Assessment of Cytospora Isolates From Conifer Cankers in China, With the Descriptions of Four New Cytospora Species

Cytospora species are widely distributed and often occur as endophytes, saprobes or phytopathogens. They primarily cause canker and dieback diseases of woody host plants, leading to the growth weakness or death of host plants, thereby causing significant economic and ecological losses. In order to reveal the diversity of Cytospora species associated with canker and dieback diseases of coniferous trees in China, we assessed 11 Cytospora spp. represented by 28 fungal strains from symptomatic branches or twigs of coniferous trees, i.e., Juniperus procumbens, J. przewalskii, Picea crassifolia, Pinus armandii, P. bungeana, Platycladus orientalis in China. Through morphological observations and multilocus phylogeny of ITS, LSU, act, rpb2, tef1-α, and tub2 gene sequences, we focused on four novel Cytospora species (C. albodisca, C. discostoma, C. donglingensis, and C. verrucosa) associated with Platycladus orientalis. This study represented the first attempt to clarify the taxonomy of Cytospora species associated with canker and dieback symptoms of coniferous trees in China.


INTRODUCTION
Coniferous trees are excellent landscaping species with a high ornamental and economic value. They are widely distributed as evergreen coniferous tree species and cultivated throughout China, except in Xinjiang and Qinghai Provinces (Ming, 2016). However, several coniferous trees are threatened by various pathogens in the process of planting and cultivation. Fan et al. (2020) reported five novel and one known Cytospora species causing canker and dieback diseases in conifers, with detailed descriptions and illustrations. Armillaria spp., Heterobasidion annosum, and Phellinus spp. have been reported to cause root and butt rot (Shaw and Kile, 1991;Hansen and Goheen, 2000). Moreover, leaf blight and Phytophthora diseases (Tucker and Milbrath, 1942;Phillips and Burdekin, 1992;Schlenzig et al., 2014) are also destructive to conifers.
Cytospora is one of the most important pathogenic fungi of hardwoods and coniferous trees with a worldwide distribution and large host range (Adams et al., 2005(Adams et al., , 2006Fan et al., 2014aFan et al., ,b, 2015aAriyawansa et al., 2015;Liu et al., 2015;Maharachchikumbura et al., 2015Maharachchikumbura et al., , 2016Hyde et al., 2016;Li et al., 2016;Lawrence et al., 2017Lawrence et al., , 2018Norphanphoun et al., 2017Norphanphoun et al., , 2018. Dieback and stem canker caused by Cytospora leads to the growth weakness or death of host plants, thereby causing significant economic and ecological losses (Adams et al., 2005). In conifers, Cytospora canker commonly occurs in the lowermost branches of mature trees, and stops spreading at the trunk (Adams et al., 2005). The asexual morph of Cytospora is characterized by the pycnidial stromata immersed in the bark with a single or multiple locule(s), with or without conceptacle. The conidia are aseptate, hyaline, allantoid, eguttulate, and smooth (Adams et al., 2005). The sexual morph is characterized by the ascomata immersed in the substrate with an erumpent pseudostroma, with or without necks. Asci are unitunicate, clavate to cylindrical. Ascospores are biseriate or multi-seriate, elongate-allantoid, thin-walled, hyaline, aseptate (Adams et al., 2005).
