Abstract: In this study, we reported sequencing and molecular analysis in detail of central part of 2BMP gene (approximately 270 nts) in case of Grapevine fanleaf virus (GFLV) and Arabis mosaic virus (ArMV). For more complex analysis of genetic variability we added reference sequences of evaluated nepoviruses and also sequence of phylogenetically very similar Nepovirus GDefV. Range of sequence analysis showed 74.5-99.3% identity at the nucleotide level and 86.5-100% identity at the amino acid level for partial gene 2BMP among Czech, Italian, German, Swiss sequences and references of all isolates. In this study eleven isolates were sequenced consisting of 5 GFLV and 6 ArMV isolates. As expected, GFLV and ArMV sequences were distinguished based on sequence identities but differences were not significant. Surprisingly, GDefV sequence was almost identical with the majority of GFLV sequences, even more than some GFLV sequences between each other. Local complexity of nucleotide sequences was also evaluated and it was in average of 0.530 for GFLV and GDefV, for ArMV the average value was 0.533. This study provides detail insights for the genetic variability of GFLV, ArMV and GDefV in the central part of the 2BMP coding region.
INTRODUCTION
Grapevine fanleaf virus (GFLV) is one of the oldest known viruses on the grapevine (Vitis vinifera L.) (Martelli, 1986). It is thought that GFLV has coexisted with grapes since their earliest cultivation and has spread with vegetatively propagated crop. The GFLV disease is spread less in the Czech Republic than another Nepovirus, family Secoviridae, ArMV (Kominek and Holleinova, 2003). The ArMV is assigned as quarantine virus in the Czech Republic but GFLV not yet. The GFLV and ArMV induce significant yield reduction and lowering of the quality of grapevine fruit and must. They cause leaves, shoots and fruits malformation, whereas some strains cause yellow discoloration of the leaves. Plant-to-plant spread of GFLV in the vineyard occurs only by the ectoparasitic dagger nematode Xiphinema index Thorne and Allen (Esmenjaud et al., 1993), vector of ArMV is Xiphinema diversicaudatum. Viruses are also transmitted efficiently by grafting and via distribution of infected vegetative propagation materials.
GFLV and ArMV virus particles are very similar on the molecular level. These viruses are typical for their segmented genome. The genome is the main component of virus particle. Virus particles have angular outline roughly 30 nm in diameter, containing a single protein species of Mr 56,000. The genome consists of two positive-sense ssRNA molecules (RNA1, RNA2) which are encapsidated separately. Both genomic RNAs are covalently linked to their 5’ ends to small Viral Protein (VPg) and they are polyadenylated in their 3’ ends. The 3’-NCR of both RNAs are identical for many nepoviruses (Le Gall et al., 1995). The first complete sequenced macromolecule of GFLV RNA1 and RNA2 belonged to F13 strain (Serghini et al., 1990; Ritzenthaler et al., 1991). The final length was 7342 and 3774 nucleotides for RNA1 and RNA2, respectively.
RNA2 of ArMV was completely sequenced for the first time in 2001 by Wetzel et al. (2001) and RNA1 of ArMV was completely sequenced in 2008 (Dupuis et al., 2008).
