Plant Cell Rep (2008) 27:903–909
DOI 10.1007/s00299-008-0519-8
GENETIC TRANSFORMATION AND HYBRIDIZATION
Agrobacterium tumefaciens-mediated transformation of taro
(Colocasia esculenta (L.) Schott) with a rice chitinase gene
for improved tolerance to a fungal pathogen Sclerotium rolfsii
Xiaoling He Æ Susan C. Miyasaka Æ
Maureen M. M. Fitch Æ Paul H. Moore Æ
Yun J. Zhu
Received: 25 September 2007 / Revised: 2 February 2008 / Accepted: 10 February 2008 / Published online: 27 February 2008
Ó Springer-Verlag 2008
Abstract Taro (Colocasia esculenta) is one of the most
important crops in the Pacific Islands, however, taro yields
have been declining in Hawaii over the past 30 years partly
due to diseases caused by oomycete and fungal pathogens.
In this study, an efficient Agrobacterium tumefaciens-
mediated transformation method for taro is first reported. In
total, approximately 200 pieces (8 g) of embryogenic cal-
luses were infected with the super-virulent A. tumefaciens
strain EHA105 harboring the plant transformation plasmid
pBI121/ricchi11 that contains the rice chitinase gene ric-
chi11. The presence and expression of the transgene
ricchi11 in six independent transgenic lines was confirmed
using polymerase chain reaction (PCR) and reverse tran-
scription-PCR (RT-PCR). Southern blot analysis of the six
independent lines indicated that three out of six (50%) had
integrated a single copy of the transgene, and the other
three lines had two or three copies of the transgene.
Communicated by P. Ozias-Akins.
X. He
Department of Tropical Plant and Soil Sciences,
University of Hawaii, 3190 Maile Way,
Honolulu, HI 96822, USA
S. C. Miyasaka (&)
Department of Tropical Plant and Soil Sciences,
University of Hawaii, 875 Komohana St,
Hilo, HI 96720, USA
e-mail: miyasaka@hawaii.edu
M. M. M. Fitch
P. H. Moore
USDA ARS, PBARC, 99-193 Aiea Heights Drive,
Aiea, HI 96701, USA
Y. J. Zhu
Hawaii Agriculture Research Center,
99-193 Aiea Heights Drive, Aiea, HI 96701, USA
Compared to the particle bombardment transformation of
taro method, which was used in the previous studies, the
Agrobacterium-mediated transformation method obtained
43-fold higher transformation efficiency. In addition, these
six transgenic lines via Agrobacterium may be more
effective for transgene expression as a result of single-copy
or low-copy insertion of the transgene than the single line
with multiple copies of the transgene via particle bom-
bardment. In a laboratory bioassay, all six transgenic lines
exhibited increased tolerance to the fungal pathogen Scle-
rotium rolfsii, ranging from 42 to 63% reduction in lesion
expansion.
Keywords Agrobacterium tumefaciens

Transformation
Taro
Colocasia esculenta
Chitinase

Fungal disease tolerance
Abbreviations
BA Benzyladenine
NAA a-Naphthaleneacetic acid
MS Murashige and Skoog plant culture medium
AS Acetosyringone
(3, 5-dimethoxy-4-hydroxy-acetophenone)
YEB Yeast extract broth
Introduction
Taro (Colocasia esculenta (L.) Schott) is a monocotyle-
donous crop belonging to the Araceae family. It is one of
the most important staple food crops in the Pacific Islands,
and is widely cultivated throughout the South Pacific,
South America, Asia, Africa, and the Caribbean (Kreike
et al. 2004; Wang 1983). Taro corms or cormels have much
123
904
smaller starch grains that are fairly rich in the soluble
starch called amylose. Amylose is an excellent starch for
people with digestive problems (Perez et al. 2005). Further,
the taro corms are good sources of protein, calcium, and
phosphorus compared to other starchy crops (Hussain et al.
