Identification of Three Viruses Infecting Mulberry Varieties

Viruses-mediated genome editing in plants is a powerful strategy to develop plant cultivars with important and novel agricultural traits. Mulberry alba is an important economic tree species that has been cultivated in China for more than 5000 years. So far, only a few viruses have been identified from mulberry trees, and their application potential is largely unknown. Therefore, mining more virus resources from the mulberry tree can pave the way for the establishment of useful engineering tools. In this study, eight old mulberry plants were gathered in seven geographic areas for virome analysis. Based on transcriptome analysis, we discovered three viruses associated with mulberries: Citrus leaf blotch virus isolate mulberry alba 2 (CLBV-ML2), Mulberry-associated virga-like virus (MaVLV), and Mulberry-associated narna-like virus (MaNLV). The genome of CLBV-ML2 was completely sequenced and exhibited high homology with Citriviruses, considered to be members of the genus Citrivirus, while the genomes of MaVLV and MaNLV were nearly completed lacking the 5′ and 3′ termini sequences. We tentatively consider MaVLV to be members of the family Virgaviridae and MaNLV to be members of the genus Narnavirus based on the results of phylogenetic trees. The infection experiments showed that CLBV-ML2 could be detected in the inoculated seedlings of both N. benthamiana and Morus alba, while MaVLV could only be detected in N. benthamiana. All of the infected seedlings did not show obvious symptoms.


Introduction
Next-generation sequencing (NGS) serves as an unbiased technology for the diagnosis of plant viral diseases since no prior information about the pathogen is required [1]. Resent day NGS tools are capable of sequencing any type of nucleic acid and have emerged as the tool of choice to detect novel viral diseases from very few viral titers [2]. Many novel viruses were identified by NGS, which before were quite difficult because of either low viral titer or difficulty of RNA extractions from woody plants or the absence of any symptoms. Recently, transcriptome sequencing was used in citrus to identify a novel latent viral infection, citrus yellow vein-associated virus (CYVaV), and analysis showed that it could be an ancestral virus to the current umbravirus [3,4].
The mulberry (Morus alba) belongs to the genus Morus in the family Moraceae, which is a deciduous tree that has been grown for thousands of years in China [5] The leaves of mulberry trees are the only food source for silkworms [6] and also can be converted into mulberry tea [7] and other beverages [8]. Their fruits are rich in nutrients and can be used for preparing fruit juice and wine [9,10]. The yield and quality of mulberry leaves can be seriously affected by a virus infection, causing severe loss to the sericultural industry. Before the application of NGS, only a few viruses had been identified in mulberry trees, such as the Mulberry crinkle leaf virus (MCLV) [11] and the Paper mulberry mosaic-associated virus (PMuMaV) [12]. With the rapid development of NGS technology, much more viruses have been identified in trees of mulberry recently. Mulberry mosaic dwarf-associated virus (MMDaV) caused mosaic and dwarfing symptoms in mulberry trees [13], and mulberry The present study examined eight different species of ancient mulberry trees from Chinese provinces (Heilongjiang, Henan, Shandong, Anhui, Beijing, Xinjiang, and Xizang). The samples were leaves of ancient mulberry trees, which were asymptomatic at the time of collection. They were stored in sterile 50 mL conical tubes at −80 • C until further analysis.

Sequence Analyses
The obtained samples were sent to Majorbio (Shanghai, China) for total RNA isolation and transcriptome sequencing and analysis. Raw data were trimmed using Trimmomatic (Version 3.90) [18] and trinity (Version 2.8.5) with the default parameters used for contig assembly [19]. Then, blastx from the BLAST suite was used to search for conserved viral proteins in the assembled contigs [20]. Finally, open reading frames (ORFs) were predicted using the ORFfinder web tool (https://www.ncbi.nlm.nih.gov/orffinder/, accessed on 15 January 2022). A conserved domain search was performed using the CDD/SPARCLE tool (https://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi/, accessed on 15 January 2022) of NCBI [21]. TBtools v0.67 software was used to display the pairwise identity of the sequences aligned by the ClustalW program and prepare a correlation heatmap.

