Genetic transformation of the apple rootstock M26 with the RolB gene and its influence on rooting

Introduction

• JOU •• ALOF. PlI"'P .. ,.~.I.., © 1998 by Gustav Fischer Verlag, Jena

Genetic Transformation of the Apple Rootstock M 26 with the Ro/B Gene and its Influence on Rooting

MARGARETA WELANDERI, NATHALIE PAWLICKI I, ANNA HOLEFORSl, and FIONA WILSON2

1 The Swedish University of Agricultural Sciences, Department of Horticulture, Box-55, S-230 53 Alnarp, Sweden

2 Horticultural Research International, East Malling, West Malling, Kent ME19 6BJ, UK

Received September 24,1997 . Accepted October 25, 1997

Summary

The apple rootstock M26 (Malus x domestica Borkh.) was transformed using the binary vector pCMB­B: GUS in the Agrobacterium tumefaciens strain CS8C!. The vector contained the nptII gene under the nos promotor and a tandem gene construct of the gus and the rolB genes both under the rolB promotor. Two transformed clones were verified with PCR and Southern blot analyses. One clone contained all three genes, whereas in the other clone only rolB and nptII genes were integrated in the plant genome. Rooting experiments using both intact shoots and stem segments showed that integration of the rolB gene in­creased auxin sensitivity and rooting ability compared with untransformed control shoots. GUS activity driven by the rolB promotor was higly expressed in the root meristem but not in the extended root. In the shoot, GUS activity was confined to the procambial strands. In the bud, the leaf primordia showed rather intense uniform staining. During growth of the leaf the provascular cells of the midrib showed weak GUS expression and developed leaves were almost unstained.

Key words: Malus X domestica, Agrobacterium rhizogenes, p -glucuronidase, genetic engineering, rolB, rooting.

Abbreviations: BA = benzyladenine; CTAB = hexadecyltrimethylammoniumbromide; GUS = ~-glucu­ronidase; kb = kilobase; MS = Murashige and Skoog; MUG = 4-methylumbelliferyl glucuronide; NM = a-naphthaleneacetic acid; PCR = polymerase chain reaction; SDS = sodium dodecyl sulphate; X-glue = 5-bromo-4-chloro-3-indolyl-~-D-glucuronide.

The most limiting factor in the vegetative propagation of many agronomically valuable cultivars and rootstocks is the very poor rooting efficiency. It is well known that rooting is developmentally regulated and that the rooting capacity of cuttings declines with the increase in age of the mother plant. When cuttings are competent to root, auxins are often re­quired for root induction; however, in many recalcitrant to root species the cuttings form only callus and no roots in re­sponse to exogenously applied auxins. Thus, the question ari­ses whether lack of rooting is due to changes in active endog­enous auxin concentrations or changes of tissue sensitivity to auxins.

The poor basic knowledge of adventitious rooting in woody species encourages new approaches to study root for­mation. New techniques and increased understanding of root induction should have important applications in agriculture, horticulture and silviculture.

The infection of dicotyledonous plants with Agrobacterium rhizogenes usually results in a proliferation of morphologically distinctive roots at the site of bacterial infection (Riker et al., 1930; Bevan and Chilton, 1982). This is achieved by the transfer of a small part of the root-inducing (Ri) plasmid, called T-DNA, into the plant genome (Chilton et al., 1982). On the T-DNA, 18 open reading frames (ORFs) have been identified (Slightom et al., 1986), and the role of four genes, designated rolA, rolB, rolC and ro/O, has been well described

J Plant PhysioL W11. 153. pp. 371-380 (1998)

372 MARGARBTA WBLANDBR, NATHAUB PAWLICKI, ANNA HOLBFoRS, and FIONA WILSON

(Cardarelli et aI., 1987 a; Spena et aI., 1987; Vilaine et aI., 1987). Among these oncogenes, rolB is the most effective in promoting rooting in different host plants (White et aI., 1985; Cardarelli et aI., 1987b; Spena et aI., 1987; Capone et aI., 1989), while rolA and rolC are mainly involved in the modification of the morphology and the growth rate of the induced roots. Moreover, rolB gene expression increases auxin sensitivity in tobacco protoplasts (Shen et aI., 1988; Mawel et aI., 1990). Delbarre et aI. (1994) demonstrated that the ac­cumulation and the metabolism of exogenous auxins in trans­genic tobacco leaf protoplasts were independent of the rolB gene expression level, and the increased sensitivity to auxin was due to changes in the auxin perception pathway.

Very recendy Filippini et aI. (1996) showed that an over­production of the rolB protein in Escherichia coli has tyrosine phosphatase activity and that in transformed plant cells it is localized in the plasma membrane. This strongly suggests that the rolB protein is involved in signaI transduction of the plant hormone auxin.

The aim of our work was to study the influence of the rolB gene and the rolB promotor on rooting in Malus. By using the construct pCMB-B: GUS carrying a rolB gene and rolB:gusgene fusion in tandem (Mawel et aI., 1994), the in­fluence and specific expression of the rolB gene could be studied in different organs, tissues and cells (Altamwa et aI., 1991). As a model plant, we used the apple rootstock M26 (Malus X dommica Borkh.), a semidwarf rootstock frequendy used throughout Ewope. Studies of this rootstock include its micropropagation Oames and Thurdon, 1981; Welander, 1983), shoot regeneration from leaf explants (Welander, 1988; Predieri and Fasolo, 1989), and genetic transformation by Agrobacterium tumefaciens (Maheswaran et aI., 1992).

