PlantCell Reports

Plant Cell Reports (1996) 16:114-117

© Springer-Verlag 1996

Regeneration and Agrobacterium-mediated transformation of Forsythia x intermedia "Spring Glory" Carlo Rosati, Alain Cadic, Jean-Pierre Renou, and Michel Duron INRA C.R. Angers-Station d'Am61ioration des Esp~ces Fruiti+res et Ornementales 42, rue Georges Morel - 49071 Beaucouz6 Cedex France Received 25 September 1995/Revised version received 23 April 1996 - Communicated by G. Pelletier Key words: Forsythia, genetic transformation, regeneration, w o o d y ornamentals.

Abstract Internode explants of in vitro plants of Forsythia x intermedia "Spring Glory" were transformed with the gus and npt 11 genes after inoculation with the A. tumefaciens strain EHA 101 harbouring the plasmid pFAJ3000. Shoot organogenesis took place from callused edges of explants. The first transformed buds were detected 4 to 6 weeks after transfer on regeneration medium, containing 25 mg/l kanamycin as selective agent. An average of 1% of explants regenerated transgenic shoots. 13-glucuronidase assays and culture on kanamycincontaining medium provided the first indication of integration and expression of introduced genes in transformants. Southern blot and polymerase chain reaction amplification analyses gave molecular confirmation of genetic transformation. Transgenic plants were acclimatized in the greenhouse. Enzymatic assays on several organs of mature plants still showed 13-glucuronidase activity, thus confirming stable integration of T-DNA in the plant genome. Abbreviations. BAP: 6-benzyl-aminopurine; CaMV: Cauliflower Mosaic Virus; GUS and gus: ~-glucuronidase; IAA: indole-3-acetie acid; IBA: indole-3-butyrie acid; MS: Murashige and Skoog; NOS: nopaline synthase; NPT II and npt 11: neomycin phosphotransferas¢ II; PCR: polymerase chain reaction; SDS: sodium dodeeyl sulphate; SSC: sodium chloride-sodium citrate; X-Glue: 5-bromo-4-chloro-3-indolyl glueuronide.

Introduction Forsythia (Oleaceae) is a commonly used ornamental shrub highly appreciated for its early bright yellow flowering. All species except F. europea originated in north-eastern Asia and have been introduced to Europe since the 18th century. However, a large number of cultivated varieties derive from Forsythia x intermedia Zabel, a natural, presumed hybrid between F. suspensa and F. viridissima extensively used in hybridization programs. In addition, polyploid genotypes have been induced and mutagenic treatments gave rise to mutants. Forsythia is a vegetatively propagated crop, highly heterozygous because of its heteromorphic Correspondence to: M. Duron

autoincompatibility system of mating. Crosses are only possible between thrum and pin species or cultivars (Sampson 1971). With such a biological situation, biotechnologies offer alternative tools to cross breeding or conventional mutagenesis to introduce agronomically important characteristics into an already improved variety. However, reliable plant regeneration systems are a prerequisite for gene transfer approaches. In the literature, only Roest and Bokelman (1980) have reported shoot regeneration from leaves of F. x intermedia "Spectabilis". As for transformation, reports dealing with woody ornamentals have been few, e.g. juneberry (Hajela et al. 1993), Rhododendron (Pavingerovfi et al. 1995), and rose (Firoozabady et al. 1994). As a complement to our conventional breeding program at the INRA Station of Angers, we undertook transformation experiments finally aimed to produce changes in ornamental traits (e.g. flower colour, growth habit) in Forsythia varieties. Preliminary studies had shown a genotype and explant effect on in vitro culture response and the inadequacy of previously reported protocols for our material. In this paper, we thus report a suitable regeneration and transformation method for internode explants of Forsythia x intermedia "Spring Glory", a widespread variety of great ornamental value. Materials and Methods

Plant material and culture media. In vitro plants of Forsythia x intermedia "SpringGlory"were used as sourceof explant in this work.

