1. Introduction
Cucurbits are very important crops in Spanish agriculture. Within the vegetable sector, annual cucurbit production reached 3 million tons and yielded revenues of more than €1.9 billion in 2018 [
1]. One of the most destructive diseases that affect these crops is powdery mildew, which is an important limiting factor for cucurbit production in Spain [
2,
3]. This disease can be caused by the biotrophic fungal species
Podosphaera xanthii (Fr.) U Braun & N Shishkoff or
Golovinomyces cichoracearum (DC.) VP Galut, [
4], but in Spain, only
P. xanthii has been detected over the last three decades [
5,
6,
7,
8,
9,
10]. Despite the substantial efforts that have been invested in plant breeding programs to combat powdery mildew disease, chemical control continues to be the principal practice for managing most cucurbit crops; however, it has been hampered by the emergence of resistant populations in the field soon after the introduction of certain classes of site-specific fungicides. In southern Spain, resistance to the most popular anti-powdery mildew fungicides, such as quinone outside inhibitors (QoIs), demethylation inhibitors (DMIs) and methyl benzimidazole carbamate (MBC) fungicides, has been reported [
7,
8,
9,
10,
11]. More importantly, multiresistant isolates have been found in several areas of more intense cropping [
9].
Succinate dehydrogenase inhibitors (SDHIs; FRAC group 7) have been on the market for more than 40 years, and are within the class with the fastest growth in terms of new compounds released onto the market. To date, twenty-three SDHI active ingredients belonging to 11 chemical classes with a broader spectrum of fungal activity have been offered for fungal plant pathogen control [
12]. SDHI fungicides have a single-site mode of action, inhibiting the fungal respiration pathway by binding the ubiquinone binding site of succinate dehydrogenase (SDH; also known as complex II) and blocking mitochondrial electron transfer from succinate to ubiquinone [
13]. This target protein is formed by four subunits (A, B, C and D), but the ubiquinone-binding site only comprises amino acids from subunits B (SdhB), C (SdhC), and D (SdhD; [
14]).
SDHIs are classified as medium to high risk for resistance development. Therefore, it is not surprising that resistance to these fungicides has been documented since shortly after their registration for use against several phytopathogenic fungi [
15]. More than 40 point mutations in SdhB, SdhC and SdhD have been linked to reduced sensitivity to SDHIs. In SdhB, the changes H272L/R/T and H277L/R/Y are the most common, having been described in several fungal plant pathogens, such as
Alternaria alternata [
16,
17,
18,
19,
20,
21],
Botrytis cinerea [
22,
23,
24,
25,
26,
27],
B. elliptica [
28],
Didymella bryionidae [
29],
P. xanthii [
30],
Pyrenophora teres [
31] and
Sclerotinia sclerotiorum [
28]. Notably, the amino acid changes H133R in SdhC and D124E/N and H133P/R in SdhD have been found in pathogens such as
A. alternata [
16,
17,
18,
19,
20,
21],
A. solani [
32],
B. cinerea [
22],
P. xanthii [
33],
P. teres [
31] and
S. sclerotiorum [
28]. Although these amino acid changes are the most commonly described, other changes in the subunits SdhB (P225F/H/L/T, P230A/D/F/I/R, N230I, N235D/E/G/T and T268I), SdhC (S73P, N75S, G79R, T79N, W80S, A86V, N86S, G91R, S135R, H146R, G150R, H151R, V166M and G172D) and SdhD (A47T, S89P, G109V, S121P, H137R and D145G) have been documented in several phytopathogenic fungi, generating a pool of point mutations that confer different levels of resistance to the different SDHI fungicides [
15].
