaccine candidates

[0048] Furthermore, Discovery Studio was used to analyze the conservation of the VP2 protein across various norovirus genotypes, ensuring that the epitopes fell within relatively conserved regions during viral evolution. Finally, it was confirmed that all screened epitopes were located within the P domain of the VP2 protein, as the P domain is a major antigenic determinant region of the virus and plays a crucial role in the immune response. Sequence alignment in Discovery Studio showed that these epitopes also contained conserved regions of the VP2 sequence in circulating strains, and that linear B-cell epitopes were all located within the P domain. In addition, considering that high population coverage and high hydrophilicity are ideal characteristics for antigenic epitopes, these values ​​were predicted using lEDB population coverage and the ToxinPred tool, respectively.

[0049] Example 3: Construction of a multi-epitope vaccine The construction of the vaccine sequence includes adjuvants, polyhistidine tags (His-tags), CTL epitopes, HTL epitopes, LBL epitopes, and corresponding linkers. Specifically, the linker with the amino acid sequence EAAAK is used to link the adjuvant, the linker with the amino acid sequence RVRR is used to link the histidine tag, the linker with the amino acid sequence AAY is used to link the CTL epitope, the linker with the amino acid sequence GGPPG is used to link the HTL epitope, and the linker with the amino acid sequence KK is used to link the LBL epitope. Multi-epitope vaccines are constructed using the following four strategies, distinguished by epitope sequence arrangement, epitope repetition pattern, and adjuvant selection: Strategy 1: Constructing by splicing all tabletops TLR7 agonists (Uniprot: Q5U7J2), TLR3 agonists (Uniprot: Q74EI3), and PADRE (amino acid sequence AKFVAAWTLKAAA) were used as immune adjuvants. These adjuvants were placed at both ends of the epitope sequence, with PADRE linked to one of the agonists, and a His tag added to the end of the sequence for purification. In the middle epitope region, selected CTL, HTL, and LBL cell epitopes were grouped by category to enhance the immune effect. Specifically, as shown... Figure 1As shown in Figure a, for predicted norovirus CTL, HTL, and LBL epitopes, a TLR7 agonist, PADRE adjuvant, CTL epitope regions (C1-C5), HTL epitope regions (H1, H2), LBL epitope regions (B1, B2), TLR3 agonist, and a multihistidine His tag were linearly tandemly from start to finish in a linker. The positions of the epitopes within their respective CTL, HTL, and LBL epitope regions were comprehensively permuted and combined. This was combined with swapping the positions of the TLR3 and TLR7 agonists to form multiple candidate vaccine sequences. Here, permutation and combination refers to the combined permutation and combination of all epitopes of the same type within the epitope region (CTL, HTL, or LBL), resulting in 480 candidate vaccine sequences. Due to slight differences in their physicochemical properties, a representative sequence was selected as a typical candidate for further analysis. It should be noted that human Toll-like receptor 3 (TLR3) and human Toll-like receptor 7 (TLR7) are cellular recognition systems that can respond to viral nucleic acids. The pan HLADR-binding epitope (PADRE) is a potent immune adjuvant that can stimulate a strong T-cell immune response.

[0050] Strategy 2: Construction of Optimal Epitope Selection Strategy Two uses the same three immune adjuvants as Strategy One (TLR3, TLR7 agonists, and PADRE), but selects only the one with the best overall properties from each epitope class for combination, and adjusts the position of the adjuvants at both ends of the sequence. Specifically, for example... Figure 1 As shown in b, based on Strategy 1 above, the optimal CTL epitope, optimal HTL epitope, and optimal LBL epitope with the best conditions are retained respectively, and their positions are arbitrarily arranged and combined. Combined with the swapping of the positions of TLR3 agonists and TLR7 agonists, multiple candidate vaccine sequences are formed. The optimal conditions are: the best epitope is the one that combines high affinity (percentile rank less than 0.5), good hydrophilicity (the best among various epitopes, the higher the positive value, the higher the hydrophilicity), and population coverage (covering those restricted by the above HLA genotypes, see Table 6). The optimal CTL epitope is TLEPIFIPV, the optimal HTL epitope is PVAGGAIAA, and the optimal LBL epitope is FFRSYIPLKGGFGNTA. Twelve candidate vaccine sequences were constructed here.

