Simple sequence repeat markers useful for sorghum downy mildew (Peronosclerospora sorghi) and related species

Background A recent outbreak of sorghum downy mildew in Texas has led to the discovery of both metalaxyl resistance and a new pathotype in the causal organism, Peronosclerospora sorghi. These observations and the difficulty in resolving among phylogenetically related downy mildew pathogens dramatically point out the need for simply scored markers in order to differentiate among isolates and species, and to study the population structure within these obligate oomycetes. Here we present the initial results from the use of a biotin capture method to discover, clone and develop PCR primers that permit the use of simple sequence repeats (microsatellites) to detect differences at the DNA level. Results Among the 55 primers pairs designed from clones from pathotype 3 of P. sorghi, 36 flanked microsatellite loci containing simple repeats, including 28 (55%) with dinucleotide repeats and 6 (11%) with trinucleotide repeats. A total of 22 microsatellites with CA/AC or GT/TG repeats were the most abundant (40%) and GA/AG or CT/TC types contribute 15% in our collection. When used to amplify DNA from 19 isolates from P. sorghi, as well as from 5 related species that cause downy mildew on other hosts, the number of different bands detected for each SSR primer pair using a LI-COR- DNA Analyzer ranged from two to eight. Successful cross-amplification for 12 primer pairs studied in detail using DNA from downy mildews that attack maize (P. maydis & P. philippinensis), sugar cane (P. sacchari), pearl millet (Sclerospora graminicola) and rose (Peronospora sparsa) indicate that the flanking regions are conserved in all these species. A total of 15 SSR amplicons unique to P. philippinensis (one of the potential threats to US maize production) were detected, and these have potential for development of diagnostic tests. A total of 260 alleles were obtained using 54 microsatellites primer combinations, with an average of 4.8 polymorphic markers per SSR across 34 Peronosclerospora, Peronospora and Sclerospora spp isolates studied. Cluster analysis by UPGMA as well as principal coordinate analysis (PCA) grouped the 34 isolates into three distinct groups (all 19 isolates of Peronosclerospora sorghi in cluster I, five isolates of P. maydis and three isolates of P. sacchari in cluster II and five isolates of Sclerospora graminicola in cluster III). Conclusion To our knowledge, this is the first attempt to extensively develop SSR markers from Peronosclerospora genomic DNA. The newly developed SSR markers can be readily used to distinguish isolates within several species of the oomycetes that cause downy mildew diseases. Also, microsatellite fragments likely include retrotransposon regions of DNA and these sequences can serve as useful genetic markers for strain identification, due to their degree of variability and their widespread occurrence among sorghum, maize, sugarcane, pearl millet and rose downy mildew isolates.

analysis by UPGMA as well as principal coordinate analysis (PCA) grouped the 34 isolates into three distinct groups (all 19 isolates of Peronosclerospora sorghi in cluster I, five isolates of P. maydis and three isolates of P. sacchari in cluster II and five isolates of Sclerospora graminicola in cluster III).
Conclusion: To our knowledge, this is the first attempt to extensively develop SSR markers from Peronosclerospora genomic DNA. The newly developed SSR markers can be readily used to distinguish isolates within several species of the oomycetes that cause downy mildew diseases. Also, microsatellite fragments likely include retrotransposon regions of DNA and these sequences can serve as useful genetic markers for strain identification, due to their degree of variability and their widespread occurrence among sorghum, maize, sugarcane, pearl millet and rose downy mildew isolates.

