Occurrence of the F129L mutation in Alternaria solani populations in Germany in response to QoI application, and its effect on sensitivity

Authors


Abstract

Early blight caused by Alternaria solani is a highly destructive disease of potatoes. Control of early blight mainly relies on the use of preventive fungicide treatments. Because of their high efficacy, azoxystrobin and other quinone outside inhibitors (QoIs) are commonly used to manage early blight. However, loss of sensitivity to QoIs has previously been reported for A. solani in the United States. Two hundred and three A. solani field isolates collected from 81 locations in Germany between 2005 and 2011 were screened for the presence of the F129L mutation in the cytochrome b gene; of these, 74 contained the F129L mutation. Sequence analysis revealed the occurrence of two structurally different cytb genes, which differed in the presence (genotype I) or absence (genotype II) of an intron, with genotype I being the most prevalent (63% of isolates). The F129L mutation was detected only in genotype II isolates, where it occurred in 97%. Sensitivity to azoxystrobin and pyraclostrobin was determined in conidial germination assays. All isolates possessing the F129L mutation had reduced sensitivity to azoxystrobin and, to a lesser extent, to pyraclostrobin. Early blight disease severity on plants treated with azoxystrobin was significantly higher for A. solani isolates with reduced fungicide sensitivity in the conidial germination assay compared with sensitive isolates. Data suggest an accumulation of F129L isolates in the German A. solani population over the years 2009–2011. It is assumed that the application of QoIs has selected for the occurrence of F129L mutations, which may contribute to loss of fungicide efficacy.

Introduction

Early blight (EB), caused by Alternaria solani, is one of the most common and important foliar diseases of potato (Solanum tuberosum). EB, which is widespread in potato growing areas, mainly infects potato foliage and leads to premature and progressive defoliation (van der Waals et al., 2004). Destruction of foliage decreases plant photosynthesis and reduces storage of tuber assimilates (e.g. starch). Yield losses can reach up to 30% (Shtienberg et al., 1990; Leiminger & Hausladen, 2012), reducing the profitability of potato production. Reliable disease management is mainly based on multiple fungicide application (Pscheidt & Stevenson, 1988). Prior to the registration of azoxystrobin (which belongs to the QoI group) for potato in 2007 in Germany, EB control was mainly achieved by multiple and frequent application of protectant fungicides (Rosenzweig et al., 2008). However, these fungicides containing chlorothalonil or mancozeb were only moderately effective in the control of EB and guaranteed adequate disease control only under moderate disease pressure (Stevenson & James, 2000). In contrast, QoI fungicides turned out to be highly effective against EB (Pasche et al., 2004).

Quinone outside inhibitors (QoIs), also known as strobilurins, are an important class of fungicides in agriculture because they have a broad-spectrum activity, low rates of use and excellent yield benefits (Bartlett et al., 2002). They inhibit mitochondrial respiration in fungi by binding to the Qo site of the cytochrome b (cytb) complex, blocking electron transfer and inhibiting ATP synthesis (Bartlett et al., 2002). However, as these fungicides have a specific single-site mode of action, they are more vulnerable to the evolution of fungicide resistance than fungicides with multisite activity (e.g. chlorothalonil, mancozeb; Kuck & Mehl, 2003).

Since the registration of the first QoI fungicide azoxystrobin (Syngenta Crop Protection) for cereals in the USA in 1996 and 2 years later for potatoes, potato producers have taken advantage of the high level of disease control (Pasche et al., 2004). Ortiva® (azoxystrobin) and Signum® (pyraclostrobin + boscalid) were registered in Germany as EB-specific fungicides in potatoes in 2007 and 2008, respectively.

The occurrence of fungicide resistance is a serious problem affecting disease management in agricultural ecosystems. Fungicide resistance can be defined as the stable and acquired resistance by a pathogen to a fungicide, resulting in reduced sensitivity or tolerance of the pathogen to the fungicide (Steinfeld et al., 2001). Reduced EB control was first observed in 2000 in the USA, where inadequate control by azoxystrobin was caused by a shift in fungicide sensitivity of A. solani (Pasche et al., 2004). A survey of A. solani isolates from North Dakota, Nebraska and Wisconsin showed a 10-fold decrease in sensitivity. The main mechanism of resistance to QoIs has been identified as mutations in the mitochondrial target gene, cytb (Sierotzki et al., 2007). Single nucleotide polymorphisms (SNP) in the fungal cytb gene, which result in amino acid substitutions, can prevent the fungicide binding to the cytb protein and results in fungicide resistance or reduced sensitivity (Sierotzki et al., 2007). The most important resistance mechanism to QoI fungicides that has been described so far is related to the target site mutation glycine to alanine at position 143 (G143A) in the cytb protein (Sierotzki et al., 2000b). Beyond G143A, F129L and G147R are the only point mutations that have been detected in plant pathogenic fungi which are responsible for QoI resistance (Fisher & Meunier, 2008). In A. solani, only the F129L amino acid substitution of phenylalanine (F) to leucine (L) at position 129 has been observed (Pasche et al., 2004). This substitution is caused by one of three nucleotide mutations (TTA, CTC and TTG). In contrast to the G143A mutation, which leads to a complete loss of sensitivity to QoI fungicides, the F129L and G137R mutations result in only moderate levels of resistance (Bartlett et al., 2002; Kuck & Mehl, 2003; Yamada & Sonoda, 2012).

