Introduction

Ebola virus disease is the deadliest infectious disease caused by infection with Ebola virus (EBOV), an enveloped, nonsegmented negative-sense RNA virus (Feldmann, Jones, Klenk, & Schnittler, 2003). The 19-kb viral genome comprises seven genes encoding the nucleoprotein (NP), virion protein 35 (VP35), VP40, glycoprotein (GP), VP30, VP24 and RNA-dependent RNA polymerase (L) (Mahanty & Bray, 2004). Inclusion bodies (IBs) formed in EBOV-infected cells are specialized intracellular compartments that serve as sites for EBOV replication and the generation of progeny viral RNPs (Hoenen et al., 2012; Nanbo, Watanabe, Halfmann, & Kawaoka, 2013). In IBs, the EBOV genome is replicated and transcribed by viral polymerase complexes (Misasi & Sullivan, 2014). VP35 serves as a cofactor of RNA-dependent RNA polymerase and contributes to viral replication by homo-oligomerization through a coiled-coil domain (Reid, Cardenas, & Basler, 2005), in addition to its recently discovered presence of phosphorylation and ubiquitination (van Tol et al., 2022; Zhu et al., 2020).

Innate interferon responses constitute the first lines of host defense against viral infection. Retinoic acid-inducible gene I (RIG-I)-like receptors, including RIG-I and melanoma differentiation-associated protein 5, play pivotal roles in the response to RNA virus infection. After the recognition of RNA virus infection, RIG-I is recruited to the mitochondrial antiviral adaptor protein (MAVS) through the caspase activation and recruitment domain. The activation of MAVS recruits multiple downstream signaling components to mitochondria, leading to the activation of inhibitor of κ-B kinase ε (IKKε) and TANK-binding kinase 1 (TBK1), which in turn phosphorylate IFN regulatory factor 3 (IRF3). Phosphorylated IRF3 forms a dimer that translocates to the nucleus, where it activates the transcription of type I interferon (IFN-I) genes (Fitzgerald et al., 2003; Liu et al., 2015).

To promote viral replication and persistence, viruses have evolved various strategies to evade or subvert host antiviral responses. For example, severe fever with thrombocytopenia syndrome virus (SFTSV) has developed a mechanism to evade host immune responses through the interaction between nonstructural proteins and IFN-I induction proteins, including TBK1, IRF3 and IRF7 (Hong et al., 2019; Lee & Shin, 2021; Ning et al., 2014; Wu et al., 2014), sequestering them inside SFTSV-induced cytoplasmic structures known as IBs. In addition to inhibiting IFN-I induction, SFTSV nonstructural proteins can hijack STAT1 and STAT2 in IBs to suppress IFN-I signaling (Ning et al., 2015). These studies highlight the role of viral IBs as virus-built “jails” that sequester some crucial host factors and interfere with the corresponding cellular processes.

EBOV uses various approaches to evade the host immune response, including antagonizing IFN production, inhibiting IFN signaling, and enhancing IFN resistance (Basler et al., 2000; McCarthy et al., 2016; Reid et al., 2006). VP35 is an IFN-I inhibitor that antagonizes host innate immunity by interacting with TBK1 and IKKε (Basler et al., 2003; Prins, Cardenas, & Basler, 2009), suppressing RNA silencing and inhibiting dendritic cell maturation (Haasnoot et al., 2007; Yen, Mulder, Martinez, & Basler, 2014). Here, we report that viral IBs in EBOV trVLPs-infected cells appear to play a role in immune evasion by sequestering IRF3 into IBs and preventing the interaction of IRF3 with TBK1 and IKKε.

Results

IRF3 is hijacked into cytoplasmic IBs in EBOV transcription and replication-competent virus-like particles infected cells

When HepG2 cells were infected with EBOV transcription and replication-competent virus-like particles (trVLPs) (Hoenen, Watt, Mora, & Feldmann, 2014), which authentically model the complete virus life cycle, IBs with a unique structure and viral particles formed in the cytoplasm (Fig. 1 – figure supplement 1A-B). Surprisingly, we found that a substantial percentage of endogenous IRF3 was trapped in viral IBs in EBOV trVLP-infected cells with large IBs (Fig. 1A), while no detectable TBK1 or IKKε, the essential upstream components of IRF3 signaling (Fitzgerald et al., 2003), was sequestered in the viral IBs (Fig. 1B and 1C). These results suggested that IRF3 was specifically compartmentalized in viral IBs, and this compartmentalization spatially isolated IRF3 from its upstream activators TBK1 and IKKε.

