Engineered extracellular vesicles for delivering functional Cas9/gRNA to eliminate hepatitis B virus cccDNA and integration

ABSTRACT

The persistence of HBV covalently closed circular DNA (cccDNA) and HBV integration into the host genome in infected hepatocytes pose significant challenges to the cure of chronic HBV infection. Although CRISPR/Cas9-mediated genome editing shows promise for targeted clearance of viral genomes, a safe and efficient delivery method is currently lacking. Here, we developed a novel approach by combining light-induced heterodimerization and protein acylation to enhance the loading efficiency of Cas9 protein into extracellular vesicles (EVs). Moreover, vesicular stomatitis virus-glycoprotein (VSV-G) was incorporated onto the EVs membrane, significantly facilitating the endosomal escape of Cas9 protein and increasing its gene editing activity in recipient cells. Our results demonstrated that engineered EVs containing Cas9/gRNA and VSV-G can effectively reduce viral antigens and cccDNA levels in the HBV-replicating and infected cell models. Notably, we also confirmed the antiviral activity and high safety of the engineered EVs in the HBV-replicating mouse model generated by hydrodynamic injection and the HBV transgenic mouse model. In conclusion, engineered EVs could successfully mediate functional CRISPR/Cas9 delivery both in vitro and in vivo, leading to the clearance of episomal cccDNA and integrated viral DNA fragments, and providing a novel therapeutic approach for curing chronic HBV infection.

KEYWORDS:

• Extracellular vesicles
• CRISPR/Cas9
• gene editing
• antiviral therapy
• hepatitis B virus

Introduction

Hepatitis B virus (HBV) infection is a major global health problem. Despite the ambitious goal of eliminating viral hepatitis as a significant public health threat by 2030, there are currently over 296 million individuals worldwide affected by chronic HBV infection, resulting in approximately 820,000 annual deaths attributed to end-stage liver disease [1]. The persistence of covalently closed circular DNA (cccDNA) in the nucleus of infected hepatocytes, serving as the template for viral replication, represents the primary obstacle to curing chronic HBV infection [2]. Moreover, HBV double-strand linear DNA, produced during viral replication, can permanently integrate into the host genome. Viral integration increases the risk of developing hepatocellular carcinoma and mediates the expression of hepatitis B surface antigen (HBsAg) independent of the HBV replication cycle [3]. While nucleos(t)ide analogues (NAs) and interferon-α (IFN-α) can effectively suppress viral replication and delay the progression of liver diseases [4], they fail to eliminate HBV cccDNA and the integrated viral DNA fragments. Consequently, it is challenging for patients with chronic hepatitis B (CHB) to achieve a functional cure characterized by HBsAg loss [5,6].

Over the past decade, Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR)/CRISPR-associated protein 9 (Cas9) has been widely utilized for gene editing [7]. Lin et al. [8] and Seeger et al. [9] demonstrated that CRISPR/Cas9 could induce cleavage of the viral genome and inhibit HBV replication. Our previous studies further revealed that the combination of dual guide RNA (gRNA) not only significantly suppressed viral transcription but also reduced the reservoir of cccDNA [10,11]. Furthermore, CRISPR/Cas9 exhibited the ability to suppress HBV replication in cell lines harbouring viral integration [12,13], indicating that it holds promise as a next-generation targeted antiviral therapy capable of eliminating both episomal cccDNA and integrated HBV DNA fragments.

However, the current challenge lies in the lack of an ideal delivery tool for CRISPR/Cas9. While viral vectors such as adeno-associated virus have been approved for clinical trials to deliver CRISPR/Cas9 for genetic disease treatment, concerns have been raised regarding the risks of integrated mutagenesis and off-target editing [14,15]. Synthetic material-based methods like lipid nanoparticles are cost-effective and suitable for large-scale production [16], but the lower efficiency and potential immunogenicity of in vivo delivery restrict their accessibility to target organs. Extracellular vesicles (EVs), as natural nanovesicles, demonstrate promising characteristics such as mediating transient delivery of biological cargo, good biocompatibility, and low immunogenicity. Our previous work showed that endogenous exosomes facilitated intercellular transfer of functional Cas9 protein and gRNA with confirmed antiviral activity, although the efficiency still required improvement [17].

Exogenous cargo loading strategies, such as co-incubation, electroporation, and ultrasonic treatment, may compromise the integrity of EVs and have limitations in payload capacity. In contrast, endogenous strategies have successfully achieved the natural loading of macromolecular cargo by genetically modifying cells to exploit the components involved in EVs biogenesis [18]. However, directly fusing cargo to the EVs-sorting proteins could potentially affect their therapeutic function in target cells, and cargo delivery is also limited due to low endosomal escape. To address these challenges, we explored the potential of light-induced protein interaction using the heterodimerization of Cryptochrome 2 (CRY2) protein and its ligand protein CIBN [19]. Previous studies showed that a simple membrane-anchored short peptide (the N-terminal acylation tag MGCINSKRKD-) could target cytoplasmic proteins to vesicle budding sites and subsequently into EVs through protein myristoylation and palmitoylation [20,21]. Combining this approach with light-induced interaction can further enhance the delivery of CRISPR/Cas9.

In this study, we developed engineered EVs with improved loading efficiency of Cas9/gRNA through light-controlled reversible heterodimerization and membrane-anchored acylation tag. Furthermore, to enhance Cas9 nuclear translocation in recipient cells, we decorated the membrane of engineered EVs with the fusogenic VSV-G. These modified EVs demonstrated the ability to suppress viral antigens and load by eliminating episomal cccDNA and integrated viral DNA fragments in HBV-replicating and infected models, suggesting their potential to cure chronic HBV infection.

