﻿Endophytic Colletotrichum (Sordariomycetes, Glomerellaceae) species associated with Citrusgrandis cv. “Tomentosa” in China

﻿Abstract Colletotrichum species are well-known plant pathogens, saprobes, endophytes, human pathogens and entomopathogens. However, little is known about Colletotrichum as endophytes of plants and cultivars including Citrusgrandis cv. “Tomentosa”. In the present study, 12 endophytic Colletotrichum isolates were obtained from this host in Huazhou, Guangdong Province (China) in 2019. Based on morphology and combined multigene phylogeny [nuclear ribosomal internal transcribed spacer (ITS), glyceraldehyde-3-phosphate dehydrogenase (gapdh), chitin synthase 1 (chs-1), histone H3 (his3) actin (act), beta-tubulin (β-tubulin) and glutamine synthetase (gs)], six Colletotrichum species were identified, including two new species, namely Colletotrichumguangdongense and C.tomentosae. Colletotrichumasianum, C.plurivorum, C.siamense and C.tainanense are identified as being the first reports on C.grandis cv. “Tomentosa” worldwide. This study is the first comprehensive study on endophytic Colletotrichum species on C.grandis cv. “Tomentosa” in China.


Introduction
Citrus grandis cv. "Tomentosa" is an important traditional medicinal plant which contains essential oils, flavonoids and polysaccharides. In traditional Chinese medicine, Citrus grandis cv. "Tomentosa" has been used for treatments due to its anti-inflammatory effect (Zhao et al. 2017). It has also been used in the treatment of coughs, asthma, food stagnation, vomiting and other symptoms (Peng et al. 2019). Current research on C. grandis cv. "Tomentosa" is still focused on medicinal components, with a relatively long timescale needed to accumulate the effective ingredient. It is likely that the endophytic community living inside the host affects the metabolites of the plant. Dai et al. (2017) found that nine species of Taxus endophytic fungi could produce paclitaxel. Hasan et al. (2022) found endophytic fungi, Penicillium crustosum from Annona muricata L. has anti-cancer activity against HeLa cells. Therefore, it is necessary to study the effects of the endophytic community associated with these traditional medicinal plants. The findings of this research can help in finding potential new natural medicines and form the basis for subsequent screening of strains.
Colletotrichum Corda (1831), belongs to Glomerellaceae (Sordariomycetes), which comprises plant pathogens, endophytes and saprobes on a wide range of hosts (Christy et al. 2020;Jayawardena et al. 2021). They are one of the most often isolated endophytic fungal groups encompassing a wide range of hosts. These endophytic Colletotrichum species have some advantages to the host, such as providing disease resistance, drought tolerance and promoting growth of the host (Hacquard et al. 2016;Dini-Andreote 2020). Endophytic species can also change their lifestyle and become pathogenic (Photita et al. 2004). Liu et al. (2022) accepted 280 Colletotrichum species, from which 23 species have been identified from Citrus spp. Therefore, studying diversity and clarifying taxonomic affinities of isolates can answer a range of important ecological and evolutionary questions. Although there have been several studies on Colletotrichum species associated with Citrus Huang et al. 2013;Guarnaccia et al. 2017), there is still imprecise identification of endophytes of Colletotrichum species on C. grandis cv. "Tomentosa".
Species delineation of Colletotrichum is challenging because there are few distinctive morphological characters available . Colletotrichum is characterised as an intricate genus with 16 species complexes and 15 singleton species (Liu et al. 2022). Although host specificity was the most used character for identification in early studies, current taxonomic classifications and species delineations are based on morphology alongside multi-locus phylogeny Jayawardena et al. 2021;Liu et al. 2022). Phylogenetic analyses of Colletotrichum have been based on ITS, gapdh, chs-1, act and β-tubulin and multi-loci phylogeny. However, some complexes that cannot be distinguished by five loci required additional loci for identification Jayawardena et al. 2021;Liu et al. 2022). Therefore, the selection of gene combinations depends on the species complex .
The objectives of this study were to isolate and identify the dominant endophytic Colletotrichum species associated with healthy C. grandis cv. "Tomentosa" in Huazhou, Guangdong, China. Morphology, molecular phylogeny and recombination analysis were used for the species characterisation. This resulted in two new species and six new host records. Detailed descriptions and coloured illustrations have been given for the novel taxa identified.

