Introduction

The species of Colletotrichum comprise important plant pathogens that cause anthracnose disease on many crops including fruits, vegetables, cereals and ornamental plants. In chili production (Capsicum annuum L.), anthracnose is a serious problem especially in tropical and sub-tropical areas including India, Korea and Thailand (Sharma et al. 2005; Kim et al. 2008; Than et al. 2008). Chili anthracnose is caused by several Colletotrichum species: C. capsici, C. acutatum, C. gloeosporioides, C. coccodes and C. dermatium (Manandhar et al. 1995; Hong and Hwang 1998; Ivery et al. 2004; Sharma et al. 2005). In Thailand C. capsici, C. gloeosporioides, and C. acutatum have been reported to affect this crop (Than et al. 2008). To date, the species identification of Colletotrichum spp. has relied mainly on morphological characteristics such as colony characteristics, conidial shape and size, shape of appressoria, conidial mass color, and the presence and absence of setae (Sutton 1992). However, the effects of growth conditions, such as light and temperature, and re-subculturing can cause morphological changes. Nowadays, DNA-based molecular techniques have been developed and are widely used for taxonomic purposes using DNA sequencing from internal transcribed spacer region (ITS) and coding sequences for fungal identification. Phylogenetic analysis using a single DNA dataset or multilocus analysis which was generated from loci such as actin, calmodulin and chitin synthesis genes has been used to support the morphology characteristics.

Recently, Weir et al. (2012) has divided the Colletotrichum gloeosporioides complex into 22 species by using a multilocus phylogenetic analysis and has reported that some isolates of Colletotrichum spp. from Capsicum annuum in Thailand belong to C. siamense. In this study, we report the characterization of isolates of Colletotrichum spp. from chili pepper exhibiting symptoms of anthracnose, using morphological characteristics, pathogenicity and DNA markers.

Materials and methods

Isolation of fungi

Chili fruit showing anthracnose symptoms were collected from chili producing fields in twelve provinces of five regions of Thailand (Fig. 1). Tissue from the margins of anthracnose lesions was cut into small pieces, sterilized with 10% sodium hypochlorite for 5 min and dried on sterile paper. Tissue samples were placed on Petri dishes containing potato dextrose agar (PDA) and incubated at 25 °C under 12-h light:12-h dark for 5–7 days. Single spores were isolated and maintained on the same medium for further studies.

Fig. 1
figure 1

Samples of chili anthracnose were collected from different provinces in Thailand, indicated by the orange dots

Full size image

Morphological identification

Intact fruits were inoculated with each isolate by spraying with a conidial suspension at approximately equal to 800 conidia/cm2 of fruit surface area. After symptoms developed and a conidial mass was produced on the lesion, conidia were transferred to PDA plates. Approximately 100 conidia were mounted on glass slides to determine the conidial shape and size and then isolates were identified using Sutton’s key (Sutton 1980, 1992). To investigate the colony morphology, mycelium agar plugs were cut with a cork borer (0.5 cm in diameter) from 5-day-old cultures of Colletotrichum and transferred to the center of a Petri dish which contained 20 ml potato dextrose agar (Difco, USA). Three plates per isolate were used. All plates were incubated at 25 °C under 12-h light:12-h dark for 9 days. Then, the colony characteristics were observed. Appressoria were produced by using a slide culture technique according to Johnston and Jones (1997) and were observed under a microscope.

DNA extraction

Fungal mycelium was cultured in malt broth (MB; malt extract 20 g/L) for 4 days. The mycelium was harvested with forceps and then suspended in extraction buffer (Alexander et al. 2007). After that chloroform:isoamyl alcohol (24:1) was added and the suspension was incubated on ice for 5 min and centrifuged at 14,000g for 10 min. The clear supernatant was transferred to a new microtube and 400 μl of isopropyl alcohol was added for nucleic acid precipitation. That tube was centrifuged at 14,000g for 5 min and the isopropanol discarded. The pellet was washed 2 times with 70% of ethyl alcohol. After washing, the DNA was suspended with 25 μl RNase water and stored at −20 °C.

