A lipid nanoparticle delivery system for in vivo generation of CAR-T cells, and methods of making and using the same

Abstract

This invention belongs to the technical field of biomedicine, specifically relating to a lipid nanoparticle delivery system for in vivo CAR-T cell generation, its preparation method, and its applications. The delivery system of this invention first utilizes a high-affinity CD3ε targeting peptide to enable lipid nanoparticles (LNPs) to precisely target and efficiently transfect T cells. Secondly, it employs circular RNA (circRNA) as the CAR encoding vector; its closed circular structure resists exonuclease degradation. Finally, the aforementioned peptides and circRNA-encapsulated LNPs are integrated into a targeted delivery system using microfluidic preparation technology. The results of the embodiments demonstrate that the delivery system of this invention possesses advantages such as extremely short treatment cycles, simplified preparation processes, highly efficient targeted delivery, durable CAR expression, and high safety.

Description

Technical Field

[0001] This invention belongs to the technical field of biomedicine, specifically relating to a lipid nanoparticle delivery system for generating CAR-T cells in vivo, its preparation method, and its application. Background Technology

[0002] Gastric cancer is a highly prevalent and deadly malignant tumor worldwide. Current treatments have limited effectiveness for advanced-stage patients, necessitating new treatment strategies. While chimeric antigen receptor T-cell (CAR-T) therapy has achieved revolutionary success in hematological malignancies, its application in solid tumors (such as gastric cancer) still faces numerous challenges. Although CAR-T therapies targeting solid tumors (such as Claudin 18.2) have entered clinical trials and shown some potential, their overall development and widespread application are severely constrained by current technological limitations.

[0003] Current CAR-T cell production primarily relies on the in vitro genetic modification of patient T cells using viral vectors (such as lentiviruses and retroviruses). This approach has several inherent drawbacks: 1) The production process is complex, time-consuming, and costly, typically taking several weeks from cell collection to reinfusion, making it difficult to meet urgent clinical needs; 2) There are potential safety risks, as random genomic integration of viral vectors may lead to long-term risks such as insertional mutations and activation of oncogenes; 3) The induced CAR expression is permanent, potentially leading to over-activated immune responses or uncontrollable toxicity.

[0004] To mitigate the risks associated with viral vectors, non-viral delivery technologies, particularly transient expression strategies based on mRNA and lipid nanoparticles (LNPs), have garnered significant attention. The success of mRNA-based COVID-19 vaccines validates the effectiveness and feasibility of LNP delivery systems. However, applying this platform to in vivo CAR-T cell generation still faces key obstacles: 1) Lack of targeting: Traditional four-component LNP systems exhibit strong liver tropism, failing to effectively deliver genetic material to circulating T cells; 2) Insufficient mRNA stability: Linear mRNA has a short half-life in vivo and is easily degraded, resulting in short-duration and low-level CAR protein expression, which may not maintain a sustained anti-tumor effect.

[0005] In recent years, circular RNA (circRNA) has been shown to be more stable than linear mRNA due to its natural resistance to exonucleases caused by its closed circular structure, enabling more persistent and higher-level protein expression, making it a superior gene expression vector. Meanwhile, phage display technology has become a mature platform for screening high-affinity and high-specificity targeting peptides.

[0006] In summary, there is an urgent need in this field for diverse delivery systems for CAR-T cell construction. Summary of the Invention

[0007] The purpose of this invention is to provide a lipid nanoparticle (LNP) delivery system for in vivo generation of CAR-T cells, its preparation method, and its application. The delivery system provided by this invention can reprogram ordinary T cells into CAR-T cells in situ and in one step in the patient's body via intravenous injection, eliminating all complex in vitro operations, greatly shortening the treatment cycle, and avoiding the risk of integration mutation of viral vectors.

[0008] To achieve the above objectives, the present invention provides the following technical solution: The present invention provides a lipid nanoparticle delivery system, the delivery system comprising a targeting component, an effector component, and a delivery carrier; the targeting component is a polypeptide that specifically binds to T cells; the effector component is a circular RNA (circRNA) encoding a chimeric antigen receptor (CAR); and the delivery carrier is a lipid nanoparticle, the lipid nanoparticle encapsulating the effector component.

[0009] Preferably, the amino acid sequence of the peptide in the targeting component is as shown in SEQ ID NO: 1 or SEQ ID NO: 2. More preferably, the peptide is covalently coupled to the linker molecules on the surface of the lipid nanoparticles (LNPs).

[0010] Preferably, the chimeric antigen receptor in the effector component is the Anti-Claudin 18.2 chimeric antigen receptor.

[0011] Preferably, the lipid composition of the lipid nanoparticles comprises, in molar percentage: Ionizable lipids: accounting for 30%-60% of the total molar amount of lipids, preferably 40%-50%; Neutral cofactor phospholipids: accounting for 5%-25% of the total molar amount of lipids, preferably 10%-20%. The neutral cofactor phospholipid is preferably 1,2-distearyl-sn-glycerol-3-phosphocholine (DSPC). Cholesterol: accounting for 25%-40% of the total molars of lipids, preferably 30%-35%; PEGylated lipids: accounting for 0.5%-5% of the total molar amount of lipids, preferably 1%-2%; the PEGylated lipids are preferably DMG-PEG2000.

[0012] Preferably, the core-shell structure of the lipid nanoparticle delivery system includes a core, an outer layer, and targeted modification; the core is a complex core formed by ionizable lipids and circRNA through electrostatic interaction; the outer layer is a stable lipid bilayer composed of ionizable lipids, helper phospholipids, cholesterol, and PEGylated lipids, and the outer layer encapsulates the core; the targeted modification involves anchoring the targeted component to the outer surface of the lipid nanoparticle.

