Phylogenetic diversity and antioxidant activities of culturable fungal endophytes associated with the mangrove species Rhizophora stylosa and R. mucronata in the South China Sea

Mangrove endophytic fungi can produce impressive quantities of metabolites with promising antioxidant activities that may be useful to humans as novel physiological agents. In this study, we investigated the phylogenetic diversity and antioxidant potential of 46 fungal endophytes derived from the mangrove species Rhizophora stylosa and R. mucronata from the South China Sea. The fungal isolates were identified using a combination of morphological characteristics and phylogenetic analysis of the internal transcribed spacer (ITS) sequences. Seventeen genera belonging to 8 taxonomic orders of Ascomycota were discovered, specifically, Botryosphaeriales, Capnodiales, Diaporthales, Eurotiales, Glomerellales, Hypocreales, Pleosporales, and Xylariales. The most abundant fungal orders included Xylariales (35.49%) and Diaporthales (27.61%), which were predominantly represented by the culturable species Pestalotiopsis sp. (34.54%) and Diaporthe sp. (18.62%). The stems showed more frequent colonization and species diversity than the roots, leaves, hypocotyls, and flower tissues of the host plant. The antioxidant activities of all the isolated fungal extracts on four different culture media were assessed using improved 2,2′-diphenyl-1-picrylhydrazyl (DPPH) and 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulphonicacid) (ABTS) methods. A relatively high proportion (84.8%) of the isolates displayed antioxidant capacity (%RSA > 50%). Further research also provided the first evidence that HQD-6 could produce flufuran as a significant radical scavenger with IC50 values of 34.85±1.56 and 9.75±0.58 μg/mL, respectively. Our findings suggest that the utilization of a biotope such as that of the endophytic fungal community thriving on the mangrove plants R. stylosa and R. mucronata may be suitable for use as a sustainable resource for natural antioxidants.


Sampling site and plant material
Healthy roots, stems, leaves, hypocotyls and flowers of R. stylosa (Rhizophoraceae) and roots, stems, leaves and flowers of R. mucronata were collected from a specific location (110˚32 0 -1103 7 0 E, 19˚51 0 -20˚01 0 N) in Dong Zhai Gang-Mangrove Garden on Hainan Island, China. The collected plants were authenticated by Prof. Dr. Xiaobo Yang (Hainan University), with a voucher specimen preserved in the herbarium of College of Landscape and Horticulture, Hainan University.

Isolation of endophytic fungi
To obtain the endophyte within different parts of the plant, surface sterilization was carried out following an isolation protocol described by Proksch's group [24] with some modifications. Samples were initially washed with tap water, then surface sterilized by dipping in 75% ethanol for 1 min, in bleach solution containing 1.3% sodium hypochlorite for 3 min, in 75% ethanol for 30 s and finally were rinsed three times (3 min each) in sterile Millipore water. The surface sterilized samples were cut into small segments using a sterile scalpel and were incubated on PDA plates supplemented with streptomycin to suppress bacterial growth at 28˚C in darkness for 2-4 weeks and were checked daily for hyphal growth. Single isolate were transferred to fresh PDA plates using the hyphal tip method. Culture purity was determined from colony morphology. Voucher strains are deposited at one of the authors' laboratory (J. X.). http://dx.doi.org/10.17504/protocols.io.jhmcj46 [PROTOCOL DOI]

Identification of endophytic fungi
The endophytic fungi were identified in combination of morphologic characteristics and p internal transcribed spacer (ITS) sequences.

DNA extraction, PCR amplification, sequencing and phylogenetic analysis
Genomic DNA was extracted from fresh fungal mycelium collected in a PDA plate using cetyltrimethylammonium bromide (CTAB) method [25]. The internal transcribed spacer (ITS1-5.8S-ITS2) regions of the fungi were amplified with the universal ITS primers, ITS1F (5 0 CTTGGTCATTTAGAGGAAGTAA 3 0 ) and ITS4 (5 0 TCCTCCGCTTATTGATATGC 3 0 ) [26], using the polymerase chain reaction (PCR). The amplified products were submitted for sequencing (Invitrogen, Shanghai) and aligned with the sequences in the GenBank by Basic Local Alignment Search Tool (BLAST) programs to find out the sequence homology with closely related organisms. When the top three matching BLAST hits were from the same species and were ! 98% similar to the query sequence, this species name was assigned to the selected isolate. Fungal ITS-rDNA sequences of 46 representative isolates were deposited in GenBank under the accession number KX618209 and numbers ranging from KX631698 to KX631742 are provided in Table 1. Phylogenetic analyses of the endophytes were carried out using MEGA 5.1 software. The neighbor-joining (NJ) method was used to infer the evolutionary history of the fungal isolates, and the bootstrapping was carried out using 1,000 replications [27]. http://dx.doi.org/10.17504/protocols.io.jinckde [PROTOCOL DOI]