1984). In addition, taro leaves are consumed as a vegetable,
which provides good sources of dietary fiber and vitamin C
(Ferguson et al. 1992). However, many taro cultivars are
susceptible to various pathogens (Ooka 1994). In Hawaii,
taro yields have been declining over the past 30 years
partially as a result of diseases caused by oomycete and
fungal pathogens. According to the Hawaii Agricultural
Statistics Service (2006), Hawaiian taro production in 2005
was the lowest in records kept since 1946. Oomycete and
fungal diseases were estimated to decrease taro yields up to
50% (Miyasaka et al. 2001). Sclerotium or southern blight,
caused by Sclerotium rolfsii Sacc., is a major fungal dis-
ease of dryland-grown (non-flooded) taro. It results in rots
of the petiole and corm, as well as an overall stunting of
plants (Ooka 1994).
Genetic engineering offers a new method to improve
crop yields, quality or disease resistance via transforming a
single gene or several genes of interest into target plants.
Genetic transformation of several plants with a rice chiti-
nase gene increased fungal resistance in cucumber
(Kishimoto et al. 2002), grapevine (Yamamoto et al. 2000),
rice (Lin et al. 1995), Italian ryegrass (Takahashi et al.
2005) and tobacco (Zhu et al. 1994).
There are only two previous reports of the genetic
transformation of taro. Fukino et al. (2000) transformed a
reporter gene b-glucuronidase (gus) gene using particle
bombardment into a Japanese taro cultivar Eguimo. How-
ever, the efficiency of transformation was very low,
yielding only two transgenic lines from a total of 192 g of
regenerative calluses. In addition, at least five copies of the
transgene were inserted into the taro genome (Fukino et al.
2000). A similar particle bombardment transformation
method was applied to introduce a rice chitinase gene into
a Chinese taro cultivar, Bun Long (He 2006). The trans-
formation efficiency was also very low, yielding only one
transgenic line from a total of 1,350 pieces (60 g) of
regenerative calluses. Thirteen copies of the transgene were
inserted into the taro genome (He 2006). That could result
in gene rearrangement and gene silencing (Luo et al. 2004;
Veluthambi et al. 2003).
Agrobacterium-mediated transformation transfers a
defined segment of DNA (T-DNA) from Agrobacterium
and integrates it into the target plant genome. Agrobacte-
rium-mediated transformation has many unique advantages
over particle bombardment transformation. It is simpler
and less expensive than particle bombardment transfor-
mation (Veluthambi et al. 2003). In addition, Agrobacte-
rium-mediated transformation usually results in insertion of
123
Plant Cell Rep (2008) 27:903–909
a single-copy or a low-copy number of the transgene and
presumably lowers incidences of gene rearrangement and
gene silencing (Luo et al. 2004; Veluthambi et al. 2003). In
a survey of recorded instances of gene silencing in
monocots (Iyer et al. 2000), 19 instances of gene silencing
were found in transgenic monocot plants including rice,
barley, maize, oats, pearl millet, ryegrass, sugarcane
and wheat via particle bombardment transformation, but
only three instances of gene silencing were found in
transgenic rice and wheat via Agrobacterium-mediated
transformation.
Taro is a monocotyledonous plant. Previously, a major
limitation of Agrobacterium-mediated transformation
originated from the fact that most monocotyledonous
plants are not natural hosts of Agrobacterium and do not
secrete the molecular signal phenolics, such as acetosy-
ringone, which are needed to induce vir genes (Veluthambi
et al. 2003). Now in advanced transformation systems,
acetosyringone is applied (Veluthambi et al. 2003). In
addition, more efficient Agrobacterium strains (e.g. super-
virulent Agrobacterium strain EHA 105) are used along
with more efficient transformation vectors and transfor-
mation conditions (Veluthambi et al. 2003). To date, a
wide range of monocotyledonous plants including rice,
maize, wheat, and barley have been successfully trans-
formed by Agrobacterium-mediated transformation (Luo
et al. 2004; Veluthambi et al. 2003).
In this study, we developed an efficient Agrobacterium-
mediated transformation method of monocotyledonous
taro, inserted the rice chitinase gene ricchi11, and dem-
onstrated increased disease tolerance to the fungal
pathogen Sclerotium rolfsii. To our knowledge, there is no
previous report of taro being transformed via Agrobacte-
rium tumefaciens.
Materials and methods
Plant material
Chinese taro cultivar Bun Long was chosen as the plant
transformation material. Cultivar Bun Long is the most
important cultivar grown in Hawaii for consumption of
leaves and for production of chips from starchy corms. The
plant material was obtained from the University of
Hawaii’s Waiakea Experiment Station.