To avoid field control failure, and for the efficient use of the fungicides that are available on the market, it is important to have good knowledge about the resistance situation in the field. For that reason, the most commonly used methods are based on mycelial growth or conidial germination in vitro assays in culture medium or plant material supplemented with different fungicide concentrations [
34]. However, these methods are time consuming, especially when studying biotrophic fungi, and, therefore, molecular methods based on the detection of single-nucleotide polymorphisms are gaining in importance due to their quicker response times. Among the most commonly used approaches are polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP), allele-specific PCR (AS-PCR), cleaved amplified polymorphic sequences (CAPS) and high-resolution melt (HRM) analysis [
16,
17,
35,
36,
37,
38,
39]. Although all these methods are faster than in vitro assays, they require specific equipment that not all laboratories can afford. In recent years, the loop-mediated isothermal amplification (LAMP) technique developed by Notomi and collaborators [
40], has become an excellent alternative due to its cost, speed, and accuracy in fungicide resistance monitoring studies [
41,
42,
43,
44]. This technique, which is based on the combination of the
Bst polymerase and four primer pairs, which hybridize with six regions in the target DNA, can amplify the product of interest under isothermal conditions [
40]. In addition, the amplification product can be visualized with the naked eye using DNA-intercalating reagents such as SYBR-Green I [
45], metal-ion indicators such as hydroxy naphthol blue (HNB) [
46] or calcein [
47], and even pH-sensitive dyes [
48]. Recently, the LAMP technique has been successfully used to detect two-point mutations involved in SDHI resistance, namely H272R in SdhB in
B. cinerea [
49] and the change N75S in the SdhC in
C. cassiicola [
50].
In Spain, there are seven chemical classes of fungicides (aryl-phenyl-ketones (FRAC group 50), DMI (FRAC group 3), hydroxy-(2-amino-) pyrimidines (FRAC group 8), MBC (FRAC group 1), phenyl-acetamide (FRAC group U 06), QoIs (FRAC group 11) and SDHIs) registered for cucurbit powdery mildew control, with SDHI fungicides being one of the most frequently applied classes. Boscalid was the first SDHI registered in 2008, followed by fluopyram (2016), penthiopyrad (2017), isopyrazam (2018) and fluxapyroxad (2019). To date, P. xanthii resistance to SDHIs has not been documented in Spain, but evidence has started to emerge that resistance is developing in commercial cucurbit fields. For this reason, in the current study, cucurbit samples affected by powdery mildew symptoms from the primary cucurbit production areas in south-eastern Spain were collected during the 2018 and 2019 growing seasons. The fungal pathogen was isolated and identified as P. xanthii, and its sensitivity to boscalid and fluopyram was characterized using an in vitro leaf-disc bioassay and in planta analysis. The molecular alterations in the target gene subunits (SdhB, SdhC and SdhD) and the possible associated fitness costs were also studied. In addition, a LAMP assay for the rapid and reliable detection of two-point mutations related to P. xanthii SDHI resistance in our country was developed.
4. Discussion
The study of the fungicide resistance phenomenon is an essential step to avoiding the losses associated with fungal diseases in the field. The purpose of this work was to study, for the first time in Spain, the fungicide resistance situation of P. xanthii populations to SDHI fungicides, one of the families with relatively more active ingredients registered in recent years. For this purpose, several experiments, such as in vitro and in vivo fungicide sensitivity studies, analyses of the point mutations involved in resistance to boscalid and fluopyram, including its rapid molecular diagnosis through the LAMP technique and, lastly, the possible fitness cost associated with SDHI resistance, were performed.
The SDHI resistance monitoring studies performed during the 2018 and 2019 cucurbit production seasons showed that almost half of the isolates analysed had reduced sensitivity to boscalid (37.9%) and fluopyram (44.0%). The results varied in each test province: Granada and Malaga showed an absence or low levels of resistance, while Almeria and Murcia revealed high frequencies of resistant isolates (Almeria, 51.1% to boscalid and 57.4% to fluopyram; Murcia, 55.6% and 66.9% to boscalid and fluopyram, respectively). Moreover, the level of SDHI resistance in these two provinces usually increased from one year to another. In Almeria, the frequency of boscalid resistance increased from 42.9% to 59.4%, and for fluopyram it increased from 44.4% to 70.3%. With respect to Murcia, except for boscalid-resistant isolates, which remained at approximately 55%, the frequency of fluopyram resistance increased from 56.3% to 74.5%. According to the information provided by the growers, this increase could be due to the use of these fungicides to control other fungal diseases (anthracnose, Alternaria leaf blight, grey mould, leaf spot, and Sclerotinia stem rot). In addition, in the provinces of Granada and Malaga, which showed an absence or very low levels of resistance, a good fungicide management application, meaning the alternation between fungicides with different modes of action, and between single- and multisite fungicides, was performed.
