Identification and Characterization of Fungal Pathogens Associated with Boxwood Diseases in the Republic of Korea

Fungal isolation

Collected symptomatic leaves and stem cuttings (10 mm long) were surface sterilized by soaking in 1% sodium hypochlorite for 1 min, 70% ethanol for 1 min, and were then rinsed with sterile water three times (Park et al., 2020). Eight or nine sterilized samples were placed on 90-mm-diameter Petri plates containing potato dextrose agar (PDA; Difco, Sparks, MD, USA) and incubated at 25°C for 3–5 days. Then, fungal colonies were transferred to complete media (CM) (Leslie and Summerell, 2006). All isolates were pure-cultured using the single-spore isolation method (Ho and Ko, 1997) and stored in 20% glycerol at −80°C until use. Representative isolates of P. buxi and N. buxi (NL1-1-1, B-2, B3L6, and B2L2-1) were deposited to the Korean Collection for Type Cultures (KCTC) under deposit numbers KCTC56889-56892, respectively.

Morphological identification

Each fungal isolate was grown on PDA and CM agar at 25°C for 10 days. Spores were collected from cultures grown on oatmeal agar (OA) or pine needle agar (PNA) at 25°C under near-UV light for 21 days. The structure of spores was observed with light microscopy at 400× (Leica DM500, Jena, Germany). The average dimensions of spores were recorded from 100 spores per isolate. Pictures of fungal colonies were taken 10 days after inoculation. ImageJ was used to measure the length and width of spores (Schneider et al., 2012).

Molecular identification and phylogenetic analyses

Fungal genomic DNA was extracted following the standard protocol (Leslie and Summerell, 2006). Initially, the internal transcribed spacer (ITS) region of 18S–28S nuclear DNA (rDNA) was amplified using the primers ITS4 and ITS5 (White et al., 1990). Polymerase chain reaction (PCR) was performed in 20 μl reactions of AccuPower ProFi Taq PCR PreMix (Bioneer, Daejeon, Korea), 20 pmol of each primer, and approximately 20 ng of template DNA followed by previously described parameters (White et al., 1990). Amplified PCR products were purified using the MEGAquick-spin Plus Total Fragment DNA Purification Kit (Intron Biotechnology, Seongnam, Korea) and were then sequenced (Macrogen, Seoul, Korea). The consensus sequences of each isolate were used as queries to conduct a BLASTn search against the NCBI GenBank database to obtain preliminary identification.

Additional loci, including β-tubulin and the D1/D2 region of the large subunit of the rDNA gene cluster (LSU), were sequenced for phylogenetic analyses using the primer pairs Btub4Rd/Btub2Fd (Woudenberg et al., 2009) and LR0R/LR5 (Rehner and Samuels, 1994; Vilgalys and Hester, 1990), respectively. The sequences of the representative isolates were deposited in GenBank (GenBank accession nos. M169273-OM169276, OM176571-OM176574, ON340774-ON340791). The reference strains of P. buxi (CBS 114049 and CBS 324.53), P. foliicola (CBS 122566 and CBS 123190), and N. buxi (CBS 113714, CBS 116.75, and CBS 214.31) were obtained from the culture collection of the Westerdijk Fungal Biodiversity Institute (Utrecht, The Netherlands). A consensus sequence was manually generated for each set of forward and reverse sequences in MEGA-X (Kumar et al., 2018). Additional sequences of N. parvum strains (CBS 123652, CBS 115186, and CBS 121486) were obtained from the GenBank database (Yang et al., 2017). Sequences of Sarocladium kiliense and Pseudofusicoccum ardesiacum were used as the outgroups in the analysis of Pseudonectria species and N. buxi, respectively.

