Universal method for preparing circular RNA (Ribonucleic Acid) and application of circular RNA

Abstract

The invention discloses a universal method for preparing circular RNA (Ribonucleic Acid) and application of the circular RNA. Specifically, an optimized ribozyme recognition site is determined based on a specific local secondary structure, and a ribozyme sequence is designed correspondingly, so that the circular RNA precursor molecule is constructed, and the invention provides a method for preparing the circular RNA based on the precursor molecule and application of the circular RNA prepared by the method. The cyclic RNA is high in cyclization efficiency, any sequence except the target cyclic RNA sequence is not introduced into the cyclization product, the cyclization efficiency is not influenced by different sequences of the cyclized RNA, and the method is a universal method for preparing the cyclic RNA. The circular RNA prepared by the method can be used in various biological medicine fields such as diagnosis, prevention and treatment of human or animal diseases.

Description

Technical Field

[0001] The present invention relates to the fields of molecular biology and genetic engineering, and in particular to a universal method for preparing circular RNA and its application. Background Art

[0002] Circular RNA (also known as circular RNA) is a type of molecule without 5' and 3' ends that is "covalently connected head to tail". Due to its special "circular" structure, circular RNA molecules can resist degradation by RNases and show higher stability than linear RNA molecules in vivo and in vitro. In addition, recent studies have found that synthetic circular mRNA can efficiently express proteins in mammals. These excellent properties make circular RNA have broad application prospects in fields such as biomedicine and engineering.

[0003] Studies have shown that circular RNA molecules have many functions and potential uses, such as functional non-coding RNA, circular mRNA that translates proteins (such as vaccines, therapeutic proteins, etc.) in cells as mRNA translation templates, circular aptamers for diagnosis, treatment and imaging, circular guide RNA (circ-gRNA) for DNA and RNA editing, circular RNA that recruits ADAR editing systems, circular RNA for in vivo and in vitro cell therapy (such as CAR-T, CAR-NK, CAR-M cell therapy, stem cell therapy, etc.), etc. Circular RNA has a wide range of potential uses in fields such as biomedicine and engineering.

[0004] The preparation of circular RNA is the basic condition for expanding the potential uses of circular RNA in various fields. The cyclization method for preparing circular RNA needs to take into account the following urgent concerns: First, the cyclization efficiency should be as high as possible. Secondly, the cyclization process must ensure the accuracy of the sequence. It is necessary to avoid introducing additional sequences during the cyclization preparation process as much as possible to ensure that the sequence of the RNA molecule remains unchanged after cyclization, thereby avoiding potential risks caused by sequence changes. Finally, the cyclization platform must be versatile to prepare a variety of circular RNAs for various applications.

[0005] At present, the most commonly used circular RNA preparation strategy is the permuted intron-exon (PIE) ribozyme autocatalytic strategy, namely the PIE strategy. The PIE strategy has obvious advantages in the cyclization of long-chain RNA, which is crucial for the biomedical use of circular RNA. The PIE method has a simple reaction system, and its cyclization efficiency mainly depends on whether the design of the self-splicing ribozyme circular RNA precursor molecule is efficient.

[0006] Methods for preparing short-chain RNA cyclization usually include chemical ligation, enzyme-catalyzed ligation and other methods. Compared with PIE, the cyclization reaction system of these methods is more complicated, for example, it is necessary to introduce additional catalytic enzymes or splint sequences, etc. In addition, the sequence of the cyclized product may change, and it is easy to form intermolecular polymerization connections. The currently known PIE strategy can avoid the above defects, but it is rarely used to prepare short-chain circular RNA. The main reasons include: the need to introduce an additional cyclization auxiliary sequence, or the need to rely on the degree of sequence matching to screen suitable splicing sites, or the need to rely on the matching of larger splicing site structure modules to screen suitable splicing sites. Therefore, it is difficult for the existing PIE method to achieve the design of a better self-splicing ribozyme circular RNA precursor molecule on a short-chain RNA sequence.

[0007] In summary, the currently known PIE methods still have the defect that they cannot universally circularize various types of RNA sequences and RNA sequences of various total lengths.

[0008] Therefore, there is an urgent need in this field to develop a universal cyclization method with high cyclization efficiency, accurate sequence and the ability to cyclize various types of RNA, so as to promote and expand the application of circular RNA in fields such as biomedicine and engineering. Summary of the invention

[0009] The purpose of the present invention is to provide a universal cyclization method which has high cyclization efficiency and can cyclize various types of RNA.

[0010] In a first aspect of the present invention, a polynucleotide for preparing circular RNA is provided, comprising, from the 5' end to the 3' end, an operably connected module Z1, a module to be cyclized L, and a module Z2, wherein the polynucleotide is a linear structure;

[0011] The modules Z1 and Z2 are reconstructed into a ribozyme with self-splicing function, and self-splicing (cutting and connecting) is performed under appropriate conditions, so that the module L to be circularized forms a circular RNA, wherein the splicing site is located in the distal loop region (distal loop region) in the stem-loop structure (stem-loop structure) formed by the module L to be circularized after the circular RNA is formed.

[0012] In another preferred embodiment, the suitable conditions refer to conditions suitable for self-splicing.

[0013] In another preferred embodiment, the module Z1 and the module Z2 are reconstructed into a ribozyme with self-splicing function under appropriate conditions (or annealing conditions).

[0014] In another preferred embodiment, the module L to be cyclized is a linear counterpart corresponding to the target sequence.

[0015] In another preferred embodiment, the linear counterpart corresponding to the target sequence is a linear RNA sequence formed by disconnecting the target circular RNA at an appropriate site. In the same target circular RNA, there may be multiple appropriate disconnection sites, which can form multiple linear counterparts. After cyclization, the multiple linear counterparts can obtain a target circular RNA with the same sequence.

[0016] In another preferred embodiment, the target sequence is a circularized circular RNA.

[0017] In another preferred embodiment, the stem-loop structure is an endogenous stem-loop structure in the target sequence.

[0018] In another preferred embodiment, the stem-loop structure is a single stem-loop structure or a poly stem-loop structure.

[0019] In another preferred embodiment, the stem-loop structure is a single stem-loop structure.

[0020] In another preferred embodiment, the single stem-loop structure has a stem region (also called an "arm region") and a terminal loop region located at the distal end of the stem region.

[0021] In another preferred embodiment, in the single stem-loop structure, the stem region may or may not have one or more internal loop regions.

[0022] In another preferred embodiment, in the single stem-loop structure, the stem region may or may not have one or more bulge regions.

[0023] In another preferred embodiment, the stem-loop structure is a poly stem-loop structure.

[0024] In another preferred embodiment, the poly stem-loop structure has 2-5, preferably 2-4, more preferably 3-4 terminal loop regions.

[0025] In another preferred embodiment, the poly stem-loop structure has a stem region (also called an "arm region") and a terminal loop region located at the distal end of the stem region.

[0026] In another preferred embodiment, in the multi-stem-loop structure, the number of stem regions and terminal loop regions are the same or substantially the same.

[0027] In another preferred embodiment, in the multi-stem-loop structure, each of the stem regions independently has or does not have one or more internal loop regions.

[0028] In another preferred embodiment, in the multi-stem-loop structure, each of the stem regions independently has or does not have one or more bulge regions.

[0029] In another preferred embodiment, the splicing site is a splicing site for circularization.

[0030] In another preferred embodiment, the terminal loop region contains 2-15 nucleotides, preferably 2-10, and more preferably 3-8 nucleotides.

[0031] In another preferred embodiment, the length Y of the stem region is not less than 2 nt, preferably 4≤Y≤100 nt, and more preferably 10≤Y≤100 nt.

[0032] In another preferred embodiment, the module Z1 and / or module Z2 are directly connected to the splicing site respectively.

[0033] In another preferred embodiment, the module Z1 contains the 3' end section h1 of the ribozyme, and the module Z2 contains the 5' end section h2 of the ribozyme; or the module Z1 contains the 5' end section h2 of the ribozyme, and the module Z2 contains the 3' end section h1 of the ribozyme.

[0034] In another preferred embodiment, the polynucleotide comprises RNA, DNA, or RNA / DNA hybrid form.

[0035] In another preferred embodiment, the polynucleotide is RNA.

[0036] In another preferred embodiment, the polynucleotide is single-stranded.

[0037] In another preferred embodiment, the polynucleotide is L-RNA composed of L-ribonucleotides.

[0038] In another preferred embodiment, the polynucleotide has a structure shown in Formula I:

[0039] Z1-L-Z2 (Formula I)

[0040] Z1 contains the 3′ end section h1 of the ribozyme;

[0041] L is the linear counterpart of the target sequence;

[0042] Z2 contains the 5' end section h2 of the ribozyme.

[0043] In another preferred embodiment, the Z1 further comprises a 3' end section h1 of the ribozyme and a random sequence r1, the Z2 further comprises a 5' end section h2 of the ribozyme and a random sequence r2, and the structural formula of the polynucleotide is as shown in Formula II:

[0044] r1-h1-L-h2-r2 (Formula II)

[0045] Wherein r1 and r2 are reverse complementary sequences.

[0046] In another preferred embodiment, the number of reverse complementary base pairs is ≥4, preferably ≥5.

[0047] In another preferred embodiment, the number of reverse complementary base pairs is ≤100 nt, preferably ≤50 nt, and more preferably ≤30 nt.

[0048] In another preferred embodiment, r1 and r2 are completely complementary or substantially complementary (eg, homology ≥ 80%, preferably ≥ 90% or ≥ 95%, or 100%).

[0049] In another preferred embodiment, the 3' end of the module L comprises a ribozyme recognition site i, and the ribozyme recognition site i is the first n nucleotides near the end of the 3' end of the module L, n≥1, preferably 1≤n≤7;

[0050] The 5' end of L comprises a ribozyme recognition site ii, and the ribozyme recognition site ii is the first m nucleotides of the 5' end of L, where m≥1, preferably 1≤m≤7.

[0051] In another preferred embodiment, the polynucleotide can undergo a self-catalytic splicing reaction to generate a circular RNA molecule.

[0052] In another preferred embodiment, the ribozyme is a self-catalytic splicing ribozyme, and the ribozyme can recognize a specific site in module L.

[0053] In another preferred embodiment, the L contains a specific secondary structure, that is, a specific splicing site and its surrounding sequences constitute the secondary structure to be spliced; when the ribozyme matches the specific secondary structure, a high-efficiency splicing reaction can occur at the specific splicing site. The high-efficiency splicing reaction activity is related to the specific secondary structure, and is not limited to the one-dimensional sequence constituting the specific secondary structure.

[0054] In another preferred embodiment, the ribozyme is a group I intron (or type I intron).

[0055] In another preferred embodiment, the ribozyme is selected from the following group: Anabaena pre-tRNA-Leu gene intron, T4 bacteriophage Td gene intron, Tetrahymena intron, etc.

[0056] In another preferred embodiment, the ribozyme is the intron of the pre-tRNA-Leu gene of Anabaena.

[0057] In another preferred embodiment, the ribozyme is an intron of the Td gene of T4 bacteriophage.

[0058] In another preferred embodiment, the ribozyme is a Tetrahymena intron.

[0059] In another preferred embodiment, the ribozyme is a ribozyme in which the key region is mutated. In another preferred embodiment, the ribozyme is a ribozyme in which the internal guide sequence (IGS) is mutated.

[0060] In another preferred embodiment, the ribozyme is a ribozyme modified from the IGS intron of the pre-tRNA-Leu gene of Anabaena.

[0061] In another preferred example, the ribozyme is modified to mutate the IGS in the intron of the Anabaena pre-tRNA-Leu gene, so that the mutated IGS can form n base complementary pairings with the ribozyme recognition site on the target sequence, n≥1, preferably, 1≤n≤7.

[0062] In another preferred embodiment, the ribozyme is a modified ribozyme formed by partially or completely mutating the 10th to 16th positions of the 5' end of the intron of the pre-tRNA-Leu gene of Anabaena. Among them, the mutation rule is: the base at the 10th position at the 5' end of the intron mutates to the complementary base of the 2nd base at the 3' end of the cyclized sequence, the base at the 11th position mutates to the complementary base of the 3rd base at the 3' end of the cyclized sequence, the base at the 12th position mutates to the complementary base of the 4th base at the 3' end of the cyclized sequence, the base at the 13th position mutates to the complementary base of the 5th base at the 3' end of the cyclized sequence, the base at the 14th position mutates to the complementary base of the 6th base at the 3' end of the cyclized sequence, the base at the 15th position mutates to the complementary base of the 7th base at the 3' end of the cyclized sequence, and the base at the 16th position mutates to the complementary base of the 8th base at the 3' end of the cyclized sequence.

[0063] In another preferred embodiment, the ribozyme modification is to mutate the 10th and 11th positions of the intron of the Anabaena pre-tRNA-Leu gene at the 5' end.

[0064] In another preferred embodiment, the ribozyme modification is to mutate the 10th, 11th and 12th positions of the intron of the Anabaena pre-tRNA-Leu gene at the 5' end.

[0065] In another preferred embodiment, the ribozyme modification is to mutate the 10th, 11th, 12th and 13th positions of the intron of the Anabaena pre-tRNA-Leu gene at the 5' end.

[0066] In another preferred embodiment, the ribozyme modification is to mutate the 10th, 11th, 12th, 13th and 14th positions of the intron of the Anabaena pre-tRNA-Leu gene at the 5' end.

[0067] In another preferred embodiment, the module L comprises:

[0068] (L1) fragment i of the noncoding region;

[0069] (L2) mRNA coding region; and

[0070] (L3) fragment ii of the noncoding region;

[0071] Among them, L1 and L3 are formed by cutting a non-coding region at a specific site.

[0072] In another preferred embodiment, the module L comprises:

[0073] (L1a) fragment i of a single stem-loop structural element;

[0074] (L2a) a sequence other than the single stem-loop structure; and

[0075] (L3a) fragment ii of a single stem-loop structural element;

[0076] Among them, L1a and L3a are formed by a single stem-loop structural element being cut at a specific site.

[0077] In another preferred embodiment, the module L is mRNA.

[0078] In another preferred embodiment, the single stem-loop structure is located in the coding region or non-coding region of the mRNA.

[0079] In another preferred embodiment, the single stem-loop structure is located in the non-coding region of the mRNA.

[0080] In another preferred embodiment, the single stem-loop structure is located in the translation initiation element of the mRNA.

[0081] In another preferred embodiment, the translation initiation element includes but is not limited to: internal ribosome entry sites (IRES), Kozak sequences, methylation modification sites, cap-independent translation enhancers (CITE), 5' untranslated region (5'-UTR), 3' untranslated region (3'-UTR), and other elements that can regulate the translation activity of mRNA.

[0082] In another preferred embodiment, the translation initiation element is IRES.

[0083] In another preferred embodiment, the single stem-loop structure is located in the regulatory element of the mRNA.

[0084] In another preferred embodiment, the mRNA regulatory element is an element capable of regulating the behavior of circular mRNA molecules in cells, and the mRNA regulatory element includes but is not limited to: an aptamer sequence element for binding to an aptamer substrate, a binding element for binding to microRNA (microRNA or miRNA), and an element for interacting with other RNA molecules to change the secondary structure or spatial conformation of the circular mRNA.

[0085] In another preferred embodiment, the non-coding region comprises a spacer sequence for avoiding the formation of a secondary structure between the translation initiation element and the coding region.

[0086] In another preferred embodiment, the non-coding region comprises a spacer sequence for preventing the formation of a secondary structure between the regulatory elements.

[0087] In another preferred embodiment, the non-coding region comprises a spacer sequence that prevents IRES from forming a secondary structure with the rest of the circular RNA sequence.

[0088] In another preferred example, the mRNA coding sequence contained in the L2 module is used to encode 1 target protein.

[0089] In another preferred embodiment, the mRNA coding sequence contained in the L2 module is used to encode n (n≥2) target proteins.

[0090] In another preferred embodiment, the n target proteins are the same protein or different proteins.

[0091] In another preferred embodiment, a spacer sequence is inserted between the mRNA coding sequences of the n target proteins, wherein the product encoded by the spacer sequence can be degraded under specific conditions, thereby separating the multiple proteins.

[0092] In another preferred example, the spacer sequence is an RNA sequence encoding a 2A peptide.

[0093] In another preferred embodiment, the single stem loop structure is an SL structure (Stem Loop structure).

[0094] In another preferred embodiment, the non-coding region or the coding region can form a SL structure.

[0095] In another preferred embodiment, the multi-stem-loop structure is a TLSL structure (TLS-like structure).

[0096] In another preferred embodiment, the non-coding region or the coding region can form a specific secondary structure.

[0097] In another preferred embodiment, the secondary structure comprises a TLSL structure (TLS-like structure) module.

[0098] In another preferred embodiment, the module L is determined by the following method:

[0099] (1) directly connecting the 5' and 3' ends of the linear target sequence to be circularized to form a target circular sequence;

[0100] (2) Determine the splicing site (SS) based on the local secondary structure of the target sequence. The polynucleotide sequences on both sides of the splicing site are ribozyme recognition sites.

[0101] (3) Cutting the target circular sequence at the selected splicing site SS to form the module L.

[0102] In another preferred embodiment, the splicing site is located in the SL structure.

[0103] In another preferred embodiment, the SL is composed of an arm connection and a distal loop.

[0104] In another preferred embodiment, the arm may include a bulge and / or an internal loop.

[0105] In another preferred embodiment, the splicing site is located on the distal loop of the SL structure.

[0106] In another preferred embodiment, the splice site is located between the nucleotide U (or T) and the adjacent nucleotide on the distal loop, and the nucleotide U (or T) is located at the 3' end of the module L formed after cleavage at the splice site.

[0107] In another preferred embodiment, for the distal loop region in the stem-loop structure, the size of the distal loop is ≥ 3 nt, preferably 4-25 nt, more preferably 4-20 nt, and most preferably 5-15 nt.

[0108] In another preferred embodiment, the splicing site is located in the SL structure of the translation initiation element.

[0109] In another preferred embodiment, the translation initiation element is a cap-independent translation initiation element.

[0110] In another preferred embodiment, the translation initiation element is an internal ribosome entry site (IRES).

[0111] In another preferred embodiment, the splicing site is located in the SL structure of the mRNA coding region.

[0112] In another preferred embodiment, the splicing site is located in the SL structure of the 5'-UTR region of the mRNA.

[0113] In another preferred embodiment, the splicing site is located in the SL structure of the 3'-UTR region of the mRNA.

[0114] In another preferred embodiment, the coding region encodes a protein product (polypeptide, protein or fragment thereof) or a non-protein product (such as miRNA precursor, ncRNA, etc.).

[0115] In another preferred embodiment, the module L is non-coding RNA (ncRNA).

[0116] In another preferred embodiment, the splicing site is located in the SL structure of non-coding RNA (ncRNA).

[0117] In another preferred embodiment, the splicing site is located in the TLSL structure.

[0118] In another preferred embodiment, the TLSL is composed of a multi-arm junction and multiple distal loops, wherein one arm of the multi-arm junction is the terminal arm, the terminal arm is not connected to the distal loop, and the remaining arms are each closedly connected to a distal loop, forming a multi-distal loop continuous structure with exposed 5' and 3' ends only on the distal arm.

