Molecular characterization of Colletotrichum species causing Begonia anthracnose in Sri Lanka

Anthracnose disease is known to affect many tropical and subtropical fruits, vegetables and also certain cut-flowers and foliage plants. The disease was known to be caused by Colletotrichum gloeosporioides or C. acutatum which are presently accepted as species complexes. The present study was conducted with the main objective of identifying Colletotrichum species causing begonia anthracnose using morphological and molecular data. Begonia is an ornamental foliage plant grown worldwide. Anthracnose symptoms appear in begonia leaves as brownish, irregular, necrotic lesions. Colletotrichum was isolated from three Begonia species showing anthracnose symptoms, collected from three Provinces of Sri Lanka. Thirty isolates were obtained in the study of which 29 formed oblong conidia and the remainder produced falcate conidia. Six randomly selected isolates forming oblong conidia and the isolate with falcate conidia were sequenced for Internal Transcribed Spacer (ITS) and β-tubulin 2 (TUB2) regions. Considering >98% similarity with NCBI GenBank database for both sequences, the isolates with oblong conidia were identified as C. siamense and the isolate with falcate conidia as C. truncatum. Newly generated sequences were subjected to phylogenetic analysis with the closely related extype and authenticated isolates sequences. Phylogenetic analysis confirmed the species as C. siamense and C. truncatum. Koch’s postulates were performed to establish whether the fungi isolated from anthracnose lesions were actually causing anthracnose disease in Begonia leaves. This is the first report of C. siamense causing Begonia anthracnose.


INTRODUCTION
Begonia, belonging to the Family Begoniaceae, is one of the largest angiosperm genera comprising over 1500 species (Hughes and Hollingsworth, 2008). They are native to moist tropical and subtropical climates and cultivated as indoor or outdoor houseplants and as edgings in gardens for their elegant, succulent, often varicolored leaves and waxy flowers.
Begonias are susceptible to several diseases caused by fungi, bacteria, and viruses, as well as abiotic disorders. Among fungal diseases, the grey mould caused by Botrytis cinerea, powdery mildew incited by Erysiphe cichoracearum and Rhizoctonia crown rot are common and important (Benson and Cartwright, 1996). A leaf rust caused by Coleosporium solidaginis has also been reported which produces orange-red pustules on the under surface of leaves resulting in shedding them prematurely. Leaf and stem rot, vein yellowing, and wilting of Begonia elatior hybrids (Begonia x hiemalis), caused by Fusarium foetens, were reported from Japan (Sekine et al., 2008). One of the most common and wide-spread diseases in Begonia is bacterial leaf spot or blight with small blisterlike spots on leaves caused by Xanthomonas campestris pv. begonia (Vauterin et al., 1990). Affected leaves may shed prematurely.
Anthracnose is a common foliage disease in potted begonias and is more serious in nurseries, often resulting in outbreaks (Park et al., 2006). The disease affects mainly the foliage, directly reducing the plant's horticultural value. Anthracnose in Begonia in South Korea was reported to be caused by C. acutatum using RAPD analysis and morphological characteristics (Park et al., 2006). Colletotrichum truncatum was more recently identified as the cause of anthracnose in China using morphological and multigene sequence data (Zhai et al., 2018). Here the anthracnose symptoms appeared as large, sunken and brown lesions, mostly at the leaf margin, with numerous black acervuli on Begonia stems.
In recent years, a considerable progress has been made in delineating the taxonomy of different Colletotrichum species that cause anthracnose. Molecular phylogeny and genomics data have enabled distinguish Colletotrichum species complexes and identify genetically distinct subgroups. Anthracnose disease in Begonia in Sri Lanka has not been subject to any detailed studies. This present study was conducted to characterize the Colletotrichum species causing Begonia anthracnose in Sri Lanka by DNA sequence analysis.