Fig. 1

Fig. 1– figure supplement 1.

IBs sequestration of IRF3 was further investigated at different hours post infection (hpi) of EBOV trVLPs. Detectable IRF3 puncta colocalized with viral proteins were apparent at 36 hpi in infected cells and correlated significantly with the size and shape of the viral IBs (Fig. 2A and 2B). As the size of IBs increased at 48 hpi, nearly all IRF3 colocalized with viral IBs, whereas the IRF3 distribution was completely different in the uninfected cells nearby (Fig. 2A and 2B). Using a fluorophore line of interest analysis, we assessed the intensity profiles of cytoplasmic IRF3 intensity in IBs as well as the increase in the diameter of the aggregates (Fig. 2C). As infection proceeded, the intensity of the IRF3 signal in the puncta increased as the level of cytoplasmic-dispersed IRF3 decreased (Fig. 2A), indicating that the total amount of IRF3 in the cells did not dramatically change during infection (Fig. 2D) and that only its subcellular localization changed. Taken together, the results above showed that IRF3, but not TBK1 or IKKε, was sequestered in viral IBs.

Fig. 2

EBOV trVLPs infection attenuates the TBK1-IRF3 association and IRF3 nuclear translocation

Upon virus infection, IRF3, as a critical transcription factor in the IFN induction pathway, can be phosphorylated and activated by TBK1, and then phosphorylated IRF3 translocates from the cytoplasm into the nucleus, eliciting the expression of antiviral IFNs. Given the sequestration of IRF3 by EBOV trVLPs in IBs, the TBK1-IRF3 association in EBOV trVLPs-infected cells was accessed by an in situ Duolink proximity ligation assay (PLA). Cytoplasmic complexes consisting of endogenous TBK1 with IRF3 (the red signal) were observed in HepG2 cells treated with poly(I:C), which induces the activation of the RIG-I signal cascade and IRF3 phosphorylation, and poly(I:C)-induced TBK1: IRF3 complexes were significantly reduced by EBOV trVLPs infection (Fig. 3A and 3B). Decreased TBK1-IRF3 association was further demonstrated by immunoprecipitation (Fig. 3C). Moreover, as shown in Fig. 3D, SeV infection-induced IRF3 phosphorylation and nuclear translocation were significantly inhibited by EBOV trVLPs (Fig. 3D, 4A and 4B). Importantly, IRF3 was also recruited into IBs-like compartments in the cytoplasm in the cells infected with live EBOV (Fig. 4C). These data collectively suggested that the EBOV-mediated sequestration of IRF3 in IBs blocks IRF3 phosphorylation and nuclear translocation in the TBK1-IRF3 signaling cascade, which is critical for IFN induction.

Fig. 3

Fig. 4

IBs-like structures formed by the viral proteins VP35 and NP play a key role in inducing IRF3 sequestration

Ectopic expression of NP alone (Noda, Watanabe, Sagara, & Kawaoka, 2007) or NP and the VP35 protein (Noda, Kolesnikova, Becker, & Kawaoka, 2011) in cells was sufficient to form IBs-like structures. To investigate the viral protein(s) involved in the sequestration of IRF3 in IBs, HepG2 cells were transiently transfected with plasmids encoding NP/VP35, NP/VP35/L, NP/VP35/L/VP30, NP/VP35/L/VP30/VP24, or NP/VP35/L/VP30/vRNA-RLuc/T7 and stained with anti-IRF3 and anti-NP at 48 hpi. Coexpression of NP and VP35 resulted in substantial sequestration of IRF3 in the IBs-like structure, which in turn resulted in a significant reduction of IRF3 in the nucleus, as observed in the cells transfected with vectors only and treated with poly(I:C) (Fig. 5A and 5B). Little if any VP35 or NP was demonstrated to interact with IRF3 by immunoprecipitation (Fig. 5 – figure supplement 1A-B). Compared to NP/VP35 coexpression, the presence of protein L, VP30 and VP24 showed little, if any, effect on IBs-like structure formation, IRF3 sequestration and nuclear IRF3 levels (Fig. 5A and 5B). These results suggested that IBs-like structures as well as VP35 expression were indispensable for IRF3 sequestration.

Fig. 5

Fig. 5– figure supplement 1.

