Self-circularization constructs based on Bacillus subtilis phage β-22 type I intron ribozyme and their applications

Abstract

This invention provides a self-circularization construct based on the Bacillus subtilis phage β-22 type I intron ribozyme and its applications, belonging to the fields of nucleic acid therapeutics and molecular biology. A universal T2IC backbone and a candidate T2EC circularization backbone are constructed, and the target gene is circularized seamlessly through two consecutive transesterification reactions mediated by an internal guide sequence (IGS). Furthermore, the introduction of a single-base deletion in the 3' intron of the β-22 thy ribozyme inhibits co-transcriptional side reactions during in vitro transcription, significantly improving product uniformity. A process for preparing a dT7 promoter restorer PCR template adapted for industrial scale-up is also provided. The resulting construct can successfully circularize various macromolecular payloads, exhibits significantly lower immunogenicity than the classic Anabaena PIE system, and experimental evidence demonstrates its long-term expression advantages in both in vitro and in vivo, making it particularly suitable for the preparation of low-immunogenic, seamless circRNAs and nucleic acid drugs.

Description

Technical Field

[0001] This invention belongs to the field of nucleic acid therapeutics and molecular biology technology, specifically relating to a self-cyclized construct based on Bacillus subtilis phage β-22 type I intron ribozyme and its application. Background Technology

[0002] Linear messenger RNA (mRNA) drugs have demonstrated their therapeutic potential as nucleic acid drugs in clinical practice, but they still have inherent pharmacokinetic limitations in delivering large functional protein-coding sequences, establishing long-term stable expression, and circumventing innate immunogenicity. Covalently closed circular RNA (circRNA), without a free 5' / 3' end, has become an ideal delivery medium for long-acting nucleic acid vaccines, protein replacement therapies, large molecular gene editing tools, and engineered cell therapies, thanks to its natural resistance to exonucleases, significantly prolonged intracellular half-life, more persistent protein translation ability, and significantly reduced endogenous immunogenicity compared to linear mRNA.

[0003] However, the core bottleneck in realizing the clinical application of circRNA lies in the efficient, precise, and scalable in vitro circularization process. Among existing in vitro circularization strategies, the permuted intron-exon (PIE) system based on the group I intron self-splicing mechanism (using the T4 phage td system and Anabaena pre-tRNA) is the most common. Leu PIE systems (represented by PIE) have become the gold standard for current industrial applications due to their lack of the need for protein cofactors, mild reaction conditions, and high catalytic fidelity. However, PIE systems still face the following bottlenecks:

[0004] (1) Physically splitting and topologically rearranging the complete intronic catalytic core can easily disrupt the three-dimensional folding stability of ribozymes, especially when cyclizing macromolecular loads, leading to a sharp drop in cyclization efficiency.

[0005] (2) The classic PIE architecture inevitably leaves non-functional exon “scar” sequences at the splicing interface of the final cyclized product. These exogenous scars not only occupy the effective payload capacity, but also activate host pattern recognition pathways such as RIG-I and PKR, which can easily induce innate immune responses and seriously restrict its potential for clinical translation.

[0006] (3) The high Mg²⁺ concentration required for in vitro transcription (IVT) can prematurely activate ribozyme self-splicing activity, producing co-transcriptional circularization, nicked and multi-tandem byproducts, which can disrupt the homogeneity of the product.

[0007] To address the aforementioned problems, researchers have attempted to "hide" splicing sites within naturally compatible motifs within the target gene. However, such approaches are highly dependent on the target gene sequence, requiring rescanning and redesigning of the internal guide sequence (IGS) for each new payload, thus limiting their versatility. Furthermore, the existing toolkit for engineered type I introns is extremely limited. Therefore, simultaneously addressing the three major challenges of "scar immunogenicity," "payload capacity ceiling," and "co-transcriptional side effects" at the mechanistic level, and establishing a truly "plug-and-play" universal scarless circularization manufacturing platform, are key technical challenges that urgently need to be overcome in this field. Summary of the Invention

[0008] The purpose of this invention is to provide a self-cyclization construct based on the type I intron ribozyme of Bacillus subtilis phage β-22 thymidine synthase gene and its application. The core of this invention lies in the rational design guided by structural domains. Based on the systematic screening and verification of key splitting sites, the optimal framework that maintains the topological integrity of the catalytic core of Bacillus subtilis phage β-22 type I intron ribozyme was established. Based on this framework, the Thy series cyclization system was developed.

[0009] To achieve the above objectives, the present invention adopts the following technical solution:

[0010] On the one hand, the present invention provides a self-circulating construct based on Bacillus subtilis phage β-22 type I intron ribozyme, wherein the structure of the construct is optionally one of the following:

[0011] (a) The 5' end to the 3' end includes: homologous arm 1, 3' intron, exon E2, internal homologous region 1, spacer region 1, CVB3-IRES, target gene, spacer region 2, internal homologous region 2, exon E1, 5' intron, and homologous arm 2; the nucleotide sequences of exon E1 and exon E2 are shown in SEQ ID NO:11 and SEQ ID NO:12, respectively; the nucleotide sequences of internal homologous region 1 and internal homologous region 2 are shown in SEQ ID NO:17 and SEQ ID NO:18, respectively; the nucleotide sequences of spacer region 1 and spacer region 2 are shown in SEQ ID NO:19 and SEQ ID NO:20, respectively;

[0012] (b) The 5' end to the 3' end includes: homologous arm 1, 3' intron, 3' EGFP, spacer region 3, CVB3-IRES, 5' EGFP, 5' intron, and homologous arm 2; the nucleotide sequence of the 3' EGFP is shown in SEQ ID NO:28, the nucleotide sequence of the 5' EGFP is shown in SEQ ID NO:27, and the nucleotide sequence of the spacer region 3 is shown in SEQ ID NO:21;

[0013] (c) The 5' end to the 3' end includes: homologous arm 1, 3' intron, 3' IRES, target gene, 3' UTR, 5' IRES, 5' intron, and homologous arm 2; the nucleotide sequence of the 3' IRES is shown in SEQ ID NO:23, the nucleotide sequence of the 5' IRES is shown in SEQ ID NO:24, and the nucleotide sequence of the 3' UTR is shown in SEQ ID NO:25;

[0014] The nucleotide sequences of homologous arm 1 and homologous arm 2 are shown in SEQ ID NO:15 and SEQ ID NO:16, respectively; the nucleotide sequence of CVB3-IRES is shown in SEQ ID NO:22.

[0015] Furthermore, the 3' and 5' introns in the constructed structure are derived from the decomposition of the Bacillus subtilis phage β-22 type I intron ribozyme sequence, selected from any of the following groups:

[0016] (a) The nucleotide sequences of the 5' intron and the 3' intron are shown in SEQ ID NO:3 and SEQ ID NO:4, respectively;

[0017] (b) The nucleotide sequences of the 5' intron and the 3' intron are shown in SEQ ID NO:5 and SEQ ID NO:6, respectively;

[0018] (c) The nucleotide sequences of the 5' intron and the 3' intron are shown in SEQ ID NO:7 and SEQ ID NO:8, respectively;

[0019] (d) The nucleotide sequences of the 5' intron and the 3' intron are shown in SEQ ID NO:9 and SEQ ID NO:10, respectively.

[0020] Furthermore, in the construct (b), the 3'EGFP and the spacer 3 sequence also contain an internal guide sequence ATGTAG; in the construct (c), the 5'IRES and the 5' intron sequence also contain an internal guide sequence TTGGAG.

[0021] Furthermore, the target gene in the constructed structure is selected from genes with a size of 1.8kb to 6.3kb.

[0022] On the other hand, the present invention provides a self-circularized mutant construct, which is obtained by site-directed mutagenesis of the above-mentioned self-circularized construct. The site-directed mutagenesis is a deletion of G at position 345 of the type I intron ribozyme sequence in the self-circularized construct. The type I intron ribozyme sequence is shown in SEQ ID NO:2.

[0023] On the other hand, the present invention provides an expression vector comprising the above-mentioned self-circularized construct or self-circularized mutant construct, wherein the expression vector further comprises a T7 promoter or a catalytically inactivated T7 promoter, the nucleotide sequence of the T7 promoter is shown in SEQ ID NO:13, and the nucleotide sequence of the catalytically inactivated T7 promoter is shown in SEQ ID NO:1.

[0024] On the other hand, the present invention provides a method for preparing circular RNA in vitro from the above-mentioned self-circularized construct or self-circularized mutant construct, comprising the following steps:

[0025] (1) Using the self-circularized construct or the self-circularized mutant construct, a linear nucleic acid molecule is transcribed;

[0026] (2) Circulation reaction was carried out under the conditions of incubation at 55℃ for 1 h or incubation at 37℃ for 30 min followed by incubation at 55℃ for 30 min to obtain circularized RNA.

