Antifungal activity of metyltetraprole against the existing QoI‐resistant isolates of various plant pathogenic fungi

Abstract

BACKGROUND

Metyltetraprole is a novel quinol oxidation site of Complex III inhibitor (QoI) fungicide that inhibits mitochondrial electron transport at the Qo site of the cytochrome bc1 complex. Previous reports have demonstrated that it is also active against the QoI‐resistant (QoI‐R) isolates of Zymoseptoria tritici and Pyrenophora teres with the mutations G143A and F129L in their cytochrome b gene, respectively. Further studies on cross‐resistance between metyltetraprole and existing QoIs were performed using an increased number of isolates of Z. tritici, P. teres, Ramularia collo‐cygni, Pyrenophora tritici‐repentis, and several other plant pathogenic fungi.

RESULTS

Differences in the EC₅₀ values between the wild‐type and QoI‐R isolates with the mutations G143A or F129L were always smaller for metyltetraprole compared to those for the existing QoIs, and they were never greater than five in terms of resistance factor. The 2‐year field experiments showed that the metyltetraprole treatment did not increase the percentage of QoI‐R isolates likely to harbor the G143A mutation in a Z. tritici population.

CONCLUSION

The unique behavior of metyltetraprole against the existing QoI‐R isolates was confirmed for all tested pathogen species. Our results provide important information to establish a fungicide resistance management strategy using metyltetraprole in combination or alternation with other fungicides. © 2019 The Authors. Pest Management Science published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry.

Box‐whisker plot of the EC₅₀ values (mg L^‐1) for metyltetraprole and pyraclostrobin expressed in a log10 scale for each of the wild‐type and G143A mutant isolates of Zymoseptoria tritici.

1. INTRODUCTION

The quinol oxidation site of Complex III inhibitor (QoI) is one of the most important classes of agricultural fungicides with the advantages of high efficacy against a wide range of pathogens, good crop safety, and some favorable physiological effects on crops.1 However, the efficacy against many pathogens was drastically reduced because of the development of resistance, especially by the mutation corresponding to the amino acid substitution of G143A in the cytochrome b gene (CytB).1, 2 Before the discovery of metyltetraprole, all previously existing commercial QoIs lost their effectiveness against pathogens having a high population of resistant isolates harboring the G143A mutation.1, 2 For example, Zymoseptoria tritici, one of the representative plant pathogens that affects wheat in Western Europe, has lost sensitivity to existing QoIs because of the presence of the G143A mutation.1, 2, 3 Overcoming this problem of QoI resistance through small structural modification of existing QoIs was believed to be difficult because of direct steric interference between the central linking‐ring of QoIs, which is essential for retaining enough stability under sunlight, and the methyl group of the alanine residue at position 143 of CytB substituted in resistant mutants.2, 4 Furthermore, the resistant strains harboring the F129L mutation in CytB are known to be a second problematic type of QoI‐resistant (QoI‐R) mutants, although their strength of resistance is weaker than that of the strains harboring the G143A mutation.1, 2 The F129L mutation is estimated to obstruct the interaction between the CytB protein and the pharmacophore of QoIs. This second type of QoI‐R mutants have been reported to show partial resistance to QoI in pathogenic fungi such as Pyrenophora teres etc.1, 2

Metyltetraprole, 1‐(2‐{[1‐(4‐chlorophenyl)‐1H‐pyrazol‐3‐yl]oxymethyl}‐3‐methylphenyl)‐1,4‐dihydro‐4‐methyl‐5H‐tetrazol‐5‐one, is a novel fungicide with the unique tetrazolinone pharmacophore found by Sumitomo Chemical Co. Ltd.5, 6, 7, 8 Its molecule has a side chain very similar to that of pyraclostrobin, a representative of the existing QoIs, and it also binds to the Qo site of the cytochrome bc1 complex.5, 9 A competitive binding assay using radiolabeled metyltetraprole and pyraclostrobin showed that both fungicides bound to the same target pocket in the cytochrome bc1 complex.5 Nevertheless, it was previously demonstrated to be highly active even against the Z. tritici and P. teres strains harboring the mutations G143A and F129L, respectively.7 Therefore, metyltetraprole is considered to be the first molecule of a new innovative generation of QoIs. Although metyltetraprole itself is a highly promising agricultural fungicide, which is expected to be commercialized in the near future (https://sumitomo-chem-agro.com/blog/sumitomo-chemical-submits-a-registration-application-in-the-eu-for-metyltetraprole-a-new-fungicide/), its discovery has paved the way to overcome the mutations G143A and F129L in the history of synthetic QoIs. However, the numbers of tested isolates and pathogen species were limited in our previous reports.7, 8

