Gray mold caused by Botrytis cinerea Pers.Fr (teleomorph Botryotinia fuckeliana (de Bary) Whetzel) is one of the most destructive diseases of high-value crops throughout the world. The pathogen can infect more than 1,400 plant species, among which major agricultural crops are included such as grapes, greenhouse-grown vegetables, ornamental plants, and stone, pome, and small berry fruits. It causes mostly pre- or postharvest fruit decays and, in addition, leaf blights or leaf spots, stem cankers, flower rots, and damping off of seedlings (Molly and Grant-Downton, 2015; Shaw et al., 2016). Based on its extremely broad host range and its scientific and economic significance, B. cinerea is ranked as the second most important plant pathogenic fungus (Dean et al., 2012). Currently, efforts to develop new disease management methods that are highly effective and environmentally friendly are intensive and include RNA interference, biological control, or plant resistance inducers (Roca-Couso et al., 2021; Samaras et al., 2021; Spada et al., 2023). However, none of these methods can provide control efficacy similar to that provided by fungicides, and thus, chemical control remains the major gray mold management method under conditions favorable for the development of the disease. A wide array of fungicides that belong to several target-site- specific chemical classes, such as the anilinopyrimidines (Aps; FRAC 9), sterol biosynthesis inhibitors class III and SBIs class III (FRAC 17), phenylpyrroles (PPs; FRAC 12), quinone outside inhibitors (QoIs; FRAC 11), succinate dehydrogenase inhibitors (SDHIs; FRAC 7), benzimidazoles (MBCs; FRAC 1), and dicarboximides (DCs; FRAC 2), are available against the pathogen (Fillinger and Walker, 2016).

However, successful control of gray mold with the use of fungicides is hampered by resistance development, one of the major restricting factors in the success of chemical control because of reduction or even complete loss of control efficacy that may lead to failures in disease management. Such problems have been faced in the past in B. cinerea populations that have developed resistance worldwide to all site-specific fungicides, associated with target site modifications due to mutations’ presence in genes encoding the target site of the fungicides (Brent and Hollomon, 2007; Hahn, 2014; Fernandez-Ortuno et al., 2017; Fillinger and Walker, 2016; Shao et al., 2021). More recent studies have shown that in pathogen populations throughout the world, strains have been selected possessing more than one target-site alternation conferring resistance to the respective fungicides and leading to multiple resistance (MLR) to fungicides (de Miccolis Angelini et al., 2014; Rupp et al., 2017; Naegele et al., 2022).

In addition to target site modifications, more mechanisms have been associated with fungicide resistance development in B. cinerea. Among them, overexpression of efflux transporters located in the plasma membranes plays a crucial role and leads to resistance development in several chemically and structurally unrelated compounds (multidrug resistance (MDR)). Two major families of efflux transporters have been recognized and associated with resistance to fungicides in fungal species: ABC transporters (ATP-binding Cassette superfamily) and MFS transporters (Major Facilitator Superfamily transporters) (de Waard et al., 2006). The evidence for MDR due to increased fungicide efflux in plant pathogenic filamentous fungi is relatively limited. So far, field strains overexpressing ABC and MFS transporters have been described in B. cinerea, Zymoseptoria tritici, Penicillium digitatum, and Penicillium expansum (Kretschmer et al., 2009; Omrane et al., 2015; Samaras et al., 2020). In B. cinerea, four different types of MDR strains have been recognized (MDR1, MDR1h, MDR2, and MDR3) differing in both their spectrum and level of resistance to fungicides and linked with overexpression of atrB- and mfsM2- encoded efflux transporters (Kretschmer et al., 2009; Mernke et al., 2011; Leroch et al., 2013). In addition to differences in fungicide sensitivity levels, the various MDR types may also differ in their fitness level, although further investigations are required to clarify this issue (Kretschmer et al., 2009; Chen et al., 2016).

In Greece, B. cinerea has developed resistance to almost all the target- site-specific botryticides that have been used for its control on several crops (Bardas et al., 2008; Samuel et al., 2011; Veloukas et al., 2011; Chatzidimopoulos et al., 2013). Recently, it became evident that in fungal populations originating from major crops heavily treated with botryticides, subpopulations with multiple resistance have been built up, but information related to their frequencies remains scarce (Bardas et al., 2010; Konstantinou et al., 2015; Makris et al., 2022). However, nothing is known related to the presence of fungal strains possessing resistance mechanisms that are associated with MDR.

