Mechanisms inuencing the evolution of

resistance to Qo inhibitor fungicides

Ulrich Gisi,
1
* Helge Sierotzki,
2
Alison Cook
3
and Alan McCaffery
3
1
SYNGENTA Crop Protection, Research, Product Biology, WRO-1060, CH-4002 Basel, Switzerland
2
SYNGENTA Crop Protection, Research Biology, WST-540, CH-4332 Stein, Switzerland
3
SYNGENTA Crop Protection, Research, Product Biology, Jealotts Hill International Research Centre, Bracknell, UK
Abstract: Fungicides inhibiting the mitochondrial respiration of plant pathogens by binding to the
cytochrome bc1 enzyme complex (complex III) at the Qo site (Qo inhibitors, QoIs) were rst intro-
duced to the market in 1996. After a short time period, isolates resistant to QoIs were detected in eld
populations of a range of important plant pathogens including Blumeria graminis Speer f sp tritici,
Sphaerotheca fuliginea (Schlecht ex Fr) Poll, Plasmopara viticola (Berk & MA Curtis ex de Bary) Berl
& de Toni, Pseudoperonospora cubensis (Berk & MA Curtis) Rost, Mycosphaerella jiensis Morelet
and Venturia inaequalis (Cooke) Wint. In most cases, resistance was conferred by a point mutation in
the mitochondrial cytochrome b (cyt b) gene leading to an amino-acid change from glycine to alanine
at position 143 (G143A), although additional mutations and mechanisms have been claimed in a
number of organisms. Transformation of sensitive protoplasts of M jiensis with a DNA fragment of a
resistant Mjiensis isolate containing the mutation yielded fully resistant transformants, demonstrat-
ing that the G143A substitution may be the most powerful transversion in the cyt b gene conferring
resistance. The G143A substitution is claimed not to affect the activity of the enzyme, suggesting that
resistant individuals may not suffer from a signicant tness penalty, as was demonstrated in B
graminis f sp tritici. It is not known whether this observation applies also for other pathogen species
expressing the G143A substitution. Since fungal cells contain a large number of mitochondria, early
mitotic events in the evolution of resistance to QoIs have to be considered, such as mutation frequency
(claimed to be higher in mitochondrial than nuclear DNA), intracellular proliferation of mitochondria
in the heteroplasmatic cell stage, and cell to cell donation of mutated mitochondria. Since the cyt b
gene is located in the mitochondrial genome, inheritance of resistance in lamentous fungi is expected
to be non-Mendelian and, therefore, in most species uniparental. In the isogamous fungus Bgraminis f
sp tritici, crosses of sensitive and resistant parents yielded cleistothecia containing either sensitive or
resistant ascospores and the segregation pattern for resistance in the F1 progeny population was 1:1. In
the anisogamous fungus V inaequalis, donation of resistance was maternal and the segregation ratio
1:0. In random mating populations, the sex ratio (mating type distribution) is generally assumed to be
1:1. Therefore, the overall proportion of sensitive and resistant individuals in unselected populations is
expected to be 1:1. Evolution of resistance to QoIs will depend mainly on early mitotic events; the
selection process for resistant mutants in populations exposed to QoI treatments may follow mechan-
isms similar to those described for resistance controlled by single nuclear genes in other fungicide
classes. It will remain important to understand how the mitochondrial nature of QoI resistance and
factors such as mutation, recombination, selection and migration might inuence the evolution of QoI
resistance in different plant pathogens.
# 2002 Society of Chemical Industry
Keywords: strobilurin fungicides; fungicide resistance; Qo inhibitors; cytochrome b; mitochondrial inheritance;
uniparental inheritance
1 INTRODUCTION
The risk of resistance to fungicides and the mode of
resistance in plant pathogens against anti-fungal
agents should ideally be determined separately for
each fungicide/pathogen combination. Such studies
include investigations at the molecular, genetic,
biochemical, physiological and population levels.
They are the basis for sound product-use recommen-
(Received 8 June 2002; accepted 17 June 2002)
* Correspondence to: Ulrich Gisi, SYNGENTA Crop Protection, Research, Product Biology, WRO 1060, CH-4002, Basel, Switzerland
E-mail: ulrich.gisi@syngenta.com

Dedicated to the memory of Steve Heaney

Based on a presentation at the Meeting Resistance 2001, organised by IACR, SCI and BCPC and held at IACR, Rothamsted, Harpenden,
UK, on 2426 September 2001
# 2002 Society of Chemical Industry. Pest Manag Sci 1526498X/2002/$30.00 859
Pest Management Science Pest Manag Sci 58:859867 (online: 2002)
DOI: 10.1002/ps.565
dations in disease-control programmes and allow
successful strategies to delay the evolution of resis-
tance in pathogen populations. As a result of intensive
product use on a large area, naturally occurring
resistant individuals may be selected and may compete
with the wild-type population. Therefore, resistance of
plant pathogens to fungicides can be dened
1
as a
combination of:
(1) the presence of naturally occurring resistant
individuals, initially at very low frequency, origi-
nating from recurrent mutations conferring resis-
tance and from sexual and asexual recombination;
(2) the increase in frequency of resistant individuals
over time (during the season, from year to year)
resulting in resistant sub-populations caused by
the selection process of fungicide applications and
through migration of resistant individuals;
(3) reduction of disease control compared to earlier
results and to standard treatments with products
of a different mode of resistance class.
