Plant Cell Reports
Plant Cell Reports (1988) 7:229-232
© Springer-Verlag 1988
Genetic transformation of 9 in vitro clones of Alnus and Betula by Agrobacterium tumefaciens John M a c k a y , Armand Srguin, and Maurice Lalonde Centre de Recherche en Biologie Foresti~re, Facult~ de Foresterie et de G6od6sie, Universit6 Laval, Sainte-Foy, Quebec, Canada G1K 7P4, Canada Received January 7, 1988/Revised version received March 30, 1988 - Communicated by F. Constabel
ABSTRACT Crown gall tumorigenesis, integration and expression of T-DNA encoded genes from Agrobacterium tumefaciens were investigated in 9 clones of Alnus glutinosa, A. incana and Betula papyrifera. Tumor formation on in vitro shoots was frequent in all clones with strain Ach5 and present in 8 clones with strain C58. Tumors excised from shoots were selected for autotrophic growth in vitro and axenic cultures were established. Octopine or nopaline, respective of the strain type used for inoculation, was detected in tumorous cultures. Southern blot analyses demonstrated T-DNA integration by hybridization of DNA from tumors with tmr and no._.~s gene probes. One clone of B. papyrifera produced tumors with a morphogenic character, unusual in calli of this species, generating viable shoots which did not synthesize opine. ABBREVIATIONS Cb, Carbenicillin; Cf, Cefotaxime; 2,4-D, 2,4-Dichloro-phenoxyacetic acid INTRODUCTION
Agrobacterium mediated gene transfer has been reported in a variety of agronomically-importantcrop plants (Goodman et al., 1987). Various Ti plasmid vectors and transformation procedures have been developed for use with plant model systems such as tobacco (Bevan, 1984) and tomato (Horsch et al., 1985). According to host range studies (De Cleene and De Ley, 1976), Agrobacterium can infect a wide spectrum of woody plants including angiosperms and some gymnosperms, however until recently there were few reports demonstrating transformation of forest trees. T-DNA integration into the plant genome was demonstrated with Populus (Parsons et al., 1986), Pinus (Stomp et al., 1987) previously excluded from Agrobacterium's host range (De Cleene and De Ley, 1976) and Pseudotsuga (Dandekar et al., 1987). Pathogenicity has been shown in Salix (Vahala et al., 1987), Abies and Picea (Clapham and Ekberg, 1986) and a few others. The genus Alnus (alder) is comprised of woody plants that can fix atmospheric nitrogen through a symbiotic association with the actinomycete Franla'a. Nitrogen fixing woody plants are of interest for intensive forest production as well as land reclamation (Chatarpanl and Carlisle, 1983; Hall and Maynard 1979). As of yet, little is known about the plant symbiotic genes and their regulation. The development of vector systems for gene transfer and expression in plant cells could aid in determining the regulation and specific functions of the genes involved (Herrera-Estrellaet al., 1985). Transformation vector systems for plants could lead to transfer of plant symbiotic genes to non nitrogen fixing species, as well. Betula (birch) which is closely related to Alnus is a candidate for such gene transfers (Normand and Laionde, 1986). We report here on crown gall tumorigenesis and the genetic transformation of 9 in vitro propagated clones of the genera Alnus and Betula by two wild type strains of Agrobacterium Offprint requests to." M. Lalonde
tumefaciens. Transformation is demonstrated by phyto-hormone independant growth, opine production and T-DNA integration. We conclude that Agrobacterium mediated gene transfer with engineered Ti plasmids may be applied in these species, such as in Populus (Fillati et al., 1987). M A T E R I A L S AND METHODS
Plant Material and Bacterial Strains Nine plant clones, 1 of Alnus glutinosa (L.) Gaertn., 4 ofAlnus incana (L.) Moench. and 4 ofBetula papyrifera Marsh. (Table 1) were used to test Agrobacterium infectivity. These clones are from varied geographic origins and two (AI-41, AI-45) were initiated from mature individuals ofA. incana, 22 and 20 years old (Prrinet et al., 1988). All in vitro shoots were obtained through micropropagation as described by Tremblay et al. (1986). The multiplication medium used in our assays consisted of MS salts, with 3% sucrose or 1.5% glucose for A. incana, 5 ktM BAP (6Benzyl-anainopurine) and was solidified with 0.7% to 0.8% agar (Difco Laboratories). In the roofing medium, macro-nutrients were reduced to half and the growth regulator was 1 pM IBA (Indole-3butyric acid). Two wild type strains of Agrobacterium tumefaciens (kindly provided by Dr P. Dion, Universit6 Laval, Qu6bec) were used: the octopine-type Ach5 and nopaline-type C58. The bacteria were grown at 29°C for 2 days on LB plates.
Inoculation with Agrobacterium A general method using whole shoots was devised to assay Agrobacterium tumefaciens infectivity on the clones listed above (Table 1). Different growth media and inoculation methods were tested to determine optimal conditions. Whole shoots 4-5 cm tall were preconditioned for 5-6 days either on multiplication or rooting medium and returned to preconditioning medium after inoculation (no antibiotics). For each inoculation, bacteria were introduced onto a sharp blade by directly collecting a few colonies of a freshly grown plate. The shoots were then inoculated either by decapitation or superficially wounding the stem. The controls received identical treatments but without the introduction of bacteria.
