See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/202168524

Stability of Cytoplasmic Male Sterility in Maize under Different Environmental

Conditions

Article  in  Crop Science · January 2009

DOI: 10.2135/cropsci2007.12.0694

CITATIONS READS
50 2,449

7 authors, including:

Nikolai Kirilov Christov

Agrobioinstitute
52 PUBLICATIONS   581 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Chromosomal and DNA alterations in barley mutant lines produced by gamma-irradiation View project

Genetic variability in mutant barley lines induced by ionizing radiation View project

All content following this page was uploaded by Nikolai Kirilov Christov on 16 June 2016.

The user has requested enhancement of the downloaded file.


Stability of Cytoplasmic Male Sterility in Maize
under Different Environmental Conditions
Christophe Weider, Peter Stamp,* Nikolai Christov, Alexandra Hüsken,
Xavier Foueillassar, Karl-Heinz Camp, and Magali Munsch
ABSTRACT
Cytoplasmic male sterility (CMS) is a maternally
transmitted trait, whereby a plant is unable to
produce viable pollen. Studies have revealed
that this trait is a tool for enabling efficient and
reliable coexistence between genetically modi-
fied (GM) and non-GM cultivation by biocontain-
ment of GM maize (Zea mays L.) pollen. Maize
has three types of male-sterile cytoplasm (T, S,
and C), the fertility of which can be restored by
nuclear rf genes or by interactions with the envi-
ronment. Twenty-two CMS versions of modern
European maize hybrids were evaluated in 17
environments in Switzerland, France, and Bul-
garia, with two or three sowing dates, in 2005
and 2006. Stable and unstable male sterility
occurred in all three CMS types. T-cytoplasm
hybrids were the most stable, while S-cyto-
plasm hybrids often showed partial restoration
of fertility. C-cytoplasm was similar to T-cyto-
plasm with regard to maintaining male steril-
ity. Climatic factors, especially air temperature,
evapotranspiration, and water vapor, during the
10 d before anthesis as well as during anthesis,
were correlated positively or negatively with the
partial reversion to male fertility of CMS hybrids,
indicating an interaction between genetic and
climatic factors. This study illustrates that T- and
C-cytoplasm in particular open up viable pros-
pects for containing transgenic pollen, espe-
cially for Bt-maize.
CROP SCIENCE, VOL. 49, JANUARY– FEBRUARY 2009
RESEARCH
C. Weider and P. Stamp, ETH Zurich, Institute of Plant Sciences,
CH8092 Zurich, Switzerland; K.-H. Camp and M. Munsch, Delley
Seeds and Plants, CH1567 Delley, Switzerland; N. Christov, AgroBio-
Institute, Dragan Tzankov Blvd. 8, 1164 Sofi a, Bulgaria; A. Hüsken,
Biologische Bundesanstalt für Land- und Forstwirtschaft, Messeweg
11/12, D-38104 Braunschweig, Germany; X. Foueillassar, ARVALIS
Institut du Végétal, Chemin de Pau 21, F-64121 Montardon, France.
Received 7 Jan. 2008. *Corresponding author (peter.stamp@ipw.
agrl.ethz.ch).
Abbreviations: CMS, cytoplasmic male sterile/sterility; GM, geneti-
cally modified; MKS, mean kernel set; rf gene, fertility restorer gene.
C ytoplasmic male sterility (CMS) is defined as a form of male
sterility induced by the complementary action of nuclear and
cytoplasmic genes. Cytoplasmic male sterility, a natural trait, causes
a plant to produce nonfunctional pollen (Laser and Lersten, 1972;
Schnable and Wise, 1998). Specific mutations in mitochondrial DNA
(maternally inherited) are responsible for a dysfunction in the respira-
tory metabolism that can lead to deleterious phenotypes (Cilier et al.,
2004; Chase, 2007), resulting in the abnormal production of male
gametes (Budar et al., 2003). Therefore, plants either produce no
pollen or pollen that is not viable. However, female fertility is not
affected by CMS that is used in the breeding process, and male-
sterile plants can set seeds if viable pollen is available.
Maize (Zea mays L.) has three major types of male-sterile cyto-
plasm: T-cytoplasm of maize (Texas) (Rogers and Edwardson, 1952),
S-cytoplasm (USDA) (Jones, 1957) and C-cytoplasm (Charrua)
(Beckett, 1971). These three CMS types are defined according to
the specific nuclear restorer genes (rf genes) that are able to coun-
teract the CMS trait and restore fertility in the first-generation (F1)
Published in Crop Sci. 49:77–84 (2009).
doi: 10.2135/cropsci2007.12.0694
© Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
All rights reserved. No part of this periodical may be reproduced or transmitted in any
form or by any means, electronic or mechanical, including photocopying, recording,
or any information storage and retrieval system, without permission in writing from
the publisher. Permission for printing and for reprinting the material contained herein
has been obtained by the publisher.
77
progeny (Schnable and Wise, 1998). T-cytoplasm was the spontaneously to fertility and produce variable amounts of
most extensively used type of cytoplasm between the 1950s viable pollen.