[0013] Preferably, the targeted modification involves a covalent reaction between the active groups in the targeted component and the functional PEGylated lipids embedded in the outer layer of the lipid bilayer, thereby anchoring the targeted component to the outer surface of the lipid nanoparticles.

[0014] Preferably, the particle size of the lipid nanoparticle delivery system is 80 nm to 200 nm; the particle size is the hydrodynamic diameter measured by dynamic light scattering (DLS). The particle size is preferably 100 nm to 150 nm.

[0015] Preferably, the surface potential (Zeta potential) of the lipid nanoparticle delivery system is -10 mV to +10 mV under physiological pH conditions.

[0016] Preferably, the encapsulation efficiency of circRNA in the lipid nanoparticle delivery system is greater than 80%, more preferably greater than 90%. The encapsulation efficiency is determined using a fluorescent dye (such as RiboGreen) method.

[0017] The present invention also provides a method for preparing a lipid nanoparticle delivery system, the method comprising: S1, screening and validating peptides that specifically bind to T cells; S2, construct and prepare circRNA encoding CAR; In step S3, a microfluidic mixing technique is used to mix the lipid mixture with the aqueous solution of circRNA prepared in step S2, which self-assembles to form LNPs carrying circRNA. Subsequently, the peptides obtained in step S1 are linked to modify the surface of the LNPs to construct a lipid nanoparticle delivery system with T cell targeting function.

[0018] Preferably, S2 in the method is as follows: the coding sequence of Anti-Claudin 18.2 CAR is cloned into an in vitro transcription vector containing 5' and 3' inverted repeat sequences and self-cleaved intron sequences; complete circular RNA (circRNA) is generated by in vitro transcription and circularization reaction, and its circular structure and sequence correctness are verified by gel electrophoresis, RNase R resistance experiment and sequencing.

[0019] The present invention also provides the application of the lipid nanoparticle delivery system described above in the preparation of medicaments for treating tumors. Preferably, the tumor is a solid tumor, and the solid tumor is one or more of colorectal cancer, lung cancer, esophageal cancer, gastric cancer, prostate cancer, kidney cancer, cervical cancer, and myeloma.

[0020] The present invention also provides a kit comprising a lipid nanoparticle delivery system as described above.

[0021] Beneficial Effects: This invention utilizes phage display technology to screen for peptides Cys Asn Ala Ile Lys Phe Leu His Cys (CNAIKFLHC) and Cys Tyr Asn Asn His AlaPro Asn Cys (CYNNHAPNC) that specifically bind to the CD3-epsilon antigen. These peptides are then conjugated to the surface of lipid nanoparticles (LNPs) encoding circular RNA (circRNA) for Anti-Claudin 18.2 CAR, constructing a targeted delivery system (circRNA-LNP-CD3εpep). This system can precisely deliver circRNA to T lymphocytes, generating Anti-Claudin 18.2 CAR-T cells in situ in a one-step process in vivo, and effectively killing Claudin 18.2-positive gastric cancer cells. The peptides and delivery system exhibit high specificity and affinity, enabling rapid and safe construction of CAR-T cells.

[0022] Compared to existing technologies, the outstanding advantage of the products and methods provided by this invention lies in achieving safe, efficient, and convenient in vivo in situ CAR-T cell preparation. Specifically, its core advantage is that it can revolutionize the complex CAR-T cell preparation process (including T cell isolation, viral transduction, amplification, and reinfusion) that traditionally requires several weeks to complete in vitro, into a "one-step" automated assembly that can be completed in the patient's body with a single dose. This fundamental improvement stems primarily from the synergistic effect of three key technologies in the technical solution: First, by utilizing phage display to screen and select high-affinity CD3ε targeting peptides, LNPs can be precisely located and efficiently transfected into T cells, overcoming the critical obstacle of traditional non-targeting LNPs accumulating in the liver and failing to be effectively delivered to T cells. Second, circular RNA (circRNA) is used as the encoding vector for CARs. Its closed circular structure resists exonuclease degradation and is more stable than linear mRNAs commonly used in existing technologies, thereby achieving more persistent and higher-level CAR protein expression in T cells, which is the foundation for maintaining in vivo efficacy. Finally, the aforementioned peptides and circRNA-encapsulated LNPs are integrated into a targeted delivery system using microfluidic preparation technology, forming a complete "factory" capable of intelligently targeting, releasing, and expressing CARs in vivo. Therefore, this innovative technical system of "specific targeting peptides (precise guidance) + stable circRNA (long-acting instructions) + integrated LNPs (efficient delivery)" gives it the advantages of extremely short treatment cycles, simplified preparation processes, efficient targeted delivery, persistent CAR expression, and high safety (non-viral vector). Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 The image shows the ELISA results of the binding ability of CD3ε-targeting cyclic heptapeptide monoclonal phage.

[0025] Figure 2 The image shows the fluorescence labeling results of CD3ε targeting peptides (C5-1, C5-2) on Jurkat cells and 293T cells.

[0026] Figure 3 Figure 1 shows the flow cytometry results of the binding of CD3ε targeting peptide to PBMC-derived T cells.

[0027] Figure 4 This is a graph showing the agarose gel electrophoresis results of circular RNA and linear RNA precursors.

[0028] Figure 5 The figure shows the results of circRNA stability verification after RNase R digestion.

[0029] Figure 6 The expression results of EGFP encoded by circRNA in 293T cells and Jurkat cells.

[0030] Figure 7 Flow cytometry results of circRNA-encoded CAR in Jurkat cells.

[0031] Figure 8 The image shows the results of Western blotting analysis of circRNA-encoded CAR protein expression.

[0032] Figure 9 The flow cytometry results show the expression of CAR in primary T cells after transfection with circRNA-LNP-CD3εpep.