Background
Sorghum (Sorghum bicolor) is the world's 5 th most important grain crop [1], and with a genome much smaller than maize, is a model for tropical grasses of worldwide importance. It is a major source of food, feed, fiber, fuel, and chemical feed stocks across a range of environments and production systems. However, diseases and insects, in addition to abiotic stresses, are major impediments to sorghum production. Among the diseases, sorghum downy mildew (SDM), caused by Peronosclerospora sorghi, [2] is a disease of sorghum and maize (Zea mays) [3]. SDM can cause severe epidemics, resulting in heavy yield loss. In the past P. sorghi has been subdivided into 'sorghum/ maize' and 'maize' infecting strains. As more information has become available, the Peronoscleosporas have been designated as separate species, with names that indicate the preferred host (P. maydis on maize and P. sacchari for sugarcane). Although P. sorghi can produce symptoms on both maize and sorghum, it typically does not complete sexual reproduction on maize, so no oospores are formed. Sorghum and maize downy mildew (P. sorghi) reported in 44 countries and pearl millet downy mildew (Sclerospora graminicola) reported in 51 countries around the globe are found in most tropical and subtropical regions [4,5]. These oomycetes are both seed and soil-borne pathogens, thus rendering crop rotation less effective in controlling the disease. A sudden re-emergence and disease outbreak of sorghum downy mildew resulting in significant yield loss was noted in Wharton and Jackson counties of Texas during the spring of 2001 and again in 2002 [6]. P. sorghi isolates from the outbreak were found to be resistant to metalaxyl [methyl N-(2, 6-dimethylphenyl)-N-(methoxyacetyl)-DL-alaninate] fungicide, which had been used as an effective seed treatments for many years. Subsequently, isolates collected from previously resistant hosts revealed the evolution of a new pathogenic race of P. sorghi and demonstrated a need for constant monitoring of the pathogen populations.
Philippine downy mildew of maize, caused by Peronosclerospora philippinensis (Weston) C.G. Shaw, has not been reported in the western hemisphere, but it is a serious problem in the Philippines where disease losses in some fields have been 40-60% [7]. Although maize varieties resistant to P. philippinensis have been developed in Asia, maize hybrids currently grown in the United States are highly susceptible to P. philippinensis and American breeding lines that are highly resistant to P. sorghi in the United States are generally susceptible to P. philippinensis (M.R. Bonde, personal communications). Besides maize and sorghum [8], P. philippinensis has been reported to infect Avena sativa L., Saccharum officinarum L., Sorghum halepense (L.) Pers, S. propinquum (Kunth) Hitchc. [9] and Saccharum spontaneum L. [10]. Consequently, P. philippinensis is the subject of quarantine in many countries and was listed by USDA APHIS under CFR 7 part 331 as a "select agent" in the US, meaning it is considered to be a major threat to agricultural security. Because of the threat posed by P. philippinensis, a DNA sample available from the USDA, ARS Foreign Disease-Weed Science Research Unit at Ft. Detrick, MD was included in this study. The information gained, besides having epidemiologic significance for maize and sorghum production, also could help resolve confusion over the taxonomy of Peronosclerospora spp.
The population structure of P. sorghi has been previously addressed with physical characters, such as spore size or pathotype, which require comparative disease responses on a series of host differential cultivars [11,12]. Conventional methodologies for the detection of different pathotypes of downy mildews of cereals have not been satisfactory particularly for the identification of organisms at extremely low propagule numbers and are cumbersome for use with obligate parasites (such as plant-pathogenic rusts and mildews), which cannot be cultured. Hence, in tropical climates, symptoms of Peronosclerospora can be confused with those from related downy mildew pathogens, including P. maydis, P. sacchari and P. philippinesis. Yao et al. [13] showed that clones of P. sorghi DNA crosshybridized well on Southern blots with DNA from the other Peronosclerospora species. While few differences were detected among the RFLP patterns within P sorghi, different patterns were seen for the different species. Similarly, ITS-2, an internal transcribed spacer of ribosomal genes was different in size [14]. While potentially useful in species identification, more easily scored and highly variable markers from locations throughout the genome are needed for differentiating among closely related pathotypes, and to tag genes that contribute to host specificity, race specificity, and virulence. Although polymorphisms were detected in SDM pathotypes using RAPD-PCR, many of the short (10 mer) primers resulted in monomorphic banding patterns. A larger problem involved inconsistently amplified bands that varied greatly in intensity, even in duplicate experiments [15]. As a consequence, a more consistent procedure was sought in the present investigation for the development of microsatellite markers.
Simple sequence repeat (SSR), or microsatellites, are hypervariable and dispersed in the form of long arrays of short tandem repeat units throughout the genome [16,17]. These SSRs provide codominant Mendelian markers, much more powerful than dominant markers and can be used to determine population genetic structure, kinship, reproductive mode, and genetic isolation [18,19]. When compared with several other marker types, SSR markers were superior for genetic characterization in Aspergillus fumigatus, and in Saccharomyces cerevisiae [20,21]. Microsatellite markers have been developed for several fungi including Magnaporthe grisea [22,23], and another oomycete, Phytophthora infestans [24] among others. However, based on sequence searches using genome data P. infestans and two other plant pathogenic Phytophthoras, potential SSRs have been reported to be rare [25].
In the present study, many new microsatellite sequences comprised of dinucleotide and trinucleotide repeat structures were retrieved from genomic DNA of Sorghum Downy Mildew (Peronosclerospora sorghi) pathotype 3 using an optimized and highly simplified biotin capture protocol. On the basis of sequence analysis of these captured fragments, appropriate primers were designed for amplification of these microsatellite loci and the developed microsatellites for P. sorghi were tested for the ability to differentiate pathogenically diverse isolates of P. sorghi and for their utility for cross-species amplification in P. philippinensis, P. maydis (maize), P. sacchari (sugar cane), Sclerospora graminicola (pearl millet) and Peronospora sparsa (rose) downy mildew. Clustering analysis was performed to assess the genetic diversity and close relatedness among 34 isolates including the different species of downy mildews.

