SEARCH

SEARCH BY CITATION

Keywords:

  • Cassava brown streak virus;
  • Ipomovirus;
  • RT-PCR detection

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

An RT-PCR based detection method for Cassava brown streak virus (CBSV)-infected cassava has been developed. The RT-PCR detection method described includes RNA extraction methods for cassava leaves, a distinct primer set for the virus and RT-PCR conditions. The primers were designed to the virus coat protein gene and generate a virus-specific product of 231 bp from infected cassava. The test can detect the virus in the new growth of cassava sticks before any disease symptoms are visible. This test was used successfully with infected cassava from both Tanzania and Mozambique. Three isolates from Tanzania were found to exhibit different symptoms on the secondary host plants Nicotiana benthamiana and N. tabacum SR1. They have nucleotide sequence variation within the coat protein region of up to 8% and amino acid differences of up to 6%.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Cassava originates from Central and South America but the plant's drought-hardy properties mean it is grown in greater quantities in Sub-Saharan Africa (SSA). Of the total 168 million metric tonnes (MMt) of cassava grown world-wide in 1999, 92 MMt were in SSA (FAO, 1999). Compared with other crops grown in this region [35 MMt of yam, 27 MMt of maize, 14 MMt of wheat and 9 MMt of sweet potato (FAO, 1999)], the importance of cassava is clear.

Cassava is propagated through cuttings and hence virus infection can accumulate. Eight diseases caused by viruses have been observed from cassava in SSA (Thresh et al., 1994). After the African cassava mosaic diseases caused by geminiviruses the next economically important virus disease of cassava in East Africa is cassava brown streak disease (CBSD) (Legg & Raya, 1998). CBSD is caused by Cassava brown streak virus (Monger et al. 2001). This disease is found in Kenya, Tanzania, Uganda, Malawi and Mozambique (Nichols, 1950; Bock, 1994; Legg & Raya, 1998). The symptoms include yellowing of leaves, brown lesions on the stems, reduction in root size and necrotic lesions within the roots. These lesions result in spoiling of the crop after harvesting (Hillocks et al., 1996).

The 3′ terminal region of the genome of CBSV has been sequenced including the coat protein ORF (Monger et al., 2001). Sequence data suggests that CBSV may be a member of the Ipomovirus genus of the Potyviridae family. This genus has two other sequenced members, Sweet potato mild mottle virus (SPMMV) (Colinet et al., 1996) and Cucumber vein yellowing virus (CVYV) (Lecoq et al., 2000). CBSD was previously thought to be caused by two viruses (Lennon et al., 1985) but most of the evidence for two viruses can now be accounted for by this unusual ipomovirus. One observation that has not yet been addressed is the symptom differences that have been reported when the virus infects a secondary host plant (Bock, 1994); two variants of the virus were isolated and subsequently maintained on Nicotiana debneyi, where they induced readily distinguishable symptoms. The variants were both filamentous particles of 650 nm in length. The results reported in this paper offer an explanation for such symptom variation.

Diagnosis of CBSD in the field through symptoms is difficult, with immature leaves of infected cassava often appearing symptomless, and symptoms of the disease varying greatly with the variety of cassava and environmental conditions (Lister, 1959; Hillocks et al., 1996). An antiserum has been raised to purified virus particles. This antiserum, although reproducibly detecting the virus in N. benthamiana, gave erratic and insensitive results for CBSV-infected cassava (Lennon et al., 1985). A sensitive detection method would be a valuable tool for cassava breeding programmes in Africa to detect material that is infected latently with CBSV.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Source and maintenance of plant material

All infected cassava plants were maintained at NRI Chatham. Three infected cultivars (cv. Kibaha, cv. Vumbi and cv. Mukukumkuku) were collected by Mrs Mtunda in 1996 from the Kibaha region of Tanzania. Infected cassava sticks were also collected in 1999 by Dr R. J. Hillocks and Professor J. M. Thresh from Tanzania and the Zambesia province of northern Mozambique.

Nicotiana benthamiana and N. tabacum SR1 were used as secondary host plants for CBSV. The plants were inoculated mechanically. Infected cassava leaves were macerated in water containing carborundum powder and the extract rubbed onto the leaves. Inoculated plants were maintained at 24°C with 16 h of light in a 24-h period.

