A taxogenomic view of the genus Torulaspora: an expansion from ten to twenty-two species

© 2025 Westerdijk Fungal Biodiversity Institute You are free to share - to copy, distribute and transmit the work, under the following conditions: Attribution: You must attribute the work in the manner specified by the author or licensor (but not in any way that suggests that they endorse you or your use of the work). Non-commercial: You may not use this work for commercial purposes. No derivative works: You may not alter, transform, or build upon this work. For any reuse or distribution, you must make clear to others the license terms of this work, which can be found at https://creativecommons.org/licenses/by-nc-nd/4.0/. Any of the above conditions can be waived if you get permission from the copyright holder. Nothing in this license impairs or restricts the author’s moral rights. ISSN (Online) 1878-9080 https://doi.org/10.3114/persoonia.2025.54.08 Persoonia 54, 2025: 265–283 https://www.persoonia.org RESEARCH ARTICLE A taxogenomic view of the genus Torulaspora: an expansion from ten to twenty-two species M.R. Silva1,2#, F. Paraíso1,2#, J. Al-Oboudi3,4#, M. Abegg5,6, A. Aires1,2,7, K.O. Barros3,4,5, P.H. Brito1,2, M. Jarzyna3,4, K. Sylvester3,4, Q.K. Langdon3, D.A. Opulente3,4, F. Carriconde8, J.W. Fell9, T.A. Hofmann10, M.-A. Lachance11, J.-L. Legras12, D. Libkind13, A. Pontes1,2, P. Gonçalves1,2, C.A. Rosa5, M. Groenewald14, C.T. Hittinger3,4*, J.P. Sampaio1,2,7* 1UCIBIO, Departamento de Ciências da Vida, Faculdade de Ciências e Tecnologia, Universidade Nova de Lisboa, 2829-516 Caparica, Portugal 2Associate Laboratory i4HB - Institute for Health and Bioeconomy, Faculdade de Ciências e Tecnologia, Universidade Nova de Lisboa, 2829-516 Caparica, Portugal 3Laboratory of Genetics, Wisconsin Energy Institute, Center for Genomic Science Innovation, J. F. Crow Institute for the Study of Evolution, Microbiology Doctoral Training Program, University of Wisconsin-Madison, Madison, WI 53726, USA 4DOE Great Lakes Bioenergy Research Center, University of Wisconsin-Madison, Madison, WI 53726, USA 5Departamento de Microbiologia, ICB, C.P. 486, Universidade Federal de Minas Gerais, Belo Horizonte, MG, 31270-901, Brazil 6Instituto de Ciências Exatas e Tecnologia - ICET, Universidade Federal do Amazonas - UFAM, Itacoatiara - AM, 69100-000, Brazil 7PYCC – Portuguese Yeast Culture Collection, Departamento de Ciências da Vida, Faculdade de Ciências e Tecnologia, Universidade Nova de Lisboa, 2829-516 Caparica, Portugal 8Equipe “Sol & Végétation” (SolVeg), Institut Agronomique néo-Calédonien (IAC), 98000 Nouméa, New Caledonia, France 9Rosenstiel School of Marine and Atmospheric Science, University of Miami, Miami, FL 33149, USA 10Mycological Research Center (CIMi), Herbarium UCH, Universidad Autónoma de Chiriquí - UNACHI, 0427 David, Chiriquí province, Panama 11Department of Biology, University of Western Ontario, London, Ontario N6A 5B7, Canada 12SPO, Université de Montpellier, INRAE, Institut Agro, Montpellier 34060, France 13Centro de Referencia en Levaduras y Tecnología Cervecera (CRELTEC), Instituto Andino Patagónico de Tecnologías Biológicas y Geoambientales (IPATEC), Universidad Nacional del Comahue, CONICET, CRUB, Quintral 1250, San Carlos de Bariloche, 8400, Río Negro, Argentina 14Westerdijk Fungal Biodiversity Institute, Uppsalalaan 8, 3584CT Utrecht, The Netherlands #These authors contributed equally *Corresponding authors: C.T. Hittinger, [email protected]; J.P. Sampaio, [email protected] Abstract: The yeast genus Torulaspora (subphylum Saccharomycotina, family Saccharomycetaceae) is mostly known from its type species, T. delbrueckii, a frequent colonizer of wine and sourdough bread fermentations. The genus currently contains 10 species that are typically found in various natural terrestrial environments in temperate and tropical climates. Here we employ taxogenomic analyses to investigate a large collection of Torulaspora strains obtained in multiple surveys we carried out in Asia, Australasia, North America, South America, and Europe, and to which we added several strains maintained in culture collections. Our analyses detected twelve novel species that are formally described here, thereby more than doubling the species diversity of Torulaspora. We also sketch a genotype-phenotype map for the genus and show how the complex relationship between key genes and the physiological traits they control both between and within species. This remarkable increase in the number of species in the genus Torulaspora highlights how limited the current inventory of fungal taxa is. It also shows how integrated taxogenomic approaches can foster the assessment of species circumscriptions in fungi. Citation: Silva MR, Paraíso F, Al-Oboudi J, Abegg M, Aires A, Barros KO, Brito PH, Jarzyna M, Sylvester K, Langdon QK, Opulente DA, Carriconde F, Fell JW, Hofmann TA, Lachance M-A, Legras J-L, Libkind D, Pontes A, Gonçalves P, Rosa CA, Groenewald M, Hittinger CT, Sampaio JP (2025). A taxogenomic view of the genus Torulaspora: an expansion from ten to twenty-two species. Persoonia 54: 265–283. doi: 10.3114/persoonia.2025.54.08 Received: 27 September 2024; Accepted: 2 May 2025; Effectively published online: 23 May 2025 Corresponding editor: J. Houbraken Key words: new taxa phenotype phylogenomics taxogenomics Torulaspora yeast INTRODUCTION The genus Torulaspora currently contains 10 species that are typically found in various natural terrestrial environments, most frequently soil and plant material in temperate and tropical climates. The genus belongs to the family Saccharomycetaceae of the yeast subphylum Saccharomycotina (Groenewald et al. 2023) and is sister to Zygotorulaspora and Zygosaccharomyces (Shen et al. 2018), with Zygotorulaspora being intermediate between the other two genera. Salient features of Torulaspora are the formation of tapered protuberances resembling conjugation tubes in vegetative cells and the formation of persistent asci with up to four, but frequently two or one, ascospores (Kurtzman Persoonia – Volume 54, 2025266 2011). Physiologically, species of the genus Torulaspora are adapted to ferment simple sugars and exhibit the Crabtree effect; in other words, they ferment sugars when they are in high concentrations, even if oxygen is present. Torulaspora thus resembles Saccharomyces, although it is less efficient at sugar fermentation and ethanol production (Hagman et al. 2013). In contrast to Saccharomyces, their ancestor did not undergo the whole genome duplication documented in the lineage that includes Saccharomyces (Wolfe & Shields 1997). Torulaspora delbrueckii is the most emblematic species in the genus because it has been frequently isolated in association with commercial fermentations, particularly those related to the production of wine and sourdough bread, but also from other sources like for example cocoa and coffee fermentations. It is also very frequently isolated from the wild and has been described as cosmopolitan (Spurley et al., 2022). This species is considered an important microorganism for biotechnology due to its fermentative abilities and tolerance to various stresses such as freezing and freeze-thawing, ethanol toxicity, osmotic pressure, and acidic conditions (Pacheco et al., 2012). Besides the more conventional utilizations of T. delbrueckii mentioned above, novel applications have been proposed for brewing (Canonico et al. 2016) and cider fermentations (Wei et al. 2019), to aroma enhancement in co-fermentations (Zhang et