Environmental Risk Assessment of Genetically Engineered ...
Transcript of Environmental Risk Assessment of Genetically Engineered ...
207
Herbicide-tolerant Zoysia grass (Zoysia japonica Steud.) has been generated previously through Agrobacterium tumefaciens-mediated transformation. Th e genetically modifi ed (GM) Zoysia grass survived Basta spraying and grew to maturity normally while the wild-type (WT) grass stopped growing and died. GM Zoysia grass will permit more effi cient weed control for various turf grass plantings such as home lawns, golf courses, and parks. We examined the environmental/biodiversity risks of herbicide-tolerant GM Zoysia before applying to regulatory agencies for approval for commercial release. Th e GM and WT Zoysia grass’ substantial trait equivalence, ability to cross-pollinate, and gene fl ow in confi ned and unconfi ned test fi elds were selectively analyzed for environmental/biodiversity eff ects. No diff erence between GM and WT Zoysia grass in substantial traits was found. To assess the potential for cross-pollination and gene fl ow, a non-selective herbicide, Basta, was used. Results showed that unintended cross-pollination with and gene fl ow from GM Zoysia grass were not detected in neighboring weed species examined, but were observed in WT Zoysia grass (on average, 6% at proximity, 1.2% at a distance of 0.5 m and 0.12% at a radius of 3 m, and 0% at distances over 3 m). On the basis of these initial studies, we conclude that the GM Zoysia grass generated in our laboratory and tested in the Nam Jeju County fi eld does not appear to pose a signifi cant risk when cultivated outside of test fi elds.
Environmental Risk Assessment of Genetically Engineered Herbicide-Tolerant Zoysia japonica
T. W. Bae Cheju National University
E. Vanjildorj and S. Y. Song Chungnam National University
S. Nishiguchi Cheju National University
S. S. Yang Chonnam National University
I. J. Song, T. Chandrasekhar and T. W. Kang Cheju National University
J. I. Kim Chonnam National University
Y. J. Koh Sunchon National University
S. Y. Park Cheju Halla College
J. Lee Cheju National University
Y.-E. Lee Dongguk University
K. H. Ryu Seoul Women’s University
K. Z. Riu, P.-S. Song,* and H. Y. Lee Cheju National University
Turf grasses are commercially important species. As a perennial
monocot species, Zoysia grass is one of the most popularly
cultivated grasses for sports and recreational environments,
particularly in East Asia, because of its relatively high drought
tolerance, disease tolerance, and relatively slow growth habit. To
further improve the turf grass through plant biotechnology, the
transformation of this species (Inokuma et al., 1998; Toyama et
al., 2002; Ge et al., 2006; Li et al., 2006) has been investigated as a
prerequisite for the generation of several transgenic lines including
herbicide-tolerant grass (Toyama et al., 2003).
In a continuing eff ort to realize the biotechnology-based agro-
nomic potential of turf grass, we investigated (Toyama et al., 2003)
the herbicide tolerance of Zoysia grass by introducing a bar gene that
codes for phosphinothricin N-acetyltransferase (PAT) (Th ompson
et al., 1987) which catalyzes acetylation of the amino group of phos-
phinotricin (phosphinothricyl-L-alanyl-L-alanine). Th e N-acetylated
peptide can no longer inhibit the key enzyme in the nitrogen as-
similation pathway, glutamine synthetase (Bayer et al., 1972). Th e
bar gene confers tolerance to the broad-spectrum glufosinate-based
herbicide Basta in transgenic crops. Glufosinate is not only a non-se-
lective herbicide, but it is also quite readily biodegraded under natu-
ral conditions. Th us, we consider Basta as the herbicide of choice
Abbreviations: GM, genetically modifi ed; PAT, phosphinothricin N-acetyltransferase;
WT, wild type.
T.W. Bae, S. Nishiguchi, I.J. Song, T. Chandrasekhar, K.Z. Riu, P.-S. Song, and H.Y. Lee,
Faculty of Biotechnology, Cheju National Univ., Jeju 690-756, Korea. E. Vanjildorj and
S.Y. Song, Dep. of Horticulture, Chungnam National Univ., Daejeon 305-764, Korea. S.S.
Yang and J.I. Kim, Dep. of Biotechnology (BK21 Program) and Kumho Life Science Lab.,
Chonnam National Univ., Gwangju 500-757, Korea. T.W. Kang, Applied Radiological
Science Research Inst., Cheju National Univ., Jeju 690-756, Korea. Y.J. Koh, School of
Environmental and Agricultural Science, Sunchon National Univ., Sunchon 540-742,
Korea. S.Y. Park, Dep. of Clinical Pathology, Cheju Halla College, Jeju 690-708, Korea.
J. Lee, School of Medicine, Cheju National Univ., Jeju 690-756, Korea. Y.E. Lee, Dep. of
Biotechnology, Dongguk Univ., Kyungju, Kyongbuk 780-714, Korea. K.H. Ryu, Division of
Environmental and Life Sciences, Seoul Women’s Univ., Seoul 139-774, Korea.
Copyright © 2008 by the American Society of Agronomy, Crop Science
Society of America, and Soil Science Society of America. All rights
reserved. No part of this periodical may be reproduced or transmitted
in any form or by any means, electronic or mechanical, including pho-
tocopying, recording, or any information storage and retrieval system,
without permission in writing from the publisher.
Published in J. Environ. Qual. 37:207–218 (2008).
doi:10.2134/jeq2007.0128
Received 13 Mar. 2007.
*Corresponding author ([email protected]).
© ASA, CSSA, SSSA
677 S. Segoe Rd., Madison, WI 53711 USA
TECHNICAL REPORTS: ECOLOGICAL RISK ASSESSMENT
208 Journal of Environmental Quality • Volume 37 • January–February 2008
in terms of minimal environmental impact. Another important
reason for our choice of the bar gene for turf grass biotechnology
application is that it enables the use of herbicide tolerance as a se-
lectable marker for development of transgenic turf grass cultivars
having multiple genes (i.e., herbicide tolerance plus other traits
by gene pyramiding) currently in our development pipeline.
In the present study, we characterized the phenotypic perfor-
mance of bar-gene transgenic Zoysia grass in the test fi eld and
used the marker gene in preliminary assessments of the environ-
mental/biodiversity concerns arising from GM Zoysia grass. In
view of the widely expressed concerns about the ecological and
biodiversity implications of GM crops and plants, releasing a GM
plant to agronomic habitats entails prior assessments of its risks
to the environment as well as to human and animal health. Th e
herbicide-tolerant GM crops that underwent such risk assessments
include creeping bentgrass (not currently commercially available
from Scotts), soybean (Monsanto and Bayer CropScience), cotton
(Monsanto, Calgene, Dow AgroSciences, and Bayer CropScience),
maize (Monsanto, Syngenta, DuPont, Bayer CropScience, and
Pioneer Hi-bred), rice (Bayer CropScience), chicory (Bejo Zaden
BV), Argentine canola (Bayer CropScience and Monsanto), Polish
canola (Bayer CropScience and Monsanto), and sugar beet (No-
vartis, Monsanto, and Bayer CropScience). Yaneshita et al. (1997)
studied the outcrossing or self-pollination potential of Zoysia ja-ponica, and reported evidence of interspecifi c hybridization within
the genus Zoysia (Z. matrella, Z. sinica, Z. tenuifolia, and Z. macro-stashya) on the basis of RFLP and morphological characterization.
In this report, we focused our attention on similar ecological and
environmental concerns arising from the release of GM Zoysia
grass to natural environment.
Materials and Methods
Plant MaterialsUnless stated otherwise, all plant materials used for the risk
assessment study reported here have been generated by Toyama
et al. (2003). Th e seeds of Zoysia grass (Zoysia japonica Steud.)
were obtained as described previously (Bae et al., 2001; Toyama
et al., 2002, 2003). Th e Zoysia grass stolons produced were
vegetatively propagated in Cheju National University-approved
confi ned vinyl houses as well as in a test fi eld in Nam Jeju
County, Jeju, Korea, expressly approved for environmental risk
assessments of GM plants by the Rural Development Adminis-
tration/Korea Ministry of Agriculture and Forestry.
Transformation of Zoysia japonicaTh e Agrobacterium-mediated transformation of Zoysia japonica
was established by our laboratory. Th e bar gene introduced, the
promoter used, and the selection markers and the vector chosen
have been reported in detail elsewhere (Becker, 1990, Becker et al.,
1992; Toki, 1992; Lee et al., 1998; Toyama et al., 2002, 2003).