Highly regenerative calluses were induced from the
apical meristems of corms using Murashige and Skoog
(MS) medium (Murashige and Skoog 1962) supplemented
with 2 mg L-1 BA and 1 mg L-1 NAA (M5 medium) that
was adjusted to pH 5.8 and autoclaved at 121°C for 30 min
(He 2006). These calluses were selected as the transfor-
mation explant tissues.
Plant Cell Rep (2008) 27:903–909
Agrobacterium strain and plasmid
The super-virulent Agrobacterium strain EHA105 and the
plant transformation plasmid pBI121/ricchi11 (He 2006)
were used to introduce the ricchi11 gene into taro. The
plasmid pBI121/ricchi11 was generated by inserting the
rice chitinase gene with its own terminator and the CaMV
35S promoter into the unique HindIII site of pBI121.
Plasmids were selected where the ricchi11 gene had been
cloned in the same transcriptional direction as that of other
genes on pBI121. In addition to the ricchi11 gene, pBI121/
ricchi11 also contains the selectable marker gene neomycin
phosphotransferase II (nptII) and the reporter gene b-glu-
curonidase (gus). The pBI121 plasmid was obtained from
Clontech (Mountain View, CA. http://www.clontech.com).
The rice chitinase gene ricchi11 was obtained from Dr. S.
Muthukrishnan at Kansas State University.
Agrobacterium-mediated transformation method
Agrobacterium-mediated transformation of taro cultivar
Bun Long with the ricchi11 gene followed the method of
Chen and Kuehnle (1996) for anthurium except that
explants were wounded, the Agrobacterium inoculation
period was extended from 8 to 30 min, and kanamycin and
carbenicillin were replaced with geneticin and cefotaxime
to select transformants and eliminate Agrobacterium over-
growth, respectively. A single colony of the Agrobacterium
strain EHA105 carrying the binary vector pBI121/ricchi11
was cultured in 3 mL YEB (Yeast extract broth) medium
containing 50 lg mL-1 kanamycin and 25 lg mL-1 rif-
ampicin (for EHA105). The medium was adjusted to pH 7.2
and autoclaved at 121°C for 30 min before adding antibi-
otics. The medium was shaken at 250 rpm and 28°C for 24–
48 h until the bacterial suspension became turbid
(OD600 = 0.5–1.0). Two micro liters of 0.3 M acetosyrin-
gone (AS) was added into the turbid agrobacterial
suspension and mixed well. The bacteria containing AS
suspension was diluted tenfold with YEB liquid medium
(final concentration of 20 lM AS). Agrobacterium-medi-
ated transformation experiments were repeated four times.
In each experiment, approximately 2 g of highly regener-
ative calluses were chopped and cut into 50 pieces to make
cutting wounds, and immersed into the Agrobacterium
suspension solution for 30 min, then transferred to the co-
cultivation hormone-free MS medium plates (10 pieces/
plate) and co-cultivated with Agrobacterium for four days at
25 ± 2°C in the dark. Then calluses were blotted on sterile
tissue papers to remove excess Agrobacterium from the
surface. The calluses were transferred to selection medium
(M5 medium containing 50 mg L-1 geneticin and
250 mg L-1 cefotaxime) and selected for a total of 90 days
with sub-culturing every 30 days at 25 ± 2°C in the dark.
905
Calluses were transferred monthly onto fresh selection
medium, and examined at least once a week for re-growth
of Agrobacterium. The explants were rinsed with sterile
water and cultures with over-grown Agrobacterium were
discarded or sub-cultured in an attempt to prevent the re-
growth of Agrobacterium. After 90 days, healthy, light
yellow or green calluses were selected for shoot and root
induction on M15 medium (MS medium supplemented
with 4 mg L-1 BA) (He 2006) with 50 mg L-1 geneticin
and 250 mg L-1 cefotaxime at 25 ± 2°C with a 16 h
photoperiod, at a light intensity of 15 lmol m-2 s-1. Then,
multiple shoots were sub-cultured on fresh M15 selection
medium with 50 mg L-1 geneticin and 125 mg L-1 cefo-
taxime every 30 days at 25 ± 2°C with a 16 h
photoperiod, at a light intensity of 15 lmol m-2 s-1.
