Reduced fitness cost and increased aggressiveness in fenhexamid-resistant Botrytis cinerea field isolates from Chile

Copyright: © 2021 M. Esterio, C. OsorioNavarro, M. Azócar, C. Copier, M. Rubilar, L. Pizarro, J. Auger. This is an open access, peer-reviewed article published by Firenze University Press (http://www.fupress.com/pm) and distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.


INTRODUCTION
Gray mold (caused by Botrytis cinerea Pers.: Fr.) is the most economically important disease in Chilean table grape production. Botrytis cinerea infection is favoured under wet conditions with temperatures below 22°C; it is a cool-season disease. Environmental conditions between late winter and spring in Chile usually provide the requirements for B. cinerea infection in the table grape growing area, causing blossom blight during the bloom period at the beginning of the season. Botrytis cinerea infections may also remain latent (Keller et al., 2003;Viret et al., 2004), leading to disease appearance after harvest either during storage or after purchase by consumers.
Control of B. cinerea on diverse crops is commonly achieved with combinations of pesticide and agronomic practices. Agronomic practices alone cannot prevent the disease in central Chile, so chemical treatments must be applied (Esterio et al., 2011). Because of the epidemiological traits of B. cinerea, disease forecasting models are not commonly used. Instead, treatments are applied at fixed phenological plant stages: bloom, bunch closure, veraison, and pre-harvest. However, more sprays may be scheduled under specific weather events that increase the risks of disease outbreaks. Among the wide range of fungicides registered for use against B. cinerea, fenhexamid, a hydroxyanilide derivate, has become a key component of gray mold management in Chilean table grape vineyards.
The sterol-3-ketoreductase enzyme (3-KR) encoded by the Erg27 gene is the biological target of fenhexamid. This enzyme is required for C4 demethylation during ergosterol biosynthesis (Debieu et al., 2001). Inhibition of 3-KR leads to ergosterol depletion and accumulation of cytotoxic-ergosterol precursors, triggering defects in central cellular processes (Akins, 2005). Resistance to fenhexamid in B. cinerea has been reported in vineyards from Europe and the United States of America (Fillinger and Walker, 2016), and is linked to several mutations in the Erg27 gene. High level of resistance occurs in isolates carrying single point mutations in codon 412 (Fillinger et al., 2008).
Since the introduction of fenhexamid in 1999, this fungicide has been widely used to B. cinerea control in table grape vineyards in Chile, being applied mainly during the grapevine bloom period. Fenhexamid resistance was reported in B. cinerea isolates from the Central Valley of Chile in the 2006-2007 growing season (Esterio et al., 2007;Esterio et al., 2011). Therefore, alternation of fungicides with different modes of action has been the strategy widely used for B. cinerea chemical control, in order to reduce the selection pressure.
Acquisition of high-level specific resistance to fenhexamid in B. cinerea has been described in isolates carrying mutation in codon 412 of Erg27, and has been associated with important decreases in pathogen fitness, including reduced conidium germination, myceliuml growth, and sclerotium development. Consequently, field problems associated with loss of efficacy of fenhexamid have not been reported to date (Ziogas et al., 2003;De Guido et al., 2007;Billard et al., 2012). In recent years, however, loss of sensitivity to fenhexamid has progressively and persistently increased in table grape fields in central Chile (Esterio et al., 2017). In order to maintain and promote fenhexamid effectiveness, the fitness cost of fenhexamid-resistance isolates from central Chile must be determined, and these should be included during construction of comprehensive and updated B. cinerea control programmes. For this purpose, B. cinerea isolates were recovered from six 'Thompson Seedless' vineyards managed with at least two fenhexamid applications per growing season to: (i) assess their sensitivity of the isolates to fenhexamid; (ii) determine their Erg27 genotype; and (iii) evaluate fitness parameters.

Botrytis cinerea isolation and culture media
Botrytis cinerea isolates were recovered during the 2013-2014, 2014-2015 and 2015-2016 growing seasons, from grapevine flowers collected at the full bloom and berries with 16.5° Brix stages, from six cv. Thompson Seedless vineyards located in the Chilean Central Valley, covering the three most important table grape production areas of Valparaíso Region (VR), Metropolitan Region (MR) and O'Higgins Region (OR). These vineyards had been undergoing field programmes with high fungicide pressure, being sprayed at least twice with fenhexamid per growing season. Botrytis cinerea singleconidium cultures isolated from these three regions were grown on malt yeast agar (20 g L -1 malt extract, 5 g L -1 Bacto yeast extract, 12.5 g L -1 agar) maintained at 20°C in constant darkness until conidiation. In total, 132 isolates from VR, 118 from MR, and 158 from OR were used in this study.