Long before resistant individuals reach a signicant
proportion of the eld population, some fundamental
processes will have occurred in the target organisms,
including recurrent mutations and repair of mutated
DNA, transcription of mutated DNAto target protein,
intracellular selection for sensitivity and resistance (to
mitochondrial encoded target), as well as segregation
of wild-type and mutated offspring individuals after
sexual recombination. The early processes are often
overlooked because appropriate detection methods are
not available in most cases. Nevertheless, so as to
understand early events in resistance evolution, the
genetic and biological background of the fungus/
fungicide interaction has to be investigated in highly
accurate and sensitive experimental approaches.
Strobilurin fungicides have a single-site mode of
action. They inhibit mitochondrial respiration by
binding to the Qo site (the outer, quinone oxidizing
pocket) of cytochrome bc1 enzyme complex (complex
III), thus blocking electron transfer in the respiration
pathway and leading to energy deciency due to a lack
of ATP,
2
and are known as Qo inhibitors (QoIs). The
enzyme is encoded by the cytochrome b (cyt b) gene
located in the mitochondria. Following the commer-
cial introduction of QoIs in 1996 resistant isolates
were detected in eld populations in several plant
pathogens including Blumeria graminis Speer f sp tritici
on wheat in Germany, France and the UK, Sphaero-
theca fuliginea (Schlecht ex Fr) Poll and Pseudoperono-
spora cubensis (Berk & MA Curtis) Rost on cucumber
in Spain and Japan, Plasmopara viticola (Berk & MA
Curtis ex de Bary) Berl & de Toni on grape in Italy,
Mycosphaerella jiensis Morelet on banana in Costa
Rica and Venturia inaequalis (Cooke) Winter on apple
in experimental elds in Switzerland and northern
Germany.
35
Since resistant isolates collected from
eld populations were available, a range of basic
studies was possible which contributed to the elucida-
tion of mechanisms of resistance to the QoI class of
fungicides. This paper will review current knowledge
on QoI resistance with special reference to mitochon-
drial biology and inheritance.
2 CHEMICAL CLASSES OF Qo INHIBITORS
Based on structural similarities, seven chemical classes
can be distinguished within the Qo inhibitors. They all
have a common mode of action by binding to the Qo
site of the cytochrome bc1 complex and are generally
cross-resistant when tested in different assays such as
in planta tests, spore germination, mycelial growth,
cell-free enzyme tests and on articial mutants,
3,68
although their spectra and intrinsic levels of biological
activity are quite different.
9
The seven classes and
representative compounds (Fig 1) are:
(I) methoxyacrylates: natural compounds such as
strobilurin A and B, and synthetic analogues
such as azoxystrobin and picoxystrobin;
(II) methoxycarbamates: pyraclostrobin (BAS 500);
(III) oximinoacetates: kresoxim-methyl, trioxy-
strobin;
(IV) oximinoacetamides: metominostrobin (SSF
126);
(V) oxazolidinediones: famoxadone
(VI) imidazolinones: fenamidone (RPA 407213)
(VII) antibiotics: myxothiazol.
The toxophore is similar in all compounds and
always carries a carbonyl oxygen moiety (Fig 1) that is
thought to be responsible for binding to the enzyme.
The toxophore is connected to a benzene ring, linker,
(except for strobilurin B and myxothiazol which have
an open carbon chain) which carries a long side chain
(except for famoxadone and fenamidone). Based on
published crystal structures of the cytochrome bc1
enzyme complex
1012
and co-crystallisation of the
bovine heart enzyme with different Qo inhibitors, a
model was constructed showing their most likely t in
the binding pocket (Fig 2). The carbonyl oxygen
moiety of the toxophore may bind with a hydrogen
bond to the amide group of glutamine at position 272
(Fig 2) or to proline 271
12
of the ef-protein helix (Fig
2). The ester methoxy oxygen of the toxophore may
come close to alanine 128 and tyrosine 274, while the
acrylate methoxy group may be close to phenylalanine
129 and tyrosine 132. The benzene ring linker moiety
(or the C=C bond in open carbon chain of the natural
strobilurins and in myxothiazol; see arrow in Fig 1)
appears to come close to glycine 143 of the cd-protein
helix (Fig 2), whereas the side chain is in a rather large
pocket containing isoleucine 147, phenylalanine 275
and other amino acids (Fig 2).
12
This binding pattern
appears to be true for molecules carrying aliphatic or
aromatic toxophore linkers such as the methoxyacry-
lates (eg azoxystrobin or the strobilurins) as well as the
oxazolidinediones and imidazolinones (eg famoxa-
done and fenamidone) and may be the biochemical
explanation for a general cross-resistance reaction
when the inhibitors are tested against whole organisms
860 Pest Manag Sci 58:859867 (online: 2002)
U Gisi et al
in vitro or in planta.
3
A somewhat different binding
model than that presented here and in the cited
literature was published recently for metominostrobin
claiming histidine 161 of the ironsulfur subunit of
cytochrome bc1 complex as binding site.
