Molecular Diversity of Anthracnose Pathogen Populations Associated with UK Strawberry Production Suggests Multiple Introductions of Three Different Colletotrichum Species

Fragaria × ananassa (common name: strawberry) is a globally cultivated hybrid species belonging to Rosaceae family. Colletotrichum acutatum sensu lato (s.l.) is considered to be the second most economically important pathogen worldwide affecting strawberries. A collection of 148 Colletotrichum spp. isolates including 67 C. acutatum s.l. isolates associated with the phytosanitary history of UK strawberry production were used to characterize multi-locus genetic variation of this pathogen in the UK, relative to additional reference isolates that represent a worldwide sampling of the diversity of the fungus. The evidence indicates that three different species C. nymphaeae, C. godetiae and C. fioriniae are associated with strawberry production in the UK, which correspond to previously designated genetic groups A2, A4 and A3, respectively. Among these species, 12 distinct haplotypes were identified suggesting multiple introductions into the country. A subset of isolates was also used to compare aggressiveness in causing disease on strawberry plants and fruits. Isolates belonging to C. nymphaeae, C. godetiae and C. fioriniae representative of the UK anthracnose pathogen populations showed variation in their aggressiveness. Among the three species, C. nymphaeae and C. fioriniae appeared to be more aggressive compared to C. godetiae. This study highlights the genetic and pathogenic heterogeneity of the C. acutatum s.l. populations introduced into the UK linked to strawberry production.

Introduction route of entry and establishment of C. acutatum in the UK strawberry production. Over this period, various nurseries were importing planting material into the UK, and at least 55 cases of infested material that was planted in the field through imports that were not intercepted by the border inspection posts, were identified [20].
The focus of the present study was to assess the extent of the genetic and pathogenic diversity of these introduced pathogen populations mainly utilising a unique collection of C. acutatum s. l. isolates established through the plant health inspection surveys from the early 1980s onwards. We focused on C. acutatum s.l. because previous reports from France, Israel, UK, Bulgaria and Spain had described this taxa as a major widely distributed pathogen, compared with other species such as C. gloeosporioides s.l. that occur less frequently in Europe [2][3][4][5]12]. A range of historic and contemporary C. acutatum s. l. isolates including those from worldwide strawberry crops, other plant hosts in the UK, as well as worldwide representatives from different hosts building on our previous work were accessed as reference sources for determining the genetic and species identities of isolates associated with UK strawberry anthracnose phytosanitary control work. Based on multi-locus phylogenetic analysis, we have identified 12 different haplotypes that belong to three different species C. nymphaeae, C. godetiae and C. fioriniae suggesting multiple introductions of the strawberry anthracnose pathogen. Pathogenic and growth characteristics of these haplotype representatives further highlight the heterogeneity of the introduced pathogen populations.

Fungal isolates and culture conditions
A diverse collection of C. acutatum s.l. was assembled for this study including: 67 isolates associated with strawberry production in the UK (obtained from the UK Food and Environment Research Agency, or FERA responsible for plant health within the Department for Environment, Food and Rural Affairs), 27 C. acutatum s.l. isolates collected from strawberry in other countries, and 13 isolates collected from other host species in the UK. For further comparison, 33 isolates were added to represent other genetic groups, and novel species from previous studies [6,17,18]. This included two isolates of C. fruticola, two isolates of C. aenigma (belonging to C. gloeosporioides species complex [8]) associated with strawberry, two UK isolates of C. spinaciae and one isolate each of C. graminicola, C. higginsianum [27] and C. fioriniae [28]. Sequence data of the markers was retrieved from the reference genome sequences available from Genbank for C. graminicola and C. higginsianum (accession numbers: ACOD01000000 and CACQ02000000, respectively) used among out-groups in the phylogenetic analysis (Fig 1). Details of the isolate collection used in the present study are provided in Table 1.
Cultures were maintained at 25°C on potato dextrose agar medium (PDA, Difco Laboratories, USA) for up to ten days under a 12 h light/ 12 h dark cycle. Long-term storage at 4°C involved cutting mycelial plugs from the edge of actively growing cultures on PDA and suspending them in sterile water.