Environmental Risk Assessments
Preparation of Plants
In T3 generation, the stolons of the herbicide-tolerant Zoysia
grass (GM Zoysia grass hereafter) were subjected to various tests.
Th e growth and propagation of the grass were investigated during
hardening and vegetative propagation of the stolons in one of the
isolated greenhouses. Wild-type Zoysia grass (WT) plants were
used as the control for the test. Th e grass stolons thus obtained
were transplanted in the confi ned test fi eld. Th e grass plants were
transplanted in a set of porcelain pots with each pot containing
GM and WT plants separated by 25-cm radii (1 pot = 1 unit).
Genetic Stability
Th e GM Zoysia grasses are tolerant to the non-selective
commercial herbicide Basta (Bayer CropScience, Australia) at
a fi nal concentration of 0.1% (w/v) glufosinate. Th e tolerance
to these herbicides sprayed to GM Zoysia grass was monitored
periodically throughout the T0 and T
1 generations. Th e effi -
cacy of herbicide spraying was assessed under optimal growth
conditions for the grasses. Th e growth and the herbicide ac-
tion on GM Zoysia grasses and naturally occurring weed spe-
cies were investigated 2 wk after Basta was applied.
Phenotypic Characterization
Th e growth and morphology (stems, leaves, seedlings, etc.)
of the GM Zoysia grass were compared with those of the WT
cultivated under greenhouse conditions, according to previous
methods (Honda and Kono, 1963; Yu et al., 1974; Hong and
Yeam, 1985; Hwang and Choi, 1999; Kim et al., 1996; Choi
and Yang, 2004). Table 1 lists the physicochemical properties
of the soils used for the greenhouse habitat. Th e morphological
comparisons included the plant height, the length of the blade,
Table 1. Physicochemical properties of soil mixture used to grow genetically modifi ed (GM) and wild-type (WT) Zoysia grasses. Each indicates the mean ± standard error of three replicates.
Soil
sample pH† EC‡ OM§
Available
P
Exchangeable cations 0.1 N HCl extractable
K Ca Mg Fe B Zn Mn Cu
dS m−1 g kg−1 mg kg−1 ——–––—cmol kg−1¶——–––— ——————––––––——mg kg−1#———––––––————-GM 4.86 ± 0.17 0.032 ± 0.00 46.9 ± 6.8 13.5 ± 2.57 0.78 ± 0.13 0.45 ± 0.14 0.36 ± 0.13 19.1 ± 2.08 0.85 ± 0.13 1.58 ± 0.32 27.8 ± 1.39 0.83 ± 0.07
WT 4.97 ± 0.06 0.036 ± 0.00 40.8 ± 0.6 14.8 ± 2.34 1.00 ± 0.40 0.55 ± 0.07 0.43 ± 0.05 21.5 ± 1.20 0.65 ± 0.21 1.69 ± 0.23 29.0 ± 0.98 1.29 ± 0.66
t-test NS†† NS NS NS NS NS NS NS NS NS NS NS
† pH of soil: water (1:5).
‡ EC, electrical conductivity.
§ OM, organic matter.
¶ cmol kg-1, centimols of positive charge per kilogram of soil.
# mg kg-1, cation concentration.
†† NS, statistically insignifi cant.
Bae et al.: GM Zoysia Grass and Environmental Risk 209
width, and the angle of a plant leaf, the
third-youngest leaf of each erect stem was
chosen for measuring the leaf parameters
to minimize the variations due to envi-
ronmental factors described by Youngner
(1961) and Hong and Yeam (1985), the
distance between the shoot base and the
lowest leaf blade, and the dry weight (up
to the third leaf of the plant) after 48 h
in a drying oven. Th e chlorophyll con-
centration was measured using a portable
chlorophyll analyzer (SPAD-502; Mi-
nolta Co., Japan). For seed morphology
(number, length, and width) compari-
sons, the seeds were harvested from one
spike, and the average weight of the seeds
was measured based on those harvested
from 45 individual plants.
Intra-species Hybridization Potentials
To investigate pollination-induced
hybrid formation between the GM
and the WT grasses, three test plots
each containing both the GM Zoysia
grass and the synchronously fl owering
WT Zoysia grass planted in a 25-cm
diam. by 20 cm deep porcelain pot
were distributed within the test fi eld
in Nam Jeju County, Jeju. Each pot in two of the three plots
had three pots each of GM and WT Zoysia plants, and the
remaining plot contained fi ve pots each. Th e seeds harvested
from the WT Zoysia grass were germinated and the grasses
grown until three leaves appeared, then were screened for
their herbicide tolerance by spraying Basta. Th e herbicide-
tolerant lines screened were then subjected to PCR analysis
based on bar primers. Th e primers for the detection of bar gene were 5’-GGTCTGCACCATCGTCAACC-3’ and 5’-
ATCTCGGTGACGGGCAGGA-3’. Th e Z-A2 actin primers
for the expression in Zoysia japonica were 5’-GTCAACCCTG
TGCAGCAGTA-3’ and 5’-ATTCAGGTTGGTTGCTC-
CAC-3’. Th irty fi ve cycles of PCR were performed under the
following conditions: denaturation at 94°C for 30 s, annealing
at 61°C for 30 s, and elongation at 72°C for 45 s.
Hybridization Potentials in Other Species
Th e GM Zoysia grass and the native weeds grew within the
same confi ned test fi eld during 2003–2005. In May 2004, Bas-
ta was sprayed both inside and within a 5-m radius outside the
test fi eld to investigate cross-pollination between the GM grass
and the weeds mediated by wind. One year later, in May 2005,
hybridization in weed species was then examined on the plants
having identical fl owering time by means of PCR analysis.
Nearest Neighbor (0.5 m) Cross-pollination
During the 2-yr study (2004 through 2005) performed in both
test fi elds in Cheju National University and Nam Jeju County, we
found that the GM and the WT Zoysia grasses fl owered in late
April, but the GM Zoysia grasses fl owered 5 to 7 d later than WT.
Th e GM and the WT Zoysia grasses were transplanted and distrib-
uted according to a completely randomized plot design (Nakayama
and Yamaguchi, 2002; Belanger et al., 2003b) and the randomized
block and crossing block design (Belanger et al., 2004) or an alter-
nating population combination design (Song et al., 2003).
In the alternating population combination design test, fi ve
blocks each (1 × 12 m2) of GM and WT Zoysia grasses were
distributed alternatively (Fig. 1). Figure 2 illustrates the distri-
bution patterns of pots (25-cm diam. and pot-to-pot distance
?0.5 m) containing GM and WT Zoysia grasses. After allow-
ing growth for 10 wk under natural fi eld conditions, hybrid-
ization results were scored for WT samples in each plot as a
function of distance and plot design.
Next-nearest Neighbor (≤3 m) Cross-pollination
Figure 3 illustrates the test for the next-nearest neighbor cross-
fertilization showing 3-m intervals of four GM grass pots (25-cm
diam.) surrounded by one 6 by 16 m2 patch of WT grass. Hybrid
formation within the 3-m separation was determined after 2 mo
of growth under natural fi eld conditions. Mature WT seeds were
harvested from 96 fractions of a 1 by 1 m2 area around a GM pot
and dried naturally under sunlight. Fresh seeds were stored in ice
box at −10°C until used. Th e seeds were dehusked mechanically
and kept at 4°C for 1 wk. Th e seeds were sterilized in 2% sodium
hypochrorite solution for 15 min and rinsed fi ve times in distilled
water. Th e seeds were then allowed to imbibe on wet fi lter papers
at 35°C under 4200 cd sr–1 m−2 light for 72 h, and placed at 25°C
under 3500 cd sr–1 m−2 light for germination. Th e resulting plants
were then sprayed with Basta to screen for hybridization.
Fig. 1. Field testing (a) and schematic illustration (b) for cross-hybridization between genetically modifi ed (GM) and wild-type (WT) Zoysia grass at 0.5-m separation. Grass lanes, GM Zoysia grass; pots (25- cm diameter each), WT Zoysia grass.
210 Journal of Environmental Quality • Volume 37 • January–February 2008
Neighbor (3 m to 9 m) Cross-pollination
Figure 4 shows a 9 by 9 m2 hexagonal test
plot (initiated July 2004) enclosing the GM
grass (1-m radius) surrounded by cold- and
warm-season grass species at distances from 3
to 9 m. Each plot included fi ve grass species,
Z. japonica, Z. sinica, Z. matrella, perennial
ryegrass (Lolium perenne L.), and Kentucky
bluegrass (Poa pratensis L., data not shown).