The development of resistance to fungicides by B. cinerea is a major threat to modern agriculture, and its successful management is a major challenge. With varying levels of resistance to the main fungicides now present in the fungal populations throughout the world, coupled with the expected decrease in the number of fungicide active ingredients available to European farmers as a result of the European Green Deal, a better understanding of how resistance develops and evolves becomes increasingly important. Furthermore, MLR or MDR is hard to manage with the most common tactics of fungicide resistance management that include alternations or use of fungicide mixtures. Therefore, knowledge of the composition of B. cinerea populations in terms of MLR or MDR frequencies is a prerequisite for the development and application of resistance management strategies. In addition, although the effect of MDR is considered to be low for plant pathogens since it is in general associated with low/moderate resistance levels, the effect on fungal sensitivity of its combination with target-site modifications remains largely unknown. Thus, this study was initiated aiming to i) determine frequencies of MLR and MDR in pathogen populations originating from different hosts, ii) identify mutations associated with either target site modifications or increased efflux activity of membrane transporters, and iii) determine the effect on the levels of resistance to fungicides of the simultaneous presence in the same strains of different resistance mechanisms. A preliminary report on parts of this study has already been presented at the 1st Joint Botrytis-Sclerotinia Symposium 2022 (Sofianos et al., 2022).

Monitoring data obtained in this study showed a widespread presence of resistance to several fungicide classes in B. cinerea populations in the most heavily treated crops in Greece. Such data confirm previous reports suggesting the development of high resistance frequencies in fungal populations originating from several hosts to APs, QoIs, SDHIs, DCs, or SBIs class III in several regions of Greece (Myresiotis et al., 2007; Avenot et al., 2008; Bardas et al., 2010; Samuel et al., 2011; Konstantinou et al., 2015). The resistance frequencies in fungal populations originating from strawberry, greenhouse- grown tomato, or rootstock seedlings were high for almost all fungicide classes, with the exception of PPs. All these crops are grown in closed environments such as greenhouses or high height tunnels that facilitate the development of high relative humidity conditions favoring the development of the disease and necessitating a high number of fungicide spray applications to successfully combat it. In contrast, lower resistance frequencies were observed in the populations originating from wine grapes. The dry climate of the Mediterranean region contributes to a relatively low risk for gray mold development on grapes and, thus, in a limited number of botryticide applications in contrast to northwestern Europe where grapevines are heavily treated with botryticides. Similarly low resistance frequencies have been observed in B. cinerea populations from grapes in Australia, while in northwestern European countries, resistance frequencies in the grape populations are higher (Leroch et al., 2011; Walker et al., 2013; Harper et al., 2022). Differences in the frequencies of fungicide resistance within B. cinerea populations originating from different hosts have been observed on several crops and several places of the world and have been correlated with differences in the level of crop susceptibility to the pathogen, the total number of fungicide spray applications applied on the crop, and the availability of cultural control methods as an alternative to chemicals (Cosseboom and Hu, 2021.)

In addition to the observed high resistance frequencies to specific fungicide classes in heavily treated crops, the results of this study showed that in the same crops, high frequencies of isolates showing resistance to multiple modes of action (MLR) have been selected. The observed frequencies of MLR strains in the populations originating from strawberries and tomatoes were higher compared to the respective frequencies reported in a previous survey from the same region (Konstantinou et al., 2015). Thus, in the current study, we documented a further increase in MLR frequencies within the fungal population. Multiple resistance to fungicides in B. cinerea is a rapidly expanding problem in the fungal populations worldwide, reported in several studies during the last decade (Leroch et al., 2011; de Miccolis Angelini et al., 2014; Fernandez-Ortuno et al., 2014; Li et al., 2014; Lu et al., 2016; Campia et al., 2017; Rupp et al., 2017; Cosseboom et al., 2020; Makris et al., 2022; Naegele et al., 2022). The observed increased frequencies of MLR strains in populations from crops heavily treated with fungicides are obviously associated with the high number of sprays required to successfully control the disease, the application of pre-packed mixtures of different active ingredients that are of high risk for resistance development, or the absence of effective alternative control methods that could contribute to a reduction of selection pressure. Moreover, in addition to botryticides applied against gray mold, the fungal population on some of these hosts may be exposed to fungicides that are applied against other foliar or fruit diseases. Such an exposure increases the selection pressure on the fungal population and may contribute to the increased MLR frequencies observed in our study.

