Molecular Characterization of Tomato-Infecting Begomoviruses in Central America and Development of DNA-Based Detection Methods
M.K. Nakhla, and A. Sorenson, University of Wisconsin-Madison; L. Mejía, Universidad de San Carlos, Guatemala; P. Ramírez, and J.P. Karkashian, Universidad de Costa Rica, San Jose; and D.P. Maxwell, University of Wisconsin-Madison.
Keywords: Tomato severe leaf curl virus, Tomato golden mottle virus, Tomato mild mottle virus, Tomato yellow mottle virus, Pepper golden mosaic virus, Tomato mosaic Havana virus, Tomato leaf curl Sinaloa virus, Pepper huasteco yellow vein virus

Begomoviruses cause great losses of tomato crops in Central America. Eight begomoviruses were identified by sequencing PCR fragments. Four of these were new viruses: Tomato severe leaf curl virus (ToSLCV, AF130415), Tomato golden mottle virus (ToGMoV, AF132852), Tomato mild mottle virus (ToMiMoV, AF131071) and Tomato yellow mottle virus (ToYMoV, AF112981) and four were previously characterized: Pepper golden mosaic virus (PepGMV, AF136404) Tomato mosaic Havana virus (ToMHV, AF139078), Tomato leaf curl Sinaloa virus (ToLCSinV, AF131213) and Pepper huasteco yellow vein virus (PHYVV). A general probe, consisting of the most conserved region of the CP gene of Bean yellow golden mosaic virus (BGYMV, M91604), was used in non-radioactive hybridization methods to detect these viruses. Specific probes, which consisted primarily of the common region for each virus, were developed. Specific primers for PCR identification were designed for each virus, and the PCR fragments obtained from plant samples with these begomovirus-specific primer sets were sequenced to confirm the primers’ specificity. ToSLCV, ToMHV and ToMiMoV were present in tomatoes from Comayagua, Honduras. ToYMoV and ToLCSinV were present only in samples from Costa Rica. ToSLCV, ToGMoV, ToMiMoV, ToMHV, PepGMV, ToLCSinV and PHYVV were detected in tomatoes in one valley in Guatemala. ToSLCV only existed in mixed infections with another bipartite begomovirus, and no DNA-B was ever identified for this virus. These detection methods were used to monitor begomoviruses in tomato-breeding lines being developed in Guatemala for resistance to these viruses.


Tomato-infecting begomoviruses cause serious losses throughout Central America (Morales and Anderson, 2001; Jones, 2003). In many locations, incidence is 100% during the dry season and losses may exceed 60%. These epidemics are associated with the changing agricultural practices, such as continuous cropping of tomatoes and increased cultivation of hosts for the natural vector, Bemisia tabaci (Polston and Anderson, 1997). Since resistant cultivars are not available, the major control measures have involved numerous applications of insecticides, e.g., applications every third day. Losses have been so extensive in some areas of Guatemala and Nicaragua that tomatoes are no longer grown. The begomoviruses associated with tomatoes in Central America have not been molecularly characterized. Thus, this study was undertaken to provide sequence data for the begomoviruses associated with tomatoes collected in Central America, and these sequences were used to develop specific molecular tools for detection of these begomoviruses and to apply these tools in a tomato breeding program in Guatemala for begomovirus resistance (Mejía et al., 2004).