Development of fungicide resistance in populations of Alternaria species has become an increasingly important issue. As a result of the intensive use of azoxystrobin and other QoI fungicides on commercially grown potato crops in the United States, the selective advantage of QoI resistant mutants in A. solani populations has increased dramatically (Rosenzweig et al., 2008). Similar to other pathogens, A. solani represents a ‘high-risk’ pathogen in terms of resistance development as a result of its high genetic variability, abundant sporulation, polycyclic nature and the repeated application of sprays required to control the disease (Pasche et al., 2004; van der Waals et al., 2004; Rosenzweig et al., 2008). Accordingly, understanding the evolutionary basis of the emergence and dispersal of fungicide resistance is of primary interest. Monitoring the emergence of resistant genotypes at an early stage would thus support the adaptation of control strategies and might be a useful tool in the prediction of further resistance development. Additionally, the implementation of anti-resistance strategies would contribute to an increased life expectancy for fungicides.

Up to now, little is known about the prevalence of A. solani isolates with reduced sensitivity to QoI fungicides outside the USA. The present study was carried out in order to determine whether a reduction in sensitivity to azoxystrobin has already occurred within German A. solani populations. The objectives of this research were to determine: (i) the prevalence of the F129L substitution among A. solani isolates collected from conventional potato fields throughout Germany between 2005 and 2011; (ii) the in vitro sensitivity of A. solani isolates to azoxystrobin and pyraclostrobin with and without a history of QoI exposure (sampled before and after QoI registration in potato); and (iii) the effect of reduced sensitivity to azoxystrobin of A. solani isolates on early blight disease control.

Materials and methods

Isolate sampling

Two hundred and three field isolates of A. solani were sampled between 2005 and 2011. All isolates were collected from commercial potato crops naturally infected with EB. Table 1 lists their places of origin. Infected leaflets were sampled at random during EB disease epidemics between July and September of each year. Material was surface sterilized in 5% NaOCl for 1 min and then washed in sterile distilled water. One single conidium was transferred to synthetic low nutrient (SN) media (1 g KH2PO4, 1 g KNO3, 0·5 g MgSO4.7H2O, 0·5 g KCl, 0·2 g glucose, 0·2 g saccharose, 0·6 mL 1 m NaOH, 20 g agar, in 1 L double distilled water) and only one isolate was collected per diseased plant. Alternaria solani cultures were first identified on the basis of morphological characteristics and spore size (Simmons, 2007). Isolates grew on SN agar for 10 days to colonize and to form conidia. For long-term storage, small sections (0·5 cm) were cut from colonies and placed in 2% malt extract mixed with 10% glycerol. Stocks were stored at −80°C for further analysis in order to minimize any effects of subculturing. Isolates collected in 2005 and 2006 were defined as the baseline group. Isolates sampled in 2007, the year when azoxystrobin was registered for use in potato in Germany, were defined as the non-baseline group. Seven A. solani strains (1174-1, 13-1, 526-3, 1172-8, 1178-W1, 1189-7 and 1190-16), kindly provided by Neil Gudmestad (North Dakota State University) were included in the investigations and designated as US reference isolates.

Table 1. Collection data of Alternaria solani isolates collected from fields (2005–2011) from different sites
Isolate characterizationNumber of isolates
State of originNo. of different locationsAzoxystrobin exposureTotalWildtype identification by:Mutant identification by sequencing
Real-time PCR LightCyclerSequencing
  1. a

    Isolates collected before registration of azoxystrobin, isolates collected in subsequent years were not necessarily collected from the same sites as baseline isolates.

  2. b

    Isolates collected after azoxystrobin received emergency registration.

2005 (baseline group)a
Bavaria1No633 
2006 (baseline group)a
Bavaria16No17116 
North Rhine-Westphalia1No1 1 
Lower Saxony3No321 
Schleswig-Holstein1No11  
Mecklenburg-West Pomerania2No22  
2007 (non-baseline group)b
Bavaria9Yes17512 
Baden-Wuerttemberg1Yes1 1 
Mecklenburg-West Pomerania1Yes11  
Lower Saxony1Yes11  
2008 group
Bavaria5Yes541 
2009 group
Bavaria14Yes3110192
Lower Saxony5Yes725 
Saxony-Anhalt1Yes1 1 
2010 group
Bavaria4Yes6312
North Rhine-Westphalia1Yes22  
Lower Saxony2Yes211 
2011 group
Bavaria11Yes95 2570
Lower Saxony1Yes2 2 
Schleswig-Holstein1Yes2 2 
 Total81 203488174

DNA extraction

Genomic DNA of A. solani isolates was extracted from mycelia cultivated on V8 medium under near-UV light for 14 days at 21°C. Mycelium and spores were carefully scraped off with a spatula and ground in liquid nitrogen. Genomic DNA extraction was carried out using the Wizard Genomic DNA Purification Kit (Promega Corp.) according to the manufacturer's instructions for isolating genomic DNA from plant tissue, followed by a phenol–chloroform extraction. The DNA was resuspended in sterile H2O. The DNA content was measured with a NanoDrop Spectrophotometer (ND-1000) and adjusted to 2·5 ng μL−1.