When 2–3 leaves emerged, the geneticin level was reduced
to 35 mg L-1. Geneticin-resistant plantlets were trans-
ferred into Magenta boxes with liquid M15 selection
medium maintained at 25 ± 2°C with a 16 h photoperiod,
at light intensity of 15 lmol m-2 s-1, and shaken at
95 rpm. Multiple plantlets were sub-cultured on fresh M15
selection medium every 30 days. The roots of plantlets
were immersed in the medium and whole plantlets were
floating on the medium without rafts. Leaves were ana-
lyzed at this stage for the presence of the transgene by
PCR, RT-PCR and Southern blot analyses. All M5 and
M15 media were adjusted to pH 5.8 and autoclaved at
121°C for 30 min before adding antibiotics.
GUS histochemical assay
The method of Jefferson (1987) was used for a GUS his-
tochemical assay. Non-transformed shoots and the
secondary multiplied shoots from the transformed shoot
lines were randomly selected for GUS assay. These trans-
formed shoot lines were induced from the transformed
calluses and then multiplied and selected for three months
in the selection medium. In a 96-well tissue culture plate,
the shoot cuttings were injured by chopping or cutting
with a sterile scalpel and submerged into 50 mM
sodium phosphate buffer (pH 7.0) containing the substrate
X-Gluc (5-bromo-4-chloro-3-indolyl-beta-D-glucuronide).
The plate was incubated overnight at 37°C. The shoot
cuttings were observed for the blue color that indicates the
presence of GUS activity. Shoot cuttings were soaked in
70% ethanol for 2 days to remove the chlorophyll.
Verification of presence of the ricchi11 gene
using PCR analysis
Non-transformed plantlets and transformed plantlets in the
magenta boxes were randomly selected for PCR analysis.
These transformed plantlets were multiplied from
123
906
transformed shoot lines, which had been cultured in
selection medium for 6 months with sub-culturing every
30 days. Genomic DNA was extracted from approximately
30 mg (fresh weight) of plantlet leaves using the modified
sodium dodecyl sulfate (SDS) method described by Lin
et al. (2001). PCR analysis followed the method described
by He (2006). The primer pair used to amplify a 540-bp
fragment of the integrated ricchi11gene was: G11U (21
mer) 50 -CCGCGCTAAGGGCTTCTACAC-30 and G11L
(20 mer) 50 -CACTCCACACCGCCGTTGAT-30 ). The PCR
reactions were performed in 50 ll volume consisting of
1 ll template DNA (20–30 ng), 2 ll of each primer
(20 lmol), 5 ll dNTPs (2 mM), 5 ll 10x Taq buffer
(Promega, Madison, WI), 1 unit Taq polymerase (Promega,
Madison, WI), and 36.5 ll H2O. Amplification of ricchi11
fragments was performed for 30 cycles at 94°C for 30 s,
55°C for 45 s and 72°C for 45 s, for denaturing, annealing
and primer extension, respectively.
Verification of expression of the ricchi11 gene using
reverse transcription-PCR (RT-PCR) analysis
One-year-old non-transformed plants and PCR-positive
plants were randomly selected for RT-PCR analysis. Total
RNA was isolated from approximately 100 mg of fresh
plant leaves ground to a fine powder with liquid N2
according to a method described by Bugos et al. (1995).
Reverse transcription (RT) was conducted with oligo-dT
primers and reverse transcriptase in the RT-PCR kit
(Promega, Madison, WI). The reverse transcripts were used
as templates for PCR. The PCR primers for the ricchi11
gene and the PCR reaction system and conditions were
described above in the section on PCR analysis. As a
control, PCR analysis also was conducted using total RNA
without reverse transcription to confirm that the PCR
product was derived from mRNA and not from the con-
taminating DNA.
