Current Status of Soybean Anthracnose Associated with Colletotrichum truncatum in Brazil and Argentina

Brazil and Argentina have a combined soybean area of 53.6 million hectares, which accounts for over half of the total global production. The soybean crop in South America extends from latitude 8–10° S to 32–36° S. Such a vast, almost contiguous area imposes a serious sanitary risk to the crop. Currently, the prevalence of anthracnose is increasing, with recurring reports of severe epidemics and expressive yield losses. Soybean anthracnose is mainly associated with Colletotrichum truncatum, although other Colletotrichum species have also been reported as causal agents of this disease. Knowledge about the morphological, cultural, and molecular variability of C. truncatum in South America is crucial for disease management. Here, we present data on the molecular, morphological, biological, cultural, and pathogenicity of C. truncatum isolates collected in Brazil and Argentina. Light microscopy and randomly-amplified polymorphic DNA (RAPD) analysis were used for estimating the variability of isolates. Colletotrichum truncatum displayed three types of conidiogenesis, viz. conidial formation from conidiogenous cells on hyphal extremities, in conidiomas in acervuli, and directly from fertile setae (a mechanism yet-unreported for C. truncatum). RAPD profiling was effective in revealing the genetic diversity among C. truncatum isolates. The intra-group similarity was greater among the Argentinian isolates when compared to the Brazilian group. Furthermore, the results indicated a strong correlation between geographical origin and molecular grouping, with the exclusive or semi-exclusive assembling of Brazilian and Argentinian isolates in distinct clades. Finally, a preliminary account of the reaction of soybean accessions to C. truncatum is also included.

Soybean anthracnose was first reported in Brazil in 1961 in the southern subtropical state of Rio Grande do Sul (lat. 28-31 • S) and has been prevalent in the midwestern Savannah-like "Cerrado" region (lat. 8-19 • S) since the 1980s [14], often causing severe damage [6]. With the introduction of the soybean rust pathogen (Phakopsora pachyrhizi Sydow) to Brazil in 2001, soybean crops began to be routinely sprayed, with anthracnose persisting as a secondary, sporadic disease. However, in the 2010s soybean anthracnose re-emerged in Brazil [6,15]. The current reports of significant losses may be associated with the pathogen's intrinsic (inter-specific or otherwise) variability, climatic, and cultural factors, as well as due to changes in the genetic makeup of the leading commercial soybean varieties.
Re-emergence of anthracnose in Brazil prompted several efforts to elucidate the nature and distribution of the causal agent(s) [11,12,15,16]. Rogério et al. [16] studied a group of Brazilian isolates collected from 1992 to 2007, initially finding only isolates of C. truncatum. However, C. cliviae was later reported as a soybean anthracnose agent in the northern states of Mato Grosso [11] and Tocantins [12], which indicated a more complex nature of this disease in the country. Most recently, Rogério et al. [15] reported that C. truncatum populations of two north-central Brazilian states were grouped into three genetic clusters, with no evidence of intraregional gene flow.
The present study provides additional information on the Colletotrichum-soybean pathosystem in South America, based on a pool of isolates collected more recently from an extensive area. We report morphometrical and molecular analyses on isolates obtained from symptomatic plants in soybean-growing fields in Brazil and Argentina encompassing parallels 8 • S to 32 • S. The current study also illustrates in detail morphological features of the C. truncatum isolates from soybean and reports the presence of distinct levels of reaction to this disease in G. max germplasm. Potentially useful sources of genetic resistance in the G. max germplasm, for genetic control of anthracnose, are also presented.

Results
Collection sites are depicted in Figure 1 and the C. truncatum isolates are listed in Table 1. Sequencing analysis of the CHS-1 (chitin-synthase gene 1) region indicated high identity levels (ranging from 97% to 100%) of a subset of 18 representative isolates with the C. truncatum reference isolate CTM33 ID:gb/KC109571.1.