13
There may
be some differences in binding between single Qo
inhibitors (S Pember, pers comm, 2001), but the most
likely t to the enzyme pocket seems to be very similar
for all QoIs. All models presented so far involve animal
enzyme (bovine, chicken, cow, rabbit) and it remains a
challenge to co-crystallize a fungal enzyme with
different Qo inhibitors.
3 MUTATIONS IN cyt b GENE
3.1 Mutations conferring resistance
At least eleven single or combined mutations in the cyt
b gene with single nucleotide polymorphisms (SNPs)
have been described as conferring resistance to Qo
inhibitors in two extra-membrane regions of the
cytochrome bc1 enzyme complex: at amino acid
positions from 127 to 147 (cd loop) and from 275 to
296 (ef loop) in different organisms including bacteria,
algae, yeast, protozoa and animals.
14,15
Individuals
resistant to QoIs usually express high resistance
factors, especially when changes occur in the aa
127147 loop (Table 1).
14
The mutations leading to
the substitution of phenylalanine for leucine at posi-
tion 129, F129L, and glycine for alanine at position
143, G143A, resulted in high resistance factors but
were claimed to have no signicantly negative effect on
enzyme activity
15
and thus on tness of individuals. In
contrast, the G137R/E/V/S substitution was reported
to cause a strong effect on the respiratory competence
of mutants, which may not survive in nature. When the
amino-acid sequences of the cytochrome bc1 complex
of wild type isolates in fungi like Aspergillus nidulans
(Eidam) Winter and Saccharomyces cerevisiae Meyer ex
Hansen were compared to basidiomycete fungi pro-
Figure 2. Binding pocket with amino acids (yeast numbering) of the
cytochrome bc1 enzyme complex (of bovine heart) based on the co-crystal
structure with the Qo inhibitor azoxystrobin (adapted from Link et al,
10
Zhang et al,
11
Iwata et al.
12
Figure 1. Chemical classes and
structures of Qo site respiration
inhibitors (QoIs). The carbonyl oxygen
(highlighted with small circle) of the
toxophore (large circle) binds with a
hydrogen bond to the amide group of
glutamine at position 272, and the
linker moiety (arrow) comes close to
glycine 143 of the bc1 enzyme complex
(see Fig 2).
Pest Manag Sci 58:859867 (online: 2002) 861
Evolution of resistance to Qo inhibitor fungicides
ducing strobilurins as metabolic end-products, such as
Strobilurus tenacellus (Pers ex Fr) Singer, Mycena
galopoda and M viridimarginata, ve signicant differ-
ences in the amino-acid sequence were detected in the
latter group of fungi: a change from threonine to
isoleucine, T127I, from alanine to serine, A153S,
from serine to glutamine, S255Q, from asparagine to
aspartic acid, N262D, and the already mentioned
G143A (Table 2).
16
These ve changes are thought to
protect the strobilurin producers from being poisoned
by their own metabolite. Since the cytochrome bc1 of
these fungi is obviously functional, the amino-acid
changes cannot result in any signicant tness penalty,
at least not in that group of fungi, otherwise they
would not have survived in evolution.
It may not be surprising that the principle mutation
detected so far in plant pathogenic fungi that confers
eld resistance to QoIs is one of the ve substitutions
detected in basidiomycetes, namely G143A. This
mutation has been detected in resistant isolates of B
graminis f sp tritici, B graminis f sp hordei, S fuliginea, P
cubensis, P viticola and M jiensis.
1,3,4,7,17,18
In V
inaequalis, only some of the resistant isolates carried
G143A, while those which do not must possess an as
yet unknown mode of resistance.
5
There is no
evidence that the size of the cyt b gene and the number
of introns in the gene (six in V inaequalis, one in M
jiensis, none in B graminis and P viticola)
1
would affect
susceptibility to mutation. Recently, an additional
amino-acid change from phenylalanine to leucine at
position 129, F129L, was detected conferring resis-
tance to QoIs in Pyricularia grisea (Sacc)
19
and Pythium
aphanidermatum (Edson) Fitzp (G Olaya, pers comm,
2001), both important pathogens of turf grass. As Qo
Table 1. Amino-acid (aa) positions and
changes in the cytochrome bc1 enzyme
complex conferring resistance (RF) to
myxothiazol in different species
aa Position and change Species Resistance factor RF
F129 L Saccharomyces cerevisiae 930
F129 L Rhodobacter capsulatus 530
F129 L Chlamydomonas reinhardtii (alga) 400
F129 S Rhodobacter capsulatus 530
G137 R Saccharomyces cerevisiae 4
G137 E Saccharomyces cerevisiae 20
G137 V Saccharomyces cerevisiae 4
G137 S Rhodobacter capsulatus 37
G143 A Mus musculus (mouse) 2000
G143 A Paracentrotus lividus (sea urchin) 1990
G143 D Rhodobacter capsulatus 10000
G137 !T
G143 !T
N256 !F
9
=
;
Paramecium aurelia (protozoan) 22000
N256 Y Saccharomyces cerevisiae 6
L275 S Saccharomyces cerevisiae 6
L275 T Saccharomyces cerevisiae 4
V292 A Rhodobacter capsulatus 5
L295 F Mus musculus (mouse) 4
Data adapted from Degli Esposti et al.
