Association of satellite RNA with Grapevine fanleaf virus in its geographical origin and sequence characteristics

The region between The Caspian Sea and Black Sea has been hypothesized as the origin of grapevine. Likewise, an extensive study on Grapevine fanleaf virus (GFLV) from this region suggests this region as a possible origin of the virus. However, as yet there is no information as to whether or not the virus variants from this region accompanied with the virus satellite RNA (satRNA) and if so, how diverse theses satRNAs can be. To answer these questions, Grapevine fanleaf virus isolates were collected from different vineyards in the northwest region of Iran for the occurrence of the satellite RNA. A total of 421 samples including Vitis vinifera , Chenopodium quinoa , C. amaranticolor , Cynodon dactylis , Medicago sativa were initially screened against GFLV by RT-PCR with the coat protein primers (Cp433, Cp912). When the GFLV-infected plants (36 samples) were screened by RT-PCR with the newly designed satRNA primers Gf750 and Gr750, Phylogenetic studies revealed distinctness of the satRNAs from Iran. This may suggest coevolution of the satRNA with the helper virus because GFLV isolates from this region also form a distinct branch. Among previously reported corresponding satRNA sequences to those from Slovenia and South Central Europe were the closest. This study also reports C. amaranticolor as the natural host for GFLV, which was previously reported only as an experimental host.


Introduction
Grapevine fanleaf virus (GFLV) is responsible for one of the important diseases of grapevine (Vitis vinifera L.), fan leaf, worldwide and causes economic losses (Martelli & Boudon-Padieu 2006). This virus has two genomic RNA molecules each coding for a viral polyprotein (Andret-Link et al 2004; Sokhandan Bashir et al 2009). Some GFLV isolates have an extra RNA molecule known as satellite RNA (satRNA). SatRNAs are thought to be originated from similar helper viruses through mutation events such as deletion, insertion, recombination or reassortment (Cepin et al 2016). GFLV satRNA was initially reported in GFLV isolate F13 which caused severe symptoms on Chenopodium quinoa (Pinck et al 1988). Later on, satRNA was also detected in other nepoviruses such as Arabis mosaic virus and Tomato black ring virus (Cepin et al 2016).
SatRNAs associated with nepoviruses are divided into three groups: large linear, small linear and circular satRNAs (Roossinck et al 1992;Palukaitis 2016). The large satRNAs usually encode nonstructural proteins.
They all encode a protein which is necessary for the replication of satRNA associated with nepoviruses (Hu et al 2009;Cepin et al 2016). In contrast, the small satRNAs are highly structural and have internal base pairing. Small satRNAs of nepoviruses are packaged as linear RNAs; however, during replication in plants they form a circular RNA (Roossinck et al 1992;Palukaitis 2016).
SatRNAs of nepoviruses vary in size from 1.1 to 1.5 kb and have end structures (5′-Vpg and 3′-poly A) like that of the helper virus genome. These satRNAs encode a 37-48 kDa non-structural protein which is necessary, but not sufficient, for the satRNA replication (Roossinck et al 1992). SatRNAs might affect the symptoms incurred by the helper virus (Simon et al 2004); which could be due to satRNA's potential in multiplication and accumulation level in host plant (Betancourt et al 2011). This role of satRNAs involves some aspects of RNA silencing and its suppression as described by Palukaitis (2016). A comparative analysis of GFLV isolates containing satRNA or not, showed no obvious effect on virus accumulation and symptom expression in C. quinoa (Gottula et al 2013).
This study was carried out for the first time to determine presence of satRNA in the GFLV isolates from vineyards in the northwest region of Iran. Vineyards in this region have previously been shown to be widely infected with GFLV. Also, phylogenetic analyses have shown that GFLV variants from this region form a distinct clade. This region is located between The Caspian Sea and Black Sea and hypothesized to be the origin of the grapevine and GFLV (Raski et al 1983). Therefore, we were interested to find out if the GFLV isolates from this region are accompanied with the satellite and, if so, what would be the differences between satRNA of theses isolates compared to that of GFLVs reported from elsewhere. Association of satRNA with viruses may have an important implication as to the biology of helper virus and its potential use in the virus control.