The rooting capacity was studied in complete shoots and aIso with a test syst~ consisting of thin stem slices of 1 mm thickness (Welander and Pawlicki, 1993; Van der Krieken et aI., 1993). The disc system has severaI advantages. From one shoot about 7-9 stem discs can be obtained and used in the rooting test: all discs from the same stem have a comparable rooting response and show no positionaI effects, rooting is fast and aImost synchronous, and the rooting process occurs without interference of endogenous auxins and other com­pounds produced in the leaves.

Materials and Methods

Bactmal strains and plasmitis

In preliminary transformation experiments to study the influence of the bacterial background on transformation frequency, the binary vector pBI121 with the marker genes nptII and gus was used in two

different backgrounds C58Cl (pGV3850) and LBA4404 (pAL4404).

In later transformation experiments we used the binary vector pCMB-B : GUS in the Agrobacterium tumifacims strain C58Cl, kindly provided by C. Maurel (CNRS, GiflYvette, France). A sche­matic drawing of the vector showing position of the genes and re­striction sites is shown in Fig. 1. The vector contains the nptII gene under the nopaline synthase (nos) promoter, and a tandem gene construct consisting of the gus gene under the rolB promotor and the rolB gene under its own promotor. The Agrobacterium strain was cultured at 28 ·C in LB medium (1O giL tryptone, 5 gIL yeast ex­tract, 10 giL NaCI and 1 giL glucose) supplemented with 200 mg/L kanamycin and 75 mglL neomycin.

Transformation and "Ktntration proctdurt

The bacteria used for inoculation were grown overnight at 28 ·C in a shaking waterbath (150 rpm) to a density of A420 = 1.5 to 2.0. Bacterial cells were pelleted at 5126 g for 15 min, then resuspended to a final density of A420 = 0.5 to 1.0 in liquid MS (Murashige and Skoog. 1962) medium supplemented with 2 % sucrose. Young ex­panded leaves, excised from 5-week-old micropropagated shoots, were wounded with a scalpel and gently shaken in the bacterial sus­pension for 5 min. Leaf explants were then co-cultivated on a shoot regeneration medium at 24 ·C in the dark. The shoot regeneration medium consisted of MS basal medium supplemented with 5 mglL (22 l1fIlollL) BA, 0.2 mglL (OJ ~moIlL) NAA, 3 % sucrose and 0.25 % gelrite. After 3 days, the explants were rinsed in liquid MS medium containing 500 mglL cefotaxime, dry blotted on a sterile filter paper and transferred onto the regeneration medium supple­mented with 500 mglL cefotaxime. Ten days after co-cultivation, the explants were transferred, under identical cultural conditions, onto selection medium consisting of regeneration medium plus 250 mglL cefotaxime and 25 mglL kanamycin. When shoots appeared the ex­plants were transferrred to light and subcultured every 4 weeks on the selection medium. Shoots from individual leaves were excised and propagated separately on multiplication medium consisting of MS basal medium, 1 mglL (4.411f1l0IlL) BA, 0.1 mglL (0.5 ~moIlL) IBA, 3 % sucrose, 0.35 % agar (Bacto Difco) , 0.12 % gelrite, 100 mglL cefotaxime and 50 mglL kanamycin. Eight Petri dishes (A­H) each containing 7 leaves were used in the experiments.

Histochemical and jluoromttric GUS assays

Leaves and roots from regenerated kanamycin resistant plants were analysed for gus gene expression with X-Gluc (5-bromo-4-chlo­ro-3-indolyl-P-D-glucuronide) according to Jefferson et al. (1987). Plant materials were incubated with 1 mmol/L X-Gluc in 100 mmoll L phosphate buffer (pH 7.0) at 37·C overnight. GUS activity was also quantified in leaves, stems, and the basal part of the roots using 4 mmollL 4-methyl umbelliferyl-p-D-glucuronide (MUG) as the substrate Uefferson et al., 1987). The extraction procedure and fluo­rometric determinations of GUS activity (expressed in pmoll min/g. fw) were according to James et al. (1993).

RB !!!!!! p rol8 p rol8 GUS

I' Ii i H~~i:;: iiiiiii.-: <I LB HE BBH B(Hp) HB E

> 1.4 kb < > 6.S kb ----< >------------ 10.7 kb --------- <

FlJ. 11 A schematic drawing of the binary vec­tor pCMB-B: GUS showing the position of the genes and restriction sites. Lines below de­note the length of T-DNA and anticipated size of internal fragments using EcoR! and HindIII.

Histochemical studies were performed using X-gluc treated mate­rial. Both freeze microtome sections and sections from paraffin em­bedded material were used. Plant tissues were soaked in 70 % etha­nol, to clear chlorophyll, and fixed in FM. Subsequently, samples were quick frozen according to the method described by Gahan (1984), and then sectioned with a Brights' cryostat (30 - 50 JI.). For paraffin sections (15-20 JI.) the material was fixed, dehydrated and embedded as described by Welander and Pawlicki (1993).