In the case of QoIs, it has been reported that the orthologous amino acid mutations result in similar resistance profiles across different organisms with respect to the strength of resistance; the G143A mutation causes very high level of resistance (the resistance factor RF, [EC₅₀ of the resistant mutant]/[EC₅₀ of the wild type], is always greater than 100) to all commercial QoIs in any known cases of more than 20 pathogen species.1, 2 However, recently some cases of different resistance profiles within the orthologous mutations of different pathogen species were reported in cases of resistance against succinate dehydrogenase inhibitor (SDHI), another important agricultural fungicide class.10, 11, 12, 13 Considering the unique behavior of metyltetraprole in the G143A mutants of Z. tritici as well as in the F129L mutants of P. teres, its consistency in orthologous mutations in different species should be investigated.

Furthermore, in recent years, a fungicide resistance mechanism different from the target‐site mutations has been reported in some species, including Z. tritici. It is a multi‐drug resistance (MDR) phenotype resulting from the overexpression of a multidrug transporter, such as the major facilitator superfamily (MFS) transporter in the plasma membrane.14, 15, 16, 17 In the case of Z. tritici, the overexpression of MgMFS1 associated with the insertions to the genomic sequence upstream of its open reading frame (ORF) was detected in field isolates that showed an MDR phenotype.14 Such isolates have been reported to show a decreased sensitivity to various classes of fungicides, regardless of their structures.14, 15, 16, 17 Such a sensitivity shift due to the lower uptake of chemicals into the cells in an MDR phenotype should be checked for metyltetraprole because its target site is assumed to be located in the mitochondria inside the cell.

We report on the cross‐resistance studies between metyltetraprole and previously existing QoIs in various pathogenic fungi. At first, each representative pair of the wild‐type (WT) and QoI‐resistant isolates harboring the G143A or F129L mutation was used to compare the RFs of metyltetraprole and existing QoIs. Thereafter, the influence of the QoI sensitivity on the metyltetraprole sensitivity was investigated in field populations of Z. tritici, Pyrenophora tritici‐repentis, Ramularia collo‐cygni, and P. teres in Western Europe. Furthermore, the effect of an MDR phenotype on the antifungal activity of metyltetraprole was investigated in Z. tritici.

2. MATERIALS AND METHODS

2.1. Chemical materials

Salicylhydroxamic acid (SHAM) and 2,3,‐dihydroxybenzaldehyde (DHBA) were purchased from Tokyo Kasei Kogyo Co., Ltd (Tokyo, Japan). Pyraclostrobin, azoxystrobin, fluxapyroxad, prothioconazole‐desthio, tolnaftate, and tebufenpyrad were purchased from Sigma‐Aldrich Japan (Tokyo, Japan). Metyltetraprole (>99% purity) was synthesized as previously described.6 For conducting in vitro experiments, all chemical compounds were dissolved in dimethyl sulfoxide as stock solutions. For carrying out the metyltetraprole treatment in fields, an emulsifiable concentrate (EC) formulation of metyltetraprole was prepared by Sumitomo Chemical.

2.2. Fungal materials

The Z. tritici, R. collo‐cygni, P. teres, and P. tritici‐repentis strains were isolated from wheat or barley leaves in the fields of European countries (Table S1). Detached leaves were kept in humid conditions to induce spore formation. A single spore was collected under the microscope and grown on potato dextrose agar (PDA) medium (39 g PDA in 1 L water). The origins of the strains of Parastagonospora nodorum, Microdochium majus, Colletotrichum gloeosporioides, Pyricularia oryzae, Venturia inaequalis, Alternaria solani, Passalora fulva, and Pestalotiopsis longiseta are shown in Table S1.

2.3. Antifungal tests

The antifungal activity of all tested compounds against Z. tritici, R. collo‐cygni, P. teres, P. tritici‐repentis, P. nodorum, M. majus, C. gloeosporioides, P. oryzae, V. inaequalis, A. solani, P. fulva, and P. longiseta was evaluated by two different methods under the incubation conditions detailed in Table S2. Synergic agents such as alternative oxidase (AOX) inhibitors (e.g. SHAM) were added only for the species in which growth was not completely inhibited by QoI in a low nutrient medium (e.g. YBG) because those agents are themselves harmful to fugal growth. The RF was calculated using the following formula:

$$\mathrm{RF}=\left({\mathrm{EC}}_{50}\text{of resistant isolate}\right)/\left({\mathrm{EC}}_{50}\text{of}\mathrm{WT}\right).$$

