Plant Cell Reports

Plant Cell Reports (1987) 6:204-207

© Springer-Verlag 1987

Transformation of Stylosanthes spp. using Agrobacterium tumefaciens J. M . M a n n e r s CSIRO Division of Tropical Crops and Pastures, CSIRO-UQ Plant Pathology Unit, Department of Botany, University of Queensland, St. Lucia 4067, Australia Received October 30, 1986 / Revised version received February 3, 1987 Communicated by J. M. Widholm

ABSTRACT Tumours

MATERIALS AND METHODS were

incited

on

leaf

sections

of and

Si#~o~anIAes Aam~igisp S, ha~ai~ S. guianan~ia $. ~cal~a following infection by Ag~o~aci~nium iamafacie~a. The suitability of 2 binary vectors (pGA472, BIN6) for gene transfer in $. h u m ~ s was tested and kanamycin-resistant tumour tissue was obtained from infected leaf pieces. The presence and expression of the neomycin phosphotransferase (NPT II) gene in the plant cells was demonstrated by hybridization of the coding region of the NPT II gene of the transposon Tn5 to DNA and RNA of kanamycin resistant tumours and by detection of significant NPT II activity in tissue extracts. Tumours also produced teratomatous shoots expressing the NPT II gene, but these could not be rooted. INTRODUCTION In recent years the natural gene transfer system of Agro~a~ ~u.rf~faciarLs(Hookyaas and Schilperoort, 1984) has been successfully adapted to provide methods for the transfer of cloned genes to some plant species (Fraley a£ a£. 1986). Several vectors containing dominant selectable marker genes have been constructed to permit the isolation of transformed plant cells expressing resistance to specific antibiotics (Herrera-Estrella at a£. 1983; Bevan at d ° 1983). Initially the transfer of these vectors to the plant genome was effected via integration into the T-DNA of the Ti plasmid. More recently binary vector systems have been developed and these vectors replicate independently of the Ti plasmid in A, ~faa/a~ (Bevan, 1984; An e£ a£. 1985). If gene transfer technology using A. lamafacian~ is to have an impact on agricultural productivity it is necessary that methods be developed for the transformation of crop and pasture species of economic importance. S b j £ o s a ~ spp. are pasture legumes grown widely in many tropical and subtropical regions (Stace and Edye, 1984). StglosanY_4ga spp. do not appear to have been reported as hosts for A g r o ~ ~ spp. (DeCleene and De Ley, 1976) and have not previously been studied in molecular gene transfer experiments. The aim of the present investigation was to assess the susceptibility of Stg£o~ spp. to A, £u2uzfac/ena and investigate the potential of binary vectors for the transfer of foreign genes to cells of this legume.

Plant Material Plants (Table i) were grown in a peat-sand (1:3) mixture in either a glasshouse or a controlled environment chamber (16h daylength, 30°C day temp, 25°C night temp) for 4-5 months, and freshly emerged leaves were removed for experiments. Bacterial Strains and plasmids Strains C58 (Hamilton and Fall, 1971) and A281 (An ~/ af. 1985) of A. ta~efaciaq~ carrying the plasmids pTiC58 and pTiBo542 respectively were chosen because bacterial strains carrying these plasmids have been reported to infect other legumes (Hamilton and Fall, 1971; Hood a/ a/. 1984). The binary vectors pGA472 (An ag a/o 1985) and BIN6 (Bevan, 1984) were the kind gifts of Dr G. An, Institute of Biological Chemistry, Washington State University, Pullman, U.S.A. and Dr M. Bevan, Plant Breeding Institute, Trumpington, U.K., respectively. Binary vectors were transferred from £ ~ c ] ~ a A / ~ c o l strain DHI to A. tumefaaieqs by triparental mating (An at d , 1985). Transformation and culture of transformed tissue Bacterial cultures were grown in 5 ml of Luria broth for 2 days at 28°C. Leaflets were detached, surface-sterilised and cut into approximately 5 mm sections and these were suspended in the bacterial culture for 2-5 minutes. The leaf pieces were then blotted dry and placed on MSO (the media of Mnrashige and Skoog, 1962 containing ig/l casein acid hydrolysate with 0.7% agar and lacking phytohormones) and coincubated with the bacteria at 28°C in the dark for 2 days. The leaf pieces were then washed in sterile water and placed on MSO containing 500 pg/ml of either cefotaxime or carbenicillin to kill the bacteria. The incidence of tumour formation was estimated after 7-8 weeks incubation at 28°C in the dark. In transformation experiments using A. tum~facien~ strain A281 carrying the binary vectors, the bacterial culture medium was removed by centrifugation and the bacteria washed with agar-free MSO and resuspended in this medium at their original cell density prior to inoculation of leaf sections of S. h u ~ cv. Paterson as described above. Inoculated leaf pieces were placed on MS0 for 2 days and then transferred to MSO containing 500 ~g/ml cefotaxime and 50 ~g/ml kanamycin. Transformed calli were subcultured every 2-3 weeks during maintenance. Cultures were transferred to a 16-h photoperiod after

205

3

Fig.l A) Formation of white tumours on leaf sections of $o Aum~ six weeks following inoculation with J. t~afacien~ strain A281. B) Agropine assay of extracts of tumours of $° Aum~a (lanes 3 & 4), $. Aom~o/a (lane 5) and $o gai~aia (lane 6) incited by strain A281 of A. lumafacian~. The positions of authentic ldeoxymannityl glutamine (lane I) and agropine (lane 2) and the origin (o) of electrophoresis are shown. C) Teratomatous-shoot formation on a kanamycinresistant tumour of $. Aum///a 4 months after transformation. D) Excised teratomatous shoots of S, hum///a at 3(1), 5(2) and 10(3) weeks after transfer to I/4 strength MSO.

