Lipid nanoparticles targeting liver sinusoidal endothelial cells and use thereof

Abstract

Lipid nanoparticles (LNPs) targeting liver sinusoidal endothelial cells and the use thereof. The LNPs consist of a cationic ionized lipid, an auxiliary lipid, a ligand-modified PEGylated lipid, and cholesterol. The auxiliary lipid and the ligand-modified PEGylated lipid target surface receptors of the liver sinusoidal endothelial cells (LSECs). The LNPs enhance, by means of a multivalent targeting design, specific uptake by LSECs, and encapsulate an RNA drug for treating an allergic disease or an autoimmune disease. The RNA drug can encode at least one epitope of an antigen causing an allergic or autoimmune disease. The LNPs can efficiently target LSECs to induce immune tolerance, providing a new antigen-specific immunomodulatory strategy for the treatment of allergic diseases or autoimmune diseases.

Description

A lipid nanoparticle targeting hepatic sinusoidal endothelial cells and its application

[0001] This application claims priority to a prior Chinese application, application number 2026101924847, filed on February 10, 2026; all of its contents are part of this invention.

[0002] This application claims priority to a prior Chinese application, application number 2025102240116, filed on February 27, 2025; all its contents are part of this invention.

[0003] This application claims priority to the earlier Chinese application, application number 2025102826781, filed on March 11, 2025; all its contents are part of this invention.

[0004] This application claims priority to the earlier Chinese application, application number 2025103613072, filed on March 26, 2025; all its contents are part of this invention.

[0005] This application claims priority to the earlier Chinese application, application number 2025103758531, filed on March 27, 2025; all its contents are part of this invention.

[0006] This application claims priority to a prior Chinese application, application number 2025103759587, filed on March 27, 2025; all of its contents are part of this invention.

[0007] This application claims priority to the earlier Chinese application, application number 2025103940638, filed on March 31, 2025; all its contents are part of this invention.

[0008] This application claims priority to a prior Chinese application, application number 2025105843929, filed on May 7, 2025; all of its contents are part of this invention.

[0009] This application claims priority to a prior Chinese application, application number 2025106749125, filed on May 23, 2025; all its contents are part of this invention.

[0010] This application claims priority to the earlier Chinese application, application number 2025109003751, filed on July 1, 2025; all its contents are part of this invention. Technical Field

[0011] This invention belongs to the field of drug delivery, and more specifically, relates to a lipid nanoparticle that targets hepatic sinusoidal endothelial cells and its application. Background Technology

[0012] Beyond current treatments employing anti-inflammatory, immunosuppressive, targeted monoclonal antibody, or immunomodulatory approaches, the need for developing novel therapies for autoimmune and allergic diseases remains unmet. While most of these therapies offer symptom relief and temporary reduction in disease activity, they do not offer the prospect of long-term suppression of chronic disease activity or a cure. However, there is a growing recognition of the power of regulatory T-cell (Treg) biology and its importance in providing antigen-specific immune tolerance for autoimmune diseases such as rheumatoid arthritis, lupus, and type 1 diabetes, and allergic diseases such as food allergies, anaphylactic reactions, and asthma. Inducing antigen-specific tolerance is one approach that uses biodegradable nanoparticles to initiate and maintain immunomodulatory responses. Based on the ability of these carriers to encapsulate disease-related antigens, these antigens are delivered to antigen-presenting cells (APCs), thereby inducing antigen-specific tolerance.

[0013] The liver's tolerance is well-known because this organ is effective in preventing immune responses to exogenous food antigens from the gastrointestinal and portal venous systems, and in promoting the persistence of tumor metastasis to this organ. Furthermore, the liver enjoys immune privilege during organ transplantation, requiring less immunosuppressive therapy compared to kidney or heart transplants. Studies have also shown that simultaneous kidney or heart and liver transplantation is less likely to result in immune rejection compared to isolated organ transplants.

[0014] The immunosuppressive effect of the liver can be attributed in part to its unique APC system, including naturally tolerant APCs such as Kupffer cells (KCs), dendritic cells (DCs), and hepatic sinusoidal endothelial cells (LSECs). These tolerogenic APCs constitute a component of the hepatic reticuloendothelial system, which plays a crucial role in clearing foreign substances, degradation products, and toxins from sinusoidal blood through phagocytosis and endocytosis. Furthermore, while specialized phagocytes (KCs and DCs) preferentially eliminate circulating microscale particulate matter through phagocytosis, LSECs are more adept at eliminating soluble macromolecules and nanoparticles in the 200 nm range through clathrin-mediated endocytosis. From an immunomodulatory perspective, LSECs play a key role in inducing immunosuppression in both CD8+ and CD4+ populations by generating antigen-specific Tregs, producing TGF-β, and upregulating programmed cell death protein 1 (PD-1) receptor ligand (PD-L1). Therefore, the ability of LSECs to control antigen-specific Treg function should be considered when treating autoimmune and allergic diseases.

[0015] Hepatic sinusoidal endothelial cells (LSECs), as core cells regulating hepatic immune tolerance, express a variety of specific receptors on their surface. The biological characteristics and targeting potential of these receptors have been extensively studied in this field, making them key targets for nanocarrier targeted delivery: SR-E family receptors: SR-E (Scavenger Receptor Class) The E family is an important pattern recognition receptor subset on the surface of LSECs. Among them, SR-E1 / LOX-1 (lectin-like oxidized low-density lipoprotein receptor 1) and SR-H1 / STABILIN-1 (stabilin-1) can specifically recognize anionic lipid ligands such as phosphatidylserine (PS) and dioleoyl-sn-glycerol-3-phosphate-L-serine (DOPS), mediating ligand-receptor-dependent endocytosis. Studies have confirmed that this type of receptor plays a key role in regulating the uptake of nanoparticles by LSECs. SR-E3 / mannose receptor (CD206) mainly recognizes sugar ligands such as mannose, trimannose, and L-fucose, promoting intracellular delivery through clathrin-mediated endocytosis, and is a classic target for targeted drug delivery by LSECs. LSECtin / CLEC4G: As a member of the C-type lectin family, LSECtin / CLEC4G is specifically expressed on the surface of LSECs and can bind to sugar ligands such as GlcNAc and mannose. Its unique intracellular signaling domain can regulate the antigen presentation process and participate in the induction of immune tolerance. It has been studied as a target for LSEC-targeting vectors. LYVE-1 (Lymphatic Endothelial Hyaluronic Acid Receptor 1): LYVE-1 is a hyaluronic acid receptor specifically expressed on the surface of LSECs. It can mediate receptor-dependent endocytosis by recognizing hyaluronic acid ligands. Its high expression and ligand binding specificity make it an important candidate target for LSEC-targeted delivery.

[0016] However, efficient delivery of antigens to LSECs remains a challenge. The use of lipid nanoparticles (LNPs) to induce immune tolerance is an active research area, including the surface decoration of peptides / major histocompatibility complexes (MHCs) as alternative antigen presentation platforms for immune tolerance in the absence of co-stimulation. Recent studies have shown that by adding mannose ligands to the surface of LNPs, they can specifically target LSECs and enhance immune tolerance by increasing cellular uptake, surface epitope presentation, and inducing Tregs. However, this design involves only one ligand and may still suffer from imprecise targeting, leading to uptake by hepatocytes or other non-solid cells.

[0017] Therefore, there is an urgent need for a targeted and specific method of LNPs to efficiently deliver antigens to LSECs. Summary of the Invention

[0018] This invention discloses lipid nanoparticles (LNPs) targeting hepatic sinusoidal endothelial cells and their applications. The LNPs consist of cationic ionized lipids, helper lipids, ligand-modified PEGylated lipids, and cholesterol. The helper lipids and ligand-modified PEGylated lipids target receptors on the surface of hepatic sinusoidal endothelial cells (LSECs). The LNPs enhance specific uptake by LSECs through multivalent targeting design and encapsulate RNA drugs for treating allergic or autoimmune diseases. These RNA drugs encode at least one epitope of an antigen that causes allergies or autoimmune diseases. The LNPs can efficiently target LSECs to induce immune tolerance, providing a novel antigen-specific immunomodulatory strategy for the treatment of allergic or autoimmune diseases.

[0019] On one hand, the present invention provides a lipid nanoparticle comprising cationic ionized lipids, auxiliary lipids, ligand-modified PEGylated lipids, and cholesterol; wherein the ligand in the ligand-modified PEGylated lipids is capable of targeting receptors on the surface of hepatic sinusoidal endothelial cells, and the auxiliary lipids are selected from lipids capable of targeting receptors on the surface of hepatic sinusoidal endothelial cells or lipids without targeting.

[0020] Existing technologies suffer from two major drawbacks: First, lipid nanoparticles exhibit poor targeting, often resulting in non-specific delivery or targeting with only a single ligand, leading to low LSEC uptake efficiency and a high proportion of uptake by non-targeted cells (such as hepatocytes). Second, the component design is vague, lacking a clear understanding of the synergistic targeting logic between the auxiliary lipids and ligands (e.g., targeting CD206 with a single mannose ligand results in inaccurate targeting). The lipid nanoparticles provided by this invention allow for the selection of auxiliary lipids that target receptors on the surface of LSECs, or auxiliary lipids without targeting function, combined with at least one PEGylated lipid modified with a ligand targeting receptors on the surface of LSECs to achieve monovalent or multivalent targeting. When the auxiliary lipids have targeting function, or when they lack targeting but are combined with at least two PEGylated lipids modified with ligands, the prepared LNPs can multivalently target receptors on the surface of LSECs, rather than targeting the entire liver, thus solving the problem of low LSEC targeting efficiency despite LNPs being enriched in the liver in existing technologies.

[0021] Further, the cationic ionized lipid is SM102; the auxiliary lipid includes any one of PS, DOPS, and DPPS; the ligand-modified PEGylated lipid includes at least any one of DSPE-PEG2K-mannose, DSPE-PEG2K-GlcNAc, DSPE-PEG2K-trimannose, DSPE-PEG2K-GalNAc-4-sulfate, DSPE-PEG2K-L-fucose, and DSPE-PEG2K-hyaluronic acid.

[0022] The PS (phosphatidylserine) targets SR-E1 / LOX-1 and SR-H1 / STABILIN-1 on the surface of LSECs.

[0023] The DOPS (1,2-dioleoyl-sn-glycerol-3-phosphate-L-serine) targets the SR-E1 / LOX-1 and SR-H1 / STABILIN-1 receptors of LSECs.

[0024] The DPPS (1,2-dipalmitoyl-sn-glycerol-3-phosphate-L-serine) targets the SR-E1 / LOX-1 receptor of LSECs.

[0025] The GalNAc-4-sulfate targets the SR-E3 / mannose receptor (CD206) on the surface of LSECs; the mannose, trimannose, L-fucose and GlcNAc target the SR-E3 / mannose receptor (CD206) and LSECtin / CLEC4G on the surface of LSECs; the hyaluronic acid targets LYVE-1 on the surface of LSECs.

[0026] Preferably, the auxiliary lipid is PS.

[0027] In some embodiments, the types of target-specific auxiliary lipids were screened, and when the auxiliary lipid was PS, the prepared LNPs showed the best effect in targeting the liver.

[0028] Furthermore, the ligand-modified PEGylated lipid is any two of DSPE-PEG2K-mannose, DSPE-PEG2K-GlcNAc, DSPE-PEG2K-trimannose, DSPE-PEG2K-GalNAc-4-sulfate, DSPE-PEG2K-L-fucose, and DSPE-PEG2K-hyaluronic acid.

[0029] Preferably, the ligand-modified PEGylated lipids are DSPE-PEG2K-trimannose and DSPE-PEG2K-GlcNAc.

[0030] In some methods, the types and amounts of ligand-modified PEGylated lipids were screened, and the prepared LNPs showed the best liver-targeting effect when the ligand-modified PEGylated lipids were a combination of DSPE-PEG2K-trimannose and DSPE-PEG2K-GlcNAc.

[0031] Furthermore, the molar ratio of the cationic ionized lipid, auxiliary lipid, ligand-modified PEGylated lipid, and cholesterol is 20–70:1–15:1–5:25–45, and the sum of the molar ratios of each component is 100%; the N / P ratio of the cationic ionized lipid is 2–6.

[0032] Preferably, when the cationic ionized lipid is SM102, the N / P ratio of SM102 is 6.

[0033] Furthermore, the lipid nanoparticles have a particle size of 80–200 nm; the lipid nanoparticles encapsulate RNA drugs for treating allergic or autoimmune diseases.

[0034] The pore diameter in LSECs typically ranges from 50 to 200 nanometers, thus the size of the lipid nanoparticles provided by this invention can be maintained within the sinus space, thereby facilitating the interaction between LSECs and lipid nanoparticles and the opportunity for LSECs to take up lipid nanoparticles.

[0035] Furthermore, the mass ratio of the RNA drug to the lipid nanoparticles is 1–10:20–100.

[0036] Preferably, the mass ratio of the RNA drug to the lipid nanoparticles is 1:40.

[0037] Furthermore, the RNA is mRNA, which includes a 5'-cap structure, a 5'-UTR, a coding region, a 3'-UTR, and a poly A tail.

[0038] The mRNA is unstable and carries a negative charge, while the cell membrane surface also carries a negative charge. Electrostatic repulsion makes it difficult for mRNA molecules to pass through the cell membrane and enter the cell. Therefore, the encapsulation of the lipid nanoparticles is required to achieve mRNA delivery and intracellular expression.

[0039] The 5'-UTR is a non-coding polypeptide mRNA region located directly upstream (5') of the start codon (the first codon in the mRNA transcript translated by the ribosome). The 3'-UTR is a non-coding polypeptide mRNA region located directly downstream (3') of the stop codon (the codon in the mRNA transcript that signals the termination of translation). The polyA tail is the 3' end of most eukaryotic mRNAs and helps regulate mRNA stability, transport, and translation. Both the 5'-UTR and 3'-UTR are typically transcribed from genomic DNA and are elements of pre-mature mRNA. The characteristic structural features of mature mRNA (5'-cap structure and polyA tail) are usually added to the transcribed mRNA during mRNA processing.

[0040] Furthermore, the sequences of the 5'UTR and 3'UTR are independently derived from at least one of natural and synthetic proteins.

[0041] Preferably, the natural protein includes any one of α-globulin, β-globulin, and heat shock protein HSP70.

[0042] Furthermore, the 5'-UTR contains a Kozak sequence.

[0043] In this invention, the Kozak sequence is a nucleotide sequence located after the 5'-cap structure of mRNA, which can bind to the promoter and mediate the translation initiation of mRNA containing the 5'-cap structure.

[0044] Furthermore, the coding region includes at least one epitope encoding an antigen that causes the treatment of allergic or autoimmune diseases.

[0045] Epitopes are specific chemical groups in antigen molecules that determine antigen specificity, also known as antigenic determinants. They are the basic units for TCR / BCR and antibody-specific binding. Epitopes can be divided into continuous epitopes (linear epitopes) and discontinuous epitopes (conformational epitopes). In the immune response, based on the different TCRs and BCRs recognized by the antigenic epitope, they are divided into T-cell epitopes and B-cell epitopes. Epitopes are generally no more than 20 amino acids in size, can be recognized by the body, and can stimulate the body to produce antibodies. They are the basis of protein antigenicity and the basic structure for inducing the body to produce an immune response. Naturally occurring immune responses cannot recognize all epitopes, but rather concentrate on a relatively small number of epitopes.

