INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Marginal chlorosis of strawberry (Fragaria × ananassa) has affected strawberry production in France for more than 10 years. A phloem-restricted bacterium was shown to be associated with the disease (25). Attempts to grow the bacterium were unsuccessful, as reported for other phloem-restricted bacteria, such as "Candidatus Liberibacter" (12), and the proteobacteria of the papaya bunchy top and cucurbit yellow vine diseases (4, 9). Phylogenetic characterization of the marginal chlorosis bacterium was based on 16S rRNA gene (rDNA) sequence analysis and showed the bacterium to be a new "genus" in the 3 subdivision of the Proteobacteria. It was named "Candidatus Phlomobacter fragariae" (28). Among 20 or so phloem-restricted walled eubacteria reported to be associated with plant diseases, only 5 have been phylogenetically characterized. All turned out to be new bacteria belonging to different phyla within the clade Proteobacteria. "Candidatus Liberibacter asiaticus" and "Candidatus Liberibacter africanus," two species associated with citrus Huanglongbing (ex greening disease) are members of a new phylum in the  subdivision of the Proteobacteria (19-21). The bacterium responsible for yellow vine disease of cucurbits is a Serratia marcescens-related member in the 3 subdivision, whereas papaya bunchy top is caused by a rickettsia-related bacterium in the  subdivision (3, 10). All of the phloem-restricted bacteria studied so far are transmitted by sap-feeding insects, i.e., leafhoppers, planthoppers, or psyllids. Strategies for disease control are based on the production of pathogen-free plant material, eradication of infected plant sources, and insect vector control. Thus, sensitive and specific detection of the bacterium and identification of the insect(s) vector(s) are required for disease management. The psyllids Trioza erytreae and Diaphorina citri were demonstrated to be vectors of the citrus Huanglongbing bacteria (5, 24), and the leafhoppers Empoasca papayae and E. stevensi were shown to transmit the papaya bunchy top agent (1, 15) using exhaustive insect inventory and extensive experimental transmission trials, as no alternative techniques were available. Today, experimental transmission assays can be greatly reduced by preliminary identification of insect carriers of the causal agent. In this way, we identified the planthopper Hyalesthes obsoletus as the vector of the stolbur phytoplasma (11). In the case of "P. fragariae," no flying insect vector has been identified. Transmission of "P. fragariae" by an insect vector is, however, strongly suspected as insecticide treatments reduce the incidence of strawberry marginal chlorosis and natural infection of potted healthy in vitro-propagated strawberry plants was obtained after exposure in the field (L. Zreik, J. L. Danet, X. Foissac, J. G. Nourrisseau, J. Gandar, E. Verdin, J. M. Bové, and M. Garnier, unpublished data).

"P. fragariae" can be efficiently detected in plants by a PCR test based on the sequence of its 16S rDNA (16S-PCR) (28). However, when the 16S-PCR test was used to detect "P. fragariae" in field-collected homopteran insects, many insects gave positive reactions. Sequencing of some of the amplicons indicated that they corresponded to the 16S rDNAs of phylogenetically related bacteria (Zreik et al., unpublished). Indeed, the 3 proteobacterial subgroup includes enteric bacteria of insects, as well as insect symbionts and parasites. Because restriction fragment length polymorphism (RFLP) profiles of the 16S rDNA amplicons from "P. fragariae" and from the cross-reacting bacteria were identical, cloning of "P. fragariae" genes other than the 16S rDNA was necessary.

In the work reported here, we have cloned and sequenced part of the spoT gene of "P. fragariae" by using comparative randomly amplified polymorphic DNA (RAPD) analysis, a method that we have already used to isolate genes of "Liberibacter sp." (16). A PCR assay is described which, in combination with RFLP, allows specific identification of "P. fragariae" in insects. Results from detection tests conducted on sap-sucking insects collected in and around production tunnels over a 4-year period, as well as on whiteflies proliferating on "P. fragariae"-infected strawberry plants under confined greenhouse conditions, are also presented.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Plant materials and greenhouse conditions. Field-collected strawberry plants were insecticide treated (one application of systemic imidaclopride) and maintained in individual pots in a greenhouse compartment at 22 ± 2°C, under 16 h of light.