Morphological and Culture Observations
Stroma, setae, acervuli, conidiophores, conidiogenous cells, conidia and appressoria of a representative subset of C. truncatum isolates associated with soybean anthracnose are depicted in  Figure 3C). The colony growth rate varied as recorded in Table 1. The colony color ranged from light to dark grey with spore masses orange or beige/orange. Some isolates, especially from Argentina, were orange and roseate (Table 1). Colony texture varied from non-cottony to cottony. Slow, intermediate, and fast growth rates were recorded in the collection, corresponding to seven-day colony diameters of 3.0-5.0; 5.1-6.5; and 8.8-9.0 cm, respectively.

Types of Conidiogenesis Recorded in South American C. truncatum Isolates
Three types of conidiogenesis were observed: (i) Conidia were frequently produced in phialidic conidiogenous cells at the extremities of hyphae; (ii) conidia were also frequently produced in phialidic conidiogenous cells in setose acervular conidiomata; and, (iii) conidia were sometimes produced in fertile setae, differentiated at the distal end, forming conidiogenous cells. Although setae were observed in all colonies, production of conidia from fertile setae was isolate-dependent. Production of conidia in fertile setae by C. truncatum is recorded for the first time. The three types of conidiogenesis are depicted in Figures 2-6. Examples of conidiogenesis in setae are illustrated in detail in Figure 7.

Population Structure of C. Truncatum in South America
Variability of the 54 isolates of C. truncatum was analyzed based on 104 informative randomly-amplified polymorphic DNA (RAPD) amplicons, obtained with 12 Operon primers. Primers OPD-3 and OPG-17 produced the greatest numbers of polymorphic bands (12 each). The number and sizes of amplicons per primer are presented in Table 3. Genetic distances among isolates, based on Jaccard's similarity coefficients, varied from 16% to 100% (Supplemental Table S1). Three main groups, with subgroups, were detected, as illustrated in the dendrogram of

Population Structure of C. Truncatum in South America
Variability of the 54 isolates of C. truncatum was analyzed based on 104 informative randomlyamplified polymorphic DNA (RAPD) amplicons, obtained with 12 Operon primers. Primers OPD-3 and OPG-17 produced the greatest numbers of polymorphic bands (12 each). The number and sizes of amplicons per primer are presented in Table 3. Genetic distances among isolates, based on Jaccard's similarity coefficients, varied from 16% to 100% (Supplemental Table 1). Three main groups, with subgroups, were detected, as illustrated in the dendrogram of Figure 8. Group I was composed of seven isolates, two from Argentina (11.2AR and 11.1AR), two from the Southern Brazilian state of RS (11BMM and 11B2MM), and three from the central and northern states of GO and MA (10.1MM, 2172MM, and 2171MM). Group II was composed of five Brazilian and Argentinian isolates (17.4MM, 17.3MM, 17.1MM, 5.1AR, and 5.3AR). The largest group was Group III, with 42 isolates, subdivided into five subgroups, discriminating according to geographic origin: The genetic structure of the collection of C. truncatum isolates is also depicted in graphic form in Figure 9, where the grouping of isolates broadly parallels with corresponding geographic origin. Here, greater genetic distances are observed among isolates separated by largest greatest differences in latitude, whereas shorter genetic distances are seen among isolates obtained in similar latitudes.  Table 1.
The genetic structure of the collection of C. truncatum isolates is also depicted in graphic form in Figure 9, where the grouping of isolates broadly parallels with corresponding geographic origin. Here, greater genetic distances are observed among isolates separated by largest greatest differences in latitude, whereas shorter genetic distances are seen among isolates obtained in similar latitudes.  Table 1.   Table 1.

