Mechanisms of Resistance to an Azole Fungicide in the Grapevine Powdery Mildew Fungus, Erysiphe necator.

Page 1 of 40

1

Mechanisms of resistance to an azole fungicide in the grapevine powdery

2

mildew fungus, Erysiphe necator

Phytopathology "First Look" paper • http://dx.doi.org/10.1094/PHYTO-07-14-0202-R • posted 10/01/2014 This paper has been peer reviewed and accepted for publication but has not yet been copyedited or proofread. The final published version may differ.

3 4

Omer Frenkel*, Lance Cadle-Davidson*, Wayne F. Wilcox and Michael G. Milgroom

5 6

*The first two authors contributed equally to this work

7 8

First and fourth authors: School of Integrative Plant Science, Section of Plant Pathology

9

and Plant-Microbe Biology, Cornell University, Ithaca, NY 14853; second author: United

10

States Department of Agriculture, Agricultural Research Service Grape Genetics

11

Research Unit, Geneva, NY 14456; and third author: School of Integrative Plant Science,

12

Section of Plant Pathology and Plant-Microbe Biology, Cornell University, Geneva, NY

13

14456

14 15

Accepted for publication xx xxxxx 2014

16 17

Corresponding author: Michael Milgroom; E-mail address: [email protected]

18 19

Current address of first author: Department of Plant Pathology and Weed Research,

20

Institute of Plant Protection, Agricultural Research Organization (ARO), The Volcani

21

Center, Bet Dagan, 50250, Israel

22

1


Page 2 of 40

23

Mention of trade names or commercial products is solely for the purpose of providing

24

specific information and does not imply recommendation or endorsement by the U.S.

25

Department of Agriculture. USDA is an equal opportunity provider and employer.

26

2


Page 3 of 40

27

-----------------------------------------------------------------------------------------------------------

28

Abstract

29

Frenkel, O., Cadle-Davidson, L., Wilcox, W.F., and Milgroom, M.G. 2014. Mechanisms

30

of resistance to an azole fungicide in the grapevine powdery mildew fungus, Erysiphe

31

necator. Phytopathology xx:xxx-xxx.

32 33

We studied the mechanisms of azole resistance in Erysiphe necator by quantifying the

34

sensitivity to myclobutanil (EC50) in 65 isolates from the eastern U.S. and 12 from Chile.

35

From each isolate, we sequenced the gene for sterol 14α-demethylase (CYP51), and

36

measured the expression of CYP51 and homologs of four putative efflux transporter

37

genes, which we identified in the E. necator transcriptome. Sequence variation in CYP51

38

was relatively low, with sequences of 40 U.S. isolates identical to the reference sequence.

39

Nine U.S. isolates and five from Chile carried a previously identified A to T nucleotide

40

substitution in position 495 (A495T), which results in an amino acid substitution in codon

41

136 (Y136F) and correlates with high levels of azole resistance. We also found a

42

nucleotide substitution in position 1119 (A1119C) in 15 U.S. isolates, whose mean EC50

43

value was equivalent to that for the Y136F isolates. Isolates carrying mutation A1119C

44

had significantly greater CYP51 expression, even though A1119C does not affect the

45

CYP51 amino acid sequence. Regression analysis showed no significant effects of the

46

expression of efflux transporter genes on EC50. Both the Y136F mutation in CYP51 and

47

increased CYP51 expression appear responsible for azole resistance in eastern U.S.

48

populations of E. necator.

49

-----------------------------------------------------------------------------------------------------------

3

Page 4 of 40


50 51

Pests and pathogens of all taxonomic groups invariably evolve resistance to pesticides or

52

drugs aimed at their management (30). Human-pathogenic fungi have evolved resistance

53

to antifungal drugs (3, 40), and in agriculture, plant-pathogenic fungi have evolved

54

resistance to fungicides used to manage plant diseases (28). Understanding the process by

55

which fungicide resistance evolves is essential for delaying its development, or managing

56

it subsequently (38, 39). Most modern fungicides are considered to have single-site

57

modes of action, to which high levels of resistance can evolve by a single mutation,

58

sometimes resulting rapidly in disease control failures (4, 28). In contrast to the

59

distinctively bimodal distributions of fungicide sensitivity to many fungicides—fungal

60

individuals are either sensitive or resistant—sensitivity to azole fungicides is a more

61

continuous phenotype, and high levels of resistance evolve more slowly (13, 22). Azole

62

fungicides, which include triazoles and imidazoles, are the most widely used group of

63

sterol 14α-demethylation inhibiting (DMI) fungicides for treating both human- and plant-

64

pathogenic fungi. Quantitative variation in azole sensitivity is thought to result from

65

multiple mechanisms and multiple genes contributing to resistance (8, 13, 24). To fully

66

understand and manage resistance to azoles, we need to understand the mechanisms and

67

underlying genetics affecting these fungicides.

68

Reduced sensitivity to azoles occurs by three known mechanisms (13). The first

69

involves target site mutations in the enzyme sterol 14α-demethylase (CYP51), a

70

cytochrome P450 (CYP) monooxygenase essential for ergosterol biosynthesis in fungi.

71

Several mutations in CYP51 affecting the amino acid sequence are associated with

72

reduced sensitivity to different azoles; the successive accumulation of these mutations in

4


Page 5 of 40

73

some fungi affects the range of different azoles to which sensitivity is reduced (12, 25).

74

The second mechanism that reduces sensitivity is the constitutive overexpression of

75

CYP51. Overexpression has been shown in several fungi to be caused by mutations in the

76

promoter region upstream of CYP51 (27, 33). The third mechanism involves the

77

overexpression of genes that control the efflux of azoles out of fungal cells. Genes

78

associated with efflux of fungicides are mainly ATP binding cassette (ABC) transporters

79

or major facilitator superfamily (MFS) transporters (8, 24). The types of mutations and

80

mechanisms operating in a particular fungus have different effects on the azole sensitivity

81

phenotype depending on which azole fungicide is used. Moreover, these genes/mutations

82

may have epistatic effects depending on the genetic background of the fungal individual

83

(12). Therefore, it would be helpful to understand resistance mechanisms for avoiding

84

and managing azole resistance.

85

The grapevine powdery mildew fungus, Erysiphe necator (syn. Uncinula

86

necator), is one of the first plant-pathogenic fungi in which mutations to CYP51 were

87

shown to be associated with azole resistance (15). Resistance to azoles and other

88

demethylation-inhibiting (DMI) fungicides has been described in North American and

89

European populations of E. necator, and in some cases has been associated with

90

decreased disease control (11, 15, 17, 18, 29, 45). As found in other fungi, the

91

distribution of sensitivities (expressed as EC50) was mostly continuous, and in some cases

92

with distinctly higher mean EC50 values in populations exposed to applications of azoles

93

(11, 18, 43, 45). Délye et al. (15) found a point mutation in CYP51, causing an amino

94

acid substitution from tyrosine (Y) to phenylalanine (F) in codon 136 (denoted Y136F),

95

to be associated with a significant degree of azole resistance. This same amino acid

5


Page 6 of 40

96

substitution was later shown to account for a major portion of triadimenol resistance in

97

the barley powdery mildew fungus, Blumeria graminis f. sp. hordei (44). Y136F has

98

been found in azole-resistant isolates of E. necator in Virginia (11), and has been used for

99

surveying some populations in France and Italy for azole resistance (17, 29). However,

100

not all of the highly resistant isolates examined carried this mutation (15, 29), nor is

101

resistance always high in isolates with Y136F (29). Such limitations in the correlation

102

between the Y136F mutation and high levels of azole resistance indicate that monitoring

103

populations only for this mutation is not adequate for estimating the frequency of azole

104

resistance. It is more likely that multiple mechanisms are conferring resistance, and the

105

distribution of azole sensitivity may vary among populations as a function of the

106

frequencies of these mechanisms. Furthermore, knowing which mechanisms are

107

operating might be useful in choosing which chemicals to apply because different CYP51

108

mutations or efflux transporters may affect cross resistance relationships (13, 24).

109

As a nonmodel system, studies on the biology and genetics of E. necator are

110

relatively challenging. For example, we have limited ability to obtain ascospores from

111

controlled crosses (21, 35), preventing us from associating azole resistance with

112

segregation of specific mutations, as done with the cereal powdery mildew fungus,

113

Blumeria graminis (44); and a robust transformation system is not yet available for most

114

powdery mildew fungi (9), precluding the option of gene-knockout experiments to

115

determine the effects of specific genes on azole resistance phenotypes. Moreover, a

116

genome sequence of E. necator is not yet available, as it is for two formae speciales of B.

