Present address: Leimenweg 81, 4493 Wenslingen, Switzerland.
Generation of pathogenic F1 progeny from crosses of Phytophthora infestans isolates differing in ploidy
Version of Record online: 2 JUL 2012
© 2012 The Authors. Plant Pathology © 2012 BSPP
Volume 62, Issue 3, pages 708–718, June 2013
How to Cite
Hamed, B. H. and Gisi, U. (2013), Generation of pathogenic F1 progeny from crosses of Phytophthora infestans isolates differing in ploidy. Plant Pathology, 62: 708–718. doi: 10.1111/j.1365-3059.2012.02655.x
- Issue online: 22 APR 2013
- Version of Record online: 2 JUL 2012
- diploid progeny;
- flow cytometry;
- sexual recombination;
- simple sequence repeats;
- SNP quantification;
- triploid progeny
Ten Phytophthora infestans isolates were analysed for maximal growth rate (μmax), sporulation capacity (K), ploidy and the ability to produce infective F1 progeny when combined in different crosses on tomato leaflets. Ploidy was determined using three different methods: simple sequence repeats (SSR, microsatellites), allele quantification at F382Y in the RNA polymerase, and flow cytometry. Eight out of 10 isolates were classified as diploid, the other two as triploid. There was no correlation between ploidy, growth rate and sporulation capacity although the latter was rather low for triploid isolates. Ten crosses were investigated in tomato leaflets (five 2n × 2n; four 2n × 3n; one 3n × 3n). Oospore production was observed in all crosses independent of parental ploidy. Germination and pathogenicity of oospores were investigated by measuring the number of infection sites (pathogenic F1 progeny) over a time period of 72 days. All crosses containing a triploid parent produced many fewer F1 isolates than 2n × 2n crosses. The generation of pathogenic F1 progeny isolates was best for 2n × 2n crosses. A total of 319 F1 offspring were produced and analysed for mating type and ploidy; a subset was tested for ploidy with all three methods. For 2n × 2n crosses, the majority of offspring were diploid, whereas in crosses containing one or two triploid parents, diploid, triploid and trisomic offspring were observed. The results suggest that less pathogenic F1 progeny isolates are produced if at least one parent is triploid.
Phytophthora infestans, the causal agent of potato and tomato late blight belongs to the Oomycetes phylum in the Straminipila clade (Hawksworth et al., 1995; Yoon et al., 2002). Differences in ploidy were first reported by Sansome (1977), who detected that two British isolates of P. infestans had twice as high a chromosome number when compared to Mexican isolates, the latter consequently being claimed as diploid. Ten years later, Tooley & Therrien (1987), using Feulgen staining of zoospores, described a mean DNA content of 0·59 pg for 23 Mexican isolates (claimed as diploid) and 0·92 pg for 18 non-Mexican isolates (claimed as triploid). With DAPI staining, both A1 and A2 isolates showed a range of DNA contents from a basic 2C level (presumed diploid) through intermediate to 4C (presumed tetraploid) stages (Whittaker et al., 1991a). The same authors (Whittaker et al., 1991b) counted chromosomes in gametangia during mitotic metaphase; diploid isolates had a chromosome number of 8–12, tetraploid isolates 14–20. They also performed different crosses in vitro and observed that oospore germination and establishment of colonies were generally lower in 2n × 4n and 4n × 4n crosses than in the 2n × 2n cross. Progeny of 2n × 2n crosses were all diploid, whereas those from 2n × 4n and 4n × 4n ranged from 2n to >3n and from 2n–4n, respectively (Whittaker et al., 1991b). Ritch & Daggett (1995) counted chromosomes in zoospore and hyphal tip nuclei of field isolates and found 2n, 3n and 4n isolates. Newer studies with molecular markers (restriction fragment length polymorphisms, RFLPs, and amplified fragment length polymorphisms, AFLPs) describe a few triploid but a lot of trisomic offspring and chromosomal rearrangement in F1 progeny (Carter et al., 1999; Van der Lee et al., 2004). Catal et al. (2010) described a new method for counting nuclei in P. infestans with laser flow cytometry. They found nuclear DNA contents in field isolates varying from 1·75 to 0·75× that of nuclei of a standard (diploid) isolate from central Mexico, suggesting a DNA aneuploidy. The P. infestans genome has been sequenced recently, resulting in a monoploid genome size of c. 240 Mb which is two to four times larger than other sequenced Phytophthora species (Haas et al., 2009). However, there are few results available about the number of chromosomes and ploidy in P. infestans.
To get a better understanding of P. infestans epidemiology and to explain and predict population changes, it is important to investigate how ploidy may affect mating and production of F1 progeny in the field. In this study, field isolates of P. infestans were tested for differences in ploidy, growth rate and sporulation capacity. Ploidy of F1 and field isolates was determined with molecular markers, such as simple sequence repeats (SSR; Knapova & Gisi, 2002; Lees et al., 2006; EUCABLIGHT [http://www.eucablight.org/EucaBlight.asp]) and allele quantification of a single nucleotide polymorphism (SNP; Syngenta internal data) and compared to flow cytometry (Vercauteren et al., 2011). With selected isolates differing in maximal growth rate, sporulation capacity and ploidy, crosses in tomato leaflets were made (2n × 2n, 2n × 3n, 3n × 3n) and the number of oospores and infection sites produced by oospores analysed to evaluate whether ploidy had an impact on the number of pathogenic F1 progeny isolates produced per cross, and to predict if such isolates may have equal chances to compete in field populations.