Our results are supported by other studies in which high levels of SDHI resistance were described in several fungal pathogens, including
P. xanthii,
B. cinerea,
A. alternata,
A. solari,
D. bryoniae,
C. cassiicola and
P. teres collected from fields where these fungicides have been frequently applied. In relation to
P. xanthii, similar frequencies were described for boscalid in the American and Japanese populations of this pathogen, with 44 and 45.96% resistant isolates, respectively [
30,
58]. Different studies on
B. cinerea showed approximately 50% boscalid-resistant isolates for the total population in Greece, Germany, and Spain [
23,
27,
59]. With regard to a study developed in several strawberry fields in Spain, an increase from 5.3% to 10.4% in fluopyram-resistant isolates was observed in the
B. cinerea population collected between 2015 and 2016 [
27]. With reference to
A. alternata, several studies in pistachio orchards in the USA also documented high percentages of SDHI-resistant isolates [
19,
60] and, in addition, an increasing trend over the years in farms where boscalid had been used extensively [
16,
61]. High frequencies of boscalid-resistant isolates have been described for
A. solani (75%), D. bryoniae (79.6%), and
C. cassiicola (48.9%) in monitoring studies performed in SDHI-treated fields from the USA and Japan [
62,
63,
64]. Regarding
P. teres, in a very complete study performed in several European countries, similar levels of resistance to those presented in this study, namely 44% for boscalid and 47% for fluopyram, were obtained for the isolates sampled from Germany in 2013 and 2014 [
31]. Moreover, the overall percentage of SDHI-resistant isolates in all the studied countries increased from 1.2% to 25% in 2012 and 2013, respectively [
31].
According to their growth in in vitro leaf-disc assays, our results showed that the Spanish
P. xanthii population could be divided into four different levels of resistance; however, when some representative isolates were tested in planta, all the resistant phenotypes, regardless of the resistance category, were able to develop disease in plants sprayed with the field doses of boscalid and fluopyram (100 mg/L), showing the same colony development for all of them. The differences in SDHI applications may explain the different results observed in in vitro and in vivo assays. For
P. xanthii, the laboratory approach for fungicide sensitivity tests is based on the use of leaf discs in direct contact with the fungicide solution during a period, which creates a larger exposure to the fungicide. This characteristic could make the final concentration of the fungicide in the plant tissue higher and make it possible to distinguish different categories of phenotypes. However, in the field, when these fungicides are applied, after entering the plant fungicides are transferred to different parts by the xylem due to acropetal phytomobility, making the concentration lower than that in leaf discs; therefore, all phenotypes (LR, MR, R and HR) were capable of developing powdery mildew disease under the conditions of these experiments [
65]. Similar discrepancies have also been documented for another biotrophic fungus, the grape powdery mildew
Erysiphe necator [
66]. In that study, a leaf disc sporulation assay was conducted to establish sensitivity to quinoxyfen, with some isolates showing decreased sensitivity; however, when the results were contrasted with a quantitative assay based on germ tube elongation inhibition, the same isolates were completely inhibited by quinoxyfen [
66]. Differences between resistant phenotypes in in vitro and in vivo assays were also observed in experiments on
B. cinerea. Isolates that were considered moderately resistant and resistant to cyprodinil in the in vitro assay developed grey mould disease at the same levels in infected cyprodinil-treated fruits [
67]. These results were confirmed in other studies in which
B. cinerea isolates, which had moderate and higher resistance levels in in vitro assays for cyprodinil and iprodione, infected fruits with the same degree of virulence when treated with these fungicides [
68,
69].
