Journal of Applied Microbiology 2004, 96, 579–587 doi:10.1111/j.1365-2672.2004.02193.x Internal transcribed spacer 2 amplicon as a molecular marker for identification of Peronospora parasitica (crucifer downy mildew) S. Casimiro1,2, M. Moura1,2, L. Zé-Zé2, R. Tenreiro2 and A.A. Monteiro1 1 Instituto Superior de Agronomia, Universidade Técnica de Lisboa, Tapada da Ajuda, Lisboa, Portugal, and 2Departamento de Biologia Vegetal and Centro de Genética e Biologia Molecular, Faculdade de Ciências da Universidade de Lisboa, Campo Grande, Lisboa, Portugal 2002/0250: received 23 June 2003, revised 20 July 2003 and accepted 25 November 2003 ABSTRACT S . C A S I M I R O , M . M O U R A , L . Z É - Z É , R . T E N R E I R O A N D A . A . M O N T E I R O . 2004. Aims: The purpose of the study was to characterize the internal transcribed spacer (ITS) regions of Peronospora parasitica (crucifer downy mildew) in order to evaluate their potential as molecular markers for pathogen identification. Methods and Results: PCR amplification of ribosomal RNA gene block (rDNA) spacers (ITS1 and ITS2) performed in 44 P. parasitica isolates from different Brassica oleracea cultivars and distinct geographic origins, revealed no length polymorphisms. ITS restriction analysis with three endonucleases, confirmed by sequencing, showed no fragment length polymorphisms among isolates. Furthermore, ITS amplification with DNA isolated from infected host tissues also allowed the detection of the fungus in incompatible interactions. The combination of the universal ITS4 and ITS5 primers, for amplification of full ITS, with a new specific forward internal primer for ITS2 (PpITS2F), originates a P. parasitica specific amplicon, suitable for diagnosis. Conclusions: As ITS2 regions of P. parasitica, B. oleracea, other B. oleracea fungal pathogens and other Peronospora species are clearly distinct, a fast and reliable molecular identification method based on multiplex PCR amplification of full ITS and P. parasitica ITS2 is proposed for the diagnosis of crucifer downy mildew. Significance and Impact of the Study: The method can be applied to diagnose the disease in the absence of fungal reproductive structures, thus being useful to detect nonsporulating interactions, early stages of infection on seedlings, and infected young leaves packed in sealed plastic bags. Screening of seed stocks in sanitary control is also a major application of this diagnostic method. Keywords: ARDRA, ITS, molecular identification, Oomycetes, Peronospora parasitica, Peronosporales, rDNA. INTRODUCTION Peronospora parasitica is an exclusively biotrophic oomycete responsible for crucifer downy mildew, one of the most important diseases of brassica crops worldwide (Channon 1981). Brassicas are economically very important all over the world, Brassica oleracea being the most cultivated species in Correspondence to: Roge´rio Tenreiro, Departamento de Biologia Vegetal, Faculdade de Cieˆncias da Universidade de Lisboa, R. Ernesto de Vasconcelos, Edificio C2, Piso 4, Campo Grande, 1749-016 Lisboa, Portugal (e-mail: rptenreiro@fc.ul.pt). ª 2004 The Society for Applied Microbiology the Western Hemisphere, whereas B. campestris predominates Asia (Monteiro and Lunn 1999). Between 1989 and 1991, world brassica crops represented 16 69 000 ha, increasing to 19 83 000 ha in 1999 (FAO 1999). Although field plants are also severely affected, crucifer downy mildew damages are particularly important in the nursery, when the infection may kill the seedlings, retard their development or cause lack of uniformity and quality (Coelho et al. 1999). As the identification of P. parasitica is based on morphological characteristics of conidia and conidiophores 580 S . C A S I M I R O ET AL. (Dickinson and Greenhalgh 1977), downy mildew diagnosis on infected seedlings is delayed until sporulation occurs. The ineffectiveness of conventional identification methods is also a major concern in incompatible interactions with resistant hosts, where no pathogen reproduction occurs but tissue damage is observed (Leckie et al. 1999), and in the detection of infected cabbage seeds, embedding intact or germinating oospores, mycelium and conidia (Kluczewski and Lucas 1983; Badul and Achar 1998). Conventional methods also cannot detect infected young brassica leaves packed in sealed plastic bags, which may develop sporulating lesions before reaching the consumer. Therefore, new sensitive and reliable diagnostic methods are needed to reduce seedling losses, detect pathogen reservoirs and perform an efficient sanitary control. In fungal genomes, the highly conserved rRNA genes are separated by two less conserved internal transcribed regions, the internal transcribed spacers 1 and 2 (ITS1 and ITS2), which are therefore suitable for polymorphism studies among species or even at infra-specific level (Duncan et al. 1998; Mills et al. 1998). Amplified ribosomal DNA restriction analysis (ARDRA), using PCR primers based on conserved regions of the rRNA genes (White et al. 1990), followed by restriction with