Plant cultures were initiated by mefistem culture (Duron 1977) and propagated as rootedcuttings on a mediumcontainingMS (Murashige and Skoog 1962) macro- and microelementsat half and full strength respectively,vitaminsof Morel and Martin (1955), 15 g/l sucrose,7 g/l agar (Merck)and 0.12 mgo IBA. The basal regenerationmedium(M6) was an MS supplementedwith 30 g0 sucrose, 7 g/i agar (Merck), 100 mg/I myoinositol,0.5 mg/i thiamin, 0.75 mg/l nicotinic acid and 0.75 mg/1 pyridoxirL The regeneration medium subsequently adopted was completed by the additionof 1 mg/IBAP and 0.5 mg/IIAA (M6 105). The pH of all culturemedia was adjustedto 5.8 beforeautoclavingat 120°Cfor 20 min.Culturesweremaintainedat 22-24°C,70 ttmolm2 s" photosyntheticphotonfluxdensitywith a day lengthof 16 h. Bacterial strain and cultur~ Transformation experiments were performed with a disarmed derivative of the supervirulent Agrobacterium tumefaciensstrainEHA 101 (Hoodet al. 1986)carrying the 15-kbpbinaryvectorpFAJ3000(de Bondtet al. 1994).Thisbinary

115 vector contains two genes within its T-DNA: 35S CaMV-GUS intron (Vaneanneyt et al. 1990) and NOS-npt 11. Fig. 1 shows the T-DNA with the two HindlIl sites (distance 2.9 kbp), the gpnl site beside the left border and no EcoRl sites.

Hind III

Hind III

pFAI3000 Fig. 1.

Kpn I

T-DNA

1 kbp

Schematic representation of the T-DNA of pFAJ3000 plasmld (adapted from de Bondt et al. 1994; p: promoter;, t: terminator).

The bacterial culture was initiated by inoculating 20 ml of LB (Sigma L-3522) liquid medium supplemented with the appropriate antibiotics (50 mg/l kanamycin, 50 mg/l rifampicin, 100 mg/l streptomycin, 300 mg/l spectinomycin) with 100 p.I of stock suspension kept at -80°C. The density of the overnight culture was measured at OD . Bacteria were spun down (2500g for 5 rain) and resuspended in steri~wate¢ to obtain an inocuhim at 107 cells/mi. Regeneration and transformation. Intemodes of 2-month-old plants were excised and cut into 5-ram segments. For regeneration experiments, explants were placed directly onto regeneration media based on M6 supplemented with combinations of BAP and IAA at 0.5, 1 and 2.5 mg/I and 0, 0.25, 0.5 and 1 mg/l respectively. Leaf explants were subjected to the same treatments, of which 1 mg/I BAP and 1 mg/i IAA corresponded to the hormonal balance of Roest and Bokelman (1980), and were used as controls. Before transtbrmation assays, preliminary tests had indicated that kanamyein at 25 mg/i inhibited regeneration on control explants; preliminary GUS assays were also performed to determine the optimal length of coculture. The following transformation protocol was adopted. Explants were cut with a scalpel blade previously dipped into the inoculum. After cutting,excess suspension was drained on sterile filter paper and explants were placed onto M6 105 for a 4-day eoculture period. After cocultivation, explants were transferred for 3 days onto M6 105 supplemented with 300 mg/l cefotaxim for disinfection and then subcultured for 1 week on the same medium supplemented with kanamycin (25 mg/l) as selective agent. Cefotaxim concentration was then lowered to 200 mg/i for 4-week subcultures on kanamycincontaining (25 mg/l) M6 105. All culture and regeneration steps are shown in fig. 2. day

0

4 |

,,

7

14

I

I

42

,

70

I

I

cefotaxim

0 300

300

200

200

kanamycin

0

25

25

25

Fig. 2.

0

Time course of regeneration and selection of transgenic plants on MS 105-based media supplemented with antibiotics (mg/i)o

Regenerant offshoots were excised from explants when they had reached 10 mm in length (i.e. 2 months after infection), and grown and rooted on micropropagation medium supplemented with 25 mg/l kanamycin and 200 mg/l cefotaxim. Subcultures were carried out every 4 weeks. Ten- to 15-era-long rooted putative transformants were transplanted into perlite:peatmoss (1:1) and grown to maturity in greenhouse conditions. GUS assay. The histochemieal GUS assay was carded out according to the protocol of Jefferson (1987) with some modifications. A wide range of tissues (i.e. sliced edges of inoculated explants, young regenerating leaves, as well as portions of roots, leaves, petals, sepals and pollen grains) of putative transgenie plants were incubated overnight at 37°C in a solution containing 3 mM X-Gluc, 500 mM Na EDTA and 9.5% sodium phosphate buffer (pH 7.2). When neeessary, Zplant material was soaked in 95% (v/v) ethanol to flee tissues from chlorophyll prior to observation under the microscope. GUS activity was recorded as the number of blue spots/explant for hand-sliced edges of explants 2-4 days after inoculation, and by the presence of blue areas for the other explants.