[0027] On the other hand, the present invention provides circular RNA prepared by the above-described method for preparing circular RNA in vitro.

[0028] On the other hand, the present invention provides the use of the above-mentioned self-cyclized construct, self-cyclized mutant construct or expression vector in expressing target peptides.

[0029] On the other hand, the present invention provides the application of the above-mentioned self-circulating construct, self-circulating mutant construct or expression vector in the preparation of nucleic acid drugs and their delivery formulations.

[0030] Compared with the prior art, the present invention has the following beneficial effects:

[0031] 1. The self-circulating construct of this invention can break through the 5 kb payload ceiling, successfully circulating a 6.3 kb ultra-large payload, achieving absolutely seamless linkers and significantly reducing innate immunogenicity. It achieves efficient circulating of six representative payloads (SB, VSV-G, dRfxCas13d, Cas9-NLS-Flag fusion protein, ABE8e-Cas9, and Src-EABR-ABE8e-Cas9) without significant polymerization or degradation, providing a universal industrial-grade framework for the nucleic acid delivery of macromolecular gene editing tools and multivalent vaccines. The T2IC construct exhibits significantly lower activation levels of four innate immune markers (IL-6, RIG-I, IFN-α, IFN-β) than the classic Anabaena PIE (Ana) and the Thy2 conformation containing exon residual sequences, with T2ICm further showing the lowest immunogenicity baseline. Based on the above-mentioned excellent traceless circularization ability and low immunogenicity, the self-circulating constructs of the present invention are particularly suitable for preparing low immunogenic traceless circular RNA, as well as for preparing nucleic acid drugs and their delivery formulations.

[0032] 2. Based on the Thy2, T2EC, and T2IC constructs, single-base deletions inhibit co-transcriptional side reactions, improving product homogeneity. The structural remodeling of the three mutants Thy2m, T2ECm, and T2ICm in this invention effectively prevents premature circularization during the IVT stage, resulting in highly uniform single-band precursor RNA.

[0033] 3. It is compatible with high-throughput PCR transcription technology. The dT7 promoter restorer PCR template strategy is fully compatible with the T2IC / T2ICm system. Its RNA yield and intracellular translation function are equivalent to traditional linearized plasmid templates, laying the technological foundation for GMP-level large-scale manufacturing.

[0034] 4. In an in vitro cell-free translation system, the self-circulating construct of this invention can achieve long-lasting in vitro translation, with the target protein expression duration significantly superior to that of optimized linear mRNA. Furthermore, in vivo LNP-IM administration demonstrates excellent translation durability. This confirms that the self-circulating construct of this invention possesses the outstanding advantage of long-lasting protein expression in both in vivo and in vitro environments, providing a crucial performance foundation for gene therapy drugs and vaccines requiring continuous supply of therapeutic proteins. Attached Figure Description

[0035] Figure 1 This is a schematic diagram of the wild-type secondary structure of Bacillus subtilis phage β-22 Thy ribozyme and the first-generation Thy1 rearrangement configuration in Example 1 of the present invention; wherein, Figure 1 In this context, A represents the distribution of 5' / 3' splice sites in the β-22 thy intron (GenBank L31962.1, position 820–1211); Figure 1B in the figure represents the prototype secondary configuration of Thy1, which is split based on the nt 1127 site of the Thy gene sequence. Figure 1 In the figure, C represents the agarose gel electrophoresis verification results of the in vitro transcription and cyclization products of the Thy1 construct and the control Ana; Figure 1 The diagram below, labeled D, illustrates the construction of the Thy1RNA precursor.

[0036] Figure 2 The diagram shows the initial system validation of Thy2 / Thy3 / Thy4 PIE and the structure of the T2IC / T2EC traceless system. Figure 2 A in the diagram represents three different site splitting and recombination configurations based on wild-type Thy introns (Thy2, Thy3, Thy4). Figure 2 B in the figure represents the agarose gel electrophoresis verification results of the in vitro transcription and cyclization products of Thy2, Thy3, and Thy4. Figure 2 The figure shows the fluorescence expression results of HEK293T cells 24 h after transfection with the in vitro cyclization products of Thy2, Thy3, Thy4 and the control Ana system. Figure 2 In the figure, D represents the Sanger sequencing results of the in vitro circularized products of the Thy2, Thy3, and Thy4 constructs. Figure 2 In the text, E represents the Thy2 construct and a comparison of the architectures of two traceless circular constructs based on the Thy2 construct, T2EC and T2IC.

[0037] Figure 3 For the in vitro ringing validation of two traceless systems, T2IC and T2EC, among which, Figure 3 A in the figure represents the agarose gel electrophoresis verification results of the in vitro transcription and cyclization products of the Thy2 construct and the Thy2-based traceless cyclization systems T2EC and T2IC, the control Ana, and the Ana traceless system (Ana-c). Figure 3 B in the figure is a comparison of the translational equivalence of in vitro transcription and cyclization products of the T2EC, T2IC and Ana-c systems in HEK293T without trace cyclization. Figure 3 In the figure, C represents the Sanger validation result of the seamless splicing sequence of the in vitro circularization products of T2EC and T2IC constructs.

[0038] Figure 4 To construct and evaluate Thy2, T2EC, and T2IC mutation systems based on single-base deletion mutations, among which, Figure 4 In the diagram, A represents the secondary structure of the three mutants: Thy2m, T2ECm, and T2ICm. Figure 4 B in the figure represents the agarose gel electrophoresis verification results of the in vitro transcription and cyclization products of Thy2m, T2ECm, and T2ICm. Figure 4The figure shows the fluorescence expression results of the in vitro transcription and cyclization products of Thy2m, T2ECm and T2ICm in HEK293T cells 24 h after transfection.

[0039] Figure 5 The diagram shows the construction and evaluation of T2IC and T2ICm circularized constructs compatible with high-throughput PCR. Figure 5 In the diagram, A represents the high-throughput dT7-PCR template preparation process and cyclization schematic. Figure 5 B in the figure represents the in vitro cyclization product validation results of the T2IC and T2ICm cyclization constructs compatible with high-throughput PCR; Figure 5 C in the figure represents the evaluation of the significant differences in the expression levels of the in vitro transcription and circularization products of the T2IC and T2ICm circularized constructs compatible with high-throughput PCR in HEK293T cells.

[0040] Figure 6 This is a graph showing the evaluation results of the large payload delivery capability of the T2IC construct, where... Figure 6 A in the diagram represents the six target gene payloads (T1-T6) containing 1.8–6.3 kb. Figure 6 Figure B in the figure shows the results of in vitro transcription and circular electrophoresis verification of T2IC with different loads.

[0041] Figure 7 An in vitro immunogenicity validation diagram of circRNA generated for the T2IC construct, in which, Figure 7 In this figure, A represents the RT-qPCR quantitative results of the expression of innate immune activation-related genes (IL-6, RIG-I, IFN-α, IFN-β) in HEK293T cells after transfection of the in vitro circularized products of the four constructs Ana, Ana-c, Thy2, and T2IC into the cells. Figure 7 B in the figure represents the RT-qPCR quantitative results of the expression of in vitro cyclization products and mRNAs of the three cyclization systems T2EC, T2IC, and T2ICm in HEK293T cells after transfection with HEK293T cells, which are related to the expression of innate immune activation genes (IL-6, RIG-I, IFN-α, and IFN-β).

[0042] Figure 8 The in vitro long-term translation of circRNA generated by T2IC and the in vivo translational kinetics of LNP-luciferase loaded with NeOLNP™-LNP were evaluated. Figure 8 In the figure, A represents a comparison of translation kinetics within 9 days after transfection of HEK293T with the circular EGFP-circRNA product generated in vitro by the T2IC construct and the capped modified EGFP-mRNA linear product; Figure 8 B in the figure represents the agarose gel electrophoresis verification results of the luciferase cyclic products generated in vitro in the linear luciferase, T2IC cyclization system, and Ana control system. Figure 8 The figure in Figure C represents the evaluation results of the expression activity of cyclic luciferase generated in vitro by the linear luciferase and T2IC cyclization system in HEK293T cells. Figure 8 D in the figure represents the statistical results of IVIS in vivo imaging of C57BL / 6 mice at 1, 3, 5, 9, 11, 15, and 19 days after intramuscular injection of linear luciferase and T2IC cyclization system.