peR analysis

Genomic DNA was extracted from in vitro leaf material using a CTAB method modified by H. Y. Yang (Institut fur Gartenbauliche Pflanzenziichtung, D-2070 Ahrensburg, Germany). The material was ground in 1.5 mL microfuge tubes containing pre-warmed 2 % CTAB buffer (2 % CTAB, 1.4 mollL NaCl, 20 mmollL EDTA, 100 mmol/L TRIS-Cl, pH 8.0, 2 % PVP-40, 1 % 2-Mercaptoetha­nol) and incubated for 10 min at 65 'C, then emulsified with 500 Jl.L dichloromethane:isoamyl alcohol (24: 1) and microfuged for 2 min. The top phase was transferred to a fresh microfuge tube and DNA was precipitated by adding 300 Jl.L isopropanol and microfuging for 2 min. The pellet was washed in 500 J1l wash buffer (76 % ethanol, 10mmollL ammonium acetate) for 2min and microfuged for 3 min before discarding the supernatant. The resultant pellet was dried briefly and resuspended in water. PCR reactions were carried out in 25 Jl.L volume containing 200 Jl.mollL of each dNTP, 1 Jl.mollL of each oligonucleotide primer, 2.5 U Taq Polymerase (Appligene) per 100 Jl.L and 25 ng DNA, using the following programme: 1 cycle at 94 ·C for 6 min, 30 cycles of 94 ·C for 1 min, 65'C for 1 min and 72 ·C for 1 min, and 1 cycle at 72 'C for 6 min. Amplified DNA fragments were electrophoresed on a 1 % agarose gel and visualised by staining with ethidium bromide.

The primers used for amplification of fragments of the genes were: rolB, position 1861 5'-ATGGATCCCAAATTGCTATTCCTTC­CACGA-3' and position 2637 5' -TTAGGCTTCTTTCTTCAGGT­TTACTGCAGC-3', yielding a 776bp product; nptII, position 1716 5' -GCCCTGAATGAACTGCAGGACGAGGC-3' and position 2127 5'-GCAGGCATCGCCATGGGTCACGACGA-3', yielding a 411 bp product; gus, position 275'-CCTGTAGAAACCCCAACC­CGTG-3' and position 392 5'-CCCGGCAATAACATACGGC­GTG-3', yielding a 365 bp product.

Southern blot analysis

Plant DNA was isolated from shoots of young in vitro propagated apple plants using a modification of the CTAB method described by Aldrich and Cullis (1993). Plasmid DNA was isolated using the al­kaline lysis method according to Sambrook et al. (1989). Ten JI.g DNA was digested in each reaction using 10 units of restriction en­zyme per Jl.g DNA over 2-3h. BeoRl was used for subsequent hy­bridisation with rolB and nptII probes, and HindIII for subsequent hybridisation with the gus probe. Digested DNA was separated on a 1 % agarose gel and transferred to a Hybond-N membrane (Amers­ham). RolB fragment, for use as hybridisation probe, was prepared in 100 Jl.L volume PCR reactions using the method desribed above. The PCR products were purified using the QIAquick PCR Purifica­tion Kit (Qiagen), The gus gene was excised from the plasmid pSCV1.6 (Glyn Edwards, Shell Forestry Unit, personal communica­tion) using restriction enzymes Kpn1 and Sma1, and purified from an agarose gel using the QIAEX Gel Extraction Kit (Qiagen). The nptlI probe was prepared according to Hamill et al., 1991. The pro­bes were labelled using the (Readi to Go, kit (Pharmacia) using 50 ng DNA fragment and 50 Jl.Ci 2p' Unincorporated nucleotides were removed using NICK columns (Pharmacia). Pre-hybridisation and hybridisation were carried out at 55 ·C for approximately 24 h and 15h, respectively, in 6xSSC, 5x Denhardt's reagent, 0.5% (w/v)

Transformation and Expression of RolB Gene in Apple 373

sodium dodecyl sulphate (SDS), and 100 JJ.gImL denaturated frag­ment salmon sperm DNA.

Rooting experiments

In the rooting experiments, we used transformed and untrans­formed shoots. Defoliated and debudded sterns from 4-week-old micropropagated axillary shoots were cut into discs of 1 mm thick­ness, using a set of razor blades. The discs were cultured with the basal part upwards on a rooting medium according to Pawlicki and Welander (1995). In a first experiment, the stem discs were in­cubated in the dark for 1, 3, 4, 5, 7 or 9 h on the rooting medium supplemented with 5 mglL (24.6 Jl.moIlL) IBA. In a second experi­ment, the stem discs were incubated in the dark for 4, 5 or 7 h on the rooting medium supplemented with 1 mglL (4.9 Jl.moIlL), 2.5 mglL (12.3 Jl.moI/L) or 5 mglL (24.6 Jl.moIlL) IBA. In both ex­periments the stem discs were transferred, afrer the IBA treatment period, to the basal medium without IBA under a 16: 8 h photope­riod. The rooting percentage and the number of roots per disc were recorded afrer 3 weeks of culture. Each treatment consisted of two Petri dishes, each containing 18 stem discs, and repeated once. Anal­ysis of variance (ANOVA) and Duncan's Multiple Range Test were conducted using the SAS program (GLM procedure, SAS Institute Inc., 1990). Percentage data were subjected to arcsine transforma­tion before statistical analysis and then transformed back for presen­tation. In addition, complete shoots from the transformed and un­transformed clones were tested for their rooting ability with or with­out IBA. Twenty-five shoots from each material were used. The rooting procedure was according to Welander (1983) except that shoots were cultured on a rooting medium with IBA at a concentra­tion of 0.25 mglL (1.23 Jl.moIlL) for the whole rooting period. The rooting percentage and number of roots per rooted shoot were re­corded afrer 3 weeks.