[0051] Strategy 3: Epitope Repetition Enhanced Construction Strategy 3 retains the adjuvant types and layout from Strategy 2, selecting one of the three optimal epitopes screened in Strategy 2 and repeating it six times consecutively. This repetitive design significantly enhances the humoral immune response and overall immunogenicity of the vaccine, and reduces mutation occurrence during subsequent plasmid expression, aiming to strengthen the immune effect through repetitive sequences. Specifically, such as... Figure 1 As shown in c, six candidate vaccine sequences were constructed here.

[0052] Strategy 4: Heat shock protein adjuvant alternative construction Strategy four, while retaining the epitope repeat structure, replaced the original TLR agonist adjuvant with heat shock protein HSP65 (amino acid sequence shown in SEQ ID NO.1) and the 407-426 amino acid fragment of HSP70 (amino acid sequence shown in SEQ ID NO.2) as immune-enhancing factors, with the HSP70 (407-426) fragment repeated twice. Specifically, as follows... Figure 1 As shown in d, based on strategy three above, the TLR7 agonist at one end is replaced with the heat shock protein HSP65, and the TLR3 agonist at the other end is replaced with HSP70 (407-426), and the replaced HSP70 (407-426) is repeated twice consecutively (i.e., Figure 1 (d in M) to form multiple candidate vaccine sequences. Here, three candidate vaccine sequences were constructed.

[0053] Example 4: Screening of Multiepitope Vaccines 1. Prediction of the antigenicity, sensitization, and physicochemical properties of vaccines Protein antigenicity was predicted using the VaxiJen 2.0 server (https: / / www.ddg-pharmfac.net / vaxijen / VaxiJen / VaxiJen.html), and proteins with antigenicity <0.5 were removed. Sensitization was predicted using AllerTop2.0 (https: / / www.ddg-pharmfac.net / vaxijen / VaxiJen / VaxiJen.html).

[0054] The remaining candidates were used to predict protein solubility on Protein-Sol (https: / / protein-sol.manchester.ac.uk / ), with a threshold of 0.45. A score > 0.45 was considered to indicate good solubility.

[0055] A series of indicators, including half-life, stability index, aliphatic index, and average GRAVY, were estimated using ProtParam (https: / / web.expasy.org / protparam / ). Candidate vaccine sequences with longer half-lives, stability indices less than 40, higher aliphatic indices, and negative GRAVY values ​​were selected.

[0056] All 480 candidates constructed using Strategy 1 as described in Example 3 were largely similar. Therefore, 7PC12345H12B123 was retained as representative for subsequent studies. Sequences with short half-lives, such as 3PCHB7 and 3PCCCCC7, were excluded, with half-lives of 1.1 hours, 2 minutes, and 2 minutes in mammalian reticulocytes, yeast, and E. coli, respectively. Sequences with solubility below 0.45 were also excluded. The stability of the vaccine candidates was assessed by predicting the instability index, aliphatic index, and large average hydrophilicity (GRAVY) using the ProtParm server. The results showed that the remaining candidates had an instability index below 40 and a favorable aliphatic index, while negative GRAVY values ​​indicated their hydrophilicity. Finally, six vaccine sequences met all screening criteria (Table 7) and were designated NV1-NV6, with their full names shown in parentheses.