Extraction of RNA from cassava leaves

A number of methods were evaluated for extracting RNA from cassava leaves. These included the SDS/phenol method (Chattopodhay et al., 1993), the CTAB methods (Verwoerd et al., 1989; Lodhi et al., 1994) and the guanidinium thiocyanate method (Chomczynski & Sacchi, 1987). The following commercial RNA extraction kits were also evaluated: the RNeasy Plant Mini kit (QIAGEN Ltd, Crawley, UK); RNA isolatorTM Total RNA isolation reagent (Sigma-Genosys, Pempisford, UK); and TrizolTM reagent (Gibco BRL, Paisley, UK). All appropriate precautions against RNA degradation were taken.

RT-PCR

cDNA was generated using the reverse transcriptase polymerase chain reaction (First-strand RT-PCR kit, Stratagene Ltd, Cambridge, UK) using the manufacturer's instructions, with 7 µg of total RNA in a 50-µL volume and using the primer oligo d(T).

The 20-µL PCR reactions contained 1·8 µL of 11× buffer [500 mm Tris pH 8·8, 125 mm (NH4)2 SO4, 50 mm MgCl2, 75 mm 2-mercaptoethanol, 50 µm EDTA pH 8·0, 11 mm each dNTP, 1·26 mg mL−1 BSA], 80 ng of each primer, 1 µL of cDNA and 0·5 U Taq polymerase (Promega UK Ltd, Southampton, UK). The PCR program was 94°C for 1 min, 50°C for 1 min and 72°C for 1 min, repeated for 30 cycles then held at 72°C for 10 min. Products were separated by agarose gel electrophoresis containing ethidium bromide and viewed under UV light.

A range of PCR primer sets were designed to the published sequence of CBSV (GenBank accession No. AY007597) and a series of PCR parameters tried in order to optimize the RT-PCR test. The cDNA generated with the RT-PCR kit from Stratagene was evaluated undiluted, as well as serially diluted down to 1 : 30 with water. A range of PCR machines, PCR buffers and magnesium chloride concentrations were evaluated. Uninfected cassava leaves were used as a control for the RT-PCR test (provided by Dr R. M. Cooper, University of Bath, UK).

All cDNA was checked using a control primer set designed to the cassava ribulose 1-5-bisphosphate carboxylase small subunit precursor (rbc1 5′-CTA CTA TGG TGG CTC CGT TC-3′ and rbc2 5′-CCG TTC AGT CGG AGA AAC TC-3′) (Mak & Ho, 1995). The purpose of these primers was to check that the cDNA had been made and prevent false negative results. The primer set generates a 619-bp product with cassava cDNA (contaminating genomic DNA can be distinguished because it produces a larger PCR product of approximately 800 bp).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

RNA extraction

RNA extracted from cassava leaves with the SDS/phenol method (Chattopodhay et al., 1993), the CTAB method (Verwoerd et al., 1989) and the guanidinium thiocyanate method (Chomczynski & Sacchi, 1987) were not suitable for RT-PCR amplification due to either a high degree of RNA degradation or an insoluble RNA pellet. In contrast, an adaptation of Lodhi et al. (1994), a CTAB method developed by Dr R. A. Mumford whilst at NRI Greenwich UK, was found to be very successful at extracting high-quality RNA from cassava leaves.

Cassava leaf (0·3 g) was ground to a powder with liquid nitrogen in a sterile pestle and mortar. Three millilitres of grinding buffer was added (2% CTAB, 2% polyvinylpolypyrrolidone-40, 100 mm Tris pH 8·0, 20 mm EDTA, 1·4 m NaCl and 20 mm DTT added fresh). The suspension (800 µL) was transferred to a microfuge tube and incubated at 65°C for 10 min After incubation 600 µL of chloroform : IAA (24 : 1) was added and mixed by inverting the tube. The phases were separated by microfuge at maximum speed for 10 min. The upper aqueous phase was removed and the chloroform extraction repeated. Ethanol precipitation of the nucleic acid was performed with 0·5 volumes of 5 m NaCl and 2 volumes of ice cold ethanol at − 20°C for 30 min. The nucleic acid was collected by microfuge at half maximum speed for 10 min and resuspended in 0·5–1·0 mL of 2 m LiCl. The nucleic acid was left in the LiCl overnight at 4°C. The RNA was pelleted in a microfuge at maximum speed for 30 min at 4°C. The LiCl was removed, the pellet washed with 70% ethanol, dried and resuspended in 50–100 µL of RNase-free water.

