Mycologia, 98(6), 2006, pp. 1006–1017.
`# 2006 by The Mycological Society of America, Lawrence, KS 66044-8897
`
`Phylogenetics of Saccharomycetales, the ascomycete yeasts
`
`Sung-Oui Suh
`Meredith Blackwell1,2
`Department of Biological Sciences, Louisiana State
`University, Baton Rouge, Louisiana 70803
`
`Cletus P. Kurtzman
`Microbial Genomics and Bioprocessing Research,
`National Center for Agricultural Utilization Research,
`ARS/USDA, Peoria, Illinois 61604-3999
`
`Marc-Andre´ Lachance
`Department of Biology, Western Ontario University,
`London, Ontario, Canada N6A 5B7
`
`Abstract: Ascomycete yeasts (phylum Ascomycota:
`subphylum Saccharomycotina: class Saccharomycetes:
`order Saccharomycetales) comprise a monophyletic
`lineage with a single order of about 1000 known
`species. These yeasts live as saprobes, often in
`association with plants, animals and their interfaces.
`A few species account
`for most human mycotic
`infections, and fewer than 10 species are plant
`pathogens. Yeasts are responsible for important
`industrial and biotechnological processes, including
`baking, brewing and synthesis of recombinant pro-
`teins. Species such as Saccharomyces cerevisiae are
`model organisms in research, some of which led to
`a Nobel Prize. Yeasts usually reproduce asexually by
`budding, and their sexual states are not enclosed in
`a fruiting body. The group also is well defined by
`synapomorphies visible at the ultrastructural
`level.
`Yeast identification and classification changed dra-
`matically with the availability of DNA sequencing.
`Species identification now benefits from a constantly
`updated sequence database and no longer relies on
`ambiguous growth tests. A phylogeny based on single
`gene analyses has shown the order to be remarkably
`divergent despite morphological similarities among
`members. The limits of many previously described
`genera are not supported by sequence comparisons,
`and multigene phylogenetic studies are under way to
`provide a stable circumscription of genera, families
`and orders. One recent multigene study has resolved
`species of the Saccharomycetaceae into genera that
`differ markedly from those defined by analysis of
`morphology and growth responses, and similar
`changes are likely to occur in other branches of the
`yeast tree as additional sequences become available.
`
`Accepted for publication 13 August 2006.
`1 Corresponding author. E-mail: mblackwell@lsu.edu
`2 Junior authors are listed alphabetically.
`
`Key words: Hemiascomycetes, rDNA, systematics
`
`INTRODUCTION
`
`The subphylum Saccharomycotina contains a single
`order, the Saccharomycetales (FIG. 1, TABLE I). These
`ascomycetes have had a significant role in human
`activities for millennia. Records from the Middle
`East, as well as from China, depict brewing and
`bread making 8000–10 000 years ago (McGovern et al
`2004, http://www.museum.upenn.edu/new/exhibits/
`online_exhibits/wine/wineintro.html). These early
`‘‘biotechnologists’’ had no idea why bread dough rose
`or why beer fermented. The explanation awaited
`Antonie van Leeuwenhoek (1680), who showed with
`his microscope that small cells that were to become
`known as yeasts were present and further for Louis
`Pasteur (1857) to demonstrate conclusively that yeasts
`caused the fermentation of grape juice to wine.
`
`What is a yeast? —Yeasts eventually were recognized
`as fungi, prompting description of
`the ‘‘sugar
`fungus’’
`from beer to be named Saccharomyces
`cerevisiae. Yeasts usually grow as single cells with
`division by budding (FIGS. 2, 3) or less frequently by
`fission (FIG. 4), and asci and ascospores (FIGS. 5–13)
`are not produced in fruiting bodies as is common for
`many other kinds of ascomycetes. In addition both
`meiotic nuclear divisions occur within an intact
`nuclear envelope and the enveloping membrane
`system associated with each ascospore during delimi-
`tation has an independent origin. It initially was
`believed that all yeasts were ascomycetes. The
`presence of ballistoconidia in such yeasts as the
`asexual genus Sporobolomyces however led Kluyver
`and van Niel (1927) to suggest that some yeasts might
`be basidiomycetes. A clear connection was provided
`by Nyland’s (1949) discovery of the teliospore-form-
`ing genus Sporidiobolus, which was followed by
`descriptions of teliosporic life cycles for a number
`of other genera (e.g. Banno 1967, Fell et al 1969,
`Kwon-Chung 1975, Boekhout et al 1991). In addition
`to ascomycete and basidiomycete yeasts, the term
`‘‘yeast-like’’ also has been extended to the cellular
`phase of dimorphic members of
`the zygomycete
`genus Mucor (Flegel 1977), the ‘‘black yeasts’’ (de
`Hoog 1999), which comprise diverse pigmented
`ascomycete genera such as Aureobasidium, Fonsecaea
`and Phaeococcomyces, and even certain achlorophyl-
`lous algae in the genus Prototheca (Kurtzman and Fell
`1998). Dimorphic (e.g. species of Ajellomyces and
`
`1006
`
`LCY Biotechnology Holding, Inc.
`Ex. 1046
`Page 1 of 12
`
`