Southern blot analysis
Six-month-old non-transformed plantlets and PCR-positive
plantlets were randomly selected for Southern blot analy-
sis. Genomic DNA was extracted from approximately 80–
100 mg of fresh leaves using the modified SDS method
described by Lin et al. (2001). DNA (20 lg) was digested
with the enzyme EcoRI that has a single digestion site in
the T-DNA region of the plasmid. Digested DNA was
fractionated by electrophoresis in a 1% agarose gel, then
alkali-blotted onto a Hybond N + membrane (Amersham,
UK) according to the manufacturer’s instructions. The PCR
amplification product (a 540-bp fragment of the ricchi11
gene from the plasmid pBI121/ricchi11) was labeled with
AlkPhos Direct (Amersham, UK) according to the
123
Plant Cell Rep (2008) 27:903–909
manufacturer’s instructions, and used as a chemifluorescent
probe. Hybridization and detection were performed
according to the manufacturer’s protocol (Amersham, UK).
Bioassay of transgenic plantlets challenged
by the fungal pathogen Sclerotium rolfsii
Six-month-old transgenic plantlets multiplied from the
independent lines were used in the bioassay with the fungal
pathogen Sclerotium rolfsii. The method was similar to that
described by He (2006). One mature, brownish black
sclerotium was placed on the shoot base of a plantlet. One
inoculated plantlet was placed on moistened filter paper in
a Petri dish sealed to maintain 100% humidity. The Petri
dishes were incubated at room temperature with a 10 h
photoperiod, at light intensity of 15 lmol m-2 s-1. Three
non-transformed plantlets and three transgenic plantlets per
transgenic line (6 lines total) were inoculated for each trial
and each trial was repeated three times. After inoculation,
the inoculated plantlets were observed daily for lesion
initiation and expansion in 7 days. The lengths of the
necrotic lesions on the shoots of plantlets were measured
on the fourth day after inoculation and data were analyzed
using the general linear model (GLM) program of SAS
software (Statistical Analysis System, Cary, NC).
Results
Geneticin selection for transformants and GUS
histochemical assay
In total, approximately, 200 calluses (8 g) were co-culti-
vated with the super-virulent strain EHA105: pBI121/
ricchi11. After 90 days on the 50 mg L-1 geneticin
selection medium, 45% of calluses survived. Twenty
independent lines produced shoots after the calluses were
transferred to the shoot-inducing selection medium for
60 days. Of these 20 shoot lines, 10 lines showed uniform
blue colouration throughout the shoots, suggesting inte-
gration into the plant genome and stable expression of the
gus gene (Fig. 1).
Verification of presence and expression of ricchi11
gene in transformed taro using PCR analysis
and RT-PCR analysis
PCR analysis was conducted using leaves from the 20
surviving geneticin-resistant shoot lines and the non-
transformed control. The expected 540 bp PCR product
specific for the ricchi11 gene fragment was observed in
total DNA extracted from six independent GUS-positive
shoot lines and in the plasmid pBI121/ricchi11 positive
Plant Cell Rep (2008) 27:903–909 907

M P H C1 C2 C3 C4 C5 C6 NT

1353bp

1078bp

872bp
603bp
540bp

310bp

Fig. 2 PCR analysis of transgenic lines transformed with

EHA105:pBI121/ricchi11. M molecular weight marker AX174/Hae-
III, P Plasmid pBI121/ricchi11 control, H Water control, NT Non-
transformed plant control, C1–C6 Transgenic lines 1–6

M -C1 C1 -C2 C2 -C3 C3 -C4 C4 -C5 C5 -C6 C6 NT

1353bp
1078bp
872bp
603bp 540bp
310bp
Fig. 1 GUS histochemical assay for a shoot from an independent
shoot line induced from callus transformed with pBI121/ricchi11 (T)
after a 90 day selection period and a shoot from non-transformed Fig. 3 RT-PCR analysis of transgenic lines transformed with
control (NT) EHA105: pBI121/ricchi11. M molecular marker AX174/HaeIII, NT
Non-transformed plant control, C1–C6 Transgenic lines 1–6, -C1–-C6
Negative RT of Transgenic lines 1–6
control, indicating that the ricchi11 gene had been suc-
cessfully transformed into these 6 lines (Fig. 2). Also, the
expected 540 bp PCR product specific for the ricchi11 C1 C2 C3 C4 C5 C6 NT P
gene fragment was amplified in the reverse-transcription
products of total RNA extracted from these 6 independent 23.1kb
shoot lines (Fig. 3). Total DNA of the non-transformed
control plant did not exhibit this 540 bp PCR product.