VP35: STING interactions play an important role in isolating IRF3 into viral IBs

TBK1 and IKKε were spatially separated from VP35 upon infection by EBOV trVLPs (Fig. 1B and 1C), and IRF3 itself was demonstrated not to interact with VP35 (Fig. 5 – figure supplement 1A-B), implying that other IRF3-interacting proteins might be involved in IRF3 sequestration in IBs upon viral infection. Stimulator of IFN genes (STING), an endoplasmic reticulum adaptor associated with IRF3 (Petrasek et al., 2013), was observed to interact with VP35 (Fig. 6A) and be recruited into IBs when the cells were infected by EBOV trVLPs (Fig. 6B and 6C). A substantial portion of STING was found to be recruited into IBs at 36 hpi in EBOV trVLPs-infected cells (Fig. 6D, 6E and Fig. 6 – figure supplement 1). STING knockdown by siRNA inhibited IRF3 sequestration in viral IBs (Fig. 6F and 6G). These results suggested that STING played important roles in the sequestration of IRF3 in viral IBs, possibly by interacting with VP35.

Fig. 6

Fig. 6– figure supplement 1.

Viral IBs-induced IRF3 sequestration suppresses IFN-β production

EBOV trVLPs could hijack IRF3 and sequester IRF3 into IBs and thus block the nuclear translocation of IRF3, which suggested that EBOV trVLPs may suppress IRF3-driven IFN-β production. As reported previously (Basler et al., 2000), expression of VP35 (Fig. 7A), but not NP, resulted in a mild inhibition of SeV-induced IFN-β-Luc expression (Fig. 7B). Coexpression of VP35 and NP, which led to the formation of IBs and the sequestration of IRF3 (Fig. 5A), suppressed IFN-β-Luc expression much more potently than VP35 expression alone (Fig. 7B). Coexpression of NP/VP35/L/VP30 was more potent in the inhibition of SeV-induced IFN-β-Luc expression than NP/VP35 (Fig. 7B). Moreover, coexpression of NP/VP35/VP30/L almost completely suppressed poly(I:C)-induced IFN-β transcription (Fig. 7C). IRF3 depletion showed little, if any, effect on IFN-β transcription upon NP/VP35/L/VP30 coexpression (Fig. 7C), which suggested that NP/VP35/L/VP30 coexpression was similarly powerful as IRF3 depletion in antagonizing IFN-β expression. These results strongly suggested that viral IBs sequestration of IRF3 was substantially more powerful than that of VP35 expression.

Fig. 7

We next assessed the effect of IRF3 hijacking and sequestration by viral IBs on EBOV trVLPs replication. Compared with wild-type cells, IRF3 depletion showed little, if any, effect on EBOV replication, as indicated by luciferase activity, suggesting that trVLPs efficiently blocked IRF3 signaling (Fig. 7D). Moreover, overexpression of IRF3/5D (a phospho-mimic of activated IRF3), but not IRF3, inhibited EBOV trVLPs replication in IRF3-depleted cells (Fig. 7D). Taken together, these results suggest that the hijacking of IRF3 and sequestration into IBs by EBOV trVLPs can be significantly more potent in the inhibition of IFN-I production, thereby antagonizing the inhibitory effect of IFN-I on viral replication.

Discussion

Accumulating evidence suggests that EBOV has established multiple ways to antagonize host innate immune responses to maintain viral replication. Several EBOV proteins (VP35, VP24, GP, VP30 and VP40) are known to participate in host immune evasion to facilitate viral replication and pathogenesis (Audet & Kobinger, 2015; Bhattacharyya, 2021; Cantoni & Rossman, 2018). VP35 was demonstrated to suppress IFN-I production by inhibiting IRF3/7 phosphorylation, disrupting DC maturation, and facilitating the escape of immune sensation by dsRNA (Basler, 2015; Cardenas et al., 2006; Messaoudi, Amarasinghe, & Basler, 2015; Prins et al., 2009). VP30 and VP40 suppress RNA silencing by interacting with Dicer and modulating RNA interference components via exosomes, respectively (Fabozzi, Nabel, Dolan, & Sullivan, 2011; Pleet, DeMarino, Lepene, Aman, & Kashanchi, 2017). VP24 and GP are also known to block IFN-I signaling by hiding MHC-1 on the cell surface and counteracting tetherin or interfering with established immune responses by adsorbing antibodies against GP, respectively (Audet & Kobinger, 2015; Bhattacharyya, 2021).