Characterization of genetic variation
Genomic DNA was extracted according to the Chelex 100 protocol [29], with some modifications [30]. DNA was quantified using a NanoDrop ND-1000 spectrophotometer (Thermo Scientific, DE, USA).
Various target regions were used to characterise genetic diversity amongst the fungal isolates including: ITS region, partial sequence of the beta-tubulin 2 gene (TUB) (exons 3 through 6, including introns 2 through 4), partial sequence of the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene, and partial sequence of the mating type gene (MAT1-2) (the intron included in the conserved HMGbox region). Target regions were amplified using PCR reaction mixes (20 μl) that contained 1 μl of DNA, 1 μl each of primer (20 μM), 7 μl of H 2 0 and 10 μl of ReadyMix RedTaq (Sigma).
PCR amplification of the target regions for sequencing was carried out as described below using previously published primers under conditions standardised for routine work. For ITS, primers ITS1Ext and ITS4Ext [31] were used. The amplification program consisted of 2 min of initial denaturation (95°C), 30 cycles of amplification (1 min at 94°C, 1 min at 55°C, and 1 min at 72°C) and a final extension at 72°C for 5 min. For TUB, primers TB5 and TB6 [31] were used. The amplification program consisted of 2 min initial denaturation (95°C), 30 cycles of amplification (1 min at 94°C, 1 min at 65°C and 1 min at 72°C) and a final extension at 72°C for 5 min. For GAPDH, primers GDF1 and GDR1 [32] were used. The amplification program consisted of 2 min initial denaturation at 95°C, 35 cycles of amplification (1 min at 94°C, 1 min at 60°C and 30 sec at 72°C) and a final extension at 72°C for 3 min. For MAT1-2, primers HMGacuF2 and HMGacuR [21] for C. acutatum s.l. and primers HMGgloeF1 and HMGgloeR1 for C. gloeosporioides s. l. [33] were used. The amplification program consisted of 5 min initial denaturation at 95°C, 40 cycles of amplification (1 min at 95°C, 1 min between 48°C and 55°C and 30s at 72°C) and a final extension of 20 min at 72°C. PCR products were separated using gel electrophoresis and purified using the QIAquick PCR purification kit (Qiagen, USA).
Sequencing of PCR products was carried out at the University of Warwick Genomics Centre, using an ABI Prism 7900HT or ABI3100 sequence detection system (Applied Biosystems, UK). PCR products were cleaned up and then quantified with reference to a ladder (Bioline EasyLadder I) containing DNA fragments of known concentration. One to five microliters of each sample (depending on DNA concentration) were used in sequencing reactions with the BigDye Terminator v3.1 cycle sequencing kit (Applied Biosystems, UK). ABI trace files were analyzed and consensus sequences were generated using Geneious 7.1.6 [34]. All the sequences were aligned using MUSCLE (http://www.ebi.ac.uk/Tools/msa/muscle/) and were manually edited to optimise the alignment, as required. Multiple alignments were end trimmed in order to have comparable nucleotides.
Multiple sequence alignments were exported to MEGA5 [35] where best-fit substitution models were calculated for each separate sequence dataset. In order to evaluate whether the four sequenced loci were congruent and suitable for concatenation, tree topologies of 50% Neighbour-Joining bootstrap and maximum parsimony analysis (100,000 replicates) were separately performed for each gene and visually compared [36]. The multilocus concatenated alignment (ITS, TUB2, MAT1-2 and GAPDH) was performed with Geneious 7.1.6 [34]. A Markov Chain Monte Carlo (MCMC) algorithm was used to generate phylogenetic trees with Bayesian probabilities using MrBayes 3.2.1 [37] for combined sequence datasets. Models of nucleotide substitution for each gene determined by MEGA5 were included for each locus. The analysis in MrBayes ran for 5000000 of generations to reach a P value lower than 0.01 with two parallel searches using three heated and one cold Markov chain sampled every 100 generations; 25% of generations were discarded as burn-in. Further phylogenetic analysis was performed by Multilocus phylogenetic analysis of the Colletotrichum isolates used in this study. Bayesian MCMC analysis tree constructed from the alignment based on the concatenation of rRNA, TUB, MAT1-2 and GPDH partial sequences of 140 Colletotrichum acutatum sensu lato isolates used in this study. The tree was rooted with sequences from C. graminicola and C. higginsianum retrieved from whole genome sequences and sequences of four C. gloeosporioides sensu lato and two C. spinaciae obtained experimentally. Isolates used to investigate variation in aggressiveness are highlighted in bold.

