INTRODUCTION

Improvement of citrus by conventional method is

hampered by polyembryony, sexual incompatibility and

male or female sterility (Guo and Deng, 2001; Grosser and

Gmitter, 2005). Genetic transformation is an alternative to

overcome these difficulties. For successful transformation,

regeneration of whole plants from the transformed cells

is a prerequisite. Cut modes and hormones may affect

in vitro citrus shoot regeneration. Transversal cut, the

most popular cut mode (Moore et al., 1992; Pena et al.,

2004), is simple to manipulate but produces the fewest

adventitious buds. Longitudinal cut, a newly developed but

infrequently used cut mode (Yu et al., 2002; Kayim et al.,

2004), producing the most adventitious buds, is laborious

and difficult to manipulate especially when the epicotyl

Botanical Studies (2007) 48: 165-171.

explants are thin or weak. For hormones, the effect of

auxin on shoot regeneration was rarely concerned, though

the main hormone effect on bud formation was due to the

addition of BA (Garcia-Luis et al., 1999). Till now, little

was known on the effect of IAA and its interaction with

BA in citrus regeneration.

The green fluorescent protein gene (gfp) from the

jellyfish Aequorea victoria as a vital marker has attracted

increasing interest and is considered to have several

advantages over other visual marker genes (Ghorbel

et al., 1999). Expression of GFP5 results in greatly

improved levels of fluorescence and using the modified

gene (mgfp5-ER) has potential to regenerate intensely

fluorescent and fertile plantlets (Siemering

et al., 1996;

Haseloff et al., 1997). The goal of this study was: 1) to

find a moderate and efficient cut mode and optimize a

hormone combination for multiple shoot induction and

regeneration of citrus seedling epicotyls, and 2) to rapidly

achieve transgenic plants of an elite citrus cultivar with

the GFP gene, which will be promising for further genetic

transformation with agronomic traits.

phySIOlOgy

Multiple shoot induction from seedling epicotyls and

transgenic citrus plant regeneration containing the

green fluorescent protein gene

Yan-Xin DUAN

1

, Xin LIU

1, 2

, Jing FAN

1

, Ding-Li LI

1

, Reng-ChaoWU

1

, and Wen-Wu GUO

1,

*

1

National Key Laboratory of Crop Genetic Improvement, National Center of Crop Molecular Breeding, Huazhong

Agricultural University, Wuhan 430070, China

(Received May 8, 2006; Accepted October 3, 2006)

ABSTRACT.

This research aimed to optimize the organogenesis of epicotyl segments and to efficiently

obtain transgenic plants of ¡¥Bingtang¡¦ sweet orange (Citrus sinensis L. Osb.), an elite citrus cultivar in China.

Organogenesis induction was induced in epicotyl segments of 2 weeks old seedlings of this cultivar. Two

important factors influencing organogenesis in vitro viz.

hormone combination (IAA and BA) and cut modes

were characterized. IAA had a positive effect on bud formation only when BA was used at the concentration

of 2.0 mg/l, and an inhibitive effect was observed with higher or lower concentration of BA. The number

of regenerated buds reached up to 8.9 per explant with the combination of IAA 0.2 mg/l and BA 2.0 mg/l.

Among cut modes, oblique cut performed the best for its effect on the number and quality of regenerated

shoots and its convenience to manipulate. With the optimized hormone combination and oblique cut mode,

citrus transformation with green fluorescent protein (GFP) gene was performed and twelve independently

transformed plant lines were achieved. Southern blot hybridization confirmed the stable integration of GFP

gene into the citrus genome. The successful transformation of this cultivar revealed that it is possible to

introduce other genes with agronomic traits into it. Furthermore, these GFP expressing transgenic plants could

serve as a visual marker material for citrus somatic fusion and sexual hybridization.

Keywords: Auxin; Citrus; Genetic transformation; Green fluorescent protein; Oblique cut; Shoot formation.

Abbreviations: AS, acetosyringone; BA, 6-benzyladenine; Cef, cefotaxime; GFP, green fluorescent protein

gene; Km, kanamycin; IAA, indole-3-acetic acid; LB, (Luria-Bertani) medium; MBI medium, solid MT

medium containing 2.0 mg/l BA, 0.2 mg/l IAA and 30 g/l sucrose; MT medium, Murashige and Tucker.

2

The first two authors contributed equally to this paper.