Phylogenetic Analyses
The amino acid sequences of the predicted RNA-dependent RNA polymerase (RdRp) regions from MaVLV and MaNLV were used for phylogenetic analyses in this study. The sequences were obtained from National Center for Biotechnology Information (NCBI, https: //www.ncbi.nlm.nih.gov/, accessed on 15 January 2022), compared with the related viral sequences using ClustalW and analyzed using the Maximum likelihood (ML) algorithm with the LG and WAG substitution model to construct phylogenetic trees in MEGA X with 1000 bootstrap replications, respectively [22]. The whole sequence of CLBV-ML2 was used for phylogenetic analyses. The sequences obtained from NCBI and subject to ClustalW multiple sequence alignments. Then, the maximum likelihood method based on the TN93 matrix-based models with 1000 bootstrap replications was separately conducted.

Completion of Genomes of Viruses
The full genome sequence of the candidate CLBV-ML2 was obtained by rapid amplification of cDNA ends (RACE) using a SMARTer ® RACE 5 /3 kit (Takara, Beijing, China). The 5 RACE CDS Primer A and the 3 RACE CDS Primer A was used to synthesize the cDNAs, which could include specific amplification of 5 and 3 end sequences, respectively. The universal primer mix and the gene-specific primers (GSP , Table S1) were used to complete the 5 -terminal and 3 -terminal sequences with 2×Hieff Canace ® Plus PCR Master Mix (Yeasen, Shanghai, China). The purified PCR product was cloned into a pESI-T vector (Yeasen, Shanghai, China), and the M13-F/R primer was used for positive clone detection and sequenced commercially (Youkang, Hangzhou, China). Finally, the whole sequence was assembled with DNAMAN software (version 9.0), and observed that an additional 66 nt at the 5 terminus and 528 nt at the 3 terminus was identified.

Mechanical Inoculation and Virus Detection by RT-PCR
Mechanical inoculation was used for virus infection. The basic procedure was to grind up the samples from ancient mulberry trees and add the appropriate amount of phosphate buffer (250 mL 0.2 mol/L KH 2 PO 4 , 118 ml 0.2 mol/L NaOH, add H 2 O to 1000 ml, pH 6.8), then rub the extracted sap onto the leaves of 5 N. benthamiana and 5 Morus alba. Total RNA was extracted from system leaves after 10 days. RT-PCR was used for the RdRp sequence detection from the samples' total RNA and inoculated plants' total RNA. Following the manufacturer's instructions, the RT reaction was performed with Hifair ® III 1st Strand cDNA Synthesis Kit (Yeasen, Shanghai, China) with Oligo (dT) 18 primer. One microliter of cDNA was used as a template for the PCR reaction in the presence of primers of RdRp detection. The reaction procedure was 94 • C for 3 min; 94 • C for 30 s; 55 • C for 30 s; 72 • C for 1 min (30 s/kb); 72 • C for 5 min, and end of 4 • C, which produced 1295 bp, 729 bp, and 1298 bp fragments, respectively.
MaNLV has a 4097 nt genome, excluding the 5 and 3 UTR sequences. Its sequence shares 79.21% nt similarity with the Narnaviridae sp. isolate 174-k141_105200 (MZ680060), which was found in pond sediment. Two ORFs were found by the ORFfinder web tool. ORF1 is positioned at 73-789 nt in the genome, and ORF2 is positioned at 725-3991 nt in the genome and possibly encodes RdRp protein. However, no conserved domain was found by the CDD/SPARCLE search ( Figure 2C).

Phylogenetic Placement of Identified Viruses
The phylogenetic tree is based on the RNA-dependent RNA polymerase (RdRp) amino acid sequences of representative members of the Virgaviridae family. The phylogenetic analysis result shows that MaVLV falls in the clade with Grapevine-associated virgalike virus (QXN75450) and belongs to the unclassified Virgaviridae ( Figure 3A). We propose that MaVLV was a member of unclassified virgaviridaes because its RdRp amino acid sequence showed low similarity with isolates of GaVLV in the correlation heatmap ( Figure 3B).