GM and these grass species were trans-
planted and arranged according to a plot
design (Belanger et al., 2003a). Th e seeds
were harvested in August 2005, dried, ger-
minated, and screened for hybrid forma-
tion by herbicide application.
Long Distance Cross-Fertilization
Potential gene fl ow from the total
936 m2 GM grass fi eld (14 × 16 m2, 16
× 40 m2, and 6 × 12 m2) to WT Zoysia
grasses in the surrounding wilderness (119
Zoysia japonica and 2 Zoysia sinica sampling
sites as shown in Fig. 5) within a 3-km
radius was tested based on Basta screen-
ing and PCR analysis. Th e majority of the
sampling sites are east and northeast biased
relative to the GM grass test fi eld because
of the local land topography—namely high
hills, woody forests, seaside, bushy valleys
south, and southeast of the test fi eld.
Unintended gene fl ow and seed propa-
gation from the GM creeping bentgrass
fi eld established in 2003 were tested in
2005. For this purpose, both herbicide
screening and PCR methods were used, as
described previously (see Intra-species Hy-
bridization Potentials).
Skin Prick Tests
With informed consent, we performed a
similar study with pollen extracts from GM
and WT grasses on chronic allergy patients
admitted at the allergy clinic of Cheju Na-
tional University Hospital and on healthy
volunteers over the period from October
2005 through April 2006. For the skin
prick tests, twenty common inhalent aller-
gens and pollen extracts of GM and WT
grasses were used with positive (histamine
1 mg mL−1) and negative (0.9% NaCl)
controls. Th e sensitization was defi ned
when each wheal size showed more than 3
mm. One hundred twenty seven subjects
(55 males, mean age of 38) were included.
Among the 127 subjects, 87 individuals
were sensitive to inhalent allergens.
Fig. 2. Field testing (a) and schematic illustration (b) for cross hybridization between genetically modifi ed (GM) and wild-type (WT) Zoysia grass according to a randomized complete block design. Black circle, GM Zoysia grass; white circle, WT Zoysia grass.
Fig. 3. Field testing for gene fl ow from genetically modifi ed (GM) to wild-type (WT) Zoysia grass within a 3-m radius. Each GM grass pot is surrounded by WT grass patches of 6 by 16 m2 area.
Bae et al.: GM Zoysia Grass and Environmental Risk 211
ResultsTwo copies of the bar gene introduced
in GM-Zoysia japonica retained its stable
integration in the host plant in the T1 to
T6 generations, exhibiting a 15:1 segrega-
tion ratio in accordance with Mendelian
genetics, and also showing the transgenic
line’s tolerance to ammonium glufosinate
throughout the culture period. Th e geno-
type was retained through the multiple
generations in both transgenic Zoysia lines
and their WT hybrids. Before subjecting
the GM Zoysia japonica grass to the envi-
ronmental risk assessment study reported
here, we observed that the GM Zoysia
grass survived application of the non-se-
lective herbicide spray, whereas WT grass
did not, indicating the stable inheritance
of bar gene in the transgenic grass.
Th e GM Zoysia grasses were cultivated
in a greenhouse and periodically checked
for the herbicide tolerance at various
stages of growth and development. Th e
grass lines hardened during repeated
vegetative propagation were then used
for all the studies reported hereafter. Both
the WT and GM Zoysia plants displayed
essentially identical germination and growth rates, and morpho-
logical and physiological characteristics. One interesting diff erence
between the two types of Zoysia grass was that locusts and other
insects preferred to reside in the Basta-sprayed GM fi eld, com-
pared to the WT fi eld, most likely avoiding the higher amounts of
selective herbicide residuals such as Pyrazosulfuron [ethyl 5-(4,6-
Fig. 4. Field testing and schematic illustration for cross hybridization between genetically modifi ed (GM) and wild-type (WT) Zoysia grass and it relative weed species as a function of distance. (a) GM Zoysia japonica tillers (illustrated with white circle). (b) WT grass group containing Zoysia japonica (Zj), Zoysia matrella (Zm), Zoysia sinica (Zs), Lolium perenne (Lp), and Poa pratensis (Pp). (c) Hexagonal arrangement illustrating a GM test area shown in (a); orange, 3-m radius; blue, 6-m radius; black, 9-m radius).
Fig. 5. Test for the potential gene fl ow from genetically modifi ed (GM) grass to wild-type (WT) grasses within a 3-km radius during a 2-yr period from 2003 to 2005. The GM grass fi eld is centrally located in the Wimi-Ri test fi eld in Nam Jeju County. The sampling sites shown were randomly chosen where Zoysia grasses grew. The sampling site distribution is biased in the north easterly direction from the GM grass site, whereas other directions are less favorable for grass growth due to geo-topographic factors (volcanic rocks, bushy jungles, forest, etc.).
212 Journal of Environmental Quality • Volume 37 • January–February 2008
dimethoxypyrimidin-2-ylcarbamoyl)sulfamoyl-1-methylpyrazole-
4-carboxylate], Alachlor (2-chloro-2′,6′-diethyl-N-methoxymethyl-
acetanilide), and Triclopyr (3,5,6-trichloro-2-pyridyloxyacetic acid)
in the latter. More frequent applications (fi ve to seven times) of the
selective herbicide spray were required to keep the WT grass fi eld
free of weeds, whereas the non-selective herbicide (Basta), just once
or twice was suffi cient for the GM grass fi eld.
Th e herbicide tolerance of the bar transgenic Zoysia grass was
stably preserved for the testing period spanning more than 2 yr.
During the same period that the Basta tolerance of the GM grass
was sustained stably, they remained susceptible to non-selective
herbicides such as paraquat and glyphosate. Th us, for any reason
if it is necessary to terminate the cultivation and spread of the
GM grass within and beyond the test fi eld, the GM plants can be
readily killed by applying a herbicide spray other than Basta.
Is GM Grass Environmentally Risky? Comparative
Characterizations of GM and WT GrassesConventional environmental risk assessments of GM crops
have been performed in four categories, namely, (i) establishment
of the substantial equivalence between the GM and the WT
plants, (ii) determination of pollen fl ight and potential gene fl ow,
(iii) biodiversity eff ects of GM plants on unintended or non-tar-
get and target plants in their ecological habitats, and (iv) health
risk assessments for animals including humans. We adopted the
four-category protocol for an initial evaluation of GM Zoysia japonica before its release to agronomic habitats. Category (i) vis.
substantial equivalence, is described and further discussed here
along with the remaining aspects (categories ii–iv). Th e substan-
tial equivalence between the GM and the WT grasses has been
established on the basis of their essentially identical reproduc-
tion rate, morphology of leaves and seeds, germination rate, and
chemical composition.
To ascertain the grassy characteristics of the GM Zoysia plants
in comparison to those of WT plants, the plants were grown
under identical conditions with respect to soil composition, ir-
rigation, and fertilization, etc. Results from the various studies
described here indicate that GM grass displays morphological,
physiological, and genetic characteristics virtually indistinguish-
able from WT grass, except for the bar-gene transgenicity of the
former, which imbues it with tolerance to a herbicide (Basta)
spray. Both GM and WT seeds showed relatively low germina-
tion rates, so vegetative propagation through the spreading of
grass stolons over the soil surface became the preferred method
for the hardening and vegetative propagation of the Zoysia grass
cultivars. Th e following summaries provide further details of the
studies including some of the subtle diff erences in grassy features
observed between the GM and the WT lines.
Reproduction and Genetic Traits
Flowering Time
During the 2-yr study conducted in 2004 and 2005, we ob-
served WT grasses starting to fl ower in late April, whereas the GM
plants fl owered about 5 to 7 d later. Both types of grasses fully
fl owered within 5 d of 10 May, formed dried pollens (i.e., inactiva-
tion of anthers), and full seed formation by mid-July, consistent
with the fl owering times reported by Kitamura (1967). Since GM
grass began fl owering about 5 to 7 d after WT plants, the fre-
quency of formation of intraspecifi c GM hybrids could have been
reduced. To circumvent this possibility arising from the fl owering
time diff erence, both GM and WT grass cultivars that fl ower at
about the same time in the greenhouse were replanted in the plots
of the test fi eld. However, after 2 yr this became unnecessary as
both types of grass fl owered simultaneously (see below).
Pollination
Pollen formation was maximal around 10 May, and intra-
species hybrid formation induced by pollination was most
prevalent thereafter, 15 to 18 May. We decided to introduce
WT plants having similar fl owering time into the GM grass
greenhouse and outdoor test fi eld on 15 May, 2004, and
2005. Results showed that both GM and WT grasses cross-
pollinated at an average rate of 6% at close proximity.