[0046] The epitopes mentioned are preferred epitopes. Through AI-assisted design, epitopes with high affinity and hydrophilicity that induce antigens for treating allergic or autoimmune diseases are selected as preferred epitopes.

[0047] Furthermore, when there are multiple tabletops, they are connected in series via connectors.

[0048] The linker links multiple antigenic epitopes together to enable peptide cleavage upon release of the epitopes into the cell.

[0049] Furthermore, the connector is a flexible connector.

[0050] Preferably, the linker is a glycine-serine (Gly-Ser) linker or a GGPPG linker.

[0051] Furthermore, a target sequence is inserted upstream of the coding region.

[0052] The targeting sequence enables the antigen epitope to enter the MHC-II endosome compartment for peptide presentation to Treg precursor cells. In this invention, the targeting sequence is preferably a 1-80 amino acid fragment of the invariant chain (Ii), abbreviated as Ii(1-80). The protein subdomain of this fragment (a molecular chaperone protein of MHC-II, which facilitates peptide loading into MHC-II for antigen presentation) allows the polypeptide epitope to enter the MHC-II from the cytoplasm for CD44 expression. + T cell presentation. The target sequence can be Ii(1-80) or transferrin receptor.

[0053] This invention also optimizes the codons of the template cDNA transcribed into mRNA to achieve optimal gene expression of a non-human cell tRNA library compared to human cells. This was accomplished using the GenScript online codon optimization tool. Furthermore, during transcription, uridine was replaced with N1-methylpseudouridine; after transcription, a 5' cap and a Poly A tail were added to the mRNA, wherein the 5' cap is either CleanCap or ARCA, and the Poly A tail is preferably 100-120 nucleotides in length.

[0054] Furthermore, the present invention provides a method for preparing the mRNA, comprising the following steps:

[0055] (1) Design and synthesize cDNA containing 5'-UTR, coding region and 3'-UTR;

[0056] (2) Transcribe the cDNA from step (1) into mRNA;

[0057] (3) Add a 5'-cap structure and a poly A tail to the mRNA transcribed in step (2).

[0058] Due to the fragility of RNA drugs, using multiple covalent linking steps may lead to instability of LNPs and mRNA. Therefore, this invention employs a "one-pot" synthesis of LNPs and their encapsulated RNA drugs to bind multiple ligands to the same LNP, ensuring the specific targeting of LNPs to LSECs.

[0059] In some methods, microfluidic methods are used to prepare LNPs.

[0060] This invention utilizes an innovative multivalent targeting design to precisely bind to multiple receptors on the surface of LSECs, efficiently delivering RNA drugs encoding disease-associated antigenic epitopes. The core mechanism of action is inducing antigen-specific immune tolerance, thereby fundamentally regulating abnormal immune responses. This mechanism determines that its therapeutic scope is not limited to allergic diseases such as birch pollen allergy and artemisia allergy, or autoimmune diseases such as type 1 diabetes, myasthenia gravis, toxic diffuse goiter, systemic lupus erythematosus, primary cholangitis, multiple sclerosis, and Sjögren's syndrome, but rather covers all allergic and autoimmune diseases that meet the core pathological characteristics of "immune tolerance disruption." For allergic diseases, regardless of the type of allergen (pollen, dust mites, food proteins, pet dander, etc.), the essence is an excessive IgE-mediated immune response to harmless exogenous substances. This invention can induce regulatory T cell (Treg) proliferation and inhibit Th2 cell activation by loading immunodominant epitope RNA corresponding to the allergen, thus blocking the allergic reaction cascade. For autoimmune diseases, regardless of the pathogenic target (pancreatic β cells, myelin sheath, thyroid receptors, bile duct epithelial cells, etc.), the core is the abnormal activation of autoreactive T / B cells attacking the body's own tissues. This invention can specifically induce immune tolerance by loading optimized epitope RNA corresponding to the autoantigen, inhibiting the production of autoantibodies and the release of pro-inflammatory factors, thus alleviating tissue damage. As long as the disease meets the core characteristics of "the presence of a clear pathogenic antigen and an imbalance in immune tolerance," targeted therapy can be achieved by screening for immunodominant epitopes of the corresponding antigen, optimizing the RNA drug coding region sequence, and using the LNPs delivery system of this invention, demonstrating broad applicability and scalability.

[0061] Preferably, the allergic diseases include birch hay fever and artemisia allergy, and the autoimmune diseases include type 1 diabetes, myasthenia gravis, toxic diffuse goiter, systemic lupus erythematosus, primary cholangitis, multiple sclerosis, and Sjögren's syndrome.

[0062] The RNA encapsulated in lipid nanoparticles for treating birch hay fever has a coding region that encodes at least one amino acid sequence as shown in SEQ ID No. 2-7; the RNA encapsulated in lipid nanoparticles for treating artemisia allergy has a coding region that encodes at least one amino acid sequence as shown in SEQ ID No. 185-186; the RNA encapsulated in lipid nanoparticles for treating type 1 diabetes has a coding region that encodes at least one amino acid sequence as shown in SEQ ID No. 22-41; the RNA encapsulated in lipid nanoparticles for treating myasthenia gravis has a coding region that encodes at least one amino acid sequence as shown in SEQ ID No. 64-69, 77-79; the RNA encapsulated in lipid nanoparticles for treating toxic diffuse goiter has a coding region that encodes at least one amino acid sequence as shown in SEQ ID No. 86-93; the RNA encapsulated in lipid nanoparticles for treating systemic lupus erythematosus has a coding region that encodes at least one amino acid sequence as shown in SEQ ID No. 103-112; the RNA encapsulated in lipid nanoparticles for treating primary cholangitis has a coding region that encodes at least one amino acid sequence as shown in SEQ ID No. 185-186. The coding region of the RNA encapsulated in the lipid nanoparticles for treating the multiple sclerosis encodes at least one of the amino acid sequences shown in SEQ ID No. 143-150; the coding region of the RNA encapsulated in the lipid nanoparticles for treating the Sjögren's syndrome encodes at least one of the amino acid sequences shown in SEQ ID No. 164-173.

[0063] Furthermore, the reverse-translated cDNA sequences corresponding to the amino acid sequences shown in SEQ ID No. 2-7 are shown in SEQ ID No. 8-13.

[0064] In some embodiments, the present invention provides a preferred mRNA for treating birch hay fever, wherein the coding region of the mRNA encodes all preferred epitope amino acid sequences as shown in SEQ ID No. 2 to 7, and the reverse-translated cDNA sequence corresponding to the coding region amino acid sequence is shown in SEQ ID No. 16, with a GC content of 56.74%.

[0065] Furthermore, the reverse-translated cDNA sequences corresponding to the amino acid sequences shown in SEQ ID No. 185-186 are shown in SEQ ID No. 187-188.

[0066] In some embodiments, the present invention provides a preferred mRNA for treating artemisia allergy, wherein the coding region of the mRNA encodes all preferred epitope amino acid sequences as shown in SEQ ID No. 185-186, and the reverse-translated cDNA sequence corresponding to the coding region amino acid sequence is shown in SEQ ID No. 21.

[0067] Furthermore, the reverse-translated cDNA sequences corresponding to the amino acid sequences shown in SEQ ID No. 22-41 are shown in SEQ ID No. 42-61.

[0068] In some embodiments, the present invention provides a preferred mRNA for treating type 1 diabetes (T1D), wherein the coding region of the mRNA encodes all preferred epitope amino acid sequences as shown in SEQ ID No. 22-41, and the reverse-translated cDNA sequence corresponding to the coding region amino acid sequence is shown in SEQ ID No. 62, with a GC content of 55.88%.

[0069] Furthermore, the reverse-translated cDNA sequences corresponding to the amino acid sequences shown in SEQ ID No. 64-69 and 77-79 are shown in SEQ ID No. 70-75 and 80-82.

[0070] In some embodiments, the present invention provides a preferred mRNA for myasthenia gravis (MG), wherein the coding region of the mRNA encodes all preferred epitope amino acid sequences as shown in SEQ ID No. 64-69, 77-79, and the reverse-translated cDNA sequence corresponding to the coding region amino acid sequence is shown in SEQ ID No. 84, with a GC content of 55.08%.

[0071] Furthermore, the reverse-translated cDNA sequences corresponding to the amino acid sequences shown in SEQ ID No. 86–93 are shown in SEQ ID No. 94–101.

[0072] In some embodiments, the present invention provides a preferred mRNA for toxic diffuse goiter (GD disease), wherein the coding region of the mRNA encodes all preferred epitope amino acid sequences as shown in SEQ ID No. 86-93, and the reverse-translated cDNA sequence corresponding to the coding region amino acid sequence is shown in SEQ ID No. 102, with a GC content of 58.63%.

[0073] Furthermore, the reverse-translated cDNA sequences corresponding to the amino acid sequences shown in SEQ ID No. 103-112 are shown in SEQ ID No. 113-122.

[0074] In some embodiments, the present invention provides a preferred mRNA for systemic lupus erythematosus (SLE), wherein the coding region of the mRNA encodes all preferred epitope amino acid sequences as shown in SEQ ID No. 103-112, and the reverse-translated cDNA sequence corresponding to the coding region amino acid sequence is shown in SEQ ID No. 124, with a GC content of 57.48%.

[0075] Furthermore, the reverse-translated cDNA sequences corresponding to the amino acid sequences shown in SEQ ID No. 126–131 are shown in SEQ ID No. 132–137.

[0076] In some embodiments, the present invention provides a preferred mRNA for primary cholangitis (PBC), wherein the coding region of the mRNA encodes all preferred epitope amino acid sequences as shown in SEQ ID No. 126-131, and the reverse-translated cDNA sequence corresponding to the coding region amino acid sequence is shown in SEQ ID No. 139, with a GC content of 59.87%.

[0077] Furthermore, the reverse-translated cDNA sequences corresponding to the amino acid sequences shown in SEQ ID No. 143-150 are shown in SEQ ID No. 151-158.

[0078] In some embodiments, the present invention provides a preferred mRNA for multiple sclerosis (MS), wherein the coding region of the mRNA encodes all preferred epitope amino acid sequences as shown in SEQ ID No. 143-150, and the reverse-translated cDNA sequence corresponding to the coding region amino acid sequence is shown in SEQ ID No. 159, with a GC content of 60.37%.

[0079] Furthermore, the reverse-translated cDNA sequences corresponding to the amino acid sequences shown in SEQ ID No. 164-173 are shown in SEQ ID No. 174-183.

[0080] In some embodiments, the present invention provides a preferred mRNA for Sjögren's syndrome (SS), wherein the coding region of the mRNA encodes all preferred epitope amino acid sequences as shown in SEQ ID No. 164-173, and the reverse-translated cDNA sequence corresponding to the coding region amino acid sequence is shown in SEQ ID No. 184, with a GC content of 57.03%.

[0081] On the other hand, the present invention provides the use of the lipid nanoparticles as described above in the preparation of formulations that enhance the ability to target hepatic sinusoidal endothelial cells.

[0082] In some embodiments, the multivalent targeted LNPs provided by this invention can improve the ability to target hepatic sinusoidal endothelial cells compared to monovalent targeted LNPs and non-targeted LNPs.

[0083] In another aspect, the present invention provides the use of the lipid nanoparticles described above in the preparation of formulations that enhance the therapeutic effects of allergic or autoimmune diseases.

[0084] In some embodiments, the LNPs provided by this invention treat hay fever caused by birch pollen in allergic diseases. Monovalent, bivalent, trivalent, tetravalent, and pentavalent targeted LNPs carrying mRNAs that treat hay fever (the mRNAs encode six preferred epitopes of the Bet v1 protein, a major allergen in birch pollen) are intravenously injected into a mouse model of hay fever. Compared to hay fever mouse models that have not been injected with LNPs, the treated mice show increased levels of Foxp3 in their lymph nodes and mucosal tissues. + The increased proportion of regulatory T cells and decreased levels of IL-4, IL-5, and IL-13 indicate that monovalent or multivalent targeting of LNPs can induce Foxp3. + The generation of regulatory T cells (Tregs) allows these cells to migrate to draining lymph nodes and mucosal tissues, thereby inducing tolerance responses to allergens. On the other hand, it can inhibit Th2 cells, reduce IgE production, lower IL-4, IL-5, and IL-13 levels, and limit the degranulation of mast cells and eosinophils, thus effectively controlling allergic reactions. Specifically, Foxp3 mice injected with trivalent LNPs (components including PS + DSPE-PEG2K-trimannose + DSPE-PEG2K-GlcNAc) showed improved allergic response. + The highest proportion of regulatory T cells and the lowest levels of IL-4, IL-5, and IL-13 resulted in the best therapeutic effect. Therefore, the LNPs provided by this invention, which encapsulate allergen peptide epitopes, can prevent or treat allergic diseases.

[0085] In some embodiments, the LNPs provided by this invention treat type 1 diabetes mellitus, an autoimmune disease. Monovalent, bivalent, trivalent, tetravalent, and pentavalent targeted LNPs, carrying mRNAs encoding T1D (20 preferred human epitopes of T1D-related antigens), are intravenously injected into accelerated T1D NOD mouse models. Compared to accelerated T1D NOD mouse models without LNP injection, the mice treated showed reduced weight loss and decreased blood glucose levels. However, only mice injected with trivalent targeted LNPs (components including PS+DSPE-PEG2K-trimannose+DSPE-PEG2K-GlcNAc) achieved a blood glucose level close to normal, while exhibiting the lowest pancreatic islet tissue damage score. Therefore, the LNPs provided by this invention, containing peptide epitopes carrying autoantigens, can prevent or treat autoimmune diseases.

[0086] In some embodiments, the LNPs provided by this invention are also used to treat artemisia allergy, myasthenia gravis, toxic diffuse goiter, systemic lupus erythematosus, primary cholangitis, multiple sclerosis, and Sjögren's syndrome. The LNPs respectively encapsulate mRNAs encoding all preferred epitopes of antigens that trigger various allergic or autoimmune diseases. The therapeutic effects of injecting trivalent targeted LNPs (components containing PS+DSPE-PEG2K-trimannose+DSPE-PEG2K-GlcNAc) into mouse models of various diseases are significantly better than those of monovalent or other multivalent targeted LNPs.

[0087] The present invention has the following beneficial effects:

[0088] 1. The LNPs provided by this invention have high targeting precision and uptake efficiency: innovative multivalent targeting design (targeting auxiliary lipids + at least one ligand-modified PEGylated lipids) accurately binds to multiple receptors on the surface of LSECs, resulting in a higher proportion of liver-specific uptake.

[0089] 2. This invention has screened out a better composition of LNPs, further improving targeting and therapeutic efficacy for allergic and autoimmune diseases.

[0090] 3. The present invention optimizes the mRNA drug loaded with LNPs, wherein the coding region of the mRNA includes at least one preferred antigenic epitope;

[0091] 4. This invention screens preferred epitopes of antigens that trigger various allergies or autoimmune diseases, and provides the corresponding reverse-translated cDNA sequences of the epitopes. It also provides the most preferred mRNAs for treating birch hay fever, artemisia allergy, type 1 diabetes, myasthenia gravis, toxic diffuse goiter, systemic lupus erythematosus, primary cholangitis, multiple sclerosis, and Sjögren's syndrome.

[0092] 5. The LNPs provided by this invention can be developed into vaccines and immunomodulators for the prevention or treatment of allergic diseases and autoimmune diseases, and have the characteristics of strong antigen specificity and low side effects. Attached Figure Description

[0093] Figure 1 shows the fluorescence distribution of mice injected with LNPs prepared by different cationic ionized lipids in Example 2 using an in vivo imaging system (IVIS).