Culture of Tumors and Differentiated Shoots Tumors were excised from shoots 30-40 days after inoculation with Agrobacterium and grown on NT (Nagata and Takebe, 1971) medium supplemented with 500 mg L -1 CB and 250 mg L -1 Cf. 2,4-D was later used with this medium at concenlxations of 2.5 and 5 gM, to increase the growth rate. To eliminate all bacteria, cultured tumors larger than - 7 mm in diameter were surface sterilised in a 1% solution of NaOC1 (commercial Javex) for 5-10 minutes, surface cell layers were removed and the inner core was plated on NT medium with 2,4-D, without antibiotics. Smaller tumors were simply maintained on media with antibiotics. Shoots
230 TABLE 1. Description and origin of Alnus and Betula clones. Species Alnus glutinosa Alnus incana Betula
papyrifera
Clone acronyme AG-8 AI- 1, AI-2 AI-41, AI-45 BP-15, BP-16
Reference Tremblay and Lalonde, 1984 Idem P6rinet et al., 1988 Thispaper 1
BP-28, BP-34
1. Betulapapyrifera cultures were initiated from seed (Berthier- ville, QuEbec). Seeds were sterilized with 30% peroxide for 7 minutes followed by 1% solution of NaOCI (commercial Javex) also for 7 minutes, thoroughly rinsed in distilled water and germinated in vitro on water solidified with 1% agar, containing 3% (w/v) sucrose.
from morphogenic tumors were collected and propagated individually on the multiplication medium supplemented with 100 mg L -1 Cb and 50 mg L -1 Cf. The shoots were placed on standard rooting medium, with antibiotics, in solid (0.7% agar) and liquid (no agar) forms, for root induction. All tumorous calli and shoots differentiated from tumors were sub-cultured at 3-4 week intervals. Opine Detection Crude ethanol extracts were electrophoresed on paper and visualised with phenanthrenequinone as described by Parsons et al. (1986) with the following modifications. A low pH buffer was used (formic acid: acetic acid: water, 3:6:91), 20 laL of each extract was spotted onto 3MM chromatographic paper (Whatman®) and electrophoresed at 800V for 1 hour. The effect of incubating tumorous calli on 5 mM arginine for five days prior to extraction was tested. Opines were determined by comigration with octopine and nopaline standards, arginine served as a standard for negative discrimination. DNA Extraction and Southern Blot Analysis Agrobacterium tumefaciens Ti-plasmid was obtained by a partial purification (Currier and Nester, 1976). DNA was extracted from tumorous and non transformed calli following the procedure described by Malmberg et aI. (1984). In order to yield highly digestable DNA, further purification on a CsC1 gradient was performed. The DNA was suspended in 4.0 ml of 10 mM Tris pH 8.0, 1 mM EDTA, 4.0 g of CsC1 were added to the solution and the DNA banded by equilibrium density centrifugation (Maniatis et aL, 1982). DNA was digested with restriction enzymes under conditions specified by the supplier (BRL). Digested DNA was subsequently separated by gel electrophoresis in 0,8% agarose and blotted onto Gene Screen Plus membranes (New England Nuclear) following the protocols recommended by the manufacturer. 32p_ labelled probes: pRA123 containing a eDNA clone of the tmr gene from pTiA6 (provided by Dr E. W. Nester, University of Washington, Seafle) and a cloned 2.3 kb fragment including the nos gene of pTiT37 (Bevan et al., 1983; provided by Dr B. Miki, Agriculture Canada, Ottawa) were produced by nick translation to 5-10x108 cpmAtg as described by Maniatis et al. (1982). The probes were hybridized to the membrane-bound genomic DNA (following membrane manufacturer's protocol) and washed in stringent conditions (0.2X SSC, 65°C) as described in Maniatis et al. (1982), and visualized by exposure to X-ray film (Kodak XAR) for 24 hours with an intensifying screen (Lightening Plus, DuPont) at -70°C. RESULTS A high incidence of tumorigenesis was observed in shoots superficially wounded on the stem, but only when these were preconditioned on and returned to rooting medium, containing IBA. This treatment incited tumors in all clones tested (Figure la,
Table 2). Macroscopic tumorous growth was detected as early as 7 to 9 days after inoculation in B. papyrifera and I1 to 13 days in Alnus clones. Multiplication medium containing BAP completely inhibited tumorigenesis and wounding of the shoot apex was also unfavorable to tumor development (data not shown). No tumors or calli were observed on control plants (Table 2). Morphology of tumorous outgrowth as observed early on resulted from multiple transformation events at a given wounding site (data not shown). Growth was more rapid in tumors of B. papyrifera (Figure la), which were observed to reach 10 mm in diameter after 30-40 days. With the clones of Alnus and Betula tested, strain Ach5 was generally more effective in causing tumorigenesis (Table 2). Excised tumors were cultured as caUi (Figure lb). Large tumors were surface sterilized and dissected: four tumorous bacteria-free cultures of B. papyrifera and A. glutinosa were thus obtained. In addition, several tumors yielded axenic cultures without surface sterilization, after 3-4 subcultures on antibiotics. In absence of growth regulators and with antibiotics (500 mg L -1 Cb and 250 mg L -1 Cf), growth was slow but sustained and most tumors remained compact. Cell proliferation was sufficient for subdivision of cultures at 4 week intervals. Phyto-hormone independent cultures were selected after 7 months. After selection, the growth rate was doubled by the addition of 5 gM 2,4-D. Opines were detected in a number of tumors randomly selected among those which grew in absence of phyto-hormones (Table 2). Tumors induced by C58 all synthesized nopaline regardless of arginine incubation, however octopine was detected in 50% to 75% of the Ach5 tumors and only when incubated with 5 mM arginine. Neither nopaline or octopine were present in non transformed caUi after incubation with arginine (not shown). Axenic tumorous calli were screened for T-DNA integration using Southern blot analysis. DNA extraction from these caUi proved to be a limiting factor throughout the work. Calli yielded non digestable DNA with a number of extraction methods, including CTAB methods (Parsons et al., 1986; Dellaporta et al., 1983) nuclei isolation methods (FiUati et al., 1987) and maize DNA mini-prep (Malmberg et al., 1987). DNA digestable by BamHI and EcoRI was recovered from overnight CsC1 purification, but with a very low yield. Hence, 6-8 gm fresh weight was used to obtain sufficient quantities of DNA for each tumorous culture. The tmr and nos probes hybridized with tumorous DNA fragments the same size as in pTi, as shown for a tumor induced by strain C58 on clone BP-34 (Figure 2). Southern blot analysis showed T-DNA integration in DNA extracts for a total of 12 tumorous calli of BP-28, BP-34 or AG-8 (data not shown). No hybridization occured with DNA of non transformed calli each of
TABLE 2. Tumour induction 1 and opine detection.