*

Corresponding author: E-mail: guoww@mail.hzau.edu.cn;

Tel: 86-27-87281543; Fax: 86-27-87280016.

166

Botanical Studies, Vol. 48, 2007

MATERIAlS AND METhODS

Etiolated seedlings of ¡¥Bingtang¡¦ sweet orange (Citrus

sinensis L. Osb.) were used as explant sources. Seeds were

decontaminated as described by Moreira-Dias et al. (2001)

and germinated at 26¢XC in the dark in test tubes containing

MT medium (Murashige and Tucker, 1969) with 25 g/l

sucrose, 7.5 g/l agar, and pH 5.7. Epicotyl cuttings (1 cm

long) from 2 weeks old seedlings were plated horizontally

on the culture medium. Twenty-five explants were

evenly distributed in 90 ¡Ñ 20 mm Petri plates containing

approximately 20 ml medium. Hundred cuttings from

20 seedlings were measured per treatment with three

repetitions. At the end of the incubation period (45

d), the

number of visible buds and shoots were counted.

Effect of hormones on bud formation

The epicotyl explants were cut obliquely and incubated

on solid MT medium (containing 30 g/l sucrose),

supplemented with BA (1.0, 2.0, 3.0 mg/l) alone or

combined with 0.2 mg/l IAA, in the dark/light condition

(two weeks of darkness followed by 31 d in the light).

Effect of cut modes on the formation of buds

and shoots

In addition to longitudinal and transversal cut modes,

oblique cut was used in the present study. To manipulate

the oblique cut mode, the epicotyls were cut obliquely

with 3-4 mm end. Segment cuttings were incubated on

MBI medium (solid MT medium containing 2.0 mg/l BA,

0.2 mg/l IAA and 30 g/l sucrose, PH 5.7) and cultured in

the dark/light condition.

Transformation and regeneration

The disarmed A. tumefaciens strain EHA-105 harboring

a binary vector plasmid pBin-mgfp5-ER (Haseloff et al.,

1997) was used. Two experiments each with 60 epicotyl

segments were performed. The transformation procedure

was: a fresh single colony of disarmed A. tumefaciens

strain EHA105 was selected and multiplied on solid LB

(Luria-Bertani) medium with 50 mg/l kanamycin (Km)

for 48 h at 28¢XC. The bacteria were collected, transferred

to liquid MT medium in an orbital shaker at 28¢XC and

180 rpm for 2 h, and then adjusted to an absorbance A

600

=

0.5-0.8. Epicotyl cuttings were infected with the adjusted

A. tumefaciens for 30 min, and then dried by sterile filter

paper and co-cultured on solid MT medium containing

2 mg/l AS (acetosyringone) for 3 days in the dark at 23

¢XC. After co-cultivation, the explants were screened on

selective medium [MBI medium supplemented with

cefotaxime (Cef) 400 mg/l

and 50 mg/l Km] under dark/

light condition. Explants were screened every four weeks

on the same medium till resistant shoot regenerated. GFP

expressing shoots were excised from the cut end and

enlarged on the Km free (but containing Cef 200 mg/l)

medium. Well-developed shoots were either induced to

root as described by Guo et al. (2002)

or grafted onto

rootstocks to obtain whole transgenic plants as described

by Deng et al. (1993).

Detection of gFp by fluorescence stereomi-

croscopy and molecular analysis

After co-cultivation, the explants were screened for

GFP transient fluorescence and periodical examination

was done under a fluorescent stereomicroscope equipped

with a Leica fluorescence stereomicroscope

(MZFLIII)

comprising a 480/40 nm exciter filter, a 505 nm LP

dichromatic beam splitter and a 510 nm LP barrier filter.

Buds and shoots expressing green fluorescence were

considered as putative transgenics.

Southern blot analysis was performed to confirm the

stable integration of the GFP in the transgenic plants.

DNA was extracted from leaves according to Cheng

et al. (2003). Genomic DNA of three GFP expressing

plants and one non-transformed plant was digested with

EcoRI, separated on 0.8% (w/v) agarose gels and blotted

onto nylon membranes (Hybond-N

+

, Amersham). Films

were probed with a P

32

-labelled fragment of the GFP

prepared by PCR. The PCR specific primers of GFP were

PL: 5¡¦-TGGCCAACACTTGTCACTAC-3¡¦ and PR: 5¡¦-

AGGACCATGTGGTCTCTCTT-3¡¦, which resulted in

a 500 bp fragment amplified from the plasmid template.

