Genome composition analysis of multipartite BNYVV reveals the occurrence of genetic re-assortment in the isolates of Asia Minor and Thrace

Beet necrotic yellow vein virus (BNYVV) is the cause of rhizomania, an important disease of sugar beet around the world. The multipartite genome of the BNYVV contains four or five single-stranded RNA that has been used to characterize the virus. Understanding genome composition of the virus not only determines the degree of pathogenicity but also is required to development of resistant varieties of sugar beet. Resistance to rhizomania has been conferred to sugar beet varieties by conventional breeding methods or modern genome engineering tools. However, over time, viruses undergo genetic alterations and develop new variants to break crop resistance. Here, we report the occurrence of genetic reassortment and emergence of new variants of BNYVV among the isolates of Thrace and Asia Minor (modern-day Turkey). Our findings indicate that the isolates harbor European A-type RNA-2 and RNA-3, nevertheless, RNA-5 is closely related to East Asian J-type. Furthermore, RNA-1 and RNA-4 are either derived from A, B, and P-types or a mixture of them. The RNA-5 factor which enhance the pathogenicity, is rarely found in the isolates studied (20%). The creation of new variants of the virus emphasizes the necessity to develop new generation of resistant crops. We anticipate that these findings will be useful for future genetic characterization and evolutionary studies of BNYVV, as well as for developing sustainable strategies for the control of this destructive disease.


Continued
Interestingly, some of the isolates (TR-S86, TR-S91, and TR-S125) were identical to both of the I-12 and F2 isolates ( Table 2). According to RNA-4-based similarity assessments, TR-S19, TR-S61, TR-S49, TR-S2, TR-S58, TR-S79 and TR-S105 isolates were designated as B-type, however, TR-S67 and TR-S5 isolates were A-type ( Table 2). According to phylogenetic studies, some of the isolates were in a close relationship with the French isolate F2 and some others were related to the Italian isolate I-12. This finding was consistent with the results of the matrix analysis of nucleotide sequence similarities (Fig. 3d).

RNA-5.
Sequence analysis of RNA-5 (P26) in the coding region (420-1160 nt) showed that TR-S49 and TR-S79 isolates were notably different from previous ones, unlike TR-S5. Despite the presence of polymorphism in TR-S5, the nucleotide sequence is the closest (99.58%) to the Chinese isolate CH2 (AB018614.1). Nevertheless, the genetic distance between TR-S49 and TR-S79 isolates and previously reported sequences was higher (97.11%).
The results showed that the studied isolates harbored P26 similar to East Asian J-type (Table 2), but there were a significant number of nucleotide variations. More than 66% of these nucleotide variations lead to amino acid replacements, emphasizing the genetic difference of TR-S49 and TR-S79 (Fig. 2e). Phylogenetic analyses revealed a close genetic relationship between these isolates and those previously reported. Therefore, all isolates were placed in one cluster and the highest relationship was noted with East Asian isolates with bootstrap value over 80% ( Fig. 3e; Sup. 1).

