sntoxa, sntox1 and sntox3 originated in parastagonospora ...22 crescent, a region in the middle east...
TRANSCRIPT
SnToxA, SnTox1 and SnTox3 originated in Parastagonospora nodorum in the Fertile 1
Crescent 2
3
Fariba Ghaderi1, Bahram Sharifnabi1, Mohammad Javan-Nikkhah2, Patrick C. Brunner3, Bruce 4
A. McDonald3 5 6 1Department of Plant Protection, Isfahan University of Technology, Isfahan, Iran 7 2Department of Plant Protection, College of Agriculture and Natural Resources, University of 8
Tehran, Karaj, Iran 9 3Plant Pathology Group, Institute of Integrative Biology, ETH Zurich/LFW, Universitätstrasse 2, 10
CH-8092 Zurich, Switzerland 11 12
Correspondence to: Bruce A. McDonald, Plant Pathology Group, Institute of Integrative 13
Biology, ETH Zurich/LFW, Universitätstrasse 2, CH-8092 Zurich, Switzerland. E-mail: 14
16
17
ABSTRACT 18
19
The center of origin of the globally distributed wheat pathogen Parastagnospora 20
nodorum has remained uncertain because only a small number of isolates from the Fertile 21
Crescent, a region in the Middle East where wheat was domesticated from wild grasses, were 22
included in earlier population genetic and phylogeographic studies. We isolated and genetically 23
analyzed 193 P. nodorum strains from three naturally infected wheat fields distributed across 24
Iran, a country located within the Fertile Crescent, using eleven neutral microsatellite loci. 25
Compared to previous studies that included populations from North America, Europe, Africa, 26
Australia and China, the populations from Iran had the highest genetic diversity globally and also 27
exhibited greater population structure over smaller spatial scales, patterns typically associated 28
with a species' center of origin. Genes encoding the necrotrophic effectors SnToxA, SnTox1 and 29
SnTox3 were found at a high frequency in the Iranian population. By sequencing 96 randomly 30
chosen Iranian strains, we detected new alleles for all three effector genes. Analyses of allele 31
diversity showed that all three effector genes had higher diversity in Iran than in any population 32
included in previous studies, with Iran acting as a hub for the effector diversity that was found in 33
other global populations. Taken together, these findings support the hypothesis that P. nodorum 34
originated either within or nearby the Fertile Crescent with a genome that already encoded all 35
three necrotrophic effectors during its emergence as a pathogen on wheat. Our findings also 36
suggest that P. nodorum was the original source of the ToxA genes discovered in the wheat 37
pathogens Phaeosphaeria avenaria f. sp. tritici 1, Pyrenophora tritici-repentis and Bipolaris 38
sorokiniana. 39
40
Keywords: microsatellites, SSRs, necrotrophic effectors, Stagonospora nodorum, population 41
genetics, center of origin 42
43
INTRODUCTION 44
45
Parastagonospora nodorum (syn. Phaeosphaeria nodorum) is the causal agent of 46
Stagonospora nodorum leaf and glume blotch (SNB) on durum and bread wheat (Quaedvlieg et 47
al. 2013). This disease is found in most wheat-growing regions of the world (Wiese, 1987) and 48
can cause yield losses of up to 31% (Bhathal et al. 2003). The pathogen infects mainly leaves 49
and ears, reducing both grain quality and yield (Eyal 1987, 1999). 50
The genetic structure of P. nodorum populations has been analyzed at field, regional, 51
continental, and global scales using several types of neutral genetic markers, including restriction 52
fragment length polymorphisms (RFLPs) (McDonald et al. 1994; Keller et al. 1997a, 1997b), 53
amplified fragment length polymorphisms (AFLPs) (Bennett et al. 2005) and microsatellites 54
(also called simple sequence repeats or SSRs) (Stukenbrock et al. 2005). Populations of P. 55
nodorum exhibited high levels of genetic diversity in North America, Europe, Africa, Australia 56
and China. Migration rates between continents were high, resulting in a shallow population 57
structure even on continental and global scales (Stukenbrock et al. 2006). 58
Based on the findings of higher private allelic richness at eight microsatellite loci and a 59
higher number of private multilocus haplotypes across four sequence loci in Iran compared to 60
global populations, it was hypothesized that P. nodorum's center of origin coincides with its 61
wheat host in the Fertile Crescent (McDonald et al. 2012). However, the origin of P. nodorum 62
remained uncertain because only 24 strains from the Fertile Crescent region were analyzed 63
previously. McDonald et al. (2013) examined the evolutionary histories of three necrotrophic 64
effectors (NEs) in P. nodorum encoded by the genes SnToxA, SnTox1, and SnTox3. Contrary to 65
expectations, the 24 Iranian isolates did not exhibit the highest genetic diversity for any of these 66
NEs. Instead, the highest diversity for SnToxA based on rarefaction analyses and the number of 67
private alleles was observed in South Africa while the highest diversity for SnTox1 was in 68
Europe and the highest diversity for SnTox3 was in North America. These findings, coupled with 69
the absence of all three NE-encoding genes in all but one close relative of P. nodorum called 70
Phaeosphaeria avenaria f. sp. tritici 1 (Pat1), led to the hypothesis that P. nodorum acquired 71
these NEs through three independent horizontal gene transfers (McDonald et al. 2013). 72
Here we report new findings based on analyses of 193 new P. nodorum strains sampled 73
from three naturally infected wheat fields located in different regions of Iran. We aimed to 74
specifically test the hypothesis that the Fertile Crescent, represented by Iran, is the center of 75
origin of P. nodorum by analyzing population genetic structure using the same neutral SSR loci 76
that were used previously to analyze globally distributed P. nodorum populations. Measures of 77
genotype diversity, disequilibrium and mating type frequencies were combined to determine the 78
importance of sexual recombination in these Iranian populations. We also analyzed sequence 79
diversity for SnToxA, SnTox1 and SnTox3 to test the hypothesis that all three genes were 80
acquired by P. nodorum through independent horizontal gene transfers. Our findings support an 81
origin for P. nodorum in the Fertile Crescent and suggest that all three necrotrophic effectors 82
emerged in P. nodorum populations at their center of origin, contradicting the earlier hypothesis 83
of independent origins through a series of horizontal gene transfers after the pathogen moved to 84
different continents. These new data strengthen the hypothesis that P. nodorum was the ultimate 85
source of the ToxA genes horizontally acquired by Pat1, Pyrenophora tritici-repentis and 86
Bipolaris sorokiniana. 87
88 MATERIALS AND METHODS 89
90
Comparisons with earlier studies. To make comparisons with previous P. nodorum 91
studies as compatible as possible, we conducted the same analyses using the same software 92
whenever possible. To increase the overall sample size and to obtain a more detailed portrait of 93