[0119] In another preferred embodiment, the arms of the multi-arm connection may include a bulge and an internal loop, and the connection of the multi-arm connection may include a multiloop.

[0120] In another preferred embodiment, the TLSL structure is a TLSL structure containing 2-5 distal loops.

[0121] In another preferred embodiment, the splicing site is located on any distal loop of the TLSL structure.

[0122] In another preferred embodiment, the splice site is located between the nucleotide U (or T) and the adjacent nucleotide on any distal loop of the TLSL structure, and the nucleotide U (or T) is located at the 3' end of the module L formed after cutting at the splice site.

[0123] In another preferred embodiment, the size of the distal loop is ≥ 3 nt.

[0124] In another preferred embodiment, the splicing site is located in the TLSL structure of the translation initiation element.

[0125] In another preferred embodiment, the translation initiation element is a cap-independent translation initiation element.

[0126] In another preferred embodiment, the translation initiation element is an internal ribosome entry site (IRES).

[0127] In another preferred embodiment, the splicing site is located in the TLSL structure of the mRNA coding region.

[0128] In another preferred embodiment, the splicing site is located in the TLSL structure of the 5'-UTR region of the mRNA.

[0129] In another preferred embodiment, the splicing site is located in the TLSL structure of the 3'-UTR region of the mRNA.

[0130] In another preferred embodiment, the splicing site is located in the TLSL structure of non-coding RNA (ncRNA).

[0131] In another preferred embodiment, the method for determining the splicing site is as follows:

[0132] ① Obtain the secondary structure of the target circular RNA;

[0133] ② Determine the SL structure or TLSL structure based on the secondary structure information;

[0134] ③ Select the splicing site SS on the SL structure or TLSL structure;

[0135] The SL structure or TLSL structure has ribozyme recognition activity.

[0136] In another preferred embodiment, the method for obtaining the secondary structure of the target circular RNA includes prediction using a computer algorithm and determination through experiments.

[0137] In another preferred embodiment, the computer algorithm prediction method includes RNAfold, Linearfold, etc.

[0138] In the second aspect of the present invention, a vector is provided, which comprises the polynucleotide as described in the first aspect of the present invention.

[0139] In another preferred embodiment, the vector contains a nucleic acid sequence (such as a DNA sequence) for producing the polynucleotide as described in the first aspect of the present invention.

[0140] In another preferred embodiment, the vector is a plasmid.

[0141] In another preferred embodiment, the vector is a virus-like particle.

[0142] In another preferred embodiment, the vector is a non-virus-like particle.

[0143] In another preferred embodiment, the vector is a viral vector.

[0144] In another preferred embodiment, the vector is a non-viral vector.

[0145] In the third aspect of the present invention, a method for preparing circular RNA is provided, wherein the polynucleotide as described in the first aspect of the present invention is obtained, and the polynucleotide undergoes self-splicing reaction under appropriate conditions to generate circular RNA.

[0146] In another preferred embodiment, the method specifically comprises the steps of:

[0147] (1) determining the splicing site based on the local secondary structure of the target sequence, and obtaining the module L in the polynucleotide as described in the first aspect of the present invention, i.e., the linear counterpart of the target sequence;

[0148] (2) by appropriate methods, the 3' end section of the ribozyme, the linear counterpart of the target sequence and the 5' end section of the ribozyme are located on the same polynucleotide to obtain a circular RNA precursor molecule;

[0149] (3) The circular RNA precursor molecule obtained in step (2) undergoes self-splicing reaction under appropriate conditions to produce the target circular RNA.

[0150] In another preferred embodiment, the reaction condition is that the reaction system contains magnesium ions.

[0151] In another preferred embodiment, the reaction condition is that the reaction system contains GTP and magnesium ions.

[0152] In another preferred embodiment, the reaction system contains 0.01 mM-10 mM GTP.

[0153] In another preferred embodiment, the reaction system contains 1 mM-100 mM magnesium ions.

[0154] In another preferred embodiment, the pH value of the reaction system is 5-8.

[0155] In another preferred embodiment, the temperature of the reaction system is 25-65°C.

[0156] In another preferred embodiment, the reaction is carried out in vitro for 5 min to 60 min.

[0157] In another preferred embodiment, the reaction conditions are 2±1 mM GTP and a buffer containing magnesium ions (such as 50 mM TrisHCl, 10 mM MgCl2, 1 mM DTT, pH 7.5), 55±3°C for 15±5 minutes.

[0158] In the fourth aspect of the present invention, a circular RNA is provided, which is formed by self-splicing of the polynucleotide described in the first aspect of the present invention.

[0159] In another preferred embodiment, the circular RNA is prepared by the method described in the third aspect of the present invention.

[0160] In another preferred embodiment, the circular RNA is selected from the following group: circular mRNA, circular guide RNA, circular RNA molecular sponge, circular interfering RNA, circular RNA aptamer, circular RNA probe, circular RNA molecular sensor, circular RNA molecular diagnosis, functional non-coding circular RNA, or a combination thereof.

[0161] In another preferred embodiment, the circular mRNA comprises a coding region and a non-coding region.

[0162] In another preferred embodiment, the circular mRNA comprises a coding region and a non-coding region, and the non-coding region comprises a translation initiation element.

[0163] In another preferred embodiment, the non-coding region comprises a spacing element for preventing the formation of a secondary structure between the translation initiation element and the coding region.

[0164] In another preferred embodiment, the polynucleotide for preparing circular RNA or the circular RNA of the present invention contains one or more modified nucleotides.

[0165] In another preferred embodiment, the nucleotides in the RNA molecule are modified in ways that include but are not limited to: N6-methyladenosine (m6A), pseudouridine (ψ), N1-methylpseudouridine (m1ψ) and 5-methoxyuridine (5moU).

[0166] In another preferred embodiment, the nucleotides in the RNA molecule are modified in a manner selected from one or more of N6-methyladenosine (m6A), pseudouridine (ψ), N1-methylpseudouridine (m1ψ) and 5-methoxyuridine (5moU).

[0167] In another preferred embodiment, the circular RNA molecule can express a polypeptide or protein in a cell.

[0168] In another preferred embodiment, the cell is in vivo.

[0169] In another preferred embodiment, the cells are mammalian cells.

[0170] In another preferred embodiment, the mammalian cell is derived from human.

[0171] In another preferred embodiment, the production of the polypeptide or protein is tissue-specific.

[0172] In another preferred example, the tissue-specific localization is in a tissue selected from the group consisting of muscle, heart, spleen, liver, kidney, brain, lung, skin, pancreas, blood and lymph node.

[0173] In another preferred embodiment, the cells are in vitro.

[0174] In another preferred embodiment, the polypeptide or protein is a therapeutic polypeptide or protein. In another preferred embodiment, the polypeptide or protein is a disease-preventive polypeptide or protein.

[0175] In the fifth aspect of the present invention, a pharmaceutical composition is provided, which contains (a) the polynucleotide for preparing circular RNA described in the first aspect of the present invention, the vector described in the second aspect of the present invention, the circular RNA described in the fourth aspect of the present invention, or a combination thereof; and (b) a pharmaceutically acceptable carrier.

[0176] In another preferred embodiment, the pharmaceutical composition includes a vaccine composition.

[0177] In another preferred embodiment, the pharmaceutical composition further contains other additional drugs.

[0178] In another preferred embodiment, the other drugs include but are not limited to one or more selected from the following groups: additional circular RNA, mRNA vaccines, mRNA drugs, vaccine adjuvants, siRNA, small molecule drugs, polypeptide drugs, protein drugs, antibody drugs, cell therapy drugs, gene editing drugs, gene therapy drugs, and microbial therapy.

[0179] In another preferred embodiment, the pharmaceutical composition contains: lipid nanoparticles, liposomes, exosomes, cationic liposomes, targeted lipid nanoparticles, micelles, polymer materials, nano-drug delivery systems, mesoporous materials, adeno-associated viruses, Sendai viruses, protamine carrier systems, and living cell carriers.

[0180] In another preferred embodiment, the pharmaceutically acceptable carrier is a modified lipid nanoparticle.

[0181] In another preferred embodiment, the modified lipid nanoparticles have a targeted delivery function.

[0182] In another preferred embodiment, the pharmaceutical composition is a liquid preparation, a solid preparation, or a gel preparation.

[0183] In another preferred embodiment, the pharmaceutical composition is administered by a method selected from the group consisting of gene gun injection, intravenous injection, intratumoral injection, intramuscular injection, myelin injection, intradermal injection, subcutaneous injection, intranodular injection, nasal inhalation, mucosal infiltration, and microneedle.

[0184] In the sixth aspect of the present invention, a cell is provided, which contains the polynucleotide described in the first aspect of the present invention, or contains the vector described in the second aspect of the present invention, or contains the circular RNA described in the fourth aspect of the present invention, or contains the pharmaceutical composition described in the fifth aspect of the present invention, or a coding sequence for producing the polynucleotide described in the first aspect of the present invention is integrated into the genome.

[0185] In another preferred embodiment, the cell is a mammalian cell.

[0186] In another preferred embodiment, the cells are stem cells.

[0187] In another preferred embodiment, the cells are tumor cells.

[0188] In another preferred embodiment, the cells are non-tumor cells.

[0189] In another preferred embodiment, the cells are human immune cells.

[0190] In another preferred embodiment, the cell is a human T cell.

[0191] In another preferred embodiment, the cells are human macrophages.

[0192] In another preferred embodiment, the cells are human NK cells.

[0193] In another preferred embodiment, the cells are human tumor infiltrating lymphocytes (TIL).

[0194] In the seventh aspect of the present invention, a cell population is provided, characterized in that it contains the polynucleotide described in the first aspect of the present invention, or contains the vector described in the second aspect of the present invention, or contains the circular RNA described in the fourth aspect of the present invention, or a coding sequence for producing the polynucleotide described in the first aspect of the present invention is integrated into the genome.

[0195] In another preferred embodiment, the cell population contains multiple types of mammalian cells.

[0196] In another preferred embodiment, the cell population contains stem cells.

[0197] In another preferred embodiment, the cell population contains tumor cells.

[0198] In another preferred embodiment, the cell population contains non-tumor cells.

[0199] In another preferred embodiment, the cell population contains human immune cells.

[0200] In another preferred example, the cell population contains human T cells.

[0201] In another preferred embodiment, the cell population contains human macrophages.

[0202] In another preferred embodiment, the cell population contains human NK cells.

[0203] In another preferred example, the cell population contains human tumor infiltrating lymphocytes (TIL).

[0204] In the eighth aspect of the present invention, a method for producing a cell population as described in the seventh aspect of the present invention is provided, the method comprising contacting cells of the cell population with a composition, the composition containing (a) the polynucleotide described in the first aspect of the present invention, the vector described in the second aspect of the present invention, the circular RNA described in the fourth aspect of the present invention, or a combination thereof; and (b) a delivery vector.

[0205] In another preferred example, the delivery vector is used to deliver the circular RNA and / or the vector to a cell.

[0206] In the ninth aspect of the present invention, a method for producing a polypeptide or protein in a cell is provided, the method comprising: under the condition that the polypeptide or protein encoding nucleic acid sequence of the circular RNA is translated and the polypeptide or protein is produced in the cell, contacting the cell with: (i) the circular RNA according to the fourth aspect of the present invention, (ii) the vector according to the second aspect of the present invention; and / or (iii) the pharmaceutical composition according to the fifth aspect of the present invention;

[0207] Alternatively, the method comprises: culturing the cell according to the sixth aspect of the present invention or the cell population according to the seventh aspect of the present invention under conditions suitable for expression, thereby producing a polypeptide or a protein.

[0208] In another preferred embodiment, the cell is in vivo.

[0209] In another preferred embodiment, the cells are mammalian cells.

[0210] In another preferred embodiment, the mammalian cell is derived from human.

[0211] In another preferred embodiment, the production of the polypeptide or protein is tissue-specific.

[0212] In another preferred example, the tissue-specific localization is in a tissue selected from the group consisting of muscle, heart, spleen, liver, kidney, brain, lung, skin, pancreas, blood and lymph node.

[0213] In another preferred embodiment, the cells are in vitro.

[0214] In another preferred embodiment, the polypeptide or protein is a therapeutic polypeptide or protein.

[0215] In another preferred embodiment, the polypeptide or protein is a disease-preventive polypeptide or protein.

[0216] In the tenth aspect of the present invention, there is provided a use of the circular RNA as described in the fourth aspect of the present invention, or the vector as described in the second aspect of the present invention, or the pharmaceutical composition as described in the fifth aspect of the present invention, or the cell as described in the sixth aspect of the present invention, or the cell population as described in the seventh aspect of the present invention, or the method for producing a polypeptide or protein in a cell as described in the eighth aspect of the present invention, for preparing a pharmaceutical composition or a cosmetic composition.

[0217] In another preferred embodiment, the pharmaceutical composition is used for a purpose selected from the following group: anti-aging, medical cosmetology, medical care, prevention, diagnosis and treatment of human or animal diseases.

[0218] In another preferred embodiment, the circular RNA is a circular mRNA, and the circular mRNA encodes various types of proteins, including but not limited to: functional proteins, preventive vaccines, therapeutic vaccines, monoclonal antibodies, bispecific antibodies, multispecific antibodies, intracellular antibodies, chimeric antigens, fusion proteins, and cytokines.

[0219] In another preferred embodiment, the circular mRNA is used to prepare mRNA cell therapy drugs in vitro, and the mRNA cell therapy drugs include but are not limited to: chimeric antigen receptor T cell immunotherapy (CAR-T), chimeric antigen receptor NK cell immunotherapy (CAR-NK), tumor infiltrating lymphocyte therapy (TIL), chimeric antigen receptor macrophage immunotherapy (CAR-M), induced pluripotent stem cell therapy (iPS), embryonic stem cell therapy (ESC), adult stem cell therapy (ASC), mesenchymal stem cell therapy (MSC), and hematopoietic stem cell therapy (HSC).

[0220] In another preferred embodiment, the circular mRNA is used to generate mRNA cell therapy drugs in vivo, wherein the circular mRNA is transported to specific cell types in the body through appropriate delivery vectors, and expresses proteins required for cell therapy in the cells, thereby converting the cell types into cell drugs with therapeutic functions, and the cell drugs include but are not limited to chimeric antigen receptor T cell immunotherapy (CAR-T), chimeric antigen receptor NK cell immunotherapy (CAR-NK), tumor infiltrating lymphocyte therapy (TIL), chimeric antigen receptor macrophage immunotherapy (CAR-M), induced pluripotent stem cell therapy (iPS), embryonic stem cell therapy (ESC), adult stem cell therapy (ASC), mesenchymal stem cell therapy (MSC), and hematopoietic stem cell therapy (HSC).

[0221] It should be understood that within the scope of the present invention, the above-mentioned technical features of the present invention and the technical features specifically described below (such as embodiments) can be combined with each other to form a new or preferred technical solution. Due to space limitations, they will not be described one by one here. BRIEF DESCRIPTION OF THE DRAWINGS

[0222] Figure 1 Design principles for a classical PIE cyclization construct are shown.

[0223] Figure 2 A brief design principle of Hi-scarless-PIE is shown.

[0224] Figure 3 The basic secondary structural elements that make up the TLS-like structure (TLSL) are shown.

[0225] Figure 4 A schematic diagram of the secondary structure of a typical TLS-like structure (TLSL) is shown.

[0226] Figure 5 The process and results of determining the TLSL structural elements, modifying the Ana-PIE intronic ribozyme, and forming the "ribozyme-TLSL structural module" to compose the Hi-Scarless-PIE circular RNA construct are shown. Figure 5 A shows the TLSL structural elements, and arrows indicate the locations of splice sites; Figure 5 B shows the capillary electrophoresis detection results of the self-splicing reaction of the natural Ana-PIE intron and the modified Ana-PIE intron.

[0227] Figure 6 The effect of different self-splicing ribozyme recognition sites on circularization efficiency is shown. Figure 6 A shows the location of the first splice site; Figure 6 B shows the location of the second splice site; Figure 6 C shows the capillary electrophoresis detection results.

[0228] Figure 7 It shows whether the product of the present invention is the target circular RNA molecule. Figure 7 A shows agarose gel electrophoresis analysis of the products; Figure 7 B shows agarose gel electrophoresis analysis of the products after RNase H treatment; Figure 7 C shows the sequencing results of the splice sites.

[0229] Figure 8 The effects of different ribozyme mutations on the Hi-Scarless-PIE cyclization efficiency are shown. Figure 8 A shows the secondary structure of the native Anabaena intron with the default numbering of nucleotides and the regions where the ribozyme was engineered; Figure 8 B shows the agarose gel electrophoresis analysis of circular RNA synthesized by different modified ribozymes.

[0230] Fig. 9 Shown are the construct design based on the PV3 IRES-derived TLSL structure and the experimental results for validation of the cyclization results.

[0231] Fig.10 Shown are the experimental results of construct design based on HRV-B3 IRES- and EV-B107 IRES-derived TLSL structures and validation of the circularization results.

[0232] Fig.11Shown are the construct design based on the SimianV4 IRES-derived TLSL structure and the experimental results for validation of the circularization results.

[0233] Fig.12 Shown are the construct design based on the BEV1 IRES-derived TLSL structure and the experimental results for the validation of the cyclization results.

[0234] Fig.13 The cyclization results of different "ribozyme-TLSL cyclization modules" are shown. Fig.13 A shows the detection profile of capillary electrophoresis; Fig.13 B shows the electrophoresis of the crude RNA product.

[0235] Fig.14 The cyclization effect of EMCV after insertion of TLSL is shown. Fig.14 A shows the secondary structures of TLSL, EMCV IRES, and EMCV+TLSL; Fig.14 B shows the agarose gel electrophoresis detection results of the products; Fig.14 C shows the protein expression results.

[0236] Fig.15 The circularization effect of CrPV after insertion of TLSL is shown. Fig.15 A shows the secondary structures of TLSL, CrPV IRES, and CrPV+TLSL; Fig.15 B shows the agarose gel electrophoresis detection results of the products; Fig.15 C shows the protein expression results.

[0237] Fig.16 The results of the analysis of the distribution of TLSL structural elements in various types of RNA molecules are shown. Fig.16 A shows the typical structural elements of the trefoil TLSL; Fig.16 B shows the universal Hi-scarless-PIE cyclization platform; Fig.16 Some representative trefoil TLSL structural elements from IRES, human 5'-UTR, human 3'-UTR, human CDS, and non-coding RNA sequence sets are shown in CG, respectively.

[0238] Fig.17 Shown is the expression of immunogenic cytokines induced by circ-GLuc-mRNA prepared using the classical PIE and Hi-Scarless-PIE methods in A549 cells.

[0239] Fig.18 The results of detecting the protein expression levels of circular GLuc mRNA with different IRES sequences in HEK-293T and A549 cells are shown.

[0240] Fig.19 Shown are protein expressions of cyclic hEPO and cyclic EGFP mRNAs prepared using the Hi-Scarless-PIE method.

