Introduction

The Black Pine, Pinus thunbergii Parl., native to South Korea and Japan, is adapted to coastal areas and can tolerate high salt environments [33, 47]. The important role of P. thunbergii forests as a bioshield protecting arable land and buildings from large waves and strong winds was highlighted after recent tsunamis in Indonesia [35] and Japan [50]. For this reason, efforts have been made to understand the ecology of P. thunbergii to aid in recovering and maintaining coastal pine forests in Korea [7, 44].

Endophytes are broadly defined as microbes that reside entirely in plant tissues of stems, leaves, or roots [6]. Fungal endophytes are ubiquitous, with a worldwide estimate of one million species [12]. This host-endophyte relationship can take on several forms, including mutualism, commensalism, or parasitism. Most endophytes are classified in the fungal phylum Ascomycota [1, 61], but some Zygomycota [10, 14, 19, 20] and Basidiomycota [40, 41] species have been identified.

Endophytic fungi can be beneficial to plants by producing growth-promoting metabolites, enhancing nutrient uptake, increasing tolerance to harsh environmental conditions, and enhancing defenses to herbivores and pathogens [2, 3, 40, 55]. Environmental stressors, such as salinity, determine the fungal community structure and distribution [29]. Endophytes can confer habitat-adapted tolerance to host plants, while the endophyte may receive benefits such as nutrients and protection from the host plant [40, 41]. A study examining the mutualistic relationship between a coastal dunegrass (Leymus mollis) and its endophyte (Fusarium culmorum) found rapid growth and salt tolerance when together, but slowed growth when grown separately [40, 41].

Isolating and culturing endophytic fungi are essential in studying their physiological properties and applying them for biotechnological purposes [34, 51]. However, these methods are difficult for identification because many fungal species lack distinguishing morphological characteristics when cultured [8]. Recently, the use of DNA sequencing and DNA databases has enhanced our ability to accurately classify endophytic fungi [16, 28]. We take this combined culture-sequence approach to identify the diversity and understand the ecological function of endophytic fungi associated with P. thunbergii.

This study represents the first work concerning the endophyte diversity in P. thunbergii roots. Owing to the unique microhabitat of a coastal environment, it is important to identify the fungal community of salinetolerant fungi that have a potential of enhancing the survival and proliferation of P. thunbergii. To understand the geographical and ecological variation, we compared across three regions along the eastern coast of South Korea, sampling tree roots adjacent to the shoreline and inland.

Materials and Methods

Study Site and Collection

In the fall of 2012, we sampled three regions spanning the eastern coast of Korea: Goseong, Uljin, and Busan (Fig. 1). Except for Busan, where woody and herbaceous plants are syntopic with P. thunbergii, the sites are primarily occupied by P. thunbergii. In all regions, the P. thunbergii community is composed of mature trees estimated to be ~60 years old, located at least 200 m inland from the shoreline, and several hundred meters wide. A total of 18 root sections were taken at each site: three root sections from six trees (three trees adjacent to the shoreline and three trees inland). Inland samples were >200 m from the samples adjacent to the shoreline. In total, 54 roots from 18 mature trees were collected. The average length of fine root samples was approximately 30 cm. Additionally, soil from around the roots of all trees was sampled (18 total) to test soil properties. Salinity, pH, total nitrogen, and total phosphorus were analyzed for each sample at the National Instrumentation Center for Environmental Management (Seoul National University, Korea).

Fig. 1.The three study regions along the eastern coast of Korea in this study. Δ indicates each collection point. At each study site, root samples were collected from six Pinus thunbergii individuals. Three collection points were located adjacent to the shoreline, while the other three points were ~200 m inland. SB = Sandy Beach, PF =Pin e Forest, UA =U rban Area, MF =M ixed Forest, C= Cliff.

Endophyte Isolation and Saline Resistance Test

Root surfaces were sterilized using standard procedures [46]. These roots were cut to approximately 5 mm in length for isolation of endophytic fungi. Under sterile conditions, the root fragments were plated on potato dextrose agar (PDA) medium with 100 ppm streptomycin, incubated at 20℃ in the dark for several days, and subcultured from mycelia margins to a new plate to obtain pure cultures. All pure culture isolates were grouped based on culture characters and microscopic features (Nikon 80i microscope). A representative isolate of each group was selected for DNA sequencing and evaluating salt resistance. Testing for salt resistance was performed by comparing growth rates on PDA with and without 3% NaCl, each performed in triplicate and averaged. We set 3% NaCl for the growth media to approximate the average seawater salinity of 3.5% [22]. Results are expressed in a ratio of mean growth rate with NaCl/mean growth rate without NaCl.