From the commercial RNA extraction kits examined (Sigma-Genosys, Gibco BRL and QIAGEN) the RNeasy Plant Mini kit from QIAGEN gave the best results, with the following modifications. Two buffers are provided at step 2 of the protocol; the RLC buffer proved better than the RLT buffer. No incubation at elevated temperatures was performed at step 2. All centrifuge steps were performed at maximum speed in a microfuge and these were performed for longer than 15 s if the material was found not to have passed through the column.

The CTAB and QIAGEN RNeasy methods produced RNA that was successful at generating products with RT-PCR. The CTAB method described above was the method used for subsequent experiments.

Optimization of the RT-PCR technique

Several sets of primers were designed from the published sequence (GenBank accession No. AY007597) and are illustrated in Table 1. One primer set proved most consistent in generating a strong product band, when analysed on ethidium bromide stained agarose gels: CBSV 10 the forward primer (5′-ATC AGA ATA GTG TGA CTG CTG G-3′) and CBSV 11 the reverse primer (5′-CCA CAT TAT TAT CGT CAC CAG G-3′). These primers were designed using the genejockey ii computer program (BioSoft, Cambridge, UK) and together generate a 231-bp product from within the coding region of the virus coat protein (Fig. 1a).

Table 1.  The oligonucleotide primer sets used in this study. The primers were designed to the published sequence (Monger et al. 2001; accession No. AY007597). The primer sets CBSV 6 and 7, CBSV 5 and 2 and CBSV 10 and 11 were used as part of an RT-PCR test for the virus. The primer set CBSV 9 and 11 were used to generate products from different virus isolates for sequencing
 Primer  setsPrimer sequence 5′ [RIGHTWARDS ARROW] 3′Primer position on reference sequenceLength of expected fragment (bp)
 CBSV 6GTATACAAGCATTGAAAATAA749[RIGHTWARDS ARROW]769 
 CBSV 7CTCCTGAATATATCTTGGC1062[RIGHTWARDS ARROW]1044313
 CBSV 5CGGAGTTGAAGTTGAGAATTG410[RIGHTWARDS ARROW]430 
 CBSV 2GATACATTAAGATTGCTTGG693[RIGHTWARDS ARROW]674283
 CBSV 10ATCAGAATAGTGTGACTGCTGG691[RIGHTWARDS ARROW]712 
 CBSV 11CCACATTATTATCGTCACCAGG912[RIGHTWARDS ARROW]891231
 CBSV 9ATGCTGGGGTACAGACAAG1[RIGHTWARDS ARROW]19 
 CBSV 11CCACATTATTATCGTCACCAGG912[RIGHTWARDS ARROW]891912
image

Figure 1. (a) Agarose gel of the RT-PCR test for CBSV, with infected (I) and uninfected (U) cassava, and three primer sets: CBSV 6&7 (313-bp product), CBSV 5&2 (283-bp product) and CBSV 10&11 (231-bp product) (see Table 1). RNA was extracted from the cassava using the CTAB method described in the results. cDNA was generated with the Oligo dT primer in the RT-PCR kit (Stratagene) using 7 µg of RNA. 1 µL of cDNA was used in each 20-µL PCR reaction described in the methods. (b) Agarose gel of RT-PCR test of CBSV-infected cassava with the primer set CBSV 10 and 11 which generates a 231-bp product. RNA was extracted from the cassava using the CTAB method described in the results. cDNA was generated with the Oligo dT primer in the RT-PCR kit (Stratagene) according to manufacturers instructions using 7 µg of total RNA in a 50-µL volume. The cDNA was added undiluted (1 : 1) and serial diluted with sterile distilled water to a 1 in 30 concentration. 1 µL of cDNA was used in each 20 µL PCR reaction as described in the methods section.

Download figure to PowerPoint

The 10× Taq polymerase buffer (Promega) was tested with the primer set CBSV 10 and CBSV 11 and a range of magnesium chloride concentrations from 0·0 to 4·5 mm. The RT-PCR test was found to be successful within the range 0·5–3·0 mm, the optimum concentration was 2·0 mm. However, if no magnesium chloride was added no product was produced (results not shown).

A variety of PCR machines was used: the MJ Research MiniCyclerTM, the Bio-Rad Gene CyclerTM and Hybaid Omn-E all proved suitable; the MJ Research MiniCyclerTM was routinely used.

A serial dilution of the cDNA generated with the first strand RT-PCR kit (Stratagene) was performed (Fig. 1b). The greatest concentration of product was produced with the undiluted cDNA. However, a product band was still visible with a 1 in 30 dilution of the cDNA.