Sequence analysis of the target genes of boscalid (SDHI), cyprodinil (AP), and fenhexamid (SBIs class III) uncovered a variety of mutations associated with target-site resistance. In the boscalid-resistant fraction of the populations originating from all the sampled hosts, H272R was found to be, by far, the most common mutation followed by N230I, while other mutations such as H272Y or P225F were found in very low frequencies. These results are in line with the results of several other studies conducted worldwide, suggesting that H272R is the most common sdhB mutation conferring resistance to some of the SDHI fungicides (Hu et al., 2016; Harper et al., 2022). Interestingly, in our study, an increase of the N230I mutation frequency accompanied by a decrease in the frequency of H272Y mutation was observed, compared to frequencies reported in previous monitoring studies conducted in Greece on the same hosts and the same sampling regions (Veloukas et al, 2011; Konstantinou et al., 2015). This type of change in the frequencies of these mutations is most probably associated with the introduction into the spray programs of a new SDHI fungicide, fluopyram, which has been registered for use on tomatoes, strawberries, or grapes against several diseases (Veloukas and Karaoglanidis, 2012). It is well established that N230I confers resistance to several SDHIs including fluopyram, while H272Y is associated with hypersensitivity to fluopyram (Veloukas et al., 2013; Laleve et al., 2014; Fernandez-Ortuno et al., 2017). Thus, the introduction of fluopyram into the spray programs led to a reduction of the fluopyram-hypersensitive H272Y strains and favored the selection of fluopyram-resistant N230I strains. Overall, these data confirm the necessity of continuous monitoring for changes in the sdh gene mutation frequencies within the fungal populations to adopt the appropriate resistance management strategies since it is well established that the different sdh mutations affect in different ways the level of sensitivity to various SDHIs and the fitness of the resistant strains (Veloukas et al., 2013; Laleve et al., 2014; Veloukas et al., 2014).

Sequencing of erg27 gene in the fenhexamid-resistant fraction of the population revealed the presence of a wide array of mutations associated with resistance to SBI class III fungicides. In total, within the fenhexamid-resistant fraction of the population, six different mutations were identified, with F412S being predominant in the populations originating from tomatoes and strawberries. In contrast in the population originating from stone-fruit rootstock seedlings, the predominant genotype was that of isolates possessing a combination of three different erg27 mutations (F412S, E263D, and P250F). All these six mutations have been reported previously in several monitoring studies (Grabke et al., 2013; Avenot et al., 2020; Alzohairy et al., 2021). Mutations located in codon 412 of erg27 strongly affect the binding of SBI class III fungicides to their target site and are associated with high resistance levels to fenhexamid. F412S has been reported as the most common erg27 mutation conferring resistance to this fungicide class, wherever resistance to SBIs class III has been developed; thus, our results are in agreement with these reports (Fillinger et al., 2008; Grabke et al., 2013).

Resistance to APs is also widespread throughout the world (Latorre et al., 2002; Leroch et al., 2013; Shao et al., 2021), while in Greece, the first cases of APs resistance in B. cinerea populations were documented in 2007 (Myresiotis et al., 2007). Despite the fact that APs is a relatively old fungicide class introduced against gray mold during the 1990s and the first cases of resistance were reported in the late 1990s (Chapeland et al., 1999), it was only in 2017 when Mosbach et al. (2017) identified several AP resistance-conferring mutations in genes encoding proteins that are involved in mitochondrial processes, providing evidence that APs are primarily targeting the mitochondrial metabolism. Among several mutations identified as AP resistance-conferring mutations, L412F in Bcpos5 and E407K in Bcmdl1 were included. The former has been found as the predominant mutation in the AP-resistant subpopulation tested by Mosbach et al. (2017), while, more recently in an AP-resistant subpopulation originating from China, the latter was found to be predominant (Fan et al., 2023).