Molecular Approach for Identification of Endophytic Fungi

DNA was extracted using a modified CTAB extraction protocol [42]. The internal transcribed spacer region (ITS) was amplified using primer set ITS1F/ ITS4 [57]. PCR was performed on a C1000 thermocycler (Bio-Rad, USA) using the Maxime PCR PreMix-StarTaq (Intron Biotechnology Inc., Korea) in a final volume of 20 μl containing 10 pmol of each primer and 1 μl of DNA. The PCR conditions were 95℃ for 10 min, followed by 35 cycles of 95℃ for 40 sec, 55℃ for 40 sec, and 72℃ for 1 min, and a final extension step at 72℃ for 10 min. The PCR products were electrophoresed through a 1% agarose gel stained with loading STAR (Dyne Bio, Korea) and purified using the Expin PCR Purification Kit (GeneAll Biotechnology, Korea) according to the manufacturer’s instructions. DNA sequencing was performed at the DNA Synthesis and Sequencing Facility, Macrogen (Korea), using an ABI3700 automated DNA sequencer. The sequences were deposited in GenBank (Table 1).

To identify representative endophytes, we use BLAST to compare sequences against type strains [25]. In GenBank, we searched for sequences that have the identifiers “type strain”, “extype”, “holotype”, or “isotype” in the description. Since type strains available in GenBank are limited, we also included the top matching BLAST match for each sample. A phylogenetic tree was constructed to assign species names to samples. When a sequence clustered with a single species, either a type sequence or not, this identification was chosen. When the sequence was highly similar to several different species, it was only classified to the genus level.

Table 1.aClassification: A/D = Ascomycota/Dothideomycetes; A/E = Ascomycota/Eurotiomycetes; A/L = Ascomycota/Leotiomycetes; A/S = Ascomycota/Sordariomycetes; B/A = Basidiomycota/Agaricomycetes; Z/I = Zygomycota/Incertae sedis. bStrain: SFC = Seoul National University Fungus Collection. cNo. of isolates: S = shoreline; I = inland. dGrowth rate with 3%NaCl/growth rate without 3% NaCl. eTaxa in bold are those that are the most similar to type strains.

Sequence data were assembled, aligned, and edited using MEGA 5 [48]. A maximum likelihood phylogenetic analysis was conducted with RAxML [45], using the GTRGAMMA model of evolution for tree inference and 1,000 bootstrap replicates.

Statistical Analyses

The endophyte community structure between regions and between shoreline/inland samples was compared using Mothur ver. 1.31.2 [43]. Mothur commands used for analyses are written in italics. A distance matrix describing the similarity in community membership and structure was generated using the dist.shared command under the theta similarity measure [60]. A non-metric multidimensional scaling (NMDS) ordination graph was plotted (nmds) and an AMOVA (amova) was performed to determine whether the groupings based on region and shoreline/inland were statistically significant.

To analyze soil variables, a nested 2-way ANOVA was performed for each variable using R ver. 3.0.1 [37] comparing the effect of region (Goseong, Uljin, Busan) and position (shoreline, inland). When significant effects were found, a TukeyHSD post-hoc test was used to identify the significant pairs.

Results

Isolation and Identification of Endophytic Fungi

A total of 259 endophytes were isolated from P. thunbergii roots (Table 1). In aggregate across all regions, fungal isolates were grouped into 136 distinct morphotypes based on morphological features of cultures. For individual regions, 43 morphotypes (53 isolates) were present in Goseong, 60 morphotypes (98 isolates) in Uljin, and 58 morphotypes (108 isolates) in Busan. With morphological characters alone, we were able to classify most morphotypes to the genus level: Fusarium, Penicillium, Trichoderma, and Umbelopsis. These genera accounted for a majority of the isolates (76%).

The phylogeny used for identification of endophytic fungi was pruned of distantly related taxa for clarity (Fig. 2). Using ITS sequences, 58 species representing the three major fungal phyla were identified: 44 Ascomycota, 8 Zygomycota, and 6 Basidiomycota (Table 1). Sequences from 14 morphotypes were highly similar to type strains (Table 1, Fig. 2), and these identifications should be considered reliable, whereas the remaining sequences similar to non-type strains should be seen as tentative identifications. For Ascomycota, the dominant species were Fusarium oxysporum (17 isolates), Hypocrea sp. (14 isolates), Penicillium canescens (24 isolates), and Trichoderma koningiopsis (18 isolates). Umbelopsis sp. 2 (39 isolates) and Bjerkandera adusta (5 isolates) were the most frequently isolated species for Zygomycota and Basidiomycota, respectively (Table 1).