Sensitivity of the technique

Sensitivity of the RT-PCR was assessed with cassava leaves from an infected cassava cutting from cv. Vumbi that was not yet showing symptoms. The upper most, second and third leaves were selected for testing. RNA was extracted using the CTAB method (R. A. Mumford, adapted from Lodhi et al., 1994) described in this paper. cDNA was made using the RT-PCR kit (Stratagene) with 7 µg of total RNA and the PCR was performed with the primers CBSV 10 and CBSV 11 as described. All three leaves generated a strong PCR product with the diagnostic test (Fig. 2).

image

Figure 2. RT-PCR of a cassava cutting that was known to be infected with CBSV but was not showing any symptoms. The uppermost (1), second (2) and third (3) leaves were tested. RNA was extracted from the cassava using the CTAB method described in the results. cDNA was generated with the Oligo dT primer in the RT-PCR kit (Stratagene) using 7 µg of RNA. 1 µL of cDNA was used in each 20-µL PCR reaction as described in the methods section. A negative control which includes all components of the PCR reaction except the cDNA is included (4).

Download figure to PowerPoint

Verification of the RT-PCR test

In 1998 cassava sticks were collected from Tanzanian field trials. Some of these sticks were chosen because they showed atypical symptoms not previously associated with CBSV and others which were without symptoms were chosen to see if the material was virus-free or latently infected. Of the 71 original samples received from Tanzania, 37 grew in the glasshouse at NRI Chatham. The RT-PCR test was performed at NRI Chatham and 12 of the cassava sticks were tested blindly at Bristol for verification of the results. The two laboratories agreed that eight of the Tanzanian samples were infected with the virus (results not shown).

Symptom variation

The three infected cultivars of cassava from the Kibaha region of Tanzania (cv. Kibaha, cv. Vumbi and cv. Mukukumkuku) were used to infect the secondary host plant N. benthamiana. There were differences in symptoms on the inoculated plants which related to which cultivar of cassava was used. More than 60 N. benthamiana plants were treated for each cultivar over a 1-year period. No sign of infection occurred on plants inoculated with sap from cv. Kibaha, and the CBSV in these was designated Type A (Fig. 3a). Sap from cv. Vumbi plants gave rise to systemically infected N. benthamiana plants displaying leaf-curling and mosaic symptoms; the CBSV in these plants were designated Type B (Figs 3a and b). Inoculation of N. benthamiana with sap from cv. Mukukumkuku produced plants that wilted due to thinning of the petioles. These plants died rapidly and the isolates of CBSV from these plants were designated Type C (Figs 3a and c). Clear symptom differences were also present on individual inoculated leaves, with sap from cv. Kibaha (Type A) showing no sign of infection, cv. Vumbi (Type B) producing chlorotic regions and cv. Mukukumkuku (Type C) producing clear local lesions (Fig. 3d). The three cultivars were also used to inoculate the tobacco variety N. tabacum SR1. No visual sign of infection was seen with either cvs Kibaha or Vumbi but cv. Mukukumkuku produced local lesions.

image

Figure 3. (a) Nicotiana benthamiana plants 12 days postinoculation with extracts from the three CBSV-infected cassava cultivars Kibaha (designated Type A), Vumbi (designated Type B) and Mukukumkuku (designated Type C). (b) A Nicotiana benthamiana plant 12 days postinoculation with an extraction from the CBSV-infected cassava cultivar Vumbi. The systemic leaves are curled and show strong mosaic symptoms. (c) A Nicotiana benthamiana plant 12 days postinoculation with an extract from the CBSV-infected cassava cultivar Mukukumkuku. Mosaic symptoms are seen in the newly emerged leaves but petiole thinning has resulted in the leaves wilting and subsequent plant death. (d) Inoculated leaves of the secondary host plant Nicotiana benthamiana, 9 days postinoculation with extracts from the three CBSV-infected cassava cultivars Kibaha (Type A), Vumbi (Type B) and Mukukumkuku (Type C).