`

`SUH ET AL: SACCHAROMYCETALES PHYLOGENETICS
`
`1007
`
`Coccidioides immitis) (Bowmann et al 1992) and
`‘‘yeast-like’’ fungi are ascomycetes (e.g. Symbiotaph-
`rina and unnamed endosymbionts of plant-hoppers)
`that have a yeast growth form in their life history (Suh
`et al 2001). They are derived however from within
`several lineages of mycelial ascomycetes and now are
`excluded from the Saccharomycetales. One more
`striking finding on the way to defining a monophyletic
`group was the discovery that members of Schizosac-
`charomyces are not in the Saccharomycetales clade
`(see Taylor et al 1993).
`
`In search of a monophyletic group.—From the moment
`that Banno (1967) demonstrated the formation of
`clamped hyphae and basidiospores by zygotes arising
`from crosses between strains of a Rhodotorula species,
`a growing list of deep-seated differences between
`ascomycete and basidiomycete yeasts gradually be-
`came obvious: (i) Cell wall polysaccharide composi-
`tion is dominated by chitin in the basidiomycetes and
`b-glucans in the ascomycetes; (ii) nuclear DNA
`guanine + cytosine (G + C) composition tends to be
`higher than 50% in basidiomycetes and lower than
`50% in ascomycetes (Kurtzman and Fell 1998);
`(iii) bud formation is typically enteroblastic in
`basidiomycetes and holoblastic in ascomycetes;
`(iv) ascomycete yeasts are generally more fermenta-
`tive, more copiotrophic and at
`the same time
`specialized nutritionally, more fragrant and mostly
`hyaline. This is in contrast to basidiomycete species,
`which more often form mucoid colonies, display
`intense carotenoid pigments and tend to use
`a broader range of carbon compounds more effi-
`ciently at lower concentrations (Kurtzman and Fell
`1998); (v) in terms of diagnostic tests, the diazonium
`blue B reaction is almost always positive in basidio-
`mycete yeasts and negative in ascomycete yeasts (van
`der Walt and Hopsu-Havu 1976); and (vi) ascomycete
`yeasts are profoundly different ecologically and often
`found in specialized niches involving interactions
`with plants and insects or other invertebrate animals
`that they rely upon for dispersal. These niches tend to
`be liquid and rich in organic carbon. In contrast
`basidiomycete yeasts would seem to be adapted to the
`colonization of nutrient-poor, solid surfaces and
`might not rely to the same extent on animal vectors
`for their dispersal (Lachance and Starmer 1998).
`The relationship of Saccharomycetales to other
`fungi has been the subject of many hypotheses.
`Because yeast cellular morphology is rather simple,
`it was believed that yeasts are primitive organisms.
`This idea gained some support because yeast gen-
`omes tend to be smaller than those of many other
`fungi and they have fewer introns in their gene
`sequences. The philosophy of yeast relationships
`
`changed in the 1970s (e.g. Cain 1972) with the
`proposal that yeasts represent morphologically re-
`duced forms of filamentous fungi and, using this
`reasoning, some yeasts were classified into families
`with molds (Redhead and Malloch 1977). Gene
`sequence analyses have shown many of these ideas
`to be incorrect.
`
`Identification of yeasts.— Earlier classifications of
`yeasts at
`lower taxonomic levels were based on
`presence or absence of a sexual state, type of cell
`division, presence or absence of hyphae and pseudo-
`hyphae, fermentation of simple sugars and growth on
`various carbon and nitrogen compounds. Traditional
`genetic crosses however showed that strains differing
`in morphological and metabolic characters could be
`members of the same species, which cast considerable
`doubt on the importance of these commonly used
`taxonomic and phylogenetic characters. These doubts
`prompted yeast taxonomists to turn to DNA-based
`methods for species delineation.
`The transition from phenotypic identification of
`yeasts to molecular identification began with de-
`termination of the mol% G + C ratios of nuclear
`DNA. These analyses demonstrated that ascomycete
`yeasts have a range of ca. 28–50 mol% G + C whereas
`basidiomycete yeasts have a range of ca. 50–70 mol%
`G + C. Strains that differed by 1–2 mol% were
`recognized as separate species (Price et al 1978,
`Nakase and Komagata 1968). Quantitative assessment
`of genetic similarity between strains and species
`subsequently was determined by the technique of
`nuclear DNA re-association or hybridization (i.e. the
`extent of heteroduplex formation between the DNAs
`compared. DNA from the species pair of interest is
`sheared, made single-stranded, and the degree of
`heteroduplex formation between the pair is deter-
`mined from the extent of re-association [Price et al
`1978, Kurtzman 1993]). On the basis of shared
`phenotype, strains with 80% or greater re-association
`were proposed to represent members of the same
`yeast species (Martini and Phaff 1973, Price et al
`1978). Correlation of this measure with the biological
`species concept has been examined from genetic
`crosses using both heterothallic and homothallic
`species. These results also lead to the conclusion that
`strains showing ca. 70% or greater heteroduplex
`formation are likely to be members of the same
`species (Kurtzman et al 1980a, b; Smith et al 2005).
`Despite the remarkable impact
`that DNA re-
`association experiments have had on yeast systemat-
`ics, the re-association technique is slow and labor
`intensive and resolution does not extend beyond
`closely related species. Consequently DNA sequenc-
`ing has been widely adopted to understand species
`
`LCY Biotechnology Holding, Inc.
`Ex. 1046
`Page 2 of 12
`
`