Materials and Methods

Characterization of tomato-infecting begomoviruses from Central America
Tomatoes exhibiting begomoviral symptoms (leaf curl, leaf crumple, yellow mottle and dwarfing) were collected from Guatemala, Honduras, Costa Rica, and Nicaragua. Tissues from young leaves for analysis were used for DNA extraction. Tissues were stored either dried or frozen. Discs were cut from fresh leaf tissue with a flame-sterilized, number 5 cork borer, placed in sterilized 1.5-ml microfuge tubes and stored at -80°C. Dry samples were stored as 0.5-mm-wide strips cut with a flame-sterilized blade, dried at room temperature, placed in paper envelopes, and stored at room temperature. DNA from plant tissue was extracted by a modified Dellaporta method (Rojas et al., 1993) or by the Puregene DNA Isolation Kit, Cell and Tissue Kit (Gentra Systems, Minneapolis, MN).
PCR products containing the common region and part of the Rep and CP genes of DNA-A were obtained using degenerate geminiviral primer set (PRepv1978: 5’ GCCCACATYGTCTTYCCNGT 3’/PCPc715: 5’ TTDATRTTYTCRTCCATCCA 3’; Potter et al., 2003). These fragments were cloned in pCR2.1 vector (Original TA CloningR Kit, Invitrogen Corporation, Carlsbad, CA) and sequenced using Big Dye Sequencing Kit™ (Biotechnology Center, Madison, WI). Sequence analysis was accomplished by comparison with known DNA sequences through the National Center for Biotechnology Information BLAST program, the GCG program (Wisconsin Package Version 10.2, Genetics Computer Group, Madison, WI) and by DNAMAN Version 5.2.9 software, Lynnon BioSoft, Canada.
Full-length PCR fragments were obtained for ToGMoV and ToSLCV. The PCR overlapping primer set (PToGMoVAvf-NcoI: 5’ TAT CCC ATG GTT CTG GAG CCT TTG CG 3’/PToGMoVAcf-NcoI: 5’TAT TCC ATG GGT TCC TCC ATT TCC ACT CTC C 3’) was used to amplify the full-length fragment of ToGMoV DNA-A, and the primer set PToGMoVBvf-ApaI: 5’ AAC AGG GCC CAT AAA AAA TGA CCC GCG C 3’ / PToGMoVBcf-ApaI 5’ TTG TGG GCC CGG GTA GGT AAA AAA TCG C 3’ was used to amplify the DNA-B. Also, the overlapping primer set PToSLCVAvf-SacI: 5’ AAA AGA GCT CTC TCT AAA ACT CTA TGT TGC TGG 3’ / PToSLCVAcf-SacI: 5’ AAA AGA GCT CCC CCT GGT GTC CTG GC 3’ used to amplify DNA-A of ToSLCV. A full-length clone of DNA-A of Pepper huasteco yellow vein virus (PHYVV, AY044162) was provided by R. L. Gilbertson (University of California-Davis) and was used as positive control for virus detection experiments. Bean golden yellow mosaic virus full-length DNA-A clone pGAA1 (BGYMV, M91604) (Gilbertson et al., 1991), was used as positive control target DNA for the degenerate primer set PRepv1978/PCPc715. Tomato yellow leaf curl virus (TYLCV) infectious clone, pTYEG14 (AY594174), was used as positive control target DNA for this virus.