Pathogen samples and DNA isolation
A total of 19 isolates of Peronosclerospora sorghi including P1 (metalaxyl susceptible), P3 and P6 (both metalaxyl resistant and susceptible) pathotypes collected in different years from different parts of Texas where the disease out-burst occurred were included in this study. In addition, DNA samples from P. philippinensis (one), P. maydis (five), P. sacchari (three) were obtained from Dr. Douglas G. Luster, USDA/ARS, Foreign Disease-Weed Science Research Unit, Ft. Detrick, Maryland, USA and DNA samples of Sclerospora graminicola (five) and Peronosopora sparsa (one) available in Dr. Magill's lab, Dept. of Plant Pathology & Microbiology, Texas A&M University, College Station, Texas, USA, were also included in the present study (Table  1). For DNA isolation, conidia were collected from infected leaves as follows: leaves were washed with cold sterile water and placed abaxial side down on a screen in a petri plate, covered with moist paper towels and allowed to sporulate in the dark overnight at 23°C [26]. Spores collected in sterile water were used for DNA extraction. After removing excess water, conidial suspensions were washed a few times with 70% ethanol and frozen for lyophilization. The dried sample was powdered in liquid nitrogen using a mortar and pestle. DNA was extracted using a MasterPure Yeast DNA Purification kit (EPICEN-TRE Biotechnologies, Madison, WI) and the DNA was diluted to working concentration of 2.5 ng/μL by adding 1× TE (10 m M Tris-HCl, 1 mM EDTA, pH 8.0).

Microsatellite Isolation
A microsatellite-enriched library was prepared by a simplified protocol based on the biotinylated-oligonucleotide capture methods of [27]and [28]. Further, several steps in the biotin capture protocol were modified as detailed by Reddy et al. [29] to optimize the frequency of microsatellite repeats among captured genomic DNA fragments. In our protocol, no size fractionation steps or radioactive hybridizations were employed. A 500 ng sample of genomic DNA from SDM pathotype 3 was digested for 3 h in a single reaction mixture containing restriction endonucleases HaeIII, RsaI, and DraI (20 units of each), as well as 50 ng of RNaseA. This digestion resulted in a diverse population of blunt-ended restriction fragments with an average size of ~550 bp. Digested DNAs were purified using a QIA-quick PCR purification column (Qiagen, Valencia, CA), eluted with 50 μl of 5 mM Tris-pH 8.0, then dried completely under vacuum. The doublestranded adaptor molecule AP11/12 was prepared by mixing equal molar amounts of oligonucleotides AP11 (5'CTCTTGCTTAGATCTGGACTA3') and AP12 (5'pTAGTCCAGATCTAAGCA-AGAGCACA3', where p = 5' phosphate), heating to 94°C, then cooling to 25°C over a period of 5 h. Digested genomic DNA fragments were resuspended in a 30-μL ligation reaction containing 100 ng of AP11/12 double-stranded adaptor and 30 Weiss units of T4 DNA ligase. Ligation was carried out at 14°C for 16 h. Preamplification of adaptor-ligated products was performed using 2 μL of the ligation reaction as a template for 10 cycles of PCR in a 50-μL reaction volume using the Approximately 100 ng of the preamplified product was then added to a single reaction mixture containing 6× SSC   . Assembly of double-stranded DNA sequence contigs from each clone, and identification of redundancy and overlaps between clones was performed using Sequencher 3.0 (Gene Codes, Ann Arbor, MI). Microsatellite analysis for the identification of sequences with perfect and imperfect repeats and simple and compound repeat motifs were identified using Simple Sequence Repeat Identification Tool (SSRIT) database [30].