Phylogenetic Placement of Identified Viruses
The phylogenetic tree is based on the RNA-dependent RNA polymerase (RdRp) amino acid sequences of representative members of the Virgaviridae family. The phylogenetic analysis result shows that MaVLV falls in the clade with Grapevine-associated virga-like virus (QXN75450) and belongs to the unclassified Virgaviridae ( Figure 3A). We propose that MaVLV was a member of unclassified virgaviridaes because its RdRp amino acid sequence showed low similarity with isolates of GaVLV in the correlation heatmap ( Figure 3B).
The phylogenetic trees based on the whole sequences of representative members of the Citrivirus genus show that CLBV-ML2 identified in the present study falls in the clade where sequences of CLBV-PY2 and -PY1 ( Figure 4A). The correlation heatmap shows that the whole sequence of CLBV-ML2 is conserved among citriviruses ( Figure 4B). The nucleotide and putative amino acid sequences of CLBV-ML2 were compared with those of the CLBV isolates and unassigned viruses. The whole genome of CLBV-ML2 showed nucleotide identity higher than 75% (Table 2). At the nucleotide sequence level, each ORFs of CLBV-ML2 showed nucleotide identities ranging from 53.35 to 88.81%, 58.07 to 98.07%, and 78.51 to 95.32%. At the amino acid sequence level, they showed similar sequence identities ranging from 75.44 to 82.44%, 79.76 to 84.02%, and 83.12 to 84.59% (Table 2). Based on the species demarcation criteria for Betafexiviridae [23], we propose that CLBV-ML2 is a new isolate of Citrivirus because it has similar polyprotein processing programs and share more than 70% amino acid identity in the polyprotein with any other citriviruses.
The phylogenetic tree is based on the RdRp amino acid sequence of representative members of the Narnaviridae family. The phylogenetic analysis result shows that MaNLV falls in the clade with Neofusicoccum parvum narnavirus, which was found in grapevine wood tissues and is a member of unclassified narnaviruses ( Figure 5A). Based on the correlation heatmap, we considered that MaNLV belongs to the genus Narnavirus ( Figure 5B).  The phylogenetic trees based on the whole sequences of representative members of the Citrivirus genus show that CLBV-ML2 identified in the present study falls in the clade where sequences of CLBV-PY2 and -PY1 ( Figure 4A). The correlation heatmap shows that the whole sequence of CLBV-ML2 is conserved among citriviruses ( Figure 4B). The nucleotide and putative amino acid sequences of CLBV-ML2 were compared with those of the CLBV isolates and unassigned viruses. The whole genome of CLBV-ML2 showed nucleotide identity higher than 75 % (  [23], we propose that CLBV-ML2 is a new isolate of Citrivirus because it has similar polyprotein processing programs and share more than 70 % amino acid identity in the polyprotein with any other citriviruses.   The phylogenetic tree is based on the RdRp amino acid sequence of representative members of the Narnaviridae family. The phylogenetic analysis result shows that MaNLV falls in the clade with Neofusicoccum parvum narnavirus, which was found in grapevine  wood tissues and is a member of unclassified narnaviruses ( Figure 5A). Based on the correlation heatmap, we considered that MaNLV belongs to the genus Narnavirus ( Figure  5B). (B) The pairwise identity plots of the RNA-dependent RNA polymerase amino acid sequences aligned by ClustalW and displayed by sequence demarcation using TBtools software. The trees were constructed using the maximum likelihood method, and the statistical significance of the branches was evaluated by bootstrap analysis (1000 replicates). Newly discovered viruses are marked in red, unclassified narnaviridaes in black, unclassified narnaviruses in bule, and narnaviruses in green. Relationship: pink background.

The Detection of Viruses from Samples and Complete Sequence of CLBV-ML2
To further confirm the transcriptome sequencing and analysis results and identify the complete virus genomic sequence. Total RNA was extracted from the diseased leaves for the subsequent PCR and RACE experiments. The results of RT-PCR showed that the RdRp of MaVLV and CLBV-ML2 can be detected, but the passage of MaNLV was negative. Therefore, we determined that MaVLV and CLBV-ML2 were latent in G026 and G039 samples ( Figure 6A). The 5′-terminal and 3′-terminal sequences of CLBV-ML2 were amplified by the RACE kit ( Figure 6B). Then, the purification of PCR products was cloned (B) The pairwise identity plots of the RNA-dependent RNA polymerase amino acid sequences aligned by ClustalW and displayed by sequence demarcation using TBtools software. The trees were constructed using the maximum likelihood method, and the statistical significance of the branches was evaluated by bootstrap analysis (1000 replicates). Newly discovered viruses are marked in red, unclassified narnaviridaes in black, unclassified narnaviruses in bule, and narnaviruses in green. Relationship: pink background.