PCR protocols and begomovirus-specific primer design
A specific PCR primer pair was developed for each of eight tomato-infecting begomoviruses (Table 1) using unique conserved regions of the viral genome identified by sequence alignments for the Rep gene and common region (data not shown). The specificity and robustness of the specific PCR primer pairs were tested by PCR with standard viral DNAs from sequenced clones.
PCR parameters for the eight sets of specific primers were optimized for 25-µl reactions containing: 2.5 µl of 2.5 mM deoxynucleotide triphosphates (dNTPs), 2.5 µl of 10x buffer, 2.5 µl of 30 mM MgCl2, 0.2 µl of Taq DNA polymerase, 2.3 µl of Taq enzyme diluent (Idaho Technology), 2.0 µl of each complementary and viral sense primer at 10 µM, 5.0 µl of DNA extract, and H20. Taq DNA polymerase, MgCl2, 10x buffer, and dNTPs were purchased from Promega (Madison, WI) and used as indicated by the manufacturer. The target DNAs from infected plants were diluted 10 times and the cloned viral DNAs were used at 20 ng/µl. PCR cycle parameters for specific fragment amplification with the specific primers were as follows: denaturation at 96°C for 30 sec, annealing temperature as in Table 1 for 30 sec, and extension at 72°C for 35 sec for 30 cycles. The slope was set to 2.0. Annealing temperature was dependent on the primer set (Table 1). All reactions were in thin-walled 0.2-ml Labsource® PCR tubes. PCR-amplified fragments were separated by gel electrophoresis using 1.0% to 1.5% GenePure LE agarose (IscBioExpress, Kaysville, UT). PCR reactions were performed in a RapidCycler™ (Idaho Technology, Idaho Falls, ID). To confirm the specificity of the PCR-based diagnostic tools, the PCR fragments from reactions with the specific primer pairs were directly sequenced using Big Dye Sequencing Kit™. Sequences of PCR fragments from reactions with specific primer pairs that had a 91% or greater nucleotide identity for 400 nt of the Rep gene and the common region to a known sequence were considered to be the same begomovirus (Fauquet et al., 2003).
Each sample was amplified with degenerative primer pair set for DNA-A (PRepv1978/PCPc715) with the same PCR conditions used for the specific PCR primer pairs except that the PCR cycle parameters for fragment amplification were as follows: denaturation at 94°C for 1 min, annealing at 55°C for 2 min, and extension at 72°C for 2 min for 30 cycles. These cycles were followed by a final cycle of 94°C for 1 min, 61°C for 2 min, and 72°C for 7 min, and then held at 18°C. Samples that tested positive with the PRepv1978/PCPc715 primer pair were then tested with the begomovirus-specific PCR primer pairs. The latter PCR cycle parameters were used also with the TYLCV-specific PCR primer pair (PTYIRv21: 5’ TTG AAA TGA ATC GGT GTC CC 3’/PTYIRc287: 5’ TTG CAA GAC AAA AAA CTT GGG ACC 3’) to monitor the existence of this virus in Central America. These PCR reactions were performed in a MJ DNA Engine PT200 Thermocycler™ (MJ Research Inc., Waltham, MA).

Nucleic acid hybridization probes and hybridization protocols
Nucleic acid hybridization techniques were developed with non-radioactive probes for dot blot detection of the eight tomato-infecting begomoviruses. The specific probes for hybridization were designed using the common region of the viral genomes. This region differs among begomovirus species but is highly conserved within a given species. A general probe for hybridization with begomoviruses was based on the highly conserved region of the CP gene for Western Hemisphere begomoviruses. PCR primer pairs PBGGTv647 (5’ TAT GTG TAT ATC CGA TGT CAC ACG TGG 3’) and PBGGTc1048 (5’ CGA ATT TTC AAT GTC GCA TAT ACA GGG 3’) were developed to produce the DNA for the general probe from DNA-A of clone pGAA1 of BGYMV (Table 2). Virus-specific DNA for probe labeling was produced by PCR with primer pairs listed (Table 2) and with cloned viral DNAs as targets. PCR conditions for DNA probe amplification were the same as for the degenerate PCR primer pair.
The Alk Phos Direct Hybridization Kit™ (Amersham Pharmacia, Pisscatway, NJ) was used for the non-radioactive labeling of probes according to the manufacturer’s instructions. Hybridization techniques for the specific probes used high stringency with positively-charged nylon membranes (Biodyne A, Pall-Gelman, Ann Arbor, MI) at 65ºC according to the manufacturer’s instructions. Low stringency was used with the CP-general probe at 55ºC according to manufacturer’s instructions. The specificity of the hybridization technique was confirmed by using full-length cloned viral DNA of DNA-A from representative of different phylogenetic clusters (Faria et al., 1994) of begomoviruses: Bean dwarf mosaic virus (pBDA1), Bean golden mosaic virus (pBZA1), and Bean golden yellow mosaic virus (PGAA1) were obtained as infectious clones (Gilbertson et al., 1991). Bean calico mosaic virus was a full-length clone (pBCaMV-A2) and TYLCV was an infectious clone (pTYEG14). These viral DNA targets were obtained by preparation of the plasmids of the full-length clones and spotting 25 ng DNA on the membrane.
Also, viral DNA from PCR fragments prepared from the clones of the eight tomato-infecting begomoviruses with the primer pair PRepv1978/PCPc715 served as positive control DNAs for evaluation of the hybridization methods. Twenty-five ng of DNA of each of the eight targeted viruses were spotted on each membrane as control targets. DNA from plant tissue was extracted using the Dellaporta method and 5 µl of the DNA extract was spotted on the membrane as the target for hybridization. The results of the hybridization and PCR experiments were compared with those of the direct sequencing of the PCR-amplified fragments to confirm the specificities of both techniques.