PCR reactions were the same as described by Shi et al.

(2002).

RESUlTS

Effect of growth regulators and cut modes on

indirect bud and shoot organogenesis

Under dark/light condition, the buds differentiated from

the callus formed at the cut end. Bud formation increased

when BA concentration was enhanced (Figure 1).

Meanwhile, the number of quiescent shoots regenerated

increased. When combined with 0.2 mg/l IAA, the additive

effect appeared at 2.0 mg/l of BA. The mean number

of buds reached a maximum of 8.9 per explant, among

which about five buds could elongate to shoots. Therefore,

MBI medium was chosen as the optimal medium for use

during transformation of ¡¥Bingtang¡¦ sweet orange epicotyl

explants.

Figure 1. Influence of BA and IAA (mg/l) on adventitious bud

formation in epicotyl segments of ¡¥Bingtang¡¦ (Citrus sinensis L.

Osb.) sweet orange.

12

10

8

6

4

2

0

IAA 0 mg/1

IAA 0.2 mg/1

1

2

3

The concentration of BA (mg/1)

DUAN et al. ¡X Citrus multiple shoot induction and transformation

167

Concerning easiness of manipulation for different cut

modes, oblique cut proved as simple as transversal cut

but more efficient and convenient, longitudinal cut was

difficult and laborious. Both oblique and longitudinal

cuts could increase the wound area of epicotyl explants,

resulting in more callus formation and shoot regeneration

than transversal cut (Figure 2, A1-C3). Longitudinal cut

gave the highest number of buds per explant followed by

oblique cut and transversal cut (Table 1). However, most

buds regenerated by longitudinal cut were weaker and

smaller than those regenerated by oblique cut (Figure 2,

C3). In addition, regarding shoots with at least two leaves

regenerated, there was no significant difference between

those produced by oblique and longitudinal cuts, while

transversal cut still resulted in the least (Table 1). Taken

together, oblique cut performed the best in vitro.

gFp as a visual marker to localize the sites

of transgene expression at early stages after

transformation

GFP expression could be detected transiently during

co-cultivation. The cut end of the non-transformed and

transformed epicotyl showed red autofluorescence (Figure

3A) and green fluorescence (Figure 3B), respectively. Four

weeks later, that small green fluorescent buds sprouted

and more than 2 transgenic buds expressing the GFP

gene formed at one cut end were often observed (Figure

3C). Simultaneously, non-transformed shoots could also

regenerate at the same cut end on selective medium (Figure

3E). Well-grown non-transformed shoots exhibited strong

red fluorescence after one more month of growth on MBI

medium (Figure 3D). GFP expressing shoots (longer than

0.5 cm) were physically separated in vitro and elongated

on Km

free medium (MBI medium containing 200 mg/l

Cef) (Figure 3F). Transgenic plants recovered by rooting

or by grafting onto trifoliate orange (Poncirus trifoliata)

rootstock were transferred to soil in the greenhouse

and showed normal growth like the controls after six

months (Figure 4). Totally, twelve independent transgenic

plant lines were obtained and transplanted; the ultimate

transformation efficiency was 10%.

Different integration patterns of gFp

Genomic DNA from randomly selected GFP expressing

Figure 2. Comparison of three different cut modes on callus and bud formation with transversal cut (A1-A3), oblique cut (B1-B3) and

longitudinal cut (C1-C3). A1, B1, C1, Callus formation after 2 weeks culture under darkness; A2, B2, C2, Bud formation one week

after being transferred to light; A3, B3, C3, Shoot elongation under light conditions. Scale bars, 1.0 mm.

Table 1.

Effect of different cut modes on buds and shoots regeneration.

Different cut modes

Longitudinal cut Oblique cut Transversal cut

The mean number of buds per explant

19.5 a

8.9 b

2.5 c

The mean number of developed shoots with two leaves per explant

5.5 a

5.2 a

1.8 b

Data are the means of 100 epicotyl segments from three independent experiments. Numbers with same letters were not significantly

different at £\=0.05 according to Duncan¡¦s multiple range tests.

168

Botanical Studies, Vol. 48, 2007

plants and a non-transformed plant was analyzed by

Southern blot. The GFP gene fragment (500 bp) was

used as a probe to confirm the presence of the transgene.