Real-time PCR and probe hybridization analysis for the detection of the F129L mutation

The presumptive presence of the F129L mutation in the cytochrome b gene was determined for all sampled isolates. In total, 203 German plus five reference isolates (including sensitive wildtype (13-1, 1174-1) and reduced-sensitive isolates containing the F129L mutation (526-3, 1172-8, 1178-W1, provided by N. Gudmestad and J. Pasche, North Dakota State University, USA) were screened. All sampled isolates were first investigated according to a protocol established by Pasche et al. (2005) by real-time PCR with hybridization probes. For this, a LightCycler 2 thermocycler (Roche) and a LightCycler FastStart DNA Master HybProbe Kit with glass capillaries (Roche) were used. Amplification of a 207 bp fragment containing the possible mutation site using the oligonucleotide primer pair As-5F and As-5R (Pasche et al., 2005) was followed by a melting curve analysis with the hybridization probes Asol-FL and Asol-R640 (Table 2). Reference wildtype (13-1) and a reduced-sensitive strain (526-3) were used as internal controls.

Table 2. Primers and probes used in PCR assays to detect single nucleotide polymorphisms among different cytochrome b genes of Alternaria solani
Primers and probesSequence (5′–3′)Primer set for amplification ofReference
As-5fForward primerAGAACTCTAGTATGAACTATTGGGenotype IIPasche et al. (2005)
As-5rReverse primerACTTCTTGTAGAATATCCTCTTTGenotype IIPasche et al. (2005)
Asol-FLSensor probeGATGGCTACACAGCTTTCCTG-fluoresceinGenotype IIPasche et al. (2005)
Asol-R640Anchor probeLC Red 640-TTACCAACATAGCCCAAAATGGTTT-phosphateGenotype IIPasche et al. (2005)
As-Gf CGGGGACTAATATTTTGATAGenotype IThis work
As-Gr TGTTATTTAACCAAGAATGAAAGenotype IThis work

PCR amplification and sequencing of a cytochrome b fragment for the detection of the F129L mutation

For most of the isolates, the expected 207 bp amplicon could not be amplified with the As-5F and As-5R oligonucleotide primer pair. Therefore, a new set of primers, As-Gf and As-Gr (Table 2), was designed to amplify a 214 bp fragment containing the F129L mutation based on the sequence of the cytochrome b gene (GenBank database, accession no. DQ209284; Grasso et al., 2006).

Standard PCR was performed in a total volume of 25 μL containing 10 × PCR buffer, 1·5 mm MgCl2, 200 μm each dNTP (Fermentas), 0·6 μm each primer, 1 U Taq DNA polymerase (SupraTherm, 5 U μL−1, GeneCraft) and 50 ng genomic DNA as template. PCR conditions consisted of an initial denaturation step at 95°C for 10 min, followed by 31 reaction cycles consisting of denaturation at 95°C for 1 min, primer annealing at 54°C for 30 s and DNA extension at 72°C for 30 s. After a final extension step of 72°C for 10 min, samples were cooled at 4°C. PCR was carried out with a T-Personal thermoycler (Biometra). The obtained product was separated by gel electrophoresis on a 1% agarose gel with 1 × TBE and stained with ethidium bromide. The 214 bp DNA fragments were excised from the gel, extracted with the QIAquick Gel Extraction kit (QIAGEN) and sent to LGC Genomics (Berlin, Germany) for sequencing.

In vitro QoI fungicide sensitivity assay

Fungicide sensitivity was determined using an in vitro plate assay, which is based on germination of conidia on fungicide-amended agar relative to non-amended media (Olaya & Köhler, 1999).

Spores of A. solani were produced by growing isolates on SN media at 20°C under near-UV light (Philips LTD 36W/80) with an alternating 12 h photoperiod. Conidia were dislodged using a sterile glass rod and distilled H2O. The conidial suspension was adjusted to 3 × 104 conidia mL−1 using a haemocytometer. Fifty microlitres of the conidial suspension of each isolate was spread onto the surface of agar plates amended with fungicides at different concentrations (two replications) as well as on fungicide-free control plates (three replications). Plates were incubated under continuous light for 5 h at 28°C. After incubation, the germination of 100 conidia was assessed using a microscope at ×100 magnification (Zeiss).