Morphology
Th e morphology of Zoysia grass can be classifi ed in terms
of leaf width and length, according to Kitamura’s horticultural
classifi cation method (Kitamura, 1967). We examined other
appearance indices of the plants, as presented in Table 2, which
shows morphological features of GM and WT
grasses. Results show that the two types of grasses
are essentially indistinguishable and any diff erences
observed were statistically insignifi cant. Table 3
compares the seed morphologies, again indicating
that the two types of grasses are indistinguishable.
Other phenotypic traits also showed no signifi cant
diff erences between the GM and WT Zoysia culti-
vars (Tables 2, 3, and 4).
In summary, after 16 mo of cultivation in the test
fi eld, morphological analyses were performed and both
WT and GM Zoysia grass displayed a plant height of
20 cm, leaf length of 17 to 19 cm, and leaf width of
0.5 cm. Th e length of the lowest leaf blade and the
chlorophyll content were 4 cm and 1 g kg−1, respec-
tively (Table 2). After 3 mo of planting, the number
Table 2. Morphological characteristics of genetically modifi ed (GM) and wild-type (WT) Zoysia grass leaves observed under fi eld growing conditions. Each value indicates the mean ± standard error of fi fteen replicates.
Plants
Plant
height
Leaf
blade†
First leaf
height‡
Leaf
width‡
Leaf
angle§
Chlorophyll
contents¶
Leaf
weight#
–––––––––––––––––cm––––––––––––––––– A° g kg−1 FW g
GM 20.4 ± 2.5 19.7 ± 3.5 4.0 ± 0.7 0.54 ± 0.06 24.9 ± 6.4 1.06 ± 0.17 0.20 ± 0.05
WT 19.9 ± 3.9 17.8 ± 3.3 3.8 ± 0.7 0.55 ± 0.08 24.5 ± 5.1 1.04 ± 0.19 0.20 ± 0.09
t-test NS†† NS NS NS NS NS NS
† Values measured from the third leaf.
‡ Length from basal zone of the shoot to fi rst leaf blade.
§ Angle between leaf blade axis and vertical axis.
¶ Chlorophyll contents measured from SPAD values; the values were calculated from
the relation curve between the UV spectrophotometer and the Chlorophyll meter
(SPAD-502; MINOLTA, Japan).
# Leaf weight; the three of fresh leaves were dried in an oven at 80°C for 3 d.
†† NS, considered statistically insignifi cant at 0.05 level by t test.
Bae et al.: GM Zoysia Grass and Environmental Risk 213
of stolons (ca. 5), its length (approx. 30.4–33.0 cm), and the leaf-
node length (approx. 3.2–3.7 cm) were also statistically equivalent
for the both types (Table 4). In addition, a seed’s morphological
characterizations performed included the number of seeds per
spike (ca. 49); length of fl owering culms (ca. 12 cm) and the rachis
length (ca. 4.8 cm), seed length (ca. 3.1 mm) and width (ca. 1.5
mm), 1000 seeds weight (ca. 0.58 g), and the rate of germination
(approx. 3.7–4%, Table 3). Th e chemical and mineral composi-
tions of seeds harvested from both types were also performed and
no signifi cant diff erences were found (data not shown).
Hybridization
As evident from the above result (see Pollination), no
signifi cant diff erence in pollination rates was found between
the GM and the WT Zoysia grasses. We also examined the
pollination from GM to WT Zoysia grasses, and in the reverse
direction. Table 5 shows that the minimum cross-pollina-
tion rate was 3% and maximum was 9%, with an average of
6%, at the nearest distance (>0 m). At a 0.5-m distance in
both randomized and completely randomized plot designs,
cross-pollination was approximately 1.2%, which declined to
0.12% at a 3-m distance, and to 0% at distances greater than
3 m (Table 5). Figure 6 graphically illustrates the distance
dependence of GM-to-WT Zoysia gene fl ow. Th e best fi t for
this distance dependence from a regression analysis of the data
described in the results section is an exponential function as
shown in the fi gure. Similar distance dependence has been
reported for wild rice (Oryza rufi pogon) (Song et al., 2003).
Winter Dormancy
Both WT and GM Zoysia grass showed essentially identi-
cal dormancy profi les in the Nam Jeju County test fi eld, turn-
ing brown, wilting by late November, and
staying dormant until the next March.
Eff ects of GM Zoysia Grass on
Neighboring Weeds: Potential Weediness
Gene Flow from GM Zoysia to Weeds
Table 6 lists 14 co-habitant weed species
within the GM grass test plot facility. Neither
Basta nor PCR evidence was obtained to
indicate bar-gene fl ow from the GM plant’s
pollen to these neighboring weed species dur-
ing the study conducted from 2003 to 2005.
GM Zoysia’s Dominance over Weeds
Th e Zoysia grass propagates reproductively both from seeds
and vegetatively. Th e weight of the 1000 seeds is approx. 0.57
to 0.59 g. Even if wind carries the seeds over some distance,
the germination rate is less than 4% under natural conditions.
Th us, compared to germination, Zoysia can spread itself more
eff ectively through vegetative propagation. However, the Zoysia
grass is not a dominant species and does not spread into weedy
areas easily. In fact, the Zoysia grass fi eld is completely domi-
nated by the weeds within 2 to 3 yr of cohabitation. Figure 7 il-
lustrates the eff ects of dominant weeds on the GM Zoysia grass,
showing the dominance of the weeds over the Zoysia grass.
Disease Tolerance and Pathogenic OrganismsTh e eff ects of GM Zoysia grass on the population of several
pathogenic soil fungi were investigated. Table 7 lists the soil patho-
gens distributed within the soil layer and the plant segment of ap-
proximately 3 cm length from the soil surface layer in the test fi eld.
No signifi cant diff erences in the population of the four major turf
grass pathogens (Rhizoctonia spp., Pythium spp., Curvularia sp.,
and Colletotrichum sp.) between WT and GM grass planted soils
were found, with all diff erences within experimental and statistical
margins of error. However, the soil samples contained relatively
high levels of Fusarium spp. in both WT and GM grass plots. Th e
relatively dense population of this fungus is attributable to low
soil pH and electrical conductance (Kwon et al., 1998; Suh et al.,
2003). Fusarium spp. is a common fungal pathogen in soil, but
turf grass plants are apparently unaff ected. In fact, the Fusarium
spp. stimulates a plant’s growth by suppressing several co-habitat
pathogens (Meera et al., 1993, 1994; Liu et al., 1995; Yun, 1996;
Park and Yu, 2005). Th e higher density of Fusarium spp. in the
Table 3. Seed characteristics of genetically modifi ed (GM) and wild-type (WT) Zoysia grass using eight morphological traits; length of fl owering culms, length of spike without rachis; length/width ratio; germination, and weight.
PlantsNo. of seeds
per spikeLength of
rachisLength of
fl owering culms Seed length
(SL)Seed width
(SW)SL/SWratio
Frequency ofgermination
Weight of1000 seeds
––––––––––––cm–––––––––––– ––––––––mm–––––––– % g
GM 49.4 ± 7.6† 4.8 ± 0.6‡ 12.1 ± 2.4‡ 3.2 ± 0.3‡ 1.5 ± 0.2‡ 2.2 ± 0.2‡ 3.7 ± 1.2 0.6 ± 0. 03§
WT 49.1 ± 7.3 4.9 ± 0.6 11.7 ± 2.7 3.1 ± 0.3 1.4 ± 0.2 2.2 ± 0.3 4.0 ± 1.0 0.6 ± 0.02
t-test NS¶ NS NS NS NS NS NS NS
† Mean ± standard error of forty-fi ve replicates.
‡ Mean ± standard error of fi fteen replicates.
§ Mean ± standard error of fi ve replicates.
¶ NS, considered statistically insignifi cant at 0.05 level by t test.
Table 4. Growth characteristics of genetically modifi ed (GM) and wild-type (WT) Zoysia grass using fi ve morphological traits. Each value indicates mean ± standard error of triplicates.
Coverage† Stolon No. Stolon length Leaf-node length Density§
Plants 90DAP‡ 150DAP 90DAP 90DAP 90DAP 150DAP
–––––––––m2––––––––– ––––––––––––cm–––––––––––– no. cm−2
GM 0.07 ± 0.01 0.13 ± 0.06 5.2 ± 0.8 33.0 ± 7.4 3.7 ± 0.9 0.55 ± 0.04
WT 0.08 ± 0.01 0.13 ± 0.05 5.4 ± 1.1 30.4 ± 6.8 3.2 ± 0.7 0.57 ± 0.07
t-test NS¶ NS NS NS NS NS
† Coverage, about 10-cm diameter of GM and WT Zoysia grass plugged after 10 May.