Results

Transformation and regeneration

Kanamycin resistant calli were observed after 3 weeks of culture in the dark on the leaf explants maintained on the re­generation medium. Transformed shoots were obtained after 2 months (Fig. 2A). The presence of the gus gene was verified in leaf and stem explants of the different kanamycin resistant shoots by staining with X-glue. In the preliminary transfor­mation experiments, using the binary vector pBI121, in two different backgrounds the obtained transformation frequency (expressed as GUS positive shoots per explant) was 2.5 % for C5SC1 and 0 % for the LBA4404 The transformation fre­quency using the vector pCMB: GUS in C58C1 was 7.1 %. Kanamycin resistant shoots obtained using the latter vector were transferred to the multiplication medium supplemented with 50 mglL kanamycin and kept separate. Because of vitri­fication and chimeras (Fig. 2 B), some of these shoots died af­ter 2 to 3 subcultures on the multiplication medium. Only axillary shoots originating from three adventitious shoots re­generated from two different leaves survived. These clones were labelled C3.2, Fl.l. and Fl.2., referring to Petri dish and leaf number. Transformed shoots multiplied and devel­oped in the same way as untransformed ones except that sometimes a pronounced callus formation was observed at the base of the shoots, with some roots emerging from the callus.

374 MARGARETA WELANDER, NATHAUE PAWLICKI, ANNA HOLEPORS, and FIONA WILSON

f& ~

~

;F"

:i . • ~Ii ?, . ' , " ~"" Ift' ~'-,~ .. : . . -. . . :.

I

L

.'

:. .. -~--:,./ .

' ..

) /' r "I , . ,

'"

+ I ,

peR and Southern blot analyses

The presence of the different genes was first investigated by PCR analysis (Fig. 3). Untransformed shoots (used as neg­ative controls), shoots from the C3.2 done, from the five F1.1 subdones (a,b,c,d,e) and from the two F1.2 subdones (a,b) were analysed. The nptII and rolB genes were detected in all plants tested, but the gus gene was only observed in the C3.2 done. The Southern blot analysis (Fig. 4) performed with untransformed M26 shoots, the done C3.2 and the subdones F 1.1 a, b, c and F 1.2 a, b confirmed the PCR results. EcoRl-digested DNA (yielding T-ONA border fragments)

Lod.C F F F F F F F F Co.Co.PI I Icb 3.2 1.1. I.Ib 1.10 I.Id 1.1. 1.20 I.lb 1.20 I 2

--NPTU -

ROLB

GUS

p---------

• ~ ----a- ---- ... ......

Fig. 3: PCR analysis of putative transformed shoots of the apple rootstock M26. The pictures show gd electrophoresis of nptII, rolB and gus gene products from two control plants (co 1,2) plasmid DNA (PI) and 9 transformed clones (C3.2, F1.1 a,b,c,d,e, F1.2 a,b,c).

Transformation and Expression of RolB Gene in Apple 375

hybridised with the nptII probe gave two bands of 5.25 and 3.9 kb in the C3.2 done, indicating the presence of 2 copies of the nptII gene, and 1 band of 5.75 kb in the subdones of F 1.1 and F 1.2, indicating 1 copy. EcoRl-digested DNA hyb­ridised with the rolB probe gave the expected band of 6.5 Kb for the C3.2 done; however, a larger fragment of 9 kb was de­tected in the F 1.1 and F 1.2 subdones, probably because the gus gene, as well as the EcoRI restriction site at the right bor­der, were not integrated. HindIII-digested DNA hybridised with the gus probe gave 2 bands in the C3.2 done: one strong band of 4.4 kb and a weak band of 3.5 Kb correspond­ing to 2 individual integrations of the gus gene. No band was detected in the subdones F 1.1 and F 1.2, confirming that the gus gene was not integrated in the F-dones. All transformed shoots analysed from the F 1.1 and F 1.2 subdones revealed the same banding pattern, indicating that all the regenerated F-subdones were genetically identical.

Enzymatic assay ofp-glucuronidase (GUS) activity

The GUS activity, determined by MUG assay, in leaves, stems and the basal part of the roots of untransformed M26 plants and transformed plants of the dones C3.2 and F 1.1 is presented in Table 1. The results show that very low activity was detected in the different plant parts of the untransformed plants and the F 1.1 clone, which lacked the gus gene. The C3.2 clone showed high GUS activity but at various levels in the different plant parts. Very little activity was detected in the leaves and the highest activity was recorded in the stems.

Histological analysis of the expression of the rolB-gus gene GUS activity was highly expressed in callus formed on

stem discs during rooting (Fig. 2 C, D); however, there was no activtiy in the cortex cells of the stem or in the root ex­tending from the stem (Fig. 2 D). The gus gene was also highly expressed in the root tip but confined to the root me­ristem (Fig. 2 C, E, F). In the elongation zone of the root the GUS activity was no longer expressed (Fig. 2 G). In the shoot tip GUS activity was detected in the procambial strands but not in the shoot meristem (Fig. 2 H). Figure 2 J shows a transverse section of a stem with an axillary bud. In the bud, leaf primordia showed an almost uniform staining as long as

