Loss of Meiosis in Aspergillus
David A4. Geiser,” William E. Timberlake,?
*Department
of Genetics,
University
of Georgia;
and Michael L. Arnold*
and tMyco
Pharmaceuticals,
Inc., Cambridge,
Massachusetts
If strictly mitotic asexual fungi lack recombination,
the conventional view predicts that they are recent derivatives
from older meiotic lineages. We tested this by inferring phylogenetic relationships among closely related meiotic
and strictly mitotic taxa with Aspergillus conidial (mitotic) states. Phylogenies were constructed by using DNA
sequences from the mitochondrial small ribosomal subunit, the nuclear ribosomal internal transcribed spacers, and
the nuclear 5.8s ribosomal gene. Over 920 bp of sequence was analyzed for each taxon. Phylogenetic analysis of
both the mitochondrial
and nuclear data sets showed at least four clades that possess both meiotic and strictly
mitotic taxa. These results support the hypothesis that strictly mitotic lineages arise frequently from more ancient
meiotic lineages with Aspergilhs conidial states. Many of the strictly mitotic species examined retained characters
that may be vestiges of a meiotic state, including the production of sclerotia, sclerotium-like
structures, and htille
cells.
Introduction
Asexual
lineages
are believed to be susceptible to
the progressive accumulation
of deleterious mutations in
the absence of recombination
(Muller 1964; Lynch et al.
1993). This view leads to the prediction that asexual
species are recently derived from sexual lineages, because asexual species should be more susceptible to extinction (Maynard Smith 1978). However, possibly ancient lineages have been identified based on inferences
from mitochondrial
DNA sequences in unisexual vertebrates (Hedges, Bogart, and Maxson 1992; Quattro,
Avise, and Vrijenhoek 1992; Spolsky, Phillips, and Uzzell 1992). Other organisms not known to undergo sex,
such as Archaebacteria
and Euglenas,
may represent
very ancient asexual lineages as well. Fungi are uniquely suited for the study of the evolution of sexual and
asexual modes of propagation within a lineage, because
many fungal species reproduce strictly mitotically (asexually), whereas their close relatives are capable of meiotic (sexual) reproduction.
Establishing
evolutionary
patterns between meiotic and strictly mitotic species
should provide evidence as to the evolutionary
fate of
strictly mitotic lineages.
The genus Aspergillus
is a diverse and familiar
group of ascomycetes which is mostly saprobic, but includes human pathogens, plant pathogens, and species
useful in industrial processes and genetic research. Of
the 186 species with Aspergillus asexual states listed by
Samson (1994), 114 are not known to produce a sexual
meiotic spore (ascospore). These include some of the
species most frequently encountered
by humans: A. j&zvus, A. niger, A. terreus, and A. fumigatus. These species
propagate solely via asexual mitotic spores (conidia).
The remaining
72 species do produce meiotic spores,
and are placed in nine sexual genera, the most common
being Emericella, Eurotium, and Neosartorya.
All but
two meiotic species are known to produce conidia as
well as ascospores (fig. 1).
Key words: Aspergillus,
meiosis,
fungi, asexuality.
Address for correspondence
and reprints: David M. Geiser, Department of Plant Biology, 111Koshland Hall, University of California, Berkeley, California
94720. E-mail: [email protected].
edu.
Mol. Biol. Evol. I3(6): 809-S 17. 1996
0 1996 by the Society for Molecular Biology and Evolution. ISSN: 0737.4038
The roles of clonal and recombinant
modes of
propagation in nature are unknown. Evidence for sexual
recombination
has been identified in the meiotic species
Emericella
niduluns (Geiser, Arnold, and Timberlake
1994), while strictly mitotic species such as A. jkzvus
are believed to propagate in mostly or completely clonal
fashion (Cotty et al. 1994). The production of recombinant mitotic spores is possible in the laboratory (the
parasexual cycle; Pontecorvo et al. 1953), but it is not
known to occur in natural populations. No studies have
been attempted to quantify the level of recombination
that occurs in meiotic or strictly mitotic species.
If strictly mitotic lineages are clonal, the conventional view (Maynard Smith 1992) predicts that they
should be short-lived and recently derived from older
meiotic lineages. Asexuality may provide a short-term
benefit to fungi, with a long-term cost evident in such
a phylogenetic
pattern (Berbee and Taylor 1993~). Phylogenetic analysis of Penicillium, a genus closely related
to Aspergillus, showed close relationships between sexual and asexual taxa, suggesting
recent derivation of
asexual lineages (LoBuglio, Pitt, and Taylor 1993). We
used mitochondrial
and nuclear ribosomal
DNA sequences to infer phylogenetic
relationships
among meiotic and strictly mitotic taxa to determine whether strictly mitotic Aspergilli represent multiple independent derivatives of meiotic lineages.
