Characterization of Colletotrichum Isolates from Strawberry and Other Hosts with Reference to Cross-Inoculation Potential

Colletotrichum is an important phytopathogenic fungus that causes anthracnose disease in diverse agronomically important tropical food crops. Accurate pathogen identification is critical for early diagnosis and efficient management of anthracnose. ITS is not a reliable marker for this fungal genus due to its failure to phylogenetically resolve cryptic species. In this study, 36 Colletotrichum isolates belonging to the Acutatum, Boninense and Gloeosporioides species complexes were characterized using multigene phylogenetic analyses, morphology and pathogenicity assays. Additionally, the cross-inoculation potential of a representative subset of isolates was evaluated revealing that cross-infection potential is possible among the isolates belonging to the same species complex.


Introduction
Colletotrichum (Glomerellaceae, Glomerellales, Sordariomycetes) is a ubiquitous and important plant pathogenic filamentous fungal genus [1]. Colletotrichum species have high genetic variability and diverse survival lifestyles according to their hosts and the environment [2][3][4]. Members of the species can exhibit endophytic, hemibiotrophic or necrotrophic lifestyles on their host plants [2,5]. Hemibiotrophy is a common mode of survival for this species, allowing an initial interaction with the host plant before infection takes place [6,7]. As a hemibiotroph, a quiescent infection is initiated via melanized appressoria that penetrate the host tissues to form infection hyphae, eventually leading to necrotrophy. Pathogenic Colletotrichum species usually exhibit hemibiotrophic and necrotrophic modes of survival and infection.
This genus is considered a "catalogue of confusion" [8], comprising more than 248 accepted species names [9]. Previously, the Colletotrichum species were named based on hostassociation, which created a lot of nomenclatural ambiguities. Due to the presence of morphologically similar cryptic species within each species complex, genealogical concordance for phylogenetic species recognition [10] is an important approach applied to designate and characterize a Colletotrichum species. Colletotrichum taxonomy has undergone multiple amendments based on the use of multi-locus molecular phylogenetic analyses [11][12][13][14][15][16][17][18]. Along with the genetic diversity of a fungal population, it is also fundamental to assess their pathogenic potential and epidemiology to understand completely the pathogenic nature of the fungi associated with a certain crop. Thus, the use of DNA-sequence data is vital in plant pathology, as it assists in monitoring plant health, pathogen risk assessment and devising pathogen management practices [19].
The Colletotrichum isolates used in this study were originally characterized based on the banding patterns obtained from RFLPs (restriction fragment length polymorphisms) and AP-PCRs (arbitrarily primed polymerase chain reactions) [20][21][22]. However, considering the recent nomenclatural updates in Colletotrichum taxonomy, in this study, we utilized current

C. godetiae ALM-KSH10
Acutatum cottony, grey aerial mycelium; reverse dark grey to pale orange in center complex, and three Colletotrichum isolates were assigned to the C. gloeosporioides species complex. One isolate originally isolated from passiflora in Colombia, was found to be related to C. sydowii, which is a singleton species with limited reference sequence data. Thus, the identity of PASS35 according to the percent sequence identity of the ITS, act, chs1, gapdh, his3 and tub2 sequences corresponds to the type strain sequences of C. sydowii. The phylogenies of the three isolates belonging to the C. gloeosporioides species complex have been described in the supplementary information (Supplementary Figure S1).