frequently cutting endonucleases, allows the easy assessment of sequence differences in ITS regions without length polymorphisms (Buscot et al. 1996; Lanfranco et al. 1998). Although fungal ITS1 has been shown to be more polymorphic at sequence level than ITS2 (Duncan et al. 1998), analysis of P. parasitica ITS1 sequences showed no differences among isolates collected from the same host species (B. oleracea and A. thaliana) and only 85% similarity between isolates from different hosts (Rehmany et al. 2000). In fact, host range must be associated with genetic differences of the isolates, classified as belonging to the same species, and these data point to the potential of ITS regions as molecular markers, both at species and forma specialis levels. There are, however, no available data on ITS2 variability and taxonomic relevance. The objective of this work was to characterize the ITS regions of P. parasitica isolates, from different B. oleracea crops and distinct geographic origins, in order to evaluate their potential as molecular markers for identification purposes. M A T E R I A LS A N D M E T H O D S P. parasitica isolates and conidia isolation Forty-four P. parasitica isolates, collected from B. oleracea plants, were grown on seedlings of B. oleracea hosts
1 or
2 (Table 1). For each isolate, ca 50 1-week-old seedlings were inoculated with two droplets of a conidia suspension (5 · 104conidia ml)1) per cotyledon and maintained in the dark, at 16C for 24 h. Then the inoculated seedlings were transferred to a growth room and maintained at 20 ± 1C, under a 20-h photoperiod. After 6 days of incubation, the seedlings were transferred to a dark room for 24 h, to induce sporulation. Cotyledons with sporulation were harvested and shaken in 50 ml of sterile distilled water to dislodge conidia. The conidial suspension was gauze filtered and centrifuged at 2600 g for 3 min. The pellet was resuspended in 7Æ5 ml of sterile distilled water, aliquoted in 1Æ5-ml fractions and stored at )20C until use. B. oleracea fungal pathogens selection and growth Ten isolates of B. oleracea fungal pathogenic species or related species of the same genera were selected, namely Fusarium culmorum, Trichoderma sp., Alternaria sp., Phoma sp., Phytophtora cinnamomi, Sordaria sp., from our collection, and from CECT (Collecion Espanola de Cultivos Tipo) F. oxysporum (CECT 2154), Scerotinia sclerotiorum (CECT 2882), Mycosphaerella tassiana (CECT 2665) and Diaporthe phaseolorum (CECT 2022). All fungi were grown on potato dextrose agar medium, with exception of Ph. cinnamomi which was grown on corn meal agar, at 28C for 7 days. DNA isolation DNA was isolated using an adaptation of the Ferreira and Grattapaglia (1995) method. An aliquot of each P. parasitica conidial suspension was centrifuged at 6400 g for 3 min. The pellet, or 100 mg of each fungal mycelium (obtained by colony scraping), was macerated with 200 ll of glass beads (425–600 microns), and 500 ll of extraction buffer (CTAB 2%, 1Æ4 mol l)1 NaCl, 0Æ02 mol l)1 EDTA, 0Æ01 mol l)1 Tris-HCl pH 8Æ0, 1% PVP, 0Æ2% b-mercaptoethanol, 0Æ1% Proteinase K), at 65C, were added. The suspension was incubated at 65C for 45 min, with mixing by inversion each 15 min. After cooling to room temperature, 500 ll of chloroform : isoamyl alcohol (24 : 1) were added, the tube was mixed by inversion and centrifuged at 16 700 g for 10 min. The upper aqueous phase was collected and the DNA was precipitated with 600 ll of isopropanol ()20C) for 1 h at )70C. After a 10-min centrifugation at 16 700 g, the pellet was washed with 500 ll of washing buffer (ethanol 70%, 0Æ15 mol l)1 NaCl) and centrifuged at 16 700 g for 5 min. The pellet was re-suspended in 25 ll of TE (0Æ01 mol l)1 Tris-HCl pH 8Æ0, 0Æ001 mol l)1 EDTA) and stored at 4C until utilization. After maceration with liquid nitrogen and using the method above, DNA was extracted from 100 mg of shortcycle B. oleracea CrGC3Æ1 and cabbage
Coração-de-boi seedling tissue, which were not infected with P. parasitica, ª 2004 The Society for Applied Microbiology, Journal of Applied Microbiology, 96, 579–587, doi:10.1111/j.1365-2672.2004.02193.x A MOLECULAR MARKER OF P. PARASITICA Table 1 Peronospora parasitica isolates used in this study 581 Isolate Original host crop type (Brassica oleracea) Geographical origin Lab host* P501, P522 P502 Tronchuda cabbage Tronchuda cabbage 1 1 P503 Kale cabbage P504 Tronchuda cabbage P505 P506 P507 P508, P515 P509 P510 P511 P512, P516 Tronchuda cabbage Unknown Unknown Tronchuda cabbage Tronchuda cabbage Tronchuda cabbage Tronchuda cabbage Cauliflower, Broccoli, Tronchuda cabbage Tronchuda cabbage Galega kale Cauliflower, Broccoli and Tronchuda cabbage
Murciana Tronchuda cabbage and Broccoli Cauliflower Broccoli Broccoli, Tronchuda cabbage Tronchuda cabbage
Murciana Broccoli, Tronchuda cabbage
Murciana Broccoli Tronchuda cabbage