Table 1 sections

S

6

Frequency of tumour formation on leaf of $iy£osanIAas spp. following inoculation

by A g ~ o ~ a c i ~ iam~acien~. Genotype

% of leaf sections with tumours W Strain A281

5. Aum/f/a 5, Ao~ata

S. 9 n i s e i s

$. ~ c o ~ a

cv. Paterson cv. Verano cv. Graham cv. Cook cv. Endeavour cv. Oxley cv. Seca cv. Fitzroy QI0042 ~

33 6 3 4 0 0 0 0 2

(270) (138) (127) (48) (46) (129) (47) (133) (82)

Strain C58 18 16 5 5

(91) (87) (94)

6 weeks culture in the dark. Opine analysis Agropine was analysed by high-voltage electrophoresis as described by Ellis a/ a/. (1984). Authentic agropine and l-deoxymannityl glutamine (manopine) were the kind gifts of Dr M. Tate, Waite Agricultural Institute, Adelaide.

~ pooled data from up to 6 experiments, the total number of leaf sections tested is in parenthesis. {~ Queensland Department of Primary Industries No.

Neomycin phosphotransferase II (NPT II) assays The activity of NPT II in transformed tissue was assayed by the methods of An a£ o/. (1985) and Schrier at a/. (1985).

~fa~

Hybridization analysis High molecular weight DNA was isolated from callus by the method of Weeks ai el. (1986) digested with Eco R I, electrophoresed on a 0.8% (w/v) agarose gel and transferred to a nylon filter (Hybond N. Amersham Int.). Total RNA was isolated, electrophoresed and transferred to nitrocellulose filters by described methods (Higgins a/ o~f. 1985). The hybidization probe for part of the coding region of the Tn5 neomycin phosphotransferase II gene was as described by Hain ai o/. (1985). Hybridization and autoradiography were carried out using standard techniques (Maniatis a/ d . 1982).

(95)

RESULTS Susceptibility

of

S~t£o~

genotypes

to

A.

In preliminary experiments, attempts to incite tumours on the stems of $. g a i ~ s i ~ cv. Graham and cv. Oxley and $. / u m ~ s cv. Paterson by stabbing the stems with a needle previously dipped in a culture of A./am~fac/en~ strain A281 were unsuccessful. In contrast to this, similar inoculations were made on the leaf midribs and small tumourous growths of approximately 2-3 mm diameter developed over a 2 month period at inoculation sites on the leaves of two of these species at the following frequencies: 1 from l0 sites on S. guian~nsi~5, 3 from I0 sites on $. ~ur~. Because of the higher susceptibility of leaf tissue the frequency of tumonr induction on surface sterilized leaf sections of several genotypes was determined (Table i, Fig i) using two strains of A. t~efacZe~. In these experiments $. / ~ and 5. Aamonia showed the highest

206 1.0

A

~: 0.8

g E

v

~O.6 O w i-

n-

0.4

u_ Z w

0.2 W nO

Z

I

I

i

I

I

20

40

60

80

100

KANAMYCIN

(~g/ml)

Fig. 2 The effect of kanamycin on the growth over 2 months of tumour tissue obtained with A. £amafac/aaa strain A281 (o) or strain A281 carrying the binary vector pGA472 ( e ) . The mean (±SD) of 4 independently isolated tumours is shown. Cultures were initiated with about 15mg fresh wt. of tissue on MSO. frequencies

of

tumour

formation. Leaves of S. were much less susceptible and some cultivars of these species did not produce any tumours when tested with A281. Non-inoculated leaf sections of all genotypes did not produce any callus on hormone free medium. Tumour tissue of S. h ~ s , ha~aga and g u i ~ i s cv. Graham produced with strain A281 was removed for analysis of the opine agropine and its precursor ldeoxymannityl glutamine. Compounds co-electrophoresing with these standards were identified in extracts of all freshly isolated tumours obtained with A281 (Fig i). These compounds were absent from normal leaf tissue of each plant species and also absent from leaf callus of S, hum/£i~ produced by the method of Meijer (1982) (data not shown). The tumour tissue of S. h u m ~ s was again analysed after 4 months culture on MSO, and at this stage no opines were detected.

guio~enmi~ and S,

Transfer

of the NPT

~c~a

II gene to

tumour cells

tumour cells obtaiDed with strain A281 only and 328 cpm (SD_+I07) for the extracts of transformed calli assayed without kanamycin in the assay mix. NPT II activity was also detected on polyacrylamide gels following electrophoresis of cell extracts by the method of Snhrier ~t o~f. (1985). Extracts of transformed calli obtained using.the binary vectors pGA472 and BIN6 each produced a single band of NPT II activity whilst extracts of control tumours showed no activity (Fig. 3A). As expected the NPT II enzyme of calli transformed with BIN6 was of identi