‡ DAP, days after plugging.
§ Density, number of tiller per cm2.
¶ NS, considered statistically insignifi cant at 0.05 level by t test.
214 Journal of Environmental Quality • Volume 37 • January–February 2008
GM grass soil may account for the lower levels of Curvularia sp.
and Colletotrichum sp., although the pathway by which the GM
grass soil stimulates Fusarium growth remains unknown (Table 7).
In fact, our study showed that other pathogenic populations were
suppressed by Fusarium spp. (data not shown).
Potential Gene-induced and Allergic Hazards
bar Gene
Th e bar gene described earlier was originally isolated from
S. hygroscopicus. Its coded amino acid sequence showed an
84% identity with the 183-amino acid polypeptide encoded
by pat gene from S. viridochromogenes (National Center for
Biotechnology Information BLAST 2; Wehrmann et al.,
1996). On a 12% sodium dodecyl sulfate-polyacrylamide gel
electrophoresis (SDS-PAGE) gel, both proteins were identi-
cally Western blotted at 21 to 23 kDa (Herouet et al., 2005).
When the amino acid sequence of PAT was matched
against the known allergenic sequences using BLAST 2.2.15
algorithm, SwissProt, PDB, PIR, PRF in NCBI, and Food
Allergy Research and Resource Program (FARRP)-FASTA
Version 6, 80 or more amino acid-peptide sequences of the
enzyme showed less than 35% homology. In addition, no
homology was found between the protein sequences and the
eight amino acid allergen epitopes (FAO/WHO, 2001; Codex
Alimentarius Commission, 2003; Bae, 2007). Th e pat gene is
inactivated at a pH below 4 or by heating for 30 min (Weh-
rmann et al., 1996; ANZFA, 2001; Herouet et al., 2005).
Previous animal and human studies showed that the pat gene
posed no signifi cant health risks (Jones and Maryanski, 1991;
Sjoblad et al., 1992; Schmidt, 1994; Mossinger and Dietrich,
1998; WHO, 1998; OECD, 1999; Institute of Science in
Society, 2003; Th omas et al., 2004; Herouet et al., 2005). Th e
bar or pat gene used for the generation of GM Zoysia grass
has been introduced into commercial crops such as rapeseed
(Bayer CropScience), corn (Bayer CropScience), and cotton
(Bayer CropScience). Th e gene introduced showed no appar-
ent risks to human and animal health in terms of cytogenetic
toxicity and allergenic reactions (FDA, 1992; CFSAN, 1995,
1998; FAO/WHO, 1996, 2000, 2001; Health Canada,
2001; OECD, 2001, 2002; Codex Alimentarius Commis-
sion, 2003; European Commission, 2003).
Allergic Reactions
Kim et al. (1987) performed skin prick tests with pollen
extracts of Zoysia grass and found that 5% of respiratory al-
lergic patients were sensitized. Table 8 summarizes the test
results. Six cases each of a wheal reaction to the WT and GM
pollen extracts were found. Among the six, three subjects
had respiratory allergic disorders. Th us, six subjects (4.7% of
the 127 test subjects) developed a positive allergic reaction
Table 5. Number of the germinated seeds tested and the hybrids identifi ed (number in parentheses) at distances from genetically modifi ed (GM) Zoysia grass in each plot design.
Distance CRD† RCBD‡ R-3§ R-9Zj¶ R-9Zs# R-9Zm†† R-9Lp‡‡
m
m > 0 746(45) -§§ – – – – –
0.5 491(6) 967(12) – – – – –
1 – 245(4) 72(1) 243(2) – – –
2 – – 660(2) 145(1) – – –
3 – – 2547(3) 231(0) 83(0) 0 89(0)
6 – – – 209(0) 152(0) 79(0) 176(0)
9 – – – 214(0) 104(0) 58(0) 93(0)
† CRD, completely random design.
‡ RCBD, randomized complete block design.
§ R-3, WT Zoysia japonica within 3-m radius from GM Zoysia japonica pot
(0.25-m diameter).
¶ R-9Zj, WT Zoysia japonica within 9-m radius from GM Zoysia japonica
(1.5-m diameter).
# R-9Zs, WT Zoysia sinica within 9-m radius from GM Zoysia japonica (1.5-m
diameter).
†† R-9Zm, WT Zoysia matrella within 9-m radius from GM Zoysia japonica
(1.5-m diameter).
‡‡ R-9Lp, WT Lolium perenne within 9-m radius from GM Zoysia japonica
(1.5-m diameter).
§§ Blank boxes represent no plot design data.
Table 6. Test for the potential outcrossing between genetically modifi ed (GM)-Zoysia grass and weed plants grown within the Wimi-Ri test fi eld.
No Scientifi c name Common name
Flowering
season Outcross
1 Spergula arvensis Corn spurry Mar.-June –
2 Cerastium holosteoides Common mouse ear Mar.-June –
3 Stellaria media Chickweed Apr.-June –
4 Trigonotis peduncularis Cucumber herb Apr.-June –
5 Taraxacum offi cinate Dandelion Apr.-June –
6 Veronica arvensis Corn speedwell Apr.-June –
7 Vicia angustifolia Garden vetch Apr.-June –
8 Erigeron annuus Daisy fl eabane June-Sept. –
9 Mazus pumilus Japanese mazus Apr.-Aug. –
10 Youngia japonica Japanese youngia Apr.-June –
11 Cardamine impatiens Narrow leaf bitter cress Mar.- May –
12 Gnaphalium affi ne Cudweed May-July –
13 Alopecurus aequalis † Orange Foxtail Apr.-June –
14 Poa annua † Annual Bluegrass Dec.-June –
† Family of Gramineae with no hybridization from GM Zoysia grass.
Fig. 6. Distance dependence for gene fl ow from the genetically modifi ed (GM) to wild-type (WT) Zoysia grass within 9-m radius in fi eld. The observed data can be best fi t by an exponential equation resulting from a regression analysis of the data. Bars refer to standard error.
Bae et al.: GM Zoysia Grass and Environmental Risk 215
to both WT and GM Zoysia pollens. However, no diff erence
between the two types of pollens was observed.
DiscussionWe fi rst reported the successful establishment of GM herbicide/
Basta-tolerant Zoysia japonica Steud. by transforming the plant calli
with the transgene bar (Toyama et al., 2003) under greenhouse
habitats. In the present study, we confi rmed that the transgenic
Zoysia grass contained two copies of the bar gene. We further char-
acterized the phenotypic performance and the transgene introgres-
sion in the natural ecological environment of Jeju Island, Korea.
All morphological and biochemical analyses suggested that the
GM Zoysia grass developed is indistinguishable from its WT plant,
except for its Basta tolerance, under greenhouse, and fi eld habitat
conditions (Fig. 1–6 and Tables 1–7). In addition, no diff erence in
the incidence of allergic skin reactions to both GM and WT Zoysia
grass was observed (Table 8). Th us, the focus of the discussion will
be on the concerns about transgene fl ow from the GM Zoysia grass
to compatible WT Zoysia and other weed species within and out-
side the test fi eld in Jeju.
Table 7. Test for the fungal infection of genetically modifi ed (GM) and wild-type (WT) Zoysia grass within the test fi eld.
Fungi
Fungal infection
Disease
GM Zoysia grass WT Zoysia grass
Base
line†
Rhizosphere
soil‡
Base
line
Rhizosphere
soil
–––––––––––––––––%–––––––––––––––––
Rhizoctonia spp. ND§ ND ND ND Large patch
Pythium spp. ND ND ND ND Pythium blight
Curvularia sp. 2.8 3.0 4.5 8.3 Leaf blight
Colletotrichum sp. ND ND 1.5 ND Anthracnose
Fusarium spp. 75.8 19.4 22.7 11.1 Unknown
† Shoot base.
‡ Stem’s rhizosphere soil zone.
§ ND, fungi not detected.
Table 8. Results from the skin prick tests for common and Zoysia grass allergens.
