FEMS Yeast Research 4 (2004) 557^561
www.fems-microbiology.org
Reinstatement of Rhodotorula colostri (Castelli) Lodder and
Rhodotorula crocea Shifrine p Pha¡, former synonyms of
Rhodotorula aurantiaca (Saito) Lodder
Ł lvaro Fonseca
Joa‹o Ina¤cio, A
Centro de Recursos Microbiolo¤gicos (CREM), Biotechnology Unit, Faculdade de Cie“ncias e Tecnologia, Universidade Nova de Lisboa, Quinta da Torre,
2829-516 Caparica, Portugal
Received 22 August 2003 ; received in revised form 20 October 2003; accepted 23 October 2003
First published online 19 November 2003
Abstract
Rhodotorula aurantiaca (Saito) Lodder is an anamorphic basidiomycetous yeast species that belongs to the so-called ‘Erythrobasidium
lineage’ of the Urediniomycetes, according to molecular phylogenetic studies based on nucleotide sequence analyses of different ribosomal
DNA regions. In the most recent editions of the yeast taxonomy treatises the species Rhodotorula colostri (Castelli) Lodder and
Rhodotorula crocea Shifrine p Phaff were listed as synonyms of R. aurantiaca. Taxonomic heterogeneity within R. aurantiaca was
demonstrated in a study based on whole-cell protein profiles and is also hinted at by the observed differences in physiological and
biochemical characteristics among the different strains under that species name. We determined partial nucleotide sequences of the 26S
rRNA gene (D1/D2 domains) of strains maintained in the CBS culture collection under R. aurantiaca, including the type strains of its
synonyms. The results showed that R. colostri and R. crocea are clearly distinct from R. aurantiaca and from any other currently
recognised basidiomycetous yeast species. Furthermore, phylogenetic analysis of the sequence data placed the former two species in
separate lineages of the Microbotryomycetidae: R. colostri in the ‘ruineniae clade’ (Sporidiobolus lineage or Sporidiobolales) and R. crocea
loosely linked to Rhodotorula javanica (Microbotryum lineage).
? 2003 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved.
Keywords : Rhodotorula ; Urediniomycetous yeasts; rDNA sequencing ; Molecular systematics
1. Introduction
The type strain of Rhodotorula aurantiaca (Saito) Lodder (CBS 317) was originally isolated from the atmosphere
in Tokyo by Saito who described it in 1922 as Torula
aurantiaca Saito [1,2]. This species was later transferred
to the genera Torulopsis and Chromotorula, and ¢nally
into Rhodotorula by Lodder in 1934 (see also [1,2]). The
type strain of Rhodotorula colostri (Castelli) Lodder (CBS
348) was isolated from human colostrum by Castelli who
described it in 1932 as Mycotorula colostri [1,2]. Lodder
transferred it to Rhodotorula in 1934 and Lodder and
Kreger van Rij later treated it as a strain of R. aurantiaca
[1,2]. Since then it has been maintained as a synonym of
the latter species [3,4]. Rhodotorula crocea was described in
* Corresponding author. Tel.: +351 212948500; Fax: +351 212948530.
E-mail address : [email protected] (A. Fonseca).
1956 by Shifrine and Pha¡ who isolated the type strain
(CBS 2029) from a bark beetle of Pinus je¡reyi in northern
California [2]. Since the second edition of ‘The Yeasts’
R. crocea has been placed in synonymy with R. aurantiaca
[2,3]. Many physiological test responses are reported as
variable for the latter species in the latest editions of the
most comprehensive treatises on yeasts [3,4]. This physiological heterogeneity had already been mentioned by Pha¡
and Ahearn who reported di¡erent results for the assimilation of 5-keto-gluconate: positive for the type cultures
of R. aurantiaca and R. crocea and negative for the type of
R. colostri [2]. Additional di¡erences between the three
species were reported by Vancanneyt and co-workers
based on whole-cell protein pro¢les [5]. These authors
found that the type strains of each species belonged to
di¡erent clusters in a dendrogram resulting from the analysis of the SDS^PAGE pro¢les : the type strain of R. aurantiaca clustered with CBS 8074, a strain isolated from a
leaf of Callistemon viminalis in Australia; the type strain
1567-1356 / 03 / $22.00 ? 2003 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved.