117

graminis and two powdery mildew species that infect Arabidopsis (34, 42). However, the

118

transcriptome of one North American isolate was sequenced recently and includes

6


Page 7 of 40

119

homologs of ABC and MFS transporter genes associated with azole sensitivity in other

120

fungi (L. Cadle-Davidson and M. G. Milgroom, unpublished results). Identifying these

121

homologous sequences makes it possible to test whether their expression correlates with

122

azole sensitivity in E. necator. Similarly, we can determine the expression levels of

123

CYP51, even though the promoter sequence is not available. To study these relationships,

124

we looked for associations between azole sensitivity phenotypes and mutations in

125

CYP51and expression of candidate efflux transporters in samples from a diverse, sexually

126

reproducing population of E. necator.

127

The primary objectives for this study were to: 1) Determine the distribution of

128

sensitivity to the azole fungicide myclobutanil in a diverse, native population of Erysiphe

129

necator in the eastern U.S.; and 2) Determine whether mutations and expression of

130

CYP51, and expression of ABC transporter and MFS genes significantly contribute to

131

variation in myclobutanil sensitivity. A secondary objective was to look for correlations

132

between these same traits and myclobutanil sensitivity in a sample E. necator isolates

133

collected from vineyards in Chile where resistance to azoles was suspected.

134 135

Materials and Methods

136

Sampling of E. necator. We analyzed azole sensitivity in samples of E. necator collected

137

from the eastern U.S., where it is native, and from Chile where it was introduced. A

138

complete list of isolates and their hosts and geographic origins are found in the

139

supplemental Table S1. Of the 65 isolates from the eastern U.S., 28 were collected from

140

wild hosts (V. aestivalis, V. labrusca, V. riparia and V. rotundifolia), and 37 were from

141

cultivated grapevines (V. vinifera, V. rotundifolia, Vitis × labrusca and Vitis interspecific

7


Page 8 of 40

142

hybrids). The eastern U.S. population is phenotypically diverse (19), reproduces sexually

143

(6) and shows very little genetic structure (20). In contrast, populations outside the

144

eastern U.S., including California, are much less diverse and are composed of one or two

145

reproductively isolated genetic groups (A and B) (5, 20). Twelve isolates of E. necator

146

were sampled in 2007 and 2008 from commercial vineyards in northern Chile (Region

147

IV, vicinity of La Serena), in which the control of grapevine powdery mildew provided

148

by azole fungicides appeared to be substandard. Grapevine powdery mildew was first

149

reported in Chile in 1862 on imports of V. vinifera from Europe (14) and, therefore, we

150

predict that this population of E. necator has low diversity as do other introduced

151

populations. Because of the genetic differences between native and introduced

152

populations, samples from the eastern U.S. and Chile were analyzed separately. All

153

isolates were maintained in the laboratory on surface-disinfested grape leaves and

154

cultured for quantifying azole sensitivity and nucleic acid purification as described

155

previously (19).

156

Azole sensitivity assay: We used the assay described by Erickson and Wilcox

157

(18) to quantify the sensitivity of E. necator to the azole fungicide myclobutanil. This

158

fungicide is commonly used to manage grapevine powdery mildew and sensitivity of E.

159

necator to it is highly correlated with sensitivity to other azole fungicides (11, 18, 45).

160

The assay is based on inoculating leaf disks treated with different fungicide

161

concentrations and quantifying colony growth. Technical grade myclobutanil was

162

dissolved in acetone and diluted with Nanopure Type 1 water containing 0.05% Tween

163

20 to concentrations of 1, 0.5, 0.25, 0.06 and 0.016 µg/ml. The concentrations of acetone

164

in these solutions did not affect growth and conidiation of E. necator (18). Disks (9 mm

8


Page 9 of 40

165

diameter) were cut from surface-sterilized, young leaves of V. vinifera ‘Riesling’ grown

166

in a walk-in biotron. Each leaf disk was soaked for 60 min in a myclobutanil solution,

167

dried and placed on 1.5% water agar. Leaf disks soaked in Nanopure Type 1 water with

168

0.05% Tween 20 were used as the control. Leaf disks were inoculated by transferring a

169

conidial chain from the inoculum source as described previously (18), and incubated at

170

room temperature (25 C) for 7 days with 12h light/12h dark cycles, at which time the

171

maximum colony radius was measured under a dissecting microscope at 32×

172

magnification. Four replicate disks were used for each concentration per isolate. The

173

effective concentration to reduce colony growth by 50% (EC50) was determined by

174

regressing the percent of colony growth relative to the control treatment on the logarithm

175

of fungicide concentration as described previously (18).

176

Cyp51 sequencing. We PCR-amplified CYP51 and Sanger-sequenced PCR

177

products at the Cornell University Life Sciences Core Laboratories Center (CLC). We

178

used sequences of CYP51 in GenBank accessions EF649776.1 and U72657 to design

179

three sets of primers using Primer3 (37) to amplify 1682 bp of the coding region (Table

180

1). Because the beginning of the coding region was not suitable for primer design, we

181

used accession U72657 (15) to design the first primer; the rest were designed from

182

accession EF649776.1. The sequence in U72657 includes 203 bp in the upstream

183

noncoding region and allowed us to design a forward primer for amplifying the beginning

184

of the coding region. Sequences from the three parts of the gene were concatenated using

185

Vector NTI (Life Technologies, Grand Island, NY) and a multiple alignment was

186

constructed in ClustalW (36) to determine single-nucleotide polymorphisms (SNPs) and

187

haplotypes.

9

Page 10 of 40


188

Identification of homologs of ABC and MFS transporters in E. necator. ABC

189

and MFS transporter genes associated with azole resistance in ascomycetes (8, 23, 24)

190

were used to search for homologs in an early draft of an E. necator transcriptome

191

sequence (C. Cadle-Davidson and M. G. Milgroom, unpublished results). Amino acid

192

sequences of the following genes were used in this query: MFS gene Bcmfs1 (DHA14-

193

like major facilitator, GenBank no. AAF64435.2), BcatrD protein (GenBank

194

CAC41639.1) and BMR1 (GenBank BAA93677.1) from Botrytis cinerea (teleomorph

195

Botryotinia fuckeliana); and CDR1 (GenBank no. CAA54692.1) from Candida albicans.

196

cDNA sequences in the E. necator transcriptome were translated in all reading frames

197

into amino acid sequences using tBLASTn software at the Cornell University

198

bioinformatics high-performance computing (BioHPC) facility for query using amino

199

acid sequences of the above efflux transporters. Sequences in the E. necator

200

transcriptome were identified as putative homologs based on sequence similarity and then

201

blasted against the NCBI database using PBLAST to confirm their homology with

202

additional fungi. Sequences with high sequence similarity were considered homologs.

203

The same strategy was used to identify homologs of actin. Homologs to histone H3 and

204

β-tubulin were identified in E. necator previously (1, 41).

205

Quantitative reverse-transcription PCR (qRT-PCR) for quantification of

206

gene expression. We quantified the expression of genes by qRT-PCR using methods

207

described previously for E. necator (41). Isolates were cultured on surface-sterilized

208

detached leaves of V. vinifera ‘Cabernet Sauvignon’ for 8 days at 23 C in 12 hr light/12

209

hr dark conditions. Mycelia and conidia were harvested by applying fingernail polish

210

over the colony and leaf surface and peeling it off when it was dry (7). Total RNA was

10


Page 11 of 40

211

isolated using RNeasyTM (Qiagen, Valencia, CA, USA), as described by Cadle-Davidson

212

et al. (7). We used the iScript cDNA Synthesis Kit (Bio-Rad, Carlsbad, CA) to synthesize

213

cDNA. This process was repeated for all isolates and constituted biological replicates;

214

three replicates were obtained for some isolates.

215

For each RNA isolate, we performed qRT-PCR using a MyiQ thermal cycler

216

(Bio-Rad, Carlsbad, CA). Reactions included 1.5 µM Syto 9 for product quantification

217

(Invitrogen, Grand Island, NY), and 10 nM fluorescein (Sigma-Aldrich, St. Louis, MO)

218

for well-factor normalization. qRT-PCR was performed twice for each RNA sample

219

(technical replicates) and estimates were averaged for statistical analyses. PCR primers

220

were designed from transcriptome sequences using Primer3 for CYP51 and all homologs

221

(Table 1). We normalized expression data for candidate genes by the expression of three

222

constitutively expressed reference genes: actin, β-tubulin and histone H3 (Table 1). The

223

control for estimating relative expression was calculated as the average of three E.