Phylogenetic Analyses of the Isolates Belonging to the C. acutatum Species Complex
A concatenated dataset of six genes (ITS, act, chs1, gapdh, his3 and tub2) was used for the phylogenetic analyses of 24 C. acutatum species complex members used in this study. Monilochaetes infuscans (CBS 869.96) was used as the outgroup, and 65 species in the Acutatum complex were used as a reference in the analysis. The multigene sequence alignment contained 2173 characters [ITS: 1-554; act: 555-817; chs1: 818-1094; gapdh: 1095-1368; his3: 1369-1755; tub2: 1756-2173], including gaps. Sixty ambiguous characters were excluded from the alignment, and, of the remaining 2113 included characters, 1431 characters were constant; 283 variable characters were parsimony-uninformative; and 399 characters were parsimony-informative. The parsimony analysis yielded 147 trees that were equally the most parsimonious; the topology of one is shown in Figure 1 [tree length (TL) = 1285, consistency index (CI) = 0.661, retention index (RI) = 0.905, rescaled consistency index (RC) = 0.598, homoplasy index (HI) = 0.339]. The Bayesian analysis of the combined alignment lasted 5,000,000 generations, and the resulting trees were used to calculate the posterior probabilities. The bootstrap support values of the MP analysis (MP > 50%), ML analysis (ML > 50%) and the BI posterior probabilities (PP > 0.90) are depicted at the branch nodes ( Figure 1).
However, the isolates from almond, anemone and strawberry were placed within or next to the clades of C. godetiae, C. fioriniae or C. nymphaeae; that are already reported to possess high genetic diversity [12]. Thus, an additional phylogenetic analysis was performed incorporating more sequences from non-type strains of C. godetiae, C. fioriniae and C. nymphaeae. The multigene sequence alignment contained 2159 characters [ITS: 1-551; act: 552-822; chs1: 823-1096; gapdh: 1097-1361; his3: 1362-1748; tub2: 1749-2159], including gaps. Fifty ambiguous characters were excluded from the alignment and of the remaining 2109 included characters: 1564 characters were constant; 300 variable characters were parsimony-uninformative and 245 characters were parsimony-informative. The parsimony analysis yielded 205 equally most parsimonious trees, the topology of one of which is shown in Figure 2 [tree length (TL) = 703, consistency index (CI) = 0.869, retention index (RI) = 0.984, rescaled consistency index (RC) = 0.855, homoplasy index (HI) = 0.131]. Based on the Bayesian, MP and ML phylogenetic analyses (Figure 2), the isolates from almond were identified as C. godetiae and the isolates from anemone and strawberry were identified as C. nymphaeae. The NCBI-BLAST percent sequence identity of C. godetiae and C. nymphaeae with their closely related taxa is mentioned in the Supplementary Table S1. The number of parsimony-informative characters for each gene dataset is described in Supplementary  Table S2.

Phylogenetic Analyses of Isolates Belonging to the C. boninense Species Complex
A concatenated dataset of six genes (ITS, act, chs1, gapdh, his3 and tub2) was used for phylogenetic analyses of eight C. boninense species complex members in this study. Monilochaetes infuscans (CBS 869.96) was used as the outgroup, and 50 species in the Boni-

Phylogenetic Analyses of Isolates Belonging to the C. boninense Species Complex
A concatenated dataset of six genes (ITS, act, chs1, gapdh, his3 and tub2) was used for phylogenetic analyses of eight C. boninense species complex members in this study. Monilochaetes infuscans (CBS 869.96) was used as the outgroup, and 50 species in the Boninense complex were used as references in the analyses. The multigene sequence alignment contained 2246 characters [ITS: 1-569; act: 570-849; chs1: 850-1126; gapdh: 1127-1427; his3: 1428-1821; tub2: 1822-2246], including gaps. Eighty-one ambiguous characters were excluded from the alignment and of the remaining 2165 included characters: 1383 characters were constant; 216 variable characters were parsimony-uninformative and 566 characters were parsimony-informative. The parsimony analysis yielded 63 equally most parsimonious trees, the topology of one of which is shown in Figure 3  The overall branch support for the observed tree was high. Based on the MP and ML phylogenetic analyses, PASS33, PASS55, PASS62, PASS67 (passiflora) and TOM6 (tamarilla) isolates clustered with the ex-type sequences of C. colombiense CBS129818 (MP 100%, ML 100%, PP 1.0); PASS65 (passiflora) clustered with the ex-type sequences of C. brassicicola CBS101059 (MP 100%, ML 100%, PP 1.0); MAN 76 (mango) and PASS52 (passiflora) clustered with the sequences of C. karsti (MP 100%, ML 100%, PP 1.0). These isolates were also recovered as monophyletic with strong bootstrap support in four out of six individual gene trees (act, chs1, his3 and tub2) (data not shown). The phylogenetic analyses of the Boninense complex identified isolates of C. brassicicola, C. colombiense and C. karsti as pathogenic to non-host strawberry fruits.