Materials and Methods
We examined seven strictly mitotic and seven meiotic Aspergillus
species (table 1). Talaromyces jhvus
was chosen as an outgroup species based on the evidence that Aspergillus is part of a monophyletic
group
derived from it (Berbee et al. 1995). Tularomyces is sexual, and it was previously shown that many Penicillium
asexual states are independently
derived from Talaromyces ancestors (LoBuglio, Pitt, and Taylor 1993). The
14 species were chosen to represent a diverse array of
species with Aspergillus
asexual states, representing
5
subgenera (Samson 1994) and 13 sections or species
groups (Samson 1994; Raper and Fennel1 1965). The
internal transcribed spacers (ITS1 and ITS2) + the 5.8s
ribosomal DNA (nuclear rDNA), and part of the mitochondrial ribosomal
small subunit (mtrssu), were se809
810
Geiser et al.
fess resistant?
f
(- 18 to 100 hours)
-
myEum
ascospore
vwre resistant?
t
cleistothecium
FE. 1.-Generalized
life cycle for fungi with Aspergillus
conidial states. Both asexual mitotic spores (conidia) and sexual meiotic
spores (ascospores)
can germinate
and grow to form a mycelium.
In the 70 named species that are known to produce ascospores,
both
conidium-bearing
structures (conidiophores)
and ascospore-bearing
structures (cleistothecia)
are produced from the mycelium. One hundred fourteen species are not known to produce ascospores,
and are believed to reproduce strictly mitotically.
The life cycle is probably
primarily
haploid, with only a short diploid state prior to meiosis. Generally,
the production
of mature ascospores
takes weeks, while
mature conidia can be produced
in a few days. Ascospores
are generally
more resistant to desiccation
than conidia, but conidia are
generally produced in greater abundance.
.
quenced in each species. In keeping with the current
taxonomic schemes (Klich and Pitt 1988; Samson 1994),
meiotic species with Aspergillus asexual states will be
referred to by their sexual names in this paper (Emericella, etc.), while strictly mitotic species will be referred
names (table 1). Both meiotic
to by their Aspergillus
and strictly mitotic species as a whole will be referred
to as “Aspergillus”
or “Aspergilli.”
The nuclear rDNA
sequence from Neosartorya jischeri and most of the
mtrssu sequence from Tularomyces jCzvus, which was
used as an outgroup, were kindly provided from the laboratory of Dr. John Taylor. The mtrssu sequence from
E. nidulans has been published
(Kuntzel and Hochel
1981).
DNA Extraction
and Polymerase
Chain Reaction
Mycelium
was grown by inoculating
dry spores
and mycelium
into 25 ml of liquid minimal medium
(MM; Barratt, Johnson, and Okata 1965) in petri plates.
The cultures were incubated at either 30°C or 37°C for
1 day to 1 week, until a floating mass of mycelium was
present. Mycelium was harvested, followed by a DNA
extraction
protocol adapted from Yelton, Hamer, and
Timberlake
(1984) and suspension in 100 p.1 of water.
Although
10X or 100X dilutions were used in some
cases, 1 ~1 of genomic template was used as template.
Reactions were performed in 50-p,l volumes, using 5
units of Tuq DNA polymerase,
15-30 pmol of each
primer, 0.2 mM dATP, dCTP, dGTP, and TTP, and 2 mM
MgCl,, according to the enzyme manufacturer’s
instructions (Promega).
For the nuclear rDNA, primers designed for use in higher plants were used (Hodges and
Arnold 1994). These primers differ slightly from those
designed specifically for fungi (White et al. 1990). We
designed new primers that amplify an approximately
715-bp region of the Aspergillus mtrssu: AspMS5 (5’-
AAGAAAGGAAAAAGAAGGTC-3’)
or AspMS3 (5’CGGTGAGTAGTAGTAAAGGT-3’)
and AspMS4 (5’AAATAACATACTTCACTACT-3’)
were used as external primers, and AspMS6 (5’-TTCTTCCTTCTTACATAAA-3’)
and AspMS7 (5’-TTTATGTGAGAAGGAAGAA-3’)
were used as internal primers. The primers
were designed from regions of similarity identified between the aligned mtrssu sequences from E. rzidulans
and Saccharomyces
cerevisiae, with standardized
conditions for automated sequencing taken into consideration.
DNA Sequencing
and Analysis
PCR products were purified on Wizard PCR Prep
columns (Promega) and both strands were sequenced
directly using an Applied Biosystems
automated
sequencer (Model 373A) at the University
of Georgia
Molecular
Genetics
Instrumentation
Facility.
Sequences were edited using SeqEd (v. 1.0.3, Applied
Biosystems),
and initially
aligned using the PILEUP
option in GCG (Wisconsin
Package, version 8). Only
sequences
that could be read from two strands were
kept for analysis.
Alignments
were then optimized
manually. Phylogenetic
analysis was performed using
PAUP (Phylogenetic
Analysis Using Parsimony)
version 3.1.1 (Swofford
1993), with general heuristic
searches and bootstrap analyses using 500 replicates.
Random
addition
and tree-bisection-reconnection
(TBR) branch-swapping
were used. Indels of two or
more nucleotides
were encoded
according
to their
presence or absence (O/I) if the inserted sequence was
conserved.
Gaps were treated as missing data. Additional branch support analysis was performed by using
the Bremer support index, saving all trees of increasing length in a step-wise manner, increasing
one step
at a time (Bremer 1994). For example, for the nuclear
Table 1
Isolates Used in This Study
Asexual
A. compafihilis
Cams
State”
Strain
Number
Sexual State”
Samson
Emericella heterothallica
(Kwon et al.) Raper and
Fennel1
222
A. ustus (Bain.) Thorn &
Church
None
H611
A. nidulnns
Gams
Emericella
Vuill.