P. nodorum population structure in Iran, we included the data set of the previously described 94
population from the northern Golestan province (McDonald et al. 2012) in our microsatellite 95
analyses. To account for possible differences in binning of SSR alleles between the two studies, 96
randomly chosen Golestan isolates were PCR-amplified and included in the same runs as the 97
new Iranian isolates. 98
Earlier publications that analyzed the same mating types, SSR loci and necrotrophic 99
effectors included 693 strains from 17 wheat fields located in Australia (n=73), China (n=101), 100
Europe (n=301), Mexico (n=31), North America (n=132), and South Africa (n=55) (Stukenbrock 101
et al. 2006), as well as 57 strains from two wheat fields sampled in 2005 and 2010 in the 102
Golestan Province in Iran (McDonald et al, 2012). Most of these wheat fields were sampled 103
using the same hierarchical transect sampling method that was used for the new collections from 104
Iran reported here. Iran is considered a proxy for the Fertile Crescent because it is the only 105
country from this region where field populations were analyzed. Among the 57 Iranian isolates 106
analyzed earlier, more than half were shown to be closely related to P. nodorum (providing 107
additional evidence for a center of origin in the Fertile Crescent, as described in McDonald et al., 108
2012), but 29 of these isolates were P. nodorum, with complete SSR and mating-type datasets 109
obtained from the 24 Golestan strains included in the new analyses described here. 110
Sampling and DNA extraction. A total of 193 new P. nodorum isolates were made from 111
infected leaves and ears collected from three naturally infected wheat fields representing the 112
major wheat-growing areas of southern Iran in the Kohgiluyeh, Khuzestan and Fars provinces 113
(Figure 1). Including the Golestan population, these Iranian wheat fields were separated by 250 - 114
800 km and differed with regard to climate, wheat cultivars and wheat-growing seasons. Isolates 115
were obtained using hierarchical sampling (McDonald et al. 1995) from six to eight spots 116
separated by 10 m within each field. Isolates from Kogiluyeh and Khuzestan were collected from 117
infected leaves while isolates from Fars were collected from infected ears. Only one isolate was 118
collected from each plant. Single-spore isolation and other culturing procedures were performed 119
as described by Halama and Lacoste (1991). Pure cultures of each isolate were stored on 120
lyophilized filter paper strips at -80°C. 121
122
123
124
125
126
Figure 1. Sampling locations for Parastagonospora nodorum populations. A. The world map 127
shows the locations of global populations described in earlier studies (Stukenbrock et al. 2006; 128
McDonald et al. 2012). B. The map of Iran shows the newly sampled locations of Khuzestan, 129
Kohgiluyeh and Fars that form the new Iranian data set as well as the earlier described 130
population of Golestan (McDonald et al. 2012) that forms the old Iranian data set. 131
132
133
Isolates were grown on Petri dishes containing yeast sucrose agar (YSA, 10g/L yeast 134
extract, 10g/L sucrose, 1.2% agar) amended with 50 μg of kanamycin. Single colonies were 135
transferred to flasks containing 50 ml yeast sucrose broth (YSB, 10g/L yeast extract, 10g/L 136
sucrose) and grown on an orbital shaker for 5 to 7 days at 120 rpm and 18°C. Genomic DNA was 137
extracted as described previously (Murray and Thompson, 1980). 138
Mating type determination. Mating type idiomorphs for each isolate were determined 139
using the mating type primers described previously (Bennett et al. 2003). These primers 140
amplified a 510 bp PCR product for MAT1-2 isolates and a 360 bp PCR product for MAT1-1. 141
Multiplex PCR amplifications were performed as described previously (Sommerhalder et al. 142
Global RegionsAF - AfricaAS - AsiaAU - AustraliaCD - CanadaCA - Central AsiaEU - EuropeLA - Latin AmericaNA - North America
35°N
30°N
25°N
45°E45°E 50°E 55°E 60°E
CaspianSea
PersianGulf
Gulf of Oman
40°N
AF
AS
AU
CAEUCD
NA
LA
Iranian PopulationsFA - FarsGO - GolestanKZ - KhuzestanKG - Kohgiluyeh
FA
GO
KZ KG
Iran
A
B
2006). We assessed the ratio of MAT1-1 to MAT1-2 alleles and tested for deviations from a 1:1 143
ratio of the two mating types in each field using χ 2 statistics. 144
Microsatellite analysis. Eleven previously described SSR loci, SNOD1, SNOD3, 145
SNOD5, SNOD8, SNOD11, SNOD15, SNOD16, SNOD17, SNOD21, SNOD22 and SNOD23 146
(Stukenbrock et al. 2005) were amplified in each isolate. PCR amplifications were performed in 147
20 μl reactions containing 0.05 μM of each primer (Microsynth, Balgach Switzerland), 1X 148
Dream Taq Buffer (MBI Fermentas), 0.4 μM dNTPs (MBI Fermentas) and 0.5 units of Dream 149
Taq DNA polymerase (MBI Fermentas). The PCR cycle parameters were: 2 min initial 150
denaturation at 96°C followed by 35 cycles at 96°C for 30 sec, annealing at 56°C for 45 sec, and 151
extension at 72°C for 1 min. A final 7 min extension was made at 72°C. Amplicons were 152
separated in a 3730xl ABI Genetic Analyzer capillary sequencer (Life Technologies, Applied 153
Biosystems). The software Genemapper (Life Technologies, Applied Biosystems) was used for 154
genotyping. 155
Population genetic analyses. Genetic variation was quantified using measures of gene 156
and genotype diversity at the SSR loci. A clone-corrected data set containing a single 157
representative of each multilocus haplotype based on the program Genodive 2.0 (Meirmans and 158
Van Tienderen 2004) was used for subsequent analyses. Genetic diversity for each locus and for 159
each population across all loci was assessed using the program POPGENE32 (Yeh et al. 1999). 160
Genotype diversity and clonal fractions for all populations were calculated according to Stoddart 161
and Taylor (1988) as implemented in the R package poppr (Kamvar et al. 2014). The percentage 162
of maximum possible diversity (G/N) was calculated by dividing the genotypic diversity by the 163
number of isolates (Chen et al. 1994). Allelic richness as a measure of within population genetic 164
diversity was calculated using the program Fstat (Goudet 2001). Measures of recombination 165
including the indices of association IA and rd were estimated with the program MULTILOCUS 166
version 1.2.2 using 103 simulations (Agapow and Burt 2001). 167
Pairwise genetic differentiation among populations was estimated as RST using 168
ARLEQUIN 3.5 (Excoffier and Lischer 2010). In contrast to measures of FST that are based on 169
infinite allele models, RST considers stepwise mutations to better model the evolutionary 170
relatedness of microsatellites by counting the sum of the squared number of repeat differences 171
between two haplotypes (Weir and Cockerham 1984; Michalakis and Excoffier 1996). P-values 172
were obtained using 103 permutations. 173
We analyzed the population structure of P. nodorum using STRUCTURE (Pritchard et al. 174