Utilizing the diagnostic tools in a tomato breeding program for resistance to begomoviruses in Guatemala
Young tomato leaves were collected in March 2004 from different tomato breeding lines and commercial lines expressing various disease severity indexes (DSI, see Mejía et al., 2004). DNA was extracted from plant tissue by Puregene DNA Isolation Kit, Cell and Tissue Kit (Gentra Systems, Minneapolis, MN) and examined with both PCR and hybridization techniques.


Molecular characterization of tomato-infecting geminiviruses in Central America
PCR products containing the common region and part of the Rep and CP genes of DNA-A were obtained (?1.4 kb) using degenerate geminiviral primer set PRepv1978/PCPc715. These fragments were cloned and sequenced (Table 3). Seven different tomato-infecting begomoviruses were identified in tomatoes growing in Central America. Four of these were new viruses: Tomato severe leaf curl virus (ToSLCV), Tomato golden mottle virus (ToGMoV), Tomato mild mottle virus (ToMiMoV) and Tomato yellow mottle virus (ToYMoV); and three had been previously characterized: Pepper golden mosaic virus (PepGMV) Tomato mosaic Havana virus (ToMHV) and Tomato leaf curl Sinaloa virus (ToLCSinV). Based on the obtained sequences, specific primers were designed and used to amplify full-length clones of DNA-A for ToGMoV and ToSLCV (Table 3). Specific primers were also used to amplify a full-length clone of the DNA-B of ToGMoV. No DNA-B was ever identified for this for ToSLCV and it always exited in mixed infections with another bipartite begomovirus (data not shown). DNA-A of ToSLCV was also isolated in 1997 from cucumber growing near Sanarate, Guatemala (AF131735).
The primer set PPHYVVv/PPHYVVc specific for PHYVV was used to examine tomato samples collected from Guatemala. Sequence analysis of the PCR-amplified 700-bp DNA fragment showed 97% nucleotide identity to PHYVV. This proved the existence of PHYVV in Guatemala.
Sequences of the eight different tomato-infecting geminiviruses detected in Central America in this study were analyzed together with twelve previously identified, whitefly-transmitted geminiviruses. The relationships among the 20 whitefly-transmitted geminiviruses were based on 805 nucleotides of the N-terminus of the Rep gene. Sequence analysis showed that the eight identified viruses are from different species. A phylogenetic tree (DNAMAN software) indicated (Fig. 1) that ToYMoV and ToGMoV are not related and they are not closely related to any of the identified geminiviruses. ToLCSinV is in the same cluster as Potato yellow mosaic Trinidad virus (PYMTriV), ToMiMoV grouped with Tomato rugose mosaic virus (ToRMV). ToSLCV is most closely related to Squash leaf curl virus (SLCV). Using the Phylogenetic tree (Fig. 1) we can put the eight detected begomovirus in Central America in six different groups. Group one has ToYMoV. Group two has PYMTriV, ToLCSinV, and ToMHV. Group three has PHYVV. Group four has ToRMV, Potato yellow mosaic virus (PYMV) and ToMiMoV. Group five has ToGMoV. Group six has ToSLCV, SLCV and PepGMV.

Evaluation of PCR primer pairs designed for detection of begomoviruses using cloned DNAs as the target DNAs
The begomovirus-specific PCR primer pairs developed for each of the seven isolated viruses produced the expected fragments of ca. 400 bp. The specific primers designed for PHYVV amplified the expected 700-bp fragment. All primer pairs amplify only their respective viral DNAs. The specificity of the primer pair for each virus was obtained by adjusting the annealing temperature for each set of primers (Table 1).