Fig. 2.Pruned phylogenetic tree (only including closely related sequences) of endophytic fungi isolated from Pinus thunbergii roots based on Maximum Likelihood of ITS sequences. Taxa denoted with a “T” indicate type strains, while texa denoted with a “S” indicate endophyte samples from this study. Bootstrap values ≥90% are denoted with an “*”. Note the scale change for the two subtrees.

The frequency of isolation differed between species. Ascomycota species were the most common, accounting for 71.4% (185/259) of all isolates. Within Ascomycota, representation was dominated by two genera, Penicillium (76 isolates) and Trichoderma (47 isolates). The representation of Zygomycota isolates (64 isolates, 24.7%) and Basidiomycota isolates (10 isolates, 3.9%) was limited.

Regional Variation

The number of species identified from each region was similar (Goseong = 23, Uljin = 27, Busan = 29). Five species were shared across the three regions (P. atramentosum, Trichoderma hamatum, T. koningiopsis, Umbelopsis ramanniana, and Umbelopsis sp. 2), and there were many species unique to Goseong (12), Uljin (13), and Busan (17) (Table 2). Fig. 3 is a NMDS plot comparing endophyte community membership and structure across all regions, and an AMOVA analysis comparing endophyte communities across regions was not significant (F[2,3]= 1.47, p = 0.063). Comparing between shoreline/inland samples, 44 total and 29 unique species were identified from shoreline habitats, whereas 28 total and 14 unique species were identified from inland habitats (Table 2). AMOVA analyses showed that endophyte communities between shoreline and inland were not significantly different (F[1,4]= 1.07, p = 0.39).

Fig. 3.Non-metric multidimensional scaling (NMDS) plot depicting the similarity of endophyte community membership and structure across all regions. G =G oseong, U= Uljin, B=Bu san, S=sh oreline, I=in land.

Data on soil characteristics at the three regions are in Table 3. The nested ANOVA for phosphorus levels were not significant (region: F[2,12]= 0.03, p > 0.05; regionposition: F[3,12]= 1.31, p > 0.05). For pH, there was a significant effect for region (F[2,12]= 7.63, p = <0.05), but none for region-position (F[3,12]= 0.39, p > 0.05). Post-hoctests identified significant differences between Uljin-Busan and Uljin-Goseong, where Uljin (5.94) had a higher pH than Busan (5.13) and Goseong (5.1). For salinity, there was a significant effect for region (F[2,12]= 5.32, p = <0.05), but none for region-position (F[3,12]= 1.56, p > 0.05). Post-hoc tests identified significant differences between Busan-Uljin and Busan-Goseong, where Busan (3.5%) had a higher level of salinity than Uljin (1.5%) and Goseong (2%). The same pattern was seen in nitrogen levels, with a significant effect for region (F[2,12]= 6.79, p < 0.05), but not for regionposition (F[3,12]= 0.9, p > 0.05). Post-hoc tests identified significant differences between Busan-Uljin and Busan-Goseong, where Busan (35.4%) had a higher level of nitrogen than Uljin (12.2%) and Goseong (8%).

Table 2.Summary of unique species richness and saline-tolerant species from each of the regions.

Table 3.Summary of soil properties at the collection regions.

Growth Rate on PDA with 3% NaCl

Growth rates of isolates on PDA media with and without NaCl were compared (Table 1). A ratio >1 indicates higher growth rate on PDA with NaCl. Species with a growth ratio >1 are subsequently referred to as “saline-tolerant.” Among the 58 species, 18 species were saline-tolerant. Of these 18 species, 17 species were in Ascomycota and one in Zygomycota. Comparing across the three regions, Busan had the highest total number of saline-tolerant endophytes (14 species), followed by Uljin (10 species), and Goseong (6 species) (Table 2). Several saline-tolerant endophytes were unique to a single region (Busan = 7, Goseong = 1, Uljin = 1). Of the species with the greatest reduction in growth rate (≤0.1), three species were from Basidiomycota (Irpex lacteus, Merulius tremellosus, Phanerochaete sordida) and one from Ascomycota (Simplicillium lamellicola).