Fig. 2: A: Putative transformed shoots (arrow) regenerated from leaf explants after 2 months in the dark. B: Complete transformed shoot (right) and variegated chimeric shoot (left) after several subcultures on medium containing kanamycin (50 mglL). C: Rooted stem disc showing GUS activity in callus around the disc, and the root tipx6.4. D: Longitudinal section of a rooted stem disc showing GUS activity in the callus (arrows). No activitiy is visible in the stem conex or in the extended root X 16. E: Longitudinal section of a root tip showing high GUS activity in the root meristemx312. F: Transverse sectionx312. G: Transverse section of the root in the elongation zone showing the absence of GUS activity X 312. H: Longitudinal section of the shoot tip showing GUS activity in the procambial strands. No activity was detected in the shoot meristemx312. I: Transverse section of a stem with an axillary bud showing almost uniform staining in young leaf pri­mordia. In the devdoping leaves GUS activity is expressed only in the procambial strands and is absent in mature leaves X 79. J: longitu­dinal section of the stem showing GUS activity in the xylem parenchymax312. K: Transverse secclonx312. L: Transformed (n and un­transformed (Nn M26 shoots cultured in rooting medium with (+) or without (-) IBA at 0.25 mglL. NT shoots did not root without IBA wheras T shoots produced normal thin long roots. In the presence of IBA both T and NT-shoots formed roots but the T shoots were more sensitive to the auxin resulting in profuse callus at the base and inhibited root growth. M: Rooting of stem discs from two transformed clones (F1.1 and C3.2) and the negative control (Co), exposed to three different IBA concentrations (5, 2.5 and 1 mglL) for 5 h before being transferred to hormone-free medium. The picture shows that the F 1.1 clone has the highest root number and requires the lowest auxin concentration to induce rooting.

376 MARGARBTA WBLANDER, NATHALIE PAWLICKI, ANNA HOLEPORS, and FIONA WILSON

NP11I ROLB GUS

C F F F F F Coat C F F F F F Coat C F F F F F Coal ~b U 1 .la 1.lb 1.10 1.2a l.lb U 1 .la 1.lb 1.10 1.2a l.lb U 1.la I.Ib 1.10 1.2a 1.2b

F.,4: Southern blot analysis of putative transformed shoots of the apple rootstock M26. The genomic DNA was isolated from the untrans­formed control and six clones (C3.2, F1.1a,b,c and F1.2 a and b), digested with either EcoRl or HindIII and probed with p32 labelled frag­ments of nptII, ro/B and gus.

100 -y----------.

o 1h 3h 4h 5h 7h

ISA Expoeure time

14

12 J 10 "

1 8

6

4

2

0 1 h 3h 4h 5h 7h

ISA Expoeure time

they exhibited radial symmetry. When the leaf obtained dor­siventral symmetry the GUS expression was confined to the cells in the procambial strands. During growth of the leaf the provascular cells of the midrib showed weak GUS expression and expanded leaves were almost unstained. In the stem, GUS expression was localized in the conducting tissue, mainly in the xylem parenchyma (Fig. 2J Isn, Fig. 2K tvsn).

9h

9h

Fig. 5: Rooting ability of stem discs from two differ­ent transformed clones F 1.1 (0) and C3.2 (II) and the negative control (II) of the apple rootstock M26. The stem discs were cultured on a rooting medium containing IBA (5 mglL) for 1, 3, 4, 5, 7 and 9 h in the dark and then transferred to a hor­mone free rooting medium. Rooting percentage and number of roots were recorded after 3 weeks.

Rooting ability of transformed and untransformed shoots

Figure 5 shows the rooting ability of stem segments from the two transformed clones, F 1.1 and C3.2, as well as the un­transformed control exposed to IBA at 5 mglL for 1, 3, 4, 5, 7 and 9 h. Adventitious root formation occurred only in the presence of exogenous auxin for both transformed and un-

Table 1: GUS-activity expressed as pmol MUlmin/g fw in leaf, stem and root tissue from two transformed clones (Fl.l and C3.2) and the negative control of the apple rootstock M26. The C3.2 clone contains the nptIl, rolB and gus genes whereas in the Fl.l clone the gus gene is absent. Values are the means ±S.D. from two repetitions.

Clone

M26 (control) F1.l C3.2

Leaf

20.4±10.1 24.5±16.8

303.1±211.2

Stem

18.7±14.7 8.3±8.2

6917±1700

Root

28.6±32.5 27.1±25.7 1238±838

Table 2: Rooting ability of intact shoots from two different trans­formed clones (Fl.l and C3.2), and the negative control of the apple rootstock M26. The shoots were cultured on a rooting medium con­taining 0.25 mgl!. IBA in the dark for 5 days and then transferred to light. The rooting percentage and the number of roots per rooted shoot (mean ±S.D.) were recorded after 3 weeks.

Clone % rooting No of roots/rooted shoot

-IBA +IBA -IBA +IBA M26 (control) 0 64 0 2.7± 1.2 F1.l 100 67 5.5±2.1 4.1±2.8 C3.2 94 65 5.2±3.5 2.9± 1.9

transformed shoots (data not shown). The rwo transformed clones showed a significantly (::;;0.001) higher rooting capac­ity than the untransformed control clone. The F 1.1 clone ex­hibited the highest sensitivity to auxin and rooted at a fre­quency of 92 % after 1 h of exposure to auxin. The untrans­formed control segments required 9 h of auxin treatment to reach 100 % rooting whereas the F 1.1 and C3.2 clones reached 100 % rooting after 4h. The F 1.1 clone produced the highest number of roots per disc (12.9 ± 1.2) followed by the C3.2 clone (8.2 ± 0.9) and the control clone (6.2 ± 0.4).

Callus formation was also observed on the discs during rooting. The frequency of callus formed on the stem discs in­creased with the length of exposure time to auxin especially for the transformed clones (data not shown).

Figure 6 shows the rooting ability of stem segments ex­posed to three IBA concentrations (1, 2,5 and 5 mglL) in combination with different lengths of auxin treatment (4, 5 or 7h). IBA at 1- and 2.5 mglL were not able to induce root­ing in the untransformed shoots within the chosen exposure time. The more auxin sensitive F 1.1 clone rooted to a fre­quency of 64 % on 1.0 mg/L IBA after a 5 h exposure time whereas the C3.2 clone required 2.5 mg/L IBA and the con­trol clone 5mglL to reach the same level. The F 1.1 clone pro­duced significantly (p ::;;0.001) more roots per rooted disc compared with the other clones. There was no significant in­fluence of the lengths of auxin treatment. Figure 2 M shows the rooting of stem segments from the control and the rwo transformed clones exposed to IBA at 1, 2.5 and 5 mglL for 5h.