Murciana Unknown Tronchuda cabbage Batalha, Portugal Póvoa do Varzim, Portugal Oliveira do Hospital, Portugal Castelo Branco, Portugal Évora, Portugal Odemira, Portugal Batalha, Portugal Condeixa, Portugal Vila Real, Portugal Faro, Portugal Lourinhã, Portugal Batalha, Portugal 1 2 2 2 2 2 2 2 Pombal, Portugal Condeixa, Portugal Batalha, Portugal 2 2 1 Batalha, Portugal 2 Batalha, Portugal Batalha, Portugal Batalha, Portugal 2 1 1 Batalha, Portugal 1 Batalha, Portugal 1 Batalha, Portugal Batalha, Portugal 2 2 German Ameal, Coimbra, Portugal Ameal, Coimbra, Portugal Casconha, Coimbra, Portugal Casconha, Coimbra, Portugal Coimbra, Portugal Coimbra, Portugal Eira Pedrinha, Coimbra, Portugal France HRI-England 1 2 P513 P514 P517 P518 P519 P520, P521, P527 P523 P524 P525 P526 P528 P529 P531 P532 P533, P534 Tronchuda cabbage
Algarvia Tronchuda cabbage P535 cabbage
Coração-de-boi P536 P537, P538 P539 Broccoli Broccoli Broccoli FP06, FP09 P005a, P005b, P006a, P006b Unknown Cauliflower *Host 1 – CrGC3Æ1, short-cycle B. oleracea, Crucifer Genetics Cooperative, University of Wisconsin, Madison, WI, USA; Host 2 – Cabbage
Coração-de-boi (B. oleracea). Mixture of isolates collected from different hosts. ª 2004 The Society for Applied Microbiology, Journal of Applied Microbiology, 96, 579–587, doi:10.1111/j.1365-2672.2004.02193.x 2 2 2 2 2 1 2 2 1 1 582 S . C A S I M I R O ET AL. and from 100 mg of Tronchuda cabbage
Algarvia seedling tissue, either infected with the isolate P501 or uninfected. using the arithmetic average and standard error of the 44 isolates. For host ITS regions, molecular size estimations were based on two replications. ITS amplification To amplify ITS1, ITS2 and full ITS regions, either from the fungus or the hosts, the following primers were used (White et al. 1990): ITS2 and ITS5 to amplify ITS1 region; ITS3 and ITS4 to amplify ITS2 region; and ITS4 and ITS5 to amplify full ITS. Each reaction mixture contained 2 ll DNA, PCR buffer 1X (GibcoBRL, Paisley, UK), 0Æ0025 mol l)1 MgCl2, 0Æ05% W1 (GibcoBRL), 0Æ0002 mol l)1 of each dNTP (GibcoBRL), 0Æ001 mol l)1 of each primer and 2 U of Taq DNA Polymerase (GibcoBRL), in a final volume of 50 ll. To each PCR tube, ca 50 ll of mineral oil were added and amplification occurred in a RoboCycler 96 (Stratagene, La Jolla, CA, USA), according to the following amplification programme: 4 min at 95C; 35 cycles of 1 min at 95C, 1 min at 56C and 2 min at 72C; 4 min at 72C. Each reaction sample was run on a 1Æ5% agarose gel, in 0Æ5 X TBE (0Æ05 mol l)1 Tris, 0Æ045 mol l)1 boric acid, 0Æ001 mol l)1 EDTA) at 90 V for 2 h 30 min, using 1 kb Plus standard (GibcoBRL) as molecular size marker. After ethidium bromide staining, the gels were analysed with KODAK 1D 2Æ0 software (GibcoBRL). For each isolate or host ITS regions, amplification was performed two to three times in order to assess the reproducibility of the method. The molecular sizes of P. parasitica ITS regions were estimated using the arithmetic average and standard error of the 44 isolates. Molecular sizes of host ITS regions were calculated as the average value of two replicates. ITS restriction assay To perform restriction digestion of amplified ITS regions, 10 ll samples of each PCR product, not purified, were digested with 5 U of each one of three restriction endonucleases, RsaI (Nbl, Northumberland, UK), HaeIII (Biolabs, Beverly, MA, USA) and Sau3AI (GibcoBRL), in a final volume of 15 ll, according to manufacturer instructions. After a 3-h incubation period at 37C, 1Æ5 ll of bromophenol blue solution (0Æ25% bromophenol blue, 0Æ25% xylene cyanol, 15% Ficoll in water) was added to each sample to stop the reaction. Each reaction sample was run on a 1Æ5% agarose gel, in 0Æ5 X TBE at 90 V for 3 h, using a 100 bp standard (GibcoBRL) as a molecular size marker. After ethidium bromide staining, the gels were analysed with KODAK 1D 2Æ0 software. Reproducibility of the method was assessed with duplicate reactions. Molecular sizes of individual restriction fragments produced from P. parasitica ITS regions were estimated ITS2 sequencing In order to sequence the ITS2 region, 15 ll of the PCR reaction of isolate P524 were run in 1% agarose gel, in 0Æ5 X TBE at 90 V for 1 h 30 min. After ethidium bromide staining, the ITS2 band was extracted with a sterile scalpel and purified with the Concert Rapid Gel Extraction Systems kit (GibcoBRL). The purified product was cloned using the pGEM-T Easy Vector Systems kit (Promega, Madison, WI, USA), with the following adaptations: 3 ll of the purified PCR product in the ligation reaction; JM109 competent cells, after inoculation in TSS medium (1 X LB (1% tryptone, 0Æ5% yeast extract, 0Æ5% NaCl, pH 7Æ0), 10% PEG 6000, 5% DMSO, 0Æ05 mol l)1 MgSO4, pH 6Æ5); SOC medium replaced by LB medium in JM109 transformation. The recombinant cells were plated in LB medium with 0Æ15 g l)1 ampicilin, 0Æ04 g l)1 IPTG and 0Æ04 g l)1 XGAL. Screening of recombinant white colonies was performed after an overnight incubation of each colony in 2 ml of LB medium with 0Æ15 g l)1 ampicilin. Each cell suspension was centrifuged at 18 000 g for 1 min and the pellet was resuspended in 150 ll TEG (0Æ05 mol l)1 glucose, 0Æ025 mol l)1 Tris-HCl, 0Æ01 mol l)1 EDTA, pH 8Æ0), followed by the addition