RT-PCR-based detection and partial genome sequencing indicate high genetic diversity between CBSV and UCBSV in Kenya

Cassava is grown by small-scale farmers and consumed by over 200 million people in sub-Saharan Africa mainly as staple food. An emerging viral disease Cassava brown streak disease (CBSD) caused by Cassava brown streak virus (CBSV) and Ugandan cassava brown streak virus (UCBSV), has severely affected cassava production and resulted in poor quality tubers. The disease severely affecting root quality both for domestic use and marketing and has become a real threat to the livelihoods of the millions of poor in both eastern and southern African countries. There is also increase in the number of CBSV and UCBSV sequences available now, and a need to have primers that would detect all isolates of CBSV and UCBSV is required, particularly for field surveys. Two sets of virus-specific primers, CBSVF2 & CBSVR7 and CBSVF2 & CBSVR8, were developed to specifically detect CBSV and UCBSV respectively on six glass house maintained isolates and field collected samples from coastal Kenya. Positive samples produced expected PCR products of 345 bp and 441 bp using CBSVF2 & CBSVR7 and CBSVF2 & CBSVR8, respectively. In addition, our study further confirmed the genetic diversity of these viruses by sequencing over 41 new samples from Kenya as well as Tanzania and Zanzibar. The coat protein, partial HAM1 and 3' UTR regions of CBSV and UCBSV were amplified using CBSVF2 & CBSVR1 and CBSVF2 & CBSVR2, respectively. The PCR products of ~1.6 Kb obtained were extracted from gel, cloned and sequenced. The sequences of the UCBSV and CBSV isolates varied in length; 1678 nucleotide for UCBSV isolates and 1615 nucleotide for CBSV isolates. CBSV and UCBSV sequences were used to the construct phylogenetic trees by comparing with other sequences from GenBank. The phylogenetic tree clustered the CBSD isolates in two groups reflecting the two virus species causing CBSD. Based on the available sequences (~1600 bases), the CBSV group shared 93.7% nucleotide identities, UCBSV 93.1%, and there was ~70% identity between the two groups. However, the percentage nucleotide identities for coat protein nucleotide sequences were slightly greater than those observed for the sequences involving partial HAM1 and 3'UTR region. CBSV group shared 94.4% nucleotide identity, UCBSV 93.5% and ~74% identity between the two groups. In conclusion, this study has shown the genetic diversity between CBSV and UCBSV isolates collected from Kenya with those from Tanzania and Zanzibar.


Introduction
(SSA) after maize while third in Asia and Latin America after rice and maize (FAO, 2010).Cassava roots are the Cassava (Manihot esculenta Crantz) is a main source of dietary calories for over half of both the perennial woody shrub that produces edible rural and urban populations in Sub-Saharan Africa tuberous roots (Cock, 1985).It is the second most (SSA), most parts of Latin America, while it's also important staple food crop in sub-Saharan Africa commercially used for the production of animal feed, starch and starch-based products in Asia and some More importantly, the disease also causes parts of Latin America (Nassar & Ortiz, 2007; FAO yellow/brown, corky necrotic lesions on the tuberous 2010).
roots, which accounts for the quantitative and According to Food and Agricultural qualitative reduction in the total yield (Nichols et al., Organisation (FAO) statistical production data 1950;Mtunda et al., 2003).Previous experiments module, cassava production has increased from a conducted in Tanzania have shown that CBSD is total of 205 to 228 million tonnes during 2004 -2007 associated with loss of root weight in cassava plant, up period in the world.Africa produces more cassava to 50% in most sensitive varieties (Hillocks, 2003).Such than rest of the world combined; production loss of weight is caused by the need for removal of exceeds 118 million tonnes annually, while Asia and necrotic areas which render them unmarketable Latin America produced 78 and 34 million tonnes, (Hillocks, 2003;Mtunda et al., 2003).Similarly, Mtunda respectively (FAO, 2010).However, cassava in et al. (2003) reported that the average production of Africa is harvested on 11.9 million Ha which on cassava in Tanzania has been fluctuating over the past average is 9.8 t/ha, compared to 19.8 t/ha in Asia years.and 13.7 t/ha in South America.Thus, despite the The productivity per hectare was 10 t/ha in high production of cassava in Africa, low yields were 1996 but dropped to 7 t/ha in 2001 which is significantly obtained per hectare.This means increase in yield low. CBSD is spread through mechanical propagation of in Africa has been attributed to an increase in area infected cuttings (grafting), sap inoculation and rather than increase per unit area.This low naturally transmitted by the whitefly Bemisia tabaci production has been attributed to the many (Maruthi et al., 2005;Ntawuruhunga and Legg, 2007).constraints including biotic and a biotic factors.
Two viral species, Cassava brown streak virus (CBSV) The most important factors that severely and Ugandan cassava brown streak virus (UCBSV), limit cassava production in Africa include pests and were revealed causing the disease.Both viruses belong diseases particularly viral diseases.Cassava mosaic to the genus Ipomovirus, family Potyviridae and consist disease (CMD) and Cassava brown streak virus of (+) ssRNA genome (Monger et al., 2001; Mbanzibwa disease (CBSD) are the most important viral et al., 2009a, b;Winter et al., 2010).The viruses are diseases that severely affect cassava cultivation in often called together as Cassava brown streak viruses SSA (Thresh et al., 1994;Hillocks & Jennings, 2003; (CBSVs).Thresh & Cooter, 2005;Legg et al., 2006).These The genetic diversity of CBSV and CBSUV has diseases spread rapidly and threaten cultivation of been previously analyzed.Initially, nucleotide cassava in the growing regions in SSA.CMD is the sequences of the partial coat protein (CP) region were most economically important and widespread, determined for three isolates of CBSV from Kibaha found throughout cassava growing areas of SSA region in Tanzania (Monger et al., 2001).Then, thus it received much more research than CBSD.
complete virus genome and CP-encoding sequences of CBSD was the most poorly understood until recently the six isolates of CBSUV and CBSV from the Lake but considered more damaging than CMD in the Victoria basin in Uganda and Tanzania were also coastal zones from Kenya to Zambezi River in analyzed (Mbanzibwa et al., 2009b).Genetic diversity of Mozambique (Hillocks, 1997;Hillocks et al., 2001, CBSV andCBSUV was also determined for complete 2002;Abarshi et al., 2010).The first reports of genomes of seven isolates from Kenya, Tanzania, CBSD were from the foothills of Usambara Mozambique, Uganda and Malawi (Winter et al., 2010).

