Molecular Identification of Fusarium Species in Gibberella fujikuroi Species Complex from Rice, Sugarcane and Maize from Peninsular Malaysia

The objective of this study was to identify Fusarium species in the Gibberella fujikuroi species complex from rice, sugarcane and maize as most of the Fusarium species in the species complex are found on the three crops. Isolates used were collected from the field and obtained from culture collection. The Fusarium isolates were initially sorted based on morphology and identifications confirmed based on the DNA sequence of the translation elongation factor 1-α (TEF-1α) gene. Based on the closest match of BLAST analysis, five species were recovered, namely, F. sacchari, F. fujikuroi, F. proliferatum, F. andiyazi and F. verticillioides. This is the first report regarding F. andiyazi from rice in Malaysia and Southeast Asia. The phylogenetic tree generated by using the neighbor joining method showed that isolates from the same species were grouped in the same clade. The present study indicated that Fusarium species in the G. fujikuroi species complex are widespread in rice, sugarcane and maize in Peninsular Malaysia. The findings also suggest that the use of morphological characters for identification of Fusarium species in the G. fujikuroi species complex from the three crops will lead to incorrect species designation.


Introduction
Diseases caused by Fusarium species in the Gibberella fujikuroi species complex are among the most common diseases reported on agricultural crops worldwide. In Peninsular Malaysia, several species of Fusarium in the species complex such as F. verticillioides, F. sacchari, F. subglutinans, F. proliferatum and F. fujikuroi have been implicated in diseases of several agricultural crops including rice, sugarcane and maize [1][2][3].
Studies of Fusarium species on agricultural crops in Peninsular Malaysia has been based mainly on morphological characters which could lead to incorrect species designation. There are limits on the use of morphological characters for the identification of species in the G. fujikuroi species complex as some species have very similar morphology, for example F. proliferatum and F. fujikuroi, F. sacchari and F. subglutinans as well as F. verticillioides and F. andiyazi. In spite of the limitations associated with morphological characters, these traits still play an important role in sorting isolates into smaller groups before other methods of identification are applied [4].
Differences in DNA sequences of genes have been used to support morphological identification of Fusarium species. Phylogenetic analysis on DNA sequences has been used to distinguish and to evaluate the genetic relationship among closely related Fusarium species [5][6][7][8][9][10][11].
The DNA sequences encoding translation elongation factor 1-α gene (TEF-1α) has been widely used for species identification. Sequences for this gene are available through GenBank and through the more focused FUSARIUM-ID [12]. The FUSARIUM-ID database contains sequences that can be associated with a strain that is publicly available through either USDA NRRL collection or the collection maintained at the Pennsylvania State University.
The correct species name of a plant pathogenic fungus is important for the development of effective disease control management, quarantine purposes and as a basis for making decisions to protect agricultural crops as well as other natural resources from fungal pathogens [13]. Therefore, the present study was conducted to identify Fusarium species in G. fujikuroi species complex isolated from rice, sugarcane and maize, by using morphological characteristics and DNA sequencing of TEF-1α gene. Based on similar studies conducted, we expected that diverse species of G. fujikuroi species complex would be identified.