DNA extractiatt Total DNA was isolated from leaves of either in vitro or greenhouse control and transformed plants according to the method of Dellaporta et al. (1983), with some modifications. Briefly, tissues were fmely ground in liquid nitrogen, and cell wails disrupted in 1.2% SDS for 15-20 rain. Following K acetate treaUnent and isopropanol precipitation, the DNA suspension was RNase-treated (30 rain at 37°(2). Phenolchloroform and chloroform purifications, and ethanol precipitation and washes completed the extraction prooedure. _PCR. PCR analyses were performed using a MiniCyclerTM (MJ Research) thermocycler. Gus and npt 11 primers, bordering 590-bp and 390-bp fragments respectively, were used. Samples (5-10 ng plasmid and 50-100 ng plant DNA in a final vol of 100 ~1) were subjected to the following programs: gas 93°C (3 rain), 35 cycles of 93°C (1 rain), 60°C (30 s), 72°C (2 rain); npt1194°C (3 rain), 30 cycles ot"94°(2 (1 rain), 60°(2 (1 min), 72°(2 (1 rain 30 s); both programs ended with a final elongation phase at 72°C (15 rain). All P e R products were purchased from Eurogentee (scraing, Belgium). Amplified products were run on a 1% agarose gel and visualized by staining with ethidium bromide. Southern blot analysis. Leaf DNA (10 p.g) of control and transformed plants was digested overnight with 6-10 units/J.tg of either HindlII or EcoRI, run on a 0.8% agaros¢ gel along with HindllI-digested pFAJ3000 DNA (100 pg) and blotted to nylon membrane (Hybondr~-N, Amersham) following standard procedures (Sambrook et ai. 1989). For probe preparation, 1 lag ofpFAJ3000 DNA was digested by HindllI and run on a 1% agaro~ gel. The 2.9-kbp band containing the gus gene was purified (Geneclean~ kit, Bit 101); the 12.l-kbp band (i.e. the rest of the plasmid) was purified and digested by Kpn I; the resulting ~3.2-kbp fragment, purified as before, contained the npt H gene. Fillers were prehybridized for 3-4 h at 62°C in 5x SSC, 0.5% SDS, 5x Denhardt's, 50 mg/l salmon sperm DNA. Fifty ng of gus or gus+npt 11 DNA from pFAJ3000 were 32P-labelled with the "Ready to Go" kit (Pharmacia), purified on TEMIDI SELECT%D, G-50 columns (5 Prime --~ 3 Prime, Inc.'*) and used as probe. Hybridization was carried out for 16 h at prehybridization conditions. Filters were washed according to the membr',~e's manufacturer instructions and exposed to Kodak XOMAT films with one intensifying screen at -80°C.

Results Plant regeneration. Regeneration tests performed on

leaves of "Spring Glory" resulted in callus proliferation and shoot regeneration mainly from petiole cut extremities. However, 50 to 75% (for BAP concentrations of 0.5 and 1-2.5 mg/l respectively, irrespective of IAA concentration) of the regenerated shoots were vitrified. Therefore, the in vitro response of internode explants was tested, leading to reduced callus proliferation and rare appearance of vitrified shoots. Internode regeneration frequencies (fig. 3) clearly indicate that IAA is essential for plant regeneration. Analysis of variance performed on such data revealed highly significant differences within treatments (1% probability level); thus, M6 supplemented with BAP 1 mg/l and IAA 0.5 mg/1, which had ensured 100% regeneration rate, was retained for further transformation experiments. 100

~

8o

~

60

I°BnP°SmgI'

I~

IBAP nBAP

40

~

1.0 mg/I 2.5 mg/I

20 0

"

0

"

0,25

'.