and the 1970s because suitable restorer genotypes were found Cytoplasmic male sterility is a reliable solution to the
easily and there was no pollen. problem of coexistence between genetically modified (GM)
Two Rf alleles, Rf1 and Rf2 (Duvick, 1965), acting and non-GM maize because producers of non-GM maize
jointly, are essential to fully restore pollen fertility to CMS- are usually concerned about the potential risks of outcross-
T. These two Rf alleles are, however, unable to restore pollen ings by pollen from GM maize. Indeed, suitable transgenic
fertility to CMS-S and CMS-C. A single major allele (Rf3) CMS plants would not release pollen to the environment and,
is required for pollen fertility in the S-system (Duvick, 1965), therefore, would not fertilize non-GM adjacent plants. It is
while Rf4, Rf5, and Rf6 are involved in the restoration of fer- particularly important to avoid the release of pollen by GM
tility in the C-type (Kheyr-Pour et al., 1981; Gabay-Laugh- plants that are bred for the production of pharmaceuticals.
nan et al., 2004). Nevertheless, other unknown rf genes may Environmental influences on male sterility were noted
be involved (Gabay-Laughnan and Laughnan, 1994). It was for several plant species. In CMS cotton (Gossypium spp.),
estimated that 85% of hybrid maize grown in the United wind velocity, air temperature, global radiation, and pan
States in 1970 was produced with CMS-T (Ullstrup, 1972). evaporation appeared to influence, either positively or neg-
Breeders used CMS-T even though the CMS trait and sen- atively, the expression of male sterility 2 to 3 wk before
sitivity to the pathogen Bipolaris maydis were closely linked anthesis (Sarvella, 1966; Marshall et al., 1974). The effect
(Duvick, 1959; Levings, 1993). As a consequence, an epi- of different day/night temperature regimes on two CMS
demic of southern corn leaf blight occurred during the sum- systems of rape seed (Brassica napus L.) was investigated
mer of 1970, which led to the immediate decision to use (Fan and Stefansson, 1986). The lowest temperature regime
normal cytoplasm and/or S- and C-cytoplasm. However, the (22/16°C) resulted in the maintenance of sterility in both
two latter cytoplasms are not always reliable and may revert systems, while the highest temperature regime (30/24°C)
promoted anther development. High humid-
Table 1. Maize hybrids evaluated in 2005 and 2006 at two locations in Switzerland ity and low temperature were associated with
(Eschikon = CH-E, Delley = CH-D), two locations in France (Le Rheu = F-R, Cou- increases in the fertility of maize, whereas long
blucq = F-C), and one location in Bulgaria (B).† photoperiods at high temperature reduced
2005 2006 fertility (Duvick, 1965). Peterson and Foskett
Cytoplasm‡ Hybrid Breeder§ (1953) found that, like maize, CMS onions
CH-E CH-D F-R CH-E CH-D F-C F-R B
CMS-T DSP1 DSP ■ ■ ■ ■ ■ ■ ■ ■
(Allium cepa L.) had more mature pollen at
DSP2 DSP ■ ■ ■ ■ ■ ■ ■ ■
low temperature, probably due to a delay in
DSP4 DSP ■ ■ ■ ■ ■ ■ ■ ■
or absence of pollen abortion.
DSP5 DSP ■ ■ ■ ■ ■ ■ ■ ■
Environmental factors influence the ste-
DSP7 DSP ■ ■ ■ … … … … … rility patterns of various plant species. Our
DSP9 DSP ■ ■ ■ … … … … … research focused on the impact of the following
MAD1 Maïsadour ■ ■ ■ … … … … … climatic factors: air temperature (°C), cumu-
CMS-C LIM1 Limagrain ■ ■ ■ ■ ■ ■ ■ ■
lative temperature (°C), rainfall (mm), global
LIM4 Limagrain ■ ■ ■ ■ ■ ■ ■ ■
radiation (Wh m–2), water vapor (%), and
LIM3 Limagrain ■ ■ ■ ■ ■ ■ ■ ■
evapotranspiration (mm d–1) on the male-ster-
EUR1 Euralis ■ ■ ■ … … … … … ile expression in CMS maize. The objectives
DSP10 DSP ■ ■ ■ … … … … … of this study were (i) to investigate the stability
LIM6 Limagrain ■ ■ ■ … … … … … of the male-sterile trait of a large number of
LIM5 Limagrain ■ ■ ■ … … … … … modern European CMS hybrids under vari-
MAD2 Maïsadour ■ ■ ■ … … … … … ous conditions in the field; (ii) to ascertain the
CMS-S KWS1 KWS ■ ■ ■ ■ ■ ■ ■ ■
environmental and genetic impacts of spon-
DSP5 DSP ■ ■ ■ ■ ■ ■ ■ ■
taneous reversion to fertility; and (iii) to pro-
DSP6 DSP ■ ■ ■ ■ ■ ■ ■ ■
vide an efficient field-adapted system for using
DSP8 DSP ■ ■ ■ ■ ■ ■ ■ ■
CMS as a reliable tool to enable the coexis-
SYN1 Syngenta ■ ■ ■ ■ ■ ■ ■ ■
tence of GM and non-GM maize.
SYN2 Syngenta ■ ■ ■ ■ ■ ■ ■ ■
KWS2 KWS ■ ■ ■ ■ ■ ■ ■ ■ MATERIALS AND METHODS
■, …: tested or not tested at this location, respectively.
† Plant Material and
‡
Cytoplasmic male-sterile (CMS) type T, S, or C. Experimental Design
§
DSP, Delley Seeds and Plants, Delley, Switzerland; Maïsadour, Maïsadour Semences SA, Mont De Marsan, Twenty-two early- to late-maturing modern
France; Limagrain, Limagrain Genetics, Riom, France; Euralis, Euralis Saaten GmbH, Norderstedt, Ger-
many; KWS, KWS Saat AG, Einbeck, Germany; Syngenta, Syngenta Seeds S.A.S, Saint-Sauveur, France. cytoplasmic male-sterile dent-flint hybrids from