Mountains and a few years later from lower
There is need for sequencing more CBSD isolates in elevations in Tanzania (Storey, 1936(Storey, , 1939)).In farmer's field and compare their sequences to reveal the 1950, Nichols reported that the disease was diversity that exist.Also, CBSV and CBSUV genome endemic in all coastal areas of Kenya and, Tanzania contained HAM1 gene located between NIb and CP in and low elevations of Malawi (Hillocks and the 3' end of the genome (Mbanzibwa et al., 2009a). Jennings, 2003).Later years, the disease spread to The function of HAM1-like protein is not yet described coastal area of Mozambique (Hillocks et al., 2002; and little information was known about CBSV HAM1 and Hillocks and Jennings, 2003).In Kenya, Kathurima CBSUV HAM1 diversity.For these reasons, HAM1 et al. ( 2016) recently conducted disease incidence sequence data was also amplified to study CBSV and surveys in the Coast, Western and Nyanza areas CBSUV diversity.In this study, the CP, HAM1 and 3'UTR which are the major cassava-growing areas of sequences of CBSD isolates from the coastal areas of Kenya.Their survey results showed that the disease Kenya were obtained and compared with the sequences incidence in some fields in the Coast, Western and obtained from Uganda and other CBSD endemic areas.Nyanza of Kenya areas reached 33-56%.
CBSD causes distortion and foliar chlorosis on leaves as well as brown streak lesions on the stem of the cassava plant (Shaba et al., 2003).