14
Table 2. Position and changes (in italic) of amino
acids in the cytochrome bc1 enzyme complex of
Saccharomyces cerevisiae (S c) mutants and in
fungi sensitive and naturally tolerant to QoIs
Position Sensitive fungi
a
Sc mutants St
b
Mg
c
Mv
d
Vi
e
127 Thr Thr Ile Ile Ile Thr
129 Phe Leu Phe Phe Phe Phe
137 Gly Arg Gly Gly Gly Gly
143 Gly Gly Gly Ala Gly Gly
147 Ile Phe Ile Ile Ile Ile
153 Ala Ala Ala Ser Ala Ala
255 Ser(Pro) Pro Gln Ser Pro Ser
257 Asn Tyr Asn Asn Asn Asn
262 Asn Asn Asp Asn Asn Asn
275 Tyr Asn Tyr Tyr Tyr Tyr
276 Leu Ser Leu Leu Leu Leu
a
Aspergillus nidulans, Neurospora crassa, Saccharomyces cerevisiae.
b
Strobilurus tenacellus.
c
Mycena galopoda.
d
Mycena viridimarginata.
e
Venturia inaequalis (sensitive wild type).
Data adapted from Zheng and Ko ller.
16
862 Pest Manag Sci 58:859867 (online: 2002)
U Gisi et al
inhibitor fungicides are now widely used in many
crops, it will be important to follow whether or not
additional SNPs will evolve, and how frequently they
will appear in eld populations.
3.2 Molecular diagnosis of resistance
Several molecular methods have been developed for
the detection of resistance based on the presence of
G143A, all being based on PCR amplication of a
DNAfragment that spans the site of mutation.
20
In the
rst approach, allele-unspecic but species-specic
primers are used for PCR, followed by G143A-specic
restriction digest of the mutated allele (PCR-RFLP).
In the second approach, PCR is performed with allele-
unspecic primers in combination with allele-specic
hybridisation probes known as molecular beacons
(MB-PCR). In a third approach, selective amplica-
tion of either the wild-type or the mutated cyt b is used
for allele-specic primers using the amplication
refractory mutation system (ARMS) accomplished
by means of a Scorpion primer in which uorescence is
specic for the amplicon.
3,20,21
As an alternative,
allele-specic primers using the uorescent dye SYBR
Green I have been developed recently in which
uorescence is unspecic for the amplicon.
22,23
The
rst approach allows only qualitative identication of
SNPs, whereas the second two methods allow the
quantitative detection of both alleles in real-time PCR
assays. For MB-PCR and ARMS Scorpions, a simul-
taneous detection of alleles using several uorescent
dyes is possible (multiplexing). For B graminis, the
correlation between the molecular and the biological
tests are in most cases almost perfect,
3
in other
pathogens (eg P viticola, S fuliginea) resistant pheno-
types cannot always be correlated to the presence of
G143A.
3
In the case of P aphanidermatum and V
inaequalis, where other mutations in the cyt b gene or
other resistance mechanisms outside the cytochrome
bc1 complex may confer resistance, new detection
tools may be developed once the mechanisms are
elucidated.
3.3 Complementation experiments
The ultimate proof that the G143A mutation is
responsible for resistance is given by a functional
genetic study: protoplast transformation of sensitive
eld isolates of M jiensis with a 570-bp fragment
amplied from a resistant M jiensis eld isolate
containing the mutation yielded fully resistant trans-
formants of a sensitive isolate, conrming that G143A
can confer full resistance.
24
The transformation of the
mitochondrial DNA was performed with the classical
PEG/spermidine method as it is known for nuclear
DNA, using a very high density of protoplasts; the rate
of transformation was rather low (10
9
).
24
3.4 Resistance in eld populations of plant
pathogens
Three groups of plant pathogen species can be
distinguished in regard to QoI eld resistance. In the
rst group, widespread detection of G143A in eld
populations and resistant individuals without signi-
cant tness penalty have been established for B
graminis f sp tritici on wheat in northern Europe, and
S fuliginea on cucumber in Spain.
3,4
In a second group,
G143A can be detected in localized areas, such as M
jiensis on banana in Costa Rica, P viticola on grapes in
Italy and France, P cubensis on cucumber in Japan, B
graminis f sp hordei on barley in northern France and V
inaequalis in experimental elds in Switzerland and
Germany (Syngenta internal observations; H Steva
[for P v in France] and H Ishii [for P c in Japan], pers
comm, 2001). Many important pathogens such as
Mycosphaerella graminicola (Fuckel) Schroter on wheat
(thousands of isolates tested), Puccinia recondita Rob
ex Desmon wheat, Pyrenophora teres Dreschs on barley
and Uncinula necator (Schwein) Burr on grapes have
not evolved resistance so far. They have been exposed
to QoIs from the outset, and may infect the same host
as those pathogens which have developed high
resistance levels in eld populations. It remains an
open question what the genetic background is for the
different evolution of resistance to QoIs in these and
other pathogens.