Materials and methods

Cell cultures and transfection

HEK293T cells (ATCC), Huh7 cells (Chinese Academy of Sciences), and HepG2-NTCP [22] (a gift from Professor Kuanhui Xiang) were maintained in Dulbecco’s modified Eagle medium (DMEM) high glucose (Gibco) supplemented with 10% fetal bovine serum (FBS) (PAN) and penicillin/streptomycin (Gibco). HepAD38 cells [23] (a gift from Professor Ningshao Xia) were cultured in DMEM supplemented with 10% FBS, penicillin/streptomycin (Gibco), and 400 µg/mL G418 sulfate (Amerrosco). The cells mentioned above were cultured in a 37 °C incubator with 5% CO₂, and transfection of all plasmids was performed using polyethylenimine (PEI) (PolyScience).

Plasmid constructs

To construct PX458-Cas9-gHBV1 and PX458-Cas9-gHBV2, gHBV1 and gHBV2 were synthesized by Shanghai Sangon Biotech corporation and inserted into pSpCas9 (BB)-2A-GFP (PX458) respectively after annealing. CRY2 was custom synthesized by Shanghai Generay Biotech corporation and cloned downstream of Cas9 with a flexible linker consisting of glycine and serine to construct PX458-Cas9-CRY2, PX458-Cas9-CRY2-gHBV1, and PX458-Cas9-CRY2-gHBV2. The vector expressing Cas9-CRY2-eGFP fusion protein was constructed by removing the T2A sequence in PX458-Cas9-CRY2. Nluc was PCR amplified from the pNL1.1[Nluc] vector and cloned between Cas9 and CRY2 to generate PX458-Cas9-Nluc-CRY2.

With a myristoylation and palmitoylation (MyrPalm) sequence at the N-terminus, MyrPalm-CIBN was custom synthesized by Shanghai Generay Biotech corporation and inserted into the pcDNA3.1-Flag vector to construct pcDNA3.1-MyrPalm-CIBN. The DNA sequence of MyrPalm is 5’-ATGGGCTGCTGCAACCAACCAAGAAGAAAAGGAC-3’.

pBB4.5-HBV1.2 (1.2×HBV, genotype C) was constructed previously and preserved in the laboratory. pMD2.G, which expresses VSV-G, was purchased from Addgene.

DNA extraction

The total DNA of cells or tissues was extracted using QIAamp DNA Mini Kit (Qiagen) according to the manufacturer’s instructions and stored at -20°C until use.

PCR and sequencing

The primers used to amplify HBV DNA fragments from 1.2×HBV-transfected cells or mice injected with 1.2×HBV were HBV-251-F (5’-GACTCGTGGTGGACTTCTCTCAA-3’) and HBV-2137C-R (5’-CTGACTACTAATTCCCTGGATGCT-3’). HBV-251-F and HBV-2137C-R (5’-CTGACTACTAATTCCCTGGATGCT-3’) were used to amplify HBV DNA fragments from HBV-infected cells. Primers used to amplify integrated HBV DNA fragments from HBV transgenic mice were 1.28×HBV-F (5’-TCCTGCCTTAATGCTTTGTATG-3’) and 1.28×HBV-R (5’-ATCTCTGACGGAGAGAGGAAGAAGTC-3’).

PCR was performed with HiFi Taq (Genstar), and amplification products were detected by agarose gel electrophoresis. Fragments of the expected length were purified using NucleoSpin Gel and PCR Clean-up Kit (Macherey-Nagel) and then cloned into pEASY-T1 Simple vectors (TransGen) for sequencing.

EVs purification

HEK293T cells were transfected with plasmids and cultured in the medium containing 4% exosome-free FBS. After 48 h, the cell culture medium was collected and centrifuged at 4°C (800×g for 5 min followed by 2,000×g for 10 min and 10,000×g for 35 min) to remove cell debris. The supernatant was filtered with 0.22 μm filters (Merck Millipore) and then ultracentrifuged at 100,000×g for 2 h at 4°C using the SW32Ti swing-out rotor (Beckman Coulter). The pellets containing EVs were carefully washed with PBS, followed by ultracentrifugation at 100,000 g for 2 h. Finally, the EVs pellets were resuspended in PBS for further characterization and treatment.

Nanoparticle tracking analysis

The isolated EVs were diluted in filtered PBS to an appropriate concentration. The size distribution and concentration of EVs were measured using the Nanosight NS300 (Malvern Instruments). For each sample, three recordings were captured and analyzed using the NTA 3.3 software.

Transmission electron microscopy

The purified EVs were fixed in 4% osmium tetraoxide at 4°C for 30 min. Following the fixation step, the EVs were applied to a copper grid and stained with 1% phosphotungstic acid. After air drying, samples were observed using the JEM-1400 transmission electron microscopy (JEOL) operating at 60 kV to analyze the morphology of the purified EVs.

Western blot

Cell lysates or isolated EVs were lysed in RIPA lysis buffer for 20 min, and protein concentrations in the lysates were measured using the BCA Protein Assay Kit (Thermo Fisher Scientific). After protein quantification, the lysates were subjected to Western blotting under standard procedures and analyzed using the Odyssey Infrared Imager System (LI-COR) and the Chemiluminescence Imaging System (Tanon). The following primary antibodies and secondary antibodies were used as recommended: rabbit anti-Cas9 (Abcam), mouse anti-flag (Sigma), mouse anti-β actin (Servicebio), mouse anti-CD81 (Santa Cruz), mouse anti-HBcAg (a gift from Professor Ningshao Xia), fluorescence labelling anti-rabbit/mouse IgG (LI-COR) and HRP labelling anti-rabbit/mouse IgG (Cell Signaling Technology).