doi:10.1016/S1567-1356(03)00223-X
FEMSYR 1625 18-12-03
J. Ina¤cio, A. Fonseca / FEMS Yeast Research 4 (2004) 557^561
558
of R. colostri clustered with CBS 2201, a strain deposited
in the CBS by Montemartini (a possible subculture of
Castelli’s original isolate [2]); and the type strain of
R. crocea clustered with CBS 5950, a strain isolated
from Bantu beer in South Africa. The same authors presented data on the major ubiquinone (CoQ) and the molar
content of G+C of genomic DNA for the same strains,
which supported the observed clustering based on the
whole-cell protein pro¢les [5]. More recently, studies based
on the analysis of nucleotide sequences of di¡erent rDNA
regions [6^8] demonstrated the phylogenetic a⁄liation of
R. aurantiaca to the so-called ‘Erythrobasidium lineage’ of
the Urediniomycetes, a well-supported clade which contains other deeply pigmented species in the genera Rhodotorula and Sporobolomyces as well as a few teleomorphic
taxa, e.g. Occultifur externus and Sakaguchia dacryoides
[8]. However, the type strains of R. colostri and R. crocea
were not included in those studies and their phylogenetic
position remains uncertain. Here we report the determination of the nucleotide sequences of the D1/D2 domains of
the 26S rRNA gene of the above-mentioned strains
studied by Vancanneyt et al. [5] in order to re-assess the
taxonomic status of the R. aurantiaca synonyms.
2. Materials and methods
prior to sequencing employed forward primer ITS5 (5PGGA AGT AAA AGT CGT AAC AAG G) and reverse
primer LR6 (5P-CGC CAG TTC TGC TTA CC) using a
Uno II Thermal Cycler (Biometra, Go«ttingen, Germany)
and the resulting amplicon was puri¢ed with the GFX
Band Puri¢cation Kit (Amersham Biosciences Piscataway,
NJ, USA). Cycle sequencing of the D1/D2 variable domains of the 26S rDNA employed forward primer NL1
(5P-GCA TAT CAA TAA GCG GAG GAA AAG) and
reverse primer NL4 (5P-GGT CCG TGT TTC AAG ACG
G), following standard protocols. Sequences were obtained with an ALFexpress II DNA Analyser (Amersham
Pharmacia Biotech, Uppsala, Sweden), aligned with MegAlign (DNAStar software package) and visually corrected.
Phylogenetic trees were computed with PAUP version
4.0b8 (Sinauer Associates, Inc., Sunderland, MA, USA)
using the neighbour-joining method and the Kimura
two-parameter model for calculating distances, or the
maximum-parsimony analysis (full heuristic search with
the following options: random stepwise addition with 10
replications, branch swapping using tree bisection-reconnection and maximum of 100 trees). Gaps were treated as
missing data. Nucleotide sequences were deposited in GenBank under the accession numbers listed in Table 1. Additional sequences were retrieved from GenBank (accession
numbers are indicated on the phylogenetic tree).
2.1. Cultures
3. Results and discussion
A list of the strains used in this study is presented in
Table 1. Collection strains were obtained from the ‘Centraalbureau voor Schimmelcultures’, Utrecht, The Netherlands (CBS). Strains were grown on plates containing malt
extract 0.7%, yeast extract 0.05%, Soytone 0.25%, agar
1.5% (MYP) at 20‡C and maintained on MYP agar slants
at 4‡C. Physiological properties of the strains in Table 1
were retrieved from the CBS Yeast Database (http://
www.cbs.knaw.nl/yeast/webc.asp).
