In 2019 and 2020, a total of 14 fields from nine cities among the major watermelon production areas in Korea were selected for Alternaria spp. survey. Watermelon leaves and stems showing Alternaria-induced leaf blight lesions were harvested. Small pieces of tissue (10 × 10 mm) were removed from the margins between the healthy and symptomatic tissues, and their surfaces were disinfected with 70% ethanol for 1 min and rinsed with sterilized dH₂O three times. The pieces were then placed on potato dextrose agar (PDA) (MBcell, Seoul, Korea) plates and incubated at 28°C in light for seven days. Multiple fungal colonies were obtained from each symptomatic tissue, and only those with Alternaria morphologies were selected as isolates. The isolates were sub-cultured onto new PDA plates for single spore purification.

Genomic DNA was extracted following a previously described protocol (Chen and Ronald, 1999) For the extraction, Alternaria mycelia were scrapped from the surface of the seven-day cultures grown on PDA and placed in a 2 ml microcentrifuge tube containing 700 μl cetyltrimethl-ammonium bromide (CTAB) extraction buffer (3% (w/v) CTAB, 20 mM EDTA pH 8.0, 1.42 M NaCl, 100 mM Tris-HCl pH 8.0, 2% (w/v) polyvinylptrrolidone, and 0.2% β-mercaptoethanol). After 10 min incubation at 65°C, the addition of 570 μl of chloroform:isoamyl alcohol (24:1) to the tube, and 10 min in a centrifuge at 13,000 rpm, genomic DNA was precipitated with 0.7 volume isopropanol. After centrifuging for 10 min at 13,000 rpm, the genomic DNA pellet was washed with 75% ethanol and then dried at room temperature. The genomic DNA was dissolved in 20 μl sterilized dH₂O and stored at −20°C.

The ITS region of ribosomal DNA, GAPDH gene, RPB2 gene, and HIS3 gene were amplified using ITS1 (5′-TCCGTAGGTGAACCTGCGG-3′)/ITS4 (5′-TCCTCCGCTTATTGATATGC-3′) (White et al., 1990), gpd1 (5′-CAACGGCTTCGGTCGCATTG-3′)/gpd2 (5′-GCCAAGCAGTTGGTTGTGC-3′) (Berbee et al., 1999), RPB2-5F2 (5′-GGGGWGAYCAGAAGAAGGC-3′)/ fRPB2-7cR (5′-CCCATRGCTTGTYYRCCCAT-3′) (Liu et al., 1999), and H3-1a (5′-ACTAAGCAGACCGCCCGCAGG-3′)/H3-1b (5′-GCGGGCGAGCTGGATGTCCTT-3′) (Glass and Donaldson, 1995) primers, respectively. Polymerase chain reaction (PCR) was performed in a 20 μl reaction mixture containing 0.1 μl FIREPol DNA Polymerase (0.5 units) (Solisbiodyne, Tartu, Estonia), 1 μl each primer (10 μM), 0.2 μl dNTP mix (200 μM), 2 μl 10× buffer, 1 μl template (approximately 100 ng), 4 μl 5× Q-solution, and 10.7 μl sterilized dH₂O. Amplifications were performed with an initial denaturation step at 94°C for 3 min followed by 35 cycles of 15 s at 94°C, 15 s at 53°C and 60 s at 72°C for ITS region, 35 cycles of 15 s at 94°C, 15 s at 57°C and 60 s at 72°C for GAPDH, 35 cycles of 15 s at 94°C, 15 s at 56°C and 2 min at 72°C for RPB2, 35 cycles of 15 s at 94°C, 15 s at 67°C and 60 s at 72°C for HIS3, and final extension at 72°C for 7 min. Amplified fragments were loaded on an agarose gel (0.8% w/v or 1.5% w/v) stained with RedSafe Nucleic Acid Staining Solution (20,000×) (Intronbio, Seongnam, Korea) and visualized under UV light.

The morphology of Alternaria isolates was examined as previously described (Andersen et al., 2001; Ma et al., 2021) with minor modification. For colony type, color, margin, and diameter (Andersen et al., 2001), Alternaria isolates were incubated on PDA and V8 juice agar (200 ml/l Campbell V8 juice, 3 g CaCO₃, 15 g agar, and 800 ml dH₂O) plates at 28°C dark condition for seven days. For the pattern of sporulation (Ma et al., 2021), the isolates were cultured at 22°C on PDA and V8 juice agar plates under UV light with light/dark cycle of 12 h photoperiod for seven days. The sporulation patterns and spore were examined using an inverted microscope Eclipse Ti (Nikon, Tokyo, Japan) fitted with an objective (200×).