Comparison of fungal growth in culture
The 67 fungal isolates collected from strawberry in the UK were compared with a subset of other isolates (chosen based on genetic, host and geographic diversity) including 49 isolates of C. acutatum s.l. and four isolates of C. gloeosporioides s.l. for in vitro growth studies on PDA (Potato Dextrose Agar, BD Difco). For experiments, a 7 mm diameter mycelial plug excised from the edge of an actively growing PDA culture was placed at the centre of a fresh PDA plate. In the growth experiment, two perpendicular colony diameters were measured daily and colony radius was calculated from cultures incubated at four different temperatures (15°C, 20°C, 25°C and 30°C) in darkness. Data corresponding to the linear growth phase were subjected to analysis of variance of regression in order to create growth curves for each isolate at each temperature. In both tests three plates were used as replicates. Statistical analysis was performed by SIGMAPLOT 10 program (Sigmaplot Software, USA). Colony characters were recorded after 15 days of incubation at 25°C under 12 h light/ 12 h dark cycle.

Pathogenicity tests
Representative isolates (highlighted with asterisks in Table 1) of each C. acutatum s.l. group isolated from strawberry in UK, together with reference isolates from other hosts, were used for pathogenicity tests on the generally susceptible strawberry cultivar Elsanta [38]. A conidial suspension was prepared for each isolates by flooding 10-day-old PDA culture plates with sterile deionised water. Spore concentration was adjusted to 10 5 spores ml -1 and 10 6 spores ml -1 for fruit and crown inoculation, respectively [7,38]. Unripe fruits (white fruit beginning to turn pink, as shown in Fig 3A) [39] were inoculated with a 5μl drop of conidial suspension. Before inoculation, fruit surfaces were disinfected for 5 min using NaClO (1% active chlorine) in 50% EtOH, washed three times in sterilized water, blotted dry and placed in a tray with moist sand on the bottom to prevent movement of the fruits during further procedures. After inoculation, fruits were incubated at 25°C under 12h light/ 12h in dark cycle. Disease symptoms were evaluated 7 days after inoculation (d.a.i.) (Fig 3B) by recording the incidence of disease (% of infected fruits), and the aggressiveness of lesion development using the following severity scale: 0, no visible lesions; 1, lesions on less than 33% of fruit surface; 2, lesions covering 33-66% of fruit surface; and 3, lesions covering more than 66% of fruit surface. Three fruits inoculated with sterile distilled water (SDW) as well as fresh fruits served as noninoculated controls. Four independent replicates were tested for each fungal isolate, consisting of three inoculated fruits for each replicate. At the end of the experiment, Colletotrichum isolates were re-isolated from infected fruits and cultured on PDA to confirm colony characteristics.
The capability of the isolates to produce crown rot symptoms was evaluated by injecting the crowns of three-months-old strawberry plants (Fig 3C) with 0.2 mL conidial suspension using a syringe [4,7]. Plants were placed in glasshouse at 23°C with 16h light / 8h darkness. After 24 days (d.a.i.), plants were evaluated for the presence of crown tissues with red-brownish discoloration, wilting and collapse of the plant, typical symptoms of Colletotrichum crown rot, according to the following severity scale: 0, no lesions; 1, crown tissues discoloration but no wilting or collapse; 2, wilting or collapse of part of the plant; and 3, plant death. Crowns of all plants were sectioned and examined for the presence of red-brownish lesions (Fig 3D). Crown infection was confirmed by re-isolation of the pathogen. Three plant crowns injected with SDW as well as untouched plants served as negative controls for each replicates. The experiment was independently replicated three times, with six plants for each replicate.