[0043] It should be noted that in the above figure, I refers to the IVT reaction; S1 refers to the cyclization reaction incubated at 55℃ for 1 h; S2 refers to the cyclization reaction incubated at 37℃ for 30 min followed by incubation at 55℃ for 30 min; S refers to the optimal reaction conditions selected after screening, incubated at 37℃ for 30 min followed by incubation at 55℃ for 30 min; R refers to the RNase R digestion reaction; the control samples in the above figure are all synthesized in vitro from the Ana cyclization system. Detailed Implementation

[0044] To make the objectives, technical solutions, and unexpected technical effects of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that these embodiments are for illustrative purposes only and not for limiting the scope of protection of this invention. Equivalent substitutions, combinations, or modifications made by those skilled in the art based on the spirit of this invention all fall within the scope of protection defined by the appended claims.

[0045] Unless otherwise stated, the molecular cloning, in vitro transcription, cell culture, RNA purification, and real-time quantitative PCR techniques used in the following examples are all methods well-known in the art. Unless otherwise specified, the materials and reagents used in the following examples are all commercially available products.

[0046] Example 1 The first generation was based on the naturally occurring type I intron (named Thy1) in the Bacillus subtilis phage β-22 thymidine synthase gene thy. The thy gene sequence was split into two halves at nucleotide position 1127, and the halves were placed on both sides of the target sequence to construct a circular RNA precursor expression cassette and the in vitro transcription circularization function was verified.

[0047] This embodiment establishes with reverse evidence that cyclization activity cannot be obtained by arbitrarily splitting the β-22 thy intron using PIE. This negative result lays the foundation for subsequent screening of Thy2, Thy3, and Thy4 splitting sites.

[0048] (1) Target intron identification: Based on GenBank accession number L31962.1, type I introns of the thymidylate synthase (thy) gene from Bacillus subtilis phage (nucleotides 820–1211, SEQ ID NO:2) were extracted. The nucleotide sequence of the thy gene is shown in SEQ ID NO:1. Sequence alignment and secondary structure mapping analysis confirmed that the intron has typical type I intron self-splicing characteristics. At the posttranscribed RNA level, the 5' splice site is UAA, the 3' splice site is GCG, and it exhibits a highly conserved compact catalytic core fold (specifically, A in Figure 1).

[0049] (2) Construction of Thy1 prototype clone: ​​Referring to the cleavage site criteria of Thy ribozyme in the programmable trans-splicing riboglycolic regulatory element, based on the wild-type β-22 thy type I intron, the non-core sequence at positions 1006-1121 was first deleted, and then cleaved at nucleotide position 1127. The 5' intron containing exon 1 (E1, sequence as shown in SEQ ID NO:11, SEQ ID NO:11: TTGTTGGAGTCCGT) was shifted to the 3' end of the target sequence, and the 3' intron containing exon 2 (E2, sequence as shown in SEQ ID NO:12, SEQ ID NO:12: AGTAATG) was shifted to the 5' end of the target sequence, thus constructing a Thy1 ribozyme cyclized construct with an Anabaena PIE-like topology (specifically, A in Figure 1). The target gene insertion sequence contains a 0.7 kb EGFP coding sequence (SEQ ID NO:26) and a 0.7 kb Coxsackievirus B3 internal ribosome entry site (CVB3-IRES) sequence, as shown in SEQ ID NO:22. Complementary homologous arms 1 and 2 (sequences shown in SEQ ID NO:15 and SEQ ID NO:16) are configured at both ends. Internal homologous regions 1 and 2 (sequences shown in SEQ ID NO:17 and SEQ ID NO:18) are added upstream and downstream of the target gene to help disassemble fragments and reassemble them in three-dimensional space (schematic diagram shown in D in Figure 1).

[0050] The complete nucleic acid construct designed above was cloned into the pUC57 vector backbone using the seamless cloning enzyme One Step Seamless CloningMix (CWBio, catalog number CW3034M). After transformation into E. coli, Sanger sequencing was performed to verify that the sequences and orders of all elements were consistent with those described above. The target gene sequence was outsourced to Universal Biosynthesis. The above-described EGFP circular template plasmid was cleaved into a linearized plasmid using the endonuclease Xba I (this enzyme is not present in the backbone of the type I intron ribozyme sequence of the EGFP encoding gene, but is located on the pUC57 plasmid).

[0051] (3) Thy1 inactivation verification: The nucleic acid construct cloned into the pUC57 vector backbone in step (2) was transcribed in vitro using an IVT kit (T7 High Yield RNA Transcription Kit, Novoprotein, catalog number E131) and then ribozyme-mediated circularization was performed. The RNA precursor after in vitro transcription was added to 2× circularization buffer (100 mM Tris-HCl, 20 mM MgCl2, 2 mM DTT, and GTP to a final concentration of 2 mM) and incubated at 55℃ for 1 h (S1) or 37℃ for 30 min, and 55℃ for 30 min (S2). The two circularization reaction conditions were reacted separately. After the reaction, pre-cooled LiCl was added and the RNA was precipitated at –20℃ for at least 2 h. The precipitate was centrifuged at 12,000 rpm at 4℃ for 15 min. The precipitate was washed twice with 75% ethanol, dried at room temperature, and resuspended in RNase-free water to obtain the circularized product. The circularized product was purified and recovered using an RNA Cleanup Kit. Gel was prepared using 1% agarose and 1× MOPS buffer (0.2 M MOPS, 0.05 M NaOAc, 0.01 M EDTA, pH 7.0). RNA samples were mixed with 2× loading buffer (containing ethidium bromide), denatured at 75°C for 5 min, immediately incubated on ice for 5 min, and then loaded. Electrophoresis was performed at 5 V / cm. Denaturing agarose gel electrophoresis confirmed that the IRES-EGFP precursor linear RNA was successfully transcribed at the expected molecular weight (C in Figure 1). The results showed that in the presence of GTP + Mg²⁺, the Thy1 construct could generate the target precursor RNA after in vitro transcription, but no detectable circularization product bands were generated under the two circularization conditions S1 and S2. The classic Anabaena PIE system was used as a positive control under the same conditions, and the latter generated circularization products normally, forming a clear contrast between the two.

[0052] In this embodiment, the inactivation of Thy1 and the catalytic activity of type I introns are highly dependent on the three-dimensional precise folding of core structural domains such as the P4-P6 and P3-P9 helices—a consensus in the field. Arbitrary splitting at non-optimal sites (such as nt 1127) easily disrupts the fragile catalytic microenvironment, resulting in the complete loss of ribozyme self-splicing function. This negative result clearly reveals the decisive significance of "maintaining the topological integrity of the core critical structural domains of ribozyme catalysis" for the self-splicing activity of β-22thy ribozyme, providing a direct basis for the subsequent strategy of selecting splitting sites outside non-critical structural domains near the P6 helical region in this invention.

[0053] Example 2: Construction, screening, and establishment of the construct skeleton of three Thy intron substitution split sites: Thy2, Thy3, and Thy4.

[0054] This embodiment details the domain-guided rational design based on the secondary structure characteristics of the β-22 thy intron. Through the construction of three innovative alternative splitting configurations to avoid the key catalytic core of the intron, in vitro cyclization verification, joint fidelity determination and comprehensive performance comparison, Thy2 was finally established as the optimal chassis framework for the subsequent traceless cyclization system.

[0055] (1) Rational design of three split configurations: Based on the secondary structure map and gene sequence information of the wild-type β-22thy intron, and without avoiding the P4-P6 core catalytic layer, the following three split-and-permute configurations were designed (see Figure 2 As shown in A and Table 1, the splitting site sequence positions are defined with reference to the start of the wild-type nucleotide SEQ ID NO:1. Based on the wild-type intron, the region from positions 1006 to 1121 is a non-essential surface loop region, so the sequence in this region is deleted. Subsequently, the wild-type intron is split and recombinated to form new 5' and 3' introns. Among them, the splitting and recombination of Thy2 is located near the P6 helix (offset from the catalytic core), preserving the integrity of the P4–P6 core and P3–P9 catalytic architecture; the splitting and recombination of Thy3 is located near the P9.0 helix, utilizing the self-assembly tendency of the P9.0 double-stranded structure to assist recombination; the splitting and recombination of Thy4 is located in the P9.2 helical region, utilizing the relative structural independence of the outer helix.

[0056] Table 1. Rearrangement based on the splitting of the β-22thy wild-type intron

[0057] The three components in Table 1 were seamlessly assembled sequentially from the 5' end to the 3' end in the pUC57 vector backbone via seamless cloning. The 5' intron and 3' intron sequences used for different configurations are shown in Table 1: T7 promoter (SEQ ID NO:13) → homologous arm 1 (SEQ ID NO:15) → 3' intron and corresponding E2 fragment (SEQ ID NO:12) → spacer region 1 (SEQ ID NO:19) - target gene (in this example, EGFP and CVB3-IRES, sequences shown as SEQ ID NO:26 and SEQ ID NO:22, respectively) → spacer region 2 (SEQ ID NO:20) → corresponding E1 fragment (SEQ ID NO:11) and 5' intron → homologous arm 2 (SEQ ID NO:16). Although the three configurations are split in different regions, they can all be reconstructed to produce alternative folding states that can mediate transesterification reactions.