Evaluation of hybridization probes designed for detection of begomoviral DNAs with cloned DNAs and plant extracts
The PCR primer pairs (Table 2) designed to amplify specific DNA probes from cloned viral DNAs gave the expected sizes, which ranged from 120 to 700 bp. All probes were specific to their respective viral targets and did not give hybridization signals with symptomless tomatoes or from the other twelve different cloned viral DNAs (seven characterized tomato viruses from this study plus the five standard viruses from different phylogenetic clusters). The general probe gave strong signals with all five different cloned, full-length viral DNAs from different clusters of begomoviruses (BDMV, BGMV, BGYMV, BCaMV and TYLCV) and did not react with the healthy tomato extract.

Field survey of tomato plants collected in Central America
The general and specific detection tools developed were applied to field-collected tomato tissues from 1992-2004 in Guatemala, Honduras, Costa Rica, and Nicaragua (Table 4). All 88 tomato plants tested exhibited leaf curl, mottle, and/or golden mottle symptoms. DNA was extracted by the Dellaporta method (Rojas et al., 1993) and tested with PCR and hybridization methods. Seventy-three samples gave hybridization signals with the CP-general probe. The degenerate DNA-A primer pair (PRepv1978/PCPc715) detected begomoviruses in 82 samples. PCR primer pairs and hybridization probes for the specific viruses detected the presence of tomato-infecting begomoviruses in 82 of the 88 tomato samples collected from Central America (Table 4). The PHYVV-specific probe and PHYVV-specific primers were used only on samples collected from Guatemala in 2004, and were useful in documenting the existence of this virus in Guatemala. Six of the 88 tomato samples did not give PCR fragments with any of the primer pairs or hybridization signals with any of the probes. Their symptoms may have been due to viruses other than begomoviruses or environmental factors. Our results are consistent with the many reports (Potter et al., 2003) that have shown that PCR methods are more sensitive than hybridization methods for detection of viruses. ToSLCV, ToGMoV, ToMHV, PepGMV, ToLCSinV, ToMiMoV and PHYVV were detected in tomatoes in one valley near Sanarate in Guatemala. ToSLCV, ToMHV and ToMiMoV were present in tomatoes from Comayagua, Honduras. Only ToYMoV and ToLCSinV were present in samples from Alajuela, Costa Rica. Only ToMiMoV and ToLCSinV were present in Nicaragua in samples collected from Sebaco Valley in 1992. TYLCV was never identified in any of the tomato samples examined in this study.

Confirmation of specificity of the PCR and hybridization diagnostic tools
Positive and negative controls were included in each experiment to confirm the specificity of the developed diagnostic tools. In each experiment, the positive and negative controls reacted as expected. One hundred eleven PCR-amplified fragments from thirty-nine tomato samples were directly sequenced to confirm the specificity of the designed PCR-primer pairs. One hundred six PCR-amplified fragments were obtained from tomato samples collected in Guatemala in 2002, 2003 and 2004, and five PCR-amplified fragments with primers PToYMoVC1v/PToYMoVC1c were from tomato samples collected in Costa Rica in 1994 and 1996 (Table 4). All fragments obtained with virus-specific PCR primers had a 91% or greater nucleotide identity with their respective viruses for 400 nt or more for the Rep gene and the common region. Twenty-nine PCR fragments obtained with the ToSLCV-specific primer pair were identified as ToSLCV, twenty-eight as ToGMoV, thirteen as ToMiMoV, another thirteen as ToMHV, eight as ToLCSinV, seven as PepGMV and eight more as PHYVV. These PCR results agreed with the hybridization results.
The PCR and hybridization diagnostic tools developed were useful in monitoring tomato-infecting begomoviruses in Central America. These diagnostic tools are fast, sensitive, and accurate and can be used to detect these viruses in many samples. With the non-radioactive hybridization method, the begomoviruses can be identified and their concentration in the plants estimated. These detection methods will have applications in epidemiological studies and will be useful for monitoring begomoviruses in tomato germplasm in breeding programs for resistance.

