«PCR-based identiﬁcation of Erysiphe pulchra and Phyllactinia guttata from Cornus ﬂorida using ITS-speciﬁc primers By M. T. Mmbaga1,4, N. B. ...»
For. Path. 34 (2004) 321–328
Ó 2004 Blackwell Verlag, Berlin
PCR-based identiﬁcation of Erysiphe pulchra and Phyllactinia
guttata from Cornus ﬂorida using ITS-speciﬁc primers
By M. T. Mmbaga1,4, N. B. Klopfenstein2, M.-S. Kim2 and N. C. Mmbaga3
Nursery Crop Research Station, Tennessee State University, McMinnville, TN 37110, USA; 2USDA
Forest Service, RMRS, 1221 S. Main St, Moscow, ID 83843, USA; 3Middle Tennessee State University,
Murfreesboro, TN 37132, USA; E-mail: firstname.lastname@example.org (for correspondence) Summary The internal transcribed spacer (ITS) regions of rDNA and the intervening 5.8S rRNA gene for the powdery mildew fungi Erysiphe (sect. Microsphaera) pulchra and Phyllactinia guttata were ampliﬁed using standard polymerase chain reaction (PCR) protocols and the universal primer pairs, ITS1 and ITS4.
PCR products for ITS were analysed by electrophoresis in a 1.5% agarose gel and sequenced. The size of the ampliﬁed ITS products (approximately 650 bp) were not sufﬁciently different to allow reliable differentiation of E. pulchra and P. guttata; however, their sequences were distinct. Speciﬁc primers for E. pulchra and P. guttata were developed and evaluated for use as diagnostic tools. The diagnostic band size from E. pulchra-speciﬁc primer pair was 568 bp while the P. guttata band was 597 bp; the two primer pairs were highly speciﬁc to E. pulchra and P. guttata. Comparison of ITS sequences with information in the GenBank showed a very close similarity between sequences of E. pulchra isolates from Cornus ﬂorida in the USA and isolates collected on Cornus kousa in Japan. BLAST analysis of the sequence of the 650-bp band from P. guttata revealed a close alignment with sequences of P. moricola (92%), P. kakicola (94%), and P. fraxini (92%). The sequence of P. guttata in C. ﬂorida also had a 98% identity with P. guttata in Calycanthus occidentalis and 94% identity with P. guttata in Corylus cornuta.
1 Introduction Powdery mildew is one of the most problematic diseases encountered during dogwood production (McRitchie 1994; Ranney et al. 1994; Hagan et al. 1995; Hagan and Mullen 1995; Hartman 1998; Mmbaga 1998; Daughtrey and Hagan 2001). The sudden emergence of the disease in the southeastern USA and the swift spread of the disease further northward is a cause of growing concern because of economic impacts on dogwood as a landscape plant and ecological impacts on dogwood as a food source for wildlife.
Although powdery mildew was previously believed to cause only superﬁcial damage (Hagan et al. 1997) with no lasting growth effects, recent studies demonstrated that this powdery mildew signiﬁcantly reduces plant growth (Mmbaga 1998). A study by Hartman (1998) showed that infected ﬂowering dogwoods showed reduced ﬂowerbud set and diminished blooms during the season following heavy infection. Only two fungi, Phyllactinia guttata (Wallr.) Lev. [syn. Phyllactinia corylea (Pers.) P. Karst.] and Erysiphe (sect. Microsphaera) pulchra (Cook & Peck, Braun & Takamatsu) have been associated with this disease (Leigh et al. 1998; Braun and Takamatsu 2000). Both pathogens produce similar symptoms, and signs of the disease during the growing season are produced by the anamorph (asexual) stage (Oidium sp. for E. pulchra and Ovulariopsis sp.
for P. guttata). Because the anamorph stages are similar for these pathogens, they are not easily distinguished during the time of plant infection and associated damage (Barnet and Hunter 1998). However, P. guttata and E. pulchra can be readily identiﬁed at the teliomorph stage because each produces very distinct ascocarps. Thus, the ascocarp is the Received: 20.10.2003; accepted: 19.07.2004; editor: C. G. Shaw U. S. Copyright Clearance Center Code Statement: 1437–4781/2004/3405–0321 $15.00/0 www.blackwell-synergy.com M. T. Mmbaga, N. B. Klopfenstein, M.-S. Kim and N. C. Mmbaga primary diagnostic feature used for identiﬁcation of these pathogens (Leigh et al. 1998).
Unfortunately, ascocarps are formed late in the growing season (September–November), and are not found during the growing season when disease symptoms are most prominent (Williamson and Blake 1999; Mmbaga 2002).
The distribution of these two pathogens in dogwood growing areas and their associated virulence are not well known. While E. pulchra and P. guttata have been observed in North Carolina and Kentucky (Hartman 1998), the occurrence of P. guttata is relatively rare in Tennessee where E. pulchra is the predominant powdery mildew pathogen (Mmbaga 2000, 2002).