The RNA1 and RNA2 are monocistronic and each encodes a single polyprotein that is processed proteolytically into functional proteins required to complete the virus life cycle. The RNA2-encoding P2 polyprotein which contains (from the N-to C-terminus) the domains for the homing protein (2AHP), the movement protein (2BMB) and the coat protein (2CCP) (Margis et al., 1993). The 2AHP is localized in the replication site and has been implicated in RNA1-dependent replication of RNA2 (Gaire et al., 1999). The 2BMP is a movement protein and is found in tubules observed in the plasmodesmata (Ritzenthaler et al., 1995). The 2CCP is a multifunctional coat protein, important in specific transmission by X. index Thorne and Allen, encapsidation of genomic RNA and systemic spread in plants (Hewitt et al., 1958; Belin et al., 2001; Callaway et al., 2001; Andret-Link et al., 2004). The GFLV was observed in various molecular variants in many countries of Europe, Africa, Middle East, North and South America (Naraghi-Arani et al., 2001; Vigne et al., 2004; Fattouch et al., 2005; Bashir et al., 2007; Pompe-Novak et al., 2007; Liebenberg et al., 2009; Mekuria et al., 2009), occurence of ArMV is also global (Wetzel et al., 2002; Lockhart, 2006; Rakhshandehroo et al., 2006). The GFLV was described in the Czech Republic only by Kominek et al. (2006) who described the mild isolate HV5 (Eichmeier et al., 2010, 2011). The ArMV is a quarantine virus in relation to species Fragaria L. and Rubus L. Molecular information about ArMV as viral agent of grapevine has not been published yet in the Czech Republic.
The majority of the studies listed above are focused on special regions on the RNA2 molecule of these two nepoviruses. Studies focused on characterisation of gene variability localizated on RNA1 molecule are still rather exceptional (Ritzenthaler et al., 1991; Wetzel et al., 2004; Dupuis et al., 2008; Mekuria et al., 2009; Oliver et al., 2010; Eichmeier et al., 2011; Lopez-Fabuel et al., 2013). The spatial spread of both viruses in the Czech Republic is not the same. The GFLV is not so frequent as ArMV (Kominek and Holleinova, 2003). Thereby ArMV is a bigger threat than GFLV in the Czech Republic. It can be caused by the fact that vector of ArMV X. diversicaudatum is natural in the Czech Republic but the vector of GFLV X. index Thorne and Allen hasn’t been discovered in the Czech Republic yet (Kumari et al., 2005). We suppose that spread of GFLV in the Czech Republic is realized mainly by vegetative propagation. Another theoretical danger can be the fact of possible existence of recombinants between ArMV and GFLV (Mekuria et al., 2009) which may be spread by nematodes that are natural in the Czech Republic.
Nucleotide sequence coding 2BMP gene is the most homologous genomic portion of the RNA2 strand between GFLV and ArMV (Wetzel et al., 2002). Mekuria et al. (2009) further stated that 2BMP sequences seem to be “hot spots” for interspecies genetic exchanges between closely related grapevine-infecting nepoviruses.
Recombination can be an important factor in viral evolution (Garcia-Arenal et al., 2001; Moury et al., 2006) and in the case of examined nepoviruses, recombination has been reported to occur within RNA2 of ArMV and GFLV among distinct genetic variants (Vigne et al., 2004; Boulila, 2007; Pompe-Novak et al., 2007; Mekuria et al., 2009; Vigne et al., 2009) and Grapevine deformation virus (GDefV) (Mekuria et al., 2009).
MATERIALS AND METHODS
Plant material: The isolates were collected since 2008 according to screening tests by RT-PCR.
| Table 1: | RNA2 sequences of Grapevine fanleaf virus (GFLV), Arabis mosaic virus (ArMV) and Grapevine deformation virus (GDefV) used in this study |
Five infected grapevines of two grape cultivars (Kodrjanka and Pamjati Negrula) planted in the South Moravia (Czech Republic) and two infected Italian grapevines of cultivars Dimrit and URS were included to this study. Cultivars Kodrjanka and Pamjati Negrula have origin in Moldova and are grown in the South Moravia (Czech Republic) more than 20 years. Two additional isolates were provided by Mediterranean Agronomic Institute of Bari (Italy) to this collection (URS and Dimrit). Six infected grapevines (Pinot Noir and unknown cultivars) and Nicotiana clevelandii plants infected by ArMV were used for this study (Table 1).