The taxonomy of the genus Cytospora is rather confusing. Ehrenberg (1818) established Cytospora and described four species simultaneously. Eighteen Cytospora species were proposed by Fries (1823), but the genus was recorded as Cytispora due to a misspelling. Thereafter, Saccardo (1884) revised the name to Cytospora and introduced 144 species. Due to the controversy in the corresponding relationship between sexual and asexual morphs, there were several synonyms, which had caused difficulties in the identification of Cytospora (Adams et al., 2005). Adams et al. (2005) officially reported that the sexual genera Leucocytospora, Leucostoma, Valsella, and Valseutypella are synonyms of Valsa. The traditional identification of Cytospora species is based heavily on their host affiliations. Nevertheless, the species occurrence may be related to geographical and environmental factors rather than host specificity (Fan et al., 2014a,b;Fan et al., 2015a,b). To more accurately identify Cytospora species, several species have been described based on morphological observations and multilocus phylogeny in recent studies (Yang et al., 2015;Lawrence et al., 2017Lawrence et al., , 2018Zhu et al., 2018Zhu et al., , 2020Fan et al., 2020;Jiang et al., 2020;Shang et al., 2020). Norphanphoun et al. (2017) used four loci to describe 14 new species isolated from Rosa, Salix, and Sorbus. Lawrence et al. (2018) reported 15 Cytospora species that infected fruit trees and crops using a multiphasic approach. Pan et al. (2020) assessed 23 species of Cytospora associated with canker and dieback disease of Rosaceae members in China by six-locus phylogeny.
There are only a few relative taxonomic studies of Cytospora canker or dieback disease of conifers. Thus, there is an urgent need for studies to clarify the pathogens causing dieback and stem canker in coniferous trees. In this study, we aimed to reveal the diversity of Cytospora species associated with canker and dieback diseases of coniferous trees in China. As a part of an investigation of pathogens that cause canker or dieback disease in China, 28 Cytospora strains in coniferous trees with obvious symptoms were evaluated. Morphological characters in conjunction with multilocus phylogenetic analyses provided valuable information to identify the phylogenetic position of these isolates. Herein, we also introduced Cytospora albodisca, C. discostoma, C. donglingensis, and C. verrucosa as four new species with descriptions and illustrations, and compared them with other species in the genus.

Sample Collection and Isolation
Fresh specimens with typical Cytospora fruiting bodies were collected from the infected twigs and branches of coniferous trees during collecting trips in China. A total of 12 isolates were obtained by removing a mucoid spore mass from conidiomata on the twigs and branches, spreading the suspension over the surface with standard potato dextrose agar (PDA) in a Petri dish, and incubating at 25 • C for up to 24 h. Single germinating conidia were transferred on to fresh PDA plates. All specimens and isolates were deposited in the Beijing Museum of Natural History (BJM) and the working Collection of X.L. Fan (CF) housed in Beijing Forestry University (BJFU). Axenic living cultures were deposited at China Forestry Culture Collection Centre (CFCC).

Morphological Analyses
Species identification was based on morphological characteristics of the ascomata or conidiomata produced on infected host materials. The macro-morphological characteristics including structure and size of stromata; the size, color, and shape of discs; number and diameter of ostioles per disc; presence and absence of conceptacle were determined under a Leica stereomicroscope (M205). The micro-morphological characteristics including size and shape of conidiophores and conidia were determined under a Nikon Eclipse 80i microscope equipped with a Nikon digital sight DS-Ri2 high-definition color camera with differential interference contrast (DIC). Over 10 ascomata/conidiomata were sectioned, and 10 asci and 30 ascospores/conidia were selected randomly for measurement. Colony morphology and growth rates were recorded and colony colors were described after 1 or 2 weeks according to the color charts of Rayner (1970). Adobe Bridge CS v.6 and Adobe Photoshop CS v.5 were used for the manual editing. Taxonomic novelties and descriptions were deposited in MycoBank (Crous et al., 2004).