Reaction of Soybean Commercial Cultivars and Breeding Lines to Anthracnose
Reactions among soybean accessions to inoculation with a mixture of anthracnose-causing Colletotrichum isolates varied extensively. Analysis of variance of each separate assay indicated a significant accession × assay interaction (p > 0.05) for both variables, so the results are presented separately for each assay. Nevertheless, reactions were stable across assays (Table 4). Mock-inoculated plants showed only minor infection levels, which can be ascribed to either cross-contamination during the inoculation process or residual seed contamination by the pathogen. No accession was immune to the mixture of isolates. BRS Pintado and W877 consistently displayed the best performance against the Colletotrichum mixture, with significantly lower incidences of symptomatic cotyledons and stems. The remaining accessions either expressed some degree of partial resistance to cotyledon infection or to stem infection, or contrarily, presented significant higher disease levels on both cotyledons and stems (e.g., W891). Interestingly, some genetic materials with superior levels of stem resistance were present in the group of those most susceptible to cotyledon infection (e.g., W875, W731, W787, and W712). Therefore, it is possible that the genetic resistance at cotyledons and stems might be under control of genetically independent mechanisms.

Discussion
Morphological and culture observations are in agreement with previous observations of C. truncatum [16][17][18]. Variation of the cultural characteristics (colony color and growth rate) displayed in Table 1 were also noted by Rogério et al. [16] for another subset of C. truncatum isolates associated with soybean anthracnose in Brazil, including a minority of isolates exhibiting orange-colored colonies. Rogério et al. [16] also described variation in colony growth rates among Brazilian C. truncatum from soybean, with half of the isolates considered slow-growing (4.2-4.8 cm, in seven-day-old cultures) and half as fast-growing (5.2-6.2). Even so, colony growth rate of even the fast-growing isolates was inferior to that of the novel soybean anthracnose agent in Brazil, C. cliviicola (= C. cliviae), as recorded by Dias et al. [12]. Morphometrics of conidia and appressoria of the South American C. truncatum isolates also were within the range of 19-26.5 × 3.5-4.5 µm reported by Sutton [17], and in agreement with the epitype description by Damn et al. [18] of 21.8 × 3.8 µm. Some isolates presented chlamydospores and microsclerotia in PDA, as observed in previous studies [19,20], although this has not been mentioned in the epityfication of C. truncatum [8]. Production of conidia in fertile setae by C. truncatum is recorded here for the first time and implies that the setae may be functionally equivalent to conidiophores. The feature of spore-producing setae has been reported in other species of Colletotrichum [21,22].
Our results on the population structure of C. truncatum isolates from South America were based on a broad collection of samples, from widely different growing areas and environmental conditions. Nevertheless, the conclusions corroborate recently published research using microsatellite markers that also showed evidence of close identity among C. truncatum isolates from the same geographical regions. Rogério et al. [15] studied a collection of C. truncatum from a more geographically restricted soybean-growing region, reporting that fungal populations in the north-central states of GO and MT could be grouped in three genetic clusters, conserving syntopy and with no evidence of intraregional gene flow. A greater similarity was found among Argentinian isolates (c. 80%), much higher than the similarity indexes of the Brazilian isolates. A higher magnitude of the genetic diversity among Brazilian isolates was somewhat expected, given the greater geographic breadth of the collection of isolates in the country (Figure 1), which included growing regions with widely different soybean varieties and climatic conditions. Soybean cultivars are distributed in accordance with the latitude because the crop cycle and blooming of a given cultivar are dependent on the photoperiod. Thus, the genetic variability of the host cultivars could be reflected in the variability of the C. truncatum isolates across distinct Brazilian regions. On the other hand, the soybean cultivars cultivated in the Provinces of Entre Ríos, Santa Fe, and Córdoba (with similar latitudes) are presumably more genetically related, inducing less variability in the C. truncatum populations from those regions. The presence of distinct levels of reaction in G. max germplasm to a mixture of C. truncatum isolates, indicated the potential presence of large-spectrum sources of genetic resistance to this pathogen. Sources of resistance with this type of reaction are of extreme value from the breeding for resistance standpoint, especially in a scenario of high intra-specific pathogen variability such the one observed in Brazil and Argentina.
In summary, the RAPD analysis efficiently described the genetic diversity of the C. truncatum populations from a wide region in the two major soybean-producing countries in South America. Specifically, this molecular marker system was able to establish the degree of similarity between and within populations from either Argentina or Brazil, indicating greater levels of similarity among the Argentinian populations. The larger diversity among the Brazilian samples is probably related to the wider sampling area in the country (which includes broader variations of the physical environment, latitudes, and host genetic makeup). In addition, potential sources of genetic resistance were identified in the G. max germplasm making them useful for genetic control of anthracnose.