of 200 ll 0Æ2 mol l)1 NaOH, 1% SDS. The suspension was mixed by inversion and chilled on ice, 200 ll 3 mol l)1 potassium acetate (pH 4Æ8) were added and the suspension was centrifuged at 18 000 g for 10 min. To the collected upper phase, 500 ll of isopropanol ()20C) were added. After a 30 min centrifugation at 18 000 g, the pellet was washed with 500 ll of 70% ethanol and centrifuged at 18 000 g for 5 min. The final pellet was resuspended in 50 ll TE with RNase (0Æ05 g l)1). Restriction analysis of putative recombinants with the endonuclease PvuII (Biolabs) occurred for 2 h at 37C, according to the manufacturer instructions, in a final volume of 30 ll. Restriction products were resolved by electrophoresis in a 1% agarose gel, in 0Æ5 X TBE at 90 V for 1 h 30 min. The gel was stained with ethidium bromide and fragment molecular size was estimated with KODAK 1D 2Æ0 software. Recombinant colonies containing the insert were reinoculated in LB medium with 0Æ15 g l)1 ampicilin and incubated overnight at 37C. Recombinant plasmid DNA was extracted with the Concert High Purity Plasmid Miniprep System kit (GibcoBRL). Sequencing was performed using the CEQ2000 Dye Terminator Cycle Sequencing kit (Beckman, Fullerton, CA, USA) and a capilary electrophoresis CEQ2000-XL (Beck- ª 2004 The Society for Applied Microbiology, Journal of Applied Microbiology, 96, 579–587, doi:10.1111/j.1365-2672.2004.02193.x A MOLECULAR MARKER OF P. PARASITICA man) sequencer, both directly from purified PCR product and from the cloned fragment. Both DNA strands were sequenced, with the primers T7 and SP6. BLASTN (Altschul et al. 1997) of the two sequences was performed in the GenBank database. Internal primer design and multiplex PCR ITS sequences of Peronospora (26 from P. parasitica and 15 from other Peronospora spp.), Albugo (5), Botrytis (2), Alternaria (3), Leptosphaeria (2), Plasmodiophora (4), Fusarium (3), Cladosporium (2), Trichoderma (1), Phoma (1), Diaporthe (1), Phytophthora (1), Sclerotinia (1), Mycospharella (2) and B. oleracea (4) available in the GenBank database (http://www.ncbi.nlm.nih.gov) were aligned with hierarchical clustering (Corpet 1988) at INRA website (http://prodes.toulouse.fr/multialign/multialign.html). Based on this alignment, internal primers were designed for specific amplification of full ITS and ITS2 regions of P. parasitica: PpITS1F (5¢-CAAYTWTAATTGGGGG TCGTGATCTT-3¢), PpITS2F (5¢-AAGCGTGACG ATACTAATTTG-3¢) and PpITS2R (5¢-TGAAGTG CGGCCGAAGCTT-3¢. Three multiplex PCR amplifications were performed using the following combinations of primers: ITS3 and ITS4 (to amplify any ITS2 region) plus PpITS2F and PpITS2R (to specifically amplify P. parasitica ITS2 region); ITS5 and ITS4 (to amplify any full ITS region) and PpITS1F and PpITS2R (to specifically amplify P. parasitica full ITS region); and ITS5 and ITS4 plus PpITS2F to amplify any full ITS region and P. parasitica specific ITS2 region. Selectivity of internal primers was tested with samples corresponding to uninfected
Algarvia cabbage; the same host infected with P. parasitica; the infected host DNA combined with Alternaria sp. and Ph. cinnamomi DNA; P. parasitica DNA free from host DNA contamination and each one of the 10 B. oleracea fungal pathogens. Each reaction mixture contained 2 ll DNA, PCR buffer 1X (GibcoBRL), 0Æ0025 mol l)1 MgCl2, 0Æ05% W1 (GibcoBRL), 0Æ0004 mol l)1 of each dNTP (GibcoBRL), 0Æ001 mol l)1 of each primer and 2 U of Taq DNA Polymerase (GibcoBRL), in a final volume of 50 ll. To each PCR tube, ca 50 ll of mineral oil were added and amplification occurred in a RoboCycler 96 (Stratagene), according to the following amplification programme: 4 min at 95C; 35 cycles of 1 min at 95C, 1 min at 54C and 2 min at 72C; 4 min at 72C. Each reaction sample was run on a 1Æ5% agarose gel, in 0Æ5 X TBE (0Æ05 mol l)1 Tris, 0Æ045 mol l)1 boric acid, 0Æ001 mol l)1 EDTA) at 90 V for 2 h 30 min, using 1 kb Plus standard as molecular size marker. After ethidium bromide staining, the gels were analysed with KODAK 1D 2Æ0 software. 583 For each sample, amplification was performed two to three times in order to assess the reproducibility of the method. Molecular sizes of ITS regions were calculated as the average value of two replications. RESULTS ITS analysis The amplification of ITS1, ITS2 and full ITS regions (Fig. 1) revealed a common PCR product for all the 44 P. parasitica isolates with 323 ± 0Æ9 bp, 684 ± 2Æ1 bp and 987 ± 3Æ0 bp, respectively. Isolates P512, P516, P517, P518, P523 and P525, representing mixtures collected from different hosts, also have equivalent mean amplicon sizes. Other amplification products, usually in lower abundance, were observed in most of the isolates (Fig. 1). The number of these additional amplicons was higher for ITS1, pointing to a better specificity of the primers used to amplify the ITS2 region. Comparison of amplification reactions of isolates with amplifications from B. oleracea noninfected seedlings allowed the identification of host ITS amplicons among the additional products. PCR products of ITS1 and (a) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 (b) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 (c) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Fig. 1 Internal transcribed spacer amplification profiles of Peronospora parasitica isolates. (a) ITS1. (b) ITS2. (c) Full ITS. Lanes 1, 16: 1 kb Plus DNA ladder. Lanes 2–15: isolates P501, P502, P505, P517, P519, P520, P521, P522, P523, P524, P525, P526, P527 and P528 ª 2004 The Society for Applied Microbiology, Journal of Applied Microbiology, 96, 579–587, doi:10.1111/j.1365-2672.2004.02193.x 584 S . C A S I M I R O ET AL. ITS2 regions of plant hosts had the same length that was estimated as 388 ± 0Æ1 bp and 380 ± 0Æ1 bp for CrGC3Æ1 and cabbage