Evaluating between shoreline and inland positions, 17 endophytic fungi isolated from shoreline were salinetolerant, with 11 being unique, compared with seven total and one unique saline-tolerant species from inland habitats (Tables 1 and 2). The three species with the highest growth ratios were unique to the shoreline of Busan (Penicillium citreonigrum, Penicillium roseopurpureum, Myxotrichum stipitatum) (Tables 1 and 2).

Discussion

Endophytic Diversity

A wide diversity of fungal endophytes was isolated from P. thunbergii roots. Representatives of all three fungal phyla were represented, with the majority of species being in Ascomycota (75.8%), followed by Zygomycota (13.8%), and Basidiomycota (10.2%). Our results mirror previous studies where most endophytes identified are in Ascomycota [1, 10, 14].

In Ascomycota, the majority of isolated endophytes were Penicillium and Trichoderma species. Penicillium and Trichoderma are usually saprophytes, but are also common as plant symbionts, playing roles in plant growth, nutrient absorption, resistance to harsh environments, defense response against pathogens, and resistance from attack of herbivores [18, 21, 23]. For example, a Penicillium species isolated from sand dune plants has the ability to promote growth in the host plant by producing secondary metabolites [5, 23]. Zygomycota diversity in our samples was limited to seven species, but dominated by Umbelopsis sp. 2. Umbelopsis species are common in forest soil and have potential ecological interactions with plant communities [9, 27]. Six Basidiomycota species were isolated (Table 1), all which are known to be present on dead trees as wood-rotting fungi [13] and known to infect cuts in roots [26]. Owing to the ecology of these species, we believe the presence of Basidiomycota in our samples is not an endophytic relationship, but instead an opportunistic infection of the roots.

Saline-Tolerant Endophytes

The ability to resist high salinity stress is essential for survival in coastal environments. Some endophytes can improve the survival and growth of their hosts by enhancing their tolerance to environmental stresses [40, 49]. Of the saline-tolerant endophytes identified in this study, several were previously documented. Acremonium variecolor, Aspergillus sydowii, and Myxotrichum stipitatum were previously identified as part of fungal communities in hypersaline environments [15, 24, 56]. In particular, A. variecolor is a marine-derived fungi tolerant to high salt concentrations [4].

Eighteen saline-tolerant endophytes were isolated in this study, and all but one were members of Ascomycota. Eight of these saline-tolerant Ascomycota species were in the genus Penicillium. Penicillium species are widely known to be salt-tolerant and fast growing [11, 31]. The top two saline-tolerant endophytes based on growth rate ratio were Penicillium species: P. citreonigrum (ratio = 2.3) and P. roseopurpureum (ratio = 1.9). Penicillium citreonigrum has not been identified as an endophyte, but rather the producer of a potent toxin (citreoviridin) [38, 52, 53]. It is not a commonly isolated species, but is widely distributed [39]. Penicillium roseopurpureum, on the other hand, has been found as an endophyte of coffee berries, but with unknown function [54].

Comparison Between Regions and Positions

Although the overall statistical analyses on endophyte community structure, both by region and position, were not significant, there are noteworthy similarities and differences when comparing endophyte communities. All regions shared five species, of which three species (P. atramentosum, T. hamatum, T. koningiopsis) are saline-tolerant. Previous research found some plant species benefit from increased growth and protection from plant pathogens when associated with T. hamatum [17, 32] and T. kongingiopsis [30, 36]. The presence of shared endophytes across regions, especially saline-tolerant ones, suggests an important relationship with P. thunbergii roots.

When comparing between regions, Busan had the highest proportion of saline-tolerant species, including the top three species with the highest growth ratios (Table 2). Previous studies have found that differences in soil properties reflect differences in endophytic diversity [58, 59]. ANOVA analyses of soil properties detected statistical differences between the three regions. Busan soils had higher salinity and total nitrogen levels, which may be factors leading to the increased number of saline-tolerant endophytes in the region. It should also be noted that the plant community in Busan, with a combination of woody and herbaceous plants present, was more diverse than the other two regions, which may also have contributed to the differences in endophyte communities.

Comparing between positions, although there were no statistical differences in soil properties, shoreline habitats harbored higher saline-tolerant endophytic diversity. Of the 18 total saline-tolerant endophytes, 17 were isolated from shoreline soils. Moreover, the three saline-tolerant species with the highest growth ratio were unique to shoreline habitats. These observations suggest that endophytic fungi living closer to the coast may have a higher resistance to salinity and potentially have specific relationships with P. thunbergii.