78 WWW.CROPS.ORG CROP SCIENCE, VOL. 49, JANUARY– FEBRUARY 2009

different European breeders and tested in 2005 and Table 2. Sowing dates at all locations in 2005 and 2006.
2006 under field conditions were used in this study Switzerland France Bulgaria
Sowing
(Table 1). These hybrids are recommended for grain Year date
Eschikon Delley Coublucq Le Rheu Slavianovo
maize production in the midlands of Switzerland, i.e.,
2005 1 30 Apr. 2005 10 May 2005 7 May 2005
in a moderate maritime climate. The three types of
2 15 May 2005 28 May 2005
cytoplasm (S, C, and T) with different hybrid nuclear
backgrounds were investigated for the stability of their 3 29 May 2005 12 June 2005
male-sterile trait using seven CMS-T, eight CMS-C, 2006 1 3 May 2006 3 May 2006 5 May 2006 11 May 2006 5 May 2006
and seven CMS-S hybrids. 2 7 June 2006 18 May 2006 17 May 2006 23 May 2006 19 May 2006
The CMS hybrids were grown in a random-
ized complete block design with five replicates, each
similar to fertile plants. To determine whether or not viable pollen
arranged into subplots, each of which contained one CMS hybrid
was released, self-pollination was performed from 0900 to 1100
in one row, 6 m long. The single-row subplots, 0.80 m apart,
h, that is, the optimal time frame (Jarosz et al., 2003). Before silk-
were sown randomly inside each block at plant density of 6.25
–2 ing, the young ear shoots were isolated by means of plastic bags
m . Two rows of a commercial hybrid (Goldenso) were sown at
to avoid contamination by foreign pollen. On the day preceding
the edges to avoid border effects. Fertilization and plant protec-
self-pollination, the silks were shortened (1 cm) to optimize acces-
tion measures were performed according to local recommenda-
–1 sibility to the pollen and the tassels were put into paper bags to
tions to give a grain yield of 10 t ha .
collect the pollen. Self-pollination was performed according to the
method used by the breeders. Fifteen plants per hybrid and per
Site Description replication were assessed. Thus, 75 plants per hybrid, per sowing
In 2005, the experiments were conducted at Eschikon (canton date, and per location were tested for male sterility. Furthermore,
Zurich) in northeastern Switzerland (Eutric Cambisol), at Del- tests of pollen germination were performed at the same time as
ley (canton Fribourg) in western Switzerland (Chromic Luvi- the self-pollination in Environments 1 and 2 in Eschikon. In both
sol), and at Le Rheu (Brittany) in northwestern France (Humic environments, fresh pollen was gathered from 10 individuals of
Cambisol). In 2006, field trials were conducted at the same hybrids DSP4, DSP6, and SYN2.
locations as well as at two other locations: in Slavianovo (Pleven At each location, daily meteorological data (air temperatures
Province) in western Bulgaria (Chernozem) and in Coublucq [mean, min., and max.; °C], cumulative temperature [°C], rainfall
(Aquitaine) in southwestern France (clay loam Cambisol). [mm], water vapor [%], global radiation [Wh m–2], and evapotrans-
Eschikon lies between a western maritime temperate cli- piration [°C]) were assessed. Meteorological stations from METEO
mate and an eastern continental climate. This location is the only France, AGROMETEO Suisse, and PLEVEN Bulgaria provided
one that is considered to be marginal for grain maize production. the daily measurements in both seasons at the three sites.
Delley has a semicontinental climate. Le Rheu’s climate is similar
to an oceanic climate with little variation in the mean daily tem-
Harvesting and Data Analysis
perature; this climate is dominated year-round by a polar front,
All self-pollinated ears per replication were harvested. Total kernel
leading to changeable and often overcast weather. Coublucq has
number was assessed because kernel production indicates viable
an attenuated oceanic climate with mild humid conditions and
pollen. The percentage of seed set was determined for each ear
low temperature amplitudes. Slavianovo has a moderate conti-
relative to a mean kernel set of 10 fully pollinated ears (hand-pol-
nental climate with high temperature amplitudes.
linated with viable pollen collected from fertile plants) for each
To increase variation in the environmental conditions,
CMS hybrid. The statistical analysis was conducted using a gen-
the experiments were sown on one to three dates per location,
eralized linear mixed model (PROC GLIMMIX in SAS 9.1.3
about 2 wk apart (Table 2). Each sowing date–location combi-
[SAS Institute, 2002]):
nation was considered to be an environment.
Yijk = μ + α i + β j + (α β)ij + γ k( j) + ε ijk
Control of Male Sterility and
Investigated Parameters
A molecular analysis consisting in a rapid, efficient, and inex-
pensive single-seed multiplex polymerase chain reaction assay
(Liu et al., 2002) was applied to 30 kernels per genotype to
determine or to confi rm the three major types of CMS and
to ensure the quality of the seed sets provided by the breed-
ers. Observations were performed during the flowering period
which extended from July to August in both years, depend-
ing on the location and the sowing date. Tassels were grouped
according to one of three classes (Fig. 1).
Class A includes all hybrids, which did not release anthers and Figure 1. Classification of the cytoplasmic male-sterile hybrid maize
therefore did not release pollen. Classes B and C include all hybrids, tassels. (A) Sterile tassel without anthers. (B) Partially restored tassel
with variable number of anthers released. (C) Fertile tassel with
which may release pollen: partially restored, releasing anthers in
anthers on all the branches. Source: ARVALIS Institut du Végétal,
a variable range; and fertile, morphologically and physiologically
Paris, France.

