﻿Two new phyllospheric species of Colacogloea (Colacogloeaceae, Pucciniomycotina) identified in China

﻿Abstract During our ongoing survey of basidiomycetous yeasts associated with plant leaves in virgin forest, five Colacogloea strains were isolated in the Baotianman Nature Reserve, Henan Province, central China. Phenotypes from cultures and a phylogeny based on the internal transcribed spacer (ITS) regions and the D1/D2 domains of the large subunit (LSU) rRNA gene were employed to characterize and identify these isolates. As a result, two new species, namely Colacogloeaceltidissp. nov. and C.pararetinophilasp. nov., are introduced herein. In the phylogeny of combined ITS and LSU dataset, the new species C.celtidissp. nov. formed a clade with the unpublished Colacogloea strain (KBP: Y-6832), and together these formed the sister group to C.armeniacae, while C.pararetinophilasp. nov. was retrieved as a sister to C.retinophila. A detailed description and illustration of both new species, as well as the differences between them and their closest relatives in the genus are provided. Results from the present study will add to our knowledge of the biodiversity of Colacogloea in China.


Sample collection and yeast isolation
Leaf samples collected from Baotianman Nature Reserve were stored in sterile flasks and transported to the laboratory within 24 h.Yeast strains were isolated from leaf surfaces by the improved ballistospore-fall method as described in previous papers (Nakase and Takashima 1993;Hu et al. 2022).Vaseline was used to affix the semi-withered leaves onto the insides of Petri dishes filled with yeast extract-malt extract (YM) agar (0.3% yeast extract, 0.3% malt extract, 0.5% peptone, 1% glucose, and 2% agar).The dishes were then incubated at 20 °C until visible colonies had formed.Different yeast morphotypes were selected from these colonies and purified by streaking on separate YM agar plates.After purification, yeast strains were suspended in YM broth supplemented with 20% (v/v) glycerol and stored at −80 °C.Cultures of all obtained isolates were preserved at the Microbiology Lab, Nanyang Normal University, Henan, China.All isolates used in this study and their origins are presented in Table 1.

Morphological and physiological characterization
Morphological and physiological characteristics of yeast strains were defined according to methods established by Kurtzman et al. (2011).Colony characteristics were observed and recorded on YM agar after two weeks of incubation at 20 °C.To investigate mycelium formation, colonies were transferred to corn meal (CM) agar (2% cornmeal infusion and 2% agar) slide cultures and incubated at 20 °C for two weeks.Tests of sexual reproductive potential were conducted for individual strains and strain pairs on potato dextrose agar (PDA) (20% potato infusion, 2% glucose, and 1.5% agar), CM agar, and yeast carbon base plus 0.01% ammonium sulphate (YCBS) agar for two months and observed at weekly intervals (Sampaio et al. 2011;Li et al. 2020).The inverted-plate method (do Carmo-Sousa and Phaff 1962) was used to observe the ballistoconidium-forming activity of all yeasts after two weeks of incubation on CM agar at 20 °C.Glucose fermentation was carried out in liquid medium using Durham fermentation tubes.Carbon and nitrogen source assimilation tests were conducted in liquid medium and starved inoculum was used for the nitrogen test (Kurtzman et al. 2011).Cycloheximide resistance was performed in liquid medium, while urea hydrolysis was conducted on agar slants.Acid production and diazonium blue B (DBB) reactions were investigated using Petri dishes with solid medium.Growth at different temperatures (15, 20, 25, 30, 35, and 37 °C) was determined by the amount of cultivation on YM agar.Cell morphology was examined using a Leica DM 2500 microscope (Leica Microsystems GmbH, Wetzlar, Germany) and a Leica DFC295 digital microscope color camera under bright field, phase contrast, or differential interference contrast (DIC) environment.All novel taxonomic descriptions and proposed names were deposited in the MycoBank database (http://www.mycobank.org;29 October 2023).

DNA extraction, PCR amplification, and sequencing
The total genomic DNA was extracted from yeast strains using the Ezup Column Yeast Genomic DNA Purification Kit according to the manufacturer's  (Wang et al. 2014).The PCR products were purified and sequenced at Sangon Biotech Co., Ltd (Shanghai, China) with the same primers.We determined the identity and accuracy of the newly-obtained sequences by comparing them to sequences in GenBank and assembled them using BioEdit 7.1.3.0 (Hall 1999).All newly generated sequences were deposited in the GenBank database (https://www.ncbi.nlm.nih.gov/genbank/), and the accession numbers are listed in Table 2.