Download figure to PowerPoint

Sequence variation

The specific primers CBSV 9 and CBSV 11 (Table 1), were used with the cassava cultivars Kibaha, Vumbi and Mukukumkuku, designated Types A, B and C respectively, to generate PCR products 912 bp in length at the 3′ end of the virus genome. Two independently generated products were sequenced for each cassava cultivar. The PCR products were found to vary from one another by as much as 8% at the nucleotide level and 6% at the amino acid level. Differences were not evenly distributed along the length of the coat protein but were localized in regions. The nucleotide and corresponding amino acid sequences of the most diverse region was found at the 5′ end of the RT-PCR products (nucleotides 63–188 of the published sequence: GenBank accession No. AY007597) (Figs 4a and b). The full sequences for each can be found on the database: Type A (from cv. Kibaha) GenBank accession No. AY008442, Type B (from cv. Vumbi) GenBank accession No. AY008441and Type C (from cv. Mukukumkuku) GenBank accession No. AY008440. The original isolate of CBSV that was sequenced (Monger et al. 2001: GenBank accession No. AY007597) was found to share more than 98% homology at the RNA level with Type C. A summary of the identities between the three isolates generated, with the specific primers CBSV 9 and CBSV 11, is given in Table 2.

image

Figure 4. (a) Partial nucleotide sequence from 3 Tanzanian isolates of cassava brown streak virus, A (Kibaha), B (Vumbi) and C (Mukukumkuku). The region represents the most diverse region between the three sequences (nucleotides 63–188 of the published sequence: GenBank accession No. AY007597). The boxed nucleotides indicate where differences occur between at least two of the three sequences. (b) The amino acid sequences corresponding to the nucleotide sequences in (a). The boxed amino acids indicate where differences occur between at least two of the three sequences.

Download figure to PowerPoint

Table 2.  Sequence comparison of CBSV isolates A, B and C from nucleotide 20–890 and amino acid 7–296 of CBSV (GenBank accession No. AY007597). The comparison was made using the Clustal program of genejockey ii (BioSoft, Cambridge, UK)
 Amino Acid %
RNA %Type BType CType BType C
Type A92·391·994·693·6
Type B 93·4 94·9

The RT-PCR test with the primers CBSV 10 and CBSV 11 was used on the three infected cultivars from Tanzania, Kibaha, Vumbi and Mukukumkuku, along with two cultivars from Mozambique, Mulaleia and Fernando Po, which were showing symptoms of CBSD. All gave a positive result with the test (results not shown).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

From the work of Lister (1959) it is known that within the cassava plant the virus is at lowest concentration within developing leaves. The developing leaves of infected cassava do not show symptoms and some cassava varieties are reported not to show leaf symptoms at all (Hillocks et al., 1996). A diagnostic test for this virus must be able to detect the virus in the new young symptom-free leaves of the cassava plant to be an effective tool. An RT-PCR test has been developed that can detect the virus even in the young leaves of cassava that are not yet showing symptoms of the virus. This is the first effective test available for this virus. An RNA extraction method that gives high-quality undegraded RNA from cassava leaves is described. The PCR test includes a primer set which gave consistent results. This primer set proved reliable under a range of conditions and variables including a range of PCR machines, PCR buffers, magnesium concentration and cDNA concentrations.

The 11× PCR buffer described in the protocol has a magnesium chloride concentration of 4·5 mm. However, the Promega 10× PCR buffer was found not to generate a product at this concentration and was only successful within the range 0·5–3·0 mm. The PCR reaction requires different magnesium chloride concentrations depending on the components of the buffer.

The movement of cassava between countries has been restricted to prevent the spread of disease. This is particularly true in the movement of cassava from East to West Africa because of the lack of a diagnostic test for CBSD. A reliable diagnostic test enables cassava varieties to be screened by cassava breeding programs. Although very few currently have PCR diagnostic facilities, access to such facilities will become more common in the near future.

This paper has identified three isolates of CBSV based on sequence differences between their coat proteins and symptom variations. These isolates came from three different cultivars of cassava, Kibaha, Vumbi and Mukukumkuku, grown in the Kibaha region of Tanzania. These isolates differ from one another in the region sequenced by up to 8% at the nucleotide level and 6% at the amino acid level, but more importantly they show quite distinct symptoms on the secondary host plant N. benthamiana. Nicotiana benthamiana has been reported as a systemic host for this virus and N. tabacum SR1 has been reported as a local lesion host (Lister, 1959). The Kibaha isolate, designated Type A, did not show symptoms on any N. benthamiana plants despite numerous attempts at manual inoculation. Only the Type C isolate from the cultivar Mukukumkuku produced clear local lesions with both tobacco varieties. Although such indicator plants can give some idea of the disease status of the cassava, a negative result cannot be taken as a virus-free status. Symptom differences on a secondary host plant were previously reported (Bock, 1994), which were attributed either to more than one virus involved with the disease or to different isolates of the same virus. Symptom and sequence results reported in this paper support the latter theory. The isolate of CBSV previously sequenced from Tanzania (Monger et al. 2001) had a high degree of identity with Type C and exhibited the same symptoms on N. benthamiana. Hence this isolate is considered to be a Type C isolate. Only a small portion of the genomes of the isolates have been sequenced and might not be representative of the genomes as a whole. It is unlikely that the sequence differences found between the isolates are causally connected to the symptom differences. Different cassava cultivars are reported to have different symptoms when infected with CBSD (Hillocks et al., 1996). These differences may not be wholly the result of the cassava cultivar but may also depend on the isolate of CBSV with which it is infected.

