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