Collection of diseased material
Leaves showing symptoms of anthracnose disease were mostly collected from a nursery of the Royal Botanical Gardens, Peradeniya. Diseased leaves were also collected from Hantana (Kandy District), Kothmale (Nuwara Eliya District) in the Central Province, Kurunegala (Kurunegala District) in the North Western Province and Kegalle (Kegalle District) in the Sabaragamuwa Province of Sri Lanka during the period from July -December 2018. Four different Begonia varieties, Begonia rex var. Rex, B. rex var. Glass begonia, B. peltata (unknown variety) and B. masoniana var. Iron cross were encountered in the study. Diseased leaves from each variety were brought to the Plant Pathology laboratory at the Department of Botany, University of Peradeniya in punch-aerated polyethylene bags within a few hours after collection.

Isolation of Colletotrichum
Small segments (5x5 mm 2 ), cut from the diseased areas of the leaf, were surface sterilized in sodium hypochlorite (Clorox © , USA) for 3 min. After rinsing the segments in sterilized distilled water (SDW), the excess liquid was removed by placing them on sterilized filter papers before transferring aseptically on to Potato Dextrose Agar (PDA) plates (4 segments per plate), supplemented with Tetracycline to suppress bacterial growth. The plates were incubated at 28 -30 o C or 5-7 d to allow fungal growth. A small portion of the mycelium from each colony was mounted in lactophenol on a glass slide and examined under light microscope for the presence or absence of conidia typical to Colletotrichum. The isolates, tentatively recognized as Colletotrichum, were sub-cultured by transferring 6 mm diameter discs of mycelium on to fresh PDA plates and the plates were incubated at 28 -30 o C for 7-14 d.

Preparation of mono-conidial cultures
A suspension of conidia was prepared by scraping mycelia from two weeks old cultures and suspending in sterile distilled water (SDW) in 50 ml flasks. The suspension was filtered through sterile glass wool, the concentration of conidia was adjusted to 1 x 10 6 conidia ml -1 . A loopful of the suspension from each isolate was streaked on tap water agar plates and the plates were incubated at 26-28 o C for 18 h. A single germinated conidium was located by moving the objective lens (x 25) of a light microscope (Olympus BX53 with Olympus DP72 digital camera) across the streak line of the inverted agar plate. A small piece of agar with the germinated conidium was cut and transferred on to fresh PDA plates. The plates were incubated for seven days at 28-30 o C and the isolates were sub-cultured on PDA to be used for morphological and molecular studies.

Morphological study
Two weeks old mono-conidial cultures were used in triplicate to study the colony morphology viz. colony color, texture, pigmentation underneath, the presence or absence of concentric rings, sectoring and reproductive morphology, acervuli or the conidia masses. A drop (20 µl) of suspension of conidia was mounted on a clean slide and, after placing a coverslip, the conidial characteristics such as shape and color of conidia, presence of septa etc. were recorded (Olympus BX53 with Olympus DP72 digital camera and Olympus cellSens software). The dimensions were measured in µm on 50 randomly selected conidia with eyepiece graticule connected to a light microscope. The average length and width were calculated.

DNA extraction
Six isolates, randomly selected from among the 29 isolates that produce oblong conidia, and the single isolate with falcate conidia, were taken for molecular studies. DNA was extracted using Promega Wizard Genomic DNA Purification Kit. Cell lysis solution (300 µl) and the neucleilysis solution (300 µl) were added into each eppendorf tube. The mycelium, scraped from the colonies of each fungal isolate using sterilized micropipette tips, were placed in respective eppendorf tubes. They were vortexed for 2 min vigorously and incubated at 65 in a water bath for 1 h. After cooling down to room temperature, 20 µl of proteinase K solution was added to each tube. The eppendorf tubes were shaken vigorously and incubated at 65 for 1 h 20 µl of protein precipitation solution was added to each eppendorf tube and vortexed vigorously and centrifuged at 14,000 rpm for 5 min. The supernatants were transferred into separate tubes respectively to avoid mycelia. Aliquot (600 µl) of cold isopropanol was added into each tube. They were kept in ice for 5 min and centrifuged at 14,000 rpm for 5 min. The supernatants were discarded and 600 µl of 70% ethanol was added to the pellet. The tubes were centrifuged at 14,000 rpm for 1 min and the supernatants were discarded. The tubes were kept inverted on a sterile filter paper to airdry. Finally, 100 µl of DNA rehydration solution was added to each tube and the DNA samples were stored at -20 o C.
PCR products were separated by electrophoresis in 1% agarose gel and visualized with a UV trans-illuminator. PCR products, confirmed for amplification, were P. Wickramasinghe et al.

