Universal "scar"-free RNA circularization method

Abstract

Provided are a universal "scar"-free RNA circularization method, and a recombinant nucleic acid molecule for preparing a circular RNA. The recombinant nucleic acid molecule comprises, in a 5' to 3' direction, the following sequentially arranged elements: an optional 5' homology arm, a 3' half-intron fragment, a circularization fragment, a 5' half-intron fragment and an optional 3' homology arm, wherein the 5' end of the circularization fragment contains a 3' circularization recognition fragment, and the 3' end thereof contains a 5' circularization recognition fragment.

Description

A general and "scar"-free RNA circularization method This invention claims the priority of a prior application with the patent application number 2024100463312 and the invention title "A general and'scar'-free RNA circularization method", which was filed with the State Intellectual Property Office of China on January 11, 2024. The full text of this prior application is incorporated into this invention by reference. Technical Field This disclosure belongs to the field of biochemistry and specifically relates to methods for synthesizing circular RNAs. Background Art There are generally three methods for preparing circular RNAs: chemical method, ligase method, and ribozyme method. The chemical method has been phased out due to its extremely low circularization efficiency; the ligase method often has problems of easy intermolecular ligation, which reduces the circularization rate and also poses great challenges to product purification. In addition, the fragments that can be circularized by the ligase method are often very short, and these disadvantages limit the use of the ligase method. The ribozyme method is a method that uses the self-catalytic action of ribozymes to produce circular RNAs. Due to its own structure, ribozymes have an enzyme-like catalytic activity, and after back-splicing, circular RNAs are produced. After years of development, the ribozyme method has undergone a re-design of its structure, and the positions of introns and exons in the original ribozyme sequence are swapped, so it is called the PIE method (permuted Intron-Exon). The PIE method is currently the most efficient method for in vitro preparation of circular RNAs. The PIE method was initially proposed by Michael D. Been in 1992 [1] , who successfully prepared circular RNAs through the back-splicing process by modifying the type I intron from Anabeana. In 2018, Anderson et al. modified the PIE system [2] , adding elements such as homology arms and "spacers" that improve the circularization efficiency, greatly enhancing the efficiency of the PIE method for preparing circular RNAs and enabling successful circularization of RNAs up to 5 kb in length. However, in recent years, with the attention given to this technology, some researchers have found that circular RNAs prepared by the PIE method have strong immunogenicity. For example, Howard Y. Chang et al. successively in 2017 [3] and 2019 [4]It was found that the circular RNAs prepared by the PIE method generated strong immunogenicity in cells. By measuring cytokines such as RIG-I, it was found that the immunogenicity induced by the circular RNAs prepared by the PIE method was comparable to that of unmodified linear mRNAs. It was found that the immunogenicity was caused by the inevitable insertion of a foreign Exon sequence into the RNA during the preparation of circular RNAs by the PIE method; in addition, Ling-Ling Chen [5] et al. also found the same phenomenon, and also found that the source of immunogenicity was the "scar" sequence introduced during the preparation by the PIE method. Further studies showed that the immunogenicity was induced by the formation of an unconventional duplex structure by the "scar" sequence. Therefore, from the above studies, although the PIE method has a high cyclization efficiency, it has the disadvantage of inducing strong immunogenicity, which will greatly hinder the application of circular RNAs. The immunogenicity is mainly caused by the "scar" sequence. In addition, more and more literature has found that circular RNAs have regulatory functions and can be used as potential therapeutic molecules [6] . For the synthesis of therapeutic circular RNA molecules, if the PIE method is used, it will introduce additional sequences of nearly 200 nt, including the "scar" fragment and spacer sequence, which will obviously affect the biological activity of such therapeutic circular RNA molecules [7] . Therefore, by re-engineering the PIE method and developing a "scar"-free cyclization method, it is possible to maintain the excellent characteristics of the PIE method, avoid high immunogenicity, and at the same time can be used as a method for in vitro preparation of therapeutic circular RNAs. At the same time, in subsequent designs, this system is designed as a general system, and different sequences to be cyclized do not require special design, which is more convenient Circular nucleic acids are formed by the back splicing of linear plasmids (cDNA templates) through biological reactions to form a closed-loop structure by covalent bonds. The main impurities mainly come from the in vitro transcription (IVT) stage: T7 enzyme, DNA residues, dNTPs, precursor linearity, introns, open rings, polymers and circular isomers, etc. It can be known from animal cell experiments that precursor linearity, introns, polymers and circular isomers affect cytotoxicity; open rings affect cell activity, and at the same time, too many open rings have strong immunogenicity. By controlling the reaction vessels (bioreactors and wave reactors, etc.), reaction temperature, stirring speed and reaction time used in the in vitro transcription (IVT) process, as well as optimizing the DNA template, T7 enzyme, dNTP and PEG concentrations added during the IVT reaction process, the production of Naked RNA can be controlled and the yield of circular nucleic acids can be determined Circular RNAs (circRNAs) are a class of non-coding RNA molecules that do not have a 5′-terminal cap and a 3′-terminal poly(A) tail and form a circular structure through covalent bonds. CircRNAs are a novel type of RNA distinct from traditional linear RNAs, with a closed circular structure and are abundantly present in eukaryotic transcriptomes. Most circular RNAs are composed of exon sequences, are conserved in different species, and exhibit expression specificity in different tissues and developmental stages. Since circular RNAs are insensitive to nucleases, they are more stable than linear RNAs, which gives circular RNAs obvious advantages in the development and application of novel clinical therapeutic drugs. The laboratory level reported in the literature uses the A-Tailing / LiCl-buffer RNase R digestion method to obtain higher linear RNA removal and circRNA enrichment efficiency, resulting in relatively pure circRNAs. However, the total yield of circRNAs is <8%, which is not suitable for large-scale production in industrial processes. [8] . Summary of the Invention Technical Problems to be Solved by the Invention By exploring the molecular mechanism of type I ribozymes derived from Anabaena, it was found that scarless cyclization can be achieved by modifying the IG sequence, and the IG sequence and splicing sites need to meet the following requirements: ① The IG sequence needs to meet the sequence feature of "GNN". ② When the IG sequence recognizes the 5′ splice site, it is necessary to ensure base complementary pairing or "GU" wobble pairing. ③ The 3′ splice site is not conserved, but when it is base complementary paired with the sequence near the 5′ splice site, it will greatly affect the cyclization efficiency. ④ The two flanking splicing sites need to have homologous arms with a minimum length of 3 nt to stabilize the P1 duplex. Based on the above design principles, Firefly Luciferae (1653 nt) and eGFP (780 nt) were successfully cyclized to construct scarless circular RNAs containing only "IRES + ORF". At the same time, 2% EX-gel electrophoresis separation proved that they still retained a high cyclization efficiency, and large fragment RNAs can be effectively cyclized. Both cell experiments and animal experiments proved that the constructed scarless circular RNAs were successfully expressed in vivo. This proves the universality and feasibility of this scarless cyclization method. In addition, since the selection of splicing sites and the modification of IG sequences are required for circularization of different ORFs, which is not convenient. At the same time, in the experiment, we found that some spacer sequences can not only improve the circularization efficiency, but also enhance the translation of circular RNAs. We call these spacer sequences UTR sequences. We screened the UTR sequences and found that the sequence with the best enhancement effect on the translation of circular RNAs enhanced the translation by 3.7 times compared with the original circular RNA sequence of the PIE method. By modifying the IG sequence through the previous scarless circularization method, a scarless universal circularization system containing only "IRES + ORF + UTR" was constructed. The circular RNAs constructed by this method had a circularization efficiency comparable to that of the PIE method and stronger translation ability, which has been verified at the cellular and animal levels. The present invention is also based on the modification of the CBV3 IRES sequence. First, the present invention explored the optimal position for inserting the IRES element into the CVB3 IRES sequence. The present invention designed 11 positions for inserting gene sequences, and it was found that 4 insertion positions had translation effects in the experiment. The 4 specific insertion positions are respectively that the IRES element is inserted between domain I and domain II, named IRES-1; the IRES element is inserted into the stem-loop structure of domain II, named IRES-2; the IRES element is inserted into the stem-loop structure (Distal loop) of domain IV, named IRES-3; the IRES element is inserted into the stem-loop structure (Proximal loop) of domain IV, named IRES-4. The IRES element can enhance the binding of the CVB3 IRES sequence to ribosomes, thereby promoting the translation of circRNAs. The IRES element exists in circRNAs, folds into a structure similar to the initial tRNA, recruits more ribosomes and binds translation regulatory factors such as ITAF, and then introduces the ribosomes into the interior of circRNAs to bind and initiate protein translation. Ribosomes are highly diverse protein structures found in all cells. Under the action of translation regulatory factors such as ITAF, the ribosomes are introduced cThe internal structure of ircRNA for binding and initiation. Secondly, four IRES elements were screened in the present invention, and then the effects of IRES elements with different sequences having the same enhancing effect on the translation effect of cicRNA were explored. These four IRES elements all have the function of recruiting ribosomes, thus affecting the translation of circRNA. The four IRES element sequences screened were respectively cloned into a DNA template, and the composition of the DNA template was the same as above, and the optimal sequence was screened to obtain a gene sequence with high protein expression. Finally, based on the optimal position of the IRES element and the basis of screening the IRES element, a series of repeat sequence designs were carried out on the IRES element to obtain a sequence for the IRES element to achieve its function. Compared with the original sequence, IRES-1-A1-X2 in the screened sequence was increased by about 2.3 times. Finally, the conditions for in vitro IVT reaction of Circular RNA were optimized to make the circularization rate ≥ 75%, and the content of Naked RNA ≤ 3%; the purification method of circular nucleic acid was improved to ensure that the purity of circular nucleic acid ≥ 95%, the content of aggregates, precursor linear, intron and isomers < 1%, the total purification yield ≥ 30%, and the theoretical total yield of circular nucleic acid ≥ 70%, ensuring that the in vitro IVT reaction and purification of Circular RNA are suitable for industrial scale-up production. Specific solutions of the invention To solve the deficiencies of the prior art, the present disclosure specifically provides: The present disclosure first provides a recombinant nucleic acid molecule for preparing circular RNA. Along the 5' to 3' direction, the recombinant nucleic acid molecule includes elements operably linked in sequence: Optional 5' homologous arm, 3' half-intron fragment, circularization fragment, 5' half-intron fragment and optional 3' homologous arm; The 5' end of the circularization fragment contains the 3' circularization recognition fragment of the intron, and the 3' end contains the 5' circularization recognition fragment of the intron; The 5' half-intron fragment and 3' half-intron fragment are derived from group I introns, preferably group I introns from Anabeana, preferably derived from Anabaena tRNA leu . The 5' half-intron fragment and 3' half-intron fragment are used to form an intron sequence along the 5' to 3' direction; the nucleotide sequence of the 5' half-intron fragment contains a partial sequence of the intron sequence close to the 5' direction, the nucleotide sequence of the 3' half-intron fragment contains the remaining part of the intron sequence close to the 3' direction, and the 5' half-intron fragment contains an internal guide (IG) sequence; Wherein: (1) The IG sequence of the intron is 5'-GNN-3' with 3 nt, and the 5'-loop recognition fragment is 5'-NNY-3' with 3 nt, where Y is C or T / U; N is an optional nucleotide, and the nucleotides of the IG sequence corresponding to the 5'-loop recognition fragment have strict base complementary pairing or GU wobble pairing; (2) A 3'-loop recognition fragment of at least 2 nt, and it is not complementary to the 5'-loop recognition fragment and its adjacent sequence; (3) The adjacent sequences of the 5'-loop recognition fragment and the 3'-loop recognition fragment contain at least 3 nt of internal homology sequences. In certain specific embodiments of the present disclosure, the recombinant nucleic acid molecule satisfies at least one of the following conditions: 1) The second nucleotide of the IG sequence is not A and / or the third nucleotide is not G; 2) When the 3'-loop recognition fragment is located in the spacer sequence, it is not AAAA, AA, UUUU, UAAA, CAAAA, or GAAAA. In the specific embodiments of the present disclosure, the 3'-loop recognition fragment does not have obvious homology with the adjacent sequence of the 5'-loop recognition fragment, that is, it is not complementary to the adjacent sequence of the 5'-loop recognition fragment. In the specific embodiments of the present disclosure, the 5' end of the circularization fragment has a 3'-loop recognition fragment - internal homology sequence fragment, and the 3' end has an internal homology sequence fragment - 5'-loop recognition fragment. Optionally, there are 0, 1, 2, 3, 4, or 5 unpaired nucleotides between the loop recognition fragment and the internal homology sequence fragment. In a further preferred embodiment, the homology sequence is 3 nt, 4 nt, or 5 nt. In a further preferred embodiment, the length of the 3'-loop recognition fragment is 2 nt. In certain specific embodiments of the present disclosure, the circularization fragment contains the first part of the target polypeptide coding region or non-coding region, a translation initiation element, and the second part of the target polypeptide coding region or non-coding region. When designing the circularization fragment, the 5'-loop recognition fragment and the 3'-loop recognition fragment divide the target polypeptide coding region or non-coding region into two parts, that is, the target polypeptide coding region or non-coding region is cut as follows: the second part (including the 5'-loop recognition fragment) | (including the 3'-loop recognition fragment) the first part (see Figures 8a and 8c), and then the following circularization fragment is formed: the 3'-loop recognition fragment and the first part of the target polypeptide coding region or non-coding region - translation initiation element - the second part of the target polypeptide coding region or non-coding region and the 5'-loop recognition fragment. The further formed recombinant nucleic acid molecule structure is: Optional 5' homology arm, 3' half intron fragment, 3' loop recognition fragment and target polypeptide coding region or the first part of non-coding region - translation initiation element - target polypeptide coding region or the second part of non-coding region and 5' loop recognition fragment, 5' half intron fragment and optional 3' homology arm (Figure 8c); The 3' loop recognition fragment of the first part and the 5' loop recognition fragment of the second part are reconnected into a complete target polypeptide coding region or non-coding region after intron splicing. In certain specific embodiments of the present disclosure, the cyclization fragment comprises a first portion of a translation initiation element, a target polypeptide coding region or a non-coding region, and a second portion of a translation initiation element, i.e., a 5' looping recognition fragment and a 3' looping recognition fragment divide the translation initiation element into two portions, and after intron splicing, the 5' portion of the translation initiation element and the 3' portion of the translation initiation element are connected to form a complete translation initiation element region. In certain specific embodiments of the present disclosure, the circularized fragment comprises a 5' spacer portion, an optional translation initiation element, a target polypeptide coding region or a non-coding region, and a 3' spacer portion. In certain specific embodiments of the present disclosure, the 5' spacer sequence is a 5' UTR sequence, which has the functions of both "spacer" and improving translation efficiency. In certain specific embodiments of the present disclosure, the 3' spacer sequence is a 3' UTR sequence, which has the functions of both "spacer" and improving translation efficiency. The 5' looping recognition fragment is located in the 5' spacer and is recognized by the modified IG sequence, and the 3' looping recognition fragment is located in the 3' spacer, thereby forming a universal cyclization system without the need to edit or recognize the IRES sequence or the target polypeptide coding region or non-coding region. In certain specific embodiments of the present disclosure, the number of exogenously introduced nucleotides on the 5' spacer sequence and the 3' spacer sequence is less than 10. In certain specific embodiments of the present disclosure, three nucleotides are inserted at the third or fourth nucleotide of the 5' spacer sequence as internal homologous sequences, and the two or three nucleotides of the 5' spacer sequence itself are used as the 3' loop recognition fragment; 6 nucleotides are inserted at the 3' end of the 3'UTR, wherein the 1st to 3rd nucleotides are complementary to the nucleotides inserted in the 5' spacer sequence, the 4th to 6th sequences are NNY, and the bases corresponding to the NNY and IG sequences have strict base complementary pairing or GU wobble pairing. In certain specific embodiments of the present disclosure, 5 nucleotides are inserted at the 5' end of the 5' spacer sequence, wherein the 1st to 2nd nucleotides serve as 3' looping recognition fragments, the 3rd to 5th nucleotides serve as internal homologous sequences, and the two or three nucleotides of the 5' spacer sequence itself serve as 3' looping recognition fragments; 6 nucleotides are inserted at the 3' end of the 3' spacer sequence, wherein the 1st to 3rd nucleotides are complementary to the nucleotides inserted in the 5' spacer sequence, the 4th to 6th sequences are NNY, and the bases corresponding to the NNY and IG sequences have strict base complementary pairing or GU wobble pairing. Preferred 5' spacer moieties are selected from: Preferred 3' spacer moieties are selected from: Preferably, the 5' spacer sequence and the 3' spacer sequence are used alone or together; preferably, the 5' spacer sequence and the 3' spacer sequence are both polyA+8CA, preferably the polyA fragment contains 10-100 A, more preferably 60-80 A, most preferably 61 A at the 5' end and 71 A at the 3' end. Most preferably, it is a combination of 5'61A8CA and 3'71A8CA. In another embodiment, if present, the IRES sequence is an IRES sequence of: Taura syndrome virus, Triatoma virus, Theiler's encephalomyelitis virus, simian virus 40, Solenopsis invicta virus 1, Rhopalosiphum padi virus, reticuloendotheliosis virus, Forman poliovirus 1, Pseudoplusia includens virus, Kashmir bee virus, human rhinovirus 2, Homalodisca coagulata virus-1, human immunodeficiency virus type 1, Homalodisca coagulata virus-1, louse P virus, hepatitis C virus, hepatitis A virus, hepatitis GB virus, foot-and-mouth disease virus, human enterovirus 71, equine rhinovirus, Ectropis obliqua-like virus, encephalomyocarditis virus (EMCV), Drosophila C virus, tobacco mosaic virus, cricket paralysis virus, bovine viral diarrhea virus 1, black queen cell virus, aphid lethal paralysis virus, avian encephalomyelitis virus, acute bee paralysis virus, Hibiscus chlorotic ringspot virus, classical swine fever virus, human FGF2, human SFTPA1, human AML1 / RUNX1, Drosophila antennapedia, human AQP4, human AT1R, human BAG-1, human BCL2, human BiP, human c-IAP1, human cmyc, human eIF4G, mouse NDST4L, human LEF1, mouse HIF1α, human n.myc, mouse Gtx, human p27kipl, human PDGF2 / c-sis, human p53, human Pim-1, mouse Rbm3, Drosophila reaper, canine Scamper, Drosophila Ubx, human UNR, mouse UtrA, human VEGF-A, human XIAP, saliva virus, Coxsackie virus, Echovirus, Drosophila hairless, Saccharomyces cerevisiae TFIID, Saccharomyces cerevisiae YAP1, human c-src, human FGF-1, picornavirus, turnip crinkle virus, aptamer of eIF4G, Coxsackie virus B3 (CVB3) or Coxsackie virus A (CVB1 / 2). In another embodiment, the IRES is the IRES sequence of Coxsackie virus B3 (CVB3). In another embodiment, the IRES is the IRES sequence of encephalomyocarditis virus. In a specific embodiment of the present disclosure, the recombinant nucleic acid molecule comprises a CVB3 IRES sequence into which an IRES enhancer element is inserted. The insertion positions of the IRES enhancer element are: between domain I and domain II of the IRES sequence (named IRES-1), at the stem-loop structure of domain II of the IRES sequence (named IRES-2), at the stem-loop structure (Distal loop) of domain IV of the IRES sequence (named IRES-3), and at the stem-loop structure (Proximal loop) of domain IV of the IRES sequence (named IRES-4). The IRES enhancing element is: or a repeat sequence of the above sequences, for example: Optionally, the translation initiation element sequence comprises one or a combination of two or more of the following sequences: IRES sequence, 5'UTR sequence, Kozak sequence, sequence containing m6A modification, complementary sequence of ribosomal 18S rRNA. In certain specific embodiments of the present disclosure, the 5' spacer sequence is a 5'UTR sequence, which has the functions of "spacer" and improving translation efficiency. Optionally, the target polypeptide coding region or non-coding region is a protein coding region encoding a human protein or a non-human protein. Optionally, the protein coding region encodes an antibody. Optionally, the human protein or non-human protein is selected from hFIX, SP-B, VEGF-A, human methylmalonyl-CoA mutase (hMUT), CFTR, cancer autoantigen, and gene editing enzymes such as Cpf1, zinc finger nuclease (ZFN), and transcription activator-like effector nuclease (TALEN). Optionally, the protein is a protein for therapeutic use. Optionally, wherein the antibody is a human anti-HIV antibody. Optionally, wherein the antibody is a bispecific antibody. Optionally, wherein the bispecific antibody binds CD3 and CLDN6 or binds CD19 and CD22. Optionally, wherein the protein is a protein for diagnostic use. Optionally, wherein the protein coding region encodes Gauss luciferase (Gluc), firefly luciferase (Fluc), enhanced green fluorescent protein (eGFP), human erythropoietin (hEPO), or Cas9 endonuclease. In a specific embodiment of the present disclosure, the protein coding region comprises at least two coding regions, wherein a linker is connected between any two adjacent coding regions; preferably, the linker is a polynucleotide encoding a 2A peptide. Optionally, a translation initiation element is connected between any two adjacent coding regions; optionally, the translation initiation element located between any two adjacent coding regions comprises one or a combination of two or more of the following sequences: IRES sequence, 5'UTR sequence, Kozak sequence, sequence containing m6A modification, complementary sequence of ribosomal 18S rRNA. Wherein, the recombinant nucleic acid molecule further comprises an insertion element, and the insertion element is located upstream of the translation initiation element; the insertion element is selected from at least one of the following groups (i)-(iii): (i) transcriptional level regulatory element, (ii) translational level regulatory element, (iii) purification element; Optionally, the insertion element comprises a sequence of one or a combination of two or more of the following: untranslated region sequence, polyA sequence, aptamer sequence, riboswitch sequence, sequence that binds to a transcriptional regulatory factor. In a specific embodiment of the present disclosure, the recombinant nucleic acid has the homologous arms, wherein the length of each homologous arm is about 5-50 nucleotides; preferably, the length of each homologous arm is about 9-19 nucleotides. In a specific embodiment of the present invention, the recombinant nucleic acid molecule is a precursor RNA molecule for preparing circular RNA. In another specific embodiment of the present invention, the recombinant nucleic acid molecule is a DNA molecule that can obtain the above precursor RNA molecule by transcription. The second aspect of the present disclosure provides a recombinant expression vector, wherein the recombinant expression vector comprises the above-mentioned recombinant nucleic acid molecule. The third aspect of the present disclosure provides a circularized precursor RNA molecule, which is obtained by transcribing the recombinant expression vector, and the circularized precursor nucleic acid molecule fragment includes: an optional 5' homologous arm, a 3' half-intron fragment, a circularized fragment, a 5' half-intron fragment, and an optional 3' homologous arm; The 5' end of the circularized fragment contains a 3' circularization recognition fragment, and the 3' end contains a 5' circularization recognition fragment. The fourth aspect of the present disclosure provides a method for preparing circular RNA in vitro, Transcription step: transcribing the circularized precursor RNA molecule according to the above recombinant expression vector; Circularization step: the circularized precursor nucleic acid molecule undergoes a circularization reaction to obtain circular RNA; Optionally, the method further comprises a step of purifying the circular RNA. The present disclosure also provides a purification method for preparing circular RNA, which includes ultrafiltration, and a combination of the same or different means of affinity chromatography, molecular sieve chromatography, and CHT; Preferred combinations are: two affinity chromatographies, affinity chromatography and molecular sieve chromatography, or affinity chromatography and CHT chromatography; Preferably, arginine and PEG are added to the ultrafiltration reagent or chromatography reagent; preferably, the reagent is a buffer; preferably, the PEG is PEG2000-8000. The fifth aspect of the present disclosure provides a circular RNA molecule obtained from the above-mentioned recombinant nucleic acid molecule or vector, or the above-mentioned method. The sixth aspect of the present disclosure provides a circular RNA, which, along the 5' to 3' direction, comprises elements arranged in the following order: A translation initiation element, a target polypeptide coding region or a non-coding region; Optionally, the circular RNA comprises a 5' spacer sequence and a 3' spacer sequence located between the 5' end of the translation initiation element and the 3' end of the coding element; Each of the above-mentioned elements is as defined above. Preferably, the circular RNA comprises the above-mentioned spacer sequence and IRES sequence; Preferably, the circular RNA comprises a modification; preferably, the modification is m6A. The seventh aspect of the present disclosure provides a composition, wherein the composition comprises the above-mentioned recombinant nucleic acid molecule, the above-mentioned recombinant expression vector, or the above-mentioned circular RNA; preferably, it comprises the above-mentioned circular RNA; Optionally, the composition further comprises one or more pharmaceutically acceptable carriers; Optionally, the pharmaceutically acceptable carrier is selected from lipids, polymers or lipid-polymer complexes. The eighth aspect of the present disclosure provides a method for expressing a target polypeptide in a cell, wherein the method comprises the step of introducing the above-mentioned circular RNA, or the above-mentioned composition into the cell. The ninth aspect of the present disclosure provides a method for preventing or treating a disease, wherein the method comprises administering the above-mentioned circular RNA, or the above-mentioned composition to a subject. The present disclosure also provides a design method for a recombinant nucleic acid molecule for preparing circular RNA, the method comprising: 1) Providing a candidate target polypeptide coding region or non-coding region, a translation initiation element, and optionally a 5' spacer sequence and a 3' spacer sequence; 2) Searching for or setting a 5' cyclization recognition fragment, a 3' cyclization recognition fragment and an internal homology sequence in the target polypeptide coding region or non-coding region, the translation initiation element, or the 5' spacer sequence and 3' spacer sequence to design a cyclization fragment: The searching operation is: S-1) Searching for the following sequence in the above-mentioned elements: 5' internal homology sequence - 5' cyclization recognition fragment - 3' cyclization recognition fragment - 3' internal homology sequence; S-2) Splitting the corresponding element at the 3' junction of the 5' cyclization recognition fragment and the 3' cyclization recognition fragment according to the search result to form a cyclization fragment; The sequence is arranged at both ends of the 5' spacer sequence-translation initiation element-target polypeptide coding region or non-coding region-3' spacer sequence, respectively, and a 5' internal homologous sequence-5' looping recognition fragment sequence and a 3' looping recognition fragment-3' internal homologous sequence are arranged to form a circularized fragment; in: A) the 5' internal homologous sequence and the 3' internal homologous sequence contain at least 3 nt of reverse complementary sequence; B) the 3' looping recognition fragment is a sequence of at least 2 nt and has no significant homology with the adjacent sequence of the 5' looping recognition fragment; C) the 5' looping recognition fragment is a 3 nt sequence and is NNY, wherein N is an optional nucleotide and Y is C or T; wherein "-" represents a phosphodiester bond or 1-5 (e.g., 1, 2, 3, 4 or 5) optional nucleotides; 3) According to the circularized fragment obtained in step 2), the intron IG sequence to be used is modified, wherein the IG sequence is 3 nt "GNN" and has strict base complementary pairing or GU wobble pairing with the bases corresponding to the 5' circularization recognition fragment; 4) dividing the modified intron sequence into a 3' half intron fragment and a 5' half intron fragment, and arranging them in the following order to form a recombinant nucleic acid molecule for preparing circular RNA: an optional 5' homology arm, a 3' half intron fragment, a circularized fragment, a 5' half intron fragment, and an optional 3' homology arm; The 5' end of the circularization fragment comprises a 3' circularization recognition fragment, and the 3' end comprises a 5' circularization recognition fragment. Preferably, the