The Detection of Viruses from Samples and Complete Sequence of CLBV-ML2
To further confirm the transcriptome sequencing and analysis results and identify the complete virus genomic sequence. Total RNA was extracted from the diseased leaves for the subsequent PCR and RACE experiments. The results of RT-PCR showed that the RdRp of MaVLV and CLBV-ML2 can be detected, but the passage of MaNLV was negative. Therefore, we determined that MaVLV and CLBV-ML2 were latent in G026 and G039 samples ( Figure 6A). The 5 -terminal and 3 -terminal sequences of CLBV-ML2 were amplified by the RACE kit ( Figure 6B). Then, the purification of PCR products was cloned into the pESI-T vector to screen positive clones and sequenced commercially ( Figure S1). Finally, the 66 nt 5 -UTR (1-66 nt) and 528 nt 3 -UTR (8265-8792 nt) were obtained. into the pESI-T vector to screen positive clones and sequenced commercially ( Figure S1). Finally, the 66 nt 5′-UTR (1-66 nt) and 528 nt 3′-UTR (8265-8792 nt) were obtained.

Viruses Infectivity in N. benthamiana and Morus alba
In order to investigate the infectivity of this novel virus, mechanical inoculation was used for virus infection by grinding up the samples from mulberry trees and adding an appropriate amount of phosphate buffer. We then rubbed the extracted sap onto the leaves of five Nicotiana benthamiana and five Morus alba. Ten days later, there were no symptoms on the upper leaves of the inoculated seedlings ( Figures 7A and S2). Total RNA was extracted from the infection leaves. RT-PCRs with specific primers of RdRp (Table S1) were performed against the part infection leaves. There were weak stripes for MaVLV RdRp sequences in the cDNA from N. benthamiana and nothing in mock or cDNA from Morus alba compared to the result of the G026 sample with the mock and infection. Both N. benthamiana and Morus alba lanes had obvious stripes in the detection of CLBV-ML RdRp sequences, and their sizes were consistent with the sample detection results ( Figure  7B). In addition, the test results for MaNLV infectivity were negative ( Figure S2). We concluded that of the possible reasons for this phenomenon, one possibility is viral titers in collected mulberry were low, and few plants were inoculated. The other possibility is inoculation buffer is inappropriate, especially the pH-value.

Viruses Infectivity in N. benthamiana and Morus alba
In order to investigate the infectivity of this novel virus, mechanical inoculation was used for virus infection by grinding up the samples from mulberry trees and adding an appropriate amount of phosphate buffer. We then rubbed the extracted sap onto the leaves of five Nicotiana benthamiana and five Morus alba. Ten days later, there were no symptoms on the upper leaves of the inoculated seedlings ( Figures 7A and S2). Total RNA was extracted from the infection leaves. RT-PCRs with specific primers of RdRp (Table S1) were performed against the part infection leaves. There were weak stripes for MaVLV RdRp sequences in the cDNA from N. benthamiana and nothing in mock or cDNA from Morus alba compared to the result of the G026 sample with the mock and infection. Both N. benthamiana and Morus alba lanes had obvious stripes in the detection of CLBV-ML RdRp sequences, and their sizes were consistent with the sample detection results ( Figure 7B). In addition, the test results for MaNLV infectivity were negative ( Figure S2). We concluded that of the possible reasons for this phenomenon, one possibility is viral titers in collected mulberry were low, and few plants were inoculated. The other possibility is inoculation buffer is inappropriate, especially the pH-value. into the pESI-T vector to screen positive clones and sequenced commercially ( Figure S1). Finally, the 66 nt 5′-UTR (1-66 nt) and 528 nt 3′-UTR (8265-8792 nt) were obtained.