Identification of isolates
The species affiliations and identities were determined through similarity-based searches of the NCBI GenBank Database (htttp://www.ncbi.gov). Based on the identifications that resulted from the BLAST search, a combined phylogenetic analysis for ITS and TUB2 was conducted using the authenticated sequences (members of the C. gloeosporioides complex and C. truncatum complex) obtained from the GenBank. Chromatograms of all the sequences were edited and aligned using Bio Edit v7.2.5 software (Hall, 1999) and ClustalX 2.1 (Larkin et al., 2007). Bayesian Inference analysis was performed for the combined matrix. The best fitting substitution model was determined with jModelTest v.2 (Darriba et al., 2012) using the Akaike information criterion (AIC). The nucleotide substitution model Hasegawa-Kishino-Yano (HKY) was selected. Bayesian inference was conducted to obtain posterior probabilities using MrBayes ver. 3.2.6 (Huelsenbeck and Ronquist, 2001;Ronquist and Huelsenbeck, 2003) with 10,000,000 generations Markov chain Monte Carlo (MCMC) chains with a sampling frequency of every 1,000 generations. The initial 25% samples from each run were discarded as burning. A majority rule consensus tree was calculated using the remaining trees to obtain the posterior probabilities for each node. The resulting tree was visualized and edited in FigTree ver. 1.4.3 (http://tree.bio.ed.ac.uk/software/ figtree/, last accessed 2018-12-04). Colletotrichum hippeastri was used as the out group. All newly generated sequences were deposited in GenBank and the accession numbers are given in the Table 1.

Koch's postulates
Begonia leaves showing anthracnose symptoms were examined and the symptoms were recorded. Scrapings obtained from anthracnose lesions were observed under light microscope to see the presence of the identifiable pathogen structures in all diseased leaves. The two pathogens, isolated from diseased tissues and grown in pure culture as described previously, were used for inoculation. For artificial inoculation, uniform-sized, healthy plants of Begonia masoniana var. Iron cross, devoid of any blemishes or visible symptoms of disease, and grown in mediumsized plastic pots without any fungicide treatment were used. A suspension of conidia of each isolate was prepared as described previously by scraping mycelium from 2-3 weeks old cultures of the C. siamense and C. truncatum and the concentration was adjusted to 1 x 10 6 conidia ml -1 . Drops (20 µl) of conidia suspension were applied on to wounded or unwounded sites of leaves (6 drops per leaf) of Begonia. Six replicate leaves in two plants were used for inoculation per isolate and another three leaves were treated with drops of sterile distilled water as controls. Fifty percent of the sites to be inoculated in the leaves were superficially wounded using a fine needle and the rest were left unwounded. Inoculated plants, covered with perforated polythene bags, were incubated at 28 -30 o C and after 24 h the polythene bags were taken off and the pots were transferred to a plant house for further incubation. Inoculated leaves were examined when symptoms appeared and compared with those of the plants used for original isolations. The pathogens were re-isolated from the symptomatic leaves on PDA. The colonies of the isolate were compared with the original isolate used for inoculation for colony and asexual reproductive morphology.

Symptoms
Anthracnose disease was observed in the four Begonia varieties belonging to three species, Begonia rex var. Rex, B. rex var. Glass begonia, B. peltata (unknown variety) and B. masoniana var. Iron cross, encountered in the study. Anthracnose symptoms in foliage varied slightly with the variety (Figure 1). In the var. Iron cross (B. masoniana), small, dark brown and circular lesions appeared initially, scattered over the mature leaves. The lesions gradually enlarged into irregular and large lesions up to about 1-4 cm diameter. The center of the lesions was light brown and the margin of the lesions was dark brown. In B. rex var. Rex, oval shaped or irregular, dark brown and 3 -4 cm diameter lesions with concentric markings appeared towards the periphery of the leaf blade. In some plants, light brown, about 2 cm wide and up to 4 -8 cm long lesions with wavy inner margin were also observed on the margin of the leaf. The outer margin of the lesion was often shredded. In heavily infected leaves, numerous pleomorphic lesions of varied size could be seen. Anthracnose lesions were not observed in younger leaves or any of the petioles.