Materials and methods
Origin of P. infestans isolates and mating type determination
Isolates were collected from potato fields in the United Kingdom, Netherlands and Denmark in 2008. Four A1 and four A2 isolates were selected for this study (Table 1) and compared to two diploid standard isolates, PiMex-3-1, a central Mexican field isolate (Catal et al., 2010) and F80029, a Dutch field isolate (Van der Lee et al., 2004). Isolation and purification were carried out as described by Knapova & Gisi (2002). Pure cultures were incubated on rye decoction agar (RDA) plates in the dark at 19°C. Mating type determination was carried out on agar plates (Knapova & Gisi, 2002) or by a standard molecular assay using a modified version of the cleaved amplified polymorphic sequence assay (Mazakova et al., 2006).
|Isolate||Country||Mating type||SSR genotypea||Clustera|
Genome size analysis of P. infestans isolates
Flow cytometry was used to measure the genome size (pg) of parental and F1 progeny isolates (C-value = ‘holoploid genome size’ = DNA content of the whole chromosome complement; Greilhuber et al., 2005). Several methods were evaluated, but reliable results were obtained only with the method developed by Vercauteren et al. (2011) for P. ramorum, using the ‘Cystein PI absolute P’ kit (Partec) and nuclei of P. infestans isolates extracted from hyphae. As internal standard a leaflet piece (7 mm diameter) of young tomato plants (cv. Baby) was chopped for 60 s with a sharp razor blade together with a piece of mycelium (c. 1 cm2) of P. infestans (10–14-day-old cultures on RDA) in 400 μL extraction buffer. The suspension was filtered after 90 s through a 10 μm Partec CellTrics® filter into a 5 mL polystyrene round-bottom tube (12 × 75 mm; BD Falcon) and 800 μL staining solution (50 μg mL−1 propidium iodide (PI), 12 μg mL−1 DAPI and 200 μg mL−1 DNase-free RNase A) were added. Nuclei samples were incubated at 4°C overnight and analysed with a BD FACS Canto™ II Flow cytometer (BD Biosciences) equipped with an argon laser (488 nm). Histograms were obtained using WinMDI v. 2.9 (Joe Trotter, TSRI). Genome sizes were calculated from the ratios between the mean peak position of the P. infestans sample and the internal standard (nuclei of tomato plants, calibrated with Raphanus sativus cv. Saxa plants, 2C = 1·11 pg; Dolezel et al., 2007). Three nuclei samples per isolate were measured, mean values calculated and statistically analysed.
Simple sequence repeats
For DNA extraction, mycelium was taken either from RDA plates or directly from infected tuber tissue. Approximately 10–100 mg was stored in MagAttract® tubes at −80°C until use. The material was lyophilized overnight and DNA extraction was performed with the MagAttract® 96 DNA Plant Kit 24 (QIAGEN) according to manufacturer’s instructions. For further tests DNA was diluted to 10 ng μL−1.
To determine the ploidy of parental and F1 isolates and to check whether sexual recombination was successful (‘true’ offspring excluding clones in the F1 progeny), the following seven different microsatellite (SSR) loci were used: PiG11, Pi4B (with seven and three alleles, respectively; Knapova & Gisi, 2002), Pi63, Pi02, Pi04 (with three, three and four alleles, respectively; Lees et al., 2006), SSR4 and SSR11 (with six and three alleles, respectively; EUCABLIGHT). The microsatellite reaction was performed in a volume of 15 μL containing the following ingredients: 1 μL DNA (10 ng μL−1), 3 μL 5× colourless Go Taq™ reaction buffer (Promega), 1·2 μL dNTPs (0·01 mm), 0·3 μL of each forward and reverse primer, 0·14 μL Go Taq™ DNA polymerase containing 0·7 U (Promega) and 9·06 μL H2O. Forward primers of the loci Pi02, Pi04, PiG11 and SSR4 were marked at the 5′ end with the fluorescence colour HEX, whereas forward primers of the loci Pi63, Pi4B and SSR11 were coloured with FAM (Microsynth). The PCR reaction was performed in a PTC-225 Peltier Thermal Cycler (MJ Research) with the following parameters: 94°C for 2 min, followed by 26 cycles of denaturation at 94°C for 15 s, annealing at 58°C for 1 min 30 s, extension at 72°C for 10 s, and a final extension at 72°C for 10 min. To analyse the PCR products, 10 μL HiDi Formamide (Applied Biosystems), 0·3 μL GeneScan-500 ROX size standard (Applied Biosystems) and 1 μL of the diluted PCR product were filled in a Micro AMP™ Optical 96-Wellplate (Applied Biosystems) and then denatured at 94°C for 5 min. The following analysis was performed on a 3130 Genetic Analyser (Applied Biosystems). Peak patterns were analysed with the GeneMapper v. 3.7 software (Applied Biosystems). Isolates were assigned as triploid if three alleles were present in at least one SSR locus.