Pathogenicity tests on boxwood leaves

Pathogenicity tests using boxwood leaves were performed as previously described (Shi and Hsiang, 2014b). Detached leaves of 1-year-old boxwood plants were surface-sterilized by soaking in 1% sodium hypochlorite for 1 min, placing in 70% ethanol for 1 min, rinsing with sterilized distilled water three times, and drying. For each fungal isolate, 5-mm agar plugs containing fungal culture were placed on a wounded spot on the adaxial side of ten detached leaves. Then, inoculated leaves were placed on moist filter papers in transparent plastic boxes incubated at 25°C with 12 h light/12 h dark for 7 to 21 days. Fungal pathogens were reisolated from symptomatic leaves to fulfill Koch’s postulates.

Survey of fungal distribution on boxwood

Diseased tissues of boxwood that had symptoms ranging from yellow leaves to entirely dead shoots were found in some regions of Seoul, Republic of Korea. Leaves and branches were dry and dead, but the whole plants were still alive (Fig. 1A). On some symptomatic leaves, black fruiting bodies appeared as small black dots on the lower leaf surface (Fig. 1B). To isolate fungal strains associated with these disease symptoms, dozens of samples were collected from symptomatic boxwood hedges during July and August 2020. A total of 142 fungal isolates including previously reported four P. buxi strains (B2S72-1, B2S72-3, B3B2, and B4L3-3) (Shin et al., 2022) were successfully obtained and identified as 34 species using ITS sequence analysis (Table 1). The most frequently isolated species was Diaporthe eres (29 isolates), followed by Alternaria alternata (24 isolates) and various Fusarium species (13 isolates). Thirteen isolates were identified as putative pathogens of boxwood; P. buxi was found the most often, with eight isolates, along with two isolates of P. foliicola and three isolates of N. buxi.

Fig. 1

Blight disease symptoms of boxwood. (A) Blight symptoms on boxwood shrub. Diseased leaves were dried and yellowed. (B) Black sporodochia were observed on some boxwood leaves associated with blight symptoms in the growing season.

Table 1

Identification of 142 fungal isolates from boxwood samples

Morphological characterization

Fungal isolates of P. buxi, P. foliicola and N. buxi with corresponding reference strains (CBS 114049 and CBS 324.53 for P. buxi, CBS 122566 and CBS 123190 for P. foliicola, CBS 113714, CBS 116.75, and CBS 214.31 for N. buxi) (Lombard et al., 2015; Yang et al., 2017) were inoculated on PDA, OA, CM, and PNA to compare colony and spore morphologies (Fig. 2, Supplementary Fig. 1). Both P. buxi and P. foliicola isolates grown on CM showed pink and white fluffy mycelia with pink conidia (Fig. 2A and B). However, the conidia of P. foliicola were markedly smaller than those of P. buxi, as previously reported (Figs. 2 and 3) (Bezerra, 1963). P. buxi field isolates identified from this study produced distinctly shorter conidia compared to those of CBS reference strains CBS 114049 and CBS 324.53 (Fig. 3A), and the conidia of all tested P. buxi strains showed slightly variable widths (Fig. 3B). The average spore length of P. foliicola isolates (B4L1f and B4L2f) was longer than that of the CBS 123190 strain and shorter than that of the CBS 122566 strain (Fig. 3C). Both P. foliicola CBS reference strains produced slightly wider conidia than P. foliicola strains isolated from this study (Fig. 3D).

Fig. 2

Morphological characterization of Pseudonectria buxi, P. foliicola, and Neofusicoccum buxi. Morphological characteristics of P. buxi (A), P. foliicola (B), and N. buxi (C) strains isolated from symptomatic boxwood leaves and reference strains (CBS strains). Mycelial growth is shown on potato dextrose agar (PDA), complete media (CM) and oatmeal agar (OA) at 18 days post inoculation (dpi). Microscopic pictures of conidia are on the right side. Scale bars = 20 μm.

Fig. 3

Conidia size measurement of Pseudonectria buxi and P. foliicola. Measurement of conidia length (A, C) and width (B, D) from 100 conidia of each isolate of P. buxi (A, B) and P. foliicola (C, D). Conidia were collected at 18–21 dpi on oatmeal agar. The ImageJ program was used for conidia measurement. Different letters above values indicate significant differences (P < 0.05, Tukey’s test in one-way ANOVA).