Discussion
To control losses incurred by rhizomania, cultivation of resistant sugar beet cultivars developed through conventional breeding methods and transgenic techniques is considered an effective approach. Emergence and evolution of novel variants of BNYVV threaten sustainable crop production and yield in resistant crops, however, these variants are not widespread yet 3 . ELISA, RT-PCR and restriction fragment length polymorphisms (RFLP) analyses have been widely used in detection and studies of BNYVV structure. ELISA is a routine serological assay used in diagnosis of BNYVV in plants, although concentration of the virus, soil temperature and the type of the sampling tissue affect the efficiency of this method 27,28 . Therefore, the ELISA test is less efficient particularly when virus concentration is low in infected plants 22,29 . The accuracy of the ELISA is less than molecular-based detection methods, however, being fast, cheap and the presence of commercial kits make the ELISA a preferred assay for the routine detection of BNYVV. RT-PCR is a more accurate tool compared to ELISA and is considered a more reliable method in virus detection studies 22,[30][31][32][33][34] hence, it is increasingly used in the detection of BNYVV [35][36][37] . The use of nucleic acid-based methods such as RT-PCR and qRT-PCR has increased the sensitivity of the virus detection by up to 100-10,000 times 27 .
Absorbance values of ELISA are closely related to rhizomania disease index score, so the higher the disease index scores, the greater the absorbance values. The ELISA result is considered positive only if the absorbance values of the samples are more than three times the value in the negative control samples 38 . In this study, the ELISA values recorded for negative controls were below 0.02 (0.012-0.019) therefore values over 0.06 were accepted as positive samples ( Table 1). The ELISA results of bait plants indicate the infection of BNYVV in 57.6% of the sampling regions. A wide range of absorbance values among the positive samples of different regions (from 0.068 to 2.160) could be due to the different severity index of the disease in these areas. RT-PCR analyses of bait plants verified 89% of the ELISA results which could be relevant to the higher accuracy of the RT-PCR. According to RT-PCR results, 51% of the samples are positive for BNYVV and the majority of the positive samples lack the RNA-5 species thus indicating the rare distribution of the higher virulent BNYVV pathotypes in the sampling regions ( Table 1). The results of this study corroborate previous studies reporting the prevalence of BNYVV in sugar beet fields of Turkey 39 .
Since the first record of BNYVV infection in Turkey 40 several research works have diagnosed the disease in sugar beet fields of the country; however, only a few studies have partially characterized the virus 39,41-43 . Partial sequence analyses of RNA-3 isolated from Tokat province in the mid Black Sea region of Anatolia described  www.nature.com/scientificreports www.nature.com/scientificreports/ some differences in the amino acid sequence of the Turkish isolates 43 . Previous nucleotide analyses of RNA-3 and RNA-5 of Turkish isolates assigned these species as A-type and East Asian J-type 39,43 respectively. RFLP-based analyses of the isolates from the northern and central parts of Turkey designated RNA-2 and RNA-3 as A-type species 42,43 nevertheless, RFLP results of some isolates did not match the expected band profile which was the www.nature.com/scientificreports www.nature.com/scientificreports/ inspiration for the current study to examine the presence of polymorphism among Turkish isolates. RNA-1-5 analyses of BNYVV isolated from different regions of Thrace and Anatolia (modern-day Turkey) revealed a unique polymorphism which could be the probable causes of the failure of RFLP analysis in the previous study 42 . Our results are in accordance with previous studies that assigned RNA-2-3 as A-type and RNA-5 as East Asian J-type. Furthermore, this study found that RNA-1 was derived either only from A type BNYVV or from mixture of P and A-types as some of the isolates were harboring RNA-1 highly resembling both European and Japanese P and A-types (Table 2). However, the majority of RNA-1 species were closely related to the Japanese A-type strains. Based on RNA-4 analysis, isolates were mostly designated as either European A or B-types. However, there were some isolates very similar or identical to both A and B-types that seems to be derived from mixture of A and B-types. The finding that multipartite genome of BNYVV could be comprised of RNA species with different sources has rarely been reported 12,44,45 . Studying the isolates from the areas located in a borderline region located between the distribution areas of different BNYVV types indicated the incident of genome re-assortments 45 . Although mixed infection of different virus types is rarely observed 46 it might provide the required condition for this genetic alteration. Genome re-assortments lead to the emergence of new variants of the virus that may overcome the resistance of cultivars.
Among different pathogenic types of BNYVV, strains containing RNA-5 are more destructive than those lacking the fifth RNA species 20,21 . The worldwide distribution of RNA-5-containing isolates is less common than those lacking it. Studying the isolates of 24 provinces in Turkey revealed the rare occurrence of BNYVV strains harboring RNA-5 in infected fields which is in agreement with the earlier report 42 . RNA-5 was detected only in four provinces of Iğdır, Eskişehir, Çorum and Konya and based on RT-PCR only 7 out of 34 soil samples infected with BNYVV were harboring RNA-5 species (20%). Moreover, Iğdır province in Eastern Anatolia had the highest RNA-5 occurrence (57%) among the studied regions. Analysis of RNA-5 in the coding sequences indicated that although all the TR-S5, TR-S49 and TR-S79 isolates were closely related to East Asian J-type, there was a notable nucleotide and amino acid difference between them (Table 2).
In conclusion, this study aimed to carry out sequence analyses of the BNYVV genomic compartments and typology of RNA-1-5 based on pairwise identity and similarity assessments along with phylogenetic relationships. We detected the B type RNA species in the Anatolian region where BNYVV isolates were designated as A type virus in previous studies. Furthermore, our study reveals the occurrence of genomic re-assortments between P, B and A-type viruses which is the first report. The emergence of new variants of the virus threatens the sustainable production of sugar beet in the world, therefore, these findings will contribute to the sustainable control of BNYVV.

