Plant materials and virus inoculation
During field survey in August 2020, soybean samples exhibiting crinkle and stay-green symptoms were collected from Zhengzhou, Henan province of China. N. benthamiana plants were grown in pots in a growth room under a 16 h light/8 h dark photoperiod at 25 °C with 60% humidity. For agroinfiltration, Agrobacteria strain GV3101 carrying infectious viral clones were suspended in infiltration buffer (10 mM MgCl2, 10 mM MES, and 200 μM acetosyringone, pH 5.6) at an OD600 of 1, kept at room temperature for 2 to 4 h and infiltrated into N. benthamiana leaves using a 1-mL needleless syringe.
Primers used for plasmid construction are listed in Table S3. All the available constructs were sequenced.
PVX recombinant vectors was constructed by introducing coding sequences of SGVA encoded proteins into potato virus X (PVX) vector pGR106 via ClaI and SalI digestion, followed by ligation with T4 DNA ligase (NEB). The fragments used were amplified using primer pairs PVX-V2/V1/C1/C2/C2/C4-F/R.
To construct vectors for RNA silencing analysis, the coding sequence of corresponding genes were amplified and inserted into PstI/BamHI digested pGD vectors via homologous recombination using Clonexpress II one step cloning kit (Vazyme, China) , the homologous arm length was 20 bp. The fragments used were amplified using primer pairs PGD-V2 -F/R.
To construct full-length cDNA clone of SGVA, a recombinant plasmid containing 1.8 copies of the full-length fragment of the viral genome was constructed. First, a 2762 bp full length was amplified from total DNA extracted in symptomatic leaves using Q5 high-fidelity polymerase (NEB, Beijing,China) with primer pair SGVAF1F/ SGVAF1R and ligated into binary vector pGD vector between the cauliflower mosaic virus (CaMV) 35S promoter and the nopaline synthase terminator (t-Nos) via the restriction sites BglII/ HindIII to produce construct pGDSGVAF1. After sequencing, a 2377 bp partial fragment was amplified using primer pair SGVAF2F/SGVAF2R and ligated into pGDSGVAF1 vector via the restriction sites HindIII/SalI to produce construct pGDSGVA.
To construct V2 mutant vectors, a 2762 bp full length was amplified from pGDSGVA-ZZ using primer pairs SGVAV2STOPF/ SGVAF1R, SGVAF1F/ SGVAV2STOPR and ligated into pGD vector via the restriction sites BglII/ HindIII to produce construct pGDSGVAv2stopF1, then a 2377 bp partial fragment was amplified using primer pair SGVAV2STOPF/ SGVAF2R, SCVF2F/ SGVAV2STOPR and ligated into pGDSGVAv2stopF1 vector via the restriction sites HindIII/SalI to produce construct pGDSGVAv2stop. SGVAV2-3738AA mutant was constructed using primer pairs SGVA3738AAF and SGVA3738AAR in the same way.
Nucleic Acid Extraction, Metatranscriptomic Sequencing, and Data analysis
The total RNA of samples was subjected to an rRNA removal procedure using a Ribo-zero Magnetic kit according to the manufacturer’s instructions (Epicentre, an Illumina® company). Next, cDNA libraries were constructed using a TruSeq™ RNA sample prep kit (Illumina). Barcoded libraries were paired-end sequenced on an Illumina HiSeq X ten platform at Shanghai biotechnology Co., Ltd. (Shanghai, China) according to the manufacturer’s instructions (www.illumina.com).
To obtain clean reads, the Fastax online software (version: 0.0.13, http://hannonlab.cshl.edu/fastx_toolkit/index.html) was used to screen out unqualified reads from the raw reads; this step removed joint sequences, reads with low overall quality (mainly reads with lengths less than 20 bp), reads with base N (base with uncertain identity), reads with more than 20% of the bases possessing Q-values ≤ 10, and or low end quality. Then, the reads were trimmed to remove bases with a quality score of soybean. Clean reads were de novo assembled using CLC Genomics Workbench (version:6.0.4) according to the scaffolding contig algorithm (word-size = 45, minimum contig length = 300) [25,26,27]. These various steps produced the primary unigenes. These were then assembled for a second time using CAP3 EST software  to acquire the final unigene sequence set. This unigene set was used for further exploration of the transcriptome. The final unigene set was compared against the NCBI non-redundant (Nr) database using BLASTX , with an E-value < 1e−5. Then the final unigenes annotated with virus were used for the virus analysis.The assembled contigs were queried by homology search tools (BLASTn and BLASTx) against public database (GenBank) in the NCBI (the National Center for Biotechnology Information).
Sequences alignments were conducted using the ClustalW method, and phylogenetic trees were constructed by the neighbor joining method using MEGA version X . The tree was evaluated with 1000 bootstrap replicates. Sequences of geminiviruses were retrieved from GenBank and used for comparison and phylogenetic analysis in this study.
Western blotting analysis
Agro-infiltrated leaves were harvested for western blotting assay. Total protein was extracted from 0.2 g leaves using the extraction buffer containing 20% glycerol, 20 mM Tris–HCl (pH 7.5), 1 mM EDTA, 150 mM NaCl, 1 mM PMSF, 1 × Protease inhibitor cocktail (Sigma, China). Total protein was separated in SDS–polyacrylamide gel electrophoresis, followed by transfer to nitrocellulose membranes. The membranes were probed using anti-PVX CP polyclonal antibodies or SGVA CP polyclonal antibodies followed by an HRP-conjugated secondary antibody. Antigens of PVX CP and SGVA CP was acquired via prokaryotic expression. The detection signals were developed using an ECL reagent as instructed. PVX CP and SGAV CP accumulation were photographed under a chemiluminescence apparatus (Amersham imager 680). CBB staining of the large subunit of RuBisCo served as a loading control. Due to the figure size the original size image of the blots was supplied in the supplementary information.
Total RNA was extracted from harvested N. benthamiana leaves using Trizol reagent (invitrogen) and treated with RNase-free DNase I. First strand cDNA was synthesized using 500 ng total RNA, an oligo d(T) primer, random primer, and M-MLV reverse transcriptase as instructed. Ten-fold diluted cDNA product was used for qPCR on an Eppendorf Real-Time PCR system using a SYBR Green master mix (Takara). The NbUBC genes (GenBank accession number: AB026056.1) was used as internal controls. All the primers used for qRT-PCR are listed in Table S3. The relative gene expression levels were calculated using the 2−△△CT method.