Different integration patterns in the transgenic plants,

with one to 4 copies at different loci were identified when

EcoRI (only one enzyme site in the T-DNA) was used.

As was shown in Figure 5, all the three GFP expressing

plants contained the target gene (Lanes G2, G3, and G5).

No hybridization signal was detected in non-transformed

control plant (Lane G28). No correlation between copy

number and levels of GFP expression was revealed in this

study.

DISCUSSION

Establishment of highly efficient regeneration protocol

for existing cultivars will help to improve and broaden the

usage of genetic transformation technique in citrus. In the

present study, bud formation increased with the enhanced

BA concentration, and similar results were previously

demonstrated by several authors (Goh et al., 1995; Costa

et al., 2004). However, IAA enhanced bud formation at 2.0

mg/l concentration of BA, and inhibited bud formation at

1.0 or 3.0 mg/l. This was different from previous report

in which the contribution of auxins was marginal on bud

formation (Garcia-Luis et al., 1999) and different from the

result of Moreira-Dias et al. (2000). Using different citrus

genotypes or different concentrations of hormones might

be responsible for the different results. The results here

show that an additive effect might have produced between

BA and IAA when they were combined at the proper

proportions. Such results have not been reported in citrus

previously and need further validation. Though giving

approximately same number of buds, elongated shoots

produced by 3.0 mg/l

BA were less than those produced

by 2 mg/l of BAP + 0.2 mg/l of IAA (data not shown).

This may be correlated with the concentration of BA: the

lower concentration of BA, the larger the size of the shoots

produced (Gutierrez-E et al., 1997).

Bud formation was significantly affected by different

cut modes (i.e. transversal cut, oblique cut and longitudinal

cut); and the number of regenerated buds increased

with the enlarged cut area (Yu et al., 2002). In addition,

transformation mostly occurred in dividing cells such as

cambium callus formed at the cut ends of the explants

(Cervera et al., 1998; Pena et al., 2004). Longitudinal

and oblique cuts are more efficient than transversal

cut in producing wound area and cambial callus, and

should perform better in transgenic shoot regeneration.

Several papers have clarified that longitudinal cut gave

Figure 3. GFP expression in trans formed and non-transformed epicotyl cut end, buds and shoots. A, Red fluorescence in non-

transformed epicotyl cut end; B, Green fluorescence in transformed epicotyl cut end after co-cultivation; C, Three transformed buds

sprouted along the cut end 4 weeks after cultivation on selective medium (arrow indicates the transformed buds); D, Non-transformed

shoots with red fluorescence; E, Two transgenic shoots and one escape regenerated from the cut end 8 weeks after cultivation (arrow

indicates the escape); F, Well-developed transgenic shoot with bright green fluorescence. Scale bars, 1.0 mm.

Figure 4. Transgenic citrus plant growing in the greenhouse.

DUAN et al. ¡X Citrus multiple shoot induction and transformation

169

higher transformation efficiency in Carrizo citrange

than transversal cut (Yu et al., 2002; Kayim et al.,

2004). However in this study, except for difficulty in

manipulation, most buds regenerated by longitudinal cut

remained quiescent and poor, which might be correlated

with the nutrition competition between buds and shoots;

the more buds formed, the fewer shoots elongated. Taken

together, oblique cut performed the best in this study.

Citrus transformation based on the optimized

regeneration protocol was successful, and totally twelve

putative transgenic lines were obtained which grew

normally in the greenhouse. The transformation efficiency

of Bingtang sweet orange (C. sinensis) obtained in this

study (10%) was much higher than that of ¡¥Xuegan¡¦ (4.3%)

(Yu et al., 2002) and Ridge pineapple (2%) (Gutierrez-E

et al., 1997), lower than that of Hamlin (15%) (Mendes et

al., 2002) and Valencia (23.8%) (Boscariol et al., 2003).

These results revealed that the transformation efficiency

was genotype-dependent, and the regeneration protocol

in this study was suitable for transformation and it may

be possible to introduce other agronomic traits into this

elite cultivar. Using nptII as a selectable marker, more

than 60% regenerated shoots were escapes (Costa et al.,

2002; Almeida et al., 2003). Contrastingly, GFP as a non-

destructive and vital marker gene could discriminate

escapes and chimerical shoots at an early stage, thus saved

time, significantly improved efficiency and expedited the

molecular breeding process.