`

`1008
`
`MYCOLOGIA
`
`FIG. 1. Relationships of selected members of the Saccharomycetales. Consensus of 12 most parsimonious trees based on
`the combined dataset of SSU and LSU rDNA sequences with all missing and ambiguous characters deleted. Species of basal
`ascomycetes were used as outgroup taxa. The species shown in boldface are type species of each genus. The sequences of
`Dekkera bruxellensis, Pichia kudriavzevii and Candida tropicalis were combined with those from different strains of the species,
`
`LCY Biotechnology Holding, Inc.
`Ex. 1046
`Page 3 of 12
`
`

`

`SUH ET AL: SACCHAROMYCETALES PHYLOGENETICS
`
`1009
`
`relationships because of its rapidity and resolution of
`both close and distant relationships (Kurtzman and
`Robnett 1998). Nonetheless DNA relatedness studies
`provided a strong foundation upon which to interpret
`sequence analyses.
`
`MATERIALS AND METHODS
`
`A total of 95 taxa were used in the analyses for this study,
`including three basal ascomycetes and five members of
`Pezizomycotina (SUPPLEMENTARY TABLE I). For the Saccha-
`romycetales the type species of each genus listed in
`Kurtzman and Fell (TABLE I, 1998) was chosen (see FIG. 1
`for further details). In addition ecologically important taxa
`that have appeared in monophyletic groups in other studies
`also were included in the analyses. DNA sequences at five
`loci (i.e. nuclear small subunit [SSU] ribosomal RNA gene
`[rDNA], D1/D2 region of nuclear large subunit [LSU]
`rDNA, elongation factor 1a gene (EF-1a), and the largest
`and the second largest subunits of RNA polymerase II gene
`[RPB1 and RPB2]) were obtained from GenBank and the
`AFTOL database. Based on the availability of sequence data
`all genes except LSU rDNA were compared with a limited
`number of taxa (i.e. 73 taxa for SSU rDNA, 27 taxa for
`RPB2, 30 taxa for EF-1a and 13 taxa for RPB1) (TABLE II,
`online supplement).
`Initially DNA sequences were aligned with the multi-
`alignment program Clustal X (Thompson et al 1997) and
`were optimized visually. Phylogenetic analyses were con-
`ducted with parsimony, Bayesian and distance analyses with
`individual genes as well as concatenated datasets. Maximum
`parsimony analyses were performed with PAUP 4.0b10
`(Swofford 2002). Heuristic tree searches were executed with
`the tree bisection-reconnection branch-swapping algorithm
`with random sequence analysis. Bootstrap values for the
`most parsimonious tree were obtained from 1000 replica-
`tions. Bayesian Markov chain Monte Carlo (B-MCMC)
`analyses were performed with MrBayes v3.0b4 (Ronquist
`and Huelsenbeck 2003). The analysis consisted of 1 000 000
`generations of four chains sampled every 10 generations;
`the first 100 000 generations were discarded as burn-in,
`and the remaining trees were used to obtain a majority
`rule consensus tree for estimating the posterior proba-
`bility of
`the branches. Neighbor joining analyses were
`conducted using PAUP 4.0b10 with the Kimura 2 parameter
`option.
`
`RESULTS AND DISCUSSION