[0241] Fig. 20 The change trend of protein expression levels over time in HEK-293T and A549 cells using traditional capped and tailed linear mRNA, the classic PIE system and the circular mRNA prepared by Hi-Scarless-PIE in the present disclosure is shown.

[0242] Fig.21 Shown are the preparation and characterization of circular FLuc-mRNA prepared using the classical PIE method and the Hi-Scarless-PIE method for LNP lipid nanoparticle encapsulation.

[0243] Fig. 22 Shown are the protein expressions of circular FLuc-mRNA-LNP lipid nanoparticles prepared using the classical PIE method and the Hi-Scarless-PIE method in mice.

[0244] Fig.23 The brief design principle of the Uni-cRNA of the present invention is shown.

[0245] Fig.24 A schematic diagram of splice site selection in the IRES region during the design of Uni-cRNA constructs is shown. Fig.24 The blue part in B is the terminal loop, the green part is the complementary paired bases in the stem region, the yellow part is the bulge or internal loop in the stem region, and the arrow points to the splicing site.

[0246] Fig.25 The splice site selected during the design of the Uni-cRNA construct is shown to be located at the terminal loop of the CDS region. The arrow points to the splice site.

[0247] Fig.26 The splice sites selected after sequence optimization of the CDS region are shown. The arrows point to the splice sites.

[0248] Fig. 27 The splice sites selected in the inner loop of the IRES region during the design of the Uni-cRNA construct are shown. The arrows point to the splice sites.

[0249] Fig.28 The results of capillary electrophoresis detection of the circularization efficiency of uni-cRNA in preparing circular RNA are shown.

[0250] Fig.29Characterization of circular RNA products prepared from Uni-cRNA is shown.

[0251] Fig.30 The results of detecting proteins expressed by circular mRNA prepared by Uni-cRNA are shown.

[0252] Fig.31 The distribution of the single stem-loop structure in the "ribozyme-SL cyclization module" in various RNA sequences is shown. DETAILED DESCRIPTION

[0253] After extensive and in-depth research, through a large number of screenings, the inventors have developed a cyclization technology based on the secondary structure of circular RNA for the first time, including a precursor polynucleotide for constructing circular RNA, and provide a method for preparing circular RNA based on the precursor polynucleotide and the use of the circular RNA prepared by the same. Specifically, the inventors determine the optimized ribozyme recognition site based on a specific local secondary structure (i.e., a stem-loop structure), and combine wild-type or improved RNA ribozymes to construct a circular RNA precursor polynucleotide comprising an operably connected module Z1, a module L to be cyclized, and a module Z2. The circular RNA precursor polynucleotide of the present invention, under annealing conditions, the module Z1 and the module Z2 are reconstructed into a ribozyme with a self-splicing function, thereby splicing (cutting and connecting) the module L to be cyclized, so that the module L to be cyclized forms a circular RNA. The inventors unexpectedly found that the cyclization technology based on a single stem-loop structure and a multi-stem-loop structure of the present invention has excellent cyclization efficiency, and the cyclization efficiency is not affected by the different sequences of the cyclized RNA, and is a general method for cyclizing RNA. In addition, no additional sequence may be introduced into the circular RNA product of the present invention, which is a true scarless circularization without the introduction of additional nucleotides. On this basis, the present invention was completed.

[0254] RNA cyclization strategy of the present invention

[0255] RNA cyclization can be achieved by using the self-splicing properties of Group I intron ribozymes or Group II intron ribozymes. In bacteria and non-metazoan eukaryotes, some introns in rRNA, tRNA, and mRNA have autocatalytic activity, which can cut out the introns in the "5'-exon-intron ribozyme-exon-3'" arrangement by back-splicing, and at the same time, the two exons are directly connected (Past, present, and future of circRNAs.Patop IL, Wüst S, Kadener S.EMBO J.2019Aug 15; 38(16):e100836.). According to the above principle, the permuted intron exon (PIE) system was proposed as early as 1992. The PIE system divides the intron ribozyme into two fragments, the 5' end and the 3' end, and then places the 3' end intron fragment upstream of the exon 5' and the 5' end intron fragment downstream of the exon 3', forming an arrangement of "5'-intron ribozyme 3' fragment-exon-intron ribozyme 5' fragment-3'" (Group I permuted intron-exon (PIE) sequences self-splice to produce circular exons. M Puttaraju, MD Been. Nucleic Acids Res. 1992 Oct 25; 20 (20): 5357-64.). At this time, the two reverse splicing sites of the ribozyme are respectively located at the two ends of the exon (3'SS and 5'SS), so the self-catalytic reverse splicing reaction will produce circularized exons. When the sequence between the two splice sites is replaced with the target gene, the circularization preparation of the target gene can be achieved. The design principle of the classic PIE circularization construct is as follows: Figure 1 shown.

[0256] In the present invention, an improved high-efficiency and high-precision PIE cyclization system (High-efficient and Scarless PIE, hereinafter referred to as Hi-scarless-PIE) based on endogenous multi-stem-loop secondary structure is provided. Based on the secondary structure characteristics of the target sequence, a suitable splicing site is selected, and the I group intron ribozyme is modified accordingly to construct a circular RNA precursor molecule, which performs autocatalytic splicing under suitable conditions, and the formed circular RNA does not contain any sequence other than the target sequence (for example, an exon sequence or a spacer sequence from the ribozyme), ensuring that the sequence remains unchanged during the RNA cyclization process. Experiments have shown that the system has a high efficiency in cyclizing RNA, and the cyclization effect is not affected by the target sequence, and can be used for the cyclization of any RNA. Figure 2 The brief design principle of Hi-scarless-PIE is shown.

[0257] In the present invention, a universal, high-efficiency, high-fidelity RNA circularization platform (Universal & High-efficient & High-fidelity RNA Circularization Platform, i.e. Uni-cRNA) is also provided. Uni-cRNA is based on the endogenous single stem-loop secondary structure characteristics of the target sequence, selects a suitable splicing site, and transforms the I group intron ribozyme accordingly, thereby constructing a circular RNA precursor molecule, which performs autocatalytic splicing under suitable conditions, and the formed circular RNA does not contain any sequence other than the target sequence (for example, exon sequence or spacer sequence from ribozyme), ensuring high sequence fidelity during RNA circulization (i.e., the sequence of the circular RNA product is completely consistent with the target sequence). Experiments have proved from multiple angles that the system can prepare target circular RNA, has high circulization efficiency, and the circulization efficiency is not affected by the target sequence, and can circulize any RNA. Figure 2 A brief design principle of Uni-cRNA is shown.

[0258] Circular RNA precursor molecules of the present invention

[0259] Based on the above Hi-scarless-PIE or Uni-cRNA strategy, the circular RNA precursor molecule of the present invention is designed, which comprises operably connected module Z1, module L, and module Z2 from the 5' end to the 3' end, as shown in Formula I:

[0260] Z1-L-Z2 (Formula I)

[0261] Z1 contains the 3′ end section h1 of the ribozyme;

[0262] L is the linear counterpart of the target sequence;

[0263] Z2 contains the 5' end section h2 of the ribozyme.

[0264] The circular RNA precursor molecule can undergo a self-catalytic splicing reaction to generate the circular RNA molecule of the present invention.

[0265] In some embodiments, the Z1 further comprises a 3' end section h1 of the ribozyme and a random sequence r1, the Z2 further comprises a 5' end section h2 of the ribozyme and a random sequence r2, and the structural formula of the polynucleotide is as shown in Formula II:

[0266] r1-h1-L-h2-r2 (Formula II)

[0267] Wherein r1 and r2 are reverse complementary sequences.

[0268] The random sequences r1 and r2 are used to promote the ribozyme slice to fold into the correct secondary structure to have a more stable splicing activity. During the self-splicing process, the random sequences r1 and r2 will be cut off together with the ribozyme slice and will not remain in the target sequence.

[0269] In the present invention, the module L is determined by the following method:

[0270] (1) First, the 5' and 3' ends of the linear target sequence to be circularized are directly connected to form a target circular RNA sequence.

[0271] (2) Secondly, determine the splicing site (SS) of the target circular sequence. The splicing site is located on the terminal loop of the SL structural element or the TLSL structural element, and the polynucleotide sequences on both sides of the splicing site are ribozyme recognition sites.

[0272] (3) The target circular sequence is cut at the selected splicing site SS to form module L.

[0273] TLSL structural elements

[0274] In the Hi-scarless-PIE circularization strategy of the present invention, the splicing site is located at the TLSL structural element of the target sequence.

[0275] The TLSL structure of the present invention is as follows Figure 3 and Figure 4As shown, it is composed of a multi-arm junction and multiple distal loops, wherein one arm of the multi-arm junction is a distal arm, the distal arm is not connected to the distal loop, and the remaining arms are respectively connected to a distal loop to form a multi-distal loop continuous structure with only exposed 5' and 3' ends on the distal arm. The arms of the multi-arm junction may include a bulge and an internal loop, and the junction of the multi-arm junction may include a multiloop.

[0276] In one embodiment, the TLSL structure is a TLSL structure containing 2-5 terminal rings.

[0277] In one embodiment, the splice site is located at any terminal loop of the TLSL structure.

[0278] In one embodiment, the size of the terminal loop is ≥ 4 nt.

[0279] The TLSL structural element of the present invention is endogenous in the target sequence, and the endogenous refers to a structure that already exists within the target sequence, rather than a structure formed by an exogenously inserted sequence.

[0280] In some embodiments, the TLSL structural module is located at the translation initiation element site, and TLSL may be naturally present at the translation initiation element site, or may be artificially inserted into the translation initiation element site. The translation initiation element is a cap-independent translation initiation element, including but not limited to: internal ribosome entry site (Internalribosome entry site, IRES), Kozak sequence, translation initiation element containing methylation modification site, cap-independent translation enhancer (cap-independent translation enhancers, CITE), 5' untranslated region (5'untranslated region, 5'-UTR), 3' untranslated region (3'untranslated region, 3'-UTR), other elements with cap-independent translation initiation activity, and other elements with cap-independent translation initiation activity newly discovered in the future. Among them, mammalian mRNA genes containing CITE include but are not limited to: human Fibroblast growth factor 9 (FGF-9).

[0281] In some embodiments, the cap-independent translation initiation element is an internal ribosome entry site (IRES). The IRES sequence usually contains a sequence that can fold into a structure similar to the initiator tRNA. This structural element is a typical TLSL structure that can mediate ribosomal RNA binding, thereby enhancing the initiation of protein translation. A large number of IRES sequences have a conserved TLSL structural module that can be recognized by the ribozyme of the present invention.

[0282] In some embodiments, the IRES sequence is a non-natural sequence. In some embodiments, the IRES sequence is an IRES sequence artificially inserted with a TLSL module.

[0283] SL structural elements

[0284] In the Uni-cRNA circularization strategy of the present invention, the splicing site is located at the SL structural element of the target sequence.

[0285] As used herein, the terms "SL structure", "Stem loop structure" and "single stem loop structure" are used interchangeably, all referring to a single stem loop secondary structure contained in a target circular RNA molecule. The SL structure is composed of a stem region (also called an "arm") and a distal loop connected together. The stem region may include a bulge and / or an internal loop.

[0286] In some embodiments, the size of the terminal loop is ≥ 4 nt, preferably 4-25 nt, more preferably 4-20 nt, and most preferably 5-15 nt.

[0287] In some embodiments, for an SL structure, the stem region is composed of two completely complementary or partially complementary nucleotide segments, the distal end of each nucleotide segment is connected to the terminal loop, and the proximal end of each nucleotide segment is respectively connected to the main body of the polynucleotide for preparing circular RNA of the present invention (i.e., the remaining part of the target circular RNA after removing the SL structure). When the lengths of the two nucleotide segments are Y1 and Y2, respectively, the length Y of the stem region = (Y1 + Y2) / 2. Generally, the length Y is not particularly limited. Typically, Y is 4-100nt, more preferably, 6-80nt, and optimally, 6-40nt. Generally, when the two nucleotide segments are completely complementary, Y can be shorter, such as 3-10nt; when the two nucleotide segments are not completely complementary, Y can be longer, such as 4-20nt or 20-60nt.

[0288] In some embodiments, the number of complementary base pairs in the stem region is P, which is usually ≥ 3, preferably P is 5-100, and more preferably, 10-100.

[0289] In some embodiments, the ratio of the number of complementary base pairs (P) to the length of the stem region (Y) (ie, P / Y) is 40% to 100%, preferably 50% to 100%, and more preferably 60% to 95%.

[0290] A schematic SL structure is shown below: Figure 3 As shown in B, the lengths of the two nucleotide segments in the stem region, Y1 and Y2, are 23 and 23 respectively, the length of the stem region Y = (23 + 23) / 2 = 23, the number of complementary base pairs P = 13, and the ratio of P / Y is 57%.

[0291] In some embodiments, in an SL structure of the present invention, the number of the convex rings and / or inner rings is 0-4, preferably 0-3, and more preferably 0-2.

[0292] The SL structural element of the present invention is endogenous to the target sequence, that is, a structure that already exists within the target sequence, rather than a structure formed by an exogenously inserted sequence.

[0293] The SL structural element of the present invention is widely present in different types of RNA molecules. The stem-loop structure exists in almost any natural RNA molecule and is one of the most basic elements for forming a complex RNA secondary structure and further forming a three-dimensional conformation. The present invention creatively designs a self-splicing ribozyme system that can efficiently splice RNA sequences with secondary structural characteristics of the stem-loop structure.

[0294] In the present invention, splice sites are selected on the SL structural element based on the secondary structure of the target circular sequence.

[0295] In some embodiments, the splice site is located on the terminal loop of the SL structure.

[0296] In some embodiments, the splice site is located between the nucleotide U (or T) and the adjacent nucleotide on the terminal loop, and the nucleotide U (or T) is located at the 3' end of the linear RNA sequence formed after cleavage at the splice site.

[0297] In some embodiments, the target circular RNA molecule is a target circular mRNA molecule, and the SL structure is located in the coding region or non-coding region of the target circular mRNA molecule.

[0298] In some embodiments, the SL structure is located at the translation initiation element site of the target circular mRNA molecule. The SL structure may be naturally present at the translation initiation element site, or may be artificially inserted into the translation initiation element site. The translation initiation element is a cap-independent translation initiation element, including but not limited to: an internal ribosome entry site (Internal ribosome entry site, IRES), a Kozak sequence, a translation initiation element containing a methylation modification site, a cap-independent translation enhancer (cap-independent translation enhancers, CITE), a 5' untranslated region (5'untranslated region, 5'-UTR), a 3' untranslated region (3'untranslated region, 3'-UTR), other elements with cap-independent translation initiation activity, and other elements with cap-independent translation initiation activity newly discovered in the future. Among them, mammalian mRNA genes containing CITE include but are not limited to: human Fibroblast growth factor 9 (FGF-9).

[0299] In some embodiments, the translation initiation element is an internal ribosome entry site (IRES). IRES is a highly structured region with abundant secondary structures formed inside. Almost any IRES contains SL structures that can be recognized by the ribozyme of the present invention.

[0300] In some embodiments, the IRES sequences of the present invention include, but are not limited to, IRES sequences derived from the following species or genes: coxsackievirus, echovirus, poliovirus, rhinovirus, human rhinovirus, cricket paralysis virus, bovine enterovirus, enterovirus, human poliovirus, encephalomyocarditis virus, hepatitis C virus, hepatitis A virus, simian virus, aphid lethal paralysis virus, reticuloendotheliosis virus, murine encephalomyelitis virus, human paravirus, bovine viral diarrhea virus, foot-and-mouth disease virus, Kashmir bee virus, acute paralysis virus, gramineous aphid virus, Taura syndrome virus, GB virus type C, bovine crest virus, Kubu virus, picornavirus, swine vesicular disease virus. Aptamers of human FGF2, human SFTPA1, human AML1 / RUNX1, human AQP4, human AT1R, human BAG-1, human BCL2, human BiP, human c-IAP1, human c-myc, human eIF4G, human LEF1, human n-myc, human p27kipl, human PDGF2 / c-sis, human p53, human Pim-1, human UNR, human VEGF-A, human XIAP, human c-src, human FGF-1, Drosophila antennapedia, mouse NDST4L, mouse HIF1α, mouse Gtx, mouse Rbm3, mouse UtrA, canine Scamper, Drosophila reaper, Drosophila Ubx, Drosophila hairless, Saccharomyces cerevisiae TFIID, Saccharomyces cerevisiae YAP1, or eIF4G.

[0301] In some embodiments, the IRES sequence is a non-natural sequence.

[0302] In some embodiments, the IRES sequence is an artificially synthesized IRES sequence containing a SL structure.

[0303] In some embodiments, the SL structure is located in the coding region of the target circular mRNA sequence.

[0304] In some embodiments, the SL structure is located in the 5'-UTR region of the target circular mRNA sequence.

[0305] In some embodiments, the SL structure is located in the 3'-UTR region of the target circular mRNA sequence.

[0306] In some embodiments, the target circular RNA sequence is a non-coding RNA (ncRNA).

[0307] In some embodiments, the target circular RNA sequence is an RNA aptamer, a riboswitch, a toehold switche, a gene editing or RNA editing guide-RNA, snRNA, snoRNA, siRNA, miRNA, piRNA, lncRNA, tRNA, or rRNA.

[0308] Ribozyme of the present invention

[0309] As used herein, the term "ribozyme" refers to RNA with catalytic activity. The ribozyme in the present invention is a ribozyme that can self-catalyze splicing. The chemical nature of self-splicing ribozymes for cyclization is the breaking and recombination of the chemical bonds between the ribozyme splicing active region and the splicing site, and the activity of the ribozyme and the kinetic process of chemical bond breaking and recombination depend on the relative spatial position of each functional module involved in the reaction (Joining the two domains of a group I ribozyme to form the catalytic core. Tanner MA, Cech TR. Science. 1997 Feb 7; 275 (5301): 847-9.). The ribozyme design of the present invention is based on group I introns.

[0310] A large number of natural group I introns have been discovered. Although different group I introns have great differences in sequence, all group I introns contain the same core secondary structure skeleton, which constitutes the self-catalytic splicing ribozyme (Crystal structure of a group I intron splicing intermediate. Adams PL, Stahley MR, Gill ML, et al. RNA. 2004 Dec; 10 (12): 1867-87.). The self-splicing activity of group I intron ribozymes depends on their conserved secondary structure. Therefore, all ribozyme systems with self-splicing function derived from group I introns rely on the same self-splicing reaction mechanism (Joining the two domains of a group I ribozyme to form the catalytic core. Tanner MA, Cech TR. Science. 1997 Feb 7; 275(5301): 847-9. Crystal structure of a self-splicing group I intron with both exons. Peter L Adams, Mary R Stahley, Anne B Kosek, et al. Nature. 2004 Jul 1; 430(6995): 45-50.).