2000; Falush et al. 2003). The number of populations tested ranged from K = 1 to K = 7. 175
STRUCTURE runs were performed using the admixture model, sampling locations as priors, 106 176
iterations and a burn-in period of 30,000. Ten independent simulations were performed. 177
CLUMPAK (Kopelman et al. 2015) was used for summation and graphical representation of the 178
STRUCTURE runs. The implemented deltaK approach of Evanno et al. (2005) was used to 179
identify the optimal number of populations in the data set. 180
Sequence analyses of necrotrophic effectors. All strains were screened for the presence 181
of SnToxA, SnTox1, and SnTox3 using PCR assays. PCR amplifications were carried out in 20 182
μL reaction mixtures including 0.04 μM of each primer (Microsynth, Balgach Switzerland), 1X 183
Dream Taq Buffer (MBI Fermentas), 0.4μM dNTPs (MBI Fermentas) and 0.4 units of Dream 184
Taq DNA polymerase (MBI Fermentas). The PCR cycle parameters were: initial denaturation of 185
2 min at 96°C, 36 cycles of denaturation at 96°C for 40 sec, annealing temperature specific for 186
each primer for 45 sec, extension at 72°C for 1 min followed by a final extension at 72°C for 7 187
min. PCR products were purified to remove unincorporated nucleotides and primers using 188
NucleoFast® 96 PCR plates (Macherey-Nagel, Oensingen, Switzerland). Details of the annealing 189
temperatures and primers were published previously (Friesen et al. 2006 for SnToxA; Liu et al. 190
2009 for SnTox3). The SnTox1 primers were newly designed in this study. Sequences and 191
annealing temperatures specific for each primer pair are listed in Supplementary Table S1. 192
Sequencing reactions were produced in both directions using the BigDye® Terminator 193
v3.1 Sequencing Standard Kit (Life Technologies, Applied Biosystems). The PCR cycle 194
parameters were 2 min at 96°C followed by 55 or 99 cycles for 10 sec at 96°C, for 5 sec at 50°C 195
and for 4 min at 60°C. PCR products were cleaned with the Illustra Sephadex G-50 fine DNA 196
grade column (GE Healthcare) according to the manufacturer’s recommendations and sequenced 197
with a 3730xl Genetic Analyzer (Life Technologies, Applied Biosystems). Forward and reverse 198
sequences were aligned using the program Sequencer 5.1 (Gene Code, Ann Arbor, MI). Final 199
alignments were performed using MAFFT ver.7 (Katoh and Standley 2013). 200
201
202
203
204
RESULTS 205
206
Sources of isolates. We obtained new data from 193 P. nodorum isolates sampled from 207
wheat fields located in three major provinces (Kohgiluyeh, Fars and Khuzestan) in the south of 208
Iran (Figure 1; Table 1). We will refer to these isolates collectively as the “new Iranian” 209
population. We added data from 24 isolates originating from the Golestan province in northern 210
Iran that were described in earlier studies (McDonald et al. 2012; McDonald et al. 2013) and will 211
refer to these isolates collectively as the "old Iranian" population. We will refer to the 212
combination of the new and old Iranian populations as the “combined Iranian” population. The 213
combined Iranian population came from four provinces, separated by mountains and deserts, that 214
are characterized by different climates, cropping systems, wheat cultivars and growing seasons, 215
with distances among Iranian field populations ranging from 250-800 km. Measures of genetic 216
diversity in the combined Iranian populations of P. nodorum were compared with identical 217
measures made in an earlier analysis that included 693 P. nodorum isolates sampled from 17 218
wheat fields coming from nine regions distributed across five continents, with none of these 219
earlier collections coming from the Fertile Crescent (Stukenbrock et al. 2006). We will refer to 220
these 693 isolates as the “global populations”. 221
222
Table 1. Collection sites, sample sizes and host source for Parastagonospora nodorum isolates 223
from Iran. The population from Golestan was described in earlier studies (McDonald et al. 224
20012; McDonald et al., 2013). 225
226
Region Year
collected
Collectors
Sample
size
Host source Climate
Kohgiluyeh 2015 F. Ghaderi 64 wheat
leaves
temperate,
mild winters
Fars 2015 F. Ghaderi 66 wheat ears temperate,
mild winters
Khuzestan 2015 F. Ghaderi 63 wheat
leaves
warm, dry
Golestan 2005, 2010 R. Sommerhalder/M.
Razavi
24 wheat ears warm, humid
227
Measures of genetic diversity. Diversity parameters for the individual SSR loci are 228
summarized in Supplementary Table S2. All eleven SSR loci were successfully amplified for all 229
P. nodorum isolates in the new Iranian population. All loci were polymorphic, with the number 230
of alleles ranging from 3 to 22. Gene diversity ranged from 0.18 (SNOD15) to 0.94 (SNOD1) 231
with an average across all loci of 0.60. 232
Overall levels of gene and genotype diversity were high in each Iranian field population 233
(Table 2). A total of 213 different multilocus genotypes were detected among the 217 combined 234
Iranian isolates, corresponding to an overall clonal fraction of only 2%, with no field population 235
showing a clonal fraction higher than 3%. By way of comparison, clonal fractions among the 236
global populations averaged 6% and ranged from 2% in Switzerland to 33% in Mexico 237
(Stukenbrock et al. 2006). Nei’s measure of gene diversity for the combined Iranian data set 238
ranged from 0.56 in Golestan to 0.66 in Fars and averaged 0.60 across the combined Iranian field 239
populations. Gene diversity among the global populations was generally lower, ranging from 240
0.44 in Mexico to 0.57 in Texas with an average of 0.58 when including all 693 isolates in the 241
global population (Stukenbrock et al. 2006). Allelic richness among the combined Iranian 242
populations was estimated by resampling 103 datasets from each field population. The allelic 243
richness varied between 4.61 in Golestan to 5.72 in Khuzestan, with an average allelic richness 244
across the combined Iranian population of 4.85 (Table 2). We compared the allelic richness and 245
private allelic richness at SSR loci with the global populations using rarefaction to account for 246
differences in sample size. In agreement with a previous study that included only the old Iranian 247
population from Golestan (McDonald et al. 2012), private allelic richness in the combined 248
Iranian population was higher than other regions around the world. Adding the new Iranian 249
population to the old Iranian population also led to an increase in mean allelic richness, giving 250
Iran the highest level of neutral gene diversity on a global scale (Figure 2). 251
252
253
Table 2. Measures of gene and genotypic diversity and estimates of linkage disequilibrium based 254
on eleven microsatellite loci and mating type ratios in the Iranian Parastagonospora nodorum 255
populations 256
IA, rd; Association indices (measures of linkage disequilibrium). Population estimates in 257
brackets are after removing the linked locus SNOD8. The "All populations" estimates in 258