Insect collection. Insects were captured using a D-Vac aspirator at four different locations in southwestern France. The first location was a strawberry farm at Monpazier (from 1994 to 1998) where natural transmission had been demonstrated to occur (Zreik et al., unpublished). The other three locations were a strawberry farm at Villefranche du Queyran (in 1998), an experimental strawberry field at Lanxade (from 1994 to 1997), and a strawberry nursery at Siorac (in 1994). Insects were collected twice a month from May to October on strawberry plants and on wild plants around production tunnels. Homopteran insects (except aphids) were grouped in homogeneous batches according to external morphologic characteristics, and specimens from each batch were kept dried for taxonomic identification. The remaining insects were kept frozen until DNA extraction. Insects were identified in accordance with common handbooks for taxonomy of planthoppers, leafhoppers (14, 22, 26, 27), and psyllids (17) and morphological description of whiteflies (23). When a species could not be determined with certainty because of the lack of a male specimen (species criteria for leafhoppers and planthoppers often require observation of male genital morphology after dissection) or when a mixture of closely related species was suspected, only the genus was indicated. Control whiteflies were from tobacco plants grown in the laboratory greenhouse.

DNA extraction. Genomic DNA was extracted from strawberry leaf midveins and petioles after grinding in liquid nitrogen as described by Gawel and Jarret (13). Insect DNA was extracted by the same procedure from batches of 1 to 10 insects crushed in Eppendorf tubes. DNAs were solubilized in 30 µl of DNase-free water.

RAPD amplification. Decamer RAPD primers (Operon Technologies Inc., Alameda, Calif.) with 60 or 70% GC contents were used for random PCR on 100 ng of DNA extracted from three healthy and three "P. fragariae"-infected strawberry plants (cultivar Gariguette). Samples were amplified through 40 cycles of 30 s at 94°C, 30 s at 37°C, and 1 min at 72°C using a single primer at 0.4 µM and 1 U of Taq polymerase in a 25-µl reaction mixture containing 78 mM Tris-HCl (pH 8.8), 17 mM (NH₄)₂SO₄, 2 mM MgCl₂, 10 mM -mercaptoethanol, 0.05% W-1 detergent (Gibco BRL, Gaithersburg, Md.), 0.2 mg of bovine serum albumin per ml, and each deoxynucleoside triphosphate at 200 µM. Amplification results were analyzed on ethidium bromide-stained 1.5% agarose gels. DNA size markers were 1-kb and 100-bp ladders from Gibco BRL and a 100-bp DNA ladder plus from MBI-Fermentas (Vilnius, Lithuania).

Cloning and sequence analysis. When a DNA band was consistently observed in RAPD profiles of "P. fragariae"-infected plants but not in healthy plant controls, a pipette tip was inserted into the DNA band in order to collect traces of DNA from the agarose gel. The collected DNA was reamplified, purified from the agarose gel using the Cleanmix Kit (Talent, Trieste, Italy), and ligated into pGEMt easy vector (Promega, Madison, Wis.). Plasmids were electroporated into Escherichia coli XL1-blue cells, and inserts were sequenced by automated fluorescent sequencing. spoT amplicons were either cloned in pGEMt easy vector or directly sequenced using automated fluorescent sequencing. The software programs used for computer analysis were BLAST (2) and LALIGN (18).