Sequence homology search
The sequences of SDM microsatellites were subjected to homology searches using BLASTN against nucleotide collection (nr/nt) and Expressed Sequence Tags (est_others) at National Center for Biotechnology Information (NCBI; [31]).
Under Algorithm parameters the following options were chosen to improve the results.

General parameters
Max target sequences -100; Short queries -automatically adjust parameters for short input sequences; Expected threshold -10 & Word size -11

Filters and Masking
Filter -low complexity regions; species-specific repeats for fungi Mask -Mask for lookup table only options.

Design of primers and PCR amplification
Primers were designed to the flanking region of the SSR using PRIMER 3 software [32]. Primers were designed without repetitive DNA and with a base composition of greater than 40% G+C with annealing temperature between 52°C and 58°C to yield amplification product between 100 and 250 bp. Oligonucleotide primers were synthesized by Integrated DNA Technologies, Coralville, IA, USA. Primers selected by these criteria were evaluated further for melting temperature, internal structure, and propensity for primer-dimer formation using publicly accessible Worldwide Web resources (Sigma-Genosys, The Woodlands, TX). PCR amplification was carried out in 10 μL containing 5.5 ng of genomic DNA, 1 μL 10× PCR buffer (Promega), 0.8 μL of 2.5 mM dNTP mix, 1 μL each of 1 pM/μL forward and reverse primers, 1 μL of 25 mM MgCl 2 , and 1 U of Taq polymerase. PCR was performed for an initial denaturation of 2 min at 95°C followed by 35 cycles of 45 s at 95°C, 45 s at the appropriate annealing temperature for each microsatellite primer pair as highlighted in Table 1 and 60 s at 72°C, and a final extension of 20 min. at 72°C.

Gel Analysis
The developed SSR primers with 4% SFR (super fine resolution) agarose (AMRESCO, Solon, Ohio, USA) gel electrophoresis system running with 1× TBE buffer in 4°C using circulatory cool buffer system was used for better resolution of polymorphic differences between downy mildew isolates. The SSR amplification products were also analyzed using a LI-COR-NEN ® Model 4300 -dual-dye automated DNA Analyzer. Following amplification, an equal volume (5 μl) of PCR products of two sets of SSRs one labeled using the IRD-700 nm and another with IRD-800 nm fluorogenic forward primer were mixed. Basic fusion dye (2 μl) (LI-COR) was added to each pooled sample and the samples were denatured for 5 min at 95°C. Each sample (1 μl) was loaded on a 6.5% polyacrylamide gel containing 7 M urea. Gels were cast using LI-COR 25-cm plates with 0.25-mm-thick spacers and comb.
Electrophoresis was performed at a constant power of 40 W and a constant temperature of 47.5°C for 2 h.

Statistical analysis
Microsatellite fragments were scored as present (= 1) or absent (= 0) across the 34 Peronosclerospora, Peronospora and Sclerospora spp. isolates. The resulting binary matrix was used to construct a similarity matrix using the Jaccard coefficient, GS(ij) = a/(a + b + c) (17), where a is the number of fragments shared by i and j, b is the number of fragments present in i and absent in j, and c is the number of bands present in j and absent in i. Cluster analysis was performed using the unweighted pair group with arithmetic means method (UPGMA) [33]. PAUP 4.0* was used to generate 1000 bootstrap replicates for testing the reliability of the dataset and to draw a consensus tree [34]. Ordination analysis was performed to study the relatedness within a matrix by converting the pairwise distance into Eigen vectors and values. Cluster analyses, ordination analyses and the Mantel test were performed using NTSYSpc (NTSYS -for Numerical Taxonomy SYStems) version 2.1 [35].