al. 2018), and utilization as starter in frozen doughs (AlvesAraújo et al. 2004). Besides T. delbrueckii, available evidence suggests that T. microellipsoides might be also able to colonize anthropic niches. First, the few strains presently known were mostly isolated from artificial environments and, secondly, it has been extensively documented that wine strains of Saccharomyces cerevisiae horizontally acquired a 165-kb genomic region from T. microellipsoides (Marsit et al. 2015, Peter et al. 2018), thus suggesting co-existence in the same environment. The remaining Torulaspora species have been almost exclusively found in natural environments and in relatively low frequencies. For example, T. franciscae was found only once, in 1955, in a sample of Spanish soil, while T. maleeae, described in 2007, was based mostly on strains isolated from mosses collected in Thailand and from soil samples collected in tropical Japan (Kurtzman 2011). Although some strains of T. quercuum have been isolated from the arboreal and soil niches in temperate climates, other strains have been isolated from humans and dairy products (Wang et al. 2009). At present, T. delbrueckii is the only species with many isolates that have been obtained both in anthropic and arboreal niches. Thus, it is the only species in the genus for which more solid inferences on ecology and population structure have been made (Albertin et al. 2014, Silva et al. 2022). The use of genomics to evaluate yeast species circumscriptions and to inform taxonomy is gaining momentum (Libkind et al. 2020, Groenewald et al. 2023), and the number of novel species described under a taxogenomics approach is increasing (Libkind et al. 2011, Lopes et al. 2016, Haase et al. 2017, Morais et al. 2017, Čadež et al. 2019, David-Palma et al. 2020, Čadež et al. 2021, Santos et al. 2021, Brysch-Herzberg et al. 2023, Opulente et al. 2023, Santos et al. 2023). It is thus expected that, in the near future, taxogenomics will be the standard procedure for new species descriptions. Over the past two decades, we have conducted various yeast surveys mostly in wild but also in anthropic environments. Several strains were molecularly identified as belonging to Torulaspora based on comparisons of D1/D2 and ITS rDNA sequences. Here we analyse with genomics this diverse set of Torulaspora isolates collected in Asia, Australasia, North America, South America, and Europe, together with several collection strains collected as far back as 1930’s. Our taxogenomic analyses detected twelve novel species, thereby more than doubling the species diversity of the genus, which we propose to expand from ten to twentytwo species with formal taxonomic descriptions. We also sketch a genotype-phenotype map for the genus and show how key genes have evolved in concert with the physiological traits they control. MATERIALS AND METHODS Yeast isolation Given that each novel species was obtained in a different and independent survey, the details of the isolation procedures were specific for each of them. All information available on the isolations is shown in Supplementary Data S1. Phenotypic characterization Morphological and physiological tests were conducted according to the methods described by Kurtzman et al. (2011). In brief, the evaluation of the ability to utilize different compounds as carbon or nitrogen sources was carried out in test tubes with 5 mL of liquid media, incubated at 25 °C with gentle agitation for three wk. The ability to grow at different temperatures was carried out in yeast-peptonedextrose (YPD) broth (1 % yeast extract, 2 % peptone, 2 % glucose), in test tubes incubated in water baths at the desired temperatures. Ascospore formation was investigated on acetate agar (0.25 % yeast extract, 0.1 % glucose, 1 % potassium acetate, 1.5 % agar) and on corn meal agar (Difco). Dalmau plates for investigating formation of pseudohyphae and true hyphae were prepared with corn meal agar. DNA barcode sequence analysis (D1/D2 domain and complete ITS region of the rDNA) Using our dataset of assembled genomes, we retrieved using BLASTn (Altschul et al. 1990) the D1/D2 domain of the nuclear large subunit (LSU) 26S rDNA gene and the complete internal transcribed spacer region (ITS1–5.8S–ITS2). D1/ D2 and ITS sequences from the type strains of each new species served as queries in BLASTn searches against the nucleotide database (nt) of NCBI. The top 100 sequences from each BLASTn search were retrieved, and duplicates were removed. Sequences were confirmed with ITS1 and NL4 primers and aligned using MAFFT online service (Katoh et al. 2019). A Maximum Likelihood (ML) tree was constructed with IQ-TREE v. 2.2.6, using automatic detection of the best-fitting model and 1000 ultrafast bootstrap replicates. Alignments and phylogenetic trees have been deposited in figshare (https://doi.org/10.6084/m9.figshare.28794995.v1). Silva et al.: Novel species in Torulaspora 267 Whole-genome sequencing, phylogenetic analyses, and gene content analyses Paired-end Illumina NextSeq sequencing (300 cycles) was conducted at the Genomics Unit of Instituto Gulbenkian Ciência (Oeiras, Portugal) and at the University of Wisconsin Biotechnology Center. The publicly accessible reads for other Torulaspora species and the raw sequencing data underwent preprocessing two different ways using trimmomatic v. 0.39 (Bolger et al. 2014): just eliminating adapter sequences and with an optimized trimming step (LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:36) to eliminate low-quality bases. Subsequently, whole genome de novo assembly was carried out on all sets of pre-processed reads using SPAdes v. 3.13.1 (Bankevich et al. 2012). SPAdes was executed in ‘careful’ mode with k-mer sizes automatically determined based on the read length. The quality assessment of all resulting genome assemblies for each strain was performed using quast v. 5.0 (Gurevich et al. 2013), and the assembly with the highest genome size and N50 value was selected as the best. Small contigs (< 1 kb) were discarded from final assemblies. For T. delbrueckii PYCC 2477 and T. obscura PYCC 8933, long-read data was additionally acquired using Oxford Nanopore Technology (ONT), with a MinION flowcell. For de novo assembly, Canu v. 2.2 (Koren et al. 2017) was utilized with default parameters, only adjusting the genome size flag to 10m. The resulting contigs were corrected with two rounds of Racon v. 1.5.0 (Vaser et al. 2017), one using Nanopore reads and the other using Illumina reads. Afterwards, several rounds of Pilon v. 1.24 (Walker et al. 2014) were executed with Illumina data until no further modifications were observed in the change file. To enhance assembly contiguity, LINKS v. 1.8.7 (Warren et al. 2015) was implemented. For all genomes, including publicly available Torulaspora genome assemblies, ab initio prediction of protein coding genes and annotation was performed with Yeast Genome Annotation Pipeline (YGAP) (Proux-Wéra et al. 2012). To reconstruct the species phylogeny, we obtained singlecopy orthogroups (SCOs) using Orthofinder v. 2.5.4 (Emms & Kelly 2019) from the predicted proteomes of the Torulaspora genus. These SCOs were independently aligned using mafft v. 7.407 (Katoh & Standley 2013). Subsequently, the aligned SCO sequences were concatenated into a single dataset. The concatenated alignment, comprising 3766 SCOs, was then utilized to construct a Maximum Likelihood (ML) tree using IQ-TREE v. 2.2.6 (Nguyen et al. 2015) with a partition flag (-spp), an automatic detection of the