PCR test and enzymatic digestion. The spoT-based PCR test was carried out with 100 ng of plant DNA or 1 µl of insect DNA using primers Pfr1 (5'-AATGGGTGTCGCTGCCATT-3') and Pfr4 (5'-AGCAATGAAATTGTTATTAACGC-3') for 40 cycles of 10 s at 92°C, 10 s at 60°C, and 1 min at 72°C in a 50-µl reaction mixture containing 2 U of Taq polymerase (Gibco BRL) under the buffer conditions described above. After amplification, a 10-µl sample of the amplified products was analyzed on a 1.5% agarose gel and visualized by ethidium bromide staining. Amplicons were digested using 10 U of restriction enzymes (Gibco BRL) in a 40-µl volume for 2 h at the recommended temperature and analyzed on a 2.5% agarose gel.

Nucleotide sequence accession number. The nucleotide sequence of RAPD fragment A20-26 of "P. fragariae" has been deposited in the GenBank database under accession number AF191253. The sequences of the Pfr1-Pfr4 amplicons of the two bacteria from Trialeurodes vaporariorum have been deposited in the GenBank database under accession numbers AF220418, for the bacterium regularly found (BTVr; see Fig. 2), and AF220419, for the bacterium occasionally found (BTVo; see Fig. 2).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Random amplification and cloning of the "P. fragariae" spoT gene. Two strawberry plants (cultivar Gariguette) exhibiting strong marginal chlorosis symptoms and previously found to be infected by "P. fragariae" using the 16S-PCR assay were selected. Decamer-primed RAPD patterns obtained from these two plants were compared to those obtained from two healthy strawberry plants of the same cultivar. The A20 decamer (GTTGCGATCC) allowed amplification of a 1.1-kbp DNA fragment from "P. fragariae"-infected strawberry plants (Fig. 1A, lanes 3 and 4) but not from healthy strawberry plants (Fig. 1A, lanes 1 and 2). This amplicon, called A20-26, was cloned and sequenced (Fig. 2). The sequence was 1,164 bp long and shared 71% homology with the E. coli spoT gene, encoding the guanosine-3',5'-bis(diphosphate)3'-pyrophosphohydrolase (ppGppase). The A20-26 fragment encoded a 388-amino-acid partial open reading frame with 82% identity with E. coli ppGppase and 49% identity with that of Haemophilus influenzae. Primers Pfr1 and Pfr4 were defined on the "P. fragariae" sequence, and a PCR test (spoT-PCR) was developed for "P. fragariae" detection. According to the sequence, an 895-bp amplicon is expected. To verify the specificity of Pfr1 and Pfr4, they were used on DNAs extracted from two healthy (Fig. 1B, lanes 2 and 3), three "P. fragariae"-infected (Fig. 1B, lanes 4 to 6), and two stolbur-infected (Fig. 1B, lanes 7 and 8) strawberry plants. An 895-bp fragment was obtained only with "P. fragariae"-infected strawberry DNA, confirming that the A20-26 DNA fragment is part of the "P. fragariae" genome.

View larger version (32K): [in this window] [in a new window]   FIG. 1.   (A) A20-RAPD profiles obtained from healthy (lanes 1 and 2) and "P. fragariae"-infected (lanes 3 and 4) strawberry plants of cultivar Gariguette. (B) Agarose gel electrophoresis of PCR products obtained with primers Pfr1 and Pfr4 from water (lane 1), healthy strawberry plants (lanes 2 and 3), and strawberry plants infected by "P. fragariae" (lanes 4 to 6) or by the stolbur phytoplasma (lanes 7 and 8). The DNA molecular size markers in lanes M are the 1-kb ladder from Gibco BRL (A) and the 100-bp DNA ladder plus from MBI-Fermentas (B).

