Surveys reveal the occurrence of phytoplasmas in plants at different geographical locations in Peru

Source of isolates and extraction of DNA

Details of the origins of material used in the two surveys and the locations of these sites within Peru are given in Table 1 and Fig. 1[link] respectively. In survey 1, plant material from both suspected phytoplasma-infected plants and healthy looking plants was sampled during February 2007, stored at 4°C after sampling [apart from time in transport at ambient temperature (25–30°C) for a duration of 1 day for Mantaro Valley and 3 days for San Martin] and transferred to the laboratories at CIP, Lima for extraction of DNA. Plants from on-site were processed on the day of sampling, those from papaya (Carica papaya) and coconut (Cocos nucifera) within 18 days of sampling and all remaining samples were processed within 12 days of sampling. In survey 2, plant material from suspected phytoplasma-infected plants was collected during November 2007 from two areas of the Mantaro Valley and stored at 4°C after sampling (apart from time in transport at ambient temperature) at the SENASA laboratories, before being transported to Rothamsted Research, UK. DNA was extracted from 300 to 500 mg of leaf tissue using the cetyl trimethyl ammonium bromide method of Doyle & Doyle (1990).

16S ribosomal DNA analysis

Amplifications of the 16S rRNA gene were performed in 25 μL reactions for all samples using ‘illustra PuRe Taq Ready-To-Go™ PCR beads’ (GE Healthcare Life Sciences, Amersham, UK). Survey 1 used 15-ng template DNA and 100 ng of each primer in an MJ Research PTC200 thermocycler (Waltham, MA, USA), and the phytoplasma universal primer pairs P1/P7 and R16F2/R16R2 were used for first and second PCR rounds, respectively, under conditions previously described (Hodgetts et al., 2007). Survey 2 used 20-ng template DNA and 0.4 μM of each primer in a TC-412 TECHNE thermocycler (Burlington, NJ, USA) and the phytoplasma universal primer pairs R16mF2/R1 (Gundersen & Lee, 1996) for first round PCR, and fU5/rU3 (Lorenz et al., 1995) for the second round PCR with conditions as stated in the respective papers. In addition, restriction endonuclease digestion with Sau3AI and HpaII, on the nested-PCR product was performed via the manufacturer’s procedure in a 10 μL final volume. Reference controls of 16SrI [aster yellows dwarf, Germany], 16SrII [Tomato big bud, Australia] and 16SrIII [Vaccinium witches’ broom, Germany] from the Rothamsted Research phytoplasma collection were used for RFLP comparisons. Following PCR (or RFLP), 5 μL (10 μL) of resultant products were separated on 1% (1.5%) agarose gels in 1× TBE (Tris Borate EDTA) buffer containing ethidium bromide and visualised under UV light.

Cloning and sequence analysis

PCR products (16S rRNA) obtained from surveyed material were cloned using the Promega pGEM^®–T easy Vector System (Southampton, UK) via the manufacturer’s protocol. Colonies corresponding to samples from survey 1 were amplified by PCR with the universal primers M13For and M13Rev described in the kit, and these PCR products were processed on both strands using a Beckman Quickstart kit technology that uses WellRed Dye chemistry (infrared dyes), with a CEQ 8000 Genetic Analysis System, manufactured by Beckman Coulter (High Wycombe, UK). Colonies corresponding to samples from survey 2 were confirmed by PCR with fU5/rU3 primer pair. Plasmid DNA of recombinant colonies was purified (QIAprep^® Spin Miniprep; Qiagen, Crawley, UK) and sequenced on both strands with M13 forward and reverse primers.

Terminal-restriction fragment length polymorphism

Samples from survey 1 were subjected to T-RFLP analyses using primers for the 23S rRNA gene as previously described (Hodgetts et al., 2007). After amplification, PCR products were transported to the University of Nottingham, digested with either MseI (New England Biolabs, Hitchin, UK) or Bsh12361 (Fermentas Life Sciences, York, UK) in 10 μL reaction volumes containing 1 U of restriction enzyme for 2 h at 37°C, and analysed on a CEQ 8000 DNA analysis system as previously described (Hodgetts et al., 2007).

Phylogenetic analysis

Complete sequences from primers R16F2 to R16R2 were obtained for all the phytoplasmas from survey 1, but samples from survey 2 were only sequenced between primers fU5 and rU3. BLAST searches (Altschul et al., 1990) were performed at the National Center for Biotechnology Information web site at http://www.ncbi.nlm.nih.gov/. Phylogenetic trees were constructed with partial 16Sr sequences obtained by trimming all sequences to the same region flanked by primers fU5 and rU3. Sequence alignments were performed using ClustalW (Thompson et al., 1994), and phylogenetic and molecular evolutionary analyses were performed using the MEGA version 3.1 neighbour-joining method and all default values, with 1000 replications for bootstrap analysis (Kumar et al., 2004).