best-fitting model of amino acid evolution (Kalyaanamoorthy et al. 2017) and 1000 ultrafast bootstrapping replicates (Hoang et al. 2018). For the gene content analyses, protein sequences were extracted with a tBLASTN search using T. delbrueckii sequences as queries with an expect value (e) cutoff of 1e-50. Only full-length hits, based on the expected length of the query sequence, were considered present to exclude fragmented or partial hits and draw more direct correlations between the presence of functional genes and positive phenotypes. When no hits were returned for a tBLASTN search at e < 1e-50, the gene was assumed to be either absent or fragmented beyond standard recognition for that genome. Substrate-specificities for alpha-glucosidase genes were inferred based on previous observations in various yeast species (Viigand et al. 2018), including multiple Torulaspora species (Silva et al. 2023). Alignments and phylogenetic trees have been deposited in figshare (https://doi.org/10.6084/m9.figshare.28788923.v1). Average nucleotide identities (ANI) Average nucleotide identities (ANI) values were estimated from whole genome assemblies using OrthoANI (Lee et al. 2016) with an improved algorithm that employees USEARCH (OrthoANIu), as implemented in the Orthologous Average Nucleotide Identity Tool (OAT) (Yoon et al. 2017). RESULTS Detection of novel Torulaspora species with taxogenomics The strains upon which the novel species descriptions are based (Table S1) were obtained from various isolation programs and surveys whose details are given in Supplementary Data S1. They were preliminarily identified using the sequences of the D1/D2 or ITS regions of the rDNA. After genome sequencing of strains suspected to represent novel species, single-copy orthogroups (SCOs) were identified from a predicted proteome dataset that included representatives of all known Torulaspora species. The resulting maximum likelihood phylogeny is shown in Fig. 1. The putative novel species were distributed across the genus and generally have as closest relatives already known Torulaspora species. To further assess discontinuities at the species level, we calculated average nucleotide identity (ANI) values within and between representatives of the main clades of Fig. 1. We took into consideration earlier studies suggesting 95 % or lower identity as a guideline for the separation of yeast (with et al. in italics: (Lachance et al. 2020, Libkind et al. 2020), microsporidian (de Albuquerque & Haag 2023), and prokaryotic (Jain et al. 2018) species, and expanded our earlier calculations for species in this genus (Silva et al. 2022). In the present study, all proposed novel species had ANI values that were equal or lower than 91 % when compared with their closest relatives, whereas intraspecific values were equal or higher than 96.5 % (Fig. 1 and Fig. S1). Taken together, our analyses suggested the existence of an additional twelve novel species, which would add to the ten species currently recognized in the genus Torulaspora, more than doubling its size. Recognition of the novel species using DNA barcode sequences To investigate if the new species could be recognized using DNA barcode sequences and to determine if additional representatives could be found among sequences deposited in the NCBI archive, we prepared extensive phylogenies based on the two DNA barcode regions most frequently used for the delineation of yeast species, the D1/D2 region of the LSU rDNA and the complete ITS region. The corresponding phylogeny for the D1/D2 region is shown in Fig. S2, the phylogeny based on the ITS region is shown in Fig. S3, and the phylogeny combining both regions is shown in Fig. S4. We included sequences from representative genomes used Persoonia – Volume 54, 2025268 Zygosaccharomyces bailii CBS 680 PYCC 9028 PYCC 9542 NRRL Y-1549T PYCC 8102T PYCC 8933T UFMG-CM Y5068 PYCC 5189 NCYC 2473 UCD657 CBS 11403T yHJA138 PYCC 9324 CLIB 1064 TBRC 10639T CBS 10694T UFMG-CM Y6992T PYCC 9889T PYCC 9890 PYCC 9027 PYCC 9026 PYCC 8418 PYCC 9025 PYCC 8417T CBS 16004T UFMG-CM Y750 UFMG-CM Y714 UFMG-CM Y7138 PYCC 9056 NRRL YB-1738 CBS 764T PYCC 2995T CBS 2947 PYCC 2997 PYCC 2996 CBS 12408T NRRL Y-2021 NRRL YB-3602 NRRL YB-3598 CBS 2952 NRRL YB-1481 CBS 5216 yHJA112T PYCC 8308 PYCC 9561 CLIB 1032 UWOPS 91-920.1 UWOPS 91-901.1 PYCC 9382 PYCC 8101T PYCC 8100 PYCC 8099 CBS 5080T CBS 2926T UWOPS 85-398.1-1T CBS 11123 CBS 11100T CBS 11124 CBS 11121 UWOPS 91-683.2 CBS 2785 CBS 9333 ZP 1874 UWOPS 83-1046.2 NRRL Y-17251T PYCC 9054 PYCC 9055 PYCC 9053 PYCC 9894 PYCC 9057 PYCC 9893T PYCC 2916 PYCC 8413 PYCC 8419 PYCC 9322 PYCC 2477T PYCC 6792 PYCC 5321 Tree scale: 0.1 T. delbrueckii T. incommunis T. franciscae T. cukson T. pahokee T. pretoriensis T. mapucheana T. microellipsoides T. nezacadeziae T. obscura T. quercuum T. nypae T. maleeae T. floroides T. gathmanii T. kanakia T. jiuxiensis T. globosa T. ventriculi T. indica T. markjohnstonii T. asahi 89 82 91 87 91 85 85 83 77 88 87 ANI Fig. 1. Phylogenomic placement of the novel taxa within the genus Torulaspora. The phylogeny is based on a concatenated alignment of 3766 single-copy orthologs from 77 Torulaspora genomes and was rooted with Zygosaccharomyces bailii. The JTT+F+I+G4 model of sequence evolution and the maximum-likelihood method were employed. Black bars highlight the novel species, and white bars represent the known ones. Nodes with black dots represent ≥ 95 % bootstrap support. The scale bar corresponds to the estimated number of substitutions per site. For the proposed novel species, numerals in black circles depict ANI values to their closest relatives. Silva et al.: Novel species in Torulaspora 269 in Fig. 1, as well as sequences retrieved from the GenBank database, whose accession numbers are listed in Table S2. Although, we observed that the species delimitation obtained with whole-genome data (Fig. 1) is not completely supported by the phylogenies based on those regions, all novel species could be recognized based on the DNA barcode sequences. The D1/D2 region (Fig. S2) appeared to have less resolution than the ITS region (Fig. S3) and thus we recommend the later for species identifications in the absence of whole genome sequences. In Fig. 3, a simplified ITS phylogeny including a single representative from each species summarizes these analyses. It also depicts the number of nucleotide substitutions observed between the more closely related species. Among these, species pairs showing the lowest number of nucleotide substitutions were T. delbrueckii - T. mapucheana sp. nov. and T. pretoriensis – T. cukson sp. nov. that differed from their closest relative by five and four substitutions each, respectively (Fig. 3). These substitutions were consistently found when multiple representatives were tested and are seen as apomorphies. i.e. unique for each species. Our extensive analyses using DNA barcode sequences shown in Figs S2, S3 and S4 allowed us to identify additional representatives of the novel species. For example, a considerable number of isolates of T. ventriculi sp. nov. was identified among strains isolated from soil in Cameroon and previously identified as T. globosa (Aljohani et al. 2018) (Fig. S3 and Table S2). In the case of T. incommunis sp. nov. for which a single isolate was initially available, we detected one D1/D2 (Fig. S2) and one ITS (Fig. S3) sequence in the pool of GenBank sequences analysed that appear to belong to two additional strains of this species (Table S2). In fact, our analyses allowed to enlarge the