Agrobacterium tumefaciens strain Plant clone
Controls Tumors2
AG-8 AI-1 AI-2 AI-41 AI-45 BP-15 BP-16 BP-28 BP-34
0/12 0/12 0/10 0/10 0/10 0/12 0/12 0[10 0/10
Ach5
C58
Tumors 2 Octopine3 21/224 9/12 11/12 11[12 12112 21/224 17/194 13/194 10/10
3/5 2/4 2/3 3/3 3/4 2]4 3/4 2]4 2/5
Tumors Nopaline3 3/224 0/10 3/12 8/12 5112 19/224 11/184 12/184 6/10
4/4 4/4 4/4 3/3 4/4 4/4 4/4 3]4
1.Whole shoots were preconditioned on rooting medium for 5-6 days, superficiallywounded on the stem and returned to rooting medium. 2.Nurnber of shoots with one tumour or more per total number of shoots wounded (controls) or wounded and inoculated. 3.Nurnber of tumours with opines (octopine or nopaline) detected per number screened. Octop'mewas detected only after incubation on medium containing 5 mM arginine. 4.Combined results of two separate assays, others are from a single sssay.
231 F I G U R E 1. a. Tumor induced by Agro-
bacterium tumefaciens (strain Ach5) on a shoot ofBetula papyrifera (clone BP-34), shown - 2 0 days after inoculation, b. Transformed cell culture established with tumors induced by C58 on shoots of A. glutinosa ; addition of 5 I.tM 2,4-D promoted rapid growth rate. These cells produced nopaline and DNA hybridized with T-DNA probes.
Betula (Figure 2) and Alnus (Figure 3). In some tumors, hybridization revealed a major band different than the fragment observed with Agrobacterium pTi (Figure 3). In BamHI digests, tmr probe hybridized with a 4.85 kb fragment of pTiAch5 corresponding to fragment BamHI 17a (Hille et al, 1984), however in tumors the probe hybridized with a 2.85 kb fragment (Figure 3). Furthermore, in half the tumors which hybridized with either tmr or no_..~sprobe, 1 to 4 additional minor bands, which were distinct from the Agrobacterium bands, were detected on autoradiographs (data not shown). Controls showed that bands not present in Agrobacterium were not due to different migration rates between plant and bacterial DNA (data not shown). The presence of bands specific to transformed cells thus suggests that different integration patterns or rearragements of T-DNA occured. With B. pap3%~era clone P.P-28, 13 shoots inoculated with Ach5 developed tumors, on 3 of these, morphogenic differentiation into leaves and shoots occurred in planta. Tumors bearing shoots were placed on multiplication medium (as described above) for 21 days and yielded several viable shoots. The shoots grew normally and bore normal leaves (Figure 4). Twenty individual shoots were collected and micropropagated with multiplication rates varying from 1:1 to 1:3 over a 4 week period (Figure 4a). Micropropagated shoots would not root on the standard agar rooting medium, as normal shoots, but produced abundant callus; however rooting was obtained after 15 days in a liquid version of the same medium (Figure 4b; Bertrand and Lalonde, 1985). Leaves from the above shoots were incubated with 5raM arginine (as described) and did not synthesize octopine at a detectible level. We concluded that these shoots had not regenerated from transformed ceils. However in callus cultures established from the initial tumor, octopine synthesis remained a constant feature,
F I G U R E 2. Southern blot analysis of DNA from a tumor induced on BP-34 by Agrobacterium tumefaciens, strain C-58. 50 ng of partially purified Ti plasmid of strain C-58 digested with EcoRI (Lanes A, A'), 5 Ixg of CsC1 purified DNA of non transformed B. papyrifera callus (Lanes B, B') and tumorous culture of BP-34, digested with EcoRI (Lanes C, C') and BamHI (Lanes D, D'), was hybridized with probes for tm_..~r(A, B, C, D) and nos (A', B', C', D') genes.
A 23
9.4
¸
66
4.2
B
C
D
A"
B'
C"
D'
and T-DNA integration was verified (as described in above paragraphs). Finally, morphogenic tumors continued to produce a few shoots for - 2 months, then lost this potential whether in presence of 2,4-D, BAP or no growth regulator. DISCUSSION The results presented here clearly demonstrate the infectivity of Agrobacterium tumefaciens and the capaci~ for gene transfer and expression of both nopaline and octopine strains in cells of Alnus and Betula, The only previous report on infection by Agrobacterium in these genera indicated tumorigenesis to be a rather slow process, taking 37 days in Betula pendula and 343 days in f* l a t',..... g o rubra kX.~,~, ~'-'~ '-" ,,..lea, ....l i e and De Ley, ,9;u). ' ~ ~" ,aur ^ results show that rapid tumor development follows infection of in vitro shoots; tumors were detected as early as 7 to 14 days after inoculation. Genetic transformation was demonstrated in tumorous tissues according to three criteria in addition to the fact that tumors were formed only on plants infected by Agrobacterium. (i)Tumors grew in absence of phyto-hormones, even though better growth was obtained by the addition of 2,4-D. (ii)Specific opines were detected in tumorous cells only, in accordance with respective opine encoded by T-DNA of the Agrobacterium strains. (iii)Hybridization of probes for the tm_...~rgene and nos gene with DNA of tumorous cultures which were exempt of Agrobacterium. Hence, our results show that sequences encoding for tmr and nos genes were integrated into plant genomes and, that T°DNA encoded opine synthetases and oncogenes were expressed in the plant cells. According to the Southern blot analyses, different patterns of T-DNA integration did occur, with one type dominating for each single tumorous culture. It is well known that inserted T-
F I G U R E 3. 50 ng of partially purified pTiAch5 (Lane A), 4-8 ~g of CsC1 purified DNA from non transformed A. glutinosa callus (Lane B) and from 4 tumorous cultures (Lanes C, D, E and F) hybridized with the tm__.Zprobe. The probe hybridizes with a 4.85 Kb fragment in pTiAch5 corresponding to fragment BamHI 17a (I-Iille et al., 1984), and with a 2.85 Kb fragment of tumorous plant DNA, not present with Ti plasmid.