Analytical standards of azoxystrobin and pyraclostrobin (Sigma–Aldrich) were added to water agar. Stock solutions of active ingredients (100 mg mL−1) were prepared and dissolved in 1 mL acetone. The final concentration of acetone in all media was 0·1% (v/v). Fungicides were added to cooling water agar (15 g agar dissolved in 1 L distilled H2O) after autoclaving and adjusted to different concentrations (0, 0·01, 0·1, 1 and 10 μg mL−1). Fungicide sensitivity was determined by comparing spore germination on water agar amended with azoxystrobin or pyraclostrobin at different concentrations. Fungicide sensitivity assays were performed in the presence of 100 mg L−1 salicylhydroxamic acid (SHAM) dissolved in methanol. Petri dishes with SHAM but without fungicides were used as controls. Fungicide sensitivity was measured as the concentration at which spore germination was inhibited by 50% relative to the untreated control (EC50 value) and was determined for each isolate. Using this spore germination assay, 22 wildtype strains (including eight baseline isolates from the year 2006) and 10 isolates containing the F129L mutation were evaluated for their fungicide sensitivity. US reference isolates were included as internal controls (see 'Real-time PCR and probe hybridization analysis for the detection of the F129L mutation').

In vivo QoI fungicide sensitivity assay

Azoxystrobin was evaluated for its efficacy against a subset of six wildtype and five F129L A. solani isolates in glasshouse trials. In each 15 cm pot, one eye cutting of the early blight susceptible potato cv. Kuras was grown in Einheitserde Typ T (Uetersen) for 3 weeks. Azoxystrobin was applied 24 h prior to inoculation in the greenhouse. Analytical standard of azoxystrobin (99·9%, Sigma-Aldrich) was first dissolved in acetone and the stock solution was further diluted with water to final fungicide concentrations of 0, 10 and 100 μg mL−1. Plants were sprayed at different fungicide concentrations until run-off. Conidia from 14-day-old cultures were used for spray inoculation at a concentration of 4 × 104 spores mL−1. Inoculated potato plants were held in a mist chamber at 98% relative humidity (RH), 22–24°C and a 16 h photoperiod for 48 h, and then at 60–70% RH. Disease severity was rated visually by estimating the infected leaf area of three fully expanded leaves at 7 days after inoculation, and recorded as percentage diseased tissue. Three replications (three pots) were analysed for each isolate and fungicide concentration. In vivo glasshouse tests were carried out twice.

Results

Detection of German A. solani field isolates carrying the F129L mutation

According to their cytb gene structure, two different genotypes were found among German A.  solani isolates. The cytb gene structure of genotype I matched that of the publication of Grasso et al. (2006), consisting of five exons and four introns, with the putative F129L mutation site being located at the beginning of exon 2 (Fig. 1a). For this genotype I, a 214 bp fragment, containing the potential mutation site, could only be amplified with newly designed primer pair As-Gf/As-Gr. Genotype I was the most prevalent among all isolated strains; 127 out of 203 analysed isolates were genotype I (Table 3). Subsequent sequencing of the PCR fragments showed that all genotype I isolates lacked the F129L mutation.

Table 3. Occurrence of genotype I and II among German Alternaria solani isolates between 2005 and 2011 and their association with F129L mutation
YearTotal number of isolatesWildtypeF129L
Genotype IGenotype IIGenotype IGenotype II
200566   
200624231  
20072020   
200855   
20093937  2
2010108  2
201199231 70
Total2031272 74
Figure 1.

Schematic representation of a part of the cytochrome b gene of different German Alternaria solani isolates. (a) Genotype I or ‘European genotype’ (b) genotype II or ‘US genotype’. Arrows indicate the base triplet (underlined and in italics) which is the site of mutation in F129L mutants. A126: alanine at amino acid position 126; G131, G143: glycine at amino acid positions 131 and 143. Horizontal arrows in (a) indicate the position of primers AS-Gf and As-Gr used to amplify a 214 bp PCR fragment for genotyping. Horizontal arrows in (b) indicate the position of primers As-5f and As-5r used to amplify a 207 bp fragment.

In genotype II, intron 1 was missing, and the first exon of this genotype was composed of exon 1 and the first part of exon II as far as codon G131 (Fig. 1b). The sequenced part of the following intron (139 bp) showed 90% similarity with intron bi3 of the Pyrenophora teres cytb gene (Sierotzki et al., 2007). For genotype II a 207 bp fragment containing the putative F129L mutation site could only be amplified with the primer pair As-5f/As-5r. Seventy-six German isolates were genotype II and resembled those described by Pasche et al. (2005), and 74 of these contained the F129L mutation.

In the German A. solani isolates, the F129L mutation was identified only in genotype II, although the F129L mutation was found in some US reference strains, which were genotype I (1189-7, 1190-16).