Pathogenicity Testing
The pathogenicity of all the Colletotrichum isolates used in this study from their original host of isolation were already studied and published elsewhere [20,21,23]. In this study, we assessed the cross-inoculation potential of the Colletotrichum isolates to non-hosts. Strawberry was used as an experimental non-host, and symptoms were verified for the different representative Colletotrichum isolates (C. godetiae-ALM-KSH10, ALM-BZR82; C. nymphaeae-TUT-5954, STR-101, ANE-NL4, ANE-HV83C; C. tamarilloi-TOM-10, C. acutatum-ANE-NL12, C. fioriniae-APL2, ALM-US4; C. karsti-MAN76, PASS33; C. colombiense-TOM6, C. brassicicola-PASS65, C. sydowii-PASS35, C. gloeosporioides s. l.-Cg-Sc-A1, APL7). The C. gloeosporioides s. l. isolate Litchi-Cg2 could not be assessed for pathogenicity due to inadequate production of conidia. The Colletotrichum species belonging to the Acutatum and Gloeosporioides species complexes exhibited anthracnose symptoms on strawberry fruits after 7 days of inoculation, whereas isolates belonging to the Boninense species complex and C. sydowii did not exhibit any noticeable anthracnose disease symptoms on inoculated fruit. Particularly, C. nymphaeae is an aggressive pathogen of strawberry in Israel, exhibiting 100% disease incidence and 75-100% disease severity in wounded inoculations. In unwounded inoculations, the disease incidence remained low. However, C. nymphaeae caused increased disease incidence and severity in unwounded inoculations as well. Interestingly, the C. acutatum sensu stricto and C. nymphaeae isolates from anemone were able to cross-infect strawberry. Similarly, C. godetiae and C. fioriniae isolates from almond, C. fioriniae isolate from apple and C. tamarilloi isolates from tamarillo were capable of infecting strawberry under wounded, as well as unwounded, conditions. This implies that the members of the Acutatum and Gloeosporioides species complexes used in this study are key pathogens of strawberry fruits and possess the potential for cross-infection. Detailed results of the pathogenicity tests are presented in Figures 4 and 5 and in Supplementary Tables S3 and S4.
The Colletotrichum species belonging to the Acutatum and Gloeosporioides species complexes exhibited anthracnose symptoms on strawberry fruits after 7 days of inoculation, whereas isolates belonging to the Boninense species complex and C. sydowii did not exhibit any noticeable anthracnose disease symptoms on inoculated fruit. Particularly, C. nymphaeae is an aggressive pathogen of strawberry in Israel, exhibiting 100% disease incidence and 75-100% disease severity in wounded inoculations. In unwounded inoculations, the disease incidence remained low. However, C. nymphaeae caused increased disease incidence and severity in unwounded inoculations as well. Interestingly, the C. acutatum sensu stricto and C. nymphaeae isolates from anemone were able to cross-infect strawberry. Similarly, C. godetiae and C. fioriniae isolates from almond, C. fioriniae isolate from apple and C. tamarilloi isolates from tamarillo were capable of infecting strawberry under wounded, as well as unwounded, conditions. This implies that the members of the Acutatum and Gloeosporioides species complexes used in this study are key pathogens of strawberry fruits and possess the potential for cross-infection.

