​RCA-based Detection of Begomoviruses in Weed Genera Associated with Legumes in Southern Karnataka

Global food production is required to be doubled as the total population is estimated to reach 10 billion by 2050 (Willet et al., 2019). Grain legumes are important source of dietary proteins, particularly in India where a significant population (about 43%) are vegetarians. India accounts for about 23.46% of global legume production (17-22 MT) with an average area of 23.9 MH (Rao 1982). Despite such a vast area under legume cultivation, shortage of pulses was encountered at times due to sudden drop in production. Among the various factors, the unexpected burst of Yellow Mosaic Disease caused by begomoviruses accounts for a sudden drop in overall legume yield. In India, approximately an annual yield loss of about USD 300 million was estimated in crops such as blackgram, mungbean and soybean (Varma and Malathi 2003; Hema et al., 2014). Therefore, studies related to the characterization and management of begomoviruses are important towards avoiding loss in overall legume yield.
       
The begomoviruses belong to family Geminiviridae with either monopartite or bipartite genome assembly and transmitted through the whitefly vector, Bemisia tabaci, in a circulative persistent manner in dicotyledonous crops. The first report of YMD was from western India in Lima bean and later in mungbean in northern India (Capoor and Varma 1948; Nariani 1960). YMD induces yellowing of veins, yellow mosaic and leaf curling with 100 per cent yield loss (Dasgupta et al., 2003; Borah and Dasgupta 2012). The yield loss due to begomovirus infection is not only limited to legumes but also in other crops like tomato, chilli, okra and papaya (Varma and Malathi 2003; Leke et al., 2015; Inoue-Nagata et al., 2016; Sharma et al., 2018).
       
The weeds of legume fields were reported to serve as alternative hosts and virus reservoirs during the offseason. The weed, A. conyzoides was shown to harbor mungbean yellow mosaic India virus and tomato leaf curl virus (Naimuddin et al., 2014; Pandey et al., 2017). Thus, increasing evidence suggest that the weed plants may serve as permanent source of infection and may have a role in disease epidemiology (Saikia and Muniyappa, 1986; Ansari and Tewari 2005; Wisler and Norris 2005). Therefore, the identification and characterization of weed hosts in the agricultural fields will prove crucial in weed management and as well as in the disease management.

The incidence of YMD was assessed by roving surveys carried out in major legume crop growing regions of Tumakuru, Chikkaballapura, Mandya, Chamarajanagara, Bengaluru rural and urban districts during 2017-18. The infected leaf samples with symptoms and asymptomatic weeds around each legume cropping systems were collected for molecular analysis. The per cent disease incidence of YMD was calculated by counting the number of plants infected out of total number of plants.
       
The molecular analysis of the leaf samples was carried out in Plant Virology Laboratory, Department of Plant Pathology, college of Agriculture GKVK, Bengaluru. Total nucleic acid from the leaf samples of legume crops and weeds collected during the surveys was extracted by CTAB method (Maruthi et al., 2002). PCR was performed in 25 μL reaction mixture containing 12.5 μL master mix, 2 μL of DNA template, 6.5 μL of sterile distilled water and 2 μL each of forward and reverse primers. The genomic regions of YMV were amplified using begomovirus specific Deng primer pair (5' TAATATTACCKGWKGVCCSC 3' and 5' TGGACYTTRC AWGGBCCTTCACA 3') (Deng et al., 1994), HgMYV-CP gene specific primer pairs (5'ATGCTTGCAATTAAGTACTT GCA3' and 5'TAGGCGTCATTAGCCATAGGCA3¢) (Anburaj et al., 2010) and MYMV-CP gene specific primer pairs (5'ATGGGKTCCGTTGTATGCTTG3' and 5'GGCGTCATTA GCATAGGCAAT3') (Naimuddin and Akram, 2010) amplifying around 500 bp and 900 bp regions, respectively.
       
The RCA was carried out using bacteriophage Φ29 DNA polymerase included in the “IllustraTempliPhi 100 Amplification Kit” (GE Healthcare) to enrich YMV template in samples that failed to amplify with direct PCR. The product obtained by RCA was subjected to PCR to detect trace amount of YMV in the samples. The Deng primers and CP specific primers were used to amplify the RCA product with the expected amplicon size of 500 and 900 bps, respectively. PCR with 2 μL of RCA product and mixture explained above was done. After PCR confirmation, the amplicons were sequenced in both directions using CP gene specific forward and reverse primers at Eurofins Genomics India Private Limited and the sequences were submitted to NCBI database with accession numbers [MK391938-MK391952; MK409375-MK409378].
       