4 NUCLEAR AND CYTOPLASMIC INHERITANCE
In contrast to all previous fungicide classes, QoIs
inhibit a target site, cytochrome bcl, encoded by a
mitochondrial rather than a nuclear gene, cyt b. There
are several important differences between nuclear
(Mendelian) and cytoplasmic (extra-nuclear, non-
Mendelian) inheritance. The segregation of nuclear
encoded traits is a result of meiosis during sexual
recombination and pairs of alleles are assorted
independently in unlinked genes, whereas segregation
of mitochondrial genes always occurs during mitotic
cell division. Inheritance of nuclear genes is biparental,
whereas for mitochondrial genes, it is mostly uni-
parental, often maternal, and intracellular selection
can lead to a loss of alleles or establishment of mutated
alleles. The rearrangement in mitochondrial genomes
mostly does not affect metabolic processes and thus
tness, but may interfere with cell ageing (at least in
animal cells). It is assumed that the mutation
frequency is lower in nuclear than mitochondrial
DNA due to more effective repair mechanisms; this
was observed in animal cells, although it has been
reported that it might be the reverse in some fungi
(yeasts).
25
In addition, there are always interactions
between nuclei and mitochondria, and nuclear en-
coded mutators may lead to an increased mutation
frequency in mitochondria. Genetic diversity is often
signicantly lower in mitochondrial than in nuclear
genomes. In eld populations only a few mitochon-
drial, although a large number of nuclear, genotypes
were identied in M graminicola
26
and Phytophthora
infestans (Mont) de Bary.
27
In such species, mutated
genotypes may be eliminated more easily during
evolution from populations than in species with higher
Pest Manag Sci 58:859867 (online: 2002) 863
Evolution of resistance to Qo inhibitor fungicides
diversity in the mitochondrial genome (eg Stagono-
spora nodorum).
26
As a consequence, basic questions
with regard to the evolution of QoI resistance have to
be asked: what are the mutation frequencies, repair
mechanisms and occurrence of G143A in the mito-
chondrial genome, and what are the intracellular
frequencies, inheritance and selection of mutated
mitochondria in fungal species both at the individual
and population level?
5 UNIPARENTAL INHERITANCE
5.1 Options for mitochondrial inheritance
The parental contributions in mitochondrial inheri-
tance can vary quite signicantly, depending on the
biology of organisms. In all uniparental cases, the
offspring population receives mitochondria from only
one parent, either male (parental) or female (maternal
inheritance). If sex organs can be differentiated
(antheridia in male, oogonia or ascogonia in female),
the organism is anisogamous and inheritance of
mitochondrial genes can be either maternal as in
oomycetes (eg Pythium, Phytophthora) and most
ascomycetes (eg Neurospora, Venturia, Mycosphaerella)
(and mammals and most angiosperms); or it is
paternal as in the chytridiomycete Allomyces (and
many gymnosperms).
28
If sex organs cannot be
differentiated morphologically ( and indivi-
duals), the organism is isogamous and inheritance can
be maternal as in A nidulans and Basidiomycetes (eg
Coprinus, Agaricus, Armillaria) or hermaphroditic, ie
either parent can act as female, as in B graminis.
29
It is
not well understood whether an isogamous fungus
inherits mitochondria in a maternal or hermaphroditic
manner. Finally, inheritance can also be biparental (or
a combination of uni- and biparental or with leakage
of opposite mating type) where offspring receive
mitochondria from both parents in different pro-
portions, but may undergo partitioning in later cell
divisions as in S cerevisiae (and Drosophila and
crickets).
30
It is assumed that in most organisms there
is always some leakage of the opposite mating partner
in addition to the major type of inheritance.
28
In uniparental inheritance, there are four stages
where mitochondria of the opposite mating type can
be eliminated (Fig 3):
28
(1) pre-zygotic stage, formation of gametangia (an-
theridia and oogonia/ascogonia) or gametes: ex-
clusion of mt DNA from (male) gametes through
partitioning or preferential degradation (eg in
green algae and sh; also known as male sterility
in maize);
(2) fertilization (mating) stage: male mitochondria fail
to penetrate the female cell (only male nuclei
invade female), thus mt DNA of male is lost (eg in
oomycetes, many ascomycetes, basidiomycetes);
(3) zygotic (deterministic) stage: selective silencing or
preferential degradation of mitochondria or mt
DNA (eg in Aspergillus, but also in green algae);
partitioning of parental mitochondria into sepa-
rate cells or novel membrane formation (in some
fungi);
(4) Post-zygotic, mitotic stage (stochastic): random
replication (drift) or partitioning through intra-
cellular selection (eg in yeast).
A simplied ow chart of mitochondrial inheritance
is shown in Fig 3 describing the possible steps for an
uniparental anisogamous case (eg haploid asco-
mycetes like V inaequalis and M jiensis). In step 2,
male mitochondria will be excluded and the zygote is
therefore homoplasmic; after meiosis the offspring
population segregates in a 1:0 (and 0:1) pattern.