[0058] (2) All three configurations can achieve spontaneous circularization without helper: Thy2, Thy3, and Thy4 constructs were treated according to the method described in Example 1 to transcribe linear precursor RNA in vitro: Under the condition of no exogenous helper protein, spontaneous circularization occurred in both S1 and S2 circularization reaction conditions for all three resolving ribozyme configurations; agarose gel electrophoresis analysis was performed according to the method described in Example 1, and the results showed that: compared with the precursor RNA generated by in vitro transcription, due to the topological compactness of the covalently closed circular product, the circularized product showed a significantly faster migration rate than the corresponding linear precursor, forming a clear and specific circular RNA characteristic band ( Figure 2 (B in the middle).

[0059] (3) Validation of circularized adapter fidelity (BSJ precision): The circularized products of Thy2, Thy3, and Thy were reverse transcribed using HiScript III RT SuperMix to obtain DNA products. Then, the PCR products containing the circularized interfaces were amplified using divergent primers. The PCR reaction system is shown in Table 2, and the reaction conditions are shown in Table 3. After gel recovery, the amplified products were sent to Sanger sequencing by Qingke Biotechnology. The sequencing peaks were compared with the preset head-to-tail junction sequence to confirm seamless circularization. The Sanger sequencing results showed that the back-splice junction (BSJ) sequences of the Thy2, Thy3, and Thy4 circularized products were completely consistent with the preset “CGTAGT” E1-E2 interface (D in Figure 2). This confirmed that the three variants of this engineered β-22 thy ribozyme can achieve precise, error-free covalent closure at the 3' and 5' splice sites.

[0060] In this embodiment, the upstream and downstream primers used to amplify the PCR product containing the circularization interface are:

[0061] qJunction-F (SEQ ID NO:36):CATGGTCCTGCTGGAGTTCGTG;

[0062] qJunction-R (SEQ ID NO: 37): GTACCGTGATACCAGAGTGCT.

[0063] Table 2 High-fidelity PCR reaction system

[0064] Table 3 High-fidelity PCR reaction conditions

[0065] (4) Preliminary evaluation of translation function: Thy2, Thy3, Thy4, and Ana circularized products were added to RNase R (E224-01A) at a dose of 20 U / μg RNA and treated at 37℃ for 30 min to completely degrade the linear precursor, the cleaved linear ribozyme, and the nick / multimer byproducts. Then, high-purity monomeric circular RNA was recovered using the RNA Cleanup Kit (Tiangen) according to the instructions. 500 ng of each of the purified and enriched circular RNAs were transfected into HEK293T cells using GP-mRNA transfection reagent (pop medium, p89143), and fluorescence microscopy was performed after 48 h. The results showed that the circ-EGFP generated by the three engineered β-22 Thy ribozyme systems (Thy2, Thy3, and Thy4) exhibited intracellular functional stability consistent with that of the circulated products synthesized in vitro in the Ana control system. These results indicate that the circulated products generated in vitro by the Thy2, Thy3, and Thy4 systems all drive high-abundance, 5'-cap-independent EGFP protein expression via the CVB3-IRES element. Figure 2 (C in the middle).

[0066] (5) Comprehensive performance comparison and establishment of the Thy2 construct framework: The denatured agarose gel strips were integrated with ImageJ grayscale and calculated using the following formula:

[0067] Circulation efficiency (%) = [circular RNA integral gray level / (circular RNA + linear RNA integral gray level)] × 100.

[0068] A comprehensive evaluation of three feasible splitting strategies based on three dimensions: cyclization efficiency, byproduct profile, and intracellular biological translation function. Figure 2(C in Table 4): The Thy2 configuration exhibits the best overall performance among the three and is therefore selected as the non-marking circumferential system for further development in this invention.

[0069] Table 4. In vitro ringing efficiency of different constructs

[0070] In this embodiment, through domain-guided rational design based on the secondary structure features of the β-22thy intron, the catalytic core (P4-P6) was successfully circumvented, resulting in the construction of three novel split-rearrangement conformations (Thy2, Thy3, and Thy4). In vitro circularization verification showed that all three conformations could undergo spontaneous covalent circularization without helper proteins, forming circular RNA bands with significantly faster migration rates than the linear precursors. Sanger sequencing confirmed that the backsplicing junctions (BSJs) of the circularization products of all three conformations perfectly matched the pre-defined E1-E2 interface (CGTAGT), achieving zero-misjoint, traceless circularization. Translational function evaluation showed that the circular RNA purified by RNase R could be efficiently expressed in HEK293T cells via the CVB3-IRES element, comparable to the control system Ana. After comprehensive comparison of circularization efficiency, byproduct profiles, and biological functions, the Thy2 conformation was determined to have the best overall performance among the three splitting strategies.

[0071] Example 3: T2EC target-specific scarless circularization system based on hidden splicing sites within the target gene.

[0072] This embodiment describes a first-generation EGFP scarless circularization prototype, T2EC (Thy2 EGFP-specific Circularization), based on the Thy2 backbone constructed in Example 1 and inspired by Clean-PIE. This system achieves scarless closure by mining naturally available cryptic splicing sites within the target gene (EGFP in this embodiment), serving as an alternative sub-implementation of the general approach.

[0073] (1) Target motif mining and pairing IGS design:

[0074] Structural scanning and translational non-essentiality assessment within the EGFP coding sequence identified the following engineered element combination: CTATAT (CUAUAU at the RNA level), naturally present at position I152 of EGFP, serves as the P1 binding motif (endogenous in the target gene). This position is located in a region non-essential for EGFP fluorescence activity, and modification does not impair reporter gene function; it can serve as the upstream P1 recognition region of the 5' splice site (UAA). ATGTAG (AUGUAG post-transcriptionally), as an engineered internal guide sequence (IGS), contains thermodynamically stable G·U wobble base pairs, ensuring specific recognition under in vitro conditions, and can form a P1 duplex with CUAUAU.

[0075] (2) T2EC Reorganization Topology Construction:

[0076] The following key modifications were made to the Thy2 backbone: complete deletion of internal homology regions in the Thy2 construct; complete deletion of exogenous E1 / E2 exon sequences; segmentation of the EGFP gene along the natural CTATAT motif into 5' and 3' segments; placement of the EGFP 5' and 3' segments at opposite ends of the precursor RNA, together with the retained β-22 thy ribozyme catalytic core, to form the recombinant precursor (see...). Figure 2 (E in the example). The following elements were seamlessly connected in the pUC57 backbone in the order of 5'→3' using One Step Seamless Cloning as described in Example 1: T7 promoter → homologous arm 1 → 3' intron → 3' EGFP (SEQ ID NO:28) → spacer region 3 → CVB3-IRES → 5' EGFP (SEQ ID NO:27) → 5' intron → homologous arm 2.

[0077] (3) Verification of T2EC non-marking ringing function:

[0078] Following the in vitro transcription and circularization procedure described in Example 1, and using the classic Anabaena (Ana) PIE system and its traceless derivative Ana-c system as parallel controls, the in vitro agarose gel electrophoresis results are as follows: Figure 3 As shown in A, the T2EC traceless construct can generate precursor RNA and partially self-circulated circular RNA after in vitro transcription. After circularization treatment, circular RNA is significantly generated. After treatment with RNase R as described in Example 2, the precursor RNA is significantly eliminated, and the circular RNA is resistant to RNase R, thus achieving product enrichment.

[0079] Subsequently, following the method described in Example 2, the precursor RNA of Ana-c and T2EC, purified and enriched by RNase R, was transfected into HEK293T cells. The results were as follows: Figure 3 As shown in Figure B, the translation level of T2EC circular RNA in HEK293T cells was not significantly different from that of the classic Anabaena scarless system (Ana-c). The circularization interface of T2EC circular RNA was verified according to the method described in Example 2. Sanger sequencing showed that the adapter was perfectly reconstructed into a CTATATCAT sequence, without any foreign sequence insertion or deletion, achieving scarless closure. Figure 3 (C in the middle).

[0080] In this embodiment, the upstream and downstream primers used to amplify the PCR product containing the circularization interface are:

[0081] qEGFP-F (SEQ ID NO:38)TCGGCGAGCTGCACGCTGCCGT,

[0082] qEGFP-R (SEQ ID NO:39)TGGTGAACCGCATCGAGCTG,

[0083] In this embodiment, based on the Thy2 backbone, the natural CTATAT sequence at position I152 of EGFP was mined as the P1 motif, and an engineered IGS (ATGTAG) containing G·U wobble was designed to construct a traceless circularization prototype, T2EC. This construct deleted the internal homologous region and exogenous E1 / E2 from the Thy2 backbone, placing the split EGFP 5' and 3' segments at opposite ends of the precursor RNA. In vitro validation showed that T2EC could spontaneously circularize, and after RNase R enrichment and transfection into HEK293T cells, its translation level was not significantly different from the Ana-c system. Sanger sequencing confirmed that the circularization interface was precisely reconstructed as CTATATCAT, with no exogenous sequence insertion or deletion.