9.4kb
Similarly, the reverse-transcription product of total RNA
extracted from the non-transformed control plant did not 6.5kb
result in this 540 bp PCR product. In addition, total RNA
without reverse transcription extracted from these six
4.4kb
independent shoot lines did not exhibit the 540 bp PCR
product, confirming that there was no DNA contamination
in the RNA extractions. 2.3kb

Southern blot analysis
EcoRI restriction digests of genomic DNA extracted from
the six PCR-positive transgenic lines showed various band
patterns larger than the 1.9 kb of ricchi11 (Fig. 4). A single
band was found in three transgenic lines (C1, C2 and C4),
two bands in two transgenic lines (C3 and C6), and three
bands in one transgenic line (C5) (Fig. 4). Since the
T-DNA region containing the ricchi11 gene has only one
EcoRI digestion site, a single band in the Southern blot
indicated a single-copy transgene insertion into the taro
genome. Two and three bands indicated two and three-copy
Fig. 4 Southern blot analysis of six independent lines (C1–C6) and a
non-transformed control (NT). A 20 lg aliquot of the genomic DNA
extracted from leaves and 200 pg plasmid pBI121/ricchi11 DNA (P)
were digested with EcoRI, which cuts only once within the T-DNA
region. Blots were hybridized with the PCR product of the rice
chitinase gene labeled with the AlkPhos direct labeling and detection
system
insertions of the transgene, respectively. Therefore, three
independent lines contained a single-copy insertion with no
apparent gene rearrangement and three other independent
lines contained two or three-copy insertions. In addition,
the six bands of different sizes provide evidence that the

123
908
transgene was integrated into different sites in the taro
genome.
Bioassay of the transgenic plantlets challenged
by the fungal pathogen Sclerotium rolfsii
Necrotic lesions of non-transformed plantlets were visible
on day 1 after inoculation with the fungal pathogen Scle-
rotium rolfsii. In contrast, necrotic lesions on all six
transgenic plantlet lines were visible only on day 3 after
inoculation. The lengths of lesions measured on day 4 after
inoculation was significantly shorter (P \ 0.05) in all 6
transgenic lines with the chitinase gene ricchi11 than those
of the non-transformed plantlet controls (Figs. 5, 6). On
average, the lesions of transgenic lines were approximately
42–63% shorter than those of the non-transformed controls.
The reduction in lesion length of these 6 lines (42–63%)
was greater than that of the 1 transgenic line (35%)
achieved via particle bombardment in an earlier study (He
2006). The non-transformed plantlets died within 3.5 days
after inoculation, whereas, the transgenic plantlets died
within 7 days after inoculation.
Discussion
In this study, we have developed an efficient transforma-
tion method for the monocotyledonous taro cultivar Bun
Long using Agrobacterium tumefaciens. The effects of
Agrobacterium strain LBA4404 and EHA105 on transfor-
mation were evaluated in a preliminary test (He 2006). We
found that Agrobacterium strain LBA4404 appeared to be
ineffective, and the super-virulent Agrobacterium strain
EHA105 was able to infect taro calluses in the presence of
20 lM AS. In addition, several important transformation
parameters were optimized, such as the wounding
Fig. 5 Laboratory bioassay of
plantlets of the transgenic lines
C1–C6 and two non-
transformed plantlet controls
(NT) challeged with Sclerotium
rolfsii, 4 days after inoculation.