setting is to insert 3 nucleotides as internal homologous sequences at the third or fourth nucleotide of the 5' spacer sequence, and use the two or three nucleotides of the 5' spacer sequence as the 3' loop recognition fragment; insert 6 nucleotides at the 3' end of the 3'UTR, wherein the 1st to 3rd nucleotides are complementary to the nucleotides inserted in the 5' spacer sequence, the 4th to 6th sequences are NNY, and have strict base complementary pairing or GU wobble pairing with the bases corresponding to the IG sequence; Also preferably, the setting is to insert 5 nucleotides at the 5' end of the 5' spacer sequence, wherein the 1st to 2nd nucleotides serve as 3' looping recognition fragments, the 3rd to 5th nucleotides serve as internal homologous sequences, and the two or three nucleotides of the 5' spacer sequence itself serve as 3' looping recognition fragments; and insert 6 nucleotides at the 3' end of the 3' spacer sequence, wherein the 1st to 3rd nucleotides are complementary to the nucleotides inserted in the 5' spacer sequence, the 4th to 6th sequences are NNY, and have strict base complementary pairing or GU wobble pairing with the bases corresponding to the IG sequence. The present disclosure also provides a computer module that stores a program for implementing the above design method. The present disclosure also provides a CVB3 IRES sequence into which an IRES enhancing element is inserted; Preferably, the insertion positions of the IRES enhancing element are: between domain I and domain II of the IRES sequence (designated IRES-1), at the stem-loop structure of domain II of the IRES sequence (designated IRES-2), at the stem-loop structure (Distal loop) of domain IV of the IRES sequence (designated IRES-3), and at the stem-loop structure (Proximal loop) of domain IV of the IRES sequence (designated IRES-4); Preferably, the IRES enhancing element is: or a repeat sequence of the above sequence: The present invention also provides mutants of Anabaena tRNA leu class I introns or fragments combinations with self-splicing activity derived from the mutants, and comprising at least one of the following groups of mutations: 1) The second base in the original IG sequence 5'-GAG-3' is mutated to a base other than A; or / and 2) The third base (G at 3') in the original IG sequence 5'-GAG-3' is mutated to a base other than G; Preferably, the mutant comprises at least a mutation of the third base (G at 3'); Preferably, the mutant does not contain the original E1 and E2 sequences; Preferably, it is a combination of SEQ ID No. 44 and 45 with the above mutations. Beneficial technical effects 1. By screening and modifying the sequence of Anabaena type I intron, it is verified that by modifying the IG sequence to recognize the circularization sequence, circularization without exogenous sequences can be achieved. Among them, the original IG sequence is "GAG", where the first "G" is strictly conserved and can only recognize the 5' splice site through "GC" pairing or "GU wobble" pairing; the second "A" and the third "G" are not conserved and can be modified according to the specific nucleotides of the circularization sequence, as long as strict base complementary pairing or wobble pairing relationships are ensured. In summary, the IG sequence of Anabaena needs to satisfy "GNN" and can be modified by the required circularization sequence. In addition, the 3' splice site of the circularization sequence is not conserved. The Anabaena ribozyme system is different from other type I ribozyme systems, and its IG sequence does not recognize the 3' splice site. The same conclusion is also found in this disclosure, but this disclosure further finds that at least 3 nt of nucleotides at the 3' splice site need to be non-homologous to the 5' splice site, otherwise the circularization efficiency will be affected. In addition, since the Anabaena ribozyme is derived from tRNA, its original Exon sequence is the stem structure of "clover" and naturally has 5 nt nucleotide length homology. When circularizing without exogenous sequences, a certain length of homology around the splice site also needs to be retained. This disclosure finds that a homology length of at least 3 nt will not affect the circularization efficiency. This disclosure also screened the UTR sequence. Here, the UTR refers to the spacer sequence in the original PIE technology, and its main function is to play a "spacer" role between the ribozyme and the IRES, thereby improving the circularization efficiency. This disclosure finds that the UTR sequence also has the functions of "spacing" and improving translation efficiency. In this disclosure, 10 sequences were first screened at both the 5'UTR and 3'UTR. It was found that 41A8CA and 51A8CA at the 5'UTR can both improve translation efficiency, and 15A8CA at the 3'UTR can also improve translation efficiency. At the same time, it was found that the improvement of translation efficiency is proportional to the number of "A". Therefore, after increasing the number of "A", the translation efficiency was further improved, and the optimal sequences of 5'61A8CA and 3'71A8CA were determined. This disclosure screened different 5' and 3'UTRs, and the optimal combination of the two ends of the UTR can improve the translation efficiency of circular RNA by about 5.7 times. In addition, the present disclosure also modifies based on the CVB3 IRES sequence. First, the present disclosure explores the optimal position for inserting the IRES enhancer element into the CVB3 IRES sequence. The present disclosure designs 11 positions for inserting the gene sequence, and it is experimentally found that there are 4 insertion positions with translation effects. The 4 specific insertion positions are respectively: the IRES enhancer element is inserted between domain I and domain II, named IRES-1; the IRES enhancer element is inserted into the stem-loop structure of domain II, named IRES-2; the IRES enhancer element is inserted into the stem-loop structure (Distal loop) of domain IV, named IRES-3; the IRES enhancer element is inserted into the stem-loop structure (Proximal loop) of domain IV, named IRES-4. The IRES enhancer element can enhance the binding of the CVB3 IRES sequence to ribosomes, thereby promoting the translation of circRNA. The IRES enhancer element exists in circRNA, folds into a structure similar to the initial tRNA, recruits more ribosomes, and binds translation regulatory factors such as ITAF, and then introduces the ribosomes into the interior of circRNA to bind and initiate protein translation. Ribosomes are highly diverse protein structures found in all cells. Under the action of translation regulatory factors such as ITAF, the ribosomes are introduced into the internal structure of circRNA to bind and initiate. Secondly, the present disclosure screens 4 IRES enhancer elements, and then explores the influence of IRES enhancer elements with different sequences having the same enhancement effect on the translation effect of cicRNA. These 4 IRES enhancer elements all have the function of recruiting ribosomes, thus affecting the translation of circRNA. The 4 screened IRES enhancer element sequences are respectively cloned into the DNA template, and the composition of the DNA template is the same as above, and the optimal sequence is screened to obtain the gene sequence with high protein expression. Finally, based on the optimal position of the IRES enhancer element and the basis of screening the IRES enhancer element, the present disclosure conducts a series of repetitive sequence designs on the IRES enhancer element to obtain the sequence for the IRES enhancer element to achieve its function. Among the screened sequences, IRES-1-A1-X2 has increased by about 2.3 times compared with the original sequence. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1: Design diagram of Example 1 (Figure 1a: Main components of the PIE method and the structure of Design1; Figure 1b: Design2 added "GU" wobble pairing; Figure 1c: Design2 generated circular RNA by 2% EX-gel electrophoresis analysis; Figure 1d: The circular RNA generated by Design2 was introduced into Hela cells to measure expression; Figure 1e: Reverse transcription PCR sequencing of the product of Design2) Figure 2: Effect of the number of bases in the 3'-circularization recognition fragment on the circularization efficiency (Figure 2a: Trunc-Exon; Figure 2b: Reverse transcription PCR sequencing of Trunc-Exon products; Figure 2c: Fluorescent electrophoresis of products after single nucleotide deletion in the 5nt sequence; Figure 2d: Cellular expression of products after single nucleotide deletion in the 5nt sequence) Figure 3: Conservation test of the 2nt sequence in the 3'-circularization recognition fragment (Figures 3a-b: Construction of possible combinations based on the 2nt structure; Figures 3c-d: Almost all 2nt sequences can be successfully circularized without a decrease in circularization efficiency; Figures 3e-f: Adjacent homology between the 2nt and 5'-circularization recognition fragments) Figure 4: Partial conservation of the IG sequence and recognition of the 5'-circularization recognition fragment through base pairing or "GU" wobble pairing (Figure 4a: Original Exon1 sequence; Figure 4b: Mutant Exon1 sequence; Figure 4c: The IG sequence needs to meet the sequence feature of "GNN"; Figure 4d: Effect of "GU" wobble pairing on splicing; Figure 4e: Failure to meet base pairing completely disrupts splicing) Figure 5: Effect of internal homology (Figure 5a: Internal homologous sequence of the original Exon of Anabaena; Figure 5b: 19nt-long homology arm in the studied sequence; Figure 5c: Effect of removing the 19nt; Figures 5d-f: Effect of homologous sequences with different GC ratios and lengths on the circularization efficiency) Figure 6: Overall design principle of the present invention Figure 7: Design and results of circularized IRES+Fluc (Figure 7.a: Circular RNA containing only "IRES+Fluc"; Figure 7b: Splice site setting; Figure 7c: Circularization efficiency; Figure 7d: Translation efficiency; Figure 7e: Reverse transcription PCR sequencing in vitro) Figure 8: Design and results of circularized IRES+eGFP (Figure 8a: Splice site and internal homologous sequence setting; Figure 8b: Circularization efficiency; Figure 8c: Reverse transcription PCR sequencing in vitro; Figure 8d: Translation efficiency) Figure 9: Effect of UTR on circularization efficiency and translation efficiency (Figure 9a: 5'UTR screening; Figure 9b: 3'UTR screening; Figures 9c-d: PolyA sequence optimization; Figures 9e-g: "IRES+ORF+UTR" scarless circularization system) Figure 10: Insertion position of CVB3 IRES Figure 11: Electrophoresis diagram of the IRES enhancer element Figure 12: Transfection effect of the IRES sequence with the inserted IRES enhancer element in Example 4 of the present invention Figure 13: Production process flow of circular RNA Figure 14: Gluc - Gel electrophoresis diagram of the cRNA production process Figure 15 NHS Affinity Purification of the First Step of Gluc-cRNA Figure 16 SDS-PAGE Diagram of Sampling for the Affinity Purification of Gluc-cRNA (Sample Loading Amount: 12.2 mg; Total Recovery of cRNA: 10.23 mg; Recovery Rate of cRNA: 83.85%; Purity: 88.2%) Figure 17 Molecular Sieve Purification of Gluc-cRNA-Core400 Figure 18 Agilent 5200 Capillary Electrophoresis Diagram in the Two-Step Purification Method of Gluc-cRNA (Sample Loading Amount: 9.4 mg; Total Recovery of cRNA: 8.01 mg; Recovery Rate of cRNA: 85.21%; Purity: 98.9%) Detailed Implementation Modes Aiming at the current technical problems of circular RNA, taking the ribozyme of Anabaena as the research object, by modifying the internal guide sequence in the ribozyme to make it recognize the required cyclization sequence, the synthesis of circular RNA without the incorporation of foreign sequences is achieved. At the same time, this disclosure has screened multiple UTR sequences, and the screened UTR sequences have an enhancing effect on the translation of circular RNA. The optimal combination increases the translation efficiency by 5.7 times. In addition, the IG sequence in this disclosure recognizes the UTR sequence, making this system a universal cyclization system. For different cyclization sequences, no other designs are required. And the IRES sequence is screened and modified. Through the research on the cyclization mechanism of the PIE method, this disclosure finds that the occurrence of the cyclization process needs to meet the following principles: ① The 3 nucleotides of the IG sequence need to meet the characteristic of "GNN"; ② The IG sequence needs to ensure base complementary pairing or GU wobble pairing with the 3 nucleotides at the 5' splice site; ③ The 3' splice site should not have obvious homology with the 5' splice site and its adjacent nucleotides. Homology is not conducive to splicing and will greatly reduce the cyclization efficiency; ④ At least 3 nucleotides adjacent to the 5' splice site need to have homology with at least 3 nucleotides adjacent to the 3' splice site. In addition, through the modification of the internal guide sequence to make it recognize the UTR sequence, a method for RNA cyclization without "scar" sequence has been successfully developed, and the advantage of high cyclization efficiency is retained. At the same time, due to the recognition of the spacer sequence, this method is universal for different sequences to be cyclized and no additional design is required. Secondly, the present disclosure screened 4 IRES enhancer elements, and then explored the influence of IRES enhancer elements with different sequences having the same enhancing effect on the translation effect of cicRNA. These 4 IRES enhancer elements all have the function of recruiting ribosomes, thus having an impact on the translation of circRNA. The 4 screened IRES enhancer element sequences were respectively cloned into a DNA template, and the composition of the DNA template was the same as above, and the optimal sequence was screened to obtain a gene sequence with high protein expression. Finally, based on the optimal position of the IRES enhancer element and the basis of screening the IRES enhancer element, a series of repeat sequence designs were carried out on the IRES enhancer element to obtain the sequence for the IRES enhancer element to achieve its function. Among the screened sequences, IRES-1-A1-X2 increased by about 2.3 times compared with the original sequence. Technical terms When used in the claims and / or the specification in conjunction with the term "comprising", the words "a" or "an" can mean "one", but can also mean "one or more", "at least one", and "one or more than one". As used in the claims and the specification, the words "comprising", "having", "including", or "containing" mean inclusive or open-ended and do not exclude additional, unrecited elements or method steps. Throughout the application, the term "about" means: a value includes the standard deviation of the error of the device or method used to measure that value. Although the disclosed content supports the definition of the term "or" as only alternatives and "and / or", unless expressly stated to be only alternatives or mutually exclusive between alternatives, the term "or" in the claims means "and / or". The terms "polypeptide", "peptide", and "protein" are used interchangeably in the present disclosure and are amino acid polymers of any length. The polymer can be linear or branched, it can contain modified amino acids, and it can be separated by non-amino acids. In a specific embodiment of the present disclosure, the nucleic acid or peptide sequence of the present invention comprises a sequence with a homology or sequence identity of more than 90%, preferably more than 95%, more preferably 96%, 97%, 98%, 99%. Methods for determining sequence homology or identity known to those of ordinary skill in the art include, but are not limited to: Computational Molecular Biology, Lesk, A.M. ed., Oxford University Press, New York, 1988; Biocomputing: Informatics and Genome Projects, Smith, D.W. ed., Academic Press, New York, 1993; Computer Analysis of Sequence Data, Part I, Griffin, A.M. and Griffin, H.G. eds., Humana Press, New Jersey, 1994; Sequence Analysis in Molecular Biology, von Heinje, G., Academic Press, 1987 and Sequence Analysis Primer, Gribskov, M. and Devereux, J. eds. M Stockton Press, New York, 1991 and Carillo, H. and Lipman, D., SIAM J. Applied Math., 48: 1073 (1988). Preferred methods for determining identity obtain the maximum match between the sequences being tested. Methods for determining identity are compiled in computer programs available to the public. Preferred computer program methods for determining identity between two sequences include, but are not limited to: the GCG program package (Devereux, J. et al., 1984), BLASTP, BLASTN, and FASTA (Altschul, S, F. et al., 1990). The BLASTX program (BLAST Manual, Altschul, S. et al., NCBI NLM NIH Bethesda, Md. 20894; Altschul, S. et al., 1990) is available to the public from NCBI and other sources. The well-known Smith Waterman algorithm can also be used to determine identity. Group I ribozyme or Group I intron As used in the present disclosure, a Group I ribozyme or Group I intron refers to a "Group I Intron" that has a self-splicing system that self-splices into a loop under the conditions of the presence of GTP and Mg 2+ and the presence of Mg. Group I introns are a class of very large ribozymes that can undergo self-splicing reactions and are usually widely present in many species, mainly participating in the catalytic excision of precursors of mRNA, tRNA, and rRNA. According to the secondary structure, group I introns can be divided into 10 domains, named P1 - P10 respectively. Their catalytic center is located at the junction of the P4 - P6 (P4, P5, P6) and P3 - P9 (P3, P7, P8, P9) domains. P1 and P10 are the 5' and 3' splice sites respectively, while the other sequences surrounding the conserved catalytic center are responsible for maintaining the stability and correct folding of the ribozyme. Its core secondary structure includes paired regions (P) and corresponding loop regions (L). The splicing of Group I Intron proceeds through two consecutive transesterification reactions. Exogenous guanosine or guanosine nucleotide (G) first docks at the active G-binding site located in P7, and its 3'-OH aligns to attack the phosphodiester bond at the 5′ splice site in P1, resulting in a free 3′-OH group at the upstream exon, and the exogenous G is linked to the 5′ end of the intron. Then the terminal G (omega G) of the intron exchanges with the exogenous G, occupying the G-binding site, organizing the second transesterification reaction: the 3′-OH group of the upstream exon in P1 aligns to attack the 3′ splice P10 site, resulting in the ligation (circularization) of the adjacent upstream and downstream exons and the release of the catalytic intron. In the present disclosure, intron fragment I and intron fragment II, or the 5' half-intron fragment and the 3' half-intron fragment have the same meaning, referring to sequences derived from group I introns and respectively containing partial sequences of group I introns close to the 5' direction and partial sequences close to the 3' direction. The internal cleavage sites of group I introns can be determined by those skilled in the art. For example, they can be determined with reference to the following literature: Puttaraju M., et al., (1992). Group I permuted intron-exon (PIE) sequences self-splice to produce circular exons; Puttaraju M., et al., (1996). Circular ribozymes generated in Escherichia coli using group I self-splicing permuted intron-exson sequences. For example, for group I introns, particularly for Anabaena group I introns, they can usually be cleaved at specific sites within their P6 region to form a PIE system. Therefore, in some embodiments, the 3′-end portion of a native group I intron (e.g., Anabaena group I intron) is the portion from a specific site within the P6 region of the native group I intron (e.g., Anabaena group I intron) to the 3′ terminus. In some embodiments, the 5′-end portion of the native group I intron (e.g., Anabaena group I intron) is the portion from a specific site within the P6 region of the native group I intron (e.g., Anabaena group I intron) to the 5′ terminus. For group I introns, particularly for Anabaena group I introns, the cleavage site can also be located within their P2, P5, P8, or P9 region to form a PIE system, as shown in WO2021236855A1. The group I introns also include various truncated forms or other mutants with self-splicing activity, such as the corresponding variants disclosed in WO2020 / 237227A1 (Tables 4 and 5), WO2023046153A1, or CN115997018A, etc. In some embodiments, the self-splicing intron is a Group I intron of Anabaena, for example, the Group I intron of the Anabaena pre-tRNA-Leu gene. Accordingly, in some embodiments, the 3′ self-splicing intron fragment (3′ Group I intron fragment) and the 5′ self-splicing intron fragment (5′ Group I intron fragment) are derived from the Group I intron of Anabaena, for example, the Group I intron of the Anabaena pre-tRNA-Leu gene. In some embodiments, the native Group I intron of the Anabaena pre-tRNA-Leu gene has the nucleotide sequence of SEQ ID NO: 43. The P6 region corresponds to positions 98 to 157 of SEQ ID NO: 43. The cleavage site can be any position between positions 122 and 138 of SEQ ID NO: 43. In some embodiments, the 3′ self-splicing intron fragment (3′ Group I intron fragment) is derived from the Group I intron of the Anabaena pre-tRNA-Leu gene and comprises the nucleotide sequence of SEQ ID NO: 44 or a nucleotide sequence having at least 75%, for example, at least 80%, at least 85%, at least 90%, at least 95%, 100% identity to SEQ ID NO: 44, or consists of the same. In some embodiments, the 5′ self-splicing intron fragment (5′ Group I intron fragment) is derived from the Group I intron of the Anabaena pre-tRNA-Leu gene and comprises the nucleotide sequence of SEQ ID NO: 45 or a nucleotide sequence having at least 75%, for example, at least 80%, at least 85%, at least 90%, at least 95%, 100% identity to SEQ ID NO: 45, or consists of the same. Internal guide sequence or IGS The internal guide sequence (IGS) generally refers to a nucleotide sequence in a Group I intron that pairs with the corresponding exon sequence through Watson-Crick pairing or wobble pairing, and is usually located in the P1 stem of the Group I intron. The original IG sequence of the Group I intron of the pre-tRNA-Leu gene is "GAG". 5’ loop recognition fragment and 3’ loop recognition fragment The 5’ loop recognition fragment is derived from the exon sequence (Exon 1, E1) linked to the 5’ end of the Group I intron, and the 3’ loop recognition fragment is derived from the exon sequence (Exon 2, E2) linked to the 3’ end of the Group I intron, wherein the internal guide sequence recognizes the 5’ loop recognition fragment through Watson-Crick pairing or wobble pairing. Wherein E1 is the adjacent exon sequence adjacent to the 5' splice site, and its length is at least 1 nucleotide (for example, the length is at least 5 nucleotides, the length is at least 10 nucleotides, the length is at least 15 nucleotides, the length is at least 20 nucleotides, the length is at least 25 nucleotides, the length is at least 50 nucleotides); wherein E2 is the adjacent exon sequence adjacent to the 3' splice site, and its length is at least 1 nucleotide (for example, the length is at least 5 nucleotides, the length is at least 10 nucleotides, the length is at least 15 nucleotides, the length is at least 20 nucleotides, the length is at least 25 nucleotides, the length is at least 50 nucleotides). WO2023046153A1 discloses alternative E1 fragments such as CUU or CUC of the class I intron of the pre-tRNA-Leu gene; and alternative E2 fragments such as AAAA, AA, UUUU, CAAA or GAAA. In the prior art, usually E1 and E2 (including optimized E1 and E2) are retained in the loop-forming sequence to form a "scar" of 5'-E1-E2-3’, also known as the residual loop-forming sequence; in the present disclosure, since the IG sequence has been modified, the selection of the 5' loop-recognizing fragment is thus expanded, and the corresponding 5' loop-recognizing fragment can be set in the translation initiation element, coding region or spacer sequence; similarly, the present disclosure also optimizes the number and type of bases of the 3' loop-recognizing fragment, expanding the selectivity of the 3' loop-recognizing fragment. In the present disclosure, the 3' loop-recognizing fragment does not have obvious homology with the 5' loop-recognizing fragment and its adjacent nucleotides so as not to reduce the cyclization efficiency, and the obvious homology means that two or more, such as three nucleotide bases, are complementary to each other. Homologous sequence or homologous arm The homologous sequence or homologous arm has the same meaning in the present disclosure, including the 5' homologous arm located at the 5' end of the recombinant nucleic acid molecule and the 3' homologous arm located at the 3' end of the recombinant nucleic acid molecule, and the nucleic acid sequence of the 5' homologous arm is complementary to the nucleotide sequence of the 3' homologous arm. Internal homology or internal homologous sequence In E1 and E2, or in the nucleotides adjacent to the 5' loop-recognizing fragment and the 3' loop-recognizing fragment in the artificially inserted cyclization fragment, several complementary fragments are required to form a "splicing bubble". It is found in the present disclosure that at least 3 nucleotides adjacent to the 5' loop-recognizing fragment need to have homology with at least three nucleotides adjacent to the 3' loop-recognizing fragment. In certain embodiments of the present disclosure, the internal homology is provided by a spacer sequence. The term "adjacent region" or "adjacent sequence" refers to a region or sequence of 1-20 bases, preferably 1-10 bases, that is connected to and located upstream or downstream of a described fragment such as a 5' loop recognition fragment or a 3' loop recognition fragment in the recombinant nucleic acid molecule; in a specific embodiment of the present disclosure, the adjacent sequence or adjacent region of the 5' loop recognition fragment or the 3' loop recognition fragment refers to the sequence or region that is adjacent to the above loop recognition fragment and located in the sequence or region of the circularized fragment, that is, it remains in the final circular RNA molecule together with the loop-forming fragment. Spacer sequence: In the present disclosure, a "spacer sequence" refers to any of the following continuous nucleotide sequences: 1) predicted to avoid interfering with proximal structures, such as those from an IRES, coding or non-coding regions or introns, 2) at least 7 nucleotides in length (optionally not exceeding 100 nucleotides), 3) located downstream and adjacent to the 3' intron fragment and / or upstream and adjacent to the 5' intron fragment, and / or 4) containing one or more of the following: a) an unstructured region at least 3 nt long, b) a region predicted to base pair with a distal (i.e., non-adjacent) sequence at least 3 nt long, which includes another spacer sequence, and / or c) a structured region at least 7 nt long, the scope of which is limited to the sequence of the spacer sequence. In certain embodiments, the spacer sequence can be, for example, at least 10 nucleotides in length, at least 15 nucleotides in length, or at least 30 nucleotides in length. In certain embodiments, the length of the spacer sequence is at least 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 25, or 30 nucleotides. In certain embodiments, the length of the spacer sequence does not exceed 100, 90, 80, 70, 60, 50, 45, 40, 35, or 30 nucleotides. In certain embodiments, the length of the spacer sequence is 20 to 50 nucleotides. In certain embodiments, the length of the spacer sequence is 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, or 50 nucleotides. The spacer sequence can be a polyA sequence, a polyA-C sequence, a polyC sequence, or a poly-U sequence, or the spacer sequence can be specifically modified according to the IRES. The spacer sequence described in the present disclosure can have two functions: (1) promoting cyclization and (2) improving translation efficiency. More specifically, the spacer sequence described in the present disclosure is designed to have the following functions: 1) being inert to the folding of the proximal intron and the IRES structure; 2) sufficiently separating the intron and the IRES secondary structure; 3) carrying a splice site; 4) containing a spacer sequence-spacer sequence complementary region to promote the formation of a "splicing bubble"; 5) improving translation efficiency. The present disclosure preferably selects three types of 5' spacer sequences: The first type is the protienbingding sequence, which can improve translation efficiency by binding and recruiting proteins, and the two sequences are 31A8CA and 41A8CA respectively; the second type is the IRES enhancer sequence, and the IRES binds to 18s rRNA, eIF4G, and eIF4A; the third type is the full-length UTR sequence. The 5' spacer sequence and the 3' spacer sequence of the present disclosure are preferably used together. Preferably, both are polyA+8CA. Preferably, the polyA fragment contains 10-100 As, more preferably 60-80 As, and most preferably 61 As at the 5' end and 71 As at the 3' end. In a specific embodiment of the present disclosure, splice sites and internal homologous sequences are set on the 5' spacer sequence and the 3' spacer sequence, thereby forming a general cyclization system. The "splicing bubble" refers to the region between the homologous arms and the internal homologous region, which contains a splicing ribozyme, see CN 112399860 A. The term "coding region" refers to a gene sequence that can transcribe messenger RNA and ultimately be translated into a target polypeptide or protein. The term "expression" includes any step involved in polypeptide production, including but not limited to: transcription, post-transcriptional modification, translation, post-translational modification, and secretion. Unless otherwise defined or clearly indicated by the context, all technical and scientific terms in the present disclosure have the same meaning as commonly understood by those of ordinary skill in the art to which the present disclosure belongs. The present disclosure provides a general and "scar"-free RNA cyclization method, and the method includes but is not limited to a DNA construct for preparing circular RNA, a recombinant expression vector including the DNA construct, a circular precursor RNA molecule obtained by in vitro transcription using the recombinant expression vector, etc. The present disclosure first provides recombinant nucleic acid molecules for preparing circular RNAs. Exemplarily, the recombinant nucleic acid molecules can be the above-mentioned DNA constructs or circular precursor RNA molecules for preparing circular RNAs. Circularized fragment In the present disclosure, the finally circularized RNA molecular fragment or its corresponding DNA fragment is referred to as a circularized fragment, which contains a translation initiation element, a target polypeptide coding region or a non-coding region, and other optional regulatory and insertion elements such as 5' or 3' UTRs, and also contains a spacer fragment between intron fragments. In some specific embodiments of the present disclosure, the circularized fragment contains the 5' part of the target polypeptide coding region or non-coding region, a translation initiation element, and the 3' part of the target polypeptide coding region or non-coding region, that is, the 5' circularization recognition fragment and the 3' circularization recognition fragment divide the target polypeptide coding region or non-coding region into two parts, and after intron splicing, they are reconnected into a complete target polypeptide coding region or non-coding region. In some specific embodiments of the present disclosure, the circularized fragment contains the 5' part of the translation initiation element, the target polypeptide coding region or non-coding region, and the 3' part of the translation initiation element, that is, the 5' circularization recognition fragment and the 3' circularization recognition fragment divide the translation initiation element into two parts, and after intron splicing, the 5' part and the 3' part of the translation initiation element are