None of the isolates collected between 2005 and 2008 had the F129L mutation. Within F129L isolates, the TTC to TTA nucleotide exchange was predominantly detected. A CTC mutation was detected for only one isolate. Two isolates which contained the F129L mutation were observed in 2009 (= 39; 2009), both derived from the same field in Laberweinting, Bavaria. In 2010, two further F129L isolates were sampled from two different locations (Kirchheim and Ismaning), whilst in 2011 six different locations (Aiterhofen, Hagelstadt, Mintraching, Laberweinting, Stephansposching and Sallach) yielded F129L isolates. In 2011, the number of isolates carrying the F129L mutation increased strongly. Seventy out of 99 isolates (71%) contained the F129L mutation (Table 1). However, isolates sampled from seven other locations (Donaueschingen, Emsbühren-Mehringen, Freising, Kirchheim (two sites), Lüchow and Straubing), lacked the F129L mutation (= 28). The frequency of isolates lacking F129L was 100% between 2005 and 2008, 95% in 2009, 80% in 2010 (based on only a small number of isolates, = 10) and 28% in 2011 (Table 3).

In vitro assay of isolate sensitivity to QoI fungicides

Fungicide sensitivity was determined by assessing the spore germination for a range of isolates on fungicide-amended and non-amended water agar in the presence of SHAM. All tested isolates could be classified into two groups. One group of 22 isolates with EC50 values <0·10 μg mL−1 azoxystrobin was clearly sensitive. Azoxystrobin sensitive isolates were found in all years and all lacked the F129L mutation.

EC50 values of A. solani baseline isolates collected in 2006 ranged from 0·058 to 0·1 μg mL−1 for azoxystrobin (Table 4). EC50 values tested for A. solani isolates collected after registration of azoxystrobin between 2007 and 2011, ranged from 0·054 to 0·099 μg mL−1 (Table 4). This value was not significantly different from the mean EC50 of tested baseline isolates collected before azoxystrobin registration. In general, sensitivity to azoxystrobin of A. solani isolates lacking F129L did not change over years and was comparable to EC50 values of wildtype strains obtained from the USA (Fig. 2).

Table 4. EC50 values of Alternaria solani isolates and characterization of the cytochrome b genotype collected from fields between 2006 and 2011 from different locations
LocationStateIsolate characterizationGenotypeWildtype isolates EC50 (μg mL−1)F129L isolates EC50 (μg mL−1)
AzoxystrobinPyraclostrobinAzoxystrobinPyraclostrobin
2006 (baseline group)
1 Berg im GauBavariaWildtypeGenotype I0·0590·041  
2 BuirNorth Rhine-WestphaliaWildtypeGenotype I0·0820·046  
3 KirchheimBavariaWildtypeGenotype I0·1000·157  
4 MoosburgBavariaWildtypeGenotype I0·0650·043  
5 NeustadtBavariaWildtypeGenotype I0·0690·041  
6 OtzingBavariaWildtypeGenotype I0·0580·045  
7 ScheyernBavariaWildtypeGenotype I0·0580·045  
8 StraßkirchenBavariaWildtypeGenotype I0·0660·050  
Mean   0·0700·058  
Standard deviation   0·0150·040  
2007 (non-baseline group)
1 KirchheimBavariaWildtypeGenotype I0·0750·049  
2 OberdingBavariaWildtypeGenotype I0·0700·044  
3 StraubingBavariaWildtypeGenotype I0·0780·047  
4 ThonstettenBavariaWildtypeGenotype I0·0600·037  
Mean   0·0710·044  
Standard deviation   0·0080·005  
2008 group
1 DonaueschingenBavariaWildtypeGenotype I0·0800·060  
2 WeihenstephanBavariaWildtypeGenotype I0·0990·053  
Mean   0·0890·056  
Standard deviation   0·0130·005  
2009 group
1 HagelstadtBavariaWildtypeGenotype I0·0650·042  
2 LaberweintingBavariaWildtypeGenotype I0·0550·055  
3 Laberweinting, A.s. 1BavariaF129L isolateGenotype II  0·3260·122
4 Laberweinting, A.s. 2BavariaF129L isolateGenotype II  0·3250·081
5 OberwaltingBavariaWildtypeGenotype I0·0660·046  
6 WeicheringBavariaWildtypeGenotype I0·0710·050  
Mean   0·0640·0480·3260·101
Standard deviation   0·0070·0060·0010·029
2010 group
1 IsmaningBavariaF129L isolateGenotype II  0·7290·370
2 Kirchheim A.s.1BavariaF129L isolateGenotype II  0·3120·122
3 Kirchheim A.s.2BavariaWildtypeGenotype I0·0540·040  
Mean   0·0540·0400·5210·246
Standard deviation   0·2950·175
2011 group
1 AiterhofenBavariaF129L isolateGenotype II  0·1830·061
2 LaberweintingBavariaF129L isolateGenotype II  0·2050·109
3 LüchowSchleswig-HolsteinWildtypeGenotype II0·0540·036  
4 MintrachingBavariaF129L isolateGenotype II  0·2410·079
5 Sallach A.s. 2BavariaF129L isolateGenotype II  0·1080·083
6 Sallach A.s. 6BavariaF129L isolateGenotype II  0·5720·060
7 StephansposchingBavariaF129L isolateGenotype II  0·2620·116
8 Straubing A.s. 1BavariaWildtypeGenotype I0·0710·053  
9 Straubing A.s. 2BavariaWildtypeGenotype I0·0490·044  
Mean   0·0580·0440·2620·085
Standard deviation   0·0120·0080·1610·024
Reference strains
1 1174-1United StatesWildtypeGenotype II0·0660·042  
2 13-1United StatesWildtypeGenotype II0·1120·049  
3 526-3United StatesF129L isolateGenotype II  0·4120·068
4 1172-8United StatesF129L isolateGenotype II  0·6660·079
5 1178-W1United StatesF129L isolateGenotype II  1·6560·082
6 1189-7United StatesF129L isolateGenotype In.a.n.a.n.a.n.a.
7 1190-16United StatesF129L isolateGenotype In.a.n.a.n.a.n.a.
Mean   0·0890·0450·9110·076
Standard deviation   0·0330·0050·6570·008
Figure 2.