224

necator isolates with EC50 values close to the mean baseline sensitivity of 0.01 - 0.02

225

ppm (18). To minimize any possible selection by exposure to azoles or other fungicides,

226

or recombination with resistant isolates, we chose control isolates that were collected

227

from wild Vitis species in Panther Creek State Park (Blood Mountain, GA), and Brevard,

228

NC, more than 10 km from any known vineyards (19). To calculate relative gene

229

expression we used the formula given by Pfaffl (32): =

( )∆ ( ) ( )∆

230

( )

(1)

231

where Etarget and Eref are the amplification efficiency of the target genes and the reference

232

(or normalizing) genes, respectively, and ∆CPtarget and ∆CPref include the differences in 11


Page 12 of 40

233

expression between the test isolates and the mean of three control isolates for the target

234

and reference genes, respectively.

235

Statistical analyses. We used stepwise regression to determine which factors

236

were associated with variation in the logarithm of EC50 values. Variables used in the

237

model included mutations in CYP51 sequences, expression of CYP51, expression of each

238

of the ABC and MFS transporter homologs, and all possible two-way interactions

239

between these variables. The probability for a variable to enter was set at α = 0.2, and the

240

probability to leave the model was set at α = 0.1. All analyses were done using JMP5

241

(SAS, Cary, NC).

242 243

Results

244

Distribution of azole sensitivity. EC50 values for myclobutanil varied over more than

245

two orders of magnitude (0.007 to 3.54 µg/ml) among the isolates of E. necator

246

examined, and therefore we present their distribution on a log scale (Fig. 1). Although

247

phenotypes were found throughout the range of values encountered, the distribution of

248

log EC50 values within this range was distinctly bimodal in both the U.S. and Chile, with

249

the two peaks at approximately 0.03 and 0.56 µg/ml for the more- versus less-sensitive

250

groupings, respectively, consistent with previously-documented distributions in E.

251

necator populations with significant DMI fungicide exposure versus those from baseline

252

populations (18, 43)

253

Sequence variation in CYP51. We found five haplotypes among the 65 CYP51

254

sequences determined for isolates from the eastern U.S. (Table 2). The most common

255

haplotype was found in 40 isolates; we refer to this as the wild type (CYP51+) because it

12


Page 13 of 40

256

is the most common and is not associated with myclobutanil resistance (see below). Nine

257

isolates carried haplotype CYP51A495T/A1170C, which is characterized by mutations at

258

nucleotides 495 and 1170 relative to the wild type. The mutation at nucleotide 495 causes

259

a tyrosine to phenylalanine substitution in codon 136 (Y136F) and is known to confer

260

azole resistance in E. necator (15), whereas the mutation at nucleotide 1170 (A1170C) is

261

a synonymous mutation in codon 343. Two other haplotypes also share the A1170C

262

mutation. We found an additional mutation, A1119C, which is a synonymous mutation in

263

codon 326, in haplotypes CYP51A1119C and CYP51A1119C/A1170C (Table 2).

264

Only two haplotypes were found among the 12 isolates from Chile. Seven carried

265

haplotype CYP51C608T, which is identical to the reference haplotype in accession

266

EF649776.1 (accession number KM077176) and five carried CYP51A495T/C608T

267

(KM077177), which has the Y136F amino acid substitution. Mutation C608T results in

268

an amino acid difference from isoleucine to threonine in codon 156 (I156T). This amino

269

acid substitution is characteristic of a known difference between genetic groups A and B

270

in Europe but is not associated with azole resistance (16). Microsatellite genotyping of a

271

sample of Chilean isolates confirmed they were in group B (data not shown).

272

All isolates carrying the Y136F mutation, from both the U.S. and Chile, were at

273

the higher end of the EC50 distribution (Fig. 1, Table 2). Similarly, U.S. isolates carrying

274

the mutation A1119C also had high EC50 values, even though this mutation does not

275

change the CYP51 amino acid sequence. EC50 values of isolates with either the Y136F or

276

A1119C mutation were significantly greater (P < 0.001) than those without these

277

mutations. There was no significant difference in mean EC50 values between U.S. isolates

278

carrying the A1119C versus the Y136F mutation (P = 0.23). Of the 20 isolates carrying

13


Page 14 of 40

279

mutation A1170C, 18 had haplotypes with either amino acid substitution Y136F or

280

mutation A1119C. However, the two CYP51A1170C isolates, in which no other mutations

281

to the gene occurred, had low EC50 values (Table 2), suggesting this mutation alone is not

282

associated with azole resistance.

283

Expression of CYP51. We obtained expression data for CYP51 from 47 E.

284

necator isolates from the eastern U.S. and seven from Chile. Several isolates were lost

285

before we could culture them for RNA isolation, therefore, our sample sizes are smaller

286

for CYP51 expression than for nucleotide sequences. Overall, the logarithm of CYP51

287

expression was positively correlated with log EC50 for myclobutanil in the U.S. (r = 0.50,

288

P < 0.001); Fig. 2). Relative to isolates carrying the wild-type haplotype, expression was

289

significantly (P < 0.001) higher for isolates carrying either the Y136F mutation or the

290

A1119C mutation, even though the amino acid sequence for the latter group was identical

291

to the wild type (Fig. 3). CYP51 expression was also significantly greater in isolates

292

carrying the A1119C mutation than in isolates carrying Y136F (P < 0.001). The mean log

293

expression of the Chilean isolates carrying the Y136F mutation was two times greater

294

than those without Y136F but was not significantly different (P = 0.19) due to small

295

sample sizes.

296

Identification and expression of ABC transporter and MFS homologs in E.

297

necator. We identified three ABC transporter homologs in the subgroup ABCG and one

298

MFS homolog in the E. necator transcriptome (Table 3). We refer to these homologs as

299

ABCG-1, -2, -3 and MFS1, respectively. To estimate the contribution of these four

300

homologs to reduced azole sensitivity, regression analyses were conducted on log EC50

301

values using expression data from 47 isolates from the eastern U.S. Because of the

14


Page 15 of 40

302

markedly bimodal distribution of log EC50 values, with isolates carrying mutations

303

Y136F and A1119C more resistant than the others (Fig. 1), we conducted regressions on

304

each group separately to see if any other factors could explain the remaining variance. No

305

significant factors were found by regression within either group. Similarly, nearly all of

306

the variance in log EC50 in the small sample from Chile was explained by mutation

307

Y136F (data not shown).

308 309

Discussion

310

Azole sensitivity in E. necator can be reduced by three known mechanisms: variation in

311

the amino acid sequence of CYP51, overexpression of CYP51 and efflux transporters that

312

pump azoles out of cells. The first two of these mechanisms were responsible for reduced

313

sensitivity to myclobutanil in E. necator isolates from the eastern U.S. and Chile. Amino

314

acid substitution Y136F alters the target site in CYP51 and is well known for its effect on

315

azole resistance in E. necator (15) and other fungi (28, 44). Y136F appeared to have a

316

major effect on the myclobutanil sensitivity of E. necator isolates investigated in this

317

study. Approximately half of the most resistant isolates from the eastern U.S. and all of

318

the most resistant isolates from Chile carry this mutation, whereas none of the most

319

sensitive isolates do. Interestingly, the other half of the most resistant isolates from the

320

eastern U.S. carry the synonymous mutation A1119C in CYP51. Isolates carrying this

321

mutation have EC50 values equal to those with Y136F (Fig. 1, Table 2).

322

The significant correlation of nucleotide substitution A1119C in CYP51 with

323

azole resistance cannot be explained by an alteration of the target site because this

324

mutation does not affect the amino acid sequence. However, isolates carrying this

15


Page 16 of 40

325

mutation overexpress CYP51 (Figs. 2 and 3), which is an important mechanism known to

326

contribute to azole resistance (27, 33). In other fungi, overexpression of CYP51 can be

327

caused by mutations in the promoter region. For example, a transposable element inserted

328

into the promoter of CYP51 in Monilinia fructicola alters gene expression (26). However,

329

we can only speculate on the cause of overexpression of CYP51 in E. necator because the

330

promoter sequence has not been isolated reproducibly (despite several attempts in our

331

laboratories). The simplest explanation for this correlation is that a mutation in the

332

promoter conferring overexpression originally occurred by coincidence in an individual

333

carrying the A1119C mutation, increased in frequency because of selection by fungicide

334

applications and later migrated widely throughout the eastern U.S. However, because the

335

two mutations are tightly linked (on the order of about 2 kbp), overexpression and

336

A1119C are in linkage disequilibrium. In this scenario, linkage disequilibrium is transient

337

and will decay over time, albeit very slowly because recombination must occur between

338

the promoter and nucleotide site 1119. In the long run, however, A1119C would

339

eventually become an unreliable marker for this type of myclobutanil resistance.