Values of disease severity were used to calculate a Disease Index (DI, average severity) according to the following formula: Svn/N, where v represents the numeric value of the class, n is the number of plants or fruits assigned to the class, N is the total number of the plants or fruits assessed. Data for pathogenicity tests on both fruits and plants were subjected to analysis of variance ANOVA and means compared using Tukey's multiple range test by Systat11 (Systat Software, USA).
Molecular characterisation of 67 Colletotrichum isolates collected from strawberry in the UK along with the reference isolates representing the host and geographic diversity (Figs 1 and 2) suggests that there have been multiple introductions of the anthracnose pathogen belonging to different Colletotrichum species into the country. Three different species C. nymphaeae, C. godetiae and C. fioriniae were identified based on sequence from four loci [6,17,18]. Incidence of these species is shown in Fig 2, where C. nymphaeae corresponds to 73%, followed by C. godetiae (18%) and C. fioriniae (9%). GAPDH is the locus that shows the highest variability across the nucleotide dataset, with 24.1% identical sites for the entire set of data (out-group included) and 59.3% within C. acutatum s.l. The MAT1-2 gene also shows a high variability with 34.4% identical sites of which 78.6% in C. acutatum s.l. TUB and ITS loci show lower percentage of variable sites. In detail, TUB has 58.1% of identical sites in the final alignment and 80.7% only considering C. acutatum s.l. While ITS has 77.8% and 92.4% of conserved nucleotides, respectively with and without out-groups. Based on the nucleotide variability referred to above, four haplotypes of C. nymphaeae, three haplotypes of C. fioriniae, and five haplotypes of C. godetiae were identified further highlighting the multiple introductions of the pathogens belonging to these species into the UK.

Fungal growth in plate culture
Radial growth data of C. acutatum s.l. and C. gloeosporioides s.l. isolates were subjected to analysis of variance of regression in order to obtain growth curves that were all statistically significant (R 2 !0.9447 and P<0.0001), with the only exception of one isolate showing a R 2 = 0.770 (C. nymphaeae CSL224 at 30°C). The slope for each isolate (three replicates for each isolate) belonging to the same species were averaged, in order to detect the hypothetical optimal growth temperature, and results are shown in Table 2. Almost all species, particularly those containing isolates from strawberry in the UK namely C. nymphaeae, C. fioriniae, and C. godetiae had highest growth rates at 25°C that was considered as optimum temperature. It is pertinent to mention that higher levels of strawberry anthracnose incidence in the UK have been reported in the southwest and southeast regions, where relatively high temperatures are most often reached [20]. However, C. phormii, C. kinghormii and C. rhombiforme showed the highest growth rate at the temperature of 20°C and they were not able to grow at 30°C. Interestingly, these three species are evolutionarily closely related, suggesting a specific adaptation to different environmental conditions compared to other members of the same complex. With respect to C. gloeosporioides s.l. isolates (C. aenigma CSL780 and CSL 869; C. fruticola CSL 311 and CSL386), used as out-groups, all the four isolates showed the highest growth rate at all the tested temperatures when compared with all the other isolates.
C. nymphaeae isolates developed white cottony aerial mycelium, light brownish conidial masses with peculiar colony colour from dark grey to dark brown. Twelve isolates belonging to C. godetiae were characterized by white aerial mycelium, and yellow pigmentation to white colour on the reverse side of the culture. C. fioriniae isolates were dark red on the reverse side of the cultures with orange conidial masses in large drops on the colony surface, and conidiomata formed directly on the hyphae. However, these characters are often difficult to describe reliably, and can change following sub-culturing or based on the length and type of storage. Thus, there is a need for further development of molecular methods for reliable and rapid diagnosis and monitoring of the pathogen populations belonging to different species associated with strawberry production in a specific geographic location.