[0084] In summary, T2EC, as the first EGFP-specific, scarless circularization prototype based on the Thy2 backbone, provides a feasible alternative implementation scheme for the general system of this invention. Although T2EC achieves perfect scarless closure under EGFP payloads, the circularization mechanism of this system strictly relies on splicing target motifs naturally present within the target gene that are compatible with the P1 duplex structure. For any new therapeutic payload, it is necessary to re-scan for available P1-compatible motifs within it; redesign paired IGS; and re-validate splicing activity and translational functional integrity.

[0085] Example 4: A universal, traceless circularization system for T2IC based on hidden splice sites within IRES.

[0086] This embodiment describes a universal, traceless circularization T2IC (Thy2 IRES-split Circularization) construct based on the Thy2 backbone and inspired by Clean-PIE. This system overcomes the dependence of T2EC on the endogenous sequence of the target gene by pre-embedding self-splicing target motifs within a highly conserved CVB3-IRES structure, achieving a completely decoupled, "plug-and-play" modular circularization architecture from the target gene sequence.

[0087] (1) Target motif mining and pairing IGS design:

[0088] A systematic secondary structure scan was performed on the full sequence of CVB3-IRES (as shown in SEQ ID NO:22) to screen for 5'-NNNUNN-3' motifs compatible with P1duplex. Finally, the naturally occurring CTCTAA motif (which becomes CUCUAA after transcription) located at position 403 of CVB3-IRES (i.e., a complex structure containing NNUA + UAA) was identified as the self-splicing target. At the post-transcriptional RNA level, the core molecular element of T2IC contains a 5' splice site and a 3' splice site. The 5' splice site is located at the end of the CUCU segment and serves as the cleavage site for the first transesterification reaction. The 3' splice site is located at the beginning of the downstream AAUAC segment and serves as the splicing site for the second transesterification reaction. The engineered internal guide sequence (IGS) of the construct is set to the TTGGAG sequence (UUGGAG after transcription), which can form a P1 duplex with the splice site and spatially anchor the ribozyme catalytic center. In addition, the T2IC construct retains the β-22 thy intron active domain in the Thy2 backbone as the ribozyme catalytic core, deletes the internal homologous region but retains the original Thy2 outer homologous arm to maintain the spatial proximity of the precursor RNA folding.

[0089] (2) T2IC Reconstruction Topology:

[0090] like Figure 2As shown in E, unlike T2EC's cleavage of the target gene, T2IC splits the CVB3-IRES itself along the CTCTAA motif into 5'IRES (SEQ ID NO:24) and 3'IRES (SEQ ID NO:23), placing them upstream and downstream of the payload, respectively, while the target gene (Gene of Interest, GOI) remains intact between the two IRES fragments. Following the procedure described in Example 1, the pUC57 backbone is seamlessly ligated in the order 5'→3' using One Step Seamless Cloning Mix: T7 promoter → homologous arm 1 → 3' intron → 3'IRES (SEQ ID NO:23, containing the 3' splice site AAU) → target gene (GOI, any substitution is acceptable, EGFP is used as an example in this example) → 3'UTR (SEQ ID NO:25) → 5'IRES (SEQ ID NO:24) → 5' intron → homologous arm 2.

[0091] BamHI and HindIII restriction sites were introduced between CVB3-IRES elements to facilitate rapid replacement of GOI.

[0092] (3) Kinetics of T2IC cyclization reaction and product characterization:

[0093] Based on the in vitro transcription and circularization process of Example 1, and using the classic Anabaena (Ana) PIE system and its traceless derivative Ana-c system as parallel controls, the performance verification results of the T2IC system are as follows: Using 1,458 nt CVB3-IRES-EGFP as the loading for circularization testing, in vitro agarose gel electrophoresis results showed that after in vitro transcription to generate precursor RNA, the T2IC system can achieve highly efficient circular RNA generation through in vitro circularization. Its circularization efficiency did not decrease after deletion of the exogenous helper sequence, successfully overcoming the efficiency cliff problem commonly seen in previous traceless systems, and is comparable to the control group Ana-c. Figure 3 (A) The circular RNA of T2EC was validated by circularization interface according to the method described in Example 2. Sanger sequencing results of the reverse splice adapter showed that the adapter was accurately reconstructed into a CTCTAATAC sequence, with no foreign nucleotide insertions or deletions at single nucleotide resolution, achieving precise and seamless closure. Figure 3 (C in the middle).

[0094] In this embodiment, the upstream and downstream primers used to amplify the PCR product containing the circularization interface are:

[0095] qIRES-F (SEQ ID NO:40)TAACCCCAGTGTAGATCAGGTCGAT;

[0096] qIRES-R (SEQ ID NO:41)TGTGTGCTCCGCAGTTAGGATT.

[0097] For functional evaluation, HEK293T cells were treated with RNase R as described in Example 2, and the enriched circular RNA and control precursor RNA were transfected. The results are as follows: Figure 3 As shown in B, the translation level of T2IC circular RNA in HEK293T cells was not significantly different from that of the classic Anabaena traceless system (Ana-c), indicating that it completely preserved the spatial conformation required for IRES to recruit ribosomes.

[0098] The aforementioned BSJ sequencing and translation function verifications jointly demonstrate that the rational design of hiding the splice site within the CVB3-IRES preserves both the robust catalytic kinetics of the ribozyme and the complete spatial conformation required for IRES to recruit ribosomes and drive efficient translation. This modular architecture allows for the free replacement of downstream therapeutic GOIs without altering the circularization backbone. The T2IC circularization system designed and constructed in this embodiment achieves a truly "plug-and-play" process flow, laying the core technological foundation for the universal T2IC platform described in this invention.

[0099] Example 5: Construction and validation of the Thy2m / T2ICm / T2ECm co-transcriptional circularization repressor mutant.

[0100] This embodiment details a sequence optimization strategy for the side effect of co-transcriptional circularization during the in vitro transcription (IVT) stage.

[0101] (1) Design of a repressive mutant based on the co-transcriptional circularization of the active domain of the β-22 thy intron in the Thy2 backbone:

[0102] The high Mg²⁺ environment necessary for the IVT reaction can prematurely activate the endogenous splicing activity of type I introns, causing splicing and transcription to occur simultaneously. This results in unidentifiable intron cyclization byproducts, nicked RNA, and multimeric concatenators, all of which affect product purity. These abnormal transcripts are easily degraded further during the subsequent 55°C high-temperature cyclization step, causing significant interference to the chromatographic purification of circRNA. Therefore, through site-directed mutagenesis screening of the engineered β-22thy ribozyme domain, the following key mutation was identified: a G base deletion at position 345 of the β-22thy ribozyme intron (SEQ ID NO:2) in the Thy gene. This deletion alters the local complementary base pairing, leading to a corresponding change in the secondary conformation of this region. This single-base deletion at this position does not impair the overall autocatalytic ability of the ribozyme and can precisely inhibit premature co-transcriptional cyclization during the IVT stage.

[0103] In this embodiment, the upstream and downstream primers used to construct the mutant construct during point mutation are:

[0104] ThydG-F (SEQ ID NO:42): TCATGGAAACATGAGCAGCGTAAGCGGGCAGGGAG;

[0105] ThydG-R (SEQ ID NO:43): CTTACGCTGCTCATGTTTCCATGAAGGCCAGACTAT;

[0106] (2) Construction of the three mutants Thy2m, T2ICm, and T2ECm and evaluation of product homogeneity:

[0107] The aforementioned single-base deletions were inserted into the constructs of the Thy2, T2IC, and T2EC backbones, respectively, forming three mutant series, whose secondary structures are as follows: Figure 4 As shown in A in the diagram.

[0108] Following the in vitro transcription and circularization process described in Example 1, products of the wild-type backbone (Thy2 / T2EC / T2IC) and its corresponding mutants (Thy2m / T2ECm / T2ICm) were prepared. The precursor RNA and circular RNA of the wild-type backbone and mutants were separated and compared using agarose gel electrophoresis. The evaluation results are as follows: Figure 4As shown in B in the diagram. Regarding premature co-transcriptional circularization during the IVT stage, the wild-type scaffold clearly exhibits both precursor RNA and prematurely circularized circular RNA, while the mutant scaffold more uniformly generates precursor RNA. In terms of the purity of linear precursor RNA bands, the wild-type scaffold shows multiple bands, mixed with precursor and co-transcriptional byproducts; the mutant scaffold, on the other hand, shows highly uniform, single, and clear bands.