Sample collection and isolation
Healthy leaves and twigs of Citrus grandis cv. "Tomentosa" were randomly collected from a Citrus orchard in Huazhou, Guangdong Province, China (21°66'N, 110°63'E). A total of 20 trees were randomly selected for the collection. Ten samples were collected from the upper, middle and lower parts of each plant. Asymptomatic samples were packed into zip-lock bags in a foam box with ice and were then brought to the plant pathology laboratory of Zhongkai University of Agriculture and Engineering where they were preserved at 4 °C before processing. Isolation was undertaken within 48 h after collection, following the procedure by Dong et al. (2021).
Endophytic fungi were isolated following the methods described by da Silva et al. (2020). The samples were initially washed with running tap water followed by sterile water. The leaves were cut into 3 mm × 3 mm segments, while the twigs were cut into 3 mm long pieces. Each piece was then surface sterilised by being dipped sequentially into 75% ethanol for 30 s, 2.5% NaClO (sodium hypochlorite) for 30-60 s (leaves for 30 s, twigs for 60 s), before being rinsed three times with sterilised water. They were then dried on sterilised filter paper. The cuttings were then placed on potato dextrose agar (PDA: 200 g potato, 20 g dextrose, 20 g agar per 1 litre of water). Plates were incubated at 25 °C with 12 h of dark and 12 h of fluorescent light. Pure cultures were cultured on PDA for 7 to 14 days at 25 °C. All the pure cultures obtained in this study were deposited in the Culture Collection of Zhongkai University of Agriculture and Engineering (ZHKUCC). The living cultures (ex-type) of new species identified in this study were deposited in the Culture Collection of the Chinese Academy of Sciences (CGMCC, C. guangdongense for the holotype with CGMCC 3.24127 and C. tomentosae with CGMCC 3.24128). Herbarium materials as dry cultures of novel species were deposited in the Herbarium of Zhongkai University of Agriculture and Engineering (ZHKU). The strain numbers belonging to all isolates (from ZHKUCC 21-0095 to 21-0106 and 22-041 to 22-0042) for this study are presented in Appendix 1.

Morphological studies
For macro-and micro-morphological characterisation, 5 mm diameter agar plugs were cut from all the actively growing pure cultures on PDA and were then transferred on to new PDA. The colony diameter was measured daily for 5-9 d to determine the growth rate (mm/day) on the PDA at 25 °C under 12 h of dark and 12 h of fluorescent light. Appressoria formation was observed following Johnston and Jones (1997) and Cai et al. (2009). The cultures were incubated for 2-4 weeks and morphological characters (appressoria, ascomata, asci, ascospores, conidiophores and conidia) were observed. Macro-morphological characters were photographed using a SteREO Discovery.V20 (Zeiss, Germany) stereomicroscope. Fruiting bodies were cut into thin sections by a CM1860 freezing sliding microtome (LEICA, Germany). Digital images were captured with an Eclipse 80i photographic microscope (Nikon, Japan). Measurements were taken using NIS Elements BR 3.2 (Nikon, Japan). The mean values were calculated with their standard deviations (SDs).

DNA extraction, PCR amplification and sequencing
Total genomic DNA was extracted from mycelium grown on PDA and incubated for approx. seven days at 25 °C using the CTAB method (Sun et al. 2009). The ITS region was amplified and sequenced. The resulting sequences were subjected to BLASTn searches in GenBank (https://blast.ncbi.nlm.nih.gov) to identify them to the genus level. Once the BLAST results had confirmed isolates as being Colletotrichum species, an additional six gene regions, namely gapdh, chs-1, his3, act, β-tubulin and gs, were amplified and sequenced. The PCR conditions for each primer pair are given below ( Table 1). The amplicons were observed on 1% agarose electrophoresis gel and positive amplicons were sequenced by Tianyi Huiyuan Biotechnology Co., Ltd., Guangzhou, China. The initial sequence quality was checked using BioEdit v. 7.25 (Hall 2006). A total of 66 sequences generated in this study were submitted to GenBank (Appendix 1).