connected into a complete translation initiation element region. Translation initiation element In the present disclosure, the translation initiation element can be any type of element that can initiate the translation of a target polypeptide. In some embodiments, the translation initiation element is an element containing any one or more of the following sequences: IRES sequence, 5'UTR sequence, Kozak sequence, a sequence containing m6A modification (N(6)-methyladenosine modification), and a complementary sequence of ribosomal 18S rRNA. In some other embodiments, the translation initiation element can also be any other type of cap-independent translation initiation element. In some embodiments, the translation initiation element is an IRES element, and the sources of the IRES element include, but are not limited to, viruses, mammals, Drosophila, etc. In some alternative embodiments, the IRES element is derived from a virus. Exemplarily, the IRES enhancer element contains an IRES sequence from a picornavirus. In some alternative embodiments, the IRES element is derived from Taura syndrome virus, Triatoma virus, Theiler's encephalomyelitis virus, simian virus 40, Solenopsis invicta virus 1, Rhopalosiphum padi virus, reticuloendotheliosis virus, Forman poliovirus 1, Autographa californica multiple nucleopolyhedrovirus, Kashmir bee virus, human rhinovirus 2, Homalodisca coagulata virus-1, human immunodeficiency virus type 1, Homalodisca coagulata virus-1, louse P virus, hepatitis C virus, hepatitis A virus, hepatitis GB virus, foot-and-mouth disease virus, human enterovirus 71, equine rhinovirus, Ectropis obliqua-like virus, encephalomyocarditis virus (EMCV), Drosophila C virus, tobacco mosaic virus, cricket paralysis virus, bovine viral diarrhea virus 1, black queen cell virus, aphid lethal paralysis virus, avian encephalomyelitis virus, acute bee paralysis virus, Hibiscus chlorotic ringspot virus, classical swine fever virus, human FGF2, human SFTPA1, human AML1 / RUNX1, Drosophila antennapedia, human AQP4, human AT1R, human BAG-1, human BCL2, human BiP, human c-IAP1, human cmyc, human eIF4G, mouse NDST4L, human LEF1, mouse HIF1ɑ, human n.myc, mouse Gtx, human p27kipl, human PDGF2 / c-sis, human p53, human Pim-1, mouse Rbm3, Drosophila reaper, canine Scamper, Drosophila Ubx, human UNR, mouse UtrA, human VEGF-A, human XIAP, salivary virus, Coxsackievirus, Echovirus, Drosophila hairless, Saccharomyces cerevisiae TFIID, Saccharomyces cerevisiae YAP1, human c-src, human FGF-1, simian picornavirus, turnip crinkle virus, an aptamer of eIF4G, Coxsackievirus B3 (CVB3) or Coxsackievirus A (CVB1 / 2); further preferably, the IRES is the IRES sequence of Coxsackievirus B3 (CVB3); still preferably, the IRES is the IRES sequence of encephalomyocarditis virus. Preferably, it can be selected from the specific sequences shown in Table 3 of WO2020 / 237227A1. Target polypeptide coding region or non-coding region In some embodiments, the target polypeptide coding region or non-coding region is not a naturally occurring nucleotide sequence. In some embodiments, the target polypeptide coding region encodes a natural or synthetic protein. In some embodiments, the coding or non-coding region can be a natural or synthetic sequence. In some embodiments, the coding region can encode a chimeric antigen receptor, an antibody, an immunomodulatory protein, and / or a transcription factor. In some embodiments, the non-coding region can encode a sequence that can alter cell behavior (e.g., lymphocyte behavior). In some embodiments, the non-coding sequence is antisense to the cellular RNA sequence. In certain specific embodiments of the present disclosure, the circularized fragment comprises a 5'UTR, a target polypeptide coding region, and a 3'UTR, and the 5'-circularization recognition fragment and the 3'-circularization recognition fragment are located in the 5'UTR and the 3'UTR, respectively. Insertion element In some embodiments, the recombinant nucleic acid molecule comprises an insertion element, which can be used to regulate the transcription of the recombinant nucleic acid molecule, to regulate the translation of the circular RNA, to achieve the specific expression of the circular RNA between different tissues, or to purify the circular RNA, etc. Exemplarily, the insertion element is located between the coding element and the translation initiation element. In some embodiments, the insertion element is selected from at least one of the following groups (i)-(iii): (i) a transcriptional level regulatory element, (ii) a translational level regulatory element, (iii) a purification element. Exemplarily, the insertion element comprises a sequence of one or any combination of two or more of the following: an untranslated region (UTR) sequence, a polyN sequence, an aptamer sequence, a riboswitch sequence, a sequence that binds to a transcriptional regulatory factor; in the polyN sequence, N is selected from at least one of A, T, G, and C. In some alternative embodiments, the translation regulatory element comprises an untranslated region sequence, which can be used to regulate properties such as the stability, immunogenicity of the circular RNA, and the efficiency of the circular RNA to express the target polypeptide. The present disclosure does not specifically limit the untranslated region sequence, which can be selected from any type of sequence in the art that has properties such as regulating the transcription, translation, intracellular stability, and immunogenicity of the circular RNA. Further, the untranslated region sequence is not limited to the 5'UTR sequence or the 3'UTR sequence. In some alternative embodiments, the untranslated region sequence contains one or more miRNA recognition sequences, for example, 1, 2, 3, 4, 5, 6, 7, etc. By adding one or more miRNA recognition sequences, the specific expression of the circular RNA in different tissues and cells can be achieved, and the targeted delivery of the circular RNA molecule can be achieved. In some alternative embodiments, the translation regulatory element comprises a polyN sequence, where N can be at least one of A, T, G, and C. By increasing the translation regulatory element comprising the polyN sequence, the efficiency of expressing the target polypeptide by the circular RNA can be improved, the immunogenicity, stability, etc. can be improved, or it can be used for the purification of circular RNA. The present disclosure does not specifically limit the length of the polyN sequence, the selection types of N in the polyN sequence, and the composition manner, as long as it is beneficial to improving the performance of circular RNA. Exemplarily, the polyN sequence is a polyA sequence, a polyAC sequence, etc. In some alternative embodiments, the translation regulatory element comprises a riboswitch sequence. The riboswitch sequence is a class of untranslated sequences that have regulatory functions on the transcription and translation of RNA. In the present disclosure, the riboswitch sequence can affect the expression of circular RNA, including but not limited to transcription termination, translation initiation inhibition, mRNA self-cleavage, and alteration of splicing pathways in eukaryotes. In addition, the riboswitch sequence can also control the expression of circular RNA by triggering the binding or removal of molecules. Exemplarily, the riboswitch sequence is a cobalamin riboswitch (also called B12-element), an FMN riboswitch (also called RFN element), a glmS riboswitch, a SAM riboswitch, a SAH riboswitch, a tetrahydrofolate riboswitch, a Moco riboswitch, etc. The present disclosure does not restrictively limit the type and sequence of the riboswitch sequence, as long as it can achieve the regulation of the transcription and translation levels of the target polypeptide expressed by circular RNA. In some alternative embodiments, the translation regulatory element comprises an aptamer sequence. In the present disclosure, the aptamer sequence can be used to regulate the transcription and translation of circular RNA, or for the in vitro purification and preparation of circular RNA. A recombinant expression vector comprising a recombinant nucleic acid molecule As used in the present disclosure, a "vector" refers to a piece of DNA that is synthetic (e.g., using PCR) or extracted from a virus, plasmid, or cells of a higher organism, into which an exogenous DNA fragment can be inserted or has been inserted for cloning and / or expression purposes. In certain embodiments, the vector can be stably maintained in an organism. The vector can include, for example, an origin of replication, a selectable marker or a reporter gene, such as antibiotic resistance or GFP, and / or a multiple cloning site (MCS). The term includes linear DNA fragments (e.g., PCR products, linear plasmid fragments), plasmid vectors, viral vectors, cosmids, bacterial artificial chromosomes (BACs), yeast artificial chromosomes (YACs), etc. In some embodiments, the vector allows the production of translatable and / or biologically active circular RNA in eukaryotic cells. In some embodiments, the recombinant nucleic acid molecule exists as part of a recombinant expression vector for preparing circular RNA. During in vitro transcription and cyclization processes, circular RNA expressing a target polypeptide can be prepared. In some other embodiments, the recombinant nucleic acid molecule can also exist as a pre-circularized RNA molecule or a part thereof obtained after linearization treatment and transcription reaction of the recombinant expression vector. That is, the recombinant nucleic acid molecule only needs to undergo a cyclization reaction to obtain circular RNA expressing the target polypeptide. In some embodiments, the steps for preparing circular RNA in vitro include: Transcription step: The recombinant nucleic acid molecule as described in the present disclosure or the recombinant expression vector according to the present disclosure is transcribed to form a pre-circularized nucleic acid molecule; Cyclization step: The pre-circularized nucleic acid undergoes a cyclization reaction to obtain circular RNA. In some alternative embodiments, the method further includes a step of purifying the circular RNA. Circular RNA or circRNA In the present disclosure, circular RNA or circRNA have the same meaning, both referring to RNA circular molecules without "scar" obtained by the method of the present application. Target polypeptide The present disclosure does not restrictively define the types of target polypeptides, which can be human proteins or non-human proteins. Exemplarily, the target polypeptide includes but is not limited to antigens, antibodies, antigen-binding fragments, fluorescent proteins, proteins with disease treatment activity, proteins with gene editing activity, etc. In the present disclosure, the term "antibody" is used in the broadest sense, referring to a protein containing an antigen-binding site, covering natural antibodies and artificial antibodies of various structures, including but not limited to monoclonal antibodies, polyclonal antibodies, multispecific antibodies (e.g., bispecific antibodies), single-chain antibodies, intact antibodies, and antibody fragments. In the present disclosure, the term "antigen-binding fragment" is a part or segment of a complete or full antibody with fewer amino acid residues than the complete or full antibody, which can bind an antigen or compete with the complete antibody (i.e., the complete antibody from which the antigen-binding fragment is derived) for binding to the antigen. Antigen-binding fragments can be prepared by recombinant DNA technology, or by enzymatic or chemical cleavage of the complete antibody. Antigen-binding fragments include but are not limited to Fv, Fab, Fab’, Fab’-SH, F(ab')2; diabodies; linear antibodies; single-chain antibodies (e.g., scFv); single-domain antibodies; bivalent or bispecific antibodies or fragments thereof; camelid antibodies (heavy-chain antibodies); and bispecific or multispecific antibodies formed by antibody fragments. In the present disclosure, proteins having disease treatment activity may include, but are not limited to, enzyme replacement proteins, proteins for supplementation, protein vaccines, antigens (such as tumor antigens, viruses, bacteria), hormones, cytokines, antibodies, immunotherapies (such as for cancer), cell reprogramming / transdifferentiation factors, transcription factors, chimeric antigen receptors, transposases or nucleases, immune effectors (such as those affecting susceptibility to immune responses / signaling), regulated death effector proteins (such as inducers of apoptosis or necrosis), non-lytic inhibitors of tumors (such as oncoprotein inhibitors), epigenetic modifiers, epigenetic enzymes, transcription factors, DNA or protein modification enzymes, DNA intercalators, efflux pump inhibitors, nuclear receptor activators or inhibitors, proteasome inhibitors, enzyme competitive inhibitors, protein synthesis effectors or inhibitors, nucleases, protein fragments or domains, ligands or receptors, and CRISPR systems or their components, etc. Coding element comprising tandem coding regions In some embodiments, the coding element in the recombinant nucleic acid molecule comprises the coding element comprising at least one coding region. Exemplarily, the coding element comprises 1, 2, 3, 4, 5, 10, 15, 20, 25, and so on. In some alternative embodiments, optionally, the coding element comprises at least two coding regions, and each coding region independently encodes any type of target polypeptide. In some alternative embodiments, the coding element comprises 2 or more coding regions. The recombinant nucleic acid molecule comprises elements arranged in the following order along the 5' to 3' direction: intron fragment II, translation initiation element truncated fragment II, at least 2 coding regions, translation initiation element truncated fragment I, intron fragment I. In some other alternative embodiments, the recombinant nucleic acid molecule consists of the elements arranged in the above order. In some preferred embodiments, the coding element further comprises a linker located between any two adjacent coding regions. The linker separates adjacent coding regions, enabling the circular RNA prepared from the recombinant nucleic acid molecule to achieve the expression of 2 or more target polypeptides. In the present disclosure, the linker may be a polynucleotide encoding a 2A peptide, or other types of polynucleotides encoding a linker peptide for spacing target polypeptides. Among them, the 2A peptide is a short peptide (~18 - 25 amino acids) derived from a virus, and they are usually referred to as "self-cleaving" peptides, which can generate multiple proteins from one transcript. Exemplarily, the 2A peptide is P2A, T2A, E2A, F2A, and so on. In some embodiments, the coding element comprises at least two coding regions, wherein a translation initiation element is connected between any two adjacent coding regions. Exemplarily, the coding element comprises 1, 2, 3, 4, 5, 10, 15, 20, 25, etc. Moreover, within the coding element, a translation initiation element is disposed between any two adjacent coding regions. By the above method, a translation initiation element can be connected upstream of each coding region in the circular RNA prepared in vitro of the recombinant nucleic acid molecule, so as to achieve the expression of two or more target polypeptides. In some alternative embodiments, the coding element comprises two or more coding regions. The recombinant nucleic acid molecule comprises elements arranged in the following order along the 5'-to-3' direction: intron fragment II, truncated translation initiation element II, at least two coding regions, truncated translation initiation element I, intron fragment I. Wherein, a translation initiation element is connected between any two adjacent coding regions. In some other alternative embodiments, the recombinant nucleic acid molecule is composed of the elements arranged in the above order. In the present disclosure, the translation initiation element can be any type of element capable of initiating the translation of the target polypeptide. In some embodiments, the translation initiation element is an element comprising any one or more than two of the following sequences: IRES sequence, 5'UTR sequence, Kozak sequence, sequence containing m6A modification (N(6)-methyladenosine modification), complementary sequence of ribosomal 18S rRNA. In some other embodiments, the translation initiation element can also be any other type of cap-independent translation initiation element. In the present disclosure, each coding region comprised by the coding element independently encodes any type of target polypeptide. Wherein, the target polypeptides encoded by any two coding regions can be the same or different. In the circular RNA prepared by using the above recombinant nucleic acid molecule, a translation initiation element is correspondingly connected to the 5'-end of each coding region, and multiple coding regions are connected in series by multiple translation initiation elements, so as to achieve the expression of at least two target polypeptides. Drug composition / Administration In the embodiments of the present disclosure, the circRNA products described and / or produced by using the vectors and / or methods described in the present disclosure can be provided in the form of a composition (such as a drug composition). Accordingly, in certain embodiments, the present disclosure also relates to compositions, such as compositions comprising a circRNA (circRNA product) and a pharmaceutically acceptable carrier. In one aspect, the present disclosure provides a pharmaceutical composition comprising an effective amount of the circRNA described herein and a pharmaceutically acceptable excipient. The pharmaceutical compositions of the present disclosure may comprise the circRNA described herein, as well as one or more pharmaceutically or physiologically acceptable carriers, excipients, or diluents. In certain embodiments, the pharmaceutical compositions of the present disclosure may comprise circRNA-expressing cells, e.g., a variety of circRNA-expressing cells as described herein, in combination with one or more pharmaceutically or physiologically acceptable carriers, excipients, or diluents. In certain embodiments, a pharmaceutically acceptable carrier can be a component of a pharmaceutical composition other than the active ingredient that is non-toxic to a subject. Pharmaceutically acceptable carriers can include, but are not limited to, buffers, excipients, stabilizers, or preservatives. Examples of pharmaceutically acceptable carriers are physiologically compatible solvents, dispersion media, coatings, antibacterial and antifungal agents, isotonic agents, and absorption delaying agents, etc., such as salts, buffers, sugars, antioxidants, aqueous or non-aqueous carriers, preservatives, wetting agents, surfactants, or emulsifying agents, or combinations thereof. The amount of the pharmaceutically acceptable carrier in a pharmaceutical composition can be determined experimentally based on the activity of the carrier and the desired properties of the formulation, such as stability and / or minimal oxidation. In certain embodiments, such compositions may comprise buffers, such as acetic acid, citric acid, histidine, boric acid, formic acid, succinic acid, phosphoric acid, carbonic acid, malic acid, aspartic acid, Tris buffer, HEPPSO, HEPES, neutral buffered saline, phosphate buffered saline, etc.; carbohydrates, such as glucose, sucrose, mannose, or dextran, mannitol; proteins; polypeptides or amino acids, such as glycine; antioxidants; chelating agents, such as EDTA or glutathione; adjuvants (e.g., aluminum hydroxide); antibacterial and antifungal agents; and preservatives. In certain embodiments, the compositions of the present disclosure can be formulated for a variety of modes of parenteral or non-parenteral administration. In one embodiment, the composition can be formulated for infusion or intravenous administration. The compositions disclosed herein can be provided, for example, as a sterile liquid preparation, such as an isotonic aqueous solution, emulsion, suspension, dispersion, or viscous composition, which can be buffered to the desired pH. Formulations suitable for oral administration can include liquid solutions, capsules, sachets, tablets, lozenges, and troches, liquid suspensions, powders, and emulsions in a suitable liquid. As used herein, the terms "complementary" or "hybridizing" are used to refer to "polynucleotides" and "oligonucleotides" (which are interchangeable terms referring to nucleotide sequences) related to the base pairing rules. For example, the sequence "CAGT" is complementary to the sequence "GTCA". Complementary or hybridization can be "partial" or "complete". "Partial" complementarity or hybridization means that one or more nucleic acid bases are mismatched according to the base pairing rules, and "total" or "complete" complementarity or hybridization between nucleic acids means that each nucleic acid base is paired with another base under the base pairing rules. The degree of complementarity or hybridization between nucleic acid strands has an important impact on the hybridization efficiency and strength between nucleic acid strands. This is particularly important in amplification reactions and detection methods that depend on the binding between nucleic acids. The term "recombinant nucleic acid molecule" refers to a polynucleotide having sequences that are not linked together in nature. The recombinant polynucleotide can be included in a suitable vector, and the vector can be used to transform into a suitable host cell. Then the polynucleotide is expressed in the recombinant host cell to produce, for example, "recombinant polypeptide", "recombinant protein", "fusion protein", etc.; RNA molecules can also be obtained by reverse transcription in vivo or in vitro. The term "recombinant expression vector" refers to a DNA construct used to express, for example, a polynucleotide encoding a desired polypeptide. The recombinant expression vector can include, for example, a transcriptional subunit containing i) a set of genetic elements that regulate gene expression, such as promoters and enhancers; ii) a structure or coding sequence that is transcribed into mRNA and translated into a protein; and iii) appropriate transcriptional and translational start and stop sequences. The recombinant expression vector is constructed in any suitable manner. The nature of the vector is not important, and any vector can be used, including plasmids, viruses, phages, and transposons. Possible vectors for the present disclosure include, but are not limited to, chromosomal, non-chromosomal, and synthetic DNA sequences, such as viral plasmids, bacterial plasmids, phage DNA, yeast plasmids, and vectors derived from combinations of plasmid and phage DNA, DNA from viruses such as lentivirus, retrovirus, vaccinia, adenovirus, fowlpox, baculovirus, SV40, and pseudorabies. The term "host cell" refers to a cell into which an exogenous polynucleotide has been introduced, including progeny of such a cell. Host cells include "transformants" and "transformed cells", which include primary transformed cells and progeny derived therefrom. A host cell is any type of cell system that can be used to produce the antibody molecules of the present invention, including eukaryotic cells, such as mammalian cells, insect cells, yeast cells; and prokaryotic cells, such as Escherichia coli cells. Host cells include cultured cells, and also include cells inside transgenic animals, transgenic plants, or cultured plant tissues or animal tissues. The term "recombinant host cell" covers a host cell that is different from the parental cell after introduction of a recombinant nucleic acid molecule, a recombinant expression vector, or a circular RNA. A recombinant host cell is specifically achieved by transformation. The host cells of the present disclosure can be prokaryotic cells or eukaryotic cells, as long as they are cells capable of introducing the recombinant nucleic acid molecules, recombinant expression vectors, circular RNAs, etc. of the present disclosure. As used herein, the terms "individual", "patient", or "subject" include mammals. Mammals include, but are not limited to, domestic animals (such as cows, sheep, cats, dogs, and horses), primates (such as humans and non-human primates like monkeys), rabbits, and rodents (such as mice and rats). As used herein, the terms "transformation, transfection, transduction" have the meanings generally understood by those skilled in the art, which refer to the process of introducing exogenous DNA into a host. The methods of said transformation, transfection, transduction include any method of introducing nucleic acids into cells, and these methods include, but are not limited to, electroporation, calcium phosphate (CaPO4) precipitation, calcium chloride (CaCl2) precipitation, microinjection, polyethylene glycol (PEG) method, DEAE-dextran method, cationic liposome method, and lithium acetate-DMSO method. As used herein, "treatment" means: after a subject has contracted a disease, contacting (such as administering) the circular RNA, circular precursor RNA, composition, etc. of the present invention to the subject, so that the symptoms of the disease are alleviated compared to when not contacted, and it does not necessarily mean complete suppression of the symptoms of the disease. Contracting a disease means: the body shows symptoms of the disease. As used herein, the term "effective amount" refers to such an amount or dose of the recombinant nucleic acid molecule, recombinant expression vector, circular precursor RNA, circular RNA, vaccine, or composition of the present invention that, when administered to a patient in a single or multiple doses, produces the desired effect in a patient in need of treatment or prevention. The effective amount can be easily determined by an attending physician who is skilled in the art by considering various factors such as: the species of the mammal; its size, age, and general health; the specific disease involved; the degree or severity of the disease; the response of the individual patient; the specific antibody administered; the mode of administration; the bioavailability characteristics of the administered formulation; the selected dosing regimen; and the use of any concomitant therapies. As used herein, the terms "individual", "patient" or "subject" include mammals. Mammals include, but are not limited to, domestic animals (e.g., cows, sheep, cats, dogs, and horses), primates (e.g., humans and non-human primates such as monkeys), rabbits, and rodents (e.g., mice and rats). Unless otherwise defined or clearly indicated by the context, all technical and scientific terms used in this disclosure have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure pertains. Examples The sequence modification method used in this disclosure is basically based on the NEB kit as HiFi DNA Assembly Master Mix, and the operation is carried out according to the kit instructions. A small amount of plasmid construction is carried out through a small amount of sequence modification by the Gibson method, and the kit used is HiFi DNA Assembly Master Mix. The linearized template for in vitro transcription is prepared by digestion with the enzyme BspQI. After digestion, it is recovered by the Sangon PCR purification kit, and agarose gel electrophoresis or capillary electrophoresis is used to determine whether the digestion is complete. The in vitro transcription kit used for in vitro transcription includes T7 polymerase Mix, reaction buffer, and the four nucleotides AGCU, which are purchased from Shanghai Zhaowei Company. The original plasmid template is purchased from GenScript. The preparation of LNP is carried out by encapsulation with cationic liposomes, and expression verification is carried out in Hela cells or 293T cells. 100 ng of RNA is transfected into each well of a 96-well plate. The electrophoretic separation of circular RNA is carried out by E-gel EX 2% Agarose from Thermo Fisher. All RNAs used in this disclosure are prepared by in vitro transcription. The in vitro transcription templates used are purchased from GenScript, and the templates are linearized by BspQI. The in vitro transcription kit includes T7 polymerase Mix, reaction buffer, and the four nucleotides AGCU, which are purchased from Shanghai Zhaowei Company. Sequence modification is mainly based on NEB's Site-Directed Mutagenesis Kit, and a small amount of sequence modification is carried out by the Gibson method. The kit used is HiFi DNA Assembly Master Mix. The gel electrophoresis separation of circular RNA is carried out by E-gel EX 2% Agarose from Thermo Fisher. Example 1: Feasibility experiment of modifying IG sequence to achieve scar-free cyclization The main elements of the PIE method are shown in the upper part of Figure 1.a. By adding homology arms on both sides, the splicing sites are brought closer to each other, which is more conducive to splicing; by adding spacer sequences to the Intron and IRES sequences, the two sequences with complex secondary structures are separated to avoid mutual influence and thus improve the cyclization efficiency; the Intron sequences on both sides play the role of ribozymes, and splice and cyclize after recognizing the heterologous Exon sequence through the IG sequence in the Intron. Anabaena tRNA used in the present invention leu The IG sequence of the original type I ribozyme is 3nt in length, and these 3nt are used for recognition and splicing of splicing sites. Therefore, we first planned to modify the IG sequence so that it can recognize a sequence located in the ORF, and then perform the cyclization reaction independently of the heterologous Exon sequence. Firefly Luciferase (1653nt) was used as a reporter gene for the experiment here, and two designs were made. Design 1 first simulated the secondary structure and screened sites that were close to each other in space as splicing sites (Figure 1a). This feature is considered to be conducive to cyclization; Design 2 is based on Design 1 and adds the "GU" wobble pairing feature (Figure 1b). This feature is conserved in many other type I ribozymes [9] , so it is considered to be a more important feature for the shear reaction. After in vitro transcription and cyclization, it was found that design1 did not produce circular RNA, while design2 did. However, 2% EX-gel electrophoresis (Figure 1c) showed that compared with the PIE method, the circular RNA yield produced by design2 was very low. After being introduced into Hela cells for expression measurement (Figure 1d), it was found that it was only about 10% of the PIE method. In addition, reverse transcription PCR sequencing was performed on design2, and it was found that precise splicing was performed (Figure 1e). This proves that design2 does produce circular RNA and undergoes precise splicing, but the current yield is too low, which may be caused by the unclear molecular mechanism of ribozyme action, so it is planned to explore the molecular mechanism. We found that in the PIE system, polyAC spacer can well prevent Intron from forming a complex secondary structure with its adjacent sequences. This secondary structure may affect the formation of the natural secondary structure of the ribozyme or destroy the P1 structure, which is not conducive to splicing. Therefore, in subsequent experiments, in order to study the molecular mechanism, the constructed sequence is similar to the original PIE method, but the heterologous Exon sequence is removed. Example 2: Molecular mechanism of Anabaena type I ribozyme cyclization 2.1 The 3' splice site is not recognized and cannot base-pair with the nucleotides near the 5' splice site The structures of type I ribozymes are very conserved. Almost all of them have a segment of IG sequence to recognize the splice sites on both sides. The IG sequence forms P1 with the 5' splice site recognition and P10 with the 3' splice site recognition. However, compared with other type I ribozymes, the type I ribozyme from Anabaena has not been found to have P10. There is only a P1 structure formed by a 3-nt-long IG sequence and the 5' splice site, which seems to indicate that the type I ribozyme from Anabaena does not need to recognize the 3' splice site and only needs to be adjacent to it for splicing. However, there is controversy about this. Some researchers believe that if the type I ribozyme from Anabaena has P10, it may be the "AU" base pair