Origin of Isolates
Fifty-four isolates were collected from soybean plants exhibiting anthracnose symptoms on leaves, stems petioles and pods from geographically distinct commercial fields in Brazil and Argentina in 2011/2012 (Table 1). Twenty-one isolates were collected in the Brazilian midwestern and northern states of Mato Grosso, Mato Grosso do Sul, Goiás, and Maranhão and three were collected in the southern Brazilian state of Rio Grande do Sul. Thirty additional isolates were collected from the Provinces of Entre-Ríos, Córdoba, and Santa Fe in Argentina (Figure 1). All isolates were obtained from symptomatic soybean leaves, stems, petioles, and pods. Prior to the studies, monosporic cultures were obtained for all isolates, which were maintained in sterile water, at room temperature, in 5 mL Eppendorf tubes, according to Castellani [23].

Morphological and Cultural Studies
Isolates were initially characterized via morphologic and morphometric traits according to Sutton [17] and Hyde et al. [8]. The cultural patterns of seven-day-old colonies were observed in potato dextrose agar (PDA) culture medium, incubated at 25 • C ± 2 • C, 12 h photoperiod. Colony color was annotated by visual observation and mycelium growth rate by measurements taken with a graduated ruler from the seven-day-old cultures. In addition, five isolates were studied in more detail by light microscopy. For these isolates, conidium and appressorium shape, length, and width, shape of conidiogenous cells and presence of fertile setae were noted. The formation of appressoria was induced as described by Cai et al. [24]. Microscopic measurements were carried out in semi-permanent lactoglycerol slides, and photodocumented by light microscopy coupled to an electronic photoimaging device (Leica DM 2500 with camera Leica DFC 490). Measurements of fungal structures were conducted using Leica Qwin software.