Phylogenetic analysis
A total of 57 nucleotide sequences that belonged to 27 taxa were included in the phylogenetic analyses.Except for 10 sequences recognized in this study, the other sequences were obtained from previous studies (Li et al. 2020;Wang et al. 2021) and GenBank (Table 2).Udeniozyma ferulica CBS 7416 T was used as the outgroup.The phylogenetic relationships of the new Colacogloea species and their relatives were determined using a combined ITS and LSU sequence dataset.Sequences of the individual loci were aligned with Clustal X 1.83 (Thompson et al. 1997) or MAFFT 7.110 (Katoh and Standley 2013) using default settings.PhyloSuite V1.2.2 (Zhang et al. 2020) was used to concatenate the aligned sequences of the different loci.The few ambiguously aligned regions of the ITS and LSU alignments were removed with Gblocks v.0.91b, by keeping the default settings but allowing all gap positions when not ambiguous and manually adjusted in Sequencher 5.4.5 (Katoh et al. 2019;Castresana 2000).Phylogenetic analyses were carried out using maximum likelihood (ML) and Bayesian inference (BI) methods.The ML analysis was conducted with RAxML v. 8.2.3 (Stamatakis 2014) using a GTRGAMMA substitution model.ML bootstrap values (MLBS) of the nodes were evaluated using 1,000 rapid bootstrap replicates.For the BI approach, ModelFinder (Kalyaanamoorthy et al. 2017) was used to determine the appropriate substitution model that would best fit the DNA evolution for the combined dataset.MrBayes 3.2.7a(Ronquist et al. 2012) in the CIPRES Science Gateway version 3.3 was used to analyze the BI data.Best-fit evolution models were determined as GTR+I+G for the ITS and LSU partitions.Six simultaneous Markov chains were run for 50 million generations and trees were sampled every 1,000 th generation.The first 25% of created sample trees were discarded as they represent the burn-in phase of analysis.The remaining trees were used to calculate the Bayesian posterior probabilities (BPP) of the clades.
The resulting trees were viewed in FigTree v. 1.4.3 (Andrew 2016) and processed with Adobe Illustrator CS5.Branches that received MLBS ≥ 50% and BPP ≥ 0.95 were considered significantly supported.

Phylogenetic analysis
During this study, five strains of Colacogloea were discovered in the Baotianman Nature Reserve.To reveal the phylogenetic position of the specimens, we performed phylogenetic analyses with combined ITS and LSU sequence data.
The dataset consisted of 1,272 characters (674 characters from ITS and 598 characters from LSU), of which 715 were constant, 536 were variable, 366 were parsimony-informative, and 164 were singleton.ML and BI analyses generated similar topologies, with the BI analysis reaching an average standard deviation of split frequencies of 0.009922.The consensus topology from the ML analysis with MLBS (≥ 50%) and BPP (≥ 0.95) labeled on branches is shown (Fig. 1).In the phylogenetic trees, five strains isolated in this study formed two strongly supported groups (100% MLBS/1 BPP), and were clearly distinct from other known species of Colacogloea.The two strains NYUN 2210184 T and NYUN 221136 possess identical sequences in both the D1/D2 domains and ITS regions, indicating they belong to same species.The NYUN 2210184 T group formed a well-supported clade and then grouped with the unpublished strain Colacogloea sp.KBP: Y-6832 and C. armeniacae, with strong support (100 MLBS/1 BPP; Fig. 1).The D1/ D2 sequences of this group differed by only 3 nt substitutions (~0.5%) from Colacogloea sp.KBP: Y-6832; however, there were 16 nt (~2.9%) differences in the ITS regions, which indicates that the isolate KBP: Y-6832 may represent a different species.Similarly, the NYUN 2210184 T group differed from the type strain of the closest known species C. armeniacae by 6 nt (~1%) substitutions in the D1/D2 domains and by more than 15 nt (~2.5%) mismatches in the ITS regions.According to the basidiomycetous yeast species thresholds proposed by Fell et al. (2000), Scorzetti et al. (2002), and Vu et al. (2016), strains that differ by two or more nucleotide substitutions in the D1/D2 domains or 1-2% nucleotide differences in the ITS regions may represent different taxa.Therefore, the differences in both the D1/D2 and ITS sequences were significant enough for the NYUN 2210184 T group to be considered a distinct Colacogloea species.
Three strains NYNU 2110393 T , NYNU 2110421, and 2211185 formed a well-supported clade (100% MLBS/1 BPP; Fig. 1).They shared a 100% of nucleotide identity based on their D1/D2 and ITS sequences, indicating that they are conspecific.The closest relative of the NYNU 2110393 T group is C. retinophila, but differed from the type strain of the latter by six nt (~1%) substitutions in the D1/D2 domains and 31 nt (~5%) mismatches in the ITS regions, respectively.According to the criteria mentioned above, this data clearly supports the distinction between the NYNU 2110393 T group and C. retinophila at the species level.Etymology.The specific epithet "celtidis" refers to Celtis, the plant genus, from which the type strain was isolated.