`
`Phylogenetic relationships among yeasts.—The tree
`(FIG. 1) is based on a combined SSU rDNA and D1/
`D2 LSU rDNA dataset. The D1/D2 LSU rDNA region
`has been sequenced for almost all known yeasts as an
`identification tool and also to estimate phylogenetic
`relationships in the Saccharomycetales (Kurtzman and
`Robnett 1998). About 20 complete yeast genome
`sequences have been determined, but genome se-
`quences for species in many basal clades, such as
`Zygoascus, Trichomonascus and Kodamaea (FIG. 11),
`are lacking. A stable tree based on phylogenetic
`analyses including representatives of the deep lineages
`is needed to provide a stable phylogenetic classification
`system. We will present information on 12 clades that
`are well supported in analyses based on rDNA
`sequences and discuss the support for the clades.
`Support for a few lineages is increased by sequences
`from protein-coding genes, but protein-coding genes
`are not available for clades 2, 3, 4, 5, 6, 9, 10, 11 and 12
`(FIG. 1; TABLE II, online supplement).
`As in many groups of fungi, the use of a morpho-
`logical form concept has resulted in the circumscrip-
`tion of many genera and families that are not
`monophyletic. However by using phylogenetic analy-
`sis of a multigene dataset one clade is well supported
`(FIG. 1, clade 1) and corresponds with Saccharomy-
`cetaceae (TABLE I) (Kurtzman 2003, Kurtzman and
`Robnett 2003). In other cases groups are almost
`certainly polyphyletic and work is under way to
`redefine them and stabilize the nomenclature based
`on a concept of monophyly. For example the genus
`Pichia currently extends across the full phylogenetic
`spectrum of ascomycetous yeasts because species are
`characterized by budding cells that form hat-shaped
`or spherical ascospores, but relationships cannot be
`determined based on this simple, convergent pheno-
`type. Thus separation of
`the new monophyletic
`genera on the basis of phenotype probably will not
`be possible. Several studies using multigene analyses
`soon will divide Pichia into about 20 genera. The
`Pichia membranifaciens clade (FIG. 1, clade 10), in-
`cluding species of Issatchenkia, will comprise the
`
`r b
`
`ut the strains in each species were conspecific based on the sequences in the D1/D2 or ITS regions. Tree length 5 3232;
`consistency index 5 0.3583; homoplasy index 5 0.6417; retention index 5 0.5710; rescaled consistency index 5 0.2046.
`Numbers above branches or to the left of slashes (/) indicate support above 50% in 1000 bootstrap replicates with parsimony
`analysis. Numbers below branches or to the right of slashes represent probability of nodes in Bayesian analysis. Clade numbers
`1–12 in the tree correspond to those in the text. SSU, SSU rDNA; LSU, LSU rDNA; T, type strain; NT, neotype strain; A,
`authentic strain; NRRL, Agricultural Research Service Culture Collection, National Center for Agricultural Utilization
`Research, Peoria, Illinois.
`
`LCY Biotechnology Holding, Inc.
`Ex. 1046
`Page 4 of 12
`
`