In vitro azoxystrobin sensitivity assay of Alternaria solani wildtype and F129L isolates collected between 2006 and 2011. Columns represent mean EC50 values, i.e. the effective fungicide concentration that inhibited spore germination by 50%. Bars represent standard deviations. Columns with the same letter are not significantly different (Tukey's b test, = 0·05).

The 10 F129L isolates collected between 2009 and 2011 had significantly higher EC50 values compared to the baseline populations (Table 4). All isolates with the F129L mutation were less sensitive to azoxystrobin. EC50 values varied between 0·108 and 0·729 μg mL−1 (Table 4) and were designated as ‘azoxystrobin reduced-sensitive’. The three F129L US reference strains had EC50 values of 0·412 up to 1·656 μg mL−1.

Altogether, comparing mean EC50 values, isolates lacking F129L were c. 4-fold more sensitive compared to F129L isolates (Fig. 3). The loss in sensitivity to azoxystrobin was less expressed among German A. solani F129L isolates compared to reduced-sensitive US reference isolates (Fig. 2).

Figure 3.

Mean EC50 values (effective fungicide concentration that inhibited spore germination by 50%) for sensitive (= 22) and reduced-sensitive (= 10) Alternaria solani isolates obtained from the in vitro cross-sensitivity assessment of azoxystrobin compared to pyraclostrobin. Within fungicides, columns with the same letter are not significantly different (Tukey's b test, = 0·05). Vertical bars indicate standard deviation for all tests performed on each isolate group.

The use of pyraclostrobin similarly showed reduced efficacy as a result of cross-sensitivity. Exposure to pyraclostrobin resulted in mean EC50 values of 0·051 μg mL−1 for wildtype and 0·12 μg mL−1 for F129L isolates. The shift towards reduced sensitivity was less pronounced compared to azoxystrobin (Fig. 3) and showed a 2·5-fold loss.

Effect of QoI fungicide sensitivity on disease control

To test whether this reduced fungicide sensitivity had an impact on disease control, greenhouse experiments were carried out. A subset of six sensitive and five reduced-sensitive isolates were tested to determine disease control provided by azoxystrobin. Results showed that control of reduced-sensitive isolates of A. solani was significantly less effective than that of sensitive isolates and reduced by almost half at any fungicide concentration. The data show that the efficacy of azoxystrobin (QoIs) has declined for isolates carrying the F129L mutation (Fig. 4).

Figure 4.

Disease severity of azoxystrobin against German wildtype (= 6) and F129L (= 5) Alternaria solani isolates investigated in in vivo greenhouse tests (mean of two replications). Disease severity was rated 1 week after artificial inoculation. Columns with the same letter are not significantly different (Tukey's b test, = 0·05). Vertical bars indicate standard deviation.

Discussion

Resistance and reduced sensitivity to fungicides among fungal plant pathogens are significant problems in the area of chemical pest management. QoI fungicides, which are widely used to protect crops, have been characterized as ‘high risk’ for the development of resistance (Kuck & Mehl, 2003). In Germany, QoIs were registered for the control of EB in potatoes in 2007 for the first time. Fungicide resistance has become increasingly important as loss of sensitivity has previously been reported for A. solani against QoIs (Pasche et al., 2004). To the authors' knowledge, this is the first report of the identification of the F129L mutation in the cytb gene in A. solani populations in Germany and Europe. The first F129L isolates of A. solani were found in 2009, 2 years after the first registration of azoxystrobin for potatoes in Germany. In the present study, isolates containing the F129L substitution displayed a shift in sensitivity to azoxystrobin as well as to pyraclostrobin in in vitro spore germination assays. Furthermore, the in vivo studies here revealed that disease control was affected. According to the present investigations, the frequency of reduced-sensitive isolates increased over the years 2009 to 2011. This is in line with findings by Pasche et al. (2004) and Rosenzweig et al. (2008) who showed that the use of QoIs promotes the selection of A. solani isolates carrying the F129L substitution. It is likely that the application of QoI fungicides between 2007 and 2011 has led to the increased occurrence of F129L mutants in Germany.