Single nucleotide polymorphism F382Y in the RNA polymerase I enzyme
Pyrosequencing was used for allele quantification of the SNP F382Y representing an exchange of phenylalanine (F) for a tyrosine (Y) at position 382 in the RNA polymerase I enzyme (Syngenta internal data) and performed according to a modified method of Zhou et al. (2006). Pyrosequencing PCR was performed in a 50 μL reaction volume which contained 1 μL DNA (20 ng μL−1), 0·2 μm of forward (AQF382Yfw: 5′- ACGCGCAGAACTCGCACTT-3′) and reverse (AQF382Yrev: 5′- AGATTCACCTGCTCCTCCTTCTC-3′) primers, 0·2 μm dNTPs (Promega), 5 μL Mango Taq 5 × PCR buffer, 0·2 mm MgCl2 (Bioline) and 2 U Mango Taq DNA polymerase (Bioline). The primers were obtained from Microsynth. PCR cycling conditions were 94°C for 4 s followed by 44 cycles of denaturation at 94°C for 20 s, annealing at 52°C for 20 s, extension at 72°C for 20 s, and a final extension at 72°C for 2 min. PCR was performed using a PTC-225 Peltier Thermal Cycler (MJ Research). For the pyrosequencing reaction, single-stranded biotinylated PCR products were prepared using the pyrosequencing Vacuum Prep Tool (Biotage AB). Five microlitres of streptavidin sepharose™ (High Performance, Healthcare Bio-Sciences AB) was added to 40 μL binding buffer (QIAGEN) and mixed with 20 μL PCR product and 15 μL water for 10 min at 1400 rpm and room temperature using a MixMate PCR 96 shaker (Eppendorf). Beads containing the immobilized templates were captured on the filter probes after the vacuum was applied and then washed with 70% ethanol for 5 s, denaturation solution (0·2 m NaOH) for 5 s and washing buffer (QIAGEN) for 5 s. The vacuum was then released and the beads were released into a PSQ 96 plate Low (Biotage AB) containing 39·84 μL annealing buffer and 0·16 μL sequencing primer (AQF382Yseq: 5′- CTTGTCGAAGATTATGACTT-3′). Pyrosequencing reactions were performed according to manufacturer’s instructions using the PSQ 96 reaction kit (Biotage AB), which contained the enzyme, substrate and nucleotides. The assays were performed on the PSQ 96 MA (Biotage AB) using nucleotide dispensation orders. Results were analysed with the pyromark ID 1.0 software. Ploidy was associated with the A/T frequency (adenosine/tyrosine ratio) in the pyrogramme: a ratio of 0:100 was considered as wildtype (ploidy cannot be determined), whereas ratios of about 50:50 (±10) and 33:67 (±10), or 67:33 (±10) were assumed to represent diploid and triploid individuals, respectively.
Growth rate and sporulation capacity
Growth rate of parental isolates was determined on RDA in 90 mm Petri dishes. Diameters of cultures were measured after 5 and 7 days. Three plates were analysed for each isolate and mean values calculated. Maximal growth rate μmax (mm day−1) was calculated as ; statistical significance was evaluated with Tukey’s test by sigmaplot v. 11.0 software. Sporulation of parental isolates was measured on RDA in 90 mm plates incubated in the dark at 19°C for 7 and 10 days. Plates were washed with 5 mL water; the sporangia concentration was determined with a Neubauer counting chamber. For each isolate three plates were evaluated and mean values calculated. Sporulation capacity K (sporangia mm−2 day−1) was calculated as and analysed by Tukey’s test with the statistical software sigmaplot v. 11.0.
The number of oospores was determined in 10 A1 × A2 crosses (110 × 181, 110 × 171, 110 × 157, 110 × 101, 80 × 181, 80 × 171, 80 × 157, 76 × 171, 74 × 171 and 74 × 101) according to a modified method developed by Rubin & Cohen (2006). Three leaflets of tomato cv. Baby (6-week-old plants, chosen from leaf numbers 5–8 from below) were placed upside down in Petri dishes (145 mm diameter) containing moist filter paper (6 mL water). Sporangia suspension was produced by adding cold water to 10–13-day-old P. infestans colonies on RDA plates (50 mm diameter). The suspension was filtered through three layers of hydrophilic fabric into an Erlenmeyer flask on ice. Concentrations of sporangia were determined with a Neubauer counting chamber and adjusted to 2 × 104 mL−1. For each cross, 1 mL of each A1 and A2 sporangia suspension was mixed (1:1), vortexed for 5 s (Genie2™; Bender & Hobein AG) and sprayed with an atomizer on tomato leaflets (three leaflets per cross). As a control, two leaflets were inoculated with sporangia of each single mating type separately. For each cross, three replicates were produced with fresh cultures. Petri dishes were then placed in Plexiglas boxes (34 × 30 × 10 cm) and incubated at 19°C and 70% RH for 14 days with a day length of 14 h. At harvest, three discs (12 mm) from each of the three leaflets (total of nine discs per cross) were cut and homogenized in 5 mL water for approximately 30 s with an Ultra turrax T25-Mixer (Janke & Kunkel, IKA Lab techniques) in 50 mL Cellstar tubes (Greiner BioOne). Homogenizing facilitated oospore release from the rotten leaf material. One millilitre of homogenate was bleached with 500 μL of 10% sodium hypochlorite solution (Sigma-Aldrich) for 15 min to facilitate the detection of oospores. Ten samples of oospores (10 μL droplets) were counted microscopically (Zeiss Axiolab, Carl Zeiss GmbH; Fig. 1), mean values calculated, and analysed by Tukey’s test with the sigmaplot v. 11.0.