RNA extraction and reverse transcription

TRIzol (Invitrogen) was used to extract total RNA from cells, and EasyPure miRNA Kit (TransGen) was used to extract RNA from purified EVs. The Reverse transcription was performed using Transcriptor First Strand cDNA Synthesis Kit (Roche). The random primer was used for the reverse transcription of β-actin, and the specific primer used for gRNA reverse transcription was 5’-CGGCATAAGGAGTTCAAAAAAAGCACCGACTC-3’. The relative level of gRNA was analyzed by qPCR, and β-actin was used as an internal control.

Detection of luciferase activity

To measure the luciferase activity of EVs released by HEK293T cells post-transfection, EVs containing Cas9-Nluc-CRY2 protein were purified from the culture supernatants and resuspended in PBS. The purified EVs were mixed with Nano-Glo substrate diluted in buffer (Promega) and then subjected to luciferase activity detection using EnVision Multimode Plate Reader (PerkinElmer).

To detect the luciferase activity of EVs uptake by recipient cells, isolated EVs were added to Huh7 cells that had been seeded the day before and cultured for 6 h at a 37°C incubator with 5% CO₂. After incubation, Huh7 cells were washed with PBS and lysed in 0.1% TritonX-100 in PBS, and then incubated on a shaker for 10 min at room temperature. The luciferase activity of the cell lysates was measured as detailed above.

To determine the distribution of EVs containing Cas9-Nluc-CRY2 in mice after tail vein injection, tissues were harvested and lysed in 0.1% TritonX-100 using the Qiagen Tissue Lyser II according to the manufacturer’s instructions. The tissue lysates were diluted to a suitable concentration and mixed with buffer-diluted Nano-Glo substrate (Promega). The luciferase activity of the tissue lysates was measured as detailed above [24].

Confocal microscopy

To examine the uptake of EVs by recipient cells, the EVs purified from HEK293T cells were stained with the red fluorescent dye PKH26 (Sigma) for 5 min at room temperature, and neutralized using exosome-free FBS. The PKH26-labelled EVs were washed with PBS and centrifuged at 100,000 g for 2 h at 4°C. Huh7 cells were incubated with the resuspended PKH26-labelled EVs for 3 h, followed by washing with PBS. 4% paraformaldehyde (PFA) was used to fix cells at room temperature for 15 min, and Hoechst 33342 (Beyotime) was used to stain the nucleus of cells. The uptake of EVs by Huh7 cells was observed using the TCS-SP8 STED 3X confocal microscope (Leica).

For immunocytofluorescence analysis, cells were fixed in 4% PFA for 15 min and permeabilized with 0.5% TritonX-100 for 15 min, followed by incubation with anti-CD63 antibody (Abcam) overnight at 4°C. Cells were stained with Hoechst 33342 (Beyotime) and observed using the TCS-SP8 STED 3X confocal microscope (Leica).

Viral antigens and DNA levels

The culture supernatants of cells or sera from mice treated with EVs were collected, and chemiluminescence immunoassay was conducted using MAGLUMI X3 (Snibe) to measure the levels of HBeAg and HBsAg.

To remove plasmids, the samples were treated with DNase I for 1 h. The viral nucleic acid was then extracted using the EasyPure Viral DNA/RNA Kit (TransGen), and HBV DNA levels were quantified using quantitative PCR (qPCR).

Total DNA from the cells was extracted, and T5 exonuclease was utilized to digest the rcDNA at 37°C for 30 min. Subsequently, qPCR analysis was performed to measure the levels of intracellular cccDNA, and PRNP was used as an internal control [25].

HBV infection

HBV inoculum (genotype D) was purified from culture supernatants of HepAD38 cells using PEG8000 (Merck) precipitation as previously described [26]. The HepG2-NTCP cells were seeded for 24 h and then cultured in the medium containing 3% FBS, 2% dimethylsulfoxide (DMSO), and 1% non-essential amino acid (NEAA) for 24 h. The cultured cells were incubated with HBV inoculum in the presence of 8% PEG8000 and 2% DMSO for 24 h.

Mouse in vivo imaging experiment

C57BL/6 mice, with an average weight of 20 g, were injected with VSV-G coated EVs containing Cas9-Nluc-CRY2 or PBS (as negative control) via the tail vein. After 1 h, mice were injected with Furmazine diluted in PBS and anesthetized using isoflurane gas. The distribution of EVs was observed using IVIS Spectrum (PerkinElmer) in luminescence mode with a 60-second exposure time.

HBV-replicating mouse models

The HBV-replicating mouse model was established through the hydrodynamic injection of 1.2×HBV replicon (genotype C) into C57BL/6 mice. HBV transgenic mice carrying the HBV genome at a 1.28-fold length (genotype A) were purchased from Beijing Vitalstar Biotechnology Co., Ltd. All mice were accommodated in specific pathogen-free (SPF)-class housing with the laboratory. The animal experiments conducted complied with the animal welfare legislation and adhered to the principles outlined in the Guide for the Care and Use of Laboratory Animals.

Immunohistochemistry

After injection of EVs, the mice were euthanized, and their livers were collected and fixed in 4% PFA. Subsequently, the livers were embedded in paraffin, and sections of 5 μm thickness were prepared. Following dewaxing and rehydration, immunohistochemical (IHC) staining of the liver sections was performed using standard procedures. Anti-HBcAg (MXB) was used to detect the levels of intrahepatic HBcAg. Digital images of the stained sections were captured using the WS-10 scanning image system (Wisleap) and analyzed by NDP. View 2.7.52 software.