The black arrow indicates one
mature, brownish black round
sclerotium placed on the shoot
base of each plantlet. Necrotic
brown lesions of NT were
visible and extended throughout
whole plantlet at day 4 after
inoculation. Necrotic brown
lesions of C1–C6 were visibly
shorter than NT at day 4 after
inoculation
123
Plant Cell Rep (2008) 27:903–909
30
a
25
Lesion (mm)
20
b b
15 b b b
b
10
5
0
NT C1 C2 C3 C4 C5 C6
Fig. 6 Lesion length (mm) on Sclerotium rolfsii-inoculated plantlets
of non-transformed control (NT) and transgenic lines C1–C6
measured 4 days after inoculation. Graph bars represent the mean ±
SD values of three replicates (each replicate consisted of three NT
plantlets and three plantlets of C1–C6). Different letters (a, b) indicate
significant differences (P \ 0.05) using the least significant differ-
ence (LSD) test of SAS (Statistical Analysis System, Cary, NC)
between NT and C1–C6 challenged by Sclerotium rolfsii
pre-treatment of explants, the Agrobacterium inoculation
period, the AS concentration, and the co-cultivation period
(He 2006). Also, antibiotics for transgenic line selection
and Agrobacterium elimination were assessed in the pre-
liminary experiments (He 2006). Geneticin and cefotaxime
were more effective than kanamycin and carbenicillin to
select transformants and eliminate Agrobacterium growth
respectively. Using the present optimal protocol, starting
from 200 calluses (8 g), we produced six independent
transgenic taro lines, or an efficency rate of 3%. Compared
to the particle bombardment transformation method of taro
described by Fukino et al. (2000) and He (2006) that had an
efficiency rate of 0.07%, the Agrobacterium-mediated
transformation method obtained 43-fold higher transfor-
mation efficiency.
In addition, Southern blot analysis of the six indepen-
dent lines obtained in this study indicated that three out of
Plant Cell Rep (2008) 27:903–909
six lines (50%) had integrated a single copy of the trans-
gene, and the other three lines had low copy numbers (2 or
3 copies) of the transgene. In contrast, multiple copy
insertions (13 copies) of the rice chitinase gene were
obtained via particle bombardment (He 2006). As a result,
transformation of taro via Agrobacterium may be more
effective for transgene expression due to lower incidence
of gene silencing and a greater stability in transmission of
genetic information to the next generation (Luo et al. 2004;
Veluthambi et al. 2003).
Finally, in the laboratory bioassay, all six transgenic
lines exhibited increased tolerance to the fungal pathogen
Sclerotium rolfsii of taro. The reduction in lesion expansion
in these six lines ranged from 42 to 63%. In contrast, the
one transgenic line achieved by particle bombardment had
a lesion reduction rate of only 35% (He 2006). In the lit-
erature, several other transgenic plants with a rice chitinase
gene were reported to have significantly increased resis-
tance to fungal diseases. For example, transgenic rice with
class I rice chitinase gene chi11 showed increased resis-
tance to Rhizoctonia solani, the rice sheath blight pathogen
(Lin et al. 1995). Also, transgenic grapevine with class I
rice chitinase gene rcc2 showed a slight resistance against
Elisinoe ampeina that induces anthracnose (Yamamoto
et al. 2000). In addition, Kishimoto et al. (2002) developed
a transgenic cucumber with rice chitinase gene rcc2, and
demonstrated increased disease resistance to gray mold
caused by the fungal pathogen Botrytis cinerea. Recently,
Takahashi et al. (2005) introduced the rice chitinase gene
rcc2 into Italian ryegrass, and transgenic plants showed
enhanced resistance to crown rust disease caused by the
fungal pathogen Puccinia coronata. Apparently, transfor-
mation of various plants with a rice chitinase gene is an
effective method for enhanced resistance against a broad
range of fungal pathogens. In conclusion, Agrobacterium-
mediated transformation of taro with a rice chitinase gene
is a promising method to increase disease tolerance and
yields of taro. Future plans involve testing of all transgenic
taro cultivar. Bun Long lines in the greenhouse for con-
firmation of increased disease tolerance.
References
Bugos RC, Chiang VL, Zhang XH, Campbell ER, Podila GK,
Campbell WH (1995) RNA isolation from plant tissues
recalcitrant to extraction in guanidine. Biotechniques 19:734–
737
Chen F-C, Kuehnle AR (1996) Obtaining transgenic Anthurium
through Agrobacterium-mediated transformation of etiolated
internodes. J Amer Soc Hortic Sci 121:47–51
Ferguson LR, Roberton AM, Mckenzie RJ, Watson ME, Harris PJ
(1992) Adsorption of a hydrophobic mutagen to dietary fiber
from taro (Colocasia esculenta), an important food plant of the
South Pacific. Nutr Cancer 17(1):85–95
909
Fukino N, Hanada K, Ajisaka H (2000) Transformation of taro
(Colocasia esculenta Schott) using particle bombardment. JARQ
34(3):159–165
Hawaii Agricultural Statistics Service (2006) Taro production hits
record low. NASS Fact Finding for agriculture.
http://www.nass.usda.gov/hi/vegetable/taro.htm
He X (2006) Transformation and regeneration of taro with two plant
disease resistance genes: a rice chitinase gene and a wheat
oxalate oxidase gene. PhD Dissertation. University of Hawaii.