`

`1010
`
`MYCOLOGIA
`
`TABLE I. Currently recognized families and genera of the
`subphylum Saccharomycotina: class Saccharomycetes: order
`Saccharomycetales.1,2 The classification provided is
`a revision of the current classification from ‘‘Outline of
`Ascomycota—2006’’ (Vol. 12, 22 Mar 2005) available at
`Myconet http://www.fieldmuseum.org/myconet/outline.
`asp#sub_sacch
`
`Saccharomycotina
`Saccharomycetes
`Saccharomycetales Kudryavtsev
`Ascoideaceae J. Schro¨ ter
`Ascoidea Brefeld & Lindau (T)
`Cephaloascaceae L. R. Batra
`Cephaloascus Hanawa (T)
`Dipodascaceae Engler & E. Gilg
`Dipodascus Lagerheim (T)
`Galactomyces Redhead & Malloch (T)
`Geotrichum Link:Fries (A)
`Endomycetaceae J. Schro¨ ter
`Endomyces Reess (T)
`Helicogonium W. L. White (T)
`Myriogonium Cain (T)
`Phialoascus Redhead & Malloch (T)
`Eremotheciaceae Kurtzman
`Coccidiascus Chatton emend. Lushbaugh, Rowton
`& McGhee (T)
`Eremothecium Borzi emend. Kurtzman (T)
`Lipomycetaceae E. K. Novak & Zsolt
`Babjevia van der Walt & M. Th. Smith (T)
`Dipodascopsis Batra & Millner (T)
`Lipomyces Lodder & Kreger van Rij (T)
`Myxozyma van der Walt, Weijman & von Arx (A)
`Zygozyma van der Walt & von Arx (T)
`Metschnikowiaceae T. Kamienski
`Clavispora Rodrigues de Miranda (T)
`Metschnikowia T. Kamienski (T)
`Pichiaceae Zender
`Brettanomyces Kufferath & van Laer (A)
`Dekkera van der Walt (T)
`Kregervanrija Kurtzman (T)
`Pichia Hansen (pro parte) (T)
`Saturnispora Liu & Kurtzman (T)
`Saccharomycetaceae G. Winter
`Kazachstania Zubkova (T)
`Kluyveromyces Kurtzman, Lachance, Nguyen &
`Prillinger (T)
`Lachancea Kurtzman (T)
`Nakaseomyces Kurtzman (T)
`Naumovia Kurtzman (T)
`Saccharomyces Mayen ex Reess (T)
`Tetrapisispora Ueda-Nishimura & Mikata (T)
`Torulaspora Lindner (T)
`Vanderwaltozyma Kurtzman (T)
`Zygosaccharomyces Barker (T)
`Zygotorulaspora Kurtzman (T)
`Saccharomycodaceae Kudryavtsev
`Hanseniaspora Zikes (T)
`Kloeckera Janke (A)
`Saccharomycodes Hansen (T)
`
`TABLE I. Continued
`
`Saccharomycopsidaceae von Arx & van der Walt
`Saccharomycopsis Schio¨ nning (T)
`Saccharomycetales incertae sedis
`Aciculoconidium King & Jong (A)
`Ambrosiozyma van der Walt (T)
`Ascobotryozyma J. Kerrigan, M. Th. Smith & J. D.
`Rogers (T)
`Blastobotrys von Klopotek (A)
`Botryozyma Shann & M. Th. Smith (A)
`Candida Berkhout (A)
`Citeromyces Santa Marı´a (T)
`Cyniclomyces van der Walt & Scott (T)
`Debaryomyces Lodder & Kreger-van Rij (T)
`Hyphopichia von Arx & van der Walt (T)
`Kodamaea Y. Yamada, T. Suzuki, Matsuda & Mikata
`emend. Rosa, Lachance, Starmer, Barker, Bowles
`& Schlag-Edler (T)
`Komagataella Y. Yamada, Matsuda, Maeda &
`Mikata (T)
`Kuraishia Y. Yamada, Maeda & Mikata (T)
`Lodderomyces van der Walt (T)
`Macrorhabdus Tomaszewski, Logan, Snowden,
`Kurtzman & Phalen (A)
`Nadsonia Sydow (T)
`Nakazawaea Y. Yamada, Maeda & Mikata (T)
`Ogataea Y. Yamada, Maeda & Mikata (T)
`Pachysolen Boidin & Adzet (T)
`Phaffomyces Y. Yamada, Higashi, S. Ando & Mikata
`(T)
`Schizoblastosporion Ciferri (A)
`Sporopachydermia Rodrigues de Miranda (T)
`Starmera Y. Yamada, Higashi, S. Ando & Mikata (T)
`Starmerella Rosa & Lachance (T)
`Sugiyamaella Kurtzman & Robnett (T)
`Trichomonascus Jackson (T)
`Trigonopsis Schachner (A)
`Wickerhamia Soneda (T)
`Wickerhamiella van der Walt (T)
`Yamadazyma Billon-Grand emend. M. Suzuki,
`Prasad & Kurtzman (T)
`Yarrowia van der Walt & von Arx (T)
`Zygoascus M. Th. Smith (T)
`
`1 (A) 5 anamorphic genus, (T) 5 teleomorphic genus.
`2 Phylogenetic relationships of many yeast genera are
`unclear and these are placed in Saccharomycetales incertae
`sedis until family relationships become known.
`
`residual, much reduced Pichia. Other groups to be
`extracted from Pichia include the methanol-assimilat-
`ing species (FIG. 1, clade 2), the Pichia anomala clade
`(FIG. 1, clade 3), the xylose-fermenting species in the
`Pichia stipitis clade (FIG. 1, clade 5) and numerous
`other smaller clades.
`Another continuing source of disparity is the use of
`the genus Candida as a dumping ground for most
`
`LCY Biotechnology Holding, Inc.
`Ex. 1046
`Page 5 of 12
`
`