Results
A collection of 513 colonies, picked at random from primary transformation plates, was inoculated into 96-well culture plates for high throughput sequencing. Of the 513 clones sequenced, 170 (33%) were redundant with other clones in the same collection. A total of 73 out of 343 clones (21%) did not contain repeat motifs. Of the remaining 270 nonredundant clones, Sequence analysis showed that virtually all of the inserts contained microsatellite repeat motifs matching one or more of the biotinylated oligonucleotides used in the selection process. Out of 270 clones, 144 insert fragments (53%) were truncated within, or at the end of, a microsatellite repeat. Of the remaining 126 non redundant clones with unique sequences flanking both ends of the microsatellite repeats, 55 were selected for further analysis on the basis of the following criteria: (i) the length of the microsatellite was five or greater in dinucleotide and four in trinucleotide repeat units and; (ii) the unique 5' and 3' flanking sequences were both of suitable structure and composition for the design of efficient primers. Microsatellites with flanking sequences that were highly repetitive in nature and/or had high A+T rich nucleotides were eliminated.
Thirty six microsatellite loci containing a simple repeat type made up 66% of the DM collection, including 28 (55%) with dinucleotide repeat types and 6 (11%) with trinucleotide repeat types (three CAA and one each of GTT, AGG and ACA). Of the microsatellites containing a simple repeat type, 94% had "perfect" repeats (except DM9 and DM16), uninterrupted by nonrepeat nucleotides. The remaining 19 SSR loci (39%) of the collection were composed of "compound repeats" consisting of more than one repeat type at a single locus and all were interrupted by nonrepeat nucleotides and hence considered as "imperfect" repeats ( Table 2). All the compound repeats were combinations of dinucleotide repeat motifs except DM45 (trinucleotide) and DM49 (combination of tetra and dinucleotide) repeat motifs. A total of 22 microsatellites with the CA/AC or GT/TG repeat type were the most abundant (40%) and the GA/AG or CT/TC type contributes 15% in our collection.
Of the 55 cloned SSR sequences queried in BLASTN searches of the NCBI non-redundant genomic and EST databases(as of July 4, 2008), only four clones DM21, DM28, DM53 and DM55 were found to show similarity to previous entries and these were to sequences of Phytophthora and Pythium species which are also plant pathogenic oomycetes. Details are presented in table 3. These results indicate that two of the cloned sequences (DM53 and DM54) isolated from P. sorghi contain incomplete portions of genes found in known retrotransposon elements.
Primers were designed for the 55 selected microsatellite loci using the criteria described in materials and methods. These primers (for Table 4 see additional file 1) had an average length of 20.7 nucleotides, with an average G+C content of 49% and annealing temperature in a range of 50-55°C. Predicted cloned products ranged in size from 106 (DM52) to 249 (DM21) bp. These 55 new microsatellite loci described in   Imperfect repeat TG TG TG TG TG TG TG TG TG TATG TG TG TG TG TG TG TG TG TG TG TG TG TG  TG TG TG TG TG TG TG TG   DM13 compound and imperfect CACACACACACACACATCTGTATATATA Bold and underlined are interrupted non repeat nucleotides makes the SSR loci imperfect the distinct banding differences for amplification products obtained with 34 Peronosclerospora, Peronospora and Sclerospora spp. isolates using the DM9 microsatellite primer pair are shown in Fig. 1.
A presence (1) or absence (0) binary matrix containing 260 clear amplicons from 54 SSRs was used to generate the genetic similarity estimates. The similarity co-efficient ranged from 0.77 to 0.98 indicating narrow genetic distance among the isolates of the different species studied.
The dendrogram constructed using UPGMA (Fig. 2)   *E value or the Expect value is a parameter that describes the number of hits one can "expect" to see just by chance when searching a database of a particular size Primers in bold letters are cross-amplified in all species * Primer combinations expressed unique bands differences from other species were distinctly separated from others with a confidence limit of 79% and 76%. Five isolates of P. maydis and three isolates of P. sacchari were grouped in two sub clusters with confidence limits of 83 and 74% respectively in cluster II and five isolates of Sclerospora graminicola were grouped in cluster III with confidence limit of 70%. The remaining P. sparsa and P. philippinensis isolates were not grouped into any of these three clusters and showed their distinct uniqueness from other species. Principal coordinate analysis (PCA) was also performed to show the relationship among the 34 isolates as a three dimensional display. In this analysis the first two principal components (having eigen value >1) explained 86% of the total variation. Like the UPGMA clustering dendrogram, PCA analysis placed isolates of different species of downy mildew into distinct positions (Fig. 3).