Assuming the sex ratio in a randomly mating popula-
tion is 1:1, the overall distribution of a mitochondrion-
encoded trait is 1:1. Because oomycetes are diploid,
meiosis will occur during zygote formation (step 2),
otherwise they will follow the same pattern as
described for the anisogamous ascomycetes. Iso-
gamous fungi such as basidiomycetes and some
ascomycetes like B graminis probably produce hetero-
plasmic zygotes and convert mitochondrial popula-
tions into homoplasmic offspring only after a certain
number of cell divisions. It is not known howmany cell
divisions are needed to express the mutated trait
phenotypically in plant pathogens.
5.2 Case studies of mitochondrial inheritance in
two plant pathogens
5.2.1 Venturia inaequalis
Two examples illustrate the mechanisms described
above in regard to inheritance of QoI resistance in F1
progeny isolates produced from sr crosses. In the
anisogamous pathogen V inaequalis, all randomly
collected F1 progeny isolates produced from one
specic cross were sensitive to QoIs and were not
statistically different from the sensitive parent (Fig 4).
5
This 1:0 segregation pattern for sensitivity conrms
the mitochondrial inheritance of QoI resistance and
indicates that for this cross the sensitive parent was
obviously the female.
Figure 3. Mitochondrial (non-Mendelian) inheritance in an uniparental,
anisogamous case (eg Venturia inaequalis, Mycosphaerella jiensis).
864 Pest Manag Sci 58:859867 (online: 2002)
U Gisi et al
5.2.2 Blumeria graminis f sp tritici
When cleistothecia produced in a cross of the iso-
gamous, hermaphroditic pathogen B graminis f sp
tritici were collected and single ascospores tested for
their sensitivity to QoIs, all ascospores per cleistothe-
cium were either sensitive or resistant to QoIs (Fig 5),
whereas a 1:1 segregation was observed when a
randomly selected population of ascospores were
analysed (Fig 6).
29
This observation demonstrates
again the mitochondrion-encoded resistance for QoIs.
The same progeny isolates were also tested for their
sensitivity against the DMI fungicide triadimenol. A
continuous sensitivity prole for the F1 progeny
intermediate to both parents was observed (Fig 6)
suggesting a polygenic, nuclear controlled background
for resistance to DMIs.
5.3 Evolution of resistance
Inheritance of QoI resistance is non-Mendelian and
results in a 0:1 segregation at single progeny and a 1:1
segregation at population level. After homoplasmic
resistant individuals have emerged, the selection
process for QoI resistance in populations is expected
to follow similar mechanisms, as it is known to occur
for other fungicide classes with nuclear encoded
resistance. The evolution of resistance to QoI fungi-
cides at the individual level probably depends on early
mitotic events such as mutation rate and repair
mechanisms as well as the transition process from
hetero- to homoplasmic cells as a result of intracellular
selection. However, the evolution of resistance to QoIs
in populations is the result of recurrent mutation,
recombination and migration of the pathogen as well
as the selection process imposed by the fungicide.
6 OPEN QUESTIONS
Although important insights to the mechanisms
inuencing the evolution of resistance to Qo inhibitor
fungicides have been gained during the last years,
many basic questions have not yet been answered in
detail, such as:
. What are the mutation frequencies and repair
mechanisms for QoI resistance and interactions
between nucleus and mitochondria in plant patho-
gen species?
. What is the transition from mutated single- to
multiple-copy genome in one mitochondrium and
level of transcription of the cyt b gene, and what is
the turnover of the cytochrome bc1 enzyme com-
plex in sensitive versus resistant mitochondria in
plant pathogen species?
. How does the genetic background of a pathogen
species inuence the expression of the cytochrome b
encoded resistance?
. How many mutated mitochondria per cell are
required to produce a resistant phenotype, and
what impact does mitochondrial organisation as a
network rather than separate organelles have on the
expression of resistance?
. Do pathogen species allow heteroplasmy and, if yes,
Figure 4. Sensitivity to trioxystrobin of F1 progeny isolates of Venturia
inaequalis produced with a sr cross (means with standard deviations for
ANOVA test; adapted from Steinfeld et al
5
).
Figure 5. Sensitivity to QoI fungicides of single ascospore F1 progeny
isolates hatched from individual cleistothecia from a sr cross (r with
G143A change) of Blumeria graminis f sp tritici (adapted from Robinson et
al
29
).
Figure 6. Sensitivity (EC 50) to DMI (triadimenol) and QoI fungicides (eg
trioxystrobin) of F1 progeny isolates produced with a sr cross of
Blumeria graminis f sp tritici on detached leaves of wheat (adapted from
Robinson et al
29
).
Pest Manag Sci 58:859867 (online: 2002) 865
Evolution of resistance to Qo inhibitor fungicides
for how many cell divisions? What are the transition
mechanisms in mitochondrial inheritance from
heteroplasmic to homoplasmic stages and the
intracellular partitioning into sensitive and resistant
individuals in absence and presence of fungicide
selection in different plant pathogen species?
. How does mitochondrial diversity inuence the
likelihood of QoI resistance evolving in different
plant pathogen species?
. Why do related pathogen species behave differently
in regard to QoI resistance evolution, eg Myco-
sphaerella jiensis with detection of G143A versus M
graminicola without resistance; or Blumeria graminis f
sp tritici with widespread detection of G143A versus
B graminis f sp hordei with localized resistance; or
pathogens on the same crop plant such as B graminis
f sp graminis with widespread detection of G143A
versus Puccinia recondita without resistance?