[0311] Specifically, the main mechanism of the self-splicing reaction of group I introns is as follows: the intron located between the upstream exon Exon1 and the downstream exon Exon2 can cut out the intron fragment and connect Exon1 and Exon2 through a self-catalyzed splicing reaction. The splicing reaction mechanism of group I introns mainly includes two-step esterification reactions. A sequence at the 3' end of Exon1 forms a base complementary pairing with a sequence at the 5' end of the intron, and this pairing is called helix P1. In helix P1, the sequence on the ribozyme P1 is called the internal guide sequence (IGS). The formation of a GU mismatch between IGS and the first nucleotide of Exon1 is crucial for the first step of the esterification reaction. The ribozyme relies on this GU mismatch to accurately identify the splicing site (5'splicing site, 5'SS) of the upstream Exon. In the first step of the esterification reaction, the 3'-OH of free guanosine or guanylate attacks the 5'SS, so that the exon and intron are cut at the 5'SS, and the exon 5'SS exposes the free 3'-OH for the second step of the esterification reaction. In the second step, the free 3'-OH of the 5'SS attacks the phosphodiester bond on the splicing site (3'splicing site, 3'SS) at the 5' end of Exon2 as a nucleophilic group, so that the intron is cut off from Exon2, and a new phosphodiester bond is formed between the 3' end of Exon1 and the 5' end of Exon2, so that Exon1 and Exon2 are connected. The accurate recognition of 5'SS must rely on the formation of helix P1 between introns and exons. In contrast, whether a complementary pairing is formed between introns and Exon2 is not a necessary condition for accurate recognition of 3'SS. Therefore, the design of the ribozyme in the present invention mainly focuses on the modification of P1 of group I introns. On the exon, a sequence that forms a complementary pair with ribozyme P1 (i.e., helix P1) at the 5'SS end is also called the ribozyme recognition site. The ribozyme recognition site includes the first nucleotide at the 5' end of the exon to the last complementary nucleotide on helixP1. In different group I introns, the length of the ribozyme recognition site required to cause a high-efficiency autocatalytic splicing reaction varies.

[0312] In some embodiments, the ribozyme of the present invention is a ribozyme obtained by mutation modification of key regions based on natural ribozymes.

[0313] In some embodiments, IGS in a natural ribozyme is mutated to form an engineered ribozyme.

[0314] In certain embodiments, the IGS in the natural ribozyme is mutated so that the mutated IGS can form multiple base complementary pairs with the ribozyme recognition site. In the described embodiments, the present invention first discovered that when the IGS in the anabaena intron forms a 2-7nt complementary pairing with the ribozyme recognition site (i.e., the length of the ribozyme recognition site is within the range of 2-7nt), the autocatalytic splicing reaction can be efficiently achieved. However, when the length of the ribozyme recognition site is 8nt, the cyclization efficiency is significantly reduced, which may be related to the structure of the P2 module being affected so that the ribozyme activity is affected.

[0315] In some embodiments, the ribozyme of the present invention is a modified ribozyme formed by partially or completely mutating the 10th to 16th positions of the 5' end of the pre-tRNA-Leu gene intron of Anabaena (Ana-PIE for short). Among them, the mutation rule is: the base at the 10th position at the 5' end of the PIE intron mutates to the complementary base of the 2nd base at the 3' end of the cyclized sequence, the base at the 11th position mutates to the complementary base of the 3rd base at the 3' end of the cyclized sequence, the base at the 12th position mutates to the complementary base of the 4th base at the 3' end of the cyclized sequence, the base at the 13th position mutates to the complementary base of the 5th base at the 3' end of the cyclized sequence, the base at the 14th position mutates to the complementary base of the 6th base at the 3' end of the cyclized sequence, the base at the 15th position mutates to the complementary base of the 7th base at the 3' end of the cyclized sequence, and the base at the 16th position mutates to the complementary base of the 8th base at the 3' end of the cyclized sequence.

[0316] In some embodiments, positions 10 and 11 of the 5' end of the intron Ana-PIE from Anabaena are mutated.

[0317] In some embodiments, positions 10, 11 and 12 of the 5' end of the intron Ana-PIE from Anabaena are mutated.

[0318] In some embodiments, positions 10, 11, 12 and 13 at the 5' end of the intron Ana-PIE from Anabaena are mutated.

[0319] In some embodiments, positions 10, 11, 12, 13 and 14 of the 5' end of the intron Ana-PIE from Anabaena are mutated.

[0320] Circular RNA of the present invention

[0321] As used herein, the terms "circular RNA", "circular RNA", "circular RNA" and "circRNA" are used interchangeably to refer to a class of molecules without 5' and 3' ends that are covalently linked head to tail. Due to the special "circular" structure, circular RNA molecules can resist degradation by RNases and show higher stability than linear RNA molecules in vivo and in vitro.

[0322] The circular RNA of the present invention is formed by the circular RNA precursor of the present invention through a self-splicing reaction. In some embodiments, the circular RNA is a circular mRNA, and the circular mRNA can express proteins in mammals with high efficiency.

[0323] In some embodiments, the circular RNA is a circular noncoding RNA (ncRNA).

[0324] The circular RNA of the present invention is prepared by the following method:

[0325] (1) The circular RNA precursor molecule of the present invention can be obtained by chemical synthesis or transcription from a vector.

[0326] In some embodiments, the circular RNA precursor molecule of the present invention is directly obtained by chemical synthesis.

[0327] In some embodiments, the circular RNA precursor molecule of the present invention is transcribed from a vector. The steps are:

[0328] ① constructing a plasmid vector containing the circular RNA precursor molecule sequence;

[0329] ② Use a plasmid vector to transcribe and produce circular RNA precursor molecules.

[0330] (2) Exposing the circular RNA precursor molecule obtained in step (1) to certain reaction conditions to cause a self-splicing reaction to obtain the target circular RNA.

[0331] In some embodiments, the reaction condition is that the reaction system contains GTP and magnesium ions.

[0332] In some embodiments, the reaction conditions are that the reaction system contains 0.01mM-10mM GTP, 1mM-100mM magnesium ions, the pH value of the reaction system is between 5-8, the temperature at which the cyclization reaction occurs is between 25-65°C, and the cyclization reaction takes place in vitro for about 5min-60min.

[0333] In some embodiments, the reaction conditions are 2±1 mM GTP and a buffer containing magnesium ions (50 mM TrisHCl, 10 mM MgCl2, 1 mM DTT, pH 7.5) at 55±3°C for 15±5 minutes.

[0334] The main advantages of the present invention include:

[0335] (1) The cyclization platform has excellent versatility, and the cyclization effect is not affected by different target sequences, and can achieve high-fidelity and high-efficiency cyclization for diversified RNA sequences. The cyclization module at the core of the circular RNA construct of the present invention is a single stem-loop structure or a multi-stem-loop structure (commonly referred to as a stem-loop structure). The stem-loop structure of the present invention can be widely present in any type of RNA molecule (including various types of long-chain and short-chain RNA molecules), especially the single stem-loop structure is more widely distributed, which makes the Uni-cRNA cyclization platform of the present invention have strong versatility.

[0336] (2) The cyclization reaction does not introduce any additional sequences. In the final product of circular RNA, the sequence of the linear RNA to be cyclized is completely consistent with the sequence of the generated circular RNA. In biomedical applications, ensuring that the sequence of the RNA molecule remains unchanged after cyclization can avoid potential risks caused by sequence changes.

[0337] (3) High cyclization efficiency. The cyclization efficiency of the Uni-cRNA with a single stem-loop structure of the present invention can be as high as about 80% or higher, while the cyclization efficiency of the Hi-scarless-PIE with a multi-stem-loop structure of the present invention is about 90-95% or higher. The cyclization efficiency of both stem-loop structures is unexpectedly high, which is beneficial to the downstream process after cyclization and has good practicality.

[0338] (4) The prepared circular mRNA has prolonged stability, a long half-life in vivo, and significantly increased protein expression efficiency. Circular RNA molecules exist in the body for a longer time and translate proteins for a longer time, so they show better drug properties than linear mRNA, laying the foundation for the preparation of circular RNA products for medical use.

[0339] (5) The circular RNA molecules prepared by the present invention have lower immunogenicity because they do not carry additional sequences. The circular RNA of the present invention has high purity, which is beneficial to reducing its immunogenicity.

[0340] The present invention will be further described below in conjunction with specific examples. It should be understood that these examples are intended to illustrate the present invention only and are not intended to limit the scope of the present invention. The experimental methods in the following examples where specific conditions are not specified are usually performed under conventional conditions, such as those described in Sambrook et al., Molecular Cloning: A Laboratory Manual (New York: Cold Spring Harbor Laboratory Press, 1989), or under conditions recommended by the manufacturer. Unless otherwise indicated, percentages and parts are weight percentages and weight parts.

[0341] Example 1. Design and verification of Hi-Scarless-PIE circular RNA construct

[0342] This embodiment involves the process and method of determining the TLSL cyclization element, modifying the Ana-PIE intron ribozyme, forming a "ribozyme-TLSL cyclization module", and further forming a Hi-Scarless-PIE circular RNA construct, and experimentally verifying the effect of the construct in synthesizing circular RNA.

[0343] The design principle of the Hi-Scarless-PIE circular RNA construct in the present invention is as follows Figure 2 shown.

[0344] The TLSL cyclization element consists of the ribozyme recognition site and the secondary structure surrounding it.

[0345] First, the TLSL cyclization module was obtained by the following steps: In CVB3 IRES (SEQ ID NO: 1), the nucleotide sequence of the TLSL cyclization element located at positions 1 to 94 of CVB3 IRES was determined based on the secondary structure and named TLSL-CVB3-IRES (SEQ ID NO: 2), and its secondary structure was as follows: Figure 5 As shown in A.

[0346] Secondly, the splicing site is selected: the splicing site is located between the 22nd U and the 23rd C of the TLSL-CVB3-IRES sequence, and the splicing site is located on the loop of the first stem loop close to the 5' end in the secondary structure of TLSL-CVB3-IRES. Figure 5 The arrows in A show the locations of the splice sites.

[0347] In this example, the target circular RNA sequence (SEQ ID NO: 4) is composed of CVB3-IRES (SEQ ID NO: 1) and the mRNA coding region of Gaussia Luciferase (GLuc) (SEQ ID NO: 3). According to the selected splicing site, the Seq4 sequence is separated from the splicing site to form a linear counterpart of the target circular RNA sequence (SEQ ID NO: 5).

[0348] Design of a GLuc circular RNA precursor molecule based on the modified Ana-PIE ribozyme: The 3' end slice of the Ana-PIE ribozyme (SEQ ID NO: 6), the linear counterpart of the target circular RNA sequence (SEQ ID NO: 5) and the 5' end slice of the modified Ana-PIE ribozyme (SEQ ID NO: 7) were connected to each other to form a GLuc circular RNA precursor molecule based on the modified Ana-PIE ribozyme, named circ-CVB3-IRES-GLuc-precursor-1 (SEQ ID NO: 8).

[0349] Design of a GLuc circular RNA precursor molecule based on the unmodified Ana-PIE ribozyme: The 3' end slice of the Ana-PIE ribozyme (SEQ ID NO: 6), the linear counterpart of the target circular RNA sequence (SEQ ID NO: 5) and the 5' end slice of the Ana-PIE ribozyme (SEQ ID NO: 9) were connected to each other to form a GLuc circular RNA precursor molecule based on the unmodified Ana-PIE ribozyme, named circ-CVB3-IRES-GLuc-precursor-2 (SEQ ID NO: 10).

[0350] Synthesis of the circular RNA precursor molecules and constructs designed above: First, synthesize the DNA sequences corresponding to SEQ ID NO: 1, SEQ ID NO: 3, SEQ ID NO: 6, and SEQ ID NO: 9 (Beijing Qingke Biotechnology Co., Ltd.). These DNA sequences are cloned into the linear plasmid vector pUC57 containing the T7 promoter generated by PCR by molecular cloning (GenBuilder Plus Cloning Kit, L00744, GenScript Biotechnology Co., Ltd.) to obtain the corresponding construct (i.e., the construct formed using the natural Ana-PIE intron). The following plasmid is obtained: pUC57-CVB3-GLuc-1. On the basis of this construct, the sequence of the Ana-PIE ribozyme 5' end section (SEQ ID NO: 9) is transformed, wherein the A at the 10th position of the 5' end of SEQ ID NO: 9 is mutated to T, and the G at the 11th position is mutated to C, forming the transformed Ana-PIE ribozyme intron construct pUC57-CVB3-GLuc-2.

[0351] Preparation of linear plasmid template: Transform the plasmid into Top10 competent cells, plate, and select clones to expand culture in LB medium containing Amp resistance the next day (37°C / 200rpm / overnight). Then perform plasmid extraction (endotoxin-free plasmid mini-extraction kit, DP118, Tiangen Biochemical Technology (Beijing) Co., Ltd.). Then use EcoRI (1040, Takara) single enzyme digestion method to prepare linearized plasmid template (37°C, 1 hour). The cleavage products were recovered by PCR&DNA Cleanup Kit (T1030, New England Biolabs), the concentration was determined by Nano-Drop (Thermo), and the cleavage products were identified by 1.5% agarose gel electrophoresis. The purified linear plasmid was used as a template for in vitro transcription.

[0352] Preparation of circRNA precursor by in vitro transcription: T7 High Yield RNA Synthesis Kit (E2040S, New England Biolabs) was used to perform in vitro transcription (37°C, 2 hours) on the linearized plasmid DNA template to synthesize the cicrRNA precursor. After in vitro transcription, the linear DNA template was digested with DNaseI (M0303S, New England Biolabs) (37°C, 15 min). Column purification was performed using RNA Cleanup Kit (T2050, New England Biolabs).

[0353] circRNA synthesis: The circular RNA precursor molecule was added to a buffer (50 mM Tris-HCl, 10 mM MgCl2, 1 mM DTT, pH 7.5) containing GTP (final concentration of 2 mM), and the solution was heated at 55 °C for 15 min and then column purified ( RNA Cleanup Kit, T2050, New England Biolabs).

[0354] Identification of circular RNA using capillary electrophoresis: The circularized RNA precursor and the crude circularized RNA product were measured for concentration by Nano-Drop (ThermoFisher Scientific), and then processed using QIAxcel RNA High Sensitivity Cartridge Kit (929112) according to the instructions, and then detected and analyzed using QIAxcel Connect capillary electrophoresis instrument according to the operating instructions. This experiment used a high-sensitivity RNA marker (QIAxcel RNA High-Sensitivity Marker Set, 929112, QIAGEN) to indicate different RNA molecular bands.

[0355] Research findings: Figure 5 A shows the local secondary structure of the ribozyme recognition site. Figure 5 B shows the self-splicing reaction of the natural Ana-PIE intron and the modified Ana-PIE intron. Figure 5 As shown in B, there is no obvious difference between the bands in capillary lanes 1 and 2, and lane 2 does not produce obvious bands at the position of circular RNA, indicating that the natural Ana-PIE intron (Natural PIE) cannot effectively undergo self-splicing circularization reaction. On the contrary, compared with lanes 3 and 4, most of the bands in lane 4 are located at the position of circular RNA, with only a small amount of linear RNA bands, and there are obvious bands of intron fragments, which indicates that the modified Ana-PIE intron group (Modified PIE) has undergone a highly efficient self-splicing circularization reaction.

[0356] Example 2. Effect of the selection of splicing sites of self-splicing ribozymes on the cyclization efficiency

[0357] This example tests the effect of different ribozyme splicing sites on the cyclization efficiency.

[0358] To verify the effect of splicing sites at different positions on the cyclization efficiency, in addition to the splicing site determined in Example 1 (located between the 22nd U and 23rd C of the TLSL-CVB3-IRES sequence, located at the end loop of the first stem loop near the 5' end in the TLSL-CVB3-IRES secondary structure, Figure 5 In addition to the position indicated by the arrow in A, this example selected two other splicing sites on TLSL-CVB3-IRES for experimental verification.

[0359] Among them, the first splice site is selected on the stem (specific location such as Figure 6A), correspondingly, the 10th and 11th positions at the 5' end of the Ana-PIE ribozyme were mutated accordingly, and the sequence of the final circular RNA precursor molecule was sequence 11 (SEQ ID NO: 11). The second splice site was selected on the smaller end loop (the end loop contains 3 nucleotides), and the specific position of the splice site is as follows Figure 6 As shown by the arrows in B, correspondingly, the 10th and 11th positions at the 5' end of the Ana-PIE ribozyme are mutated accordingly, and the final circular RNA precursor molecule formed is sequence 12 (SEQ ID NO: 12).

[0360] The experimental results are as follows Figure 6 As shown in Figure C, lanes 0, 1, and 2 are RNA molecule markers, linear precursor molecules of circular RNA, and products of linear precursor molecules of circular RNA after cyclization reaction, respectively. The bands in capillary lane 2 have bands at the position of circular RNA, but the brightness is weak, while the bands at the position of linear RNA are very obvious, proving that only a small amount of circular RNA molecules are formed. Similarly, Figure 6 Only a small amount of bands at the position of circular RNA were observed in lane 2 in D, indicating that the cyclization efficiency of the splicing site was also low. These results collectively show that when the splicing site of the self-splicing ribozyme is on the stem or the end loop of the TLSL cyclization element but the selected end loop is small, cyclization can be achieved, but the cyclization efficiency is significantly reduced. Therefore, the ribozyme splicing site on the TLSL cyclization element is preferably located on the end loop of the stem loop, and the size of the end loop should be greater than or equal to 4.

[0361] Example 3. The Hi-Scarless-PIE cyclization system can produce sequence-accurate circular RNA molecules

[0362] In this embodiment, a variety of methods are used to identify whether the product of the present invention is a target circular RNA molecule, and the identification method includes:

[0363] Method 1. Identification by RNaseR. RNaseR is a 3'-5' exonuclease that can cut and degrade RNA from the 3'-5' direction. It can degrade linear RNA molecules, but circular RNA without 3' and 5' ends cannot be degraded.

[0364] The specific steps are as follows: the purified circular RNA or linear circRNA precursor was heated at 65°C for 3 minutes and then immediately placed on ice for 2 minutes. The linear RNA was then digested with RNaseR (E049, Applied Biological Materials) (37°C, 30 min). RNA digested with RNase R was purified by column purification using RNA Cleanup Kit (T2050, New England Biolabs), and then 1.5% agarose gel was prepared for agarose gel electrophoresis analysis.

[0365] The results are as follows Figure 7 As shown in A, lanes 1 and 3 are linear circRNA precursors (precursor-mRNA, SEQ ID NO: 8) that have not undergone cyclization, lanes 2 and 4 are the products after cyclization, and lanes 3 and 4 are the results after RNaseR treatment. Figure 7 As shown in A, after cyclization treatment, the main band of lane 2 is located below the band of lane 1, and an obvious band appears at the bottom, which indicates that the cyclization reaction produces circular mRNA molecules with reduced molecular weight and intron fragments with very small molecular weight. The RNaseR treatment group shows that no band appears in lane 3, and the main band of lane 4 is not affected by RNaseR, so it is inferred that the band is a band of circular RNA.

[0366] Method 2: Add Poly(A) fragments for identification. Linear mRNA molecules can add hundreds of adenylic acids (i.e. Poly(A) fragments) to increase their molecular weight. However, circular RNA is closed at both ends and has no 3' end, so it cannot be tailed. Therefore, agarose gel electrophoresis can be used to distinguish the size change of RNA before and after tailing to determine whether it is circular RNA.

