The passion fruit is one of the most important crops in Brazil. It belongs to Passiflora genus and Passifloraceae family. Viral diseases are the most severe in terms of passion fruit yield loss (Sokhandan et al., 1997; Gibbs et al., 2008b). This is mainly due to the lack of both quick and accurate diagnosis as well as effective methods for disease control (Andrade and Pio-Ribeiro, 2001). In Brazil and Africa woodiness disease is caused by Cowpea aphid-borne mosaic virus (CABMV) (McKern et al., 1994; Nascimento et al., 2006; Maciel et al., 2009). This is in contrast to Australia where the passion fruit woodiness virus (PWV) is the most common causal agent (Sokhandan et al., 1997) and also to Asia where the same disease is caused by the East asia passiflora virus (EAPV) (Iwai et al., 2006). Its genus Potyvirus, family Potyviridae, contains more agriculturally important plant viruses than any other (Adams et al., 2005). Although in the past 15 years molecular studies about Brazilian CABMV have been carried out (Novaes and Rezende, 2003; Nascimento et al., 2004; Barros et al., 2011; Cerqueira-Silva et al., 2012; Nicolini et al., 2012), surprisingly none of them has phylogenetically investigated the virus from the region of largest passion fruit production in Brazil, the Livramento de Nossa Senhora County. Exploring the molecular variation of the coat protein (CP) in CABMV helps to better understand variability among strains, isolates and related species, to elucidate the evolutionary history of CABMV as well as assists to shed light on disease outbreaks in the course of passion fruit dispersion (Iwai et al., 2006; Gibbs et al., 2008a; Gibbs et al., 2008b). Bahia State is the place where passion fruit woodiness disease was firstly identified in Brazil during the 70’s (Chagas et al., 1981) before its dissemination to most of the countries’ fields. This led us to hypothesize whether Bahia State might host phylogenetically different isolates than the rest of Brazil that could be due to genomic mutation, recombination with other viral species, selective pressure or other evolutionary events. In this study we report a phylogenetic characterization of a population of CABMV from the State of Bahia, in comparison to other viruses causing fruit woodiness disease in passion fruit vines around the world.

Virus and plant resources

Ten CABMV isolates were collected from different passion fruit crop fields in the farming region of Livramento de Nossa Senhora Country, Bahia State, Brazil (13° 38′ 34″ S, 41° 50′ 27″ W). Ground extract from symptomatic leaves was used for inoculation on healthy passion fruit plants for maintenance, following the method described by Novaes and Rezende (2003). The 10 CABMV isolates here reported were named: CABMV-Lns1 to CABMV-Lns10. Several sequences from passion fruit woodiness disease-related viruses were acquired from public databases (Supplementary Table 1 and Figure 1).

RNA extraction, PCR, cloning and sequencing

RNA extraction was performed following the Arabidopsis Functional Genomics Consortium Trizol protocol (Arabidopsis Functional Genomics  Consortium – AFGC, 2013).  cDNA  synthesis and  PCR   (Promega), DNA purification (PCR Purification Kit – NORGEN, Wizard SV Gel and PCR Clean-Up System – Promega) and cloning (pGEM-T Easy Vector – Promega, Max Efficiency DH5α Competent Cell – Invitrogen) were performed as essentially as described in the manufacturer’s protocol. Newly designed primers for both cDNA synthesis and PCR were: CABMV8364-F (5’-CCTTTCCTTCTACGATG-3’), CABMV9389-R (5’-CAACCGGGGTATGGCCTC-3’), CABMV8359–F (5’-GGCATCCTTTCCTTCTATG-3’) and CABMV9402–R (5’-GGCATCCTTTCCTTCTATG-3’).

Sequencing and phylogenetic analysis

Sequencing of at least three independent clones from each isolate was performed twice by Macrogen (South Korea). The sequences were manipulated using DNA Baser Sequence Assembler v2 (DNA Baser Sequence Assembler v2.x, 2010). Phylogenetic tree was constructed using MEGA5 (Tamura et al., 2011). For nucleotide-based tree, Neighbor-Joining (NJ) and BioNJ pairwise distances matrix were estimated by maximum composite likelihood (MLC), estimated under bootstrap values higher than 70% (1,000 replications).