0,5

I

I

I A A concentration (rag/l) Fig. 3. Regeneration frequencies of Forsythia x intermedia "Spring Glory" internodes 6 weeks after subculture on different regeneration medi~

116 When BAP was associated with IAA, steady callus proliferation occurred; microscopic observations on transversal sections of explants revealed actively dividing sub-epidermal ceils as early as 4 days after the beginning of culture. As a consequence of this, buds differentiated in the outer portions of cut edges of explants. Explants exhibited a striking polarity: the internode edge distal to the apex exhibited an earlier and markedly higher callus proliferation and shoot regeneration than the proximal one. Buds were observed as early as 2-3 weeks after the beginning of the culture. Though several buds differentiated on the same explant, only 1 or 2 shoots per explant eventually elongated (fig. 4).

high number of blue spots per explant, only 1% of explants regenerated transformed shoots. The transformed clones were detected among regenerants by GUS assays performed on the first leaves elongated from the regenerated shoots, which strongly expressed GUS activity, turning completely blue (fig. 5B). Escapes were also observed. Both escapes and control plants were unable to elongate, their leaves turned pale green and eventually bleached during subculture on selective medium. On the other hand, putative transformed shoots had deep green leaves and were able to grow and root on the propagation medium with a range of kanamycin concentrations from 25 to 80 mg/1. Internode explants of transgenic plants were able to regenerate adventitious buds on M6 105 supplemented with 25 mg/1 kanamycin, giving additional evidence of genetic transformation. Finally, GUS assays on transformants were repeated also after several in vitro subcultures and on greenhouse plants. In both cases, enzymatic GUS expression was still present in all organs tested (fig. 5C, 5D).

Fig. 4. Plant regeneration from internode explants of Forsythia x/ntermed/a "Spring Glory" 6 weeks after inoculation. Bar = 10 lnllk

Transformation and GUS assays. The length of the coculture period was fixed to 4 days by comparing GUS activity of internode sliced edges after 2 or 4 days of coculture (tab. 1). Tab. 1. GUS activity of internodes o f Forsythia x intermedia "Spring Glory" 8 days after inoculation. days of

explants with blue spots °

spots/explant '

coeulture

(%)

(n)

2

40

2.3

4

100

9.7

• Mean value of three independent experiments (10 explants/experiment). b Data refer only to explants with spots.

Doubling the length of the coculture period actually led to a 2.5-fold and a more-than-4-fold increase of the frequency of explants with blue spots and the number of spots per explant, respectively. GUS activity spots were mainly located near the cambial zone. When the number of GUS spots was recorded separately for the proximal and the distal edge of explants, the distal edges turned out again to be the most reactive, thus confirming the polarity already displayed for regeneration ability (fig. 5A). Five successive transformation experiments were then performed on 100 to 200 internode explants each according to the protocol shown in fig. 2. Despite the

Fig. 5. GUS assays on different plant material of Forsythia x intermedia "Spring Glory". A) pro~-nal (]eft) and distal (right) callused explant edges 4 days after inoculation; B) first leaf of a transgenle shoot; C) leaf and D) anther and petiole explants of adult transgenic plants. Bar beside letters = 1 nun.

PCR. Fig. 6 shows the results of GUS and NPT II PCRs. Samples of pFAJ3000 plasmid and the 4 analyzed kanamycin-resistant, putative transgenic clones showed amplification of the expected bands. Such bands were not present in the untransformed control, thus giving proof of foreign gene presence. Southern blot analysis. Two of the 4 clones examined by PCR (and obtained from different transformation experiments) were also subjected to Southern blot hybridization analysis. Fig. 7 shows the autoradiography patterns of DNA digested by Hind III and EcoR I, hybridized with gus+npt H and gus probes, respectively. For Hind III patterns, both transformed clones showed the 2.9-kbp band corresponding to the gus gene. The other visible bands should correspond to the npt H gene and allow to distinguish the two clones. For the clone 3, the hybridization pattern indicates a double insertion of the T-DNA. EcoR I patterns were consistent with Hind III ones for transformation proof, difference between restriction patterns, as well as

117 multiple insertion of T-DNA in clone 3.