2.2. Molecular methods
Genomic DNA for rDNA sequencing was isolated from
1-week-old cultures on MYP agar plates by a simpli¢ed
method using glass beads for cell disruption following the
protocol used by Sampaio et al. [9]. PCR ampli¢cation
Nucleotide sequences of the D1/D2 domains of the 26S
rDNA determined for the strains in Table 1 were compared to those available for the same region at GenBank
using BLAST. Strains CBS 317 and CBS 8074 had identical sequences that di¡ered from the sequence available at
GenBank for strain CBS 317 (AF189921) at three nucleotide positions (one substitution and two indels). These
di¡erences were found to correspond to edition mistakes
in the sequence already deposited in GenBank, upon rechecking the original sequencing gel images by the authors
of [7] (G. Scorzetti, pers. comm.; a corrected version
for AF189921 is now available at GenBank). Strains CBS
348 and CBS 2201 had identical sequences and matched
two sequences deposited at GenBank (AF406922 and
AF406923) corresponding to strains CBS 9002 and CBS
Table 1
List of cultures used in this study
Species
Strain
R. aurantiaca
CBS
CBS
CBS
CBS
CBS
CBS
R. colostri
R. crocea
a
317T
8074
348T
2201
2029T
5950
Origin
D1/D2 seq.a
Air (Japan)
Leaf of Callistemon viminalis (Australia)
Colostrum of woman (Italy?)
Unknown
Bark beetle (USA)
Bantu beer (S. Africa)
AY372175
AY372176
AY372177
AY372178
AY372179
AY372180
GenBank (NCBI) accession numbers.
FEMSYR 1625 18-12-03
J. Ina¤cio, A. Fonseca / FEMS Yeast Research 4 (2004) 557^561
Table 2
Salient physiological and biochemical di¡erences between R. aurantiaca,
R. colostri and R. croceaa
Property
R. aurantiaca
R. colostri
R. crocea
D-galactose
3
3
+
3
3
3
10
55.1-55.4
+
3
3
3
+
3
9
57.6^57.9
+
+
+
+
3
+
10c
58.4^59.3
D-ribose
D-glucuronate
D-galacturonate
Ethylamine (N)
Growth at 30‡C
CoQb
Mol % G+Cb
a
Based on the available data for the strains in Table 1 ; physiological
data from the CBS Yeast Database (http://www.cbs.knaw.nl/yeast/webc.
asp): assimilation of carbon or nitrogen (N) compounds.
b
Data from [5] ;
c
Data available for CBS 2029 only.
559
9028 labelled as Rhodosporidium a¡. lusitaniae. The latter
were isolated from £oral nectaries in Germany [10]. These
results suggest that strains CBS 2201, CBS 9002 and CBS
9028 all represent the species R. colostri as typi¢ed by
strain CBS 348. Strains CBS 2029 and CBS 5950 di¡ered
at two nucleotide positions and the closest matches in
GenBank were sequences from a few Rhodotorula spp.
having more than 30 bp di¡erences. The two strains did
not di¡er signi¢cantly in physiological tests (mismatching
responses were only found for the assimilation of D-arabinose, arbutin and L-lysine; data not shown) and it is therefore very likely that strains CBS 2029 and CBS 5950 are
conspeci¢c and represent the species R. crocea. Full con¢rmation of this hypothesis would bene¢t from DNA/
DNA reassociation experiments and/or additional se-
Fig. 1. Phylogenetic tree of the strains of R. aurantiaca and former synonyms listed in Table 1, and of selected urediniomycetous yeasts, obtained by
neighbour-joining analysis of 26S rRNA gene (D1/D2 domains) sequences using PAUP 4.0b8. The numbers given on the branches are the frequencies
( s 50%) with which a given branch appeared in 1000 bootstrap replications. Kondoa aeria, K. malvinella and Bensingtonia subrosea that belong to the
Agaricostilbum lineage were used as outgroup. Sequences determined by the authors of the present study are typed in boldface (GenBank accession
numbers in Table 1). Additional sequences were retrieved from GenBank (accession numbers between parentheses).