Allergen No. of patients Positive reaction
%
Positive control (1mg/mL histamine) 127 100
Negative control 0 0
Dermatophagoides farinae 72 56.7
Dermatophagoides pteronyssinu 60 47.2
American cockroach 36 28.3
German cockroach 39 30.7
Cat and dog hair 18 14.2
Horse and cattle hair 4 3.2
Flag 3 2.4
Broadleaf tree 2 1.6
Acicular tree 6 4.7
Japanese cedar 5 3.9
American cedar 6 4.7
House dust-fungi 8 6.3
Outdoor fungi 6 4.7
Flowers 12 9.4
Weeds 14 11.0
Crops 10 7.9
GM Zoysia grass 6 4.7
WT Zoysia grass 6 4.7
Fig. 7. Zoysia japonica plants are overcame by dominant weed plants under natural ecological conditions. (A) Unmanaged Zoysia and weeds habitats. (B) The Zoysia lawn after weeds were removed. (C) Weeds began to overtake Zoysia grass (1 yr without weed control). (D), (E) The same as C after 2 and 3 yr without weed control, respectively.
216 Journal of Environmental Quality • Volume 37 • January–February 2008
Th e GM Zoysia grass has been developed for its eventual release
to agronomic habitats and recreational lands such as golf courses.
Th is work has been performed under a joint developmental agree-
ment with the Jeju Provincial Government. Th e local government
is particularly interested in herbicide-tolerant grass for its potential
environmental and economic implications for Jeju’s golf courses
and recreational parks. We anticipate that the use of GM Zoysia for
golf courses will substantially reduce the amount and frequency of
weed-selective herbicide sprays performed annually. Th e volcanic
island’s water supply for half a million inhabitants is underground
springs, and concerns about potential herbicide contamination can
be lessened through the use of non-selective herbicide (Basta) ap-
plications to the GM grass lands.
Recently, Reichman et al. (2006) monitored the pollen or
seed transfer from a large fi eld of GM herbicide/glyphosate-
tolerant creeping bentgrass (Agrostis stolonifera L.) developed
by Scotts and Monsanto. Watrud et al. (2004) and Reichman
et al. (2006) looked at cp4 epsps transgene fl ow and the escape
from large test production plots planted by Scotts in Oregon.
Th e establishment of transgenic plants in wild populations
was the result of unintended releases from Scotts fi elds.
Pollen-mediated introgression of herbicidal transgene cp4 epsps introduced in the herbicide-tolerant creeping bentrass has
been detected within populations of closely related grass spe-
cies at up to 3.8 km from the perimeter of GM grass habitats
(Baack, 2006; Reichman et al., 2006). When released to the
natural environment, transgene fl ow to related species could
occur, including turf grass. Th us, it is critical to consider the
ecological and societal implications of transgenic bentgrass.
To release GM grass to agronomic habitats including golf
courses and parks, we must address the concerns about trans-
gene fl ow from GM Zoysia grass to other compatible grasses and
weeds under the natural ecological conditions in Jeju. As a fi rst
step to assess the possibility of unintended environmental and
ecological risks associated with transgene (bar in the present case)
introgression, we performed several analyses primarily involving
transgene fl ow from GM Zoysia to WT Zoysia grass, as sum-
marized in the results section. Within short distances of 3 m or
less, intra-specifi c hybridization between the GM and WT Zoysia
grass tillers was signifi cant (Fig. 6). Th is observa-
tion is consistent with the documented cases of
conventional gene fl ow and hybridization between
cultivated and non-cultivated plant populations
including the transgene introgression between GM
bentgrass and compatible WT grasses (Reichman
et al., 2006). However, at distances over 3 m the
frequency of cross hybridization drops precipitously
to essentially zero, as discussed below.
As cited earlier, pollen-mediated introgression
of herbicidal transgene cp4 epsps introduced in
bentgrass has been detected within the popula-
tions of closely related grass species at up to 3.8
km from the perimeter of the GM grass habitats
(Baack, 2006; Reichman et al., 2006). Th us,
when released to the natural environment, herbi-
cide-tolerant weeds could evolve as the result of
transgene fl ow. Implications of the bentgrass case are critical for
both ecological and societal concerns. However, it should be
pointed out that the comparison of the distances of transgene
fl ow in the present study to those of Reichman et al. (2006)
is not quantitatively valid, since the scale of the pollen sources
between the two studies diff er markedly (we thank the referee
for pointing this out).
However, from 121 sampling sites beyond our test area perim-
eter up to 3-km distance (Fig. 6), we found no evidence for either
pollen-mediated hybridization of Zoysia grass or seed dispersal,
although the PCR and Basta resistance methods we used would
not have discriminated pollen-mediated hybrid Zoysia progeny
from those that grew from GM crop seeds, as pointed out by a
referee. Th is observation contrasts with the case of GM bentgrass
showing transgene fl ow at multi-kilometer distances mediated by
downwind pollen and/or seed fl ights (Reichman et al., 2006), as
concerns of such unintended gene fl ow was discussed previously by
Wipff and Fricker (2001) and Watrud et al. (2004). However, long
distance gene fl ow is of lesser concern for GM Zoysia grass since
pollen/seed-mediated hybridization was not observed at distances
greater than a few meters. Further studies are warranted to monitor
the dispersal of viable transgenic pollen over much greater distances
from the larger plots of GM Zoysia grass than those reported in this
work. Th e Zoysia japonica seeds show only a 4% germination rate
after winter dormancy in its natural ecological habitats (Niwa and
Takanashi, 1943; Bae, 2007), further contributing to the reduced
risk of transgene fl ow from the GM Zoysia grass.
Several factors can account for the lack of “long-distance”
gene fl ow from Jeju Island’s Wimi-Ri test area to the sampling
sites (Fig. 6). Th ey include inherently recalcitrant cross-pol-
lination in Zoysia japonica, low germination rate under natural
conditions, relatively small GM pollen source, land topography,
and wind variations during the month of May when Zoysia
produces pollen in Jeju. Wind mediates cross-pollination.
Th e island is known for its strong wind. Figure 8 shows wind
directions with maximum velocities of 6.7 m/s northerly, 5.6
m/s easterly and south easterly, 8.4 m/s southerly, 4 m/s south
westerly, and 5.9 m/s westerly winds. Average monthly wind
velocity in May 2005 was 5 m/s. Perhaps strong and multi-
Fig. 8. Average wind velocity and directions on Jeju Island during the month of May 2005. The wind directions, maximum velocities of 6.7 m/s northerly; 5.6 m/s easterly and south easterly, 8.4 m/s southerly; 4 m/s south westerly, and 5.9 m/s westerly winds. Average monthly wind velocity was 5 m/s during the fl owering season and gene fl ow testing.
Bae et al.: GM Zoysia Grass and Environmental Risk 217
directional winds may be counterproductive for pollination, as
both eff ective pollen fl ight and deposit are aff ected by the wind.
We examined potential transgene fl ow from GM Zoysia grass
to 14 co-habitant wild plant species within the test area. Th e
fourteen weed species (Table 6) with similar fl owering periods
were sampled for testing the transgene introgression mediated
by pollen fl ight. Th ere were no incidences of cross-hybridiza-
tion between the GM Zoysia and the co-habitant weed group.
Furthermore, no transgene fl ow from GM Zoysia to other grass
species such as perennial ryegrass, Kentucky bluegrass, tall fescue,
and cogon grass that were co-cultivated with GM Zoysia inside
the perimeter of the test area was observed. Since transgene in-
trogression has not been detected among the monocot and dicot
grasses sampled for testing, we tentatively conclude that fl ow of
the bar gene from Zoysia japonica to other plant species is rare
under ecological conditions in and around the test habitat.
In a natural ecological environment, Zoysia grasses are read-
ily overtaken by dominant wild plant species and they do not
survive well in the wild (Fig. 7). Since growth and propagation
of the Basta-tolerant Zoysia grass can be terminated by the use of
non-glufosinate herbicides, the risk of GM Zoysia grass spread-
ing to unintended areas is low and controllable. Interestingly, we
observed that locusts and other insect habitants in the test fi eld
thrived by feeding on GM Zoysia grass while avoiding the WT
Zoysia grass treated with several obligatory sprays of non-Basta
herbicides. Also, potential for horizontal gene fl ow (HGT) medi-
ated by insects’ microfl ora and soil microorganisms is unlikely
(Bae et al., 2007). Brown et al. (2001) observed the populations
of GM canola, potato, corn, and sugarcane in their natural habi-
tats for over 10 yr, and found that these GM crops were not more
adaptive to the environment than their WT plants. Similarly, for
the next 8 to 10 yr, we will continue to study the herbicidal activ-
ity, intra- and inter-specifi c transgene introgressions, and ecologi-
cal eff ects of the GM Zoysia grass described in this report.