[0057] Table 7. Set of vaccine candidates for further structural prediction

[0058] 2. Vaccine Immunization Simulation The online immune simulation server C-IMMSIM (https: / / kraken.iac.rm.cnr.it / C-IMMSIM / index.php?page=1) was used to simulate the immune response of vaccine candidates after vaccination on days 0, 28, and 56 (simulation period of 365 days). Each time step in the server was set to 8 hours, for a total of 1095 steps. Common host HLA types were selected: A02:01, A24:02, B56:01, B07:02, DRB115:01, and DRB103:01. C-IMMSIM is an agent-based immune response prediction model that uses a position-specific scoring matrix (PSSM) to simulate the immune system's response to pathogens and uses a three-dimensional cellular automaton to represent the interaction of immune cells, simulating the behavior of bone marrow, thymus, and lymphatic organs. Vaccine sequences with good immune responses after each simulated vaccination were retained, especially those that induced high antibody titers and significant increases in the levels of various immune cells such as B cells and T cells. Analysis of the immunosimulation results of six vaccine candidates showed that NV6 induced a significantly high antibody titer with an appropriate IgM to IgG level ratio, as well as significant increases in other immune factors and cellular levels. In the single-epitope construct, NV1 induced a high antibody titer with satisfactory levels of other immune cells. In contrast, NV2 and NV3 showed lower antibody titers, and the helper T cell counts after secondary vaccination did not show a better response than after primary vaccination. Furthermore, the macrophage population exhibited premature apoptosis. Therefore, NV2 and NV3 were excluded from the candidates. For NV4, its antibody titer was very high. However, IgM levels were consistently higher than IgG levels in both early and late stages. Analysis revealed that NV4 has the potential to combat acute norovirus infection and was therefore retained. The immunosimulation results for NV5 were not as good as the previous models. However, considering its excellent physicochemical properties, such as a high aliphatic coefficient, strong protein solubility, and moderate hydrophilicity, it was retained for further analysis and research.

[0059] Example 5: Vaccine Structure Prediction and Validation 1. Secondary and Tertiary Structure Prediction The secondary structure of candidate vaccines was predicted using PSIPRED 4.0 (http: / / bioinf.cs.ucl.ac.uk / psipred / ), followed by high-confidence prediction of their tertiary structures using the Robbetta online server (https: / / robetta.bakerlab.org / ). This process helps predict the stability, potential biochemical functions, and modes of action of vaccine molecules.

[0060] 2. Verification and refinement of the rationality of the structural conformation On the Robbetta online server (https: / / robetta.bakerlab.org / ), five models will be generated for each vaccine sequence. The most suitable model will be selected by analyzing the conformational rationality of each model. The evaluation criteria include small local energy fluctuations, few chemical conformational errors, and fewer residues in the forbidden region of the Ramachandran diagram.

[0061] The overall quality and local energy functions of the model were evaluated using ProSA-web (https: / / prosa.services.came.sbg.ac.at / prosa.php). Ramachandran plots were generated using SAVES v6.0 (https: / / saves.mbi.ucla.edu / ) to check the reasonable distribution of the main chain φ / ψ angles. Structural refinement was then performed using the FG-MD server (https: / / www.aideepmed.com / FG-MD / ). These servers can evaluate the conformational validity of the tertiary structure. Tertiary structure validation for four vaccine candidates (NV6, NV1, NV4, and NV5) is presented on [website name missing]. Figure 2 and Figure 3 The overall model quality was assessed using Z-scores from the ProSA-web server, which can be used to check if the Z-score of the input structure falls within the score range for similar-sized proteins natively. Local model quality was characterized by the energy function of the amino acid sequence. These results collectively indicate that the predictive model is reasonable and reliable in assessing both overall and local quality. Ramachandran plot results analyzed by SAVES v6.0 are shown... Figure 2 and Figure 3 The Ramachandran diagram depicts the permissible regions of phi (φ) and psi (ψ) torsion angles within the protein backbone. It reflects the conformational feasibility of individual amino acid residues within the protein structure. Stereochemistry is considered acceptable if more than 90% of the amino acids fall within the permissible regions, and fewer amino acids in the disallowed regions are preferred. The results showed that the vast majority of amino acids in the candidates fell within the permissible regions, with only 0.8–3.3% located in the disallowed regions.