A
B
C
D
E
F
232 FIGURE 4. Shoots differentiated from tumor induced by Ach5 on B. papyrifera, clone B-28. Shoots maintained normal growth and morphology and did not produce octopine. a.Multiplication (5 IIM BAP) was obtained at rates of 1:1 to 1:3 over 4 weeks, b.Rooting was achieved in liquid medium (1 IIM BAP) only.
DNA can be truncated or undergo rearrangements and modifications (Simpson et al., 1983; Peerbolte et al., 1987). In our work, such TDNA processing could be the source of a new major 2.85 kb fragment (Figure 2) and other minor bands to which tm..._lrprobe hybridized (Figure 3). In planta morphology and rapid growth rate of tumors suggest that many tumorous cultures originated from multiple infection. This is substantiated by the presence of several fragments homologous with the T-DNA probe. Given the possibility of multiple infection within a tumorous culture and limited information on exact genome sizes, it was not possible to determine the number of copies of integrated T-DNA per genome. Shoots proliferated in some tumors induced by strain Ach5 on clone BP-28: we suggest that these were primarily comprised of non transformed cells. According to our observations they do not fit the accepted definition of teratomous shoots (Turgeon, 1982) following three criteria: (i) An octopine utilising strain (Ach5) incited the original tumor. (ii) Stems and leaves had normal morphology and growth: identical to normal shoots. (iii) Shoots rooted without callus formation, but on liquid medium only. Futhermore they did not synthesize octopine. Nonetheless they were initiated from tumors, shown to contain transformed cells. Thus we conclude that the transformed cells present in morphogenic tumors had the effect of organizing untransformed ceils which differentiated into shoots (Turgeon, 1982). The fact that B. papyrifera are generally difficult to regenerate from calli (McCown, 1987) give this result specific interest. Agrobacterium tumefaciens "shooter" mutants (Steffen et al., 1986 ) and a trans Zeatin gene coupled to oncogenes (Fillati et al., 1987) have both been used to stimulate regeneration in transformants. The results presented here suggest that Agrobacterium genes encoding for plant growth regulators could be used to enhance regeneration ofB. papyrifera. In addition, potential for plant regeneration may be related to specific clones or cell lines since shoot differentiation from tumors occurred in only one out of four clones. We have shown that the wild type C58 and Ach5 Agrobacterium tumefaciens strains have the ability to introduce TDNA into Alnus and Betula genomes with expression of integrated genes. Hence Ti plasmid-based vectors can be used with these species. Regeneration, particularly in the case of Alnus, remains a major gap to bridge before production of transgenic plants becomes wide spread. Given the regeneration ofBetula papyrifera on tumors described here, we suggest that at least with this species, oncogenic vectors (Fillati et al., 1987) could be used to enhance the production of transgenie plants. Finally, the ability to transform cultures initiated from mature trees, e.g. clones AI-41, AI-45 (Ptrinet et al., 1988) as we have done provides the possibility to directly manipulate the genotype of subjects selected according to adult phenotype without the intermediate of controlled seed production.
ACKNOWLEDGEMENTS We are grateful to L. Simon for critical advice throughout the work, L.-J. Bertrand and M. DesLauriers for technical assistance and H. Kope for reviewing the manuscript. We also thank P. Ptrinet for Alnus incana clones. This work was made possible by a grant from the Minist~re de l'Energie et des
Ressources du Qutbec to M. L. J. M. received a graduate scholarship from the N.S.E.R.C. of Canada.
REFERENCES Bevan M 1984 Nucleic Acid Res. 12:8711-8721 Bevan M, Barnes WM, ChiltonM-D 1983 Nucleic Acid Res. 11: 369-385 Bertrand LJ, Lalonde M 1985 Plant Soil 87:143-152 Chatarpaul L, Carlisle A 1983 For. Chron. October: 249-259 Clapham DH, Ekberg I 1986 Scand. J. For. Res. 1:435-437 Currier TC, Nester EW 1976 Anal.Biochem. 76:431-441 Dandekar AM, Gupta PK, Durzan DJ, Knauf V 1987 Bin/Technology 5:587-590 De CleeneM, De Ley J 1976 The BotanicalReview 42 (4): 389-465 Dellaporta SL, Wood J, Hicks B 1983 Plant Mol. Biol. Rep. 1:19-21 Fillatti JJ, SellrnerJ, McCown JB, Haissig B, Comai L 1987 Mol. Gen. Genet. 206:192-201 GoodmanRM, Hauptli H, Crossway A, KnaufVC 1987 Science 236: 48-54 Hall RB, Maynard CA 1979 In: Gordon JC, Wheeler CT, Perry DA (eds) SymbioticNitrogenFixationin the Managementof Temporate Forests, Oregon State University, Corvallis Or., pp. 322-344. Herrera-EstrellaL, De Block M, ZambryskiP, Van MontaguM, Schell J 1985 In: Dedds JH (ed) Plant GeneticEngineering,Cambridge University Press, pp. 61-93 Hille J, Van Kan J, SchilperoortR 1984 In: Verma DPS, Holm T (eds) Genes Involved in Microbe-PlantInteractions,Plant Gene Research 1, Springer, Wien New York. pp. 287-309 Horseh RB, Fry JE, Hoffmann NL, Eichholtz E, Rogers SG, Fraley RT 1985 Science 227:1229-1231 MalmbergR, Messing J, Sussex I 1984 In: MolecularBiology of Plants - A Laboratory Course Manual, Cold Spring Harbor Laboratory, New York pp. 36-37 ManiatisT, Fritsch EF, SambrookJ 1982 MolecularCloning:A laboratory manual, Cold Spring Harbor Laboratory, New York, 545 pages McCown BH 1987 In: YPS Bajaj (ed) Biotechnologyin Agricultureand Forestry, Vol 4, Springer-VerlagBerlin Heidelberg(In press) Nagata T, Takebe I 1971 Planta 99:12-20 Nonnand P and Lalonde M. 1986 Plant Soil 90:429-453 Parsons TJ, Sinkar VP, Stettler RF, Nester EW, Gordon MP 1986 Bio/Technology 4 : 533-536 Peerbolte R, Ruigrok P, Wullems G, SchilperoortR 1987 Plant Mol. Biol. 9:51-57 Ptrinet P, Vallte G, Tremblay FM 1988 P1. Cell Tissue Organ Cult. (In press) SimpsonRB, Spielman A, MargossianL, McKnightTD 1986 Plant Mol. Biol. 6:403-415 Steffen A, Eriksson T, Schieder O 1986 Theor. Appl. Genet. 