`

`SUH ET AL: SACCHAROMYCETALES PHYLOGENETICS
`
`1011
`
`FIGS. 2–10. Ascomycete yeasts in pure culture. 2. Saccharomyces cerevisiae. Cells dividing by multilateral budding. 3.
`Saccharomycodes ludwigii. Cell division by bipolar budding. Note the ascus with four spherical ascospores above the dividing
`cell. 4. Schizosaccharomyces pombe. Cell division by fission. Once divided, the newly formed cells often will be morphologically
`indistinguishable from cells formed by budding. 5. Eremothecium (Nematospora) coryli. Free, needle-shaped ascospores with
`whip-like tails of extended wall material. Members of this group are some of the few yeasts that are plant pathogens. 6.
`Pachysolen tannophilus. A single ascus forms on the tip of an elongated refractile tube. The ascus wall becomes deliquescent
`and releases four hat-shaped ascospores. P. tannophilus was the first yeast discovered to ferment the pentose sugar D-xylose,
`a major component of hemicellulose from biomass. 7. Lodderomyces elongisporus. Persistent ascus with a single ellipsoidal
`ascospore. This is the only species of the clade, which includes Candida albicans and C. tropicalis, that is known to form
`ascospores. 8. Torulaspora delbrueckii. Asci with 1–2 spherical ascospores. Asci often form an elongated extension that may
`function as a bud conjugant. 9. Zygosaccharomyces bailii. Asci with spherical ascospores. Often there are two ascospores per
`conjugant giving rise to the term ‘‘dumbbell-shaped’’ asci. This species is one of the most aggressive food spoilage yeasts
`known. 10. Pichia bispora. Ascospores are hat-shaped and released from the asci at maturity. Hat-shaped ascospores are
`produced by species in a variety of different genera. FIGS. 2–4, phase contrast. FIGS. 5–10, bright field. Bars 5 5 mm.
`
`budding yeasts that do not form ascospores. Members
`of the current genus are found in essentially all
`teleomorphic clades. Division of Candida into a large
`number of monophyletic genera based on phyloge-
`netic analysis has little appeal to taxonomists because
`most will be unrecognized from phenotype. Perhaps
`the best descriptor for these clades is to note their
`
`association with phylogenetically defined ascosporic
`genera, some of which also may be phenotypically
`inseparable. Although the yeast identification of the
`near future will rest on phylogenetic analysis of gene
`sequences rather than from phenotypic characters,
`this is not so different from some previous circum-
`scriptions of taxa based heavily on metabolic data.
`
`LCY Biotechnology Holding, Inc.
`Ex. 1046
`Page 6 of 12
`
`

`

`1012
`
`MYCOLOGIA
`
`FIGS. 11–15. Scanning electron micrographs. 11. Ascospores of Kodamaea anthophila. Reproduced by permission from the
`Society for General Microbiology (Rosa et al 1999). 12. End of ascospores of Metschnikowia borealis released by treatment of
`the ascus with Mureinase. Reproduced by permission (Lachance et al 1998). 13. Agglutinated ascospores of Saccharomycopsis
`synnaedendrus. B. Schlag-Edler and M.-A. Lachance. 14, 15. Elongated predaceous cells of Arthroascus schoenii penetrating
`ovoid cells of Saccharomyces cerevisiae by means of narrow infection pegs (I). B. Schlag-Edler, A. Pupovac-Velikonja and M.-A.
`Lachance. Bars 5 2 mm
`
`Major yeast clades.—Although the basal branches of
`the ascomycete yeast tree are not yet well resolved due
`to lack of sampling of certain genes of some major
`groups, a number of clades however are well
`supported in most analyses. In this section we discuss
`members of the 12 clades and their habitats and
`substrates. It should be noted that because the rate of
`species discovery is so high many of the yeasts are
`poorly known and have been isolated only once or
`twice. The history of Candida tanzawaensis (see
`below) provides an example of this kind of problem.
`This species was isolated from a moss collected in the
`Tanzawa Mountains of Japan 22 y before it was
`described (Nakase et al 1988) and has never been
`recollected. It now is clear however that it is a member
`of a moderate-size clade of species that are common
`in associations with insects (see clade 6, below).
`Saccharomycetales. Analyses of DNA usually provide
`
`strong support for the traditional view of Saccharomy-
`cetales as a monophyletic group (FIG. 1). These
`include trees based on (i) RPB2 with SSU and LSU
`(D1/D2 region) rDNA in a reduced dataset, (ii) SSU
`and LSU (D1/D2 region) rDNA and EF-1a; (iii) SSU
`and LSU (D1/D2 region) rDNA, EF-1a, and RPB2; (iv)
`SSU and LSU (D1/D2 region) rDNA, EF-1a, RPB1 and
`RPB2; and (v) RPB2 only. Support was somewhat lower
`for trees based on single gene datasets (e.g. SSU rDNA
`[79%] only and D1/D2 region of LSU rDNA [87%])
`only.
`Clade 1. For many the quintessential yeast or fumo is
`Saccharomyces cerevisiae (FIGS. 2, 3), a highly special-
`ized, ethanol resistant species that contains an un-
`usually high number of chromosomes. This clade
`contains most of the yeasts with known complete
`genome sequences, and it should be noted that the
`correct name for Kluyveromyces waltii is Lachancea
`
`LCY Biotechnology Holding, Inc.
`Ex. 1046
`Page 7 of 12
`
`