340

Alternatively, we speculate that this mutation could affect mRNA stability, for example

341

by altering secondary and/or tertiary structure, although we have no evidence to support

342

that hypothesis. Based on in silico analysis using the RNAz algorithm

343

(http://rna.tbi.univie.ac.at/RNAz), it appears that nucleotide 1119 is flanked by but is not

344

within a region of secondary structure, and the alleles described here do not affect the

345

predicted structure (data not shown). To assess the risk of resistance because of

346

overexpression of CYP51 more directly, mutations in the promoter will need to be

347

identified in future studies.

16

Page 17 of 40


348

We found no evidence for a role of efflux transporters in azole resistance in E.

349

necator. The expression of E. necator ABCG and MFS transporter homologs had no

350

significant effects in the regression analysis after accounting for the effects of mutations

351

Y136F and A1119C. In other fungi, ABC transporters classified as ABCG are known to

352

function as pleiotropic drug resistance (PDR) genes (23) and have been shown

353

experimentally to affect the efflux of azoles (8). However, we tested for sensitivity to

354

only a single azole fungicide and it is possible these transporters may confer a degree of

355

resistance to other members of the group. At the time this study was conducted, these

356

four homologs were the only ones we could find in the E. necator transcriptome. A

357

search in a more complete transcriptome sequence in the future would be needed to

358

determine whether other transporter genes have more effect on azole sensitivity. For

359

example, efflux transporters in subfamilies ABCB and ABCC are involved with drug

360

resistance, including to azoles, in other ascomycetes (23). Their role in azole resistance

361

should also be tested in future studies of E. necator for resistance to myclobutanil and

362

other azole fungicides.

363

We present correlative data for making inferences on the role of mutation Y136F

364

and overexpression of CYP51 as major mechanisms of azole resistance in eastern U.S.

365

populations of E. necator. These correlations, as well as the regression analyses to

366

evaluate the effects of efflux transporter gene expression, are justified by three factors.

367

First, the mechanisms we identified are well-known to contribute to azole resistance in

368

other fungi (8, 13, 24). Second, the experimental approaches available for studying model

369

fungi are not yet available for E. necator. In other fungi, mechanisms of azole resistance

370

have been inferred from segregation of genes (44), efflux studies (24) and gene-knockout

17


Page 18 of 40

371

experiments (8). None of these approaches is possible in E. necator yet. Third, because

372

we sampled from a diverse, recombining population in the eastern U.S. with very little

373

underlying genetic structure (5, 6, 19, 20) we could conduct an association study to infer

374

the underlying genetics of the azole sensitivity phenotypes (46). In our case, high levels

375

of resistance to myclobutanil were highly correlated with mutations Y136F and A1119C

376

in CYP51, but none of the residual variance was explained by expression of efflux

377

transporter genes.

378

We observed a bimodal distribution of azole-sensitivity phenotypes (EC50 values)

379

in samples of E. necator from the eastern U.S. and Chile (Fig. 1). All isolates with Y136F

380

and all but four with mutation A1119C have EC50 values greater than 0.18 µg/ml, and

381

therefore would be considered to be resistant to myclobutanil (18). Samples of E. necator

382

from the eastern U.S. were taken from a mix of commercial vineyards and wild vines

383

(Table S1), and therefore exposure to azole fungicide applications varied markedly. In

384

previous studies, populations exposed to azole applications were composed primarily of

385

individuals with EC50 values significantly higher than those from populations that were

386

not exposed (11, 18, 43). In our study, 22 of the 24 most resistant isolates came from

387

commercial vineyards. Although we cannot rule out heterogeneous selection as a factor

388

explaining the bimodal distribution, all isolates in the more resistant group (EC50 values

389

>0.1 µg/ml) had either the altered target site mutation Y136F or overexpression of CYP51

390

(Figs. 2 and 3). Moreover, isolates carrying Y136F also had significantly greater CYP51

391

expression than wild type, which may compensate for reduced CYP51 activity, as shown

392

for mutants in Mycosphaerella graminicola (2). Although mutation Y136F and

393

overexpression of CYP51 in E. necator explain a substantial portion of the variation in

18


Page 19 of 40

394

azole sensitivity, a considerable degree of variation was also found among isolates within

395

these groups. For example, EC50 values for isolates with mutation Y136F ranged from

396

0.21 to 3.54 µg/ml. This represents about a 17-fold range, comparable to the 18-fold

397

difference in the median EC50 values for the more-resistant versus sensitive groups. And

398

whereas the distributions of phenotypes within our sampled populations of E. necator

399

from the eastern U.S. and in Chile were bimodal, the resistant phenotypes did not exhibit

400

the much larger resistance factors typically associated with resistance to benzimidazole,

401

phenylamide, or QoI fungicides, caused by single mutations that have more extreme

402

effects on resistance phenotypes. Thus, the bimodal distribution of EC50 values in E.

403

necator is characterized by a relatively small quantitative difference between the two

404

modes, with overlapping distributions. By contrast, the pattern observed for these other

405

classes of fungicides is characterized by large, qualitative differences that can be easily

406

classified into two disjunct distributions of fungicide-sensitive and fungicide-resistant

407

isolates (22).

408

Haplotype diversity in CYP51 appears to be lower in the eastern U.S. compared to

409

Europe. We found three mutations in four haplotypes (Table 2) in addition to the

410

reference haplotype first described from France (15). Délye et al. (16) found three

411

additional mutations in this gene among a collection of isolates sampled from Europe,

412

India and Australia, and Miazzi and Hajjeh (29) found four more in Italy. Conversely,

413

populations of E. necator in the eastern U.S. are markedly more diverse for most other

414

genetic markers (5, 20). The CYP51 gene is under strong selection for azole resistance

415

because of fungicide applications in commercial vineyards, which might result in

416

selective sweeps for multiple alleles conferring resistance. Under this scenario, azole

19


Page 20 of 40

417

resistance could arise multiple times, each resulting from potentially different mutations.

418

This type of process has occurred in other grapevine pathogens. For example, identical

419

mutations conferring resistance to QoI fungicides arose in the grapevine downy mildew

420

pathogen, Plasmopara viticola, in two independent lineages (10).

421

Understanding the mechanism of azole resistance may have practical

422

implications, especially with regard to monitoring for azole resistance and cross

423

resistance among azoles (13). For example, surveys of E. necator in Virginia and Europe

424

have found isolates with high resistance to azoles that do not carry Y136F (11, 15, 29).

425

Therefore, the strategy of monitoring populations of E. necator for Y136F to assess the

426

risk of azole resistance (e.g., 17) will underestimate it where other resistance mechanisms

427

are operating, as is particularly true in the eastern U.S. Sensitivities to different azoles in

428

E. necator appears to be relatively well correlated (11, 18), suggesting that the two

429

dominant CYP51 mutations affecting myclobutanil resistance may have similar effects on

430

cross resistance. However, this assumption may need further testing. Long before the

431

mechanisms of azole resistance were known, Peever and Milgroom (31) found that cross

432

resistance relationships varied among different populations of Pyrenophora teres. They

433

speculated that different alleles for resistance, and therefore different mechanisms, were

434

conferring resistance in different populations. This hypothesis can now be tested in other

435

fungi such as E. necator.

436

In conclusion, it appears that monitoring populations of E. necator in the eastern

437

U.S. for azole resistance can be done relatively well by assaying for two SNPs, A495T

438

(Y136F) and A1119C. Similarly, monitoring populations in Chile for mutation Y136F

439

might be sufficient to predict azole resistance, although this suggestion is based on a

20


Page 21 of 40

440

small sample collected from a limited area in that country. However, a monitoring

441

strategy based on detection of specific SNPs in CYP51 will not detect any new mutations

442

that arise in the future. Additional studies on azole resistance need to be done to fully

443

understand the mechanisms involved.

444 445

Acknowledgements

446

This research was supported by Vaadia-BARD Postdoctoral Fellowship Award No. FI-

447

410-2008 from the United States – Israel Binational Agricultural Research and

448

Development Fund (BARD) to O.F., a USDA-Viticulture Consortium-East award to

449

L.C.-D. and W.F.W., and Hatch project NYC-153410 to M.G.M. We gratefully

450

acknowledge technical assistance from Judith Burr in conducting myclobutanil sensitivity

451

assays; Mickey Drott for statistical analyses; and Fernando Riveros of the Instituto de

452

Investigaciones Agropecuarias (INIA) in La Serena, Chile for supplying E. necator

453

isolates from that region.