number of known representatives of most novel species and of all the already known species (Table S2). We also used these analyses to ascertain if any of the currently 15 recognized synonyms of T. delbrueckii coincides with any of the proposed novel species. These synonyms and their molecular identification are shown in Fig. S5. All synonyms were confirmed to belong to T. delbrueckii. Moreover, for eight of these synonyms, whole-genome sequences also validated their identification as T. delbrueckii. The genomic and phenotypic landscape of Torulaspora Our updated perspective of the genus Torulaspora, as revealed through whole genome sequences of 77 strains from 22 species, allowed us to take a snapshot of genome composition across all currently known species. Our first observation was that the genomic landscape of Torulaspora was highly variable, with a marked gene content oscillation both between and within species. Genomes of Torulaspora spp. contain, on average, 4978 (± 87) genes with a disparity of 409 genes between the most gene-rich, T. microellipsoides NRRL-Y-1549 (5219 genes), and the least gene-rich, T. asahi sp. nov. NBRC11086 (4810 genes). Although isolates of the same species exhibit some variability in gene-richness, this value was roughly similar within species. For example, the three isolates of T. mapucheana sp. nov. have an average of 4863 (± 43) genes, while the three isolates of T. microellipsoides contain an average of 5165 (± 46) genes. This variability was reflected in the presence or absence of genes involved in the metabolism of key carbohydrate sources and propagates to the phenotypic level (Fig. 2). For example, in T. delbrueckii and its closest relative, T. mapucheana sp. nov., the inability to grow on galactose was observed frequently and, in those cases, two of the three genes required for canonical galactose metabolism, GAL1, GAL7, and GAL10 (Hittinger et al. 2004), were absent or inactivated pseudogenes (Fig. 2). Conversely, galactose growth in T. delbrueckii, and in all other species in the genus, was linked to the presence of these three genes. Overall, variation across the genus was seen for maltose and melibiose metabolism, as well as for the presence or absence of the associated α-glucosidase-encoding genes (Fig. 2). Three classes of α-glucosidase-encoding genes were defined by analysing signature amino acids that correlate to substrate specificity (Table S4). Following Viigand et al. (2018), we tentatively grouped the α-glucosidase genes in three categories corresponding to the type of enzymes they encode: maltases, isomaltases, and mixed maltase– isomaltase activity. Our results are generally consistent with previous observations in known Torulaspora species (Silva et al. 2022) and suggest that maltose utilization is based on a widespread prevalence of α-glucosidase-encoding genes. Moreover, these genes appear to have been lost multiple times (Fig. 2, Table S4). Thus, we observed that consumption of the substrate predicted the presence of the gene coding for the canonically associated enzyme, but not vice versa. Those latter cases (i.e. gene presence but lack of the corresponding phenotype) might be explained by gene inactivation or gene expression impairments. Such variations occurred both between and within species. For example, we found that all five isolates of T. jiuxiensis contained the three GAL genes, while they were completely absent in all five isolates of T. kanakia sp. nov. Moreover, the intraspecific variation that we previously observed for T. delbrueckii (Silva et al. 2022) was observed here for other species, especially with respect to maltose and melezitose utilization (Fig. 2). However, the observed gene content variation at the species level was not universal. For traits, such as sucrose and trehalose consumption, the genes most commonly associated with these phenotypes were consistently found to be present in all Torulaspora genomes. As above, this consistent gene presence fails to act as a perfect predictor of a positive phenotype for sucrose, raffinose, and trehalose assimilation (Fig. 2). Taken together, these results paint the picture of a genomic landscape that is highly diverse, with frequent events of gene loss that appear to be recent in terms of the ancestry of the genus Torulaspora (i.e. encompass a single species or a group of closely related species). This variation is not solely associated with the phylogeny. While some genes and traits are found consistently in one species but not another, others vary substantially among isolates of the same species, and the genus Torulaspora as a whole displays substantial variation in phenotype both within and between species (Fig. S6). The correlation between assimilation and fermentation abilities varied between substrates and species, but as expected, assimilation was a prerequisite for fermentation. For example, all isolates that could assimilate sucrose could also ferment it; with just two exceptions, all isolates tested for galactose assimilation were also positive for galactose Persoonia – Volume 54, 2025270 Maltose (G) Galactose (G) Zygosaccharomyces bailii CBS 680 T. delbrueckii T. incommunis T. franciscae T. cukson T. pahokee T. pretoriensis T. mapucheana T. microellipsoides T. nezacadeziae T. obscura T. quercuum T. nypae T. maleeae T. floroides T. gathmanii T. kanakia T. jiuxiensis T. globosa T. ventriculi T. indica T. markjohnstonii T. asahi Galactose (F) Maltose (F) GAL1 GAL7 GAL10 Melezitose (G) Melezitose (F) Sucrose (F) Sucrose (G) AG (isomal) AG (mal) AG (unk) Raffinose (G) Raffinose (F) Melibiose (G) α,α-Trehalose (F) α,α-Trehalose (G) SUC2 NTH1 ATH1 Melibiose (F) MEL1 Fig. 2. Phenotype and gene content variation related to the utilization of selected carbohydrates across the genus Torulaspora. Squares and circles correspond to phenotypes and genotypes, respectively. A filled shape indicates a positive result (growth, G; fermentation, F; or gene presence), an empty shape indicates a negative result, and an absent shape indicates a lack of testing for that trait. Genes are designated using standard nomenclatures and capital letters corresponding to their encoded activities (isomal, isomaltase; mal, maltase; unk, unknown, encoding either for a maltase or isomaltase). Fermentation and assimilation results were obtained in this work or were taken from the literature as indicated in Table S3. Silva et al.: Novel species in Torulaspora 271 fermentation. A similarly strong correlation was found for maltose assimilation and fermentation. However, raffinose and trehalose did not exhibit this correlation; most isolates that exhibited assimilation of these carbohydrates did not also ferment them. These correlations between assimilation and fermentation sometimes persisted within species. For example, all three tested isolates of T. jiuxiensis both assimilated and fermented raffinose, whereas all four tested isolates of T. ventriculi sp. nov. only exhibited assimilation. Likewise, all three tested isolates of T. jiuxiensis both assimilated and fermented melibiose, whereas the two tested isolates of T. obscura sp. nov. only assimilated this carbohydrate. However, variation within species was common; for example, one T. asahi sp. nov. isolate could ferment raffinose but not trehalose, while the other two isolates could ferment trehalose but not raffinose. Taxonomy The formal descriptions of the new species are provided below. Species in which sexual structures were not observed are designated as forma asexualis, f.a. (Lachance 2012). Torulaspora asahi M. Silva, F. Paraíso, M.