2.3.1. Microtiter plate (96‐well) method

The growth of Z. tritici, R. collo‐cygni, P. nodorum, C. gloeosporioides, V. inaequalis, and P. fulva was evaluated in 96‐well microtiter plates. The inoculum of each fungal strain was harvested at more than 100 times higher density in distilled water and suspended in the appropriate medium at the density shown for each species (Table S2). A 100‐fold dilution series of the active ingredient at the designated final concentration was prepared for each test (Table S2). Aliquots (1 μL) of each fungicide were mixed with the inoculum (99 μL) prepared in medium. After incubating all fungal strains for the incubation periods listed in Table S2, their growth was determined by measuring the optical density of each culture solution at a wavelength of 600 nm using the microplate reader SH‐9000 Lab (Corona Electric, Ibaraki, Japan) with a 3 × 3 matrix of scanning points. The optical density values were corrected by the value of the blank well without inoculum. The 50% effective concentrations (EC₅₀) were calculated by probit analysis.

2.3.2. Agar plate method

Pyrenophora teres, P. tritici‐repentis, M. majus, A. solani, P. oryzae, and P. longiseta were cultured on agar medium amended with a series of designated concentrations of fungicides (see Table S2). The mean mycelial radial growth (mm) from two different inocula on each agar plate was measured at designated periods after inoculation, and the EC₅₀ values were calculated by probit analysis.

2.4. Sampling of isolates from the field trial of metyltetraprole

Leaves were collected from wheat plants treated with an EC formulation of metyltetraprole at 120 g hectare⁻¹ and untreated wheat plants from 20 fields in 2014 and 22 fields in 2015, when the field efficacy evaluation of metyltetraprole was performed. The treatments were carried out by contractor companies according to the guidelines of the European and Mediterranean Plant Protection Organization (http://pp1.eppo.int/). The sampling of leaves was also done by contractor companies 25–40 days after the metyltetraprole treatment. Leaves for sampling were selected randomly from each plot treated with metyltetraprole at 120 g hectare⁻¹ and an untreated plot (both were divided into three to four replicates in individual trial sites). Z. tritici isolations were done as previously described,7 and the sensitivity tests were performed for five isolates from each field, except for a few sites from where only four strains could be isolated.

2.5. Statistical evaluation

Statistical evaluations were performed with the statistic software BellCurve for Excel (Social Survey Research Information Co., Ltd, Tokyo, Japan). The significance of differences in the median EC₅₀ values was analyzed by the Mann–Whitney U‐test. The tests for the equality of two proportions were performed for the percentage of putative G143A mutants in each of the untreated and metyltetraprole (120 g L⁻¹)‐treated populations.

2.6. Identification of mutations in each pathogen species

The presence of mutation in each resistant isolate was checked as previously described.3, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28 The DNA of each fungi was extracted using the PrepMan Ultra Reagent (Applied Biosystems, Waltham, USA) according to the manufacturer's instructions. PCR amplification of the CytB gene was carried out using the Ex Taq Kit (Takara, Kyoto, Japan) based on the previously described methods for each pathogen species.3, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28 DNA sequencing was also performed according to the method described in previous reports.3, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28 For the identification of the genotypes of P. teres and P. tritici‐repentis isolated in this study, a PCR‐restriction fragment length polymorphism (PCR‐RFLP) was carried out as previously described, using the restriction enzymes (BsaJI, Fnu4HI, and MnlI) and primers listed in Table S3.20 Sequences were analyzed by the GENETYX ver. 12 software (Genetyx Corporation, Tokyo, Japan).

3. RESULTS

3.1. Resistance factors for the representative sets of several pathogen species

The test results for the G143A mutants are shown in Table 1 and for the F129L mutants in Table 2. The EC₅₀ values of metyltetraprole were always lower than 0.1 mg L⁻¹ for the WT isolates of all tested species. In G143A mutants, the EC₅₀ values of pyraclostrobin and azoxystrobin were significantly higher than those in the WT isolates, regardless of the species tested, as already reported.1, 2, 3, 18, 21, 22, 23, 24, 25 The mean RFs were >2000 and >400 for pyraclostrobin and azoxystrobin, respectively. On the other hand, the EC₅₀ values of metyltetraprole in the G143A mutants were less than twice of those in the WT isolates for all tested species, except for V. inaequalis. The RFs of metyltetraprole ranged from 0.17 to 5.17, and the mean was 1.6. In F129L mutants, the EC₅₀ values of pyraclostrobin and azoxystrobin were also higher than those in the WT isolates, regardless of the species tested, as those differences were smaller than those of the G143A mutants, as already reported.1, 2, 18, 19, 25, 26, 27 The mean RFs were 15.8 and 90.6 for pyraclostrobin and azoxystrobin, respectively. On the other hand, the EC₅₀ values of metyltetraprole in the F129 L mutants were not so different from those in the WT isolates among all tested species. The RFs ranged from 0.7 to 3.3, and the mean was 2.1.