Generation of F1 progeny
To produce F1 progeny isolates, 5 mL aliquots of leaf homogenate containing oospores were pipetted into Petri dishes (90 mm) and dried in a strongly ventilated hood (60–100 lfm) for 24 h to remove the liquid and destroy infective vegetative structures as described by Rubin & Cohen (2006). Then, 10 mL of water was added to each dish to resuspend the dried material and it was left to dry again for 24 h. Finally, 20 mL of water and 2 g of PePe® plantperlite (Otto Hauenstein Samen AG) were added to the dried homogenates. Two tomato leaflets per Petri dish were placed with lower leaf side touching the surface of the perlite substrate (Fig. 2a). Three Petri dishes were incubated for each cross in Plexiglas boxes under the same conditions as described above. Petri dishes were incubated until infection sites on leaflets could be detected (after 3–15 days, depending on specific cross; Fig. 2b). Then the leaflets were removed, plates dried for 24 h and 20 mL water and two fresh leaflets added and incubated again. This procedure (cycle) was repeated after the appearance of each new infection site and continued for a total of 72 days, resulting in 7–24 cycles per cross (depending on productivity of crosses).
As soon as infection sites were seen on the upper side of the leaves, a single sporangiophore per site was picked with forceps (Fig. 2c), transferred to a 10 μL droplet of 4°C sterile water and placed on the lower side of fresh tomato leaflets to produce new infection sites. As soon as sporulation was observed, leaflet pieces were put between potato halves as described by Knapova & Gisi (2002); infected tuber tissue was picked later to produce F1 progeny isolates on RDA plates. Instead of statistically analysing the cycles independently, the total number of progeny produced over 72 days per cross was taken as one bulk sample and calibrated to an initial inoculum density of 104 oospores to compare crosses, thus making calculations of standard deviations impossible.
Ploidy (genome size and molecular markers) of parental isolates
Genome size of parental isolates was measured by flow cytometry. As diploid standard isolates, PiMex-3-1 and F80029 were included in this study. Genome 2C DNA contents (pg) of the parental and diploid reference isolates are shown in Figure 3. Isolates could statistically be distinguished into two groups. The first included isolates 157, 80, 171, 76, PiMex3-1, 74, F80029 and 181 (arranged according to rising genome size); they are defined as 2n with a median of 0·68 ± 0·04 pg (ranging from 0·594 to 0·821 pg). Isolates 110 and 101 represented the second group of isolates with a median of 0·99 ± 0·06 pg (ranging from 0·874 to 1·065) and are called 3n because their genome size is approximately a third bigger than the 2n median.
SSR genotypes of parental isolates were assigned according to the nomenclature developed by Knapova & Gisi (2002) and Gisi et al. (2011; Table 1). Isolate 110 (genotype D-10) showed three alleles each at locus Pi02 (149, 157, 159) and at SSR4 (285, 293, 297) and is classified as SSR triploid. Also isolate 101 (genotype G-4) showed three alleles (150, 156, 160) but at locus PiG11 and is therefore SSR triploid (Table 2). All other isolates (genotypes B-5, B-3, D-2 and G-10) showed two alleles at all tested loci and are therefore classified as SSR diploid. Ploidy of parental isolates was also evaluated based on the presence and quantification of SNP F382Y in the RNA polymerase I gene. Isolates 110 and 101 showed an A/T frequency of about 1:2 and were classified as SNP triploid, whereas isolates 74, 76, 80 and 181 showed an A/T frequency of about 1:1 and were considered as SNP diploid (Table 2). Allele quantification was not possible for the other four isolates because they lacked the mutation and were classified as wildtype. For all other isolates, the two molecular markers (SSR and SNP F382Y) yielded identical ploidy levels and also perfectly matched the 2C DNA contents (Table 2).
|Isolate||Mating type||SSR genotype||SSR ploidy||SNP F382Y A/T frequency||SNP ploidy||Genome size (pg)||Flow cyt. ploidy|
|110||A1||D-10||3n||42/58||3n||0·95 ± 0·08||3n|
|74||A1||B-5||2n||52/48||2n||0·70 ± 0·04||2n|
|76||A1||B-5||2n||53/47||2n||0·67 ± 0·06||2n|
|80||A1||B-5||2n||53/47||2n||0·65 ± 0·02||2n|
|157||A2||B-3||2n||0/100||Wildtype||0·65 ± 0·02||2n|
|171||A2||D-2||2n||0/100||Wildtype||0·66 ± 0·06||2n|
|101||A2||G-4||3n||35/65||3n||1·03 ± 0·04||3n|
|181||A2||G-10||2n||54/46||2n||0·78 ± 0·04||2n|
|F80029||A1||nd||2n||100/0||Wildtype||0·66 ± 0·04||2n|
|PiMex-3-1||A2||nd||2n||0/100||Wildtype||0·65 ± 0·02||2n|
Growth rate and sporulation capacity of parental isolates
Maximal growth rates μmax were determined for parental isolates. Statistically significant (P < 0·05) differences between isolates were observed. Isolates 76, F80029, 181, 80 and 74 showed the highest growth rates with 9·36 ± 0·46, 9·17 ± 0·55, 9·11 ± 0·89, 8·28 ± 0·50 and 8·08 ± 0·63 mm day−1, respectively, followed by isolates Mex-3-1, 171 and the two triploid isolates 110 and 101 with clearly lower growth rates (6·61 ± 0·13, 5·78 ± 0·19, 6·44 ± 0·67 and 6·39 ± 0·24 mm day−1); the lowest growth rate was observed for isolate 157 with 2·22 ± 0·32 mm day−1 (Fig. 4).