Fermentation and extraction
The fungal strains were cultured on Petri dishes of potato dextrose agar (PDA) at 28˚C for 5 days. The agar patches with the title fungus were inoculated into Erlenmeyer flasks (1000 mL). To monitor the influences of the fermentation media towards the fungal production of antioxidant metabolites, four different media (Table 1) such as Potato Dextrose Agar (PDA), Czapek's Agar (CZA), Rice Medium (RM), Grain Medium (GM) were adopted for the stationary fermentation and grown for 40 days at room temperature. The harvested cultures were extracted twice with an equal volume of EtOAc and evaporated to dryness and residues were dissolved in dimethyl sulfoxide (DMSO) to give a stock solution (

Isolation of flufuran
HQD-6 was fermentated on rice solid medium (to 100 g commercially available rice was added 110 mL of distilled water and kept overnight prior to autoclaving, 100 flasks) at room temperature under static conditions and daylight for 40 days. The mycelia and solid rice medium were extracted with EtOAc (each 200 mL) and the combined EtOAc extracts were evaporated under reduced pressure to yield 62.0 g residue. This residue was separated into seven fractions (Fr. 1-Fr. 7) on a vacuum liquid chromatography (VLC) on a short silica gel column using a step gradient elution of CH 2 Cl 2 /MeOH (v/v 0:100-100:0). Fr. 3 (5.0 g) was rechromatographed on a silica gel column, eluted with petroleum petroleum ether/CH 2 Cl 2 / MeOH (5:5:0.1), to provide four subfractions (fractions 3. Assay for antioxidant activity DPPH radical scavenging capacity measurement. The radical scavenging ability was estimated by using adapted 2,2'-diphenyl-b-picrylhydrazyl (DPPH) method described previously [22,28]. Thus, an aliquot of extract (10 μl) was added to 195 μl of ethanolic DPPH  [29]. The assay was carried out in a 96-well microtitre plates. Extract (10 μl) or ethanol (10 μl, control) was added to 195 μl ABTS radical solution and allowed to react for 30 min until a stable absorbance was obtained. The decrease in absorbance at 734 nm was measured against a blank (ethanol). http:// dx.doi.org/10.17504/protocols.io.jirckd6 [PROTOCOL DOI]

Statistical analysis
The relative species frequency (RF) was used to estimate a specific endophytic taxon from mangrove tissue or individual plant and was calculated as the number of isolates of an individual species from the assemblage divided by the total number of isolates of all species [30]. Camargo's index (1/S) was used to determine fungal dominance, A species is defined as dominant when satisfy RF> 1/S, where S represents species richness, equals the total number of species in an assemblage [31]. The colonization frequency (CF%) can reflect the extent of endophyte infection was calculated as CF = (N col /N t )×100, where N col is the number of segments colonized by an individual species, and N t is the total number of incubated segments [32].
The fungal endophytic diversity of each mangrove variety and the different plant tissues was estimated with Shannon-Weine Diversity index(H 0 ), Simpson's Diversity index(1-D), Shannon eveness (E) and Sorensen's index of similarity (QS) using the software EstimateS version 9.10, nonparametric indices abundance-based coverage estimator (ACE), Chao1 and Chao2 were adopted [33].
Results were expressed as mean ± standard deviation (SD) of triplicate of measurements for the DPPH and ABTS assays. Data were analyzed by ANOVA (p < 0.05) using Statistical Analysis System (SAS, version 9.1).