Discussion
Traditional methods of classification for Colacogloea species are based primarily on phenotypical features, such as colony morphology, cell shape, basidia formation, details of physiological and biochemical characteristics, etc. (Sampaio et al. 2011).The classification based on these phenotypical features, however, was in many cases not consistent with the results obtained from phylogenetic analyses.For example, R. cycloclastica, R. philyla, and R. retinophila, originally classified in the polyphyletic anamorphic genus Rhodotorula, are nested within the genus Colacogloea based on phylogenetic analyses (Sampaio et al. 2011;Wang et al. 2015a).As a result, these three species were then reassigned to the genus Colacogloea, according to the International Code of Nomenclature for Algae, Fungi, and Plants (McNeill et al. 2012).Therefore, a combination of phenotypical characteristics and phylogenetic analysis has been adopted as the standard method for concretely identifying Colacogloea species (Wang et al. 2015b).
In this study, we introduce C. celtidis sp.nov.and C. pararetinophila sp.nov.as two new species of Colacogloea, and describe them in asexual morphs based on molecular analyses and morphological features.Our phylogenetic analyses indicated that the genus of Colacogloea has two subclades (Fig. 1), which are in concordance with previous studies (Wang et al. 2015a;Li et al. 2020;Wang et al. 2021;Schoutteten et al. 2023).C. celtidis sp.nov., with its sister species C. armeniacae, form a well-separated clade in subclade I, which is comprised of anamorphic species only.C. pararetinophila sp.nov., with its sister species C. retinophila, form a monophyletic lineage in subclade II, which includes eight teleomorphic species and seven anamorphic species.In previous studies, the two sub-clades were well supported in phylogenetic trees from the four protein-coding genes and the combined seven-loci analysis (Wang et al. 2015a;Wang et al. 2021;Schoutteten et al. 2023).Multi-gene phylogenetic analyses suggest that the two sub-clades of the genus Colacogloea seem to represent two genera, although that was not well supported in the phylogenetic tree produced by this study (Fig. 1).Therefore, further analyses using more molecular data or genomic data are needed to clarify the possible heterogeneity of the genus.
Colacogloea species are widely distributed and are found in different habitats.Filamentous morphs of Colacogloea species were mainly isolated from the hymenia of corticioid fungi, especially from the genera Peniophorella and Tubulicrinis (Sampaio et al. 2011;Schoutteten et al. 2023).The yeast morphs of Colacogloea species can be isolated from leaves, fruits, tree bark, plant residues, soil, insects, and tunnels (Belloch et al. 2007;Hamamoto et al. 2011;Sampaio 2011;Sampaio et al. 2011;Yurkov et al. 2016;Li et al. 2020;Wang et al. 2021), but most of them are found mostly in association with plant materials, especially leaves.Moreover, the yeast morphs of C. papilionacea and C. philyla, that were isolated from insects and insect tunnels, were also collected from plants (Kirschner and Oberwinkler 2000;Sampaio 2011;Sampaio et al. 2011).

Figure 1 .
Figure 1.Maximum likelihood (ML) phylogram of Colacogloea species based on combined ITS and LSU sequence data.Udeniozyma ferulica CBS 7416 T was used as the outgroup.Branches are labeled with MLBS ≥ 50% and BPP ≥ 0.95.New strains described in this study are shown in bold.

Table 1 .
Yeast strains and isolation sources investigated in this study.

Table 2 .
Taxon names, strain numbers, and GenBank accession numbers used for phylogenetic analyses.Entries in bold were newly generated for this study.
Phylogenetic analyses revealed that C. pararetinophila sp.nov.has a close relationship with C. retinophila with high support values (100 MLBP/1 BPP; Fig.1).C. pararetinophila sp.nov.can be physiologically differed from its closest

Table 3 .
Physiological and biochemical characteristics that differ between the new species and closely related species.
+, positive reWang et al. 2021ive reaction; d, delayed positive; s, slowly positive; w, weakly positive; n, data not available.All data from this study, except* which were obtained from the original description(Sampaio 2011;Wang et al. 2021).