Samson
A. versicolor
Tiraboschi
and
and
(Vuill.)
nidulans
None
A. &wipes (Bain and Sart.)
Thorn and Church.
Fennelliaflavipes
Simmons
A. terreus Thorn
None
A. niger van Tieghem
A. jlavus
Eurotium
gin
Smith
None
Samson
and
A. wentii Wehmer..
A. warcupii Samson
Gams
and
A.spergi1lu.s ,fischrriunus
Samson and Cams
Penicillium
Pitt
dangeardi
Subgenus Nidulantes,
section Usti
L76744
L76869
FGSC4
Fungal Genetics
Stock Center
nidulans
Subgenus Nidulantes,
section Nidulantes
L76746
JO1393
226
R. B. G. Dales
Aspergillus
versicolor
Subgenus Nidulantes,
section Versicolores
L76745
L769 14
S. Peterson
group
Aspergillus
flavipes
Subgenus Nidulantes,
section Flavipedes
L76775
L76918
D. Geiser
grow
Aspergillus
terreus
Subgenus Nidulantes,
section Terrei
L76774
L76917
R. T. Hanlin
group
Aspergillus
niger
Subgenus Circumduti,
section Nigri
L76748
L76916
R. T Hanlin
group
Aspergillus
jluvus
Subgenus Circumdati,
section Flavi
L76747
L76915
R. T. Hanlin
group
Aspergillus
glaucus
Subgenus Aspergillus,
section Aspergillus
L76777
L76920
R. T. Hanlin
group
Aspergillus
restrictus
Subgenus Aspergillus,
section Restricti
L76776
L76919
R. T Hanlin
group
Aspergillus
cremeus
Subgenus Cricumdati,
section Cremei
L76805
L76922
NRRL5.502
H38
H267
None
H631
NeosartoryaJischeri
(Wehmer) Malloch and Cain
Talaromyces fravus (var. macrosporus) Stolk and Samson
L76868
ustus
H701
(War-
L76743
group
Aspergillus
group
Chaetosartorya chrysella
(Kwon and Fennel]) Subramanian
Warcupiella spinulosa
cup) Sburamanian
Subgenus Nidulantes,
section Nidulantes
Aspergillus
H142
Man-
A.spergil1u.s nidulans
Nuclear rDNA
mtrssu
GenBank
GenBank
Accession No. Accession No.
Classification
(Samson 1994)
R. T. Hanlin
HI25
amstelodami
Source
R. B. G. Dales
group
GAO2
None
A. vitis Novobranova
A. chryseidis
Gams
Wiley and
None
Link
A. restrictus
(Eidam)
Classification
(Raper and
Fennel1 1965)
group
NRRL4376
NRRL 4161
(mtrssu),
FRR 1833
(m-DNA)”
FRR 2386’
R. T. Hanlin
Aspergillus
wentii
Subgenus
section
Circumdati,
Wentii
L76806
L7692 1
S. Peterson
group
Aspergillus
ornatus
Subgenus
Ornati
Omati,
L76807
L76923
R. T Hanlin
group
Aspergillus fumigatus
U18355
L76924
grow
-
section
Subgenus Fumigati,
tion Fumigati
sec-
5
:
%
Ul8354
L14508,
L76925
‘The names used in thi\ papw are those in boldface; “vxual states” (those referring to the m&tic state) were used in m&tic ape&s, and “asexual” AspergiIlus states were used as names for strictly mitotic species.
h The nuclear ITS + 5.8s rDNA sequence from N. /i~,hen’ isolate FRR IX13 (CSIRO culture collection, North Ryde, NSW, Australia) WBF provided by Mary Berbee and John Taylor.
‘ The nuclear ITS + 5.8s rDNA and part of the mitochondrial sequence from the outgroup species Tularom~wt, ,jaru.s isolate FRR 2386 were kindly provided by Katherine LoBuglio, Mary Berbee, and John Taylor.
,”
_.
g.
v
6’
e
B
2
=;
?
h
812
Geiser et al
data, the most parsimonious
tree had 139 steps. General heuristic searches were performed saving all trees
of length 5140 steps, then all trees of 5141 steps, up
to six steps longer than the most parsimonious
tree.
A strict consensus
tree was identified
with each increase, and the number of increased
steps necessary
to collapse each branch on the most parsimonious
tree
was determined.
When the number of trees saved at
each increment
exceeded
computer
memory
limits,
the consensi of three sets of 8,000 trees generated by
random
sequence
addition
was determined.