Results and discussion
Phylogenetic diversity of culturable fungi derived from R. stylosa A total of 135 isolates were recovered from 375 fragments of R. stylosa. Based on morphological characteristics, 25 independent representatives were selected for sequencing (S1 Fig). These isolates were identified based on ITS sequences or their closest neighbors retrieved from NCBI. BLAST analysis indicated that the majority of the isolated endophytes displayed diverse taxonomic affinities. Phylogenetic reconstruction based on the neighbor-joining (NJ) algorithm indicated that they could be classified into five taxonomic orders: Botryosphaeriales, Capnodiales, Diaporthales, Hypocreales, and Xylariales. Some representatives of the predominant order, Xylariales, including isolates from the genera Pestalotiopsis and Seiridium, were recovered. This is because they rapidly sporulate and are easily cultured. Diaporthales species were mainly isolated from stem tissues and represented by the genera Phomopsis sp., Diaporthe, Valsa, and Cytospora. Botryosphaeriales, represented by the genera Botryosphaeria, Lasiodiplodia, Neofusicoccum, and Phyllosticta, grew widely across the different tissues. The least abundant orders were Capnodiales and Hypocreales, which were represented by the genera Cladosporium and Fusarium, respectively. Among the 25 fungal isolates identified, the predominant fungi observed were Neopestalotiopsis protearum (31.11%), Pestalotiopsis spp.  Table 2). The highest endophytic fungi colonization frequency was observed in the stems (CF = 53.33%), followed by the hypocotyls (49.33%) and roots (25.33%). The lowest frequencies were observed in the flowers (6.67%) and leaves (4.00%) ( Table 3). As assessed using Camargo's index, the predominant endophytic fungal species differed between each tissue, with the exception of Neopestalotiopsis protearum and Pestalotiopsis spp., which were the predominant species in the roots, stems, and hypocotyl tissues. Some predominant endophytes exhibited high tissue specificity, e.g., Phomopsis longicolla, Lasiodiplodia pseudotheobromae, and Diaporthe eucalyptorum. These were determined to be the predominant species in the stems, but they were not detected in any other tissue types (Figs 1 and 2A).

Phylogenetic diversity of culturable fungi derived from R. mucronata
With regards to R. mucronata, 21 isolates representing 90 endophytic fungi were sequenced using the same approach as R. stylosa (S2 Fig). We were able to further categorize these isolates into 13   Antioxidant endophytic fungi from Rhizophora spp.

Comparison of culturable fungal diversity in two mangrove species
We selected the native mangrove species R. stylosa and an exotic mangrove species introduced from Australia to the South China Sea, R. mucronata and investigated their endophytic community diversity. The results indicated that the 46 independent representatives out of the 220 isolates from R. stylosa and R. mucronata that were selected for sequencing represented 17 genera belonging to eight taxonomic orders of Ascomycota. Pestalotiopsis sp. (34.54%), Diaporthe (18.62%), and Neofusicoccum (14.54%) were found to be the predominant culturable genera. The most frequent colonizers of R. stylosa and R. mucronata were Neopestalotiopsis protearum and Lasiodiplodia theobromae, respectively. These representative isolates and their best matches in the NCBI database are summarized in Table 2 Figs 1 and 2). To characterize the endophytic community, isolates from the various tissues were compared using the Shannon-Wiener and Simpson's diversity indices ( Table 4). The Shannon-Wiener diversity index (H 0 ) indicated that R. mucronata (2.63) exhibited higher fungal species diversity than R. stylosa (2.52), and similarly, Simpson's diversity index (1−D) suggested higher fungal endophyte diversity in R. mucronata (0.91) than in R. stylosa (0.86). The Shannon and Simpson diversity indices were highest in the stems than the other tissues in both mangrove species (H 0 = 1.32 and 1.54; 1−D = 0.88 and 0.90), suggesting that these endophytic fungi have a preference for stem colonization. The Shannon evenness (E) index of R. mucronata (0.85) was similar to that of R. stylosa (0.78), indicating uniform species composition across both hosts. The values obtained in these analyses indicate that the endophytic fungal diversity in R. mucronata and R. stylosis is relatively high. The Sorensen's similarity indices (QS) between the endophytes of both mangroves and various tissues ranged from 0.12% to 0.47% (Table 5), indicating significant variation in the endophytic fungal biodiversity from the different tissues of both plants. We used three different estimators based on abundance data. The observed number of operational taxonomic units (OTUs) was still lower than predicted in estimates using the nonparametric ACE, Chao1, and Chao2 indices, with values of 36.12, 33.04, and 36.29 for R. stylosa, and 25.59, 24.56, and 24.36 for R. mucronata, respectively, at 97% similarity.