-A. Lachance & J.P. Sampaio, sp. nov. MB 853717. Fig. 4A–C. Etymology: Torulaspora asahi. a.sa.hi’, N.L. app. n. asahi, the Japanese word “asahi” meaning rising sun Japan’s sobriquet (“the Land of the Rising Sun”) as several strains of this species, including the ex-type strain, were isolated in that country. Tree scale: 0.1 Zygotorulaspora mrakii 15 17 20 27 13 30 33 24 30 5 5 PAH CUK PRE FRA 6 10 4 14 12 16 20 18 16 23 T. maleeae CBS 10694T T. floroides UFMG-CM-Y6992T T. gathmanii PYCC 9889T T. nypae TBRC 10639T T. jiuxiensis CBS 16004T T. kanakia PYCC 8417T T. globosa CBS 764T T. ventriculi PYCC 2995T T. indica CBS 12408T T. obscura PYCC 8933T T. microellipsoides NRRL Y-1549T T. nezacadeziaeae PYCC 8102T T. quercuum CBS 11403T T. asahi PYCC 8100T T. markjohnstonii PYCC 9891T T. delbrueckii PYCC 2477T T. mapucheana PYCC 9893T T. pretoriensis NRRL Y-17251T T. cukson UWOPS85-398.1-1T T. pahokee CBS 11100T T. franciscae CBS 2926T T. incommunis CBS 5080T 44 46 214 284 285 594 DEL C A T G G T MAP T T C A C - 33 62 67 113 184 189 225 258 631 696 698 PAH C G C - A T T C A A C CUK C A T - A T C C G G T PRE T A T C C A C T G G T (A) (C) (B) 64 65 120 278 289 684 ASA A G G G A A MAR G A C A G - Fig. 3. Phylogeny of all the species of the genus Torulaspora based on the complete ITS region. A. Phylogenetic tree constructed with the maximum-likelihood method and the TPM2u model of sequence evolution, chosen according to the Bayesian information criterion (BIC), and rooted with Zygotorulaspora mrakii. The final dataset had a total of 813 positions. Names in bold indicate the novel species. The scale bar represents the estimated number of substitutions per site. Percentage bootstrap values of 1000 replicates equal or higher that 95 % are depicted as black circles at tree nodes. B. For the most closely related species, the number of nucleotide substitutions in the ITS alignment is given. C. For the species pairs exhibiting the lowest number of nucleotide substitutions the location and nature of those substitutions is given (position 1 is the first position after the primer). Persoonia – Volume 54, 2025272 Table 1. Standard physiological and biochemical profiles of the novel Torulaspora species. T. asahi T. cukson T. floroides T. gathmanii T. incommunis T. kanakia T. mapucheana T. markjohnstonii T. nezacadeziaeae T. obscura T. pahokee T. ventriculi Fermentation D-Glucose s + + + + +/s + + + s + + D-Galactose w + - - + - - - - s/w - - Maltose - + w - w v - - - - v - Sucrose s + + w + +/w + +/s +/s s/w +/w +/s α-α-Trehalose s - - - - v - - - - - - Lactose - - - - - - - - - - - - Melibiose - - - - - - - - +/s - - - Melezitose - - +/s - w s - - - - - - Cellobiose - - - - - - - - - - - - Raffinose s - - - - - - - +/s v - - Assimilation Carbon sources D-Glucose + + + + + + + + + + + + D-Galactose +/s + - - + - - - + + - - L-Sorbose +/s - - - - - - s w +/s - - D-Glucosamine - - - - - -/w - - - - - - D-Ribose - - - - - - - - - - - - D-Xylose - - - - - - - - - - - - L-Arabinose - - - - - - - - - - - - D-Arabinose - - - - - - - - - - - - L-Rhamnose - - - - - - - - - - - - Sucrose + + + + + + + + + + + + Maltose - + + +/w + v - - s - +/s - α-α-Trehalose +/s + s - + v - +/s - v - - Methyl α-DGlucoside - + + + + v - v - - v - Silva et al.: Novel species in Torulaspora 273 Table 1. (Continued). T. asahi T. cukson T. floroides T. gathmanii T. incommunis T. kanakia T. mapucheana T. markjohnstonii T. nezacadeziaeae T. obscura T. pahokee T. ventriculi Cellobiose - - - - - - - - - - - - Salicin - - - - - - - -/w - s/w - - Melibiose - - - - - - - - + + - - Lactose - - - - - - - - - - - - Raffinose + + + + + +/s + + + + s +/s Melezitose - + + v + +/s - - - - - - Inulin v - - - - - s v - - - - soluble Starch - w w w w - w -/w - +/w w - Methanol - - - - - - - - - - - - Ethanol -/w + w w + +/s s/w +/w w +/w +/s + Glycerol s/w - + s/w - v s +/s + - v +/s Erythritol - - - - - - - - - - - - Ribitol - w w w + v w +/w w w w +/w Xylitol v + - - s - - s/w - - v - D-Glucitol + s + - + v + + - v v - D-Mannitol + s + w + s/w + + - s v - Galactitol - - - - - - - - - - - - myo-Inositol - - - - - - - - - - - - D-Glucono-1,5lactone - w + s s - s s s + s/w - D-Gluconate - - s w + - w v - + v - D-Glucuronate s - s - - +/s - - - - - - DL-Lactate v w + + + + s + w +/w - v Succinate - - - - - - - - - v - - Citrate - - - - - - - - - - - - L-Malic acid - - - - - - - - - - - - L-Tartaric acid - - - - - - - - - - - - Protocatechuic acid - - - - - - - - - - - - Persoonia – Volume 54, 2025280 distributions despite the limited number of isolates collected so far, with different collection sites separated by oceans. The first species was collected in Japan and Hawaii, the second was collected in North America and Europe, and the third was found in Africa (Mozambique) and in Papua New Guinea. Other Torulaspora species have very different distribution ranges. For example, T. franciscae is known from a limited number of soil isolates collected once in a single locality in Spain, more than 60 years ago (Capriotti 1958). Some of the novel species described here appear to be endemic to a particular region and likely have very restricted distributions, but further yeast isolation surveys are necessary to conclusively confirm or reject this possibility. For example, T. gathmanii was only found in Wisconsin, USA, T. kanakia was only collected in New Caledonia, and T. mapucheana was only found in temperate South America, in Patagonia. Among the novel species described here, T. cukson, T. floroides, and T. nezacadeziae are based on single isolates. Although we recognize that proposing novel species based on single strains might fall short of documenting the inherent biological variability and ecological range of the species, we believe that the recognition of these novel taxa might contribute to the discovery of additional strains. Torulaspora jiuxiensis exemplifies one such case. The species was described in 2022 based on two strains isolated in China (Chu et al. 2022). During our phylogenomic analyses, we expanded the number of representatives of this species by adding several arboreal isolates collected in Brazil and in Panama (Table S1). Among the novel species, glucose and sucrose were the sugars most frequently fermented, whereas melezitose was only fermented (weakly) by three species (T. incommunis, T. floroides, and T. kanakia) (Table S5). Overall, melibiose was infrequently fermented in the genus, with positive results for T. pretoriensis and variable results for T. delbrueckii (Table S6). Assimilation of galactose as sole source of carbon and energy was positive in six of the novel species and negative in the other six, with no intraspecific variable results (Table S5). This high variation is maintained when all species in the genus are considered (Table S6). Up to now, melibiose utilization had only been observed in three species, T. jiuxiensis, T. maleeae, and T. microellipsoides. Among the novel species, T. gathmanii, T. nezacadeziae, and T. obscura also assimilated this compound. Melezitose is another compound whose spectrum of utilization is markedly increased with the expansion of Torulaspora. It was previously only utilized by T. delbrueckii, T. pretoriensis, and T. maleeae, but five of the novel species can utilize it (T. cukson, T. incommunis, T. kanakia, T. gathmanii, and T. floroides). Finally, maximum growth temperatures were found to vary markedly among the novel species. At the lower end of the temperature spectrum, most