brackets are based on excluding the admixed population Kohgiluyeh. 259
* Significant at P < 0.05; ns: non-significant. 260
ᵡ2 Value for deviation from a 1:1 mating type ratio. 261
262
263
264
Population no. of isolates
Clone corrected
Clonal fraction
Genotype diversity
IA rd Gene diversi
ty
Allelic richness
Ratio MAT1-1:MAT1-2
ᵡ2
Kohgiluyeh 64 62 0.03 97% 0.31* (0.21*)
0.032* (0.028*)
0.58
4.99
45:17
6.66*
Khuzestan 63 61 0.03 97% ns 0.068 ns 0.008 0.57 4.72 34:27 0.40 ns
Fars
66 66 0 100% 0.23* )ns(0.12
0.025* )ns15 0.0(
0.66
5.10
37:29
0.49 ns
Golestan 24 24 0 100% 0.034 ns 0.004 ns
0.56
4.61
10:14
0.34 ns
All populations
217 213
0.02 98% 0.34* (0.24 ns)
0.033* (0.027 ns)
0.60 4.85 126:87 3.59 ns
265
266
Figure 2. Comparisons of population diversity at neutral SSR loci for global Parastagonospora 267
nodorum populations. ADZE rarefaction results for the combined Iranian populations of P. 268
nodorum compared to global populations analyzed in an earlier study (McDonald et al. 2012). 269
The rarefied mean allelic richness (A) and private allelic richness (B) are shown in relation to 270
sample size. Maximum samples sizes are limited by the smallest population sample size. 271
272
273
Population differentiation. Genetic differentiation between all Iranian populations was 274
estimated using the stepwise RST mutation model (Supplementary Table S3). Differentiation 275
between southern population pairs ranged from RST = 0.08 to 0.11, while differentiation between 276
the southern populations and the northern population of Golestan was higher, ranging from RST = 277
0.26 to 0.37. These values were considered to be high given the much smaller geographic 278
sampling scale compared to the nine global populations located on different continents and 279
separated by thousands of km that showed an average pairwise RST = 0.07 for the same SSR loci 280
(Stukenbrock et al. 2006). 281
In concordance with the RST analyses, STRUCTURE clustered the Iranian isolates into 282
four distinct groups according to their geographical origins (Figure 3). The high assignment 283
probabilities for all individuals from Golestan, Fars and Khuzestan suggest limited gene flow 284
among these populations. In contrast, several isolates from Kohgiluyeh were identified as being 285
admixed mainly with the Fars population, suggesting recent gene flow between these two 286
populations. 287
288
0
1
2
3
4
5
6
7
8
0 10 20 30 40 50 60 70 80 0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
2 6 10 14 18 22 26 30
combined Iran China Australia Oregon New York South Africa Switzerland Texas
Mea
n pr
ivat
e al
lelic
rich
ness
Mea
n al
lelic
rich
ness
Sample size Sample size
A B
289
290
Figure 3. Structure analysis of the combined Iranian Parastagonospora nodorum populations. 291
Individuals were assigned to clusters based on the SSR data set. A. DeltaK shows the highest 292
likelihood at K = 4 assumed populations. B. STRUCTURE plot of assignment probabilities for 293
each individual assuming K = 4 populations. 294
295
296
Tests for random mating based on mating type frequencies and measures of linkage 297
disequilibrium. PCR amplification of the mating type idiomorphs produced single amplicons 298
corresponding to either MAT1-1 or MAT1-2 in all Iranian isolates. The frequency distribution 299
showed a bias towards MAT1-1, but it was not significantly different from the expected 1:1 ratio 300
for the entire data set, with 126 MAT1-1 vs. 87 MAT1-2. The mating type ratio was not 301
significantly different from 1:1 for three of the field populations, but Kohgiluyeh had a 302
significantly skewed distribution with 45 MAT1-1 vs. 17 MAT1-2 (Table 2). 303
Linkage disequilibrium (LD) estimates were significantly different from the distribution 304
expected under the hypothesis of no associations for Kogiluyeh and Fars (Table 2). Closer 305
inspection revealed that loci SNOD8 and SNOD17 were significantly associated. A BLAST 306
search on the genome of the P. nodorum reference isolate SN15 v2.0 (Hane et al. 2007) revealed 307
that both loci were located on scaffold 4 approximately 1000 bp apart, i.e. they were tightly 308
linked. Removing locus SNOD8 from the analysis resulted in non-significant LD for Fars, but 309
Kogiluyeh still showed significant LD. 310
Analyses of necrotrophic effectors. We PCR-screened the new Iranian population for 311
the presence of the SnToxA, SnTox1 and SnTox3 genes with gene-specific primers 312
GolestanFarsKhuzestanKohgiluyehK
10Delta
K 2030
0
2 3 4 5 6
A B
(Supplementary Table S1). SnToxA and SnTox1 were found at frequencies of 95% and 97%, 313
respectively while SnTox3 was found in 72% of isolates. The distribution of all possible multi-314
effector genotypes is shown in Supplementary Fig. S1. The combination with all three effectors 315
present (A+3+1+) was by far the most abundant (70%, compared to 66% expected under neutral 316
associations), followed by genotype A+3-1+ (25%, compared to 24% expected under neutral 317
associations). When compared to an earlier analysis of multi-effector genotypes in the global 318
population of P. nodorum, only the Australian population had a higher frequency of strains 319
carrying all three necrotrophic effectors (McDonald et al. 2013). 320
We sequenced SnToxA, SnTox1 and SnTox3 for 96 randomly chosen isolates from the 321
new Iranian population and added the previously published sequence data from the old Iranian 322
population (McDonald et al. 2013). The combined Iranian SnToxA sequences collapsed into 16 323
distinct haplotypes. Twelve of these haplotypes were already detected among the P. nodorum 324
isolates in the global data set, and haplotypes H1, H5 and H15 had been detected in the related 325
species Pat1 (McDonald et al. 2013). Importantly, we detected several SnToxA haplotypes 326
among the new Iranian isolates that were previously reported as private alleles for other regions 327
(McDonald et al. 2013). For example, SnToxA haplotypes H2, H4 and H9 were unique to South 328
Africa in the earlier global data set, but we found each of these alleles in the new Iranian field 329
populations (Figure 4). Three SnToxA haplotypes (H18 – H20) found in the new Iranian 330
population were not detected previously, resulting in Iran having the highest number of private 331
SnToxA alleles and increasing the number of distinct P. nodorum SnToxA haplotypes to 20 332
(Figure 4A). The 20 SnToxA haplotypes translated into 11 protein isoforms. It is noteworthy that 333
all 11 isoforms were detected among Iranian isolates (Figure 4B). Haplotypes 3 and 17, detected 334
in North America and South Africa, respectively, contain several stop codons each and are 335
unlikely to translate into functional proteins (Figure 4). 336
337
338
Figure 4. Haplotype network showing global SnToxA diversity. A. A TCS network connecting 339
all distinct SnToxA haplotypes. Populations organized according to global regions use the same 340