[0367] The specific steps are as follows: the purified circular RNA or linear circRNA precursor was tailed by using Escherichia coli Poly(A) polymerase (M0276, New England Biolabs) (37°C, 30 min). Column purification was performed using RNA Cleanup Kit (T2050, New England Biolabs).

[0368] The results are as follows Figure 7 As shown in A, the linear mRNA precursor, the product after the cyclization reaction of the linear mRNA precursor, and the product after adding Poly (A) were loaded into lanes 5 and 6, respectively. Compared with lane 1, the band of the linear mRNA precursor in lane 5 shifted upward after the Poly (A) experiment, indicating that the molecular weight increased and Poly (A) was successfully added. Compared with lane 4, the band in lane 6, which was presumed to be circular RNA, was in the same position and size as the band of circular RNA in lane 4 after the Poly (A) experiment, indicating that Poly (A) was not added.

[0369] Method 3. Use RNase H for identification. RNase H is an endoribonuclease that can specifically hydrolyze RNA in a DNA-RNA hybrid chain. Due to the difference in the structure of linear RNA and circular RNA, when the same probe is combined with linear RNA and circular RNA, the linear RNA will be cut into two bands, while the circular RNA will only be cut into a linear or one band.

[0370] The specific steps are as follows: add a 10-fold molar excess of DNA probe to the circular RNA obtained by RNaseR treatment, then heat at 65°C for 5 minutes and anneal at room temperature. Then digest with RNase H (M0279, New England Biolabs) at 37°C for 20 minutes, and column purify the digested RNA. The DNA probe sequence is shown in SEQ ID NO: 13.

[0371] like Figure 7 As shown in B, after RNase H treatment, the linear mRNA precursor formed two bands, while the circular mRNA molecule formed one band.

[0372] Method 4. Verify the accuracy of the ribozyme recognition site sequence by RT-PCR and Sanger sequencing. The aforementioned methods (1), (2) and (3) can confirm that a circular RNA molecule is formed after the cyclization reaction, but the accuracy of the ribozyme recognition site sequence needs to be verified. Therefore, this example also uses sequencing at the splicing site to prove that the cyclization reaction occurs at the expected position and forms the target circular RNA molecule.

[0373] The specific steps are as follows: The circular RNA obtained by RNaseR treatment is synthesized into the first strand of cDNA using the RevertAid First Strand cDNA Synthesis Kit (K1622, Thermo Fisher Scientific) and random primer. Then, two forward and reverse PCR primers are designed upstream and downstream of the splicing site, respectively. Using cDNA as a template, PCR is performed to obtain the amplified fragment and sequence verification is performed. The circular mRNA molecule will detect an RNA sequence that crosses the splicing site in accordance with the expectation near the splicing site. The primer sequences are shown in SEQ ID NO: 14 and SEQ ID NO: 15.

[0374] The results are as follows Figure 7 As shown in Figure C, the sequencing results confirmed that the sequences on both sides of the splicing site were consistent with the designed circular RNA sequence, proving that the head and tail of the circular RNA had been connected together. The circular mRNA molecule with the correct sequence was successfully obtained.

[0375] In summary, this embodiment verifies that the target circular RNA molecule is successfully synthesized in the in vitro cyclization reaction system of the present invention through four complementary methods.

[0376] Example 4. Effects of different ribozyme mutations on the cyclization efficiency of Hi-Scarless-PIE

[0377] In this example, the effects of different ribozyme modification conditions on the cyclization efficiency of Hi-Scarless-PIE were determined by performing corresponding mutations on positions 10-19 of the 5' end of the intron derived from Anabaena. Figure 8 A shows the secondary structure of the native Anabaena intron and the default numbering of nucleotides ( Figure 8 The left figure of A is quoted from the literature Distinct sites of phosphorothioate substitution interfere with folding and splicing of the Anabaena group I intron. Lupták A, Doudna JA. Nucleic Acids Res. 2004 Apr 23; 32(7): 2272-80. doi: 10.1093 / nar / gkh548.), and the enlarged view of the positions of helix P1 and P2 is shown in the left figure. Figure 8 In the right figure of A, the nucleotides in the dotted box are the regions where the natural ribozyme is mutated in this embodiment. The nucleotide numbers of this intron are all used in this article. Figure 8 The marking method shown in A.

[0378] The modified ribozymes used in the aforementioned Examples 1, 2 and 3 only mutated the 10th and 11th positions of the 5' end of the intron from Anabaena. In this example, the 10-14th, 10-17th and 10-19th positions of the 5' end of the intron from Anabaena in the circ-CVB3-Gluc-precurser-2 (SEQ ID NO: 10) sequence were mutated accordingly, and the sequences of the circular RNA precursor molecules formed in sequence are as follows: SEQ ID NO: 16, SEQ ID NO: 17, SEQ ID NO: 18. Then, the reaction was carried out according to the method for synthesizing circular RNA described in Example 1, and then 1.5% agarose gel was prepared for agarose gel electrophoresis analysis.

[0379] Research results: The results are as follows Figure 8As shown in Figure B, lanes 1, 3, and 5 are the circular RNA precursor molecules formed after the 10-14, 10-17, and 10-19 positions of the 5' end of the intron from Anabaena were mutated accordingly, and lanes 2, 4, and 6 are the cyclized products of the precursor molecules after the 10-14, 10-17, and 10-19 positions of the 5' end of the intron from Anabaena were mutated accordingly. Figure 8 As shown in Figure B, after the cyclization treatment, only the main band position of the second swimming lane is located below the band of the first swimming lane, and an obvious band appears at the position of the intron fragment at the bottom of the swimming lane. However, there is no obvious change in the band position of the fourth swimming lane and the third swimming lane, as well as the sixth swimming lane and the fifth swimming lane. This shows that a circular RNA molecule can be formed after the 5'-14th position of the intron from Anabaena is mutated in the circular RNA precursor molecule. And the 10-17th and 10-19th positions of the 5'-end of the intron may affect some core modules of the ribozyme, thereby causing the catalytic activity of the ribozyme to be affected and the circularized RNA cannot be formed. It can be determined that, preferably, in the present invention, some or all of the nucleotides in the region between the 10-16th positions of the 5'-end of the intron from Anabaena are mutated, so that the region forms a partial or complete complementary pairing with the target gene, which can ensure that the self-splicing cyclization efficiency of the modified ribozyme is at a higher level.

[0380] Example 5. Study on the universality of the Hi-Scarless-PIE cyclization system between different “ribozyme-TLSL cyclization modules”

[0381] The aforementioned Examples 1, 2 and 3 all use the nucleotide sequence from positions 1 to 94 from CVB3 IRES as the TLSL cyclization element (TLSL-CVB3-IRES, SEQ ID NO: 2). In order to verify that the "ribozyme-TLSL cyclization module" of this patent is a universal method based on the local secondary structure of the ribozyme recognition site, this example selects multiple TLSL structural elements different from TLSL-CVB3-IRES to verify whether the "ribozyme-TLSL cyclization module" constituted by them can also achieve efficient Scarless circular RNA preparation.

[0382] The TLSL structural elements used in this embodiment include five TLSL structural elements: SEQ ID NO: 19, SEQ ID NO: 20, SEQ ID NO: 21, SEQ ID NO: 22, and SEQ ID NO: 23. The nucleotide sequences of these five TLSLs are different, but all conform to the description of the TLSL structural features in this patent. Among them:

[0383] SEQ ID NO: 19 is derived from HRV-B3 IRES (SEQ ID NO: 24),

[0384] SEQ ID NO: 20 is derived from EV-B107 IRES (SEQ ID NO: 25),

[0385] SEQ ID NO:21 is derived from PV3 IRES (SEQ ID NO:26),

[0386] SEQ ID NO: 22 is derived from SimianV4 IRES (SEQ ID NO: 27),

[0387] SEQ ID NO:23 is derived from BEV1 IRES (SEQ ID NO:28).

[0388] For each TLSL structural element, the 10th and 11th positions of the 5' end of the Ana-PIE ribozyme were mutated accordingly. This example uses the same method as that described in Example 1 to construct the precursor molecule of the cyclic IRES-GLuc as shown below:

[0389] circ-HRV-B3-IRES-GLuc-precursor (SEQ ID NO: 29),

[0390] circ-EV-B107-IRES-GLuc-precursor (SEQ ID NO:30),

[0391] circ-PV3-IRES-GLuc-precursor (SEQ ID NO:31),

[0392] circ-SimianV4-IRES-GLuc-precursor (SEQ ID NO:32),

[0393] circ-BEV1-IRES-GLuc-precursor (SEQ ID NO:33).

[0394] Then, the circular RNA molecular identification method described in Example 3 was used to verify the effect of preparing circular RNA.

[0395] Research findings:

[0396] first, Fig. 9 The experimental results of construct design based on TLSL structure derived from PV3 IRES and verification of cyclization results are shown. Fig. 9A shows the TLSL structure derived from PV3 IRES and the selected splicing site (indicated by the arrow). Fig. 9 As shown in B, lanes 1 and 3 are linear circRNA precursors (precursor-mRNA) that have not undergone cyclization, lanes 2 and 4 are the products after cyclization, and lanes 3 and 4 are the results after RNaseR treatment. Fig. 9 As shown in B, after cyclization treatment, the main band of lane 2 is located below the band of lane 1, and an obvious band appears at the bottom, which indicates that the cyclization reaction produces circular mRNA molecules with reduced molecular weight and intron fragments with very small molecular weight. The RNaseR treatment group shows that no band appears in lane 3, while the band of circular RNA in lane 4 still exists without being affected by RNaseR. Lane 5 is the linear mRNA precursor after RNase H treatment, and lane 6 is the circular mRNA after RNase H treatment. After RNase H treatment, the linear mRNA precursor forms two bands, while the circular mRNA molecule forms one band. Fig. 9 The sequencing results of C confirmed that the sequences on both sides of the splicing site were consistent with the designed circular RNA sequence, proving that the head and tail of the circular RNA had been connected together. The circular mRNA was successfully obtained.

[0397] Similarly, Fig.10 , Fig.11 and Fig.12 The following diagrams respectively show the situation of searching for TLSL structures and selecting splice sites in different IRES. Fig.10 The experimental results of construct design based on TLSL structures derived from HRV-B3 IRES and EV-B107 IRES and validation of cyclization results are shown. Fig.11 The experimental results of construct design based on the SimianV4IRES-derived TLSL structure and validation of the cyclization results are shown. Fig.12 The experimental results of construct design based on the BEV1 IRES-derived TLSL structure and validation of the cyclization results are shown. The results are presented in the same way as Fig. 9 resemblance, Fig.10 , Fig.11 and Fig.12 The results all show that the present invention can successfully and efficiently synthesize circular RNA molecules for different TLSLs, proving that the Hi-Scarless-PIE system in the present invention has universality.

[0398] Example 6. Practical verification of the Hi-Scarless-PIE cyclization system between different “ribozyme-TLSL cyclization modules”

[0399] In order to further verify whether the Hi-Scarless-PIE cyclization system has stability in the cyclization efficiency between different "ribozyme-TLSL cyclization modules", this example quantitatively analyzed the changes in the cyclization efficiency of different "ribozyme-TLSL cyclization modules" by capillary electrophoresis.

[0400] Specific implementation method: After the concentration of the circularized RNA precursor and the circularized RNA crude product was determined by Nano-Drop, they were treated with QIAxcel RNA High Sensitivity Cartridge Kit (929112, QIAGEN) according to the instructions, and then detected and analyzed using QIAxcel Connect capillary electrophoresis instrument (QIAGEN) according to the operating instructions.

[0401] The nucleotide sequences involved in this embodiment are as follows: SEQ ID NO:29, SEQ ID NO:30, SEQ ID NO:31, SEQ ID NO:32 and SEQ ID NO:33.

[0402] Implementation effect: Fig.13 A shows the detection pattern of capillary electrophoresis. Wherein the 2nd, 4th, 6th, 8th, 10th channels are the circular RNA precursor molecules of SEQ ID NO:29, SEQ ID NO:30, SEQ ID NO:31, SEQ ID NO:32 and SEQ ID NO:33, and the 3rd, 5th, 7th, 9th, 11th channels are the products of these circular RNA precursor molecules after cyclization reaction. According to experience, the molecular weight of circular RNA after ringing is smaller than that of its linear precursor molecule, and the corresponding band position moves downward. In addition, the cyclized product contains intron peaks and intron dimer peaks, and may contain circularized RNA dimer peaks. Fig.13 B shows the peak graph of the RNA crude product after the circular RNA precursor (SEQ ID NO: 30) is cyclized. Fig.13 A and Fig.13B, the molecular weight of the circular RNA precursor (SEQ ID NO: 30) decreases after circularization, and the corresponding band position moves down (from 1648nt to 1330nt), and there are 177nt intron peaks and 336nt intron dimer peaks, as well as 2676nt circularized RNA dimer peaks (due to the capillary electrophoresis system itself, the molecular weight marked by the system may have a slight error with the molecular weight of the actual band). Compared with 2, 4, 6, 8, 10, most of the bands in channels 3, 5, 7, 9, and 11 are located at the position of circular RNA, while there are only a few linear RNA bands at the position of linear RNA. The circularization efficiency is calculated using the software (QIAxcel ScreenGel 2.0) that comes with the QIAxcel Connect capillary electrophoresis instrument. Wherein, circularization efficiency = 100%*circularization / (linear+circularization). The results show that the cyclization efficiencies of SEQ ID NO:29, SEQ ID NO:30, SEQ ID NO:31, SEQ ID NO:32 and SEQ ID NO:33 are all above 90%, namely 95.45%, 91.58%, 93.90%, 90.03 and 92.14%, respectively. These results collectively indicate that the Hi-Scarless-PIE system disclosed in the present invention can efficiently cyclize different target sequences (cyclization efficiency is above 90%).

[0403]

[0404] Example 7. Modular design of Hi-Scarless-PIE circular RNA constructs using pre-defined TLSL circularization elements

[0405] Considering that not all RNA molecules contain TLSL structural elements, this embodiment verifies whether the verified TLSL cyclization element can be inserted into the target RNA molecule when there is no TLSL structural element, so that the target RNA molecule can be efficiently cyclized using the Hi-Scarless-PIE system. This embodiment uses mRNA as a case for verification. It is reported that the TLSL structural element can recruit ribosomes in IRES to promote protein translation (Structural basis for cloverleaf RNA-initiated viral genome replication. Keerthi Gottipati, Sean C McNeme, Jerricho Tipo, et al. Nucleic Acids Res. 2023 Sep 8; 51 (16): 8850-8863.). Therefore, this embodiment adopts the method of inserting TLSL cyclization element into IRES. On the one hand, it verifies whether Hi-Scarless-PIE can be used for high-efficiency cyclization after inserting TLSL cyclization element. On the other hand, it verifies whether the protein translation level of circular mRNA can be enhanced after inserting TLSL structural element into IRES.

[0406] Experimental methods:

[0407] In this embodiment, the TLSL structural element (TLSL-CVB3-IRES: SEQ ID NO: 2) was inserted into the 5' end of EMCV IRES (SEQ ID NO: 34) and CrPV IRES (SEQ ID NO: 35), respectively, to form new sequences containing the TLSL structural element, named: EMCV+TLSL (SEQ ID NO: 36) and CrPV+TLSL (SEQ ID NO: 37). Then, the target circular RNA sequence (SEQ ID NO: 38) was formed by EMCV+TLSL (SEQ ID NO: 36) and the mRNA coding region of Gaussia Luciferase (GLuc) (SEQ ID NO: 3), and the target circular RNA sequence (SEQ ID NO: 39) was formed by CrPV+TLSL (SEQ ID NO: 37) and the mRNA coding region of Gaussia Luciferase (GLuc) (SEQ ID NO: 3). The 3' end slice of the Ana-PIE ribozyme (SEQ ID NO: 6), the linear counterpart of the target circular RNA sequence (SEQ ID NO: 38 and SEQ ID NO: 39) and the 5' end slice of the modified Ana-PIE ribozyme (SEQ ID NO: 7) were connected to each other to form the corresponding constructs: circ-EMCV+TLSL-IRES-GLuc-precursor (SEQ ID NO: 40) and circ-CrPV+TLSL-IRES-GLuc-precursor (SEQ ID NO. 41). Then, the circular RNA precursor molecule was synthesized and the circular RNA was prepared in vitro according to the method described in Example 1. Then, 1.5% agarose gel was prepared for agarose gel electrophoresis analysis. At the same time, circular RNA prepared by the classical PIE system (i.e., circular RNA molecules without the TLSL cyclization element inserted) was used as a control (classical PIE-EMCV-IRES-GLuc-precursor: SEQ ID NO: 42 and classical PIE-CrPV-IRES-GLuc-precursor: SEQ ID NO: 43).

[0408] Cell culture: HEK-293T cells were cultured in an incubator at 37°C and 5% CO2 (Thermo Fisher Scientific). HEK-293T cells were cultured in high-glucose DMEM medium (L110KJ, Shanghai Yuanpei Biotechnology Co., Ltd.) containing 10% fetal bovine serum (SH30396.03, Hyclone) and 1% double antibody (S110JV, Shanghai Yuanpei Biotechnology Co., Ltd.). Cells were subcultured every 2-3 days.

[0409] Cell transfection: Before transfection, HEK-293T cells were seeded at 20,000 cells / well in a 96-well plate, and then 50-100 ng (equimolar amount of each RNA) of the circular mRNA molecules prepared according to the method of Example 1 were transferred into HEK-293T cells using Lipofectamine 3000 transfection reagent (L3000001, Thermo Fisher Scientific) according to the instructions.

[0410] Protein expression analysis: 24 hours after transfection, 10 μl of culture supernatant was taken and the luciferase activity in the supernatant was detected using the Pierce Gaussia Luciferase Glow Assay Kit (Thermo Fisher Scientific). The luciferase activity was detected using a SpectraMax i3 microplate reader (Molecular Devices).

[0411] Experimental results:

[0412] Fig.14 A shows the secondary structures of TLSL, EMCV IRES and EMCV+TLSL. Fig.15 A shows the secondary structures of TLSL, CrPVIRES and CrPV+TLSL. From the secondary structure, it can be seen that the sequence after TLSL insertion forms the TLSL structural element at the expected position, and this structural element has no effect on other secondary structural regions on IRES. Fig.14 B and Fig.15 As shown in the agarose gel results of B, the first lane is the linear circRNA precursor (precursor-mRNA) that has not undergone cyclization reaction, and the second lane is the product after cyclization. After cyclization treatment, the main band position of the second lane is significantly moved down, and a clear band appears at the bottom, which indicates that the cyclization reaction produces circular mRNA molecules with reduced molecular weight and intron fragments with very small molecular weight. This shows that after inserting the TLSL cyclization element, Hi-Scarless-PIE can be used for efficient cyclization. Fig.14 C and Fig.15 As shown in the protein expression results of C, the activity value of luciferase expressed by the circular RNA molecule inserted with the preset TLSL cyclization element synthesized by the Hi-Scarless-PIE cyclization method of the present invention is significantly higher than that of the circular RNA molecule without the TLSL cyclization element (synthesized using the classic PIE method).