The sequences of both forward and reverse reactions were aligned and the non-overlapping regions were joined together to compile full length sequence using ‘Basic Local Alignment Search Tool (BLAST)’ at the National Centre for Biotechnology Information database. The sequences were compared with other respective viral sequences at the NCBI database using BLAST and the multiple sequence alignment was carried out using CLUSTALW2 tool. The phylogenetic Neighbor-Joining tree analysis was conducted using MEGA7.0 software package. Robustness of the tree was determined by bootstrap sampling of multiple sequence alignment with 1000 replications.

Survey and YMD incidence in legume crops
 
Although, the legume fields chosen for the surveys were at varying stages of growth and showed a large variation in the extent of YMD incidence typical yellow mosaic symptoms were observed in all the surveyed regions (Table 1; Fig 1). Maximum YMD incidence of 64.6% and 86.5% was observed in horsegram fields in Kharif 2017 and Summer 2018 respectively while a minimum of 9.0 and 11.1% incidence was observed in field bean. Disease incidence of 27.4%-76.1% was observed in mungbean field. Earlier, maximum disease incidence of YMD in mungbean in Tumakuru was reported 79.54% during Rabi season (Manjunath et al., 2011) and in New Delhi 35.7% in both pre Kharif and Kharif seasons (Biswas et al., 2015). Further, the total DNA extracted from the leaf samples were tested for the presence of begomovirus. Using Deng primer pair an amplicon of 500 bp was obtained in PCR reactions of all the crops (Fig 2a). While both the CP gene specific primers resulted in 900 bps size amplicon with RCA enriched DNA samples of all the crops except the field bean samples (Fig 2b). Previously, 900 bp size amplicon was also observed using HgYMV CP gene specific primers in lima bean with symptoms of YMD (Bhagyashree et al., 2017).
 

 

 

 

 
YMD incidence among associated weeds
 
Sixty weed samples collected randomly from the surveyed legume fields were identified into 20 different species representing nine different families such as Amaranthaceae, Asteraceae, Commelinaceae, Convolvulaceae, Euphorbiaceae, Fabaceae, Malvaceae, Poaceae and Rubiaceae. The weed samples were categorized as symptomatic and asymptomatic depending on visible YMD symptoms such as yellowing, yellow mosaic, leaf curl, vein clearing, leaf distortion, etc. (Table 2). A. conyzoides, A. sessilis, C. benghalensis and E. geniculata were found most abundantly across all the cropping systems (Fig 3).
 

 

       
In blackgram cropping system, six weed species including three asymptomatic C. dactylon, E. geniculata and P. hysterophorus were confirmed for the presence of virus with Deng primers. Symptomatic A. sessilis and asymptomatic A. viridis did not amplify through PCR with Deng primers but resulted in 500 bp amplicon when subjected for RCA-PCR (Table 3a-c).
 

 

 

       
In field bean cropping system, presence of virus was confirmed in symptomatic E. geniculata by PCR with begomovirus specific primers. However, asymptomatic A. aspera and D. arvensis showed no amplicon with direct PCR but resulted in 500 bp amplicon after RCA-PCR. Similarly, symptomatic A. aspera and D. arvenensis did not amplify in direct PCR but showed 500 bp amplicon after RCA-PCR (Table 3a-c). 
       
Six of ten weed samples collected from horsegram fields were symptomatic and remaining four were asymptomatic. Two of the symptomatic weed samples confirmed for the presence of YMV by direct PCR using either Deng primers or CP gene specific primers. In the remaining four symptomatic weed samples, only A. sessilis was detected positive for YMV after enrichment of virus template by RCA. Among the four asymptomatic weed samples, two were confirmed for the presence of virus by direct PCR while remaining two did not amplify despite RCA enrichment of virus template (Table 3a-c).
       
Out of 17 (9 symptomatic and 8 asymptomatic) weed samples collected from mungbean cropping system, twelve samples detected for the presence of begomovirus through direct PCR using Deng primers. Of the remaining five weed samples subjected for RCA-PCR, four were confirmed for the presence of begomovirus. Of these four, two were asymptomatic such as A. aspera and A. hispidium and remaining two were symptomatic namely, A. sessilis and E. geniculate (Table 3a-c).  
       
In the pole bean fields, asymptomatic weed, S. rhombifolia was positive for begomovirus by direct PCR while asymptomatic weed species A. viridis and E. geniculata were confirmed only after RCA-PCR (Table 3a-c).
       
Out of 10 weed hosts collected from soybean fields, eight were symptomatic and two were asymptomatic. Four of these were confirmed by direct PCR and four were confirmed by RCA-PCR. The symptomatic A.conyzoides, A. viridis, C. tora and E. geniculata were confirmed for begomovirus presence only by RCA-PCR (Table 3a-c).
       
