Abstract
Seneca Valley virus (SVV) disease is a newly emerging infectious disease of pigs caused by SVV, which seriously endangers the pig industry. This study was set out to identify the essential host factors required for SVV entering porcine cells. Using a CRISPR/Cas9 library containing 93,859 sgRNAs that were designed to target approximately 22,707 porcine genes, we generated mutated porcine cell libraries, which were subjected to SVV challenge for enrichment of cells resistant to SVV infection. These resistant cells were subsequently analyzed to identify genes essential for SVV infection. We demonstrated that ANTXR1, a type I transmembrane protein encoded by ANTXR1, heparan sulfate (HS), glycosaminoglycans modified by acetylation and sulfation of HS2ST1, and Neu5Gc, a non-human sialic acid catalyzed by CMAH, were the essential host factors for SVV entry into porcine cells. These results will be helpful to elucidate the pathogenesis of SVV and the development of prevention and control measures.
Introduction
Seneca Valley virus (SVV), belongs to the Senecavirus genus in the picornaviridae family, is responsible for a porcine idiopathic vesicular disease. Since 2014, SVV has been confirmed to be the causative agent of a newly emerging swine epidemic in the US 1–4, Brazil 5–8 and China 9, 10. To date, the global outbreak of SVV in swine has caused a sharp decline in the production of neonatal piglets and significant economic losses. However, no licensed vaccine or antiviral therapy is available yet, highlighting an urgent need for basic research on SVV.
Viruses rely on the host to complete their life cycles. Host cell entry factors play undoubtedly key roles determining the viral host range, tissue tropism, and viral pathogenesis. A thorough comprehension of interaction between host factors and virus will be the key to assess the impact of virus and will be helpful for finding better preventive and therapeutic tools. For example, as angiotensin-converting enzyme 2 (ACE2) was identified as the specific host receptor of severe acute respiratory syndrome coronavirus (SARS-CoV) 11, engineered soluble ACE2 has been used to compete with host receptor and thus prevent SARS-CoV-2 entry 12. In addition, gene editing of host factors has recently been shown to be an effective way defending against viruses in animal husbandry 13. For instance, CD163 serves as an uncoating receptor of porcine reproductive and respiratory syndrome virus (PRRSV). Based on understanding the key domains of CD163 interacting with PRRSV, pigs with precise gene editing of CD163 exhibited a full resistance to the porcine reproductive and respiratory syndrome (PRRS) 14, 15, 16.
Isolated in the cell culture of human fetal retinoblasts in 2002 17, SVV was initially reported as a nonpathogenic oncolytic virus, as SVV showed high specificity for some human tumor cells rather than normal tissues. The anthrax toxin receptor-1 (ANTXR1) of human origin, alternatively named tumor endothelial marker 8 (TEM8) highly expressing in some cancer tissue as a tumor-specific endothelial marker, was previously identified as a SVV receptor in neuroendocrine cancer 18. It explained the tropism of SVV to human tumor cells. However, the essential host factors for SVV entry are yet to be determined in pigs.
CRISPR/Cas9-based loss-function libraries have recently been widely used for characterization of host factors associated with virus infections, such as SARS-CoV-2 19, Rift Valley fever virus (RVFV) 20, and alphaviruses 21. In this work, a genome-wide CRISPR/Cas9 single-guide RNA library was designed for the pig and subsequently used to screen host factors for SVV susceptibility, and the functional role of the identified genes in mediating SVV infection was investigated to clarify the mechanism underlying infection, which allowed insights into prevention and treatment of the disease.
Results
A CRISPR/Cas9 mediated genome-wide knockout library was generated using IBRS-2 cells
To identify host factors required for SVV infection, we constructed a genome-scale loss-of-function library containing 93,859 CRISPR single-guide RNAs (sgRNAs), targeting 22,707 genes in the swine genome. To guarantee the targeting efficiency and to avoid the effects of off-targeting, we designed approximately five sgRNAs for each gene, and the target sites were mainly chosen in the first exon of the genes to improve the efficiency of the sgRNA. In addition, one non-targeting sgRNAs was included as a negative control. We subsequently integrated an all-in-one expression vector to simultaneously deliver Cas9, a sgRNA, and a puromycin resistance gene into the target cells (Fig. 1a). We measured that the 92,918 designed sgRNA oligonucleotides were successfully cloned into the plasmid library, and the relative abundance difference between the highest 10% sgRNA and the lowest 10% sgRNA was within 15 fold (Fig. 1b). The plasmid pools were packaged into lentivirus in HEK293-T cells. 1.2×108 IBRS-2 cells, which are permissive to SVV, were transduced with the lentivirus library at 0.2 MOI to ensure that the majority of cells harbored one sgRNA (Fig. 1c). The cells were cultured for 7 days after a puromycin selection was completed to make sure that the protein functions were destroyed. The genomic DNA was isolated and the sgRNA regions were amplified by PCR. Deep sequencing results showed that 85,151 guide RNAs were retained in genome, covering about 91% of the original sequences (Fig. 1d).