Fenhexamid sensitivity assay
Fenhexamid sensitivity was evaluated in vitro using colony growth tests. Colony growth tests were made on plates containing Sisler synthetic medium (2 g L -1 KH 2 PO 4 , 1.5 g L -1 K 2 HPO 4 , 1 g L -1 (NH 4 ) 2 SO 4 , 0.5 g L -1 , MgSO 4 ·7H 2 O, 10 g L -1 glucose, 2 g L -1 yeast extract and 12.5 g L -1 agar) (Leroux et al., 1999) supplemented with different concentrations of fenhexamid (0; 0.03; 0.1; 0.3; 1; 3 and 10 mg L -1 ). Four-day-old mycelium plugs were seeded on the plates, and the cultures were then kept for 5 days at 20°C in darkness. Colony growth was determined by measuring the diameter of the resulting colonies. Three replicate plates were analyzed for each fenhexamid concentration in colony growth experiments. EC 50 values (effective inhibitory dose that gave half-maximum inhibitory responses) were calculated for each isolate using the Minitab Version 12 statistical software program.

Erg27 genotyping: amplification and sequencing
Botrytis cinerea genomic DNA was isolated from 7-day-old mycelia using the DNeasy Plant mini kit (QIA-GEN). A fragment of the Erg27 gene was amplified using primers erg1800down and erg27End, which amplify a 1052 pb fragment, previously described by Fillinger et al. (2008). The PCR mix was composed of 50-100 ng genomic DNA, 1X GoTaq® Green Master Mix (Promega) and 0.2 μM each primer; 25μL volume was completed with nanopure water (Promega). The PCR product was purified and used for sequencing (Macrogen). Identification of Erg27 genotypes was performed by alignment of the sequences using BioEdit software (Hall, 1999).

Pathogenicity test
To assess the pathogenicity on grape berries B. cinerea isolates were inoculated onto wounded and unwounded berries of 'Thompson Seedless', at harvest stage based on soluble solids content (16.5°Brix). The berries were washed in 1% sodium hypochlorite solution for 0.5 min, rinsed twice with sterile distilled water and allowed to dry under a laminar flow hood. Subsequently, a 10 μL droplet of B. cinerea isolate suspension (10 6 conidia mL -1 ) was inoculated on the surface of each unwounded or wounded berries. Wounding was made by puncturing each berry with a sterile needle to a depth of 2 mm. Inoculated berries were incubated at 0 or 20°C in sealed humidity chambers (80% relative humidity) for 4 days and the diameter of the gray mold lesion on each berry was measured. Eighteen berries were used for each Erg27 genotype separated into three replicates. The experiment was repeated twice independently, firstly using table grape berries from seasons 2016-2017 and and second from 2018-2019.

Evaluations of colony growth, conidium production and sclerotium development
Six isolates for each identified Erg27 genotype were used in this study, including wild type (no mutations in the Erg27 gene). The exception was for erg27 F412V , where only three isolates were found. In each case, 4-days-old non-sporulating mycelium plugs grown in Potato Dextrose Agar (PDA) were transferred onto a fresh PDA plate for phenotype evaluation. Three plates were used for each genotype in two independent experimental repetitions.
Mycelium radial growth was evaluated for 4 or 5 days in continuous darkness, under three temperature conditions: 15, 20 or 25°C. Conidium production was evaluated after 17 days of continuous colonial growth in darkness at 20°C. For each evaluation, total sporulating mycelium was recovered in a vial with 15 mL of sterile water, which was stirred, and conidia concentration was determined using a haematocytometer. For sclerotium development, plates were incubated for 40 days in darkness at 5 or 20°C. Number and mass of sclerotia in each plate were recorded, and the Sclerotium Index was defined as the ratio of total number of sclerotia to total sclerotium mass per plate.

Statistical analyses
Statistical analyses were carried out using ANOVA and the Bonferroni post hoc test in InfoStat software (Di Rienzo et al., 2015).