[0094] Figure 2 shows the fluorescence intensity results of in vivo imaging (IVIS) of mice injected with LNPs prepared by different cationic ionized lipids in Example 2.

[0095] Figure 3 shows the different fluorescence distributions in various organs of mice injected with LNPs prepared by different cationic ionized lipids in Example 2.

[0096] Figure 4 shows the different fluorescence percentages in various organs and the different fluorescence intensities in the liver of mice injected with LNPs prepared by different cationic ionized lipids in Example 2.

[0097] Figure 5 shows the different fluorescence percentages in various organs of mice injected with SM102-LNPs (multivalent target), MC3-LNPs (multivalent target), SM102-LNPs (monovalent target 1), SM102-LNPs (monovalent target 2) and SM102-LNPs (monovalent target 3) in Example 3.

[0098] Figure 6 shows the intensity of different fluorescence in the livers of mice injected with SM102-LNPs (multivalent target), MC3-LNPs (multivalent target), SM102-LNPs (monovalent target 1), SM102-LNPs (monovalent target 2) and SM102-LNPs (monovalent target 3) in Example 3.

[0099] Figure 7 is a schematic diagram of the complete mRNA structure in Example 6;

[0100] Figure 8 shows the experimental protocol for constructing an accelerated T1D NOD mouse model in Example 7 (red arrows), as well as the experimental protocol for treating the accelerated T1D NOD mouse model and its monitoring protocol (blue and green arrows).

[0101] Figure 9 is a schematic diagram of the distribution of different epitopes in the α subunit sequence of AChR in Example 9;

[0102] Figure 10 shows the detection results of the proliferation response of different epitopes of the α subunit of AChR in four different MG patients (Pt 3 / 7 / 10 / 11) in Example 9;

[0103] Figure 11 shows the frequency at which different sequence peptides of the α subunit of AChR were recognized by seven responsive MG patients in Example 9 (indicated by light gray bars with dots).

[0104] Figure 12 shows the correlation between disease severity scores and MuSK-Ig1 epitope pattern responsiveness in 22 Italian patients with type G in Example 9;

[0105] Figure 13 shows the single-cell sequencing results of SLE patients and healthy volunteers, as well as SLE mice and wild-type mice in Example 11;

[0106] Figure 14 shows the binding of the RO60 epitope to the HLA allele in Example 14;

[0107] Figure 15 shows the binding of the TRIM21 epitope to the HLA allele in Example 14;

[0108] Figure 16 shows the binding of the lupus La epitope to the HLA allele in Example 14;

[0109] Figure 17 shows the binding of the SPTN1 epitope to the HLA allele in Example 14. Detailed Implementation

[0110] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be noted that the embodiments described below are intended to facilitate the understanding of the present invention and do not limit it in any way.

[0111] Example 1: Preparation of a lipid nanoparticle

[0112] The lipid nanoparticles (LNPs) prepared in this embodiment comprise SM102, phosphatidylserine (PS), DSPE-PEG2K-trimannose, DSPE-PEG2K-GlcNAc, cholesterol, and an encapsulated RNA drug. SM102 is a cationic ionized lipid, phosphatidylserine is an accessory lipid, and DSPE-PEG2K-trimannose and DSPE-PEG2K-GlcNAc are ligand-modified PEGylated lipids. The RNA drug is an RNA drug for treating allergic or autoimmune diseases. The RNA drug is mRNA, comprising a 5'-cap structure, a 5'-UTR, a coding region, a 3'-UTR, and a poly A tail, wherein the coding region encodes at least one epitope of an antigen causing an allergic or autoimmune disease.

[0113] The specific preparation method is as follows: The components for preparing LNPs are prepared using a microfluidic method, with an ethanol phase to aqueous phase ratio of 1:3 (the ethanol phase contains all lipid components, and the aqueous phase contains the RNA drug), a total flow rate of 12 mL / min, and a molar mass ratio of SM102:PS:Chol:DSPE-PEG2K-trimannose:DSPE-PEG2K-GlcNAc of 50:10:38.5:0.75:0.75, wherein the N / P ratio of SM102 is 6, and the mass ratio of RNA drug to lipid nanoparticles is 1–10:20–100 (preferably 1:40). The prepared LNPs are dialyzed using PBS.

[0114] In this embodiment, PS, DSPE-PEG2K-trimannose, and DSPE-PEG2K-GlcNAc all have the ability to target receptors on the surface of LSECs, so the prepared LNPs are trivalent targeting LNPs.

[0115] Example 2: Screening of cationic ionized lipids and their N / P ratios in lipid nanoparticles

[0116] In this embodiment, LNPs were prepared according to the method of Example 1, except that the cationic ionized lipid SM102 was replaced with MC3. When the cationic ionized lipid was SM102, its N / P ratio was set to 2, 3, 4, and 6, respectively; when the cationic ionized lipid was MC3, its N / P ratio was set to 4. To facilitate subsequent characterization, experiments, and in vivo tracking of LNPs, the prepared LNPs were all loaded with two mRNAs encoding enhanced green fluorescent protein (EGFP) and luciferase (Luc), respectively. DiR was added to the LNP components for labeling (the molar ratio of DiR was 0.3; since the total molar ratio was 100%, the molar ratio of Chol was reduced from 38.5 to 38.2).

[0117] First, the physicochemical properties of the different LNPs prepared above were characterized, including encapsulation efficiency (EE), particle size (Size), polydispersity index (PDI), zeta potential and acid dissociation constant (pKa). The characterization results are shown in Table 1 below.

[0118] Table 1. Characterization results of the physicochemical properties of LNPs prepared from different cationic ionized lipids.

[0119] As shown in Table 1, when the cationic ionized lipid SM102-N / P ratio is 2–6, the encapsulation efficiency of the prepared LNPs is higher than that of LNPs prepared when the cationic ionized lipid MC3-N / P ratio is 4. Therefore, the preferred cationic ionized lipid is SM102. Furthermore, comparing the physicochemical properties of LNPs prepared from SM102 with different N / P ratios, the LNPs prepared with an N / P ratio of 6 exhibit the highest encapsulation efficiency, smallest particle size, and most uniform particle size distribution. Considering all factors, an N / P ratio of 6 is the optimal choice.

[0120] Furthermore, 20 μg mRNA / mouse of LNPs prepared from different cationic ionized lipids was injected subcutaneously into the tail of mice, with an equal volume of PBS injected as a control. Six hours later, the location of LNPs and mRNA expression were observed using an in vivo imaging system (IVIS). The results are shown in Figures 1 and 2. Due to the strong background signal of EGFP, it is not of reference value. However, comparing the qualitative and quantitative fluorescence intensity results of DiR and Luc, it was found that when the cationic ionized lipid in the LNPs was SM102, and the N / P ratio of SM102 was 6, the fluorescence signals of DiR and Luc were stronger and concentrated in the mouse liver.

[0121] Further, mice in each group were sacrificed, and their organs were harvested and imaged to compare their luminescence or fluorescence intensity. Organs included the kidneys, spleen, lungs, liver, heart, and lymph nodes. The results are shown in Figures 3-4. Comparing the fluorescence percentage of DiR in each organ and the fluorescence intensity of DiR in the liver of each group of mice showed little difference in the fluorescence percentage of DiR in the liver among the groups. However, mice injected with SM102-N / P6 LNPs had the highest DiR fluorescence intensity in their livers. Comparing the fluorescence percentage of Luc and EGFP in each organ and the fluorescence intensity of Luc and EGFP in the liver of each group of mice showed that mice injected with SM102-N / P6 LNPs had the highest fluorescence percentage of both Luc and EGFP in the liver, as well as the highest fluorescence intensity in the liver. These results indicate that when the cationic ionized lipid in the LNPs is SM102, and the N / P ratio of SM102 is 6, its ability to specifically target the liver and translate mRNA is strongest.

[0122] In summary, the preferred cationic ionized lipid in LNPs is SM102, and the preferred N / P ratio of SM102 is 6.

[0123] Example 3: In vivo targeting validation of LNPs

[0124] In this embodiment, two types of LNPs prepared in Example 2 when the cationic ionized lipids were SM102-N / P=6 and MC3-N / P=4 were selected as experimental subjects and were respectively denoted as SM102-LNPs (trivalent targeting 1, i.e. PS, DSPE-PEG2K-trimannose and DSPE-PEG2K-GlcNAc targeting LSECs) and MC3-LNPs (trivalent targeting 2, i.e. PS, DSPE-PEG2K-trimannose and DSPE-PEG2K-GlcNAc targeting LSECs). Simultaneously, the combination of DSPE-PEG2K-trimannose and DSPE-PEG2K-GlcNAc in the SM102-LNPs (trivalent targeting 1) is replaced with a single DSPE-PEG2K, denoted as SM102-LNPs (monovalent targeting 1, i.e., only PS targets LSECs); the PS in the SM102-LNPs (trivalent targeting 1) is replaced with a common auxiliary lipid DSPC, and the combination of DSPE-PEG2K-trimannose and DSPE-PEG2K-GlcNAc is replaced with a single DSPE-PEG2K. - Trimannose, denoted as SM102-LNPs (monovalent targeting 2, i.e., only DSPE-PEG2K-trimannose targets LSECs); in the aforementioned SM102-LNPs (trivalent targeting 1), PS is replaced with ordinary auxiliary lipid DSPC, and the combination of DSPE-PEG2K-trimannose and DSPE-PEG2K-GlcNAc is replaced with a single DSPE-PEG2K-GlcNAc, denoted as SM102-LNPs (monovalent targeting 3, i.e., only DSPE-PEG2K-GlcNAc targets LSECs).

[0125] The five LNPs were injected subcutaneously into the tail of mice at a dose of 20 μg mRNA per mouse. Six hours later, the location of the LNPs and the expression of their mRNA were observed using an in vivo imaging system (IVIS). Subsequently, the mice in each group were sacrificed, and their organs were harvested and imaged to compare the luminescence or fluorescence intensity. The organs included the kidney, spleen, lung, liver, heart, and lymph nodes.

[0126] The fluorescence percentages of DiR, Luc, and EGFP in various organs of the five groups of mice, as well as the fluorescence intensities of DiR, Luc, and EGFP in the liver, are shown in Figures 5 and 6. Comparing the fluorescence percentage of DiR in Figure 1 and the fluorescence intensity of DiR in the liver in Figure 2, the fluorescence percentage and intensity of DiR in the liver of mice injected with SM102-LNPs (trivalent targeting 1) and MC3-LNPs (trivalent targeting 2) were significantly higher than those of mice injected with the other three monovalent targeting LNPs. Comparing the fluorescence percentages of Luc and EGFP in Figure 1 and the fluorescence intensities of Luc and EGFP in the liver of mice injected with SM102-LNPs (trivalent targeting 1) and MC3-LNPs (trivalent targeting 2) were also significantly higher than those of mice injected with the other three monovalent targeting LNPs. The above results indicate that multivalent targeting LNPs have a stronger ability to specifically target the liver and translate mRNA than monovalent targeting LNPs. At the same time, SM102-LNPs (trivalent targeting 1) have a better targeting effect than MC3-LNPs (multivalent targeting 2).

[0127] Example 4: The effect of assisting lipids on the targeting of lipid nanoparticles

[0128] In Example 1, the PS, DSPE-PEG2K-trimannose, and DSPE-PEG2K-GlcNAc in the LNPs prepared all have the ability to target receptors on the surface of LSECs, with PS serving as an accessory lipid. This example will replace the type of accessory lipid described above. Replaceable accessory lipids include DOPS (1,2-dioleoyl-sn-glycerol-3-phosphate-L-serine), DPPS (1,2-dispalmitoyl-sn-glycerol-3-phosphate-L-serine), and DSPC (distearylphosphatidylcholine). DOPS and DPPS have the ability to target receptors on the surface of LSECs; DOPS targets the SR-E1 / LOX-1 and SR-H1 / STABILIN-1 receptors of LSECs, while DPPS targets the SR-E1 / LOX-1 receptor of LSECs. Therefore, in this embodiment, when the auxiliary lipids are PS, DOPS and DPPS, the prepared LNPs are trivalent targeting LNPs, that is, the auxiliary lipids and the two ligand-modified PEGylated lipids can target LSECs; when the auxiliary lipid is DSPC, the prepared LNPs are bivalent targeting LNPs, that is, the auxiliary lipids do not target LSECs, and only the two ligand-modified PEGylated lipids can target LSECs.

[0129] Four types of LNPs were prepared by PS and its alternative auxiliary lipids according to the method in Example 1. Each LNP carried two mRNAs encoding EGFP and Luc, respectively. DiR was added to the LNP components for labeling (the molar ratio of DiR was 0.3, and since the total molar ratio was 100%, the molar ratio of Chol was reduced from 38.5 to 38.2).

[0130] The four LNPs were injected subcutaneously into the tail of mice at a dose of 5 μg / mouse. Six hours later, the mice in each group were sacrificed and their organs were harvested. The organs were imaged and their luminescence or fluorescence intensity was compared. The organs included the kidney, spleen, lung, liver, heart, and lymph nodes. The results showed that the fluorescence proportion in the liver of each group of mice was significantly higher than that in the other organs. The fluorescence proportions of DiR, Luc, and EGFP in the liver of each group of mice are shown in Table 2 below.

[0131] Table 2. Fluorescence percentages of DiR, Luc, and EGFP in the livers of mice in each group.

[0132] According to the data in Table 2, when the auxiliary lipids were targeted PS, DOPS, and DPPS, the fluorescence proportions of DiR, Luc, and EGFP in the liver of the prepared LNPs were all higher than those of LNPs prepared without targeted DSPC. This indicates that the addition of targeted auxiliary lipids improved the liver-specific targeting effect of LNPs. Furthermore, the LNPs prepared with PS showed the highest fluorescence proportions of DiR, Luc, and EGFP, indicating that these LNPs had the best targeting effect on the liver and the strongest ability to translate mRNA in the liver. Therefore, PS is the preferred targeted auxiliary lipid.

[0133] Example 5: Effect of ligand-modified PEGylated lipids on the targeting of lipid nanoparticles

[0134] This embodiment, based on Example 4, fixes the auxiliary lipid as PS (preferred) and changes the quantity and type of ligand-modified PEGylated lipids. In addition to DSPE-PEG2K-trimannose and DSPE-PEG2K-GlcNAc, the ligand-modified PEGylated lipids also include DSPE-PEG2K-mannose, DSPE-PEG2K-GalNAc-4-sulfate, DSPE-PEG2K-L-fucose, and DSPE-PEG2K-hyaluronic acid.

[0135] LNPs were prepared by combining PEGylated lipids modified with different ligands, as shown in Table 3 below.