Characterisation of variation in pathogenicity
Thirty-four C. acutatum s.l. isolates were chosen for pathogenicity tests on fruits and plants, including six representative isolates from each of the three species described above related to strawberry production in the UK (highlighted with Ã in Table 1 and in bold in Fig 1), and one or more isolates representative of all the major species of the C. acutatum complex. Four C. gloeosporioides s.l. isolates that were isolated from strawberry infected tissues from UK (CSL 780 and CSL 869, C. aenigma) and USA (CSL 311 and CSL 386, C. fruticola) were included in the experiments as an out-group.
C. acutatum s.l. isolates varied in aggressiveness on both host tissues. In the fruit assays, among the three species identified from the strawberry production systems in the UK, C. nymphaeae and C. fioriniae were more aggressive compared to C. godetiae. This was particularly noticeable for isolates originating from strawberry as reflected by the fruit disease index range for C. nymphaeae (2.08-3.00), C. fioriniae (1.92-2.75) and C. godetiae (0.75-2.08). Interestingly, with isolates originating from other hosts, C. nymphaeae isolates were less aggressive (0.67-1.67), and one or more isolates belonging to C. fioriniae (2.00-2.17) as well as C. godetiae (2.17) showed fruit disease index in the range of the strawberry isolates. Among the other species tested within the C. acutatum complex, C. acutatum s.s., C. simmondsii and Colletotrichum sp.2 included one or more isolates originating from non-strawberry hosts that showed medium level of aggressiveness with fruit disease index ranging from 1.17 to 2.08. Whereas, C. lupini (0.08-075), C. phormii (0.58), C. salicis (0.17-0.67), and C. rhombiforme (0.67) along with Colletotrichum sp.1 (0.33) isolates originating from various hosts other than strawberry were much less aggressive as reflected by the fruit disease index. The C. gloeosporioides s.l. isolates tested showed a fruit disease index ranging from 1.50 to 2.50 (Table 3).
In the in vitro assays, anthracnose fruit rot symptoms were observed (e.g. Fig 3B) for various isolates tested with different levels of aggressiveness, as shown by the disease index ranging from 0.08 to 3.0 ( Table 3). The variation in aggressiveness among different isolates was clearly reflected by the differences in incidence which ranged from 8.33 to 100% with only 4 out of 38 isolates showing 91.7 to 100% as well as the lesion type which ranged from 0.1 to 3.0 (S1 Table). When lesion morphology was evaluated, different kinds of lesions could be distinguished on fruits, ranging from brown ones containing orange drops of conidia to those entirely covered with aerial mycelium, with different lesion size. C. nymphaeae CSL899 was the most aggressive on strawberry fruits with the highest disease index (3.0, corresponding to symptoms covering more than 66% of fruit surface).
In the plant assays, varying degrees of crown rot symptoms were recorded 24 d.a.i, as reflected by the disease index range shown in Table 3. Symptom severity was generally low, with no isolate scoring higher than 2 (wilting and collapse of plant). Among the three species identified from UK strawberry production systems, C. fioriniae isolates originating from strawberry showed a higher range of disease index (0.72-1.00) compared to C. nymphaeae (0.5-0.83) and C. godetiae (0.39-0.67). The C. gloeosporioides s.l. isolate CSL 311 (C. fruticola from strawberry in USA) showed the highest disease index (1.6), this isolate was also amongst the most aggressive on fruit (Table 3). Colletotrichum isolates were recovered from all crowns showing symptoms.

Discussion
The UK strawberry industry has expanded rapidly in recent years, and this appears to correlate with increasing losses attributed to anthracnose caused by Colletotrichum spp. [6]. This study provides the first molecular characterization of C. acutatum sensu lato diversity related to strawberry production in the UK, combined with pathogenic characterization. A collection of 148 isolates representative of UK and global diversity of C. acutatum s.l. populations has been assembled. The isolates were chosen based on host association, geographic distribution, phylogenetic relationships and biological diversity.