Results and Discussion
The isolates in the stock cultures were identified solely based on morphological characteristics, which could easily result in misidentification. Based on the morphological characteristics of isolates from the field and reidentification of isolates from the stock culture, 30 isolates were identified as F. verticillioides, 20 as F. sacchari, seven as F. subglutinans, 12 as F. proliferatum and nine as F. fujikuroi. The Fusarium isolates showed white or purple cottony aerial mycelium. Various pigmentations from white to light purple and from purple to dark purple were observed. Most of the macroconidia were slender with three septate with a curved apical cell and a notched basal cell. The morphology of microconidia from all the isolates were oval shaped with flattened base and were mostly formed in false head. The length of the chain varied from short (less than 10 microconidia), to medium (10-20 microconidia) and long (more than 20 microconidia).
Although isolates of Fusarium species in G. fujikuroi species complex were difficult to distinguish, there are some morphological characteristics which could be used to differentiate these species. Isolates of F. proliferatum produced abundant aerial mycelia, pyriform and clavate conidia, polyphialides were frequent and the isolates were recovered from rice, sugarcane and maize. Isolates of F. fujikuroi produced cottony aerial mycelia, abundance of sporodochia, lesser polyphialides and only recovered from rice. The length of microconidia in chain of F. proliferatum was longer (more than 20) than F. fujikuroi (less than 20).
Isolates of F. sacchari and F. subglutinans were distinguished by their macroconidia. Isolates of F. sacchari isolated from rice and sugarcane produced pyriform macroconidia while isolates of F. subglutinans isolated from maize produced oval macroconidia. F. sacchari and F. subglutinans were differentiated from F. verticillioides, F. fujikuroi and F. proliferatum by the absence of microconidia in chain. Microconidia formation of F. subglutinans and F. sacchari were produced in false heads.
Problems with using morphological characteristics to identify Fusarium species in the G. fujikuroi species complex have been reported by many workers [14][15][16][17]. However, initial identification and characterization using morphological characters is important to sort the species into smaller groups before applying other methods of identification [4].
From PCR amplification with the EF-1 and EF-2 primers, a single band of 750 bp was successfully amplified from all 78 isolates of Fusarium spp. BLAST search for similarities using FUSARIUM-ID showed that the percentage of similarity of the isolates ranged from 97%-100%. Therefore, the species name assigned was according to the closest BLAST search (Table 1). Altogether, based on TEF-1α sequences, the closest match for all the isolates, 23 isolates were tentatively identified as F. verticillioides, 6 as F. andiyazi, 29 as F. sacchari, 11 as F. fujikuroi and 9 as F. proliferatum (Table 1). Exact match or 100% similarity with isolates in FUSARIUM-ID database was found for only 12 isolates. Possible explanations were suggested by Geiser et al. [12] who suggested that the reason for this could be the occurrence of allelic variant, existence of a new species, no sequence representative in the database or that the query sequence is poorly defined.
Six isolates (3055, 3061, 3073, 3086, 3088 and 3137) from rice were tentatively identified as F. andiyazi as the TEF-1α sequences were similar to F. andiyazi with similarity ranging from 98%-99%. This report is the first of F. andiyazi occurrence in Malaysia as well as in Southeast Asia, and the first report of F. andiyazi on rice in Malaysia. F. andiyazi was first described by Marasas et al. [16] from sorghum in Africa and the United States and has also been reported from rice seed samples from Africa and Asia [10]. In the present study, swollen cells were observed, however, the swollen cells produced by F. verticillioides and pseudochlamydospores formed by F. andiyazi were difficult to distinguish which may explain why the structures were not easily distinguishable on CLA. F. andiyazi was also reported to be morphologically similar with F. thapsinum, a species in G. fujikuroi complex which is commonly found in sorghum [18]. Thus, F. andiyazi was very difficult to identify with only morphological characters and can only be distinguished by using DNA sequences. Wulff et al. [10] also differentiated F. andiyazi and F. verticillioides isolated from rice using TEF-1α sequences and both species were grouped in different clades. Another method to differentiate between F. andiyazi and F. verticillioides is by mycotoxins production. In a study by Wulff et al. [10], F. verticillioides produced significant level of fumonisin B1 whereas F. andiyazi only produced traces of fumonisions.
F. verticillioides is widely distributed worldwide and is one of the most common pathogens of maize, associated with ear rot and stalk rot of maize as well as symptomless infections or as endophyte on maize [19]. In the present study, only four isolates of F. verticillioides were isolated from rice infected with bakanae disease. In addition to agricultural crops, F. verticillioides has been isolated from mangrove soil in Malaysia [20]. In the present study the isolates of F. verticillioides were most commonly isolated from maize.
Isolates identified as F. subglutinans from maize based on morphological characteristics were all identified as F. sacchari based on TEF-1α sequences. Most of the isolates of F. sacchari in this study were isolated from sugarcane showing pokkah boeng disease symptoms although nine isolates were obtained from rice and two isolates from maize. F. sacchari causes pokkah boeng of sugarcane and is widely distributed in sugarcane growing areas in Malaysia [3]. Isolates of F. sacchari have also been isolated from rice with bakanae disease symptoms [2]. According to Burgess et al. [21], F. subglutinans is seldom isolated in warmer countries as it was widespread in cooler temperate regions. Logrieco et al. [22] also reported that F. subglutinans occurs under colder and more humid conditions. F. subglutinans and F. sacchari are distinct biological species [23] and it is possible that most, if not all of the reports of F. subglutinans from Malaysia are actually reports of F. sacchari.
Phylogenetic relationships were analyzed by the neighbor-joining method which was based on 670 aligned positions (Figure 1). In the data set, there were 135 variable sites, 526 conserved sites and 111 parsimony informative sites. All the main clades of the tree were well supported with bootstrap values ranging from 98%-100%. The phylogenetic tree can be divided into two main clades, I and II. Clade I consisted of sub-clades A and B. Isolates of F. sacchari were grouped in sub-clade A. Isolates of F. fujikuroi and F. proliferatum were grouped separately in sub-clades B1 and B2 respectively. Isolates of F. andiyazi and F. verticillioides were grouped separately in sub-clades C1 and C2. F. oxysporum and F. inflexum were clustered separately from the isolates of G. fujikuroi species complex.
The isolates of F. sacchari formed a distinct clade that was clearly separated from the other isolates of Fusarium in the neighbor-joining tree. The taxa closest to F. sacchari were F. proliferatum and F. fujikuroi. The grouping of F. sacchari, F. proliferatum and F. fujikuroi were in accordance with the biogeographic hyphotesis formulated by O'Donnell et al. [14] which grouped the three species in the Asian clade.
Eleven isolates of F. fujikuroi and nine isolates of F. proliferatum were identified based on the closest match on of TEF-1α gene sequences. Both isolates of F. fujikuroi and F. proliferatum formed separate sub-clades in the tree (Figure 1) but are in the same main clade which indicated that F. fujikuroi and F. proliferatum are closely related. Similar results were also reported by Wulff et al. [10] in which F. fujikuroi and F. proliferatum, which both cause bakanae disease, were grouped in separate sub-clades. Both F. fujikuroi and F. proliferatum are grouped in the Asian clade and are morphologically and biologically very similar [14,24]. Moreover, F. fujikuroi and F. proliferatum are regarded as sibling species and some isolates can cross fertile [25].
Isolates of F. verticillioides were clustered in the same main clade with isolates of F. andiyazi. F. verticillioides was grouped in the African clade with several other agricultural important pathogens of G. fujikuroi species complex [14]. Although F. verticillioides is a member of the African clade, the major host of the fungus is maize which originated in Mexico or Central America. With more species recovered from different host plants and substrates in Malaysia, it could change the composition of the African clade. As for F. andiyazi, the detail species concept and phylogenetic analysis of the species remains to be determined [26]. This study presents the identification of Fusarium species of G. fujikuroi species complex from rice, sugarcane and maize based on TEF-1α gene which is a useful marker for distinguishing Fusarium species [12]. Phylogenetic analysis based on the gene clearly grouped the different species into separate groups with high bootstrap values confirming that TEF-1α gene is a good marker for identification of Fusarium species. However, additional genes such as other protein coding genes and genes involved in mycotoxin biosynthesis would support the taxonomic affinities of isolates which showed 97-99% sequence similarities as well as to study the pathogen populations. Studies on the evaluation of mycotoxins production are also needed as mycotoxins are commonly produced by Fusarium species in the G. fujikuroi species complex. Furthermore, there is also the probability that new strains will probably emerge through hybridization, which could lead to the establishment of new plant diseases.