[0136] Table 3. Combinations of PEGylated lipids with different ligand modifications

[0137] Since the auxiliary lipid in this embodiment has been fixed as PS that can target LSECs, the LNPs prepared by the different ligand-modified PEGylated lipid combinations in Table 3 above are at least bivalent LNPs. That is, when there is one type of ligand-modified PEGylated lipid, it is a bivalent LNP; when there are two types of ligand-modified PEGylated lipid, it is a trivalent LNP, and so on. PEGylated lipids with one ligand modification (molar mass ratio of one lipid is 1.5, preparing divalent LNPs) were prepared from LNPs 1 to 6; PEGylated lipids with two ligand modifications (molar mass ratio of two lipids is 0.75:0.75, preparing trivalent LNPs) were prepared from LNPs 7 to 21; PEGylated lipids with three ligand modifications (molar mass ratio of three lipids is 0.5:0.5:0.5, preparing tetravalent LNPs) were prepared from LNPs 22 to 25; and PEGylated lipids with four ligand modifications (molar mass ratio of four lipids is 0.375:0.375:0.375:0.375, preparing pentavalent LNPs) were prepared from LNPs 25 to 31. Each LNP contains two mRNAs encoding EGFP and Luc, respectively. DiR is added to the LNPs for labeling (the molar ratio of DiR is 0.3, and since the total molar ratio is 100%, the molar ratio of Chol is reduced from 38.5 to 38.2).

[0138] The above 31 LNPs were injected subcutaneously into the tail of mice at 5 μg / mouse. After 6 hours, the mice in each group were sacrificed and their organs were harvested. The organs were imaged and their luminescence or fluorescence intensity was compared. The organs included the kidney, spleen, lung, liver, heart and lymph nodes. The results showed that the fluorescence proportion in the liver of each group of mice was significantly higher than that in the other organs. The fluorescence proportions of DiR, Luc and EGFP in the liver of each group of mice are shown in Table 4 below.

[0139] Table 4. Fluorescence percentages of DiR, Luc, and EGFP in the livers of mice in each group.

[0140] Based on the data in Table 4, comparing the fluorescence proportions of DiR, Luc, and EGFP in the livers of mice injected with LNPs 1–6, 7–21, 22–25, and 26–31, it was found that LNPs 1–6, 22–25, and 26–31 were less effective at targeting the liver and less capable of translating mRNA in the liver than LNPs 7–21. This indicates that LNPs prepared with one, three, or four types of ligand-modified PEGylated lipids were less effective at targeting the liver and less capable of translating mRNA in the liver than LNPs prepared with two types of ligand-modified PEGylated lipids. Further comparison of the experimental data for LNPs 7–21 revealed that the fluorescence proportions of DiR, Luc, and EGFP in the livers of mice injected with LNP 7 were higher than those injected with the remaining LNPs 8–21, indicating that LNP 7 had the best effect at targeting the liver and the strongest ability to translate mRNA in the liver. The results above show that more PEGylated lipids modified with ligands do not necessarily lead to better targeting effects. The combination of PEGylated lipids modified with two ligands has a better targeting effect, with DSPE-PEG2K-trimannose + DSPE-PEG2K-GlcNAc being the optimal choice. Too many ligands will reduce the targeting efficiency due to steric hindrance and receptor competition.

[0141] Based on the above experimental results, when the auxiliary lipid is PS, the preferred ligand-modified PEGylated lipids are DSPE-PEG2K-trimannose and DSPE-PEG2K-GlcNAc.

[0142] Example 6: Treatment of birch hay fever

[0143] This embodiment verifies that the LNPs prepared in Example 1 can be used to treat hay fever caused by birch pollen.

[0144] 1. Prediction, design, and mRNA preparation of antigenic epitope peptides

[0145] Bet v1 protein is a major birch allergen identified in the pollen of birch and weeping birch. Its amino acid sequence is shown in SEQ ID NO.1. It induces IgE binding in more than 95% of patients with birch pollen allergies. IgE can cause type 1 hypersensitivity reactions, which manifest as allergic rhinitis, allergic conjunctivitis, hay fever, allergic asthma, bee venom allergy, and food allergy.

[0146] MHC-II epitopes were searched using the Immune Epitope Database (IEDB) and the NetMHCIIpan 4.1 EL tool. Table 5 below lists the epitopes in the Bet v1 protein that have a high binding percentile with the aforementioned MHC-II epitopes.

[0147] Table 5. Epitopes in the Bet v1 protein that have high binding percentages to the aforementioned MHC-II epitopes.

[0148] Based on HLA allele coverage, six preferred Bet v1 epitope peptides were selected from Table 5 above, as shown in Table 6 below.

[0149] Table 6. Six preferred epitope peptides of Bet v1

[0150] The six epitope peptides in Table 6 were added with flanking sequences and reverse-translated into cDNA for mRNA design, as shown in Table 7 below.

[0151] Table 7. Reverse translation of preferred epitope peptides

[0152] Further, a complete cDNA sequence is constructed, the elements of which include a 5' UTR (containing the Kozak sequence), a coding region, and a 3' UTR.

[0153] The cDNA sequence of the 5'UTR is as follows:

[0154] The cDNA sequence of the 3'UTR is as follows:

[0155] The coding region encodes the target sequence and at least one epitope from Table 7; in this embodiment, six epitopes are selected. The six cDNA sequences from Table 13 are linked together using a flexible linker (in this embodiment, the GGPPG linker is selected, whose reverse-translated cDNA sequence is GGCCCGGGCCCGGGC) as part of the coding region to facilitate peptide cleavage upon epitope release within the cell. The target sequence allows the antigenic epitope to enter the MHC-II endosomal compartment for peptide presentation to Treg precursor cells. In this embodiment, the target sequence is selected as the 1-80 amino acid fragment of the invariant chain (Ii), as shown in SEQ ID NO. 14, abbreviated as Ii(1-80), whose reverse-translated cDNA sequence is shown in SEQ ID NO. 15. The protein subdomain of this fragment (a molecular chaperone protein of MHC-II, which helps load the peptide into MHC-II for antigen presentation) allows the peptide epitope to enter MHC-II from the cytoplasm for CD4 cleavage. + T cell presentation. The target sequence can be Ii(1-80) or transferrin receptor.

[0156] Furthermore, codon optimization was performed on the coding region cDNA to achieve optimal gene expression of the non-human cell tRNA library compared to human cells. This was accomplished using the GenScript online codon optimization tool. The optimized nucleotide sequence of the coding region is shown in SEQ ID NO.16, with a GC content of 56.74%, wherein the sequence encoding Ii(1-80) is upstream of the sequence encoding the six epitopes.

[0157] Further, the optimized cDNA is transcribed into mRNA. During transcription, uridine is replaced with N1-methylpseudouridine; after transcription, a 5' cap and a Poly A tail are added to the mRNA. The 5' cap is either CleanCap or ARCA, and the Poly A tail is 100-120 nucleotides in length. The complete mRNA sequence is shown in Figure 7.

[0158] Preferably, mRNA can be transcribed by inserting the coding region cDNA into the pTNT plasmid.

[0159] 2. Preparation of LNPs

[0160] The LNPs prepared in Example 3, where the auxiliary lipid was DSPC and the ligand-modified PEGylated lipid was a single DSPE-PEG2K-trimannose, were designated as monovalent targeting LNPs. LNPs 1, 7 (same as Example 1), 22, and 26 in Table 4 of Example 5 were designated as bivalent targeting LNPs, trivalent targeting LNPs, tetravalent targeting LNPs, and pentavalent targeting LNPs, respectively. All LNPs contained the mRNA prepared in this example (the cDNA sequence of the coding region is shown in SEQ ID NO. 16).

[0161] 3. LNPs are used to treat hay fever caused by birch pollen.

[0162] Establishment of a mouse model of hay fever: A pollen allergy animal model was established by sensitization and immunization with 25 μg of birch pollen extract (SC). The challenge step was inhalation of 1% Bet v1 purified protein.

[0163] The monovalent, bivalent, trivalent, tetravalent, and pentavalent targeted LNPs were administered intravenously at 5 μg / mouse to hay fever model mice, respectively. The control group consisted of model mice injected with an equal amount of blank LNPs. After a period of treatment, Foxp3 levels in the lymph nodes and mucosal tissues of each group of mice were measured. + The percentage of regulatory T cells and the expression levels of inflammatory factors IL-4, IL-5 and IL-13 were detected, and the results are shown in Table 8 below.

[0164] Table 8, Foxp3 +The proportion of regulatory T cells and the expression levels of inflammatory factors IL-4, IL-5 and IL-13.

[0165] LNP therapy can induce Foxp3. + The generation of regulatory T cells (Tregs) allows these Tregs to migrate to draining lymph nodes and mucosal tissues, thereby inducing tolerance responses to allergens. On the other hand, it can inhibit Th2 cells, reduce IgE production, lower the levels of IL-4, IL-5, and IL-13, and limit the degranulation of mast cells and eosinophils, thus effectively controlling allergic reactions. Table 8 shows that, compared with the control group, mice injected with targeted LNPs showed increased levels of Foxp3... + The proportions of Tregs were significantly increased, while the levels of IL-4, IL-5, and IL-13 were significantly decreased. Comparing the mice injected with targeted LNPs, the trivalent targeted LNP (same as in Example 1) Foxp3... + Tregs had the highest proportion and the lowest levels of IL-4, IL-5, and IL-13, indicating that they were the most effective in treating allergic reactions.

[0166] The above results indicate that the LNPs containing allergen peptide epitopes can prevent or treat allergic diseases, and are particularly suitable for hay fever or oral allergy syndrome caused by birch pollen or birch extract, thus showing great application potential.

[0167] Example 7: Treatment of Artemisia argyi allergy

[0168] This embodiment verifies that the LNPs prepared in Example 1 can be used to treat Artemisia argyi allergy.

[0169] 1. Prediction, design, and mRNA preparation of antigenic epitope peptides

[0170] Bet v1 protein is a major birch allergen identified in the pollen of *Betula pulcherrima* and *Betula pulcherrima*, and it induces IgE binding in over 95% of patients with birch pollen allergies. Immunoglobulin E (IgE) can cause type 1 hypersensitivity reactions, manifesting as allergic rhinitis, allergic conjunctivitis, hay fever, allergic asthma, bee venom allergy, and food allergies.

[0171] This embodiment utilizes IEDB with the NetMHCIIpan 4.1 eluent (EL) predictor to predict the binding of epitopes in Art v1 to MHC-II. While binding affinity can assess the ability of a peptide to bind to MHC molecules, eluent (EL) prediction further integrates the possibility that the peptide has been naturally processed and presented, making the prediction results more likely to identify true extracellular T cell sites. Table 9 below shows the ranking of epitope prediction results:

[0172] Table 9. Binding of epitopes to MHC-II in Art v1

[0173] Next, based on the data in Table 9 above, and taking into account hydrophilicity, hydrophobicity and water solubility, the splicing epitopes in Table 10 below were selected.

[0174] Table 10. Segmentation Table Positions

[0175] Further, a complete cDNA sequence is constructed, the cDNA comprising a 5' UTR (including the Kozak sequence), a coding region, and a 3' UTR. The nucleotide sequences of the 5' UTR and 3' UTR are the same as in Example 6. The coding region encodes a target sequence and at least one epitope sequence from Tables 9-10, with multiple epitopes linked by linkers as part of the coding region. The leader sequence and linker sequences are the same as in Example 6. Further codon optimization is performed on the cDNA coding region, using the same optimization method as in Example 6. Further, the optimized cDNA is transcribed into mRNA, using the same method as in Example 6.

[0176] 2. Preparation of LNPs

[0177] The LNPs prepared in Example 3, where the auxiliary lipid was DSPC and the ligand-modified PEGylated lipid was a single DSPE-PEG2K-trimannose, are designated as monovalent targeting LNPs. LNPs 1, 7 (same as Example 1), 22, and 26 in Table 4 of Example 5 are designated as bivalent targeting LNPs, trivalent targeting LNPs, tetravalent targeting LNPs, and pentavalent targeting LNPs, respectively. All LNPs contain the mRNA prepared in this example (the cDNA sequence of the coding region encodes the two polypeptides spliced ​​to epitope 1 in Table 10 above, i.e., the coding region cDNA sequence is shown in SEQ ID NO. 21).

[0178] 3. LNPs are used to treat artemisia allergy.

[0179] A pollen allergy animal model was established using a subcutaneous (sc) sensitization method with Artemisia argyi pollen extract. The specific method was as follows: BALB / c mice were sensitized three times by subcutaneous injection of 25 μg on days 1, 8, and 15. Then, on days 22, 23, and 24, they were challenged by inhaling 1% Art v1 purified protein daily, and blood samples were collected to detect total IgE, antigen-specific IgE, and antigen-specific IgG to verify the sensitization effect. Then, on days 26, 28, 30, and 32, mice were subcutaneously injected with 2 μg of the aforementioned monovalent, bivalent, trivalent, tetravalent, and pentavalent targeted LNPs, respectively. Normal mice (blank control group) and model mice injected with an equal amount of blank LNPs (positive control group) were used for comparison. Next, on days 37, 38, 39, 40, and 41, mice were again inhaled with 1% Art v1 purified protein daily. On day 42, the animals were sacrificed, and blood, BALF, lung, and spleen samples were collected for further analysis. Paraffin sections were prepared from collected lung tissue. Hematoxylin and eosin (H&E) staining was used to assess inflammatory cell infiltration in the lungs of mice in each group. Results showed that, compared with the blank control group, the positive control group mice exhibited significant inflammatory cell infiltration in the bronchi; compared with the positive control group, mice treated with different LNPs showed significantly reduced inflammatory cell infiltration and tracheal fibrosis in the bronchi. The inflammatory infiltration of the bronchi in each group of mice was scored, and the scoring results are shown in Table 11 below.

[0180] Table 11. Inflammatory infiltration scores of bronchial pulmonary arteries in mice of each group

[0181] According to the data in Table 11, compared with the positive control, the bronchial inflammation infiltration scores of mice treated with the five LNPs all decreased, but only the scores of mice injected with trivalent targeted LNPs were closest to those of the blank control (normal mice).

[0182] Cell count, IgE level, IgG level, IgG1 level, IgG2a level, CD4 in BALF + CD25 + FoxP3 + The cell percentage and IL-10 content are shown in Table 12 below.

[0183] Table 12. Cell count, IgE level, IgG level, IgG1 level, IgG2a level, and CD4 count in BALF of mice in each group. + CD25 + FoxP3 + Cell percentage and IL-10 content

[0184] According to the data in Table 12, compared with the positive control, the total number of cells, IgE level, IgG level, and IgG1 level in the BALF of mice treated with the five LNPs decreased, while the proportion of Tregs and IL-10 content increased, and the IgG2a level remained basically unchanged. However, the total number of cells, IgE level, IgG level, and IgG1 level in the BALF of mice treated with trivalent targeted LNPs were the lowest, and the proportion of Tregs (CD4+) was also highest. + CD25 + FoxP3 + The highest proportion of cells and IL-10 content resulted in the best treatment effect.

[0185] This demonstrates that subcutaneous injection of trivalent targeting LNPs expressing Art v1 Artemisia v1 pollen protein epitopes can significantly induce the production of Tregs and IL-10 cytokines, thereby inhibiting pollen protein-induced allergic lung inflammation.

[0186] 3. Screening of spliced ​​table positions

[0187] Furthermore, trivalent targeting LNPs were prepared from the mRNAs encoding different splicing epitopes in Table 10 according to the method in Example 1.

[0188] A pollen allergy animal model was established using a subcutaneous injection (sc) of Artemisia argyi pollen extract to induce sensitization, following the same method described above. LNPs containing mRNAs with different splicing epitopes were injected subcutaneously for treatment. Normal mice (blank control group) and model mice injected with the same amount of blank LNPs (positive control group) were compared. The bronchial inflammatory infiltration was scored in each group of mice, and the number of cells, IgE level, IgG level, IgG1 level, IgG2a level, and CD4 count in the bronchoalveolar lavage fluid (BALF) of each group were measured. + CD25 + FoxP3 + The percentage of cells (regulatory T cells (Tregs)) and the IL-10 content were detected using the same methods as above.