Production of Mycelium, Extraction, and Quantification of DNA
Mycelium for DNA extraction was produced by adding four mycelial disks from seven-day-old PDA cultures of each isolate to 250 mL Erlenmeyer flasks containing potato-dextrose broth, amended with 0.02 g L −1 of chloramphenicol. Liquid cultures were shaken for 10 days, at 25 • C, 150 rpm. Mycelium was treated with Tris EDTA buffer at 4 • C, and maintained at −20 • C prior to DNA extraction. Before extraction, the mycelium was dried on sterile filter paper, and circa 100 mg was macerated into fine powder in liquid N 2 using microcentrifuge tube pestle, and transferred to Eppendorf tubes. DNA was extracted following a modified CTAB 2× protocol, at pH 8.0, followed by two rounds of chloroform/isoamyl alcohol extraction and a final precipitation step with cold isopropanol [25]. DNA was resuspended in TE buffer and quantified on a Nanodrop ND-1000 spectrophotometer (Thermo Scientific, Wilmington, DE, USA) at 230 nm (purity was verified by the ratio of the absorbance values 230/280 nm greater than 1.8). DNA integrity was assessed by 1% agarose gel electrophoresis with TBE (Tris-HCL EDTA-boric acid) buffer, at 80 V for 20 min, using rphological and Cultural Studies olates were initially characterized via morphologic and morphometric traits according to [17] and Hyde et al. [8]. The cultural patterns of seven-day-old colonies were observed in dextrose agar (PDA) culture medium, incubated at 25 o C ± 2 °C, 12 h photoperiod. Colony as annotated by visual observation and mycelium growth rate by measurements taken with uated ruler from the seven-day-old cultures. In addition, five isolates were studied in more y light microscopy. For these isolates, conidium and appressorium shape, length, and width, of conidiogenous cells and presence of fertile setae were noted. The formation of appressoria duced as described by Cai et al. [24]. Microscopic measurements were carried out in seminent lactoglycerol slides, and photodocumented by light microscopy coupled to an electronic maging device (Leica DM 2500 with camera Leica DFC 490). Measurements of fungal res were conducted using Leica Qwin software.
duction of Mycelium, Extraction, and Quantification of DNA ycelium for DNA extraction was produced by adding four mycelial disks from seven-day-A cultures of each isolate to 250 mL Erlenmeyer flasks containing potato-dextrose broth, ed with 0.02 g L -1 of chloramphenicol. Liquid cultures were shaken for 10 days, at 25 °C, 150 ycelium was treated with Tris EDTA buffer at 4 °C, and maintained at −20 °C prior to DNA ion. Before extraction, the mycelium was dried on sterile filter paper, and circa 100 mg was ted into fine powder in liquid N2 using microcentrifuge tube pestle, and transferred to dorf tubes. DNA was extracted following a modified CTAB 2× protocol, at pH 8.0, followed rounds of chloroform/isoamyl alcohol extraction and a final precipitation step with cold panol [25]. DNA was resuspended in TE buffer and quantified on a Nanodrop ND-1000 o Scientific, Waltham, MA, USA) at 230 nm (purity was verified by the ratio of the ance values 230/280 nm greater than 1.8). DNA integrity was assessed by 1% agarose gel phoresis with TBE (Tris-HCL EDTA-boric acid) buffer, at 80 V for 20 min, using ʎ DNA concentrations (Invitrogen) for standards [26]. The DNA samples were diluted to 3 ng.µL -1 -pure water for later PCR reactions.  (Table 3). Reactions were performed on a final DNA known concentrations (Invitrogen) for standards [26]. The DNA samples were diluted to 3 ng.µL −1 in ultra-pure water for later PCR reactions.

Diversity of C. truncatum Populations Based on RAPD Profiles
Genetic diversity of the C. truncatum collection was examined via RAPD. Firstly, a group of 12 primers (Operon Technologies, Alameda, CA, USA) which produced clear, consistent, polymorphic bands were selected from a larger group of 64 primers, using a six-isolate subset of the collection, in duplicate. Subsequently, PCR reactions of all 54 isolates were individually performed with the 12 primers with the most stable polymorphic patterns (Table 3). Reactions were performed on a final volume of 13 µL consisting of 3.42 µL milli-Q water, 1.3 µL PCR buffer 10×, 1.04 µL de Bovine Serum Albumin (BSA), 1.04 µL dNTPs, 0.20 µL Taq DNA polymerase (Invitrogen, Carlsbad, CA, USA), 3 µL primer (10 ng/µl), and 3 µL of DNA (3 ng/ µL). PCR was conducted on a GeneAmp ® System 9700 thermocycler with reaction conditions consisting of an initial 5 min denaturation at 92 • C, followed by 40 cycles (1 min denaturation at 92 • C, 1 min annealing at 35 • C, 2.5 min extension at 72 • C), and a 5 min final extension at 72 • C. Blue Juice ® gel loading buffer (Invitrogen, Carlsbad, CA, USA) was added to the PCR products (2.0 µL per sample) and samples electrophoresed in 1.5% agarose gels, in 1× TBE buffer, at 160 V for 90 min. Gels were stained with 1.5% ethidium bromide and photographed under UV light. Amplicon sizes were estimated using High DNA Mass Ladder 1KB (Gibco BRL, Gaithersburg, MD, USA).