[0413] According to the research results of this embodiment, it can be reasonably inferred that: this patent can be used on any IRES (natural or modified, such as inserting a preset TLSL structural element to modify the natural IRES) to perform modular "ribozyme-TLSL cyclization element" design, thereby achieving efficient circular mRNA preparation, and at the same time, the protein translation level of IRES is significantly improved after the insertion of the TLSL cyclization element. These results show that the modular design concept of the "ribozyme-TLSL cyclization element" of this patent can achieve the "killing two birds with one stone" effect of high-efficiency cyclization preparation and improving protein translation level in the preparation and application of circular mRNA.

[0414] Example 8. TLSL structural elements are abundant in natural and non-natural RNA molecules

[0415] This example comprehensively analyzes the distribution of TLSL structural elements in various RNA molecules. In this example, only the classic clover-shaped TLSL is used as a typical case for analysis.

[0416] Implementation method:

[0417] Search for TLSL elements in RNA secondary structure. First, the RNAfold algorithm is used to predict the secondary structure of the RNA sequence, and the prediction results are presented using dot-bracket notation. RNAfold predicts the corresponding secondary structure for each RNA sequence. Then, the TLSL search program built in this article is used to search for each RNA sequence in the secondary structure file to see if it contains a TLSL structure. The matching principle of the TLSL search follows the clover TLSL structural features. In this embodiment, in addition to the basic features satisfying the structural description of the classic clover TLSL, the total length of TLSL is limited to no more than 200nt.

[0418] Distribution analysis of TLSL structural elements in various major RNA sequences. IRES, 5'-UTR, 3'-UTR, mRNA coding region, non-coding RNA and non-natural RNA sequences are RNA types closely related to this patent. Therefore, for the above-mentioned types of RNA molecules, first search and download representative sequences in public databases, then perform secondary structure prediction and TLSL element search on the sequences, and finally use a bar graph to display the proportion of TLSL elements in various RNA sequences. The data collection and preprocessing process is as follows:

[0419] The IRES sequence is derived from the IRESbase database (IRESbase: A Comprehensive Database of Experimentally Validated Internal Ribosome Entry Sites. Zhao J, Li Y, Wang C, et al. Genomics Proteomics Bioinformatics. 2020Apr; 18(2): 129-139.), which contains a total of 1,328 IRES sequences from humans, viruses, and other eukaryotic organisms. The dataset also contains a portion of artificially designed IRES sequences with a length of 174 nt, which is named "IRES-IRESbase". Whether the IRES sequences from the IRESbase database can efficiently initiate protein translation in mRNA has not been experimentally verified by consistency. Therefore, this experiment further collected a set of IRES sequences that have been verified by consistency experiments from the literature. Specifically, 51 experimentally verified IRES sequences that can efficiently initiate mRNA translation were extracted from an experimental paper published in Nature Biotechnology in 2023 (Engineering circular RNA for enhanced protein production. Chen R, Wang SK, Belk JA, et al. Nat Biotechnol. 2023 Feb; 41(2): 262-272.). The data set of these 51 sequences was named "IRES-High Activity".

[0420] The 5'-UTR and 3'-UTR sequences were derived from the UTRdb (UTRdb 2.0: a comprehensive, expert-curated catalog of eukaryotic mRNAs untranslated regions. Lo Giudice C, Zambelli F, Chiara M, et al. Nucleic Acids Res. 2023 Jan 6; 51(D1): D337-D344.) database, which currently contains more than 12.9 million 5'-UTR sequences and more than 13.19 million 3'-UTR sequences from 673 species. In order to save computing time, this study used human UTR sequences as a typical case for TLSL analysis, of which UTRdb contained a total of 93,457 5'-UTR sequences and 87,932 3'-UTR sequences. The criteria for inclusion in the analysis were: 1) sequence length greater than 150 nt, 2) only the longest sequence from the same gene was retained. Finally, a total of 12,661 5'-UTR sequences and 18,414 3'-UTR sequences were included in this analysis.

[0421] The human mRNA coding region sequence is derived from the Consensus Coding Sequence (CCDS) database (Consensus coding sequence (CCDS) database: a standardized set of human and mouse protein-coding regions supported by expert curation. Pujar S, O'Leary NA, Farrell CM, et al. Nucleic Acids Res. 2018 Jan 4; 46 (D1): D221-D228.). Specifically, the file containing the CDS nucleotide sequence (file name "CCDS_nucleotide.current.fna.gz") was downloaded from the official website of the CCDS database. This version of the CCDS database contains a total of 35,624 human CDS sequences, all of which were included in the analysis.

[0422] Non-coding RNA (ncRNA) comes from the RNAChallenge benchmark dataset, which is a benchmark dataset for evaluating the classification performance of ncRNA classification tools. It contains 11,040 ncRNA sequences from multiple species (A large-scale benchmark study of tools for the classification of protein-coding and non-coding RNAs. Dalwinder Singh and Joy Roy. Nucleic Acids Research, gkac1092, 2022). The download address of the dataset is https: / / github.com / cbl-nabi / RNAChallenge. All 11,040 ncRNA sequences were included in this analysis.

[0423] Research findings:

[0424] The results are shown in the following table and Fig.16 As shown in A, in the "IRES-IRESbase" sequence set, a total of 194 clover TLSL structural elements were matched in 188 IRES sequences, accounting for 14.16% of all IRES in this analysis. In the "IRES-high activity" sequence set, a total of 46 clover TLSL structural elements were matched in 37 IRES sequences, accounting for 72.55% of all high-activity IRES in this analysis. In the human 5'-UTR sequence set, a total of 2442 clover TLSL structural elements were matched in 2327 5'-UTR sequences, accounting for 18.38% of all 5'-UTR sequences in this analysis. In the human 3'-UTR sequence set, a total of 9766 clover TLSL structural elements were matched in 6508 3'-UTR sequences, accounting for 35.34% of all 3'-UTR sequences in this analysis. In the human CDS sequence set, a total of 28,526 cloverleaf TLSL structural elements were matched in 17,896 CDS sequences, accounting for 50.24% of all CDS sequences analyzed in this study. In the ncRNA sequence set, a total of 14,104 cloverleaf TLSL structural elements were matched in 6,706 ncRNA sequences, accounting for 60.74% of all ncRNA sequences analyzed in this study.

[0425]

[0426]

[0427] In summary, using the typical cloverleaf TLSL structural features for matching, in the IRES-IRESbase, IRES-high activity, human 5'-UTR, human 3'-UTR, human CDS, and non-coding RNA sequence sets analyzed in this example, 14.16%, 72.55%, 18.38%, 35.34%, 50.24%, and 60.74% of the sequences contained typical cloverleaf TLSL structural elements, respectively ( Fig.16 A). Fig.16 Some representative trefoil TLSL structural elements from IRES, human 5'-UTR, human 3'-UTR, human CDS, and non-coding RNA sequence sets are shown in CG. Fig.16 The first line of the annotation next to the TLSL structure in C is the ID number of the IRES from which the TLSL originates, where the ID number uses the number of the IRESbase database, and the three numbers in the second line of the annotation are the starting point, the end point and the length of the TLSL in the original IRES. Fig.16 The annotation next to the TLSL structure in D is the name of the gene from which the TLSL is derived. Fig.16 The annotation next to the TLSL structure in E is the name of the gene from which the TLSL is derived. Fig.16 The first line of the annotation content next to the TLSL structure in F is the ID number of the CDS sequence from which the TLSL comes, where the ID number uses the number of the CCDS database, and the three numbers in the second line of the annotation content are the starting point, end point and length of the TLSL in the original CDS sequence. Fig.16 The first line of the annotation content next to the TLSL structure in G is the ID number of the ncRNA sequence from which the TLSL is derived, where the ID number uses the sequence ID number of the RNAChallenge dataset. The three numbers in the second line of the annotation content are the starting point, ending point and length of the TLSL in the original ncRNA sequence.

[0428] In summary, this example demonstrates that TLSL structural elements are widely present in RNA molecules such as IRES, UTR, CDS and ncRNA, which provides flexibility and broad space for the design of the "ribozyme-TLSL cyclization module" in the Hi-Scarless-PIE cyclization strategy of the present invention, and the cyclization platform of this patent has good versatility. In theory, it can be reasonably inferred from Example 5 and Example 6 that the method of this patent can prepare any RNA sequence containing TLSL structural elements by high-efficiency cyclization; at the same time, it can be reasonably inferred from Example 7 that the method of this patent can prepare any mRNA by high-efficiency cyclization using natural or modified IRES proposed in this patent as a translation initiation sequence.

[0429] Example 9. Circular RNA prepared by Hi-Scarless-PIE cyclization system has low immunogenicity

[0430] In this example, the expression of immune-related factors induced by the circular mRNA prepared by the Anabaena PIE system and Hi-Scarless-PIE in the present disclosure in A549 cells was detected.

[0431] The specific steps are as follows:

[0432] HPLC purification: The RNaseR-treated circularized RNA product was heated at 65°C for 5 minutes and then placed on ice for 2 minutes. HPLC purification conditions: silica-based aqueous gel column (Dikma Protein SEC-1000 300x8.0mm, 5μm, 99305, Dima Technology), flow rate: 0.5ml / min, nuclease-free buffer: 10mM Tris, 1mM EDTA, pH=6, HPLC instrument model: Shimadzu LC-20AD XR high-performance liquid chromatograph, using a UV detector at 260nm to detect RNA signals and collect RNA. The collected RNA was column purified.

[0433] Cell culture: A549 cells (cell bank of Chinese Academy of Sciences) were cultured in an incubator (Thermo Fisher Scientific) at 37°C and 5% CO2. A549 cells were cultured in RPMI-1640 medium (L210KJ, Shanghai Yuanpei Biotechnology Co., Ltd.) containing 10% fetal bovine serum (SH30396.03, Hyclone) and 1% double antibody (S110JV, Shanghai Yuanpei Biotechnology Co., Ltd.). Cells were subcultured every 2-3 days.

[0434] Immunogenicity detection: A549 cells were inoculated in 12-well plates at 100,000 cells / well, and then according to the instructions, after the cells reached 70-90% confluence, 1000ng / well RNaseR-treated circular RNA or further HPLC-purified circular RNA and 3p-hpRNA (tlrl-hprna, InvivoGen) were transferred into 12-well plates using Lipofectamine 3000 transfection reagent (L3000001, Thermo Fisher Scientific). After 16 hours of transfection, cells were lysed using Trizol and RNA was extracted. The first strand of cDNA was then synthesized using RevertAid First Strand cDNA Synthesis Kit (K1622, Thermo Fisher Scientific). The expression of four immune response proteins, CCL5, IL6, INF-1B and RIG-1, was then detected by fluorescent quantitative PCR.

[0435] The sequence used for quantitative PCR is as follows:

[0436] GAPDH-F:GTCTCCTCTGACTTCAACAGCG(SEQ ID NO:44)

[0437] GAPDH-R:ACCACCCTGTTGCTGTAGCCAA(SEQ ID NO:45)

[0438] IL6-F:AGACAGCCACTCACCTCTTCAG(SEQ ID NO:46)

[0439] IL6-R:TTCTGCCAGTGCCTCTTTGCTG (SEQ ID NO:47)

[0440] RIG-1-F:CACCTCAGTTGCTGATGAAGGC(SEQ ID NO:48)

[0441] RIG-1-R:GTCAGAAGGAAGCACTTGCTACC(SEQ ID NO:49)

[0442] CCL5-F:CCTGCTGCTTTGCCTACATTGC (SEQ ID NO:50)

[0443] CCL5-R:ACACACTTGGCGGTTCTTTCGG(SEQ ID NO:51)

[0444] IFNB1-F:CTTGGATTCCTACAAAGAAGCAGC(SEQ ID NO:52)

[0445] IFNB1-R:TCCCTCTTTCTGGAACTGCTGCA(SEQ ID NO:53)

[0446] Research findings:

[0447] Fig.17 The expression of immunogenic cytokines induced by circ-GLuc-mRNA prepared by the classical PIE and Hi-Scarless-PIE methods in A549 cells is shown. Among them, the circular RNA precursor molecule (SEQ ID NO: 54) prepared by the classical PIE method was used as a control for comparison of immunogenicity, and the precursor molecule of circular RNA prepared by the Hi-Scarless-PIE method used the aforementioned SEQ ID NO: 8. 3p-hpRNA is a triphosphate hairpin RNA, which is a characteristic structure recognized by RIG-I and is an effective and specific agonist of RIG-I. In this experiment, 3p-hpRNA was used as a positive control that can show high immunogenicity in A549. In the experimental groups, the PIE unpurified group was a sample of the cyclization product of the classical PIE without HPLC purification, the PIE purified group was a sample of the cyclization product of the classical PIE after HPLC purification, the Ym-11unpurified group was a sample of the cyclization product of the Hi-Scarless-PIE without HPLC purification, and the Ym-11purified group was a sample of the cyclization product of the Hi-Scarless-PIE after HPLC purification. The experiment compared the difference in immunogenicity of circular RNA synthesized by the classical PIE method and the Hi-Scarless PIE method. At the same time, the difference in immunogenicity of circular RNA purified by HPLC and not purified by HPLC was also compared.

[0448] The results are as follows Fig.17As shown, compared with the positive control, circ-GLuc-mRNA prepared by the classical PIE and Hi-Scarless-PIE methods showed lower expression levels in the four cytokines regardless of whether they were purified by HPLC. Compared with the group without HPLC purification, the samples purified by HPLC showed lower IL6, INF-B1 and RIG-1 expression, and this trend was consistent in the samples prepared by classical PIE and Hi-Scarless-PIE. At the same time, this experiment also observed that the samples prepared by classical PIE did not show a decrease in CCL5 expression levels after purification, while the samples prepared by Hi-Scarless-PIE showed a significant decrease in CCL5 expression levels after purification. In the sample group after HPLC purification, the samples prepared by Hi-Scarless-PIE showed lower expression of CCL5, IL6 and INF-B1 than the classical PIE method.

[0449] Based on the results of this experiment, the circular RNA prepared by the Hi-Scarless-PIE method showed immunogenicity comparable to or lower than that of classical PIE in cells.

[0450] Example 10. Circular mRNAs containing different IRES sequences prepared by the Hi-Scarless-PIE cyclization system can all express proteins in vitro

[0451] In this example, the protein expression levels of circular GLuc mRNAs with different IRES sequences prepared by Hi-Scarless-PIE in the present disclosure were detected in HEK-293T and A549 cells.

[0452] The nucleotide sequences involved in this embodiment are as follows: SEQ ID NO:8, SEQ ID NO:29, SEQ ID NO:30, SEQ ID NO:31, SEQ ID NO:32, SEQ ID NO:33, SEQ ID NO:55 and SEQ ID NO:56.

[0453] Implementation method:

[0454] Cell transfection: Before transfection, HEK-293T or A549 cells were seeded at 20,000 cells / well in a 96-well plate, and then 50-100 ng (equimolar amount of each RNA) of the circular mRNA molecules prepared according to the method of Example 1 were transferred into HEK-293T and A549 cells respectively using Lipofectamine 3000 transfection reagent (L3000001, Thermo Fisher Scientific) according to the instructions.

[0455] Protein expression analysis: 24 hours after transfection, 10 μl of culture supernatant was taken and the luciferase activity in the supernatant was detected using the Pierce Gaussia Luciferase Glow Assay Kit (Thermo Fisher Scientific). The luciferase activity was detected using a SpectraMax i3 microplate reader (Molecular Devices).

[0456] Experimental results:

[0457] Fig.18 A and Fig.18 B shows the protein expression levels of circular GLuc mRNA with different IRES sequences in HEK-293T and A549 cells. Taking CVB3 IRES as a reference, it can be found that the protein expression levels of the same IRES in different cells are significantly different. For example, the translation level of GLuc in A549 cells is very low for PV3 and CVA20 IRES, but it is significantly increased in HEK-293T cells, reaching about half the level of CVB3 IRES. In A549 cells, the protein expression levels of CVB3 IRES and EV-B107IRES are basically the same, but the protein translation level of EV-B107 IRES in HEK-293T cells is reduced to about 15% of CVB3 IRES. These results show that the protein translation level of IRES is cell type selective and that different IRES have different cell selectivity tendencies. This study demonstrates that circular mRNA molecules with different IRES sequences prepared by the Hi-Scarless-PIE system disclosed in the present invention can express proteins at the cellular level, and the same IRES may express different levels of proteins in different cells, and different IRES may express different levels of proteins in the same cell. These results suggest that the protein translation initiation activity of different IRESs is cell type selective.

[0458] Consistent with our findings, other research teams have also reported that the protein translation level of IRES is cell and tissue selective. For example, at the 25th Annual Meeting of the American Society of Gene and Cell Therapy (ASGCT) in 2022, Alexander Wesselhoef reported a high-throughput screening study (Discovery of Translation Initiation Elements Enabled by a Parallel Arrayed Screen of Full-length Viral UTRs in Synthetic Circular RNA, May 16, 2022), which found that the activity of IRES depends on the cell type, with different protein translation levels in different cell types.

[0459] Based on this study and the fact that IRES activity is cell and tissue selective, combined with the platform advantages of the versatility and practicality of the Hi-Scarless-PIE cyclization system demonstrated in Examples 5, 6, 7 and 8, this patent can be used to design circular mRNA products for the preparation of a wide range of medical uses based on the Hi-Scarless-PIE platform. The basic principle is as follows: Fig.18 As shown in C: The cap-independent translation initiation element (such as IRES) is used as a non-coding region. This patent can design a "ribozyme-TLSL cyclization module" in this region to achieve high-efficiency and high-fidelity circular mRNA preparation, based on the diverse tissue-selective protein translation initiation activity of the IRES library, and the flexible and changeable protein encoding ability of the mRNA coding region, for the corresponding diseased tissues of a variety of diseases, to achieve the design of diseased tissue-specific circular IRES-mRNA drugs, so as to be used for the treatment of a wide range of diseases. Under this design concept, the IRES region simultaneously assumes the functions of high-efficiency cyclization preparation and tissue-specific selection of protein expression, and the sequence of the mRNA coding region will not affect the efficiency of cyclization preparation. Therefore, this design concept can achieve high-efficiency and high-fidelity cyclization preparation and tissue-selective expression of mRNA therapy for any mRNA sequence.

[0460] In summary, the Hi-Scarless-PIE invented in this patent is a universal circular mRNA preparation platform that can be used for a wide range of biomedical applications.

[0461] Example 11. Detection of protein expression levels of cyclic hEPO and EGFP

[0462] In this example, the protein expression of cyclic hEPO and cyclic EGFP mRNA prepared by the Hi-Scarless-PIE method was detected.