EdU incorporation experiment in mice

The mice were injected intraperitoneally with EdU solution at a dose of 50 mg/kg and euthanized after 2 h. Normal saline perfusion was performed on the harvested livers, which were subsequently fixed in 4% PFA. The livers were embedded in paraffin and sectioned. The histological sections were dewaxed and rehydrated, washed in glycine solution for 10 min, and stained using Cell-Light Apollo567 Stain Kit (RiboBio) according to the instructions of the manufacturer. Finally, the incorporation of EdU in the livers was analyzed using the TCS-SP8 STED 3X confocal microscope (Leica).

Statistical analyses

The unpaired Student’s t-test was performed using SPSS version 21.0 (SPSS). All statistical tests were two-tailed, and a p-value less than 0.05 was considered statistically significant.

Results

The Cas9-CRY2 fusion protein maintained its gene editing activity and translocated into the cytoplasm upon light-induced interaction

To construct a delivery system based on the light-induced heterodimerization, CRY2 was fused to Cas9 protein through a flexible linker (termed as Cas9-CRY2), and the oligopeptide (MGCCNSKRKD-) was fused to the N-terminal of CIBN protein (termed as MyrPalm-CIBN). The membrane anchoring of CIBN protein through N-myristylation and palmitoylation allowed the localization of Cas9-CRY2 protein within membranous vesicles under blue light (Figure 1A). Upon co-transfection in HEK293T cells, the eGFP-tagged Cas9-CRY2 protein, which contains a nuclear localization signal, demonstrated a predominant redistribution into the cytoplasm under blue light (Figure 1B). To evaluate the gene editing activity of the Cas9-CRY2 fusion protein, vectors expressing 1.2×HBV, Cas9-CRY2/gHBV1, and Cas9-CRY2/gHBV2 were transiently transfected into Huh7 cells. Electrophoresis analysis of the HBV replicon PCR products obtained from the transfected cells revealed cleavage of the HBV genome and the formation of a smaller fragment, which was indicative of cut and rejoining of the HBV replicon (Figure 1C) and consistent with the previous study [27]. Sequencing analysis showed that the sequence of the smaller PCR products corresponded to the editing sites of gHBV1 and gHBV2, indicating that the Cas9-CRY2 fusion protein retained gene editing function (Figure 1D).

Figure 1. The Cas9-CRY2 protein retained its gene editing function and translocated into the cytoplasm under blue light. (A) Diagram illustrating vectors expressing Cas9-CRY2/sgRNA or MyrPalm-CIBN and their light-induced interaction. (B) HEK293T cells were co-transfected with vectors expressing Cas9-CRY2-eGFP and MyrPalm-CIBN, followed by culturing under blue light or in the dark. After 48 h, the localization of Cas9-CRY2-eGFP was observed using confocal microscopy. (C) Huh7 cells were co-transfected with vectors expressing 1.2×HBV, Cas9-CRY2/gHBV1, and Cas9-CRY2/gHBV2. Additionally, vectors expressing Cas9/gHBV1 and Cas9/gHBV2 were transfected into Huh7 cells as a positive control. After 48 h, DNA was extracted and amplified by PCR using HBV-251-F and HBV-2137C-R primers. The PCR products were detected via agarose gel electrophoresis. (D) Diagram of PCR amplification after dual gRNA cleavage and representative sequence. The smaller PCR products (indicated by the red arrow) were purified and subjected to clone sequencing analysis.

Light-induced heterodimerization system promotes Cas9 protein loading into EVs

We next characterized the morphology and size distribution of EVs obtained from the conditioned medium using differential centrifugation. HEK293T cells were transfected with Cas9/gHBV1 alone as a control, or co-transfected with a combination of Cas9-CRY2/gHBV1 and MyrPalm-CIBN under blue light irradiation. Transmission electron microscopy (TEM) revealed a typical cup shape in both EVs-Ctrl-Cas9 (Figure 2A, left) and EVs-Cas9-CRY2 (Figure 2B, left), consistent with the morphological characteristics of exosomes. Nanoparticle tracking analysis showed that the average size of EVs-Ctrl-Cas9 was approximately 102.8 nm (Figure 2A, right), while the average size of EVs-Cas9-CRY2 was about 110.6 nm (Figure 2B, right), indicating that light-induced heterodimerization did not significantly alter the properties of EVs. To explore the impact of the light-controlled interaction system on Cas9 loading into EVs, the levels of Cas9-CRY2 protein in purified EVs were detected using SDS-PAGE. The purity of the isolated EVs was confirmed by the expression of CD81 (a positive EVs protein biomarker) and the absence of β-actin (Figure 2C). Notably, the highest level of Cas9-CRY2 fusion protein was observed in EVs derived from HEK293T cells co-expressing MyrPalm-CIBN under blue light (Figure 2C), suggesting that our modification strategy efficiently facilitated Cas9 protein sorting into EVs. Moreover, the presence of gRNA was detected in the purified EVs mentioned above, and its sequence was confirmed using Sanger sequencing, meeting the prerequisites for the formation of Cas9/gRNA ribonucleoprotein complex and subsequent gene editing activity of Cas9 protein (Figure 2D).

Figure 2. Protein acylation and light-induced interaction promoted encapsulation of Cas9/gRNA into EVs. (A) The vector expressing Cas9/gHBV1 was transfected in HEK293T cells, and EVs were purified from the culture supernatants through differential centrifugation. The morphology and size distribution of EVs-Ctrl-Cas9 were examined using JEM-1400 transmission electron microscopy (TEM) and Nanosight NS300. (B) HEK293T cells were co-transfected with vectors expressing Cas9-CRY2/gHBV1 and MyrPalm-CIBN, and cultured under blue light. The morphological structure and size distribution of EVs-Cas9-CRY2, isolated through differential ultracentrifugation, were analyzed by JEM-1400 TEM and Nanosight NS300. (C) Western blot analysis was performed on HEK293T cells expressing Cas9-CRY2/gHBV1 and MyrPalm-CIBN under blue light, as well as their purified EVs. The expression levels of Cas9-CRY2 and MyrPalm-CIBN in total cell lysates and EVs were examined. β-actin served as a negative protein biomarker for EVs, while CD81 was used as a positive protein biomarker. (D) The presence of gHBV1 in the purified EVs mentioned above was detected using qRT-PCR and further confirmed by Sanger sequencing. GAPDH was utilized as an internal control. Statistical analysis was conducted using Student's t-test with ***p < 0.001.