Sensitivity of Botrytis cinerea isolates to fenhexamid
Sensitivity to fenhexamid of each isolate was evaluated through the mycelium growth EC 50 . The isolates were then classified as sensitive (Fen S ) or resistant (Fen R ) to fenhexamid, using the recommended cutoff value for field applications of fenhexamid (0.17 mg L -1 ; Teldor-Bayer). Thirty-six isolates were obtained from VR vineyards in the 2013-2014 season and 11% of these were Fen S , 72 isolates were obtained in 2014-2015 and 6% were Fen S , and 24 isolates were obtained in 2015-2016 and 13% were Fen S ( Figure 1A, Table 1). In the 2013-2014, 2014-2015 and 2015-2016 seasons, 35, 36, and 47 isolates from the MR vineyards were analyzed, and 23%, 25%, and 0% of them were Fen S ( Figure 1B, Table 1). Of the isolates from OR vineyards 40, 72, and 46 isolates were obtained in the seasons 2013-2014, 2014-2015 and 2015-2016, respectively, among them 18%, 15% and 19% presented sensitivity to the fungicide ( Figure 1C, Table  1). Resistance to fenhexamid was classified as low when 0.17 mg L -1 > EC 50 ≥ 2 mg L -1 and high when EC 50 ≥ 2 mg L -1 , considering the cutoff value described by Fillinger et al. (2008). The frequency of fenhexamid-resistant and highly resistant isolates in the B. cinerea population analysed in this study increased with time and this occurred in the three geographical regions under study.

Genetic characterization of Erg27 in Botrytis cinerea isolates
Mutations in wild type Erg27 allele in B. cinerea isolates from the field and laboratory-generated strains have been associated with different ranges of loss of sensitivity to fenhexamid (Fillinger et al. 2008;Esterio et al. 2011;Grabke et al. 2013;Amiri and Peres, 2014). In particular, mutations in 412 codon of Erg27 trigger high resistance to this fungicide (Fillinger et al., 2008;Debieu and Leroux, 2015;Fillinger and Walker, 2016).
The Erg27 genotypes of isolates from the three regions (2015-2016 season) were evaluated in order to find a genetic factor associated with resistance to fenhexamid. Of a total of 24 isolates from the VR region, 29% carried a serine substitution (erg27 F412S ) and 58% the isoleucine substitution (erg27 F412I ) at position 412, and only 13% of total isolates maintained phenylalanine at the 412  . Fenhexamid effective concentration (mg L -1 ) was evaluated by EC 50 value (effective concentration that reduces mycelial growth by 50%). Isolates were considered as low resistance when 0.17 mg L -1 > EC 50 ≥ 2mg L -1 and high resistance when the EC 50 ≥ 2 mg L -1 ; bothof these sensitivity limits are shown by dashed lines.
position (Erg27 F412 ) (Table 2, Figure 2). As expected, all isolates from this region carrying the mutations in the 412 position of Erg27 were highly resistant to fenhexamid, and the isolates without mutation in this codon were fenhexamid sensitive. From the MR region, 47 isolates were tested. Substitutions erg27 F412S was at frequency of 62%, and erg27 F412I at 13%, while 19% of the isolates had no mutation in Erg27 F412 (Table 1, Figure 2). Three isolates (6%) carried a non-common mutation of valine instead phenylalanine at 412 position (erg27 F412V , 6%) (Table 1, Figure 2). All the highly resistant isolates in this population had mutations in codon 412 of Erg27, as expected. However, nine isolates with no mutation in Erg27 F412 showed some resistance to fenhexamid, suggesting that mutations in other positions of Erg27 or on another gene could be responsible for the resistance.
In the 46 isolates from OR, 48% carried the erg27 F412S mutation and 22% the erg27 F412I mutation, exhibiting high resistance to fenhexamide (Table 1, Figure  2). In this case, 30% of the isolates had no mutation at Erg27 F412 ; two of these isolates showed high resistance to the fungicide and four had low resistance.
The low and high resistance in isolates from the MR and OR regions that lacked mutations in Erg27 F412 raised the possibility of another codon of Erg27 being mutated and conferring resistance to fenhexamid. To answer this, the sequence of the Erg27 gene was scrutinized to identify other mutations. Five other mutations were found in the Erg27 gene, including erg27 L195F , erg27 P238S , erg27 Δ298 , erg27 R330P and erg27 N369D . These mutations were found in different combinations, in the 117 isolates analyzed from the three regions. erg27 P238S , erg27 L195F/ Δ298 and erg27 Δ298/ R330P were present in isolates that lacked mutation in position 412 of Erg27 and were resistant to fenhexamid ( Figure 3A), suggesting that these mutations could lead to resistance to fenhexamide. However, erg27 Δ298 by itself possibly did not affect resistance to this fungicide. Similarly, erg27 N369D combined with erg27 P238F/N369D did not give resistance to fenhexamid, although erg27 P238S by itself correlated with fenhexamid resistance. Mutations in other positions were also detected in the isolates carrying erg27 F412S or erg27 F412I ( Figure 3B-D). However clear correlations between their presence and fenhexamid resistance were not detected, indicating that mutations in position 412 are more relevant for fenhexamid resistance.