Screening for antioxidant activity in endophytic fungal extracts
Another aim of the present study was to identify the antioxidant potential of endophytic fungi from the two mangrove species. A total of 46 representative endophytic fungal extracts from R. stylosa and R. mucronata were tested for antioxidant capacity, including the ability to scavenge DPPH and ABTS radicals using a microtiter plate assay. Of the four media used for fermentation, Czapek's agar and rice medium were determined to be more suitable for antioxidant production in fungal isolates than PDA and the grain medium for both host plants. Most of the fungal strains (39 isolates, 84.8%) exhibited antioxidant capacity to some extent (%RSA > 50%) and a concentration of 10 mg/mL was selected based on an antioxidant activity assay using different concentrations (1 mg/mL to 10 mg/mL). The DPPH and ABTS radical scavenging activities of the crude extracts showed a dose-dependent relationship. Their IC 50 values were obtained by interpolation from a linear regression analysis, which indicated that each isolate had high levels activity in one or more of the different culture media. High rates of antioxidant activity were detected from the fungal extracts belonging to Diaporthales and Xylariales. Four culture media of Diaporthe eucalyptorum (HHL53 and HHQ62) and Cytospora rhizophorae (HHL55) all exhibited moderate scavenging ability towards DPPH or ABTS radicals. Czapek's agar culture of the endophytic strains Cytospora rhizophorae (HHL55) and Seiridium ceratosporum (HHL38) isolated from R. stylosa showed the most potent radical-scavenging activity, with IC 50 values of 0.33±0.02 mg/mL and 0.37±0.02 mg/mL, respectively ( Table 6, S3 Fig). The radical scavenging activities of these fungal extracts were comparable to ascorbic acid (Vc, IC 50 of 0.10±0.01 and 0.09±0.00 mg/mL), suggesting their potency. Further research also provided the first evidence that HQD-6 could produce flufuran, which would then act as a scavenger of DPPH and ABTS radicals, with IC 50 values of 34.85 ±1.56 and 9.75±0.58 μg/mL, respectively (NMR data elsewhere), indicating that it is more potent as an antioxidant than the positive control.