CROP SCIENCE, VOL. 49, JANUARY– FEBRUARY 2009 WWW.CROPS.ORG 79


Figure 2. Average cytoplasmic male sterility (CMS) in maize of 22 S-, C- and T-cytoplasms
tested in 17 environments. †Number of hybrids per class.
where μ is the overall mean, α i the effect of the i th CMS geno-
type, β j the effect of the j th environment, and (α β)ij the effect of
the ij th CMS genotype by environment interaction. γ k( j) is the
effect of the kth repetition within the j th environment, and εijk is
the error term. For the latter two parameters we assumed that
γ k( j) ~ N(0, σ2γ ) and εijk ~ N(0, σ2 ε).
We used the restricted maximum likelihood estimation tech-
nique. Approximate degrees of freedom for the Type III test of
fixed effects were calculated according to the Satterthwaite for-
mula. The first objective was to identify a genotype × environment
interaction. To determine the environmental effect on expression
of male sterility of the different CMS hybrids, we compared the
kernel percentage of each genotype in the various environments by
a pairwise comparison of the Tukey-adjusted least square means.
When there were significant differences in kernel set
(P < 0.05) among the various environments, tests were conducted
to determine whether one or more climatic factors, characteristic
of the environment, influenced the male sterility of each hybrid,
either positively or negatively. Therefore, correlations between
the percentage of kernel set and each of the climatic factors were
analyzed according to the Spearman’s rank correlation coefficient
(CORR procedure, spearman option, SAS 9.1.3), which does not
require normally distributed variables. Two key growth stages
were found: preanthesis and anthesis. For anthesis, the average of
daily climatic data over a 6-d period (starting at flowering) was
calculated. The preanthesis period was split into three periods: a
10-d period before flowering, a 5-d period right before flower-
ing, and a 5-d period 5 d before flowering. For each period and
each factor, the average daily climatic variables were calculated.
RESULTS
Molecular Analysis and
Patterns of Male Sterility
The molecular analyses showed that all seeds were of the
CMS type, as rated by the breeders; the seed sets were not
contaminated. The average male sterility of 22 S-, C-, and
80 WWW.CROPS.ORG
T-cytoplasms, tested in 17 environments,
showed that cytoplasm S had the least reli-
able male sterility; six of seven hybrids
showed unstable male sterility (Fig. 2).
Only one S-cytoplasm was fully ster-
ile. Cytoplasms T and C were the most
reliable; both included five hybrids that
were fully sterile across all environments
(Table 3). In the T-cytoplasm, two hybrids
had partially restored male sterility. In the
C-cytoplasm, one hybrid showed inter-
mediate male sterility, while two hybrids
showed fully restored male sterility.
Partially Restored Hybrids:
A Heterogeneous Class
Among all the tested hybrids, nine hybrids
(41%) showed variable expression of
male sterility. Cytoplasm S was the most
affected by reversion to partial restoration
of fertility. However, the S and T types did not show full
reversion to fertility. The class of partially restored hybrids
was very heterogeneous and included two main groups.
The first group of five hybrids (KWS1, DSP5, DSP6,
SYN2, and DSP4) was homogeneous in all environments.
The tassels of these hybrids released various numbers of
anthers and viable pollen was released. The other group
with four hybrids (SYN1, DSP8, LIM3, and DSP2) usu-
ally contained only a few individual plants with partial
male sterility, thus enabling viable pollen to be produced.
The genotype × environment interaction was highly sig-
nificant (P < 0.001). The expression of male sterility of
the hybrids was influenced by the environment. However,
a partially restored hybrid expressed different degrees of
male fertility in the various environments.
Self-pollination was performed at the beginning of
male flowering. Therefore, the number of kernels devel-
oped on the ears provided objective and complementary
information about the potential amount of viable pollen
that was released at this time. Figure 3 presents the mean
kernel set (MKS) per ear in all environments for each of
the partially restored hybrids.
DSP4, DSP6, and SYN2 hybrids were most affected
by reversion to partial fertility. Although the tassels of
these hybrids were not morphologically comparable
to those of fertile plants, they released large amounts
of anthers and released a significant quantity of viable
pollen in some environments. High MKS per ear was
reached in some environments. However, only a few
pollen grains are necessary to reach full seed set. Tests
of pollen germination indicated that the rate of viability
was about 71% for DSP4, 52% for DSP6, and 36% for
SYN2, whereas the rate of viability of a fertile individual
ranged from 80 to 90%.
CROP SCIENCE, VOL. 49, JANUARY– FEBRUARY 2009
 Although KWS1 released a large number
of anthers in all environments, this hybrid
as well as DSP2, LIM3, and DSP8 produced
few kernels. The MKS per ear was 0.5% for
KWS1, 2.0% for DSP2, 1.1% for LIM3, and
2.5% for DSP8. DSP2 produced significantly
more kernels in two environments (P < 0.05)
and DSP8 produced significantly more kernels
in one environment (P < 0.001). The MKS
per ear of the other two hybrids was compa-
rable across all environments.