ProQuest. UMI number: 3251049
Hussain M, Norton G, Neale RJ (1984) Composition and nutritive
value of cormels of Colocasia esculenta (L.) scott. J Sci Food
Agric 35:112–119
Iyer LM, Kumpatla SP, Chandrasekharan MB, Hall TC (2000)
Transgene silencing in monocots. Plant Mol Biol 43:323–346
Jefferson RA, Kavanaugh TA, Bevan EW (1987) GUS fusion: Glu-
curonidase as a selective and versatile gene fusion marker in
higher plants. EMBO J 6:3901–3907
Kishimoto K, Nishizawa Y, Tabei Y, Hibi T, Nakajima M, Akutsu K
(2002) Detailed analysis of rice chitinase gene expression in
transgenic cucumber plants showing different levels of disease
resistance to gray mold (Botrytis cinerea). Plant Sci 162:655–
662
Kreike CM, Van Eck HJ, Lebot V (2004) Genetic diversity of taro,
Colocasia esculenta (L.) Schott, in Southeast Asia and the
Pasific. Theor Appl Genet 109:761–768
Lin W, Anuratha CS, Datta K, Potrykus I, Muthukrishnan S, Datta SK
(1995) Genetic engineering of rice for resistance to sheath blight.
Biotechnology N Y 13:686–691
Lin RC, Ding ZS, Li LB, Kuang TY (2001) A rapid and efficient
DNA minipreparation suitable for screening transgenic plants.
Plant Mol Biol Rep 19:379a–379e
Luo H, Hu Q, Nelson K, Longo C, Kausch AP, Chandlee JM, Wipff
JK, Fricker CR (2004) Agrobacterium tumefaciens-mediated
creeping bentgrass (Agrostis stolonifera L.) transformation using
phosphinothricin selection results in a high frequency of single-
copy transgene integration. Plant Cell Rep 22(9):645–652
Miyasaka SC, Hollyer JR, Kodani LS (2001) Mulch and compost
effects on yield and corm rots of taro. Field Crops Res 71:101–
112
Murashige T, Skoog F (1962) A revised medium for rapid growth and
bioassays with tobacco tissue cultures. Physiol Plant 15:473–497
Ooka JJ (1994) Taro diseases, a guide for field identification. Univ.
Hawaii, Hawaii Inst Trop Agr Human Res, Res Ext Ser 148. pp
13
Perez E, Schultz FS, de Delahaye EP (2005) Characterization of some
properties of starches isolated from Xanthosoma sagittifolium
(tannia) and Colocassia esculenta (taro). Carbohydr Polym
60:139–145
Takahashi W, Fujimori M, Miura Y, Komatsu T, Nishizawa Y, Hibi
T, Takamizo T (2005) Increased resistance to crown rust disease
in transgenic Italian ryegrass (Lolium multiflorum Lam.)
expressing the rice chitinase gene. Plant Cell Rep 23:811–818
Veluthambi K, Gupta AK, Sharma A (2003) The current status of
plant transformation technologies. Curr Sci 84:368–378
Wang JK (1983) Introduction. In: Wang JK (ed) Taro, a review of
Colocasia esculenta and its potentials, University of Hawaii
Press, Honolulu, pp 3–13
Yamamoto T, Iketani H, Ieki H, Nishizawa Y, Notsuka K, Hayashi T,
Matsuta N (2000) Transgenic grapevine plants expressing a rice
chitinase with enhanced resistance to fungal pathogens. Plant
Cell Rep 19:639–646
Zhu Q, Maber EA, Masoud S, Dixon RA, Lamb CJ (1994) Enhanced
protection against fungal attack by constitutive co-expression of
chitinase and glucanase genes in transgenic tobacco. Biotech-
nology N Y 12:807–812
123