A candidate gene list for Seneca Valley virus infection was obtained by CRISPR/Cas9 screen
The selected cell pools were challenged with SVV at 1.0 MOI for 24 h, with a clear cytopathic effect (CPE). The DNA was extracted from the surviving cells for PCR amplification and then subjected to next-generation sequencing (NGS) and analysis. A significant reduction in the diversity of sgRNAs in the surviving cells was observed, reflecting that the sgRNAs targeted essential genes (Fig. 2a, Supplementary Fig. 1a, b).
The top 7 genes that showed the highest screen scores in three independent experiments are shown in Fig. 2b and Table S1: ENSSSCG00000022032 (heparan sulfate 2-O-sulfotransferase 1, HS2ST1), ENSSSCG00000008340 (ANTXR cell adhesion molecule 1, ANTXR1), ENSSSCG00000001099 (cytidine monophospho-N-acetylneuraminic acid hydroxylase, CMAH), ENSSSCG00000027031, ENSSSCG00000018405, ENSSSCG00000028377 (crystallin zeta like 1, CRYZL1), and ENSSSCG00000011672 (RAS p21 protein activator 2, RASA2). We further performed Gene Ontology (GO) analysis (Fig. 2c) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis (Fig. 2d) on the data. From the KEGG analysis, we found that the glycosaminoglycan synthesis pathway not only showed the most significant enrichment but also contained the most abundant pathway. Of note, the genes from the enriched list highly likely to be involved in viral entry were: ANTXR1, encoding a type I transmembrane protein; HS2ST1, relating to heparan sulfate (HS) modification and CMAH, responsible for catalytic synthesis of sialic acid.
Knockout of ANTXR1 leads to a significant reduction of Seneca Valley virus permissivity
The second most significantly enriched gene in the SVV screen library were found to be the ANTXR1 gene, encoding a single-pass cell surface protein consisting of an intracellular region, transmembrane region, and an extracellular region with a von Willebrand Factor A (vWA) domain 22, 23, 24. To validate the function of ANTXR1 in SVV infection in pigs, knockout cells (PK-15 and CRL-2843) were generated using the CRISPR/Cas9 system (Supplementary Fig. 2a, b). The ANTXR1 KO cell lines were infected with SVV for 12 h. As expected, viral replication was significantly decreased as shown by qPCR (Fig. 3a, e). A reduction in the protein level was also confirmed by Western blotting (Fig. 3b, f). Cytopathic effects (CPE) were not observed in CRL-2843 after ANTXR1 knockout (Supplementary Fig. 3a), and no virus was observed in the knockout (KO) cell lines using immunofluorescence staining (Fig. 3g, Supplementary Fig. 3b), indicating that the cell lines had gained protection against the SVV infection. Moreover, exogenous expression of the ANTXR1 restored the susceptibility of gene knockout lines to viral permissivity, suggesting that the antiviral phenotype of cells is indeed caused by the deletion of ANTXR1 (Fig. 3c, d).
ANTXR1 is the cellular receptor for Seneca Valley virus
To further analyze the potential role of pig ANTXR1 in SVV entry, we examined whether viral entry into the cells was inhibited in the absence of ANTXR1. Incubating with SVV at 4 °C for 1 h or 37 °C for 30 min respectively, ANTXR1 KO cell lines displayed a significantly decreased viral attachment and internalization (Fig. 4a, b). Furthermore, ectopic expression of pig ANTXR1 enabled SVV attachment (Fig. 4c) and an 11-fold increase in virus infection for 24 h (Fig. 4d) in MDCK, a cell line which normally is not permissive for SVV infection.