Materials and Methods
species, virus-specific primers were designed; CBSVF2 (5' GG(A/G) CCA TAC AT(C/T) AA(A/G) TGG TT 3') & Sampling and virus strains CBSVR7 ((5' CCC TTT GCA AA(A/G) CT(A/G) AAA Forty one (41) cassava leaves samples TA(A/G) C 3') and CBSVF2 & CBSVR8 (5' CCA TT(A/G) were collected from farmer's fields in the Coastal TCT (C/T)TC CA(A/C) A(G/A/T)C TTC 3'), to specifically area of Kenya.Coastal area lies between Latitude amplify CBSV and UCBSV, respectively.These primers 30 00' 00" S, Longitude 390 30' 00" E. The area were then tested by carrying out PCR on cDNAs from six covered an area of 79,686.1 km² along the Indian glass house maintained isolates as previously described Ocean.Samples were collected from important by Abarshi et al., 2010.All the primers were used at an localities in the Coastal area, as shown in Figure 1.directions.In order to distinguish CBSV and UCBSV   Phylogenetic relationships of CBSV and UCBSV samples in separate reactions (Fig. 2).No amplification from this study and others from the genbank database obtained from the healthy plants.were established by analysing the nucleotide and amino acid sequences of CP, partial HAM1 and 3' UTR regions.
Amplification of complete CP, partial HAM1 and, 3' UTR Sequences obtained were edited to remove vector regions for CBSV and UCBSV sequence using the Bio-edit software package (Hall, Seventeen of the 21 Kenyan samples produced 1999).The sequence data were uploaded to MEGA 5.2 expected PCR products of ~1.6 kb using CBSVF2 & (Tamura et al., 2007), which were later aligned with CBSVR1 primers (Fig. 3A).Some samples that failed to sequences available in Genbank (http://www.ncbi.nlm.nih.gov/)amplify in the first attempt were amplified subsequently using ClustalW.The alignments obtained were used as by further optimisation of PCR conditions, except inputs for generating phylogenetic trees and calculating Mwatundo and Mwajambo samples, which did not pairwise distance matrix.Phylogenetic analysis was amplify at all.The PCR products were extracted from performed using maximum parsimony method with gel, cloned and sequenced.Similarly, Seven out of the 70% (1000 replications) bootstrapping scores using eight Tanzanian samples produced the expected ~1.6 MEGA5.2.Mean similarities within and between the Kb PCR fragments (Fig. 3B).Amplification from these CBSV and UCBSV groups and the genetic distances (psamples was comparatively difficult as only faint bands distances) were calculated.
were visible.Although increasing the amount of Taq polymerase from 0.1 to 0.2 µl per sample, however,

Results
resulted in brighter bands.