Table 2 shows the rooting ability of intact shoots from the rwo transformed clones, F 1.1 and C3.2, and of untrans­formed shoots in presence or absence of IBA (0.25 mglL).

Transformation and Expression of RolB Gene in Apple 377

Unlike the stem segments the intact shoots of the F 1.1 and C3.2 clones formed roots on a medium without IBA whereas the untransformed shoots needed IBA. Shoots of the rwo transformed clones produced significantly more roots than the control shoots but there were no significant differences berween the transformed clones. The presence of IBA in the rooting medium reduced significantly (p::;;0.05) both rooting percentage and number of roots per rooted shoot in the transformed clones. Moreover, in presence of IBA, the trans­formed shoots developed profuse callus at the base, showing the increased sensitivity to auxin. Figure 2 L shows the root­ing performance of the untransformed shoots and the clone C3.2: the transformed shoot formed thin long roots in the absence of IBA whereas in the presence of IBA a profuse cal­lus was formed at the base.

Discussion

The transformation frequency is influenced by both vector and bacterium back-ground. Co-cultivation with the vector pCMB: GUS in C58Cl resulted in a transformation fre­quence of 7.1 %, the highest value reported for the apple rootstock M26. The vector pB 1121 used in rwo different backgrounds produced only 2.5 % transformed shoots in C58 C1 and none in LBA4404. The effect of different vec­tors and bacterium strains has been shown earlier in M26 by Maheswaran et al. (1992) and in apple cultivars (Puite and Schaart, 1996). Thus, optimal combinations of vir systems and chromosomal background giving rise to efficient T-DNA transfer for a particular species or even variety can only be de­termined by empirical testing.

Shoot regeneration from apple leaves is very sensitive to kanamycin (Maheswaran et al., 1992; Sriskandarajah et al., 1994). If concentrations high enough to avoid escapes are used the number of transformed shoots will be very low be­cause supporting tissue around transformed cells will die. If lower concentrations are used the number of transformed shoots will increase but also the number of escapes. Another disadvantage is that the number of chimeric shoots will also increase. In this paper we used only 25mglL kanamycin dur­ing the first weeks until shoots appeared. In the multiplica­tion medium the concentration of kanamycin was increased to 50 mglL. Using this method we could increase the trans­formation frequency but more time had to be spent on elimi­nating escapes and chimeric shoots. In our experiments we tried to avoid callus formation and work with direct regenera­tion; however, for many apple cultivars this is not possible and the only way to obtain transformed shoots is from trans­formed calli (De Bondt et al., 1994; Puite and Schaart, 1996).

The GUS expression of the ~-glucuronidase gene was dri­ven by a region of 1177 bp upstream of the rolB start codon. This construct is similar to the long version of the rolB pro­motor used by Altamura et al. (1991). GUS expression in the different organs and callus was very similar in transgenic apple and tobacco.

Measurement of GUS activity by fluorometric assay showed higher activity in the stems than in the roots. This is due to the fact that the gus gene is only expressed in the root

378 MARGARETA WELANDER, NATHALIE PAWUCKI, ANNA HOLEFORS, and FIONA WILSON

1 mgIIlBA

100T-------------------~

4h 5h 7h

IBA ExpoMn time

2.5 mglllBA

100 ,.------------..,...,..,

o 4h 5h 7h

IBA ExpoeuN time

5 mglllBA

4h 5h 7h

IBA Expo8UN time

J 12 'a 10

n 1 mglllBA

4h 5h 7h

IBA Expoewe time

2.5 mglllBA

12T--------------------~

i 10

n o

4h 5h 7h

IBA Expoeunt time

5 mglllBA

4h 5h 7h

IBA Expoeunt time

Ftg. 6: Rooting ability of stem discs from two different transformed clones F 1.1 (0) and C3.2 (II) and the negative control (II) of the apple rootstock M26. The stem discs were cultured on rooting medium containing three different concentrations of IBA (1, 2.5 and 5 mg/L) for 4. 5 and 7h in the dark and then transferred to a hormone free rooting medium. The rooting percentage and the number of roots were re­corded after 3 weeks.

meristem. Because the rolB promotor is tissue and cell spe­cific, results of measurement of activity in whole organs were somewhat deceptive.

The advantage of using both complete shoots and stem segments in the rooting experiments is that the rooting re­sponse can be studied both in the presence and absence of compounds produced in the leaves and shoot tips. The root­ing experiments with complete shoots show clearly that shoots transformed with the rolB gene have increased rooting

capacity. Increased rooting is probably due to increased auxin sensitivity since the shoots form roots in the absence of ap­plied auxin. which is necessary for untransformed shoots. Transformed shoots also produce a profuse callus at the base of the stem in the presence of very low auxin concentrations.