Coração-de-boi, respectively. Total ITS of CrGC3Æ1 and cabbage
Coração-de-boi was estimated as 756 ± 0Æ1 bp and 740 ± 0Æ1 bp, respectively. Nevertheless, when host and P. parasitica ITS regions were co-amplified, their distinction was evident for ITS2 and full ITS (Fig. 1), because of significant differences of molecular sizes. As the biotrophic nature of P. parasitica prevents the use of axenic cultures, the remaining unidentified products presumably resulted from secondary pathogens infecting the seedlings after tissue necrosis. In fact, the presence of biological contaminants in conidial suspensions was detected by microscopic observation during this work. ARDRA analysis The restriction profiles of ITS regions, obtained with the three endonucleases (Table 2), showed that only Sau3AI recognized a restriction sequence in P. parasitica ITS1 region, producing two fragments. Conversely, all three enzymes recognized two restriction sites in the ITS2 region. No restriction fragment length polymorphisms were observed in ITS regions among P. parasitica isolates. To reinforce these results and to contribute to restriction site location within the internal transcribed spacers, ARDRA analysis was also performed with full ITS (Table 2). Although a small variation in fragment molecular size estimations was observed, ITS1 and ITS2 restriction profiles were confirmed and the majority of restriction sites could be located in the physical map displayed in Fig. 2. Only the relative order of the two 3¢ terminal RsaI and HaeIII fragments could not be assessed. Beyond the fragments resulting from P. parasitica ITS restriction, others could be seen after electrophoresis. Some fragments resulted from partial or incomplete restriction, and provided an additional tool to locate the restriction sites within the ITS regions. ARDRA analysis of brassica ITS regions (Table 2) also confirmed the host origin of other fragments. The few remaining unidentified fragments, less frequent in ITS2 restriction analysis, possibly resulted from the additional amplicons referred in ITS analysis. ITS2 sequence of isolate P524 Although sequencing was directly attempted from the PCR product, fully reliable sequence data were only obtained from the cloned PCR product. In this case, a product of 684 bp was sequenced for both DNA strands. Homology sites for ITS3 and ITS4 primers were detected in the 3¢–5¢ and 5¢–3¢ DNA strands, respectively, and the BLASTN (Altschul et al. 1997) of both sequences revealed a total complementarity between them. Homology search against 16 available GenBank sequences of Peronospora spp. (one from P. parasitica, eight from P. sparsa, two from P. destructor and one from P. farinosa, P. rumicis, P. niessleana, P. arborescens and P. manshurica) showed that nucleotides 1–128 correspond to the 5Æ8S rRNA gene terminal sequence, nucleotides 129–624 represent the complete ITS2 spacer sequence (496 bp) and nucleotides 625–684 correspond to the 28S rRNA gene initial sequence. The complete sequence of the ITS2 of P. parasitica, isolate Table 2 Internal transcribed spacer (ITS) restriction fragments obtained with the endonucleases RsaI, HaeIII and Sau3AI for Peronospora parasitica and Brassica oleracea CrGC3Æ1 and cabbage
Coração-de-boi Brassica oleracea Peronospora parasitica CrGC3Æ1 Cabbage
Coração-de-boi Enzyme ITS1 (bp) ITS2 (bp) Total ITS (bp) ITS1 (bp) ITS2 (bp) ITS1 (bp) ITS2 (bp) RsaI No restriction No restriction No restriction No restriction No restriction 85 ± 0Æ8 238 ± 1Æ8 48 127 213 68 320 66 109 213 40 348 37 83 260 54 142 184 100 ± 0Æ1 180 ± 0Æ1 Sau3AI 157 285 545 71 248 668 71 85 223 608 194 ± 0Æ1 HaeIII 157 242 285 71 248 365 71 223 390 ± ± ± ± ± ± ± ± ± 0Æ5* 0Æ6 0Æ6 1Æ5 4Æ5 3Æ7 0Æ9 1Æ3 1Æ0 ± ± ± ± ± ± ± ± ± ± 0Æ6 0Æ7 1Æ1 0Æ8 0Æ3 0Æ7 0Æ6 1Æ3 1Æ3 2Æ3 ± ± ± ± ± 0Æ1 0Æ1 0Æ1 0Æ1 0Æ1 ± ± ± ± ± 0Æ1 0Æ1 0Æ1 0Æ1 0Æ1 ± ± ± ± ± ± 0Æ1 0Æ1 0Æ1 0Æ1 0Æ1 0Æ1 36 ± 0Æ1 344 ± 0Æ1 *Values refer to average ± S.E. The number of determinations was 44 for P. parasitica and two for CrGC3Æ1 and cabbage