Discussion
The Peronosclerospora spp. that cause downy mildew of sorghum and maize and other gramineous hosts include some of the most destructive pathogens in the tropics and subtropics [36] and some of them infect more than one important crop. Of the 10 recognized species of fungi that cause downy mildew diseases of maize, 5 currently are in the genus Peronosclerospora [37,38]. The confusion over nomenclature of specific pathogens involved in these diseases is a major problem in global evaluation of the downy mildews caused by Peronosclerospora. It is difficult to determine if research results and disease control methods developed in one area of the world might be applicable to another. In several instances, taxonomy within the genus is confused because, at most, only slight differences in morphology of recognized species exist [39]. With the result of traditional approaches [40] combined with utilizing molecular techniques including isozyme analysis [41] and PCR (Polymerase Chain Reaction, [14]) the maize strain of P. sorghi from Thailand was given specific rank as P. zeae [13]. Legg [42] has outlined the advantages of some of the molecular methods for studying the Peronosclerospora spp. A maize strain of P. sorghi was reported in southern Nigeria [43]. However, Yao et al. [14] using PCR, were unable to differentiate an isolate of SDM from maize in southern Nigeria from sorghum/maize-infecting strains and the identity of this pathogen remains unknown. The problems described warrant development of simple and powerful molecular diagnostic tools for constant monitoring of this disease. Since microsatellite regions are highly mutable due to their propensity of addition and deletion of repeat copies they can be used to differentiate within and between related taxa, even at the level of individual isolates in a single species. Simple DNA-profiling methods based on microsatellite variability provide possibilities to identify individual genotypes for studies in population genetics, ecology and taxonomy [44].

Characterization of microsatellite loci
In the present study, a total of 55 microsatellite loci were isolated from Peronosclerospora sorghi -pathotype P3 DNA enriched for simple sequence motifs. We observed 33% internal redundant motifs, likely consequences of using two PCR amplification steps during the isolation process. Of the sequenced clones with an insert, over 79% contained microsatellites. This is a significant enrichment as compared to traditional microsatellite isolation protocol [45]. However, the relatively small fragment size average resulting from the use of three restriction enzymes in the initial fragmentation (HaeIII, (GGCC) targets, RsaI, (GTAC) and DraI (AAATTT)) may explain why 144 insert fragments out of 270 clones (53%) were truncated within, or at the end of, a microsatellite repeat. This result is in accordance with the CM collection [46], in which genomic DNAs were mechanically fragmented using a Licor gel image of 34 isolates of different downy mildew species using DM 9 SSR primers high-pressure nebulizer, and a substantial portion of the clones were truncated within the microsatellite. The most frequent dinucleotide microsatellite motifs detected in our libraries were (CA/AC) n or (GT/TG) n . followed by (GA/AG) n or (CT/TC) n . In earlier studies, (CA) n repeat motifs [47] and (AG) n repeat motifs [48] were also detected from the ascomycetes, Podospora anserina and Lobaria pulmonaria respectively. Approximately 39% of all isolated loci were composed of complex intermixed motifs, which were rarely arranged in a tandem way. This pattern, known as cryptic simplicity [49], is common in larger eukaryotic genomes [50]. Microsatellites identified with trinucleotide repeats (CAA, AGG and ACA) were found in this study with low frequencies, but seem to differ from those typically found in 14 fungal genomes (AAG, AAC, AGC, and ATC) [51]. One trinucleotide repeat clone (DM45) included multiple repeats in a complex pattern but no significant homology was found to other sequences currently in GenBank.
Determining if microsatellites will be polymorphic can only be assessed empirically, however, two considerations can help predict if polymorphism is likely. First, as slippage during replication increases with the number of tandem repeats [52], loci in which the motif is iterated at least eight times are desirable. Second, it may be possible to predict the stability of a given microsatellite motif by assessing the likelihood of "expandability" [53]. Microsatellites reflect a balance between expansion and contraction, but while there is evidence of a bias toward increasing microsatellite size [54], very long tandem repetitions are rare [55]. In contrast, in our study, (CA/AC)n and (GT/TG)n were commonly repeated more than 20 times, suggesting expansion to these lengths is not limited. The presence of these motifs in most fungal and oomycete genomes suggests they would be useful microsatellites to target in different species where genome data are not available.