Fungal Isolates
Fungal isolates used in this study were obtained from three hosts, namely, rice, sugarcane and maize. Abnormal growth of rice plants and sugarcane leaves were cut to approximately 1 cm 2 with a sterile scalpel and soaked in 10% sodium hypoclorite (Clorox) for 30 s, transferred into sterile distilled water and soaked for another 30 s. The pieces of the tissues were then dried on sterilized filter paper (Whatman ® No 1) and put onto peptone pentachloronitrobenzene agar (PPA) plate. The PPA plates were incubated at 25 °C until mycelia growth was observed. Mycelia from the samples were then subcultured onto potato sucrose agar (PSA). Kernels from maize were directly plated onto PPA plates with sterile forceps and incubated at 25 °C until mycelia growth were observed from the samples. The mycelia were also transferred onto PSA. To obtain pure culture of mycelia grow from the tissues and maize kernels, single spore isolation was carried out using dilution plating [4]. For identification, three different types of media were used, namely Potato Dextrose Agar (PDA), PSA and Carnation Leaf Agar (CLA) as described in The Fusarium Laboratory Manual [4]. Microscopic and macroscopic characteristics as described in the manual were used for species identification.
Thirty isolates from the field comprising 19 isolates from maize, three isolates from rice and eight isolates from sugarcane were successfully isolated and identified. Another 48 isolates from the rice, sugarcane and maize were also obtained from stock culture of School of Biological Sciences, Universiti Sains Malaysia culture collection. The samples were collected in 2007-2008. The isolates from the stock culture were also recovered from bakanae disease of rice, pokkah boeng disease of sugarcane and rotting of maize kernels and were identified based on morphological characteristics. In total, 78 isolates of Fusarium were used in this study (Table 1).