Stem segments from transformed shoots cultivated on a medium without auxin did not produce any roots. This shows that the rolB gene can not induce roots without auxin. This was also shown for carrot discs transformed with the

rolB gene (Capone et al., 1989). However, in intact shoots the amount of auxin produced in the shoot tip and young leaves is enough to induce rooting in transformed shoots but not in untransformed ones. Stem segments from rolB shoots require a lower auxin concentration for root induction than untransformed ones and also produce more roots per stem disc. This might indicate that increased auxin sensitivity could induce root primordia in normally unresponsive cells as suggested by Altamura et al. (1994) for tobacco. Our experi­ments show that auxin is necessary for root induction and strongly confirm the theory that expression of the rolB gene increases sensitivity to auxin (Maurel et al., 1991, 1994; Nils­son et al., 1993; Delbarre et al., 1994).

Interestingly, the clone F 1.1 without the gus gene showed a higher rooting response than the C3.2 clone with the gus gene. Recent experiments with the apple rootstock Jork9 (Se­dira et al. unpubl.) show similar results, indicating that not only the positional effect, that is where the T-DNA is in­serted, influence gene expression but also the surrounding genes. Our results show that the rolB gene seems to be very useful in increasing root formation in woody species. Re­search is now under way to insert the rolB gene into the very difficult to root apple rootstock M9 and the pear rootstock PB 30 selected in the Swedish breeding programme.

Acknowledgements

The authors are indebted to Mrs. Annelie Ahlman for her excel­lent technical assistance, Jan-Erik Englund for statistical treatments and the Swedish Council for Forestry and Agricultural Research for financial support. The exchange of information within the Con­certed Action on Genetic Transformation of Fruit Tree Species (EU contract number AlR3-CT93-0847) is highly acknowledged.

References

ALDRICH, J. and C. CULLlS: RAPD analysis in flax: optimization of yield and reproducibility using KlenTaq 1 DNA polymerase, Che­lex 100, and gel purification of genomic DNA. Plant Mol. BioI. Rep. 11, 128-141 (1993).

ALTAMURA, M. M., T. ARCHILETTI, I. CAPONE, and P. CONSTAN­TINO: Histological analysis of the expression of Agrobactmum rhi­zogm~s rolB-GUS gene fusions in transgenic tobacco. New Phy­tol. 118, 69-78 (1991).

ALTAMURA, M. M., F. CAPITANI, L. GAZZA, I. CAPONE, and P. CONSTANTINO: The plant oncogene rolB stimulates the forma­tion of flower and root meristemoids in tobacco thin cell layers. New Phytol. 126, 283-293 (1994).

BEVAN, M. W. and M. D. CHILTON: T-DNA of the Agrobacttrium Ti and Ri plasmids. Annu. Rev. Genet. 16, 357-384 (1982).

CAPONE, I., L. SPANO, M. CARoARELLl, D. BELLINCAMPI, A. PETIT, and P. CONSTANTINO: Induction and growth properties of carrot roots with different complements of Agrobactmum rhizogm~s T-DNA genes. Plant Mol. BioI. 13,43-52 (1989).

CARoARELLI, M., L. SPANO, D. MARIOTTI, M. L. MAURO, and P. CONSTANTINO: The role of auxin in haity root induction: Mol. Gen. Genet. 208,457-463 (1987 a).

CARoARELLI, M., D. MARIOTTI, M. POMPONI, L. SPANO, I. CA­PONE, and P. CONSTANTINO: Agrobactmum rhizogm~sT-DNA ge­nes capable of inducing haity root phenotype. Mol. Gen. Genet. 209,475-480 (1987 b).

Transformation and Expression of RolB Gene in Apple 379

CHILTON, M. D., D. A. TEPFER, A. PETIT, C. DAVID, F. CAsSE­DELBART, and J. TEMPE: Agrobactmum rhizogm~s inserts T-DNA into the genome of host plant root cells. Nature 295, 432-434 (1982).

DE BONDT, A., K. EGGERMONT, P. DRUART, M. DE VIL, I. GODE­RIS, J. VANDERLEIDEN, and W. F. BROEKAERT: Agrobacterium­mediated transformation of the apple (Malus do~stica Borkh.): an assesment of factors affecting gene transfer efficiency during early transformation steps. Plant Cell Rep. 13,587-593 (1994).

DELBARRE, A., P. MULLER, V. IMHOFF, H. BARBIER-BRYGOO, C. MAUREL, N. LEBLANC, C. PERROT-REcHENMANN, andJ. GUERN: The rolB Gene of Agrobactmum rhizogm~s Does Not Increase the Auxin Sensitivity of Tobacco Protoplasts by Modifying the Intra­cellular Auxin Concentration. Plant Physiol. 105, 563-569 (1994).

FILIPPINI, F., V. ROSSI, O. MARIN, M. TROVATO, P. COSTANTINO, P. M. DOWNEY, F. La SCHIAVO, and M. TERZI: A plant oncogene as a phosphatase. Nature 379, 499-500 (1996).

GAHAN, P. B.: Plant histochemistry and cytochemistry. Academic Press, INC. {London) LTD. ISBN 0-12-273270-7 (1984).

HAMILL, J. D., S. ROUNSLEY, A. SPENCER, G. TEDD, and M. J. C. RHODES: The use of the polymerase chain reaction in plant trans­formation studies. Plant Cell Rep. 10, 221-224 (1991).

JAMES, D. and I. J. THURBON: Phenolic compounds and other fac­tors controlling rhizogenes <in vitrq, in the apple rootstocks M9 and M26. Z. Pflanzenphysiol. 105, 11-20 (1981).

JAMES, D. J., S. URATSU, J. CHENG, P. NEGRI, P. VISS, and A. M. DANDEKAR: Acetosytingone and osmoprotectants like betaine or proline synergistically enhance Agrobactmum-mediated transfor­mation of apple. Plant Cell Rep. 12,559-563 (1993).