Coração-de-boi. Double co-migrating fragments. ª 2004 The Society for Applied Microbiology, Journal of Applied Microbiology, 96, 579–587, doi:10.1111/j.1365-2672.2004.02193.x A MOLECULAR MARKER OF P. PARASITICA ITS1 spacer 219 bp ITS2 spacer 496 bp S Fig. 2 Physical map of Peronospora parasitica rDNA cluster. R ¼ RsaI restriction sites (positions 242 and 527 within ITS2 amplicon); H ¼ HaeIII restriction sites (positions 365 and 436 within ITS2 amplicon); S ¼ Sau3AI restriction sites (position 85 within ITS1 amplicon; positions 390 and 461 within ITS2 amplicon) R HS R 25–28S ITS2 amplicon (684 bp) ITS1 amplicon (323 bp) P524, is available in the GenBank database (accession number AY029235). The ITS2 sequence obtained, confirmed the ARDRA results and the physical map presented in Fig. 2. H S 5.8S 17–18S 585 Total ITS amplicon (987 bp) (a) 1 2 3 4 5 6 (b) 1 2 3 4 5 6 ARDRA analysis in incompatible host–pathogen interactions Cabbage
Algarvia is resistant to P. parasitica isolate P501, despite some growth of intercellular mycelium. This incompatible host–pathogen system was selected to amplify ITS2 and full ITS, using DNA isolated from a small amount of cabbage tissue both infected and not infected. Amplification of P. parasitica ITS regions could be achieved from infected tissue, whereas amplification of host ITS regions occurred in both cases (Fig. 3a). As ITS regions of cabbage
Algarvia and CrGC3Æ1 have similar sizes, distinction of pathogen and host products was also more evident with ITS2 amplification. ARDRA with RsaI applied to these PCR samples (Fig. 3b) revealed that this enzyme does not recognize any sequence within ITS regions of cabbage
Algarvia, as observed with cabbage
Coração-de-boi, and produced the expected restriction fragments from P. parasitica. Multiplex PCR The amplification of ITS2 region with primers ITS3 and ITS4, simultaneously with primers PpITS2F and PpITS2R, revealed the expected amplicons of host and P. parasitica, with 380 bp and 684 bp, respectively. Also, as expected was the internal PCR product with 381 bp, amplified only in P. parasitica. The molecular sizes of ITS amplicons of the other samples matched the expected values (data not shown). The amplification of full ITS with primers ITS5 and ITS4, simultaneously with primers PpITS1F and PpITS2R, also revealed the expected amplicons of host and P. parasitica, with 760 bp and 987 bp, respectively. The internal PCR product, specific for P. parasitica, presented the predicted molecular size of 875 bp. As for ITS2 region, Fig. 3 (a) Amplification of internal transcribed spacer 2 and full ITS from Tronchuda cabbage
Algarvia infected with Peronospora parasitica isolate P501 (lanes 2 and 4, respectively) and from noninfected tissues of the same Brassica oleracea (lanes 3 and 5, respectively). P. parasitica amplicons are indicated with arrows. Lanes 1, 6: 100 bp DNA ladder. (b) Amplified ribosomal DNA restriction analysis profiles with RsaI, from noninfected Tronchuda cabbage
Algarvia (lane 2: ITS2; lane 4: full ITS) and the same B. oleracea infected with P. parasitica isolate P501 (lane 3: ITS2; lane 5: full ITS). Lanes 1, 6: 100 bp DNA ladder the molecular sizes of full ITS amplicons of the other samples matched the expected values (data not shown). The multiplex PCR with three primers (ITS5, ITS4 and PpITS2F) also revealed the expected full ITS amplicons of host and P. parasitica, with 760 bp and 987 bp, respectively (Fig. 4). The internal PCR product, specific for P. parasitica, presented the predicted molecular size of 410 bp. ITS amplicons obtained for the other samples corresponded only to the full ITS regions, matching the expected values. DISCUSSION No ITS length polymorphisms were observed in P. parasitica isolates collected on B. oleracea hosts. As shown by the standard errors associated with each amplicon, the observed ª 2004 The Society for Applied Microbiology, Journal of Applied Microbiology, 96, 579–587, doi:10.1111/j.1365-2672.2004.02193.x 586 S . C A S I M I R O ET AL. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Fig. 4 Multiplex PCR with three primers. Lanes 1 and 16: 1 kb Plus DNA ladder. Lanes 2 and 3: Tronchuda cabbage
Algarvia, not infected and infected with Peronospora parasitica isolate P501, respectively. Lane 4: host DNA (from Tronchuda cabbage
Algarvia) mixed with Alternaria sp. and Phytophtora cinnamomi DNA. Lane 5: P. parasitica isolate P501. Lanes 6–15: other Brassica oleracea fungal pathogens (Alternaria sp., Phytophthora cinnamomi, Fusarium culmorum, Trichoderma sp., Phoma sp., Sordaria fimicola, Sclerotinia sclerotiarum, F. oxysporum, Mycosphaerella tassiana and Diaporthe phaseolorum, respectively) variation is within electrophoresis variability as all the samples could not be observed in a single gel. The ITS homogeneity was also evidenced by mixed isolates (e.g. P512) that showed an unique amplicon for each ITS region. The molecular size of ITS1 region of P. parasitica isolates was identical to the one described by Rehmany et al. (2000). In fact, these authors refer to four of the isolates used in this work, namely P005 (AF241754), P006 (AF241755), P501 (AF241762) and P502 (AF241763). ITS1 ARDRA analysis revealed comparable restriction profiles in all isolates. Based on ARDRA results, validated by ITS1 sequencing (Rehmany et al. 2000), we can assume that ITS1 sequence is quite conserved among