[0189] The results showed that, compared with the blank control group, the positive control group mice had significant inflammatory cell infiltration in the bronchi; compared with the positive control group, mice treated with LNPs encoding mRNAs of different splicing epitopes showed significantly reduced inflammatory cell infiltration and tracheal fibrosis in the bronchi. The inflammatory infiltration of the bronchi in each group of mice was scored, and the scoring results are shown in Table 13 below.

[0190] Table 13. Inflammatory infiltration scores of bronchial pulmonary arteries in mice of each group

[0191] According to the data in Table 13, compared with the positive control, the bronchial inflammation infiltration scores of mice treated with LNPs were all reduced, but only the scores of mice injected with LNPs encoding splicing epitope 1 were closest to those of the blank control (normal mice).

[0192] Cell count, IgE level, IgG level, IgG1 level, IgG2a level, CD4 in BALF + CD25 + FoxP3 + The percentage of cells (regulatory T cells (Tregs)) and the IL-10 content are shown in Table 14 below.

[0193] Table 14. Cell count, IgE level, IgG level, IgG1 level, IgG2a level, and CD4 count in BALF of mice in each group. + CD25 + FoxP3 + Cell percentage and IL-10 content

[0194] According to the data in Table 14, compared with the positive control, the total number of cells, IgE level, IgG level, and IgG1 level in the BALF of mice treated with LNPs encoding spliced ​​epitopes were decreased, while the proportion of Tregs and IL-10 content were increased, and the IgG2a level remained basically unchanged. However, the total number of cells, IgE level, IgG level, and IgG1 level in the BALF of mice treated with LNPs encoding spliced ​​epitope 1 were the lowest, while the proportion of Tregs and IL-10 content were the highest, indicating the best treatment effect.

[0195] Therefore, the preferred combination of epitopes for the mRNA coding region is the one that splices epitopes 1 in Table 10.

[0196] Further, APPGAAPPPAAGGSP was ultimately selected as preferred epitope 1 (SEQ ID NO. 185) and FCYFDCSKSPPGATP was selected as preferred epitope 2 (SEQ ID NO. 186). The reverse-translated cDNA sequence corresponding to preferred epitope 1 is shown in SEQ ID NO. 187, and the reverse-translated cDNA sequence corresponding to preferred epitope 2 is shown in SEQ ID NO. 188. The optimized coding region cDNA sequences are shown in SEQ ID NO. 18-21, respectively, with GC contents of 63.53%, 57.49%, 60.39%, and 57.79%, respectively. Among them, SEQ ID NO. 18 contains epitope 1, SEQ ID NO. 19 contains epitope 2, SEQ ID NO. 20 contains epitopes 1 and 2, and SEQ ID NO. 21 also contains two preferred epitopes.

[0197] Example 8: Treatment of Type 1 Diabetes

[0198] This embodiment verifies that the LNPs prepared in Example 1 can be used to treat type 1 diabetes (T1D).

[0199] 1. Prediction, design, and mRNA preparation of antigenic epitope peptides

[0200] Type 1 diabetes (T1D) stems from the disruption of immune tolerance, where autoreactive T cells (particularly CD4+ and CD8+ T cells) recognize and attack insulin-producing beta cells. These cells target a variety of islet antigens, including insulin (INS), glutamate decarboxylase 65 (GAD65), insulinoma-associated protein 2 (IA-2), ZnT8, and IGRP. Extensive evidence supports the crucial role of CD4+ T cells in T1D. Their importance is highlighted by strong genetic associations with specific HLA class II haplotypes. CD4+ T cells not only drive the autoimmune attack but also promote the production of islet-specific autoantibodies—antibodies that often develop before clinical symptoms appear and spread as beta cell damage progresses. Furthermore, these cells are essential for generating Tregs that may suppress the disease, a mechanism that supports strategies such as targeting LSECs with tolerance nanoparticles to present antigens to naive T cells using MHC-II.

[0201] This embodiment selects 20 human epitopes from 8 types of T1D-related antigens. The related antigens include insulin, GAD65, chromogranin A, IAPP, IGRP, ZnT8, IA2 and InsB mimic epitopes, and HIPs (fusion epitopes of insulin peptide with other antigen fragments).

[0202] This embodiment focuses on epitope screening of sequences that interact with HLA-DRB1*0401—this allele pairs with DQA1*03:01 or DQB1*03:02 to form the high-risk "DR4" haplotype (Noble et al. J Autoimmunity 64, 2015:101e112). Notably, DRB1*04:01 has a unique motif that can present human T1D-related insulin antigen epitopes (Tait et al. Eur J Immunogenet. 1995; 22(4):289-97). However, the T1D risk stems from the synergistic effect of DR / DQ molecules: DRB4*01:01 (DRB4), as a minor HLA-DR product, is a component of the high-risk DR4 / DQ8 haplotype (James et al. J Immunol 2018; doi:10.4049 / jimmunol.1800723). DRB4 also shares structural homology with HLA-DR4 alleles associated with other autoimmune diseases such as multiple sclerosis, rheumatoid arthritis, and Lyme disease. This embodiment utilizes DRB1*0401 and its associated DR4 haplotype epitopes to develop a T1D treatment regimen using tolerable nanoparticles to deliver multi-epitope mRNA.

[0203] In this embodiment, the preferred peptides (epitopes) that can bind to DRB1*0401 and its associated DR4 haplotypes among the eight T1D-related antigens were predicted using the NetmHciipan_el 4.1 server. The prediction results are shown in Table 15 below.

[0204] Table 15. Predicted optimal epitopes for 8 T1D-related antigens

[0205] Further, a complete cDNA sequence was constructed, the elements of which include a 5' UTR (including the Kozak sequence), a coding region, and a 3' UTR. The nucleotide sequences of the 5' UTR and 3' UTR are the same as in Example 6. The coding region encodes a target sequence and at least one epitope sequence from Table 15, with multiple epitopes linked by linkers as part of the coding region. The leader sequence and linker sequences are the same as in Example 6. In this example, the coding region of the cDNA sequence contains all the reverse-translated cDNA sequences from Table 15, and the cDNA in the coding region is further codon-optimized using the same method as in Example 6. The optimized coding region cDNA sequence is shown in SEQ ID NO. 62, with a GC content of 55.88%.

[0206] Furthermore, the optimized cDNA is transcribed into mRNA, using the same method as in Example 6.

[0207] 2. Preparation of LNPs

[0208] The LNPs prepared in Example 3, where the auxiliary lipid was DSPC and the ligand-modified PEGylated lipid was a single DSPE-PEG2K-trimannose, were designated as monovalent targeting LNPs. LNPs 1, 7 (same as Example 1), 22, and 26 in Table 4 of Example 5 were designated as bivalent targeting LNPs, trivalent targeting LNPs, tetravalent targeting LNPs, and pentavalent targeting LNPs, respectively. All LNPs contained the mRNA prepared in this example (the cDNA sequence of the coding region is shown in SEQ ID NO. 62).

[0209] 3. LNPs are used to treat T1D

[0210] Female NOD.ShiLtJ mice typically develop hyperglycemia spontaneously between 10 and 14 weeks of age, with an incidence rate of 70-90%.

[0211] This experiment used an accelerated T1D NOD mouse model to accelerate the progression of autoimmune diabetes by disrupting immune tolerance in mice through PD-1 blockade. The specific protocol was as follows: 10-week-old female NOD mice were intraperitoneally injected with anti-PD-1 antibody (5 mg / kg) twice, on days 0 and 14 (as shown by the red arrows in Figure 8). This model rapidly induced type 1 diabetes (defined as two consecutive blood glucose levels >250 mg / dL). This method accelerates the development of diabetes by interfering with regulatory immune pathways (Fife et al., J Exp Med 2006; Nishimura et al., 2001), providing a reliable platform for T1D immunomodulatory intervention research.

[0212] To evaluate the effects of the LNPs prepared above, monovalent, bivalent, trivalent, tetravalent, and pentavalent targeted LNPs were dispersed in PBS to prepare 100 μL solutions (2.7 μg mRNA). These solutions were administered twice intravenously to accelerated T1D NOD female mice (intravenous injection times correspond to the blue arrows in Figure 8). The positive control group consisted of accelerated T1D NOD female mice injected with an equal amount of blank LNPs, and the negative control group consisted of normal mice injected with an equal amount of blank LNPs.

[0213] The body weight and blood glucose levels of the seven groups of mice were monitored. The blood glucose level monitoring indicators included glycated hemoglobin (HbA1c) and fasting plasma glucose (FPG). Fasting plasma glucose (FPG) and oral glucose tolerance test (OGTT) are the current diagnostic methods for diabetes; however, HbA1c is a stable indicator reflecting chronic blood glucose levels, which is more in line with the definition of diabetes and can better reflect long-term blood glucose levels and the risk of chronic complications.

[0214] The study of mouse body weight showed that the negative control group (normal mice) experienced a steady increase in body weight, while the positive control group (accelerated T1D NOD mice) showed a continuous decrease in body weight. Mice injected with monovalent, bivalent, trivalent, quadrivalent, and pentavalent targeted LNPs all showed an overall increasing body weight trend, with the trivalent targeted LNPs showing the closest weight gain trend to normal mice. This indicates that accelerated T1D NOD mice exhibited typical diabetic symptoms such as weight loss, and that treatment with monovalent or multivalent LNPs could alleviate this weight loss, with trivalent targeted LNPs showing the best effect.

[0215] The changes in blood glucose levels in each group of mice from week 0 to week 8 are shown in Table 16 below.

[0216] Table 16. Changes in blood glucose levels in mice from week 0 to week 8.

[0217] Table 16 shows that blood glucose levels remained stable within the normal range in negative control mice, while blood glucose levels in positive control mice rose sharply and remained high. Mice injected with monovalent or multivalent targeted LNPs showed a significant reduction in blood glucose levels compared to positive control mice from 0 to 8 weeks. However, comparing the therapeutic effects of monovalent or multivalent targeted LNPs, only mice injected with trivalent targeted LNPs showed blood glucose levels closest to those of negative control mice.

[0218] The percentage of HbA1c in total hemoglobin in each group of mice after 8 weeks is shown in Table 17 below.

[0219] Table 17 shows the percentage of HbA1c in total hemoglobin in each group of mice after 8 weeks.

[0220] Table 17 shows that after 8 weeks, the proportion of HbA1c in total hemoglobin of positive control mice was significantly higher than that of negative control mice. Comparing the treatment effects of mice injected with monovalent or multivalent targeted LNPs, it was shown that the proportion of HbA1c in all groups of mice decreased after LNP treatment, but only the mice injected with trivalent targeted LNPs had the closest HbA1c proportion to the negative control mice.

[0221] The pathological scores (degree of islet damage stained with HE) of the pancreatic tissue sections of mice in each group after 8 weeks are shown in Table 18 below.

[0222] Table 18. Pathological scores of HE-stained pancreatic islet tissue sections from mice in each group after 8 weeks.

[0223] Table 18 shows that the islets of Langerhans in positive control mice were significantly damaged compared to negative control mice. However, the degree of islet damage was reduced after mice were treated with monovalent or multivalent targeted LNPs. In contrast, only mice treated with trivalent targeted LNPs showed the least degree of islet damage and their scores were closer to those of normal mice.

[0224] The above experimental results indicate that trivalent targeted LNPs can significantly reduce blood glucose and maintain blood glucose levels close to normal in the long term when used to treat diabetes, while also significantly reducing the degree of pancreatic islet damage.

[0225] This embodiment successfully designed a multi-epitope mRNA and related products (LNPs) that can tandemly encode T1D-related antigens, which can significantly reduce blood glucose and glycated hemoglobin levels in a diabetes model and effectively control the development of diabetes.

[0226] Example 9: Treatment of myasthenia gravis

[0227] This embodiment verifies that the LNPs prepared in Example 1 can be used to treat myasthenia gravis (MG).

[0228] 1. Prediction, design, and mRNA preparation of antigenic epitope peptides

[0229] Myasthenia gravis (MG) is a chronic autoimmune neuromuscular disease caused by autoantibody-mediated disruption of neuromuscular junction (NMJ) signaling, leading to muscle weakness, fatigue, and in severe cases, respiratory failure. The main characteristic of MG is the presence of autoantibodies targeting key NMJ proteins, primarily the acetylcholine receptor (AChR), muscle-specific kinase (MuSK), and low-density lipoprotein receptor-associated protein 4 (LRP4). Approximately 85% of MG patients are AChR positive, and their CD4+... + T-cell-dependent autoreactive B-cell activation leads to the production of pathogenic anti-AChR autoantibodies. These antibodies trigger receptor internalization, complement-mediated postsynaptic membrane lysis, and neuromuscular signaling disorders, resulting in progressive muscle weakness, ptosis, and difficulty swallowing or breathing. In contrast, MuSK-positive MG (accounting for 5-10%) involves IgG4 autoantibodies disrupting AChR aggregation, while LRP4-positive MG is rarer and its pathological mechanism is unclear. Immune dysregulation in MG is closely related to HLA inheritance, especially HLA-DR3 and HLA-DQ8, which play key roles in antigen presentation and autoreactive T / B cell responses.

[0230] The main pathogenic targets of MG are AChR and MuSK, so this embodiment will screen the preferred epitopes of these two pathogenic targets.

[0231] (1) Selection of AChR epitopes

[0232] The results indicate that the α subunit of nicotinic AChR (amino acid sequence shown in SEQ ID NO. 63) is crucial in the pathogenesis of the autoimmune paralytic disease MG, as it contains both epitopes that dominate the anti-AChR antibody response and epitopes recognized by CD4+AChR-specific helper T (Th) cells. The distribution of different epitopes in the α subunit of AChR is shown in Figure 9.

[0233] Figure 10 shows the results of detecting the proliferative responses of different epitopes of the α subunits of AChR in four different MG patients (Pt 3 / 7 / 10 / 11). CD4+ enriched lymphocytes were used for Pt 3 / 7 / 10, while peripheral blood mononuclear cells (PBMCs) were used for Pt 11. The results show that, on the one hand, multiple patients share common epitopes: α118-137 induced significant proliferation in all four patients with extremely high proliferative intensity, making it the most dominant epitope; α304-322 also showed significant proliferation in Pt 3 and Pt 10, and is another important dominant epitope. On the other hand, there are significant individual differences in the immune responses of different MG patients to the α subunit epitopes of AChR: for example, Pt 3 showed extremely strong proliferation in α48-67, while Pt 10 showed significant proliferation in α387-405, etc.

[0234] In Figure 11, the light gray bars represent the recognition frequency of different peptide epitopes of the α subunit of AChR by fresh immune cells (PBMCs and CD4+ enriched cells) isolated from the peripheral blood of seven MG patients. The recognition frequency shows that α118-137 has the highest number of responders, making it the most widely recognized epitope in the peripheral blood of MG patients. α304-322 is also a frequently recognized epitope, and multiple responders also showed responses in regions such as α48-67 and α387-405. The black bars represent the frequency of recognition of these peptides by anti-AChR CD4+ T cell lines cultured long-term from four MG patients in previous studies. The black bars for α48-67 and α304-322 are the highest, indicating that these two epitopes are core targets widely recognized in the four long-term T cell lines. The black bars for α419-437 are found in approximately three cell lines, also indicating a high-frequency recognition epitope in these cell lines.