72: 135-140 Stomp A-M, Loopstra C, SederoffR, Chilton S, Fillati J, Dupper G, Tedeschi P, Kinlaw C 1987 In: Hanover J, Keathley D (eds) Genetic Manipulation of Woody Plants, Plenum Press, New York, (In press) Tremblay FM, Lalonde M 1984 P1. Ceil Tissue Organ Cult. 3: 189-199. TremblayFM, Ptrinet P, Lalonde M 1986 In: YPS Bajaj (ed) Bioteclmologyin Agricultureand Forestry, Vol 1: Trees, SpringerVerlag Berlin Heidelberg,pp. 87-100 Turgeon R 1982 In: Kahl G and Schell JS (eds) MolecularBiology of Plant Tumors, Academic Press, New York, pp. 391-414 Vahala T, Stabel P, Eriksson T 1987 In: Hanover J, Keathley D (eds) Genetic Manipulationof Woody Plants, Plenum Press, New York, NY (In press)
Plant Cell Reports (1988) 7:229-232
© Springer-Verlag 1988
Genetic transformation of 9 in vitro clones of Alnus and Betula by Agrobacterium tumefaciens John M a c k a y , Armand Srguin, and Maurice Lalonde Centre de Recherche en Biologie Foresti~re, Facult~ de Foresterie et de G6od6sie, Universit6 Laval, Sainte-Foy, Quebec, Canada G1K 7P4, Canada Received January 7, 1988/Revised version received March 30, 1988 - Communicated by F. Constabel
ABSTRACT Crown gall tumorigenesis, integration and expression of T-DNA encoded genes from Agrobacterium tumefaciens were investigated in 9 clones of Alnus glutinosa, A. incana and Betula papyrifera. Tumor formation on in vitro shoots was frequent in all clones with strain Ach5 and present in 8 clones with strain C58. Tumors excised from shoots were selected for autotrophic growth in vitro and axenic cultures were established. Octopine or nopaline, respective of the strain type used for inoculation, was detected in tumorous cultures. Southern blot analyses demonstrated T-DNA integration by hybridization of DNA from tumors with tmr and no._.~s gene probes. One clone of B. papyrifera produced tumors with a morphogenic character, unusual in calli of this species, generating viable shoots which did not synthesize opine. ABBREVIATIONS Cb, Carbenicillin; Cf, Cefotaxime; 2,4-D, 2,4-Dichloro-phenoxyacetic acid INTRODUCTION
Agrobacterium mediated gene transfer has been reported in a variety of agronomically-importantcrop plants (Goodman et al., 1987). Various Ti plasmid vectors and transformation procedures have been developed for use with plant model systems such as tobacco (Bevan, 1984) and tomato (Horsch et al., 1985). According to host range studies (De Cleene and De Ley, 1976), Agrobacterium can infect a wide spectrum of woody plants including angiosperms and some gymnosperms, however until recently there were few reports demonstrating transformation of forest trees. T-DNA integration into the plant genome was demonstrated with Populus (Parsons et al., 1986), Pinus (Stomp et al., 1987) previously excluded from Agrobacterium's host range (De Cleene and De Ley, 1976) and Pseudotsuga (Dandekar et al., 1987). Pathogenicity has been shown in Salix (Vahala et al., 1987), Abies and Picea (Clapham and Ekberg, 1986) and a few others. The genus Alnus (alder) is comprised of woody plants that can fix atmospheric nitrogen through a symbiotic association with the actinomycete Franla'a. Nitrogen fixing woody plants are of interest for intensive forest production as well as land reclamation (Chatarpanl and Carlisle, 1983; Hall and Maynard 1979). As of yet, little is known about the plant symbiotic genes and their regulation. The development of vector systems for gene transfer and expression in plant cells could aid in determining the regulation and specific functions of the genes involved (Herrera-Estrellaet al., 1985). Transformation vector systems for plants could lead to transfer of plant symbiotic genes to non nitrogen fixing species, as well. Betula (birch) which is closely related to Alnus is a candidate for such gene transfers (Normand and Laionde, 1986). We report here on crown gall tumorigenesis and the genetic transformation of 9 in vitro propagated clones of the genera Alnus and Betula by two wild type strains of Agrobacterium Offprint requests to." M. Lalonde
tumefaciens. Transformation is demonstrated by phyto-hormone independant growth, opine production and T-DNA integration. We conclude that Agrobacterium mediated gene transfer with engineered Ti plasmids may be applied in these species, such as in Populus (Fillati et al., 1987). M A T E R I A L S AND METHODS
Plant Material and Bacterial Strains Nine plant clones, 1 of Alnus glutinosa (L.) Gaertn., 4 ofAlnus incana (L.) Moench. and 4 ofBetula papyrifera Marsh. (Table 1) were used to test Agrobacterium infectivity. These clones are from varied geographic origins and two (AI-41, AI-45) were initiated from mature individuals ofA. incana, 22 and 20 years old (Prrinet et al., 1988). All in vitro shoots were obtained through micropropagation as described by Tremblay et al. (1986). The multiplication medium used in our assays consisted of MS salts, with 3% sucrose or 1.5% glucose for A. incana, 5 ktM BAP (6Benzyl-anainopurine) and was solidified with 0.7% to 0.8% agar (Difco Laboratories). In the roofing medium, macro-nutrients were reduced to half and the growth regulator was 1 pM IBA (Indole-3butyric acid). Two wild type strains of Agrobacterium tumefaciens (kindly provided by Dr P. Dion, Universit6 Laval, Qu6bec) were used: the octopine-type Ach5 and nopaline-type C58. The bacteria were grown at 29°C for 2 days on LB plates.