The alignment of the partial sequences of CBSV isolates from Tanzania found most of the differences in nucleotides to be at the 5′ end of the sequence. This region is either the 3′ end of the NIb gene or the 5′ end of the coat protein, since the cleavage site between these coding regions is not yet known with ipomoviruses. The primers for the RT-PCR test are designed within the coat protein close to the 3′ end of the coding region. The sequences of the CBSV isolates show this to be a highly conserved region of the virus.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The authors would like to thank Mrs Mtunda, Dr Frances Kimmins, Dr Rory Hillocks and Professor Mike Thresh for providing CBSV-infected cassava sticks. This publication is an output from a research project funded by the Department for International Development of the United Kingdom. However, the Department for International Development can accept no responsibility for any information provided or views expressed [DFID project number R6617, Crop Protection Programme].

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Bock KR, 1994. Studies on cassava brown streak virus disease in Kenya. Tropical Science 34, 13445.
  • Chattopodhay N, Kher R, Godbole M, 1993. Inexpensive SDS/phenol method for RNA extraction from tissues. Biotechniques 15, 246.
  • Chomczynski P & Sacchi N, 1987. Single-step method of RNA isolation by acid guanidinium thiocyanate–phenol–chloroform extraction. Analytical Biochemistry Journal 162, 1569.
  • Colinet D, Kummert J, Lepoivre P, 1996. Molecular evidence that the whitefly-transmitted sweetpotato mild mottle virus belongs to a distinct genus of the Potyviridae. Archives of Virology 141, 12535.
  • FAO, 1999. FAOSTAT Database[http://apps.fao.org;. Food and Agriculture Organisation of the United Nations.
  • Hillocks RJ, Raya M, Thresh JM, 1996. The association between root necrosis and above-ground symptoms of brown streak virus infection of cassava in southern Tanzania. International Journal of Pest Management 42, 2859.
  • Lecoq H, Desbiez C, Delecolle B, Cohen S, Mansour A, 2000. Cytological and molecular evidence that the whitefly-transmitted cucumber vein yellowing virus is a tentative member of the family Potyviridae. Journal of General Virology 81, 228993.
  • Legg JP & Raya MD, 1998. Survey of cassava virus diseases in Tanzania. International Journal of Pest Management 44, 1723.
  • Lennon AM, Aiton MM, Harrison BD, 1985. Cassava viruses from Africa. In: Annual Report of the Scottish Crop Research Institute. Dundee, UK: SCRI, 168.
  • Lister RM, 1959. Mechanical transmission of cassava brown streak virus. Nature 183, 15889.
  • Lodhi MA, Ye GN, Weeden NF, Reisch BI, 1994. A simple and efficient method for DNA extraction from grapevine cultivars and Vitis species. Plant Molecular Biology Reporter 12, 613.
  • Mak YM & Ho KK, 1995. Sequence of cassava ribulose-1,5-bisphosphate carboxylase small subunit precursor cDNA. DNA Sequence 5, 22932.
  • Monger WA, Seal S, Isaac AM, Foster GD, 2001. Molecular characterization of Cassava brown streak virus coat protein. Plant Pathology 50, 52734.DOI: 10.1046/j.1365-3059.2001.00589.x
  • Nichols RFW, 1950. The brown streak disease of cassava: distribution, climatic effects and diagnostic symptoms. East African Agricultural Journal 15, 15460.
  • Thresh JM, Fargette D, Otim-Nape GW, 1994. The viruses and virus diseases of cassava in Africa. African Crop Science Journal 2, 45978.
  • Verwoerd TC, Dekker BMM, Hoekema A, 1989. A small scale procedure for the rapid isolation of plant RNAs. Nucleic Acids Research 17, 2362.