Sporulation (sporangia production) was determined for parental isolates in 7 and 10-day-old cultures. Statistically significant (P < 0·05) differences in sporulation capacity K between isolates were observed (Fig. 5). Isolates 171 and 74 showed the highest sporulation capacity with 428·87 ± 44·08 and 293·00 ± 79·47 sporangia mm−2 day−1, respectively, followed by isolates 157, 80, 76, F80029, Mex-3-1 and the triploid isolate 110 (ranging from 130·00 ± 47·40 to 42·70 ± 29·87 sporangia mm−2 day−1). Isolate 181 and the triploid isolate 101 showed the lowest sporulation capacity (23·29 ± 10·08 and 15·52 ± 6·72 sporangia mm−2 day−1), respectively. Growth rate and sporulation capacity are generally considered as important (but independent, asexual) fitness parameters and are often multiplied to give a general indication of competitiveness in populations. Isolates 171 and 74 were strongest (fitness product, FP, about 2400), followed by isolates 76, F80029 and 80 (FP about 700); the other isolates were weaker (FP < 300).
Sporulation capacity, growth rate and genome size (pg) were plotted as three different pairs against each other. No correlations were found between sporulation capacity and growth rate (R2 = 0·04), sporulation capacity and genome size (R2 = 0·06), and growth rate and genome size (R2 = 0·00).
Oospore production was determined microscopically for 10 out of 16 possible crosses (with each four A1 and A2 isolates). The 3n × 3n cross 110 × 101 resulted in a rather low number of oospores (5 oospores mm−2, Fig. 6); about the same number was produced by the two 2n × 3n crosses 110 × 157 and 74 × 101, and the two 2n × 2n crosses 80 × 171 and 76 × 171. Significantly higher numbers of oospores were produced by cross 80 × 157 (2n × 2n), with about 10 oospores mm−2, and the two crosses 110 × 181 (3n × 2n) and 80 × 181 (2n × 2n), with about 15 oospores mm−2. Crosses 110 × 171 (3n × 2n) and 74 × 171 (2n × 2n) produced the highest number of oospores (about 24 mm−2; Fig. 6). Thus, both low and high numbers of oospores were produced in 2n × 2n and 3n × 2n crosses and in crosses with the same parents, e.g. isolates 110, 171 and 74. Control leaflets inoculated with single isolates did not show any oospore formation.
Generation of F1 progeny
Sporangiophores of sporulating infection sites on leaflets were continuously picked and transferred during a time period of 72 days, thus allowing generation of infectious, pathogenic F1 progeny isolates in pure culture; with this procedure both weakly and strongly pathogenic (aggressive) isolates were recovered because there was no competition among them during picking. Drying the perlite cultures after each cycle obviously killed sporangia potentially produced by former F1 colonies, because barely any clones were detected in F1 isolates picked from the same perlite culture. For each cross the number of offspring produced in perlite cultures (7–24 plates per cross) over 72 days was totalled. The total number of progeny isolates per cross was then calibrated to a starting inoculum density of 104 oospores. Cross 110 × 101 (3n × 3n) produced only very few infection sites and F1 isolates, followed by the 3n × 2n crosses 110 × 157, 74 × 101, 110 × 171 and 110 × 181 (<10 F1 isolates; Fig. 7). The 2n × 2n crosses (80 × 171, 80 × 157, 76 × 171, and especially 74 × 171 and 80 × 181) produced the highest numbers of pathogenic F1 progeny isolates (10–24 per 104 oospores after 72 days; Fig. 7). The production of infection sites had already started after 3–6 days (74 × 171, 76 × 171) or 9–15 days (other crosses) and progressed more or less continuously, although slowed down somewhat at the end of the 72 days of incubation.
Ploidy of F1 isolates
In total, 319 F1 progeny isolates were produced from 10 crosses, of which six were very productive: the four 2n × 2n crosses 74 × 171 (N = 96), 80 × 171 (N = 58), 76 × 171 (N = 52) and 80 × 181 (N = 29), and two of the four 2n × 3n crosses 110 × 171 (N = 43) and 74 × 101 (N = 25). Ploidy was determined with all three methods for a total of 63 progenies, i.e. 20 from 2n × 2n crosses, 39 from 2n × 3n and four from 3n × 3n crosses. For technical reasons, the remaining isolates could not be analysed. The results obtained with flow cytometry are shown in Figure 8. Overall, median 2C DNA contents for 2n (N = 59) and 3n (N = 4) isolates were 0·69 ± 0·03 pg and 1·01 ± 0·04 pg, respectively. In 2n × 2n crosses, the median for 2n isolates was 0·70 ± 0·03 pg (with 19 offspring ranging from 0·578 to 0·813 pg) and one triploid isolate with a 2C DNA content of 0·98 (0·935–1·028) pg (Fig. 8).