Inoculation with Agrobacterium A general method using whole shoots was devised to assay Agrobacterium tumefaciens infectivity on the clones listed above (Table 1). Different growth media and inoculation methods were tested to determine optimal conditions. Whole shoots 4-5 cm tall were preconditioned for 5-6 days either on multiplication or rooting medium and returned to preconditioning medium after inoculation (no antibiotics). For each inoculation, bacteria were introduced onto a sharp blade by directly collecting a few colonies of a freshly grown plate. The shoots were then inoculated either by decapitation or superficially wounding the stem. The controls received identical treatments but without the introduction of bacteria.
Culture of Tumors and Differentiated Shoots Tumors were excised from shoots 30-40 days after inoculation with Agrobacterium and grown on NT (Nagata and Takebe, 1971) medium supplemented with 500 mg L -1 CB and 250 mg L -1 Cf. 2,4-D was later used with this medium at concenlxations of 2.5 and 5 gM, to increase the growth rate. To eliminate all bacteria, cultured tumors larger than - 7 mm in diameter were surface sterilised in a 1% solution of NaOC1 (commercial Javex) for 5-10 minutes, surface cell layers were removed and the inner core was plated on NT medium with 2,4-D, without antibiotics. Smaller tumors were simply maintained on media with antibiotics. Shoots
230 TABLE 1. Description and origin of Alnus and Betula clones. Species Alnus glutinosa Alnus incana Betula
papyrifera
Clone acronyme AG-8 AI- 1, AI-2 AI-41, AI-45 BP-15, BP-16
Reference Tremblay and Lalonde, 1984 Idem P6rinet et al., 1988 Thispaper 1
BP-28, BP-34
1. Betulapapyrifera cultures were initiated from seed (Berthier- ville, QuEbec). Seeds were sterilized with 30% peroxide for 7 minutes followed by 1% solution of NaOCI (commercial Javex) also for 7 minutes, thoroughly rinsed in distilled water and germinated in vitro on water solidified with 1% agar, containing 3% (w/v) sucrose.
from morphogenic tumors were collected and propagated individually on the multiplication medium supplemented with 100 mg L -1 Cb and 50 mg L -1 Cf. The shoots were placed on standard rooting medium, with antibiotics, in solid (0.7% agar) and liquid (no agar) forms, for root induction. All tumorous calli and shoots differentiated from tumors were sub-cultured at 3-4 week intervals. Opine Detection Crude ethanol extracts were electrophoresed on paper and visualised with phenanthrenequinone as described by Parsons et al. (1986) with the following modifications. A low pH buffer was used (formic acid: acetic acid: water, 3:6:91), 20 laL of each extract was spotted onto 3MM chromatographic paper (Whatman®) and electrophoresed at 800V for 1 hour. The effect of incubating tumorous calli on 5 mM arginine for five days prior to extraction was tested. Opines were determined by comigration with octopine and nopaline standards, arginine served as a standard for negative discrimination. DNA Extraction and Southern Blot Analysis Agrobacterium tumefaciens Ti-plasmid was obtained by a partial purification (Currier and Nester, 1976). DNA was extracted from tumorous and non transformed calli following the procedure described by Malmberg et aI. (1984). In order to yield highly digestable DNA, further purification on a CsC1 gradient was performed. The DNA was suspended in 4.0 ml of 10 mM Tris pH 8.0, 1 mM EDTA, 4.0 g of CsC1 were added to the solution and the DNA banded by equilibrium density centrifugation (Maniatis et aL, 1982). DNA was digested with restriction enzymes under conditions specified by the supplier (BRL). Digested DNA was subsequently separated by gel electrophoresis in 0,8% agarose and blotted onto Gene Screen Plus membranes (New England Nuclear) following the protocols recommended by the manufacturer. 32p_ labelled probes: pRA123 containing a eDNA clone of the tmr gene from pTiA6 (provided by Dr E. W. Nester, University of Washington, Seafle) and a cloned 2.3 kb fragment including the nos gene of pTiT37 (Bevan et al., 1983; provided by Dr B. Miki, Agriculture Canada, Ottawa) were produced by nick translation to 5-10x108 cpmAtg as described by Maniatis et al. (1982). The probes were hybridized to the membrane-bound genomic DNA (following membrane manufacturer's protocol) and washed in stringent conditions (0.2X SSC, 65°C) as described in Maniatis et al. (1982), and visualized by exposure to X-ray film (Kodak XAR) for 24 hours with an intensifying screen (Lightening Plus, DuPont) at -70°C. RESULTS A high incidence of tumorigenesis was observed in shoots superficially wounded on the stem, but only when these were preconditioned on and returned to rooting medium, containing IBA. This treatment incited tumors in all clones tested (Figure la,
Table 2). Macroscopic tumorous growth was detected as early as 7 to 9 days after inoculation in B. papyrifera and I1 to 13 days in Alnus clones. Multiplication medium containing BAP completely inhibited tumorigenesis and wounding of the shoot apex was also unfavorable to tumor development (data not shown). No tumors or calli were observed on control plants (Table 2). Morphology of tumorous outgrowth as observed early on resulted from multiple transformation events at a given wounding site (data not shown). Growth was more rapid in tumors of B. papyrifera (Figure la), which were observed to reach 10 mm in diameter after 30-40 days. With the clones of Alnus and Betula tested, strain Ach5 was generally more effective in causing tumorigenesis (Table 2). Excised tumors were cultured as caUi (Figure lb). Large tumors were surface sterilized and dissected: four tumorous bacteria-free cultures of B. papyrifera and A. glutinosa were thus obtained. In addition, several tumors yielded axenic cultures without surface sterilization, after 3-4 subcultures on antibiotics. In absence of growth regulators and with antibiotics (500 mg L -1 Cb and 250 mg L -1 Cf), growth was slow but sustained and most tumors remained compact. Cell proliferation was sufficient for subdivision of cultures at 4 week intervals. Phyto-hormone independent cultures were selected after 7 months. After selection, the growth rate was doubled by the addition of 5 gM 2,4-D. Opines were detected in a number of tumors randomly selected among those which grew in absence of phyto-hormones (Table 2). Tumors induced by C58 all synthesized nopaline regardless of arginine incubation, however octopine was detected in 50% to 75% of the Ach5 tumors and only when incubated with 5 mM arginine. Neither nopaline or octopine were present in non transformed caUi after incubation with arginine (not shown). Axenic tumorous calli were screened for T-DNA integration using Southern blot analysis. DNA extraction from these caUi proved to be a limiting factor throughout the work. Calli yielded non digestable DNA with a number of extraction methods, including CTAB methods (Parsons et al., 1986; Dellaporta et al., 1983) nuclei isolation methods (FiUati et al., 1987) and maize DNA mini-prep (Malmberg et al., 1987). DNA digestable by BamHI and EcoRI was recovered from overnight CsC1 purification, but with a very low yield. Hence, 6-8 gm fresh weight was used to obtain sufficient quantities of DNA for each tumorous culture. The tmr and nos probes hybridized with tumorous DNA fragments the same size as in pTi, as shown for a tumor induced by strain C58 on clone BP-34 (Figure 2). Southern blot analysis showed T-DNA integration in DNA extracts for a total of 12 tumorous calli of BP-28, BP-34 or AG-8 (data not shown). No hybridization occured with DNA of non transformed calli each of
TABLE 2. Tumour induction 1 and opine detection.