Fig. 6. Gus and npt H PCRs on DNA of pFAJ3000 (P), untransformed "Spring Glory" (C') and 4 transformed clones (1, 2, 3 and 4)° L: BRL ladder. Fragment size is expressed in bp.

place from the subepidermal layers. This might explain the low regeneration frequency of transformed plants. The occurrence of escapes is a likely consequence of subculture onto selective medium only 7 days after inoculation. At this time the regeneration process is already established: control explant cultures on M6 105 medium clearly showed organized meristematic structures at day 10 of culture (results not shown). Several escapes can thus be formed, which may be antagonists for differentiation in transformed cells. To limit such competition, earlier addition of the selective agent could improve the transformation frequency. Another strategy may be to hasten the transformation process. In this respect, acetosyringone treatments were reported to enhance the virulence of A. tumefaciens (Godwin et al. 1991) as well as to accelerate the appearance of the first transformed cells (Gnivarc'h et al. 1993). These results encourage further transformation experiments. Work is currently underway to improve the transformation frequency and introduce foreign genes in order to change Forsythia flower colour and other ornamental traits. Acknowledgements

Fig. 7.

Southern blot hybridization of pFAJ3000 (P), untransformed "Spring Glory" (C) and two transformed clones (3 and 4) digested by H/nd III (IT) and EcoR I (E). Fragment size is expressed in kbp.

The authors wish to thank Dr. W. Broekaert (Katholieke Universiteit Leuven, Belgium)for kindly providingthe Agrobacteriumtumefaciens strain containingthe pFAJ3000 binary vector, Dr. Cardoen (PGS Gent, Belgium) for givingthe permissionto use the vector, Dr. E. Chevreau (1NRAAngers,France)for critical readingof the manuscriptand Dr. C. Lambert (ENITHP Angers, France) for her invaluable advice in early transformationexperiments. References

Discussion

In this work we report a reliable regeneration and transformation method for internode explants of Forsythia x intermedia "Spring Glory". Unlike leaf explants which were able to regenerate on M6 supplemented with BAP only (data not shown), internodes required both auxins and cytokinins for shoot organogenesis. IAA probably triggers cell division and differentiation at the wounded edges of explants. The polarity displayed by explants for shoot regeneration is often recorded in tissue culture systems and may be correlated with auxin transport: the same phenomenon for both callus proliferation and shoot regeneration was observed in Weigela, another ornamental species (Duron 1981). The percentage of transformed shoots per inoculated explant is low but comparable to other woody species systems (Schuerman and Dandekar 1993). Transgenic plant regeneration is closely dependent upon an efficient gene transfer method and an optimal regeneration medium. Furthermore, the transformation and regeneration sites in the explant must overlap. Forsythia explants did not meet the last condition: transformation sites were mainly located near the cambial zone, while regeneration events typically took

de BondtA, EggermontK, Druart P, De Vii M, GoderisI, Vanderleyden J, BroekaertW (1994) Plant Cell Reports 13:587-593 DellaportaS, WoodJ, HicksJB (1983) Plant Mol Biol Rep 4:19-21 Duron M (1977) C R Acad Sci s6rieD 284:183-185 Duron M (1981) Agronomic1:865-868 Firoozabady E, Moy Y, Courtney-Gutterson N, Robinson K (1994) Bin/Technology12:609-613 Godwin I, Todd G, Ford-Lloyd B, Newbury HJ (1991) Plant Cell Reports9:671-675 Guivarc~h A, Caissard J-C, Brown S, Marie D, Dewitte W, Van Onckelen H, Chriqui D (1993) Protoplasma174:10-18 Hajela RK, Hajela N, BolyardMG, Barnes WM, Sticklen MB (1993) HortSci28:330-332 HoodEE, HelmerGL, FrayleyRT, Chilton MD (1986) J Bacteriol 168: 1297-1301 JeffersonRA (1987) Plant Mol Biol Rep 5:387-405 Morel G, Martin C (1955) C R AcadAgric41:472-475 MurashigeT, SkoogF (1962) PhysiolPlant 15:469-497 Pavingerowl D, Biskov~iR, Kod3RekK (1995) Proc. XVIII Eucarpia Syrup,Tel Aviv, p. 23. Roest S, BokelmanGS (1980) Vakbladvoor de Bloemisterij45:34-35 Sambrook J, Fritseh EF, Maniatis T (1989) Molecular Cloning: A LaboratoryManual. 2nd edn, Cold SpringHarborLaboratoryPress, Cold SpringHarbor,NY SampsonDR (1971) Can J GenetCytol 13:368-371 SehuermanPL, DandekarA (1993) Sei Hortie55:101-124 Vaneanneyt G, SchmidtR, O' ConnoroSanchezA, WillmitzerL, RoehaSosa M (1990) Moi Gen Genet220:245-250