[0463] The nucleotide sequence involved in this embodiment is as follows:

[0464] EGFP-precursor: SEQ ID NO: 57

[0465] hEPO-precursor: SEQ ID NO: 58

[0466] Implementation method: Before transfection, HEK-293T cells were inoculated in a 96-well plate at 20,000 cells / well, and then 100 ng of the cyclic EGFP, hEPO and GLuc mRNA molecules of the Hi-Scarless-PIE system of the present invention prepared according to the method of Example 1 were transferred into HEK-293T cells using Lipofectamine 3000 transfection reagent (L3000001, Thermo Fisher Scientific) according to the instructions. After 24 hours, an inverted fluorescence microscope (OLYMPUS Olympus inverted CKX53 microscope) was used to observe the expression of green fluorescence. For HEK-293T cells transfected with circ-hEPO mRNA, 10 microliters of culture supernatant was taken after 24 hours, and the concentration of hEPO protein in the supernatant was detected using the human Erythropoietin ELISA kit (hEPO) (ab274397, Abcam) according to the instructions. ELISA detection uses a SpectraMax i3 microplate reader (Molecular Devices).

[0467] Experimental results: Fig.19 The protein expression of cyclic hEPO and cyclic EGFP mRNA prepared by Hi-Scarless-PIE method is shown. Fig.19 A is the fluorescence detection result of circ-EGFP protein expression level. Here, circ-hEPO has no fluorescence and is used as a negative control for EGFP fluorescence detection. The results show that circ-hEPO has no specific green fluorescence, while circ-EGFP shows obvious specific green fluorescence. Fig.19B is the protein expression detection result of circ-hEPO. Gluc here is used as a negative control. The results show that there is no hEPO detection signal in the Gluc group, while the hEPO protein concentration detected in the hEPO group reaches about 2000mIU / ml. This example proves that the circular RNA prepared by the Hi-Scarless-PIE method can effectively express EGFP and hEPO genes.

[0468] Example 12. Dynamic trends of protein expression levels of circular mRNA prepared by Hi-Scarless-PIE cyclization system in vitro

[0469] This example compares the protein expression levels of capped and tailed linear RNA, circular RNA prepared by the classic PIE system, and circular mRNA prepared by the Hi-Scarless-PIE in HEK-293T and A549 cells over time. The sequences involved are as follows:

[0470] Capped and tailed linear mRNA (SEQ ID NO: 59)

[0471] PIE Circular (SEQ ID NO: 54)

[0472] Circular mRNA precursor molecule SEQ-01Circular (SEQ ID NO: 8) prepared by Hi-Scarless-PIE

[0473] Implementation method:

[0474] Linear mRNA synthesis and purification: The 5'-UTR region sequence (SEQ ID NO: 60) and 3'-UTR region sequence (SEQ ID NO: 61) from the human beta-globin (HBB) gene were used in the linear mRNA. The linear mRNA obtained according to the in vitro transcription method in Example 1 was heated at 65°C for 5 minutes, and then immediately ice-bathed for 5 minutes. Then, the cap structure was added to the transcription product using vaccinia virus capping enzyme (M2080, New England Biolabs) and mRNA cap-2'-O-methyltransferase (M0366, New England Biolabs) according to the operating instructions of New England Biolabs. Then, the Poly (A) tail was added to the capped transcription product using Escherichia coli Poly (A) polymerase (M0276, New England Biolabs). The fully processed product was then purified by The RNA was purified using RNA Cleanup Kit (T2050, New England Biolabs).

[0475] Cell transfection: Before transfection, HEK-293T or A549 cells were seeded at 20,000 cells / well in a 96-well plate, and then 50-100 ng (equimolar amount of each RNA) of the circular mRNA or capped and tailed linear mRNA molecules of the Anabaena PIE system and the Hi-Scarless-PIE system in the present disclosure prepared according to the method of Example 1 were transferred into HEK-293T and A549 cells respectively using Lipofectamine 3000 transfection reagent (L3000001, Thermo Fisher Scientific) according to the instructions.

[0476] Protein expression analysis: For HEK-293T cells, 10 μl of culture supernatant was taken at 24 hours (Day 1), 48 hours (Day 2), 72 hours (Day 3), 96 hours (Day 4), 120 hours (Day 5), 144 hours (Day 6), and 168 hours (Day 7) after transfection, and the luciferase activity in the supernatant was detected using the Pierce Gaussia Luciferase Glow Assay Kit (Thermo Fisher Scientific). For A549 cells, 10 μl of culture supernatant was taken at 24 hours (Day 1), 48 hours (Day 2), and 72 hours (Day 3) after transfection, and the luciferase activity in the supernatant was detected using the Pierce Gaussia Luciferase Glow Assay Kit (Thermo Fisher Scientific). The luciferase activity was detected using a SpectraMax i3 microplate reader (Molecular Devices). After each day of testing, the culture medium was completely removed and replaced with new culture medium in the culture plate.

[0477] Research findings:

[0478] Fig. 20 The figure shows the trend of protein expression levels in HEK-293T and A549 cells over time using traditional capped and tailed linear mRNA, the classic PIE system, and the circular mRNA prepared by Hi-Scarless-PIE in the present disclosure. Fig. 20 As shown, Linear is the traditional capped and tailed linear mRNA, PIE Circular is the circular RNA prepared by the classic Anabaena PIE system, and SEQ-01 Circular is the circular mRNA prepared by Hi-Scarless-PIE in the present disclosure. Fig. 20The results showed that in HEK-293T cells, SEQ-01Circular and PIE Circular were the same, both of which could continuously and stably express Gaussia Luciferase protein at a high level, and the half-life of protein expression was 4-5 days, which was significantly better than that of the linear mRNA group. In A549 cells, SEQ-01Circular and PIE Circular had the same protein expression trend, and both were better than linear mRNA, with a half-life of 2 days.

[0479] These results indicate that, similar to the classical PIE system, the circular mRNA prepared by Hi-Scarless-PIE in the present invention has prolonged stability, a long in vivo half-life, and significantly increased protein expression efficiency.

[0480] Example 13. Circular mRNA prepared by Hi-Scarless-PIE cyclization system expresses protein in mice

[0481] In this example, the protein expression of circular mRNA prepared by the Anabaena PIE system and Hi-Scarless-PIE in the present disclosure in mice was detected.

[0482] Implementation method: According to the method described in Example 1, circular mRNA: CVB3-FLUC (SEQ ID NO: 62) and its circular RNA precursor molecules are: SEQ ID NO: 63 and SEQ ID NO: 58, respectively. At the same time, the circular RNA molecule (SEQ ID NO: 64) prepared according to the classical PIE method (its circular RNA precursor molecule is: SEQ ID NO: 65)) was used as a control group. In addition, this example also prepared CVB3-hEPO (SEQ ID NO: 66) and its circular RNA molecule precursor (SEQ ID NO: 58), as well as the circular RNA precursor molecule (SEQ ID NO: 67) used in the classical PIE preparation scheme. The prepared circular mRNA was purified according to the method described in Example 1.

[0483] Preparation of circular mRNA-LNP: The purified product was used to prepare circular mRNA-lipid nanoparticle complexes. The specific steps are as follows: DLin-MC3-DMA: DSPC: cholesterol: DMG-PEG2000 (Aiweituo (Shanghai) Pharmaceutical Technology Co., Ltd.) was dissolved in an ethanol solution at a lipid molar ratio of 50:10:38.5:1.5, and the circular RNA was dissolved in a 10mM sodium citrate buffer with a pH of 4.5. The two phases were quickly mixed using a microfluidic device (INano L / L+, Myanna (Shanghai) Instrument Technology Co., Ltd.), with a total flow rate of 12ml / min, and a flow rate ratio of water buffer (circular RNA) to ethanol lipid mixture = 3:1. After preparation, ultrafiltration was used to remove ethanol and replace it with PBS solution. The particle size, polydispersity index (PDI), and Zeta potential of the circular RNA-plasmid nanocomplex were then detected by dynamic light scattering (DLS) using a laser particle size analyzer (Zetasizer Nano-ZSP, Malvern, UK), and the content and encapsulation efficiency of circular RNA in the complex were detected using Quant-iT Ribogreen (Thermo Fisher Scientific).

[0484] In vivo administration: The prepared circ-FLuc-mRNA-LNP lipid nanoparticles were injected into the muscle of the right hind leg of Balb / c mice of about 8 weeks old at a dose of 3 μg per mouse. The fluorescence intensity of the whole body of the mice was detected on days 1, 3, 7, 14, and 21 after injection. The prepared circ-FLuc-mRNA-LNP lipid nanoparticles were intravenously injected into Balb / c mice of about 8 weeks old at a dose of 8 μg per mouse. The fluorescence intensity of the whole body of the mice was detected on days 1, 3, and 7 after injection. D-Luciferin potassium salt (D-luciferin potassium salt) (ST196-25mg, Biyuntian) was injected into the tail vein before detection at a dose of 10 μl / g body weight. After 10-20 minutes after injection, the fluorescence signal reached the strongest stable plateau, and then the fluorescence intensity was detected using a small animal fluorescence in vivo imaging system (PerkinElmer, IVIS Spectrum). The prepared circ-hEPO-mRNA-LNP lipid nanoparticles were intravenously injected into 8-week-old Balb / c mice at a dose of 8 μg per mouse. Blood was collected at 6h, 12h, 24h, 48h, 72h and 96h after injection, and the concentration of hEPO protein in plasma was detected using the human Erythropoietin ELISA kit (hEPO) (ab274397, Abcam).

[0485] Research findings:

[0486] Fig.21The preparation and characterization of LNP lipid nanoparticle encapsulation by circular FLuc-mRNA prepared by the classical PIE method and the Hi-Scarless-PIE method are shown. Fig.21 A) and Hi-Scarless-PIE method ( Fig.21 B) Characterization results of the prepared circ-GLuc mRNA encapsulated by LNP are shown in the figure. The results of this experiment prove that the circular RNA molecules prepared by the Hi-Scarless-PIE method have relatively consistent characteristics when preparing LNP encapsulation mixtures.

[0487] Fig. 22 The protein expression of circular FLuc-mRNA-LNP lipid nanoparticles prepared using the classical PIE method and the Hi-Scarless-PIE method in mice is shown. Fig. 22 A and 22B show the expression of FLuc protein after intramuscular injection of circular FLuc-mRNA-LNP and the statistical results of the expression amount. Compared with the control group, the FLuc protein can be significantly expressed after injection of circular FLuc-mRNA-LNP and the high expression of FLuc protein can last for at least one week. Fig. 22 C shows the expression of FLuc protein after intravenous injection of circular FLuc-mRNA-LNP. Compared with the control group, obvious FLuc protein expression can be detected one day after intravenous injection of circular FLuc-mRNA-LNP.

[0488] Fig. 22 D shows the expression of hEPO in mouse plasma after intravenous injection of circ-hEPO-mRNA-LNP lipid nanoparticles. The concentration of hEPO was at a high level 6 hours, 12 hours, and 1 day after injection, and decreased significantly after 2 days after injection.

[0489] The results of this experiment demonstrated that, similar to circular RNA molecules prepared using classic PIE, circular RNA molecules prepared using the Hi-Scarless-PIE method can efficiently express target proteins in mice.

[0490] Example 14. Design and verification of Uni-cRNA circular RNA construct

[0491] In this example, a single stem-loop (SL) cyclization element was determined, and the Ana-PIE intron ribozyme was modified accordingly according to the sequence near the splicing site on the SL. Further, the modified ribozyme and SL formed a "ribozyme-SL cyclization module", and finally formed a process and method of a Uni-cRNA circular RNA construct, and the effect of the construct in synthesizing circular RNA was experimentally verified.

[0492] The design principle of the Uni-cRNA circular RNA construct in the present invention is as follows Fig.23 shown.

[0493] The SL cyclization element consists of a stem-loop structure and a ribozyme recognition site located on the terminal loop.

[0494] In this example, a variety of situations were tried, including searching for Stem-loop structures and selecting SL structural elements in the IRES region and CDS (mRNA coding region) region, searching for SL structural elements in natural sequences and algorithm-optimized non-natural sequences. This example also demonstrated two situations where the splicing site is located in the terminal loop in the SL and in the internal loop. This example constructed circular RNA constructs for the above situations respectively, and tested the efficiency of the constructs in preparing circular RNA.

[0495] Searching for SL cyclization elements in the IRES region: First, the SL cyclization element was obtained by the following steps: In CVB3 IRES (SEQ ID NO. 68), its secondary structure is as follows Fig.24 As shown in A, the schematic diagram of the secondary structure of CVB3 IRES is cited from the literature (Nat Biotechnol. 2023; 41 (2): 262-272). According to the secondary structure of CVB3 IRES, the nucleotide sequence of the SL cyclization element located at positions 183 to 232 of CVB3 IRES was selected and named Stem-loop-CVB3-IRES (SEQ ID NO. 69). Its secondary structure is as follows Fig.24 B. Next, the splicing site is selected: (1) The splicing site is located between the 24th U and the 25th C of the Stem-loop-CVB3-IRES sequence. Fig.24The arrow in B shows the location of the splice site. In this example, the target circular RNA sequence (SEQ ID NO.71) is composed of CVB3-IRES (SEQ ID NO.68) and the mRNA coding region of Gaussia Luciferase (GLuc) (SEQ ID NO.70). According to the selected splice site, the Seq 4 sequence is separated from the splice site to form a linear counterpart of the target circular RNA sequence (SEQ ID NO.72).

[0496] Design of a GLuc circular RNA precursor molecule based on the modified Ana-PIE ribozyme: The 3' end slice of the Ana-PIE ribozyme (SEQ ID NO.73), the linear counterpart of the target circular RNA sequence (SEQ ID NO.72) and the 5' end slice of the modified Ana-PIE ribozyme (SEQ ID NO.74) were connected to each other to form a GLuc circular RNA precursor molecule based on the modified Ana-PIE ribozyme, named circ-CVB3-IRES-GLuc-precursor (SEQ ID NO.75).

[0497] Synthesis of the circular RNA precursor molecule and construct designed above: First, DNA sequences corresponding to SEQ ID NO.68, SEQ ID NO.70, SEQ ID NO.73 and the 5' end section (SEQ ID NO.76) of the natural Ana-PIE ribozyme were synthesized (Beijing Qingke Biotechnology Co., Ltd.). These DNA sequences were cloned into the linear plasmid vector pUC57 containing the T7 promoter generated by PCR by molecular cloning (GenBuilder Plus Cloning Kit, L00744, GenScript Biotechnology Co., Ltd.) to obtain the corresponding construct (i.e., the construct formed using the natural Ana-PIE intron). On the basis of this construct, the sequence of the 5' end section (SEQ ID NO:76) of the natural Ana-PIE ribozyme was transformed, wherein the A at the 10th position of the 5' end of SEQ ID NO:76 was mutated to G, and the G at the 11th position was mutated to C, forming the transformed Ana-PIE ribozyme intron construct.

[0498] Preparation of linear plasmid template: Transform the plasmid into Top10 competent cells, plate, and select clones the next day to expand the culture in LB medium containing Amp resistance (37°C / 200rpm / overnight). Then perform plasmid extraction (endotoxin-free plasmid mini-extraction kit, DP118, Tiangen Biochemical Technology (Beijing) Co., Ltd.). Then use EcoRI (1040, Takara) single enzyme digestion method to prepare linearized plasmid template (37°C, 1 hour). Use The cleavage products were recovered by PCR&DNA Cleanup Kit (T1030, New England Biolabs), the concentration was determined by Nano-Drop (Thermo), and the cleavage products were identified by 1.5% agarose gel electrophoresis. The purified linear plasmid was used as a template for in vitro transcription.

[0499] Preparation of circRNA precursor by in vitro transcription: T7 High Yield RNA Synthesis Kit (E2040S, New England Biolabs) was used to perform in vitro transcription (37°C, 2 hours) on the linearized plasmid DNA template to synthesize the cicrRNA precursor. After in vitro transcription, the linear DNA template was digested with DNaseI (M0303S, New England Biolabs) (37°C, 15 min). Column purification was performed using RNA Cleanup Kit (T2050, New England Biolabs).

[0500] circRNA synthesis: The circular RNA precursor molecule was added to a buffer (50 mM Tris-HCl, 10 mM MgCl2, 1 mM DTT, pH 7.5) containing GTP (final concentration of 2 mM), and the solution was heated at 55 °C for 15 min and then column purified ( RNA Cleanup Kit, T2050, New England Biolabs).

[0501] Identification of circular RNA using capillary electrophoresis: The circularized RNA precursor and the crude circularized RNA product were measured by Nano-Drop (ThermoFisher Scientific), and then processed using QIAxcel RNA High Sensitivity Cartridge Kit (929112) according to the instructions, and then detected and analyzed using QIAxcel Connect capillary electrophoresis according to the operating instructions. This experiment used a high-sensitivity RNA marker (QIAxcel RNA High-Sensitivity Marker Set, 929112, QIAGEN) to indicate RNA molecular bands of different sizes.

[0502] Find SL cyclization elements in the CDS region:

[0503] (1) Natural sequence: In Gaussia Luciferase (SEQ ID NO. 70), the nucleotide sequence of the SL cyclization element located at positions 384 to 420 of Gaussia Luciferase was selected based on the secondary structure and named SL-Gaussia Luciferase (SEQ ID NO. 77). Its secondary structure is as follows Fig.25 The splicing site is located between the 18th U and 19th C in the SL-GaussiaLuciferase sequence. Fig.25 The arrows in the figure show the positions of the splice sites in the full-length GLuc.

[0504] (2) Non-natural sequences optimized by computer algorithms: A large number of computer algorithms can optimize RNA sequences, thereby improving the properties or functions of the RNA molecule itself. Among them, Linear design is a representative algorithm for optimizing the CDS region of mRNA. This algorithm makes the sequence of the CDS region of mRNA more stable by replacing synonymous codons without affecting the encoded protein product. In this embodiment, in the Gaussia Luciferase sequence optimized by Linear design (SEQ ID NO.78), it was found based on the secondary structure that the nucleotide sequence from the 384th to the 420th position of the Gaussia Luciferase sequence optimized by Linear design is also a stem-loop (SEQ ID NO.79), and the Stem-loop was selected as the SL cyclization element. The secondary structure of the sequence that has not been optimized by Linear Design is as follows Fig.26 As shown in A, the secondary structure of the sequence after Linear Design optimization is as follows Fig.26 As shown in B, Fig.26 C shows the specific location of the splicing site as indicated by the arrow. The splicing site is located between the 18th U and the 19th G of the SL-Gaussia Luciferase sequence. The splicing site is selected, and then the corresponding circular RNA precursor molecules (SEQ ID NO.80 and SEQ ID NO.81) are synthesized according to the above-mentioned design and synthesis method of circ-CVB3-IRES-GLuc-precursor.

[0505] Splice site in the internal loop: In addition to the terminal loop in the stem-loop, common loop structures in RNA also include internal loops. Therefore, this example compares the effect of the splice site located in the terminal loop and in the internal loop on the cyclization efficiency. Fig. 27As shown in A, the schematic diagram is cited from the literature (Nat Biotechnol. 2023; 41 (2): 262-272). In CVB3 IRES (SEQ ID NO. 68), the nucleotide sequence of the SL cyclization element located at positions 183 to 232 of CVB3 IRES was selected based on the secondary structure and named SL-CVB3-IRES (SEQ ID NO. 69). Its secondary structure is as follows Fig. 27 B. Next, the splicing site is selected: the splicing site is located between the 16th U and the 17th A in the SL-CVB3-IRES sequence, as shown in Fig. 27 The position indicated by the arrow in B. The splicing site was selected, and then the corresponding circular RNA precursor molecule (SEQ ID NO.82) was synthesized according to the above-mentioned design method for synthesizing circ-CVB3-IRES-GLuc-precursor.