Morphological characteristics
Thirty Colletotrichum isolates were obtained from the symptomatic Begonia leaves collected in the study. The colony morphology among the 6 isolates forming oblong conidia used for the study varied slightly from each other. Colony color ranged from white to light grey. Most isolates formed cottony, raised colonies with concentric markings and yellowish pink conidial masses. The lower surface of the colony was cream in color ( Figure 2). Conidia were aseptate, hyaline, oblong, and the mean length ranged 12.7-18.1 µm.
The seventh isolate formed whitish, fluffy, raised colonies with yellowish pink center and dark grey to black color, the mycelium in the periphery was flat, consists of 1.5-2 cm wide radial sectors with dense conidia masses of 2-3 mm diameter. The lower surface of the colony was black color. Conidia were aseptate, hyaline and falcate with a prominent area in the center with granulate cytoplasm, the mean length varied from 18.2 -20.1 µm. Setae were straight, dark brown and septate ( Figure  2). From the morphological characteristics of acervuli and conidia (Zhai et al., 2018), the solitary isolate was tentatively identified as C. truncatum.

Identification based on molecular data
The lengths of the sequences that resulted from the PCR reactions for the ITS region were between 580-600 bp, while the TUB2 region was between 500-520 bp. Based on the BLAST search results with authenticated sequences (GenBank, htttp://www.ncbi.gov) that gave > 98% similarity, six isolates (UPBTBCS01, UPBTBCS02, UPBTBCS03, UPBTBCS04, UPBTBCS05, and UPBTBCS06) were identified as Colletotrichum siamense and the isolate with falcate conidia was confirmed as C. truncatum (UPBTBCT01) (Figure 2), based on the sequences of the two gene regions.

Phylogenetic analysis
The phylogenetic tree that resulted from the Baysian analysis is given in the Figure 3. The six isolates, identified as C. siamense during the present study, together with the authenticated ex-type culture of Colletotrichum siamense, have formed a monophyletic group with a posterior probability of 0.99 and is resolved within the CG complex. Similarly, the other isolate, identified as C. truncatum, has also formed a monophyletic group with the authenticated type isolate of C. truncatum with a posterior probability value of 1 as a separate clade (Figure 3).

Koch's postulates
The pathogen with oblong conidia, subsequently identified as C. siamense, was isolated from 29 diseased leaves of begonia indicating its consistent presence in lesions and grown on fresh PDA as pure cultures. C. truncatum with falcate conidia was isolated from only one diseased leaf. Begonia leaves artificially inoculated with C. siamense P. Wickramasinghe et al. Figure 2: A to C -Colletotrichum siamense isolated from Begonia, A -7-day old mono-conidial isolate on PDA, B -Conidia masses produced on culture, C -Conidia. D to F -C. truncatum isolated from Begonia, D -7-day old culture of mono-conidial isolate on PDA, E -Acervuli with black, straight setae projecting out, and F -Conidia. or C. truncatum developed brown color, irregular, necrotic lesions typical to begonia anthracnose observed in originally used begonia leaves for isolation, 7-14 days after inoculation. The leaves that were inoculated without wounding, with either of the two fungi, initiated lesions 2 -3 days later and took a few more days for expansion than those in wound-inoculated leaves. C. truncatum developed necrotic lesions consisting numerous black color acervuli. The lesion development by C. truncatum was much quicker compared to C. siamense showing that C. truncatum was more aggressive on Begonia (Figure 1). There were no symptoms observed in the control plants. The culture plates prepared by re-isolation were morphologically similar to the cultures used for inoculation.