Relationship between Partial

Reversion to Fertility and
Climatic Parameters
The climatic parameters measured during con-
secutive 5-d periods before anthesis did not
provide additional information compared to
the cumulative 10-d period. Therefore, only
the results from the 10-d period are presented
with regard to preanthesis climatic conditions.
The climatic factors did not affect the expres- Figure 3. Mean kernel set per ear of DSP4, DSP2, LIM3, DSP6, SYN2, SYN1,
sion of male sterility of SYN2 and DSP5. No DSP5, DSP8, and KWS1 cytoplasmic male-sterile (CMS) maize hybrids of S-,
C- and T-cytoplasms across all environments. The box defined between the first
significant correlations were found for these
quartile (Q1) and the third quartile (Q3) represents 50% of the environments.
two hybrids between the percentage of MKS The line in the box indicates the median value. Any data observation which lies
and the different climatic factors. However, for more than 1.5-IQR (the interquartile range [IQR] = Q3 − Q1) lower than the first
the other seven hybrids, there were correlations quartile or 1.5-IQR higher than the third quartile is considered to be an outlier
between MKS and the different climatic fac- (bullet point). The lowest and highest value that is not an outlier is indicated by
tors (Table 4), especially with the temperature connecting it to the box with a horizontal line or “whisker.”
parameters, water vapor, and rainfall.
There was no general rule that controlled the interac- male fertility only a few times across all environments,
tion between the different climatic factors and the expres- while the expression of male sterility of DSP4 was simi-
sion of male sterility. Each partially restored hybrid was larly unstable across all environments. No main rf genes
sensitive to one or more factors during the two investi- seemed to be responsible for this phenomenon. In plants
gated growth periods. There was a negative correlation carrying T-cytoplasm, the complementary expression of
between preanthesis male fertility and/or male fertility dominant alleles of both the rf1 and rf2 nuclear restorer
during anthesis with the temperature parameters for DSP4, loci allows normal pollen development (Laughnan and
DSP6, and DSP2; that is, when temperature increased, the Gabay-Laughnan, 1983). Dominant alleles of at least two
MKS percentage per ear decreased. In contrast, there was
a positive correlation between temperature and MKS for Table 3. Fully restored and completely sterile maize hybrids
SYN1, DSP8, KWS1, and LIM3. across all environments.
Sterile behavior Fertile behavior
DISCUSSION Name Cytoplasm Name Cytoplasm
The coexistence of GM and non-GM maize is an impor- DSP1 T DSP10† C
tant issue in Europe, but there are concerns as well in MAD1† T LIM6† C
other parts of the world. For the production of pharma- DSP9† T
ceuticals from GM maize, it is mandatory that pollen flow DSP5 T
is contained. Cytoplasmic male sterility is the only natural DSP7† T
way to prevent the release of GM pollen and to enable the LIM5† C
coexistence between GM and non-GM farming systems. EUR1† C
In this study, all three types of cytoplasm provided sterile LIM1 C
hybrids. Cytoplasm T had the most reliable male steril- MAD2† C
ity; only two of seven hybrids, DSP2 and DSP4, showed KWS2 S
unstable behavior. DSP2 male-sterile hybrids reverted to †
MAD1, DSP9, DSP7, LIM5, EUR1, MAD2, DSP10, and LIM6, tested only in 2005.