Molecular Phylogeny of Isolates of Colletotrichum
Over the past decade, the phylogeny of the genus Colletotrichum has been in flux [24]. Prior to that, ITS sequencing was used to delineate isolates to the major "species", e.g., C. gloeosporioides, C. acutatum, C. coccodes, etc. [24]. Thus, the single-gene phylogeny based on ITS sequencing did not result in reliable species diagnostics and identification [25]. However, since 2012, the main species clusters have been redefined according to multi-locus sequencing. For example, the Acutatum species complex has been defined according to phylogenetic analyses of the ITS, act, tub2, chs1, gapdh and his3 genes [12]; the Gloeosporioides species complex has been defined according to analyses of up to eight genes [18]; and the Boninense species complex has been defined according to phylogenetic analyses of the ITS, act, tub2, chs1, gapdh, his3 and cal genes [13]. Thus, the multi-locus sequencing of many populations of Colletotrichum has resulted in the renaming of undescribed cryptic species within the systematics of the genus. To date, a comprehensive systematics of the genus Colletotrichum includes a list of 16 species complexes, while an additional 15 novel single-isolate representatives were also described [26].

Previous Characterization of Strains
Past studies have characterized the current Colletotrichum isolate collection from Israel and elsewhere to species based on various molecular techniques. For example, isolates of C. gloeosporioides from avocado and other hosts were characterized according to AP-PCR, A+T-rich mitochondrial DNA, nuclear DNA and rDNA analyses; however, none of these methods were suitable in differentiating between the species from this complex [20]. Similarly, in additional studies, the C. acutatum isolates from various hosts, including anemone, olive and strawberry, were characterized to the specific complex using ITS sequencing, however, under limited conditions that did not coincide with pathogenicity tests [22]. Likewise, isolates within the species complex, Boninense, were originally identified based on ITS and rDNA sequencing [27]. In all of these past studies, the ITS sequence alone was not informative enough and did not delineate isolates to the species level within each Colletotrichum complex [20][21][22].

Colletotrichum of Almond
Under the circumstances, Colletotrichum affecting almond appears to be diverse; however, among the studied cases, the different populations were clustered within the Acutatum species complex. In Israel, the population of Colletotrichum is specific to almond and appears to be clonal and unique; however, ITS sequencing alone was not able to differentiate between the population from the US and those affecting almond from Australia [20]. The Australian populations were distinct from those found in Israel, however, and were more closely related to those affecting almond from the US [28].
As far as pathogenicity is concerned, all of the above populations of Colletotrichum affecting almond were pathogenic to the crop. In Israel, the specificity was demonstrated as follows: isolates that originated from almond were unique, and none of these isolates were detected within the populations affecting avocado, characterized within the Gloeosporioides complex; the anthracnose-affected almond orchards were cultivated adjacent, at a 5-10 m distance, to the affected avocado orchards [21]. Therefore, there is no need to be concerned that cross-inoculation from almond to avocado and vice versa takes place in the field. Thus, management protocols for disease reduction in each crop can be developed individually without any threat of cross-infection taking place between both of these crops.

Colletotrichum Affecting Anemone and Strawberry
Another case study in Israel, concerns the molecular identity of populations of Colletotrichum that affect anemone and strawberry. Anthracnose disease of anemone and strawberry is common, and both crops can be affected annually in Israel [21]. In certain instances, the same farmer can cultivate both crops in close proximity, when the infection takes place simultaneously. In the past, populations of the Acutatum complex were reported as causal agents of both crops, and it was shown that cross-inoculation of certain isolates/populations takes place under controlled conditions [21] and under field conditions (Freeman, pers. comm.).