The
“Trace
Character”
option in MacClade
(v. 3.0.1;
Maddison
and Maddison
1992) was used to superimpose the evolution
of sexual and asexual taxa on the
parsimony
trees. The tracings were performed
using
method,
which maximizes
both the “ACCTRAN”
early gains and forces subsequent
reversals,
and the
“DELTRAN”
method, which delays changes away
from the ancestors. The “constraints”
option in PAUP
was used to constrain tree topologies,
forcing asexual
taxa to be monophyletic,
and heuristic searches were
performed
as described above to find the most parsimonious
constrained
trees. Maximum-likelihood
estimates
for the most-parsimonious
and constrained
most-parsimonious
trees were calculated
using the
DNAML program in the PHYLIP package (v. 3.57;
Felsenstein
1993). To test whether constrained
trees
were significantly
less likely than unconstrained
trees,
the Kishino-Hasegawa
test was performed
using
DNAML
(Kishino
and Hasegawa
1989). Neighborjoining analyses were performed by using the DNADIST and NEIGHBOR
programs,
and the SEQBOOT
program for bootstrapping
(Felsenstein
1993). The Jukes-Cantor
(Jukes and Cantor 1969) and Kimura twoparameter
(Kimura
1980) estimates
of genetic distance were employed, both of which gave nearly idenratio
tical branching
order. A transition : transversion
of 2.0 was used in estimating
the Kimura two-parameter distance for both data sets.
Results
In each species, a minimum of 460 bp of doublestranded sequence was derived from both strands of the
nuclear rDNA region, and a minimum of 565 bp of sequence was derived from the mtrssu gene. The ITS
regions of the nuclear rDNA were highly variable
among Aspergillus species, so we removed four sections
totalling -90 bp of ITS 1 and - 12 bp of ITS2 because
alignment was problematic
due to insertions and deletions. This left a total of 354 nucleotide positions for
analysis. Two small indels were included as data, encoded as single characters. Forty informative sites were
found in the 354-bp region. The mtrssu provided another
50 phylogenetically
informative characters. A 12-bp inde1 was found in the mtrssu of A. wentii and C. chrysella, the presence of which was coded as a single character.
Heuristic searches yielded single most-parsimonious trees for the mitochondrial
(140 steps) and nuclear
data sets (139 steps) (fig. 2). The topologies of the most
parsimonious
trees from the two data sets differed
slightly, but mostly in branches that had poor (<50%)
bootstrap support and Bremer support indices of 1 or 2.
If branches with less than 70% bootstrap support are
collapsed into polytomies, trees for the two data sets are
identical with one exception. The nuclear data place A.
versicolor and Emericella nidulans together as a strongly supported clade, while the mitochondrial
data place
Emericella nidulans basal to A. versicolor. This inconsistency may reflect a real difference in the genealogies
of these two genes, as might be expected in closely related taxa (Avise and Ball 1990). Except for this inconsistency, all of the inferred trees showed three strongly
supported groupings of strictly mitotic and meiotic taxa:
A. ustus and Emericella heterothallica, A. restrictus and
Eurotium amstelodami, and A. wentii and Chaetosartorya chrysella. Since two independent
losses (or gains)
of meiosis can be inferred in the A. ustuslE. heterothallicalil. versicolor/E.
nidulans clade, at least four independent losses (or gains) of meiosis can be inferred
for the entire group.
In contrast to the parsimony
analysis, neighborjoining of both the mitochondrial
and nuclear data produced two clades grouping Emericellu heterothallicalil.
ustus and Emericella nidulan.slA. versicolor, although
the latter clade was not strongly supported by bootstrap
analysis of the mitochondrial
data (65%). With this exception, differences between the topologies of the parsimony and neighbor-joining
trees were on branches
with poor (<50%) bootstrap support on the parsimony
trees. The Jukes-Cantor
and Kimura two-parameter
distance estimates both gave trees with identical branching
order for both data sets, except for the branching order
on the A. jlavuslA. terreuslA. nigerlFennellia
juvipes
clade using the nuclear data. All of the differing branches had very weak (~32%) bootstrap support. Both data
sets show multiple independent
losses or gains of meiosis.
When the seven asexual taxa were constrained
to
form a monophyletic
ingroup, heuristic searches using
the mitochondrial
data set produced nine most-parsimonious trees of 191 steps, 51 steps longer than the
unconstrained
most-parsimonious
tree. Heuristic searches using the nuclear data set produced two trees of 191
steps, 52 steps longer than the unconstrained
most-parsimonious tree. The log likelihood scores of the unconstrained most-parsimonious
trees and the constrained
most-parsimonious
trees with the highest log likelihood
are given in table 2. Both constrained most-parsimonious trees are significantly less likely than the unconstrained trees: the nuclear constrained tree is more than seven standard deviations less likely, and the mitochondrial
constrained
tree is more than five standard deviations
less likely. Differences of more than 1.96 standard deviations are considered significant (Kishino and Hasegawa 1989).
Discussion
The inferred
phylogenies
confirm conclusions
based on morphology (Raper and Fennel1 1965; Samson
Loss of Meiosis in Aspergillus
Emericella
813
Emericella
a =.734
a=.
RI=.
RC=
Emericella nidulans
mericella nidulans
.A. restrictus
.J
Neosartorya fischeri
34
-A.