Finally, about 5% of pulmonary patients show positive al-
lergic reactions to Zoysia grass’ pollen (Kim et al., 1987). We
showed similar incidences (4.7%) of allergic skin responses to
the extracts from the pollens of both GM and WT Zoysia grass-
es. No diff erences were observed between the pollens of the two
grass types, suggesting that the bar gene introduced in GM Zoy-sia grass does not induce production of specifi c allergens. Th is
suggestion is in agreement with preliminary reports that the bar gene does not induce or encode for any allergen (OECD, 1999;
Codex Alimentarius Commission, 2003; Herouet et al., 2005).
ConclusionsTh e herbicide-tolerant GM and the WT Zoysia are substan-
tially equivalent, except for the transgene bar’s phenotypic traits
that are conferred on the former. Transgene introgression is sig-
nifi cant at close proximity between the GM and the WT Zoysia
grasses. However, no gene fl ow mediated by pollen fl ights was de-
tected among the wild species within the test habitat and the WT
Zoysia grasses at 121 sampling sites multi-kilometers away from
the perimeter of the test area in Jeju. Also, the herbicide-tolerant
grass poses no specifi c allergic/health risks associated with the bar
transgene. We conclude that the GM Zoysia grass developed is an
environmentally, ecologically, and dermatologically safe grass for
potential use in golf courses and other recreational parks on the
Island of Jeju and possibly in other Zoysia habitats elsewhere.
AcknowledgmentsTh is work was supported in part by grants from Korea
Ministry of Agriculture and Forestry/RDA Biogreen 21 Program
(20050401034689 and 20050601034857), Korea Ministry of
Science and Technology/KOSEF Environmental Biotechnology
National Core Research Center (R15-2003-012-010030), and
City of Seogwipo Dep. of Parks and Recreation, Jeju, Korea.
ReferencesANZFA. 2001. Final assessment report (inquiry-section 17)- Application A375:
Food derived from glufosinate ammonium-tolerant corn line T25. Australia New Zealand Food Authority, Canberra, Australia.
Baack, E.J. 2006. Engineered crops: Transgenic go wild. Curr. Biol. 16:R583–R584.
Bae, T.W. 2007. Development of genetically modifi ed (GM) plant and ecological risk assessment of GM herbicide-tolerant Zoysia grass (Zoysia japonica Steud.). Ph. D. thesis. Cheju Univ., Korea.
Bae, T.W., H.Y. Lee, K.H. Ryu, T.H. Lee, P.O. Lim, P.Y. Yoon, S.Y. Park, K.Z. Riu, P.-S. Song, and Y.E. Lee. 2007. Evaluation of horizontal gene transfer from genetically modifi ed zoysiagrass to the indigenous microorganisms in isolated GMO fi eld. Kor. J. Plant Biotechnol. 34:75–80.
Bae, C.H., K. Toyama, S.C. Lee, Y.P. Lim, H.I. Kim, P.-S. Song, and H.Y. Lee. 2001. Effi cient plant regeneration using mature seed-derived callus in Zoysiagrass (Zoysia japonica Steud.). Kor. J. Plant Tissue Cult. 28:61–67.
Bayer, E., K.H. Gugel, K. Hagele, H. Hagenmaier, S. Jessipow, S.A. Konig, and H. Zahner. 1972. Phosphinothricin and phosphinothricyl-alanyl-alanin. Helv. Chim. Acta 55:224–239.
Becker, D. 1990. Binary vectors which allow the exchange of plant selectable markers and reporter genes. Nucleic Acids Res. 18:203.
Becker, D., E. Kemper, J. Schell, and R. Masterson. 1992. New plant binary vectors with selectable markers located proximal to the left T-DNA border. Plant Mol. Biol. 20:1195–1197.
Belanger, F.C., S. Bonos, and W.A. Meyer. 2004. Dollar spot resistant hybrids between creeping bentgrass and colonial bentgrass. Crop Sci. 44:581–586.
Belanger, F.C., T.R. Megaher, P.R. Day, K.A. Plumley, and W.A. Meyer. 2003a. Inter-specifi c hybridization between Agrostis stolonifera and related Agrostis species under fi eld conditions. Crop Sci. 43:240–246.
Belanger, F.C., K.A. Plumley, P.R. Day, and W.A. Meyer. 2003b. Inter-specifi c hybridization as a potential method for improvement of Agrostis species. Crop Sci. 43:2172–2176.
Brown, M.J., W.S. Hail, R.S. Kohn, and M. Ree. 2001. Transgenic crops in natural habitats. Crawley. Nature 209:682–683.
CFSAN. 1995. Biotechnology consultation note the fi le BNF No. 000029: Glufosinate-tolerant Corn. CFSAN/Offi ce of Premarket Approval, Center for Food Safety and Applied Nutrition.
CFSAN. 1998. Biotechnology consultation note the fi le BNF No. 000055: Glufosinate-tolerant Soybean lines. CFSAN/Offi ce of Premarket Approval, Center for Food Safety and Applied Nutrition.
Choi, J.S., and G.M. Yang. 2004. Development of new hybrid cultivar ‘Senock’ in Zoysia grass. Kor. Turfgrass Sci. 18:1–12.
Codex Alimentarius Commission. 2003. Report on the 4th session of the Codex ad hoc intergovernmental task force on foods derived from biotechnology (ALINORM 03/34A). In Codex principles and guidelines on foods derived from biotechnology. 26th session, Rome, Italy. 30 June–7 July 2003. Joint FAO/WHO Food Standards Programme, Food and Agriculture Organization.
European Commission. 2003. Prepared for the scientifi c steering committee by the joint working group on novel foods and GMOs. Scientifi c Committees on Plants, Food, and Animal Nutrition. p. 1–26. In Guidance document for the risk assessment of genetically modifi ed plants and derived food and feed. Health and Consumer Protection Directorate-General.
FAO/WHO. 1996. Biotechnology and food safety. Report of a joint FAO/WHO
218 Journal of Environmental Quality • Volume 37 • January–February 2008
consultation. Food and Agricultural Organization, Food and Nutrition Paper 61. Rome, Italy.
FAO/WHO. 2000. Safety aspects of genetically modifi ed foods of plant origin. Report of a joint FAO/WHO expert consultation on foods derived from biotechnology. Geneva, Switzerland, WHO/SDE/PHE/FOS/00.6.
FAO/WHO. 2001. Joint FAO/WHO expert consultation on foods derived from biotechnology evaluation of allergenicity of genetically modifi ed foods. Rome, Italy.
FDA. 1992. Statement of policy: Foods derived from new plant varieties. Fed. Regist. 57:22984–23005.
Ge, Y., T. Norton, and Z.Y. Wang. 2006. Transgenic Zoysia (Zoysia japonica) plants obtained by Agrobacterium-mediated transformation. Plant Cell Rep. 25:792–798.
Health Canada. 2001. Elements of precaution: Recommendations for the regulation of food biotechnology in Canada. An expert panel report on the future of food biotechnology prepared by the royal society of Canada at the request of health Canada. Canadian Food Inspection Agency and Environment Canada, Ottawa, Canada.
Herouet, C., D.J. Esdaile, B.A. Mallyon, E. Debruyne, A. Schulz, T. Currier, K. Hendrickx, R. Van der Klis, and D. Rouan. 2005. Safety evaluation of the phosphinothricin acetyltransferase proteins encoded by the pat and bar sequences that confer tolerance to glufosinate-ammonium herbicide in transgenic plants. Regul. Toxicol. Pharmacol. 41:134–149.
Honda, M., and M. Kono. 1963. Morphological and anatomical study of grass: Specially about Zoysia japonica Steud. Tech. Bull. Fac. Hortic. Chiba Univ. 11:1–22 (In Japanese).
Hong, K.H., and D.Y. Yeam. 1985. Studies on interspecifi c hybridization in Korean lawngrass (Zoysia spp.). J. Kor. Hortic. Sci. 26:169–178.
Hwang, Y.S., and J.S. Choi. 1999. Eff ect of mowing interval, aeration, and fertility level in the turf quality and growth of zoysiagrass (Zoysia japonica Steud.). Kor. Turfgrass Sci. 13:79–90.
Inokuma, C., K. Sugiura, C. Cho, R. Okawara, and S. Kaneko. 1998. Transgenic Japanese lawngrass (Zoysia japonica Steud.) plants regenerated from protoplasts. Plant Cell Rep. 17:334–338.
Institute of Science in Society. 2003. Animals avoid GM food, for good reasons: Novotny E. (2002) Report for the Chardon LL hearing. Non-suitability of genetically engineered feed for animals. ISIS Press Release 13/12/03. ISIS, London.
Jones, D.D., and J.H. Maryanski. 1991. Safety considerations in the evaluation of transgenic plants for human foods. p. 64–82. In M.A. Levin and H.S. Strauss (ed.) Risk assessment in genetic engineering. McGraw-Hill, New York.