Table 1.

Cross‐resistance on representative sets of several pathogen species (wild types and G143A mutants)

	Metyltetraprole			Pyraclostrobin			Azoxystrobin
	Wild type EC50 (mg L−1)a	G143A EC50 (mg L−1)a	Resistancefactorb	Wild type EC50 (mg L−1)a	G143A EC50 (mg L−1)a	Resistancefactorb	Wild type EC50 (mg L−1)a	G143A EC50 (mg L−1)a	Resistancefactorb
Zymoseptoria tritici	0.00048	0.00077	1.60	0.00050	0.42	840.0	0.012	5c	>400
Parastagonospora nodorum	0.0067	0.0054	0.81	0.00085	0.43	505.9	0.0042	2.5	595.2
Microdochium majus	0.0068	0.013	1.91	0.00020	2.2	11 000.0	0.015	10c	>650
Ramularia collo‐cygni	0.0019	0.00033	0.17	0.00046	0.67	1456.5	0.011	0.82	74.5
Colletotrichum gloeosporioides	0.010	0.010	1.00	0.0040	0.52	130.0	0.029	10c	>300
Pyrenophora tritici‐repentis	0.029	0.025	0.86	0.0039	3.0	769.2	0.058	50c	>800
Pyricularia oryzae	0.047	0.056	1.19	0.0072	0.61	84.7	0.040	1c	>25
Venturia inaequalis	0.0060	0.031	5.17	0.0030	1c	>300	0.0030	10c	>1000
Mean of resistance factor			1.6			>2000			>400

^aValues represent the mean 50% effective concentration (EC₅₀) values (mg L⁻¹) of each isolate in two replicates. The 95% confidence intervals of all EC₅₀ values ranged between 50% (MIN) and 200% (MAX) of the representative values.

^bResistance factor (EC₅₀ value for the mutant isolate/EC₅₀ value for the wild‐type isolate).

^cCut‐off value.

Table 2.

Cross‐resistance in representative sets of several pathogen species (wild types and F129L mutants)

	Metyltetraprole			Pyraclostrobin			Azoxystrobin
	Wild type EC50 (mg L−1)a	F129 L EC50 (mg L−1)a	Resistancefactorb	Wild type EC50 (mg L−1)a	F129 L EC50 (mg L−1)a	Resistancefactorb	Wild type EC50 (mg L−1)a	F129 L EC50 (mg L−1)a	Resistancefactorb
Pyrenophora teres	0.0035	0.0055	1.6	0.0013	0.0086	6.6	0.0054	0.31	57.4
Pyrenophora tritici‐repentis	0.013	0.031	2.4	0.0039	0.14	35.9	0.058	4.1	70.7
Zymoseptoria tritici	0.00060	0.0017	2.8	0.00090	0.013	14.4	0.076	2.0	26.3
Alternaria solani	0.0029	0.0020	0.7	0.00073	0.0043	5.9	0.00093	0.19	204.3
Passalora fulva	0.0029	0.0097	3.3	0.00057	0.016	28.1	0.0045	0.75	166.7
Pestalotiopsis longiseta	0.011	0.018	1.6	0.011	0.043	3.9	0.065	1.2	18.5
Mean of resistance factor			2.1			15.8			90.6

^aValues represent the mean 50% effective concentration (EC₅₀) values (mg L⁻¹) of each isolate in two replicates. The 95% confidence intervals of all EC₅₀ values ranged between 50% (MIN) and 200% (MAX) of the representative values.

^bResistance factor (EC₅₀ value for the mutant isolate/EC₅₀ value for wild type isolate).

3.2. Z. tritici population in Western Europe

A total of 377 isolates were collected from five countries in June and July of 2015. In the box‐whisker plot for the EC₅₀ values of metyltetraprole and pyraclostrobin, the population was divided into two different groups based on their sensitivity to pyraclostrobin (Fig. 1). The isolates showing the pyraclostrobin EC₅₀ values greater than 0.05 mg L⁻¹ were classified as the QoI‐R isolates and were expected to harbor the G143A mutation.2, 3 The median EC₅₀ value for metyltetraprole in the QoI‐sensitive (QoI‐S) population (isolates whose pyraclostrobin EC₅₀ values were less than 0.01 mg L⁻¹) was 0.00027 mg L⁻¹ (N = 19). On the other hand, the median EC₅₀ value for metyltetraprole in the QoI‐R population was 0.0022 mg L⁻¹ (N = 358). The difference between the EC₅₀ values of the QoI‐S and QoI‐R isolates was not supported by the results of the Mann–Whitney U‐test (P = 0.92) for metyltetraprole, therefore we observed almost no difference in the metyltetraprole EC₅₀ values between the QoI‐S and QoI‐R isolates of Z. tritici.