Viral IBs are a characteristic of cellular EBOV infection and are important sites for viral RNA replication, and NP and VP35 are extremely critical proteins for the formation of IBs structures (Hoenen et al., 2012). However, whether viral IBs are involved in antagonizing IFN-I production during EBOV trVLPs infection has not yet been reported. Here, we found that IRF3 is hijacked and sequestered into EBOV IBs by viral infection (Fig. 1A), which demonstrates that viral IBs are utilized for IRF3 compartmentalization. Meanwhile, this compartmentalization resulted in the spatial isolation of IRF3 from the kinases TBK1 and IKKε (Fig. 1B and 1C). This suggests that IRF3 deprivation by viral IBs may antagonize host antiviral signaling by inhibiting IFN-I production signaling.

As expected, the expression of NP/VP35/VP30/L, which is involved in the composition of IBs, was significantly more antagonistic to SeV-induced IFN-β production than the expression of VP35 alone (Fig. 7B). In addition, the expression of NP/VP35/VP30/L can significantly antagonize the promoting effect of poly(I:C) on IFN-β transcription, and IRF3 knockout could not further inhibit the transcription of IFN-β (Fig. 7C), which may be because viral hijacking of IRF3 into IBs nearly completely antagonized its function of promoting IFN-β production. Furthermore, knockout of IRF3 in cells cannot further promote EBOV trVLPs replication compared to wild-type cells (Fig. 7D), which may be because IRF3 is hijacked into viral IBs and cannot be phosphorylated into the nucleus to regulate IFN-I production. These results suggest that viral IBs act as virus-built ‘jails’ to imprison transcription factors and present a novel and possible common mechanism of viral immune evasion in which the critical signaling molecule IRF3 is spatially segregated from the antiviral kinases TBK1 and IKKε.

Although almost all IRF3 could be sequestered to viral IBs formed by VP35 and NP (Fig. 5A and 5B), we found that neither VP35 nor NP interacted with IRF3 (Fig. 5 – figure supplement 1). Here, we found that VP35 interacts with STING and colocalizes in IBs and that knockdown of STING inhibits the sequestration of IRF3 in IBs (Fig. 6A-G). These results suggest that VP35 may hijack IRF3 into IBs through STING. However, whether other host proteins are involved in this process and the role of NP in the recruitment of IRF3 by VP35 remains unclear. In addition, we found that VP35 may hijack IRF3 into IBs via STING association (Fig. 6A-6C); however, whether VP35 activates the STING-IRF3 pathway in a cGAS-independent manner by interacting with STING and the molecular mechanism remain to be further investigated.

In summary, EBOV VP35 sequesters IRF3 into viral IBs and inhibits the association of IRF3 with TBK1 and IKKε, preventing IRF3 from entering the nucleus and thereby inhibiting IFN-I production (Fig. 8). Therefore, this study reveals a new strategy by which EBOV escapes the innate immune response and provides new ideas for Ebola virus disease treatment.

Fig. 8

Materials and Methods

Cell lines and transfections

HEK293, HeLa and IRF3-knockout HeLa cells (ABclonal, RM02113) were grown in Dulbecco’s modified Eagle’s medium (DMEM, Gibco). HepG2 cells were grown in minimum essential medium (MEM, Gibco) supplemented with a 1% nonessential amino acid solution (NEAA, Gibco). All media were supplemented with 10% heat-inactivated fetal bovine serum (Gibco), 2 mM L-glutamine, 100 units/ml penicillin and 100 units/ml streptomycin, and cells were grown at 37°C under an atmosphere with 5% CO₂. Transient transfection was performed with Lipofectamine 3000 (Invitrogen) according to the manufacturer’s instructions.

Vectors and viruses

Flag-tagged VP35, NP, STING, TBK1 and IRF3 vectors were constructed by cloning the corresponding gene fragments into a pcDNA3.0-based Flag-vector (Invitrogen). Myc-tagged VP35, IRF3 and IRF3/5D vectors were constructed by inserting the corresponding gene fragments into the pCMV-Myc vector (Clontech). All the constructs were validated by Sanger DNA sequencing.

SeV was amplified in 9- to 11-day embryonated specific pathogen-free (SPF) eggs. Live EBOV (Mayinga strain) is preserved by the BSL-4 Lab at the Wuhan Institute of Virology, Chinese Academy of Sciences.