[0062] Example 6: Molecular docking and interaction analysis with TLR3 Protein-protein docking was performed using the online server ClusPro 2.0 (https: / / cluspro.bu.edu / ). The remaining candidate vaccines were docked with TLR3 (PDB ID: 7C76) to evaluate docking quality and interaction levels. TLR3 is a cell recognition system that recognizes viral nucleic acids and is suitable for docking analysis of norovirus antigens. The docking results of the vaccine candidates with TLR3 were shown... Figure 4 In the middle, candidates NV6, NV1, and NV4 interact closely and deeply with TLR3, while candidate NV5 has a rather limited interaction area and appears unstable.

[0063] Based on the docking results, the interaction interface of the docking model was further analyzed using PDBePISA (https: / / www.ebi.ac.uk / pdbe / pisa / ), and LigPlot+ software was used to draw an interaction diagram to visually demonstrate information such as hydrogen bonds and hydrophobic contacts. The results are shown in Table 8. It can be observed that the interface-in-ligand ratio of candidate NV5 is lower than that of the other three candidates. Therefore, its interaction with the TLR3 docking complex is relatively weak.

[0064] Table 8. Interaction analysis results of the docking complex on PDBePISA

[0065] Example 7 Molecular Dynamics (MD) Simulation Molecular dynamics simulations are used to examine molecular motion and conformational stability in a solvent environment (simulating the in vivo environment). Amber was used to perform molecular dynamics simulations on four candidate vaccine-TLR3 complexes to examine the stability of their molecular structures over a certain period. The complexes were placed in a TIP3P water tank, and Na was added. + With Cl - Ions were used to simulate a physiological saline environment (0.9% NaCl). The system was then subjected to a two-stage energy minimization, heating, and equilibrium process, culminating in a 150 ns production simulation. Trajectory analysis was performed, and stability was verified using root mean square deviation (RMSD). If the complex maintained a small and stable RMSD range over a longer period, it indicated stable molecular conformation and no significant topological changes, which is crucial for subsequent molecular docking and interaction studies. Results are as follows... Figure 5 As shown, the RMSDs of the candidate vaccine molecules tend to stabilize, with average RMSD values ​​of 8.12, 4.53, 4.54, and 10.05 Å over the last 50 nanoseconds. These results indicate that the simulation has reached equilibrium, and the complex molecules maintain a highly stable conformation. However, the RMSD values ​​of NV6 and NV5 and their docking complexes are relatively large, at 8.12 and 10.05 Å, respectively, suggesting that their conformations underwent some changes in the simulated in vivo environment, and their stability is not as good as the other two candidate vaccines.

[0066] Example 8: Codon Optimization and In Situ Cloning Simulation Codon optimization of gene sequences to match Escherichia coliThe expression system enhances protein expression and synthesis efficiency by reducing the use of rare codons and aligning with abundant tRNAs. GC content and codon fitness index (CAI) are two important indicators for codon optimization. Specifically, GC content is controlled within the 45-60% range to ensure gene sequence structural stability, and the codon fitness index (CAI) target is set to >0.7 to improve host translation efficiency. Four candidates achieved good CAI and GC content levels after codon optimization on Jcat (https: / / www.jcat.de / ), as shown in Table 9.

[0067] Table 9. Codon fitness index (CAI) and GC content of candidate genes after codon optimization.

[0068] After optimization, the codon-optimized sequence was inserted into the pET-28a(+) E. coli expression plasmid using SnapGene as a design preparation for subsequent cloning and expression.