of the novel species could grow at 30 °C, but not at 35 °C. The two exceptions were T. ventriculi, which could grow at 40 °C, and T. cukson, which could grow at 42 °C. The first species is notable in being found in the gut of some mammals, while the second one was found in the Arizona desert. These habitats might relate to the higher temperature adaptations seen in these two species. Torulaspora ventriculi is sister to T. indica that can grow even at 42 °C, while the closest relative of T. cukson, T. pahokee, has a maximum growth temperature around 30 °C. Of the 12 novel species, nine produced a sexual morph and formed asci with ascospores. Typically, asci contained less than four ascospores, normally one or two. We speculate if this pattern of low number of ascospore formation per ascus might be evolutionary connected with the conditions of low nutrient availability in environments like soil or tree bark. This might perhaps parallel what has been observed in S. cerevisiae, in which nutritional scarcity leads to the formation of less than four ascospores (Taxis et al. 2005). The remarkable increase in the number of species in the genus Torulaspora reported here highlights the limited status of the current inventory of fungi. It also shows how integrated taxogenomic approaches can foster the assessment of species circumscriptions in fungi. ACKNOWLEDEGMENTS This study was supported by Fundação para a Ciência e a Tecnologia (Portugal) grants 2021.08108.BD (M.R. Silva), UIDB/04378/2020 (UCIBIO), and LA/P/0140/2020 (i4HB). Research in the Hittinger Lab is funded by the National Science Foundation (DEB-2110403), USDA National Institute of Food and Agriculture (Hatch Project 7005101), in part by the DOE Great Lakes Bioenergy Research Center (DOE BER Office of Science DE–SC0018409), and an H.I. Romnes Faculty Fellowship (Office of the Vice Chancellor for Research with funding from the Wisconsin Alumni Research Foundation). Research conducted by CA Rosa was supported by project “INCT Yeasts: Biodiversity, preservation and biotechnological innovation”, funded by Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), Brazil, grant #406564/2022-1, and was also supported by CNPq grants 435040/2018 and 408733/2021-7. Research conducted by J.W. Fell was supported by grants from the National Science Foundation. Research conducted by M.- A. Lachance was supported by grants from the Natural Science and Engineering Research Council of Canada. The computational work was carried out with support of INCD funded by FCT and FEDER (Portugal) under grants 2215301/SAICT/2016, and FCT Advanced Computing Projects 2023.09581.CPCA.A1 (to AP) and 2022.15871.CPCA.A2 (to PHB). We gratefully acknowledge B. Turchetti (DBVPG) and H.-M. Daniel (BCCM/MUCL) for providing cultures, K. Wolfe for making available beforehand genome data of several T. indica strains, K. Bensch for assistance with the nomenclature of the new species, K. Shibayama for providing complementary information on the original isolation of T. incommunis, N. Fernandez and Y. Imanishi, for assistance in sample collection in Yuco, Lanin National Park, Argentina, and in Chiba, Japan, respectively, and citizen scientist L. Shown for contributing the sample from which T. markjohnstonii was isolated. Declaration on conflict of interest The authors declare that there is no conflict of interest. REFERENCES de Albuquerque NRM, Haag KL (2023). Using average nucleotide identity (ANI) to evaluate microsporidia species boundaries based on their genetic relatedness. Journal of Eukaryotic Silva et al.: Novel species in Torulaspora 281 Microbiology 70(1): e12944. https://doi.org/10.1111/jeu.12944 Aljohani R, Samarasinghe H, Ashu T, et al. (2018). Diversity and relationships among strains of culturable yeasts in agricultural soils in Cameroon. Scientific Reports 8: 1–11. https://doi. org/10.1038/s41598-018-34122-2 Albertin W, Chasseriaud L, Comte G, et al. (2014). Winemaking and bioprocesses strongly shaped the genetic diversity of the ubiquitous yeast Torulaspora delbrueckii. PLoS ONE 9(8): e94246. https://doi.org/10.1371/journal.pone.0094246 Altschul SF, Gish W, Miller W, et al. (1990). Basic local alignment search tool. Journal of Molecular Biology 215(3): 403–410. https://doi.org/10.1016/S0022-2836(05)80360-2 Alves-Araújo C, Almeida MJ, Sousa MJ, et al. (2004). Freeze tolerance of the yeast Torulaspora delbrueckii: Cellular and biochemical basis. FEMS Microbiology Letters 240(1): 7–14. https://doi.org/10.1016/j.femsle.2004.09.008 Bankevich A, Nurk S, Antipov D, et al. (2012). SPAdes: A new genome assembly algorithm and its applications to single-cell sequencing. Journal of Computational Biology 19(5): 455–477. https://doi.org/10.1089/cmb.2012.0021 Bolger AM, Lohse M, Usadel B (2014). Trimmomatic: A flexible trimmer for Illumina sequence data. Bioinformatics 30(15): 2114– 2120. https://doi.org/10.1093/bioinformatics/btu170 Brysch-Herzberg M, Jia GS, Sipiczki M, et al. (2023). Schizosaccharomyces lindneri sp. nov., a fission yeast occurring in honey. Yeast 40(3): 237–253. https://doi.org/10.1002/yea.3857 Čadež N, Bellora N, Ulloa R, et al. (2019). Genomic content of a novel yeast species Hanseniaspora gamundiae sp. nov. from fungal stromata (Cyttaria) associated with a unique fermented beverage in Andean Patagonia, Argentina. PLoS ONE 14(1): e0210792. https://doi.org/10.1371/journal.pone.0210792 Čadež N, Bellora N, Ulloa R, et al. (2021). Hanseniaspora smithiae sp. nov., a novel apiculate yeast species from Patagonian forests that lacks the typical genomic domestication signatures for fermentative environments. Frontiers in Microbiology 12: 679894. https://doi.org/10.3389/fmicb.2021.679894 Canonico L, Agarbati A, Comitini F, et al. (2016). Torulaspora delbrueckii in the brewing process: A new approach to enhance bioflavour and to reduce ethanol content. Food Microbiology 56: 45–51. https://doi.org/10.1016/j.fm.2015.12.005 Capriotti A (1958). Neue Hefen, isoliert aus Böden Spaniens II: Torulaspora franciscae nov. spec. Archives of Microbiology 28(4): 338–343. Chu SB, Hu WT, Hui FL (2022). Torulaspora jiuxiensis sp. nov., a novel yeast species isolated from rotting wood. International Journal of Systematic and Evolutionary Microbiology 72(11): 005629. https://doi.org/10.1099/ijsem.0.005629 Criscuolo A, Gribaldo S (2010). BMGE (Block Mapping and Gathering with Entropy): A new software for selection of phylogenetic informative regions from multiple sequence alignments. BMC Evolutionary Biology 10: 210. https://doi.org/10.1186/1471-214810-210 David-Palma M, Libkind D, Brito PH, et al. (2020). The untapped Australasian diversity of astaxanthin-producing yeasts with biotechnological potential—Phaffia australis sp. nov. and Phaffia tasmanica sp. nov. Microorganisms 8(10): 1518. https://doi. org/10.3390/microorganisms8111651 Emms DM, Kelly S (2019). OrthoFinder: Phylogenetic orthology inference for comparative genomics. Genome Biology 20(1): 238. https://doi.org/10.1186/s13059-019-1832-y Groenewald M, Hittinger CT, Bensch K, et al. (2023). A genomeinformed higher rank classification of the biotechnologically important fungal subphylum Saccharomycotina. Studies in Mycology 101: 1–20. https://doi.org/10.3114/sim.2023.105.01 Gurevich A, Saveliev V, Vyahhi N, et al. (2013). QUAST: Quality assessment tool for genome assemblies. Bioinformatics 29(8): 1072–1075. https://doi.org/10.1093/bioinformatics/btt086 Haase MAB, Kominek J, Langdon QK, et al. (2017). Genome sequence and physiological analysis of Yamadazyma laniorum