RT-PCR detection of CBSV and UCBSV
The primer set also amplified (~1.6 Kb) from The primer sets, CBSVF2 & CBSVR7 and the six glass house maintained isolates (Nampula, CBSVF2 & CBSVR8, distinguished CBSV and UCBSV Naliendele, Kabanyolo, Zanzibar, Mwalumba, and on six glass house maintained isolates.CBSVF2 & Kibaha) and from field-collected samples from Zanzibar CBSVR7 amplified the expected PCR fragment of (Fig. 3C).16 samples out of the 18 were amplified using ~345 bp size and detected CBSV in Mwalumba, these primers and the remaining samples were Zanzibar, Naliendele and Nampula samples while amplified subsequently.Multiple bands were observed CBSVF2 & CBSVR8 primer amplified ~441 bp PCR in some samples.fragment from UCBSV in Kibaha and Kabanyoro  1 shows the names and acronyms of ipomoviruses and genbank accession numbers of their sequences used in multiple alignments.Two clones from each sample were sequenced to ensure sequence identity and reliability.Sequences obtained were highly reproducible and the consensus sequences were selected for phylogenetic analysis (Table 2).The sequence length of 1678 nucleotides (nt) were obtained from Kenya CBSD samples, while those of Zanzibar, presence of an additional aa at position 114 in HAM1 region, which was verified in three independent clones of Kilifi 18-2:08.This was further confirmed in two other clones (Kilifi 20-1 and Kilifi 20-3).Similarly, one aa was missing at position 200 in all six clones of Denyenye (Denyenye 1-2).Sequencing of additional clones from the same isolates (Denyenye 1-1, Deny 1 and Deny 2) further confirmed the absence of one aa.
Nampula, Naliendele and Tanzania CBSD samples had a deletion of 63 bases therefore they were 1615 nt long.Sequences from Tanzanian samples had also 63 bases deletion except Namarebe Phylogenetic analyses sample which had 1678, similar to Kenyan samples.The deletion was found in the 3' UTR region similar to CBSV grouping based on CP and partial HAM1 and 3' that observed for the previously published sequences UTR sequences of Tan70 (accession number: FN434437), Tanzanian Based on CP, partial HAM1 and 3' UTR sequence (GQ329864) and Mo_83 (FN434436) in the GenBank.
comparisons with the reference viruses from GenBank, The translation of the nt sequences revealed 492 aa the phylogenetic analysis clustered all CBSD isolates for Kenyan samples and 503 aa for Zanzibar CBSD into two major groups; CBSV and UCBSV (Fig. 5).The isolate.Sequence comparisons revealed some unique CBSV group consisted of isolates originating from features such as the Tanzania and Mozambique, and the UCBSV group from Kenya, Uganda, Malawi and Tanzania.Therefore, the Table 1: Names and Acronyms of Ipomoviruses CBSV group comprises mainly of Naliendele, Zanzibar and genbank accession numbers of the sequences and Nampula isolate along with other reference viruses used in multiple alignments Tan_70, Tanz and Mo_83 from genbank.Tanzanian samples from Gallu, Kigala, Masahunga, Namabaza and Guta were also in CBSV group with the exception of Namarebe which grouped into UCBSV.The UCBSV group consisted mainly of Kabanyolo, Kibaha and all Kenyan field-collected isolates in addition to the reference sequences from the gene bank Ug, Ug_23, Ke_54, Ke_125, MLB3, Ma_42 and Ma_43.SPMMV, CVYV and SqVYV clustered separately to form an out-group.Based on the available sequences (~1600 bases), the CBSV group shared 93.7% nucleotide identities, UCBSV 93.1%, and there was ~70% identity between the two groups.
The phylogenetic tree deduced based on the alignment of only CP nucleotide was similar to the grouping observed with CP, partial HAM1 and 3' UTR sequence.The percentage identities for the sequences were slightly greater than those observed for the sequences involving partial HAM1 and 3'UTR region (Table 3).CBSV group shared 94.4% nucleotide sets, CBSVF2 & CBSVR7 and CBSVF2 & CBSVR8 were identity, UCBSV 93.5% and ~74% identity between found to be highly specific and reliable for the the two groups.
amplification of CBSV and UCBSV, respectively.The RT-PCR assays for the detection of CBSV and UCBSV Table 3: The intra-group and inter-group individually was also highly successful.Our results are in nucleotide identities and amino acid similarities for agreement with Mbanzibwa et al. (2011) that also CBSV and UCBSV sequences developed a RT-PCR for the CBSV and UCBSV, and found that 24% of the 114 samples examined from Uganda and Tanzania were infected with CBSV and UCBSV.Therefore, these primers will be ideal for the diagnosis of CBSV and UCBSV in field surveys.
The genetic diversity of CBSV and UCBSV based on CP sequences has been previously analyzed.Initially, nucleotide sequences of the partial CP region were determined for three isolates of CBSV from Kibaha region in Tanzania (Monger et al., 2001).Based on nucleotide and amino acid sequence comparisons, the three isolates varied from one another by up to 8% and 6%, respectively.Complete virus genome and CPencoding sequences of the six isolates of UCBSV and CBSV from the Lake Victoria basin in Uganda and Tanzania were also analyzed (Mbanzibwa et al., 2009b).These isolates showed 90.7-99.5 and 93.7-99.5% identities at the nucleotide and amino acid levels, respectively, for isolates in UCBSV and CBSV groups.Genetic diversity of CBSV and UCBSV was also determined for complete genomes of seven isolates from Kenya, Tanzania, Mozambique, Uganda and Malawi (Winter et al., 2010).
This was further confirmed in our study by comparing the sequences of over 41 new samples especially from Kenya for which the virus diversity was not known, and also from Tanzania, Uganda and Mozambique.CP gene and the 3' UTR are the most common markers used for studying the genetic variability of potyviruses (Shukla et al., 1994).Thus, present study assessed the genetic variability of CP and partial HAM1 and 3' UTR regions of CBSV and UCBSV.Fragments of ~1.6 kb of both viruses were amplified using degenerate primer set (CBSVF2 & CBSVR1), though some multiple bands were observed in some Figure 5: The most parsimonious tree illustrating samples.The appearance of multiple bands in some the grouping of CBSV and UCBSV based on samples may have resulted from the presence of complete Coat Protein, partial HAM1 and 3' UTR inhibitors in the cDNA samples that could block the DNA sequences.
template during PCR amplification.This may increase the amount of partially amplified products, which can Discussion then serve as primers in the next cycles and thereby Two distinct species (CBSV and UCBSV) are increasing the number of chimeric amplicons in the PCR shown to cause CBSD in east African countries product (Kalle et al., 2014).(Mbazibwa et al., 2009b;Monger et al., 2010; The RT-PCR products from tested samples were Winter et al., 2010).Recently research efforts have sequenced and used for the construction of been focused on CBSD in which there was increase phylogenetic trees by comparing with reference in the number of CBSV and UCBSV sequences sequences from GenBank.The sequences of the available now in genbank, and therefore there was UCBSV and CBSV isolates varied in length; 1678 nt for a need to have primers that would distinguish CBSV UCBSV isolates and 1615 nt for CBSV isolates.A from UCBSV, particularly for field surveys.Primer deletion of about 63 nt was found in the 3' UTR region of CBSV isolates which was also found in previously developed RT-PCR protocol for CBSV and UCSBV sequenced isolates (Monger et al., 2010;Winter et detection separately.In addition, the study shows the al., 2010).Such deletions appear to be having no genetic diversity between CBSV and UCBSV isolates deleterious effect on the virus because all three NRI collected from Kenya and Uganda with those from CBSV isolates with deletions expressed typical Tanzania and Zanzibar.Isolates of UCBSV collected CBSD symptoms on cassava plants.The 3'UTRs are from Uganda from a CBSD epidemic region are highly variable in terms of length and sequence homology similar to those from Kenya, Malawi and Tanzania, and as well as its secondary structure (Riechmann et al., also to the previously reported sequences from the 1992).Rodriquez-Cerezo et al. (1991) reported a same region in the UCBSV (from Kenya, Uganda and viral determinant of symptom severity mapped in Malawi).
Similarly, isolates from Tanzania and Zanzibar the 3' UTR of the TVMV (Tobacco vein mottling were similar with CBSV isolates (Coastal areas of virus) genome and suggested its involvement in Tanzania and Mozambique).These findings can be symptom induction of disease by RNA viruses.
powerful and cost-effective tools for typing and sub-Based on these results, some degree of typing of virus strains in different epidemiological geographical pattern was associated with the studies.In addition, our results on genetic study can phylogenetic grouping where all strains of CBSV have huge implications for cassava breeding programs were from Mozambique and Tanzania and all strains most especially in cassava-growing African countries.of UCBSV were from Kenya, Uganda, Malawi and The Cassava breeders can breed varieties of cassava Tanzania, which is consistent with results of lines that are resistant to these viruses.previous studies on genetic diversity of partial CP (Mbanzibwa et al., 2009b) and complete genome