Transgenic citrus plants obtained in this study could

serve as mesophyll parent (citrus mesophyll protoplasts

are not regenerable) for protoplast isolation and then fused

with protoplasts from embryogenic calluses of another

parent using GFP as a visual marker. Somatic hybrid

cells at an early developmental stage during somatic

hybridization could be screened and monitored (Guo and

Grosser, 2005). Furthermore, since most citrus species

are polyembryonic, it is hard to obtain sexual hybrids by

crossing; using transgenic citrus plants expressing the

GFP gene as the pollen parent, sexual embryos will be

easily detected and separated for further in vitro culture

to recover sexual hybrid progenies facilitated by in vivo

fluorescence of GFP expression. The transgenic plants

expressing GFP gene produced herein are valuable

materials for these studies in the future.

Acknowledgments. This research was financially

supported by the National Natural Science Foundation of

China (No. 30571288), the Ministry of Education of China

(705037, IRT0548, NCET-04-0734), the Huoyingdong

Education Foundation (91030), and the International

Foundation for Science in Stockholm, Sweden (IFS,

D/2895-3).

lITERATURE CITED

Alm ei da, W.A.B ., F.A.A. Mourao Fi lho, L .E. P ino, R.L .

Boscariol, A.P.M. Rodriguez, and B.M.J. Mendes. 2003.

Genetic transformation and plant recovery from mature

tissues of Citrus sinensis L. Osbeck. Plant Sci. 164:

203-211.

Boscariol, R.L., W.A.B. Almeida, M.T.V.C. Derbyshire, F.A.A.

Mourao Filho, and B.M.J. Mendes. 2003. The use of the

PMI/mannose selection system to recover transgenic sweet

orange plants (Citrus sinensis L. Osbeck). Plant Cell Rep.

22: 122-128.

Cervera, M., J.A. Pina, J. Juarez, L. Navarro, and L. Pena. 1998.

Agrobacterium-mediated transformation of citrange: factors

affecting transformation and regeneration.

Plant Cell Rep.

18: 271-278.

Cheng, Y.J., W.W. Guo, H.L. Yi, X.M. Pang, and X.X. Deng.

2003. An efficient protocol for genomic DNA extraction

from Citrus species. Plant Mol. Biol. Rep. 21: 177a-177g.

Costa, M.G.C, V.S. Alves, E.R.G. L ani, P.R. Mosquim, C.R.

Carvalho, and W.C. Otoni. 2004. Morphogenic gradients of

adventitious bud and shoot regeneration in epicotyl explants

of citrus. Sci. Hort. 100: 63-74.

Costa, M.G.C, W.C. Otoni, and G.A. Moore. 2002. An evaluation

of fac tors a ffe cting the efficie ncy of Agrobacterium-

m ediated transformation of Citrus paradisi (Macf.) and

production of transgenic pla nts containing carotenoid

biosynthetic genes. Plant Cell Rep. 21: 365-373.

Deng, X.X., S .Y. Xiao, and W.C. Zhang. 1993. Inters pecific

s omatic hybrid of Ichang P apeda with Valencia orange.

Chinese J . Biotech. 9: 128-131 (in Chinese with English

summary).

Figure 5. Southern-blot analysis of the GFP from three green

fluorescent protein-expressing plants and one non-transformed

plant. Genomic DNA digested with EcoRI, which has only one

enzyme site in the plasmid. Lane G28, non-transformed plant;

Lanes G2, G3, G5, GFP expressing plants; M, £f DNA/HindIII

Molecular weight marker, Molecular weights are indicated in

kilobases on the right.

170

Botanical Studies, Vol. 48, 2007

Garcia-Luis, A., Y. Bordon, J.M. Moreira-Dias, R.V. Molina,

and J.L. Guardiola. 1999. Explant orientation and polarity

determine the morphogenic response of epicotyl segments

of Troyer citrange. Ann. Bot. 84: 715-723.

Ghorbel, R., A. Dominguez, L. Navarro, and L. Pena. 1999.

Green fluorescent protein as a screenable marker to increase

the efficiency of generating transgenic woody fruit plants.

Theor. Appl. Genet. 99: 350-358.

Goh, C.J., G.E. Sim, C.L. Morales, and C.S. Loh. 1995. Plantlet

regeneration through different morphogenic pathways in

pommelo tissue culture. Plant Cell Tis s. Org. Cult. 43:

301-303.