VSV-G enhances the nuclear translocation of Cas9 protein in recipient cells after EVs internalization

Since EVs exhibit distinct tropism for tissues under different conditions [28,29], we investigated whether recipient Huh7 cells could uptake EVs containing Cas9-CRY2/gHBV (Figure 3A). Purified EVs were labelled with PKH26 and subsequently incubated with Huh7 cells. Confocal microscopy analysis revealed prominent red fluorescence signals in the cytoplasm of Huh7 cells, indicating successful EVs internalization (Figure S1). To trace and quantify Cas9 within the EVs, a nano-luciferase (Nluc) tag was integrated into Cas9-CRY2 (Figure S2A). Luciferase activity was assessed in cell lysates from HEK293T donor cells, and no significant differences were observed between the groups (Figure S2B). However, HEK293T-derived EVs from the light-on and MyrPalm-CIBN co-transfected group exhibited the highest luciferase activity (Figure 3B). Moreover, the luciferase activity in Huh7 cells incubated with EVs described above surpassed that of any other group (Figure 3C), suggesting that the uptake of Cas9-CRY2 enriched by protein acylation and light-induced interaction with the EVs was remarkably efficient.

Figure 3. VSV-G enhanced the translocation of Cas9 protein into the nucleus of recipient cells following the uptake of EVs. (A) Experimental schematic for producing EVs containing Cas9-Nluc-CRY2. HEK293T cells were co-transfected with vectors expressing Cas9-Nluc-CRY2 and MyrPalm-CIBN, and cultured under blue light. EVs were purified from the conditioned medium and incubated with Huh7 cells in the dark for 6 h. (B-C) Luciferase activity was measured in EVs derived from HEK293T donor cells and total lysates of Huh7 recipient cells after incubation. RLU: relative luminescence units. Data were statistically analyzed using Student's t-test with *p < 0.05 and ***p < 0.001. (D) HEK293T cells were transiently co-transfected with vectors expressing Cas9-CRY2-eGFP and MyrPalm-CIBN, and cultured under blue light. EVs were purified and incubated with Huh7 cells in the dark for 24 h. The subcellular localization of Cas9-CRY2-eGFP in Huh7 recipient cells was observed under confocal microscopy. (E) Vectors expressing Cas9-CRY2/gHBV1, MyrPalm-CIBN, and VSV-G were co-transfected into HEK293T cells. After culturing the cells under blue light, EVs were isolated using differential centrifugation. The morphology and size distribution of EVs-Cas9-CRY2 + VSV-G were analyzed using JEM-1400 TEM and Nanosight NS300. (F) HEK293T cells were co-transfected with vectors expressing Cas9-CRY2-eGFP, MyrPalm-CIBN, and VSV-G and cultured under blue light. EVs containing Cas9-CRY2-eGFP and VSV-G were purified and incubated with Huh7 cells in the dark. The subcellular localization of Cas9-CRY2-eGFP in Huh7 recipient cells was observed under confocal microscopy at 6, 12, and 24 h post-treatment with EVs.

Proper nuclear localization of the Cas9 protein is crucial for its gene editing function. We examined the intracellular distribution of eGFP-tagged Cas9-CRY2 after 24 h incubation of EVs and Huh7 cells. Surprisingly, the green fluorescence signals were predominantly distributed in the cytoplasm of recipient cells, with minimal signal observed in the nucleus (Figure 3D). Previous studies demonstrated that endosomal entrapment significantly hinders the delivery of cargo via EVs [30,31]. Double staining of CD63, a classic marker for endosomal compartment, revealed a substantial overlap between the green fluorescence of eGFP tagged Cas9-CRY2 and the red fluorescent puncta of CD63, indicating that the internalized Cas9-CRY2 was trapped in CD63-positive compartments (Figure S3). To address this issue, we explored the use of hydroxychloroquine (HCQ), which increases endo-lysosomal pH and inhibits its degradation function [32,33]. However, despite pretreating Huh7 cells with HCQ, there was no significant improvement in the situation. Although the intensity of the eGFP signal seemed to be strengthened, the majority of Cas9-CRY2 remained entrapped (Figure S3).

Given that VSV-G can mediate membrane fusion in the acidic environment of endo-lysosomes, we hypothesized that the incorporation of VSV-G onto the EVs’ membrane could enhance the endosomal escape of Cas9 protein. Notably, the modification with VSV-G did not alter the average sizes or morphology of EVs (Figure 3E). Whereafter, Huh7 cells were treated with VSV-G coated EVs containing Cas9-CRY2-eGFP/gHBV and monitored under confocal microscopy at different time points. The results demonstrated a gradual accumulation of nuclear Cas9-CRY2-eGFP signals over time. At 6 h post-incubation, only faintly green fluorescence signals were observed in the nucleus of Huh7 cells, while at 12 and 24 h post-incubation, the presence of positive dots became more apparent, indicating successful translocation of Cas9-CRY2 into the recipient nucleus, facilitated by the fusogenic activity of VSV-G (Figure 3F). Taken together, our results highlight the successful delivery of Cas9-CRY2 into the nucleus of recipient Huh7 hepatoma cells using EVs modified with light-controlled interaction and VSV-G-mediated fusion.