View larger version (47K): [in this window] [in a new window]   FIG. 2.   Nucleotide sequence of the A20-26 RAPD fragment and comparison with the other Pfr1-Pfr4 amplicon sequences obtained from T. vaporariorum maintained on "P. fragariae"-infected strawberry plants (BTVa), obtained regularly from T. vaporariorum (BTVr), or obtained occasionally from tunnel-collected T. vaporariorum (BTVo) or from field-collected M. laevis (BMLo). Sequences of the A20, Pfr1, and Pfr4 oligonucleotides and restriction sites are underlined.

spoT-PCR and RFLPs on field-collected homopteran insects. PCR with primers Pfr1 and Pfr4 was carried out on 31 homopteran insect batches collected in southwestern France from 1994 to 1996 that tested positive with primers Fra4 and Fra5 for amplification of 16S rDNA (Zreik et al., unpublished) and on additional batches of insects captured in 1997 and 1998. Table 1 summarizes the results of the spoT-PCR tests. Four leafhopper species, namely, Balclutha punctata, Euscelis incisus, Mocydia crocea, and Psammotettix confinis, occasionally gave a positive signal, whereas three Macrosteles species were frequently found positive. An 895-bp amplicon was also observed in the case of two planthopper species, Conomelus anceps and Laodelphax sp., as well as with the psyllid Trioza urticae and the whitefly T. vaporariorum.

View larger version (69K): [in this window] [in a new window]   FIG. 3.   RsaI and AluI restriction profiles of Pfr1-Pfr4-amplified DNAs from "P. fragariae"-infected strawberry plants (lane 1), batches of T. vaporariorum whiteflies collected on "P. fragariae"-infected strawberry plants maintained in a laboratory greenhouse (lane 2), T. vaporariorum whiteflies from strawberry production tunnels (lanes 3 to 5), T. urticae psyllids (lane 6), M. sexnotatus (lanes 7 and 8), M. viridigriseus (lanes 9 and 10), P. confinis (lane 11), C. anceps (lane 12), and M. crocea (lane 13).

Detection and identification of "Candidatus P. fragariae" in T. vaporariorum whiteflies proliferating on "Candidatus P. fragariae"-infected strawberry plants. In order to assess the presence of "P. fragariae" in whiteflies from the greenhouse, five groups of 20 T. vaporariorum from our infected strawberry collection were tested by PCR-RFLP with RsaI, AluI, and HincII (Fig. 4, lanes 3 to 7) in comparison with a "P. fragariae"-infected plant control (Fig. 4, lanes 1 and 2). The characteristic RsaI and AluI profiles of the "P. fragariae" amplicon were clearly evident in two out of the five whitefly groups tested (Fig. 4, lanes 3 and 4) and faintly evident in a third one (Fig. 4, lane 5). Digestions with HincII showed that the amplicons from batches of T. vaporariorum that had RsaI and AluI profiles identical to those of "P. fragariae" also had the same HincII profiles (lanes 1, 3, 4, and 5). However, two DNA fragments of 290 and 605 bp could be found on the "P. fragariae" profile, indicating the presence of only one HincII site (position 290), while the sequence of the A20-26 fragment (Fig. 2) predicted a second HincII site at position 565. As the "P. fragariae"-infected strawberry plant from which the A20-26 fragment was cloned and sequenced had died, we could not check whether this site was present in the original isolate or had been mistakenly introduced by Taq polymerase during the A20-26 amplification process. In order to confirm that "P. fragariae" was present in T. vaporariorum, the amplicon was directly sequenced. As expected, the amplicon had a sequence identical to that of the "P. fragariae" A20-26 fragment except for two nucleotide changes, including one on the second HincII site (Fig. 2, lanes BTVa). The Pfr1-Pfr4 amplicon of type IV regularly obtained with T. vaporariorum whitefly batches was also sequenced (Fig. 2, lanes BTVr). This amplicon exhibited 94% homology (73 nucleotide changes over 852 bp) with the corresponding sequence on the "P. fragariae" A20-26 fragment and thus corresponded to a third bacterial type, as it was also different from type III amplicons, i.e., from M. laevis and T. vaporariorum insects (75 nucleotides different over 852 bp).