[0109] (3) Intracellular functional evaluation of the in vitro RNA products of Thy2, T2EC, T2IC and their corresponding mutants:

[0110] Following the procedure described in Example 2, the purified Thy2 and Thy2m IVT reaction products were directly transfected into HEK293T cells and cultured for 24 hours. The results showed that the background fluorescence of the precursor product in the Thy2m group was significantly lower than that in the Thy2 group, directly verifying that the fluorescence signal mainly originated from co-transcriptional byproducts, rather than the target precursor. Simultaneously, when enriched and purified Thy2 and Thy2m circular RNA products were transfected into HEK293T cells, the Thy2m circular RNA product maintained an equivalent and high abundance of protein translation compared to the purified Thy2 product. Figure 4 (C in the middle).

[0111] Following the above procedure, in vitro agarose gel electrophoresis was also performed on the T2EC and T2IC systems, as well as the T2ECm and T2ICm mutants. Similarly, as... Figure 4 As shown in B, T2ECm and T2ICm exhibit highly uniform, single, and clear bands in the IVT stage compared to the wild-type scaffold; however, the circularization product bands show that the T2ECm and T2ICm systems can still effectively generate circular RNA. Despite this... Figure 4 The results of the circularization lanes of Thy2m, T2ICm, and T2ECm showed that the final in vitro circularization efficiency was slightly reduced. However, the reporter gene fluorescence intensity mutant driven by HEK293T cells was not significantly different from its wild-type counterpart. This indicates that the source inhibition of co-transcriptional byproducts (including immunostimulatory nicks) significantly improved product homogeneity. This improvement in homogeneity effectively compensated for the corresponding decrease in circularization efficiency, thereby ensuring translational fidelity.

[0112] This embodiment is the first to discover a G deletion mutation at position 345 in the intron region of the β-22thy gene, which can precisely suppress the co-transcriptional circularization side reaction during the IVT stage while maintaining self-splicing activity, resulting in highly uniform precursor bands. This strategy improves the purity and translational fidelity of circRNA products from the source and has process adaptability for clinical-grade preparation.

[0113] Example 6: IVT template preparation process for the dT7 promoter T2IC construct.

[0114] Traditional IVT processes rely on plasmids linearized with restriction endonucleases as transcription templates. This approach suffers from several industrial limitations: lengthy preparation cycles; limited batch yields; high scale-up costs; difficulty in ensuring batch-to-batch consistency; and restriction sites limited by the target gene sequence. Recent studies have shown that PCR-amplified transcription templates can rival or even surpass plasmid templates in terms of RNA yield and integrity. This embodiment details an alternative transcription template preparation process based on the T2IC construct that meets the requirements for large-scale industrial production.

[0115] (1) dT7 promoter restorer PCR strategy:

[0116] This invention adapts the T2IC whole transcription cassette to a T7 promoter-inactivated recovery PCR strategy. The main steps, specific details, and objectives of the dT7 promoter-inactivated PCR template preparation process are as follows: Figure 5 As shown in A in Example 2. First, in the plasmid construction step, the complete T2IC transcription cassette was cloned into a pUC57-derived vector containing a catalytically inactivating T7 (dT7) promoter (SEQ ID NO: 14), so that the plasmid itself could not initiate any in vitro transcription reaction. Second, in the PCR amplification step (as shown in Example 2), high-fidelity amplification was performed using an upstream forward primer (SEQ ID NO: 46) carrying the active T7 promoter sequence and a corresponding downstream primer (SEQ ID NO: 47) to ensure that only the linear PCR amplification product contained the active T7 promoter, thereby enabling in vitro transcription.

[0117] Subsequently, in the purification step, the amplification product was purified using the Tiangen Universal DNA Purification Kit (DP214) to remove residual plasmids and PCR byproducts. Finally, in the in vitro transcription step, the purified linear PCR template was used instead of the conventional linearized plasmid, and transcription was performed according to the standard procedure described in Example 1. At this point, any residual plasmid template could not interfere with transcription specificity due to promoter inactivation.

[0118] In this embodiment, the upstream and downstream primers used to construct the dT7 mutant construct via point mutation are:

[0119] dT7-F (SEQ ID NO:44): GAAATGCTAGCTAATACGACACACTATAGGGATAATCCGTCGA;

[0120] dT7-R (SEQ ID NO:45): GTCGTATTAGCTAGCATTTCGATAAGCCAGTAAG.

[0121] In this embodiment, the upstream and downstream primers used to amplify the linear template carrying the active T7 promoter sequence are:

[0122] 57GsmK_IVT_F (SEQ ID NO:46):CGAAATGCTAGCTAATACGACTCACTATA;

[0123] 57GsmK_IVT_R (SEQ ID NO:47):TCCAAGCTAGAGTCGAGGCTGATC.

[0124] (2) Performance verification of the PCR template preparation pathway:

[0125] The in vitro transcription and circularization performance evaluation results of the linearized plasmid template and the dT7-PCR template were compared according to the procedure described in Example 2. Figure 5 As shown in B. Regarding the transcription yield of T2IC and T2ICm linear precursors, both the dT7-PCR template and the linearized plasmid template achieved equivalent levels. In terms of in vitro traceless circularization efficiency, the dT7-PCR template and the baseline template showed comparable circularization efficiency; the change in template form did not significantly affect the circularization efficiency. Regarding EGFP translation levels in HEK293T cells, the dT7-PCR template also completely preserved translational function, consistent with the linearized plasmid template. Figure 5 (C in the text). In summary, the dT7-PCR template is equivalent to the plasmid linearization template in all evaluation dimensions.

[0126] This embodiment is based on the molecular mechanism of improving circRNA purity through the deletion of a single base in the 3' intron of a ribozyme. It also utilizes the good compatibility between the T2IC traceless platform and the high-throughput PCR transcription process, making it suitable for the large-scale preparation of RNA therapeutic drugs.

[0127] Example 7: Verification of the T2IC platform of the present invention's 1.8–6.3 kb large load universal ringization capability.

[0128] This embodiment systematically evaluates the cyclization compatibility of the T2IC platform with ultra-large molecular loads. The results show that the platform can break through the 5 kb load limit of the traditional PIE system.

[0129] (1) Technical background and verification logic:

[0130] When developing circRNA therapeutics targeting large functional proteins (such as CRISPR-Cas gene editing systems or large viral antigens), the payload capacity of the preparation platform is a key indicator for evaluating its clinical translational potential. Currently, the PIE system splits group I introns in two and places them on either side of the payload. Its circularization process highly depends on the precise pairing and folding of the homologous arms at both ends, which are separated by the payload itself, in three-dimensional space. Therefore, when the RNA payload exceeds 5 kb, the extremely complex secondary structure of the internal sequence can severely hinder the cleavage of the split ribozyme core, causing a sharp decrease in circularization efficiency. The T2IC platform of this invention retains a complete, full-length β-22 thy ribozyme core and strictly positions the cis-splicing reaction at one end of the payload. The ribozyme folding is completely independent of the internal payload sequence, giving it the ability to circularize ultra-large RNA molecules.

[0131] (2) Load gradient series design:

[0132] Based on the CVB3-IRES-driven T2IC backbone, six representative precursor linear RNAs were constructed according to the general cloning procedure described in Example 1, with length gradients covering 1.8–6.3 kb (including CVB3-IRES and 3'UTR). Figure 6 The actual clinical demand ranges in A and Table 5): Table 5. List of 6 representative genes used in the high-load circularization capability validation of the T2IC platform.

[0133] (3) Validation of T2IC external looping for 6 loads:

[0134] The six linear precursors were processed according to the general process described in Example 1: The looping verification results of the T2IC platform for six different length loads (1.8–6.3 kb) are as follows: Figure 6 As shown in B in the figure. Regarding the electrophoretic band characteristics, all six precursors of different lengths were successfully catalyzed under mild conditions. Compared with the precursor RNA bands transcribed in vitro, circularization produced circular RNA product bands of the expected size with faster migration efficiency. For the ultra-long payload challenge (5.8 kb ABE8e-Cas9 and 6.3 kb Src-EABR-ABE8e), despite the extremely complex internal secondary structures involved, the electrophoretic spectra still showed clear target circularized bands. Regarding byproduct formation, no significant nonspecific multi-tandem formation was observed.

[0135] The results show that the successful and efficient circularization of 6.3 kb complex functional RNA demonstrates that the T2IC platform can break through the traditional PIE technology's payload limit of about 5 kb, while its circularization products maintain a traceless characteristic.

[0136] Example 8: Parallel immunogenicity evaluation of five groups: Ana / Ana-c / Thy2 / T2IC / T2ICm.