[0033] Figure 10 CD4+ of transfected T cells + / CD8 + Figure showing the results of the subgroup proportion analysis.

[0034] Figure 11 The figure shows the results of in vitro killing experiments of CAR-T cells on Claudin 18.2 positive and negative tumor cells.

[0035] Figure 12The flow cytometry results show the expression of CD69 and CD107a, markers of CAR-T cell activation.

[0036] Figure 13 The image shows the detection results of IFN-γ and TNF-α secreted by CAR-T cells.

[0037] Figure 14 This figure shows the comparison results of CAR expression efficiency between the method of this invention and the lentivirus transduction method. Detailed Implementation

[0038] This invention provides a targeted LNP-circRNA delivery system, which is a composite nanostructure whose core function is to specifically deliver the genetic material (circRNA) it carries to T cells in vivo and complete the in situ expression of CAR protein.

[0039] (1) This delivery system consists of the following three essential parts: A. Targeting Component: A cyclic heptapeptide that specifically binds to human CD3ε protein. This peptide acts as a targeting head, guiding the entire system to recognize T cells. The most preferred amino acid sequence is CNAIKFLHC (SEQ ID NO: 1) or CYNNHAPNC (SEQ ID NO: 2). This peptide is covalently coupled to linker molecules on the surface of LNPs via its terminal portion (e.g., Cys residues).

[0040] B. Effector Component: A circular RNA (circRNA) encoding the Anti-Claudin 18.2 chimeric antigen receptor (CAR). This circRNA contains, sequentially from 5' to 3', the coding sequences for the Anti-Claudin 18.2 single-chain antibody (scFv), the CD8 hinge region, the CD28 transmembrane region, the 4-1BB co-stimulatory domain, and the CD3ζ signaling domain. This RNA molecule forms a covalently closed circular structure through a self-cleaving intron-mediated circulization reaction (e.g., a PIE structure based on the Anabaena pre-tRNA intron).

[0041] C. Delivery Vector: A lipid nanoparticle (LNP) core encapsulating and protecting the circRNA. Its lipid composition includes (in molar percentage): Ionizable cationic lipids: accounting for 30%-60% of the total molar lipids, preferably 40%-50%. Their function is to protonate under acidic conditions, bind to negatively charged circRNAs, and promote endosome escape.

[0042] Neutral accessory phospholipids: accounting for 5%-25% of the total molar amount of lipids, preferably 10%-20%. For example, 1,2-distearate-sn-glycerol-3-phosphocholine (DSPC) is used to stabilize the LNP bilayer structure.

[0043] Cholesterol: accounting for 25%-40% of the total molar lipids, preferably 30%-35%. Used to regulate membrane fluidity and stability.

[0044] PEGylated lipids: accounting for 0.5%-5% of the total molar amount of lipids, preferably 1%-2%. For example, DMG-PEG2000 is used to improve the colloidal stability and cycling time of nanoparticles and to provide them with chemical groups for coupling to target peptides (such as maleimide groups in Mal-PEG-DSPE).

[0045] (2) This delivery system has a defined core-shell structure: Core: The core of the complex formed by the ionizable lipids and circRNA through electrostatic interaction is encapsulated in a lipid bilayer.

[0046] Outer layer (lipid bilayer): A stable lipid bilayer shell composed of ionizable lipids, cofactor phospholipids, cholesterol, and PEGylated lipids.

[0047] Targeted modification: The CD3ε-targeting cyclic heptapeptide covalently reacts with the terminal active group (such as maleimide) of a functional PEGylated lipid (such as Mal-PEG-DSPE) embedded in the outer layer of the lipid bilayer, thereby anchoring itself to the outer surface of LNP particles in a "grafting" manner, exposing it to the external environment, and achieving specific binding to the CD3ε receptor on the surface of T cells.

[0048] (3) The key physicochemical properties of this delivery system directly affect its in vivo behavior and function: Particle size: The hydrodynamic diameter, measured by dynamic light scattering (DLS), is in the range of 80 nm to 200 nm, with the optimal range being 100 nm to 150 nm. This particle size range is beneficial for the long circulation of the system in vivo and avoids rapid clearance by the reticuloendothelial system, while facilitating permeation through blood vessels and effective uptake by cells.

[0049] Surface potential (Zeta potential): Under physiological pH conditions (~7.4), its surface Zeta potential is -10 mV to +10 mV, close to electroneutrality. This near-neutral potential helps reduce non-specific adsorption with negatively charged proteins in the blood, improving targeting efficiency.

[0050] circRNA encapsulation efficiency: Measured using a fluorescent dye (e.g., RiboGreen) method, the encapsulation efficiency of circRNA is greater than 80%, preferably greater than 90%. High encapsulation efficiency ensures effective utilization of the active payload.

[0051] Integrity of circular RNA: As verified by agarose gel electrophoresis, the encapsulated circRNA maintained an intact circular structure in LNP and was resistant to RNase R digestion, while the corresponding linear precursor was completely degraded.

[0052] This invention also provides a method for constructing the targeted LNP-circRNA delivery system, and its application method for one-step in vivo generation of Anti-Claudin 18.2 CAR-T cells. The specific process is as follows: (1) Screening and validation of CD3ε-targeted cyclic heptapeptide Raw materials and sources: Phage display cyclic heptapeptide library (such as the Ph.D.-C7C™ library from New England Biolabs), Escherichia coli ER2738 strain, and recombinant human CD3ε protein (Bepsys).

[0053] Process flow and parameters: a. Solid-phase coating: Dilute CD3ε protein to a concentration of 1-100 μg / mL, preferably 10 μg / mL, with coating buffer, add 50-200 μL to each well, and coat overnight (8-16 hours) at 4°C.