CROP SCIENCE, VOL. 49, JANUARY– FEBRUARY 2009 WWW.CROPS.ORG 81

Table 4. Spearman rank correlation coefficients between the percent-
that of S- and T-cytoplasm. The restoration
age of mean kernel set per maize ear and the climatic factors across all
environments. system is still unclear, and the number of rf
genes involved have not been defi ned ( Joseph-
Climatic Hybrid
Period son and Arnold, 1978; Kheyr-Pour et al.,
factor † DSP4 DSP2 SYN1 DSP8 DSP6 KWS1 LIM3
1981; Gabay-Laughnan and Laughnan, 1994;
Tmean –0.73** ns‡ ns 0.50§ ns ns ns
Hu et al., 2006). In this study, five of the eight
Tmin –0.62* ns ns 0.57* ns ns 0.52§
hybrids tested were fully sterile. This indicates
Tmax –0.57* ns 0.50§ ns ns 0.59* ns
that the European breeding pool is still rela-
Preanthesis R 0.72** ns ns –0.68** ns ns ns
tively free of rf genes. Two hybrids were fully
GR ns ns 0.72** ns ns ns 0.52§
restored, which clearly indicates the presence
WV –0.68** ns ns ns ns ns 0.72**
of one or more major Rf alleles. Male steril-
ETP ns ns –0.56* ns ns –0.56* ns
ity of LIM3 was unstable in a few plants only.
Tmean –0.75** –0.59* ns ns –0.056* 0.55* ns
In this cytoplasm, partial restoration of fertil-
Tmin –0.64* –0.52§ 0.65* ns ns 0.72** ns
ity linked to minor rf genes is known to be
Anthesis Tmax –0.73** –0.61* ns ns –0.54* 0.58* ns
highly influenced by the environment (Tracy
R 0.52§ ns ns ns ns ns ns
et al., 1991). In this study, some correlations
WV –0.59* ns ns ns ns ns 0.59*
between climatic factors and the partial rever-
*Significant at 0.05 probability level.
sion to male fertility of LIM3 were found,
**Significant at 0.01 probability level.
†
particularly with water vapor.
Climatic factors: air temperature (Tmean, Tmin, and Tmax) (°C); rainfall (R) (mm); global radiation (GR) (Wh
m –2); water vapor (WV); evapotranspiration (ETP) (°C). For S-cytoplasm, the situation is more
‡
ns, not significant. complex, probably due to a large number of
§
Tendency at the 0.1 probability level. alleles at the Rf3 locus but also due to spon-
taneous nuclear and cytoplasmic mutations at
other genes, rf8 and rf*, can substitute for the Rf1 allele loci other than Rf3, all of which condition restoration of
to partially restore fertility in the presence of Rf2 (Dill et fertility (Gabay-Laughnan et al., 1995). A second major
al., 1997). These two alleles are rare in maize germplasm restorer gene has been characterized recently (K.J. New-
and their expression is affected by the environment. In ton and S. Gabay-Laughnan, personal communication,
DSP2 CMS-T, only a few individuals were affected by 2007). A dominant allele at this locus has been officially
reversion to partial fertility; thus, it was hypothesized that designated Rf9 and is environmentally sensitive. This
minor Rf alleles may be involved. The case of DSP4 is dif- genetic instability of S-cytoplasm, coupled with its sen-
ferent because this hybrid showed similar morphological sitivity to environmental conditions, makes the S system
behavior across all environments, and the implication of generally unreliable for hybrid seed production and for
an environmentally sensitive restorer allele, such as Rf8, is coexistence as well. The correlations found between cli-
consistent. Indeed, tassels of these partially restored plants matic factors and the MKS of several S hybrids clearly
released variable but always lower numbers of anthers that indicate that their restorer genes are sensitive to the
produced different amounts of viable pollen. Further- environment. Under very specific conditions, an rf gene
more, their anthesis-silking interval was short and anthers might be activated or inactivated, which could explain
were often released 2 to 5 d after silk emergence rather the extent of heterogeneity in grain set, especially in the
than 2 to 3 d before silk emergence, which is more usual case of DSP6 and SYN2. After germination tests, the rate
(Tracy et al., 1991). Similar behavior was reported by Dill of pollen vitality of the two latter hybrids was around 50
et al.(1997) about CMS-T lines, inbreds, or hybrids car- and 40%, respectively, which is in agreement with the
rying the dominant allele of the rf8 gene. Because the theoretical value of 50% vitality expected for a game-
restoration system by T-cytoplasm is sporophytic (Dill et tophytic restoration system (Wise et al., 1999). Stable S
al., 1997), all existing pollen is viable. Pollen germina- hybrids, such as KWS2, exist and breeders should be able
tion tests in Environments 1 and 2 indicated that the pol- to produce S hybrids with stable genetic backgrounds.
len viability rate of DSP4 was 75%, whereas viability rate Different backgrounds may be required for different
for a fertile individual ranged between 80 and 90%. This growing regions. More information is required about the
high rate of pollen viability resulted in a high MKS in exact number of rf genes involved and about the different
both environments. The important heterogeneity of MKS alleles and their sensitivity to environmental factors to
definitely reflects the sensitivity of DSP4 to environmen- develop an efficient and simple molecular tool to identify
tal conditions, which was illustrated by the correlations the genes. Many of the tested S hybrids (DSP8, KWS1,
found, especially with temperature parameters. SYN1, and DSP5) produced only a very small amount of
Many breeders focus on C-cytoplasm. The stability viable pollen. On average across all environments, <5%
of male sterility in this group is known to be between kernel set was obtained. Therefore, it is possible to use