[0506] The research results are as follows Fig.28 As shown, capillary lanes 1 and 2 are circular RNA precursor molecules and circularized products with splicing sites on the inner loop, respectively. Compared with lanes 1 and 2, most of the bands in lane 2 are located at the position of linear RNA, with only a small amount of circular bands. The circularization efficiency was calculated using the software (QIAxcel ScreenGel 2.0) that comes with the QIAxcel Connect capillary electrophoresis instrument. Among them, circularization efficiency = 100%*circularization / (linear+circularization). The quantitative results show that the circularization efficiency of this group is 6.77%. Capillary lanes 5 and 6 are circular RNA precursor molecules and circularized products synthesized at the end loop of the stem loop of CVB3 IRES, respectively. Compared with lanes 6 and 5, the bands in lane 6 are mainly distributed at the position of circular RNA, and the quantitative results show that the circularization efficiency of this group is 62.41%. The above two groups of experimental results show that the circularization efficiency of the splicing site on the end loop is significantly higher than that on the inner loop.

[0507] like Fig.28As shown, capillary lanes 3 and 4 are the circular RNA precursor molecules and the circularized products synthesized at the terminal loop of the natural Gaussia Luciferase sequence with the splicing site, respectively, and capillary lanes 7 and 8 are the circular RNA precursor molecules and the circularized products synthesized at the terminal loop of the Gaussia Luciferase sequence optimized by Linear Design, respectively. Compared with the circular RNA precursor molecules, the main band of the circularized product is located at the position of the circular RNA, and the circularization efficiency of the natural Gaussia Luciferase sequence and the Gaussia Luciferase sequence optimized by Linear Design are comparable, which are 78.55% and 79.21%, respectively. These results show that the "ribozyme-SL circularization module" can achieve high-efficiency circularization preparation in both natural and non-natural RNA sequences.

[0508] The peak data and cyclization efficiency of capillary electrophoresis used for quantitative calculation in this experiment are shown in the following table.

[0509] Lane 2 Lane 4 Lane 6 Lane 8 Circular (pg / μl) 2589.88 38577.59 23284.58 30942.76 Linear (pg / μl) 35670.35 10533.99 14026.36 8122.15 Cyclization efficiency (%) 6.77 78.55 62.41 79.21

[0510] Example 15. Uni-cRNA cyclization system can produce sequence-accurate circular RNA molecules

[0511] In this embodiment, a variety of methods are used to identify whether the product of the present invention is a target circular RNA molecule, and the identification method includes:

[0512] 1. Identification by RNaseR. RNaseR is a 3'-5' exonuclease that can cut and degrade RNA from the 3'-5' direction. It can degrade linear RNA molecules, but circular RNA without 3' and 5' ends cannot be degraded.

[0513] The specific steps are as follows: the purified circular RNA or linear circRNA precursor was heated at 65°C for 3 minutes and then immediately placed on ice for 2 minutes. The linear RNA was then digested with RNaseR (E049, Applied Biological Materials) (37°C, 30 min). RNA digested with RNase R was purified by column purification using RNA Cleanup Kit (T2050, New England Biolabs), and then 1.5% agarose gel was prepared for agarose gel electrophoresis analysis.

[0514] The results are as follows Fig.29As shown in Figure A, lanes 1 and 3 are linear circRNA precursors (precursor-mRNA, SEQ ID NO: 80) that have not undergone cyclization, lanes 2 and 4 are the products after cyclization, and lanes 3 and 4 are the results after RNaseR treatment. As shown in the figure, the main band position of lane 2 after cyclization is located below the band of lane 1, and an obvious band appears at the bottom, which indicates that the cyclization reaction produces circular mRNA molecules with reduced molecular weight and intron fragments with very small molecular weight. The RNaseR treatment group shows that no band appears in lane 3, while the band in lane 4, which is presumed to be circular RNA, still exists without the influence of RNaseR. Lanes 8 and 10 are linear circRNA precursors (precursor-mRNA, SEQ ID NO: 81) that have not undergone cyclization, lanes 9 and 11 are the products after cyclization, and lanes 10 and 11 are the results after RNaseR treatment. As shown in the figure, after cyclization treatment, the main band of lane 9 is located below the band of lane 8, and an obvious band appears at the bottom, which indicates that the cyclization reaction produces circular mRNA molecules with reduced molecular weight and intron fragments with very small molecular weight. The RNaseR treatment group shows that no band appears in lane 10, while the band of lane 11, which is presumed to be circular RNA, is still present without the influence of RNaseR.

[0515] 2. Use RNase H for identification. RNase H is an endoribonuclease that can specifically hydrolyze RNA in a DNA-RNA hybrid chain. Due to the difference in the structure of linear RNA and circular RNA, when the same probe is combined with linear RNA and circular RNA, the linear RNA will be cut into two bands, while the circular RNA will only be cut into a linear or one band.

[0516] The specific steps are as follows: add a 10-fold molar excess of DNA probe to the circular RNA obtained by RNaseR treatment, then heat at 65°C for 5 minutes and anneal at room temperature. Then digest with RNase H (M0279, New England Biolabs) at 37°C for 20 minutes, and column purify the digested RNA. The DNA probe sequences are shown in SEQ ID NO:83 and SEQ ID NO:84.

[0517] The experimental results are as follows Fig.29As shown in A, in the Seq-13 group, lane 5 is a linear RNA precursor molecule, lane 6 is a product after a cyclization reaction, and after the samples of lanes 5 and 6 are treated with RNase H, lane 5 forms two bands, while lane 6 forms a single band, which indicates that lane 6 contains circular RNA. Similarly, in the Seq-14 group, lane 12 is a linear RNA precursor molecule, and lane 13 is a product after a cyclization reaction. After being treated with RNase H, lane 12 forms two bands, while lane 13 forms a single band, which indicates that lane 13 contains circular RNA.

[0518] 3. Verify the accuracy of the ribozyme recognition site sequence by RT-PCR and Sanger sequencing. The aforementioned methods 1 and 2 can determine that a circular RNA molecule is formed after the cyclization reaction, but whether the connection near the splicing site is accurate requires further sequencing verification. Therefore, this embodiment also uses sequencing at the splicing site to prove that the cyclization reaction occurs at the expected position and forms the target circular RNA molecule.

[0519] The specific steps are as follows: The circular RNA obtained by RNaseR treatment is synthesized into the first strand of cDNA using the RevertAid First Strand cDNA Synthesis Kit (K1622, Thermo Fisher Scientific) and random primer. Then, two forward and reverse PCR primers are designed upstream and downstream of the splicing site, respectively. Using cDNA as a template, PCR is performed to obtain the amplified fragment and sequence verification is performed. The circular mRNA molecule will detect an RNA sequence that crosses the splicing site in accordance with the expectation near the splicing site. The primer sequences are shown in SEQ ID NO:85, SEQ ID NO:86, SEQ ID NO:87 and SEQ ID NO:88.

[0520] The results are as follows Fig.29 B and Fig.29 As shown in Figure C, the sequencing results of Seq-13 and Seq-14 confirmed that the sequences on both sides of the splicing site were consistent with the designed circular RNA sequence, proving that the head and tail of the circular RNA have been connected together. The correct circular mRNA molecule was successfully obtained.

[0521] In summary, this embodiment verifies that the target circular RNA molecule is successfully synthesized in the in vitro cyclization reaction system by three complementary methods.

[0522] Example 16. Circular mRNAs prepared by the Uni-cRNA cyclization system can all express proteins in cells

[0523] In this example, the protein expression level of circular GLuc mRNA prepared by Uni-cRNA in the present disclosure was detected in HEK-293T and A549 cells.

[0524] The nucleotide sequences involved in this example are as follows: SEQ ID NO: 75, SEQ ID NO: 89 (this sequence is a circular RNA construct circ-CVA20-IRES-GLuc-precursor synthesized by searching for stem-loop in CVA20 IRES, determining the splicing site and synthesizing the circular RNA construct circ-CVB3-IRES-hEPO-precursor according to the method described in Example 14).

[0525] Implementation method:

[0526] Cell transfection: Before transfection, HEK-293T or A549 cells were seeded at 20,000 cells / well in a 96-well plate, and then about 100 ng (equimolar amount of each RNA) of circular mRNA molecules prepared according to the method of Example 1 were purified by RNaseR and transferred into HEK-293T and A549 cells respectively according to the instructions using Lipofectamine 3000 transfection reagent (L3000001, Thermo Fisher Scientific).

[0527] Protein expression analysis: 24 hours after transfection, 10 μL of culture supernatant was taken from cells transfected with circ-GLuc, and the activity of luciferase in the supernatant was detected using the Pierce Gaussia Luciferase Glow Assay Kit (Thermo Fisher Scientific). The activity of luciferase was detected using the SpectraMax i3 microplate reader (Molecular Devices). For HEK-293T cells transfected with circ-hEPO mRNA, 10 μL of culture supernatant was taken after 24 hours, and the concentration of hEPO protein in the supernatant was detected using the Human Erythropoietin ELISA Kit (hEPO) (ab274397, Abcam) according to the instructions. The ELISA assay was performed using the SpectraMax i3 microplate reader (Molecular Devices).

[0528] Experimental results: Fig.30 A and Fig.30B shows the protein expression levels of circular GLuc mRNA containing different IRES sequences in HEK-293T and A549 cells, respectively. Fig.30 C shows the protein expression level of circular hEPO mRNA containing CVB3 IRES sequence in HEK-293T cells. These results demonstrate that the circular RNA prepared by the Uni-cRNA method of the present invention can effectively express GLuc and hEPO proteins.

[0529] Example 17. SL structural elements are widely present in different types of RNA molecules

[0530] In order to confirm the scope of application of the "ribozyme-SL cyclization module" approach in diverse RNA molecules, this example comprehensively analyzes the distribution of SL structural elements in various RNA molecules. Stem-loop structure is one of the most basic RNA secondary structures and is widely present in various RNA molecules. This example only uses several common RNA categories for exemplary analysis, where a representative data set is selected for analysis for each type of RNA. The present invention mainly involves two major types of RNA molecules: 1) mRNA (mRNA molecules serve as translation templates, and translation produces corresponding protein products as effector molecules), and 2) non-mRNA (not used to encode polypeptides and proteins, but RNA molecules themselves as effector molecules). Therefore, in this example, mRNA selects the human mRNA reference set as the analysis object, and non-mRNA selects non-coding RNA (ncRNA) as the analysis object. In addition, the IRES element is the most commonly used translation initiation regulatory element in circular mRNA, so this example also analyzes the distribution of stem-loop in IRES. For each type of RNA, the source and acquisition method of the representative data set selected in this example are as follows:

[0531] The 5'-UTR and 3'-UTR sequences were derived from the UTRdb (UTRdb 2.0: a comprehensive, expert-curated catalog of eukaryotic mRNAs untranslated regions. Lo Giudice C, Zambelli F, Chiara M, et al. Nucleic Acids Res. 2023 Jan 6; 51(D1): D337-D344.) database, which currently contains more than 12.9 million 5'-UTR sequences and more than 13.19 million 3'-UTR sequences from 673 species. In order to save computing time, this study used human UTR sequences as a typical case for TLS analysis, of which UTRdb contained a total of 93,457 5'-UTR sequences and 87,932 3'-UTR sequences. The criteria for inclusion in the analysis were: 1) sequence length greater than 150 nt, 2) only the longest sequence from the same gene was retained. Finally, a total of 12,661 5'-UTR sequences and 18,414 3'-UTR sequences were included in this analysis.

[0532] The human mRNA coding region sequence is derived from the Consensus Coding Sequence (CCDS) database (Consensus coding sequence (CCDS) database: a standardized set of human and mouse protein-coding regions supported by expert curation. Pujar S, O'Leary NA, Farrell CM, et al. Nucleic Acids Res. 2018 Jan 4; 46 (D1): D221-D228.). Specifically, the file containing the CDS nucleotide sequence (file name "CCDS_nucleotide.current.fna.gz") was downloaded from the official website of the CCDS database. This version of the CCDS database contains a total of 35,624 human CDS sequences, all of which were included in the analysis.

[0533] The IRES sequence comes from the IRESbase database (IRESbase: A Comprehensive Database of Experimentally Validated Internal Ribosome Entry Sites. Zhao J, Li Y, Wang C, et al. Genomics Proteomics Bioinformatics. 2020Apr; 18(2): 129-139.), which contains a total of 1,328 IRES sequences from humans, viruses, and other eukaryotic organisms. The dataset also contains a portion of artificially designed IRES sequences with a length of 174 nt.

[0534] Non-coding RNA (ncRNA) comes from the RNAChallenge benchmark dataset, which is a benchmark dataset for evaluating protein-coding RNA and non-coding RNA classification algorithms. It contains 11,040 ncRNA sequences from multiple species (A large-scale benchmark study of tools for the classification of protein-coding and non-coding RNAs. Dalwinder Singh and Joy Roy. Nucleic Acids Research, gkac1092, 2022). The download address of the dataset is https: / / github.com / cbl-nabi / RNAChallenge.

[0535] Experimental methods:

[0536] Search for stem-loop structure in RNA secondary structure. First, the RNAfold algorithm is used to predict the secondary structure of the RNA sequence, and the prediction results are presented using dot-bracket notation. Then, the stem-loop search program built in this article is used to search for each RNA sequence in the secondary structure file to see whether it contains a stem-loop structure and the number of different stem-loops. In this embodiment, the length of the stem-loop arm is limited to ≥2 consecutive complementary paired nucleotides, and the size of the loop is set to contain n (3≤n≤10) unpaired nucleotides.

[0537] Research findings:

[0538] The results are shown in the following table and Fig.31As shown in A, using stem-loop structural features for matching, in the human 5'-UTR, human CDS, human 3'-UTR, IRES, and non-coding RNA sequence sets analyzed in this example, the proportions of sequences containing at least one stem-loop are 99.69%, 99.99%, 99.85%, 98.72%, and 100%, respectively ( Fig.31 A) In addition, Fig.31 BF shows the percentage of sequences containing 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 and more than 10 stem-loop structures in the same sequence. The results show that in the human CDS region sequence, human 3'-UTR sequence and ncRNA sequence set, the percentage of sequences containing more than 10 stem-loop structures is very high, accounting for the vast majority of the sequences, which provides a large number of optional splicing sites for the design of the "ribozyme-SL structural module" of the present invention. The number of stem-loops contained in the IRES sequence set is mainly distributed between 1 and 5, and the number of stem-loops contained in the 5'-UTR sequence set is mainly distributed between 2 and 6. This may be due to the shorter average sequences of IRES and 5'-UTR.

[0539]

[0540] discuss

[0541] Example 5, Example 6, Example 7 and Example 8 together illustrate that in the present invention, the Hi-Scarless-PIE cyclization system based on the modular design concept of "ribozyme-TLSL cyclization module" has good versatility and practicality. Fig.16 B), for the target mRNA sequence, when the sequence contains TLSL structural elements, the "ribozyme-TLSL cyclization module" can be directly designed and high-efficiency cyclization preparation can be achieved. When the target mRNA molecule does not contain TLSL structural elements, the pre-designed TLSL structural element module can be inserted in the appropriate region to design the "ribozyme-TLSL cyclization module" and achieve high-efficiency cyclization preparation. This fully demonstrates that the Hi-Scarless-PIE cyclization platform proposed in the present invention is a universal circular mRNA preparation platform.

[0542] The uses of circular RNA mainly include two categories: mRNA and non-coding RNA. In circular mRNA, when using the IRES+CDS sequence combination, the SL structural element can be matched in almost 100% of the combined sequences, thereby constructing a "ribozyme-SL cyclization module" to achieve efficient circular mRNA preparation. The SL structural element exists in almost all non-coding RNAs, so the present invention can design a "ribozyme-SL cyclization module" accordingly to achieve efficient non-coding RNA cyclization preparation.

[0543] In summary, Example 14, Example 15 and Example 17 collectively illustrate that the Uni-cRNA cyclization platform based on the "ribozyme-SL cyclization module" proposed in the present invention can achieve efficient in vitro cyclization preparation of almost any RNA molecule. Example 16 demonstrates that circular mRNA prepared by Uni-cRNA can efficiently express protein. These examples collectively demonstrate that the Uni-cRNA of the present invention is a universal, efficient and high-fidelity RNA cyclization technology suitable for preparing circular mRNA and circular non-coding RNA for a wide range of uses.

[0544] All documents mentioned in the present invention are cited as references in this application, just as each document is cited as reference individually. In addition, it should be understood that after reading the above teachings of the present invention, those skilled in the art can make various changes or modifications to the present invention, and these equivalent forms also fall within the scope defined by the claims attached to this application.

Claims

1. A polynucleotide for preparing circular RNA, characterized in that: From the 5' end to the 3' end, the polynucleotide comprises an operably linked module Z1, a module to be cyclized L, and a module Z2, wherein the polynucleotide is a linear structure; The modules Z1 and Z2 are reconstructed into a ribozyme with self-splicing function, and self-splicing (cutting and connecting) is performed under appropriate conditions, so that the module L to be circularized forms a circular RNA, wherein the splicing site is located in the distal loop region (distal loop region) in the stem-loop structure (stem-loop structure) formed by the module L to be circularized after the circular RNA is formed.

2. The polynucleotide according to claim 1, wherein The module L to be cyclized is a linear counterpart corresponding to the target sequence, and the stem-loop structure is an endogenous stem-loop structure in the target circular RNA.

3. The polynucleotide according to claim 1, wherein The stem-loop structure is a single stem-loop structure or a poly stem-loop structure.

4. The polynucleotide according to claim 3, wherein The single stem-loop structure comprises a stem region and a terminal loop region located at the distal end of the stem region.

5. The polynucleotide according to claim 3, wherein The poly stem-loop structure has 2-5 terminal loop regions.

6. The polynucleotide according to claim 1, wherein The polynucleotide has a structure shown in Formula I: Z1-L-Z2 (Formula I) Z1 contains the 3′ end section h1 of the ribozyme; L is the linear counterpart of the target sequence; Z2 contains the 5' end section h2 of the ribozyme.

7. The polynucleotide according to claim 6, wherein The Z1 further comprises a 3' end section h1 of the ribozyme and a random sequence r1, and the Z2 further comprises a 5' end section h2 of the ribozyme and a random sequence r2. The structural formula of the polynucleotide is as shown in Formula II: r1-h1-L-h2-r2 (Formula II) Wherein r1 and r2 are reverse complementary sequences.

8. The polynucleotide according to claim 1, wherein The ribozyme is a self-catalytic splicing ribozyme, and the ribozyme can recognize a specific site in the L module.

9. The polynucleotide according to claim 8, wherein When the ribozyme matches a specific secondary structure, a high-efficiency splicing reaction can occur at a specific splicing site.

10. The polynucleotide according to claim 1, wherein The ribozyme is a modified ribozyme formed by partially or completely mutating the 10th to 16th positions of the 5' end of the pre-tRNA-Leu gene intron of Anabaena.

Images