[0010] . In the previous design, only the IG sequence was modified, and the cyclization rate was too low. Therefore, it was speculated whether there is a recognition mechanism for the 3' splice site. So, first, the 5-nt sequence of the original Exon2, which serves as the 3' splice site in the Ana system, was retained for in vitro cyclization. This sequence is called "Trunc-Exon" (Figure 2a). This sequence was successfully cyclized, and the cyclization efficiency did not decrease compared with the PIE method. After RT-PCR and sequencing, it was found that this sequence also underwent correct splicing (Figure 2b). Then, single nucleotides were deleted from this 5-nt sequence. We found that when 4 nucleotides were deleted, the cyclization efficiency decreased significantly. When all 5 nucleotides were deleted, almost no circular RNA was produced (Figure 2c), and the cell expression experiment also reached the same conclusion (Figure 2d). Therefore, the 2-nt (AT) conserved region was first determined, which is consistent with the previous "AU" base pair conjecture However, to further determine whether the nucleotides in this 2-nt conserved region are "AT" conserved, on the 2-nt structure (Figure 3.a), 15 (2 4 -1) other possible 2-nt nucleotide combinations were constructed. The results showed that almost all 2-nt sequences could be successfully cyclized, and the cyclization efficiency did not decrease (Figure 3.c d). Only the cyclization efficiencies of "CG", "GG", and "TG" decreased significantly, which is the same as the result of Exon2Δ5 before. Because the 2-nt sequence of it as the 3' splice site is also "CG". And in the sequences we constructed, the 2-nt sequence of Exon2Δ5 as the 3' splice site is actually located in the 19-nt "GC"-rich internal homology. That is to say, the nucleotides that should have been the 3' splice site base-paired strongly with the nucleotides adjacent to the 5' splice site, which is not conducive to the subsequent splicing process and may be the reason for the decrease in cyclization efficiency To verify the above conjecture, we planned two experiments, one positive and one negative. We performed sequence modifications on Exon2-AA with a high cyclization rate and Exon2Δ5 that could not cyclize. We added the "TT" sequence to the 5' splice site to pair it with the "AA" of Exon2-AA. At the same time, we also added a segment of "AA" to the 5' splice site of Exon2Δ5 to disrupt the pairing characteristics of the original "CG" and "GC" (Figure 3.e). The results showed that for Exon2Δ5, when this homology was disrupted, cyclization immediately recovered and exhibited the same cyclization efficiency as the original sequence. In contrast, the originally normally cyclizable Exon-AA lost its cyclization ability completely after mutation (Figure 3.fg). This experiment demonstrated that for the Ana system, the 3' splice site is non-conservative and there is no other sequence for recognition. However, when there is a pairing characteristic between the 3' splice site and the adjacent sequence of the 5' splice site, the cyclization efficiency will be greatly reduced. This may be because when the 3' splice site forms such a base complementary characteristic, a relatively tight secondary structure will be formed, making the second splicing process less likely to occur. 2.2 The IG sequence is partially conserved and recognizes the 5' splice site through base pairing or "GU" wobble pairing After a detailed study of the 3' splice site, we then studied the 5' splice site. The 5' splice site is recognized by the "GAG" of the IG sequence of Aha, and the original Exon1 sequence is "CTT". The Exon1 sequence was sorted from 1 to 3 in the 5'-3' order (Figure 4a). First, in the first experiment, with the IG sequence unchanged, we mutated its recognition of the Exon1 sequence. For example, the "G" at the first position recognizes the "T" at the first position of Exon1, so we mutated the "T" into the other three nucleotides "A, G, C". This experiment was to explore the recognition mechanism of the IG sequence for the Exon1 sequence. The electrophoresis results (Figure 4.b) showed that the IG sequence recognizes the splice site through base complementary pairing and "GU" wobble pairing. The first and third nucleotides strictly follow this characteristic, while the second nucleotide does not fully comply with this rule. For example, the combination of "A-C" in the experiment can also be spliced and still retains a high cyclization efficiency. This may mean that the second position of the IG sequence is a non-conservative base position, and the cell expression experiment also reflects the same rule. The second experiment was a study on the conservation of the IG sequence. From the previous experiment, it was known that ensuring base complementary pairing, i.e., satisfying the recognition relationship, allows for normal cyclization. Then, when the IG sequence changes and base complementary pairing is simultaneously ensured, will it affect the cyclization process, that is, whether the IG sequence itself has a certain degree of conservation. Thus, a series of subsequent mutations were constructed. For example, the first nucleotide of the IG sequence is recognized through "G-U" wobble pairing. Then, this pair of nucleotides was modified to "C-G", "A-T", "T-A", and "G-C" to explore whether it would affect cyclization. The electrophoresis results showed (Figure 4.c) that when strictly following base complementary pairing, for the second base, the mutation of the IG sequence did not reduce the cyclization efficiency. Even when mutated to the base "GC", the cyclization efficiency increased. From the experimental results of the mutation of the first base, it can be seen that the modification of the IG sequence had little effect on the cyclization efficiency. Even if there was some decrease, it was not significant. However, the mutations of the IG sequence at the third base all led to a significant decrease in the cyclization efficiency. In this experiment, except for the mutation at the third base, where obvious differences could be seen in the electrophoresis pattern, the effects of mutations at other positions were not obvious enough in the electrophoresis results. Therefore, the cell expression experiment is crucial for characterizing this cyclization efficiency and can more clearly quantify the cyclization efficiency. In addition, "GU wobble pairing" is also a possible recognition rule during the recognition of the IG sequence. However, in the original sequence, "GU wobble pairing" only exists at the first base. Therefore, we studied the effect of "GU" wobble pairing on the cyclization efficiency at all other positions. Thus, the base pairs at different positions were also mutated. We found that when the "T-G" at the first position was mutated to "G-T", the impact on cyclization was huge, and almost no circular RNA was produced. However, such mutations at other positions could all undergo cyclization, but the cyclization efficiency decreased to varying degrees (Figure 4.d). In addition, we believe it is necessary to clarify that in the above research, we found that most mutations only reduced the cyclization efficiency, rather than completely eliminating the production of circular RNA. We believe there is an essential difference. To verify this theory, we constructed a set of mutant sequences, denoted as Mut1-4. This set of sequences did not follow the above recognition rules. Except for maintaining pairing at the non-conserved second base, the other two positions did not undergo base complementary pairing. After in vitro cyclization, we found that such sequences that did not follow the recognition rule did not produce any circular RNA at all (Figure 4.e), which is different from the results of reducing the cyclization efficiency in the above experiments. Therefore, we believe that when maintaining Watson-Crick base pairing or "G-U" wobble pairing, the decrease in cyclization efficiency is caused by the decrease in cleavage rate, which is different from the complete absence of circular RNA production when the recognition rules are not met. This decrease in cyclization efficiency should be optimized by exploring the optimal cyclization temperature, metal ion concentration, and cyclization reaction time. However, since nicking RNA is generated during cyclization, obviously a faster cleavage rate is more beneficial. Therefore, based on the above experiments, we believe that the IG sequence should satisfy the "GNN" feature and be recognized through Watson-Crick base pairing and partial "G-U" wobble pairing. 2.3 Sequence homology near the splicing site affects the cyclization effect The type I ribozyme system derived from Anabaena was previously studied and found that there is 5-nt homology in the adjacent sequences of the two splicing sites at its ends. This homology is derived from the original Exon of Anabaena (Figure 5a). The study pointed out