all P. parasitica isolates pathogenic to B. oleracea. No length or restriction polymorphisms were detected in ITS2 as the ARDRA analysis revealed the same profiles in all isolates. The sequence of ITS2 region obtained in this work (AY029235) confirmed the restriction profiles of the three enzymes used. The comparison of this sequence with ITS2 spacer sequences of isolates from other crucifers will be helpful in considering ITS regions as being distinct among formae speciales. Comparison of GenBank database available sequences shows that the ITS2 spacer is larger in P. parasitica than in other Peronospora species. Eight isolates from P. sparsa have an ITS1 spacer with 217 bp and an ITS2 spacer with 423– 425 bp (Cooke et al. 2000), P. manshurica ITS1 spacer has 219 bp whereas ITS2 spacer has 423 bp (GenBank accession number AB021711) and P. destructor ITS1 and ITS2 spacers have 220 bp and 421 bp, respectively (AB021712). As P. parasitica isolates from B. oleracea have an ITS1 spacer with 219 bp (Rehmany et al. 2000) and an ITS2 spacer with 496 bp, the size of ITS1 spacer seems to be more conserved within the genus and ITS2 spacer could be used as a molecular marker for P. parasitica. The availability of a molecular marker based on ITS2 spacer provides a PCR-based method that can be applied as a reliable identification tool of P. parasitica, allowing its detection in early stages of crucifer downy mildew in seedlings and field plants, and in the screening of seed stocks. The method can be very useful to detect infected young brassica leaves, which do not show any damaging lesions at the time of packing in sealed plastic bags, but may develop sporulating lesions under high humidity and darkness inside the bags before reaching the consumer. Moreover, this procedure will open up silent reservoirs of this pathogen both in the period from infection to conidia release and in interactions with resistant hosts, where no morphological identification characteristics are available for diagnosis. As an example, in incompatible reactions of P. parasitica with B. oleracea cabbage
Algarvia, only superficial lesions are observed. The P. parasitica mycelium grows in the intercellular spaces of the host tissues but is unable to complete its life cycle and produce conidiophores and conidia. The lesions resulting from this infection are not diagnosed, but the PCR amplification of ITSs directly from DNA isolated from the infected tissue, as shown in this work, allows the identification of the pathogen. Although infected plant tissue is used, the size difference between host and P. parasitica ITS regions is sufficiently distinct, particularly for ITS2. Despite the starting biological material for DNA extraction, conidia or host tissues, it is quite dificult to ensure the absence of contamination by other microorganisms. In fact, P. parasitica multiplication is not performable in a sterile environment, because of growth composts and watering, and disease development is also prone to the appearance of secondary pathogens as a consequence of tissue necrosis. The presence of this kind of contaminant was previously detected and determined by Rehmany et al. (2000). They identified, by sequencing, co-amplified ITS products from Fusarium, Cladosporium and Alternaria species. Contaminant ITS spacers are more evident for ITS1 and full ITS, making their amplification profiles less clear, although P. parasitica amplicons, and sometimes the host ones, are more intense. However, in ARDRA analysis the contaminant restriction products are highly diluted when compared with P. parasitica or host products. Nevertheless we opted to include ITS2 and full ITS in the search for a specific PCR amplification method that enables a reliable detection of P. parasitica in contaminated samples. The design of internal primers was easier for ITS2 region as ITS1 region is more conserved and, therefore, has a higher consensus. Although the specificity was achieved with any studied multiplex PCR, the localization of the annealing region of the new specific primers governed the molecular size of resulting amplicons, affecting the discrimi- ª 2004 The Society for Applied Microbiology, Journal of Applied Microbiology, 96, 579–587, doi:10.1111/j.1365-2672.2004.02193.x A MOLECULAR MARKER OF P. PARASITICA nation ability within the genus Peronospora and, mostly, between P. parasitica and other Brassica pathogens. In fact, the P. parasitica specific amplicon of 381 bp, obtained in ITS2 multiplex PCR with four primers, was difficult to distinguish from the host ITS2 amplicon (380 or 388 bp) and very similar to the ITS2 amplicon of other Brassica fungal pathogens. Regarding full ITS multiplex PCR with four primers, distinction of the amplicons obtained for the host (740 or 756 bp) and P. parasitica (875 and 987 bp) may eventually be impaired by electrophoresis resolution. Another relevant aspect of multiplex PCR with four primers was the relative intensity of amplification products, as universal ITS primers seemed to be more effective than the designed