`

`SUH ET AL: SACCHAROMYCETALES PHYLOGENETICS
`
`1013
`
`waltii, a member of a segregate genus from Kluyver-
`omyces. Some members of clade 1, including Saccha-
`romyces cerevisiae, underwent a genome duplication in
`the past (see Genomics Contributions to Phylogenetics,
`below). Of note the natural habitat of these well
`known model organisms remains to be established in
`spite of claims to the contrary (Pennisi 2005).
`Saccharomyces and several related genera once labeled
`‘‘Saccharomyces sensu lato’’ are difficult
`to define
`ecologically. Most species are found only sporadically
`in nature, which might mean that we have yet to
`determine their principal habitats. One exception
`might be species of the basal genus Hanseniaspora,
`a predictable component of naturally fermenting fruit
`and other sugar-rich materials,
`including certain
`nectars and sap fluxes. These species use few carbon
`the b-
`compounds but usually vigorously ferment
`glucoside cellobiose (as well as glucose). A clear
`synapomorphy cannot be identified for the entire
`clade. The clade was supported by bootstrap analysis
`(80%) and posterior probability (100%) in analyses of
`SSU and LSU (D1/D2 region) rDNA (FIG. 1) but not
`when the RPB2 gene sequence was included with SSU
`and LSU rDNA in a reduced dataset. Analyses of other
`reduced datasets including SSU and LSU rDNA and
`EF-1a; SSU and LSU rDNA, EF-1a and RPB2 genes;
`SSU and LSU rDNA, EF-1a, RPB1 and RPB2 genes;
`and RPB2 gene only (analyses not shown) also were
`well supported.
`Clade 2. This clade of ascomycete yeasts contains
`many species endowed with the interesting property of
`methanol assimilation (FIG. 1). These include Candida
`boidinii, Komagatella (Pichia) pastoris, Ogataea (Han-
`senula) polymorpha and related anamorphs. These
`species are intimately associated with the decaying
`wood of trees or the necrotic soft tissues of succulent
`plants and may serve as agents of detoxification for
`invertebrates that colonize these materials. Komaga-
`tella pastoris and O. polymorpha are widely used in
`biotechnology for expression of recombinant proteins.
`Clade 3. This clade (FIG. 1),
`typified by Pichia
`anomala, contains many species that frequently are
`isolated from trees that suffer insect damage. Many of
`the species formerly were assigned to the genus
`Hansenula on the basis of nitrate use. The latter
`characteristic however cannot serve as a reliable
`synapomorphy.
`Clade 4. The Saccharomycopsis clade (FIG. 1) groups
`a wide variety of morphologies and physiologies. Some
`species are purely cellular (Arthroascus spp.) and
`others are nearly exclusively hyphal (Saccharo-
`mycopsis selenospora). Some produce powerful extra-
`cellular hydrolases. All share a deficiency in sulfate
`uptake, which could afford them an accrued resistance
`to toxic ions that share the same transport pathway
`
`(Lachance et al 2000). Possibly related to these
`unusual properties is the widespread ability of the
`species to penetrate and kill other fungi by means of
`infection pegs (FIGS. 14, 15). Taken together these last
`two properties constitute a clear synapomorphy for the
`clade.
`Clade 5. Xylose fermentation is a relatively rare trait
`among the yeasts that have been tested. The Pichia
`stipitis clade (FIG. 1) is of interest because many of
`these taxa have the ability to ferment xylose (Jeffries
`and Kurtzman 1994). Members of the clade have been
`found in wood, often in association with wood-
`ingesting beetles (Nguyen et al 2006, Suh et al
`2003). It should be noted that other yeasts (e.g.
`Pachysolen tannophilus) have similar physiological
`profiles including the ability to ferment xylose
`although they are not members of the P. stipitis clade.
`Clade 6. The relatively obscure Candida tanzawaen-
`sis clade (FIG. 1) has grown from a single described
`species (Nakase et al 1988) to a total of 23 (Kurtzman
`2001, Suh et al 2004). The new taxa were isolated from
`a variety of mycophagous beetles, notably in the family
`Erotylidae, and other insects.
`Clade 7. The notorious human commensal, Candi-
`da albicans combines extracellular lipase activity, the
`ability to form invasive hyphae and the ability to grow
`at 37 C, which may have earmarked this species to be
`the bane of many a human. Other members of the
`clade (FIG. 1) share these properties. This includes C.
`tropicalis isolated from clinical samples, soil, fodder,
`fermentation vats and rotten pineapples, and Lodder-
`omyces elongisporus (FIG. 7) recovered from fingernails,
`baby cream and orange juice. Support for the clade
`was 76% with bootstrap analysis and 100% posterior
`probability in analyses of SSU and LSU (D1/D2
`region) rDNA (FIG. 1). Bootstrap support with re-
`duced datasets (12–27 taxa) including protein-coding
`genes was increased with SSU and LSU (D1/D2
`region) rDNA and RPB2 gene; SSU and LSU rDNA,
`EF-1a and RPB2 genes; SSU and LSU rDNA, EF-1a,
`RPB1 and RPB2 genes. Support for reduced datasets
`including SSU and LSU rDNA, and EF-1a genes; and
`RPB2 gene only were essentially similar to SSU and
`LSU rDNA (not shown).
`Clade 8. The large and diverse clade (FIG. 1)
`containing the genera Metschnikowia and Clavispora
`exhibits a remarkably uniform nutritional profile
`shared with certain less closely related species such as
`Candida sake, Candida oleophila and a few others.
`These species often are found in association with
`herbivorous invertebrates. The carbon compounds
`favored by these yeasts include plant sugars such as
`sucrose, maltose and other a-glucosides and b-gluco-
`sides, as well as sorbose, mannitol, glucitol, and N-
`acetyl-D-glucosamine. Lipolytic activity and the utiliza-
`
`LCY Biotechnology Holding, Inc.
`Ex. 1046
`Page 8 of 12
`
`