[0054] b. Blocking: Discard the coating solution, add 150-300 μL of TBS solution containing 1-5% BSA to each well, and block at 4℃ for 1-2 hours.

[0055] c. Phage panning: Phage peptide libraries (initial input approximately 1×10⁻⁶) are selected... 11 Add PFU to the coated wells and incubate at room temperature (20-25℃) for 30-60 minutes. Wash 6-10 times with TBST containing 0.1%-0.5% (v / v) Tween-20 to remove unbound phages.

[0056] d. Elution and Amplification: Add 0.1-0.2 M glycine-hydrochloric acid (pH 2.0-2.5) elution buffer, 50-100 μL per well, and elute with shaking at room temperature for 5-15 minutes. Immediately neutralize with 1 M Tris-HCl (pH 8.0-9.5). Infect the eluted product with logarithmic growth phase ER2738 bacterial culture (OD600≈0.5), and incubate at 37℃ and 200-250 rpm with vigorous shaking for 4-5 hours. Centrifuge the culture at 4℃ and 10,000-12,000 g for 10-15 minutes, collect the supernatant, add 1 / 4-1 / 6 volume of 20% PEG8000 / 2.5 M NaCl solution, and incubate at 4℃ overnight (12-16 hours). Centrifuge to collect the precipitated phages, and resuspend to obtain the secondary library.

[0057] e. Repeat the screening: Repeat step ad for 3-5 rounds of screening, increasing the Tween-20 concentration in the washing solution in each round (e.g., from 0.1% to 0.5%) to increase the screening pressure.

[0058] f. Single-clone identification: ≥50 individual blue phage plaques are randomly selected from the titration plates after the final round of selection, amplified, and single-stranded DNA is extracted and sequenced. For peptide sequences that appear repeatedly (frequency ≥2 times), phage ELISA verification is performed using the original phage clone: ​​Equal amounts of phage (e.g., 10...) are... 9 -10 10 PFU was added to the wells to coat them with CD3ε protein. After incubation, HRP-labeled anti-M13 antibody and TMB substrate were used for color development, and the optical density (OD) was measured at 450 nm. Clones with an OD value more than 2.1 times higher than the negative control (BSA-coated wells only) and a higher value (e.g., OD450 > 1.0) were considered positive high-affinity clones. The final preferred sequences were determined to be CNAIKFLHC (named C5-1) and CYNNHAPNC (named C5-2).

[0059] (2) Construction and preparation of Anti-Claudin 18.2 circRNA

[0060] Raw materials and sources: pGEM-T Easy vector (Sangon Biotech), restriction endonucleases, T7 in vitro transcription kit (NEB), RNA cyclase / kit (e.g., Bio-Rad), DNA purification kit, RNA purification kit.

[0061] Process flow and parameters: a. Vector construction: The complete coding sequence of Anti-Claudin 18.2 CAR was inserted into the modified pGEM-T Easy vector containing 5' and 3' inverted repeat sequences and self-splicing intron sequences (such as those derived from Anabaenapre-tRNA) by molecular cloning (i.e., homologous recombination) to construct the transcription plasmid pGEM-Ana-CAR.

[0062] b. Linearization template preparation: Using the recombinant plasmid described above as a template, perform PCR amplification using specific primers (such as Transcript-F / R), or linearize the plasmid using a single restriction endonuclease (such as ApaI) to obtain a DNA template for in vitro transcription. Verify and purify the template by agarose gel electrophoresis. The recommended template concentration is 0.1-1 μg / μL.

[0063] c. In vitro transcription: Following the T7 transcription kit instructions, establish the reaction system in a 1.5 mL RNase-free centrifuge tube: 0.5-2 μg linearized template, 10-20 μL NTP mixture (containing cap analogue), 5-10 μL T7 RNA polymerase, reaction buffer, and RNase-free water to a total volume of 20-50 μL. Incubate at 37°C for 1-3 hours.

[0064] d. DNase I treatment and purification: After the reaction, add 1-2 μL of DNase I (RNase-free) and incubate at 37°C for 15-30 minutes to remove the DNA template. Then add 1 / 10 volume of 5 M ammonium acetate and 2.5 volumes of anhydrous ethanol, precipitate at -20°C for 30-60 minutes, centrifuge at 4°C and ≥12,000 g for 15-30 minutes to collect the RNA precipitate, wash with 70% ethanol, dry, and dissolve in RNase-free water.

[0065] e. RNA cyclization: Use an RNA cyclization kit. Mix purified linear RNA precursor (0.1–1 μg / μL) with cyclase and reaction buffer (containing GTP and MgCl2), to a total volume of 20–50 μL. Incubate at 37°C for 30–60 minutes, then raise the temperature to 60–65°C and hold for 10–15 minutes to inactivate the enzyme and promote the stabilization of the cyclization product.

[0066] f. circRNA purification and validation: The circularized reaction product was purified using LiCl precipitation or a commercial RNA purification kit. Its slower migration rate compared to the linear precursor was verified by 1%–2% agarose gel electrophoresis; its resistance was verified by RNase R digestion (37°C, 30 minutes); and sequence correctness was verified by reverse transcription-PCR targeting the circularization junction and sequencing.

[0067] (3) Preparation of a targeted LNP-circRNA delivery system (circRNA-LNP-CD3εpep)

[0068] Raw materials and sources: ionizable lipids (such as DLin-MC3-DMA), DSPC, cholesterol, DMG-PEG2000, Mal-PEG-DSPE (Xi'an Ruixi / AVANTI), circRNA prepared in step (2) above, synthetic cyclic heptapeptide determined in step (1) above (Qiangyao Biosynthesis), microfluidic mixer (such as NanoAssemblr).