454 455 456

Literature Cited 1.

Amrani, L., and Corio-Costet, M. F. 2006. A single nucleotide polymorphism in

457

the beta-tubulin gene distinguishing two genotypes of Erysiphe necator

458

expressing different symptoms on grapevine. Plant Pathol. 55:505-512.

459

2.

Bean, T. P., Cools, H. J., Lucas, J. A., Hawkins, N. D., Ward, J. L., Shaw, M. W.,

460

and Fraaije, B. A. 2009. Sterol content analysis suggests altered eburicol 14α-

461

demethylase (CYP51) activity in isolates of Mycosphaerella graminicola adapted

462

to azole fungicides. FEMS Microbiol. Lett. 296:266-273.

21

Page 22 of 40


463

3.

Becher, R., and Wirsel, S. R. 2012. Fungal cytochrome P450 sterol 14α-

464

demethylase (CYP51) and azole resistance in plant and human pathogens. Appl.

465

Microbiol. Biotech. 95:825-840.

466

4.

Brent, K. J., and Hollomon, D. W. 2007. Fungicide Resistance: The Assessment

467

of Risk, 2nd revised edition Fungicide Resistance Action Committee, Croplife

468

International, Brussels.

469

5.

Brewer, M. T., and Milgroom, M. G. 2010. Phylogeography and population

470

structure of the grape powdery mildew fungus, Erysiphe necator, from diverse

471

Vitis species. BMC Evol. Biol. 10:268.

472

6.

Brewer, M. T., Frenkel, O., and Milgroom, M. G. 2012. Linkage disequilibrium

473

and spatial aggregation of genotypes in sexually reproducing populations of

474

Erysiphe necator. Phytopathology 102:997-1005.

475

7.

Cadle-Davidson, L., Wakefield, L., Seem, R. C., and Gadoury, D. M. 2010.

476

Specific isolation of RNA from the grape powdery mildew pathogen Erysiphe

477

necator, an epiphytic, obligate parasite. J. Phytopathol. 158:69-71.

478

8.

Cannon, R. D., Lamping, E., Holmes, A. R., Niimi, K., Baret, P. V., Keniya, M.

479

V., Tanabe, K., Niimi, M., Goffeau, A., and Monk, B. C. 2009. Efflux- mediated

480

antifungal drug resistance. Clin. Microbiol. Rev. 22:291-321.

481

9.

482 483

Chaure, P., Gurr, S. J., and Spanu, P. 2000. Stable transformation of Erysiphe graminis, an obligate biotrophic pathogen of barley. Nat. Biotechnol. 18:205-207.

10.

Chen, W. J., Delmotte, F., Richard-Cervera, S., Douence, L., Greif, C., and Corio-

484

Costet, M. F. 2007. At least two origins of fungicide resistance in grapevine

485

downy mildew Populations. Appl. Environ. Microbiol. 73:5162-5172.

22

Page 23 of 40

486

11.

487


488

Colcol, J. F., Rallos, L. E., and Baudoin, A. B. 2011. Sensitivity of Erysiphe necator to demethylation inhibitor fungicides in Virginia. Plant Dis. 96:111-116.

12.

Cools, H. J., and Fraaije, B. A. 2013. Update on mechanisms of azole resistance

489

in Mycosphaerella graminicola and implications for future control. Pest

490

Management Science 69:150-155.

491

13.

492 493

of azole resistance in plant pathogenic fungi. Plant Pathol. 62:36-42. 14.

494 495

Cools, H. J., Hawkins, N. J., and Fraaije, B. A. 2013. Constraints on the evolution

Del Pozo, J. 2004. Historia del vino Chileno. 3rd edition ed. Editorial Universitaria, Santiago, Chile.

15.

Délye, C., Laigret, F., and Corio-Costet, M. F. 1997. A mutation in the 14-alpha-

496

demethylase gene of Uncinula necator that correlates with resistance to a sterol

497

biosynthesis inhibitor. Appl. Environ. Microbiol. 63:2966-2970.

498

16.

Délye, C., Ronchi, V., Laigret, F., and Corio-Costet, M. F. 1999. Nested allele-

499

specific PCR primers distinguish genetic groups of Uncinula necator. Appl.

500

Environ. Microbiol. 65:3950-3954.

501

17.

Dufour, M.-C., Fontaine, S., Montarry, J., and Corio-Costet, M.-F. 2011.

502

Assessment of fungicide resistance and pathogen diversity in Erysiphe necator

503

using quantitative real-time PCR assays. Pest Management Science 67:60-69.

504

18.

Erickson, E. O., and Wilcox, W. F. 1997. Distributions of sensitivities to three

505

sterol demethylation inhibitor fungicides among populations of Uncinula necator

506

sensitive and resistant to triadimefon. Phytopathology 87:784-791.

507 508

19.

Frenkel, O., Brewer, M. T., and Milgroom, M. G. 2010. Variation in pathogenicity and aggressiveness of Erysiphe necator from different Vitis species

23

Page 24 of 40

509

and geographic origins in the eastern United States. Phythopathology 100:1185-

510

1193.


511

20.

Frenkel, O., Portillo, I., Brewer, M. T., Péros, J.-P., Cadle-Davidson, L., and

512

Milgroom, M. G. 2012. Development of microsatellite markers from the

513

transcriptome of Erysiphe necator for analyzing population structure in North

514

America and Europe. Plant Pathol. 61:106-119.

515

21.

Gadoury, D. M., Cadle-Davidson, L., Wilcox, W. F., Dry, I., Seem, R. C., and

516

Milgroom, M. G. 2012. Grapevine powdery mildew (Erysiphe necator): A

517

fascinating system for the study of the biology, ecology, and the epidemiology of

518

an obligate biotroph. Mol. Plant Pathol. 1:1-16.

519

22.

520 521

inhibitors - a new challenge. Plant Dis. 71:1066-1074. 23.

522 523

Koeller, W., and Scheinpflug, H. 1987. Fungal resistance to sterol biosynthesis

Kovalchuk, A., and Driessen, A. 2010. Phylogenetic analysis of fungal ABC transporters. BMC Genomics 11:177.

24.

Kretschmer, M., Leroch, M., Mosbach, A., Walker, A. S., Fillinger, S., Mernke,

524

D., Schoonbeek, H. J., Pradier, J. M., Leroux, P., De Waard, M. A., and Hahn, M.

525

2009. Fungicide-driven evolution and molecular basis of multidrug resistance in

526

field populations of the grey mould fungus Botrytis cinerea. PLoS Path.

527

5:e1000696.

528

25.

Leroux, P., and Walker, A.-S. 2011. Multiple mechanisms account for resistance

529

to sterol 14α-demethylation inhibitors in field isolates of Mycosphaerella

530

graminicola. Pest Management Science 67:44-59.

24

Page 25 of 40


531

26.

Luo, C.-X., and Schnabel, G. 2008. The cytochrome P450 lanosterol 14α-

532

demethylase gene is a demethylation inhibitor fungicide fesistance determinant in

533

Monilinia fructicola field isolates from Georgia. Appl. Environ. Microbiol.

534

74:359-366.

535

27.

Ma, Z., Proffer, T. J., Jacobs, J. L., and Sundin, G. W. 2006. Overexpression of

536

the 14α-demethylase target gene (CYP51) mediates fungicide resistance in

537

Blumeriella jaapii. Appl. Environ. Microbiol. 72:2581-2585.

538

28.

Ma, Z. H., and Michailides, T. J. 2005. Advances in understanding molecular

539

mechanisms of fungicide resistance and molecular detection of resistant

540

genotypes in phytopathogenic fungi. Crop Prot. 24:853-863.

541

29.

Miazzi, M., and Hajjeh, H. R. 2011. Differential sensitivity to triadimenol of

542

Erysiphe necator isolates belonging to different genetic groups. J. Plant Pathol.

543

93:729-735.

544

30.

545 546

Palumbi, S. R. 2001. Humans as the world's greatest evolutionary force. Science 293:1786-1790.

31.

Peever, T. L., and Milgroom, M. G. 1993. Genetic correlations in resistance to

547

sterol biosynthesis-inhibiting fungicides in Pyrenophora teres. Phytopathology

548

83:1076-1082.

549

32.

550 551

Pfaffl, M. W. 2001. Relative quantification. Pages 63-82. in: Real-time PCR T. Dorak, ed. International University Line.