The pcDNA3.1 vector containing pig ANTXR1 was transfected into HEK293-T cells and SVV was added 24 h later. SVV-ANTXR1 co-localization was observed after 8 h (Fig. 4e). We also tested the co-localization of endogenous ANTXR1 and SVV in IBRS-2 cells (Fig. 4f). To investigate interactions between ANTXR1 and SVV structural proteins, HEK293-T cells were transfected using a Myc–ANTXR1 expressing vector, with Flag-tagged viral structural proteins VP1, VP2, VP3 expressing plasmids, respectively. Analyzed by Western blotting, we found that ANTXR1 interacts with VP1 and VP2 (Fig. 4g). These results supported the hypothesis that pig ANTXR1 is the cellular receptor for SVV infection.
Heparan sulfate is associated with Seneca Valley virus infection
A large number of genes encoding enzymes involved in heparan sulfate (HS) synthesis and modification were enriched (p<0.01) in the candidate gene list, such as B3GAT3 (ranking number 2211), EXT1 (ranking number 37), EXT2 (ranking number 2492), NDST2 (ranking number 3559), NDST3 (ranking number 257), HS2ST1 (ranked at the very top) and HS3ST5 (ranking number 1295). Most strikingly, we found that the HS synthesis pathway was upregulated in the KEGG analysis in the transcriptome sequences of wild-type cells after SVV infection, including EXTL1, EXT1, HS3ST1 and HS3ST3B1 (Fig. 5a, Supplementary Fig. 4).
Therefore, to investigate whether HS serves as an essential factor for SVV infection, we performed mutations in EXT1, HS2ST1 and HS3ST5 (Supplementary Fig. 5a-c) and then measured the effect on SVV infection. The relative amount of SVV mRNA was around 5-fold lower in EXT1 KO cells compared to wild type PK-15 cells (Fig. 5b), while the relative mRNA of SVV was reduced 5-fold to 10-fold in the HS2ST1 KO cells (Fig. 5c), 2-fold to 12-fold in the HS3ST5 KO cells compared to wild type PK-15 cells (Fig. 5d), reflecting that the KO cell lines gained resistance to SVV infection. Western blotting also confirmed the low level of SVV-VP2 protein in KO cells as compared to wild type cells (Fig. 5e, f).
Soluble heparin sodium, an analogue of heparan sulfate, is usually used as a HS competitive binding reagent. SVV was added to PK-15 cells after treatment with different concentrations of soluble heparin sodium for 1 h, which reduced viral infection substantially at 0.5 mg/mL and higher concentrations (Fig. 5g). Under the same condition, SVV infection was also reduced by soluble heparin sodium in CRL-2843 cells (Fig. 5h).
Heparan sulfate affects virus attachment
To determine whether HS could affect SVV binding to the cell surface, the ability of SVV attachment and internalization were further analyzed in HS2ST1 and EXT1KO cell lines. The SVV attachment and internalization were significantly reduced in the HS2ST1 KO cells compared with the wild type PK-15 cells (Fig. 6a, b). The same results were also observed in the EXT1 KO cells (Fig. 6c, d). In addition, heparin sodium also very potently reduced the SVV attachment and internalization (Fig. 6e-h). These data further demonstrated that HS serves as an adhesion factor for SVV.
Functional validation of CMAH in SVV infection
To validate the function of CMAH in SVV infection, knockout cells (IBRS-2) were generated using the CRISPR/Cas9 system (Supplementary Fig. 6). The CMAH KO cell lines were infected with SVV for 12 h, and the relative amount of mRNA encoding SVV was about 4-fold lower in CMAH KO cells compared to wild type IBRS-2 cells (Fig. 7a). Western blotting results indicated that viral VP2 was decreased significantly (Fig. 7b). Moreover, another CMAH KO cell line (IBRS-2) also had a similar phenotype (Fig. 7c). As demonstrated in Fig. 7d, ectopic expression of pig CMAH restored the susceptibility of CMAH KO lines, suggesting that SVV infection depends on CMAH.
Seneca Valley virus entry depends on CMAH
CMAH is the key enzyme to synthesize Neu5Ac into Neu5Gc. Previous studies have shown that Neu5Gc is a cell surface receptor for influenza virus. We thus measured viral attachment, viral entry, and internalization to determine how CMAH affects SVV infection. CMAH KO cells showed significantly reduced SVV attachment and internalization (Fig. 7e, f). The ectopic expression of pig CMAH increased SVV attachment (Fig. 7g), whereas the antibody of Neu5Gc reduced SVV infection significantly (Fig. 7h). These results demonstrated that SVV entry depends on Neu5Gc.