Figure 1.

Box‐whisker plot of the EC₅₀ values (mg L⁻¹) for metyltetraprole and pyraclostrobin expressed in a log10 scale for each of the QoI‐sensitive (QoI‐S) and QoI‐resistant (QoI‐R) isolates of Z. tritici. **P < 0.01 in the Mann–Whitney U‐test.

3.3. Putative G143A isolates of R. collo‐cygni and P. tritici‐repentis

In Fig. 2(a),(b), the box‐whisker plots show the EC₅₀ values of metyltetraprole for both pathogens. For R. collo‐cygni, 88 isolates were collected in 2015; of these 82 were highly resistant to both azoxystrobin and pyraclostrobin (QoI‐R isolates, EC₅₀ > 0.1 mg L⁻¹) and were expected to harbor the G143A mutation.21 The median EC₅₀ value of metyltetraprole in the QoI‐R isolates was 0.0029 mg L⁻¹, and the EC₅₀ values in the QoI‐S isolates were 0.0029, 0.0028, 0.0029, 0.0035, 0.0039, and 0.0051 mg L⁻¹, indicating almost no effect of QoI‐resistance on the activity of metyltetraprole against R. collo‐cygni. For P. tritici‐repentis, 64 isolates were collected in 2015; of these, 45 were highly resistant to pyraclostrobin (QoI‐R isolates showing EC₈₀ > 1 mg L⁻¹) and were expected to harbor the G143A mutation.19 In the isolates whose growth was inhibited by 1 mg L⁻¹ of pyraclostrobin, we checked the presence of the F129 L mutation by PCR‐RFLP20 as well as by sequencing of fragments. As a result, seven isolates were found to contain the F129L mutation and 12 isolates did not (QoI‐S isolates). The median EC₅₀ value for metyltetraprole in the QoI‐R isolates was 0.0065 mg L⁻¹, whereas that in QoI‐S isolates was 0.016 mg L⁻¹ (Fig. 2(b)). The result of the Mann–Whitney U‐test indicated that the difference of EC₅₀ values for metyltetraprole between the QoI‐S and QoI‐R isolates was statistically significant (P = 0.0027). Thus, we concluded that the metyltetraprole sensitivity of the QoI‐R isolates, which are likely to harbor G143A mutation, is slightly higher than that of the QoI‐S isolates of P. tritici‐repentis.

Figure 2.

Box‐whisker plot of the EC₅₀ values (mg L⁻¹) for metyltetraprole expressed in a log10 scale for each of the QoI‐sensitive (QoI‐S) and QoI‐resistant (QoI‐R) isolates of (a) R. collo‐cygni and (b) P. tritici‐repentis. **P < 0.01 in the Mann–Whitney U‐test.

3.4. F129 L mutants of P. teres and P. tritici‐repentis

Figure 3(a) shows the box‐whisker plot for both pathogens showing the EC₅₀ of metyltetraprole. For P. teres, 35 isolates were collected in 2015. We checked the presence of the F129L mutation by PCR‐RFLP20 and sequencing of mitochondrial DNA (mtDNA) fragments, and found 15 isolates to be F129L mutants and 20 to be WT. The median EC₅₀ value of the F129L mutants was 0.030 mg L⁻¹, whereas that of the WT isolates was 0.019 mg L⁻¹. The results of the Mann–Whitney U‐test indicated that the difference of EC₅₀ values between the WT and F129L mutants was statistically significant (P = 0.013) in P. teres. For P. tritici‐repentis, 12 WT and 7 F129L mutants were collected as previously mentioned. The median EC₅₀ value in the F129L mutants was 0.034 mg L⁻¹, whereas that in the WT isolates was 0.016 mg L⁻¹ (Fig. 3(b)). The result of the Mann–Whitney U‐test did not support the statistical significance of this difference (P = 0.50) of the EC₅₀ values between the WT and F129L strains of P. tritici‐repentis.

Figure 3.