Pathogenicity
The pathogenicity of Colletotrichum producing anthracnose symptoms can be evaluated on detached or attached plant structures. It has been shown, on the one hand, that detached fruit are not ideal "substrates" for pathogenicity testing; however, on the other hand, postharvest evaluations are conducted primarily on detached fruit, e.g., pathogenicity and virulence on avocado [29]. In this study, the artificial inoculation of Colletotrichum isolates delineated to specific species within each complex were artificially inoculated on strawberry fruit. Reliable results were achieved that differentiated between specific species, indicating that this method is accurate in evaluating the pathogenic and virulence specificity for the isolates tested in this study.
Since the members of the Acutatum and Gloeosporioides species complexes are known pathogens of strawberry, all of the tested representative isolates belonging to these two species complexes were able to cross-infect strawberry fruits in the wounded as well as unwounded inoculations. However, slight differences in disease symptoms and severity were observed among the different isolates, such as those from anemone (ANE-NL4) and strawberry (TUT-5954) that may be attributed to the high genetic diversity within C. nymphaeae. Similar to the present study, C. acutatum and C. fioriniae have been previously reported to be associated with strawberry anthracnose symptoms. This is the first report of C. tamarilloi causing strawberry anthracnose.
The Colletotrichum isolates belonging to the Boninense species complex did not cause anthracnose symptoms in wounded or unwounded inoculations. Thus far, there is one report of the association of C. boninense and C. karsti with strawberry in China [30,31]. However, C. brassicicola, C. colombiense and C. sydowii have not been previously associated with strawberry [32], and, as in this study, C. sydowii was found to be non-pathogenic on detached strawberry fruits. Thus, according to the pathogenicity assays reported in this study, Colletotrichum isolates possess cross-infection potential within their respective species complex.

Colletotrichum Species Affecting Anemone, Strawberry and Almond
Within the Acutatum species complex, C. acutatum, C. fioriniae, C. godetiae, C. nymphaeae and C. simmondsii have been reported as pathogens of almond and strawberry fruit, while C. acutatum, C. fioriniae and C. nymphaeae have been reported on anemone [32]. In this study, a polyphasic approach of multigene phylogenetic analysis, morphological characterization and pathogenicity assays was used to characterize 36 Colletotrichum isolates that were previously identified based on AP-PCR and ITS sequence data alone. Similar to previous studies, C. godetiae was associated with almonds, and C. nymphaeae was associated with anemone and strawberry in Israel. The genealogical concordance was also evaluated using individual gene trees; however, there is a need to improve the phylogenetically informative barcode markers for the Acutatum species complex. The percent parsimony-informative characters for the various genes used for the phylogenetic analysis of the Acutatum species complex is quite low (Supplementary Table S2). In the case of the Gloeosporioides species complex, intergenic regions of the apn2 and MAT1-2-1 gene (ApMat) markers, along with the glutamine synthase (gs) gene, provide an efficient species-level resolution [33,34].

Fungal Isolates, Growth Conditions and Morphological Characterization
The monoconidial Colletotrichum isolates used in this study are listed in Table 2. The Colletotrichum isolates associated with almond, strawberry and anemone were isolated between 1993 and 1995, in Israel [20][21][22]. The isolates associated with passiflora, tamarillo and mango were isolated between 1998 to 2000, in Colombia [27]. The isolates associated with apple, pecan and peach originated in the US [22]. The Colletotrichum isolates were retrieved from the −80 • C facility of the Freeman laboratory at the ARO, revived on potato dextrose agar (PDA, Difco, USA) plates and grown at 25 • C for morphological characterization (colony morphology and growth rate). To increase sporulation, the isolates were grown on modified Mathur's MS semi-selective (M3S) agar medium (per liter composition: 2.5 gm of MgSO 4 ·7H 2 O, 2.7 gm of KH 2 PO 4 , 1 gm of peptone, 1 gm of yeast extract, 10 gm of sucrose and 20 gm of agar) [29] and synthetic nutrient agar (SNA) medium (per liter composition: 1 gm of KH 2 PO 4 , 1 gm of KNO 3 , 0.5 gm of MgSO 4 ·7H 2 O, 0.5 gm of KCl, 0.2 gm of glucose, 0.2 gm of sucrose and 20 gm of agar) [35]. Microscopic slides were prepared in 1% lactic acid or water. For each representative isolate, the shape and size of the conidia and conidiogenous cells were measured using bright field microscopy (Nikon, Japan) and measured via NIS-Elements image analysis software (Nikon). At least 50 measurements were made for the length and width of the conidia. The growth rate was determined by measuring the colony diameter after 7 days (mm/day). Morphological characteristics are listed in Table 1.