[0069] Example 9: Expression and Immunological Evaluation of the Vaccine 1. Expression and purification The optimized gene from Example 8 was synthesized by TsingKe Biotechnology and cloned into the BamHI and SacI sites of the pET-28a(+) vector. The plasmid was chemically transformed and introduced into BL21 competent E. coli cells. The transformed BL21 cells were cultured at 37 °C and 220 rpm in a shaker until OD... 600 The concentration was 0.5–0.8. Protein expression was induced with a final concentration of 0.1 mM IPTG at 16 °C, followed by overnight incubation at 220 rpm (16–18 h). Cells were harvested by centrifugation at 5000 rpm for 5 min, and the cell pellet was resuspended in lysis buffer (20 mM Tris-HCl, pH 8.0; 500 mM NaCl). Cells were lysed using a high-pressure homogenizer (800 MPa, 3 min). The lysis buffer was clarified by centrifugation at 12000 rpm for 10 min. For soluble proteins, the supernatant was collected and purified by Ni affinity chromatography; for inclusion body proteins, the pellet (inclusion bodies) was collected and resuspended in denaturing buffer (6 M urea, 20 mM Tris-HCl, pH 8.0, 500 mM NaCl), clarified by centrifugation again, and the soluble inclusion body proteins were purified by Ni affinity column chromatography.

[0070] Gene synthesis evaluation showed that NV6 (7PC12345H12B123) has 2-3 transmembrane helices, which hinders protein expression and purification; therefore, the following three vaccine candidates were synthesized. Figure 6 a): NV5(65PCCCCCCM2): Contains a 6×CTL epitope of norovirus GII.4, consisting of one HSP65, one PADRE and two HSP70 (407~426) molecules linked together, with a 6×His tag added to the C-terminus. Its complete amino acid sequence is shown in SEQ ID NO.3; the nucleotide sequence of its encoding gene is shown in SEQ ID NO.6. NV4(7PCCCCCC3): Contains a 6×CTL epitope linked by a TLR7 agonist, a PADRE, and a TLR3 agonist, with a 6×His tag added to the C-terminus. Its complete amino acid sequence is shown in SEQ ID NO.4; the nucleotide sequence of its encoding gene is shown in SEQ ID NO.7. NV1(7PCHB3): Contains one CTL epitope, one HTL epitope and one LBL epitope, linked by one TLR7 agonist, one PADRE and one TLR3 agonist, with a 6×His tag added to the C-terminus. Its complete amino acid sequence is shown in SEQ ID NO.5, and the nucleotide sequence of its encoding gene is shown in SEQ ID NO.8.

[0071] SDS-PAGE gel electrophoresis analysis confirmed the purity and size of the expressed antigen: purified NV5 showed a dominant band at ~76 kDa. Purified NV4 showed a dominant band at ~31 kDa. Purified NV1 showed a dominant band at ~27 kDa. All observed molecular weights were generally consistent with the predicted sizes. Figure 6 (b) in the middle.

[0072] 2. Transmission electron microscope (TEM) Virus-like particles (VLPs) of type GII.4 were constructed as follows: Recombinant baculovirus was constructed using plasmid pIEXBac-GII.4 according to the product instructions (Novagen, Merck KGaA, Darmstadt, Germany). The resulting baculovirus was named Bac-GII.4 and infected Sf9 cells with a multiplicity of infection (MOI=0.1) according to the method described in reference [Ku Z, Ye X, Huang X, Cai Y, Liu Q, Li Y, et al. Neutralizing antibodies induced by recombinantvirus-like particles of enterovirus 71 genotype C4 inhibitinfection at pre-and post-attachment steps. PLOS ONE 2013;8(2):e57601]. Cells were collected 72 hours after infection and lysed with PBS buffer containing 1% NP-40. The lysis buffer was ultracentrifuged on a 20% sucrose pad for 5 hours (27,000 rpm). The resulting precipitate was resuspended in PBS buffer and then ultracentrifuged on a 10-50% sucrose gradient for 3 hours (39,000 rpm). Twelve gradient fractions (from top to bottom) were analyzed for VLP content by ELISA (method as described in reference Huang Z, Chen Q, Hjelm B, Arntzen C, Mason H. A DNA replicon system for rapid high-level production of virus-like particles in plants. Biotechnol Bioeng 2009;103(July (4)):706–14). The VLP-rich fractions were combined and ultracentrifuged on a 20% sucrose pad as described above. The final precipitate containing purified virus-like particles (VLPs) was resuspended in PBS buffer for later use. The purified protein was quantified using the Bradford method.