Box‐whisker plot of the EC₅₀ values (mg L⁻¹) for metyltetraprole expressed in a log10 scale for each of the WT isolates and F129L mutants of (a) P. teres and (b) P. tritici‐repentis. *P < 0.05 in the Mann–Whitney U‐test.

3.5. Effect of multi‐drug resistance on sensitivity to metyltetraprole

For the isolates of Z. tritici collected in 2015, we identified the isolates showing an MDR phenotype based on their sensitivity to tolnaftate.14, 15 As a result, 15 among 377 isolates showed reduced sensitivity to tolnaftate (EC₅₀ > 1.0 mg L⁻¹, which is 4‐fold greater than the median EC₅₀ values of other isolates) and were considered to exhibit the MDR phenotype. Figure 4 shows the box‐whisker plot for non‐MDR and MDR isolates. The sets of the median EC₅₀ values (non‐MDR/MDR) were 0.0021/0.0095 mg L⁻¹ for metyltetraprole, 0.027/0.13 mg L⁻¹ for fluxapyroxad, and 0.061/0.15 mg L⁻¹ for prothioconazole‐desthio. Differences in EC₅₀ values between non‐MDR and MDR isolates were statistically significant for metyltetraprole, fluxapyroxad, and prothioconazole‐desthio (P < 0.001, Mann–Whitney U‐test). The results showed a limited effect of enhanced efflux activity on sensitivity to metyltetraprole with RF values <5.

Figure 4.

Box‐whisker plot of the EC₅₀ values (mg L⁻¹) expressed in a log10 scale for each of multi‐drug resistant (MDR) and non‐MDR phenotypes in Z. tritici: (a) metyltetraprole in non‐MDR, (b) fluxapyroxad in non‐MDR, (c) prothioconazole‐desthio in non‐MDR, (d) tolnaftate in non‐MDR, (e) metyltetraprole in MDR, (f) fluxapyroxad in MDR, (g) prothioconazole‐desthio in MDR, and (h): tolnaftate in MDR.

3.6. Effect of the selection pressure of metyltetraprole on pathogen population

Table 3 shows the frequency of QoI‐R isolates in the isolates of Z. tritici obtained from the leaves of the metyltetraprole‐treated (two sprays on 20‐ to 35‐day intervals at 120 g hectare⁻¹) and untreated wheat plants of an individual trial field. No increase was observed in the frequency of QoI‐R isolates that showed resistance to pyraclostrobin (EC₅₀ > 0.05 mg L⁻¹) in both experiments conducted in 2014 and 2015 (P = 0.08 and 0.59, respectively).

Table 3.

Impact on pathogen population by the selection pressure from metyltetraprole (Z. tritici)

	2014 numbers of QoI‐sensitive/resistant isolates			2015 numbers of QoI‐sensitive/resistant isolates
	Trial site	Untreated	Metyltetraprole120 g L−1	Trial site	Untreated	Metyltetraprole120 g L−1
	France (site 1)	0/5	0/5	France (site 1)	2/3	1/4
	France (site 2)	0/5	1/4	France (site 2)	1/4	0/5
	France (site 3)	1/4	0/5	France (site 3)	2/3	0/5
	France (site 4)	0/5	0/5	France (site 4)	0/5	0/5
	France (site 5)	1/4	3/2	France (site 5)	1/4	3/2
	France (site 6)	0/5	1/4	France (site 6)	0/5	1/4
	France (site 7)	0/5	1/4	France (site 7)	0/5	0/5
	France (site 8)	2/3	4/1	UK (site 1)	0/5	0/5
	UK (site 1)	0/5	0/5	UK (site 2)	0/5	0/5
	UK (site 2)	0/5	0/5	UK (site 3)	0/5	0/5
	UK (site 3)	0/4	0/5	UK (site 4)	0/5	0/5
	UK (site 4)	0/5	0/5	UK (site 5)	0/5	0/5
	Ireland (site 1)	0/5	0/5	Ireland (site 1)	0/5	0/5
	Ireland (site 2)	0/5	1/4	Ireland (site 2)	0/5	0/5
	Germany (site 1)	0/5	0/5	Ireland (site 3)	0/5	0/5
	Germany (site 2)	1/4	0/5	Germany (site 1)	0/5	0/5
	Germany (site 3)	1/4	4/1	Germany (site 2)	0/5	1/4
	Germany (site 4)	0/5	0/5	Germany (site 3)	1/4	1/4
	Belgium (site 1)	0/5	0/4	Germany (site 4)	0/5	2/3
	Belgium (site 2)	1/4	0/5	Germany (site 5)	0/5	1/4
				Belgium (site 1)	1/4	0/5
				Belgium (site 2)	0/5	0/5
QoI‐sensitive totala		7 (7.1%)	15 (15.2%)		8 (7.3%)	10 (9.3%)
QoI‐resistant total		92 (92.9%)	84 (84.9%)		102 (92.7%)	97 (90.7%)

^aNumber and percentage of QoI‐sensitive isolates (sensitive to pyraclostrobin; EC₅₀ < 0.01 mg L⁻¹) and QoI‐resistant isolates (resistant to pyraclostrobin; EC₅₀ > 0.05 mg L⁻¹).