Discussion
We applied CRISPR/Cas9 for genome-wide screening in porcine cells in order to elucidate the host factors determining the susceptibility to SVV infection, showing that ANTXR1, heparan sulfate and Neu5Gc play important roles in SVV entry.
ANTXR1 has been confirmed as a SVV receptor in selected human tumor cells. According to previous structural studies, the R88 site and D156 site of human ANTXR1 are critical amino acid residues interacting with SVV 25, 26. However, the corresponding sites in pig ANTXR1 mutate to other amino acids. Despite the difference, we demonstrated in this study that ANTXR1 also acts a receptor mediating SVV entry into porcine cells, and it physically interacts with VP1 and VP2 of SVV. As SVV infects pigs but not humans 6, this finding suggests that more complicated factors but not a single receptor determine the species tropism of SVV.
The conservation of ANTXR1 among different species suggests that it may have other unknown important physiological functions. As reported, ANTXR1 may be associated with collagens binding and promotion of ECs migration and misaligned incisors are observed in adult ANTXR1 KO mice 27. Mutations in ANTXR1 cause progressive extracellular-matrix accumulation in patients with the GAPO syndrome, a complex phenotype consisting of growth retardation, alopecia, pseudoanodontia and progressive optic atrophy 28. Recently, ANTXR1 knockout pigs were produced, and as expected, exhibited resistance to SVV infection, while these pigs also developed GAPO-like symptoms 29. Based on the above results, more explicit structural investigations on the SVV-ANTXR1 complex are warranted in order to design accurate editing sites that destroy the function of virus receptor while maintaining its normal physiological function 30.
Heparan sulfate (HS) is a highly sulfated polysaccharide, which is widely distributed on the surface of cell membranes, basement membranes and the extracellular matrix. Its synthesis is catalyzed by a series of synthases and modifying enzymes 31. Various structures and modifications lead to the complexity and diversity of biological properties of heparan sulfate, including cell adhesion, regulation of cell growth and proliferation, and development processes 32, 33, 34. Because of its anionic characteristics and high-density negative charge, heparan sulfate possesses the ability to interact with viruses. Heparan sulfate has been identified as receptor or co-receptor for certain types of viruses, such as SARS-CoV2 35, lymphocytic choriomeningitis virus (LCMV) 36, and chikungunya virus 37. A series of HS synthetases and sulfation modifying enzymes were shown to be enriched in our experiment, and we demonstrated that heparan sulfate is also a receptor for SVV. By exogenous addition of soluble heparin sodium, competitive binding of heparan sulfate and heparin sodium to SVV was studied. We found that SVV invasion and infection were both impaired. This suggests that low-cost small molecule therapeutic drugs may block the combination of virus and HS, so as to reduce the virus infection in pigs.
Sialic acids constitute a 9-carbon monosaccharide family that is important for a wide variety of biological events. The predominant sialic acids in mammals are N-acetylneuraminic acid (Neu5Ac) and N-glycolylneuraminic acid (Neu5Gc), the latter is formed by hydroxylation by CMAH. The CMAH−/− knockout mice model was previously used to investigate the effect of Neu5Gc on influenza virus infection, showing that Neu5Gc acted as a functional receptor 38. Porcine CMAH KO lessens the severity of the PEDV infection and delays its occurrence 39. A previous study showed that sialic acid played a role in mediating SVV-GFP infectivity, but the mechanism was unclear 40. Our results indicated that SVV entry depended on CMAH, suggesting that the expression of Neu5Gc is conducive to SVV infection. Interestingly, compared with other mammals, human CMAH has been evolutionarily mutated and inactivated, which means that humans cannot produce Neu5Gc 41, and we speculate that this may also be one of the reasons for the different consequences of SVV infection in humans and pigs.
In conclusion, ANTXR1, heparan sulfate and Neu5Gc were shown to act as host factors of virus entry by our screen, and SVV disease prevention and treatment could potentially benefit from these results. These findings provide significant clues to design new vaccines, develop effective drugs, and evaluate a genetic breeding strategy.