Resistance to SDHI fungicides is conferred by point mutations in the three subunits, which conform to the ubiquinone binding site (SdhB, SdhC and SdhD) of mitochondrial complex II. Several amino acid changes have been described in different fungal species; [
15]; however, little is known about powdery mildew fungi, with only three research studies, two on
E. necator and one on
P. xanthii. In
E. necator, the amino acid change H242R/Y in ShdB correlated with boscalid and fluopyram resistance, and the amino acid change G169D in SdhD explained the low sensitivity to fluxapyroxad and fluopyram [
70,
71]. With respect to
P. xanthii, a recent study documented some mutations in the subunits SdhD and SdhC. The amino acid change S121P in SdhD provided moderate levels of resistance to isopyrazam, penthiopyrad and pyraziflumid, while high levels of resistance to the same fungicides were conferred by the changes H137R, in the same subunit, and the changes G151R and G172D in SdhC. Lastly, high levels of resistance to isofetamid were associated with the presence of the amino acid change A86V in subunit SdhC [
57]. In our study, this point mutation was observed in all
P. xanthii isolates with resistance to boscalid and fluopyram but also in isolates that were only resistant to fluopyram, independent of the resistant phenotype observed in vitro. However, the amino acid substitution G151R was also observed in two
P. xanthii isolates that presented resistance to boscalid and remained sensitive or low in resistance to fluopyram. Although boscalid and fluopyram were not tested in the work of Miyamoto and collaborators, other studies have explained the resistance to these two fungicides with the homologous position of these amino acid changes in other phytopathogenic fungi [
72]. For example, in a study on the phytopathogenic fungus
Zymoseptoria tritici, the substitution A84V was related to isofetamid and fluopyram resistance [
73]. Furthermore, in a study by Scalliet and collaborators (2012), this change interacted with the fluopyram aliphatic linker, which is a characteristic of this compound [
74]. In
B. cinerea, A86V has been associated with resistance to fluopyram, but sensitivity to boscalid [
75]. In C. cassiicola, the substitution S73P provided moderate levels of resistance to fluopyram [
50]. In relation to the amino acid change G151R, the equivalence change G150R has been described in one
Sclerotinia homoeocarpa isolate, which was resistant to several SDHI fungicides, such as boscalid, fluxapyroxad, isofetamid and penthiopyrad, but not to fluopyram [
76].
Although the most frequent described mechanism of resistance is the presence of point mutations in the corresponding target genes, other mechanisms (detoxification, overexpression of the target genes or the implication of drug efflux transporters) could be involved. In our study, most of the
P. xanthii isolates, which presented the amino acid change A86V, had cross-resistance to boscalid and fluopyram; however, eleven isolates with the same amino acid change were sensitive to boscalid and resistant to fluopyram. Therefore, could an alternative mechanism be involved in the SDHI resistance? The possibility was also raised in
Z. tritici when several fluopyram- and isofetamid-resistant isolates did not carry any point mutations in the different Sdh subunits [
73]. In other studies,
Z. tritici isolates highly resistant to DMI fungicides and poorly resistant to QoIs and SDHIs presented overexpression in the BcMFS1 gene, which encodes a major facilitator transporter (MFS), a superfamily of transporters involved in a drug efflux system [
77,
78]. In the dollar spot fungus
S. homoeocarpa, the overexpression of two ATP-binding cassette (ABC) drug efflux transporters (ShPDR1 and ShartD) explained the reduced sensitivity to nonrelated, site-specific fungicides, including boscalid (SDHI), iprodione (dicarboxamide) and propiconazole (DMI) [
79]. In the same pathogen, the amino acid substitution M853T in the transcription factor ShDR1 was responsible for the overexpression of the ABC transporter, resulting in fungicide resistance to propiconazole (DMI fungicide), iprodione (dicarboximide) and boscalid (SDHI fungicide; [
80]). In the wheat powdery mildew
B. graminis f. sp.
tritici, a BgABC1 gene was related to the overexpression of the ABC transporter in seeds treated with the DMI fungicide triadimefon [
81]. In relation to
P. xanthii, there is no information about the correlation of ABC or MFS transporter expression and fungicide resistance. However, the genome of this organism has recently been published [
52] and the implication of some of these transporter superfamilies in resistance to different fungicides, including SDHIs, could be explored in future studies.