Virus sources, extraction of total RNA and RT-PCR
First, 421 samples including grapevine and several weeds were subjected to total RNA extraction according to Rowhani et al (1993) with modifications as described elsewhere (Sokhandan et al 2007;Khabbazi et al 2017) and then screened by RT-PCR with GFLV coat protein primers to determine the virus infection.
RT-PCR was initially conducted to confirm the GFLV entity of the extracted total RNA by the use of Cp433/Cp912 primers (Table 1). Accordingly, cDNA was synthesized with the reverse primer, Cp912 following the RevertAid First Strand cDNA Synthesis Kit instructions (Thermo Fisher Scientific Inc., Waltham, USA). Subsequently, PCR was carried out using the Cp433 and Cp912 primer pairs under a thermal profile of an initial denaturation at 94 °C for 1 min, followed by 30 cycles of denaturation at 94 °C for 30 s, annealing at 50 °C for 30 s, and extension at 72 °C for 30 s, and a final polymerization step at 72 °C for 7 min. cDNA synthesis was also performed with Gsat1005r, GR1000, Rp or Gr750 primer to detect the virus satRNA, respectively. Then, the generated cDNA was used as the template to amplify the expected fragment by PCR using Gsat191f/Gsat1005r, Gf1000/Gr1000 or Fp3/Rp primer pairs, respectively. PCR amplification was performed by using one Unit of Taq DNA Polymerase (SinaClon Co., Tehran, Iran) in a total reaction mixture of 20 μl containing 1× reaction buffer, 1.5 mM MgCl2, 1 mM dNTPs and 0.25 pmol of each primer. A gradient PCR (for all the satRNA primer pairs) was initiated with an initial denaturation at 94 ⸰ C for 1 min and the reaction was progressed in 30 cycles of denaturation at 94 ⸰ C for 30 secs, annealing at 52, 54, 56, 58 and 60 ⸰ C for 30 sec and extension at 72 ⸰ C for 1 min. The amplification was ended with the elongation step of 72 ⸰ C for 7 min. In PCR with the new primer set Gf750/ Gr750, an annealing temperature at 60 °C for 30 secs was applied. A 5 μl-aliquot of the products were electrophoresed on 1% agarose.

3. Cloning and sequence analysis
DNA fragment amplified by the use of the satellite primers (Gf750, Gr750) were excised from agarose gel and purified using Thermo Fisher DNA Gel Extraction kit (Waltham, USA). The purified fragment (150 ng) from each sample was ligated into pTG19-T cloning vector (50 ng) according to the manufacturer's protocol (SinaClon Corp., Tehran, Iran) and transformed into 100 μl home-prepared competent Escherichia coli DH5α cells through heat-shock method (Chung et al 1989). The transformed bacteria were grown on LB medium containing ampicillin (100 mg/L), X-GaL (20 mg/ml) and IPTG (0.1 M). Three desired colonies each from an isolate (RT-PCR product species) were selected by colony-PCR screening (Sokhandan et al 1997) and the plasmids were extracted following the alkaline lysis method (Green & Sambrook 2012). To confirm that the satellite fragment is cloned into the vector (pTG19-satRNA), BamHI restriction analysis was conducted. Subsequently, recombinant plasmids from the desired colonies were sequenced (Refgen Corp., Ltd., Ankara, Turkey) by the use of satRNA specific primers, Gf 750 and Gr 750. Analysis of the sequences was performed to assess variation between the satRNA sequences.
Nucleotide (nt) alignment and phylogenetic trees were constructed using GeneDoc (Nicholas & Nicholas, 1997) and CLC sequence viewer 7.6.1 program (CLC Bio-Qiagen, Aarhus, Denmark), respectively. The SIAS online tool (Immunomedicine Group, UCM, Spain) was used for calculation of pairwise identity percentages. A 700 nt data from each of the three satRNA isolates were aligned with satellite sequences of GFLV and ArMV which were retrieved from GenBank. The alignment numbering was relative to the satRNA strain R2 (accession number: KC162000.1) to find out the position of highly conserved regions.

4. Glasshouse studies
Glasshouse studies were performed to gather further evidence for association of the satellite with GFLV.
It was hypothesized that if the GFLV possesses the satRNA, it may still be accompanied with the virus after seed transmission. Accordingly, C. quinoa (50 pots) and V. vinifera (20 pots) were used in this study.
One healthy control was included for every five inoculated samples. Three-week old C. quinoa and newly germinated leaves (2 cm long) of 4-month old V. vinifera were subjected to mechanical inoculation. The leaves were cleaned with a piece of wet cloth before being dusted with carborundum and rubbed with sap of GFLV-infected grapevine leaves ground in 0.1 M phosphate buffer (pH 7.2) containing 2.5% (v/v) nicotine. The inoculated leaves were rinsed with water after 10 minutes of the inoculation and the plants were kept at 24 ± 2 ֯ C for 10 to 14 days in glasshouse until appearance of the symptoms. The infected plants were tested by total RNA extraction and RT-PCR for infection with GFLV and presence of the satRNA.
The infected C. quinoa plants were grown to flowering and allowed to produce seeds. The resultant seeds were germinated and maintained under glasshouse conditions for the appearance of any viral symptom before being checked by total RNA extraction and RT-PCR for the presence of the virus and satRNA.