FEMSYR 1625 18-12-03
560
J. Ina¤cio, A. Fonseca / FEMS Yeast Research 4 (2004) 557^561
quence data (e.g. ITS region). Nucleotide di¡erences in the
D1/D2 region between the type strain of R. aurantiaca and
the types of R. colostri and R. crocea amounted to more
than 100 bp substitutions (more than 25% sequence divergence), thus suggesting the reinstatement of the two former synonyms of R. aurantiaca as separate species.
The conclusions ensuing from the sequencing results
obtained by us are fully supported by the data of Vancanneyt et al. [5]. Salient di¡erences between R. aurantiaca
and its former synonyms are summarised in Table 2.
Another striking phenotypic di¡erence between R. aurantiaca, R. colostri and R. crocea is the colour of growth on
solid media: R. aurantiaca strains are orange to orangered, those of R. colostri are pink to light red and those of
R. crocea are cream to yellow (data not shown; a colour
photograph of growth on MYP agar can be viewed
at http://www.crem.fct.unl.pt/dimorphic_basidiomycetes/
Databases/additionaldata.htm/).
A phylogenetic analysis of the sequences determined in
this study was conducted in order to assess the correct
a⁄liation of the former R. aurantiaca synonyms. Sequences of selected species belonging to di¡erent lineages of the
urediniomycetous yeasts retrieved from GenBank were included for comparison purposes. Tree topologies from
neighbour-joining and maximum-parsimony analyses
were similar and only the former is shown (Fig. 1). It is
clear that R. colostri and R. crocea are not only distinct
from R. aurantiaca at the species level but they also belong
to very distantly related phylogenetic clades. In fact
R. aurantiaca belongs to the Erythrobasidium lineage of
the Urediniomycetes, whereas both R. colostri and R. crocea fall within the subclass Microbotryomycetidae (sensu
Swann et al. [11]), which includes urediniomycetous yeasts
in the Sporidiobolus and Microbotryum lineages (sensu
Scorzetti et al. [8]) (Fig. 1). The polyphyletic nature of
the genus Rhodotorula is now well established based on
di¡erent molecular phylogenetic studies (e.g. [6,7]). However, major taxonomic changes have not yet been put forward since some yeast taxonomists have favoured nomenclatural stability over premature name changes for
anamorphic taxa. Such changes are outside the scope of
the present study but proposals of new genera have appeared recently in the literature (e.g. [12]).
R. colostri belonged to the ‘ruineniae clade’ of the Sporidiobolus lineage (this lineage has been recently given a
formal taxonomic designation: order Sporidiobolales
[12]) (Fig. 1). The closest relative of R. colostri was Rhodosporidium lusitaniae but the two taxa di¡ered at 11 nucleotide positions thus suggesting that they are separate
species. The two species also di¡ered in physiological
test responses, namely the assimilation of sucrose (positive
or delayed for R. colostri and negative for R. lusitaniae) or
galactitol (positive for R. lusitaniae and negative for R. colostri) and growth without vitamins (positive for R. lusitaniae and negative for R. colostri) [3,4]. It’s worth noting
that both species share CoQ9 as the major ubiquinone,
although CoQ10 is found in the majority of species in
the Sporidiobolales (e.g. [3]). R. colostri shares with other
members of the Sporidiobolales the presence of deeply
pigmented (orange, pink or red) colonies and the inability
to utilise D-glucuronate as sole carbon source [12]. In turn,
R. crocea falls within the Microbotryum lineage on a basal
branch loosely linked to R. javanica (Fig. 1). It shares with
other members of that lineage the absence of deeply pigmented colonies and the ability to assimilate D-glucuronate.
The molecular data obtained in the present study as well
as other relevant data mentioned above for R. aurantiaca
and its synonyms, lead us to propose the reinstatement of
R. colostri (Castelli) Lodder and R. crocea Shifrine and
Pha¡ as separate species.
3.1. Summary of taxonomic changes :
b
b
b
b
b
New list of synonyms of R. aurantiaca (Saito) Lodder :
Torula aurantiaca Saito
Torulopsis aurantiaca (Saito) Ciferri p Redaelli
Chromotorula aurantiaca (Saito) F.C. Harrison
Reinstatement of species:
R. colostri (Castelli) Lodder; synonym: Mycotorula colostri Castelli; strains: CBS 348 (type), CBS 2201, CBS
9002 and CBS 9028.