Data Analysis
Similarity and cluster analyses were performed with the Numerical Taxonomy and Multivariate Analysis System software (NTSYS) [27]. DNA profiles through the annotation of the presence or absence of a given amplicon (scored as 1 or 0), generated a binary matrix for analysis with Jaccard's coefficient (Supplemental Table S1). Doubtful or lacking bands were assigned as 9. Genetic relations among isolates were inferred from a dendrogram (built with basis on the similarity matrix) using the unweighted pair group method using the arithmetic average (UPGMA). In addition, a principal component analysis (PCA) was performed using the NTSYS software.

Reaction of Commercial Cultivars and Breeding Lines of Soybean to Anthracnose
Four cultivated varieties: TMG 115, TMG 127 (TMG), P98Y11 (Pioneer Hi-Bred International, Inc., Johnston, IA, USA), and BRS Pintado (Embrapa, Brasilia, Brazil), and 12 advanced breeding lines: W877, W811, W791, W842, W828, W801, W875, W810, W731, W787, W712, and W891 (Wehrman Agricola Ltd., Goias, Brazil) were evaluated against a pool of Colletotrichum isolates in order to assess the broad genotype reaction to anthracnose-causing agents. Experiments were conducted in a climatized greenhouse, using 128-cell polystyrene trays filled with Basaplant substrate, 16 soybean genotypes per tray, four replicates (each tray contained one replicate), and experimental units of eight plantlets, in a randomized block design. Ten-day old plantlets (stage V1/V2) were sprayed with a mixed inoculum of 10 5 conidia per mL −1 at dusk, maintained for 24 h in a humid chamber, consisting of a translucid plastic cover enclosing the trays, sprayed internally with tap water, and then transferred to greenhouse benches. Specific humidity was not measured, but condensed water droplets were present in the internal plastic walls after 24 h. Aggressiveness was evaluated at ten days after inoculation by annotating the percentage of anthracnose symptoms on cotyledons and stems (epicotyls). Experiments were repeated twice during the 2012/2013 planting season. Controls were mock-inoculated with distilled water. Data were analyzed by ANOVA and means were separated using the Scott-Knott test (p ≤ 0.05).

Concluding Remarks
Soybean anthracnose has re-emerged as a yield-limiting factor for soybean production in South America, the world's largest growing region, which has prompted several studies on the nature of the causal agents. Currently, up to nine or ten Colletotrichum species (if Damm et al. [13] identification of LFN 009 [11] as C. sojae is taken into account) have been associated with this disease worldwide. Nevertheless, among all the taxa of the causal agents, C. truncatum seems to be the most prevalent over a wide geographical region spanning from parallel 8 • S in Brazil to parallel 32 • S in Argentina. Morphology of the asexual state is essential for classification in the genus Colletotrichum [28], and morphological features are still useful for discrimination of the major soybean-infecting species complexes. For example, in the C. truncatum complex conidia are falcate, while in the C. orchidearum complex (where C. cliviicola belongs), conidia are cylindrical. Here, we described the morphological and cultural traits of C. truncatum isolates from Brazil and Argentina. Three types of conidiogenesis in the C. truncatum were recorded (one for the first time) and illustrated, which may be a trait of use in future diversity studies. The current study also examined the variability of Brazilian and Argentinian C. truncatum populations, and its correlation with the geographical origin of the isolates, in the broad sense. The RAPD marker system effectively assembled Brazilian and Argentinian isolates in distinct (exclusive or semi-exclusive) clades revealing greater variation among the Brazilian isolates, probably due to the collection over a wider area. A greater degree of intra-group similarity was established among the Argentinian isolates and much more diversity was found among the Brazilian isolates. The diversity among the C. truncatum isolates from different regions is probably related to the variation in the physical environment and in the host genetics. Conversely, the close identity found among C. truncatum isolates coming from the same geographical region is most likely related to the lack of sexual morph in this species, and conforms to an independent study carried out by Rogério et al. [15]. Finally, our assays revealed that the response of soybean accessions against anthracnose varies extensively, and that sources of partial resistance do exist in the G. max germplasm.