RT-PCR: Reaction for reverse transcription consists of water (HPLC purity), 0.25 μg random primer P(dN)6 (Roche, Indianapolis, USA) and 0.1 μg (2 μL) of total RNA in total volume 12.5 μL. This mix was denaturated at 95°C for 5 min at first. Subsequently 5 μL 1xRT buffer (Fermentas, Bourlington, Canada), 1.25 μL 10 mM dNTPs (Invitek, Berlin, Germany) and 0.5 μL 200 U reverse transcriptase M-MLV-RT (Fermentas, Bourlington, Canada) were added. The time for reverse transcription was 60 min at 42°C. The PCR master mixture was prepared according to Wetzel et al. (2002).
Sequencing of PCR products: The PCR amplicons of expected lengths were cut from the agarose gel and purified by NucleoSpin Extract II (Macherey-Nagel, Germany). Nucleotide sequencing was done by BigDye® Terminator v3.1 (Applied Biosystems, USA). The products of sequencing reaction were separated using genetic analyser ABI-PRISM 310 (Applied Biosystems, USA). The sequencing was done in 5’-3’ direction for each isolate. The amplification primers (Wetzel et al., 2002) were also used as sequencing primers.
RESULTS AND DISCUSSION
The aim of this study was to advance our knowledge of the genetic variability of GFLV and ArMV RNA2 by characterizing the partial nucleotide sequence of 2BMP gene of nine viral isolates from Czech Republic, Italy, Germany and Switzerland. For higher relevance, sequences from database GenBank NC_003623 (GFLV reference sequence), DQ386866 another known Czech isolate, NC_006056 (ArMV reference sequence) and AY291208 (GDefV) (Table 1) were added to sequence analysis. All obtained sequences were submitted to NCBI/GenBank, all accesion numbers are listed in Table 1. The isolates S10 (ArMV) and K4 (ArMV) were the most interesting, obtained ArMV isolates from the view of symptomatology. Symptoms of infections of these two isolates were absolutely indefinable from appearance of GFLV infection. Those isolates were obtained from vine cultivar Pinot Noir in different production vineyards in the Czech Republic, wine region Moravia.
Nucleotide analysis: Obtained nucleotide sequences in the lenght of 267-288 nts located in position of reference sequence GFLV (1337-1603 bp) were equivalent to position of 1403-1669 bp of ArMV reference sequence. In case of nucleotide chain of reference sequence. The GDefV the region is located in 1335-1601 bp as it is closer to GFLV than ArMV genome. These sequences were subsequently aligned by software CLC Main Workbench 5.0 (CLC bio, Denmark) (Fig. 1).
| Fig. 1: | Multiple alignment analysis of obtained
partial sequences of 2BMP GFLV, ArMV sequences supplemented
by reference sequence, GFLV, ArMV and GDefV. Dots mean identical nucleotides
and the lower line is consensus with an conservation degree |
Figure 1 shows very close relations among members of monitored nepoviruses but some differences are clear. Particularly distinct is proportion between GFLV and ArMV sequences, GDefV sequence is more similar to GFLV sequences of monitored nucleotide region. Pairwise analysis was applied and proved interesting results, all sequences together provided percent identity from 74.5-99.3% and no gaps with the exception for sequence of HV5 where 5 nucleotides of the 5’ end were missing in the alignment. Monitored sequence coding 2BMP of GFLV is slightly more conservative than the same 2BMP portion of ArMV as showed in Fig. 1.
Pairwise analysis of GFLV isolates was done. The highest percent identity was detected between sequences of isolates PN33 and KO1 relative to PN35, 99.25%. The lowest percentage identity 82.4% was detected between sequences of 55TK and HV5 isolates which is distorted because HV5 isolate contains 5 gaps in the alignment. The GDefV sequence is more similar to GFLV than to ArMV in the alignment of sequenced 2BMP part. Percent identity of GDefV sequence and GFLV isolates consensus was estimated as 87.64%, GDefV and ArMV as 77.9%. Consensus of GFLV and ArMV sequences were compared and estimated percent identity was 80.2%. Pairwise analysis of ArMV isolates proved percent identity of 83.9% between sequences of S10 and K4 isolates. This is an interesting fact because both isolates come from Moravia (Czech Republic) showed the highest differences in 2BMP region among all isolates. Maximal identity of 98.5% was between sequences AR/G2 and AR/G4 which come from the same vineyard in Germany.