Engineered EVs mediated CRISPR/Cas9 RNP delivery in HBV-replicating and infected cell models to suppress viral replication

To evaluate the antiviral effect of engineered EVs in the HBV-replicating cell model, Huh7 cells were transfected with a 1.2×HBV replicon and then incubated with VSV-G coated EVs containing Cas9/gHBV (termed as vEVs) for 48 h (Figure 4A). The levels of HBeAg and HBsAg in culture supernatants were determined using CLIA, and extracellular HBV DNA levels were quantified by qPCR. Treatment with vEVs significantly suppressed the HBeAg, HBsAg, and HBV DNA levels in supernatants of cultured Huh7 cells compared to the control groups (Figure 4B-D), suggesting that engineered EVs could inhibit viral replication in vitro. The efficiency of gene editing was evaluated through PCR and sequencing analysis. Smaller PCR products conforming to edited 1.2×HBV replicon were observed after vEVs treatment of Huh7 cells (Figure 4E). The sequence of these products was further characterized by Sanger sequencing, which demonstrated that vEVs delivering Cas9/gRNA RNP caused a precise sequence deletion in gHBV targeting sites (Figure 4F).

Figure 4. Antiviral effects of VSV-G coated EVs containing Cas9/gRNA RNP in the HBV-replicating cell model. (A) Experimental schematic for producing VSV-G packaged EVs containing Cas9/gHBV. Huh7 cells were transfected with a 1.2×HBV expression vector to construct the HBV-replicating cell model. Vectors expressing Cas9-CRY2/gHBV1, Cas9-CRY2/gHBV2, MyrPalm-CIBN, and VSV-G were transiently co-transfected into HEK293T cells. After culturing under blue light, EVs containing Cas9/gHBV and VSV-G were isolated from the conditioned medium and incubated with Huh7 cells in the dark for 48 h. (B-C) The levels of HBeAg and HBsAg in the supernatants of Huh7 cells after incubation with EVs were quantified using chemiluminescence immunoassay (CLIA). (D) HBV DNA was extracted from the culture supernatants of Huh7 cells, which had been treated with DNase I to remove cell-free DNA, and quantified using quantitative polymerase chain reaction (qPCR). Data from each group were statistically analyzed by Student’s t-test with *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. (E) Total DNA from Huh7 cells incubated with EVs was extracted and amplified using PCR with HBV-251-F and HBV-2137C-R primers. (F) The smaller PCR products (indicated by the red arrow) were purified and subjected to clone sequencing analysis.

The accessibility of the viral genome to Cas9 nuclease may differ between HBV replicon transfected cells and naturally infected cells. To simulate the viral life cycle in vitro, we utilized the human sodium taurocholate cotransporting polypeptide (NTCP) overexpressed HepG2 cell line (HepG2-NTCP) as a model for viral infection. Cells with established HBV infection (Figure S4B-C) were incubated with engineered EVs containing Cas9/gHBV and VSV-G for 48 h (Figure S4A). The intracellular HBV cccDNA was measured by qPCR. The HBcAg levels in lysates of HepG2-NTCP cells were detected by Western blot. The gene-editing capacity of VSV-G coated EVs containing Cas9/gHBV was determined by PCR and sequencing analysis. The levels of HBeAg, HBsAg, and HBV DNA in the culture supernatants were significantly reduced compared to the control groups treated with phosphate-buffered saline (PBS) or vEVs without gHBV (Figure 5A-C). Moreover, vEVs potently inhibited intracellular HBV cccDNA and HBcAg levels, as measured by qPCR and SDS-PAGE, respectively (Figure 5D-E). Notably, the engineered EVs successfully mediated the cleavage of the viral cccDNA genome in HBV-infected HepG2-NTCP cells only when Cas9 protein and gHBV were co-delivered (Figure 5F). Sequencing analysis revealed that the smaller PCR products were HBV DNA fragments with the viral sequence between the cleavage sites of the two HBV-specific gRNAs removed (Figure 5G). In summary, our engineered EVs demonstrated robust antiviral activity by targeted editing of the viral genomes in both HBV-replicating and infected cell models.

Figure 5. Antiviral effects of engineered EVs in the HBV-infected cell model. (A-C) HepG2-NTCP cells were infected with HBV and then incubated with EVs containing VSV-G and Cas9/gHBV. Subsequently, culture supernatants were collected to measure the levels of HBeAg, HBsAg, and HBV DNA using CLIA and qPCR. (D) Total DNA was extracted from HepG2-NTCP cell lysates post EVs incubation, and the relative level of cccDNA was quantified by qPCR. Data from each group were statistically analyzed by Student’s t-test with **p < 0.01, ***p < 0.001, and ****p < 0.0001. (E) The intracellular level of HBcAg in HepG2-NTCP cells was determined through Western blot, with β-actin serving as an internal control. (F) The total DNA of HepG2-NTCP cells incubated with EVs was amplified using PCR with HBV-251-F and HBV-2043D-R primers. (G) The smaller fragments (indicated by the red arrow) were purified and subjected to clone sequencing analysis.

Engineered EVs exhibited potent antiviral activity through targeted editing of episomal and integrated viral genomes in HBV-replicating mice models

After confirming the antiviral activity of engineered EVs in HBV-replicating and infected cell models, we proceeded to evaluate the gene-editing function of VSV-G coated EVs containing Cas9/gHBV in vivo. The EVs containing Cas9-Nluc-CRY2 fusion protein and VSV-G were purified from the conditioned medium of HEK293T cells and administered to mice via tail vein injection (Figure S5A). In vivo bioluminescence imaging experiment showed a significant accumulation of Cas9-Nluc-CRY2 in the liver of mice (Figure S5B). To rule out any confounding effects from substrate liver distribution, luciferase activity of major organs was quantified in vitro, confirming that EVs loaded with VSV-G and Cas9-Nluc-CRY2 mainly reached the liver after tail vein injection (Figure S5C).