1. Detection of GFLV and satRNA
When total RNA samples were tested by RT-PCR, only 36 samples (8.6%) appeared to have an infection with GFLV. The expected 350 bp and 700 bp fragments were amplified with the coat protein (Cp912, 433) and satRNA (gf 750, gr 750) primers, respectively ( Figure 1). Among the GFLV-infected plants, satRNA was detected only in three plant samples (Table 2) including V. vinifera, C. quinoa, and C. amaranticolor.
However, no satisfying result was achieved using the primers already reported in the literature.

2. Sequencing and phylogenetic analysis
When RT-PCR-amplified satRNA fragment (700 bp) was ligated into pTG19-T and the resulting recombinant plasmid was used for transformation of E. coli, the desired colonies were found among which three colonies (A2, M3, and T1) were randomly selected for further analyses. A subsequent BamHI restriction digestion also confirmed cloning of the 700 bp satRNA cDNA into the pTG19-T vector. When the recombinant plasmids were subjected to sequencing and the resultant sequence data were aligned, homologies of 91-97% were observed between the three satRNA isolates. Maximum (97%) homology was between the isolates A2 and M3 and the least homology (91%) between T1 and M3. By Comparison of these sequences to previously recorded GFLV satRNA sequences, homologies of 70-98% were revealed between them. In addition, comparison of the newly-generated GFLV satRNA isolates with ArMV satRNA sequences (accession number D00664.1 and NC_003523.1) revealed homologies of 59-61% (Table 3).
Further, there were 80-85% homologies between the new isolates and that reported from Slovenia and South Central Europe (Cepin et al 2016). This considerably big difference (15-20%) between these satRNAs could be due to the differences in lineages of the helper virus or independent events in evolutionary history such as deletion or insertion of nucleotides.
In the cladogram constructed on the basis of the satRNA sequences (Table 4), there were two clades so that the newly-generated GFLV satRNA sequences (A2, T1, and M3) were clustered in a distinct position together with previously-reported GFLV and ArMV satRNAs (Figure 2). Thus, the isolates from Iran were clustered in the same clade together with the isolates reported from Slovenia and South Central Europe.
Sequences of GFLV satRNA isolates A2, T1, and M3 were placed in a common subclade although the latter (from C. quinoa) was on a different branch in the same subclade. The first two isolates shared 95% similarity. GFLV satRNA-isolate A2 from V. vinifera was from a different geographical region far away from that of GFLV satRNA-T1 from C. amaranticolor. GFLV satRNA-M3 and T1 were from the same region, but it (M3) shared 99% similarity with satGFLV-A2 and 96% with GFLV satRNA-T1. The accession numbers provided by GenBank for the sequences of GFLV satRNA (A2, M3 and T1) are MK248516, MK248517 and MK248518, respectively. Based on the assessments of GFLV+satRNA transmission to C. quinoa, it could be concluded that the plants infected with satRNA-containing viruses indicate normal growth with local lesions only whereas the satRNA-free GFLV infection causes stunting, chlorosis, and local lesions as well (Figure 3).When the seeds of C. quinoa from the plants infected with satRNA-possessing GFLV were germinated and the progeny plants were compared to the healthy plants, obvious difference in plant height was evident, validating the virus transition to the next progeny. A subsequent RT-PCR assays confirmed the presence of GFLV and satRNA in the progeny (Figure 4).