Discussion
Mangroves are coastal biotopes that develop within the intertidal zone. They are tolerant of many types of extreme environmental stress. They harbor diverse fungal communities, which makes them potential sources of bioactive natural products. The present study showed that the endophytic fungal community of various tissues from R. stylosa and R. mucronata to be highly diverse, and a total of 46 independent representative species were isolated. The leaf endophyte composition of five Sonneratia mangrove species in the South China Sea [21], endophytic fungi of R. mucronata of Thailand [34] and two Rhizophora species (R. apiculata and R. mucronata) [35] growing in the New Caledonia mangrove forest in France all differed greatly from the present study. Phylogenetic analysis based on the ITS region identified 17 genera belonging to eight taxonomic orders within Ascomycota, with Pestalotiopsis and Diaporthe as the most prevalent endophytes. This is consistent with the findings of some previous studies [36,37]. Previous phylogenetic studies have suggested that host specificity is relatively common at higher levels of the host taxon [38,39]. In our study, a significant proportion of the predominant isolates, including eight species belonging to seven genera (Diaporthe eucalyptorum, Lasiodiplodia theobromae, Neofusicoccum mangiferae, Neofusicoccum parvum, Pestalotiopsis microspora, Phomopsis glabrae, Phomopsis longicolla, and Valsa brevispora) were found to be common in both R. stylosa and R. mucronata, which supports the conclusion that endophytic fungi may be more common at the host family level than at the host species level. The remaining species demonstrated host specificity, but not preference for any particular host species. Some of the isolates that have been previously reported as endophytes, e.g., Seiridium ceratosporum, Cytospora rhizophorae, Pseudofusicoccum stromaticum, and Paraconiothyrium hawaiiense, were observed in mangroves for the first time in this work.
The diversity indices served as evidence of the diversity within the composition of the endophytic fungal community in R. stylosa and R. mucronata, and the distribution of these endophytic fungi was tissue-specific. The relative frequency of colonization of endophytic fungi in different tissues was in the order of stems > hypocotyls > roots > flowers > leaves, which is consistent with the findings of previous reports [21]. Several endophytes, including Colletotrichum, Eutypella, and Neofusicoccum, were specifically distributed in the roots but did not colonize the stems or other tissues. The tissue-specific differences in microbial community suggest preference for distinct microenvironments. Analysis of nonparametric ACE, Chao1, and Chao2 indices indicated statistically sufficient sampling for the species composition analysis. Several factors may affect species composition, including seasonal changes, host age, and collection site. Culture-independent fungi were not considered in the present study [40] Culturefree methods including metagenomics analysis, sampling total DNA directly from the environment, and the heterologous expression of a biosynthetic gene in yeast or bacteria may be utilized as a complementary approach to conventional cultivation for the examination of fungal communities [41].
There has been little research into the antioxidants derived from mangrove endophytic fungi. In the present study, the antioxidant activities of various endophytic fungal extracts from Rhizophora plants were investigated. R. apiculata may be an excellent source of antioxidants. Various in vitro assays and in vivo studies have shown that the crude extracts of R. apiculata exhibit strong antioxidant activity, thereby imparting a protective effect on sodium nitrite-induced oxidative stress in the brains of rats [42,43]. R. mucronata has been shown to have highly effective scavenging towards DPPH, hydroxyl, nitric oxide radicals, and hydrogen peroxide radicals [44,45]. To efficiently utilize the mangrove fungal sources available from a single fungal strain, we attempted to explore the antioxidants of isolated fungi by changing the cultivation media. Four culture media, including PDA, CZA, RM, and GM, were tested using improved DPPH and ABTS assays. Among the four media used for fermentation, CZA and RM were demonstrated to have the highest suitability for antioxidant production. The Czapek's agar culture of the endophytic strains Cytospora rhizophorae (HHL55) and Seiridium ceratosporum (HHL38) isolated from R. stylosa were found to be the most potent scavengers of DPPH and ABTS radicals. Their differences might be ascribed the use of different culture media to activate biosynthetic gene clusters to produce the improved cryptic antioxidant [46]. A large proportion (84.8%) of fungal isolates possessed strong antioxidant capacities (% RSA > 50%), indicating that these plants and their endophytes might serve as sources of natural antioxidants.
Previous reports have described the antioxidant potential of some endophytic fungal metabolites including isobenzofurans pestacin and isopestacin isolated from Pestalotiopsis microspore, which was derived from Terminalia morobensis, and the strong free radical-scavenging agent graphislaetone A, which was isolated from Trachelospermum jasminoides endophytic Cephalosporium sp. IFB-E001 [47][48][49]. However, few natural antioxidants have been identified from mangrove endophytic fungi. One of the few confirmed antioxidants is xyloketals B, produced by the unidentified fungal strain Xylaria sp. No. 2508, which colonizes in Avicennia marina. It has been demonstrated to be a natural antioxidant and it could protect OGD-induced and MPP+-induced neurotoxicity in Caenorhabditis elegans and PC12 cells through the restoration of total GSH levels and through its antioxidant property [50,51]. Because of these findings, researchers have subsequently focused on identifying the antioxidants associated with mangrove endophytic fungi. This has led to the isolation of a new compound with antioxidant activities, xyloketal B, isolated from the mangrove fungus Xylaria sp., which imparts protection against oxLDL-induced endothelial oxidative injury possibly by inhibiting NADPH oxidase-derived ROS generation, promoting NO production, and restoring Bcl-2 expression, thereby rendering it a promising candidate for antioxidant discovery [14]. Racemic cyclohexenone and cyclopentenone derivatives, (±)-(4R Ã ,5S Ã ,6S Ã )-3-amino-4,5,6-trihydroxy-2-methoxy-5-methyl-2-cyclohexen-1-one and (±)-(4S Ã ,5S Ã )-2,4,5-trihydroxy-3-methoxy-4-methoxycarbonyl-5-methyl-2-cyclopenten-1-one isolated from the culture of the mangrove endophytic fungus Alternaria sp. R6, exhibited potent ABTS scavenging activities with EC 50 values of 8.19 ± 0.15 and 16.09 ± 0.01 μM, respectively [52]. Some studies have suggested that plants associated with salt tolerance and endophytes are typically rich in antioxidants. Mangrove endophytic fungus Aspergillus flavus were analyzed using various in vitro assay systems, such as iron chelating capacity, reducing power, and hydroxyl radicals/ hydrogen peroxide/DPPH radical scavenging and inhibition of lipid peroxidation using the βcarotene-linoleate model system, clearly established the antioxidant potency of extracts of mangrove plants and respective endophytic fungi and the possible roles of antioxidants in symbiotic host-endophytic relationships [53,54].

Conclusion
In the present study, a wide variety of culturable fungi from at least 17 genera belonging to eight taxonomic orders of Ascomycota was isolated from two Rhizophora species. Furthermore, these species are potential sustainable alternative resource for natural antioxidants. Of all the 46 fungal endophytes, HHL55 and HHL38 showed the most potent antioxidant properties. Chemical investigation revealed flufuran as a significant DPPH and ABTS radical scavenger isolated from HQD-6. A detailed investigation of the characterization of antioxidant compounds present in mangrove endophytes must be conducted, other specific metabolites which could explain the significant antioxidant activities of the abovementioned fungal strain must be identified, and new types of biological tests should be established.