Phylogenetic analysis
For the phylogenetic analysis, reference sequences for Colletotrichum species and related taxa were obtained from NCBI GenBank (Appendix 1). Each locus was aligned together with the sequences obtained in the present study using MAFFT (https://www.ebi.ac.uk/ Tools/msa/mafft/) (Katoh et al. 2019). Alignments were checked and manually adjusted where necessary with BioEdit v. 7.25 (Hall 2006). Alignment results were automatically trimmed using the Trimal tool in PhyloSuite (v.1.2.1) (Zhang et al. 2020). Phylogenetic analyses were conducted according to Maximum Likelihood (ML) in RAxML (Silvestro  (Ronquist and Huelsenbeck 2003). The final analyses of the Colletotrichum gloeosporioides complex were made using the concatenated dataset of act, chs-1, gapdh, ITS, β-tubulin and gs, following Liu et al. (2022). The other two complexes: Colletotrichum orchidearum complex and Colletotrichum magnum complex were analysed using act, chs-1, gapdh, his3, ITS and β-tubulin, following Liu et al. (2022).
In the MP analysis, ambiguous regions were excluded and gaps were treated as missing data. Tree stability was evaluated with 1,000 bootstrap replications. Zero-length branches were collapsed and all the parsimonious trees were saved. Tree parameters: tree length (TL), consistency index (CI), retention index (RI), relative consistency index (RC) and homoplasy index (HI) were calculated. Kishino-Hasegawa tests (KHT) were conducted to evaluate the differences between the trees inferred as being under different optimality criteria (Kishino and Hasegawa 1989). MrModelTest v. 2.3 (Nylander 2004) was used to determine the evolutionary models for each locus to be used in Bayesian and Maximum Likelihood analyses. The Maximum Likelihood analyses were conducted using RAxML-HPC2 on XSEDE (8.2.8) (Stamatakis 2014) in the CIPRES Science Gateway platform (Miller et al. 2010). The GTR + I + G evolutionary model was employed with 1,000 non-parametric bootstrapping iterations. Bayesian analysis was performed in MrBayes v. 3.1.2 (Ronquist and Huelsenbeck 2003). Posterior probabilities (PPs) were determined using Markov Chain Monte Carlo sampling (MCMC). Six simultaneous Markov chains were run for 10 8 generations, with sampling the trees at each 1000 th generation. From the 10,000 trees obtained, the first 2,500 representing the burn-in phase were discarded. The remaining 7,500 trees were then used to calculate the posterior probabilities (BPs) in a majority rule consensus tree. Taxonomic novelties were submitted to the FacesofFungi database (Jayasiri et al. 2015) and Index Fungorum (http:// www.indexfungorum.org). The final sequence alignments generated in this study were submitted to TreeBASE (http://www.treebase.org) under the submission ID 29668.

Pairwise homoplasy index (PHI) analysis
Recombination analyses were conducted to provide evidence for genetic distances for two new species identified, based on the phylogenetic analyses. The pairwise homoplasy index (Φw) (Bruen et al. 2006) was calculated in SplitsTree (version 4.1.4.4) using Kimura's two-parameter (K2P) models for low genetic distance datasets. The standard deviation of split frequencies in the PHI test results (Φw) < 0.05 indicates significant recombination within the dataset.

Results
In total, 12 endophytic Colletotrichum strains were obtained: seven from leaves and five from twigs. Based on the initial species identification undertaken through BLASTn searches, taxa isolated in this study belonged to three species complexes, namely the C. gloeosporioides, C. magnum and C. orchidearum complexes.