[0010] that this 5-nt homology has a great impact on cyclization efficiency, which may be due to the too short IG sequence of the Ana system. In other type I ribozyme systems, the median length of the IG sequence is 5 nt. The too short IG sequence leads to an unstable helix formed with the 5' splicing site and requires additional homologous sequences to stabilize it. Since this 5-nt homologous sequence is brought by the original heterologous Exon sequence, when scarless cyclization is carried out, the heterologous Exon is removed and this homologous sequence will also be deleted accordingly. In the research sequence we constructed, there is a 19-nt homology arm near the splicing site (Figure 5b), which may also be the reason why the previous experiments were not affected. After removing it, the results showed that no circular RNA was generated (Figure 5c). The experiment proved that the homology of the sequence near the splicing site is important for RNA cyclization. According to previous studies, the length of the homologous sequence of the original Exon is 5 nt. Therefore, homologous sequences with different lengths and different GC ratios were constructed on the basis of 5 nt (Figure 5d) to study their effects on cyclization efficiency. The results (Figure 5e,f) showed that when there is a 3-nt homology arm, the cyclization ability is restored and the cyclization efficiency is the same as that of the wild type. Longer homology arms such as 10 nt do not improve the cyclization ability of circular RNA, while when the homology arm is shortened to 2 nt, the cyclization efficiency is significantly reduced. And there is no obvious difference in cyclization efficiency among homologous arms of the same length with different GC ratios, which indicates that as long as homology is ensured, there is no sequence specificity. 2.4 Summarize the rules of scarless cyclization by modifying the IG sequence Based on the above experiments, we have explored in detail several principles that affect the cyclization efficiency of the Anabaena system in vitro. The following design principles for scarless cyclization are summarized: ① The IG sequence needs to satisfy the sequence feature of "GNN"; ② When the IG sequence recognizes the 5'splice site, it is necessary to ensure base complementary pairing or "GU" wobble pairing; ③ The 3'splice site is not conserved, but when it is base complementary paired with the sequence near the 5'splice site, it will greatly affect the cyclization efficiency; ④ Both sides of the splicing site need to have a homologous arm with a minimum length of 3nt to stabilize the P1 duplex The specific structure is shown in Figure 6 Example 3 Cyclize IRES+Fluc and IRES+eGFP according to the scarless cyclization strategy of modifying the IG sequence According to the design principles summarized from the previous exploration of the molecular mechanism, first find a suitable cyclization site in Firefly Luciferase, and then modify the IG sequence to construct a circular RNA containing only "IRES+Fluc" (Figure 7a). First, a sequence of "TGGTGCCTTTTCACCA" from positions 719-734 was found in Fluc. Among them, "CCT" can be used as the 5'splice site, and the IG sequence was modified to "GGG" for recognition; "TT" was used as the 3'splice site, and there was no obvious base complementary pairing with the nucleotides near the 5'splice site; and the "TGGTG" and "CACCA" on both sides could form a homologous sequence with a length of 5nt (Figure 7b). This sequence was successfully cyclized in vitro, and the cyclization efficiency did not decrease (Figure 7c). Moreover, through in vitro reverse transcription PCR sequencing, it was found that precise splicing occurred (Figure 7e). However, perhaps due to the lack of the polyAC spacer sequence, the translation efficiency decreased slightly (Figure 7d). To verify the universality of this modification method, it was planned to perform the same modification on eGFP to construct a circular RNA of "IRES+eGFP". A sequence from positions 61 to 76 in the ORF of eGFP that also met the modification rules was selected (Figure 8a). With "CGT" as the 5' splicing site, the IG sequence was modified to "GCG". The results showed that circular "IRES+eGFP" was also obtained in vitro, and the cyclization efficiency was the same as that of PIE (Figure 8b). The sequencing results also showed precise cleavage (Figure 8c). The cell expression experiment also proved that the obtained circular RNA could express eGFP in cells. However, similarly, compared with the circular RNA of the PIE method with polyA, the translation was downregulated. It is possible that the spacer sequence is very important for enhancing the translation ability of circular RNA. According to the above experiments, it was proven that circular RNA without "scar" could be successfully prepared in vitro by the above method, and the cyclization efficiency was comparable to that of the PIE method, and large-fragment circular RNA could be prepared with high efficiency. However, for different ORFs, suitable splicing sites need to be selected, and the secondary structure of the cyclized fragment may affect the function of ribozymes, which has been mentioned in previous literature. [7] , which is obviously inconvenient. In addition, in our previous study, it was found that certain spacer sequences could improve the translation ability of circular RNA, and this was also confirmed in the study by Howard Chang.