color-coding as McDonald et al. (2013) to enable easy comparisons. The new Iranian isolates (in 341
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 SnToxA protein isoforms
Bcombined Iran
North America
South Africa
Central Asia
Europe
China
Australia
B. sorokiniana
H17.12H8.9
H13.11
H5.1
H7.1H4.8
H16.2
H12.7
H6.7
H3.13H10.3H11.3
H2.5H19.10
H18.2
H1.2
H15.2
H9.4
H20.6
A
H21.14H22.15
∗∗
∗∗∗
H14.2
North AmericaSouth Africanew Iranold Iran (Golestan) Central AsiaEuropeChinaAustralia
B. sorokinianaP. tritici-repentisP. avenaria f. sp. tritici 1
P. tritici-repentisP. avenariaf. sp. tritici 1
white) are indicated separately from the old Iranian isolates (in blue). The sizes of haplotype 342
circles are proportional to their frequencies in the P. nodorum data set. Haplotypes found in other 343
species than P. nodorum are represented by a single individual. The number after the dot in the 344
haplotype designation indicates the corresponding protein isoform. Hash marks indicate single 345
nucleotide polymorphisms and asterisks indicate stop codons. B. The distribution of SnToxA 346
protein isoforms across populations. Isoforms 12 and 13 (surrounded by dashed lines) are likely 347
to be non-functional because of several stop codons in both nucleotide haplotypes. 348
349
350
Similar patterns of diversity were observed for SnTox1 and SnTox3, leading the combined 351
Iranian population to have the highest number of private alleles for all three NE-encoding genes. 352
The combined Iranian Tox1 sequences collapsed into 11 distinct haplotypes. Of these, four 353
haplotypes (H19 - H22) were new and unique to Iran, increasing the total number of SnTox1 354
haplotypes to 22 (Supplementary Figure S2). The combined Iranian SnTox3 sequences collapsed 355
into eight distinct haplotypes. Of these, two haplotypes were new and unique to Iran (H12 – 356
H13), increasing the total number of SnTox3 haplotypes to 13 (Supplementary Figure S3). The 357
rarefaction analyses also identified Iran as the region with the highest number of SnToxA alleles 358
(Figure 5), but other regions had higher numbers of SnTox1 and SnTox3 alleles (Supplementary 359
Figure S4). 360
361
362
363
Figure 5. Rarefaction analysis of global SnToxA diversity. The rarefied number of SnToxA 364
alleles found in each resampled population as a function of increasing sample size. Data for non-365
Iranian populations were taken from an earlier study (McDonald et al. 2013). 366
367
Because our sequence analyses detected many Iranian effector alleles that were shared 368
with global populations (Figure 4, Supplementary Figures S2, S3), we conducted a more detailed 369
analysis of the number of effector alleles that were shared between regional populations. Iran 370
shared the highest number of effector alleles with other populations for SnToxA and SnTox3, but 371
the pattern was less clear for SnTox1 because several populations had similar numbers of shared 372
alleles (Supplementary Fig. S5). 373
374
DISCUSSION 375
376
We analyzed population genetic diversity for neutral and selected loci in a large 377
0
1
2
3
4
5
6
7
8
9
1 3 5 7 9 11 13 15 17 19
South Africa
North America
Central Asia
Europe
China
Australia
combined Iran
Num
ber o
f SnToxA
Hapl
otyp
es
Sample Size
collection of P. nodorum isolates obtained from four different regions in Iran. We found that the 378
Iranian field populations had the highest genetic diversity detected globally to date for all genetic 379
markers. We also found significant population structure in Iran using neutral SSR markers. 380
These results add strong support to the hypothesis that P. nodorum originated in the same region 381
where wheat was domesticated (Balter 2007), similar to what was found for the wheat pathogen 382
Zymoseptoria tritici (Banke and McDonald 2005; Stukenbrock et al. 2006). We also discovered 383
that Iran is a global hotspot of diversity for all three genes encoding necrotrophic effectors and 384
that Iran appears to be a hub or source population for effector alleles shared with other P. 385
nodorum populations around the world. Taken together, these findings support the hypothesis 386
that the Fertile Crescent population of P. nodorum is the source population for all three effectors, 387
arguing against an earlier hypothesis that all three effectors were acquired by P. nodorum 388
through separate horizontal transfers after it escaped from the Fertile Crescent (McDonald et al. 389
2013). 390
P. nodorum originated in the Fertile Crescent. It is generally assumed that populations 391
at a species' center of origin (i.e. the original source for other populations that became 392
established in new locations) will display the highest genetic diversity at neutral loci due to the 393
accumulation of new mutations over many generations. In contrast, more recently founded sink 394
populations typically show lower genetic diversity compared to older source populations as a 395
result of genetic bottlenecks imposed by the founding event coupled with less residence time for 396
mutations to accumulate. 397
Consistent with a center of origin, the Iranian populations of P. nodorum showed the 398
highest diversity at neutral SSR loci compared to other global populations. The average allelic 399
richness across the combined Iranian populations was 4.85, while the highest allelic richness 400
found in global populations was 4.44 (Switzerland) and the mean allelic richness based on 401
combining all 693 global isolates was 4.81 (Stukenbrock et al. 2006). Rarefaction analyses also 402
showed that Iran has the highest allelic richness for SSRs compared to all other global 403
populations (Fig. 2). 404
Despite the relatively small spatial scale sampled in Iran, the four Iranian populations 405
showed strong signatures of differentiation, with an average RST of 0.20. For comparison, the 406
differentiation between global populations separated by similar distances were RST = 0.01 407
between Texas and Oregon, RST = 0.00 between Denmark and Switzerland, and the average 408
differentiation among all global populations was 0.07 (Stukenbrock et al. 2006). The finding of 409
greater genetic structure among geographically nearby populations is expected for older 410
populations found at a species' center of origin as a result of mutation/drift equilibrium reached 411
after many generations among populations separated by geographical features such as mountains 412
and deserts (see review by Orsini et al. 2013 and references therein). 413
Consistent with this theory, the STRUCTURE analyses showed that the four Iranian 414
sampling sites maintained genetically distinct P. nodorum populations. While Golestan, Fars and 415
Khuzestan showed only marginal signatures of admixture, several isolates from Kohgiluyeh were 416
identified as being admixed with Fars (Fig. 3). We speculate that this signature of gene flow is a 417
result of human-mediated dispersal of infected seed from the Fars region into the Kohgiluyeh 418