DNA Extraction and PCR Amplification
Genomic DNA was extracted using the modified CTAB method (Doyle and Doyle, 1990) from mycelium which was cultured on PDA with cellophane and obtained from the surface of cellophane by scraping. The extracted DNA were estimated visually by electrophoresis in 1% agarose gels by comparing band intensity with a DNA marker 1 kbp (Takara Biotech). The qualities of DNA were measured with a NanoDrop TM 2000 (Thermo, USA). Six loci including the internal transcribed spacer (ITS), the large nuclear ribosomal RNA subunit (LSU), the partial actin (act), the RNA polymerase II subunit (rpb2), the translation elongation factor 1-α (tef1-α), and the beta-tubulin (tub2) genes were amplified and sequenced using the primer pairs ITS1 and ITS4 (White et al., 1990), LROR and LR7 (Vilgalys and Hester, 1990), ACT-512F and ACT-783R (Carbone and Kohn, 1999), RPB2-5F and fRPB2-7cR (Liu et al., 1999), EF-688F and EF-1251R (Alves et al., 2008), and Bt-2a and Bt-2b (Glass and Donaldson, 1995). The PCR amplicons were electrophoresed in 2% agarose gels. DNA sequencing was carried out using an ABI PRISM R 3730XL DNA Analyzer with BigDye R Terminater Kit v.3.1 (Invitrogen) at the Shanghai Invitrogen Biological Technology Company Limited (Beijing, China). DNA sequences generated by the forward and reverse primers were used to obtain consensus sequences using Seqman v.9.0.4 (DNASTAR Inc., Madison, WI, USA).

Phylogenetic Analyses
The sequences generated from this study were analyzed with related Cytospora taxa which were obtained from GenBank and recent publications (Supplementary Table 1). To infer their phylogenetic relationship for the new sequences, the alignment based on ITS, LSU, act, rpb2, tef1-α, and tub2 sequence data was performed using MAFFT v.6 (Katoh and Standley, 2013) and edited manually using MEGA v.6.0 (Tamura et al., 2013). For individual sequences, some characters were excluded from both ends of the alignments to approximate the size of our sequences. The sequences of Diaporthe vaccinii (CBS 160.32) was included as outgroup in all analyses. Phylogenetic analyses were performed with PAUP v.4.0b10 for maximum parsimony (MP) (Swofford, 2003), MrBayes v.3.1.2 for Bayesian Inference (BI) (Ronquist and Huelsenbeck, 2003), and PhyML v.3.0 for maximum likelihood (ML) (Guindon et al., 2010).
MP analysis in PAUP v.4.0b10 was conducted using a heuristic search option of 1,000 random-addition sequences with treebisection-reconnection (TBR) as the branch-swapping algorithm (Swofford, 2003). The branches of zero length were collapsed using the command minbrlen, and all equally parsimonious trees were saved. Clade stability was assessed using a bootstrap (BT) analysis of 1,000 replicates (Hillis and Bull, 1993). Tree length (TL), consistency index (CI), retention index (RI), and rescaled consistency (RC) were calculated for all equally parsimonious trees. ML analysis in PhyML v.3.0 was performed with a general time reversible model (GTR) of site substitution following previous studies , including estimation of gamma-distributed rate heterogeneity and a proportion of invariant sites (Guindon et al., 2010). An evolutionary model for BI was estimated independently for each locus using MrModeltest v.2.3. The best-fit model was selected under the Akaike Information Criterion (AIC) (Posada and Crandall, 1998). BI analysis in MrBayes v.3.1.2 was done by a Markov Chain Monte Carlo (MCMC) algorithm with Bayesian posterior probabilities (BPP) (Rannala and Yang, 1996). Two MCMC chains started from random trees for 10 million generations, and trees were sampled each 100th generations. The first 25% of trees were discarded as the burn-in phase of each analysis, BPP were calculated to assess the remaining 7,500 trees (Rannala and Yang, 1996). Phylogram was viewed in Figtree v.1.3.1 (Rambaut and Drummond, 2010). All novel sequences derived from this study data were deposited in GenBank. The multigene sequence alignment files were deposited in TreeBASE (www.treebase.org; accession number: S27070).
Culture characteristics: Cultures on PDA are initially white, growing fast up to 7 cm in diam. after 3 days and entirely covering the 9 cm Petri dish after 5 days, becoming dark herbage green to dull green after 7-10 days. Colonies are sparse in the center and  Etymology: Named after the distinct disc of stromata on branches.