. Does the frequency of G143A differ in non-selected
populations of different plant pathogen species?
. If QoI resistant isolates (with G143A) are not less t
compared to sensitive isolates (eg in B graminis f sp
tritici), will they survive in absence of fungicide
selection, and what is the relation between the
presence of G143A (and other SNPs) and tness of
the isolates under different environmental condi-
tions?
. Why are some individuals carrying G143A appar-
ently less t than sensitive individuals, as in P
viticola,
3
whilst in other species, such as B graminis,
4
there is little or no tness penalty for resistant
individuals?
. Is migration of spores or recurrent mutation the
major driving force for dispersal of resistance in eld
populations?
. Will different mutations such as G143A and F129L
and additional resistance mechanisms be combined
in the same individual after a certain time period of
selection and what will be the contributions of these
mechanisms to the overall resistance level?
7 CONCLUSIONS
A single point mutation (single nucleotide polymorph-
ism, SNP) in the cytochrome b gene causing a G143A
substitution in the cytochrome bc1 enzyme complex
confers resistance to QoIs in many plant pathogen
species. In some resistance cases, additional SNPs and
other resistance mechanisms to QoIs can exist (eg in P
aphanidermatum, V inaequalis). The evolution of
resistance to QoIs at the individual level will depend
mainly on early mitotic events; however, in popula-
tions the selection process will probably follow similar
routes as for fungicide classes with nuclear encoded
resistance. The dynamics of resistance evolution and
mitochondrial biology in the presence and absence of
QoIs is largely unknown and should be investigated in
more detail. Strict anti-resistance strategies, including
the limitation of treatments and the use of mixtures
and alternations, should be followed in order to
maintain the high efcacy of this class of fungicide as
long as possible.
ACKNOWLEDGEMENT
We wish to express our special thanks to U Mu ller and
P Crowley for helpful discussion during the prepara-
tion of this manuscript.
REFERENCES
1 Gisi U, Chin KM, Knapova G, Ku ng Farber R, Mohr U, Parisi
S, Sierotzki H and Steinfeld U, Recent developments in
elucidating modes of resistance to phenylamide, DMI and
strobilurin fungicides. Crop Protect 19:863872 (2000).
2 Becker WF, von Jagow G, Anke T and Steglich W, Oudemansin,
strobilurin a, strobilurin b, and myxothiazol: new inhibitors of
the bc1 segment of the respiratory chain with an e-b-methoxy-
acrylate system as common structural element. FEBS Lett
132:329333 (1981).
3 Heaney SP, Hall AA, Davies SA and Olaya G, Resistance to
fungicides in the QoI-STAR cross-resistance group: current
perspectives, in Proc Brighton Crop Protect ConfPests Dis,
BCPC, Farnham, Surrey, UK, pp 755762 (2000).
4 Chin KM, Chavaillaz D, Kasbohrer M, Staub T and Felsenstein
FG, Characterizing resistance risk of Erysiphe graminis f sp
tritici to strobilurins. Crop Protect 20:8796 (2001).
5 Steinfeld U, Sierotzki H, Parisi S and Gisi U, Comparison of
resistance mechanisms to strobilurin fungicides in Venturia
inaequalis, in Modern fungicides and antifungal compounds III, ed
by Lyr H, Russell PE, Dehne H-W, Gisi Uand Kuck KH, 13th
International Reinhardsbrunn Symposium, in press (2002).
6 Jordan DB, Livingston RS, Bisaha JJ, Duncan KE, Pember SO,
Picollelli MA, Schwartz RS, Sternberg JA and Tang X, Mode
of action of famoxadone, Pestic Sci 55:105118 (1999).
7 Sierotzki H, Parisi S, Steinfeld U, Tenzer I, Poirey S and Gisi U,
Mode of resistance to respiration inhibitors at the cytochrome
bc1 complex of Mycosphaerella jiensis. Pest Manag Sci 56:833
841 (2000).
8 Chin KM, Ku ng Farber R and Laird D, Aspects of fungicide
cross-resistance and implications for strobilurins, in Proc
Brighton Crop Protect ConfPests Dis, BCPC, Farnham, Surrey,
UK, pp 415420 (2000).
9 Gisi U, Chemical control of downy mildews, in Advances in
downy mildewresearch, ed by Spencer-Phillips PTN, Gisi Uand
Lebeda A, Kluwer, Dordrecht, The Netherlands, pp 119159
(2002).
10 Link TA, Haase U, Brandt U and von Jagow G, What infor-
mation do inhibitors provide about the structure of the hydro-
quinone oxidation site of ubihydroquinonecytochrome c
oxidoreductase? J Bioenerg Biomembr 25:221232 (1993).
11 Zhang Z, Huang L, Shulmeisters VM, Chi Y-I, Kim KK, Hung
L-W, Crofts AR, Berry EA and Kim S-H, Electron transfer by
domain movement in cytochrome bc1. Nature (London)
392:677684 (1998).