Retrotransposable Elements
The presence of retrotransposon-like elements (mobile genetic elements) plays an important role in genome evolution and such elements constitute about 5 to 10% of the genome in eukaryotes [56]. DM55 (ACA) 19 showed significant sequence homology to mitochondrial sequences in several Phytophora species (Table 3). Two of the P. sorghi SSR sequences were found to have a degree of similarity to reverse transcriptases, suggesting an origin from retro-transposons possibly present in a progenitor species. In this study, P. sacchari and S. graminicola isolates were not amplified with DM 21, whereas, DM53 cross amplified in all the species studied. These results suggest the GypsyPr-2 reverse transcriptase-like gene sequence as defined in Phytopthora ramorum was not present in P. sacchari or S. graminicola isolates. However, sequences similar to the Phytopthora infestans GypsyPi-3a and GypsyPi-3b retrotransposon-like and to the GypsyPs-1A sequence from Phytopthora sojae were present in all Peronosclerospora, Peronospora and Sclerospora downy mildew isolates. Primers and probes built to take advantage of the presence of unique differences in such retrotransposon sequences can be useful for strain/race identification. Examples include the work by Sastryet al. [57] with S. graminicola and Becker et al. [58] in the basidiomycete Chondrostereum purpureum.
Principal coordinate analysis (PCA) Figure 3 Principal coordinate analysis (PCA). Three-dimensional display of 34 Peronosclerospora, Peronospora and Sclerospora isolates based on the combination of data obtained with 54 SSRs. Dimensions 1, 2 and 3 are accounted 86% of the variation observed.

Gel analysis
Resolving microsatellite amplicons in polyacrylamide gels using a LI-COR high throughput sequencing system with fluorogenic primers results in significantly more useful bands than resolving the same products in 4% SFR gels stained with ethidium bromide. For example DM12, 19, 30 and 38 gave monomorphic expression in SFR agarose gels but showed alleles with polymorphic differences in LI-COR poly-acrylamide gels. Taken together, from combined data across 34 Peronosclerospora, Peronospora and Sclerospora spp. isolates, an average of 4.8 polymorphic markers were detected per SSR in LI-COR poly-acrylamide gels versus 2.7 in SFR agarose gels. Thus the use of LI-COR poly-acrylamide gels with fluorogenic labeled primers was more efficient in separating bands which reflect clear cut polymorphisms. Fingerprinting using 54 SSR primer pairs over all the isolates resulted in 260 diagnostic alleles.
Although it is assumed that the majority of size variants (100 to 350 bp) was due to variation in the number of the repeat motifs, this has not been verified by sequencing. Insertions, deletions and base substitutions in the flanking regions [59] might also account for variation in fragment length, especially when amplicons from different species are compared.