The biodistribution pattern of EV supported our objective of targeting HBV in vivo. To establish the HBV-replicating mouse model, we performed hydrodynamic injection (HDI) of 1.2×HBV replicon. Twelve mice were randomly divided into three groups of four mice each. To determine whether vEVs could target and edit the episomal viral genome in vivo, purified EVs containing Cas9/gHBV and VSV-G were injected into the HBV-replicating mice through the tail vein, and serum levels of HBeAg and HBsAg were detected at 3 and 7 days post-treatment (Figure S6A). Seven days after injection, mice were sacrificed, and intrahepatic HBcAg levels were measured. PCR amplification and sequencing analysis were performed to assess viral genome editing in the mouse liver. The injection of vEVs resulted in a significant reduction in serum viral antigen levels (Figure 6A-B) as well as intrahepatic HBcAg levels (Figure 6C), indicating successful editing of episomal HBV replicon by vEVs carrying Cas9/gHBV (Figure 6D-E). Importantly, the injection of vEVs did not have a significant impact on the body weight of mice, confirming the safety of this novel therapeutic approach (Figure S6B).

Figure 6. Antiviral effects of engineered EVs in the HBV-replicating mouse model. (A-B) Serum levels of HBeAg and HBsAg were measured by CLIA. Data were statistically analyzed by Student’s t-test with *p < 0.05 and **p < 0.01. (C) Immunohistochemical staining (IHC) was used to detect HBcAg levels in the livers of mice 7 days post-injection. Scale bars: 100 μm. (D) To assess viral genome editing, total DNA was extracted from the liver tissue, and PCR amplification was performed using HBV-251-F and HBV-2137C-R primers. (E) Following dual-gRNA mediated cleavage, the smaller fragments (indicated by the red arrow) were purified and subjected to clone sequencing analysis.

During HBV infection, integration of the virus into the host genome is an unavoidable occurrence, leading to genomic instability and subsequent carcinogenesis in infected hepatocytes [34,35]. To evaluate the gene-editing efficacy of our engineered EVs in the integrated viral genome, we used HBV transgenic mice harbouring a 1.28×HBV genome. Twelve mice were randomly divided into two groups of six mice each. Similarly, by targeted editing of the transgenic HBV genome, both serum and intrahepatic viral antigens levels as well as serum viral loads were significantly reduced (Figure 7A-F).

Figure 7. Antiviral effects of engineered EVs in HBV transgenic mice model. (A-C) Serum levels of HBeAg and HBsAg were measured using CLIA. HBV DNA in the sera was detected by qPCR. Data were statistically analyzed by Student’s t-test with *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. (D) Intrahepatic levels of HBcAg in mice were detected by IHC. Zone 1: the portal spaces; Zone 3: the pericentral vein. Scale bars: 100 μm. (E) To detect the gene-editing of integrated viral fragments, total DNA was extracted from the liver of mice 3 days after EVs injection and amplified using PCR with 1.28×HBV-F and 1.28×HBV-R primers. (F) The smaller fragments (indicated by the red arrow) were purified and subjected to clone sequencing analysis.

The dual-gRNA mediated cleavage of the transgenic viral genome might cause genomic instability and result in cell apoptosis. It was noteworthy that all the viral replication markers rebounded at 7 days post vEVs injection compared to day 3 (Figure 7A-D), indicating a compensatory proliferation of unedited hepatocytes harbouring 1.28×HBV genome. To test the presumption that vEVs delivered Cas9-mediated genome editing contributed to hepatocyte apoptosis and regeneration, we performed the EdU incorporation assay. The proportion of cells incorporating EdU, indicating an underlying DNA duplication from liver regeneration, was higher in the treatment group than the control group after 3 days of vEVs injection (Figure S7A) but returned to normal levels at 7 days post-injection (Figure S7B).

Although embryonic integration was the situation of transgenic mice, and integrated hepatocytes only represent a small proportion of the liver in most treated CHB patients, the clearance of integrated hepatocytes by vEVs is still of benefit in clinical practice. Additionally, we found that vEVs were safe and well-tolerated, as indicated by no difference in body weight, serum aminotransferase levels, and liver hematoxylin–eosin (HE) stain in treated mice (Figure S8A-C).

Discussion

Current antiviral therapies, such as NAs and PEG-IFN-α, effectively inhibit HBV replication. However, achieving a functional cure, characterized by serum HBsAg loss, is rare in clinical practice [2]. As a powerful gene editing tool, CRISPR/Cas9 has the potential to eliminate viral infection by targeting both episomal and integrated HBV genomes [10,11,36], but the clinical application of this technology has been limited by the lack of safe and efficient delivery systems. EVs, natural nanovesicles, possess low immunogenicity and good biocompatibility, making them an ideal tool for drug delivery [28]. Here, we demonstrated that engineered EVs containing VSV-G and functional Cas9-CRY2/gHBV, relying on light-controlled heterodimerization and protein acylation, exhibited potent antiviral activity by targeted editing of episomal cccDNA and integrated HBV genome in HBV-replicating and infected models.