Evaluation of growth parameters and virulence of the Botrytis cinerea isolates carrying mutations in Erg27
Growth parameters and virulence were analyzed to evaluate the performance of fenhxamid resistant isolates from VR, OR, and MR carrying mutations in the 412 position of the Erg27 gene. Mycelium growth was evaluated under suboptimal (15°C) and optimal temperature conditions (20 or 25°C). No differences in mycelium radial growth were observed among field isolates with non-mutated Erg27 F412 and erg27 F412S , erg27 F412I or erg-27 F412V at the three growing temperature tested ( Figure  4A, 4B and 4C).
Development of sclerotia as survival structures is essential for overwintering of B. cinerea inoculum in the field. Therefore, sclerotium development was evaluated in two contrasting temperature conditions: 5 or 20°C. Numbers of sclerotia, sclerotia masses and sclerotia indices (ratio of numbers to masses) were quantified. No statistically significant differences were observed between Erg27 F412 , erg27 F412S or erg27 F412I at 5°C, but at this temperature, the restriction of sclerotium develop-
The infection capacity of the Erg27 mutants was measured on detached 'Thompson Seedless' grape berries using six isolates of each genotype. On unwounded berries incubated at 0°C, erg27 F412I isolates developed larger rot lesions than berries inoculated with Erg27 F412 , erg27 F412S or erg27 F412V isolates. In wounded berries, erg27 F412S and erg27 F412I generated larger lesions compared to the Erg27 F412 genotype, while erg27 F412V produced smaller lesions than the Erg27 F412 genotype isolates ( Figure 4H). At 20°C, unwounded berries inoculated with erg27 F412I or erg27 F412S isolates developed enhanced infection damage compared with Erg27 F412 . However, the necrotic damage observed in wounded berries infected by erg27 F412S or erg27 F412I did not differ from that caused by Erg27 F412 isolates. In contrast, erg27 F412V isolates were less aggressive than Erg27 F412 isolates in both unwounded and wounded berries (Figure 4I). Together, these results indicate that growth in the erg27 F412S , erg27 F412I and erg27 F412V isolates was not affected, although they carried a mutation in Erg27, and isolates carrying erg27 F412S and erg27 F412I were of increased aggressiveness on both unwounded and wounded berries. Table grape vines are cultivated mostly in the Central Valley of Chile because this region has favorable agroecological conditions for grape production. Grape vineyards in Chile are threatened by Botrytis outbreaks due to the frequent cool springs and wet weather conditions. Therefore, fungicides are commonly applied to vineyards in this region. Development of resistance to fungicides has been observed in B. cinerea in this area, endangering the ability to control gray mold (Latorre et al., 2015;Esterio et al., 2017).