[0069] Process flow and parameters: a. Lipid phase preparation: Ionizable lipids, DSPC, cholesterol, DMG-PEG2000 and Mal-PEG-DSPE were dissolved in anhydrous ethanol at a molar ratio of (40-50): (10-20): (30-35): (1-2): (0.5-1.5), with a total lipid concentration ranging from 5 to 20 mM, preferably about 10 mM.

[0070] b. Aqueous phase preparation: Dissolve the purified circRNA in 10-50 mM sodium citrate buffer, pH 4.0-5.0, at a concentration ranging from 0.05-0.2 mg / mL, preferably about 0.1 mg / mL. The nitrogen-to-phosphorus ratio (N / P, the molar ratio of amino groups in ionizable lipids to phosphate groups in RNA) is controlled between 3:1 and 10:1, preferably about 6:1.

[0071] c. Microfluidic mixing: Set up a microfluidic mixer. Inject the ethanol and aqueous phases of the lipids into two separate injection tubes, controlling the flow rate ratio of the two phases to 1:3 to 1:5 (aqueous phase:lipid phase), with a total flow rate of 8-15 mL / min. The mixture is instantaneously mixed within the microfluidic chip channels, where the lipids self-assemble into LNPs carrying circRNA upon contact with the aqueous phase.

[0072] d. Dialysis and concentration: Immediately transfer the mixed LNP suspension to PBS buffer at pH 7.4 and dialyze at 4°C with stirring for 4–12 hours using a dialysis bag with a molecular weight cutoff (MWCO) of 10–100 kDa to remove ethanol and free components. Subsequently, moderate concentration can be performed using ultrafiltration centrifuge tubes (MWCO 100 kDa).

[0073] e. Targeted peptide conjugation: The CD3ε-targeted cyclic heptapeptide (C5-1 or C5-2) synthesized in step (1) above was dissolved in PBS. The purified LNP (containing maleimide groups on the surface of Mal-PEG-DSPE) was mixed with the peptide solution, with a molar ratio of peptide to maleimide groups of approximately 1.5:1 to 5:1. The mixture was incubated at room temperature (20-25℃) in the dark for 2-4 hours. After the reaction was completed, unreacted free peptides were removed by dialyzing again or by using a desalting column to obtain the final targeted delivery system circRNA-LNP-CD3εpep.

[0074] f. Characterization: The particle size (target 80-200 nm, preferably 100-150 nm) and zeta potential were determined using dynamic light scattering (DLS); the morphology was observed using transmission electron microscopy (TEM); and the encapsulation efficiency of circRNA was determined using the RiboGreen fluorescent dye method (should be >80%).

[0075] (4) Application: One-step in vivo generation of Anti-Claudin 18.2 CAR-T cells

[0076] Raw materials and sources: circRNA-LNP-CD3εpep and Claudin 18.2 positive gastric cancer patients or tumor-bearing animal models prepared in step (3) above.

[0077] Process flow and parameters: a. Administration: Dilute circRNA-LNP-CD3εpep with sterile saline or PBS to a suitable injection concentration. Administer via intravenous injection or local tumor injection. The dosage, based on circRNA content, can be 0.1-5 mg / kg body weight, preferably 0.5-2 mg / kg.

[0078] b. In vivo process: After entering the systemic circulation, the CD3ε targeting peptide on its surface specifically recognizes and binds to the CD3ε receptor on the surface of T cells, and is taken up by T cells through memory action. Intracellularly, LNPs undergo structural changes in the acidic environment of the endosome, releasing circRNA into the cytoplasm.

[0079] c. In situ generation of CAR-T cells: The released circRNA initiates rolling circle translation using its internal IRES sequence, continuously expressing the Anti-Claudin 18.2 CAR protein. The CAR protein is transported to the T cell membrane surface, completing the in situ assembly of functional CAR-T cells.

[0080] d. Anti-tumor effect: In situ generated CAR-T cells recognize and bind to gastric cancer cells expressing Claudin 18.2 in the patient's body or tumor microenvironment, and kill tumor cells by releasing granzymes, perforin and inflammatory factors (IFN-γ, TNF-α, etc.) to achieve the therapeutic goal.

[0081] To further illustrate the present invention, the technical solutions provided by the present invention will be described in detail below with reference to the accompanying drawings and embodiments, but these should not be construed as limiting the scope of protection of the present invention.

[0082] Unless otherwise specified, the production processes, experimental methods, or testing methods involved in the embodiments of this invention are all conventional methods in the prior art, and their names and / or abbreviations are all conventional names in the field, which are very clear and distinct in the relevant application areas. Those skilled in the art can understand the conventional process steps based on the names and apply the corresponding equipment, and implement them according to conventional conditions or conditions recommended by the manufacturer.

[0083] The various instruments, equipment, raw materials or reagents used in the embodiments of this invention are not subject to any special restrictions on their source. They are all conventional products that can be purchased through regular commercial channels and can be prepared according to conventional methods known to those skilled in the art.

[0084] Example 1: Screening, synthesis and specificity verification of CD3ε-targeting cyclic heptapeptide

[0085] 1.1 Experimental Objective

[0086] High-affinity peptides that specifically bind to human CD3ε protein were screened from a phage-displayed cyclic heptapeptide library, synthesized, and their binding specificity to T cells was verified.