Fennellia flavipes
niger
Talaromyces flavus
Talaromyces flavus
A
heterothallica
Emericella
heterothallica
Emericella nidulans
mericella nidulans
urotium amstelodami
Fennellia flavipes
4
urotium amstelodami
Neosartorya fischeri
Talaromyces flavus
C
D
FIG. 2.-Phylogenetic
analysis of mitochondrial
and nuclear data sets. Dark lines represent inferred asexual lineages, light lines represent
sexual lineages, and striped lines represent lineages for which ACCTRAN and DELTRAN methods produced different inferences. Numbers
below branches represent bootstrap values (500 replicates), and numbers above branches are Bremer support indices. Tree branches are drawn
proportional to inferred branch lengths. Consistency indices (CI), retention indices (RI), and resealed consistency indices (RC) are given for the
parsimony trees (Farris 1989). Neighbor-joining
trees shown are based on the Jukes-Cantor measure of genetic distance (Jukes and Cantor 1969).
(A) Parsimony analysis of the mitochondrial
ribosomal small subunit (mtrssu). (B) Parsimony analysis of the nuclear ITS + 5.8s ribosomal
region. (C) Neighbor-joining
analysis of the mitochondrial
data. (D) Neighbor-joining
analysis of the nuclear data.
1994): sexual and asexual taxa are often closely related.
The data show that meiosis has been independently
lost
and/or gained at least four times in Aspergillus among
the sampled taxa. The nuclear and mitochondrial
data
sets both strongly group two meioticktrictly
mitotic species pairs, A. restrictus/Eurotium
amstelodami
and A.
wentii/Chaetosartorya
chrysella, indicating
that these
asexual Aspergillus species are derived from meiotic lineages (fig..2). Parsimony analyses of the mitochondrial
and nuclear data sets do not give a consensus as to the
origins of the asexual species A. versicolor and A. ustus.
The DELTRAN
method indicated
two independent
losses of meiosis leading to these taxa, while the ACCTRAN method indicated one loss and one gain from the
mitochondrial
data, and two gains of meiosis from the
nuclear data. Neighbor-joining
analyses of both data sets
lead to an unambiguous
inference of two independent
losses.
814
Geiser et al
Model for the loss of meiosis in Aspergillus
Meiotic (sexual) modes
Strictly mitotic (asexual) modes
irregularly shaped hiille cell masses
Aspergillus ustus
scattered hiille cells
Aspergillus we&color
Homothallism
Emericella, Chaetosartoya,
Eumtium,
no apparent vestiges
Aspeq$lus restrictus
hyphal masses resembling
immature cleistothecia
Aspergillus we&ii
FIG. 3.-Model
for the loss of meiosis in Aspergillus. Homothallism
is shown as the ancestral state, which frequently produces various
strictly mitotic derivatives. These include taxa that lack ascospores but produce masses of hiille cells or hyphae, sclerotial species, and species
that lack apparent vestiges of ascospore production. Heterothallism
is shown as a possible derivative of homothallism,
although the relationship
between heierothallism&d
homothallism
is not known.
Our sample of taxa was wide, but it was not a complete analysis of the possible 186 species with Aspergillus conidial states. We cannot be sure that a different
sample would not suggest multiple gains rather than
losses of meiosis. However, it is biologically
more feasible that the meiotic state has been lost multiply than
gained multiply.
Ascospore
production
is a complex
process that involves the development
of a specific
growth stage and formation of a specialized structure,
the cleistothecium.
The genetics of ascospore formation
suggests that ascomycetes
may harbor a large number
of genetic targets that when mutated confer a loss of
meiosis. In Neurospora
crassa, an unrelated ascomycete, over 200 different mutations have been isolated
with meiosis-specific
effects (Raju 1992). Wild N. crassa strains harbor a surprising level of lethal recessive
mutations
specific to meiotic sporulation.
Leslie and
Raju (1985) identified 106 lethal recessive diplophaseTable 2
Ln Likelihood Comparison of Unconstrained
Constrained Most-Parsimonious
Trees
Ln
Likelihood
Difference
from Most
Parsimonious Tree
Standard
Deviation
139
19 1
- I,2452
- I.5354
-290.2
36.5
140
191
- I ,639.6
- 1,902.2
-262.6
44.1
Tree
Length
(Steps)
Nuclear
and
ITS + 5.8s
Unconstrained
Constrained”
Mitochondrial
Unconstrained.
Constrained
*The constramed most-parsimonious tree with the highest log
given.
hkehhood
is
specific mutations,
most of which were unique, that
eliminated or reduced meiotic fertility from 99 wild isolates. In contrast, relatively little is known about the genetics of meiotic sporulation in Emericella nidulans, the
genetically
best-characterized
Aspergillus.
There are
many mutations known to eliminate or vastly reduce
ascospore formation in this species (Champe, Nagle, and
Yager 1994), but few known with phenotypes
exclusively related to meiosis, perhaps because no exhaustive
search has been made. Most meiotic mutants are considered to be pleiotropic, and were noted as second phenotypes in addition to vegetative or conidial defects.
Many of these mutations produce phenotypes similar to
the apparent vestiges seen in many asexual species (e.g.,
the presence of hiille cells without cleistothecium
production). For example, medusa (medA) mutants produce
aberrant conidiophores
and lack cleistothecia,
but do
produce hiille cells (Clutterbuck
1969). Likewise, A. ustus isolates do not produce cleistothecia but do produce
hiille cells. Other mutations, such as velvet (veA), have
a quantitative effect on cleistothecium
production (KBfer
1965). Similarly, a great deal of genetic variation in the
quantity of cleistothecia produced is observed among E.
nidulans isolates (Butcher 1968). Overall, these data indicate that there are a large number of potential target
genes that could confer losses of meiosis in Aspergillus,
and that variation in meiosis-specific
genes is present in
fungal populations.