Grosser, J.W. and F.G Jr. Gmitter. 2005. Applications of somatic

hybridization and cybridization in crop improvement,

with citrus as a model. In Vitro Cell Dev. Biol.-Plant 41:

220-225.

Guo, W.W. and X.X. Deng. 2001. Wide somatic hybrids of

Citrus with its related genera and their potential in genetic

improvement. Euphytica 118: 175-183.

Guo, W.W., Y.J. Cheng, and X.X. Deng. 2002. Regeneration and

molecular characterization of intergeneric somatic hybrids

between Citrus reticulata and Poncirus trifoliata. Plant Cell

Rep. 20: 829-834.

Guo, W.W. a nd J .W. Gross er. 2005. Som atic hybrid vigor

in Citrus: Direct evidence from protoplas t fusion of an

embryogenic callus line with a transgenic mesophyll parent

expressing the GFP gene. Plant Sci. 168: 1541-1545.

Gutierrez-E, M.A., D. Luth, and G.A. Moore. 1997. F actors

affecting Agrobacterium-mediated transformation in Citrus

and production of sour orange (Citrus aurantium L.) plants

expressing the coat protein gene of citrus tristeza virus .

Plant Cell Rep. 16: 745-753.

Haseloff, J., K.R. Siemering, D.C. Prasher, and H. Sarah. 1997.

Removal of a cryptic intron and subcellular localization of

green fluorescent protein are required to mark transgenic

A rabidopsis plants brightly. Proc. Natl. Acad. Sci. 94:

2122-2127.

Kayi m, M., T.L. Cec cardi, M.J .G. Be rretta , G.A. Barthe ,

and K.S . Derrick. 2004. Introduction of a citrus blight-

associated gene into Carrizo citrange [Citrus sinensis (L.)

Osb. ¡Ñ Poncirus trifoliata (L.) Raf.] by Agrobacterium-

mediated transformation. Plant Cell Rep. 23: 377-385.

Mendes, B.M.J., R.L.Boscariol, F.A.A.M. Filho, and W.A.B.

Al m e id a . 20 02 . Ag ro ba c t er i um - m ed i a te d g en e t ic

trans formation of ¡¥Hamlin¡¦ sweet orange. P esq Agropec

Brasileira 37: 955-961.

Moreira-Dias, J.M., R.V. Molina, Y. Bordon, J.L. Guardiola, and

A. Garcia-Luis. 2000. Direct and indirect shoot organogenic

pathways in epicotyl cuttings of Troyer citrange differ in

hormone requirements and in their response to light. Ann.

Bot. 85: 103-110.

Moreira-Dias, J.M., R.V. Molina, J.L. Guardiola, and A. Garcia-

Luis. 2001. Daylength and photon flux density influence

the growth regulator effects on morphogenesis in epicotyl

segments of Troyer citrange. Sci. Hort. 87: 275-290.

Moore, G.A., C.C. Jacono, J.L. Neidigh, S.D. Lawrence, and

K. Kline. 1992. Agrobacterium-mediated transformation of

citrus stem segments and regeneration of transgenic plants.

Plant Cell Rep. 11: 238-242.

Murashige, T. and D.P.H. Tucker. 1969. Growth factor

requirements of citrus tis sue culture. Proc. First Citrus

Symp. 3: 1155-1161.

Pena, L., R.M. Perez, M. Cervera, J.A. Jurez, and L. Navarro.

2004. E arly events in Agrobacterium-mediated genetic

transformation of citrus explants. Ann. Bot. 94: 67-74.

Shi, W., D.D. Li, X.X. Deng, and H.L.Yi, 2002. Transformation

of gfp in Cleopatra mandarin (Citrus reticulata) mediated

by Agrobacterium tumefaciens. Acta Horticulturae Sinica.

29: 109-112 (in Chinese with English summary).

Siemering, K.R., R. Golbik, R. Sever, and J. Haseloff. 1996.

Mutations that suppres s the thermos ensitivity of gree n

fluorescent protein. Curr. Biol. 6: 1653-1663.

Yu, C., S . Hua ng, C. Che n, Z . Deng, P. L ing, a nd F.G. Jr.

Gmitter. 2002. Factors affecting Agrobacterium-mediated

transform ation of sweet orange and citrange. P lant Cell

Tiss. Org. Cult. 71: 147-155.

DUAN et al. ¡X Citrus multiple shoot induction and transformation

171