12 Iwata S, Lee JW, Okada K, Lee JK, Iwata M, Rasmussen B, Link
TA, Ramaswamy S and Jap BK, Complete structure of the
11-subunit bovine mitochondrial cytochrome bc1 complex.
Science (Washington) 281:6471 (1998).
13 Tamura H and Mizutani A, Mode of action of strobilurin
fungicides. Nihon Noyaku Gakkaishi (J Pestic Sci) 24:189196
(1999) [in Japanese].
14 Degli Esposti M, De Vries S, Crimi M, Ghelli A, Patarnello T
and Meyer A, Mitochondrial cytochrome b: evolution and
structure of the protein. Biochim Biophys Acta 1143:243271
(1993).
15 Brasseur G, Saribas AS and Daldal F, A compilation of muta-
tions located in the cytochrome b subunit of the bacterial and
866 Pest Manag Sci 58:859867 (online: 2002)
U Gisi et al
mitochondrial bc1 complex. Biochim Biophys Acta 1275:6169
(1996).
16 Zheng D and Ko ller W, Characterization of the mitochondrial
cytochrome b gene from Venturia inaequalis. Curr Genet
32:361366 (1997).
17 Sierotzki H, Wullschleger J and Gisi U, Point-mutation in
cytochrome b gene conferring resistance to strobilurin fungi-
cides in Erysiphe graminis f sp tritici eld isolates. Pestic Biochem
Physiol 68:107112 (2000).
18 Ishii H, Fraaije BA, Sugiyama T, Noguchi K, Nishimura K,
Takeda T, Amano T and Hollomon DW, Occurrence and
molecular characterization of strobilurin resistance in cucum-
ber powdery mildew and downy mildew. Phytopathology
91:11661171 (2001).
19 Farman ML, The molecular basis of eld resistance to QoI
fungicides in Pyricularia grisea. Phytopathology 91:110 (abstr)
(2001).
20 Wille P, Sierotzki H, Tenzer I and Gisi U, Molecular tools to
diagnose strobilurin resistance in Mycosphaerella jiensis,
Annales 6th Intern Conf Plant Dis, ANPP, Tours, France, pp
843850 (2000).
21 Wille P, Sierotzki H, Stanger C, Cleere S, Burbidge J, Hall A,
Windass J and Gisi U, Qualitative and quantitative identica-
tion of SNPs in plant pathogens, in Modern fungicides and
antifungal compounds III, ed by Lyr H, Russell PE, Dehne H-
W, Gisi U and Kuck KH, Bonn, 13th International Reinhards-
brunn Symposium, in press (2002).
22 Fraaije BA, Butters J and Hollomon DW, In planta genotyping of
Erysiphe graminis f sp tritici isolates for strobilurin-resistance
using a uorimetric allele-specic PCR assay, in Proc Brighton
Crop Protect ConfPests Dis, BCPC, Farnahm, Surrey, UK, pp
401406 (2000).
23 Fraaije BA, Butters JA, Coelho JM, Jones DR and Hollomon
DW, Following the dynamics of strobilurin resistance in
Blumeria graminis f sp tritici using quantitative allele-specic
real-time PCR measurements with the uorescent dye SYBR
Green I. Plant Pathol 51:4554 (2002).
24 Sierotzki H, Schlenzig A, Wullschleger J, Windass J, Stanger C,
Burbidge J, Cleere S, Hall Aand Gisi U, Cytochrome b gene in
fungi: phylogenetic relationships and a mutation for QoI
resistance, in Modern fungicides and antifungal compounds III, ed
by Lyr H, Russell PE, Dehne H-W, Gisi U and Kuck KH,
Bonn, 13th International Reinhardsbrunn Symposium, in press
(2002).
25 Clark-Walker GD, Evolution of mitochondrial genomes in fungi.
Internat Rev Cytol 141:89127 (1992).
26 McDonald BA, Zhan J, Yarden O, Hogan K, Garton J and
Pettway RE, The population genetics of Mycosphaerella
graminicola and Stagonospora nodorum, in Septoria on cereals: a
study of pathosystems, ed by Lucas JA, Bowyer P and Anderson
HM, CABI, Wallingford, UK, pp 4469 (1999).
27 Knapova G, Tenzer I, Gessler C and Gisi U, Characterization of
Phytophthora infestans from potato and tomato with molecular
markers, in Proc 5th Congr Europ Fed Plant Path EFPP,
Biodiversity in Plant Pathology, Giardini Naxos, Italy, pp 69
(2001).
28 Birky CW, Uniparental inheritance of mitochondrial and chloro-
plast genes: mechanisms and evolution. Proc Natl Acad Sci
USA 92:1133111338 (1995).
29 Robinson HL, Ridout CJ, Sierotzki H, Gisi U and Brown JKM,
Isogamous, hermaphroditic inheritance of mitochondrion-
encoded resistance to Qo inhibitor fungicides in Blumeria
graminis fsp. tritici. Fung Gen Biol 36:98106 (2002).
30 Ro hr H, Ku es U and Stahl U, Organelle DNA of plants and
fungi: inheritance and recombination. Progr Bot 60:3987
(1998).
Pest Manag Sci 58:859867 (online: 2002) 867
Evolution of resistance to Qo inhibitor fungicides