[0073] The prepared VLPs samples were negatively stained with 0.5% uranium acetate and morphologically observed using transmission electron microscopy (TEM) (scale bar = 100 nm).

[0074] 3. Mouse immunization The antigen was emulsified with Freund's adjuvant and injected. Six BALB / c mice (5 weeks old, 19–21 g, female) were subcutaneously injected on days 0, 21, and 35, with an injection volume of 100 μL / mouse / injection and an antigen dosage of 10 μg / mouse / injection. The inoculum was either PBS (control group) or a mixture of the candidate vaccine (NV5, NV1, or NV4) and Freund's adjuvant. Specifically, in the vaccine group, on day 0, a mixture of Freund's complete adjuvant (FCA) and antigen was injected as the primary immunization (1st dose), and on days 21 and 35, booster immunizations were administered using a mixture of Freund's incomplete adjuvant (FIA) and antigen (2nd and 3rd doses, respectively). Serum was collected on days 21, 35, and 49. Figure 6 As shown in c. Fourteen days after the last immunization, mice were sacrificed and their spleens were harvested to assess interferon-γ (IFN-γ) secretion. The control group (PBS) was operated on in the same manner, except that the antigen was replaced with an equal volume of PBS.

[0075] 4. Antibody reaction specificity Indirect ELISA was used to measure antigen-specific antibodies in serum. The method was as follows: each well of a 96-well plate was coated with 30 ng of purified candidate vaccine or 10 ng of Norovirus GII.4 VLPs. The endpoint titer was determined using a serial dilution method, defined as the highest dilution that produced a positive result. Serum from the PBS control group remained negative at the lowest dilution (1:64) and was assigned a value of 32 in the GMT calculation. The results showed that the antigen-specific geometric mean titers (GMTs) against their respective immune antigens were as follows: Figure 6 As shown in d in the figure. No significant difference in antigen-binding GMT was observed between NV1 and NV4 sera, indicating comparable ability to induce antigen-specific antibodies. Serum from NV5-immunized mice showed significantly higher GMTs than the NV1 and NV4 groups, demonstrating the superior ability of NV5 to induce antigen-specific antibodies.

[0076] Before assessing the binding titer of GII.4 VLPs, purified norovirus GII.4 VLPs were characterized by transmission electron microscopy (TEM) to confirm their spherical morphology, with a diameter of ~38 nm. Figure 7 (a) Using serially diluted serum samples to determine GII.4-specific geometric mean titers (GMTs) ( Figure 7 (b) Post-primary immunization: NV5 immune serum 1109, NV1 immune serum 683, NV4 immune serum 1536; After the second immunization: NV5 immune serum 1365, NV1 immune serum 1877, NV4 immune serum 1877; After the third immunization: NV5 immune serum 1707, NV1 immune serum 1109, NV4 immune serum 4352.

[0077] Only NV4 serum showed significantly higher GII.4-VLP binding GMTs than NV5 and NV1 after primary immunization. Following booster immunization, no significant differences in GII.4-VLP-specific antibody titers were observed among the three vaccine groups, indicating that all three vaccines had comparable ability to induce GII.4-specific antibodies. However, the NV4 vaccine induced GII.4-specific antibodies more rapidly.

[0078] 5. ELISPOT Experiment Spleen cells were aseptically ground and passed through a cell filter, then dispersed by gentle pressure using a sterile syringe plunger. Cells were treated with 1× erythrocyte lysis buffer to remove erythrocytes, washed with RPMI medium containing 10% v / v FBS, and resuspended to 1.0×10⁶ cells / mL. 7 cells / mL. 1.0 × 10⁶ cells / mL was seeded per well. 5 Cells were incubated overnight in a cell culture incubator. They were then stimulated with purified Norovirus GII.4 VLPs for 24 hours. IFN-γ secretion was detected using the IFN-γ pre-coated ELISPOT kit from Dakewe (Hunan Province) according to the manufacturer's instructions. Spot-forming units (SFUs) were automatically counted using an ELISPOT plate reader.