Colletotrichum gloeosporioides complex
In the present study, eight Colletotrichum isolates were initially recognised as belonging to the C. gloeosporioides complex. Phylogenetic analyses of a combined act (1-281), chs-1 (282-573), gapdh (574-850), ITS (851-1384), β-tubulin (1385-1846) and gs (1847-2616) sequence alignment were conducted using 89 Colletotrichum strains. Colletotrichum boninense (ICMP 17904) and C. hippeastri (ICMP 17920) were used as outgroup taxa. The best-scoring MP tree is shown in Fig. 1. The dataset comprised 2,616 characters with 1,757 constant characters, 370 parsimony-informative and 489 parsimony-uninformative characters. The maximum number of trees generated was 1,000 and the most parsimonious trees had a length of 1,492 steps (CI = 0.707, RI = 0.848, RC = 0.600, HI = 0.293). The final ML tree topology was in line with the MP and BP trees. The bestscoring ML tree has a final likelihood value of −12,639.274168. The matrix consisted of 1,060 distinct alignment patterns, with 15.26% undetermined characters or gaps. For the Bayesian Inference, the TPM2uf+G model was selected for act, TIM1ef+G for chs-1, HKY+I for gapdh, TrNef+I+G for ITS, TIM3ef+G for β-tubulin and TVM+G for gs. In the phylogenetic analysis, three isolates (ZHKUCC 21-0103, ZHKUCC 21-0104 and ZHKUCC 22-0041) from this study developed a sister clade from other known species. The new species of C. tomentosae showed a close relationship to C. syzygicola (MFLUCC 10-0624) with 92% ML, 90% MP and 1.00 BP support. Three strains (ZHKUCC 21-0096, ZHKUCC 21-0097 and ZHKUCC 21-0098) from this study cluster together with C. siamense (ICMP 18578) with 0.99 BP support in the multi-locus phylogenetic tree. The strain ZHKUCC 21-0095 was clustered with C. asianum (ICMP 18580) with 100% ML, 100% MP and 1.00 BP in the phylogenetic tree. A single strain (ZHKUCC 21-0101) belongs to C. tainanense (CBS 143666) with 93% ML, 83% MP and 1.00 BP support. The PHI value indicates that there is no significant evidence for recombination amongst the species used in this analysis (p = 1.0) (Fig. 2). Based on this, we identified these isolates as novel Colletotrichum species. Species descriptions and illustrations of the new species, identified from the C. gloeosporioides complex, are presented below. Notes. The single isolate (ZHKUCC 21-0095) obtained in this study clustered with the Colletotrichum asianum ex-type strain (ICMP: 1850) with 100% ML, 100% MP and 1.0 BP values (Fig. 1). Morphologically, the isolate obtained in this study is similar to those in the original description of C. asianum (Prihastuti et al. 2009). This is the first report of C. asianum on C. grandis cv. "Tomentosa". Notes. Three isolates obtained in this study (ZHKUCC 21-0096-100) clustered with the ex-type strain of Colletotrichum siamense (ICMP: 18578) with 67% MP and 0.99 BP values (Fig. 1). Morphologically, the isolate obtained in this study is similar to those in the original description of C. siamense (Prihastuti et al. 2009). This is the first report of C. siamense on C. grandis cv. "Tomentosa".

Notes.
A single isolate obtained in this study (ZHKUCC 21-0101) clustered with the Colletotrichum tainanense (CBS 143666) ex-type strain with 93% ML, 83% MP bootstrap and 1.0 BP values (Fig. 1). Morphologically, the isolate obtained in this study is similar to those in the original description of C. tainanense (de Silva et al. 2019). To our knowledge, this is the first report of C. tainanense on C. grandis cv. "Tomentosa". Etymology. The epithet refers to the cultivar of the host plant -Citrus grandis cv. "Tomentosa".
Cultural characteristics. Colonies on PDA reach 70 mm diam. in seven days, with 10-11 mm/day (x − = 10 mm, n = 6) growth rate. Colonies flat with entire margin, floccose cottony, surface grey in the centre with glaucous margin. Reverse buff in the centre with off-white margin.