An important part of resistance analysis is the biological cost that may be associated with different processes involved in the natural survival of the pathogen. The characterization of this cost is essential to predicting the behaviour of the entire pathogen population and to implementing disease control strategies in the future [
82]. This possible biological cost is usually studied, among other approaches, based on sporulation, mycelial growth or the aggressiveness of the study isolates [
83,
84]. The results observed in the present study showed no fitness cost in a collection of representative SDHI-sensitive and SDHI-resistant
P. xanthii isolates on mycelial development and spore germination. In studies on other phytopathogenic fungi, such as
A. alternata, similar results were obtained, and no differences were observed between germination, hyphal development, sporulation or virulence when boscalid-sensitive and resistant isolates were compared [
61]. In addition, the boscalid resistance levels did not decrease after various subcultures in the absence of this fungicide, indicating the stability of resistance without selection pressure [
61]. Similar results were observed in
A. solani, with no significant differences in spore germination, mycelial expansion or aggressiveness in in vivo tests among sensitive and resistant isolates to several fungicides (anilinopyrimidine [AP], QoI and SDHI, [
84]).
Fungicide resistance monitoring studies on cucurbit powdery mildew are usually performed using bioassays with plants, meaning there is a great investment of time and material [
34]. However, when the mechanism of resistance is known and is caused by point mutations in the corresponding target gene, the detection of the different phenotypes can be performed using molecular methods, making it a better solution than time-consuming bioassays. In recent years, the LAMP technique has become an interesting alternative that offers the possibility of obtaining results from field samples within a few hours, and is an attractive tool to use in resistance monitoring studies. In the present study, this technique was developed to detect the two-point mutations (A86V and G151R) observed in the SDHI-resistant population of
P. xanthii in less than 40 min, complementing a previous study on the detection of MBC-resistant isolates carrying the E198A substitution in cucurbit powdery mildew [
10]. All this information will help to provide faster responses to growers regarding the effectiveness of these fungicides in fields affected with this disease.
The registration of SDHI fungicides, resistance to this class of fungicides in
P. xanthii has been described in several countries [
85]. Moreover, high levels of resistance to other nonrelated fungicides have been reported previously in this pathogen in Spain and in other parts of the world [
7,
8,
9,
57,
86,
87,
88,
89]. In 2020, the European Committee approved the European Green Deal, which proposes the promotion of an efficient use of resources, the restoration of biodiversity, and the reduction of pollution, toward a climate-neutral Europe by 2050. This change would be achieved through a series of objectives, including reducing pesticide use by 50% [
90]. This reduction is being reflected in Spain and, since 2017, five active substances belonging to three different chemical families [cyproconazole, flutriafol and hexaconazole (DMIs), kresoxim-methyl (QoIs) and quinoxyfen (aza-naphthalenes)] have been withdrawn from use. Due to the imminent reduction in chemical tools to control fungal diseases, the time needed to generate new substances from the phytosanitary sector (with a mean of ten years), and the rapid development of resistance by this and other fungal plant pathogens, it is fundamental to perform monitoring studies for testing the efficacy of each fungicide to increase the effectiveness of these compounds over time and to slow down the emergence of resistance. The implementation of integrated pest management (IPM) with, among others, the alternation between single-site and multisite fungicides, changes the mode of action between single-site fungicide families and is a good practice in the field. This may control the rise of resistance to SDHI and other families of fungicides, which is currently necessary.