The high identity of nucleotide and amino acid sequences amongst the ten CABMV-Lns isolates is strong evidence that they belong to the same strain of CABMV. This information is of great relevance for purposes of disease control, once most of the attempts to overcome woodiness disease in passion fruit such as premunization with mild strains of CABMV/PWV or pathogen-derived resistance have been hindered by variant strains of the same virus (Novaes and Rezende, 2003; Alfenas et al., 2005, Trevisan et al., 2006; Cerqueira-Silva et al., 2014). On the other hand, CABMV-Lns clustered separately from the isolates CABMV-Jgr and CABMV-Itb, which were also collected in Bahia State. As CABMV-Jgr and CABMV-Itb clustered together with isolates from rather far regions, the data suggest that CABMV-Lns are likely a strain restricted to Bahia State. This finding proposes either a single introduction of this strain to the area or a constraint in dissemination of this strain to other places by plant material, since there is no evidence that the virus is seed-transmitted in Passion fruit. According to the demarcation criteria for the Potyvirus genus (King et al., 2011), our data support the hypothesis that CABMV-Lns isolates are a different strain from CABMV-Prp, CABMV-Cr, CABMV-Vni, CABMV-Z, CABMV-Iba, CABMV-Monguno, CABMV-Mor, CABMV-MG-Avr, CABMV-PE-Bnt (80 to 89% amino acid identity). Additionally, our data confirm the higher sequence identity and closer phylo-genetic relationship of CABMV-Lns with Southern African CABMV rather than isolates from the North of Africa, which was also previously suggested (Nascimento et al., 2006). The slave trade between Africa and South America is the likely event for introduction of CABMV in Brazil (Gibbs et al., 2008b), by which infected biological material could have been transported from the South rather than the North of Africa. However, we are aware that the limited number of CABMV sequences currently available, associated to the little information about African CABMV infecting Passion fruit plants, significantly narrow down the possibilities of drawing more precise conclusion on this regard. With reference to the important CP conserved motifs of potyviruses, the DAG motif is related to transmissibility of potyviruses by insects (Abdullah et al., 2009). Isolate CABMV-Lns6 shows a substitution of aspartic acid by alanine it to the DAG motif (Supplementary Figure 1). Mutations in this region significantly reduced transmissibility by insects in previous studies (Abdullah et al., 2009; López-Moya et al., 1999). However, unfortunately transmissibility experiments to investigate the consequences of the mutated DAG motif on the transmissibility capacity of CABMV-Lns6 were not suitable before the submission of this work.

Our data precisely demonstrate that CABMV-Lns belong also to CABMV species. These isolates can be classified as a strain of CABMV thus far occurring only in Bahia State and phylogenetically more distant from other isolates occurring in the same region. In addition and accordingly to previous reports, they are more closely related to Southern rather than Northern African CABMV. The exploitation of a larger number of viral samples from other as yet unexplored places in the promising State of Bahia will help to shed light on the events contributing to the observed regional isolation of CABMV-Lns. The whole genome sequencing of CABMV-Lns is ongoing and will allow us to comprehend their evolutionary history with regard to other related viral species around the world.

The authors did not declare any conflict of interest.

This study consists one chapter of the first author’s Master of Science Thesis, funded by the Coordination for the Improvement of Higher Level Personnel (CAPES) in the Graduate Program of Plant Biotechnology at the Federal University of Lavras, MG, Brazil. It was partially funded by the Research Support Foundation of Bahia State (FAPESB, 0852/2006, Undergraduate Research Program-PIC/UESB Scholarship to first author) at the Southwest Bahia State University (UESB). We acknowledge the assistance of Prof. Leandro M. de Freitas (UFBA) for phylogenetic analyses, Allan S. Pereira and Claudio Benicio C. Silva (UESB) for field work and Carlos Roberto Torres (UFLA) for plant/virus maintenance. We also thank Dr. Scott Sinclair for considerably improving the text and the anonymous reviewers for their careful reading of our manuscript and their insightful comments and suggestions.