DISCUSSION
Begonias are primarily grown for their decorative foliage that offer different colors and textures and excellent alternatives to some of the flowering potted plants. They are affected by several diseases of those the anthracnose is one of the commonest and most destructive diseases adversely affecting the foliage quality. The main objective of the present study was to determine the Colletotrichum species causing Begonia anthracnose in Sri Lanka. Both morphological characteristics and multigene phylogenetic analyses were utilized for species level identification.
Of a total of 30 isolates collected in the study, 29 produced oblong conidia. Six out of 29 isolates that were subjected to multigene DNA sequence analyses were identified as Colletotrichum siamense belonging to the C. gloeosporioides species complex. C. siamense, was first described as a causal agent of the anthracnose of coffee berries from the Northern Thailand (Prihastuti et al., 2009) and subsequently from anthracnose disease in numerous plant species and fruits. The solitary isolate that produced falcate conidia from anthracnose in Begonia leaves was tentatively identified using conidial and acervuli morphology (Zhai et al., 2018) as C. truncatum. DNA sequence data and phylogenetic analysis confirmed that the isolate was indeed C. truncatum. In a previous study in China, three Colletotrichum isolates from anthracnose in Begonia leaves were identified as C. truncatum based on cultural and morphological characteristics (Zhai et al., 2018) and molecular data using ITS, TUB2 and GAPDH regions (Zhai et al., 2018). C. truncatum belongs to the C. truncatum species complex which includes C. truncatum and three other closely related species. The complex can be distinguished by their curved conidia with truncated base and acute, more strongly curved apex (Damm et al., 2009). C. truncatum was also frequently referred to as C. capsici Table 1: Accession numbers of authenticated sequences of species of the C. gloeosporioides complex and C. truncatum, obtained P. Wickramasinghe et al. Figure 3: Bayesian inference phylogenetic tree for ITS and TUB2 sequences of Colletotrichum isolates of the present study together with ex-type and authentic culture sequences of C. gloeosporioides and C. truncatum species complexes; Bayesian posterior probability values ≥ 0.5 are shown at the nodes. The ex-type cultures are denoted by an *. (Sawant et al. 2012) which is reported as an economically destructive pathogen of many tropical crops including legumes and solanaceous plants (Damm et al., 2009).
The symptoms induced following C. siamense and C. truncatum infections in Begonia leaves showed a clear difference. C. truncatum infection resulted in numerous acervuli in the dotted center of well-developed necrotic lesions. In the present study, C. truncatum was isolated from the anthracnose in Begonia leaves at a very low frequency compared to C. siamense. However, C. truncatum was observed to be more aggressive on Begonia leaves on artificial inoculations than C. siamense. Koch's postulates, performed using C. siamense and C. truncatum, firmly established that both fungi were the pathogens involved in causing anthracnose disease in Begonia. C. truncatum has been previously reported as a pathogen causing Begonia anthracnose (Zhai et al., 2018).
Colletotrichum species are well known pathogens worldwide in a wide variety of crop, ornamental and plantation plants and harvested fresh produce, fruits, vegetables and cut-flowers. They are also ubiquitous asymptomatic foliar endophytes in a vast range of hosts (Osono, 2008). Colletotrichum species appear to live as endophytes quite commonly in the leaves of Begonia species. Population of endophytic fungi has been studied in the leaves of three Begonia species in the Brazilian Atlantic Rainforest, Begonia fischeri, B. olsoniae, and B. venosa. A total of 426 fungal endophytes in 19 genera were isolated in pure culture. Among the isolates, Colletotrichum was the most abundant making 51.6% of the total isolates, followed by Diaporthe (22.5%) (Correia et al., 2018). It has been suggested that endophytic fungi may be latent pathogens, saprophytes and/or mutualists (Saikkonen et al., 2010). There is a possibility that endophytic Colletotrichum species in Begonia may become pathogenic causing anthracnose disease in the leaves. However, the interaction between most plants and endophytes are still undetermined (Rojas et al., 2010).
In summary, the Colletotrichum species causing anthracnose disease in Begonias in Sri Lanka were identified as C. siamense and C. truncatum. None of the Colletotrichum isolates isolated in this study were either C. acutatum as reported previously (Park et al., 2006;Jayawardena et al., 2016) or C. gloeosporioides sensu stricto. There were no previous reports on the association of C. siamense with the Begonia either in Sri Lanka or elsewhere in the world. The present work has shed light on taxonomic aspects of Begonia anthracnose in Sri Lanka, which may lay the foundation for future studies related to management of the disease under nursery or marketing conditions.