PCR amplification

The internal transcribed spacer (ITS) region including ITS1, 5.8sRNA gene and ITS2 was amplified by using ITS1 primer (5’GCCGTAGGTGAACCTGCGG3’) and ITS4 primer (5’TCCTCCGCTTATTGATATGC3’) (White, 1990). PCR was performed in 25 μl reaction mixtures containing 0.25 mM dNTPs, 0.5 U Taq polymerase (Thermostatic dream taq; Thermostatic), 0.1 μM of each primer and 50 ng DNA template. The PCR cycle consisted of 1 cycle of 5 min at 95 °C, 30 cycles of 45 s at 95 °C, 45 s at 60 °C, 45 s at 72 °C and a final cycle of 72 °C for 5 min. A partial actin fragment was amplified using ACT512F (5’AGTTGCAAGGCCGGTTTCG3’) and ACT738R (5’TACGAGTCCTTCTGGCCCAT3’) primer (Carbone and Kohn 1999). The PCR started with 95 °C 5 min for 1 cycle, followed by 30 cycles, with step of 30 s at 95 °C, 30 s at 52 °C, 30 s at 72 °C and a final cycle 72 °C for 5 min. All PCR products were electrophoresed in a 1.5% agarose gel, stained with a fluorescence dye (Novel juice; Genedirex) and visualized under UV light.

Phylogenetic and sequence analysis

The sequences which corresponded to each region were assembled and edited using BioEdit (Hall 1999), and consensus sequences of each isolate were created. The consensus sequences of each region were submitted to the Basic Local Alignment Search Tool (BLAST) from NCBI (http://blast.ncbi.nlm.nih.gov) for finding similar sequences. For phylogenetic analysis, data were generated from the combined dataset (ITS + ACT) and separate datasets (ITS, ACT).

The phylogenetic trees were constructed from multiple alignment from ClustalW (Thomson 1994) included in the MEGA5 software (Tamura 2011). The analysis was conducted using the neighbor-joining (NJ) likelihood method in MEGA 5.0. All positions containing gaps and missing data were eliminated. The bootstrap analysis was assessed with 1000 replicates to determine the reliability of branches. The new sequences in this study were deposited in DDBJ (DNA Data Base of Japan, Japan).

Pathogenicity test

The pathogenicity test for each species was performed on red intact chili fruits. Fruits were washed with sterilized distilled water and then dried at room temperature. The selected isolates were cultured on PDA medium at 25 °C under 12-h light:12-h dark for 9 days. Conidial suspension was prepared from 8-day-old cultures by adding sterilized distilled water to the plate. The suspension was filtered through cheesecloth. The conidia concentration was adjusted to 2 × 106 conidia/ml.

Chili fruits were inoculated by spraying the conidial suspension on one side of the fruit surface with an air brush apparatus at approximately 800 conidia/cm2 of fruit surface area. For the negative control, the other side of inoculated fruits was sprayed with sterilized distilled water. Each inoculated fruit was covered with moist plastic bags for 48 h at 25 °C. Afterwards, the bag was removed and the fruit was observed for the symptom. Three fruits were analyzed for each isolate.

Results

Morphological identification

The colonies of C. gloeosporioides were found in two types. One type showed a pale gray to dark gray mycelium covered with an orange conidial mass and the second type exhibited white to pale gray mycelium with a pinkish and orange conidial mass. The colony of C. siamense showed white to pale gray mycelium with a pinkish and orange conidial mass. The colony of C. capsici had greenish gray to light brown mycelium with a pale to a salmon orange conidial mass. The colony of C. acutatum showed gray to white puffy mycelium with a salmon to pinkish conidial mass (Fig. 2).

Fig. 2
figure 2

Colonies of Colletotrichum isolates after 9 days on PDA: a, b C. gloeosporioides NAN2002 and NAN1603, c, d C. siamense KPS2 and CN4, e, f C. acutatum NAN1103 and PR264, g, h C. capsici CC9.4 and CN2

Full size image

Conidia of the isolates were distinguished into three types (Fig. 3). The first type was hyaline, smooth, and cylindrical with a round end, 9.3–19.2 × 2.2–5.3 μm. The second type of conidia was hyaline and falcate 14.7–29 × 2–5 μm. The third type of conidia was hyaline, smooth, and fusiform 8.2–19.5 × 2–4.5 μm.

Fig. 3
figure 3

Conidia of selected Colletotrichum isolates: a, b C. gloeosporioides NAN2002 and NAN1603, c, d C. siamense KPS2 and CN4, e, f C. acutatum NAN1103 and PR264, g, h C. capsici CC9.4 and CN2

Full size image

The appressoria of C. gloeosporioides groups were clavate and brown to dark brown. C. capsici appressoria were elliptical to clavate, but some produced an irregular brown lobe. Appressoria of C. acutatum were elliptical to circular and dark brown.

These characteristics were described for C. gloeosporioides, C. capsici and C. acutatum, respectively, according to Sutton (1992).