Multiple independent
gains of meiosis are more
difficult to imagine. The sexual states associated with
Aspergillus share a great deal of morphological
similarity: all involve the formation of enclosed cleistothecia,
produce unordered, evanescent asci, and produce similarly shaped ascospores. It is unlikely that these char-
Loss of Meiosis in Aspergillus
acters evolved together independently
several times. It
is possible, however, that meiosis is an oscillating character that is lost and regained frequently. The frequency
of such oscillation
would probably have to be very
short, as asexual taxa would likely accrue additional
meiosis-specific
mutations, reinforcing the loss.
The age of strictly mitotic lineages is difficult to
address in fungi, where a paucity of fossil data makes
a reasonable calibration
of a molecular clock difficult
(Berbee and Taylor 19936). In particular, our data do
not eliminate the possibility that A. niger and/or A. Juvus represent relatively ancient, asexual lineages. While
there are no obvious candidates for closely related meiotic sister taxa to these species other than perhaps Fennellia, we did not sample all of the 72 meiotic taxa in
this study.
Of the four strictly mitotic species inferred to be
recently derived, three show apparent vestiges of ascospore production (fig. 2). Both A. ustus and A. versicolor
produce htille cells that are similar to those in their inferred meiotic sister taxa. Htille cells are associated with
cleistothecium
(meiotic fruit body) production
in subgenus Nidulantes, but are also present in many strictly
mitotic species. Some isolates of A. wentii produce
masses of hyphae that Raper and Fennel1 (1965) noted
were similar to the immature cleistothecia
of Chaetosartorya chrysella, although they classified the two species in different groups. A. restrictus does not possess
any obvious vestiges of ascospore production,
but the
fact that it is strongly xerophilic may indicate a connection to E. amstelodumi. The mitochondrial
data weakly
suggest a connection
between A. terreus and Fennellia
jluvipes. Raper and Fennel1 (1965) noted similarities between them, and the sclerotium-like
bodies present in A.
terreus may represent vestigial cleistothecia.
Sclerotia,
structures
resistant
to harsh conditions
produced
in
many strictly mitotic species (including A. niger and A.
jkzvus), are proposed to be derived from cleistothecia
(Malloch and Cain 1972), and may represent another
vestige of ascospore production.
The possibility exists that the observed variation in
meiotic reproduction
does not reflect complete losses of
meiosis in nature, but rather losses of our ability to induce meiosis in culture. However, the fact that many
single nonlethal mutations can induce losses of meiosis
in the same genetic backgrounds under the same cultural
conditions
strongly suggests that loss of meiosis is a
biologically
feasible phenomenon.
Heterothallism
in Emericella heterothallica
is similar to that observed in heterothallic Neurospora species.
An isolate must be fertilized by another isolate of opposite mating-type
for successful ascospore production
(Kwon and Raper 1967). Heterothallism
is known in
only three meiotic species with Aspergillus
mitotic
states, including E. heterothallica.
The other two species, Neosartorya
spathulata and Neosartorya fennelliae, are probably closely related to N. jischeri (KwonChung and Kim 1974; Peterson 1993). The question of
whether heterothallism
is ancestral in fungi has been
called “imponderable”
(Raper 1966), but its rarity in
Aspergillus
may suggest that it is derived, perhaps
815
caused by simple mutations that must be complemented
to allow ascospore production.
Another possibility
is
that heterothallism
is ancestral in the A. ustus/E. heterothallica clade, and A. ustus is asexual because it lost
one of the mating types. Mating-type
genes and genes
regulating sexual development have not been isolated in
Aspergillus or its sexual relatives, but extinction of one
putative mating type is another possible basis for the
loss of meiosis, particularly in the lineage leading to A.
ustus.
Aspergilli lacking meiosis are capable of efficient
propagation
via conidium production
and may not suffer short-term fitness effects from the loss of the meiotic state. The loss of meiosis may actually provide a
short-term
benefit, allowing
full investment
of resources into conidial propagation,
which is likely faster and more efficient
than ascospore
formation
(Champe, Nagel, and Yager 1994). However, in addition to recombination,
meiosis may also offer typical adaptive benefits associated with sexual reproduction (Bonner 1958), such as dormancy
and resistance
to dessication
(Cooke and Whipps 1993). Most Aspergilli can produce mature conidia within 24 h after
induction
(Timberlake
and Clutterbuck
1994), with
mature ascospore formation taking from 4 days to several weeks in different
Aspergillus-related
species
(Raper and Fennel1 1965). The loss of the resistant
qualities of the ascospore may be compensated
for by
the production
of sclerotia
in species that produce
them, such as A. jkzvus and A. niger.