On the basis of four sequence loci (ITS, TUB, GAPDH, and MAT1-2), the C. acutatum sensu lato isolates were assigned to three newly designated species C. nymphaeae, C. godetiae and C. fioriniae following a recent taxonomic re-assessment [18]. According to available literature, C. nymphaeae is the most common and C. godetiae is also often reported in European and American strawberry fields [6]. These two species were also the most representative in our dataset of isolates related to strawberry in the UK. C. fioriniae has a worldwide distribution and is common on strawberry but only a few isolates were identified in our collection, and this group was not commonly present in the fields in the UK. C. simmondsii, C. acutatum sensu stricto, C. salicis and C. miyabeana are common on strawberry in Oceania and have only been found sporadically in Europe. Isolates belonging to these species have not been detected on strawberry in the UK. The variability observed within the UK C. acutatum sensu lato species fits in part with previous reports of C. acutatum on strawberry within specific geographic regions. For example, in France, Israel, Bulgaria and Spain, the majority of strawberry anthracnose pathogen isolates clustered in the same species C. nymphaeae, and almost no intra-specific diversity was observed within each country [2][3][4][5]. A different situation has been observed on Belgian isolates, where the population represented: 33% isolates belonging to C. nymphaeae, 5% C. fioriniae, 50% C. godetiae, 3% C. acutatum s.s. and 6% C. salicis. A possible explanation to C. acutatum s.l. status in the UK might be recent introduction (late 70s) from a limited number of sources. The reason for the differences in the occurrence of various Colletotrichum species associated with strawberry production in different geographic locations still remains unclear, but the source of importation of the planting material and local trade have been heavily implicated [4,7]. The pathogenicity assays used in this work are based on a study in Belgium [7] in view of the similar molecular diversity of the anthracnose pathogen populations associated with strawberry production. These assays with the isolates representing the molecular diversity not only revealed variability in aggressiveness in different species described within C. acutatum s.l., but also complex patterns both between and within the species. For example, based on isolates originating from strawberry, C. fioriniae and C. nymphaeae appear equally aggressive on fruits with C. nymphaeae isolates indicating a degree of host-preference. Both C. fioriniae and C. godetiae included isolates originating from other hosts that showed comparable levels of aggressiveness to isolates from strawberry. Similar situation was observed with at least some nonstrawberry isolates belonging to species such as C. acutatum s.s. and C. simmondsii. Furthermore, at least one C. godetiae isolate from strawberry was much less aggressive compared to others. These patterns suggest that some Colletotrichum species such as C. fioriniae and C. godetiae include populations that are capable of infecting a wider range of hosts, also influenced by environmental conditions. Further studies using a wider set of isolates of these three species and appropriate pathological and biological assays are required to gain additional insights into the evolution of pathogenicity in relation to field symptoms as well as any differential responses to host varieties and fungicides locally used in the UK strawberry production systems.
The study has highlighted the genetic and pathogenic heterogeneity of the introduced anthracnose pathogen populations belonging to three different Colletotrichum species emphasising the need for effective phytosanitary procedures linked to pathogen monitoring and characterisation to generally limit the entry of non-native pathogens. This also underlines the requirement of reliable and rapid diagnostic tools for further research and application in strawberry anthracnose management. The recent release of a whole genome sequence of C. fioriniae isolated from strawberry [28] along with the newly characterised isolates, based on multi-locus sequence and aggressiveness information reported here, represents a useful platform for further research into the genetic basis of C. acutatum s.l.-strawberry interactions.
Supporting Information S1 Table. Variability in aggressiveness of Colletotrichum species isolates on strawberry fruits and plants. a 0, no visible lesions; 1, lesions on less than 33% of fruit surface; 2, lesions covering 33-66% of fruit surface; and 3, lesions covering more than 66% of fruit surface. b no lesions; 1, crown tissues discoloration but no wilting or collapse; 2, wilting or collapse of part of the plant; and 3, plant death. (XLSX)