View larger version (42K): [in this window] [in a new window]   FIG. 4.   RsaI, AluI, and HincII restriction profiles of Pfr1-Pfr4-amplified DNAs of "P. fragariae"-infected strawberry plants (lanes 1 and 2), batches of 20 T. vaporariorum whiteflies from "P. fragariae"-infected strawberry plants (lanes 3 to 7), and batches of 40 T. vaporariorum whiteflies grown on healthy tobacco plants (lanes 8 and 9). The DNA molecular size markers in lane M are 100-bp ladders from Gibco BRL.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Marginal chlorosis is a new disease of strawberry associated with an uncultured phloem-restricted bacterium recently identified as a new member of the 3 subdivision of the Proteobacteria (25, 28). Many insect enteric bacteria, symbionts, or parasites belong to this bacterial phylogenetic subdivision. This could explain most of the cross-reactions obtained when the 16S-PCR assay was applied to field-collected homopteran insects. For most of these amplicons, RFLP profiles were not discriminating because the few nucleotide sequence differences did not affect the restriction sites (Zreik et al., unpublished). Therefore, cloning of "P. fragariae" genes other than the 16S rDNA became necessary. Because of the low concentration of "P. fragariae" in strawberry plants and the lack of an alternative herbaceous host, such as periwinkle (Catharanthus roseus), plant DNA fractions enriched in "P. fragariae" DNA were difficult to obtain. Recently, using RAPD profiles, phytoplasma and liberibacter genes could be isolated (6, 16). This method made it possible to clone a "P. fragariae" DNA fragment homologous to proteobacterial spoT genes. Because it encodes ppGppase, an enzyme involved in a basic cellular process, i.e., the stringent response, the spoT gene is also somewhat conserved among bacteria but much less than the 16S rDNA. Indeed, fewer insects gave a positive reaction with primers Pfr1 and Pfr4 (spoT-PCR) than with primers Fra4 and Fra5 (16S-PCR) and RFLP profiles of the spoT-PCR amplicons allowed us to distinguish "P. fragariae" from other insect-associated bacteria. Different RFLP profiles of the spoT gene were also evident for various insect-associated bacteria. All of the PCR and RFLP results showed that none of the insects collected in the field over a 5-year period, including whiteflies, were contaminated by "P. fragariae." However, "P. fragariae" could be detected in several batches of T. vaporariorum whiteflies proliferating under confined conditions on "P. fragariae"-infected strawberry plants, showing that T. vaporariorum is able to acquire "P. fragariae." This was unexpected, as whiteflies, although they are known as vectors of viruses, were considered to have feeding canals too small to acquire bacteria. In the whiteflies, a bacterium different from "P. fragariae" was also regularly detected; this could possibly be one of the two symbionts previously described in T. vaporariorum (8). Indeed, the primary T. vaporariorum symbiont belongs to an intermediate cluster between the 2 and 3 subdivisions of the Proteobacteria. The secondary endosymbiont of T. vaporariorum has not been phylogenetically characterized, but that of another whitefly, Bemisia tabaci, is also a member of the 3 subdivision of the Proteobacteria (7). We have not yet been able to detect "P. fragariae" in whiteflies collected in the field; this might be because, in strawberry tunnels, whiteflies have very high proliferation rates and usually prefer to feed on healthy plants. In strawberry production tunnels, adult whiteflies were very rarely found on affected plants as these are heavily stunted. Only nymphs, which are immobile, are more likely to acquire the bacterium if they develop on an infected plant. Thus, the number of "P. fragariae"-infected whiteflies in a natural population is probably very low. Moreover, large invasions of whiteflies are observed within strawberry production tunnels when tobacco and tomato fields have been harvested, and this phenomenon certainly "dilutes" the small number of infected whiteflies with a large number of uninfected ones. In the absence of other insect candidates for the transmission of "P. fragariae," the possibility of transmission by whiteflies has to be studied further, even though at this time it must be considered only a hypothesis. Experiments meant to reproduce the acquisition of "P. fragariae" by T. vaporariorum under confined conditions and determine the ability of whiteflies to transmit "P. fragariae" are under way.