Culture characteristics: Cultures on PDA are initially white with hazel in the center, growing fast up to cover the 9 cm Petri dish after 3 days, becoming brown vinaceous after 7-10 days. Colonies are flat with a uniform texture. Pycnidia distributed irregularly on surface.
Culture characteristics: Cultures on PDA are initially white, growing slowly up to 2 cm in diam. after 3 days and entirely covering the 9 cm Petri dish after 7 days, becoming straw after 7-10 days. Colonies are flat with a uniform texture, producing pycnidia covered by sparse aerial mycelium with cream to yellowish conidial drops exuding from the ostioles after 30 days. Pycnidia aggregated on surface.
Culture characteristics: Cultures on PDA are initially white, growing up to 6 cm in diam. after 3 days and entirely covering the 9 cm Petri dish after 5 days, becoming buff to honey after 7-10 days. Colonies are flat with a uniform texture; sterile.
Additional materials examined: CHINA. Notes: Phylogenetically, our new four isolates cluster in a separate lineage (MP/ML/BI = 100/100/1) comparing to other strains included in this study (Figure 1). Morphologically, C. verrucosa has similar characteristics to C. friesii, but it can be identified by having verrucosa symptoms in branches and pycnidia with a central column, and having numerous ostioles on a large area of the ectostromatic disc. Moreover, C. verrucosa differs from the closest species C. globosa by larger size of conidia (6.5-8 × 1.5-2 vs. 4-6.5 × 1-2 µm) (Li et al., 2020).

DISCUSSION
In the present study, we utilized a polyphasic approach of molecular phylogenetic analyses of the combined alignment of ITS, LSU, act, rpb2, tef1-α, and tub2 gene suquences along with morphological observations. Eleven Cytospora species represented by 28 strains from coniferous trees, including four new species (C. albodisca, C. discostoma, C. donglingensis, and C. verrucosa), and seven known species (C. beilinensis, C. bungeanae, C. gigaspora, C. juniperina, C. piceae, C. platycladi, and C. platycladicola) were evaluated. A dubious species, C. curreyi, was reported from Abies sp. in the Sichuan Province, China (Teng, 1963), but it was not confirmed due to the nonavailability of living culture.
Consistent with the conclusions of a previous study, sexual morphs of Cytospora species associated with coniferous trees summarized herein are rarely found in nature. In the present study, among the specimens of the four new species, two new species (C. albodisca and C. donglingensis) had only sexual morphs, and the other two new species (C. discostoma and C. verrucosa) had only asexual morphs. Therefore, we observed the asexual morphs of C. albodisca and C. donglingensis after sporulation on PDA medium. Adams et al. (2005) reported that the asexual morphs of Cytospora formed naturally may be different from those formed in culture, and these morphological characteristics may not be meaningful in classification. In conidia morphology, C. discostoma resembles C. donglingensis (4.5-5.5 × 1-1.5 vs. 4.5-6 × 1-2 µm). However, C. discostoma differs from C. donglingensis in ITS (33/656), LSU (7/522), act (61/367), rpb2 (71/726), tef1-α (104/824), and tub2 (101/648). We found that the size of the conidia was distinguishable from that of other species, and this was strongly supported by DNA sequence data.