[0235] Furthermore, to clarify the epitope library of anti-AChR Th cells, this embodiment tested the response of unsorted blood CD4+ cells and / or total lymphocytes from 22 MG patients to a 20-amino acid overlapping synthetic peptide covering the full sequence of the human muscle AChRα subunit.

[0236] The test results show:

[0237] (1) Only the most severely ill patients were identified with α subunit epitopes, and these were mainly young women;

[0238] (2) In vitro detection of AChR-specific CD4+ response requires the removal of CD8+ cells: When CD8+ cell-depleted samples were used in two patients, a clear response to multiple α subunit polypeptide sequences was detected, while their total peripheral blood mononuclear cell population did not respond to any α subunit polypeptide.

[0239] (3) The patient’s peptide recognition pattern is unique to each individual, but the four most frequently recognized immunodominant regions of long-term AChR-specific CD4+ T cell lines or adjacent peptide sequences are: residues 48-67, 101-137, 293-337 and 408-437; followed by 89-105 and 320-337.

[0240] The key epitope sequences of the AChRα subunit are shown in Table 19 below.

[0241] Table 19. Key Epitope Sequences of AChRα Subunit

[0242] (2) Selection of MuSK epitopes

[0243] The full-length amino acid sequence of MuSK is shown in SEQ ID NO. 76. MuSK-Ig1 is located at positions 21-125 of the full-length amino acid sequence. This example provides an overview of the correlation between disease severity scores and MuSK-Ig1 epitope pattern responsiveness in 22 Italian MG patients. The correlation results are shown in Figure 12, indicating that the MuSK-Ig1 epitope is highly correlated with disease severity scores, while other epitopes show weaker correlations, suggesting that the Ig1 epitope is a major driver of MG disease progression. Therefore, Ig1 was selected as the target epitope for mRNA immunotherapy.

[0244] The key peptide sequences of the MuSK-Ig1 domain obtained by screening with NetmHciipan_el 4.1 in IEDB are shown in Table 20 below.

[0245] Table 20. Key peptide sequences of the MuSK-Ig1 domain

[0246] Next, based on hydrophilicity, hydrophobicity, and water solubility, the splicing epitopes in Table 21 below were selected.

[0247] Table 21. Segmentation Table Positions

[0248] (3) mRNA preparation

[0249] Further, a complete cDNA sequence is constructed, the cDNA comprising a 5' UTR (including the Kozak sequence), a coding region, and a 3' UTR. The nucleotide sequences of the 5' UTR and 3' UTR are the same as in Example 6. The coding region encodes a target sequence and at least one epitope sequence from Tables 19-21, with multiple epitopes linked by linkers as part of the coding region. The leader sequence and linker sequences are the same as in Example 6. The cDNA coding region is further codon-optimized using the same method as in Example 6. The optimized cDNA is then transcribed into mRNA using the same method as in Example 6.

[0250] 2. Preparation of LNPs

[0251] The LNPs prepared in Example 3, where the auxiliary lipid was DSPC and the ligand-modified PEGylated lipid was a single DSPE-PEG2K-trimannose, are designated as monovalent targeting LNPs. LNPs 1, 7 (same as Example 1), 22, and 26 in Table 4 of Example 5 are designated as bivalent targeting LNPs, trivalent targeting LNPs, tetravalent targeting LNPs, and pentavalent targeting LNPs, respectively. All LNPs contain the mRNA prepared in this example (the cDNA sequence of the coding region encodes the nine polypeptides spliced ​​to epitope 3 in Table 21 above, i.e., the coding region cDNA sequence is shown in SEQ ID NO. 84).

[0252] 3. LNPs are used to treat MG.

[0253] Using a transgenic mouse model expressing HLA-DR3 and HLA-DQ8, AChR-responsive CD4 was transferred via adoptive transfer. + MG mouse models were constructed by T cell or direct AChR immunization. MG mouse models were intravenously injected with trivalent targeting LNPs containing mRNAs with different splicing epitopes, while the control group was injected with blank LNPs. Foxp3 levels in the spleen of each group of mice were measured. + The efficacy was evaluated by measuring the percentage of Tregs, the levels of pro-inflammatory cytokines (IFN-γ, IL-17), the levels of autoantibodies in the blood, and the clinical score. The test results are shown in Table 22 below.

[0254] Table 22. Foxp3 lymph nodes of mice in each group + The percentage of Tregs, the levels of pro-inflammatory cytokines (IFN-γ, IL-17), the levels of autoantibodies, and clinical scores.

[0255] Table 22 shows that, compared with the control group, mice in the monovalent or multivalent LNP-targeting groups had Foxp3 levels +Increased Tregs, decreased pro-inflammatory cytokines IFN-γ and IL-17, decreased autoantibody levels, and improved clinical scores; among them, the trivalent LNPs-targeted mice showed the best performance in all test indicators, indicating that it had the strongest ability to induce immune tolerance, the best effect in controlling inflammatory response, effectively blocked autoantibody-mediated neuromuscular junction damage, and basically relieved symptoms such as muscle weakness and ptosis, which were close to normal levels.

[0256] Trivalent targeting of LNPs synergistically targets multiple receptors on the surface of LSECs. The simultaneous binding of multiple receptors significantly enhances the specific uptake efficiency of LSECs, thereby efficiently presenting antigen epitopes and inducing Foxp3. + The production of Tregs increases, while inhibiting the activation of Th1 or Th17 cells and reducing the production of pro-inflammatory factors such as IFN-γ and IL-17, as well as autoantibodies.

[0257] 4. Screening of spliced ​​table positions

[0258] Furthermore, trivalent targeting LNPs were prepared from the mRNAs containing different splicing epitopes in Table 21 according to the method in Example 1.

[0259] The MG mouse model was constructed using the same method as above. MG mice were treated with intravenous injections of LNPs containing different spliced ​​epitopes, while the control group was injected with blank LNPs. Foxp3 levels in the lymph nodes of each group of mice were measured. + The percentage of Tregs, the content of pro-inflammatory cytokines (IFN-γ, IL-17), the level of autoantibodies, and the clinical score were measured using the same methods as above. The results are shown in Table 23 below.

[0260] Table 23. Foxp3 lymph nodes of mice in each group + The percentage of Tregs, the levels of pro-inflammatory cytokines (IFN-γ, IL-17), the levels of autoantibodies, and clinical scores.

[0261] Table 23 shows that, compared with the control group, mice in the group encoding LNPs mRNA splicing epitopes had Foxp3 + Increased Tregs, decreased pro-inflammatory cytokines IFN-γ and IL-17, decreased autoantibody levels, and improved clinical scores; among them, the LNPs group of mice encoding splice epitope 3 showed the best performance in all test indicators, indicating that it had the strongest ability to induce immune tolerance, the best effect in controlling inflammatory response, effectively blocked autoantibody-mediated neuromuscular junction damage, and basically relieved symptoms such as muscle weakness and ptosis, which were close to normal levels.

[0262] Therefore, the preferred combination of epitopes for the mRNA coding region is the epitope combination of spliced ​​epitopes 3 in Table 23.

[0263] Furthermore, the mRNA coding region preferably encodes all the polypeptide sequences shown in SEQ ID NO. 64–69 and 77–79. The cDNA sequence corresponding to the mRNA coding region is shown in SEQ ID NO. 83, and the optimized cDNA sequence of the coding region is shown in SEQ ID NO. 84, with a GC content of 55.08%.

[0264] Example 10: Treatment of toxic diffuse goiter

[0265] This embodiment verifies that the LNPs prepared in Example 1 can be used to treat toxic diffuse goiter (GD disease).

[0266] 1. Prediction, design, and mRNA preparation of antigenic epitope peptides

[0267] The pathogenesis of Graves' disease (GD) involves a misrecognition by the patient's immune system, which mistakes the thyroid-stimulating hormone receptor (TSHR) on the thyroid follicular cell membrane for a "foreign antigen," thereby inducing B lymphocytes to produce specific autoantibodies against TSHR. Furthermore, patients have an imbalance between the Th1 and Th2 helper somatic cell (Th) subsets. Th2 cells in patients exhibit hyperfunction, secreting cytokines (such as IL-4 and IL-10) that promote B cell activation and the production of large amounts of autoantibodies. Simultaneously, regulatory T cells (Tregs) are deficient in function, failing to effectively suppress abnormal immune responses, further exacerbating the autoimmune disorder.

[0268] The amino acid sequence of the TSHR is shown in SEQ ID NO.85. The first to fourth 412 amino acids of the sequence are the extracellular domains of the TSHR, which are the main targets of autoantibodies. The frequency of HLA-DR3 in GD patients is 40-55% (15-30% in the general population), and those who carry HLA-DR3 have a 3-4 times increased risk of developing the disease.

[0269] (1) Investigation of the binding affinity between the extracellular domain of TSHR and HLA-DR3

[0270] The predicted affinity of the TSHR extracellular domain for HLA-DR3 based on the NIAID IEDB database is shown in Table 24 below.

[0271] Table 24. Predicted results of HLA-DR3 binding affinity between TSHR extracellular epitopes and HLA-DR3.

[0272] (2) Investigation of the binding affinity of the TSHR extracellular domain to H2-IAd / H2-IEd

[0273] Epitope prediction in animal experiments was based on the binding affinity of the TSHR extracellular domain to H2-IAd / H2-IEd, and the results are shown in Table 25 below.

[0274] Table 25. Prediction of binding affinity between TSHR extracellular epitopes and H2-IAd / H2-IEd

[0275] Based on the results in Tables 24-25 above, and further considering hydrophilicity, hydrophobicity, and water solubility, multiple groups of mRNA splicing epitopes suitable for animal models (human TSHR immunization) and patients were screened, as shown in Table 26 below.

[0276] Table 26. TSHR's preferred epitopes and their reverse-transcribed cDNA sequences

[0277] (3) mRNA preparation

[0278] Further, a complete cDNA sequence is constructed, the cDNA comprising a 5' UTR (including the Kozak sequence), a coding region, and a 3' UTR. The nucleotide sequences of the 5' UTR and 3' UTR are the same as in Example 6. The coding region encodes a target sequence and at least one epitope sequence from Tables 24-26, with multiple epitopes linked by linkers as part of the coding region. The leader sequence and linker sequences are the same as in Example 6. The cDNA coding region is further codon-optimized using the same method as in Example 6. The optimized cDNA is then transcribed into mRNA using the same method as in Example 6.

[0279] 2. Preparation of LNPs

[0280] The LNPs prepared in Example 3, where the auxiliary lipid was DSPC and the ligand-modified PEGylated lipid was a single DSPE-PEG2K-trimannose, are designated as monovalent targeting LNPs. LNPs 1, 7 (same as Example 1), 22, and 26 in Table 4 of Example 5 are designated as bivalent targeting LNPs, trivalent targeting LNPs, tetravalent targeting LNPs, and pentavalent targeting LNPs, respectively. All LNPs contain the mRNA prepared in this example (the cDNA sequence of the coding region encodes the eight polypeptides spliced ​​to epitope 1 in Table 26 above, i.e., the coding region cDNA sequence is shown in SEQ ID NO. 102).

[0281] 3. LNPs for the treatment of GD disease

[0282] The GD mouse model was constructed as follows: Susceptible strain mice (such as BALB / c) were sensitized multiple times by using plasmid DNA encoding the TSHR A subunit (TSHR-289) via muscle electroporation.

[0283] This induction method successfully stimulates the production of biologically active stimulating antibodies (TSAbs). These antibodies bind to TSHR on the surface of thyroid cells and continuously activate downstream signaling pathways, leading to elevated thyroid hormone levels, diffuse thyroid enlargement, and thyroid follicular epithelial hyperplasia.

[0284] GD mouse models were treated with monovalent, bivalent, trivalent, quadrivalent, and pentavalent targeted LNPs, respectively, while the control GD mouse model was injected with blank LNPs. The therapeutic effect was evaluated by measuring the levels of thyroid-stimulating hormone receptor antibody (TRAb), the Th1 / Th2 helper somatic cell ratio, and the levels of IL-4 and IL-10 in the blood of each group of mice. The results are shown in Table 27 below.

[0285] Table 27. Levels of thyroid-stimulating hormone receptor antibody (TRAb), the ratio of Th1 to Th2 helper somatic cells, and the levels of IL-4 and IL-10 in the blood of mice in each group.

[0286] As shown in Table 27, compared with the control group, mice in the monovalent or multivalent LNPs-targeted groups had decreased TRAb levels, increased Th1 / Th2 ratios, decreased IL-4 levels, and increased IL-10 levels. Among them, mice in the trivalent LNPs-targeted group had the lowest TRAb and IL-4 levels, the highest Th1 / Th2 ratios, and the highest IL-10 levels, indicating that it effectively blocked TSHR antibody-mediated thyroid dysfunction. The Th1 / Th2 ratio approached or even returned to the normal range, correcting the characteristic Th2 hyperactivity state of GD disease. The decrease in IL-4 levels significantly inhibited the pro-inflammatory function of Th2 cells, and the increase in IL-10 levels enhanced the anti-inflammatory immune response.

[0287] Trivalent LNPs synergistically target multiple receptors on the surface of LSECs. The simultaneous binding of multiple receptors significantly enhances the specific uptake efficiency of LSECs. After efficient uptake of LNPs, LSECs precisely present the TSHR-encoded preferred epitopes on mRNA, thereby inducing Foxp3. + Tregs proliferate and secrete anti-inflammatory factors such as IL-10; on the other hand, they inhibit Th2 cell activation, reduce IL-4 secretion and TRAb production, and ultimately achieve immune balance restoration.

[0288] 4. Screening of spliced ​​table positions

[0289] Furthermore, trivalent targeting LNPs were prepared from mRNAs containing different splicing epitopes in Table 26 according to the method in Example 1.

[0290] Construct a GD mouse model using the same method as above.

[0291] GD mouse models were treated with trivalent targeted LNPs containing mRNAs with different splicing epitopes, while control GD mouse models were treated with blank LNPs. The efficacy was evaluated by detecting the levels of thyroid-stimulating hormone receptor antibody (TRAb), the ratio of Th1 to Th2 helper somatic cells, and the levels of IL-4 and IL-10 in the blood of each group of mice. The results are shown in Table 28 below.

[0292] Table 28. Levels of thyroid-stimulating hormone receptor antibody (TRAb), the ratio of Th1 to Th2 helper somatic cells, and the levels of IL-4 and IL-10 in the blood of mice in each group.

[0293] As shown in Table 28, compared with the control group, mice in the LNPs group (encoding spliced ​​epitopes) had decreased TRAb levels, increased Th1 / Th2 ratios, decreased IL-4 levels, and increased IL-10 levels. Among them, mice in the LNPs group (encoding spliced ​​epitope 1) had the lowest TRAb and IL-4 levels, the highest Th1 / Th2 ratios, and the highest IL-10 levels, indicating that it effectively blocked TSHR antibody-mediated thyroid dysfunction. The Th1 / Th2 ratio approached or even returned to the normal range, correcting the characteristic Th2 hyperactivity state of GD disease. The decrease in IL-4 levels significantly inhibited the pro-inflammatory function of Th2 cells, and the increase in IL-10 levels enhanced the anti-inflammatory immune response.

[0294] Therefore, the preferred combination of epitopes for the mRNA coding region is the one that splices epitopes 1 in Table 26.

[0295] Further, the mRNA coding region preferably encodes all the polypeptide sequences shown in SEQ ID NO. 86–93, and the reverse-transcribed cDNA sequences corresponding to the polypeptide sequences shown in SEQ ID NO. 86–93 are shown in SEQ ID NO. 94–101. When the mRNA coding region encodes all the polypeptide sequences shown in SEQ ID NO. 86–93, the corresponding optimized coding region cDNA sequence is shown in SEQ ID NO. 102, with a GC content of 58.63%.