The cytb protein is a key determinant for the development of QoI fungicide resistance in plant pathogenic fungi as it is the target of Qo inhibitors. Concerning the structure of the cytb gene, two different genotypes were detected among the isolates investigated. Type I matched Grasso et al. (2006), with a first intron after codon 126 and the next after codon 143. In type II, a first intron was found after codon 131 and showed 90% homology to intron bi3 of Pyrenophora teres (Sierotzki et al., 2007) for the fragment analysed in this study. All type I isolates showed identical sequences, regardless of their origin (Germany or USA). The same applied for type II isolates. These results are in accordance with Latorse et al. (2010), who identified different genotypes for the A. solani cytb gene. The presence of structurally different cytb genes in P. teres (Sierotzki et al., 2007) seems to affect the occurrence of fungicide sensitivity. Yamada & Sonoda (2012) reported two types of cytb in field isolates of Pestalotiopsis longiseta, with or without an intron between codons 131 and 132. In general, the exon–intron structure of cytb genes appears to be an important factor for the evolution of resistance to QoI fungicides (Sierotzki et al., 2000b; Grasso et al., 2006). According to the present results, German isolates with genotype I did not carry the F129L mutation. The F129L substitution was observed only in German genotype II isolates, which have a cytb gene with an intron between codons 131 and 132. Yamada & Sonoda (2012) found that in P. longiseta the intron 131/132 in cytb did not affect the occurrence of QoI resistance, as the F129L mutation also occurred in cytb types which lacked the intron. However, the occurrence of the F129L mutation does not seem to be restricted to a particular genotype in A. solani. Two of the US reference strains (1189-7, 1190-16) carried the F129L mutation and were both genotype I. Up to now, it remains unclear why the F129L substitution has occurred only in German genotype II isolates. The F129L mutation has not occurred in European genotype I yet and usage of QoI fungicides has only favoured the rise of reduced-sensitive isolates of genotype II. Seventy-six out of 203 isolates were genotype II and only two of them were wildtype, whereas 74 carried the F129L substitution. Isolates of genotype I were found in each year whereas isolates of genotype II were only found from 2009 onwards, except one in 2006. It remains to be determined whether the very low frequency of wildtype strains of genotype II is associated with fitness costs.

The intron after codon 126 in genotype I could possibly be a group I intron. One characteristic of group I introns is that the last nucleotide in the upstream exon is almost always a U and the last nucleotide of the intron predominantly a G (Nielsen & Johansen, 2009), as is the case here. Group I introns are very common in fungi and are mobile elements at the DNA level (Nielsen & Johansen, 2009). It is known that in Saccharomyces cerevisiae, hot spots of point mutations are found in exons near insertion sites of optional group I intron-related sequences (Foury et al., 1998). However, according to Carbone et al. (1995), the presence or absence of an intron can differ among clonal lineages. Sequence analysis of cytb from several plant pathogenic fungi indicated the presence of varying numbers of introns (Gisi et al., 2002). Yin et al. (2012) identified six types of cytb out of 247 isolates of Botrytis cinerea containing various types of intron combinations. The occurrence of G143A and F129L mutants in P. longiseta lineages with and without intron 131/132 in cytb indicated that G143A and F129L mutants each derived from at least two origins (Yamada & Sonoda, 2012).

More investigations on the role of the cytb gene structure on fungicide sensitivity as well as on the origin of genotype II isolates are needed.

According to the observations here, A. solani isolates carrying the F129L substitution occurred for the first time in 2009. The two isolates derived from the same field. During the following years, the mutation occurred at various locations. In 2011, 70 F129L isolates were detected at six different locations. In total, F129L isolates have been sampled at nine different locations. At all these locations, QoI fungicides (azoxystrobin) had been applied in order to control EB disease. As some of the sampling locations are relatively far apart from each other (c. 180 km), the F129L mutations may have arisen independently. According to Birla et al. (2012) in Cercospora beticola, QoI resistance developed independently at several geographical regions. Kim et al. (2003) noted that QoI resistance in Pyricularia grisea occurred in five genetic backgrounds. Although A. solani spores are mainly wind dispersed, A. solani is not known for large-scale movement of inocula across wide areas. However, it is very unlikely that F129L isolates have been imported from foreign countries (e.g. USA) for which the F129L substitution has already described in A. solani populations. Therefore the F129L substitution must have occurred independently in the USA and Germany. Additionally, it must have occurred at least twice in the USA as the F129L substitution was found in US isolates of both genotypes. RAPD profiles which have been carried out for A. solani isolates within 2006 and 2008 revealed distinct genetic diversity among isolates (Leiminger, data not shown). Even isolates which were of identical geographic origin formed different subgroups showing only distant similarity. This may further indicate that QoI resistance in A. solani developed independently on multiple occasions because A. solani populations carrying the F129L mutation showed high genetic diversity.