Methods collection of soil samples and virus inoculations.
Soil samples were collected from 66 sugar beet cultivation regions in 24 provinces of Turkey and dried out at room temperature (Fig. 4). Two different soil samples known to be infested with BNYVV were supplied by the Sugar Research Institute (Ankara, Turkey) were included in the study as positive controls for RNA-1-5 components. The samples were pulverized and sieved through 2 mm-pores, mixed with sand (3:1 ratio of soil and sand) and filled into plastic pots (1.5 kg per each pot). The experiment was performed in 3 replicates. Beta vulgaris L. cv. Ansa was selected as the BNYVV-susceptible cultivar to be deployed in virus infection experiments of bait plants. In each pot, 30 seeds of bait plants were planted and kept at 25 °C and 16/8-hour light/dark period for 10 weeks. Following germination, 15 plantlets were randomly collected from each pot. The roots were detached, washed with distilled water and dried out. The root pieces collected from each soil sample were compiled and ground with porcelain mortar and pestle using liquid nitrogen. After grinding, one gram of crushed root tissue was used in the ELISA test and the remaining amount was stored at −80 °C to be utilized for RNA isolations and RT-PCR assays. primer designs and Rt-pcR assays. BNYVV nucleotide sequences were retrieved from the GeneBank and highly conserved regions were used to design PCR primers (Fig. 1a). The primer sets specific for BNYVV genomic components (RNA-1-5) were designed using the NCBI Primer Designing tool ( Table 3).
The NG dART RT kit (Eurex, Poland) was used to perform the first strand cDNA synthesis and RT-PCR assay. Prior to screening, positive controls were utilized to optimize the amplification of the RNA species. cDNA concentrations and quality were assessed by a NanoDrop-1000 Spectrophotometer. PCR amplifications were performed in a 25 µL reaction mixture containing 200 ng of cDNA, 10 pmol of each primer, 2.5 mM dNTPs, 0.5 units GoTaq DNA polymerase (Promega, Madison, WI, USA) and 1.5 mM MgCl 2 . PCR was initiated with a denaturing step of 3 min at 94 °C followed by 35 cycles including 94 °C for 1 min (denaturation), 60 °C for 2 min (annealing) and 72 °C for 2 min (extension), and the final extension period of 10 min at 72 °C. PCR products were electrophoresed on 1.5% agarose gel and the GeneGenius Gel Imaging system was used to visualize the amplicons.
Sequence analyses. Some of the isolates were selected based on ELISA and RT-PCR results to undergo further sequence and phylogenetic analyses. The resulting PCR products of RNA-1-5 species were purified from agarose gels after electrophoresis and double sequenced by the Sanger dideoxy method (BM Labosis Ltd. Co., Ankara, Turkey). Sequence analyses were performed for RNA 1-5 species in comparison with relevant sequences retrieved from the GeneBank. The sequences obtained were initially subjected to the NCBI (National Center for Biotechnology Information) Blast analysis. Sequence comparisons were conducted using the CLC Sequence Viewer 7.6.1 workbench (CLC Bio-Qiagen, Aarhus, Denmark). Phylogenetic trees were constructed to study the evolutionary relationships. CLC Sequence Viewer 7.6.1 program was used to perform the phylogenetic analysis 48 . Trees were created by the Neighbor Joining method using the Jukes-Cantor distance algorithm and bootstrap support of 100. Pairwise identity and similarity of nucleotide sequences were calculated through matrix analyses performed by a SIAS online tool (Immunomedicine Group, UCM, Spain).