Colletotrichum orchidearum complex
In the present study, a single isolate was recognised as belonging to the Colletotrichum orchidearum complex. The phylogenetic analysis of a combined ITS, gapdh, chs-1, his3, act and β-tubulin sequence alignment was constructed using 30 Colletotrichum strains. Colletotrichum magnum (CBS 519.97) and C. brevisporum (BCC 38876) were used as the outgroup. The best scoring MP tree is presented in Fig. 4. The dataset comprised 2,422 characters with 2,055 constant characters and 242 parsimony-informative and 125 parsimony-uninformative characters. The maximum number of trees generated was 1,000 and the most parsimonious trees had a length of 475 steps (CI = 0.874, RI = 0.904, RC = 0.790, HI = 0.126). The final ML tree topology was similar to the MP and BP trees. The best-scoring ML tree with a final likelihood value of -6,065.417493 is shown in Fig. 4. The matrix comprised 479 distinct alignment patterns, with 10.74% of undetermined characters or gaps. The estimated base frequencies were as follows: A = 0.214401, C = 0.319513, G = 0.254583, T = 0.211503; substitution rates AC = 0.9523776, AG = 3.421321, AT = 0.568275, CG = 0.738898, CT = 6.093168, GT = 1.000000; gamma distribution shape parameter a = 0.814817. For the Bayesian Inference, the TPM1uf+I model was selected for act, GTR+I+G for chs-1, HKY+I for gapdh, TIM2+G for his3, TIM1+I for ITS and HKY+G for β-tubulin. In the phylogenetic analysis, isolates from this study clustered together with C. plurivorum. The species description and illustration are given below.  (Fig. 4). Morphologically, the isolate obtained in this study is similar to those in the original description of C. plurivorum (Damm et al. 2019  on Capsicum annuum fruits and subsequently, has been reported as pathogens causing anthracnose or leaf spot diseases (Farr and Rossman 2022). This is the first report of C. plurivorum as an endophyte on Citrus grandis cv. "Tomentosa".

Colletotrichum magnum complex
Three of our isolates were initially recognised as belonging to the Colletotrichum magnum species complex. The phylogenetic analysis of combined act, chs-1, gapdh, his3, ITS and β-tubulin sequence alignment was conducted using 17 Colletotrichum strains. . The most parsimonious tree for Colletotrichum orchidearum complex using a combined act, chs-1, gapdh, his3, ITS, and β-tubulin sequences. The tree is rooted to Colletotrichum brevisporum and C. magnum. Bootstrap support values equal to or greater than 60% in MP and ML and BP equal to or greater than 0.95 are shown as MP/ML/BP above the respective nodes. The isolates belonging to the current study is given in blue. Ex-type strains are noted with T.
Colletotrichum orchidearum (CBS 135131) and C. cliviicola (CBS 125375) were used as outgroup taxa. The best-scoring MP tree is given in Fig. 5. The dataset consisted of 2,296 characters with 2,013 constant characters and 196 parsimony-informative and 87 parsimony-uninformative characters. The maximum number of trees generated was 1,000 and the most parsimonious trees had a length of 350 steps (CI = 0.883, RI = 0.882, RC = 0.779, HI = 0.117). The final ML tree topology was similar to the MP and BP trees. The best-scoring ML tree had a −5198.901460 final likelihood Figure 5. The most parsimonious tree of the Colletotrichum magnum complex using combined act, chs-1, gapdh, his3, ITS and β-tubulin sequences. Colletotrichum cliviicola and C. orchidearum were used as outgroup taxa. Bootstrap support values equal to or greater than 60% in MP and ML and BP equal to or greater than 0.95 are shown as MP/ML/BP above the respective nodes. The isolates of the novel taxon described in the current study are highlighted in red. Ex-type strains are noted with T.
value. The ML matrix comprised 258 distinct alignment patterns, with 6.18% undetermined characters or gaps. For the Bayesian Inference, the HKY model was selected for act, TIM2ef+G for chs-1, HKY+G for gapdh, TrN+G for his3, TIM1+I for ITS and TIM1+G for β-tubulin. In the phylogenetic analysis, isolates from this study developed to show the presence of an independent clade with high bootstrap and BP support. To confirm that these isolates belonged to novel species, the PHI index was calculated. The PHI test revealed no significant evidence for recombination (p = 1.0) amongst the taxon from this study and its closely-related taxa (Fig. 6). Etymology. The epithet refers to the Guangdong Province where the fungus was collected.