33.

Schnabel, G., and Jones, A. L. 2001. The 14 alpha-demethylase (CYP51A1) gene

552

is overexpressed in Venturia inaequalis strains resistant to myclobutanil.

553

Phytopathology 91:102-110.

25

Page 26 of 40


554

34.

Spanu, P. D., Abbott, J. C., Amselem, J., Burgis, T. A., Soanes, D. M., Stüber, K.,

555

van Themaat, E. V. L., Brown, J. K. M., Butcher, S. A., Gurr, S. J., Lebrun, M.-

556

H., Ridout, C. J., Schulze-Lefert, P., Talbot, N. J., Ahmadinejad, N., Ametz, C.,

557

Barton, G. R., Benjdia, M., Bidzinski, P., Bindschedler, L. V., Both, M., Brewer,

558

M. T., Cadle-Davidson, L., Cadle-Davidson, M. M., Collemare, J., Cramer, R.,

559

Frenkel, O., Godfrey, D., Harriman, J., Hoede, C., King, B. C., Klages, S.,

560

Kleemann, J., Knoll, D., Koti, P. S., Kreplak, J., López-Ruiz, F. J., Lu, X.,

561

Maekawa, T., Mahanil, S., Micali, C., Milgroom, M. G., Montana, G., Noir, S.,

562

O'Connell, R. J., Oberhaensli, S., Parlange, F., Pedersen, C., Quesneville, H.,

563

Reinhardt, R., Rott, M., Sacristán, S., Schmidt, S. M., Schön, M., Skamnioti, P.,

564

Sommer, H., Stephens, A., Takahara, H., Thordal-Christensen, H., Vigouroux,

565

M., Weßling, R., Wicker, T., and Panstruga, R. 2010. Genome expansion and

566

gene loss in powdery mildew fungi reveal tradeoffs in extreme parasitism.

567

Science 330:1543-1546.

568

35.

Stummer, B. E., and Scott, E. S. 2003. Detection of novel genotypes in progeny

569

from a controlled cross between isolates of Uncinula necator belonging to distinct

570

phenetic groups. Australasian Plant Pathology 32:213-218.

571

36.

Thompson, J. D., Higgins, D. G., and Gibson, T. J. 1994. CLUSTAL W:

572

improving the sensitivity of progressive multiple sequence alignment through

573

sequence weighting, position-specific gap penalties and weight matrix choice.

574

Nuc. Acids Res. 22:4673-4680.

26

Page 27 of 40


575

37.

Untergasser, A., Cutcutache, I., Koressaar, T., Ye, J., Faircloth, B. C., Remm, M.,

576

and Rozen, S. G. 2012. Primer3—new capabilities and interfaces. Nuc. Acids

577

Res. 40:e115.

578

38.

579 580

van den Bosch, F., and Gilligan, C. A. 2008. Models of fungicide resistance dynamics. Annu. Rev. Phytopathol. 46:123-147.

39.

van den Bosch, F., Oliver, R., van den Berg, F., and Paveley, N. 2014. Governing

581

principles can guide fungicide-resistance management tactics. Annu. Rev.

582

Phytopathol. 52:null.

583

40.

Verweij, P. E., Snelders, E., Kema, G. H. J., Mellado, E., and Melchers, W. J. G.

584

2009. Azole resistance in Aspergillus fumigatus: A side-effect of environmental

585

fungicide use? Lancet Infect. Dis. 9:789-795.

586

41.

Wakefield, L., Gadoury, D. M., Seem, R. C., Milgroom, M. G., Sun, Q., and

587

Cadle-Davidson, L. 2011. Differential gene expression during conidiation in the

588

grape powdery mildew pathogen, Erysiphe necator. Phytopathology 101:839-846.

589

42.

Wicker, T., Oberhaensli, S., Parlange, F., Buchmann, J. P., Shatalina, M., Roffler,

590

S., Ben-David, R., Dolezel, J., Simkova, H., Schulze-Lefert, P., Spanu, P. D.,

591

Bruggmann, R., Amselem, J., Quesneville, H., Ver Loren van Themaat, E., Paape,

592

T., Shimizu, K. K., and Keller, B. 2013. The wheat powdery mildew genome

593

shows the unique evolution of an obligate biotroph. Nat. Genet. 45:1092-1096.

594

43.

Wong, F. P., and Wilcox, W. F. 2002. Sensitivity to azoxystrobin among isolates

595

of Uncinula necator: Baseline distribution and relationship to myclobutanil

596

sensitivity. Plant Dis. 86:394-404.

27

Page 28 of 40


597

44.

Wyand, R. A., and Brown, J. K. M. 2005. Sequence variation in the CYP51 gene

598

of Blumeria graminis associated with resistance to sterol demethylase inhibiting

599

fungicides. Fungal Genet. Biol. 42:726-735.

600

45.

Ypema, H. L., Ypema, M., and Gubler, W. D. 1997. Sensitivity of Uncinula

601

necator to benomyl, triadimefon, myclobutanil, and fenarimol in California. Plant

602

Dis. 81:293-297.

603 604

46.

Zhu, C., Gore, M., Buckler, E. S., and Yu, J. 2008. Status and prospects of association mapping in plants. Plant Genome 1:5-20.

605 606

28

Page 29 of 40

607

Figure legends


608 609

Fig. 1. Distribution of myclobutanil sensitivity among isolates of Erysiphe necator,

610

expressed as log EC50 (µg/ml) in isolates from A, the eastern U.S., and B, Chile. Two

611

mutations in the CYP51 gene that correlate with reduced sensitivity to azole fungicides

612

are mapped onto the distribution, indicated by amino acid substitution Y136F (15) and a

613

synonymous nucleotide substitution A1119C. Isolates without either of these mutations

614

are considered as wild type, CYP51+.

615 616

Fig. 2. Correlation of log CYP51 expression with log EC50 for myclobutanil among

617

Erysiphe necator isolates carrying (i) the wild-type reference haplotype or mutation

618

A1170C alone (CYP51+); (ii) the Y136F amino acid substitution; or (iii) the nucleotide

619

substitution A1119C.

620 621

Fig. 3. Histogram of mean CYP51 expression for Erysiphe necator isolates carrying the

622

azole-sensitive wild-type reference haplotype or mutation A1170C alone (CYP51+), or

623

mutations Y136F and A1119C. Expression data were normalized by the expression of the

624

constitutively expressed reference genes actin, β-tubulin and histone H3 relative to the

625

average of three sensitive, wild-collected control isolates, as detailed in the Methods.

626

Significant differences were found between all pairs of mutations by Tukey’s Honest

627

Significant differences (HSD) test. Different letters above each bar denotes the significant

628

difference at P < 0.05.

629

29

Page 30 of 40

630

Table 1. Primers used for sequencing CYP51 and quantitative reverse-transcription (RT)-

631

PCR.


Use

Gene

Primer name

Primer sequence 5’→3’

Amplicon size (bp)

Sequencing CYP51

EnCYP89F

ATGCATGGGTCCAAGAATTT

EnCYP856R

ACCCTGAAAGCTAACGCAAT

EnCYP563F

GATTCGATCTGAACTGCTGGT

EnCYP1162R

TCCGAAGTTGTTCCTCATACAA

EnCYP1055F

TTCTGATGGCTGGACAACAC

EnCYP1752R

AACCCTAACACCTGCCATAAA

EnRTCYP51F

GCGTTAGCTTTCAGGGTGAG

EnRTCYP51R

CATGATCACGAGCACGATTC

En5176F

TGGAATTGGCTCAAAGGAAC

767a

(nt 1-652) a

CYP51

599

(nt 563-1162) a

CYP51

697

(nt 1055-1752) a

Quantitative RT-PCR CYP51

ABCG-1

30

219

201

Page 31 of 40


ABCG-2

ABCG-3

MSF1

β-tubulin b

actin b

histone H3 b

En5176R

AATCCCCCAACTTTTTGTCC

En9796F

AAATGGCGAAAAGGTCAATG

En9796R

TTTGCTGCTAGGCCCATATC

En14377F

GTACCAACGCCTTCTGGAAA

En14377R

CGCAAAGATCTTGTTGCTCA

EnMFSF

AACGCACCAGAAAAACAACC

EnMFSR

TACGTATCTGCTCGCCATTG

EnBTUBF

TCACAACCTTCAGCTTCACG

EnBTUBR

ATATGTTCCTCGTGCCGTTC

EnACTINR

CTCATGCAATTGCTCGTGTT

EnACTINF

AGATGACTGGCTCGCTGTCT

EnHIS3F

TAACACGCTTGGCATGGATA

EnHIS3R

TGGAACAGTTGCTCTTCGTG

632

a

633

EF649776.1. The forward primer for the first segment of the coding region (EnCYP89F)

Nucleotide sites are relative to the reference CYP51 sequence in GenBank accession

31

150

235

172

209

188

184


Page 32 of 40

634

was designed using GenBank accession U72657 (15) and includes 204 bp upstream from

635

the coding region (see text).