Phylogenetic Analyses
The forward and reverse sequences for each gene were checked for quality and assembled using MEGA-X v. 10.1.17 [42]. The different gene regions were concatenated using SequenceMatrix v. 1.7.8 [43]. Ambiguous regions from the multiple sequence alignment were not included in the analyses, and the gaps were considered missing data (N). The six-gene dataset (act, cal, gapdh, his3, ITS and tub2) was used for the phylogenetic analyses of the Colletotrichum isolates belonging to the C. acutatum and C. boninense species complex. For the remaining three isolates belonging to the C. gloeosporioides species complex, the phylogenetic analyses comprised a five-gene dataset (act, chs1, gapdh, ITS and tub2). The reference type-strain sequences for the analyses were retrieved from GenBank [22] and are listed in Supplementary Tables S5-S8. Phylogenetic analyses were conducted using a maximum parsimony (MP), maximum likelihood (ML) and Bayesian inference (BI) methods. The maximum parsimony analysis was conducted using PAUP v. 4.0b10 [44], as detailed earlier in Sharma et al. [29]. The maximum likelihood phylogeny was inferred using RAxML-HPC2 under the GTR-GAMMA model in the CIPRES portal [45,46], and the branch support was evaluated by bootstrap analysis of 100 replicates (-m GTRGAMMA -p 12345 -f a -N 100 -x 12345 -asc-corr lewis). The Bayesian inference of the phylogeny was estimated using MrBayes version 3.2.7a [47] in the CIPRES portal with four MCMC chains of 5,000,000 generations. The sample trees were recorded every 10 3 generations, and 25% of the initial trees were discarded as burn-in. For each gene dataset, a suitable model for the estimation of phylogeny was evaluated, and ML analysis was carried out using MEGA-X and the "one click mode" tree analysis method, available at www.phylogeny.fr, accessed on 27 June 2022 [48] (data not shown). The resulting trees from each analysis were viewed in FigTree v. 1.4.4 [49] and edited in Microsoft PowerPoint 2016.

Pathogenicity Assay
The pathogenic potential of the representative isolates was evaluated by artificial inoculations of unripe strawberry fruit (light green to light red in color), cv. Peles, which is susceptible to anthracnose, that originated from a disease-free organic strawberry farm in the Sharon area of Israel. Prior to inoculation, the calyx and peduncle were trimmed, and the fruits were washed under running water. The conidia were harvested as described [29] and adjusted to 2 × 10 7 conidia/mL. After air drying, the fruits were inoculated with 10 µL of conidial solution at wounded (pin-pricked) and unwounded sites. In the control fruits, sterile saline solution was used for inoculation. The inoculated fruits were maintained in a moist chamber at room temperature (25-27 • C) and observed daily for the appearance of anthracnose symptoms. The disease severity and disease incidence were calculated as described in Sharma et al. [50] and detailed in Supplementary Table S3. The disease score was recorded and indicated in Supplementary Table S4. The statistical significance for the disease severity was calculated according to Tukey's post-hoc test (p < 0.05) (Figure 5b).

Conclusions
ITS sequence data alone is not useful for identification and differentiation of species from the genus Colletotrichum. The derived classification is inaccurate and groups the isolates to a species complex but does not delineate them to species accurately. Currently, multi-locus sequence data has been adopted and is routinely used for delineating and accurately describing members of the Colletotrichum genus to new species within each complex. Additionally, the cross-infection potential of different Colletotrichum species was assessed, which provided insights into their diverse host-range and pathogenic potential on non-hosts. Plant breeders should consider the broad host range and high genetic diversity of the Colletotrichum species when developing disease resistant cultivars.