High-Throughput Sequencing Reveals Bell Pepper Endornavirus Infection in Pepper (Capsicum annum) in Slovakia and Enables Its Further Molecular Characterization

Ribosomal RNA-depleted total RNAs from a sweet pepper plant (Capsicum annuum, labelled as N65) grown in western Slovakia and showing severe virus-like symptoms (chlorosis, mottling and deformation of leaf lamina) were subjected to high-throughput sequencing (HTS) on an Illumina MiSeq platform. The de novo assembly of ca. 5.5 million reads, followed by mapping to the reference sequences, revealed the coinfection of pepper by several viruses; i.e., cucumber mosaic virus (CMV), watermelon mosaic virus (WMV), pepper cryptic virus 2 (PCV2) and bell pepper endornavirus (BPEV). A complete polyprotein-coding genomic sequence (14.6 kb) of BPEV isolate N65 was determined. A comparison of BPEV-N65 sequences with BPEV genomes available in GenBank showed 86.1% to 98.6% identity at the nucleotide level. The close phylogenetic relationship with isolates from India and China resulted in their distinct grouping compared to the other BPEV isolates. Further analysis has revealed the presence of BPEV in sweet or chili peppers obtained from various sources and locations in Slovakia (plants grown in gardens, greenhouse or retail shop). Additionally, the partial sequencing of two genomic portions from 15 BPEV isolates revealed that the Slovak isolates segregated into two molecular clusters, indicating a genetically distinct population (mean inter-group nucleotide divergence reaching 12.7% and 14.5%, respectively, based on the genomic region targeted). Due to the mix infections of BPEV-positive peppers by potato virus Y (PVY) and/or CMV, the potential role of individual viruses in the observed symptomatology could not be determined. This is the first evidence and characterization of BPEV from the central European region.


Introduction
Pepper (Capsicum sp.) is a vegetable crop of the Solanaceae family which is widely grown in temperate regions either for direct consumption or for the food and pharmaceutical industries. Both sweet and chili peppers are among the top ten most widely cultivated vegetables in the world [1,2]. However, pepper production can be often constrained by infections by viral pathogens, negatively affecting the yield or quality of the production [3]. On the other hand, persistent, nonpathogenic viruses belonging to several families are also found frequently in plants [4,5]. Bell pepper is not an exception to that, with a few partiti-and endornaviruses reported from the host; e.g., pepper cryptic viruses 1 and 2 (PCVs 1 and 2), bell pepper endornavirus (BPEV), hot pepper endornavirus (HPEV).
The Endornaviridae family, comprising the genera Alphaendornavirus and Betaendornavirus, includes viruses which have been identified in plants, fungi oomycetes and protists [6][7][8]. Members of these genera were reported from several important crops, such as rice, bean, barley, cucurbits, spinach or pepper [9]. Endornaviruses are characterized by a stable low copy number in their plant host, exhibiting no obvious symptoms or pathological effect on plants (with the exception of Vicia faba endornavirus) and efficient vertical transmission [8,10,11].
Bell pepper endornavirus (BPEV) belongs to the genus Alphaendornavirus [8]. The virus is characterized by a single linear single-stranded RNA genome of approximately 14 kb in length, containing a single open reading frame (ORF), which is translated into a large polyprotein. This polyprotein of ca. 4815-4884 aa contains several conserved functional domains, such as putative viral methyltransferase (MTR), helicase 1 (Hel-1), UDP-glycosyltransferase (UDG) and RNA-dependent RNA polymerase (RdRp) [10]. Endornaviruses seem not to form true virions and are usually found as dsRNA replicative intermediates [8,12].
In this work, the first complete genome of a European BPEV isolate from Slovakia (BPEV-N65) is reported, together with a partial molecular characterization of additional isolates from this region, contributing to the better understanding of the genetic complexity of this virus on a global scale.