636

b

637

histone H3 were designed by Wakefield et al. (41).

Genes used for normalizing expression data (see Materials and Methods). Primers for

32


Page 33 of 40

638

Table 2. Frequency of CYP51 haplotypes in populations of Erysiphe necator in the eastern United States.

639 Nucleotide sites a Haplotype

N

Deposited accession number

Mean

Mean EC50

log(EC50)

495

1119

1170

CYP51+

40

KM077178

-1.469+0.048

0.041+0.004

A

A

A

CYP51A1170C

2

KM077180

-1.699+0.301

0.025+0.015

A

A

C

CYP51A495T/A1170C

9

KM077179

-0.242+0.098

0.700+0.158

Tb

A

C

CYP51A1119C/A1170C

9

KM077181

-0.453+0.134

0.438+0.121

A

C

C

CYP51A1119C

6

KM077182

-0.385+0.143

0.587+0.149

A

C

A

640 641

a

642

U.S. wild-type CYP51+ by nucleotide substitution C608T.

643

b

644

codon 136, Y136F. Nucleotide substitutions found at positions 1119 and 1170 in the U.S. are synonymous mutations in codons 326

645

and 343, respectively.

Nucleotide sites are relative to the reference CYP51 sequence in GenBank accession EF649776.1, which differs from the eastern

Nucleotide substitution from A to T in position 495 causes an amino acid substitution from tyrosine (Y) to phenylalanine (F) at

33

Page 34 of 40

646

Table 3. Homologs of efflux transporter genes found in the transcriptome sequence of

647

Erysiphe necator using a tBLASTn search.


648 Gene

GenBank accession E. necator

E. necator

used as query

GenBank

contig

E-value

accession ABCG-1

CAA54692.1

En5176

KM217169

2e-17

En7679

KM217168

3e-22

En14377

KM217167

2e-118

En3478

KM217170

2e-62

CDR1 (Candida albicans) ABCG-2

BAA93677.1 BMR1 (Botrytis cinerea)

ABCG-3

BAA93677.1 BMR1 (B. cinerea)

MFS1

AAF64435.2 Botrytis MFS transporter Bcmfs1 (B. cinerea)

649 650

34


Page 35 of 40

Fig. 1. Distribution of myclobutanil sensitivity among isolates of Erysiphe necator, expressed as log EC50 (µg/ml) in isolates from A, the eastern U.S., and B, Chile. Two mutations in the CYP51 gene that correlate with reduced sensitivity to azole fungicides are mapped onto the distribution, indicated by amino acid substitution Y136F (15) and a synonymous nucleotide substitution A1119C. Isolates without either of these mutations are considered as wild type, CYP51+. 190x254mm (96 x 96 DPI)


Page 36 of 40

Fig. 2. Correlation of log CYP51 expression with log EC50 for myclobutanil among Erysiphe necator isolates carrying (i) the wild-type reference haplotype or mutation A1170C alone (CYP51+); (ii) the Y136F amino acid substitution; or (iii) the nucleotide substitution A1119C. 254x338mm (72 x 72 DPI)


Page 37 of 40

Fig. 3. Histogram of mean CYP51 expression for Erysiphe necator isolates carrying the azole-sensitive wildtype reference haplotype or mutation A1170C alone (CYP51+), or mutations Y136F and A1119C. Expression data were normalized by the expression of the constitutively expressed reference genes actin, β-tubulin and histone H3 relative to the average of three sensitive, wild-collected control isolates, as detailed in the Methods. Significant differences were found between all pairs of mutations by Tukey’s Honest Significant differences (HSD) test. Different letters above each bar denotes the significant difference at P < 0.05. 190x254mm (96 x 96 DPI)

Page 38 of 40

Supplementary Table S1. Isolates of Erysiphe necator, sampling locations, hosts of origin, CYP51 haplotypes and sensitivity to myclobutanil (EC50).


Isolate

Region

Eastern U.S. isolates from wild Vitis species b Portsmouth NE NCaes9

SE

Location

Vitis species/host of origin

CYP51 Haplotype a

EC50

Portsmouth, NH

V. labrusca

CYP51+

0.007

-2.173

CYP51

+

0.010

-2.000

+

Blue Ridge NY

V. aestivalis

Log EC50

Kan1

C

Lawrence, KS

V, riparia

CYP51

0.010

-2.000

PCTHT

SE

Panther creek, GA

V. rotundifolia

CYP51+

0.020

-1.703

V. rotundifolia

CYP51

+

0.020

-1.702

CYP51

+

0.020

-1.699

+

PCTH Mas4

SE NE

Panther creek, GA Lexington, MA

V. labrusca

Ith4

NE

Ulysses, NY

V. riparia

CYP51

0.020

-1.699

NClab1

SE

V. labrusca

CYP51+

0.020

-1.699

NCaes6

SE

V. aestivalis

CYP51+

0.030

-1.523

Ith8

NE

Brevard, NC Pisgah Natl. Forest, NC Lansing, NY

V. riparia

CYP51+

0.030

-1.523

+

Highland

NE

Highland, NY

V. riparia

CYP51

0.036

-1.449

NCaes7

SE

Blue Ridge NC

V. aestivalis

CYP51+

0.038

-1.417

V. aestivalis

CYP51

+

0.040

-1.398

CYP51

+

0.040

-1.398

V. aestivalis

CYP51

A1170C

0.040

-1.398

V. riparia

CYP51+

0.040

-1.398

V. labrusca

CYP51+

0.042

-1.380

V. aestivalis

CYP51+

0.042

-1.374

NCaes8d Watertown

SE NE

Blue Ridge NC

EBrunNJ

NE

Watertown, NY Blood Mountain, GA Wellesley Island, NY Blue Ridge, NC Pisgah Natl. Forest, NC E. Brunswick, NJ