Materials
Cells and viruses
IBRS-2, PK-15, MDCK, HEK293-T cells and the derived mutant cell lines were cultured in Dulbecco’s Modified Eagle Medium supplemented with 10% fetal bovine serum (FBS), and 50 μg/mL streptomycin. CRL-2843 and the derived mutant cell lines were cultured in Hyclone RPMI-1640 supplemented with 10% FBS and 50 μg/mL streptomycin. All cell lines were tested and judged free of mycoplasma contamination and maintained at 37 °C (5% CO2).
SVV strain isolated in National Foot and Mouth Diseases Reference Laboratory, Lanzhou Veterinary Research Institute, Chinese Academy of Agricultural Sciences was used for screening and validation studies. These SVV strains were propagated in IBRS-2 cells, and viral titers were determined using a doubling dilution assay; the titers were denoted as the 50% cell culture infective dose (TCID50)/mL, as determined using the Reed-Muench method. The titrated viruses were preserved at −80 °C (TCID50=7.89).
To examine the susceptibilities of cells to SVV, the cells were plated 1 to 2 days prior to inoculation, and then challenged with SVV at the indicated MOI (using the basic medium for dilution) at 37 °C for appropriate time. The viral infection state was evaluated for subsequent analysis.
Screen library construction
A pooled library encompassing 93,859 different sgRNAs targeting 22,707 pig genes was designed by the Wu laboratory, China Agricultural University, reference genome Sscrofa10.2 from https://www.ensembl.org. The sgRNAs was packaged using a lentivirus library by GenScript, NanJing, China. 1.2×108 IBRS-2 cells were transduced with lentiviruses at a multiplicity of infection (MOI = 0.2). After 2 μg/mL puromycin selection for 3 days, cells were challenged with SVV (MOI = 1) for 24 h. The experiments were performed in three independent experiments. Genomic DNA was extracted from the uninfected cells or the surviving cells. Using Specific primers located on both sides of the sgRNA frame, sgRNA sequences were amplified two rounds, and subjected to next generation sequencing. A lentivirus carried a non-sense sgRNA was transduced into 1×107 as a control cell library. All the used sgRNA amplified primers in this study are listed in Table S2.
Establishment of single gene mutant cell
The three plasmids were transfected into cell (jetPRIME, polyplus) in combination to manufacture a single gene mutant cell line - a plasmid expressing Cas9 protein regulated by 4 μg/mL doxmycin with PB transposon, a plasmid connecting sgRNA targeting gene with PB transposon and puromycin screen marker, and a vector expressing transposase. Genomic DNA was extracted from samples using the DNeasy Blood & Tissue Kit (Qiagen) and positive cells were picked. All the used sgRNA sequence and gene targeting site amplified primers in this study are listed in Table S3 and Table S4.
Overexpression experiments
The cDNA of ANTXR1 was amplified from PK-15 cells, The cDNA of CMAH was amplified from pig kidney issues and cloned into pcDNA3.1(+)-Myc vector (Invitrogen, Carlsbad, CA, USA) to yield the C-terminal Myc-tagged expression construct. All the used cDNA amplified primers in this study are listed in Table S5.
Western blotting
Cells were lysed using an immunoprecipitation (IP) lysis buffer containing protease inhibitors (Biotechnology, China). The protein concentrations of the extracts were measured with a BCA assay (Beyotime, P0012). Equal amounts of protein from each sample were analyzed by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto polyvinylidene difluoride (PVDF) membranes, which were blocked with 5% skimmed milk and 0.5% Tween-20 in Tris-buffered saline (TBST) at room temperature for one hour and incubated with primary antibodies at 4 °C. The Myc-tagged proteins were probed with a mouse anti-Myc tag monoclonal IgG antibody (1:1000 dilution, CST). The anti-VP2 polyclonal antibody was prepared in our laboratory. Subsequently, the membranes were incubated with a goat anti-rabbit IgG secondary antibody (1:10,000 dilution, Beyotime) or a goat anti-mouse IgG secondary antibody (1:10,000 dilution, Beyotime) for one hour at room temperature, followed by washes in TBST. Proteins were visualized with chemiluminescence and normalized by β-tubulin or β-actin.