DNA Extraction
Genomic DNA was extracted from mycelia grown on PSA, incubated for 7 days at 27 °C. Mycelia were ground in a mortar with liquid nitrogen until it became a fine powder. Approximately 0.23 g of the lyophilized mycelia was used for DNA extraction with a DNeasy ® Plant Mini Kit (Qiagen, Germany), according to the manufacturer's instructions. The DNA concentration was measured from all samples using UV spectrophotometry using a Nanodrop ND-1000 (Thermo Scientific, Wilmington, DE, USA) and adjusted to 20 ng/µL.

PCR Amplification
The TEF-1α gene was amplified with primers EF-1 (5'GTT AAG AGG CGC GGT GTC GGT GTG 3') and EF-2 (5'GGA AGT ACC AGT GAT CAT GTT3') [14]. Amplification reactions were carried out in a 50 µL reaction mixture which consisted of 1X PCR buffer, 3. PCR products were detected by electrophoresis in a 1.5% agarose gel in TBE buffer. The electrophoresis was run for 100 min at 80 V and 400 mA. The size of the band was estimated by comparison to 100 bp DNA marker (MBI Fermentas, Lithunia). A negative control was used to test the absence of contamination.

DNA Sequencing and Phylogenetic Analysis
The PCR products were purified using a QIAquick PCR Purification Kit (Qiagen, Germany) according to the manufacurer's instruction. The purified PCR products were directly sequence in both directions using the respective forward and reverse primers, and sent for sequencing to a service provider. DNA sequences were aligned by using ClusterW Multiple alignment in Molecular Evolutionary Genetic Analysis 4 or MEGA 4 [27] and adjusted manually. Default gap penalties were used in the pair-wise alignment. The consensus sequence was then compared with other DNA sequences by using BLAST against FUSARIUM-ID database at http://fusarium.cbio.psu.edu [12].
All sequences were aligned with gap opening penalty of 3.0, and a gap extension penalty of 1.8. Neighbor-joining analysis was then performed in MEGA 4. A neighbor-joining tree was constructed with the Jukes-Cantor model and the reliability of neighbor-joining trees was estimated by bootstrap method with 1000 replications. Fusarium oxysporum (FJ985415) and F. inflexum (AF331814) sequences from GenBank were selected as the outgroup.

Conclusions
The wide host range of Fusarium species in the G. fujikuroi species complex suggest that new species or species that have not been reported from Peninsular Malaysia are likely to be found on agricultural crops and other substrates such as from the soils. The present study suggests that many isolates of Fusarium species in the G. fujikuroi species complex in culture collections maybe misidentified which is not surprising since the isolates have been identified primarily using morphological characters. Accurate identification of Fusarium species associated with economically important crops could help in the formulation of disease control strategies, prediction of host range, climatic adaptation and mycotoxin production potential.