[0087] 1.2 Experimental Methods

[0088] Phage biopanning: A cyclic heptapeptide library was displayed using Ph.D.-C7C™ phage from New England Biolabs. Recombinant human CD3ε protein (Bepsys) was coated onto the wells of an ELISA plate at a concentration of 10 μg / mL and incubated overnight at 4°C. After BSA blocking, approximately 1 × 10⁻⁶ g of BSA was added. 11 The initial phage library of PFU was incubated at room temperature for 1 hour. It was washed 10 times with TBST containing 0.1% Tween-20 to remove unbound phages. Specifically bound phages were eluted with 0.2 M glycine-HCl (pH 2.2) and neutralized with 1 M Tris-HCl (pH 9.1). The elution product was used to infect logarithmic-phase *E. coli* ER2738, amplified, and purified by PEG / NaCl precipitation to obtain the secondary library. This process was repeated for four rounds of panning, with the Tween-20 concentration in the wash buffer increased to 0.5% in the last two rounds.

[0089] Monoclonal identification and sequencing: One hundred independent blue phage plaques were randomly selected from the titer assay plates after the fourth round of panning and amplified separately. Single-stranded DNA was extracted from each monoclonal phage using the sodium iodide method and sent to BGI Genomics in Shenzhen for sequencing. The obtained DNA sequences were translated and compared.

[0090] Binding ability verification (phage ELISA): The amplified monoclonal phages were added to the wells of CD3ε protein-coated ELISA plates and incubated at room temperature for 1 hour. After washing, HRP-labeled anti-M13 antibody was added, TMB was used for color development, and the absorbance (OD450) was measured at 450 nm. Wells containing only the supernatant of the amplified bacterial culture were used as negative controls.

[0091] Peptide synthesis and probe preparation: Peptide sequences with high sequencing repetition frequency and strong ELISA binding signals (OD450 > 1.0) were selected and synthesized on a solid-phase basis by Shanghai Qiangyao Biotechnology Co., Ltd. Fluorescein isothiocyanate (FITC) was then coupled to the N-terminus or C-terminus to form the CD3εpep-FITC probe. The purity was verified by HPLC to be >95%.

[0092] Cellular-level specificity validation: Live cell staining: Human T-lymphoblastic leukemia cells (Jurkat, CD3ε positive) and human embryonic kidney epithelial cells (293T, CD3ε negative) were seeded into 24-well plates. 1 μg / mL of CD3εpep-FITC (C5-1 or C5-2) working solution was added, and the cells were incubated at 37°C for 1 hour. After washing with PBS, the fluorescence signals of the two cell types were observed and compared under a fluorescence inverted microscope.

[0093] Flow cytometry: T cells derived from peripheral blood mononuclear cells (PBMCs) were incubated with the CD3εpep-FITC probe at 4°C in the dark for 1 hour, washed, and then analyzed. A commercially available anti-CD3 antibody was used as a T cell marker control.

[0094] 1.3 Test Results

[0095] After four rounds of screening, the phage recovery rate increased in each round (Table 1), indicating that phages that specifically bind to CD3ε were effectively enriched.

[0096] Sequencing 100 single clones yielded 24 unique cyclic heptapeptide sequences. Phage ELISA results showed that the two clones with sequences CNAIKFLHC (named C5-1) and CYNNHAPNC (named C5-2) exhibited the strongest binding signals, with average OD450 values ​​of 1.123 and 1.063, respectively, significantly higher than the background (…). Figure 1 ).

[0097] Live cell staining results showed that C5-1-FITC and C5-2-FITC probes significantly labeled Jurkat cells, while almost no fluorescence signal was observed on 293T cells. Figure 2 Flow cytometry further confirmed that the CD3εpep-FITC probe can specifically bind to the T cell population in PBMCs. Figure 3 ).

[0098] Table 1: Input and Recovery Status During Bacteriophage Panning

[0099] Example 2: Construction, preparation and functional verification of Anti-Claudin 18.2 CAR circRNA

[0100] 2.1 Experimental Objective

[0101] A circular RNA (circRNA) encoding Anti-Claudin 18.2 CAR was constructed, and its structural integrity, stability, and expression function in cells were verified.

[0102] 2.2 Experimental Methods

[0103] Transcription vector construction: Using the laboratory-preserved Anti-Claudin 18.2 CAR sequence as a template, the target fragment was amplified by PCR. Homologous recombination technology was used to insert this fragment into the modified pGEM-T Easy vector containing a self-cleaving intron sequence of Anabaena pre-tRNA, constructing the recombinant plasmid pGEM-Ana-CAR. The plasmid was transformed into DH5α competent cells, and single clones were selected for colony PCR and sequencing verification.

[0104] In vitro transcription and circularization: The validated plasmid was linearized by PCR using specific primers and used as a template for in vitro transcription. In vitro transcription was performed using the T7 High Yield RNA Synthesis Kit (NEB) at 37°C for 2 hours. After DNase I digestion, the linear RNA precursor was purified by LiCl precipitation. Using the CircRNACyclization Kit (Bio-Raybo), the purified linear RNA was circularized at 37°C for 30 minutes, followed by enzyme inactivation at 60°C for 15 minutes, and then purified again to obtain circRNA.

[0105] circRNA characterization: Gel electrophoresis: The cyclized products were subjected to 1.5% agarose gel electrophoresis, and their migration positions were compared with those of the linear precursor RNA.

[0106] RNase R digestion: Equal amounts of linear RNA precursor and circRNA were digested with RNase R (Epicentre) at 37°C for 30 minutes, followed by gel electrophoresis to observe band changes.

[0107] Linkage point sequencing: circRNA is reverse transcribed into cDNA, and specific primers are designed for circularization linkage points for PCR amplification. The products are then sent for sequencing to verify sequence correctness.

[0108] circRNA function verification (transfection): EGFP reporter gene validation: The circRNA encoding EGFP (prepared using the same vector system) was transfected into 293T cells and Jurkat cells using Lipofectamine™ 3000, and EGFP expression was observed under a fluorescence microscope 48 hours later.