[0011] , so it was planned to screen some spacer sequences that are helpful for the translation of circular RNA. Since they have the function of enhancing translation, which is similar to the UTR function of linear mRNA, they are hereafter referred to as UTR sequences. Then, the IG sequence was modified according to the UTR sequence to construct a universal cyclization method of "IRES+ORF+UTR" without "scar". Example 4: Screening of UTR sequences and construction of a universal cyclization method of "IRES+ORF+UTR" After studying the mechanism of "scar"-free cyclization in the experiment of modifying the IG sequence, it was planned to screen UTR sequences with enhanced translation effects, modify the IG sequence to recognize the UTR region, and then construct a universal "scar"-free system that does not require re-design for different cyclization sequences. The UTR sequence has two functions. First, it can separate the intron sequence from the IRES. The secondary structures of these two sequences are both relatively complex, and adjacent IRES will greatly affect the catalytic function of ribozymes. The second function is to improve the translation effect of circular RNA. Based on the above principle, the 5'UTR was screened first, and three types of sequences were selected. The first type is the IRES enhancer sequence: IRES binds to 18s rRNA, eIF4G, and eIF4A, thereby initiating translation. If this binding can be enhanced, the translation efficiency can be enhanced. Some researchers

[0012] found that there is a 9nt sequence in the 5’UTR of the homeodomain protein Gtx, and this sequence is completely paired with 1132-1124 of 18s rRNA. The article also pointed out that the tandem repeat of the 9nt sequence can greatly enhance the translation efficiency of IRES. Therefore, the 9nt sequence was repeated 6 times and denoted as "Gtx54" (54nt); in addition, the author also found in subsequent experiments

[0013] that the 7nt sequence in this 9nt sequence has a more obvious enhancing effect on the translation effect. Similarly, after repeated combination, the enhancing effect on the translation effect is more significant. Therefore, this sequence was repeated 7 times and denoted as "Gtx56" (6nt); some researchers also found a short "IRES-like" sequence

[0014] that can function as IRES to initiate translation and is denoted as "OR4F17" (56nt); some researchers also found an aptamer sequence of eIF4G, which can recruit eIF4G to improve expression. Therefore, this sequence was designed as a UTR sequence and denoted as "Apt-eIF4G" (51nt)