Baseline isolates help to identify shifts in pathogen sensitivity or failures in disease control caused by resistance in the population (Jutsum et al., 1998). Baseline sensitivity values to azoxystrobin and pyraclostrobin were established for six A. solani isolates sampled in 2006. The EC50 values of azoxystrobin in baseline isolates were comparable to EC50 values previously reported for US wildtype isolates by Pasche et al. (2005). SHAM was required to suppress alternative oxidation (AOX) which has been suggested as a possible mode of QoI resistance in vitro in different fungi (Olaya & Köhler, 1999). Previous research demonstrated that SHAM did not inhibit germination of A. solani conidia at low concentrations (Pasche et al., 2004). Accordingly, assays have been carried out in the presence of 100 μg mL−1 SHAM. All isolates lacking the F129L substitution were sensitive to azoxystrobin and pyraclostrobin. However, it cannot be ruled out that the occurrence of alternative oxidase activity may influence QoI sensitivity (Steinfeld et al., 2001). Based on the conducted in vivo experiments, all isolates lacking the F129L mutation were very effectively controlled by azoxystrobin. Although a limited number of isolates have been tested in the in vivo tests, the impact of additional resistance mechanisms has not been observed. In vitro tests revealed that mean EC50 values generated for wildtype isolates sampled between 2006 and 2011 stayed at a constant level. The significantly higher EC50 values for the QoI treated populations suggest that a shift in A. solani sensitivity to these fungicides has also occurred in Germany. As mentioned by Pasche et al. (2005), a dramatic loss in sensitivity for reduced-sensitive isolates has been observed 2 years after registration of azoxystrobin in the USA. However, reduction in sensitivity against azoxystrobin of German isolates was half that of the reduced-sensitive isolates obtained from the USA. German F129L isolates would fit in a transition group between the obviously sensitive and reduced-sensitive groups, for which Pasche et al. (2004) described EC50 values between 0·51 to 0·775 μg mL−1. Although these isolates carried the F129L mutation, loss in sensitivity to azoxystrobin was not as strongly developed as it was for the reduced-sensitive group. Greenhouse tests revealed no difference in EB control among wildtype isolates of genotype I or II. However, control of F129L isolates was less efficient when compared to wildtype strains at 100 μg mL−1 azoxystrobin concentrations. Similar to the US, EB control decreased as EC50 values increased (Pasche et al., 2005).

In the present study, fungicide cross-resistance was evaluated. Isolates with reduced sensitivity to azoxystrobin possessed reduced sensitivity to pyraclostrobin too, albeit at a lower level and without statistical significance. This is consistent with findings from previous studies which reported shifts in sensitivity of A. solani isolates to azoxystrobin, pyraclostrobin and trifloxystrobin (Pasche et al., 2004). Pyraclostrobin has higher intrinsic activity against A. solani compared to azoxystrobin (Pasche et al., 2004); however, this did not cause increased disease control in the field. Wong & Wilcox (2000) also indicated that the intrinsic activity alone is not an adequate means to assess relative fungicide efficacy.

As a result of intensive use of QoI fungicides in the USA with up to six applications per season, EB disease control has been negatively affected within a few years. A 10–12-fold shift in sensitivity (Pasche et al., 2005) has rendered azoxystrobin less effective in EB control. In the present survey, a shift towards reduced sensitivity to azoxystrobin was revealed, but at lower levels when compared to those found in the USA. The decreased application rate of EB-specific applications in Germany may be one reason why loss in sensitivity was less expressed compared to the USA. It has been shown that the number of spray applications increased the number of QoI-resistant mutants (Sierotzki et al., 2000a). Fungicide sensitivity was influenced negatively when QoI fungicides were applied too often (Pasche et al., 2004; Rosenzweig & Stevenson, 2004). Therefore the number of QoI applications has been limited, with three being the maximum number of sprays allowed per growing season in Germany. These regulations may further help to reduce the selection pressure on the population and to prevent loss of fungicide efficacy.

Unfortunately, the risk for the evolution of resistance to QoI fungicides in Germany remains high, as fungicides with alternative modes of action for use against EB are lacking, and numerous and frequent applications of fungicides are necessary for adequate disease control. As the registered fungicides are widely used, effective anti-resistance strategies have to be applied. The restricted use of QoIs, in combination with management programmes, will be essential for the continued use of azoxystrobin and other QoI fungicides. Disease control systems have been developed which support the targeted application of fungicides according to site-specific disease thresholds, thus reducing the number of sprayings to a necessary minimum (Leiminger & Hausladen, 2012). In addition, the use of products in mixtures or the alternation with multisite modes of action in spray programmes decreases the likelihood that a pathogen develops fungicide resistance (Edin & Torriani, 2012). The strict compliance with these precautions may contribute to maintenance of the high efficacy of QoI fungicides for as long as possible.

Acknowledgements

The authors thank Neil Gudmestad and Julie Pasche for providing reference isolates from the USA. They are also grateful to Ralph Hückelhoven and Ruth Eichmann for critical reading of the manuscript and helpful comments.

Ancillary