In 2n × 3n and 3n × 3n crosses, the median for 2n isolates was 0·77 ± 0·07 pg (with 36 offspring ranging from 0·603 to 0·956 pg); five triploid offspring (three and two from 2n × 3n and 3n × 3n crosses, respectively) with a median of 1·03 pg (ranging from 0·952 to 1·110 pg) 2C DNA content were found (Fig. 8). The 36 diploid isolates might be subdivided into a ‘typical diploid part’ (N = 28, median 0·74 ± 0·03 ranging from 0·603 to 0·858 pg) and an ‘upper part’ with potentially trisomic isolates (N = 8, median 0·87 ± 0·04 ranging from 0·813 to 0·947 pg) (Fig. 8, Table 3). In the 3n × 3n cross, the four offspring were classified as two triploid and one each trisomic and diploid.
|2n × 2n||N||%||2n × 3n||N||%|
Most of the infection sites produced in the 10 crosses yielded offspring that could be transferred to RDA and analysed further with molecular markers.
Inheritance of mating type and ploidy
In total, 319 offspring were analysed for mating type. The segregation was close to 1:1 with 166 A1 and 153 A2 isolates; in only two of 10 crosses, slightly more A2 than A1 isolates were produced. The inheritance pattern of ploidy was further analysed for 151 offspring. A total of 13 isolates provided identical ploidy results if all three methods were considered (Table 3). All offspring from 2n × 2n crosses were diploid, whereas in 2n × 3n crosses, 66% were diploid and 34% triploid (Table 3). Based on flow cytometry data alone, 2n × 2n and 2n × 3n crosses resulted mainly in diploid offspring, with only a few triploid. If diploid isolates are subdivided into two groups, 72% are considered as ‘true’ diploid and 20% as trisomic (Table 3). Based on analysis with the two molecular markers combined, 2n × 2n crosses resulted in diploid offspring (98%), whereas 2n × 3n resulted in two-third diploid and one-third triploid offspring.
The aim of this study was to investigate whether ploidy of P. infestans isolates has a relevant impact on the production, germination and pathogenicity of oospores produced in different crosses in plant tissue. Ploidy was measured with three different methods: (i) flow cytometry, recording the weight of nuclear DNA, (ii) molecular marker technologies including non-coding markers (presence of specific alleles in simple sequence repeats, SSR) and (iii) quantification of an SNP, F382Y, in the RNA I synthase gene possibly coding for resistance to phenylamide fungicides (e.g. mefenoxam). The study was done with 10 parental P. infestans isolates differing in their genome size. Two diploid isolates, PiMex-3-1 and F80029, served as standards with measured 2C DNA contents of 0·70 ± 0·02 and 0·71 ± 0·05 pg, respectively. Although from different sources, their genome size was identical, confirming the reliability of the flow cytometry method used. Six parental isolates did not significantly differ from these values and were therefore classified as diploid (isolates 74, 76, 80, 157, 171 and 181). In addition, isolates 110 and 101 showed a genome size of 0·95 ± 0·08 and 1·03 ± 0·04 pg, respectively, which is about one-third bigger than the median for diploid isolates, suggesting that these two isolates are triploid.
However, trisomic instead of triploid constellation cannot be excluded for these two isolates. Mobile elements such as RXLR effector genes can contribute to the dynamic nature of repetitive regions, enabling recombination events and higher rates of gene gain and gene losses observed in P. infestans (Haas et al., 2009). Microscopic observations of chromosomes may be needed to verify these features (Ritch & Daggett, 1995). However, molecular markers were used in this study to determine ploidy and compared with flow cytometry.
Interestingly, diploid and triploid (and possibly also trisomic) offspring emerged from 2n × 2n, 2n × 3n and 3n × 3n crosses otherwise seen in earlier studies only from 2n × 4n crosses (Whittaker et al., 1991b). Nevertheless, the majority of offspring in the crosses here showed diploid genome sizes. The trapping method did not (by definition) yield non-pathogenic offspring, but if these occurred at all in nature, they would not play an important role in the disease cycle anyway. However, some fitness penalty (low number and germination of oospores and low sporulation capacity) was observed in some parental isolates used for 2n × 3n and 3n × 3n crosses, probably representing an evolutionary disadvantage. To verify this hypothesis, crosses between F1 offspring isolates should be done in future investigations. Similar fitness penalties were demonstrated recently for oosporic progeny of P. infestans crosses in tomato lines (Klarfeld et al., 2009).