The relative information content of DNA can be described by its sequence complexity, a quantity obtained from DNA reassociation kinetics. It can be measured by software CLC Genomics Workbench 6.5.1 (CLC bio, Denmark). The local complexity is a measure of the diversity in the composition of nucleotides within a given range of 2BMP sequence. The K2 algorithm was used for calculating local complexity (Wootton and Federhen, 1993).
Based on results of pairwise analysis GFLV sequences were analyzed together with GDefV sequence mainly because of close sequence similarity. Local complexity of GFLV reference sequence was calculated from minimal value 0.254 to maximal value 0.596 (NC_003623), average value of local complexity for this sequence was 0.530. Average value of all GFLV and GDefV sequences based on monitoring of 2BMP portion was estimated 0.530.
Local complexity of ArMV nucleotide sequences was calculated from 0.330 to 0.596 (NC_006056) of sequenced 267 nts. Average local complexity for 267 nts of all ArMV sequences was calculated as 0.533 for all ArMV sequences was estimated.
Thus, local complexity of obtained ArMV sequences is higher by 0.003 than local complexity of average of GFLV and GDefV sequences.
Amino acids analysis: Putative amino acid chains were derived from nucleotide sequences. These chains were analyzed in detail at the level of divergence for one amino acid.
Multiple alignment (Fig. 2) depicted amino acid sequences derived from nucleotides. The GFLV sequences are very conservative apart from sequence of 55TK isolate, sequence of this Italian GFLV isolate is different in 4 amino acids, in position 132 Met it is substituted by Asp, in position 133 Glu is substituted by Gly, in position 134 Glu is substituted by Arg and in position 145 Ile is subtituted by Asn. Interesingly, sequence of GDefV is different from GFLV sequences only by one amino acid in position 135 where Tyr is substituted by Phe.
ArMV sequences are more variable in monitored 2BMP region. There is an interesting point in position 133 where sequences contain 2 types of amino acids, Glu and Asp (both amino acids have negative electric charge and are acidic), this can be supposed as a point of recombination events between mentioned nepoviruses.
| Fig. 2: | Multiple alignment of deduced amino acid sequences derived from nucleotide sequences coding 2BMP |
Position 134 also show similar variability level where the majority of ArMV sequences show Asp, majority of GFLV sequences Glu, AR/G4 show Cys and interesingly K4 (ArMV) and 55TK (GFLV) show Arg. Moreover, in this case there are described 2 aminoacids which are contiguous and this spot is very variable in the set of monitored sequences.
Isolate K4 and AR/S2 sequences show differences from the other GFLV and ArMV sequences also in position 137 where K4 shows Asn and AR/G2 Gly.
Positions 117, 175 and 184 are typical by their amino acids of monitored nepoviruses, according to these positions GFLV and ArMV sequences can be clearly distinguished. Only the sequence of isolate K4 shows a different amino acids in position 176 and 177.
According to this study it is possible to determine the sequences of K4 (ArMV) and 55TK (GFLV) are most distinct in the entire test set.
Pairwise analysis was done on all samples and the range of identity was from 86.5-100% (CLC Genomics Workbench 6.5.1), between GFLV and GDefV 93.3-100% and between ArMV 93.3-100%.