f.a. sp. nov. and a reevaluation of the apocryphal xylose fermentation of its sister species, Candida tenuis. FEMS Yeast Research 17(1): fox019. https://doi.org/10.1093/femsyr/fox019 Hagman A, Säll T, Compagno C, et al. (2013). Yeast “MakeAccumulate-Consume” life strategy evolved as a multi-step process that predates the whole genome duplication. PLoS ONE 8(5): e68734. https://doi.org/10.1371/journal.pone.0068734 Hisatomi T, Toyomura K (2021). Isolation, identification, and characterization of wild budding yeasts from rose flowers in Fukuyama City, Hiroshima, Japan, and their application in bread and wine production. Mycoscience 62(6): 382–389. https://doi. org/10.47371/mycosci.2021.10.003 Hittinger CT, Rokas A, Carroll SB (2004). Parallel inactivation of multiple GAL pathway genes and ecological diversification in yeasts. Proceedings of the National Academy of Sciences 101(39): 14144–14149. https://doi.org/10.1073/pnas.0404319101 Jain C, Rodriguez-R LM, Phillippy AM, et al. (2018). High throughput ANI analysis of 90K prokaryotic genomes reveals clear species boundaries. Nature Communications 9: 5114. https://doi. org/10.1038/s41467-018-07641-9 Hoang DT, Chernomor O, Von Haeseler A, et al. (2018). UFBoot2: Improving the ultrafast bootstrap approximation. Molecular Biology and Evolution 35(2): 518–522. https://doi.org/10.1093/ molbev/msx281 Kalyaanamoorthy S, Minh BQ, Wong TKF, et al. (2017). ModelFinder: Fast model selection for accurate phylogenetic estimates. Nature Methods 14(6): 587–589. https://doi.org/10.1038/nmeth.4285 Katoh K, Standley DM (2013). MAFFT multiple sequence alignment software version 7: Improvements in performance and usability. Molecular Biology and Evolution 30(4): 772–780. https://doi. org/10.1093/molbev/mst010 Koren S, Walenz BP, Berlin K, et al. (2017). Canu: Scalable and accurate long-read assembly via adaptive κ-mer weighting and repeat separation. Genome Research 27(5): 722–736. https:// doi.org/10.1101/gr.215087.116 Kurtzman CP (2011). Torulaspora Lindner. 1904. In: The Yeasts, A Taxonomic Study (CP Kurtzman, JW Fell, T Boekhout, eds), 5th edn. Elsevier, Netherlands: 867–874. https://doi.org/10.1016/ B978-0-444-52149-1.00007-0 Kurtzman CP, Fell JW, Boekhout T, et al. (2011). Methods for isolation, phenotypic characterization and maintenance of yeasts. In: The Yeasts, A Taxonomic Study (CP Kurtzman, JW Fell, T Boekhout, eds), 5th edn. Elsevier, Netherlands: 87–110. https://doi.org/10.1016/B978-0-444-52149-1.00007-0 Lachance MA (2012). In defense of yeast sexual life cycles: The forma asexualis–an informal proposal. Yeast Newsletter 62: 24–25. Lachance MA, Lee DK, Hsiang T (2020). Delineating yeast species with genome average nucleotide identity: A calibration of ANI with haplontic, heterothallic Metschnikowia species. Antonie van Leeuwenhoek 113(12): 2097–2106. https://doi.org/10.1007/ s10482-020-01480-9 Lee I, Kim YO, Park SC, et al. (2016). OrthoANI: An improved algorithm and software for calculating average nucleotide identity. International Journal of Systematic and Evolutionary Persoonia – Volume 54, 2025282 Microbiology 66(2): 1100–1103. https://doi.org/10.1099/ ijsem.0.000760 Libkind D, Čadež N, Opulente DA, et al. (2020). Towards yeast taxogenomics: Lessons from novel species descriptions based on complete genome sequences. FEMS Yeast Research 20(1): foaa042. https://doi.org/10.1093/femsyr/foaa042 Libkind D, Hittinger CT, Valerio E, et al. (2011). Microbe domestication and the identification of the wild genetic stock of lager-brewing yeast. Proceedings of the National Academy of Sciences 108(35): 14539–14544. https://doi.org/10.1073/pnas.1105430108 Lopes MR, Morais CG, Kominek J, et al. (2016). Genomic analysis and D-xylose fermentation of three novel Spathaspora species: Spathaspora girioi sp. nov., Spathaspora hagerdaliae f.a., sp. nov. and Spathaspora gorwiae f.a., sp. nov. FEMS Yeast Research 16(2): fow022. https://doi.org/10.1093/femsyr/fow044 Marsit S, Mena A, Bigey F, et al. (2015). Evolutionary advantage conferred by an eukaryote-to-eukaryote gene transfer event in wine yeasts. Molecular Biology and Evolution 32(7): 1695–1707. https://doi.org/10.1093/molbev/msv057 Morais CG, Batista TM, Kominek J, et al. (2017). Spathaspora boniae sp. nov., a D-xylose-fermenting species in the Candida albicans/ lodderomyces clade. International Journal of Systematic and Evolutionary Microbiology 67(10): 3798 https://doi.org/10.1099/ ijsem.0.002186 Opulente DA, Langdon QK, Jarzyna M, et al. (2023). Taxogenomic analysis of a novel yeast species isolated from soil, Pichia galeolata sp. nov. Yeast 40: 608–615. https://doi.org/10.1002/ yea.3905 Pacheco A, Santos J, Chaves S, et al. (2012). The emerging role of the yeast Torulaspora delbrueckii in bread and wine production: Using genetic manipulation to study molecular basis of physiological responses. In: Structure and Function of Food Engineering (Eissa A, ed.), IntechOpen. https://doi. org/10.5772/46024 Peter J, De Chiara M, Friedrich A, et al. (2018). Genome evolution across 1,011 Saccharomyces cerevisiae isolates. Nature 556: 339–344. https://doi.org/10.1038/s41586-018-0030-5 Proux-Wéra E, Armisén D, Byrne KP, et al. (2012). A pipeline for automated annotation of yeast genome sequences by a conserved-synteny approach. BMC Bioinformatics 13: 237. https://doi.org/10.1186/1471-2105-13-237 Santos ARO, Aires A, Pontes A, et al. (2021). Phaffia brasiliana sp. nov., a yeast species isolated from soil in a cerrado–Atlantic rainforest ecotone site in Brazil. International Journal of Systematic and Evolutionary Microbiology 71: 005080. https:// doi.org/10.1099/ijsem.0.005080 Santos ARO, Barros KO, Batista TM, et al. (2023). Saccharomycopsis praedatoria sp. nov., a predacious yeast isolated from soil and rotten wood in an Amazonian rainforest biome. International Journal of Systematic and Evolutionary Microbiology 73: 006125. https://doi.org/10.1099/ijsem.0.006125 Shen X-X, Opulente DA, Kominek J, et al. (2018). Tempo and mode of genome evolution in the budding yeast subphylum. Cell 175: 1533–1545. https://doi.org/10.1016/j.cell.2018.10.023 Silva M, Pontes A, Franco-Duarte R, et al. (2022). A glimpse at an early stage of microbe domestication revealed in the variable genome of Torulaspora delbrueckii, an emergent industrial yeast. Molecular Ecology 32: 2396–2412. https://doi.org/10.1111/ mec.16428 Spurley WJ, Fisher KJ, Langdon QK, et al. (2022). Substrate, temperature, and geographical patterns among nearly 2000 natural yeast isolates. Yeast 39: 55–68. https://doi.org/10.1002/yea.3679 Taxis C, Keller P, Kavagiou Z, et al. (2005). Spore number control and breeding in Saccharomyces cerevisiae: A key role for a selforganizing system. Journal of Cell Biology 171: 627–640. https:// doi.org/10.1083/jcb.200507168 Vaser R, Sović I, Nagarajan N, et al. (2017). Fast and accurate de novo genome assembly from long uncorrected reads. Genome Research 27: 737–746. https://doi.org/10.1101/gr.214270.116 Walker BJ, Abeel T, Shea T, et al. (2014). Pilon: An integrated tool for comprehensive microbial variant detection and genome assembly improvement. PLoS ONE 9(11): e112963. https://doi. org/10.1371/journal.pone.0112963 Wang QM, Xu J, Wang H, et al. (2009). Torulaspora quercuum sp. nov. and Candida pseudohumilis sp. nov., novel yeasts from human and forest habitats. FEMS Yeast Research 9: 1322– 1326. https://doi.org/10.1111/j.1567-1364.2009.00567.x Warren RL, Yang C, Vandervalk BP, et al. (2015). LINKS: Scalable, alignment-free scaffolding of draft genomes with long reads. Gigascience 