In this study, seven different SSR loci were used to characterize parental and F1 progeny isolates. So far it is not possible to allocate a specific locus to the corresponding chromosome; nevertheless, according to the P. infestans database of the Broad Institute [http://www.broadinstitute.org/annotation/genome/phytophthora_infestans/MultiHome.html], each locus is located on a different contig. Six field isolates and the two standard isolates showed two alleles in all tested SSR loci as expected for diploid isolates. The two (flow-cytometrically) triploid isolates, 110 and 101, showed three alleles in at least one SSR locus but two alleles in all other loci; they were considered as ‘SSR-triploid’ in this study. The fact that they do not show three alleles in every locus raises some doubts on their triploidy; they may be considered as trisomic rather than triploid, as was shown for other isolates in recent studies (Van der Lee et al., 2004; Catal et al., 2010). In an earlier study, more than three peaks were found in one locus for some P. infestans isolates (Knapova & Gisi, 2002). Therefore, ploidy measurements based on SSR analysis should be treated with some caution. As a consequence, more SSR loci, preferably associated to specific chromosomes, should be tested to receive more reliable results.
Only a few coding molecular markers are known in the P. infestans genome, of which SNP F382Y may be suitable for ploidy measurements. However, ploidy cannot be determined in isolates (e.g. 157 and 171) lacking the mutation (wildtype isolates). Some (cytometrically) diploid isolates (74, 76 and 80) showed an A/T frequency of 1:1 supporting these isolates to be diploid; isolate 101 showed an A/T frequency of 1:2 confirming its triploidy. For reliable ploidy determinations with pyrosequencing, more SNPs should be used, preferably located on different chromosomes. However, it is very hard to find conserved SNPs and the method is simply not suitable if isolates lack the mutations. Both molecular methods, SSR and pyrosequencing of SNP F382Y, tend to be less reliable than flow cytometry; flow cytometry provides the average nuclear DNA content of about 10 000–20 000 nuclei measured individually at high speed, typically at 100–1000 s−1, which facilitates representative sampling (Dolezel et al., 2007). However, the molecular results were based on ‘only’ seven loci for SSR and one SNP for allele quantification. On the other hand, flow cytometry does not provide any information on possible trisomy in certain chromosomes which may be detected only by molecular markers. Although the genome of P. infestans has been sequenced, it is important to connect this knowledge to chromosomes and find reliable molecular markers for each chromosome, enabling detection of chromosomal arrangement without the need for difficult microscopy. As many AFLP markers have already been described in P. infestans (Van der Lee et al., 2004), they could be considered as an additional molecular method to determine ploidy.
Growth rate and sporulation capacity statistically differed between parental isolates but did not correlate to each other nor to genome size. Low growth rate can be combined with high sporulation capacity as seen in isolates 157 and 171. It is interesting that the two triploid isolates did not show higher fitness than diploid isolates; their sporulation capacity was rather low and their growth rate average. Hence, fitness of an isolate in the field is more than in vitro mycelial growth rate and sporulation capacity and may be defined as the ability to produce pathogenic (sexual and asexual) progeny. The triploid isolate 101 (a ‘blue 13’ type isolate with allele 154 present in locus D13 and claimed to be more aggressive; Lees et al., 2009), showed a rather low sporulation capacity representing a potential indicator of low fitness in the field, and when used in crosses, did not produce many offspring. Therefore, ‘blue 13’ type isolates may not compete in natural selection for very long (Gisi et al., 2011).
It is striking to note that oospore formation was observed in all tested crosses including those with one or both parents being triploid. Although the number of oospores produced per square millimetre of leaf tissue differed significantly among crosses, it could not be associated to ploidy of parents. Obviously, sexual recombination is also possible among triploid P. infestans individuals, as was reported for triploid green toads (Bufo viridis; Stöck et al., 2002). Crosses with 2n × 2n or 2n × 3n parents can produce both low and high amounts of oospores depending on the fertility and match of crossing partners. This is an unexpected result because triploid P. infestans isolates were assumed to have difficulties in sexual recombination (chromosome division and multiplying) as in other organisms. Plants solved the problem of higher ploidy through tetraploidization or through apomictic reproduction (Van Baarlen et al., 2000). To get more insight into chromosomal rearrangement and inheritance of chromosomal traits during sexual recombination of triploid isolates, molecular markers located on each chromosome would be of great value.
As outlined above, higher ploidy of parental isolates did not have an impact on oospore formation, but it clearly affected the emergence of infection sites. Cross 110 × 101 (3n × 3n) produced only low numbers of infection sites. Also, all four 2n × 3n crosses produced significantly less infection sites and progeny than 2n × 2n crosses, although some 2n × 3n crosses (e.g. 110 × 171) showed a high amount of oospores. Thus, it can be concluded that crosses containing at least one triploid parent produce less infection sites (per given number of oospores) than 2n × 2n crosses. This is a remarkable result because it suggests that triploid isolates contribute very little to the sustainable evolution of F1 progeny under natural selection. The production of highly pathogenic and viable progeny is obviously best in 2n × 2n crosses supporting earlier in vitro results (Whittaker et al., 1991b). Although not measured in detail as for parental isolates, triploid offspring obviously grew much less vigorously compared to diploid offspring. Therefore, triploid genotypes may emerge occasionally as recently observed in European populations, but may be displaced by highly fit diploid individuals after short selection periods. For evaluating the importance of sexual recombination in field populations, it is much more important to investigate how many pathogenic offspring will emerge from oospores and how fit they are than just counting oospores in field samples (soil, leaves). In future studies, latent period, growth rate, sporulation capacity, aggressiveness and fertility should be tested for progeny isolates to evaluate whether these parameters can be associated to ploidy. Also, more molecular markers should be developed and used in future studies to get more information about inheritance of chromosomes and mechanisms of sexual recombination in P. infestans isolates. This may offer answers to the basic question of how exactly triploid and trisomic isolates evolve in pre- and postfusion events of sexual recombination.