Agrobacterium tumefaciens strain Plant clone
Controls Tumors2
AG-8 AI-1 AI-2 AI-41 AI-45 BP-15 BP-16 BP-28 BP-34
0/12 0/12 0/10 0/10 0/10 0/12 0/12 0[10 0/10
Ach5
C58
Tumors 2 Octopine3 21/224 9/12 11/12 11[12 12112 21/224 17/194 13/194 10/10
3/5 2/4 2/3 3/3 3/4 2]4 3/4 2]4 2/5
Tumors Nopaline3 3/224 0/10 3/12 8/12 5112 19/224 11/184 12/184 6/10
4/4 4/4 4/4 3/3 4/4 4/4 4/4 3]4
1.Whole shoots were preconditioned on rooting medium for 5-6 days, superficiallywounded on the stem and returned to rooting medium. 2.Nurnber of shoots with one tumour or more per total number of shoots wounded (controls) or wounded and inoculated. 3.Nurnber of tumours with opines (octopine or nopaline) detected per number screened. Octop'mewas detected only after incubation on medium containing 5 mM arginine. 4.Combined results of two separate assays, others are from a single sssay.
231 F I G U R E 1. a. Tumor induced by Agro-
bacterium tumefaciens (strain Ach5) on a shoot ofBetula papyrifera (clone BP-34), shown - 2 0 days after inoculation, b. Transformed cell culture established with tumors induced by C58 on shoots of A. glutinosa ; addition of 5 I.tM 2,4-D promoted rapid growth rate. These cells produced nopaline and DNA hybridized with T-DNA probes.
Betula (Figure 2) and Alnus (Figure 3). In some tumors, hybridization revealed a major band different than the fragment observed with Agrobacterium pTi (Figure 3). In BamHI digests, tmr probe hybridized with a 4.85 kb fragment of pTiAch5 corresponding to fragment BamHI 17a (Hille et al, 1984), however in tumors the probe hybridized with a 2.85 kb fragment (Figure 3). Furthermore, in half the tumors which hybridized with either tmr or no_..~sprobe, 1 to 4 additional minor bands, which were distinct from the Agrobacterium bands, were detected on autoradiographs (data not shown). Controls showed that bands not present in Agrobacterium were not due to different migration rates between plant and bacterial DNA (data not shown). The presence of bands specific to transformed cells thus suggests that different integration patterns or rearragements of T-DNA occured. With B. pap3%~era clone P.P-28, 13 shoots inoculated with Ach5 developed tumors, on 3 of these, morphogenic differentiation into leaves and shoots occurred in planta. Tumors bearing shoots were placed on multiplication medium (as described above) for 21 days and yielded several viable shoots. The shoots grew normally and bore normal leaves (Figure 4). Twenty individual shoots were collected and micropropagated with multiplication rates varying from 1:1 to 1:3 over a 4 week period (Figure 4a). Micropropagated shoots would not root on the standard agar rooting medium, as normal shoots, but produced abundant callus; however rooting was obtained after 15 days in a liquid version of the same medium (Figure 4b; Bertrand and Lalonde, 1985). Leaves from the above shoots were incubated with 5raM arginine (as described) and did not synthesize octopine at a detectible level. We concluded that these shoots had not regenerated from transformed ceils. However in callus cultures established from the initial tumor, octopine synthesis remained a constant feature,
F I G U R E 2. Southern blot analysis of DNA from a tumor induced on BP-34 by Agrobacterium tumefaciens, strain C-58. 50 ng of partially purified Ti plasmid of strain C-58 digested with EcoRI (Lanes A, A'), 5 Ixg of CsC1 purified DNA of non transformed B. papyrifera callus (Lanes B, B') and tumorous culture of BP-34, digested with EcoRI (Lanes C, C') and BamHI (Lanes D, D'), was hybridized with probes for tm_..~r(A, B, C, D) and nos (A', B', C', D') genes.
A 23
9.4
¸
66
4.2
B
C
D
A"
B'
C"
D'
and T-DNA integration was verified (as described in above paragraphs). Finally, morphogenic tumors continued to produce a few shoots for - 2 months, then lost this potential whether in presence of 2,4-D, BAP or no growth regulator. DISCUSSION The results presented here clearly demonstrate the infectivity of Agrobacterium tumefaciens and the capaci~ for gene transfer and expression of both nopaline and octopine strains in cells of Alnus and Betula, The only previous report on infection by Agrobacterium in these genera indicated tumorigenesis to be a rather slow process, taking 37 days in Betula pendula and 343 days in f* l a t',..... g o rubra kX.~,~, ~'-'~ '-" ,,..lea, ....l i e and De Ley, ,9;u). ' ~ ~" ,aur ^ results show that rapid tumor development follows infection of in vitro shoots; tumors were detected as early as 7 to 14 days after inoculation. Genetic transformation was demonstrated in tumorous tissues according to three criteria in addition to the fact that tumors were formed only on plants infected by Agrobacterium. (i)Tumors grew in absence of phyto-hormones, even though better growth was obtained by the addition of 2,4-D. (ii)Specific opines were detected in tumorous cells only, in accordance with respective opine encoded by T-DNA of the Agrobacterium strains. (iii)Hybridization of probes for the tm_...~rgene and nos gene with DNA of tumorous cultures which were exempt of Agrobacterium. Hence, our results show that sequences encoding for tmr and nos genes were integrated into plant genomes and, that T°DNA encoded opine synthetases and oncogenes were expressed in the plant cells. According to the Southern blot analyses, different patterns of T-DNA integration did occur, with one type dominating for each single tumorous culture. It is well known that inserted T-
F I G U R E 3. 50 ng of partially purified pTiAch5 (Lane A), 4-8 ~g of CsC1 purified DNA from non transformed A. glutinosa callus (Lane B) and from 4 tumorous cultures (Lanes C, D, E and F) hybridized with the tm__.Zprobe. The probe hybridizes with a 4.85 Kb fragment in pTiAch5 corresponding to fragment BamHI 17a (I-Iille et al., 1984), and with a 2.85 Kb fragment of tumorous plant DNA, not present with Ti plasmid.