Discussion
In the present study, endophytic Colletotrichum species were isolated from Citrus grandis cv. "Tomentosa" in Guangdong Province, China. Guangdong Province has a mild subtropical monsoon climate with abundant rainfall and high average annual temperatures. Vigorous fruit trees provide suitable conditions for the colonisation of Colletotrichum species . When the host is healthy, the endophyte has a symbiotic relationship with the host . However, sometimes the interaction between the plant and the endophyte can switch from mutualistic to antagonistic or pathogenic (da Silva et al. 2020). Thus, the identification and characterisation of endophytic fungi are necessary. Based on the phylogenetic analysis using a combined seven loci (ITS,gapdh,act,his3,tub2 and gs), 12 isolates from this study were identified as being six distinct species within the three Colletotrichum species complexes (Figs 1, 4, 5). These results included two new species, namely C. guangdongense, C. tomentosae and three new host records for C. asianum, C. plurivorum and C. tainanense. Colletotrichum siamense has also been identified and described as being associated with Citrus. The present study has re-affirmed that more than one Colletotrichum species can colonise a single host, which is consistent with the conclusion of Damm et al. (2012).
Species belonging to the C. gloeosporioides complex were often found as endophytes Weir et al. 2012;Jayawardena et al. 2016). Here, we identified seven strains representing four species as endophytes from the C. gloeosporioides complex. Colletotrichum siamense was previously reported as an epiphyte and an endophyte asso-ciated with coffee berries in northern Thailand (Prihastuti et al. 2009) and tea plants in China (Liu et al. 2015). Colletotrichum siamense has also been reported as a pathogen of many plants (Liu et al. 2022). In the present study, this species was isolated from leaves. Liu et al. (2015) identified six species from symptomatic and asymptomatic leaf tissue, all of which belonged to the C. gloeosporioides species complex, namely C. camelliae, C. fructicola, C. gloeosporioides, C. jiangxiense and C. siamense, providing convincing evidence that these species could switch their lifestyle from endophytic to pathogenic. Therefore, further studies are necessary to understand the pathogenicity of these endophytic strains and the factors affecting these taxa becoming pathogenic on Citrus.
Colletotrichum species belonging to the C. magnum and C. orchidearum complexes were found on tropical or subtropical plants (Damm et al. 2019). It has been proposed that some of these species might be host-and region-specific (Damm et al. 2019). Colletotrichum plurivorum is widely distributed in several hosts and most of them are pathogens. This study is the first report of the species from Citrus. Here, we introduce a new taxon belonging to the C. magnum species complex. Whether it is host-specific or not needs further confirmation.
Endophytic fungal colonisation might vary in different tissues of the same plant (Taylor et al. 1999;Huang et al. 2015). Different fungal genera could have different tissue specificities and preferences. In the present study, endophytes were isolated from leaves and twigs. Additionally, there were higher numbers of Colletotrichum species from leaves in Citrus (Hakimeh et al. 2019) and some other plants like Dendrobium (Chen et al. 2011;Ma et al. 2018). Huang et al. (2015) and Dong et al. (2021) have observed that endophytic Diaporthe species are less abundant on leaves, whereas endophytic Colletotrichum species are abundantly isolated from the Dendrobium spp. leaves (Chen et al. 2011;Ma et al. 2018). These variations may be the result of differences in the tissue organisational structure, different nutrition contents of each tissue type or the lifestyle of each genus, locality or season (Zhou et al. 2014;Huang et al. 2015). To date, the reasons for these variations are not yet known.
Overall, in the present study, two novel endophytic Colletotrichum species have been described and illustrated. Our study is the first comprehensive study on endophytic Colletotrichum species associated with Citrus grandis cv. "Tomentosa". Moreover, our molecular data and novel species introduced in this study contribute to understanding the diversity and biology of the genus Colletotrichum. These results provide an important resource and basis for plant pathologists and fungal taxonomists. However, future studies are necessary to understand the lifestyle changes of the endophytic taxa towards the pathogenicity, as well as the effects of fungus-related medicinal properties of Citrus grandis cv. "Tomentosa". Appendix I Table A1.