Application of PCR and hybridization detection methods in a tomato breeding program in Guatemala
The PCR detection and hybridization protocols developed were utilized in a program for breeding tomatoes resistant to begomoviruses in Guatemala (see Mejía et al., 2004). The begomovirus resistant lines, Gh13 and Gc16, had begomovirus resistance genes from introgressions from Lycopersicon hirsutum (Vidavsky and Czosnek, 1988) and Lycopersicon chilense (Scott et al., 1995), respectively. Young tissues (thirty days after transplanting) were collected from begomovirus susceptible line (M82) and F1 populations (susceptible x resistant; resistant x resistant), and DNA was extracted using the Gentra Extraction kit. Hybridization with CP-general probe (Fig. 2) and with the eight begomovirus-specific probes gave positive signals for seven begomoviruses (Table 5). Only ToYMoV was not detected in these samples; this results agree with the previous survey data, since this virus has not been detected in Guatemala (Table 4). Each line and F1 germplasm was infected with at least two different begomoviruses except the resistant F1 germplasm, XA175 (Gh13 x Gc16), which did not have a positive hybridization signal with any of the probes. Hybrid XA175 had no viral symptoms. However, when the degenerate PCR primers were used with samples form XA175, PCR fragments were detected, and both ToGMoV- and ToMiMoV-specific primer sets gave PCR fragments. Since no hybridization signals were detected from these plants, the viral titer is very low; and this might explain the lack of symptoms. Strong hybridization signals were noted with each of the other two hybrid lines, which had one resistant parental line crossed with M82 (Table 5). The hybrid line XA176 (G16c x M82) has resistance genes from the Lycopersicon chilense; and it had a moderate symptoms (DSI = 2.5) and positive hybridization signals with the CP-general probe and with five different begomovirus-specific probes (Table 5). The hybridization signal was strong with the CP-general probe, the ToSLCV-specific probe and the ToMHV-specific probe (Fig. 2) but the signals were weak with the ToGMoV-, ToMiMoV- and PHYVV-specific probes. Also, another hybrid line XA188 (G13h x M82), which has resistance genes from Lycopersicon hirsutum, had very slight symptoms (DSI = 1) and had positive hybridization signals with the CP-general probe and five different begomovirus-specific probes (Table 5). The CP-general probe and the begomovirus-specific probes for ToSLCV and PepGMV gave medium hybridization signals with this hybrid, whereas the hybridization signals of the specific probes for ToGMoV, ToMiMoV and PHYVV were weak. PCR-amplified viral DNA fragments of the expected sizes were detected in all plants that had positive hybridization signals (Fig. 3). In a few cases, viral DNA fragments of the expected size were successfully PCR-amplified from samples with no positive hybridization signals (Table 5). The susceptible line M82 was infected with at least six begomoviruses in one plant (ToSLCV, ToGMoV, ToMiMoV, ToMHV, ToLCSinV, and PHYVV). In another M82 plant, five begomoviruses (ToSLCV, ToGMoV, ToMHV, ToLCSinV, and PHYVV) were detected. In a third plant, only two begomoviruses (ToSLCV and ToMHV) were identified. Commercial hybrids, Elios, Marina, Silverado, and Tango, were infected with ToSLCV and ToGMoV. The specific diagnostic tools detected ToSLCV in all lines and hybrids except the highly resistant hybrid XA175, ToGMoV in all lines and hybrids, PepGMV only in the resistant line XA188 and Elios, ToMiMoV in the line M82 and four hybrids (XA175, XA176, XA188, and Silverado), but not in Tango, Elios and Marina. ToMHV was detected in M82 and all hybrids except the resistant hybrid XA175 and Silverado. ToLCSinV was found only in the susceptible line M82. PHYVV was detected in M82 and XA176, Elios, and Silverado (Table 5). For all the samples that gave positive hybridization signals with the CP-general probe and/or fragments with the degenerate PCR primers set (PRepv1978/PCPc715), there was at least one begomovirus detected with a virus-specific diagnostic tool.
These data indicate that tomato hybrids being developed for Guatemala must have resistance to multiple begomoviruses. Also, since it is expected that TYLCV will eventually be found it Central America, all begomovirus-resistant tomato hybrids for this region should have resistance to both bipartite and monopartite begomoviruses.