`

`1014
`
`MYCOLOGIA
`
`in the metabolic
`tion of alkanes are not unusual
`profile of the clade members. This clade contains
`a growing group of species associated with nitidulid
`beetles (Lachance et al 2000), which form unusually
`large ascospores ornamented with a helical array of
`barbs (FIG. 12). The well supported clade (FIG. 1) also
`obtained strong support with these reduced datasets:
`SSU and LSU (D1/D2 region) rDNA and RPB2 gene;
`SSU and LSU rDNA, EF-1a and RPB2 genes; SSU and
`LSU rDNA and EF-1a gene; and RPB2 gene only (not
`shown). No RPB1 data were available for inclusion.
`Clade 9. The rapidly expanding Starmerella clade
`(FIG. 1) (Rosa et al 2003) consists of highly specialized,
`generally small yeasts that exhibit a clear association
`with bees of all sorts. Smaller somatic cells are
`characteristic of the Wickerhamiella clade with mem-
`bers that are isolated frequently from floricolous
`drosophilids. The remaining taxa shown in clade 9
`represent a coherent assemblage of usually highly
`filamentous species (i.e. Dipodascus and Galactomyces
`spp.), in which unicellular growth is arthric and not
`blastic. The placement of
`these morphologically
`distinct members together within clade 9 probably is
`the result of long-branch attraction.
`Clade 10. Pichia membranifaciens is widely known as
`an agent of spoilage of pickled vegetables. This and
`related species frequently are encountered in a num-
`ber of substrates used by drosophilids as feeding and
`breeding sites. Many species in this clade (FIG. 1) are
`avid film formers and lie at an extreme in the
`spectrum of nutritional specialization, exhibiting a pre-
`dilection for ethanol and simple organic acids as
`carbon sources and having poor fermentative power.
`Of
`interest, Debaryomyces hansenii, although not
`a member of the clade, is similar in that it also may
`be found in pickles and certain strains for