82 WWW.CROPS.ORG CROP SCIENCE, VOL. 49, JANUARY– FEBRUARY 2009

this class of hybrids to achieve coexistence because pol-
len pressure would be drastically reduced compared to
normal fertile individuals.
In Europe, an attempt to control coexistence between
GM and non-GM crops has been made by relying on dif-
ferent, often very long isolation distances; 25 to 50 m may
be sufficient, however, to achieve the controversial thresh-
old of 0.9%, as fi xed by the European Union (Bannert and
Stamp, 2007). Cytoplasmic male sterility is an elegant way
to minimize or even eliminate the problem of GM pollen
flow of adjacent conventional or organic fields if stable T-
and C-cytoplasm is used. The Plus-Hybrid System (Feil
and Stamp, 2002; Feil et al., 2003), developed at ETH
Zurich, would be a useful tool to achieve an agricultural
biocontainment system. It consists of growing 80:20%
mixtures of cytoplasmic male-sterile GM hybrids and
male-fertile non-GM hybrids, whereby the latter compo-
nent acts as pollen donor for the entire stand. For this sys-
tem, a high level of male sterility must be maintained, as
shown by this study. Furthermore, appropriate combina-
tions of CMS hybrids and fertile pollinators could lead to
a significant gain in yield (Stamp et al., 2000; Weingartner
et al., 2002). The confirmation of a gain in yield or at least
a stable yield would definitely boost the acceptance of a
biocontainment system with cytoplasmic male sterility.
Acknowledgments
Gratitude is expressed to the European Commission, which
gave fi nancial support to this study through the Integrated
Project “Co-Extra” (contract no. 007158), within the 6th
Framework Program “Priority 5”: Food Quality and Safety.
We thank all breeders who delivered CMS hybrids and showed
interest in our study.
References
Bannert, M., and P. Stamp. 2007. Cross-pollination of maize at
long distance. Eur. J. Agron. 27:44–51.
Beckett, J.B. 1971. Classification of male-sterile cytoplasms in
maize (Zea mays L.). Crop Sci. 11:724–727.
Budar, F., P. Touzet, and R. De Paepe. 2003. The nucleo-mito-
chondrial confl ict in cytoplasmic male sterilities revisited.
Genetica 117:3–16.
Chase, C.D. 2007. Cytoplasmic male sterility: A window to the
world of plant mitochondrial–nuclear interactions. Trends
Genet. 23:81–90.
Cilier, M., D. Feruzan, and O. Göksel. 2004. Histological aspects
of anther wall in male fertile and cytoplasmic male sterile
Helianthus annuus L. Asian J. Plant Sci. 3:145–150.
Dill, C.L., R.P. Wise, and P.S. Schnable. 1997. Rf8 and Rf*
mediate unique T-urf13-transcript accumulation, revealing
a conserved motif associated with RNA processing and res-
toration of pollen fertility in T-cytoplasm maize. Genetics
147:1367–1379.
Duvick, D.N. 1959. The use of cytoplasmic male sterility in hybrid
seed production. Econ. Bot. 13:167–195.
Duvick, D.N. 1965. Cytoplasmic pollen sterility in corn. Adv.
Genet. 13:1–56.
CROP SCIENCE, VOL. 49, JANUARY– FEBRUARY 2009 WWW.CROPS.ORG
Fan, Z.G., and B.R. Stefansson. 1986. Influence of temperature on
sterility of 2 cytoplasmic male-sterility systems in rape (Brassica
napus L.). Can. J. Plant Sci. 66:221–227.
Feil, B., and P. Stamp. 2002. The pollen-mediated flow of trans-
genes in maize can already be controlled by cytoplasmic male
sterility. Available at http://www.agbiotechnet.com/Reviews.
asp?action=display&openMenu=SiteSearch&ReviewID=654
AgBiotechNet (verified 28 Oct. 2008). AgBiotechNet 4.
Feil, B., U. Weingartner, and P. Stamp. 2003. Controlling the
release of pollen from genetically modified maize and increas-
ing its grain yield by growing mixtures of male-sterile and
male-fertile plants. Euphytica 130:163–165.
Gabay-Laughnan, S., C.D. Chase, V.M. Ortega, and L. Zhao. 2004.