DISCUSSION
In the present study B. cinerea populations collected from 'Thompson Seedless' table grape vineyards from three regions of Central Chile were shown to have reduced sensitivity to fenhexamid. This reduced sensitivity increased progressively in the the three successive growing seasons tested. Nevertheless, fenhexamid is still one of the main fungicides regularly used in the local B. cinerea control programmes, with at least two applications of this chemical in each growing season.
Botrytis cinerea resistant isolates to fungicides with unisite modes of action have been widely reported in the last 10 years, after intensive application programmes, accompanied by reduced fungicide efficacy (van den Bosch et al. 2015;Fillinger and Walker, 2016). The Erg27 mutation in position 412 is one of the most common changes linked to fenhexamid resistance (Fillinger et al., 2008;Billard et al., 2012;Debieu and Leroux, 2015). We reported isolates carrying erg27 F412S , erg27 F412I and erg-27 F412V mutations; these have been previously reported in B. cinerea isolated from fields in France, Germany and the United States of America, and have been associated with high fenhexamid-resistance levels (Grabke et al., 2013;Amiri and Peres, 2014;Rupp et al., 2017). Isolates carrying mutant versions of Erg27 were predominant, including erg27 F412I from VR and erg27 F412S from MR and OR. Strong correlations were observed between the presence of mutations at codon 412 of Erg27 and high resistance to fenhexamid (EC 50 ≥ 2 mg L -1 ). The erg-27 F412I and erg27 F412V genotypes showed the greatest EC 50 values in each population, while erg27 F412S presented the lowest EC 50 among the mutants. This indicates that this mutation conferred less resistance to fenhexamid. In all the Chilean regions analyzed in this study, progressive increases of the resistant isolates were detected over the three growing seasons assessed, demonstrating the effects of constant fungicide pressure on B. cinerea population.
In addition to high fenhexamid resistance related to mutation in Erg27, particularly in the 412 position, we detected other mutations that produced moderate levels of resistance in other Erg27 codons: erg27 P238S , erg27 L195F/ Δ298 and erg27 Δ298/ R330P . Isolates carrying mutations erg27 L195F and erg27 R330P are the first reported in Chile. Particularly, erg27 R330P associated with moderate resistance to fenhexamid has not been previously reported (Debieu and Leroux, 2015). The presence of erg27 P238S and erg27 N369D together suppressed resistance to fenhexamid more than in isolates carrying erg27 P238S alone. However, the level of resistance to fenhexamid remained unchanged in strains erg27 F412S and erg27 F412I when erg27 N369D was also present, suggesting that changes close to the 3-KR transmembrane domain were more relevant in the interaction between fenhexamid and 3-KR. Our data also suggest that the presence of erg27 F412S produced a second functional change within the Erg27 sequence, in contrast to erg27 F412I and erg27 F412V .
Mutations in position 412 of Erg27 have been previously reported to reduce isolate performance (Billard et al. 2012). However, the isolates erg27 F412I and erg27 F412S , identified in the present study grew similarly to fenhexamid-sensitive strains at 15°C, 20°C and 25°C. Sclerotium development and conidium production were also not affected in erg27 F412I and erg27 F412S isolates, in contrast to previous reports that showed growth retardation in fenhexamid-resistant strains (Billard et al., 2012;Saito et al., 2014). erg27 F412I and erg27 F412S B. cinerea isolates were more pathogenic, particularly in unwounded grape berries at all the temperatures tested. It is possible that the low effects on fitness and the increase in infection capacity observed in erg27 F412I and erg27 F412S were due to accumulation of additional mutations that conferred adaptive advantages for survival under high fungicide selection pressure (Ishii, 2015), overcoming the negative effect reported in strains carrying erg27 F412I and erg27 F412S (Billard et al., 2012). Isolates carrying erg27 F412V exhibited fitness costs, producing few conidia and possessing only minor increases in infection capacity. This suggests that this mutation may be rare in the field B. cinerea populations, being found only three times in isolates obtained in the present study.
Amino-pyrazolinone fenpyrazamine was recently introduced as a Botryticide for gray mold control in Chile. Fenpyrazamine, like fenhexamid, targets 3-KR (Kimura et al., 2017), but Erg27 changes associated with resistance to fenpyrazamine have not been studied. Fenpyrazamine could potentially control fenhexamid-resistant isolates by inhibiting 3-KR, targeting the enzyme independently of the amino-acid at position 412. Therefore, experiments determining fenpyrazamine efficacy on fenhexamid-resistant isolates are required, to provide a basis for restructuring chemical control strategies to reduce occurrence of highly resistant B. cinerea populations.
The present research has highlighted the prevalence of fenhexamid resistance linked to the Erg27 genotype in B. cinera populations isolated from 'Thompson Seedless' vineyards treated with this fungicide in the Central Valley of Chile. These results show an overall reduction of fitness in fenehxamid-resistant B. cinera isolates, suggesting the appearance of adapted strains resistant to this fungicide. This poses serious risks for field control of gray mold in table grape production in Chile.