4. DISCUSSION

Our study confirmed lower RFs of metyltetraprole in QoI‐R isolates of all tested pathogen species. However, the presence of a small, but statistically significant, difference between the WT and QoI‐R isolates was not always consistent across the studied pathogen species, though its magnitude in the RFs of metyltetraprole across species was small. For example, the G143A mutation seemed to affect metyltetraprole sensitivity neither in Z. tritici nor in R. collo‐cygni (Figs 1 and 2(a)). On the other hand, the same G143A mutation seemed to confer a slight, but statistically significant, increase in metyltetraprole sensitivity in P. tritici‐repentis (Fig. 2(b)). Furthermore, the F129L mutation seemed to confer a slight, but statistically significant, reduction of metyltetraprole sensitivity in P. teres (Fig. 3(a)), although its RF value was much smaller than that of the existing QoIs. The RF calculated by the median EC₅₀ value of the WT and F129L isolates was only 1.6, while that of the existing QoIs has been reported to range from 5 to 20.1, 2, 19 On the other hand, the reduction in metyltetraprole sensitivity was not statistically significant in the F129L mutants of P. tritici‐repentis in this study, though the number of samples was limited and, therefore, more experiments might be required for solid conclusions with general applicability (Fig. 3(b)). Such inconsistency across species could be attributed to a minor difference in the original structure of the Qo site in each species due to amino acid variations in the Qo site at several positions other than positions 129 and 143 (Fig. S1). Specifically, the target Qo site structure of the CytB protein is slightly different across the studied pathogen species, even in the WT isolates, thus the structural influence of the mutation G143A/F129L could also be slightly different. Some other factors might also affect the variation in RFs across species, for example the effect of mutations on the expression level of the AOX gene that encodes a well‐known enzyme (alternative oxidase), whose overexpression can reduce the in vitro fungal sensitivity to QoIs as well as QiIs,29, 30, 31, 32, 33 to maintain the balance of homeostasis might also be variable across species. In fact, we observed that the addition of salicylic hydroxamic acid (SHAM) could increase the sensitivity of R. collo‐cygni to metyltetraprole (Table S4). However, the reduction of sensitivity conferred by AOX overexpression has been reported to play a minor role in practical efficacy reduction in vivo.2 In addition, differences in the environmental fitness of the G143A and F129L strains across species34, 35, 36, 37, 38 suggest a variable physiological effect conferred by these orthologous mutations in each species, and it could be relevant to a slight secondary effect on metyltetraprole sensitivity. Nevertheless, our results indicate that the mutations G143A and F129L had very little effect on metyltetraprole sensitivity in all tested species.

The reduction of sensitivity towards metyltetraprole by an MDR phenotype was slightly greater in terms of its magnitude than that conferred by the G143A and F129L mutations in Z. tritici. The median EC₅₀ values of metyltetraprole in MDR isolates were approximately five times bigger than those in non‐MDR isolates and thus were almost similar to those of fluxapyroxad and prothioconazole‐desthio. In general, there is no or very limited reduction or difference expected for calculated RF values <5 in terms of field efficacy. In fact, we have not observed significant reduction in the effectiveness of metyltetraprole caused by the presence of MDR isolates in field conditions (data not shown). However, these results indicate that the impact of an MDR phenotype on metyltetraprole sensitivity as well as the determination of the selection risk of an MDR phenotype caused by the use of metyltetraprole should be carefully and continuously investigated. This estimation would be also applicable for other newer fungicides that are not or are less affected by any specific mutations in the target sites of the currently available fungicides.