Indirect immunofluorescence assay
To detect SVV infection in mutant cells, an indirect immunofluorescence assay (IFA) was performed. Cells were cultured in 12-well plates and challenged by SVV (MOI = 1). At 8 hpi (hours post infection), the cells were fixed with 4% paraformaldehyde for 30 min and then permeabilized with 0.25% Triton X-100 at room temperature for 5 min. After three times washes with phosphate-buffered saline (PBS), the cells were blocked with 5% bovine serum albumin in PBS for 30 min. Thereafter, the cells were incubated with an anti-VP2 primary antibody overnight at 4 °C. The fluorochrome-conjugated secondary antibody was added in the dark for 6 h at 4 °C. After that, the cells were stained with 4’6’-diamidino-2-phenylindole (DAPI) for 10 min to reveal the nuclei. The fluorescence was visualized using a EVOS M500I imaging System (Invitrogen).
RNA extraction and quantitative real-time PCR (qRT-PCR)
Total RNA was extracted with TRIzol (Magen) according to the manufacturer’s instructions. G490 5 × All-In-One RT MasterMix (abm) was used for reverse transcription according to the manufacturer’s protocol (Promega). The qRT-PCR analysis was performed in 96-well plates using the BIO-RAD CFX96 detection system. The relative expression level of these genes was calculated using the 2-ΔΔct method, and GAPDH (glyceraldehyde-3-phosphate dehydrogenase) mRNA was used as the endogenous control. qRT-PCR was performed on each sample in triplicate. All the used qPCR primers in this study are listed in Table S6.
Coimmunoprecipitations
HEK293-T cells grown in 100 mm plates were transiently co-transfected with 8 µg of Myc-ANTXR1 and 8 µg of Flag-SVV-structure-protein plasmid. The transfectants were harvested 24 h after transfection and subjected to immunoprecipitation assay. The cells were lysed in NP-40 lysis buffer (1% NP-40, 50 mM Tris (pH 8.0), 5 mM EDTA, 150 mM NaCl, 2 mg/mL leupeptin, 2 mg/mL aprotinin, 1 mM phenylmethane sulfonyl fluoride) 30 min, 400 μL lysate incubated with anti-Myc antibody or 400 μL with control IgG antibodies, and 40 μL protein G agarose beads, then placed on a rotating wheel overnight at 4 °C. The agarose beads were pelleted and washed three times in NP-40 lysis buffer. Antibody–antigen complexes bound to the beads were eluted in SDS-PAGE sample buffer by boiling, resolved by SDS-PAGE, and analyzed by Western blotting analysis with the appropriate antibodies.
Blocking experiments with heparin sodium and an anti-Neu5Gc antibody
SVV was added (MOI = 1) to PK-15 cells after treatment with soluble heparin sodium (Yuanye) at 0.125 mg/mL, 0.25 mg/mL, 0.5 mg/mL, 1 mg/mL, and 2 mg/mL with PK-15 cells for 1 h, respectively.
An anti-Neu5Gc antibody (BioLegend) was preincubated at 1:200 dilution with IBRS-2 cells for 1h, after which SVV was added (MOI = 1). After 1h, media was replaced by DMEM. At 12 hpi, SVV mRNAs was determined using a qPCR assay.
Statistical analysis
GraphPad Prism software (GraphPad, San Diego, CA) was used to analyze the data. Unpaired two-tailed t test were performed where applicable to determine statistical significance, and the results are presented as the mean ± SEM (*p < 0.05, **p < 0.01, and ***p < 0.001).
Contributions
H.H., X.D., Z.Z., S.W., Y.Z. and H.Z. designed and supervised the project. Y.Z., H.Z., W.T. and W.Y. wrote the paper, with contributions from all authors. W.T. and X.Q. designed the CRISPR/Cas9 library. W.T. performed the screen and verification of gene function. Y.C. assisted with the establishment of single gene mutant cell and data analyses. F.G. performed coimmunoprecipitations. W.T. and K.L. performed indirect immunofluorescence assay.
Conflicts of Interest
The authors declare no conflict of interest.
Supplementary materials
Acknowledgements
This work was supported by the National Key Research and Development Program of China (Grant No.2021YFA0805905), the 2020 Research Program of Sanya Yazhou Bay Science and Technology City (202002011), the Plan 111 (B12008), the National Natural Science Foundation of China (32002180), the National Key Research and Development Program of China (Grant No. 2021YFA0805900), Gansu Provincial Major project for science and technology development (21ZD3NA001 and 20ZD7NA006), the Chinese Academy of Agricultural Science and Technology Innovation Project (CAAS-ZDRW202006), and Chinese Universities Scientific Fund.
Footnotes
↵3 These authors contributed equally to this work: Wenda Tang, Yanchao Wang, Xiaolan Qi