[0296] Example 11: Treatment of systemic lupus erythematosus

[0297] 1. Prediction, design, and mRNA preparation of antigenic epitope peptides

[0298] Systemic lupus erythematosus (SLE) is an autoimmune disease caused by the combined effects of genetic susceptibility (such as HLA class II gene polymorphism), environmental triggers (ultraviolet radiation, viral infections, etc.), and epigenetic modifications, which break the body's immune tolerance. Its core mechanism involves abnormal activation of B cells, producing various autoantibodies such as antinuclear antibodies. The resulting immune complexes deposit in multiple organs and tissues, including the skin, kidneys, and joints. This is accompanied by an imbalance in T cell subsets (such as Treg dysfunction and excessive Th17 cell proliferation), triggering a persistent inflammatory response and tissue damage, ultimately leading to multi-system involvement. Histones (divided into five classes: H1, H2A, H2B, H3, and H4), SNRNP70 (P140), SMD1, SMD3, U1A, and U1C are all core autoantigens that trigger SLE.

[0299] In this embodiment, the preferred epitopes of the above-mentioned different antigens were screened using the NetmHciipan_el 4.1 server method, and the preferred epitopes were converted into reverse transcribed cDNA sequences, as shown in Table 29 below.

[0300] Table 29. Preferred epitopes of different antigens

[0301] Further, a complete cDNA sequence was constructed, the cDNA comprising a 5'UTR (including the Kozak sequence), a coding region, and a 3'UTR. The nucleotide sequences of the 5'UTR and 3'UTR are the same as in Example 6. The coding region encodes a target sequence and at least one epitope sequence from Table 29, with multiple epitopes linked by a linker as part of the coding region. The leader sequence and linker sequence are the same as in Example 6. In this example, two cDNA sequences were designed based on the epitope sequences in Table 29, and the codons of the coding region cDNA were further optimized using the same optimization method as in Example 6. After optimization, one coding region cDNA sequence is shown in SEQ ID NO. 123, containing the preferred epitope sequence of P140, with a GC content of 54.59%; the other coding region cDNA sequence is shown in SEQ ID NO. 124, containing all preferred epitopes in Table 29, with a GC content of 57.48%.

[0302] Furthermore, the optimized cDNA is transcribed into mRNA, using the same method as in Example 6.

[0303] 2. Preparation of LNPs

[0304] The LNPs prepared in Example 3, where the auxiliary lipid was DSPC and the ligand-modified PEGylated lipid was a single DSPE-PEG2K-trimannose, are designated as monovalent targeting LNPs. LNPs 1, 7 (same as Example 1), 22, and 26 in Table 4 of Example 5 are designated as bivalent targeting LNPs, trivalent targeting LNPs, tetravalent targeting LNPs, and pentavalent targeting LNPs, respectively. All LNPs contain the mRNA prepared in this example (the cDNA sequence of the coding region is shown in SEQ ID NO. 124).

[0305] 3. LNPs for SLE treatment

[0306] Fas lpr The mouse model is a classic SLE mouse model, exhibiting MRL-Ipr / lymphoid tissue hyperplasia, the Fas lpr The mouse model is a spontaneous animal model derived from a spontaneous mutation in the Fas gene, which leads to the blockage of the Fas-mediated apoptosis pathway, resulting in a large accumulation of autoreactive lymphocytes that should be cleared through apoptosis. Pathologically, LPR mice exhibit significant lymphadenopathy, splenomegaly, and produce high titers of anti-dsDNA and other autoantibodies, ultimately leading to severe immune complex glomerulonephritis. Because this model can typically reflect autoimmune damage caused by a lack of peripheral immune tolerance, it is widely used to evaluate the effectiveness of novel immunotherapies (such as mRNA vaccines inducing regulatory T cell expansion) in restoring immune homeostasis and reversing inflammatory responses.

[0307] Fas lpr Validation of successful mouse model construction: First, the mRNA expression differences of inflammation-related genes between lupus patients and healthy volunteers were detected by single-cell RNA sequencing; the results are shown in the left figure of Figure 13. scRNAsseq data showed a widespread increase in inflammatory gene expression in lupus patients compared to healthy volunteers (yellow, orange, and red indicate increased gene expression, purple or blue indicate decreased expression); further comparisons were made with Fas... lpr The differences in mRNA expression of inflammation-related genes between mouse models and wild-type mice are shown in the right figure of Figure 13, revealing the differences in Fas... lpr The mouse model exhibits a similar pattern of inflammatory gene expression to that in human patients.

[0308] For 16-week-old Fas lprMice were intravenously injected with three doses of monovalent, bivalent, trivalent, tetravalent, and pentavalent targeted LNPs, as well as blank LNPs (control), every three weeks. Seven weeks after the third injection, mice were monitored for another 10 weeks (weeks 16–26). The monitoring parameters included: mouse survival rate from week 16 to 26; peripheral blood analysis for pro-inflammatory cytokines at week 16; and single-cell RNA sequencing of blood, bone marrow, kidneys, and spleen from sacrificed mice at week 26.

[0309] The survival rates of mice at 16–26 weeks are shown in Table 30 below.

[0310] Table 30. Survival rate of mice 16–26 weeks after treatment

[0311] As shown in Table 30, the survival rate of mice in the control group was only 25% at 26 weeks. However, after treatment with monovalent or multivalent LNPs, the survival rate of mice at 26 weeks increased. Among them, the mice injected with trivalent targeted LNPs had the highest survival rate.

[0312] Analysis of pro-inflammatory cytokines at week 16 showed that, compared with the control, the number of neutrophils, monocytes, and dendritic cells in the blood of mice injected with monovalent, bivalent, trivalent, tetravalent, and pentavalent LNPs was reduced, but the reduction in the number of these inflammatory cells in the blood of mice injected with trivalent LNPs was the most significant.

[0313] After 26 weeks, mice were sacrificed, and blood, bone marrow, kidneys, and spleens were collected for single-cell RNA sequencing. The results showed that, compared with the control group, the expression of inflammatory genes in the blood, bone marrow, kidneys, and spleens of mice injected with monovalent, bivalent, trivalent, tetravalent, and pentavalent targeted LNPs was reduced, with the trivalent targeted LNPs group showing the most significant reduction in inflammatory gene expression.

[0314] Example 12: Treatment of primary cholangitis

[0315] 1. Prediction, design, and mRNA preparation of antigenic epitope peptides

[0316] The core autoantigen of primary cholangitis (PBC) is the E2 subunit (PDC-E2) of the pyruvate dehydrogenase complex (PDC). During normal apoptosis, PDC-E2 undergoes glutathioneization modification, and the modified PDC-E2 is encapsulated in apoptotic bodies. These apoptotic bodies containing modified PDC-E2 are phagocytosed by antigen-presenting cells (APCs), which then present PDC-E2 to immune cells via MHC molecules (MHC-I / MHC-II), triggering specific immune responses, including humoral and cellular immunity. After recognizing PDC-E2, B lymphocytes differentiate into plasma cells and secrete anti-PDC-E2 antibodies (i.e., AMA-M2 type, the hallmark antibody of PBC). These antibodies bind to PDC-E2 to form immune complexes, which can deposit on the bile duct epithelium, activating the complement system (such as C3 and C5), further amplifying the inflammatory response. After CD4+Th cells are activated by antigens presented by MHC-II, they differentiate into subsets such as Th1 and Th17, and secrete cytokines (such as IFN-γ, TNF-α, and IL-17). These cytokines can not only directly damage bile duct cells, but also recruit more immune cells (such as macrophages) to aggregate. After CD8+CTLs are activated by antigens presented by MHC-I, they directly recognize and kill bile duct epithelial cells expressing PDC-E2 (bile duct cells themselves also express PDC-E2), causing specific apoptosis of bile duct cells.

[0317] HLA-DRB4*01:01 is a specific allele of the HLA-DRB4 gene, which encodes a portion of the HLA-DR molecule, an MHC class II protein involved in immune responses. HLA-DRB4*01:01 plays a crucial role in polycythemia vera (PBC) because certain HLA-DR alleles are associated with increased susceptibility to autoimmune diseases. Studies have shown that specific HLA-DRB4 alleles, particularly HLA-DRB4*01:01, are associated with an increased risk of PBC. Expression of this allele can predispose individuals to autoimmune responses targeting the PDC-E2 antigen, a key autoantigen in the pathogenesis of PBC. The HLA-DRB4 gene plays a crucial role in presenting peptide fragments of the PDC-E2 protein to CD4+ T cells. In polycystic bronchitis (PBC), PDC-E2 is one of the major autoantigens, and the HLA-DRB4*01:01 allele may present a specific epitope of PDC-E2, thereby promoting the immune activation and chronic inflammation observed in PBC. Multiple genetic studies support the association between the HLA-DRB4 allele and PBC. The detection rate of HLA-DRB4*01:01 is higher in PBC patients compared to healthy individuals, especially in those carrying antimitochondrial antibodies (AMA), a key diagnostic biomarker for PBC.

[0318] This embodiment uses the IEDB database to predict the dihydrolipoamide acetyltransferase (DLAT) epitope of PDC, whose amino acid sequence is shown in SEQ ID NO. 125 (1-86 aa are mitochondrial transport peptides). The target binding object of the epitope peptide was selected as human HLA-DR, and the allele was selected as HLA-DRB4*01:01. The epitope peptide length is 15 amino acids. The prediction parameters of the NetmHciipan_el 4.1 server were used to screen the epitope peptides. Epitopes with high affinity and located in the conserved region of the antigen were selected. The prediction results are shown in Table 31 below.

[0319] Table 31. Predicted affinity of DLAT epitopes to HLA-DRB4*01:01

[0320] Based on the affinity binding results and data in Table 31 (J Exp Med (1995) 181(5): 1835-1845; J Clin Invest. 1998; 102(10): 1831-1840; Journal of Autoimmunity 149(2024) 103327), the preferred epitopes in Table 32 below were selected.

[0321] Table 32, DLAT Preferred Table Positions

[0322] Further, a complete cDNA sequence was constructed, the elements of which include a 5' UTR (including the Kozak sequence), a coding region, and a 3' UTR. The nucleotide sequences of the 5' UTR and 3' UTR are the same as in Example 6. The coding region encodes a target sequence and at least one epitope sequence from Table 32, with multiple epitopes linked by linkers as part of the coding region. The leader sequence and linker sequences are the same as in Example 6. In this example, the coding region of the cDNA sequence contains all the reverse-translated cDNA sequences from Table 32, and the sequence of the coding region is shown in SEQ ID NO. 138. The cDNA of the coding region was further codon-optimized using the same optimization method as in Example 6. The optimized cDNA sequence of the coding region is shown in SEQ ID NO. 139, with a GC content of 59.87%.

[0323] Furthermore, the optimized cDNA is transcribed into mRNA, using the same method as in Example 6.

[0324] 2. Preparation of LNPs

[0325] The LNPs prepared in Example 3, where the auxiliary lipid was DSPC and the ligand-modified PEGylated lipid was a single DSPE-PEG2K-trimannose, were designated as monovalent targeting LNPs. LNPs 1, 7 (same as Example 1), 22, and 26 in Table 4 of Example 5 were designated as bivalent targeting LNPs, trivalent targeting LNPs, tetravalent targeting LNPs, and pentavalent targeting LNPs, respectively. All LNPs contained the mRNA prepared in this example (the cDNA sequence of the coding region is shown in SEQ ID NO. 139).

[0326] 3. LNPs for the treatment of PBC

[0327] The 2-OA-PBC model is a classic experimental animal model established by mimicking the immunopathological features of autoimmune cholangitis (PBC). This model typically uses 2-octenoic acid-modified ovalbumin (2-OA-OVA) as the core antigen, combined with specific immune adjuvants (such as Alhydrogel or high-affinity nanoadjuvants) to systematically sensitize mice. Through repeated subcutaneous injections of the sensitizing antigen, an autoimmune response against the bile duct epithelial components is induced, leading to significant periductal lymphocytic infiltration and small bile duct damage within the liver. This model effectively simulates the pathophysiological processes of hyperactive Th1 / Th17 immune responses and impaired regulatory T cell (Treg) function in PBC patients.

[0328] PBC mouse models were intravenously injected with three doses of monovalent, bivalent, trivalent, tetravalent, and pentavalent targeted LNPs, as well as a blank LNP (control), every three weeks. Seven weeks after the third injection, mice were continuously monitored for another 10 weeks (weeks 16–26). Survival rates were monitored from weeks 16–26, and at week 26, mice in each group were sacrificed, and their livers were harvested to detect the proportion of Tregs and the infiltration of inflammatory factors TNF-α, IFN-γ, and IL-17 in the liver.

[0329] The survival rates of mice at 16–26 weeks are shown in Table 33 below.

[0330] Table 33. Survival rate of mice 16–26 weeks after treatment

[0331] As shown in Table 33, the survival rate of mice in the control group was only 30% at 26 weeks. However, after treatment with monovalent or multivalent LNPs, the survival rate of mice at 26 weeks increased. Among them, the mice injected with trivalent targeted LNPs had the highest survival rate.

[0332] Foxp3 in the liver of mice after sacrifice +The percentage of Tregs and the infiltration of inflammatory factors TNF-α, IFN-γ and IL-17 are shown in Table 34 below.

[0333] Table 34. Percentage of Tregs in the liver and infiltration of inflammatory factors TNF-α, IFN-γ and IL-17.

[0334] As shown in Table 34, compared with the control group, the Foxp3 levels in mice in the monovalent or multivalent LNP-targeting groups were significantly higher. + The proportion of Tregs increased, while the levels of TNF-α, IFN-γ, and IL-17 decreased. Among these, the Foxp3 levels in mice targeting trivalent LNPs were significantly higher. + The highest proportion of Tregs indicates that it has the strongest ability to induce immune tolerance; the lowest levels of TNF-α, IFN-γ and IL-17 indicate that the liver inflammatory infiltration is effectively controlled.

[0335] Trivalent LNPs synergistically target multiple receptors on the surface of LSECs. The simultaneous binding of multiple receptors significantly enhances the specific uptake efficiency of LSECs. After efficient uptake of LNPs, LSECs precisely present antigen epitopes, thereby inducing Foxp3. + Tregs proliferate in large numbers, secrete anti-inflammatory factors to suppress abnormal immune responses; on the other hand, they inhibit the activation of Th1 / Th17 cells, reduce the release of pro-inflammatory factors such as TNF-α, IFN-γ, and IL-17, reduce bile duct epithelial cell damage and immune complex deposition, thereby prolonging the survival time of mice and improving the pathological state of the liver.

[0336] Example 13: Treatment of Multiple Sclerosis

[0337] 1. Prediction, design, and mRNA preparation of antigenic epitope peptides

[0338] Multiple sclerosis (MS) is a chronic autoimmune disease in which the body's own T cells attack myelin in the central nervous system (CNS), leading to demyelination of nerve fibers and neurodegeneration.

[0339] In this embodiment, three core proteins of myelin were selected, including human MBP protein (amino acid sequence as shown in SEQ ID NO. 140), human MOG protein (amino acid sequence as shown in SEQ ID NO. 141), and human PLP protein (amino acid sequence as shown in SEQ ID NO. 142). The binding of epitope peptides in the above three core proteins to MHC-II was predicted using the NetmHciipan_el 4.1 server. Epitopes with high affinity and located in the conserved region of the antigen were selected. The results of the preferred epitopes are shown in Table 35 below.

