1. Introduction
Endophytic fungi are ubiquitous and comprise a highly diverse polyphyletic group of microorganisms that reside internally in plant tissues, which are often asymptomatic for at least part of their life cycle [
1,
2,
3,
4]. Endophytes exist in diverse ecological relationships within their host, ranging from symbiosis to antagonism against host pathogens. Currently, endophytes have been examined in approximately 300,000 plant species, and almost all the plants were found to harbor at least one or more endophytes [
5,
6,
7]. It is estimated that 1.5 million fungal species reside in plants, but only 70,000–100,000 fungal species have been identified, which represents only 7% of the fungal endophytes [
8,
9]. These previous studies indicate the need for more intensive studies to identify the fungal endophytes associated with plant species under diverse environments and geographic locations. Fungal endophytes colonizing medicinal plants have been more commonly studied [
5,
6,
10,
11,
12,
13,
14]. Endophytes associated with ornamental plants remain largely unexplored, and this study examined an ornamental tree commonly known as flowering dogwood (
Cornus florida L.). that is native to the northeastern and southeastern region of the United States. Although
C. florida is a well-known ornamental plant with high economic importance [
15,
16], the study focused on a limited area in Mid-Tennessee where planting is in large nurseries, forest undergrowth, landscaping, and backyard gardens. Traditionally,
C. florida has been used in treatment of malaria [
16,
17] and
Cornus spp. are known to possess antimicrobial, anticancer, and antidiabetic properties [
16,
17,
18,
19,
20]. Furthermore, numerous pharmacological benefits, phytochemicals, and therapeutic applications have been associated with this
Cornus [
16,
17]. In contrast, the microbial diversity, and their biological potential of
Cornus endophytes remains largely unexplored. The association of pathogenic fungi in foliar diseases of
Cornus, such as dogwood anthracnose caused by
Discula destructive Redlin [
21,
22] and powdery mildew caused by
Erysiphe pulchra [
15,
23] have been well studied. Although reports are available on beneficial epiphytic fungi associated with
C. stolonifera (red-osier dogwood) in British Columbia [
24],
C. controversa (giant dogwood) [
25] and
C. florida in middle Tennessee [
26], information is lacking on the diversity of endophytic microorganism that reside in
C. florida and their potential role in plant protection.
Biodiversity studies have demonstrated that some endophytes are saprophytes, while others are latent or opportunistic pathogens that may become quite aggressive when the host plant is stressed and the environment is favorable to disease development [
27,
28]. It has been hypothesized that some fungal endophytes, which grow without causing disease in their host plants, may have evolved from pathogenic strains [
29,
30,
31]. Numerous studies have shown that endophytic fungi can play an important role in protecting plants against pathogens, promoting plant growth, and increasing host resistance to abiotic stress [
4,
31,
32,
33]. Thus, for plant health management, it is critical to understand the microbial diversity of endophytes and their potential as pathogens or biocontrol agents. Endophytic fungi have been shown to produce various metabolites including hydrolytic enzymes, such as cellulases, amylases, chitinase, protease, xylanases, etc. [
4,
34,
35,
36,
37]. These enzymes are known to play a role in tissue penetration/colonization, nutrient acquisition, and disease suppression that benefit both plant pathogens and biological control agents. The objectives of this study were to (a) understand the diversity of fungal endophytes that reside in
C. florida, (b) evaluate the potential role of these fungal endophytes as plant pathogens and biological control agents against selected diseases, and (c) explore the potential of these fungal endophytes to produce extracellular enzymes that may benefit plant health.
3. Discussion
Cornus florida, is one of the most economically important ornamental plants, with great ethnobotanical value [
15,
17]. In this study, diverse endophytic fungi were isolated from stems of trees that did not display disease symptoms, and the diversity of endophytic fungi within
C.
florida was revealed for the first time. The fungal genera recovered in this study that have also been previously reported as endophytes in other plant species include
Hypoxylon,
Pestalotiopsis,
Xylaria,
Cytospora,
Diaporthe,
Daldinia,
Colletotrichum,
Nigrospora,
Botryosphaeria,
Diplodia,
Alternaria,
Phoma,
Cladosporium, and
Peniophora [
12,
13,
14,
27,
41,
42,
43]. Because previous studies by Huang et al. [
10] reported that fungal endophytes are more frequent in stems as compared to other plant tissues, this study focused on fungal endophytes in the stem of flowering dogwood, and unveiled a rich and diverse community of fungal endophytes associated with
C. florida stems. More studies are needed to advance our understanding of the role of diverse endophytes in plant health.