Surveys of field crops in Peru

In the first survey, sampling was performed in two areas of Peru, the Mantaro Valley and San Martin (Fig. 1). The Mantaro Valley is a high altitude region on the side of the Andes, and San Martin is a low altitude jungle area. These areas therefore represent varied habitats, growth conditions and types of crops in cultivation. In the Mantaro Valley, carrots (Daucus carota) were sampled from two adjacent fields where the crops were 2 and 4 months old, with the latter being harvested for sale at the time of collection. The symptoms displayed by these plants are listed in Table 1, with an example of cracking and fanging of roots shown in Fig. 2a. Fanging of roots could have been attributed partially to the stony nature of the soil and there was also suspicion of nematode damage although this could not be confirmed because of lack of appropriate facilities. Maize (Zea mays) plants at the early flowering stage also showed two distinct sets of symptoms (Table 1; Fig. 2b to Fig. 2d[link]). Some of these symptoms were possibly because of the presence of viruses and the corn stunt spiroplasma that are known to be widespread in Peru (Nault et al., 1979), but these were not tested for in this study. The symptom sets were distributed throughout the fields with no obvious clustering. Symptoms on native potato (Solanum tuberosum) and on Shiri (likely Solanum tuberosum cultivar group Curtilobum or Juzepczukii; Huamán & Spooner, 2002), a native Andean root and tuber crop, are given in Table 1 and shown in Fig. 2e and Fig. 2f[link]. In certain sampling fields, evidence of the previous season’s crop was visible and there were a small number of plants regrowing within the main crop. In the carrot fields, these crops were onions and in the maize fields they were carrots, and these crops were also sampled. In San Martin, a range of papaya plantations and coconut palms were sampled (Table 1; Fig. 2g and Fig. 2h[link]).

Additional samples were obtained from CIP Lima, this being an irrigated desert habitat in which potatoes are grown as part of the organisation’s routine work. CIP also maintain some potato and tomato material in culture suspected of harbouring phytoplasmas, and these were included in the study and are listed in Table 1 under the locations from which the samples were originally obtained, that is, Tacna (irrigated desert), Andahuaylas (Andes), Chivay (Andes), Ica (irrigated desert). Those from Chivay and Ica had been transferred by grafting and then maintained in Datura spp., and those from Tacna and Andahuaylas were maintained in the original plant species. CIP’s routine screening of potatoes for export also produced some samples of improved potato variety with leaf reddening and vein necrosis that were included in the survey. The C. roseus samples were from ornamental flower beds at CIP Lima where some plants were showing phytoplasma symptoms and were growing alongside healthy looking plants.

Control plants of tomato and improved potato were provided at CIP Lima where they had been grown from disease-free stock/seed in insect proof glasshouses. Asymptomatic plant samples were also collected at each sampling site for each crop with the exception of maize, where asymptomatic plants were sampled from a region close to the sampling area (CIP, Huancayo) because of all fields at the sites showing widespread symptoms with no convincing asymptomatic plants present. In the second survey, sampling was performed within two areas of the Mantaro Valley (Fig. 1), and the symptoms displayed are listed in Table 1, with symptoms on oats shown in Fig. 2i.

16S ribosomal RNA analysis

Approximately 20% of samples produced phytoplasma PCR products following first round PCR, and the results for nested PCR are shown in Table 1. The absence of any previous molecular analysis of phytoplasmas in Peru precluded the use of positive controls in survey 1, but negative controls were used at both stages of PCR and were clear from contamination. The 89 samples from survey 1 listed in Table 1 represent extractions from 64 different individual plants, of which 52 were positive and the remainder were negative with no ambiguity. Ten carrots were sampled, and all except one had extractions performed from both the foliage and root material. No phytoplasma DNA was detected in any of the samples with bronzing of foliage, while five of the six samples that had yellowing of foliage were positive for phytoplasma DNA in either roots, leaves or both roots and leaves. Roots that were cracked and fanged were found to be both positive and negative, and thus these symptoms may not be attributed to phytoplasma infection (fanging is not characteristic of phytoplasma infection in carrots). The asymptomatic carrot was also tested and was found to be positive for phytoplasma.