In addition to target site resistances in the fungal population, the presence of isolates showing MDR phenotype was evident. These isolates were showing low-to- moderate levels of resistance to fludioxonil, a fungicide for which only rarely has been reported target site resistance in B. cinerea because of the low fitness of the resistant isolates (Dowling et al., 2021; Wang et al., 2021). The sequencing of the transcription factor gene mrr1 in these isolates confirmed their identity as MDR strains. The presence of MDR strains is confirmed for the first time in Greece, while previously their presence has been reported in Germany, France, the USA, and China (Kretschmer and Hahn, 2008; Kretschmer et al., 2009; Leroch et al., 2013; Fernandez-Ortuno et al., 2014; Shao et al., 2021). The sequencing of mrr1 gene of the MDR strains revealed the presence of three distinct genotypes. The predominant genotype was that of the MDR1h strains characterized by the 3-bp deletion at codon 497 in the mrr1 exon. Apart from this deletion, the MDR1h isolates had several target point alterations throughout mrr1 gene. MDR1h was identified for the first time in strawberry fields in Germany (Fournier et al., 2005; Leroch et al., 2013), while, later on, it was also reported in strawberry fields in the USA (Fernandez-Ortuno et al., 2015). The expression levels of atrB in MDR1h strains are generally higher compared to those of MDR1 leading consequently to higher levels of resistance to fungicides (Leroch et al., 2013; Fernandez-Ortuno et al., 2015). MDR1h strains were found to be the predominant MDR strains in the population originating from rootstock seedlings, while they were completely absent in the population originating from tomatoes. The predominance of MDR1h strains on rootstock seedlings is associated with the fact that this crop is the most heavily treated crop with fungicides among those sampled, while it is well established that the selection of MDR1h strains is stronger in most heavily treated crops than in crops where fungicide use is less common (Leroch et al., 2013). The second MDR genotype identified in the current study was a set of MDR1 isolates possessing seven different mutations in the mrr1 sequence. Most of these mutations have already been reported as mrr1 mutations associated with the overexpression of BcatrB in MDR1 isolates from Europe, Australia, and the USA (Kretschmer et al., 2009; Leroch et al., 2013; Fernandez-Ortuno et al, 2015; Harper et al., 2022). Interestingly, the third genotype identified as MDR1 was a set of isolates that possessed only the V575G substitution in mrr1 gene. All the MDR1 isolates originating from tomatoes possessed this mutation, while it was also observed in a limited number of isolates originating from strawberries. This mutation is reported for the first time; however, another mutation in the same codon (V575M) has been previously reported by Leroch et al. (2013) and was associated with the overexpression of BcatrB. All the isolates possessing this mutation have been identified as B. cinerea sensu stricto.

Interestingly, in all of our MDR1 or MDR1h isolates, the resistance mechanism associated with atrB overexpression coexisted with target site modifications in the same strains. The higher accumulation of target site resistances in MDR strains of the pathogen has been observed in the past in B. cinerea populations from several hosts in Germany (Leroch et al., 2013) and from strawberries in the USA and Norway (Fernandez-Ortuno et al., 2015; Nielsen et al., 2022). Thus, our results provide further evidence supporting the theory of a stepwise pattern of accumulation of resistance mutations in B. cinerea populations (Li et al., 2014). However, little is known about the effect of the coexistence of different resistance mechanisms on the levels of resistance of B. cinerea strains. In general, overexpression of efflux transporters as a resistance mechanism leads to low/moderate levels of resistance in plant pathogenic fungi (Leroch et al., 2013). The combination of different resistance mechanisms in the same strain may lead to increased resistance levels as has been shown in fungal species such as P. digitatum or Z. tritici (Omrane et al., 2015; de Ramon-Carbonell and Sanchez-Torres, 2020) Herein, we provided evidence that the coexistence of two different resistance mechanisms increases the level of resistance to some fungicides and thus increases the risk for the lower performance of the fungicides that are affected. However, this was not the case for fenhexamid for which the coexistence of erg27 mutations with mrr1 modifications did not lead to resistance levels higher than those obtained due to the erg27 modifications. This result is in accordance with previous studies showing that overexpression of BcatrB (and thus MDR1/MDR1h) is not correlated with increased detoxification of fenhexamid (Kretschmer et al., 2009).

In conclusion, in the current study, we reported on changes in the multiple resistance frequencies of B. cinerea populations originating from several hosts in Greece and on the frequencies of multidrug-resistant strains, which are reported for the first time in Greece. The data presented in this study highlight the practical importance of MLR/MDR in the field since resistance levels against certain fungicides might be higher than previously anticipated and might be used as a tool to design resistance management programs. The observed high frequencies of MLR isolates and the confirmed presence of MDR strains within the fungal populations present a challenge for the control of the pathogen, particularly in crops where multiple botryticide applications are required. This challenge necessitates research toward the development of fungicides with novel modes of action or the adoption of novel control strategies that will lead to the successful control of the selected MLR and MDR strains and the reduction of the selection pressure.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

GS: Data curation, Investigation, Writing – original draft. AS: Data curation, Investigation, Writing – original draft. GK: Funding acquisition, Supervision, Writing – review & editing.