The percentage of weed species detected with begomovirus from around blackgram, field bean, horsegram, mungbean, polebean and soybean is 85.71%, 66.67%, 40%, 94.11%, 71.42% and 60% respectively. Overall, out of 60 weed samples collected from six legume cropping systems, 31 weed samples (51.67%) amplified for begomovirus specific product through direct PCR using Deng primers while remaining 29 did not amplify. However, upon enrichment of viral DNA, 16 weed samples out of 29 resulted in expected amplicon (Table 3c, Fig 4).
 

       
A maximum of 94.1% of weed samples showed positive for begomovirus in mungbean crop system while a minimum of 40.0% confirmed positive in soybean (Table 4). Moreover, 13 weed species viz., A. aspera, A.conyzoides, A. hispidium, A. viridis, B. articularis, C. dactylon, C. bonplandianum, D. arvensis, E. geniculata, I. tinctoria, P. hysterophorus, S. rhombifolia and T. purpurea were found asymptomatic alternative weed hosts of  begomovirus in  legume cropping  systems. A. aspera, A. conyzoides and C. bonplandianum were confirmed as hosts for HgYMV.While, B. articularis, I.tinctoria,S.rhombifolia and T. purpurea served as alternative hosts for MYMV. Interestingly, C. dactylon, P. hysterophorus were found hosts for both HgYMV and MYMV.


       
It is important to highlight that none of the fields were free from YMD incidence despite several vector control measures were used in the farmer’s fields. This underscores the severity of the YMD in the southern Karnataka region. In fact, our findings corroborate several previous studies that have reported the YMD severity in India (Anburaj et al., 2010; Naimuddin et al., 2014; Rajkumar 2006 and Salam et al., 2011).
 
Phylogenetic tree analysis of amplicon sequences
 
The sequences of PCR products of 14 weed species and 4 legume crop samples, excluding field bean samples (did not result in PCR product) and pole bean (due to poor quality of the sequencing data), amplified using CP gene specific primers were determined. These sequences were analyzed to construct phylogenetic tree by comparing with 17 reference sequences obtained from NCBI (Fig 5). Phylogenetic analysis showed that begomoviruses from mungbean, horsegram and soybean belong to same cluster with begomoviruses from 8 weed species and the French bean isolate sequence of horsegram yellow mosaic virus (HgYMV) segment DNA-A complete sequence. Out of four common weed hosts most abundantly occurred in six leguminous crop system, A. conyzoides and E. geniculata reside in cluster I, C. benghalensis in cluster II and A. sessilis in cluster III.
 

       
The nucleotide identity matrix of CP gene sequences of begomoviruses obtained from 14 weed species and four legume crops showed 31.0 to 100.0 per cent nucleotide identity between crops and weed host samples implying that begomovirus CP gene sequences are closely related (data not shown). A. conyzoides and E. geniculata showed 98.74% nucleotide identity with French bean isolate HYMV segment DNA-A sequence. C. benghalensis showed 87.79% identity with HYMV clone pBdBg04 segment DNA-A sequence. A. sessilis showed 97.72% identity with MYMV-Soybean [Madurai] segment DNA A sequence. Thus, our molecular characterization efforts clearly indicated the presence of begomoviruses related to horsegram yellow mosaic virus, HgYMV clone pBdBg04 segment DNA-A and MYMV- Soybean [Madurai] in most legume crops and the weeds associated with the legume crop fields in southern Karnataka regions. It is pertinent to note that A. conyziodes, C. olitorius and A. sessilis, around mungbean fields were found to harbor MYMIV and were characterized as alternative hosts of MYMIV (Marabi et al., 2017).

Legumes constitute an important component of human nutrition as major source of protein. In fact, legume grains can complement cereals as affordable source of dietary protein in human and animal food. Therefore, novel research findings or methods that contribute to effective management of global legume production should further strengthen efforts towards maintaining global food security. The major insights from the current study involving the identification of asymptomatic weed species in the legume fields that carry begomoviruses and the possibility of serving sources of begomovirus inoculum underscores the need to keep legume fields free of weed plants. In fact, our findings suggest that crop rotation with non-leguminous crops, weed management either through appropriate weedicides or manual weeding, removal of volunteer weed plants, whether symptomatic or asymptomatic are very important for effective integrated management of yellow mosaic virus diseases in legume crops.

Authors are thankful to Department of Plant Pathology, UASB, Sea6Energy Private Limited Bengaluru for research facilities and financial support. Also, thankful to AICRP Weed Science, UASB for timely identification of weeds samples.