region, which enabled introgression of P. nodorum alleles from the Fars population into the 419
population at Kohgiluyeh. P. nodorum often infects ears and seed-borne infection is common 420
(e.g. Bennett et al. 2007). Infected seeds are thought to be the main mechanism responsible for 421
the global dispersal of the pathogen (Bennett et al. 2005). 422
Regular sexual recombination is a key factor that affects a pathogen's evolutionary 423
potential. The reshuffling of polymorphism allows for rapid adaptation both to the host's immune 424
response and to external stresses such as the application of fungicides (McDonald and Linde 425
2002; Linde et al. 2003). In previous studies, most populations of P. nodorum were reported to 426
exhibit the "signature of sex", including high genotypic diversity, low clonality, random 427
associations among neutral markers (i.e. gametic equilibrium), and approximately equal 428
frequencies of the two mating types (Bennet et al. 2005; Stukenbrock et al. 2006). Our findings 429
in Iran were largely in agreement with these earlier studies. After removing one of two linked 430
microsatellite markers, three of the Iranian populations were in gametic equilibrium, but the 431
Kohgiluyeh population showed gametic disequilibrium and mating type frequencies that deviated 432
significantly from the expected 1:1 ratio. We also found that many strains from Kohgiluyeh 433
showed strong signatures of population admixture, mainly with Fars. Admixture can result in 434
high levels of LD that will decline over time due to recombination (Pritchard and Rosenberg 435
1999). Based on our findings, we hypothesize that the skewed mating type ratio and linkage 436
disequilibrium in Kohgiluyeh are a result of recent admixture with P. nodorum isolates from the 437
Fars region following a recent introduction on infected seed. 438
Evidence that all three necrotrophic effectors originated in P. nodorum at its center 439
of origin. P. nodorum possesses several necrotrophic effectors that facilitate the infection 440
process (Friesen et al. 2008). These NEs show population-specific patterns of presence/absence 441
polymorphism as well as high nucleotide diversity consistent with local adaptation to the NE 442
sensitivity genes carried by the local host population (Stukenbrock and McDonald 2007; 443
McDonald et al. 2013). Earlier analyses of nucleotide diversity in the NE-encoding genes 444
SnToxA, SnTox1 and SnTox3 did not provide evidence for a single center of origin for these 445
effectors (Stukenbrock and McDonald 2007; McDonald et al. 2013). The highest nucleotide 446
diversity for SnToxA was found in South Africa, the highest diversity for SnTox1 was in Europe, 447
and the highest diversity for SnTox3 was in North America. These findings, coupled with the 448
discovery that all three effector genes were found only in the closely-related Parastagonospora 449
lineages P. nodorum and Pat1 and not in more distant Parastagonospora lineages (McDonald et 450
al. 2012; McDonald et al. 2013), led to the hypothesis that P. nodorum acquired all three NEs 451
through independent horizontal gene transfers (McDonald et al. 2013). Our detailed analyses 452
presented here, that included a much larger and more geographically diverse collection of strains 453
from Iran, indicate a very different scenario, namely an origin for all three NEs in the P. 454
nodorum population residing in the Fertile Crescent. We detected the highest number of private 455
alleles for all three NEs in the Iranian populations and we discovered that many alleles 456
previously reported to be unique in other global populations also existed in Iran. For example, 457
earlier analyses indicated that South Africa had four private alleles for SnToxA (haplotypes H2, 458
H3, H4 and H9). Apart from H9, all of these alleles were found in the new P. nodorum 459
collections from Iran (Fig. 4A). 460
It is noteworthy that many effector alleles were shared between very distant populations. 461
For example, the most frequent allele for SnToxA (H1) was found in Iran, North America, 462
Australia, Europe and also in South Africa. Some very rare alleles were also shared among 463
distant populations. For example, SnToxA alleles H2 and H9 were previously found only in 464
single isolates from South Africa (McDonald et al. 2013). In our new Iranian population, we 465
found 6 isolates with the H2 allele and 5 isolates with the H9 allele. This finding of shared rare 466
alleles can be interpreted as evidence for recent gene flow between Iran and South Africa. An 467
alternative explanation is that the same allele emerged independently in the two populations, 468
perhaps as a result of selection due to deployment of the same effector sensitivity genes in the 469
corresponding host populations. To investigate the pattern of shared alleles in more detail, we 470
conducted pairwise population analyses of normalized numbers of shared effector haplotypes 471
(Supplementary Fig. S5). Iran was identified as the region that shared the most haplotypes with 472
all other populations for SnToxA and SnTox3. The pattern for SnTox1 was inconclusive because 473
several populations shared similar numbers of haplotypes. 474
We conclude that the observed patterns of SSR diversity and necrotrophic effector 475
diversity are most compatible with the hypothesis that P. nodorum originated in the Fertile 476
Crescent region and that the original pathogen populations already carried the genes encoding 477
SnToxA, SnTox1 and SnTox3. More population samples from other regions in the Fertile 478
Crescent, eg. Turkey, Israel or Iraq, should be analyzed to further test this hypothesis. Our results 479
do not support an earlier hypothesis of independent origins for the three necrotrophic effectors 480
through horizontal gene transfer. The ToxA gene was also discovered in the genomes of the 481
wheat tan spot pathogen Pyrenophora tritici-repentis (Friesen et al. 2006) and the wheat spot 482
blotch pathogen Bipolaris sorokiniana (McDonald et al. 2018), though only one allele (H21 in 483
Figure 4A) has been found in P. tritici-repentis and only two alleles (H21 and H22 in Figure 4A) 484
have been found in B. sorokiniana. Though H21 and H22 have not yet been found in P. 485
nodorum, the closest allele in P. nodorum (H6) was found in Iran and is separated from H21 by 486
only three SNPs. Taken together, these findings support the hypothesis that P. nodorum was the 487
original source of the ToxA genes now found in Pat1 (McDonald et al. 2013), Pyrenophora 488
tritici-repentis (Friesen et al. 2006) and Bipolaris sorokiniana (McDonald et al. 2018). 489
490
ACKNOWLEDGMENTS 491
492
The authors are grateful to M. Zala for his assistance in the laboratory. DNA data were 493
collected in the Genetic Diversity Centre of ETH Zurich. FG was supported by a PhD 494
scholarship from the Iranian Ministry of Higher Education. This work was funded by the Swiss 495
Federal Institute of Technology (ETH), Zurich. 496
497
498
499
500