[0340] Table 35. Preferred Epitope Screening Results for MBP, MOG, and PLP

[0341] Furthermore, the reverse-translated cDNA sequences of the eight preferred epitopes in Table 35 above are shown in Table 36 below.

[0342] Table 36. Reverse-translated cDNA sequences of 8 preferred epitopes

[0343] Further, a complete cDNA sequence was constructed, the elements of which include a 5' UTR (including the Kozak sequence), a coding region, and a 3' UTR. The nucleotide sequences of the 5' UTR and 3' UTR are the same as in Example 6. The coding region encodes a target sequence and at least one epitope sequence from Table 36, with multiple epitopes linked by linkers as part of the coding region. The leader sequence and linker sequences are the same as in Example 6. In this example, the coding region of the cDNA sequence contains all the reverse-translated cDNA sequences from Table 36, and the cDNA in the coding region is further codon-optimized using the same method as in Example 6. The optimized coding region cDNA sequence is shown in SEQ ID NO. 159, with a GC content of 60.37%.

[0344] Furthermore, the optimized cDNA is transcribed into mRNA, using the same method as in Example 6.

[0345] 2. Preparation of LNPs

[0346] The LNPs prepared in Example 3, where the auxiliary lipid was DSPC and the ligand-modified PEGylated lipid was a single DSPE-PEG2K-trimannose, were designated as monovalent targeting LNPs. LNPs 1, 7 (same as Example 1), 22, and 26 in Table 4 of Example 5 were designated as bivalent targeting LNPs, trivalent targeting LNPs, tetravalent targeting LNPs, and pentavalent targeting LNPs, respectively. All LNPs contained the mRNA prepared in this example (the cDNA sequence of the coding region is shown in SEQ ID NO. 159).

[0347] 3. LNPs for the treatment of MS

[0348] Experimental autoimmune encephalomyelitis (EAE) induced by oligodendrocyte glycoprotein (MOG) is one of the most widely used animal models. MOG is a key antigen located in the outermost layer of myelin sheath in the central nervous system, possessing strong immunogenicity and capable of inducing a specific immune attack against the myelin sheath. In establishing a mouse model of MS, mice were subcutaneously immunized with a mixture of MOG (35-55) peptide fragments and complete Freund's adjuvant (CFA), supplemented with pertussis toxin (PTX) injection to increase blood-brain barrier permeability. This model highly mimics the core pathological features of MS, including extensive inflammatory cell infiltration in the central nervous system, widespread demyelinating lesions, and the resulting progressive limb paralysis. Because the immune response in this model is mainly driven by Th1 and Th17 cells, accompanied by impaired Treg function, it has become a core system for evaluating the ability of the mRNA-LNP platform to induce immune tolerance. By targeting the liver to deliver mRNA encoding MOG epitopes, systemic antigen-specific Tregs can be effectively induced to expand, thereby inhibiting effector T cell attacks on the nervous system and restoring immune balance.

[0349] MS mouse models were treated with intravenous injections of monovalent, bivalent, trivalent, tetravalent, and pentavalent targeted LNPs, respectively, once every three weeks for a total of three times. The control group received blank LNPs. Two weeks after the last injection, peripheral blood was collected to detect autoantibody levels, and mice in each group were then sacrificed, and spleens were collected to detect Foxp3. + The results of the Treg percentage, the detection of IFN-γ, IL-17A and TNF-α levels in central nervous system tissue, and the scoring of nerve fiber damage are shown in Table 37 below.

[0350] Table 37. Peripheral blood autoantibody levels and spleen Foxp3 levels in mice of each group + The percentage of Tregs, the levels of IFN-γ, IL-17A, and TNF-α in the central nervous system, and the nerve fiber damage score

[0351] As shown in Table 37, compared with the control group, the levels of autoantibodies in mice targeting LNPs (monovalent or multivalent) decreased, and Foxp3 levels were lower. + The proportion of Tregs increased, while the levels of IFN-γ, IL-17A, and TNF-α decreased, and the neurofibrillary damage score also decreased. Among the mice, the trivalent LNP-targeted group had the lowest levels of autoantibodies, including Foxp3. +The highest proportion of Tregs, the lowest levels of IFN-γ, IL-17A and TNF-α, and the lowest score of nerve fiber damage indicate that it has the strongest ability to induce antigen-specific immune tolerance, effectively controls inflammatory infiltration in the central nervous system, and significantly alleviates demyelination and axonal damage, bringing them close to normal levels.

[0352] Trivalent LNPs synergistically target multiple receptors on the surface of LSECs. The simultaneous binding of multiple receptors significantly enhances the specific uptake efficiency of LSECs. After LSECs efficiently uptake LNPs, they induce Foxp3. + Tregs proliferate in large numbers and secrete anti-inflammatory factors to inhibit the activation of autoreactive T cells. On the other hand, they inhibit the secretion of pro-inflammatory factors such as IFN-γ and IL-17A by Th1 / Th17 cells, reduce the production of autoantibodies, thereby blocking demyelination and nerve fiber damage and improving the pathological process of MS.

[0353] Example 14: Treatment of Sjögren's Syndrome

[0354] 1. Prediction, design, and mRNA preparation of antigenic epitope peptides

[0355] Sjögren's syndrome SSA syndrome (SS) is an autoimmune disease caused by the combined effects of genetic susceptibility (such as HLA alleles, including HLA-DRB1*03:01, HLA-DQA1*05:01, and HLA-A*03) and environmental triggers (such as viral infection and epigenetic modification) that break immune tolerance. Its core mechanism is the activation of specific T / B lymphocytes, which produce autoantibodies such as anti-SSA / Ro and anti-SSB / La. Immune cells infiltrate and damage exocrine glands such as salivary glands and lacrimal glands, leading to decreased glandular secretion function, accompanied by systemic inflammatory response and multi-organ involvement.

[0356] This embodiment provides four target proteins for treating SS: RO60 (amino acid sequence as shown in SEQ ID NO. 160), TRIM21 (amino acid sequence as shown in SEQ ID NO. 161), lupus La (amino acid sequence as shown in SEQ ID NO. 162), and SPTN1 (amino acid sequence as shown in SEQ ID NO. 163). The peptides (epitopes) of these four target proteins that can bind to MHC-II protein were predicted using the NetmHciipan_el 4.1 server.

[0357] The binding of epitopes in RO60, TRIM21, lupus La and SPTN1 to HLA alleles are shown in Figures 14-17. The preferred epitopes of four target proteins and their flanking sequences are selected from Figures 8-11 and are shown in Table 38 below.

[0358] Table 38. Preferred epitopes of four target proteins and their flanking sequences

[0359] The inverse cDNA sequences corresponding to the peptides with flanking sequences in Table 38 are shown in Table 39 below.

[0360] Table 39. Reverse cDNA sequences corresponding to peptides with flanking sequences

[0361] Further, a complete cDNA sequence was constructed, the elements of which include a 5' UTR (including the Kozak sequence), a coding region, and a 3' UTR. The nucleotide sequences of the 5' UTR and 3' UTR are the same as in Example 6. The coding region encodes a target sequence and at least one epitope sequence from Table 39, with multiple epitopes linked by linkers as part of the coding region. The leader sequence and linker sequences are the same as in Example 6. In this example, the coding region of the cDNA sequence contains all the reverse-translated cDNA sequences from Table 39, and the cDNA in the coding region is further codon-optimized using the same method as in Example 6. The optimized coding region cDNA sequence is shown in SEQ ID NO. 184, with a GC content of 57.03%.

[0362] Furthermore, the optimized cDNA is transcribed into mRNA, using the same method as in Example 6.

[0363] 2. Preparation of LNPs

[0364] The LNPs prepared in Example 3, where the auxiliary lipid was DSPC and the ligand-modified PEGylated lipid was a single DSPE-PEG2K-trimannose, are designated as monovalent targeting LNPs. LNPs 1, 7 (same as Example 1), 22, and 26 in Table 4 of Example 5 are designated as bivalent targeting LNPs, trivalent targeting LNPs, tetravalent targeting LNPs, and pentavalent targeting LNPs, respectively. All LNPs contain the mRNA prepared in this example (the cDNA sequence of the coding region is shown in SEQ ID NO. 184).

[0365] 3. LNPs for the treatment of SS

[0366] The salivary gland (SS) mouse model is established by immunizing susceptible mice with specific exocrine gland antigens or glandular extracts, supplemented by a potent adjuvant, thereby breaking peripheral immune tolerance. A commonly used antigen is salivary gland extract (SG-extract). During modeling, the antigen is typically mixed with complete Freund's adjuvant (CFA) for multiple subcutaneous sensitizations, activating effector T cells (such as Th1 and Th17 cells) and inhibiting Treg activity, triggering chronic destructive inflammation of the lacrimal and salivary glands. This model effectively mimics the typical periglandular lymphocytic infiltration and subsequent decline in glandular secretory function seen in human SS.

[0367] SS mouse models were treated with intravenous injections of monovalent, bivalent, trivalent, tetravalent, and pentavalent targeted LNPs, respectively, once every three weeks for a total of three times. The control group received blank LNPs. Two weeks after the last injection, peripheral blood was collected to detect the levels of anti-SSA / Ro and anti-SSB / La antibodies. Mice in each group were then sacrificed, and salivary glands were collected to detect Foxp3. + The percentage of Tregs, the content of IL-17 and IFN-γ, and the detection results are shown in Table 40 below.

[0368] Table 40. Levels of autoantibodies and Foxp3 in peripheral blood of mice in each group + Tregs percentage, IL-17 and IFN-γ content

[0369] As shown in Table 40, compared with the control group, the levels of autoantibodies against SSA / Ro and SSB / La in mice targeting LNPs (monovalent or multivalent) were decreased, and Foxp3 levels were also reduced. + The proportion of Tregs increased, while the levels of IFN-γ and IL-17 decreased. The trivalent LNP-targeting group of mice showed the lowest levels of two autoantibodies, Foxp3 and... + The highest proportion of Tregs and the lowest levels of IFN-γ and IL-17 indicate that it has the strongest ability to induce antigen-specific immune tolerance, effectively controls salivary gland and systemic inflammatory infiltration, and effectively reduces autoantibody-mediated exocrine gland damage.

[0370] Trivalent LNPs synergistically target multiple receptors on the surface of LSECs. The simultaneous binding of multiple receptors significantly enhances the specific uptake efficiency of LSECs. After LSECs efficiently uptake LNPs, they induce Foxp3. +Tregs proliferate in large numbers and secrete anti-inflammatory factors such as IL-10 to inhibit the activation of autoreactive T / B cells. On the other hand, they inhibit the secretion of pro-inflammatory factors such as IFN-γ and IL-17 by Th1 / Th17 cells, reduce the production of anti-SSA / Ro and anti-SSB / La antibodies, thereby blocking exocrine gland damage and systemic inflammatory response and improving the pathological process of SS.

[0371] While the present invention has been disclosed above, it is not limited thereto. Any person skilled in the art can make various modifications and alterations without departing from the spirit and scope of the invention; therefore, the scope of protection of the present invention should be determined by the scope defined in the claims.

Claims

1. A lipid nanoparticle, characterized in that, It includes cationic ionized lipids, auxiliary lipids, ligand-modified PEGylated lipids, and cholesterol; the ligands in the ligand-modified PEGylated lipids can target receptors on the surface of hepatic sinusoidal endothelial cells, and the auxiliary lipids are selected from lipids that can target receptors on the surface of hepatic sinusoidal endothelial cells or lipids without targeting.

2. The lipid nanoparticles as described in claim 1, characterized in that, The cationic ionized lipid is SM102; the auxiliary lipid includes any one of PS, DOPS and DPPS; the ligand-modified PEGylated lipid includes at least any one of DSPE-PEG2K-mannose, DSPE-PEG2K-GlcNAc, DSPE-PEG2K-trimannose, DSPE-PEG2K-GalNAc-4-sulfate, DSPE-PEG2K-L-fucose and DSPE-PEG2K-hyaluronic acid.

3. The lipid nanoparticles as described in claim 2, characterized in that, The auxiliary lipid is PS; the ligand-modified PEGylated lipid is any two of DSPE-PEG2K-mannose, DSPE-PEG2K-GlcNAc, DSPE-PEG2K-trimannose, DSPE-PEG2K-GalNAc-4-sulfate, DSPE-PEG2K-L-fucose, and DSPE-PEG2K-hyaluronic acid.

4. The lipid nanoparticles as described in claim 3, characterized in that, The ligand-modified PEGylated lipids are DSPE-PEG2K-trimannose and DSPE-PEG2K-GlcNAc.

5. The lipid nanoparticles as described in claim 1, characterized in that, The molar ratio of the cationic ionized lipid, auxiliary lipid, ligand-modified PEGylated lipid, and cholesterol is 20–70:1–15:1–5:25–45, and the sum of the molar ratios of each component is 100%; the N / P ratio of the cationic ionized lipid is 2–6.

6. The lipid nanoparticles according to claim 1, characterized in that, The lipid nanoparticles have a particle size of 80–200 nm; the lipid nanoparticles encapsulate RNA drugs for treating allergic or autoimmune diseases.

7. The lipid nanoparticles as described in claim 6, characterized in that, The allergic diseases include birch hay fever and artemisia allergy, and the autoimmune diseases include type 1 diabetes, myasthenia gravis, toxic diffuse goiter, systemic lupus erythematosus, primary cholangitis, multiple sclerosis, and Sjögren's syndrome.

8. The lipid nanoparticles as described in claim 7, characterized in that, The RNA encapsulated in lipid nanoparticles for treating birch hay fever has a coding region that encodes at least one amino acid sequence as shown in SEQ ID No. 2-7; the RNA encapsulated in lipid nanoparticles for treating artemisia allergy has a coding region that encodes at least one amino acid sequence as shown in SEQ ID No. 185-186; the RNA encapsulated in lipid nanoparticles for treating type 1 diabetes has a coding region that encodes at least one amino acid sequence as shown in SEQ ID No. 22-41; the RNA encapsulated in lipid nanoparticles for treating myasthenia gravis has a coding region that encodes at least one amino acid sequence as shown in SEQ ID No. 64-69, 77-79; the RNA encapsulated in lipid nanoparticles for treating toxic diffuse goiter has a coding region that encodes at least one amino acid sequence as shown in SEQ ID No. 86-93; the RNA encapsulated in lipid nanoparticles for treating systemic lupus erythematosus has a coding region that encodes at least one amino acid sequence as shown in SEQ ID No. 103-112; the RNA encapsulated in lipid nanoparticles for treating primary cholangitis has a coding region that encodes at least one amino acid sequence as shown in SEQ ID No. 185-186. The coding region of the RNA encapsulated in the lipid nanoparticles for treating the multiple sclerosis encodes at least one of the amino acid sequences shown in SEQ ID No. 143-150; the coding region of the RNA encapsulated in the lipid nanoparticles for treating the Sjögren's syndrome encodes at least one of the amino acid sequences shown in SEQ ID No. 164-173.

9. Use of the lipid nanoparticles as described in claims 1 to 8 in the preparation of formulations that enhance the ability to target hepatic sinusoidal endothelial cells.

10. Use of the lipid nanoparticles as described in claims 1 to 8 in the preparation of formulations that enhance the therapeutic effect on allergic or autoimmune diseases.

Images