To understand the relative roles of each fungus in disease severity or determine their prevalence in production areas, reliable techniques are needed to correctly identify these fungi at the conidial stage. The objectives of this study were to apply DNA molecular diagnostic techniques based on the internal transcribed spacer (ITS) regions of rDNA to: (1) distinguish P. guttata and E. pulchra, (2) compare DNA sequences of E. pulchra and P. guttata with other data in the GenBank, and (3) develop polymerase chain reaction (PCR)-based speciﬁc primers for diagnosis of P. guttata and E. pulchra during the anamorph stage.
2 Materials and methods
2.1 Powdery mildew sample collection and single conidiospore isolation Dogwood material infected with powdery mildew consisted of (a) samples from Tennessee and Kentucky containing conidiospores, (b) ascocarps of E. pulchra from Tennessee and Kentucky, and (c) ascocarps of P. guttata from Kentucky. Although, ascocarps of P. guttata and E. pulchra are very distinct (Fig. 1), conidiospores are similar morphologically. For this reason, single conidiospores were isolated from the ﬁeld and multiplied on detached leaves in vitro before DNA analysis. Ten ﬁeld isolates were collected from each of ﬁve counties (Cannon, Coffee, Franklin, Rutherford, and Warren) in Tennessee and one collection of mixed ﬁeld isolates was collected from Kentucky by Dr J. R. Hartman (Department of Plant Pathology Extension Service, University of Kentucky, Lexington, KY, USA).
Single conidial chains were selected for single-spore isolations under a dissecting microscope. Single conidiospore lines were established for each ﬁeld isolate collected, for a total of 50 single-spore isolates from Tennessee and 10 single-spore isolates from Kentucky.
Each isolate was cultured on the surface of dogwood leaves that were surface sterilized with 10% CloroxTM (sodium hypochlorite, Clorox company, Oakland, CA, USA) before placement in Petri dishes containing 1% water agar (WA). Cultures were maintained under ﬂuorescent lighting. Leaves used for cultures were collected from disease-free dogwoods (cv.
ÔCherokee PrincessÕ) that were maintained in a greenhouse. After 2–3 weeks, samples of mycelia were harvested from each culture using disposable micropipette tips, and placed in 0.5-ml microcentrifuge tubes. Samples were stored at )20°C until DNA extraction.
Leaf samples were collected from powdery mildew-infected trees during December and observed for the presence of P. guttata and E. pulchra ascocarps. On approximately 100 leaves collected in Tennessee, only E. pulchra was observed. However, leaf samples collected in Kentucky contained ascocarps of both fungi. After collection, all leaf samples were stored at 4°C until ascocarps were harvested.
Fig. 1. Morphology of Erysiphe pulchra and Phyllactinia guttata: (a) Ascocarp of E. (sect.
Microsphaera) pulchra; (b) dichotomously branched appendages of E. pulchra ascocarps; (c) ascospores of E. pulchra (approximately 20 lm in size); (d) larger ascocarp of P. guttata; (e) bulbous appendages of P. guttata ascocarps; (f) and P. guttata ascospore (approximately 30 lm in size) DNA samples were used directly for PCR reactions, they were held at )80°C until the PCR reagents were added and PCR cycles initiated. DNA samples not used immediately were stored at )20°C. The Lyse-N-GoTM reagent was also used to prepare DNA from samples containing 25–100 ascocarps; however, ascocarps were manually crushed and homogenized prior to each Lyse-N-GoTM procedure.
The universal primer pairs, ITS1 and ITS4, were used to amplify the internal transcribed spacer (ITS) region, including the intervening 5.8S rRNA gene (White et al. 1990). Each 50-ll reaction mixture contained the DNA template prepared by Lyse-N-GoTM reagent (or no DNA template for a negative control), 1 unit Taq polymerase (Applied Biosystems, Foster City, CA, USA), 4 mm MgCl2, 200 lm dNTPs, and 0.5 lm of each primer (ITS1 and ITS4). A Techne ProgeneTM (Techne Incorporated, Princeton, NJ, USA) thermal cycler was used. The cycling parameters were as follows: 94°C initial denaturation (5 min), M. T. Mmbaga, N. B. Klopfenstein, M.-S. Kim and N. C. Mmbaga followed by 48 cycles of denaturation at 94°C (35 s), annealing at 46°C (1 min 10 s), extension at 72°C (1 min 40 s), and one cycle of ﬁnal extension at 72°C (7 min) (Pimentel et al. 1998).
Polymerase chain reaction products were analysed by electrophoresis in a 1.5% agarose gel in 0.5X TBE [Tris–borate–ethylenediaminetetraacetic acid (EDTA)] buffer.