The PCR amplification products of ITS region, HIS3, GAPDH, and RPB2 were purified with FavorPrep GEL/PCR Purification Kit (Favorgen, Ping-Tung, Taiwan) and sequenced by Sanger sequencing (Macrogen, Seoul, Korea). Sequence similarity searches were performed using the nucleotide BLAST at National Center for Biotechnology Information (NCBI, https://blast.ncbi.nlm.nih.gov/Blast.cgi). DNA sequences of ITS region, HIS3, and concatenated (GAPDH and RPB2) gene were aligned through ClustalW and refined manually using MEGA v.7.0 (Kumar et al., 2016). Maximum likelihood based on the Tamura-Nei model (Tamura and Nei, 1993) analyses of ITS region, HIS3, and concatenated (GAPDH and RPB2) gene datasets were performed using MEGA v.7.0 with 500 bootstrap replicates. Branches corresponding to partitions were reproduced in less than 50%. The phylogenetic trees were viewed in MEGA v.7.0. All analyzed DNA sequences, ITS region, HIS3, GAPDH, and RPB2, from isolates were deposited in GenBank (https://www.ncbi.nlm.nih.gov/genbank/) and their accession numbers are listed in Tables 1 and 2.

Alternaria inoculation was performed according to the previously described detached-leaf inoculation method (Pryor and Michailides, 2002) with minor modification. Briefly, to prepare the inoculum, Alternaria isolates were pure-cultured on PDA plates at 28°C in the dark for seven days and each mycelial plug was cut from the actively growing margin of the fungal colonies. Inoculations were conducted on detached leaves of watermelon cultivar, ‘Charleston gray’ (Wu et al., 2019). Fully expanded leaves were collected from six-to seven-week-old watermelon plants grown in a growth chamber. For the inoculation of each Alternaria isolate, six leaves from ‘Charleston gray’ were placed on papers soaked with distilled water in plastic boxes (25 × 19 × 5 cm [length × width × height]) and three mycelial plugs per leaf were placed on the detached leaves. The plastic boxes were individually covered with transparent plastics to maintain 100% humidity and incubated at 25°C with a 12 h photoperiod per day for seven days. The disease index was calculated after each disease area was divided by the whole leaf area using ImageJ. Alternaria isolates were re-isolated consistently from all the inoculated leaf to fulfill Koch’s postulates

Watermelon leaves and stems showing Alternaria-induced leaf blight lesions were collected from 14 fields in nine cities of Korea for Alternaria spp. survey (Table 1, Fig. 1A). Multiple fungal colonies were obtained from each symptomatic tissue (Fig. 1B), and only those with Alternaria morphologies were selected as isolates. A total of 16 Alternaria isolates were obtained and cultured to homogeneity. All isolates developed loosely cottony and light brown to grey colonies on PDA after incubation at 28°C for seven days in the dark. (Supplementary Fig. 1).

It has been reported that the amplicon lengths of the ITS regions varied among the Alternaria genus (Blagojević et al., 2020; Woudenberg et al., 2013). The ITS regions of Alternaria isolates were amplified with primers, ITS1 and ITS4. However, there was no length variation and sequence diversity in the ITS region throughout all isolates, except for one single-nucleotide substitution (T to C) in isolate CY202-29 (Supplementary Fig. 2A and B). Phylogenetic analyses based on the ITS region was able to divide seven references retrieved from GenBank into two different clades, A. cucumerina and A. infectoria. However, our isolates and many other references were not successful in dividing into clades and all were grouped into one clade containing A. alternata, A. tenuissima, and A. gaisen (Supplementary Fig. 3). These results indicate that the ITS region cannot be used for assaying genetic diversity in the Alternaria spp. causing leaf blight on watermelons in Korea.

Partial coding sequence of HIS3 was successfully used to divide Alternaria into subgroups of based on size analysis of PCR amplicons (Zhao et al., 2016a, 2016b; Zheng et al., 2015). After PCR analysis using a primer set, H3-1a and H3-1b, all 16 Alternaria isolates were successfully divided into three subgroups (Sub I, II, and III) based on their sizes (Table 2, Fig. 2). The largest amplicon with 546 bp was produced by Sub I, consisting of CY202-11, CY202-16, CY202-25, GC191-1, GC191-14, GC192-7, GJ192-1, GJ192-2, IC191-1, and YI201-2. The second largest one with 486 bp was amplified by Sub II GC192-13. Sub III producing the smallest amplicon with 440 bp was the dominant isolates, consisting of CJ191-1, CJ191-3, CY202-18, CY202-29, and IC191-2. Two isolates from dominant Sub I (CY202-11 and GC191-1) and one isolate from each Sub II (GC192-13) and Sub III (CJ191-1) were randomly selected for further analysis.