Cross-species amplification
It has been shown that microsatellite primers developed for a distinct species can be useful for genetic analysis in closely related species [60], but successful transferability depends upon the evolutionary distance between source and target species [61]. Dutech et al. [62] reported relatively low cross-species transferability of microsatellites in fungi. Within genera, only 34% of the loci tested could be transferred, which appears much lower than in animals or plants. Here, twelve SSRs ( DM 9,12,13,14,16,19,30,31,35,49,53 and DM54) were cross amplified in samples from 6 species of downy mildew-causing oomycetes. Successful cross-amplification simply indicates that the flanking regions are conserved across the species studied, but it does not tell anything about the character and structure of the fragment and additional sequencing will be necessary to clarify the structure of these fragments. Similar findings with successful cross amplification of SSRs in closely related species were reported for lichen-associated fungi [48] and in for a variety of fungal species [51]. Another interesting feature evident from our study is the distinctness of different species as revealed by unique banding patterns with different SSRs as evinced clearly from Fig. 1 While these amplicons may be useful for diagnostics based on unique banding patterns, the information for P. philippinensis and P. sparsa is based on single isolates, so data from more independent, species-verified samples will be required to assess the degree of variation within natural populations. Based on the species-specific clustering of isolates where multiple samples were available, distinctive fingerprints can be anticipated.

Cluster analysis
In the present study, SSR markers revealed using high throughput LI-COR-NEN ® Model 4300 -dual-dye automated DNA Analyzer have been used to resolve cryptic genetic variation of closely related different downy mildew species that have been impossible to resolve with morphological systematic characters. SSR fingerprinting was conducted using 54 primer pairs and genomic similarity analyses derived from qualitative data grouped isolates of samples of downy mildew according to host specificity (sorghum, maize, sugarcane, pearl millet and rose) whose taxa had been uncertain based on morphological criteria. In this study, metalaxyl sensitive P3 isolate (#2 in Fig. 2) in cluster I grouped with all other metalaxyl resistant isolates, suggested that metalaxyl resistance may have originated in a P3 isolate to give rise to the recent outbreak, consistent with results of based on AFLP molecular characterization [15].
However, mating and recombination involving metalaxyl-resistant genotypes and other isolates, along with selection in favor of resistance would narrow the genetic base and can account for the high similarity (approximately 89%) among all 19 P. sorghi isolates in cluster I. P. maydis and P. sacchari isolates in cluster II were grouped into two sub clusters with high similarity coefficient (approx. 82%) indicates that corn and sugarcane downy mildew isolates have narrow genetic differences, showing more genetic similarities with each other than other Peronosclerospora species. The single isolates of P. philippinensis and P. sparsa were not grouped with any of the three clusters and showed their uniqueness due to their distinct banding pattern with different microsatellite primer combinations as detailed earlier. The three-dimensional comparison is compatible with cluster analysis and provides a good visual comparison of the genetic similarities and differences of the isolates. Fig. 3 clearly depicted the unique identity of P. philippinensis from other species. This display shows that the genetic distance between P. philippinensis and P. sacchai is very narrow. Similar findings with close phylogenetic relationship was recorded through isozyme analysis between P. philippinensis and P. sacchari isolates from Taiwan [26,63] and by RFLP patterns by [13]. Further study is clearly needed to confirm this relationship. Unfortunately, the select agent status of P. philipinnensis has made the obtaining of additional samples, even of DNA, impossible to date.

Conclusion
Microsatellite primer sets developed from P. sorghi sequences proved to be useful for all downy mildew species analyzed and are likely to be increasingly developed and applied to studies of pathogen epidemiology, population biology, and genomics. The diversity of microsatellite motifs gave each species a unique "signature" of repeat distributions. Subsets of the 54 newly developed SSRs may be very useful for rapid and efficient identification and genetic analysis of the natural populations and host range of these obligate oomycetes. The unique banding pattern of P. philippinensis from fifteen (DM 3,8,11,13,18,27,31,39,40,45,50,52,53,54,& DM 55) and P. sparsa from seven (DM13, 23, 24, 45, 48, 50 and DM53) SSRs make these primers useful as diagnostic markers for the respective species. In future studies it will be possible to focus on the population diversity and recombination within and between native and introduced populations of the closely related species. These distinct fingerprinting profiles can be used as diagnostic tools to formulate breeding strategies targeting host resistance to local pathogen populations and for monitoring the emergence of new virulent races.