Molecular-genetic characterization of CMS-S restorer-of-fer-
tility alleles identified in Mexican maize and teosinte. Genetics
166:959–970.
Gabay-Laughnan, S., and J.R. Laughnan. 1994. The male sterility
and restorer genes in maize. p. 418–423. In M. Freeling and V.
Walbot (ed.) The maize handbook. Springer, New York.
Gabay-Laughnan, S., G. Zabala, and J.R. Laughnan. 1995. S-type
cytoplasmic male sterility in maize. p. 395–432. In I. K. Vasil
and C.S. Levings III (ed.) Advances in cellular and molecular
biology of plants. Vol. 2. Kluwer, Dordrecht, the Netherlands.
Hu, Y.M., J.H. Tang, H. Yang, H.L. Xie, X.M. Lu, J.H. Niu, and
W.C. Chen. 2006. Identification and mapping of Rf-I an inhib-
itor of the Rf5 restorer gene for Cms-C in maize (Zea mays L.).
Theor. Appl. Genet. 113:357–360.
Jarosz, N., B. Loubet, B. Durand, A. McCartney, X. Foueillassar,
and L. Huber. 2003. Field measurements of airborne concen-
tration and deposition of maize pollen. Agric. For. Meteorol.
119:37–51.
Jones, D.F., H.T.J. Stinson, and U. Khoo. 1957. Pollen restoring
genes. Connecticut Agric. Exp. Stn. Bull. 610, New Haven.
Josephson, L.M., and M.T. Arnold. 1978. Genetics and inheritance
of fertility restoration of male sterile cytoplasms in corn. Proc.
Annu. Corn Sorghum Res. Conf. 33:48–59.
Kheyr-Pour, A., V.E. Gracen, and H.L. Everett. 1981. Genetics of
fertility restoration in the C-group of cytoplasmic male sterility
in maize. Genetics 98:379–388.
Laser, K.D., and N.R. Lersten. 1972. Anatomy and cytology of
microsporogenesis in cytoplasmic male sterile angiosperms.
Bot. Rev. 38:425–454.
Laughnan, J.R., and S. Gabay-Laughnan. 1983. Cytoplasmic male
sterility in maize. Annu. Rev. Genet. 17:27–48.
Levings, C.S. 1993. Thoughts on cytoplasmic male sterility in
cms-T maize. Plant Cell 5:1285–1290.
Liu, Z., S.O. Peter, M. Long, U. Weingartner, P. Stamp, and O.
Kaeser. 2002. A PCR assay for rapid discrimination of sterile
cytoplasm types in maize. Crop Sci. 42:566–569.
Marshall, D.R., N.J. Thomson, G.H. Nicholls, and C.M. Patrick.
1974. Effects of temperature and day length on cytoplasmic male
sterility in cotton (Gossypium). Aust. J. Agric. Res. 25:443–450.
Peterson, C.E., and R.L. Foskett. 1953. Occurrence of pollen steril-
ity in seed fields of Scott County Globe onions. Proc. Am. Soc.
Hortic. Sci. 62:443–448.
Rogers, J.S., and J.R. Edwardson. 1952. The utilization of cyto-
plasmic male-sterile inbreds in the production of corn hybrids.
Agron. J. 44:8–13.
Sarvella, P. 1966. Environmental influences on sterility in cytoplas-
mic male-sterile cottons. Crop Sci. 6:361–364.
SAS Institute. 2002. SAS user’s guide. Release 9.1.3. SAS Inst.,
Cary, NC.
83
Schnable, P.S., and R.P. Wise. 1998. The molecular basis of cyto-
plasmic male sterility and fertility restoration. Trends Plant Sci.
3:175–180.
Stamp, P., S. Chowchong, M. Menzi, U. Weingartner, and O.
Kaeser. 2000. Increase in the yield of cytoplasmic male sterile
maize revisited. Crop Sci. 40:1586–1587.
Tracy, W.F., H.L. Evrett, and V.E. Gracen. 1991. Inheritance, envi-
ronmental effects, and partial male fertility in C-type CMS in
a maize inbred. J. Hered. 82:343–346.
84 WWW.CROPS.ORG
View publication stats
Ullstrup, A.J. 1972. The impacts of southern corn leaf blight epi-
demics of 1970–1971. Annu. Rev. Phytopathol. 10:37–50.
Weingartner, U., O. Kaeser, M. Long, and P. Stamp. 2002. Com-
bining cytoplasmic male sterility and xenia increases grain yield
of maize hybrids. Crop Sci. 42:1848–1856.
Wise, R.P., K. Gobelman-Werner, D. Pei, C.L. Dill, and P.S.
Schnable. 1999. Mitochondrial transcript processing and
restoration of male fertility in T-cytoplasm maize. J. Hered.
90:380–385.
CROP SCIENCE, VOL. 49, JANUARY– FEBRUARY 2009