[0015] . The second type is the Protein-binding sequence. Some studies have found that compared with the traditional polyA, inserting cytosine at the tail of polyA will enhance protein expression. This sequence can recruit polyA-binding protein more effectively, thereby achieving the effect of enhancing translation. According to the research content of the literature, the two-terminal sequences 31A8CA (40nt) and 41A8CA (50nt) were selected for research; the third type is the full-length mRNA UTR sequence. Previous studies have screened 3 UTR sequences with obvious enhancing effects on mRNA from 12,000 sequences by high-throughput screening method. Therefore, the enhancing effects of these three sequences on circular RNA were experimentally studied and denoted as NeoUTR1 (100ntt), NeoUTR2 (100nt), and NeoUTR3 (100ntt). In total, 9 sequences were constructed above. In addition, a sequence with a GC ratio of 50% and a length of 50nt was constructed and denoted as the random sequence as a control; the original sequence is a polyAC sequence and is denoted as "polyAC". After sequence construction, the effects of these sequences on circularization were analyzed by EX-gel 2% electrophoresis. The gel image shows that these sequences have no effect on the circularization efficiency. Through cell expression experiments, it can be seen that 31A8CA and 41A8CA have obvious enhancing effects on translation (Figure 9a). After screening the 5'UTR, the 3'UTR was screened immediately. For the 3'UTR, three sequences with good effects in the 5'UTR were first selected: OR4F17, 15A8CA, and NeoUTR; then a protein-binding sequence, ARE, which can bind to the HuR protein and thus enhance the translation effect, was selected; finally, five 3'UTR sequences of highly abundant proteins in mammalian bodies were selected, including GADPH, ɑ-globin, β-globin, CYBA, and DECR1. Similarly, a random fragment with a 50% GC content was used as a control, and the original sequence was also "polyAC". The effects of these sequences on circularization were analyzed by EX-gel 2% electrophoresis. The gel diagram showed that these sequences had no effect on the circularization efficiency. Through cell expression experiments, it was found that the 15A8CA and α-globin sequences had a significant improvement in the translation effect (Figure 9b). For the two sequences, 41A8CA and 15A8CA, it was pointed out in previous reports that increasing the number of A could improve their enhancement effect on translation. Therefore, the number of A was increased in gradients, and a sequence was designed every 10 A, with a total of 9 sequences. The results showed that in the 5'UTR, as the number of A increased, the translation first increased and then decreased. Among them, 61A8CA had the most significant enhancement effect on translation. In the 3'UTR, 3'71A8CA showed the strongest enhancement effect (Figure 9.c). Finally, the screened 5’61A8CA was combined with 3'71A8CA, and the constructed sequence showed the strongest translation effect, which was nearly 3.7 times higher than the original sequence (Figure 9.d). According to the transformation principle of previous experiments, the UTR sequence and the IG sequence were transformed to make it recognize the UTR sequences at both ends. "CCC" was inserted at the third nucleotide at the 5’ end of the 5'UTR. The two "AA" in 5'61A8CA were used as the 3’ splicing site. "GGGCTT", six nucleotides, were inserted at the 3’ end of the 3'UTR. Among them, "GGG" and "CCC" at the 5’ end formed the required shortest homology, and "CTT" was recognized by the "GAG" IG sequence of the ribozyme as the 5’ splicing site to complete splicing. According to this design, the original 108nt heterologous fragment, including Exon and internal homology, could be removed, and only 9nt of the additional designed sequence was needed to complete circularization, constructing an "IRES+ORF+UTR" scarless circularization system (Figure 9e), and it was proved by 2% EX-gel separation that the circularization efficiency was the same as that of the PIE method (Figure 9.f). When this sequence was introduced into Balb / C mice, the results showed that, compared with the original sequence, the translation effect was also significantly enhanced (Figure 9g). Example 5 Modifying IRES to Enhance the Translation Effect of Circular RNA Design DNA sequence: Using PUC57-Kan as the plasmid vector, XbaLI and PciI were selected as the restriction enzyme sites on both sides of the target gene. The target gene sequence is successively the T7 promoter sequence, homology arm (HA), intron, polyAC sequence, CVB3 IRES sequence, optimized luciferase sequence, polyAC sequence, intron, and homology arm (HA). 1. Optimal position of the IRES enhancer element: The present disclosure designed 11 positions for inserting the gene sequence. The insertion positions of the IRES enhancer element are shown in Figure 10. The experimental results found that there were a total of 4 positions with translational effects. The IRES enhancer element was inserted between domain I and domain II, named IRES-1. The IRES enhancer element was inserted into the stem-loop structure of domain II, named IRES-2. The IRES enhancer element was inserted into the stem-loop structure (Distal loop) of domain IV, named IRES-3. The IRES enhancer element was inserted into the stem-loop structure (Proximal loop) of domain IV, named IRES-4. The plasmid without the inserted IRES enhancer element was named CVB3 IRES. 2. IRES enhancer elements screened from the natural gene pool: Four sequences were selected from the natural gene pool. Except that the DNA sequence of the IRES enhancer element was different from that of IRES-1, the other genes on the plasmid were the same as those of IRES-1. The four sequences were successively named IRES A1-4 [17-20] , and the specific primer sequences are shown in Table 3. The specific sequences of IRES A1-4 are shown in Table 4. 3. Exploration of the repeated design of the IRES enhancer element: The screened IRES enhancer elements were repeatedly designed to explore the optimal translation efficiency. 4. The repeated sequences were successively named IRES-Xn (n: number of repetitions), and the specific sequences are shown in Table 5. II. Insertion of the IRES enhancer element: Primers were designed, and then the plasmid with the inserted IRES enhancer element was obtained by PCR. The primer sequences for obtaining the vector by PCR are shown in Table 1, and the PCR reaction parameters are shown in Table 2. Table 1: Table 2: III. Purification of the PCR reaction: Using a PCR reaction purification kit, according to the operating steps of the instruction manual, the purified vector and IRES enhancer element were obtained. The DNA was qualitatively and quantitatively analyzed using an ultra-micro ultraviolet spectrophotometer, and agarose gel electrophoresis was performed to verify the integrity and purity of the DNA template. IV. Transformation of the PCR product into competent cells (1) The E. coli DH5α competent cells were thawed on ice before use. (2) 2 μL of PCR product was added to 50 μL of the cells, and the tube wall was flicked gently to mix evenly. (4) Incubated on ice for 30 min, quickly placed in a 42 °C water bath for heat shock for 45 s, and then quickly transferred to ice and left to stand for 2 min. (5) SOC medium was added to make up the volume to 1 mL. (6) Cultured with shaking at 37 °C for 1 hour. (7) Concentration: Centrifuged at 5000×g for 1 min, 900 μL of the supernatant was discarded, and the remaining part was pipetted and mixed evenly. (8) 100 μL of the bacterial solution was spread on an LB plate with 100 mg / L Kan resistance, and placed upright at 37 °C until the bacterial solution was absorbed. (9) Incubated overnight at 37 °C in an inverted position. (10) The next day, single colonies were picked from the plate and inoculated into LB medium with Kan resistance, and cultured with sufficient shaking at 37 °C on a shaker for 12 - 16 hours. V. In vitro transcription of circRNA: The plasmid was extracted, and the linearized DNA template was digested with enzymes. Using an in vitro transcription kit, transcription raw materials such as RNA polymerase, ribonucleoside triphosphates, and DNA template were added in vitro, incubated at 37 °C for 2 h, nuclease-free water and GTP were added, and incubated at 55 °C for 30 min to transcribe circRNA using DNA as the template. VI. Verification by precast gel electrophoresis: To detect the circularization efficiency of the synthesized circRNA, the precast gel electrophoresis pattern is shown in Figure 11. VII. Preparation of liposomes: Liposomes were prepared according to the following ratio. The molar ratio of cationic liposome:helper lipid:cholesterol:polyethylene glycol was 50:10:38.5:1.5. The cationic liposome was SM-102, the helper lipid was dioleoyl phosphatidylethanolamine, abbreviated as DOPE, and the polyethylene glycol was dimyristoyl glycerol-polyethylene glycol 2000, abbreviated as DMG-PEG 2000. VIII. Synthesis of lipid nanoparticles: According to the volume ratio of liposome to mRNA being 3:1, the required volume of nanoparticles was prepared, gently shaken and mixed evenly to obtain the lipid nanoparticles required for transfecting cells. IX. Cell transfection: One day before transfection, plate the cells and observe their status. After the cell density reaches approximately 70%-90%, transfect the synthesized circRNA containing the insertion position of the IRES enhancer element into Hela and HEK-293T cells; transfect IRES-A1-X2, IRES-A1-X5, IRES-A2-X2, IRES-A3-X1, and IRES-A4-X1 circRNA into Hela and HEK-293T cells. After transfection, gently shake and mix, and then culture the cells in an incubator. X. Detect the fluorescence intensity after 24 h: After 24 h of transfection, take out the well plate, equilibrate it to room temperature, add the luciferin substrate, and allow the cells to lyse fully for 10 minutes. Then, detect the luminescence signal using a microplate reader. The results of exploring the optimal insertion position of the IRES enhancer element, the transfection results of IRES A1-4, and the transfection results of IRES-A1-X2, IRES-A1-X5, IRES-A2-X2, IRES-A3-X2, and IRES-A4-X2 are shown in Figure 12. In addition, the present disclosure also modifies based on the CVB3 IRES sequence. First, the present disclosure explores the optimal position for inserting the IRES enhancer element into the CVB3 IRES sequence. The present disclosure designs 11 positions for inserting gene sequences, and it is experimentally found that there are 4 insertion positions with translation effects. The 4 specific insertion positions are respectively: the IRES enhancer element is inserted between domain I and domain II, named IRES-1; the IRES enhancer element is inserted into the stem-loop structure of domain II, named IRES-2; the IRES enhancer element is inserted into the stem-loop structure (Distal loop) of domain IV, named IRES-3; the IRES enhancer element is inserted into the stem-loop structure (Proximal loop) of domain IV, named IRES-4. The IRES enhancer element can enhance the binding of the CVB3 IRES sequence to ribosomes, thereby promoting the translation of circRNA. The IRES enhancer element exists in circRNA, folds into a structure similar to the initial tRNA, recruits more ribosomes, and binds translation regulatory factors such as ITAF, and then introduces the ribosome into the interior of circRNA to bind and initiate protein translation. Ribosomes are highly diverse protein structures found in all cells. Under the action of translation regulatory factors such as ITAF, the ribosome is introduced into the internal structure of circRNA to bind and initiate. Secondly, the present disclosure screens 4 IRES enhancer elements, and then explores the influence of IRES enhancer elements with different sequences having the same enhancement effect on the translation effect of cicRNA. These 4 IRES enhancer elements all have the function of recruiting ribosomes, thus affecting the translation of circRNA. The 4 screened IRES enhancer element sequences are respectively cloned into a DNA template, and the composition of the DNA template is the same as above, and the optimal sequence is screened to obtain a gene sequence with high protein expression. Finally, based on the optimal position of the IRES enhancer element and the screening of the IRES enhancer element, a series of repetitive sequence designs are carried out on the IRES enhancer element to obtain a sequence for the IRES enhancer element to achieve its function. Compared with the original sequence, IRES-1-A1-X2 in the screened sequence is increased by about 2.3 times. The IRES enhancer element is: or a repetitive sequence of the above sequence, for example: Example Six Development of Circular RNA Purification Process There are relevant literature reports on the application of arginine in the purification process of biological products. The main mechanism of action of arginine is: (1) it can inhibit the formation of nucleic acid aggregates and promote the dissociation of complexes; (2) the addition of arginine increases the overall hydrophobicity of the solution, thereby promoting the hydrophobic interaction binding between the aggregate-nucleic acid complex and the affinity probe, and reducing the non-specific adsorption of the affinity chromatography packing material.

[0021] . Related literature has reported the application of PEG in the purification process of biological products. In addition to forming missing zones around solutes, studies have found that PEG is also "repelled" from hydrophilic surfaces, such as most chromatography media: adding PEG to molecular sieve mobile phases will cause retention time to be delayed, and will also enhance the binding of substances to ion exchange and affinity chromatography fillers.

[0022] . By adding a certain concentration of arginine and PEG to the buffer during ultrafiltration and purification, the separation of the collected peaks was improved and the non-specific adsorption of nucleic acids was reduced, so that the purity and total yield of cRNA met the requirements of process scale-up. The circular nucleic acid purification process flow chart is shown in Figure 13. The selection and precision of IVT reaction equipment determines the yield of IVT in vitro reaction. Commonly used IVT reaction systems include constant temperature shaking metal bath, reactor, temperature control tank and waver shaker, etc. Common brands of constant temperature oscillating metal baths include Thermo Fisher (model: Thermo Scientific Compact), etc., reactors include Mettler reactors (models EasyMax102 and EasyMax402) and Radeys (model Mya4), etc., temperature control tanks include Le Pure and Bailinke water-cooled temperature control Le Xiaobao (model compatible with 1000mL and 500mL disposable PC square bottles, magnetic stirring, customized), etc., and Waver shakers include Satorius, Le Pure and Bailinke (model Waver10), etc. The reaction conditions of the constant temperature oscillating metal bath, reactor, and temperature controlled tank instrument are controlled as follows: The reaction conditions of Waver 10 shaker are controlled as follows: The circular nucleic acid IVT reaction system is template DNA, T7 enzyme, dNTP, pyrophosphatase (PPI) and PEG, etc. The template DNA homogeneity is required to be ≥95%, and the metal residue is less than 10PPM; the T7 enzyme brand is required to be Promega, Thermo and Kaika; PEG is PGE2000-8000 from Thermo or Sigma, which is used to increase the viscosity of the IVT reaction system and reduce the fluidity of the reaction solution, so as to prevent Mg from accumulating during the IVT reaction. 2+ Randomly attack circular nucleic acid base sequences to control the production of naked RNA. IVT reaction conditions are as follows: Circularization rate ≥ 75%, Naked RNA ≤ 3%: The cyclization steps are as follows: Step 1: Ultrafiltration buffer exchange using membrane cartridges or hollow fiber columns The ultrafiltration membrane cartridges are made of PES with pore sizes ranging from 100KD, 300KD, 500KD, 750KD to 1000KD The hollow fiber columns are made of PES with pore sizes ranging from 100KD, 300KD, 500KD, 750KD to 1000KD Manufacturers of ultrafiltration membrane cartridges and hollow fiber columns include Cytiva, Sartorius, Kebite, and Esfabo, etc. Filtration through ultrafiltration or hollow fiber columns can remove impurities such as T7 enzyme, DNA residues, dNTP, and all or part of introns. The solution for ultrafiltration replacement 1: NaCl + EDTA, pH = 7.4 ± 0.1 The solution for ultrafiltration replacement 2: NaCl + arginine + EDTA, pH = 7.4 ± 0.1 The volume of the solution for ultrafiltration replacement is 10 - 50 times (30 times), and the results are shown in Figure 14 Animal cell experiments show that precursors, linear forms, introns, polymers, and circular isomers affect cytotoxicity; open rings affect cell activity, and excessive open rings have strong immunogenicity Step 2: Coupling Oligo probes to NHS-activated beads based on 4FF or 6FF. The Oligo probes are designed according to the structures of precursors and introns generated by biological reactions (IVT), usually 1 - 3 in number. This packing material is named NHS affinity. The particle size of the affinity packing material ranges from 5um - 90um, generally for affinity anion selection, and an anion flow-through mode is adopted The NHS-activated beads based on 4FF or 6FF are from Cytiva, Chutian Microspheres, Boge Long, and Bailinke, etc. Process condition 1: NHS affinity + NHS affinity The elution solution is: NaCl + arginine + EDTA, pH = 7.4 ± 0.1 For the first NHS affinity process, discard 5 - 20% of the starting peak signal of A260 in the flow-through. The solution collected in the 5% - 20% segment mainly contains polymers and precursor linear forms Collect all of the flow-through peak of the second NHS affinity For the samples collected from the flow-through of the two affinity processes, the contents of precursors, linear forms, introns, polymers, and circular isomers can be controlled within <1%. The total yield of circular nucleic acids ≥ 30%, and the yield of circular nucleic acids with high interest rate ≥ 70%. The results are shown in Figures 15, 16, and 18 Process condition 2: NHS affinity + molecular sieve The molecular sieves include 4FF, 6FF, core400, core500, core700, core1000, SEC1000 and SEC2000, and the filler particle size ranges from 5um to 90um For the first NHS affinity process, discard 5 - 20% of the A260 flow-through starting peak signal, and the solution collected in the 5% - 20% segment is mainly polymers and precursor linear forms. Molecular sieve: Gradient elution is adopted The conditions of the equilibration solution are: PB + NaCl + EDTA, pH = 7.4 ± 0.1 The conditions of the elution solution are: PB + NaCl + PEG + EDTA, pH = 7.4 ± 0.1 The precursors, linear forms, introns, polymers and cyclic isomers in the samples collected by one affinity + molecular sieve can be controlled within <1%, and a small amount of open rings can be removed. The total yield of circular nucleic acids ≥ 30%, and the theoretical yield of circular nucleic acids ≥ 70% Process condition 3: NHS affinity + CHT For the first NHS affinity process, discard 5 - 20% of the A260 flow-through starting peak signal, and the solution collected in the 5% - 20% segment is mainly polymers and precursor linear forms. CHT: Gradient elution is adopted The conditions of the equilibration solution are: PB + NaCl + EDTA, pH = 7.4 ± 0.1 The conditions of the elution solution are: PB + NaCl + PEG + EDTA, pH = 7.4 ± 0.1 The precursors, linear forms, introns, polymers and cyclic isomers in the samples collected by one affinity + CHT can be controlled within <1%, and some open rings can be removed. The total yield of circular nucleic acids ≥ 30%, and the theoretical yield of circular nucleic acids ≥ 70%. References: [1]Group I permuted intron-exon(PIE)sequences self-spliceto produce circular exons[J]. [2]!!!INVALID CITATION!!![2], [3]!!!INVALID CITATION!!![3], [4]Chen Y G,Chen R,Ahmad S,et al.N6-Methyladenosine Modification Controls Circular RNA Immunity[J].Molecular Cell,2019,76(1):96-109.e9. [5]!!!INVALID CITATION!!![5], [6]!!!INVALID CITATION!!![6], [7]Rausch J W,Heinz W F,Payea M J,et al.Characterizing and circumventing sequence restrictions for synthesis of circular RNA in vitro[J].Nucleic Acids Res,2021,49(6):e35. [8]Beverly M,Hagen C,Slack O.Poly A tail length analysis of in vitro transcribed mRNA by LC-MS[J].Anal Bioanal Chem,2018,410(6):1667-77. [9]<Representation of the secondary and tertiary structure of group I introns.pdf>[J].

[0010] Zaug A J,Mcevoy M M,Cech T R.Self-splicing of the group I intron from Anabaena pre-tRNA:requirement for base-pairing of the exons in the anticodon stem[J].Biochemistry,1993,32(31):7946-53.

[0011] Chen R, Wang S K,Belk J A,et al.Engineering circular RNA for enhanced protein production[J].Nature Biotechnology,2022,

[0012] <A 9-nt segment of a cellular mRNA can function as an.pdf>[J].

[0013] <Biochemical and functional analysis of a 9-nt RNA.pdf>[J].

[0014] Fan X,Yang Y,Chen C,et al.Pervasive translation of circular RNAs driven by short IRES-like elements[J].2020,

[0015] Tusup M,Kundig T,Pascolo S.An Eif4g-Recruiting Aptamer Increases the Functionality of in Vitro Transcribed Mrna[J].EPH-International Journal of Medical and Health Science,2018,4(2):29-34.

[0016] Lai W C,Zhu M,Belinite M,et al.Intrinsically Unstructured Sequences in the mRNA 3′UTR Reduce the Ability of Poly(A)Tail to Enhance Translation[J].J Mol Biol,2022,434(24):167877.

[0017] Chen C K,Cheng R,Demeter J,et al.Structured elements drive extensive circular RNA translation[J].Mol Cell,2021,81(20):4300-18 e13.

[0018] Bhattacharyya S, Das S. Mapping of secondary structure of the spacer region within the 5′-untranslated region of the coxsackievirus B3 RNA: possible role of an apical GAGA loop in binding La protein and influencing internal initiation of translation[J]. Virus Research, 2005, 108(1-2):89-100.

[0019] Gao W, Li Q, Zhu R, et al. La Autoantigen Induces Ribosome Binding Protein 1(RRBP1)Expression through Internal Ribosome Entry Site(IRES)-Mediated Translation during Cellular Stress Condition[J]. International Journal of Molecular Sciences, 2016, 17(7):

[0020] Prokaryotic-like cis elements in the cap-independent internal initiation of translation on picornavirus RNA[J].

[0021] Shukla D, Zamolo L, Cavallotti C, et al. Understanding the Role of Arginine as an Eluent in Affinity Chromatography via Molecular Computations[J]. The Journal of Physical Chemistry B, 2011, 115(11):2645-54.

[0022] Levanova A, Poranen MM. Application of steric exclusion chromatography on monoliths for separation and purification of RNA molecules[J]. Journal of Chromatography A, 2018, 1574(50 - 9. Sequence information:

Claims

1. A recombinant nucleic acid molecule for preparing circular RNA, along the 5' to 3' direction, the recombinant nucleic acid molecule comprises elements operably linked in sequence: An optional 5' homologous arm, a 3' half-intron fragment, a circularization fragment, a 5' half-intron fragment and an optional 3' homologous arm; The 5' end of the circularization fragment contains a 3' circularization recognition fragment, and the 3' end contains a 5' circularization recognition fragment; The 5' half intron fragment and the 3' half intron fragment are derived from group I introns, preferably group I introns from Anabeana, preferably derived from Anabeana tRNA leu ; The 5' half-intron fragment and the 3' half-intron fragment form an intron sequence along the 5' to 3' direction; the nucleotide sequence of the 5' half-intron fragment contains a partial sequence of the intron sequence closer to the 5' direction, and the nucleotide sequence of the 3' half-intron fragment contains the remaining part of the intron sequence closer to the 3' direction, and the 5' half-intron fragment contains an internal guide (IG) sequence; Wherein: (1) The IG sequence of the intron is 3nt of 5'-GNN-3’, and the 5' circularization recognition fragment is 3nt of 5'-NNY-3’, wherein Y is C or T / U; N is an optional nucleotide, and the bases corresponding to the IG sequence and the 5' circularization recognition fragment have strict base complementary pairing or GU wobble pairing; (2) At least 2nt of the 3' circularization recognition fragment, and it has no obvious homology with the 5' circularization recognition fragment and its adjacent sequences; (3) The adjacent sequences of the 5' circularization recognition fragment and the 3' circularization recognition fragment contain at least 3nt of internal homology sequences; The obvious homology means complementary pairing of two or more nucleotides.