[0109] CAR expression validation: Jurkat cells were transfected with circRNA encoding Anti-Claudin 18.2 CAR. 48 hours after transfection, cells were collected. One portion was used for flow cytometry (using anti-G4S antibody to recognize the CAR structure), and the other portion was lysed and CAR protein expression was detected by Western Blot (WB) (primary antibody was anti-CD3ζ).

[0110] 2.3 Test Results

[0111] The pGEM-Ana-CAR plasmid was successfully constructed, and sequencing confirmed the correctness of the CAR sequence and flanking circular elements.

[0112] Agarose gel electrophoresis showed that circRNA migrated more slowly than its linear precursor ( Figure 4 After RNase R digestion, the linear RNA band completely disappeared, while the circRNA band remained intact, proving its successful circular structure. Figure 5 Connection point sequencing results confirmed that the circularized interface sequence was completely consistent with the design.

[0113] Transfection experiments showed that circRNA encoding EGFP could effectively express green fluorescence in 293T and Jurkat cells. Figure 6 Flow cytometry and Western blotting confirmed that Jurkat cells transfected with Anti-Claudin 18.2 CAR circRNA had a significantly higher CAR positivity rate than the untransfected group, and clear CAR protein bands were detected. Figure 7 , Figure 8 ).

[0114] Example 3: Preparation of a targeted LNP delivery system and in vitro functional evaluation of CAR-T cells

[0115] 3.1 Experimental Objective

[0116] We prepared a targeting LNP (circRNA-LNP-CD3εpep) loaded with Anti-Claudin 18.2 CAR circRNA and evaluated the anti-tumor function of CAR-T cells generated after transfection with it into primary T cells.

[0117] 3.2 Experimental Methods

[0118] LNP preparation and characterization: Microfluidic mixing was employed. For the lipid-ethanol phase: ionizable lipids (DLin-MC3-DMA), DSPC, cholesterol, DMG-PEG2000, and Mal-PEG-DSPE were dissolved in ethanol at a molar ratio of 50:10:38.5:1.0:0.5 (total lipid concentration 10 mM). For the aqueous phase: the circRNA prepared in Example 2 was dissolved in 25 mM sodium citrate buffer (pH 4.5) (concentration 0.1 mg / mL). The total flow rate was set at 12 mL / min, and the flow rate ratio (aqueous phase:lipid phase) was 3:1. Mixing was performed on a NanoAssemblr instrument. The resulting LNP suspension was immediately dialyzed into PBS (pH 7.4).

[0119] Peptide conjugation: The C5-1 peptide synthesized in Example 1 was incubated with dialyzed LNP at room temperature in the dark for 3 hours (peptide:maleimide molar ratio = 3:1). After the reaction, it was dialyzed again to obtain circRNA-LNP-CD3εpep.

[0120] Characterization: Particle size and zeta potential were determined using a dynamic light scattering instrument. circRNA encapsulation efficiency was determined using the RiboGreen kit.

[0121] CAR-T cell preparation: PBMCs were isolated from peripheral blood of healthy volunteers, and primary T cells were sorted using CD3 microbeads. Cells were cultured in X-VIVO 15 medium containing IL-2 (100 U / mL). circRNA-LNP-CD3εpep was added to the T cell culture system at a final circRNA concentration of 100 ng / mL, and the cells were incubated for 48 hours.

[0122] CAR expression and cell phenotype detection: 48 hours after transfection, cells were collected and the CAR positivity rate (anti-G4S antibody) and the proportion of CD4+ and CD8+ T cell subsets were detected by flow cytometry.

[0123] In vitro killing assay: Claudin 18.2-overexpressing gastric cancer cell lines AGS-OE and HGC-27-OE were used as target cells, and their empty vector control cells AGS-NC and HGC-27-NC were used as negative controls. Using the luciferase reporter gene assay, CAR-T cells with different effector-to-target ratios (0.5:1, 1.5:1, 3:1) were co-cultured with target cells for 48 hours, and changes in luciferase activity were detected to calculate the specific killing rate.

[0124] Cell activation and factor secretion assays: After co-culturing for 24 hours at an effector-to-target ratio of 3:1, the expression of CAR-T cell surface activation marker CD69 and degranulation marker CD107a was detected by flow cytometry. After co-culturing for 48 hours, the secretion levels of IFN-γ and TNF-α in the cell supernatant were detected using a flow cytometry cytokine assay kit.

[0125] 3.3 Test Results

[0126] The prepared circRNA-LNP-CD3εpep particles had a size of 128 ± 15 nm, a PDI of 0.12, a Zeta potential of -2.1 ± 0.5 mV, and an encapsulation efficiency of 92.5%.

[0127] Flow cytometry showed that primary T cells transfected with circRNA-LNP-CD3εpep had a CAR-positive expression rate of 65-80%, and the CD4+ / CD8+ subset ratio did not change significantly compared with the untreated group. Figure 9 , Figure 10 ).

[0128] In vitro killing experiments showed that this CAR-T cell could efficiently kill Claudin 18.2-positive AGS-OE and HGC-27-OE cells, with a killing rate exceeding 75% at an effector-to-target ratio of 3:1, while having almost no killing effect on Claudin 18.2-negative AGS-NC and HGC-27-NC cells. Figure 11 ).

[0129] After co-culturing with target cells, CAR-T cells highly expressed CD69 and CD107a and secreted large amounts of IFN-γ and TNF-α, indicating that they were effectively activated and possessed strong effector functions. Figure 12 , Figure 13 ).

[0130] Comparative example: Compare the key performance differences between the "one-step in vivo method" for generating CAR-T cells of this invention and the traditional in vitro viral transduction method for preparing CAR-T cells.

[0131] 4.1 Comparison Objects: This invention group: Primary T cells were treated using the circRNA-LNP-CD3epep delivery system prepared by the method in Example 3.