Four representative isolates, CY202-11, GC191-1, GC192-13, and CJ191-1, from three subgroups were subjected to morphological analysis on PDA and V8 juice agar plates. On the V8 juice agar plate, all isolates produced loosely cottony and dark grey colonies after incubation at 28°C in dark condition for seven days. On PDA plates, three isolates, CY202-11, GC191-1, and GC192-13, in Sub I and Sub II developed cotton light brown colonies, while CJ191-1 from Sub III produced a cotton dark grey colony (Table 3, Fig. 3A). In all isolates, conidiophores were short and arising singly (Fig. 3B). The conidia of them were mainly short ovoid to obclavate with a rounded apical cell, with no significant difference among the four representative isolates. A similar number of transverse septa and longitudinal septa of their conidia were also observed from 1 to 6 and from 0 to 2, respectively. All isolates had a similar size of conidiophores but slightly differed in sporulation on PDA and V8 juice agar (Table 4).

To examine whether the three subgroups can be identified into the Alternaria species, their HIS3 sequences were compared (Fig. 4, Supplementary Fig. 4). As a reference for sequence comparison, the HIS3 gene (accession no. XM_018526023) of an A. alternata isolate was used. Its full-length cDNA is 411 bp in length and the cDNA fragment amplified with primers H3-1a and H3-1b is 378 bp. In the analysis of Sub I isolates, CY202- 11 and GC191-1, the HIS3 gDNA amplicon was 546 bp in length and contained three introns 54 bp (Int A), 62 bp (Int B), and 52 bp (Int C) (Fig. 3). The HIS3 gDNA amplicon from GC192-13 isolate in Sub II was 484 bp with two introns 54 bp (Int A) and 52 bp (Int C). CJ191-1 isolate in Sub III was 440 bp in length with only one intron 62 bp (Int B).

Phylogenetic analyses based on HIS3 was performed together with available reference sequences retrieved from GenBank (eight from A. alternata. eight from A. tenuissima, one from A. gaisen, four from A. infectoria, three from A. cucumerina, two from Alternaria sp., and one outgroup Fusarium proliferatum) (Fig. 5). Sub I isolates, CY202-11 and GC191-1, were located in one clade with the known A. tenuissima from GenBank. Sub II GC192-13 isolate was together with A. gaisen and Sub III CJ191-1isolate was located in the A. alternata clade. Occurrence rates of each Alternaria sp. were relatively high with quite wide range depending on where the isolates were harvested (Supplementary Table 1). Including A. infectoria references and A. cucumerina references, the entire species-group within the Alternaria spp. was successfully split into five clades, suggesting that HIS3 can be used for classifying Alternaria spp. causing leaf blight on watermelon.

The amplicon lengths of GAPDH and RPB2 varied among some of Alternaria spp. associated with leaf spot and leaf blight disease in Solanaceae and Brassica (Bessadat et al., 2020; Blagojević et al., 2020; Lawrence et al., 2016). Although the sequences of GAPDH and RPB2 were not identical among four isolates and references retrieved from GenBank, none of the variations was specific for the Alternaria spp. (Supplementary Fig. 5). Therefore, four isolates and the reference Alternaria spp. were not in separate clades base on the phylogenetic tree of GAPDH and RPB2.

The pathogenicity of the four representatives isolates from three Alternaria species, A. tenuissima, A. gaisen, and A. alternata, was evaluated on a watermelon cultivar, ‘Charleston gray’ using the detached-leaf inoculation method (Pryor and Michailides, 2002). About seven days after inoculation, dark brown lesions around agar disks of the Alternaria spp. appeared on the detached leaves (Fig. 6), similar to those observed on naturally infected leaf samples. No symptoms were observed in the control detached leaves treated with agar disk only. The average disease incidence and average disease index caused by A. tenuissima, A. gaisen, and A. alternata ranged from 42.9% to 100% and 39.4% to 68.7%, respectively (Table 5). The Alternaria spp. were re-isolated consistently from all the leaf lesions to fulfil Koch’s postulates, while no Alternaria isolates were re-isolated from the control leaves.