Our results of the cross‐resistance studies on the isolates and populations in field conditions confirm the unique position of metyltetraprole as a promising QoI in resistance management strategies. In particular, it is notable that metyltetraprole did not increase the frequency of the G143A mutation (Table 3), although further accumulation of data would be desirable, especially for those cases where the ratio of G143A mutants is still low. On the other hand, a number of previous studies on yeast and other few model microorganisms have been published with respect to QoI‐resistant laboratory mutants.39, 40, 41, 42 Although some studies on plant pathogenic fungi and synthetic QoIs, which are used as agricultural fungicides, have reported the risk of the mutations G143A, G143S, and F129V,43, 44, 45 older studies on natural substances, like strobilurin A and myxothiazol, have reported that other mutations, such as A126T, F129L, Y132C, C133Y, G137S, G137E, G137V, G143A, N256K, N256I, N256Y, L275F, L275S, L275T, and L295F, can cause resistance to those substances in yeast or other model microorganisms.39, 40, 41, 42 The manner of interaction between these natural substances and CytB protein at the Qo site could be distinct from that between the existing synthetic agricultural QoIs and the Qo site because the former does not have a central linking ring, which is assumed to be involved in steric interference with the methyl group of the alanine residue at position 143 in CytB of the G143A mutants. Although the manner of binding of metyltetraprole to the Qo site, owing to the presence of its unique tetrazolinone pharmacophore, would be different from both existing synthetic QoIs and natural substances, we should consider the risk of cross‐resistance. Some of the above‐mentioned or other new mutations could be selected as ‘metyltetraprole‐resistant mutations’ in future. The existing synthetic agricultural QoIs have led to the selection of only G143A and F129L mutations, as well as of G137R with very low frequency, in the past two decades,1, 2 suggesting that ‘metyltetraprole‐resistant mutants’ have not been selected by the existing QoIs. It could therefore be possible to hypothesize that metyltetraprole‐resistant mutants resulting from a new mutation might not be highly resistant to the existing synthetic agricultural QoIs. This assumption could be made for major QoIs, such as azoxystrobin, pyraclostrobin, and trifloxystrobin, which have been used more frequently than others and have played a major role in the selection of the G143A and F129L mutants. Thus, in cases where the population is still sensitive to the existing QoIs, a mixture of or an alternation between metyltetraprole and major existing QoIs might effectively serve as a resistance management tool for each other. However, these hypotheses need careful evaluation and further research activities. In the past, diethofencarb was introduced as a special fungicide that controlled carbendazim‐resistant strains harboring the mutation E198A or E198G in the target site β‐tubulin. However, with time, strains possessing new mutations in the β‐tubulin protein and hence resistant to both carbendazim and diethofencarb emerged.46 Additionally, according to our data, there is a slight difference in the sensitivities between QoI‐S isolates and F129L mutants in P. teres, therefore despite having sufficient activity against F129L mutants, metyltetraprole could increase their frequency. Although its selection strength might not be as high as the existing QoIs owing to the small sensitivity difference between QoI‐S isolates and F129L mutants, the risk of F129L selection by metyltetraprole is the subject of ongoing research in our laboratory. Furthermore, in the case of the pathogen populations in which the G143A mutant isolates are already dominant, a new metyltetraprole‐resistant mutant might emerge as a strain that is resistant to both metyltetraprole and existing QoIs because of the stacking of two mutations, the G143A mutation and a new mutation causing metyltetraprole resistance. On the other hand, there is a possibility that the stacking of two mutations might reduce environmental fitness and might have less chance of emerging as a practical problem for metyltetraprole. In fact, the compound mutation of F129L with G143A has not been observed in the field isolates of any plant pathogen, and it could be difficult for such mutations to survive in field conditions. Young et al. described the mode of action of fenpicoxamid, which is also a novel fungicide for controlling Z. tritici.47 They showed that fenpicoxamid inhibited cytochrome bc1 complex at the Qi site, which is a distinct catalytic site from the Qo site. Since mutations in CytB, which confers resistance to fenpicoxamid, are located at the Qi site,47 far from the Qo site, it can be inferred that they will not directly affect the sensitivity to metyltetraprole. However, their influence on enzyme efficiency or protein folding could affect the possibility of acquiring novel mutations at the Qo site because further accumulation of mutations in CytB could result in a stronger fitness penalty. The applicability of this hypothesis could be addressed in future studies. Nevertheless, the strategy for metyltetraprole resistance management should be established based on the individual background situation in each crop and region. We hope that our report will help researchers involved in the establishment of fungicide resistance management programs in each crop and region to consider the proper adoption of metyltetraprole in local fungicide application programs.

Supporting information

Table S1 Origin of fungal isolates in this study

Table S2 Incubation conditions for antifungal activity tests^†

Table S3 Primers used for PCR‐RFLP of Pyrenophora teres and Pyrenophora tritici‐repentis

Table S4 Increased metyltetraprole sensitivity by SHAM in R. collo‐cygni

Figure S1 Alignment of partial cytochrome b amino acid sequences of the wild‐type isolates of various phytopathogenic fungi used in this study