This research was supported partially by a grant from Bean-Cowpea CRSP/USAID grant, a CDR/USAID grant no. TA-MOU-01-C21-008, Universidad de San Carlos de Guatemala, and the College of Agricultural and Life Sciences, University of Wisconsin-Madison. The authors thank Lane Milde for assistance in designing and testing the specific primers for PHYVV.

Literature Cited
Faria, J.C., Gilbertson, R.L., Hanson, S.F., Morales, F.J., Ahlquist, P., Loniello, A.O. and Maxwell, D.P. 1994. Bean golden mosaic geminivirus type II isolates from the Dominican Republic and Guatemala: Nucleotide sequence, infectious pseudorecombinants, and phylogenetic relationships. Phytopathology 84:321-329.
Fauquet, C.M., Bisaro, D.M., Briddon, R.W., Brown, J.K., Harrison, B.D., Rybicki, E.P., Stenger, D.C. and Stanley, J. 2003. Revision of taxonomic criteria for species demarcation in the family Geminiviridae, and an updated list of begomovirus species. Arch. Virol. 148:405-421.
Gilbertson, R.L., Faria, J.C., Hanson, S.F., Morales, F.J., Ahlquist, P., Maxwell, D.P. and Russell, D.R. 1991. Cloning of the complete DNA genomes of four bean-infecting geminiviruses and determining their infectivity by electric discharge particle acceleration. Phytopathology 81:980-985.
Jones, D.R. 2003. Plant viruses transmitted by whiteflies. Europ. J. Plant Path. 109:195-219.
Mejía, L., Teni, R.E., Vidavski, F., Czosnek, H., Lapidot, M., Nakhla, M.K. and Maxwell D.P. 2004. Evaluation of tomato germplasm and selection of breeding lines for resistance to begomoviruses in Guatemala. Acta Hort. (in press)
Morales, F.J. and Anderson, P.K. 2001. The emergence and dissemination of whitefly-transmitted geminiviruses in Latin America. Arch. Virol. 146:415-441.
Polston, J.E., and Anderson, P.K. 1997. The emergence of whitefly-transmitted geminiviruses in tomato in the Western Hemisphere. Plant Dis. 81:1358-1369.
Potter, J.L., Nakhla, M.K., Mejia, L. and Maxwell, D.P. 2003. PCR and DNA hybridization methods for specific detection of bean-infecting begomoviruses in the Americas and Caribbean. Plant Dis. 87:1205-1212.
Rojas, M.R., Gilbertson, R.L., Russell, D.R. and Maxwell, D.P. 1993. Use of degenerate primers in the polymerase chain reaction to detect whitefly-transmitted geminiviruses. Plant Dis. 77:340-347.
Scott, J.W., Stevens, M.R., Barten, J.H.M., Thome, C.R., Polston, J.E., Schuster, D.J. and Serra, C.A. 1995. Introgression of resistance to whitefly-transmitted geminiviruses from Lycopersicon chilense to tomato. In: Bemisia 1995: Taxonomy, Biology, Damage Control and Management, Ed. by D. Gerling and R.T. Mayer, Intercept Ltd., Andover, UK. p. 357-367.
Vidavsky, F. and Czosnek, H. 1998. Tomato breeding lines immune and tolerant to tomato yellow leaf curl virus (TYLCV) issued from Lycopersicon hirsutum. Phytopathology 88:910-914.