Gels were stained with ethidium bromide (0.5 lg/ml) and DNA was visualized using UV light. PCR products for ITS region (approximately 650 bp) were puriﬁed using ExoSAP-ITTM (USB Corporation, Cleveland, OH, USA) and sequenced using an ABI 377XL PRISM automatic sequencer (Applied Biosystems, Foster City, CA, USA). ITS sequences of E. pulchra and P. guttata were edited with BioEdit software (Hall 1999) and then compared with GenBank information using a BLAST search (Altschul et al.
Sequences of E. pulchra and P. guttata were used to develop speciﬁc primers for differentiating E. pulchra from P. guttata. The two sets of speciﬁc primers (forward and reverse: EP1, and EP2 for E. pulchra, and PG1 and PG2 for P. guttata) were prepared by the Iowa State University Biotechnology Services (Ames, IA, USA).
2.3 Evaluation of the speciﬁc primers designed for E. pulchra and P. guttata Speciﬁc primers for E. pulchra (EP1 and EP2) were evaluated for rDNA ampliﬁcation of E. pulchra and P. guttata DNA from dogwood. Mycelia and ascocarps were evaluated for E. pulchra, whereas only ascocarps for P. guttata were evaluated. All mycelial samples were collected from leaves of ﬁeld-grown trees. To test whether the primers will cross-amplify DNA from the two fungi, speciﬁc primers designed for P guttata (PG1 and PG2) were evaluated for ampliﬁcation of P. guttata and E. pulchra DNA. DNA extraction and ampliﬁcation protocols described above were used to evaluate speciﬁc primers. Because ITS1 and ITS4 successfully ampliﬁed both P. guttata and E. pulchra (650 bp), they were included as the positive control in the evaluation of P. guttata- and E. pulchra-speciﬁc primers. A negative control consisted of tissue from the leaf surface of dogwood, where powdery mildew signs were not present.
3 Results Cleistothecia of P. guttata and E. pulchra are very distinct (Fig. 1), and samples used in this study were from dry leaves collected during the previous year. Because powdery mildew spores are highly air-borne, single-spore isolation using detached leaves avoided contamination. Lyse-N-GoTM DNA preparation reagents and associated protocol produced suitable template DNA from very small amounts of mycelia or ascocarps (approximately 25–100) of E. pulchra or P. guttata. The use of very small amounts of mycelia directly from leaf tissue is of particular signiﬁcance in studies of obligate parasites, because obligate parasites cannot be grown in artiﬁcial media to produce sufﬁcient fungal material for DNA extraction.
Polymerase chain reaction products from the rDNA ITS region of E. pulchra and P. guttata were about 650 bp (Fig. 2). However, the DNA sequence of E. pulchra was distinctly different from that of P. guttata, with sequence differing by more than 50% (GenBank accession no.: AY224136 for E. pulchra and AY224137 for P. guttata). All single-spore isolates exhibited a similar-sized PCR product as that of E. pulchra ascocarps, and their ITS sequences were identical (data not shown). Thus, these mycelial samples were conﬁrmed to be E. pulchra. The ITS region sequence did not show any sequence variation between samples.
Because the sequence of P. guttata was distinctly different from that of E. pulchra, speciﬁc PCR primers could be designed for E. pulchra and P. guttata. Speciﬁc primers Powdery mildew pathogens of Cornus ﬂorida 325 Fig. 2. PCR-ampliﬁed products using ITS1 and ITS4 primers on Erysiphe (Sect. Microsphaera) pulchra and Phyllactinia guttata from Cornus ﬂorida. Lane M: Molecular size standards (bp; 1 kb plus DNA ladder), lanes 1 and 2: E. pulchra, lanes 3 and 4: P. guttata, lane 5: negative control (no DNA) designed for E. pulchra contained 18 forward bases 5¢-GTGAACCTGCGGAAGATC¢ (EP1) and 20 complementary bases (reverse) 5¢-CATGTGACTGGAACAAAAAG-3¢ (EP2). Speciﬁc primers for P. guttata had 18 forward bases 5¢-CTGAGCGTGAA GACTCTC-3¢ (PG1), and 18 complementary (reverse) bases 5¢-GGTATCCCTACCT GATTC-3¢ (PG2). These primers differentiated the two pathogens. The speciﬁc primers for P. guttata (PG1 and PG2) ampliﬁed only P. guttata DNA, and failed to amplify E. pulchra DNA. In addition, speciﬁc primers for E. pulchra (EP1 and EP2) ampliﬁed only E. pulchra DNA, and not P. guttata DNA. The diagnostic band size from E. pulchra-speciﬁc primer pair was 568 bp, while the P. guttata band was 597 bp.
Thus, the two primer pairs (EP1–EP2 and PG1–PG2) were highly speciﬁc to E. pulchra and P. guttata, respectively (Fig. 3). Speciﬁc primers for E. pulchra and P. guttata were highly selective and consistently differentiated the two fungi associated with dogwood powdery mildew.
A comparison of ITS DNA sequences along with information from the GenBank for E. pulchra isolates collected on C. ﬂorida showed a 99% match with an E. pulchra var.