Results
A total of 5,511,704 high-quality reads (with an average length of 135.5 bp) were obtained from the ribosomal RNA-depleted total RNAs of the N65 pepper sample using the Illumina MiSeq platform. Blast analyses of de novo assembled contigs (16,101 contigs with the length higher than 500 bp) indicated the presence of a complex infection, involving cucumber mosaic virus (CMV, genus Cucumovirus), watermelon mosaic virus (WMV, genus Potyvirus), pepper cryptic virus-2 (PCV2, genus Deltapartitivirus), and BPEV (genus Alphaendornavirus) ( Table 1).
Two large contigs (8851 and 5874 bp in size) were initially identified by Blast analyses to match with BPEV sequences in the GenBank. Further mapping of sequence data on the publicly available complete BPEV genomes, performed with the Geneious software, resulted in the reconstruction of the nearly complete BPEV genome sequences, lacking only 24 nucleotides at the viral 3' terminus, as compared to NC_039216.
The Phylogenetic analysis based on the full-length genome sequences showed the division of currently characterized BPEV isolates into two genetic groups ( Figure 1). The isolate BPEV-N65 resulted as most closely related to BPEV isolates from China (MH182675) and India (KU923755, KU923756). These four isolates, along with an exemplar isolate "BPEV-YW" from USA (JN019858), form a highly supported cluster representing one of the two distinct evolutionary lineages of members of this species. A similar phylogenetic topology was obtained by the analysis of the polyprotein amino acid sequences, further supporting the phylogenetic affinity of BPEV-N65 with isolates from the Far East and from the USA. depending on the AUG codon considered as an initiation codon for genome translation. The identity at the amino acid level of N65 with other isolates reached 89.8% to 98.9%. Phylogenetic analysis based on the full-length genome sequences showed the division of currently characterized BPEV isolates into two genetic groups ( Figure 1). The isolate BPEV-N65 resulted as most closely related to BPEV isolates from China (MH182675) and India (KU923755, KU923756). These four isolates, along with an exemplar isolate "BPEV-YW" from USA (JN019858), form a highly supported cluster representing one of the two distinct evolutionary lineages of members of this species. A similar phylogenetic topology was obtained by the analysis of the polyprotein amino acid sequences, further supporting the phylogenetic affinity of BPEV-N65 with isolates from the Far East and from the USA.    To investigate further the occurrence and molecular diversity of BPEV in Slovakia, leaf samples were obtained during July-August 2018 from randomly selected pepper plants grown in private gardens in Piešt'any (n = 10) and Pezinok (n = 10), in an insect-proof greenhouse in Bratislava (n = 5), and in May 2019 from a large number of young pepper plants in a retail shop in Bratislava (n = 5) (all locations in western Slovakia).
Primers targeting two different parts of the BPEV genome (nts 7776-8485 and 12,632-13,238) were designed based on the conserved regions among publicly available BPEV sequences, including HTS-generated data of isolate N65. From a total of 30 samples analyzed by reverse transcription polymerase chain reaction (RT-PCR), 15 samples tested positive for BPEV using both primer sets. The specificity of RT-PCR was checked by the direct sequencing of PCR products, resulting in the partial sequences of 669 and 567 bp, after primer removal, of 15 additional BPEV isolates from Slovakia.
Phylogenetic analyses of both genome portions resulted in congruent tree topology, separating the studied Slovak isolates into two distinct phylogenetic groups (Figure 1, Figure 2A,B). Seven Slovak BPEV isolates from Piešt'any fall into the major group I. Remaining isolates from Bratislava and Pezinok were grouped together with N65 in the group II. While the intra-group nt diversity among Slovak BPEV isolates remained low (0.2% and 0.3%, respectively), the mean inter-group divergence reached 12.7% and 14.5%, respectively, based on the genome portion analyzed. To investigate further the occurrence and molecular diversity of BPEV in Slovakia, leaf samples were obtained during July-August 2018 from randomly selected pepper plants grown in private gardens in Piešťany (n = 10) and Pezinok (n = 10), in an insect-proof greenhouse in Bratislava (n = 5), and in May 2019 from a large number of young pepper plants in a retail shop in Bratislava (n = 5) (all locations in western Slovakia).
Primers targeting two different parts of the BPEV genome (nts 7776-8485 and 12,632-13,238) were designed based on the conserved regions among publicly available BPEV sequences, including HTS-generated data of isolate N65. From a total of 30 samples analyzed by reverse transcription polymerase chain reaction (RT-PCR), 15 samples tested positive for BPEV using both primer sets. The specificity of RT-PCR was checked by the direct sequencing of PCR products, resulting in the partial sequences of 669 and 567 bp, after primer removal, of 15 additional BPEV isolates from Slovakia.
Phylogenetic analyses of both genome portions resulted in congruent tree topology, separating the studied Slovak isolates into two distinct phylogenetic groups (Figure 1, Figure 2A,B). Seven Slovak BPEV isolates from Piešťany fall into the major group I. Remaining isolates from Bratislava and Pezinok were grouped together with N65 in the group II. While the intra-group nt diversity among Slovak BPEV isolates remained low (0.2% and 0.3%, respectively), the mean inter-group divergence reached 12.7% and 14.5%, respectively, based on the genome portion analyzed.

Discussion
The persistent phytoviruses, also known as "cryptic" because of lack of obvious symptoms associated with their presence in plant hosts, have been poorly studied in the past. The use of unbiased HTS technologies for the study of the plant virome revealed their frequent occurrence in