4(1): 35. https://doi.org/10.1186/s13742-015-00763 Wei J, Wang S, Zhang Y, et al. (2019). Characterization and screening of non-Saccharomyces yeasts used to produce fragrant cider. LWT - Food Science and Technology 107: 191–198. https://doi. org/10.3390/fermentation9090824 Wolfe KH, Shields DC (1997). Molecular evidence for an ancient duplication of the entire yeast genome. Nature 387(6634): 708– 713. https://doi.org/10.1038/42711 Yoon SH, Ha S min, Lim J, et al. (2017). A large-scale evaluation of algorithms to calculate average nucleotide identity. Antonie van Leeuwenhoek 110: 1281–1286. https://doi.org/10.1007/s10482017-0844-4 Zhang BQ, Luan Y, Duan CQ, et al. (2018). Use of Torulaspora delbrueckii co-fermentation with two Saccharomyces cerevisiae strains with different aromatic characteristics to improve the diversity of red wine aroma profile. Frontiers in Microbiology 9: 606. https://doi.org/10.3389/fmicb.2018.00606 Supplementary material Data S1. Different yeast isolation procedures used in this study. Fig. S1. Average nucleotide identity (ANI) matrix for the genus Torulaspora. The analysis used the same genomes as in Fig. 1. Inter-species ANI values are shown in white or black backgrounds and represent the averages of all possible genome comparisons between two species. Numerals in black backgrounds correspond to comparisons between novel species and their closest relatives. Pairwise averages of all relevant intraspecific ANI values are shown in coloured or grey backgrounds; these values are not available (NA) when only one genome per species was studied. Fig. S2. Phylogenetic tree based on a sequence alignment of the D1/ D2 region of the LSU rDNA of representatives of all Torulaspora species, including sequences retrieved from GenBank for which genome data are not available. The phylogeny tree was constructed with the maximum-likelihood method and the Tamura–Nei model of sequence evolution, with equal base frequence, and was rooted with Zygotorulaspora mrakii. The final dataset had a total of 588 positions. Names in black indicate sequences retrieved by us from genome assemblies (bold highlights type strains). Note that, in some cases, more than one sequence was retrieved from the genome. Names in blue represent sequences retrieved from GenBank; in these cases, the original species designations were maintained. Species Silva et al.: Novel species in Torulaspora 283 names and delimitations based on taxogenomics are indicated on the right side. The scale bar represents the estimated number of substitutions per site. Percentage bootstrap values of 1000 replicates are given at each node. Fig. S3. Phylogenetic tree based on a sequence alignment of the complete ITS region of the rDNA of representatives of all Torulaspora species, including sequences retrieved from GenBank for which genome data are not available. The phylogenetic tree was constructed with the maximum-likelihood method and the Hasegawa-Kishino-Yano model of sequence evolution and was rooted with Zygotorulaspora mrakii. The final dataset had a total of 843 positions. Names in black indicate sequences retrieved by us from genome assemblies (bold highlights type strains). Note that, in some cases, more than one sequence was retrieved from the genome. Names in blue represent sequences retrieved from GenBank; in these cases, the original species designations were maintained. Species names and delimitations based on phylogenomics are indicated on the right side. The scale bar represents the estimated number of substitutions per site. Percentage bootstrap values of 1000 replicates are given at each node. Fig. S4. Phylogenetic tree based on a sequence alignment combining the D1/D2 and ITS regions of the rDNA of representatives of all Torulaspora species, including sequences retrieved from GenBank for which genome data are not available. The phylogenetic tree was constructed with the maximum-likelihood method and the TPM2u model of sequence evolution and was rooted with Zygotorulaspora mrakii. Alignment positions containing gaps and missing data were eliminated before the analysis. The final dataset had a total of 1411 positions. Names in black indicate sequences retrieved by us from genome assemblies (bold highlights type strains). Note that, in some cases, more than one sequence was retrieved from the genome. Names in blue represent sequences retrieved from GenBank; in these cases, the original species designations were maintained. Species names and delimitations based on phylogenomics are indicated on the right side. The scale bar represents the estimated number of substitutions per site. Percentage bootstrap values of 1000 replicates are given at each node. Fig. S5. Molecular validation of the synonyms of T. delbrueckii. (A) List of synonyms of T. delbrueckii and details on ITS and wholegenome sequences. (B) Phylogenetic tree based on a sequence alignment of the complete ITS region of all species in the genus Torulaspora (depicted in black) and the current synonyms of T. delbrueckii (depicted in blue). The highlighted box corresponds to the species T. delbrueckii. The phylogenetic tree constructed with the maximum-likelihood method and the TIM2 model of sequence evolution, chosen according to the BIC, and rooted with Zygotorulaspora mrakii. The final dataset had a total of 813 positions. Names in bold indicate the novel species. The scale bar represents the estimated number of substitutions per site. Percentage bootstrap values of 1000 replicates equal or higher that 95 % are depicted as black circles at tree nodes. Fig. S6. Expanded phenotype map for the genus Torulaspora. The results of physiological tests including carbon fermentation, carbon assimilation, nitrogen assimilation, temperature preference, and sensitivity to select compounds for members of the Torulaspora genus are arranged by phylogenetic relationship, indicated by the associated tree. The presence of a square indicates a tested trait while the absence of a square indicates a trait that was not tested for that isolate. A full square indicates a positive phenotype (e.g. fermentation of a substrate or growth at a specified temperature) while an empty square indicates a negative phenotype (e.g. failure to assimilate a substrate or lack of growth in the presence of certain compounds). Phenotypic results were obtained in this work or were taken from the literature as indicated in Table S3. Table S1. Strains and genomes analysed in this study and relevant information pertaining to them (strains are ordered as in the phylogeny of Fig. 1). Table S2. Sources of D1/D2 and ITS rDNA sequences used to construct Figs S2, S3, and S4. Table S3. Sources of the phenotypic information depicted in Fig. 2. Table S4. Signature amino acids of α-glucosidase-encoding genes (strains are ordered as in Fig. 2). Signature amino acids that correlate to substrate specificity for maltases and isomaltases, numbered as in Saccharomyces cerevisiae IMA1, are compared for isomaltase-, maltaseand maltase/isomaltase-like sequences. Relevant amino acid signatures for maltase and isomaltase activity are highlighted in yellow and blue, respectively, and mixed maltase–isomaltase affinity is colour-coded in green. Table S5. Standard physiological and biochemical profiles of the novel Torulaspora species described this study (results shown for each strain). Table S6. Standard physiological and biochemical profiles of all species in the genus Torulaspora (species are ordered as in the phylogeny of Fig. 1). Persoonia – Volume 54, 2025284