[0132] Existing technology group (lentiviral transduction method): Using a commercial third-generation lentiviral vector system, viral particles carrying the same Anti-Claudin 18.2 CAR gene are packaged, and activated primary T cells are infected in vitro. After culture and expansion, CAR-T cells are obtained.

[0133] 4.2 Comparison Methods and Results: a. Preparation cycle and complexity: The existing technology group requires multiple complex steps, including T cell isolation and activation, virus packaging and titration, virus transduction, in vitro amplification (usually requiring 10-14 days), and quality inspection. The process is lengthy and has extremely high requirements for production conditions (GMP workshop).

[0134] The invention group: The preparation of circRNA-LNP-CD3epep can be completed within hours. After in vivo administration, CAR-T cells are generated in situ in the patient's body, shortening the treatment preparation period of several weeks to 1-2 days, and greatly simplifying the process.

[0135] b. Security risks: Existing technology group: Lentivirals pose a potential risk of random insertion into the host genome, leading to insertional mutations and oncogenicity, and viral vectors may trigger unnecessary immune responses.

[0136] This invention utilizes non-viral LNP delivery and transient expression of circRNA, eliminating the risk of genome integration. The circRNA is eventually naturally degraded by the cell, significantly improving safety.

[0137] c. Targeting and off-target toxicity: Existing technology group: After in vitro prepared CAR-T cells are reinfused, their targeting depends entirely on the CAR itself to recognize tumor antigens, which poses a risk of "off-target toxicity", and the distribution of cells in vivo is uncontrollable.

[0138] The present invention group: LNP-CD3epep's targeting is mediated by the CD3ε peptide, ensuring that genetic material is delivered only to T cells, thereby controlling the cell type expressing CAR from the source and reducing the risk of accidentally damaging other tissues.

[0139] d. CAR expression kinetics: Existing technology group: Virus-mediated gene integration leads to permanent expression of CAR, which may cause excessive immune responses or long-term toxicity and cannot be regulated.

[0140] The present invention proposes that circRNA-mediated CAR expression is transient, and its duration can be modulated by LNP dosage and circRNA design, which is more conducive to safety management.

[0141] The comparison results of the two methods in terms of CAR expression efficiency are as follows: Figure 14 As shown

[0142] Although the above embodiments have provided a detailed description of the present invention, they are only some embodiments of the present invention, and not all embodiments. People can obtain other embodiments based on these embodiments without creative effort, and these embodiments all fall within the protection scope of the present invention.

Claims

1. A lipid nanoparticle delivery system, characterized in that, The delivery system includes a targeting component, an effector component, and a delivery carrier; the targeting component is a polypeptide that specifically binds to T cells; the effector component is a circular RNA encoding a chimeric antigen receptor; and the delivery carrier is a lipid nanoparticle that encapsulates the effector component.

2. The lipid nanoparticle delivery system according to claim 1, characterized in that, The amino acid sequence of the polypeptide in the targeted component is shown in SEQ ID NO: 1 or SEQ ID NO:

2.

3. The lipid nanoparticle delivery system according to claim 1, characterized in that, The chimeric antigen receptor in the effector component is the Anti-Claudin 18.2 chimeric antigen receptor.

4. The lipid nanoparticle delivery system according to claim 1, characterized in that, The lipid composition of the lipid nanoparticles, in molar percentage, includes: Ionizable lipids: accounting for 30%-60% of the total molar amount of lipids; Neutral cofactor phospholipids: accounting for 5%-25% of the total molar amount of lipids; Cholesterol: accounts for 25%-40% of total lipid moles; PEGylated lipids: accounting for 0.5%-5% of the total molar amount of lipids.

5. The lipid nanoparticle delivery system according to claim 4, characterized in that, The core-shell structure of the lipid nanoparticle delivery system includes a core, an outer layer, and targeted modification; the core is a complex core formed by ionizable lipids and circRNA through electrostatic interaction; the outer layer is a stable lipid bilayer composed of ionizable lipids, helper phospholipids, cholesterol, and PEGylated lipids, and the outer layer encapsulates the core; the targeted modification involves anchoring the targeted component to the outer surface of the lipid nanoparticle.

6. The lipid nanoparticle delivery system according to any one of claims 1-5, characterized in that, The lipid nanoparticle delivery system has a particle size of 80 nm to 200 nm; the particle size is a hydrodynamic diameter measured by dynamic light scattering.

7. A method for preparing the lipid nanoparticle delivery system according to any one of claims 1-5, characterized in that, The method includes: S1, screening and validating peptides that specifically bind to T cells; S2, construct and prepare circRNA encoding CAR; In step S3, a microfluidic mixing technique is used to mix the lipid mixture with the aqueous solution of circRNA prepared in step S2, which self-assembles to form LNPs carrying circRNA. Subsequently, the peptides obtained in step S1 are linked to modify the surface of the LNPs to construct a lipid nanoparticle delivery system with T cell targeting function.

8. The method for preparing the lipid nanoparticle delivery system according to claim 7, characterized in that, In the method described, S2 involves: cloning the coding sequence of Anti-Claudin 18.2 CAR into an in vitro transcription vector containing 5' and 3' inverted repeat sequences and self-cleaved intron sequences; generating complete circular RNA (circRNA) through in vitro transcription and circularization reactions; and verifying the correctness of its circular structure and sequence by gel electrophoresis, RNase R resistance experiments, and sequencing.

9. The use of the lipid nanoparticle delivery system according to any one of claims 1-5 in the preparation of a medicament for treating tumors.

10. A reagent kit, characterized in that, It comprises the lipid nanoparticle delivery system as described in any one of claims 1-5.

Images