Four maize plants were tested, and all symptomatic plants were positive in both sample types (grain and leaves). The asymptomatic control was positive in the leaves but not grains, perhaps indicating that the plant was at an earlier stage of infection. The papaya and coconut samples represent individual plants from 16 plantations, and testing revealed that all but three plantations had phytoplasma infection (Table 1). All three coconut samples and 14 of the 21 papaya samples tested positive for phytoplasma DNA. Both Shiri plants (one with and one without symptoms) were positive for phytoplasma DNA, and two of the four C. roseus samples from within CIP Lima were also positive for phytoplasmas (one symptomatic and one asymptomatic). From the samples maintained at CIP, the tomato from Ica and potato from Andahuaylas were found to be infected with phytoplasmas. Those from Tacna and Chivay had no detectable phytoplasma DNA, although this does not rule out the possibility that the original samples were positive but the phytoplasma was lost during the grafting procedures that had been used to maintain the cultures. Other potatoes at CIP that gave a positive result in testing were a ‘healthy control’ plant and a symptomatic sample taken from routine quarantine testing.

Overall, in the Mantaro Valley, the crops found to be positive for phytoplasma DNA were carrot, maize, native potato and Shiri. The symptomatic samples of alfalfa (Medicago sativa) and clover (Trifolium sp.) were negative for phytoplasma DNA, and subsequent analysis in which a region of the 23S rRNA was PCR amplified and sequenced indicated the presence of other bacterial infections (an unculturable bacterium and an Actinobacterium spp.) (results not shown). In addition, a symptomatic onion sample (found in a carrot field) was negative for all PCR tests and was therefore suspected to be a viral infection.

Samples from survey 2 represent extractions from seven different individual plants, of which five were positive for phytoplasma DNA and the remainder were negative (Table 1). Four carrots were sampled, two each from two locations, and those from Huayao were found to be positive, while those from Marcatuna were free of phytoplasma infection. Both the oat and potato samples from Huayao along with the dandelion from Marcatuna were found to be positive for phytoplasma DNA. RFLP analysis using Sau3AI (Fig. 3) and HpaII (data not shown) indicated that the carrot samples and the oat sample showed RFLP profiles identical to those of the group 16SrI reference strain. However, the restriction profiles for the phytoplasma in the potato sample were identical to those of the 16SrII reference strain, while those from the dandelion were identical to the patterns of the 16SrIII X-disease group phytoplasma.

T-RFLP results

Terminal RFLP was performed on all samples from survey 1, which were positive with the 16S rRNA primers (R16F2/R16R2) as well as a negative sample for each plant species. All the control plant samples showed a single TRF (Terminal Restriction Fragment) at 461 bp, which is the size of the product expected from plant chloroplast DNA, indicating that the PCR had been successful and that there was no phytoplasma present (Fig. 4a). With the exception of one sample, all the infected isolates were identified by the technique to be 16SrI aster yellows group phytoplasmas by the presence of a TRF of 129 bp with MseI (Fig. 4d) and 305 bp with Bsh12361 (Fig. 4e) (Hodgetts et al., 2007). The isolate Tomato 60 from Ica, was identified as a 16SrIII X-disease group phytoplasma by the presence of a TRF of 129 bp with MseI (Fig. 4b) and 450 bp with Bsh12361 (Fig. 4c). For samples where 16S rRNA PCR had failed to identify phytoplasma infection, no phytoplasmas were identified by T-RFLP.

Phylogenetic analysis

For those samples from survey 1, for which PCR and T-RFLP had confirmed the presence of phytoplasma DNA, four representative samples (papaya 34 from Viveres, native potato 55 from Andahuaylas, C. roseus 61 from CIP, Lima and Tomato 60 from Ica) were selected from which to clone and sequence the 16S rDNA. From the five positive samples from survey 2, where PCR and RFLP had confirmed the presence of phytoplasma DNA, one sample from each crop at each location (carrot A from Huayao, oat E from Huayao, potato G from Huayao and dandelion F from Marcatuna) were selected and the 16S rDNA was cloned and sequenced. Complete sequences from between the nested-PCR fU5/rU3 primers were obtained for all the samples. Sequencing and phylogenetic analysis (Fig. 5) indicated that the majority of the samples belonged to the 16SrI aster yellows group, with those from Andahuaylas, Viveres and CIP most closely aligned to the 16SrI-B subgroup, and those from Huayao to the 16SrI-A subgroup. The potato sample from Huayao had 99% sequence identity to members of the 16SrII peanut witches’ broom group, while the tomato sample from Ica and the dandelion sample from Marcatuna had 99% sequence identity with the Solanum quitoense machorreo phytoplasma from Colombia (accession number AY731819) that belongs to the 16SrIII X-disease group.