JEFFERSON, R. A., T. A. KAVANAGH, and M. W. BEVAN: GUS fu­sions: ~-glucuronidase as a sensitive and versatile gene fusion marker in higher plants. EMBO J. 6, 3901-3907 (1987).

MAHESWARAN, G., M. WELANDER, J. F. HUTCHINSON, M. W. GRA­HAM, and D. RICHARDS: Transformation of Apple Rootstock M26 with Agrobactmum tumifacims. J. Plant Physiol. 139,560-568 (1992).

MAUREL, c., J. BREVET, H. BARBIER-BRYGOO, J. GUERN, and J. TEMPE: Auxin regulates the promoter of the root-inducing rolB gene of Agrobactmum rhizogm~s in transgenic tobacco. Mol. Gen. Genet. 223, 58-64 (1990).

MAUREL, c., H. BARBIER-BRYGOO, A. SPENA, J. TEMPE, and J. GUERN: Single rol Genes from the Agrobacterium rhizogm~s TL­DNA Alter Some of the Cellular Responses to Auxin in Nicotiana tabacum. Plant Physiol. 97, 212-216 (1991).

MAUREL, c., N. LEBLANC, H. BARBIER-BRYGOO, C. PERROT­RECHENMANN, M. BOUVIER-DURAND, and J. GUERN: Alterations of Auxin Perception in rolB-Transformed Tobacco Protoplasts. Plant Physiol. 105, 1209-1215 (1994).

MURASHIGE, T. and F. SKOOG: A revised medium for rapid growth and bioassays with tobacco tissue cultures. Physiol. Plant. 15, 473-497 (1962).

NILSSON, 0., A. CROZIER, T. SCHMULLING, G. SANDBERG, and O. OLSSON: Indole-3-acetic acid homeostatis in transgenic tobacco plants expressing the Agrobactmum rhizogm~s rolB gene. Plant J. 3,681-689 (1993).

PAWLICKI, N. and M. WELANDER: Influence of carbohydrate source, auxin concentration and time of exposure on adventitious root­ing of the apple rootstock Jork 9. Plant Science 106, 167-176 (1995).

PREDIERI, S. and F. FASOLO FABRI MALAVASI: High-frequency shoot regeneration from leaves of the apple rootstock M26 (Malus pu­mila MilL). Plant, Tissue and Organ Culture 17, 133-142 (1989).

380 MARGARETA WBLANDER, NATHALIE PAWLICKI, ANNA HOLEFORS, and FIONA WILSON

PuITE, K. L. and J. ScHAART: Genetic modification of the commer­cial apple cultivars Gala, Golden Delicious and Elsrar via an Ag­robacterium tumqacimrmediated transformation method. Plant Science 119, 125-133 (1996).

RIKER, A., w. BANFIELD, W. WRIGHT, G. KEITT, and H. SAGEN: Studies on infectious hairy root of nursery apple trees: J. Agric. Res. (Washington DC) 41,507-540 (1930).

SAMBRooK, J., E. F. FRITSCH, and T. MANIATIs: A Laboratory Ma­nual, Cold Spring Harbor Laboratory Press, New York (1989).

SHEN, W. H., A. PETIT, J. GUERN, and J. TEMPB: Hairy roots are more sensitive to auxin than normal roots. Proc. Natl. Acad. Sci. USA 85, 3417-3421 (1988).

SUGHTOM, J. L., M. DURAND-TARDIF, L. JOUANIN, and D. TEPFER: Nucleotide sequence analysis ofU-DNA of Agrobacterium rhizo­gmn agropine type plasmid: identification of open-reading fra­mes. J. BioI. Chem. 261, 108-121 (1986).

SPENA, A., T. SCHMiiLuNG, C. KONCZ, and J. S. SCHELL: Independ­ent and synergistic avtivity of rol A, B and C loci in stimulating abnormal growth in plants. EMBO J. 6. 3891-3899 (1987).

SRISKANDARA]AH, S., P. B. GOODWIN, and J. SPEIRS: Genetic trans­formation of the apple scion cultivar ,Delicious) via Agrobacte­rium tumqacims. Plant Cell, TISSUe and Organ Culture 36. 317-329 (1994).

VAN DER KRIEKEN, W. M., H. BRETELER, M. H . M. VISSER, and D. MAVRlDOU: The role of the conversion of IBA into IAA on root regeneration in apple: introduction of a test system. Plant Cell Rep. 12,203-206 (1993).

VILAINE, F., c. CHARBONNIER, and F. CAsSE-DELBART: Further in­sight concerning the TL region of the Ri plasmid of Agrobacte­rium rhizogm~s strain A4: Transfer of a 1.9 kb fragment is suffi­cient to induce transformed roots on tobacco leaf fragments. Mol. Gen. Genet. 210, 111-115 (1987).

M. WELANDER: <In vitro> rooting of the apple rootstock M26 in adult and juvenile growth phase and acclimatization of the plant­lets. Physiol. Plartt. 58, 231-238 (1983).

- Plant regeneration from leaf explants and stem segments of shoots raised in vitro from mature apple trees. J. Plant Physiol. 132,738-744 (1988).

WELANDER, M. and N. PAWLICKI: A model system for studying root regeneration in woody species. Acta Hort. 336. 225 - 230 (1993).

WHITE, F. F., B. H. TAYWR, G. A. HUFFMAN, M. P. GORDON, and E. W. NESTER: Molecular and Genetic Analysis of the Transferred DNA Regions of the Root-Inducing Plasmid of Agrobactmum rhizogm~s. J. Bacteriol. 164, 33-44 (1985).