ITS sequence-based analysis has previously been reported as a reliable method for fungal identification in numerous studies. The ITS-based phylogenetic analysis presented herein represents the first such study for fungal endophytes in
C. florida. Based on the ITS sequence data, the stem samples collected from eight locations yielded 59 distinct taxa of fungal endophytes. Out of these, 39 taxa (254 isolates) were recognized to the species level, 13 taxa (84 isolates) to the genus level, 4 taxa (26 isolates) to the order level, 1 taxon (2 isolates) to the class level, and two were unidentified taxa (3 isolates). These results confirmed the diverse range of the endophytic fungal species associated with
C. florida. Most of the endophytic fungi belonged to phylum Ascomycota, and this finding is consistent with other studies on diverse plant species [
12,
13,
42,
44]. The low prevalence of the phylum Basidiomycota in
C. florida is also consistent with previous reports on other plant species [
14,
42,
45]. Sordariomycetes and Dothideomycetes were the most dominant and diverse classes of endophytes found in this study, which is similar to endophytic fungal communities recovered from
Glycine max [
41] and
Ocimum sanctum [
12]. Futhermore, fungi within the Xylariales and Pleosporales orders were major endophytic colonizers of
C. florida and comprised more than 75% of the total isolates.
In this study, the number of tissue segments used in the isolation of fungal endophytes was similar (150) among sites A–F, but lower in sites G (107) and H (75). However, the relative abundance of isolates recovered varied among sampling sites and was highest at sampling site F followed by Site E, and lowest at site A (
Table 1). The total number of tissues sampled can be used as a measure of sampling effort, and more species are likely to be encountered with more intensive sampling. However, the total number of species detected varied among the different locations, ranging between 1.34 and 3.05 as determined by Menhinick’s index (Dmn) [
12,
46]. It was observed that a higher isolation frequency does not necessarily result in higher diversity in any sampling site. Species diversity indices at different locations (
Table 2) were evaluated using the Shannon–Wiener index (H´), Simpson’s diversity index (1-D), and Margalef’s index (Dmg) [
11,
12,
44,
46,
47]. Species diversity reflects how many different types of taxa are present in communities, and considers both species richness as well as the dominance/evenness of the species. Species diversity from the Shannon–Wiener index showed moderate variation among sampling sites (1.89–2.67), while analysis of Margalef’s index showed high variation among different locations. Simpson’s Index is a measure of dominance and weights toward the abundance of the most common taxa, and Simpson’s diversity index was uniform. The fungal dominance assessed by Camargo’s index (1/Dmn), where Dmn is species richness [
12,
47] and species richness is a measure of the number of species (or other taxonomic level) present at a site (
Table 2). The Pielou evenness index ranges between 0 and 1.0, with lower values reflecting more variation in abundances among different taxa within the community. Our results show the Pielou evenness index of 0.85–0.95 (
Table 2).
The simplest measure of species richness is just the number of species that have more than one individual recorded per site. Thus, species richness was determined by the total number of different species recorded in a sample, at the sampled site. The species richness among different locations was determined by Menhinick’s index (Dmn) [
12,
46]. Overall, differences were observed in fungal endophyte frequency in
C. florida colonization, species diversity, richness indices, and similarity indices of endophytic fungal communities residing in stems of
C. florida at different locations. These results provide evidence that the host–endophyte interaction is greatly influenced by the micro-environment, even when two communities are in proximity. This finding is consistent with previous reports by Silva-Hughes et al. [
14] and Yokoya et al. [
48]; however, the fungal endophyte community could also be influenced by the host genetics and development. In addition, some species were common among all sampling sites, as reflected by Sorenson similarity indices (
Table 3), thereby suggesting host affinities.