Glasshouse studies
In this study, samples were collected in late spring, early summer and early autumn. This timing was because of the symptom disappearance in hot summer days (Francki et al 1985). In addition to different samples with various GFLV symptoms, asymptomatic samples from V. vinifera and weeds were also collected, because it has been documented that some V. vinifera varieties are tolerant to GFLV (Martelli & Boudon-Padieu 2006). Thus, they could be infected with GFLV although asymptomatically, as 9 out of 71 asymptomatic samples appeared to be carrying GFLV although none of them had the satRNA.
The distribution pattern of the satRNA among the GFLV-infected vines suggests that the satRNA could be absent in a GFLV-infected plant adjacent to others which contain the satRNA (Gottula et al 2013).
Likewise, in the present study, satRNA-containing GFLV was detected in C. amaranticolor grown in vineyards, but not in the surrounding grapevines in the same vineyard.
To confirm the detection of GFLV and its satRNA, RT-PCR with GFLV coat protein and satRNA primers were applied in this study. SatRNA detection was carried out using different primer pairs, of which three pairs had earlier been utilized in similar researches elsewhere in the world; however, detection of satRNA was not successful with the previously published primers (Gsat191f/Gsat1005r, Fp3/Rp) which might be due to the satellite genetic variation. The genetic variation in the flanking sequences of satRNA and failure in the primer annealing process might be the reason for unsuccessful RT-PCR amplification once other primers were used. Also, previous studies on Iran isolates of GFLV have revealed that they possess distinct genotypes, consequently the primers which were designed based on exotic GFLV isolates could not be efficient in annealing the local isolates (Sokhandan et al 2012). Among the newly designed primers, Gf750/Gr750 had a better amplification thus were utilized in the study. The other newly designed primers (Gf1000/Gr1000) were not useful in the detection and the expected amplified fragment (1000 bp) was very faint and the PCR conditions could not be optimized further.
Comparison of results from the two different primer sets (Gf1000/Gr1000 and Gf750/Gr750) suggested that the satRNA may possess some conserved regions internally rather than at the ends because only Gf750/Gr750 satisfactionally gave the expected band.
The main objective of this study was to determine presence of the satRNA and investigate its diversity in virus isolates of GFLV from Iran. To confirm further that the GFLV is accompanied with the satRNA, seeds from GFLV+satRNA infected C. quinoa plants were germinated and a subsequent analysis of the progeny plants revealed that the transmitted GFLV contained the satellite (Figure 4). This gave a further evidence for association of the satRNA with the GFLV isolates (causing vein banding and mosaic). The results also indicated that the satRNA did not affect severity of symptoms which is in contrast to the previous report that satRNAs impact on their helper viruses and the induced symptoms (Roossinck et al. 1992). In general, the effect of satRNA on virus symptoms is influenced by helper virus, host plant and features of satRNA (Roossinck et al 1992). As reported by Lamprecht et al (2013), helper virus was the determinative factor in the infection of C. quinoa plants when they were mechanically inoculated with satRNA-containing GFLV isolates. These GFLV isolates were infectious only when co-inoculated with GFLV as the helper virus but not when co-inoculated with ArMV-NW.
C. quinoa seeds, infected with GFLV+satRNA, germinated one month later than the control and also showed reduction in plantlet growth in 50% of the replicates when compared to the control plants which were in agreement with the results reported by Saldarelli et al (1993) and Gottula et al (2013). However, with the V. vinifera plants no obvious symptom appeared when they were inoculated with the GFLVinfected samples containing satRNA. Accordingly, the effect of satRNA on infection and symptom appearance might be dependent on the host species as well.
Among 36 infected plant samples, GFLV isolates harboring the satRNA (8.33%) were detected in three cases which demonstrates the low frequency of the satRNA in the studied region. In a similar study conducted by Saldarelli et al (1993) only five isolates from among 34 GFLV-infected samples were determined to contain the satRNA (14.70%). Likewise, in another study more than 100 plants were screened but the satRNA was found in two samples only. Therefore, it appears that the occurrence of satRNA is a rare event (Lamprecht et al 2013) which is speculated that helper virus and host plant are the factors involved in satRNA origination (Roossinck et al 1992;Saldarelli et al 1993;Lamprecht et al 2013).
This study also dealt with the variation of the satRNA compared to the previously reported GFLV satRNA.
Accordingly, phylogenetic studies demonstrated that GFLV isolates M3 (C. quinoa) and T1 (C. amaranticolor) were placed on different branches in the same subclade although they were collected from the same region (Khelejan, East Azarbaijan, Iran). As such, there appears that geographical location is not the only determinating factor for satRNA grouping in phylogenetic trees. As this is the case for the helper virus as well (Lamprecht et al 2013;Cepin et al 2016). Comparison of the satRNA sequences from Iran with the previously reported satRNA sequences suggested that the former were phylogenetically close to that reported from Slovenia and Central Europe, however, the reported satRNA sequences in this study are placed in a distinct sub clade on the phylogenetic tree. The previous studies on CP and MP genes of GFLV in Iran have shown the distinctness of Iran isolates among all the reported GFLV sequences from around the world (Sokhandan et al 2012). So, the distinct positions of satRNAs from Iran in the phylogenetic tree could be because of the distinctness of the helper virus. This also could be strengthening the previous hypothesis that Iran is the origin of GFLV and from there has spread to all over the world (Vuittenez 1970).
The present study is the first study of GFLV satRNA in the origin of grapevine, Iran, which provides new insights into the occurrence and genetic diversity of GFLV satRNA. Results achieved pave the way for future studies of satRNA and development of genetically modified GFLV-resistant varieties of grapevine.

Conclusion
In this study, GFLV satellite RNA was detected and characterized for the first time in the northwest region of Iran by RT-PCR. A heterogeneity of 2 -30% was revealed between the three newly-reported satRNAs and the previously-reported isolates and it was also demonstrated that the GFLV satRNA sequences from Iran are phylogenetically close to that reported from Slovenia and Central Europe. Though they have close relationship phylogenetically but the Iran isolates formed distinct sub clade.   GFLV satellite 870-900 R: 5 ‫ﹶ‬ -TAAWGAGCAACCAAAATCCCA-3 ‫ﹶ‬ F: 5 ‫ﹶ‬ -GTGGSCCCGCRAGTGT-3 ‫ﹶ‬