Immunoprecipitation and immunoblot analysis

Cell lysates were prepared in lysis buffer containing 1% Nonidet P-40 and protease inhibitor cocktail (Roche) (Cao, Leng, & Kufe, 2003). Soluble proteins were immunoprecipitated using anti-Flag (M2, Sigma), anti-Myc (Sigma) or IgG of the same isotype from the same species as a negative control (Sigma). An aliquot of the total lysate (5%, v/v) was included as a control. Immunoblotting was performed with horseradish peroxidase (HRP)-conjugated anti-Myc (Sigma), HRP-conjugated anti-Flag (Sigma), HRP-conjugated anti-β-actin (Sigma), anti-VP35 (Creative Diagnostics), anti-IRF3 (Cell Signaling Technology), anti-STING (Proteintech) or anti-NP (Sino Biological) antibodies. The antigen-antibody complexes were visualized via chemiluminescence (Immobilon Western Chemiluminescent HRP Substrate, Millipore). A PageRuler Western marker (Thermo) was used as a molecular weight standard.

Gene silencing using siRNA

For gene knockdown in HepG2 cells, cells maintained in 6-well plates were transfected with 100 pmol STING small interfering RNA (siRNA) (sense, 5’- GCACCUGUGUCCUGGAGUATT -3’; antisense, 5’- UACUCCAGGACACAGGUGCTT -3’) or the same concentration of scrambled siRNA (sense, 5’- UUCUCCGAACGUGUCACGUTT -3’; antisense, 5’- ACGUGACACGUUCGGAGAATT -3’) purchased from Tsingke Biotechnology (Beijing, China) with Lipofectamine 3000 (Invitrogen) according to the manufacturer’s recommendations.

Reverse transcription and quantitative RT-PCR

Total cellular RNA was prepared using an RNeasy Mini kit (QIAgen, USA) according to the manufacturer’s protocol. For cDNA synthesis, 0.5 μg of RNA was first digested with gDNA Eraser to remove contaminated DNA and then reverse transcribed using ReverTra Ace qPCR RT Master Mix with gDNA Remover (FSQ-301, Toyobo) in a 20 μL reaction volume. Then, 1 μL of cDNA was used as a template for quantitative PCR. The following primers were used in these experiments: h-IFN-β F: 5’-AGGACAGGATGAACTTTGAC-3’ h-IFN-β R: 5’-TGATAGACATTAGCCAGGAG-3’ h-GAPDH F: 5’-AAggTCATCCCTgAgCTgAAC-3’ h-GAPDH R: 5’-ACgCCTgCTTCACCACCTTCT-3’

The samples were denatured at 95°C for 2 min, followed by 40 cycles of amplification (15 s at 94°C for denaturation, 60 s at 60°C for annealing and extension). Quantitative RT-PCR (qRT-PCR) was performed using SYBR Green Real-time PCR Master Mix (QPK-201, Toyobo) with the QuantStudio 6 Flex multicolor real-time PCR detection system (ABI). Relative mRNA levels were normalized to GAPDH levels and calculated using the 2^-ΔΔCT method (Livak & Schmittgen, 2001). The means (upper limit of the box) ± SEM (error bars) of three independent experiments are presented in the figures.

In situ proximity ligation assay

Duolink in situ PLA (Sigma) was used to detect the endogenous association of IRF3 and TBK1 in cells. In brief, HepG2 cells plated on glass coverslips were transfected with EBOV minigenome plasmids. After fixation with 4% formaldehyde, the cells were permeabilized with 0.3% Triton X-100 in PBS for 15 min. After blocking with blocking buffer (Sigma, DUO82007), the cells were incubated with mouse anti-IRF3 (Cell Signaling Technology) and rabbit anti-TBK1 (Abcam) primary antibodies. The nuclei were stained with DAPI (blue). The red fluorescent spots generated from the DNA amplification-based reporter system combined with oligonucleotide-labeled secondary antibodies were detected with a Zeiss LSM 800 Meta confocal microscope (Carl Zeiss).

Immunofluorescence microscopy

Cells were transfected, fixed, permeabilized and blocked as described above. Then, after incubation with anti-TBK1 (Cell Signaling Technology), anti-IKKε (Cell Signaling Technology), anti-IRF3 (Cell Signaling Technology), anti-VP35 (Creative Diagnostics), anti-NP (Sino Biological) or anti-STING (Bioss) antibodies overnight at 4°C, the cells were washed three times with blocking solution and then incubated with FITC- or TRITC-conjugated goat anti-rabbit (or anti-mouse) IgG. The cells were then stained with DAPI after washing and imaged using a laser scanning confocal microscope (Zeiss LSM 800 Meta) with a 63× oil immersion lens.