The data presented here indicate that asexual Aspergilli are often recently derived from meiotic lineages,
and do not give any strong support to the existence of
ancient asexual lineages. The patterns of evolution in
Aspergillus and its related meiotic relatives, like those
seen in Penicillium and its meiotic Talaromyces relatives (LoBuglio, Pitt, and Taylor 1993) are consistent
with strictly mitotic lineages being deficient in recombination, and thus susceptible
to the accumulation
of
deleterious mutations through Muller’s ratchet and mutational meltdown. In Aspergillus,
although the possibility exists that significant
parasexual recombination
occurs in natural populations, the general observation is
that most isolates capable of undergoing the first step in
the parasexual
cycle (heterokaryosis)
are genetically
identical
(Croft and Jinks 1977; unpublished
data).
However, recombination
may be occurring in apparently
asexual species, and detailed population genetic analyses are still needed to establish this.
Acknowledgments
We thank M. Berbee, K. LoBuglio, and J. Taylor,
who generously
provided unpublished
sequences and
genomic DNA from Talaromycesjavus.
This work was
supported
by National
Institutes
of Health
grant
GM37886 to W.E.T. and National Science Foundation
dissertation
research grant DEB-9224224
to M.L.A.,
D.M.G., and W.E.T. D.M.G. was supported by National
Institutes
of Health
pre-doctoral
training
grant
816
Geiser et al
A107373-01 from the National
Infectious Diseases.
Institute
of Allergies
and
LITERATURE CITED
AVISE, J. C., and R. M. BALL. 1990. Principles of genealogical
concordance in species concepts and biological taxonomy.
Oxf. Surv. Evol. Biol. 7:4.5-67.
BARRATT, R. W., G. B. JOHNSON, and N. W. OKATA. 1965.
Wild-type and mutant stocks of Aspergillus nidulans. Genetics 52:233-246.
BERBEE, M. L., and J. W. TAYLOR. 1993~. Ascomycete relationships: dating the origin of asexual lineages with 18s
ribosomal RNA gene sequence data. Pp. 67-78 in D. R.
REYNOLDS and J. W. TAYLOR, eds. The fungal holomorph:
mitotic, meiotic and pleomorphic speciation in fungal systematics. CAB International, Wallingford, U.K.
BERBEE, M. L., and J. W. TAYLOR. 1993b. Dating the evolutionary radiations of the true fungi. Can. J. Bot. 71: 11141127.
BERBEE, M. L., A. YOSHIMURA, J. SUGIYAMA, and J. W. TAYLOR. 1995. Is Penicillium monophyletic?
An evaluation of
phylogeny in the family Trichocomaceae
from 18S, 5.8s
and ITS ribosomal data. Mycologia 87:210-222.
BONNER, J. T 1958. The relation of spore formation to recombination. Am. Nat. 92: 193-200.
BREMER, K. 1994. Branch support and tree stability. Cladistics
10:295-304.
BUTCHER, A. C. 1968. The relationship between sexual outcrossing and heterokaryon incompatibility in Aspergillus nidulans. Heredity 23:443-452.
CHAMPE, S. I?, D. L. NAGLE, and L. N. YAGER. 1994. Sexual
sporulation. Pp. 429-454 in S. D. MARTNELLI and J. R.
KINGHORN, eds. Aspergillus: 50 years on. Elsevier, Amsterdam.
CLUTTERBUCK, A. J. 1969. A mutational analysis of conidial
development in Aspergillus nidulans. Genetics 63:317-327.
COOKE, R. C., and J. M. WHIPPS. 1993. Ecophysiology of fungi. Blackwell Scientific Publications, London.
COTTY, F? J., P BAYMAN, D. S. EGEL, and K. S. ELIAS. 1994.
Agriculture, aflatoxins, and Aspergillus. Pp. l-27 in K. A.
POWELL, A. RENWICK, and J. E PEBERDY, eds. The genus
Aspergillus: from taxonomy and genetics to industrial application. Plenum Press, New York.
CROFT, J. H., and J. L. JINKS. 1977. Aspects of the population
genetics of Aspergillus nidulans. Pp. 339-360 in J. E.
SMITH and J. E. PATEMAN, eds. Genetics and physiology of
Aspergillus. Academic Press, London.
FARRIS, J. S. 1989. The retention index and the resealed consistency index. Cladistics 5:417-419.
FELSENSTEIN,J. 1993. PHYLIP (phylogeny inference package),
version 3.5~. Department of Genetics, University of Washington, Seattle.
GEISER, D. M., M. L. ARNOLD, and W. E. TIMBERLAKE. 1994.
Sexual origins of British Aspergillus nidulans isolates. Proc.
Natl. Acad. Sci. USA 91:2349-2352.
HEDGES, S. B., J. F? BOGART, and L. R. MAXSON. 1992. Ancestry of unisexual salamanders. Nature 356:708-710.
HODGES, S. A., and M. L. ARNOLD. 1994. Columbines: a geographically widespread species flock. Proc. Nat]. Acad. Sci.
USA 91:5129-5132.
JUKES, T. H., and C. R. CANTOR. 1969. Evolution of protein
molecules. Pp. 21-132 in H. N. MUNRO, ed. Mammalian
protein metabolism. Academic Press, New York.
KNIFER, E. 1965. The origin of translocations
in Aspergillus
nidukms. Genetics 52:2 17-232.