The authors acknowledge the support of Professor Dr Dirk Bumann’s research group of the University of Basel, Switzerland, especially Beatrice Müller, for access to the FACS machine and technical advice; Dr Jaroslav Dolezel, Institute of Experimental Botany in Olomouc, Czech Republic, for providing seeds as internal standards for flow cytometry; Dr William W. Kirk and Pavani Tumbalam, Michigan State University, USA, and Professor Francine Govers and Trudy van den Bosch, Laboratory of Phytopathology, Wageningen University, the Netherlands, for providing diploid standard P. infestans isolates; and Annelies Vercauteren, Department of Plant Production, Ghent University, Belgium, for advice on performing flow cytometry analysis.
- 1999. The detection of nonhybrid, trisomic and triploid offspring in sexual progeny of a mating of Phytophthora infestans. Fungal Genetics and Biology26, 198–208. , , , , , ,
- 2010. Heterokaryotic nuclear conditions and a heterogeneous nuclear population are observed by flow cytometry in Phytophthora infestans. Cytometry77A, 769–75. , , , , , ,
- 2007. Estimation of nuclear DNA content in plants using flow cytometry. Nature Protocols9, 2233–44. , , ,
- 2011. Changes of genotype, sensitivity and aggressiveness in Phytophthora infestans isolates collected in European countries in 1997, 2006 and 2007. Journal of Phytopathology159, 223–32. , , , , ,
- 2005. The origin, evolution and proposed stabilization of the terms ‘genome size’ and ‘C-value’ to describe nuclear DNA contents. Annals of Botany95, 255–60. , , , ,
- 2009. Genome sequence and analysis of the Irish potato famine pathogen Phytophthora infestans. Nature461, 393–8. , , et al. ,
- 1995. Ainsworth & Bisby’s Dictionary of the Fungi. Wallingford, UK: CAB International. , , , ,
- 2009. Pathogenic fitness of oosporic progeny isolates of Phytophthora infestans on late blight-resistant tomato lines. Plant Disease93, 947–53. , , ,
- 2002. Phenotypic and genotypic structure of Phytophthora infestans populations on potato and tomato in France and Switzerland. Plant Pathology51, 641–53. , ,
- 2006. Novel microsatellite markers for the analysis of Phytophthora infestans populations. Plant Pathology55, 311–9. , , , , , ,
- 2009. Phytophthora infestans population changes: implications. In: Schepers HTAM, ed. Proceedings of the 11th Workshop of the European Network for Development of an Integrated Control Strategy of Potato Late Blight. Hamar, Norway: PPO-Special Report, Vol. 13, 55–60. , , , , , ,
- 2006. Occurrence and distribution of mating types A1 and A2 of Phytophthora infestans (Mont.) de Bary in the Czech Republic. Plant Protection Science42, 41–8. , , , , , ,
- 1995. Nuclear DNA content and chromosome number in German isolates of Phytophthora infestans. Mycologia87, 579–81. , ,
- 2006. An improved method for infecting tomato leaves or seedlings with oospores of Phytophthora infestans used to investigate F1 progeny. Plant Disease90, 741–9. , ,
- 1977. Polyploidy and induced gametangial formation in British isolates of Phytophthora infestans. Journal of General Microbiology99, 311–6. ,
- 2002. A bisexually reproducing all-triploid vertebrate. Nature Genetics30, 325–8. , , et al. ,
- 1987. Cytophotometric determination of the nuclear DNA content of 23 Mexican and 18 non-Mexican isolates of Phytophthora infestans. Experimental Mycology11, 19–26. , ,
- 2000. Meiotic recombination in sexual diploid and apomictic triploid dandelions (Taraxacum officinale L.). Genome43, 827–35. , , , ,
- 2004. High-density linkage maps of Phytophthora infestans reveal trisomic progeny and chromosomal rearrangements. Genetics167, 1643–61. , , , , ,
- 2011. Aberrant genome size and instability of Phytophthora ramorum oospore progenies. Fungal Genetics and Biology5, 537–43. , , et al. ,
- 1991a. Variation in DNA content of nuclei of Phytophthora infestans as measured by microfluorimetric method using the fluorochrome DAPI. Mycological Research95, 602–10. , , ,
- 1991b. Inheritance of DNA contents in sexual progenies of Phytophthora infestans. Mycological Research95, 1094–100. , , ,
- 2002. The single, ancient origin of chromist plastids. Proceedings of the National Academy of Sciences, USA99, 15507–12. , , , ,
- 2006. Pyrosequencing, a high-throughput method for detecting single nucleotide polymorphisms in the dihydrofolate reductase and dihydropteroate synthase genes of Plasmodium falciparum. Journal of Clinical Microbiology11, 3900–10. , , et al. ,