[0079] To assess GII.4-specific interferon-γ (IFN-γ) production, spleen cells were isolated 14 days after the final immunization, stimulated in vitro with GII.4-VLPs, and evaluated using ELISPOT. Figure 7 (c) ELISPOT analysis showed that all three vaccines effectively stimulated GII.4-specific IFN-γ secretion. Notably, spleen cells from NV1-immunized mice produced significantly more IFN-γ spots than those from the NV5 and NV4 groups. Figure 7 The d in the figure indicates that the NV1 vaccine elicits a stronger cellular immune response.

[0080] This invention provides a concept and method for constructing a multi-epitope subunit vaccine for norovirus. Many methods and approaches exist for implementing this technical solution; the above description is merely a preferred embodiment of the invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of this invention, and these improvements and modifications should also be considered within the scope of protection of this invention. All components not explicitly stated in this embodiment can be implemented using existing technologies.

Claims

1. A multi-epitope subunit vaccine for norovirus, characterized in that, Its amino acid sequence is shown in any one of SEQ ID NO.3~5.

2. The gene encoding the norovirus multi-epitope subunit vaccine of claim 1.

3. The gene according to claim 2, characterized in that, Its nucleotide sequence is shown in any one of SEQ ID NO. 6 to 8.

4. The use of the norovirus multi-epitope subunit vaccine of claim 1 in the preparation of drugs for the prevention of norovirus.

5. A method for constructing a multi-epitope subunit vaccine for norovirus, characterized in that, Includes the following steps: Step 1: Obtain and screen proteomic data of norovirus genomes; Step 2: Immune cell epitope prediction; Step 3: Multi-epitope subunit vaccines were constructed using the full epitope splicing construction strategy, the optimal epitope screening construction strategy, the epitope duplication enhancement construction strategy, and the heat shock protein adjuvant substitution construction strategy, respectively. Step 4: Screen the multi-epitope subunit vaccines constructed in Step 3 through property prediction, vaccine immunization simulation, structure prediction, molecular docking analysis, molecular dynamics analysis, and immunological evaluation.

6. The construction method according to claim 5, characterized in that, In step two, the immune cell epitope prediction includes prediction of cytotoxic T lymphocyte CTL epitopes and helper T lymphocyte HTL epitopes based on human leukocyte antigen (HLA) restriction. The immune cell epitope prediction also includes prediction of linear B cell LBL epitopes.

7. The construction method according to claim 5, characterized in that, In step three, the multi-epitope subunit vaccine includes an adjuvant, an immune cell epitope, and a histidine tag, all of which are linked by a linker peptide.

8. The construction method according to claim 7, characterized in that, In the aforementioned full epitope splicing construction strategy, the adjuvants include TLR7 agonists, TLR3 agonists, and PADRE, and the immune cell epitopes are all the immune cell epitopes predicted in step two. In the optimal epitope screening and construction strategy, the adjuvants include TLR7 agonists, TLR3 agonists, and PADRE, and the immune cell epitopes are selected from each type of immune cell epitope predicted in step two and combined to form an epitope with the best overall performance. In the epitope repetition enhancement construction strategy, the adjuvants include TLR7 agonists, TLR3 agonists and PADRE, and the immune cell epitope is selected by repeating the immune cell epitope predicted in step two six times consecutively. In the aforementioned heat shock protein adjuvant replacement construction strategy, the adjuvant includes amino acid fragments 407-426 of HSP65 and HSP70 and PADRE, and the immune cell epitope is selected by selecting an immune cell epitope predicted in step two and repeating it six times consecutively.

9. The construction method according to claim 7, characterized in that, The linker peptides include any one or more combinations of EAAAK, RVRR, AAY, GGPPG, and KK.

10. A norovirus multi-epitope subunit vaccine constructed by the construction method according to any one of claims 5 to 9.

Images