Discussion
The persistent phytoviruses, also known as "cryptic" because of lack of obvious symptoms associated with their presence in plant hosts, have been poorly studied in the past. The use of unbiased HTS technologies for the study of the plant virome revealed their frequent occurrence in cultivated and non-cultivated plants [4,13]. The members of the Endornaviridae family have been reported in Plants 2020, 9, 41 5 of 9 several economically important crops; however, their effect on the host phenotype is still poorly understood [14,15]. Furthermore, the extent of their genetic variability is still not fully understood due to the paucity of data from some geographical areas.
To date, the genomes of 15 BPEV isolates have been completely sequenced. However, those reports come only from Asian or American continents. To supplement the data on BPEV genetic diversity in the European region, the complete genome of a Slovak BPEV isolate from a sweet pepper plant, referred to as N65, was determined along with partial sequence data from an additional 15 BPEV isolates from sweet to chili pepper.
The analysis of a limited number of pepper samples from different locations in Slovakia (n = 30) showed BPEV infections in half of tested plants. Based on the phylogenetic analyses and pairwise comparisons, the studied Slovak BPEV isolates are not genetically uniform and belong to two previously defined molecular groups [5] (referred here as groups I and II). Although the number of BPEV sequences is still limited, the phylogenetic grouping (Figures 1 and 2A,B) of sequenced isolates indicates an absence of geography-based clustering on a global scale. Accordingly, the separation of partially sequenced isolates from Slovakia into two groups is most likely due to different host genotype (variety) sampled rather than geography-driven divergence. Indeed, the co-evolution of endornaviruses with a specific type of plant genotype, as a consequence of a long coexistence of host and a virus, has been hypothesized recently for Cucumis melo endornavirus (CmEV; [16]), as well as for BPEV [9,10]. Accordingly, in this study, four BPEV isolates from chili peppers were grouped in the same phylogroup, suggesting adaptation and coevolution with a particular host genotype. The two primer pairs, designed on the database and HTS-based sequence determined in this work, were efficiently used in RT-PCR for the amplification of both genetic groups of isolates in the two portions targeted (spanning nts 7776-8485 and 12,632-13,238), thus showing a broad polyvalence to known BPEV variants.
Interestingly, the recombination events reported for the isolate "Penol" (NC_039216) in previous work [5] could not be confirmed in our study. It is possible that other recombination event(s) could be found if more genetically different variants of BPEV are available.
The phylogenetic analysis of Slovak isolates targeting two different genome portions did not show an incongruence in their affiliation to the respective genetic groups, which is consistent with an absence of recombination, at least in the nt 7776-13,238 region.
HTS technologies provide an enormous volume of sequence data, enabling many possible applications, including the identification of the virome, full-length genome characterization of known or emerging viruses in infected plants, or pathogen characterization without a priori knowledge [17]. Indeed, in this work, the complex virome has been identified in a symptomatic pepper plant, comprising members from the genera Cucumovirus, Potyvirus, Deltapartitivirus and Alphaendornavirus. This finding further emphasizes the intricate nature of plant viral diseases [18], indicating that the occurrence of complex viral infections in plants is rather a rule than an exception.
Different templates used for library preparation and different strategies for the enriching of viral sequences can be applied prior to the HTS analysis; e.g., virus-derived small interfering RNAs, double-stranded RNAs, ribosomal RNA-depleted total RNAs and virion-associated nucleic acids [19][20][21]. Our results confirmed the suitability of the total RNA templates, in which the virus fraction was enriched by ribosomal RNA depletion prior to library preparation for HTS. The HTS analysis of the pepper N65 sample enabled us to obtain complete or nearly complete genome sequences of several plant viruses, including acute pathogens (CMV, WMV) or persistent viruses (BPEV, PCV2).
Although the endornaviruses are considered as non-pathogenic, and their infection to be associated with no visible effect on their host, their potential contribution to plant fitness is not completely elucidated. In our work, due to the detection of the additional mixed infection of peppers by potato virus Y and/or cucumber mosaic virus, the potential role of individual viruses in the observed symptomatology (Table 2) could not be determined.

Analysis of the Virome and Determination of the BPEV Full-Length Genome Sequence by HTS
Sweet pepper plant (Capsisum annum, cv. Promotor), grown in a private garden inČachtice, western Slovakia (GPS coordinates 48 • 42'38.2" N, 17 • 47'23.2" E) and showing leaf chlorosis, mottling and deformation symptoms, was sampled in August 2017.
Total RNAs were extracted from upper leaves of pepper plants using the Spectrum Plant Total RNA Kit (Sigma Aldrich, St. Louis, MO, USA). Ribosomal RNA was removed using the Ribo-Zero rRNA Removal Kit (Illumina, San Diego, CA, USA). The sample of ribosomal RNA-depleted total RNA was used for double stranded cDNA synthesis using the SuperScript II kit (Thermo Fisher Scientific, Waltham, MA, USA). The cDNA was then purified with the 2.2 x AMPure XP beads and quantified with the Qubit 2.0 Fluorometer (Thermo Fisher Scientific, Waltham, MA, USA). Subsequently, the sample was processed with the transposon-based chemistry library preparation kit (Nextera XT, Illumina, San Diego, CA, USA). Low-cycle PCR and mutual indexing of the fragments was carried out. Fragments were purified with 1.8 x AMPure XP beads (Beckman Coulter, Brea, CA, USA) without size selection. The fragment size structure of the DNA library was assessed using the Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). Since the obtained fragment length distribution (150-700 bp) met the criteria for further sample processing, the average bp value was used for sample molarity calculation. The equimolar pool of 4 nM DNA libraries was denatured, diluted to 10 pM and sequenced (300 bp paired-end sequencing) on the Illumina MiSeq platform (Illumina, San Diego, CA, USA).
High-quality trimmed reads were used for de novo assembly and contigs aligned to the viral genomes database [22] using Geneious v.8.1.9 software. Alternatively, the reads were mapped against