Three CBS N. buxi strains (CBS 116.75, CBS 214.31, and CBS 113714) have been reported until now (Yang et al., 2017), and they showed various morphological characteristics on CM, PDA, and OA (Fig. 2C). Furthermore, field isolates (B-2, B3L6, and B2L2-1) identified as putative N. buxi showed notably different phenotypes compared to CBS strains. Colonies of N. buxi isolates showed more wrinkled mycelia and blackish gray color. Furthermore, all N. buxi isolates produced significantly larger spores than the CBS N. buxi strains (Figs. 2C and 4). Conidia of CBS strains were 18–20 × 5–7 μm in size, while conidia of isolates annotated as N. buxi were 25–27 × 10–12 μm in size (Fig. 4). The results suggest that fungal strains predicted as N. buxi in this study have different morphological and physiological characteristics compared to previously reported CBS strains (Yang et al., 2017).

Fig. 4

Conidia size measurement of Neofusicoccum buxi. Measurement of conidia length (A) and width (B) from 100 conidia of each isolate of N. buxi. Conidia were collected at 18–21 dpi on pine needle agar. Different letters above values indicate significant differences (P < 0.05, Tukey’s test in one-way ANOVA).

Phylogenetic analyses

Multigene alignment was performed using DNA barcode genes of ITS, LSU, and β-tubulin. We used the neighbor-joining and maximum likelihood methods to analyze the alignment. Sequence analysis of P. buxi and P. foliicola strains produced a total length of 1,576 bp (ITS + LSU + β-tubulin). Fungal isolates identified as P. buxi were segregated into three subclades, and two CBS P. buxi strains were separated (Fig. 5A). However, we could not distinguish distinct morphological characteristics of mycelial colonies and/or spores among the three subclades (Figs. 2 and 5A).

Fig. 5

Phylogenetic analyses of Pseudonectria spp. and Neofusicoccum buxi. Neighbor-joining phylogenetic trees based on the concatenated ITS, β-tubulin, and LSU sequences of Pseudonectria spp. (A) and N. buxi (B). Sarocladium kiliense and Pseudofusicoccum ardesiacum were used as outgroups.

We confirmed that isolates obtained in this study were closely related to N. buxi strains separated from N. parvum (Fig. 5B). Moreover, they formed a separate branch in the phylogenetic tree, distinguishable from CBS N. buxi strains (Fig. 5B). Each isolate had 30 variable sites compared to CBS N. buxi strains, showing 98.05% identity. All three field isolates of N. buxi showed identical sequences, although there were morphological variances (Figs. 2C and 5B).

Pathogenicity tests

Virulence tests were conducted by inoculation of mycelial plugs of fungal strains on healthy boxwood leaves and incubated for 7 days in the case of P. foliicola and 21 days in the case of P. buxi and N. buxi (Fig. 6). Fungal strains were reisolated from inoculated leaves to fulfill Koch’s postulates. It was previously reported that P. buxi can invade through wound sites in boxwood leaves (Shi and Hsiang, 2014b). For this reason, we predicted that P. foliicola, would have an invading mechanism similar to that of P. buxi. Detached leaves with isolates of P. buxi and P. foliicola turned dark brown, and pink sporodochia were produced on the back sides of leaves. Leaves that were inoculated with CBS strains and field isolates of P. foliicola showed faster symptom progression than those inoculated with P. buxi strains (Fig. 6A and B). All isolates of N. buxi, as well as CBS strains, produced black perithecia on the back sides of detached leaves (Fig. 6C). Moreover, all isolates of N. buxi caused straw-colored leaves.

Fig. 6

Pathogenicity tests of Pseudonectria buxi, P. foliicola, and Neofusicoccum buxi. Pathogenicity tests were conducted, and symptoms were evaluated on detached leaves using isolates of P. buxi (A), P. foliicola (B), and N. buxi (C). Detached boxwood leaves were cut and inoculated with hyphal plugs. After 7–21 days of incubation, brownish or yellowish colored leaves were detected, whereas no obvious symptoms were observed on the mock-inoculated leaves.