Viruses Infectivity in N. benthamiana and Morus alba
In order to investigate the infectivity of this novel virus, mechanical inoculation was used for virus infection by grinding up the samples from mulberry trees and adding an appropriate amount of phosphate buffer. We then rubbed the extracted sap onto the leaves of five Nicotiana benthamiana and five Morus alba. Ten days later, there were no symptoms on the upper leaves of the inoculated seedlings ( Figures 7A and S2). Total RNA was extracted from the infection leaves. RT-PCRs with specific primers of RdRp (Table S1) were performed against the part infection leaves. There were weak stripes for MaVLV RdRp sequences in the cDNA from N. benthamiana and nothing in mock or cDNA from Morus alba compared to the result of the G026 sample with the mock and infection. Both N. benthamiana and Morus alba lanes had obvious stripes in the detection of CLBV-ML RdRp sequences, and their sizes were consistent with the sample detection results ( Figure  7B). In addition, the test results for MaNLV infectivity were negative ( Figure S2). We concluded that of the possible reasons for this phenomenon, one possibility is viral titers in collected mulberry were low, and few plants were inoculated. The other possibility is inoculation buffer is inappropriate, especially the pH-value.

Discussion
This study uncovered three viruses connected to mulberry based on transcriptome sequencing of eight mulberry samples. A full-length sequence of CLBV-ML2, which exhibited high homology and clustered with citriviruses in phylogenetic trees. The other two viruses, MaVLV and MaNLV, belong to the member of unclassified virgaviridaes and unclassified narnaviruses by genome and phylogenetic analysis.
Virgaviridae family consists of plant viruses with rod-shaped virions, a singlestranded RNA genome with a 3′-terminal tRNA-like structure, and an alpha-like replication protein. Its member genera include Furovirus, Hordeivirus, Pecluvirus, Pomovirus, Tobamovirus, and Tobravirus [24,25]. In this study, we identified a novel virus, MaVLV, in mulberry plants, which was presumed to be a member of virgaviruses. Interestingly, similarly to the member of virgaviridae, MaVLV ORF1 is also an alpha-like replication protein with conserved methyltransferase (MT) and helicase (Hel) domains. The RNA-dependent RNA polymerase (RdRp) is expressed as the C-terminal part of this protein by the readthrough of a leaky stop codon. But the movement protein and coat protein have not been proven to encode in the MaVLV genome. Due to the lack of 5'UTR and 3'UTR sequences in the genome, the existence of a 3'-terminal tRNA-like structure needs future investigation.
Narnaviruses have been described as positive-sense RNA viruses with a remarkably simple genome of ~3 kb, encoding only a highly conserved RNA-dependent RNA polymerase (RdRp) [26]. They could be detected in diverse fungi, plants, protists, arthropods and nematodes [27]. In this study, we identified a novel virus MaNLV with uncomplete genome sequence of about 4097 nt in mulberry plants. It was presumed to be a member of narnaviruses. MaNLV has the same characteristics as narnaviruses, and its genomic sequence only encodes a conserved RdRp protein. The extent of the virus infection and whether it causes symptoms need to be confirmed by subsequent experiments (Table S2).
The International Committee on Taxonomy of Viruses recognizes the Citrus leaf blotch virus (CLBV) as the sole species in the genus Citrivirus [23]. The natural host range of CLBV is citrus, and it has also been reported in Prunus spp. [28], Actinidia [29], peony [30], and mulberry [16]. Studies have shown that for juvenile citrus plants, an advantage of CLBV as a viral vector for citrus is that it causes asymptomatic infection in most cultivars [31]. CLBV vectors are appropriate for gene expression or silencing in tissues that lack phloem, including meristematic regions [24], and CLBV is not transmitted by vectors. Therefore, they could be safely used in field experiments [32]. In our study, we identified a new CLBV isolate from mulberry trees named CLBV-ML2. It is highly homologous with the whole sequence and ORF1 of the previously reported CLBV-ML [16] but differs from ORF2 and ORF3. Recently, a viral vector based on the Citrus leaf blotch virus for genetics and breeding, which can used for precocious flowering of juvenile citrus. The mulberry