PCR amplification and sequencing analysis

DNA fragments from ITS and Act locus were successfully amplified from total genomic DNA of 48 isolates. A PCR product size of approximately 550 bp was obtained by using ITS1/ITS4 primer and a product fragment size with ACT738/ACT582 primer was approximately 230 bp. Based on the BLAST search of the ITS region (ITS1–5.8 s rDNA-ITS2), the result showed that the species were consistent with C. gloeosporioides, C. siamese, C. capsici, C. truncatum, C. dematium, C. acutatum, C. scoville and Colleotrichum sp. from GenBank with 98–100% nucleotide sequence identity.

Then, ITS sequences of 48 isolates of Colletotrichum species including reference sequences and out-group species were used for phylogenetic tree construction. The phylogenetic tree was clearly separated into three main clades by using NJ. All clades were separated from the out-group, Bipolaris maydis (CBS 136.29). These clades were C. capsici (Cc) clade, C. gloeosporioides (Cg) clade, and C. acutatum (Ca) clade (Fig. 4). Of the collected species 50% were in the Ca clade, followed by Cg and Cc with 20% and 30%, respectively. The Cg clade was separated into 2 sub-groups, A and B. In the Cc clade, there were 2 sub-groups: sub-groups C and D. A further phylogenetic tree was performed by using coding region sequences (partial actin gene). The tree was separated into 3 main clades as Cc, Cg and Ca (Fig. 5). In the Ca clade, there were 2 sub-clades (A and B). The Cc clade was divided into 2 groups as sub-groups C and D. A combined dataset analysis with NJ divided the isolates into 3 main clades representing the species: C. capsici, C. gloeosporioides and C. acutatum (Fig. 6). In the Cc clade, it showed 2 sub-groups with 100% bootstrap support. The Cg clade was divided into 2 groups as sub-groups C and D. In the Ca clade, there were 2 sub-groups, E and F.

Fig. 4
figure 4

The phylogenetic tree constructed by using the neighbor-joining method of the ITS region from Colletotrichum species associated with chili fruits in Thailand and reference species. The number above each branch indicates the bootstrap percentage value with 1000 replications. Branch lengths are drawn proportionally to genetic distances and the bar at the bottom of the tree indicates a length corresponding to 0.05 nucleotide substitutions per site

Full size image
Fig. 5
figure 5

The phylogenetic tree constructed by using the neighbor-joining method of the ACT gene from Colletotrichum species associated with chili fruits in Thailand and reference species. The number above each branch indicates the bootstrap percentage value with 1000 replications. Branch lengths are drawn proportionally to genetic distances and the bar at the bottom of the tree indicates a length corresponding to 0.05 nucleotide substitutions per site

Full size image
Fig. 6
figure 6

The phylogenetic tree constructed by using the neighbor-joining method of the combined dataset of ACT and ITS sequences from forty-eight isolates of Colletotrichum species associated with chili fruits in Thailand and reference species. The number above each branch indicates the bootstrap percentage value with 1000 replications. Branch lengths are drawn proportionally to genetic distances and the bar at the bottom of the tree indicates a length corresponding to 0.05 nucleotide substitutions per site

Full size image

Pathogenicity of Colletotrichum spp.

Colletotrichum isolates CC9.4, NAN1603, NAN1103, and CN4 were selected from C. capsici, C. gloeosporioides, and C. acutatum clades, and the C. siamense sub-clade, respectively, in order to test their pathogenicity. Clearly visible anthracnose symptoms developed 5 days after inoculating with C. acutatum and 7 days after inoculating with C. capsici, C. gloeosporioides and C. siamense (Fig. 7). The anthracnose symptoms typically appeared as sunken lesions at the inoculation area. The symptoms caused by C. gloeosporioides showed black sunken lesions (Fig. 7a). The necrotic spots of C. siamense were orange and surrounded with brown-black necrotic tissue (Fig. 7b). The symptom lesions of C. acutatum and C. capsici produced similar sunken lesions but fruiting structures on the center of the lesions were orange and black in color, respectively (Fig. 7c and d).

Fig. 7
figure 7

Pathogenicity test of representative isolates of each Colletotrichum species showing the disease symptoms on chili fruits. a C. gloeosporioides isolate NAN1603, b C. siamense isolate CN4, c C. acutatum isolate NAN1103 and d C. capsici isolate (CC9.4)

Full size image

Discussion

Analysis of phylogeny of partial actin (ACT) gene and combined dataset (ITS + ACT) was compared with the phylogeny of ITS region. The phylogram showed that phylogenies could separate the collected isolates into three main groups, Cc, Cg, and Ca.