Phylogenetic study: Phylogenetic analysis (Fig. 3) declassifies the relationship within the monitored isolates at the amino acid level. Phylogenetic tree is divided into 2 main clusters and clearly uncovers the basic differences between GFLV and ArMV sequences also in the set of relatively short sequences. The top cluster with GFLV contains 8 sequences which are grouped of closely identical sequences of isolates from the Czech Republic (PN35, KO1, HV5 and PN33), France (F13) and Italy (UR11). Interestingly, according to phylogenetic analysis of sequences the sequence of the isolate GDefV N66 from Turkey is more similar to sequences of GFLV isolates than sequence of 55TK isolate from Italy and also than ArMV sequences. This is a clear evidence of very variable and unstable coding region from the phylogenetic viewpoint. Second cluster contains sequences of ArMV isolates which come from various plant hosts. Identical sequences have isolates OL1 from the Czech Republic and AR/G2 from Germany followed by reference sequence of isolate NW from Germany, sequence of isolate S10 from Moravia-Czech Republic comes from naturally infected grapevine planted in production vineyard, these sequences are followed by AR/S2 from Switzerland and AR/G2 from Germany. Phylogenetically, sequence of K4 from Moravia which also come from naturally infected grapevine is the most diverse at the amino acid level.
In molecular epidemiology of infectious diseases, phylogenetic inference is also an important tool. The very fast substitution rate of RNA viruses, means that these show substantial genetic divergence over the time-scale of months and years.
| Fig. 3: | Phylogenetic tree of analyzed isolates. Algorithm UPGMA, distance measure Jukes-Cantor and bootstrap 100 replicates parametres were used (CLC Genomics Workbench 6.5.1). Scale bar represents a genetic distance of 0.018 |
Therefore, the phylogenetic relationship between the pathogens from individuals in an epidemic can be resolved and contributed valuable epidemiological information about transmission chains and epidemiologically significant events (Leitner and Albert, 1999; Forsberg et al., 2001). Very high percent identity at nucleotide level in the study for examined partial 2BMP gene of nepoviruses GFLV, ArMV and GDefV was approved. Partial gene 2BMP of all examined sequences (Table 1) at the amino acid level together showed 74.5-99.3% (GFLV and GDefV) and 86.5-100% (ArMV) percent identity. Basically, the central part of the 2BMP coding region was monitored in this study. There is a very interesting fact that sequences of GFLV and GDefV shared 93.3-100% and ArMV sequences together 99.3-100% which showed absolutely same values. Results of this study confirmed the results of Ritzenthaler et al. (1995) which indicated high percent identity of GFLV and ArMV amino acids in 2CCP coding region (69%) and in 2BMP (88%), which confirmed higher variability in coding sequence 2CCP. RNA-2-encoded proteins of GFLV and ArMV share 68-78 % similarity at the amino acid level (Vigne et al., 2008) which confirmed very close phylogenetic relationship of both examined nepoviruses. Percent identity of partial 2CCP nucleotide sequences described earlier ranged from 83-86% and from 81-91% for the amino acid residues, respectively (Eichmeier et al., 2010). Variability of 2BMP gene can also play an important role in transmissibility by X. index (Belin et al., 2001; Andret-Link et al., 2004). This is approved by the fact that chimeric virus with protein 2BMP of GFLV origin was transmitted by X. diversicaudatum as well as constructs with protein 2BMP of ArMV origin which indicates that protein 2BMP is not involved in transmission specificity (Belin et al., 1999). Belin et al. (1999) supposed that the nine C-terminal residues of 2BMP must be of the same virus origin as the proteinase for efficient proteolytic processing of polyprotein P2 and from the same virus origin as the 2CCP for systemic virus spread. It would mean that recombinations could be followed by new quasispecies only in the case that all described proteolytic activities of nucleic acids were in a proper relation to species of virus. This is the main reason why it is so important to know the variability of coding sequences where recombinations can occur. The central part of the 2BMP coding region which was desribed in this study at the molecular level clearly belongs to mentioned types of sequences with recombination potential.
ACKNOWLEDGMENT
The author would like to thank Tomás Kiss and Eva BÍezinová from Mendel University in Brno for critical reading of the manuscript. This study was supported by the Ministry of Agriculture of the Czech Republic, Project No. QH91214. Access to computing and storage facilities owned by parties and projects contributing to the National Grid Infrastructure MetaCentrum, provided under the programme “Projects of Large Infrastructure for Research, Development and Innovations” (LM2010005), is greatly appreciated. Finally, this study was supported by Project No. CZ.1.07/2.4.00/31.0089.