Previous research showed that endogenous exosomes can deliver Cas9/gRNA for gene-editing of viral genomes, but the efficiency of this delivery system still needs further improvement [17]. A variety of strategies have been explored to achieve therapeutic cargo loading based on EVs biogenesis, such as myristoylation tag, CD9, and CD63 [37,38]. Directly fusing cargo to the EVs-sorting protein terminals may impact their therapeutic function in target cells. To overcome this challenge, complex systems have been developed, such as light-induced dimerization [39,40] and chemical-induced protein interaction [41,42], allowing protein loaded in EVs to be released into the cytoplasm. Therefore, we enhanced the loading efficiency of CRISPR/Cas9 into EVs through light-induced interaction and acylation tag. However, a slight reduction in the gene-editing activity of the Cas9-CRY2 fusion protein was observed. It is worth noting that the CRY2 protein predominantly localizes in the cytoplasm [19], which may partially impact the nuclear localization of the Cas9 protein.

Besides effective protein loading of EVs, cargo delivery is also constrained by poor endosomal escape in recipient cells [43]. In our study, we observed that HCQ inhibited protein degradation by raising the pH of endo-lysosomes. However, this increase in pH also resulted in a higher retention of EVs within the endosomal compartment, which negatively affected the nuclear translocation of the Cas9 protein. In contrast, the incorporation of VSV-G onto the EVs promoted the release of cargo from EVs into the cytoplasm through low pH-induced membrane fusion, thereby enhancing the nuclear localization of Cas9 protein, similar to recent research [42].

Moreover, we conducted experiments to verify the gene-editing activity and antiviral capacity of engineered EVs in vivo and in vitro. The results showed that engineered EVs have the potential to eliminate the cccDNA reservoir and carcinogenic DNA integration. We targeted conserved regions of the viral genome using specific gRNA, leading to inhibitory effects on HBV of different genotypes [10]. In HBV-replicating and infected cell models, engineered EVs successfully disrupted the episomal viral genome of genotype C and genotype D, while also targeting integrated viral DNA fragments of genotype A in HBV transgenic mice. These findings showed that engineered EVs containing VSV-G and Cas9/gHBV have broad applicability in clinical practice for confronting HBV infection of various genotypes and mutants. However, it is important to note that the cell models used in our study cannot observe the long-term effects of engineered EVs treatment on viral replication. Additionally, there may be differences in the accessibility of gene-editing between plasmids and cccDNA. The HBV transgenic mouse model also lacks the cccDNA generated during natural viral infection. To better simulate the HBV life cycle, future studies should establish a humanized liver-chimeric mouse model with chronic HBV infection and conduct multiple injections of engineered EVs to observe their long-term therapeutic effects.

The outcome resulting from dual gRNA editing is contingent upon the efficiency and kinetics of DSBs at the target sites. Sequential DSBs lead to the introduction of point mutations at both target sites. Conversely, simultaneous DSBs cause the excision of the fragment between the two target sites [27]. If the repair system successfully re-ligates the ends created by the DSBs, smaller PCR products can be detected using specific primers. However, if the NHEJ system is inadequate or delayed in repairing the abundant Cas9-induced DSBs, the cleaved viral template will undergo degradation. Hence, the gene editing efficiency surpasses what is suggested by the PCR assay. This accounts for the more substantial reduction in viral antigens and DNA compared to the proportion of smaller PCR products. Furthermore, the smaller PCR products were subjected to clone sequencing and the results revealed that a majority of the sequences exhibited precise re-ligation of DNA ends. However, the analysis also showed a minority of cloned sequences with insertion mutations (Figure S9).

Apart from the gene-editing related antiviral activity of engineered EVs, the treatment of EVs in the absence of gHBV showed a minor inhibition of serum HBsAg in HBV-replicating mouse model, suggesting the presence of natural antiviral properties in HEK293T-derived EVs. However, it remains unclear whether the endogenous cargoes, such as microRNAs [44] and interferon-stimulated gene products [45], carried by EVs play a role in this phenomenon. Given the rapid degradation of cargo following EVs delivery, repeating dosing might be necessary for clinical applications. Therefore, it is important to consider the potential immunogenicity of VSV-G, a viral envelope protein used in this study. A recent study demonstrated that EVs containing VSV-G can enhance liver delivery efficiency through multiple injections [42]. Furthermore, different types of viral glycoproteins exhibit specific tropisms [46], and it is worthwhile to explore additional viral envelope proteins in future research to improve the safety and specificity of EVs delivery in vivo.

The destruction of viral genomes and clearance of hepatocytes containing integrated HBV DNA fragments mediated by engineered EVs is beneficial for achieving a functional cure with HBsAg loss in patients. However, it should be noted that the replenishment of cccDNA may weaken the antiviral efficacy of engineered EVs. In clinical applications, it is important to first reduce viral replication and inhibit cccDNA replenishment through efficient antiviral agents such as NAs. Afterward, engineered EVs can be utilized to deliver the gene-editing activity of Cas9 protein and gRNA, leading to clearance of the residual episomal viral genome and integrated viral DNA fragments.

In summary, we have developed a novel antiviral therapy utilizing engineered EVs carrying CRISPR/Cas9. Our engineered EVs have demonstrated potent antiviral activity and the potential of eliminating episomal HBV cccDNA and integrated HBV DNA in vivo and in vitro, paving the way for realizing a functional cure in patients with CHB.

Acknowledgments

F.L. and X.C. conceived the study. W.Z. and L.Z. designed experiments. W.Z., Y.Li, and T.Z. performed cell experiments. T.M. established the irradiation system. W.Z., J.Y., J.Z., Y.Liu, and J.N. performed mouse experiments. W.Z. and H.H. analyzed the data. W.Z. wrote the paper. F.L., X.C., and L.Z. reviewed and edited the paper.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data supporting this study’s findings are available from the corresponding author upon reasonable request.

Additional information

Funding

This work was supported by the National Key R&D Program of China [grant number 2022YFA1303600], and the National Natural Science Foundation of China [grant number 82102393, 82072280, and 82272315].