R. crocea Shifrine p Pha¡; strains : CBS 2029 (type)
and CBS 5950.
Acknowledgements
The authors wish to thank Dr. V. Robert (CBS, The
Netherlands) for providing the cultures used in this study.
J.I. receives a PhD grant (Praxis XXI/BD/19833/99) from
‘Fundac\a‹o para a Cie“ncia e Tecnologia’ (Portugal).
References
[1] Lodder, J. and Kreger-van Rij, N.J.W. (1952) The Yeasts, A Taxonomic Study. North-Holland Publ. Co., Amsterdam.
[2] Lodder, J. (1970) The Yeasts, A Taxonomic Study, 2nd edn. NorthHolland Publ. Co., Amsterdam.
[3] Kurtzman, C.P. and Fell, J.W. (1998) The Yeasts, A Taxonomic
Study, 4th edn. Elsevier Sci. Publ., Amsterdam.
[4] Barnett, J.A., Payne, R.W. and Yarrow, D. (2000) Yeasts: Characteristics and Identi¢cation. 3rd edn. Cambridge University Press,
Cambridge.
[5] Vancanneyt, M., Coopman, R., Tytgat, R., Berny, J.F., Hennebert,
G.L. and Kersters, K. (1992) A taxonomic study of the basidiomycetous yeast genera Rhodosporidium Banno and Rhodotorula Harrison
based on whole-cell protein patterns, DNA base compositions and
coenzyme Q types. J. Gen. Appl. Microbiol. 38, 363^377.
[6] Takashima, M., Hamamoto, M. and Nakase, T. (2000) Taxonomic
signi¢cance of fucose in the class Urediniomycetes : distribution of
fucose in cell wall and phylogeny of urediniomycetous yeasts. Syst.
Appl. Microbiol. 23, 63^70.
[7] Fell, J.W., Boekhout, T., Fonseca, A., Scorzetti, G. and Statzell-Tall-
FEMSYR 1625 18-12-03
J. Ina¤cio, A. Fonseca / FEMS Yeast Research 4 (2004) 557^561
man, A. (2000) Biodiversity and systematics of basidiomycetous
yeasts as determined by large subunit rDNA D1/D2 domain sequence
analysis. Int. J. Syst. Evol. Microbiol. 50, 1351^1371.
[8] Scorzetti, G., Fell, J.W., Fonseca, A. and Statzell-Tallman, A. (2002)
Systematics of basidiomycetous yeasts: a comparison of large subunit D1/D2 and internal transcribed spacer rDNA regions. FEMS
Yeast Res. 2, 495^517.
[9] Sampaio, J.P., Gadanho, M., Santos, S., Duarte, F.L., Pais, C.,
Ł . and Fell, J.W. (2001) Polyphasic taxonomy of the basiFonseca, A
diomycetous yeast genus Rhodosporidium: R. kratochvilovae and related anamorphic species. Int. J. Syst. Evolut. Microbiol. 51, 687^
697.
561
[10] Herzberg, M., Fischer, R. and Titze, A. (2002) Con£icting results
obtained by RAPD-PCR and large-subunit rDNA sequences in determining and comparing yeast strains isolated from £owers: a comparison of two methods. Int. J. Syst. Evol. Microbiol. 52, 1423^
1433.
[11] Swann, E.C., Frieders, E.M. and McLaughlin, D.J. (1999) Microbotryum, Kriegeria and the changing paradigm in basidiomycete classi¢cation. Mycologia 91, 51^66.
[12] Sampaio, J.P., Gadanho, M., Bauer, R. and WeiM, M. (2003) Taxonomic studies in the Microbotryomycetidae: Leucosporidium golubevii
sp. nov., Leucosporidiella gen. nov. and the new orders Leucosporidiales and Sporidiobolales. Mycol. Prog. 2, 53^68.
FEMSYR 1625 18-12-03