Host affiliation has been primarily used as the delimitation in Cytospora in the early stage, and this has been proven uninformative because several Cytospora species have been discovered in a wide range of hosts (Adams et al., 2005(Adams et al., , 2006Lawrence et al., 2018;Norphanphoun et al., 2018;Fan et al., 2020;Pan et al., 2020). Lawrence et al. (2018) reported that Cytospora included generalist and specialist pathogens by taking C. chrysosperma and C. punicae as examples; however, a clear elucidation of the host ranges and distribution of Cytospora species will require a more exhaustive sampling of other coniferous trees from other regions of the world. Consistent with previous findings, some Cytospora species isolated from coniferous trees occurred on different hosts (i.e., C. ampulliformis, C. gigaspora, C. melnikii) (Fan et al., 2015bNorphanphoun et al., 2017Norphanphoun et al., , 2018Lawrence et al., 2018) rather than specific hosts. In addition, it cannot be denied that some species of Cytospora have a preference for certain hosts (i.e., C. japonica with chiefly Rosaceae host record, C. mali with apple host record and C. pini with pine host record) (Teng, 1963;Tai, 1979;Wei, 1979;Zhuang, 2005;Wang et al., 2011). In our study, all species with available strains found in China were associated with a single coniferous host (mainly Juniperus, Picea, Pinus, and Platycladus), with the exception of Cytospora gigaspora, which has also been reported in Salix psammophila (Fan et al., 2015b). These findings suggest that future studies are needed to better understand the interaction between fungi and their hosts.
Coniferous trees are the main timber and greening tree species in forestry production, but they are exposed to different pathogens. Cytospora was recorded on the hosts of four coniferous families (14 Cytospora spp. infecting Cupressaceae, 14 Cytospora spp. infecting Pinaceae, three Cytospora spp. infecting Taxaceae and three Cytospora spp. infecting Taxodiaceae) based on the U.S. National Fungus Collections Fungus-Host database (Farr and Rossman, 2021), but most species are inadequately identified and lack of molecular data. In China, some new species and new records of Cytospora on conifers have been reported successively (Fan et al., 2015bPan et al., 2018), lacking a systematic study summarizing Cytospora species isolated from conifers. The present study indicated that the common families of conifers infected by Cytospora are Cupressaceae and Pinaceae, although several coniferous hosts suffering from canker disease have not been discovered. Thus, the present research is preliminary in nature, and further studies using a more intensive and wider sampling of isolates are awaited.
Cytospora canker and dieback diseases present with different symptoms in hardwoods and conifers. In hardwoods, the symptoms are characterized by elongated, slightly sunken, and discolored areas in the bark with obvious black spots . However, discoloration of the adjacent cambium in conifers has not been observed, although the fungus can be isolated from nearby xylem (Schoeneweiss, 1983). A large amount of resin flows from the infected branches, covers the bark surface surrounded by the canker, and drips onto the lower branches (Schoeneweiss, 1983). Generally, in canker disease, Cytospora begins to infect through cracks and wounds in the bark. The wounds include pruning wounds, cold injuries, leaf scars, and branches with weakened shade. If perennial cankers originating from pruning wounds occur in places critical to the strength of the trees, they can be highly destructive (Biggs, 1989;Adams et al., 2006). Therefore, the occurrence of Cytospora canker and dieback diseases can be minimized by maintaining susceptible trees as strong as possible, and by removing dead and dying branches in the dry season. All unnecessary damage should be avoided. Moreover, the occurrence of Cytospora canker diseases is not only affected by the environment and distribution, but also by transmission (Fan et al., 2015b), which may act as potential inoculum sources for other hosts in natural and artificial environments. Six pathogens, including Cytospora, have been found to pose a high risk of causing severe damage if exported to other suitable environments (Cannon et al., 2016). At present, the host specificity and pathogenicity of several Cytospora species associated with coniferous trees are poorly known. In the subsequent studies, more attention should be paid to the pathogenicity and aggressiveness of Cytospora, which can play a role in quarantine, monitoring, and early warning of forestry pathogen.
In conclusion, in this study, we focused on four Cytospora species isolated from Platycladus orientalis in China. Our study implies that many additional Cytospora species from China are still undiscovered. The keys to Cytospora species on Platycladus spp. and coniferous trees were established based on their unique morphological characteristics, and this will also help more extensive research on fungal pathogens in China. Furthermore, this study constitutes a step toward the taxonomic study of conifer pathogens, providing sustainable disease management strategies for conifers infected by Cytospora species in China.

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material.