In the analysis of Colletotrichum isolates from strawberry in Korea (Nam et al. 2013), C. fruticola which was included in the C. gloeosporioides complex was clearly distinguished from the Cg group by using a combination of ITS, partial actin, and GAPDH genes in analysis. This study suggested that the phylogenetic analysis with combined genes was useful for an accurate identification of different species of Colletotrichum. In the classification of Rhizoctonia spp. by SNPs, the use of ß-actin gene could display more allele discriminations than rDNA-ITS, while the use of rDNA-ITS sequences was not sufficient to distinguish the isolates within the subgroups (Wei et al. 2014).

In this study, the combined nucleotide sequence data (ITS+ ACT) and ITS sequence clearly divided the C. gloeosporioides group into two molecular groups which were correlated to the colony characteristics. The colony characteristic of the first group was white to light gray mycelia with orange conidia mass. In the first sub-group isolate, CN4 was closely related to C. siamense (ICMP 18575) with 73% and 89% bootstrap support in the combined analysis and ITS region analysis, respectively. C. siamense was reported in Capsicum annuum in Thailand by Weir et al. (2012). This species is a member of the C. gloeosporioides species complex. Meetum et al. (2015) reported that C. siamense was found in dragon fruits in western and central parts of Thailand, and the pathogen was identified based on morphological characteristics and ITS nucleotide sequence. In the second sub-group, Cg, the colony characteristic was gray to dark gray mycelium with an orange conidial mass. The isolate members in this group were found in various areas of the country where they could cause a serious problem in chili and other crops.

In our phylogenetic analysis, members of the C. capsici group were divided into two sub-groups with 100% bootstrap support. The phylogenetic analysis of ITS-DNA sequences showed the relationship between each group and sample collecting areas. Isolate KJN002, the only isolate from the western area of the country, was distinguished from the other C. capsici isolates. Based on the phylogenetic analysis of a partial actin and the combined dataset, the isolate members in the first sub-group were from all collection areas. In the other group, the samples were mainly from northern part of Thailand, and no isolate was obtained from the western part of the country. The morphological characteristics of the isolate members of both sub-groups were not different.

ITS region phylogenetic analysis of C. acutatum indicated that all isolates in this group were formed in a monophyletic group, whereas the analysis of partial actin and combined data exhibited some variation. In the analysis, the C. scovilli (CBS 120708) was used as a reference nucleotide sequence. The colony characteristics of isolate members in the C. acutatum group were similar, and some isolates produced fusiform conidia.

Based on these results, the use of ITS region phylogenetic analysis was not able to distinguish a group of isolates based on the collecting locations. Otherwise, combined (ACT and ITS) sequence analysis, as indicated in the Cc clade of the tree, was successful to discriminate this group. However, the discrimination in Cc clade did not correlate with morphological characteristics, but it was shown to be related to the sample collecting area. In many previous studies, the ITS region had been broadly used to define the Colletotrichum species due to its ability to separate isolates belonging to different species. However, in the work of Crouch et al. (2009), in which the group of Colletotrichum isolates with falcate conidia were distinguished from other closely related species, the ITS region alone was inadequate to separate them. Application of multi-gene or multi-locus phylogenetic analyses with ITS sequences was required for reliable analysis (Du et al. 2005). The use of the actin region, the coding sequence, has been suggested in fungal phylogenetic analysis. This gene encodes conserved actin protein found in all eukaryotic cells and a single copy in the majority of fungi tested. The gene has been used to study evolutionary relationships. In addition, it is easy to amplify and align the sequences (Daniel and Meyer 2003; Reeb et al. 2004). This is good for studying the relationships and variation in the species. In the study of Du et al. (2005), using mating-type gene sequences for resolving the Colletotrichum species complex, indicated that MAT1–2 mating type sequences provided many more alignable variable sites than rDNA regions, and the ITS2 sequence has low complexity.

In this study, the phylogenetic analyses using combined sequences of the ACT gene and ITS region were successful to resolve the species complex of C. gloeosporioides, the isolates of Thailand, since distinct phylogenetic groups were shown to be associated with their recognizable morphology groups. However, this combined multilocus analysis partially supported the morphological distinct groups in C. capsici and C. acutatum. Therefore, other DNA barcodes need to be further investigated.

In summary, this study demonstrates that four Colletotrichum species can cause chili anthracnose in Thailand, based on morphological characteristics, DNA sequence data of 48 Colletotrichum isolates from anthracnose lesions on chili fruit, and additional pathogenicity tests (Table 1).

Table 1 Colletotrichum isolates from different locations in 12 provinces of Thailand in the years 2010–2015
Full size table