∞
annealing temperature of 52 C. In addition, another set of eight CBSD isolates were collected from Tanzania and a set of six CBSD Amplification of completeCP, 3'  isolates included  in this study are Nampula, Naliendele, Kabanyolo, Zanzibar, Mwalumba, and UTR regions for CBSV and UCBSV and cloning/sequencing Kibaha.The symptoms found on the leaves were of PCR products diverse showing severe to yellowing, mosaic, veinal Two sets of primers (CBSVF2 & CBSVR1 and chlorosis, down curling, leaf distortion and severe CBSVF2 & CBSVR2) were designed in conserved stunting.Samples were stored in plastic bags atregions.CBSVF2 was within the HAM1 region abouto 20 C. 250 bp sequences upstream of the CP while CBSVR1 (5' AA(C/T) A(A/G)A AGG ATA TGG AGA AAG 3') and Nucleic acid extraction and RT-PCR detection of CBSVR2 (5' GC(A/T) A(A/T/C)C (C/T)AA AAC TCC ACC CBSV and UCBSV 3') were designed in the conserved motifs and within the 3' UTR region.These primers were used to amplify The modified cetyl trimethyl ammonium partial HAM1 coding region (3' end), CP region and part bromide (CTAB) method (Lodhi et al., 1994; Maruthi of the virus 3' UTR of the 21 Kenyan samples, eight et al., 2002) was used for the extraction of RNA fromTanzanian samples, six Zanzibar samples and six glass the cassava leaves.RNA was subsequently reverse house maintained isolates.Purified PCR products were transcribed to cDNA using ImProm-IITM Reverse R cloned using pGEM -T Easy vector system (Promega, Transcriptase kit (Promega, UK) before being UK) following the manufacturer's instructions and subjected for PCR amplification as described by sequenced at the Geneservice Ltd., Cambridge, UK.For method modified byAbarshi et al. (2010).OligodT each sample two clones were sequenced in both primer was used in the cDNA preparation.

Figure 1 .
Figure 1.Map of Kenya showing distribution of CBSD isolates used in this study.Black circles showed CBSD sample collection localities.

Figure 3 :
Figure 3: Gel electrophoresis of PCR amplified CP and partial HAM1 and 3' UTR fragments with CBSVF2 and CBSVR1 primers.A. Kenyan samples (Lanes 1-21).B. Tanzanian CBSD samples (Lanes 1-8).C. Lanes 1-12 represent the NRI CBSD isolates (two samples for each isolate), and lanes 13-18 represent Zanzibar field collected isolates.M denotes 100 bp molecular weight marker (Roche diagnostics GmbH, Germany) and 1 kb molecular weight marker (New England Biolabs).The sizes of some bands are given on the right side of the gel.'-' denotes no-DNA sterile distilled water negative control, and '+' denotes a known positive control.

Table 2 :
List of CBSD isolates sequenced with their Sequence diversity accession numbers The amplified CP and partial HAM1 and 3' UTR sequences of CBSV and UCBSV were cloned and sequenced.Table