[0137] This embodiment analyzes the low immunogenicity characteristics of the T2IC / T2ICm platform from two dimensions: sequence architecture and manufacturing process, through rigorous head-to-head comparative experiments.

[0138] (1) Experimental design logic:

[0139] Currently, there are two viewpoints on the mechanism by which in vitro synthesized circRNA induces innate immune responses. On the one hand, it is believed that the "exogenous exon scar" introduced by the traditional PIE conformation folds to form a double-stranded RNA (dsRNA) structure, which is recognized by cell pattern recognition receptors such as RIG-I and PKR. On the other hand, it is believed that the incompletely spliced ​​linear fragments or nicked byproducts generated in the cyclization reaction constitute the core immune stimuli.

[0140] To systematically analyze the immunological characteristics of the T2IC platform, five parallel control groups were set up (strictly using the Ana series as controls), as shown in Table 6: Table 6. Five experimental systems for immunogenicity evaluation

[0141] (2) Standardized processing procedures:

[0142] All five groups of samples were digested with the standard RNase R described in Example 2; column chromatography purification was performed according to the RNA Cleanup Kit; and HEK293T cells were introduced with an equimolar amount (500 ng) via GP-mRNA transfection reagent (pop medium, catalog number p89143); RT-qPCR was used to relatively quantify the expression levels of IL-6, RIG-I, IFN-α, and IFN-β genes 24 h after transfection according to the procedure in Table 7 and the reaction system in Table 2.

[0143] Table 7 RT-qPCR reaction conditions

[0144] The primer pair sequences for amplifying the GAPDH, IL-6, RIG-I, IFN-α, and IFN-β genes are shown in Table 8 below: Table 8 Primer Sequences

[0145] (3) Comparison of the four groups: Ana / Ana-c / Thy2 / T2IC:

[0146] The relative expression levels of IL-6, RIG-I, IFN-α, and IFN-β mRNA normalized by GAPDH showed the following trends:

[0147] The results of head-to-head comparison of immunogenicity in four parallel control systems are as follows: Figure 7 As shown in Figure A. First, baseline comparisons between Thy2 and Ana revealed that the innate immune response induced by Thy2 (containing exon residual sequences) was significantly lower than that of the classic Ana system, suggesting that the β-22 ribozyme backbone itself has inherently low immunogenicity. Second, comparisons between Ana-c and Ana, and between T2IC and Thy2, showed that the scarless designs (Ana-c and T2IC) exhibited further reduced immune activation levels compared to their respective parent strains containing exon residual sequences, indicating that eliminating exogenous scarring is key to circumventing RIG-I pathway recognition. Finally, comparing T2IC with all control groups, the T2IC scarless system demonstrated the lowest immunogenicity level across all testing platforms, exhibiting the dual characteristics of optimized IRES cryptic splicing sites and a low-immunogenic β-22 backbone.

[0148] (4) Synergistic immune silencing effect of T2ICm co-transcriptional repressor variants:

[0149] The results of the synergistic immune silencing effect of the T2ICm co-transcriptional repressor variant are as follows: Figure 7 As shown in B in the figure. Compared with the linear RNA control group, T2IC showed significant inhibition in all immune indicators: the relative expression of IL-6, RIG-I, IFN-β and IFN-α all decreased from a strong activation level to a significantly inhibited state. Further comparison of T2IC and T2ICm revealed that T2ICm showed significantly lower inhibition levels than T2IC in IL-6, RIG-I, and IFN-β. In IFN-α, T2ICm also showed inhibition, with a similar degree of inhibition to T2IC. This indicates that "scar-induced immunity" is eliminated by the precise, scarless sequence design of T2IC. "Byproduct-induced immunity" is inhibited by the mutation site in T2ICm, specifically position 345 of the full-length SEQ ID NO:2, corresponding to the deletion of a single base G at position 115 of the 3' intron fragment after splicing and recombination. By significantly improving the structural uniformity of the precursor and the final circular product, this fundamentally prevents trace amounts of immunostimulatory dsRNA or multi-meridian impurities from contaminating the final product.

[0150] The T2IC / T2ICm platform reduces the immunogenicity of in vitro synthesized circRNA through dual optimization of "sequence architecture and manufacturing process," making it suitable for the large-scale preparation of reusable circRNA vaccines and gene therapy products.

[0151] Example 9: In vitro long-acting translation of T2IC-circRNA and in vivo pharmacodynamics of LNP-encapsulated mouse IM administration.

[0152] This embodiment evaluates the translational durability and pharmacokinetic characteristics of T2IC-circRNA in both in vitro and in vivo dimensions by rigorously controlling clinical-grade highly optimized linear mRNA (5' cap, poly(A) tail, and stabilized 3'UTR).

[0153] (1) Comparison of in vitro long-acting translation kinetics:

[0154] The two construction types and their preparation methods used for comparison are as follows: The experimental group, T2IC-circRNA, specifically consists of CVB3-IRES-driven circ-EGFP and circ-Luc (replacing EGFP with Luc in T2IC), synthesized using the T2IC traceless system according to the procedure described in Example 2. The control group is a highly optimized linear mRNA, specifically comprising a 5' cap structure, an engineered and stabilized 3' UTR, a 3' poly(A) tail, and a homologous coding region, prepared using methods known in the art to mimic clinical-grade linear mRNA.

[0155] Following the method described in Example 2, 500 ng of purified circ-EGFP product synthesized based on the T2IC construct and engineered mRNA-EGFP product were transfected into HEK293T cells. Cell fluorescence imaging was continuously observed for 9 days to highly optimize the translation kinetics comparison between the linear mRNA group and the T2IC-circ-EGFP RNA group at different time points. Figure 8 As shown in Figure A. In the early post-transfection period (0–24 h), the linear mRNA group rapidly initiated expression, exhibiting a high initial peak, while the T2IC-circRNA group showed a relatively gradual initiation. After 72 h, the fluorescence intensity of the linear mRNA group decreased sharply due to degradation of its free ends by intracellular and extracellular restriction enzymes; the T2IC-circRNA group remained stable. On day 4 (peak day), the fluorescence intensity of the linear mRNA group had already decayed, while the T2IC-circRNA group reached its expression peak. From day 5 to 9, the fluorescence intensity of the linear mRNA group approached background levels, while the T2IC-circRNA group maintained high abundance fluorescence output, significantly superior to the linear mRNA group.

[0156] T2IC-circRNA, with its covalently closed circular topology, effectively circumvents the intrinsic susceptibility of linear mRNA to intracellular and extracellular restriction enzymes, achieving significantly prolonged intracellular translational durability.

[0157] (2) Verification of in vitro functional equivalence before in vivo biodistribution:

[0158] Following the method described in Example 1, T2IC and the control Ana construct were used to generate a circular luciferase (circ-Luc) (SEQ ID NO:29) through in vitro circularization, while a highly optimized linear luciferase (mRNA-luc) was generated using a linear control group. The in vitro functional comparison results of T2IC-circRNA and the highly optimized linear mRNA are as follows. Regarding the IVT product electrophoresis, T2IC successfully generated circular RNA, and the circularized product was enriched after RNase R digestion. Figure 8 B) in the data confirms that it forms a covalently closed ring structure.

[0159] 500 ng of purified circ-Luc and mRNA-luc were transfected into HEK293T cells as described in Example 2; after 24 h, the cells were lysed with 1× Promega Passive Lysis Buffer at room temperature with gentle shaking for 15 min; the luciferase activity of the supernatant was measured using the Promega Dual-Luciferase Reporter Assay System on a Turner Biosystems 20 / 20 nLuminometer. There was no significant difference in luciferase activity between the linear mRNA group and the T2IC-circLuc group. Figure 8 (C in the text). In summary, both constructs possess complete biological functions and are suitable as equivalent constructs for in vivo comparative studies.

[0160] (3) Animal experiments involving LNP encapsulation and intramuscular injection (IM) in mice: Six-week-old SPF-grade female C57BL / 6 mice were randomly divided into two groups (n = 6): the LNP-mRNA group (containing linear Luc mRNA); and the LNP-circRNA group (containing T2IC-circLuc). All animal experimental protocols were approved by the Experimental Animal Management and Use Committee of Zhejiang University (Approval No.: ZJU20260131). Both the linear mRNA encoding Luc and the T2IC-circRNA were encapsulated into lipid nanoparticles (LNPs) using the in vivo NeOLNP™ RNAtransfection kit (catalog number SDR8002) according to the instructions.

[0161] Dosage regimen: intramuscular injection (IM, vaccine mimic); dosage: 5 μg RNA per mouse (in LNP packaged form); Day 0 (single dose).