V. labrusca

CYP51

0.048

-1.321

Ith6

NE

Ithaca, NY

V. riparia

CYP51+

0.050

-1.301

W12

SE

Sugar Grove VA

V. aestivalis

CYP51+

0.060

-1.222

CYP51

+

0.060

-1.222

CYP51

+

0.060

-1.220

CYP51

+

0.064

-1.193

CYP51

+

0.066

-1.183

CYP51

+

0.088

-1.057

A1119C

BlMnt2

SE

Wellesley

NE

NClab4

SE

NCaes3a

SE

NClab5 BlMntS2 NCrip1 PCF2 NCaes5d

SE SE SE SE SE

Cashier, NC Blood Mountain, GA North Haven, NC Panther creek, GA Blue Ridge NC

V. riparia

V. labrusca V. aestivalis V. riparia V. rotundifolia V. aestivalis

+

WeG1

NE

Watkins Glen, NY

V. aestivalis

CYP51

0.670

-0.174

WEG4

NE

Watkins Glen, NY

V. riparia

CYP51A1119C

1.370

0.137

Blairsville, GA

V. vinifera ‘Merlot’

CYP51+

0.010

-2.000

V. vinifera ‘Merlot’

+

0.010

-2.000

Eastern U.S. isolates from cultivated vineyards GAmer1 SE VA2

SE

Boone, NC

CYP51

Page 39 of 40


VA36

Boone, NC

TXCS1

SE

Comaneche, TX

Gamer2

SE

Blairsville, GA

GAcmus

SE

Ringgold, GA

Raylen2

SE

Mocksville, NC

RoAcl2

SE

Hurdle Mills, NC

Gamer3

SE

ANYP

NE

KsFran

C

RoAwmus3

SE

Blairsville, GA Athens, NY Emporia, KS Hurdle Mills, NC

BiltCY2

SE

Ashville, NC

LNYN

NE

Lockport, NY

LW5

NE

Watkins Glen, NY

RHNCS3

SE

Tyron, NC

G14

NE

NY63

NE

Hector, NY

VA31

NE

Boone, NC

RoACS

SE

Hurdle Mills, NC

Ny19

NE

Hector, NY

WvMont

C

NY90

NE

Hector, NY

LW1

NE

Geneva, NY

MizeCS1

SE

FCon2

NE

Geneva, NY

NY72

NE

Hector, NY

Geneva, NY

Waverly, MO

Columbus, NC

VA67

NE

SHNC1

SE

HVLCY

NE

Boone, NC Pittsboro, NC

Highland, NY

V. vinifera ‘Merlot’ V. vinifera ‘Cabernet Sauvignon’ V. vinifera ‘Merlot’ V. rotundifolia ‘Carlos’ V. vinifera ‘Cabernet Sauvignon’ V. vinifera\ ‘Chardonnay V. vinifera ‘Merlot’ Vitis interspecific hybrid ‘Pinard’ Vitis interspecific hybrid ‘Frontenac’ V. rotundifolia V. vinifera ‘Chardonnay’ V. vinifera ‘Chardonnay’ unknown V. vinifera ‘Chardonnay’ Vitis interspecific hybrid ‘Rosette’ V. vinifera ‘Chardonnay V. vinifera ‘Merlot’ V. vinifera ‘Cabernet Sauvignon’ V. vinifera ‘Chardonnay’ Vitis interspecific hybrid ‘Norton’ V. vinifera ‘Chardonnay Vitis labrusca ‘Niagara’ V. vinifera ‘Cabernet Sauvignon’ Vitis x labrusca ‘Concord’ V. vinifera ‘Chardonnay V. vinifera ‘Merlot’ Vitis interspecific hybrid ‘Chambourcin’ V. vinifera ‘Chardonnay’

CYP51+

0.010

-2.000

CYP51A1170C

0.010

-2.000

CYP51+

0.030

-1.523

+

0.034

-1.465

CYP51+

0.044

-1.359

CYP51+

0.045

-1.342

+

0.048

-1.316

CYP51+

0.050

-1.301

CYP51+

0.064

-1.196

CYP51+

0.068

-1.166

+

0.064

-1.196

CYP51+

0.076

-1.118

CYP51+

0.086

-1.063

CYP51A1119C

0.090

-1.046

CYP51A1119C/A1170C

0.110

-0.959

CYP51A1119C

0.130

-0.886

CYP51A1119C

0.150

-0.824

CYP51A495T/A1170C

0.210

-0.678

CYP51A1119C/A1170C

0.260

-0.585

CYP51A495T/A1170C

0.300

-0.523

CYP51A1119C/A1170C

0.310

-0.509

CYP51A1119C/A1170C

0.330

-0.481

CYP51A495T/A1170C

0.340

-0.471

CYP51A495T/A1170C

0.420

-0.377

CYP51A1119C

0.460

-0.337

A1119C

0.590

-0.229

CYP51A1119C

0.670

-0.174

CYP51A495T/A1170C

0.700

-0.155

CYP51

CYP51

CYP51

CYP51


Page 40 of 40

SJMont

SE

Blue Ridge NC

LNYM

NE

Lockport, NY

Dresden2

NE

Dresden, NY

G9

NE

VA94

NE

Boone, NC

LICY

NE

Riverhead, NY

PUMOChn

C

Geneva, NY

Purdy, MO

V. vinifera hybrid ‘Noiret’ V. vinifera ‘Merlot’ V. vinifera ‘Chardonnay’ Vitis interspecific hybrid ‘Rosette’ V. vinifera ‘Merlot’ V. vinifera ‘Chardonnay’ Vitis interspecific hybrid ‘Chambourcin’

CYP51A495T/A1170C

0.710

-0.149

CYP51A495T/A1170C

0.720

-0.143

CYP51A1119C/A1170C

0.810

-0.092

CYP51A1119C/A1170C

0.810

-0.092

CYP51A1119C

1.160

0.064

CYP51A495T/A1170C

1.270

0.104

CYP51A495T/A1170C

1.630

0.212

V. vinifera

CYP51C608T

0.012

-1.906

V. vinifera

CYP51

C608T

0.022

-1.659

CYP51

C608T

0.024

-1.617

CYP51

C608T

0.036

-1.439

C608T

0.050

-1.296

Chilean isolates from cultivated vineyards CH19 CH32 CH36 CH18 CH31

IV IV IV IV

La Serena La Serena La Serena La Serena

V. vinifera V. vinifera

IV

La Serena

V. vinifera

CYP51

-1.221

CH2 CH33

IV

La Serena

V. vinifera

CYP51C608T

0.060

IV

La Serena

V. vinifera

CYP51C608T

0.094

-1.208

CH8

IV

La Serena

V. vinifera

CYP51A495T/C608T

1.349

0.130

V. vinifera

CYP51

A495T/C608T

1.648

0.217

CYP51

A495T/C608T

2.164

0.335

CYP51

A495T/C608T

3.200

0.505

CYP51

A495T/C608T

3.543

0.549

CH13 CH10 CH1-10 CH15

IV IV IV IV

La Serena La Serena La Serena La Serena

V. vinifera V. vinifera V. vinifera

a

CYP51 haplotypes are denoted by mutations they have relative to the wild-type haplotype (denoted ‘+’) in GenBank accession KM077178. b Eastern U.S. isolates were described previously (1-3). Literature Cited 1.

2.

3.

Brewer, M. T., and Milgroom, M. G. 2010. Phylogeography and population structure of the grape powdery mildew fungus, Erysiphe necator, from diverse Vitis species. BMC Evol. Biol. 10:268. Frenkel, O., Brewer, M. T., and Milgroom, M. G. 2010. Variation in pathogenicity and aggressiveness of Erysiphe necator from different Vitis species and geographic origins in the eastern United States. Phythopathology 100:1185-1193. Frenkel, O., Portillo, I., Brewer, M. T., Péros, J.-P., Cadle-Davidson, L., and Milgroom, M. G. 2012. Development of microsatellite markers from the transcriptome of Erysiphe necator for analyzing population structure in North America and Europe. Plant Pathol. 61:106-119.

RUN1 and REN1 Pyramiding in Grapevine (Vitis vinifera cv. Crimson Seedless) Displays an Improved Defense Response Leading to Enhanced Resistance to Powdery Mildew (Erysiphe necator).

Identification and utilization of a new Erysiphe necator isolate NAFU1 to quickly evaluate powdery mildew resistance in wild Chinese grapevine species using detached leaves.

Adaptive genomic structural variation in the grape powdery mildew pathogen, Erysiphe necator.

Effects of Bunch Rot (Botrytis cinerea) and Powdery Mildew (Erysiphe necator) Fungal Diseases on Wine Aroma.

Linkage analysis of er-1, a recessive Pisum sativum gene for resistance to powdery mildew fungus (Erysiphe pisi D.C.).

CYP83A1 is required for metabolic compatibility of Arabidopsis with the adapted powdery mildew fungus Erysiphe cruciferarum.

Quantification of the Changes in Potent Wine Odorants as Induced by Bunch Rot (Botrytis cinerea) and Powdery Mildew (Erysiphe necator).

Inheritance of powdery mildew (Erysiphe polygoni DC) resistance in mungbean (Vigna radiata L. Wilczek).

Genetic dissection of powdery mildew resistance in interspecific half-sib grapevine families using SNP-based maps.

RFLP mapping of three new loci for resistance genes to powdery mildew (Erysiphe graminis f. sp. hordei) in barley.

Inheritance of partial resistance to powdery mildew in spring wheat.

Knockdown of MLO genes reduces susceptibility to powdery mildew in grapevine.

Mapping oligogenic resistance to powdery mildew in mungbean with RFLPs.

Molecular marker analyses of powdery mildew resistance in barley.

Does farm fungicide use induce azole resistance in Aspergillus fumigatus?

Tempered mlo broad-spectrum resistance to barley powdery mildew in an Ethiopian landrace.

Nuclear Function of Subclass I Actin-Depolymerizing Factor Contributes to Susceptibility in Arabidopsis to an Adapted Powdery Mildew Fungus.

Oak powdery mildew (Erysiphe alphitoides)-induced volatile emissions scale with the degree of infection in Quercus robur.

Metrafenone resistance in a population of Erysiphe necator in northern Italy.

RFLP markers to identify the alleles on the Mla locus conferring powdery mildew resistance in barley.

Field evaluation of an expertise-based formal decision system for fungicide management of grapevine downy and powdery mildews.

Transcriptional Reprogramming of the Mycoparasitic Fungus Ampelomyces quisqualis During the Powdery Mildew Host-Induced Germination.

Erratum to: Proteomic Analysis of the Defense Response of Wheat to the Powdery Mildew Fungus, Blumeria graminis f. sp. tritici.

Genetic analysis of the obligate parasitic barley powdery mildew fungus based on RFLP and virulence loci.