Kim, H.K., K.S. Kim, Y.K. Joo, K.H. Hong, K.N. Kim, J.P. Lee, S.Y. Mo, and D.H. Kim. 1996. Variation of the morphological characteristics in the accessions of Zoysia species and their hybrid lines. Kor. Turfgrass Sci. 10:1–11.
Kim, Y.L., S.K. Lee, and S.H. Oh. 1987. Study of allergy skin tests with Korean pollen extracts. Yonsei Med. J. 28:112–118.
Kitamura, F. 1967. Lawn grass and plant for lawn. Kajimashyoten, Tokyo, Japan (In Japanese).
Kwon, J.S., J.S. Suh, H.Y. Weon, and J.S. Shin. 1998. Evaluation of soil microfl ora in salt accumulated soils of plastic fi lm house. Kor. J. Soil Sci. Fertil. 31:204–210.
Lee, H.Y., C.H. Lee, H.I. Kim, W.D. Han, W.E. Choi, J.H. Kim, and Y.P. Lim. 1998. Development of bialaphos-resistanttolerant transgenic rice using Agrobacterium tumefaciens. Kor. J. Plant Tissue Cult. 25:283–288.
Li, R.F., J.H. Wei, H.G. Wang, J. He, and Z.Y. Sun. 2006. Development of highly regenerable callus lines and Agrobacterium-mediated transformation of Chinese lawngrass (Zoysia sinica Hance) with a cold inducible transcription factor, CBF1. Plant Cell Tissue Organ Cult. 85:297–305.
Liu, L., J.W. Kloepper, and S. Tuzum. 1995. Induction of systemic tolerance in cucumber Fusarium wilt by plant growth promoting rhizobacteria. Phytopathology 85:695–698.
Meera, M.S., M.B. Shivanna, K. Kageyama, and M. Hyakumachi. 1993. Induction of systemic tolerance in cucumber plants using turfgrass rhizosphere fungi. Ann. Phytopathol. Soc. Japan 59:279 (In Japanese).
Meera, M.S., M.B. Shivanna, K. Kageyama, and M. Hyakumachi. 1994. Plant growth promoting fungi from Zoysiagrass rhizosphere as potential inducers of systemic tolerance in cucumber. Phytopathology 84:1399–1406.
Mossinger, H., and W. Dietrich. 1998. Activation of hemostasis during cardiopulmonary bypass and pediatric aprotinin dosage. Ann. Th orac. Surg. 65:45–50.
Nakayama, Y., and H. Yamaguchi. 2002. Natural hybridization in wild soybean (Glycine max ssp. Soja) by pollen fl ow from cultivated soybean (Glycine max
ssp. max) in a designed population. Weed Biol. Manage. 2:25–30.
Niwa, K., and N. Takanashi. 1943. Th e seed of Japanese lawn grass and structure. J. Jap. Inst. Landscape Arch. 10:27–32 (In Japanese).
OECD. 1999. Consensus document on general information concerning the genes and their enzymes that confer tolerance to phosphinothricin herbicide. Report No. 11. OECD, Paris. ENV/JM/MON (99) 13. Available at http://www.oecd.org/ehs/ (verifi ed 26 Oct. 2007).
OECD. 2001. Acute oral toxicity. Fixed dose procedure. OECD Guidelines for the testing of chemicals. Test No. 420. OECD, Paris.
OECD. 2002. Module II: Herbicide biochemistry, herbicide metabolism, and the residues in glufosinate-ammonium (phosphinothricin)-tolerant transgenic plants. OECD, Paris. EN/JM/MONO (2002) 13.
Park, M.S., and S.H. Yu. 2005. Plant growth promoting fungi isolated from rhizosphere of Zoysiagrass in Korea. Kor. J. Mycol. 33:30–34.
Reichman, J.R., L.S. Watrud, E.H. Lee, C.A. Burdick, M.A. Bollman, M.J. Storm, G.A. King, and C. Mallory-Smith. 2006. Establishment of transgenic herbicide-resistant creeping bentgrass (Agrostis stolonifera L.) in nonagronomic habitats. Mol. Ecol. 15:4243–4255.
Schmidt, J.O. 1994. Toxicology of venoms from the honeybee genus apis. Toxicon 33:917–927.
Sjoblad, R., J.T. McClintock, and R. Engler. 1992. Toxicological considerations for protein components of biological pesticide products. Regul. Toxicol. Pharmacol. 15:3–9.
Song, Z.P., B.R. Lu, Y.G. Zhu, and J.K. Chen. 2003. Gene fl ow from cultivated rice to the wild species Oryza rufi pogon under experimental fi eld conditions. New Phytol. 157:657–665.
Suh, J.S., J.S. Kwon, and G.H. Chon. 2003. Eff ects of parent rocks on soil microbial diversity. Kor. J. Soil Sci. Fertil. 36:127–133.
Th omas, K., M. Aalbers, G.A. Bannon, M. Bartels, R.J. Dearman, D.J. Esdaile, T.J. Fu, C.M. Latt, N. HadWeld, C. Hatzos, S.L. HeXe, J.R. Heylings, R.E. Goodman, B. Henry, C. Herouet, M. Holsapple, G.S. Ladics, T.D. Landry, S.C. MacIntosh, E.A. Rice, L.S. Privalle, H.Y. Teiner, R. Teshima, K. Th omas, R. Van Ree, M. Woolhiser, and J. Zawodny. 2004. A multi-laboratory evaluation of a common in vitro pepsin digestion assay protocol used in assessing the safety of novel proteins. Regul. Toxicol. Pharmacol. 39:87–98.
Th ompson, C.J., N.R. Movva, R. Tizard, R. Crameri, J.E. Davies, M. Lauwereys, and J. Bottermann. 1987. Characterization of the herbicide-tolerance gene bar from Streptomyces hygroscopicus. EMBO J. 6:2519–2523.
Toki, S. 1992. Expression of a Maize ubiquitin gene promoter-bar chimeric gene in transgenic rice plants. Plant Physiol. 100:1503–1507.
Toyama, K., C.H. Bae, J.G. Kang, Y.P. Lim, T. Adachi, K.Z. Riu, P.-S. Song, and H.Y. Lee. 2003. Production of herbicide-tolerant zoysiagrass by Agrobacterium-mediated transformation. Mol. Cells 16:19–27.
Toyama, K., C.H. Bae, M.S. Seo, I.J. Song, Y.P. Lim, P.-S. Song, and H.Y. Lee. 2002. Overcoming of barriers to transformation in monocot plants. Kor. J. Plant Biotechnol. 4:135–141.
Watrud, L.S., L.E. Henry, A. Fairbrother, C. Burdick, J.R. Reichman, M. Bollman, M. Storm, G. King, and P.K. Van de Water. 2004. Evidence for landscape-level, pollen-mediated gene fl ow from genetically modifi ed creeping bentgrass with CP4 EPSPS as a marker. Proc. Natl. Acad. Sci. USA 101:14533–14538.
Wehrmann, A., A. Van Vliet, C. Opsomer, J. Botterman, and A. Schulz. 1996. Th e similarities of bar and pat gene products make them equally applicable for plant engineers. Nat. Biotechnol. 14:1274–1278.
WHO. 1998. Food contamination monitoring and management system/food regional diets. Regional per capita consumption of raw and semi-processed agricultural commodities. Food Safety Dep., Global Environmental Monitoring System. WHO, Geneva, Switzerland.
Wipff , J.K., and C. Fricker. 2001. Gene fl ow from transgenic creeping bentgrass (Agrostis stolonifera L.) in the Willamette valley, Oregon. Int. Turfgrass Soc. Res. J. 9:224–242.
Yaneshita, M., R. Nagasawa, M.C. Engelke, and T. Sasakuma. 1997. Genetic variation and interspecifi c hybridization among natural populations of zoysiagrass detected by RFLP analyses of chloroplast and nuclear DNA. Genes Genet. Syst. 72:173–179.
Youngner, V.D. 1961. Growth and fl owering of Zoysia species in response to temperature, photoperiods and light intensities. Crop Sci. 1:91–93.
Yu, T., D.Y. Yeam, Y.J. Kim, and S.J. Kim. 1974. Morphological studies on Korean lawn grasses (Zoysia spp.). J. Kor. Hortic. Sci. 15:79–91.
Yun, H.K. 1996. Molecular cloning of two pathogenesis-related protein genes from Nicotiana glutinosa L. and their expression during plant disease tolerance. Ph.D. thesis. Chungnam Univ., Korea.