[0162] In vivo bioluminescence imaging (BLI):

[0163] Imaging time points: Day 1, Day 3, Day 5, Day 9, Day 11, Day 15, Day 19 (a total of 7 time points);

[0164] Pre-imaging preparation: Intraperitoneal injection of 200 μL D-luciferin solution (15 mg / mL, Macklin), followed by isoflurane anesthesia 10 min later;

[0165] Imaging system: IVIS Lumina Series III (PerkinElmer);

[0166] Quantitative analysis: The total throughput (photons / s) within the region of interest (ROI) was quantified by integration using Living Image software (PerkinElmer).

[0167] Dynamic observation results:

[0168] A comparison of the translational kinetics of highly optimized linear mRNA sequences wrapped in LNPs and LNP-T2IC-circRNA sequences in mice. Figure 8 As shown in D in the diagram. On day 1 after drug administration, the linear mRNA group showed a strong initial burst expression peak, while the circRNA group was slightly weaker than the mRNA group, with the statistical relationship being Linear mRNA > circRNA. On day 3, the signal of the linear mRNA group decayed rapidly, while the decay rate of the circRNA group was significantly slower, and the decay kinetics of circRNA were significantly better than those of linear mRNA. From day 5 onwards, the linear mRNA group continued to decay, while the fluorescence intensity of the circRNA group achieved a structural reversal, showing a statistical relationship of circRNA > Linear mRNA. On day 9, the signal of the linear mRNA group was close to baseline, while the circRNA group maintained observable translational activity, with circRNA significantly higher than mRNA. On day 11, the expression activity of the linear mRNA group was close to disappearance, while the circRNA group still maintained detectable and persistent residual fluorescence activity, with circRNA maintaining expression while mRNA was close to zero. The expression activity of the circRNA group could be maintained until day 19.

[0169] The above-mentioned in vivo pharmacokinetic "cross-kinetic" curves illustrate that highly optimized linear mRNA can provide a transient early high expression peak; the circRNA generated by the T2IC traceless platform of this invention has superior in vivo stability and translation persistence due to its covalently closed topology and significantly reduced immunogenicity; the T2IC system of this invention can be used for the preparation of long-acting nucleic acid vaccines, persistent protein replacement therapies, and macromolecular gene editing delivery vectors.

[0170] In summary, this invention establishes a highly efficient, universal, and scarless T2IC platform for in vitro circRNA synthesis by systematically mining and modifying the Thy I intron of Bacillus subtilis phage β-22. First, Thy1 negativity was used to demonstrate the crucial role of maintaining the integrity of the catalytic core in ribozyme activity, and the optimal chassis Thy2 was established through domain-guided screening. Then, based on a T2EC target-specific scarless system with a hidden splicing site within the target gene, a breakthrough was achieved by constructing a universal T2IC platform completely decoupled from the payload sequence, realizing a "plug-and-play" modular architecture. For the first time, a G deletion mutation at position 345 of the wild-type I intron was discovered, which can precisely inhibit the co-transcriptional circularization side reaction in the IVT stage while maintaining self-splicing activity, eliminating immunostimulatory impurities at the source. Simultaneously, a dT7-PCR template process suitable for industrial production was established, and the platform was verified to exceed the traditional 5kb payload limit (successfully circularizing a 6.3kb ultra-long payload). Systematic immunological evaluation confirmed that T2ICm establishes the lowest immunogenicity baseline through a dual mechanism of "scar elimination and byproduct inhibition." In vivo pharmacodynamics showed that LNP-T2IC-circRNA exhibited superior translational persistence compared to existing optimized linear mRNAs, with activity lasting up to day 19. This invention systematically established a construct backbone for the in vitro synthesis of circRNA.

[0171] Finally, it should be understood that the embodiments described in this specification are merely illustrative of the principles of the embodiments described herein. Other variations may also fall within the scope of this specification. Therefore, alternative configurations of the embodiments described herein are intended to be illustrative rather than limiting, and should be considered consistent with the teachings of this specification. Accordingly, the embodiments described herein are not limited to those explicitly introduced and described herein.

Claims

1. A self-circulating construct based on Bacillus subtilis phage β-22 type I intron ribozyme, characterized in that, The structure of the self-circulating construct can be selected from one of the following: (a) The 5' end to the 3' end includes: homologous arm 1, 3' intron, exon E2, internal homologous region 1, spacer region 1, CVB3-IRES, target gene, spacer region 2, internal homologous region 2, exon E1, 5' intron, and homologous arm 2; the nucleotide sequences of exon E1 and exon E2 are shown in SEQ ID NO:11 and SEQ ID NO:12, respectively; the nucleotide sequences of internal homologous region 1 and internal homologous region 2 are shown in SEQ ID NO:17 and SEQ ID NO:18, respectively; the nucleotide sequences of spacer region 1 and spacer region 2 are shown in SEQ ID NO:19 and SEQ ID NO:20, respectively; (b) The 5' end to the 3' end includes: homologous arm 1, 3' intron, 3' EGFP, spacer region 3, CVB3-IRES, 5' EGFP, 5' intron, and homologous arm 2; the nucleotide sequence of the 3' EGFP is shown in SEQ ID NO: 28, the nucleotide sequence of the 5' EGFP is shown in SEQ ID NO: 27, and the nucleotide sequence of the spacer region 3 is shown in SEQ ID NO: 21; (c) The 5' end to the 3' end includes: homologous arm 1, 3' intron, 3' IRES, target gene, 3' UTR, 5' IRES, 5' intron, and homologous arm 2; the nucleotide sequence of the 3' IRES is shown in SEQ ID NO:23, the nucleotide sequence of the 5' IRES is shown in SEQ ID NO:24, and the nucleotide sequence of the 3' UTR is shown in SEQ ID NO:25; The nucleotide sequences of homologous arm 1 and homologous arm 2 are shown in SEQ ID NO:15 and SEQ ID NO:16, respectively; the nucleotide sequence of CVB3-IRES is shown in SEQ ID NO:

22.

2. The self-circulating construct as described in claim 1, characterized in that, The 3' and 5' introns in the structure of the construct are derived from the splitting of the Bacillus subtilis phage β-22 type I intron ribozyme sequence and are selected from any of the following groups: (a) The nucleotide sequences of the 5' intron and the 3' intron are shown in SEQ ID NO:3 and SEQ ID NO:4, respectively; (b) The nucleotide sequences of the 5' intron and the 3' intron are shown in SEQ ID NO:5 and SEQ ID NO:6, respectively; (c) The nucleotide sequences of the 5' intron and the 3' intron are shown in SEQ ID NO:7 and SEQ ID NO:8, respectively; (d) The nucleotide sequences of the 5' intron and the 3' intron are shown in SEQ ID NO:9 and SEQ ID NO:10, respectively.

3. The self-circulating construct as described in claim 1, characterized in that, In construct (b), the 5'EGFP and 5' intron sequences also contain an internal guide sequence ATGTAG; in construct (c), the 5'IRES and 5' intron sequences also contain an internal guide sequence TTGGAG.

4. The self-circulating construct as described in claim 1, characterized in that, The target gene in the structure of the construct is selected from genes with a size of 1.8kb to 6.3kb.

5. A self-circularization mutant construct, characterized in that, The self-circularized mutant construct is obtained by site-directed mutagenesis of the self-circularized construct according to any one of claims 1-4, wherein the site-directed mutagenesis is a deletion of G at position 345 of the type I intron ribozyme sequence in the self-circularized construct, and the type I intron ribozyme sequence is shown in SEQ ID NO:

2.

6. An expression carrier, characterized in that, The expression vector comprises the self-cyclized construct according to any one of claims 1-4 or the self-cyclized mutant construct according to claim 5, and further comprises a T7 promoter or a catalytically inactivated T7 promoter, wherein the nucleotide sequence of the T7 promoter is shown in SEQ ID NO:13 and the nucleotide sequence of the catalytically inactivated T7 promoter is shown in SEQ ID NO:

1.

7. A method for preparing circular RNA in vitro using the self-circularized construct according to any one of claims 1-4 or the self-circularized mutant construct according to claim 5, characterized in that, Includes the following steps: (1) Using the self-circularized construct or the self-circularized mutant construct, a linear nucleic acid molecule is transcribed; (2) Circulation reaction was carried out under the conditions of incubation at 55℃ for 1 h or incubation at 37℃ for 30 min followed by incubation at 55℃ for 30 min to obtain circularized RNA.

8. The circular RNA prepared according to the method of claim 7.

9. Use of the self-cyclized construct according to any one of claims 1-4, the self-cyclized mutant construct according to claim 5, or the expression vector according to claim 6 in expressing the target polypeptide.

10. The use of the self-circularized construct according to any one of claims 1-4, the self-circularized mutant construct according to claim 5, or the expression vector according to claim 6 in the preparation of nucleic acid drugs and their delivery formulations.

Images