The association of fungal endophytes with latent and opportunistic fungal pathogens has been extensively studied [
27,
28,
49,
50,
51]. These latent/opportunistic fungal pathogens have the ability to asymptomatically colonize plant vascular tissue and remain quiescent while the environment is not conducive to disease development. When a host plant becomes stressed and/or the environment becomes favorable for the development or pathogenicity, the quiescent fungi can become pathogenic and cause irreversible damage [
5,
27]. Some of the fungal endophytes isolated from
C. florida in this study have been previously reported as agronomically important pathogens or saprophytes, such as species of the genera
Hypoxylon,
Diaporthe,
Diplodia,
Cytospora,
Colletotrichum,
Pestalotiopsis,
Botryosphaeria,
Phyllosticta,
Epicoccum, and
Phoma, which have been associated with leaf spots, blight, cankers, and/or dieback diseases [
27,
28,
52,
53,
54,
55,
56,
57,
58,
59]. Of these potential pathogens,
C. gloeosporioides [
60],
C. acutatum [
52], and
Botryosphaeria dothidea [
61] have been reported as destructive pathogens of
C. florida and/or other
Cornus sp.
Our results with detached leaf assays provided significant information on potential pathogens residing as endophytes or latent pathogens in healthy
C. florida. The occurrence of these pathogens in asymptomatic plants confirms the previous reports on latent phase of several endophytes before the manifestation of disease symptoms [
49,
50,
51,
59]. Some of the endophytes confirmed to have pathogenic potential on detached leaves have not been reported as pathogens of
C. florida, and the potential pathogenicity of these endophytes warrant further studies. In contrast, some fungal endophytes regarded as pathogens were unable to cause disease on detached leaves, hence supporting the interchangeable pathogen–endophytic existence of these fungi [
7,
30]. Information from this study may facilitate early response to future disease outbreaks that may adversely impact dogwood production. Isolates of fungi that are known to be pathogenic but did not display disease symptoms on pathogenicity tests indicated a possibility that they were nonpathogenic forms that existed as endophytes with potential antagonism against pathogenic forms. This phenomenon has been reported on other fungi, such as
Fusarium oxysporum, that have both pathogenic and nonpathogenic forms and the nonpathogenic forms have been developed as biological control agents against the pathogenic forms. Further pathogenicity testing using whole plants is needed to confirm and identify potential pathogens of dogwoods that exist endophytically as latent pathogens.
Some fungal endophytes isolated in this study, such as
Nigrospora spaherica,
Hypoxylon sp.,
Entonema sp., etc., have previously been shown to have biological control potential against
Fusarium solani,
F. oxysporum,
Macrophomina phaseolina,
Cercospora nicotianae, Phytophthora capsici,
P. irrigata,
P. cryptogea, and/or
P. nicotianae [
62,
63]. Furthermore, endophytic
N. sphaerica was shown to reduce the disease severity of Phytopthora blight in pepper under greenhouse conditions [
63]. Based on these results, we hypothesize that latent pathogens living endophytically within the stems of
C. florida may be suppressed by antimicrobial metabolites produced by other endophytes living in the same ecological niches.
Out of the fungal endophytes associated with
C. florida, 50 were examined for the ability to produce hydrolytic enzymes, such as cellulase, amylase, laccase, pectinase, chitinase, and protease. The results suggested that these fungal endophytes have the ability to catabolize a wide range of polymeric substrates (
Table S3). Understanding the pattern of substrate utilization and the ability to produce various extracellular enzymes may help to understand the diverse ecological roles of endophytic fungi [
64]. Production of extracellular cellulase in 75% of the isolates and protease in 96% of the isolates were prominent, as compared to other hydrolytic enzymes, such as pectinase produced by about 48% of endophytic fungi. The role of pectinase in fungal pathogenesis, tissue penetration, and plant decomposition has been well established [
65]. In this study, all fungal endophytes that caused disease symptoms on
C. florida leaves were found to produce protease, and some of these isolates also produced pectinase activity. According to Choi et al. [
35], fungal endophytes may produce pectinase and/or protease if they are latent pathogens or weak parasites. Only 55% of endophytic fungal isolates tested were able to degrade starch in this study, which contrasts with the previous report by Choi et al. [
35], where all of the endophytic fungi produced amylase. In this study, over 65% of tested fungal endophytes were able to produce extracellular laccase, which can mediate the oxidation of phenolic substrates, such as lignin. These results contrast to other studies [
35,
36,
37], which found that few or no fungal endophytes produced laccase. Many nonpathogenic fungi were able to produce chitinase enzymes, which suggests their potential in the biological control of fungal pathogens or insect pests. Extracellular enzymes secreted by endophytic fungi may have a possible role in conferring resistance to pathogen attack, tissue colonization, and/or deriving nutrition from plant as a latent pathogen [
34,
35,
37]. Thus, the results obtained from enzyme assays further support a potential role of some endophytic fungi isolated from
C. florida in plant protection against fungal pathogens.