internal specific primers. In order to overcome these problems and to maintain the ITS amplicons obtained with universal primers as positive controls in multiplex PCR, a third multiplex method was tried, that combines universal primers for full ITS amplification with one primer specific for P. parasitica ITS2 region. In this case, the amplification products are easily distinguishable, as the full ITS region of P. parasitica has, per se, a higher molecular size than the same region in B. oleracea and other studied Brassica pathogenic fungi and the specific P. parasitica ITS2 amplicon is significantly smaller (410 bp) than the above mentioned amplicons. Furthermore, the specific ITS2 amplicon will only be obtained in P. parasitica infected plants, whereas the host full ITS amplicon will always be observed, alone or in combination with full ITS amplicons from any infecting fungi (P. parasitica or others). In conclusion, we suggest the ITS multiplex PCR amplification, with primers ITS5, ITS4 and PpITS2F, as a reliable and simple method to identify P. parasitica in B. oleracea infected tissues, because of the larger size difference between the ITS 2 specific amplicon of P. parasitica and the full ITS amplicons obtained from the host and any Brassica fungal pathogen. ACKNOWLEDGEMENTS This work was partially supported by Fundação para a Ciência e Tecnologia (FCT), Portugal, Project POCTI/ 2000-AGR/33309/99. REFERENCES Altschul, S.F., Medden, T.L., Schäffer, A.A., Zhang, J., Miller, W. and Lipman, D.J. (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Research 25, 3389–3402. Badul, J. and Achar, P.N. (1998) Screening of Brassica oleracea var. capitata cultivars against Peronospora parasitica. Seed Science and Technology 26, 385–391. 587 Buscot, F., Wipf, D., Di Battista, C., Munch, J.C., Botton, B. and Martin, F. (1996) DNA polymorphism in morels: PCR/RFLP analysis of the ribosomal DNA spacers and microsatellite-primed PCR. Mycological Research 100, 63–71. Channon, A.G. (1981) Downy mildew of brassicas. In The Downy Mildews ed. Spencer, D.M. pp. 321–339. London: Academic Press. Coelho, P., Leckie, D., Astley, D., Crute, I.R., Bahccevandziev, K., Valério, L., Monteiro, A. and Buokema, I. (1999) The relationship between cotyledon and adult plant resistance to downy mildew (P. parasitica) in B. oleracea. Acta Horticulturae 459, 335–342. Cooke, D.E.L., Drenth, A., Duncan, J.M., Wagels, G. and Brasier, C.M. (2000) A molecular phylogeny of Phytophtora and related Oomycetes. Fungal Genetics and Biology 30, 17–32. Corpet, F. (1988) Multiple sequence alignment with hierarchical clustering. Nucleic Acids Research 16, 10881–10890. Dickinson, C.H. and Greenhalgh, J.R. (1977) Host range and taxonomy of Peronospora on crucifers. Transactions of the British Mycological Society 69, 111–116. Duncan, J.M., Cooke, D., Brich, P. and Toth, R. (1998) Molecular variability in sexually reproducing fungal plant pathogens. In Molecular Variability of Fungal Pathogens ed. Bridge, P.D., Couteaudier, Y. and Clarkson, J.M. pp. 19–39. Wallingford: CAB International. FAO (1999) Quarterly Bulletin of Statistics 12, 76–77. Ferreira, M.E. and Grattapaglia, D. (1995) Introduça˜o ao Uso de Marcadores RAPD e RFLP em Ana´lise Gene´tica. Brasilia: EMBRAPA-CENARGEN. Kluczewski, S.M. and Lucas, J.A. (1983) Host infection and oospore formation by P. parasitica in agricultural and horticultural Brassica species. Transactions of the British Mycological Society 81, 591–596. Lanfranco, L., Perotto, S., Longato, S., Mello, A., Cometti, V. and Bonfante, P. (1998) Molecular approaches to investigate biodiversity in mycorrhizal fungi. In Mycorrhiza Manual ed. Varma, A. pp. 353– 372. Berlin: Springer-Verlag. Leckie, D., Cogan, N., Astley, D., Crute, I.R., Boukema, I., Santos, M., Bahccevandziev, K., Silva-Dias, J.E. et al. (1999) Differential resistance to P. parasitica and Albugo candida in B. oleracea. Acta Horticulturae 459, 357–361. Mills, P.R., Sreenivasaprasad, S. and Muthumeenakshi, S. (1998) Assessing diversity in Colletotrichum and Trichoderma species using molecular markers. In Molecular Variability of Fungal Pathogens ed. Bridge, P.D., Couteaudier, Y. and Clarkson, J.M. pp. 105–120. Wallingford: CAB International. Monteiro, A. A. and Lunn, T. (1999) Trends and perspectives of vegetable Brassica breeding world-wide. Acta Horticulturae 459, 273–279. Rehmany, A.P., Lynn, J.R., Mahmut, T., Holub, E.B. and Beynon, J. (2000) A comparison of Peronospora parasitica (downy mildew) isolates from Arabidopsis thaliana and Brassica oleracea using Amplified Length Polymorphism and Internal Transcribed Spacer 1 sequence analyses. Fungal Genetics and Biology 30, 95–103. White, T., Brurns, T., Lee, S. and Taylor, J. (1990) Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In PCR Protocols: A Guide to Methods and Applications eds Innis, M.A., Gelfand, D.H., Snisky, J.J. and White, T.J. pp. 315–322. San Diego: Academic Press. ª 2004 The Society for Applied Microbiology, Journal of Applied Microbiology, 96, 579–587, doi:10.1111/j.1365-2672.2004.02193.x