LITERATURE CITED 501
502
Agapow, P. M., and Burt, A. 2001. Indices of multilocus linkage disequilibrium, Molecular 503
Ecology Notes, 1, 101-102. 504
Balter, M, 2007. Seeking agriculture's ancient roots. Science, 316, 1830–1835. 505
Banke, S., and McDonald, B. A. 2005. Migration patterns among global populations of the 506
pathogenic fungus Mycosphaerella graminicola. Molecular Ecology, 14, 1881–1896 507
Bennett, R. S., Milgroom, M. G., and Bergstrom, G. C. 2005. Population structure of seedborne 508
Phaeosphaeria nodorum on New York wheat. Phytopathology, 95, 300-305. 509
Bennett, R. S., Milgroom, M. G., Sainudiin, R., Cunfer, B. M., and Bergstrom, G. C. 2007. 510
Relative contribution of seed-transmitted inoculum to foliar populations of 511
Phaeosphaeria nodorum. Phytopathology, 97, 584-591. 512
Bennett, R. S., Yun, S. H., Lee, T. Y., Turgeon, B. G., Arseniuk, E., Cunfer, B. M., and 513
Bergstrom, G. C. 2003. Identity and conservation of mating type genes in geographically 514
diverse isolates of Phaeosphaeria nodorum. Fungal Genetics and Biology, 40, 25-37. 515
Bhathal, J. S., Loughman, R., and Speijers, J. 2003. Yield reduction in wheat in relation to leaf 516
disease from yellow (tan) spot and Septoria nodorum blotch. European Journal of Plant 517
Pathology, 109, 435–443. 518
Chen, R. S., Boeger, J. M., and McDonald, B. A. 1994. Genetic stability in a population of a 519
plant pathogenic fungus over time. Molecular Ecology, 3, 209–218. 520
Evanno, G., Regnaut, S., and Goudet, J. 2005. Detecting the number of clusters of individuals 521
using the software STRUCTURE: a simulation study. Molecular Ecology, 14, 2611–522
2620. 523
Excoffier, L., and Lischer, H. E. L. 2010. Arlequin suite ver 3.5: A new series of programs to 524
perform population genetics analyses under Linux and Windows. Molecular Ecology 525
Resources, 10, 564-567. 526
Eyal, Z. 1999. Septoria and Stagonospora diseases of cereals: A comparative perspective. Pages 527
1-25 in: Septoria on Cereals: A Study of Pathosystems. J. A. Lucas, P. Bowyer, and H. 528
M. Anderson, eds. IACR Long Ashton Research Station, Bristol, UK. 529
Eyal, Z., Scharen, A. L., Prescott, J. M., and Van Ginkel, M. 1987. The Septoria diseases of 530
wheat: Concepts and methods of disease management. CIMMYT, Mexico, D.F. 51 pp 531
Falush, D., Stephens, M., and Pritchard, J. K. 2003. Inference of population structure using 532
multilocus genoype data: linked loci and correlated allele frequencies. Genetics, 164, 533
1567–1587. 534
Friesen, T. L., Faris, J. D., Solomon, P. S., and Oliver, R. P. 2008. Host‐specific toxins: effectors 535
of necrotrophic pathogenicity. Cellular Microbiology, 10, 1421–1428. 536
Friesen, T. L., Stukenbrock, E.H., Liu, Z., Meinhardt, S., Ling, H., Faris, J. D., Rasmussen, J. B., 537
Solomon, P. S., McDonald, B. A. and Oliver, R. P. 2006. Emergence of a new disease as 538
a result of interspecific virulence gene transfer. Nature Genetics, 38, 953-956. 539
Goudet, J. 1995. Fstat version 1.2: a computer program to calculate F-statistics. Journal of 540
Heredity, 86, 485–486. 541
Halama, P., and Lacoste, L. 1991. Déterminisme de la reproduction sexuée du Phaeosphaeria 542
(Leptosphaeria) nodorum agent de la septoriose du blé. II. Action de la température et de 543
la lumière. Canadian Journal of Botany, 70, 1563-1569. 544
Hane., J. K., Lowe, R. G. T., Solomon, P. S., Tan, K.-C., Schoch, C. L, Spatafora, J. W., Crous, 545
P.W., Kodira, C., Birren, B. W., Galagan, J.E., Torriani, S. F. F., McDonald, B. A., and 546
Oliver, R. P. 2007 Dothideomycete–plant interactions illuminated by genome sequencing 547
and EST analysis of the wheat pathogen Stagonospora nodorum. The Plant Cell, 19, 548
3347-3368. 549
Kamvar, Z. N., Tabima, J. F., and Grünwald, N. J. 2014. Poppr: an R package for genetic 550
analysis of populations with clonal, partially clonal, and/or sexual reproduction. PeerJ 551
2:e281.doi:10.7717/peerj.281. 552
Katoh, K., and Standley, D. M. 2013. MAFFT Multiple Sequence Alignment Software Version 553
7: Improvements in performance and usability. Molecular Biology and Evolution, 30, 554
772–780. 555
Keller SM, Wolfe MS, McDermott JM, McDonald BA. 1997b. High genetic similarity among 556
populations of Phaeosphaeria nodorum across wheat cultivars and regions in 557
Switzerland. Phytopathology, 87, 1134-1139. 558
Kopelman, N. M., Mayzel, J., Jakobsson, M., Rosenberg, N. A., and Mayrose, I. 2015. Clumpak: 559
a program for identifying clustering modes and packaging population structure inferences 560
across K. Molecular Ecology Resources, 10.1111/1755-0998.12387. 561
Linde, C. C., Zala, M., Ceccarelli, S., and McDonald, B. A. 2003. Further evidence for sexual 562
reproduction in Rhynchosporium secalis based on distribution and frequency of mating-563
type alleles. Fungal Genetics and Biology, 40, 115–125. 564
McDonald, B. A., and Linde, C. 2002. Pathogen population genetics, evolutionary potential, and 565
durable resistance. Annual Review of Phytopathology 40, 349–379. 566
McDonald, B. A., Miles, J., Nelson, L. R., and Pettway, R. E. 1994. Genetic variability in 567
nuclear DNA in field populations of Stagonospora nodorum. Phytopathology, 84, 250-568
255. 569
McDonald, B. A., Pettway, R. E., Chen, R. S., Boeger, J. M., and Martinez, J. P. 1995. The 570
population genetics of Septoria tritici (teleomorph Mycosphaerella graminicola). 571
Canadian Journal of Botany, 73(S1), 292-301. 572
McDonald, M. C., Ahren, D., Simpfendorfer, S., Milgate, A. and Solomon, P. S. 2018. The 573
discovery of the virulence gene ToxA in the wheat and barley pathogen Bipolaris 574
sorokiniana. Molecular Plant Pathology, 19, 432-439. 575
McDonald, M. C., Oliver, R. P., Friesen, T. L., Brunner, P. C., and McDonald, B. A. 2013. 576
Global diversity and distribution of three necrotrophic effectors in Phaeosphaeria 577
nodorum and related species. New Phytologist, 199, 241-251. 578
McDonald, M. C., Razavi, M., Friesen, T. L., Brunner, P. C., and McDonald, B. A. 2012. 579
Phylogenetic and population genetic analysis of Phaeosphaeria nodorum and its close 580
relative indicate cryptic species and an origin in the Fertile Crescent. Fungal Genetics 581
and Biology, 49, 882-895. 582
Meirmans, P. G., and Van Tienderen, P. H. 2004. GENOTYPE and GENODIVE: two programs 583
for the analysis of genetic diversity of asexual organisms. Molecular Ecology Notes, 4, 584
792–794. 585
Michalakis, Y., and Excoffier, L.1996. A generic estimation of population subdivision using 586
distances between alleles with special reference for microsatellite loci. Genetics, 142, 587
1061–1064. 588
Murray, M., and Thompson, W. F. 1980. Rapid isolation of high molecular weight plant DNA. 589
Nucleic Acid Research, 8, 4321-4325. 590
Orsini, L., Vanoverbeke, J., Swillen, I., Mergeay, J., and Meester, L. 2013b. Drivers of 591
population genetic differentiation in the wild: isolation by dispersal limitation, isolation 592
by adaptation and isolation by colonization. Molecular Ecology, 22, 5983–5999. 593