Discussion
This study uncovered three viruses connected to mulberry based on transcriptome sequencing of eight mulberry samples. A full-length sequence of CLBV-ML2, which exhibited high homology and clustered with citriviruses in phylogenetic trees. The other two viruses, MaVLV and MaNLV, belong to the member of unclassified virgaviridaes and unclassified narnaviruses by genome and phylogenetic analysis.
Virgaviridae family consists of plant viruses with rod-shaped virions, a single-stranded RNA genome with a 3 -terminal tRNA-like structure, and an alpha-like replication protein.
Its member genera include Furovirus, Hordeivirus, Pecluvirus, Pomovirus, Tobamovirus, and Tobravirus [24,25]. In this study, we identified a novel virus, MaVLV, in mulberry plants, which was presumed to be a member of virgaviruses. Interestingly, similarly to the member of virgaviridae, MaVLV ORF1 is also an alpha-like replication protein with conserved methyltransferase (MT) and helicase (Hel) domains. The RNA-dependent RNA polymerase (RdRp) is expressed as the C-terminal part of this protein by the readthrough of a leaky stop codon. But the movement protein and coat protein have not been proven to encode in the MaVLV genome. Due to the lack of 5'UTR and 3'UTR sequences in the genome, the existence of a 3 -terminal tRNA-like structure needs future investigation.
Narnaviruses have been described as positive-sense RNA viruses with a remarkably simple genome of~3 kb, encoding only a highly conserved RNA-dependent RNA polymerase (RdRp) [26]. They could be detected in diverse fungi, plants, protists, arthropods and nematodes [27]. In this study, we identified a novel virus MaNLV with uncomplete genome sequence of about 4097 nt in mulberry plants. It was presumed to be a member of narnaviruses. MaNLV has the same characteristics as narnaviruses, and its genomic sequence only encodes a conserved RdRp protein. The extent of the virus infection and whether it causes symptoms need to be confirmed by subsequent experiments (Table S2).
The International Committee on Taxonomy of Viruses recognizes the Citrus leaf blotch virus (CLBV) as the sole species in the genus Citrivirus [23]. The natural host range of CLBV is citrus, and it has also been reported in Prunus spp. [28], Actinidia [29], peony [30], and mulberry [16]. Studies have shown that for juvenile citrus plants, an advantage of CLBV as a viral vector for citrus is that it causes asymptomatic infection in most cultivars [31]. CLBV vectors are appropriate for gene expression or silencing in tissues that lack phloem, including meristematic regions [24], and CLBV is not transmitted by vectors. Therefore, they could be safely used in field experiments [32]. In our study, we identified a new CLBV isolate from mulberry trees named CLBV-ML2. It is highly homologous with the whole sequence and ORF1 of the previously reported CLBV-ML [16] but differs from ORF2 and ORF3. Recently, a viral vector based on the Citrus leaf blotch virus for genetics and breeding, which can used for precocious flowering of juvenile citrus. The mulberry tree harboring CLBV-ML2 showed no symptoms at the date of collection. Therefore, it has a good application potential as a virus-mediated transgenic vector using mulberry as the host if the virus can confirm implement system infection.
In the ensuing decades, researchers focused on discovering and eliminating viral threats to plant and animal health. However, viruses are not merely agents of destruction but essential components of global ecosystems. Due to the increasing interest in organic farming to reduce the use of synthetic chemicals in agriculture, viral-based vector to manage crop pests and diseases in a sustainable way is a good strategy. A virus-mediated transgenic system is a quick system to deliver genes without any need for transformation and regeneration techniques and has currently been successfully used in apples and other woody species [33][34][35][36][37][38][39][40][41][42]. More efforts need to be made to discover new viruses in mulberry to accelerate CLBV-ML2 the application of virus-mediated transgenic in mulberry varieties.

Conclusions
In this study, we collected eight groups of mulberry leaves from different provinces in China. Three viruses were identified based on genome and phylogenetic analysis, named Citrus leaf blotch virus isolate mulberry alba 2 (CLBV-ML2), Mulberry-associated virga-like virus (MaVLV), and Mulberry-associated narna-like virus (MaNLV).

Supplementary Materials:
The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/v14112564/s1, Figure S1: The detection of positive clones; Figure S2: Symptoms on leaves of Nicotiana benthamiana and Morus alba infected by MaNLV, and virus detection; Table S1: The primers of RACE; Table S2