KIMURA, M. 1980. A simple method for estimating the evolutionary rate of base-substitution
through comparative
studies of nucleotide
sequences.
J. Mol. Evol. 16: 11 l120.
KISHINO, H., and M. HASEGAWA. 1989. Evaluation of the maximum likelihood estimate of the evolutionary tree topologies from DNA sequence data, and the branching order of
the Hominoidea. J. Mol. Evol. 29: 170-179.
KLICH, M. A., and J. I. PITT. 1988. A laboratory guide to comCSIRO,
mon Aspergillus species and their teleomorphs.
North Ryde, Australia.
KUNTZEL, H., and H. G. HOCHEL. 1981. Evolution of rRNA
and origin of mitochondria. Nature 293:751-755.
KWON, K.-J., and K. B. RAPER. 1967. Sexuality and cultural
characteristics
of Aspergillus heterothallicus. Am. J. Bot.
54:3648.
KWON-CHUNG, K. J., and S. J. KIM. 1974. A second heterothallic Aspergillus. Mycologia 66:628-638.
LESLIE, J. E, and N. B. RAJU. 1985. Recessive mutations from
natural populations of Neurosporu crassa that are expressed
in the sexual diplophase. Genetics 111:759-777.
LOBUGLIO, K. E, J. I. PITT, AND J. W. TAYLOR. 1993. Phylogenetic analysis of two ribosomal DNA regions indicates
multiple independent losses of a sexual Talaromyces state
among asexual Penicillium species in subgenus Biverticillium. Mycologia 85:592-604.
LYNCH, M., R. BURGER, D. BUTCHER, and W. GABRIEL. 1993.
The mutational meltdown in asexual populations. J. Hered.
84:339-344.
MADDISON, W. P, and D. R. MADDISON. 1992. MacClade, analysis of phylogeny and character evolution, version 3. Sinauer and Associates, Sunderland, Mass.
MALLOCH, D., and R. E CAIN. 1972. The Trichocomataceae:
Ascomycetes
with Aspergillus, Paecilomyces and Penicillium imperfect states. Can. J. Bot. 50:2613-2628.
MAYNARD SMITH, J. 1978. The evolution of sex. Cambridge
University Press, Cambridge, U.K.
-.
1992. Age and the unisexual lineage. Nature 356:661662.
MULLER, H. J. 1964. The relation of recombination
to mutational advance. Mutat. Res. 1:2-9.
PETERSON, S. W. 1993. Neosartorya pseudojischeri sp. nov.
and its relationship to other species in Aspergillus section
Fumigati. Mycol. Res. 35:547-554.
PONTECORVO, G. J., A. ROPER, L. M. HEMMONS, K. D.
MCDONALD, and A. W. J. BUFTON. 1953. The genetics of
Aspergillus nidulans. Adv. Genet. 5: 141-238.
QUATTRO, J. M., J. C. AVISE, and R. C. VRIJENHOEK. 1992.
An ancient clonal lineage in the fish genus Poeciliopsis
(Atheriniformes:
Poeciliidae). Proc. Natl. Acad. Sci. USA
89:348-352.
RAJU, N. B. 1992. Genetic control of the sexual cycle in Neurospora. Mycol. Res. 96:241-262.
RAPER, J. R. 1966. Genetics of sexuality in higher fungi. Ronald Press, New York.
RAPER, K. B., and D. I. FENNELL. 1965. The genus Aspergillus.
Williams & Wilkins, Baltimore, Md.
SAMSON, R. A. 1994. Current systematics of the genus Aspergillus. Pp. 261-276 in K. A. POWELL, A. RENWICK, and J.
F. PEBERDY, eds. The genus Aspergillus: from taxonomy
and genetics to industrial application. Plenum Press, New
York.
SPOLSKY, C. M., C. A. PHILLIPS, and T. UZZELL. 1992. Antiquity of clonal salamander lineages revealed by mitochondrial DNA. Nature 356:706-708.
Loss of Meiosis in Aspergillus
SWOFFORD, D. L. 1993. PAUP: phylogenetic
analysis using
parsimony, version 3.1.1. Illinois Natural History Survey,
Champaign, Ill.
TIMBERLAKE, W. E., and A. J. CLU~ERBUCK. 1994. Genetic
regulation of conidiation. Pp. 383-427 in S. D. MARTINELLI
and J. R. KINGHORN, eds. Aspergillus: 50 years on. Elsevier,
Amsterdam.
WHITE, T J., T. D. BRUNS, S. LEE, and J. W. TAYLOR. 1990.
Amplification
and direct sequencing of fungal ribosomal
genes for phylogenetics.
Pp. 315-322 in M. A. INNIS, D.
817
H. GELFLAND, J. J. SNINSKY, and T. J. WHITE, eds. PCR
protocols. Academic Press, San Diego, Calif.
YELTON, M. E., J. E. HAMER, and W. E. TIMBERLAKE. 1984.
Transformation
of Aspergillus nidulans by using a trpC
plasmid. Proc. Natl. Acad. Sci. USA 81:1470-1474.
RODNEY L. HONEYCUTT, reviewing
Accepted
March
12, 1996
editor