2. The recombinant nucleic acid molecule for preparing circular RNA according to claim 1, the 5' end of the circularization fragment has a 3' circularization recognition fragment-internal homology sequence fragment, and the 3' end has an internal homology sequence fragment-5' circularization recognition fragment, and there are optionally 0, 1, 2, 3, 4 or 5 unpaired bases between the circularization recognition fragment and the internal homology sequence; Preferably, the internal homology sequence is 3nt, 4nt or 5nt; and / or, the length of the 3' circularization recognition fragment is 2nt.

3. The recombinant nucleic acid molecule for preparing circular RNA according to claim 1, the circularization fragment comprises elements operably linked in sequence: The first part of the target polypeptide coding region or non-coding region containing the 3' circularization recognition fragment, a translation initiation element and the second part of the target polypeptide coding region or non-coding region containing the 5' circularization recognition fragment; That is, the target polypeptide coding region or non-coding region is cleaved as follows: the second part (containing the 5' circularization recognition fragment)|(containing the 3' circularization recognition fragment) the first part.

4. The recombinant nucleic acid molecule for preparing circular RNA according to claim 1, the circularization fragment comprises elements operably linked in sequence: The first part of the translation initiation element containing the 3' circularization recognition fragment, the first part of the target polypeptide coding region or non-coding region and the second part of the translation initiation element containing the 5' circularization recognition fragment; That is, the translation initiation element is cleaved as follows: the second part (containing the 5' circularization recognition fragment)|(containing the 3' circularization recognition fragment) the first part.

5. The recombinant nucleic acid molecule for preparing circular RNA according to claim 1, wherein the circularized fragment comprises sequentially operably linked elements: A 5' spacer portion, an optional translation initiation element, a target polypeptide coding region or a non-coding region, and a 3' spacer portion; Preferably, the exogenously introduced bases on the 5' spacer sequence and the 3' spacer sequence are less than 10; Preferably, three nucleotides are inserted at the third or fourth nucleotide of the 5' spacer sequence as internal homologous sequences, and two or three nucleotides of the 5' spacer sequence are used as 3' loop recognition fragments; six nucleotides are inserted at the 3' end of the 3' spacer sequence, wherein the first to third nucleotides are complementary to the nucleotides inserted in the 5' spacer sequence, the fourth to sixth sequences are NNY, and the bases corresponding to the NNY and IG sequences have strict base complementary pairing or GU wobble pairing; Also preferably, 5 nucleotides are inserted at the 5' end of the 5' spacer sequence, wherein the 1st to 2nd nucleotides serve as the 3' looping recognition fragment, the 3rd to 5th nucleotides serve as the internal homologous sequence, and the two or three nucleotides of the 5' spacer sequence itself serve as the 3' looping recognition fragment; and 6 nucleotides are inserted at the 3' end of the 3' spacer sequence, wherein the 1st to 3rd nucleotides are complementary to the nucleotides inserted in the 5' spacer sequence, and the 4th to 6th sequences are NNY.

6. The recombinant nucleic acid molecule for preparing circular RNA according to claim 5, The 5'-spacer portion is selected from: The 3'-spacer portion is selected from: Preferably, the 5' spacer sequence and the 3' spacer sequence are used alone or together; preferably, the 5' spacer sequence and the 3' spacer sequence are both polyA+8CA, preferably the polyA fragment contains 10-100 A, more preferably 60-80 A, most preferably 61 A at the 5' end and 71 A at the 3' end; Most preferably, it is a combination of 5'61A8CA and 3'71A8CA.

7. The recombinant nucleic acid molecule for preparing circular RNA according to claim 1, wherein the translation initiation element is an IRES sequence; Preferred are the following IRES sequences: Taura syndrome virus, Triatoma virus, Theiler's encephalomyelitis virus, simian virus 40, Solenopsis invicta virus 1, Rhopalosiphum padi virus, reticuloendotheliosis virus, Forman poliovirus 1, Autographa californica multiple nucleopolyhedrovirus, Kashmir bee virus, human rhinovirus 2, Homalodisca coagulata virus-1, human immunodeficiency virus type 1, Homalodisca coagulata virus-1, Pediculus humanus corporis P virus, hepatitis C virus, hepatitis A virus, GB hepatitis virus, foot-and-mouth disease virus, human enterovirus 71, equine rhinovirus, Ectropis obliqua-like virus, encephalomyocarditis virus (EMCV), Drosophila C virus, tobacco mosaic virus, cricket paralysis virus, bovine viral diarrhea virus 1, black queen cell virus, aphid lethal paralysis virus, avian encephalomyelitis virus, acute bee paralysis virus, Hibiscus chlorotic ringspot virus, classical swine fever virus, human FGF2, human SFTPA1, human AML1 / RUNX1, Drosophila antennapedia, human AQP4, human AT1R, human BAG-1, human BCL2, human BiP, human c-IAP1, human cmyc, human eIF4G, mouse NDST4L, human LEF1, mouse HIF1ɑ, human n.myc, mouse Gtx, human p27kip1, human PDGF2 / c-sis, human p53, human Pim-1, mouse Rbm3, Drosophila reaper, canine Scamper, Drosophila Ubx, human UNR, mouse UtrA, human VEGF-A, human XIAP, saliva virus, Coxsackievirus, Echovirus, Drosophila hairless, Saccharomyces cerevisiae TFIID, Saccharomyces cerevisiae YAP1, human c-src, human FGF-1, picornavirus, turnip crinkle virus, aptamer of eIF4G, Coxsackievirus B3 (CVB3) or Coxsackievirus A (CVB1 / 2); Further preferably, the IRES is the IRES sequence of Coxsackievirus B3 (CVB3); Still preferably, the IRES is the IRES sequence of encephalomyocarditis virus.

8. The recombinant nucleic acid molecule for preparing circular RNA according to claim 1, wherein the translation initiation element is the CVB3 IRES sequence inserted with an IRES enhancer element; Preferably, the insertion positions of the IRES enhancer element are: between domain I and domain II of the IRES sequence (named IRES-1), at the stem-loop structure of domain II of the IRES sequence (named IRES-2), at the stem-loop structure (Distal loop) of domain IV of the IRES sequence (named IRES-3), at the stem-loop structure (Proximal loop) of domain IV of the IRES sequence (named IRES-4); Preferably, the IRES enhancing element is: or a repeat sequence of the above sequences:

9. The recombinant nucleic acid molecule for preparing circular RNA according to claim 1, wherein the translation initiation element sequence comprises one or more combinations of the following sequences: IRES sequence, 5'UTR sequence, Kozak sequence, sequence containing m6A modification, complementary sequence of ribosomal 18S rRNA.

10. The recombinant nucleic acid molecule for preparing circular RNA according to claim 1, wherein the target polypeptide coding region or non-coding region is a protein coding region encoding a human protein or a non-human protein; Optionally, the protein coding region encodes an antibody; Optionally, the human protein or non-human protein is selected from hFIX, SP-B, VEGF-A, human methylmalonyl-CoA mutase (hMUT), CFTR, cancer autoantigen, and gene editing enzymes, such as Cpf1, zinc finger nuclease (ZFN), and transcription activator-like effector nuclease (TALEN); Optionally, the protein is a protein for therapeutic use; Optionally, wherein the antibody is a human anti-HIV antibody; Optionally, wherein the antibody is a bispecific antibody; Optionally, wherein the bispecific antibody binds CD3 and CLDN6 or binds CD19 and CD22. Optionally, wherein the protein is a protein for diagnostic use; Optionally, wherein the protein coding region encodes Gauss luciferase (Gluc), firefly luciferase (Fluc), enhanced green fluorescent protein (eGFP), human erythropoietin (hEPO), or Cas9 endonuclease; Optionally, the protein coding region comprises at least two coding regions, wherein a linker is connected between any two adjacent coding regions; preferably, the linker is a polynucleotide encoding 2A peptide. Optionally, a translation initiation element is connected between any two adjacent coding regions; optionally, the translation initiation element located between any two adjacent coding regions comprises one or more combinations of the following sequences: IRES sequence, 5'UTR sequence, Kozak sequence, sequence containing m6A modification, complementary sequence of ribosomal 18S rRNA.

11. The recombinant nucleic acid molecule for preparing circular RNA according to claim 1, wherein the recombinant nucleic acid molecule further comprises an insertion element, and the insertion element is located upstream of the translation initiation element; the insertion element is selected from at least one of the following groups (i)-(iii): (i) Transcription level regulatory element, (ii) Translation level regulatory element, (iii) Purification element; Optionally, the insertion element comprises a sequence of one or more combinations of the following: Untranslated region sequence, polyA sequence, aptamer sequence, riboswitch sequence, sequence binding a transcription regulatory factor.

12. The recombinant nucleic acid molecule for preparing circular RNA according to claim 1, wherein the target polypeptide coding region or non-coding region is a protein coding region encoding a human protein or a non-human protein; the recombinant nucleic acid molecule has the homologous arms, wherein, The length of each homologous arm is about 5-50 nucleotides; preferably, the length of each homologous arm is about 9-19 nucleotides.

13. A recombinant expression vector, wherein the recombinant expression vector comprises the recombinant nucleic acid molecule according to claim 1.

14. A circularized precursor nucleic acid molecule, which is obtained by transcription of the recombinant expression vector described in claim 13, and the fragment of the circularized precursor nucleic acid molecule comprises: Optional 5' homologous arm, 3' half-intron fragment, circularized fragment, 5' half-intron fragment and optional 3' homologous arm; the 5' end of the circularized fragment contains a 3' circularization recognition fragment, and the 3' end contains a 5' circularization recognition fragment.

15. A method for preparing circular RNA in vitro, comprising: 1) Transcription step: transcribing the recombinant expression vector according to claim 13 to form a circularized precursor nucleic acid molecule; 2) Circularization step: subjecting the circularized precursor nucleic acid molecule to a circularization reaction to obtain circular RNA; Optionally, the method further comprises a step of purifying the circular RNA.

16. A method for purifying circular RNA, the method comprising ultrafiltration, and a combination of the same or different means of affinity chromatography, molecular sieve chromatography and CHT; The preferred combination is: two affinity chromatographies, affinity chromatography and molecular sieve chromatography or affinity chromatography and CHT chromatography; Preferably, arginine and PEG are added to the ultrafiltration reagent or chromatography reagent; preferably, the reagent is a buffer; preferably, the PEG is PEG2000-8000.

17. A circular RNA molecule obtained by the recombinant nucleic acid molecule according to claim 1 or the vector according to claim 13, and the method according to claim 15.

18. A circular RNA, along the 5' to 3' direction, which comprises elements arranged in the following order: Translation initiation element, target polypeptide coding region or non-coding region; Optionally, the circular RNA comprises a 5' spacer sequence and a 3' spacer sequence located between the 5' end of the translation initiation element and the 3' end of the coding element; Each of the elements is as defined in claim 1; Preferably, the circular RNA comprises the spacer sequence according to claim 6; preferably, the circular RNA comprises less than 9 nt of exogenous inserted nucleotides; Preferably, the circular RNA comprises the IRES sequence according to claim 8; Preferably, the circular RNA comprises a modification; preferably, the modification is m6A.

19. A composition, the composition comprising the recombinant nucleic acid molecule according to claim 1, the recombinant expression vector according to claim 13, the circular RNA according to claim 17 or 18; preferably comprising the circular RNA according to claim 17 or 18; Optionally, the composition further comprises one or more pharmaceutically acceptable carriers; Optionally, the pharmaceutically acceptable carrier is selected from lipids, polymers or lipid-polymer complexes.

20. A method for expressing a target polypeptide intracellularly, wherein, The method comprises the step of transferring the circular RNA according to claim 17 or 18, or the composition according to claim 19 into cells.

21. A method for preventing or treating a disease, wherein, The method comprises administering to a subject the circular RNA according to claim 17 or 18, or the composition according to claim 19.

22. A design method for a recombinant nucleic acid molecule for preparing circular RNA, the method comprising: 1) Providing a candidate target polypeptide coding region or non-coding region, translation initiation element, and optional 5' spacer sequence and 3' spacer sequence; 2) Search for or set up 5'-loop recognition fragments, 3'-loop recognition fragments, and internal homology sequences in the target polypeptide coding region or non-coding region, translation initiation elements, or 5'-spacer sequence and 3'-spacer sequence to design a circularization fragment: The search operation is as follows: S-1) Search for the following sequence in the above elements: 5'-internal homology sequence - 5'-loop recognition fragment - 3'-loop recognition fragment - 3'-internal homology sequence; S-2) Split the corresponding element at the junction of the 5'-loop recognition fragment and the 3'-loop recognition fragment according to the search result to form a circularization sequence; The setting operation is as follows: Set 5'-internal homology sequence - 5'-loop recognition fragment sequence and 3'-loop recognition fragment - 3'-internal homology sequence at both ends of the sequence: 5'-spacer sequence - translation initiation element - target polypeptide coding region or non-coding region - 3'-spacer sequence to form a circularization fragment; Wherein: A) The 5'-internal homology sequence and the 3'-internal homology sequence are reverse complementary sequences containing at least 3 nt; B) The 3'-loop recognition fragment is a sequence of at least 2 nt and has no obvious homology with the adjacent sequence of the 5'-loop recognition fragment; C) The 5'-loop recognition fragment is a 3-nt sequence and is NNY, where N is an optional base and Y is C or T / U; 3) According to the circularization fragment obtained in step 2), modify the intron IG sequence to be used. The IG sequence is a 3-nt "GNN" and has strict base complementary pairing or GU wobble pairing with the nucleotide corresponding to the 5'-loop recognition fragment; 4) Split the modified intron sequence into a 3'-semi-intron fragment and a 5'-semi-intron fragment, and connect them in the following order to form a recombinant nucleic acid molecule for preparing circular RNA: Optional 5'-homologous arm, 3'-semi-intron fragment, circularization fragment, 5'-semi-intron fragment, and optional 3'-homologous arm; the 5' end of the circularization fragment contains a 3'-loop recognition fragment, and the 3' end contains a 5'-loop recognition fragment; Preferably, the setting is to insert 3 nucleotides as the internal homology sequence at the third or fourth nucleotide of the 5'-spacer sequence, and use two or three nucleotides of the 5'-spacer sequence itself as the 3'-loop recognition fragment; insert 6 nucleotides at the 3' end of the 3'UTR, where the 1st - 3rd nucleotides are complementary to the nucleotides inserted in the 5'-spacer sequence, and the 4th - 6th sequence is NNY and has strict base complementary pairing or GU wobble pairing with the nucleotide corresponding to the IG sequence; More preferably, the setting is to insert 5 nucleotides at the 5' end of the 5'-spacer sequence, where the 1st - 2nd nucleotides are used as the 3'-loop recognition fragment, the 3rd - 5th nucleotides are used as the internal homology sequence, and use two or three nucleotides of the 5'-spacer sequence itself as the 3'-loop recognition fragment; insert 6 nucleotides at the 3' end of the 3'-spacer sequence, where the 1st - 3rd nucleotides are complementary to the nucleotides inserted in the 5'-spacer sequence, and the 4th - 6th sequence is NNY and has strict base complementary pairing or GU wobble pairing with the nucleotide corresponding to the IG sequence.

23. A computer module, the module storing a program for implementing the design method recited in claim 22.

24. The CVB3 IRES sequence into which an IRES enhancing element is inserted, wherein: The insertion positions of the IRES enhancing element are: between domain I and domain II of the IRES sequence (named IRES-1), at the stem-loop structure of domain II of the IRES sequence (named IRES-2), at the stem-loop structure (Distal loop) of domain IV of the IRES sequence (named IRES-3), at the stem-loop structure (Proximal loop) of domain IV of the IRES sequence (named IRES-4); The IRES enhancing element is: or a repeat sequence of the above sequence:

25. Derived from Anabaena tRNA leu A mutant of a Group I intron or a fragment combination with self-splicing activity derived from the mutant, and comprising at least one set of mutations as follows: 1) The second base in the original IG sequence 5'-GAG-3' is mutated to a base other than A; Or / and 2) The third base (the G at 3') in the original IG sequence 5'-GAG-3' is mutated to a base other than G Preferably, it includes at least the mutation of the third base (the G at 3'); Preferably, it is a combination of SEQ ID No. 44 and 45 having the above mutations.

Images