Luciferase reporter assay

The IFN-I production assay was performed as described previously (Zhu et al., 2022). Briefly, HEK293 cells (1×10⁵ cells per well in a 24-well plate) were cotransfected with the indicated amount of pCAGGS-NP (75 ng)/pCAGGS-VP35 (75 ng)/pCAGGS-VP30 (37.5 ng)/pCAGGS-L (500 ng), 200 ng of the IFN-β reporter plasmid (Promega, USA) and 4 ng of Renilla luciferase plasmid. An empty vector was used to ensure that each well contained the same plasmid concentration. After 24 h, the cells were treated with SeV (MOI=2) or 5 μg/ml poly(I:C) for 12 h, and the luciferase activity of the cell lysates was analyzed with the dual-luciferase reporter assay system (Promega, E1960) using a GloMax 20/20 luminometer (Promega, USA). Values were obtained by normalizing the luciferase values to the Renilla values. Fold induction was determined by setting the results from the group transfected with vector without Flag-VP35 to a value of 1.

EBOV trVLPs assay

The replication of EBOV in the cells was evaluated with the minigenome system (Hoenen et al., 2014). Briefly, producer cells (p0) were cotransfected with p4cis-vRNA-RLuc (250 ng) and pCAGGS-T7 (250 ng) for T7 RNA polymerase expression and 4 plasmids for EBOV protein expression (pCAGGS-NP (125 ng), pCAGGS-VP35 (125 ng), pCAGGS-VP30 (75 ng), and pCAGGS-L (1,000 ng)), as well as the luciferase reporter vector pGL3-Promoter (Youbio, 25 ng). One day after transfection, the medium was replaced with medium containing 5% FBS, and the cells were then incubated for another 3 days. Viral replication was determined by intracellular luciferase activities using a dual-luciferase reporter assay kit (Promega, E1960) after cell lysis with passive lysis buffer (PLB, Promega). For immunofluorescence experiments, cells were harvested 48 h after transfection.

Transmission electron microscopy (TEM)

HepG2 cells transfected with EBOV minigenome p0-related plasmids were washed with PBS, fixed with 2.5% glutaraldehyde, and then prestained with osmium tetroxide. Eighty-nanometer-thick serial sections were then cut and stained with uranyl acetate and lead citrate. Images were acquired with a transmission electron microscope (Hitachi, H-7650) operating at 80 kV.

EBOV infection assay

WT HepG2 cells grown in 12-well plates were incubated with the EBOV Mayinga strain at an MOI of 10 as determined by virus titration in Vero E6 cells at 37°C for 1 h for microscopy. Then, the cells were washed 3 times with PBS, and fresh medium was added to the cells, which were incubated at 37°C for 72 h. Subsequently, the cells were fixed with 4% formaldehyde for immunofluorescence microscopy as required by the BSL-4 laboratory. All work with live EBOV was performed with BSL-4 containment.

Statistical analyses

Graphical representation and statistical analyses were performed using Prism 8 software (GraphPad Software). Unless indicated otherwise, the results are presented as the means (upper limit of the box) ± SEM (error bars) from three independent experiments conducted in duplicate. An unpaired two-tailed t-test was used for the analysis of two groups. Data were considered significant when P < 0.05 (*), P < 0.01 (**), and P < 0.001 (***).

Data availability

All the data shown in this paper are available from the corresponding authors upon reasonable request.

Acknowledgements

This work was supported by the Advanced Customer Cultivation Project of Wuhan National Biosafety Laboratory, Chinese Academy of Sciences [2022ACCP-MS04] and the National Major Science and Technology Projects of China [2018ZX09711003-005-005 to T.G., 2022YFC2600704 to H.L.].

Competing Interests statement

The authors declare no competing interests.

Author Contributions

L.Z., X.L., and C.C. designed and supervised the study. L.Z., J.J., and T.W. performed the experiments. T.G., H.L., Q.D., Y.H., Y. J., Z.L, and P. L. analyzed the data. L.Z, X.L., and C.C. wrote the manuscript. All authors have read and approved the article.