Pritchard, J. K., and Rosenberg, N. A. 1999. Use of unlinked genetic markers to detect 594
population stratification in association studies. The American Journal of Human Genetics. 595
65, 220–228. 596
Pritchard, J. K., Stephens, M., and Donnelly, P. 2000. Inference of population structure using 597
multilocus genotype data. Genetics, 155, 945–959. 598
Quaedvlieg, W., Verkley, G. J., Shin, H. D., Barreto, R. W., Alfenas, A. C., Swart, W. J., 599
Groenewald, J. Z., and Crous, P. W. 2013. Sizing up Septoria. Studies in Mycology, 75, 600
307–390. 601
Sommerhalder, R. J, McDonald, B. A., and Zhan, J. 2006. The frequencies and spatial 602
distribution of mating types in Stagonospora nodorum are consistent with recurring 603
sexual reproduction. Phytopathology, 96, 234–239. 604
Stoddart, J. A., and Taylor, J. F. 1988. Genotype diversity: estimation and prediction in samples. 605
Genetics, 118, 705–711. 606
Stukenbrock, E. H., Banke, S., and McDonald, B. A. 2006. Global migration patterns in the 607
fungal wheat pathogen Phaeosphaeria nodorum. Molecular Ecology, 15, 2895–2904. 608
Stukenbrock, E. H., Banke, S., McDonald, B. A., and Oliver, R. 2005. Isolation and 609
characterization of EST-derived microsatellite loci from the fungal wheat pathogen 610
Phaeosphaeria nodorum. Molecular Ecology Notes, 5, 931–933. 611
Stukenbrock, E. H., and McDonald, B. A. 2007. Geographic variation and positive diversifying 612
selection in the host specific toxin SnToxA. Molecular Plant Pathology, I8, 321–332. 613
Weir, B. S., and Cockerham, C. C. 1984. Estimating F-statistics for the analysis of population 614
structure. Evolution, 38, 1358–1370. 615
Wiese, M. V. 1987. Compendium of Wheat Diseases. St Paul, MN: American Phytopathological 616
Society. 617
Yeh, F. C., Yang, R. C., Boyle, T. B. J., Ye, Z. H., and Mao, J. X. 1999. POPGENE, the user-618
friendly shareware for population genetic analysis. Molecular Biology and Biotechnology 619
Centre, University of Alberta, Canada. 620
621
622
Appendix: Supplementary Tables 623 624
Table S1. PCR primers used for the detection of the necrotrophic effectors SnToxA, SnTox1, and 625
SnTox3 among P. nodorum isolates. 626
627
628
629
630
Primer pair Sequence Annealing
Temp °C
Citation
ToxA1.F 5’-CGTCCGGCTACCTAGCAATA 56 Friesen et al. 2006
ToxA1.R 5’-TTGTGCTCCTCCTTCTCGA 56 Friesen et al. 2006
Tox1_12.F 5’-AACAGCGACATCCCTACGAC 56 This paper
Tox1_741.R 5’-ATTGCCAGAACACCTGCGTA 56 This paper
Tox3_8981C.F 5’-ATGCATTTTACAAAGTTCCT 55 Liu et al. 2012
Tox3_8981C.R 5’-CTACTCCCCTCGTGGGATTGCCCCAT 55 Liu et al. 2012
Table S2. Measures of microsatellite diversity for Iranian samples of Parastagonospora 631 nodorum 632 633
634
†Calculated according to Nei (1973) 635
636
637
Locus Size range Fars (N = 66)
Kohgiluyeh (N = 64)
Khuzestan (N = 63)
Golestan (N = 23)
Total no. of alleles
Gene diversity†
SNOD1 253-439 13 14 20 13 22 0.94 SNOD3 278-318 3 3 2 2 4 0.28 SNOD5 409-472 12 8 8 5 16 0.88 SNOD8 325-425 7 7 1 4 10 0.71 SNOD11 231-236 3 2 2 2 4 0.5 SNOD15 156-174 2 2 2 3 3 0.18 SNOD16 188-208 5 5 6 6 9 0.5 SNOD17 92-158 3 4 7 5 10 0.68 SNOD21 191-228 7 6 5 5 7 0.79 SNOD22 225-267 6 6 4 5 8 0.73 SNOD23 294-402 4 7 4 1 7 0.33
Table S3. Population differentiation measured by pairwise RST (above diagonal) and 638
corresponding p-values (below diagonal) among the four Iranian Parastagonospora nodorum 639
populations. 640
641
642 643 644
Population
Kohgiluyeh
Khuzestan
Fars
Golestan
Kohgiluyeh
---
0.076
0.042
0.362
Khuzestan 0.005 --- 0.114 0.257 Fars 0.014 < 0.001 --- 0.368 Golestan < 0.001 < 0.001 < 0.001 ---
Appendix: Supplementary Figures 645
646
647 648
Figure S1. Frequency distribution of multi-effector genotypes in the new Iranian population of 649
Parastagonospora nodorum. Each isolate was screened for the presence of SnToxA, SnTox1 650
and/or SnTox3 using gene-specific PCR primers. The X-axis shows all possible combinations of 651
presence (+) and absence (-) for each effector. The legend shows the total number of individuals 652
assayed and the percentage carrying each effector. 653
654
655
140
120
100
80
60
40
20
0
ToxA 193 95.3Tox3 193 72.5Tox1 193 97.4
N % Present
A+3+1+
A+3+1-
A+3-1+
A+3-1-
A-3-1+
A-3-1-
A-3+1+
A-3+1-
Necrotrophic effector combinations
Freq
uenc
y
656
657
658 Figure S2. Haplotype network showing global SnTox1 diversity. The TCS network connects all 659
distinct SnTox1 haplotypes. Populations organized according to global regions use the same 660
color-coding as McDonald et al. (2013) to enable easy comparisons. The new Iranian isolates (in 661
white) are indicated separately from the old Iranian isolates (in blue). The sizes of haplotype 662
circles are proportional to their frequencies. Hash marks indicate single nucleotide 663
polymorphisms. 664
665
NorthAmerica_AA42
10 samples
1 sample
North AmericaSouth Africanew Iranold Iran (Golestan)Central AsiaEuropeChinaAustralia
H3
H15 H9
H5
H4
H8
H7
H11 H6
H17
H10
H12
H2
H13
H18H14
H1
H19
H20
H21
H22
666
667 Figure S3. Haplotype network showing global SnTox3 diversity. The TCS network connects all 668
distinct SnTox3 haplotypes. Populations organized according to global regions use the same 669
color-coding as McDonald et al. (2013) to enable easy comparisons. The new Iranian isolates (in 670
white) are indicated separately from the old Iranian isolates (in blue). The sizes of haplotype 671
circles are proportional to their frequencies. Hash marks indicate single nucleotide 672
polymorphisms. 673
674
675
10 samples
1 sample
North AmericaSouth Africanew Iranold Iran (Golestan)Central AsiaEuropeChinaAustralia
H10 H9
H11
H5
H3
H2
H7H6
H8
H12
H4
H1
H13
676 677 Figure S4. Rarefaction analysis of global SnToxA diversity. A. The rarefied number of SnTox1 678
alleles. B. The rarefied number of SnTox3 alleles found in each resampled population as a 679
function of increasing sample size. Data for non-Iranian populations were taken from an earlier 680
study (McDonald et al. 2013). 681
0
1
2
3
4
5
6
7
8
9 11 13 15 17 19
combined Iran North America South Africa Central Asia Europe China Australia
Num
ber o
f SnTox1
Hap
loty
pes
Sample Size
0
1
2
3
4
5
6
9 11 13 15 17 19
combined Iran North America South Africa Central Asia Europe China Australia
Num
ber o
f SnTox3
Hap
loty
pes
Sample Size
A
B
682 683
684
Figure S5. Heatmaps and corresponding cluster dendrograms showing the normalized numbers 685
of shared effector alleles among global populations of Parastagonospora nodorum. The lighter 686
the color, the more alleles are shared. The combined Iranian population appears to act as a hub 687
(source population) for globally-shared alleles of SnToxA and SnTox3, but the pattern is not clear 688
for SnTox1. 689
690
691
2.0
3.5
5.0
Cluster Dendrogram SnTox3
1.0
3.0
5.0
Cluster Dendrogram SnToxASh
ared
alle
les
1.53.0
4.5
Cluster Dendrogram SnTox1
3 1 2 7 4 6 5
AustraliaNorth Americacombined IranEuropeCentral AsiaChinaSouth Africa
1234567
3
1
2
7
4
6
5
3 7 6 5 2 1 4 7 5 1 2 6 4 3
3
7
6
5
2
1
4
7
5
1
2
6
4
3
Shar
ed a
lleles
Shar
ed a
lleles