A
B
C
D
E
F
232 FIGURE 4. Shoots differentiated from tumor induced by Ach5 on B. papyrifera, clone B-28. Shoots maintained normal growth and morphology and did not produce octopine. a.Multiplication (5 IIM BAP) was obtained at rates of 1:1 to 1:3 over 4 weeks, b.Rooting was achieved in liquid medium (1 IIM BAP) only.
DNA can be truncated or undergo rearrangements and modifications (Simpson et al., 1983; Peerbolte et al., 1987). In our work, such TDNA processing could be the source of a new major 2.85 kb fragment (Figure 2) and other minor bands to which tm..._lrprobe hybridized (Figure 3). In planta morphology and rapid growth rate of tumors suggest that many tumorous cultures originated from multiple infection. This is substantiated by the presence of several fragments homologous with the T-DNA probe. Given the possibility of multiple infection within a tumorous culture and limited information on exact genome sizes, it was not possible to determine the number of copies of integrated T-DNA per genome. Shoots proliferated in some tumors induced by strain Ach5 on clone BP-28: we suggest that these were primarily comprised of non transformed cells. According to our observations they do not fit the accepted definition of teratomous shoots (Turgeon, 1982) following three criteria: (i) An octopine utilising strain (Ach5) incited the original tumor. (ii) Stems and leaves had normal morphology and growth: identical to normal shoots. (iii) Shoots rooted without callus formation, but on liquid medium only. Futhermore they did not synthesize octopine. Nonetheless they were initiated from tumors, shown to contain transformed cells. Thus we conclude that the transformed cells present in morphogenic tumors had the effect of organizing untransformed ceils which differentiated into shoots (Turgeon, 1982). The fact that B. papyrifera are generally difficult to regenerate from calli (McCown, 1987) give this result specific interest. Agrobacterium tumefaciens "shooter" mutants (Steffen et al., 1986 ) and a trans Zeatin gene coupled to oncogenes (Fillati et al., 1987) have both been used to stimulate regeneration in transformants. The results presented here suggest that Agrobacterium genes encoding for plant growth regulators could be used to enhance regeneration ofB. papyrifera. In addition, potential for plant regeneration may be related to specific clones or cell lines since shoot differentiation from tumors occurred in only one out of four clones. We have shown that the wild type C58 and Ach5 Agrobacterium tumefaciens strains have the ability to introduce TDNA into Alnus and Betula genomes with expression of integrated genes. Hence Ti plasmid-based vectors can be used with these species. Regeneration, particularly in the case of Alnus, remains a major gap to bridge before production of transgenic plants becomes wide spread. Given the regeneration ofBetula papyrifera on tumors described here, we suggest that at least with this species, oncogenic vectors (Fillati et al., 1987) could be used to enhance the production of transgenie plants. Finally, the ability to transform cultures initiated from mature trees, e.g. clones AI-41, AI-45 (Ptrinet et al., 1988) as we have done provides the possibility to directly manipulate the genotype of subjects selected according to adult phenotype without the intermediate of controlled seed production.
ACKNOWLEDGEMENTS We are grateful to L. Simon for critical advice throughout the work, L.-J. Bertrand and M. DesLauriers for technical assistance and H. Kope for reviewing the manuscript. We also thank P. Ptrinet for Alnus incana clones. This work was made possible by a grant from the Minist~re de l'Energie et des
Ressources du Qutbec to M. L. J. M. received a graduate scholarship from the N.S.E.R.C. of Canada.
REFERENCES Bevan M 1984 Nucleic Acid Res. 12:8711-8721 Bevan M, Barnes WM, ChiltonM-D 1983 Nucleic Acid Res. 11: 369-385 Bertrand LJ, Lalonde M 1985 Plant Soil 87:143-152 Chatarpaul L, Carlisle A 1983 For. Chron. October: 249-259 Clapham DH, Ekberg I 1986 Scand. J. For. Res. 1:435-437 Currier TC, Nester EW 1976 Anal.Biochem. 76:431-441 Dandekar AM, Gupta PK, Durzan DJ, Knauf V 1987 Bin/Technology 5:587-590 De CleeneM, De Ley J 1976 The BotanicalReview 42 (4): 389-465 Dellaporta SL, Wood J, Hicks B 1983 Plant Mol. Biol. Rep. 1:19-21 Fillatti JJ, SellrnerJ, McCown JB, Haissig B, Comai L 1987 Mol. Gen. Genet. 206:192-201 GoodmanRM, Hauptli H, Crossway A, KnaufVC 1987 Science 236: 48-54 Hall RB, Maynard CA 1979 In: Gordon JC, Wheeler CT, Perry DA (eds) SymbioticNitrogenFixationin the Managementof Temporate Forests, Oregon State University, Corvallis Or., pp. 322-344. Herrera-EstrellaL, De Block M, ZambryskiP, Van MontaguM, Schell J 1985 In: Dedds JH (ed) Plant GeneticEngineering,Cambridge University Press, pp. 61-93 Hille J, Van Kan J, SchilperoortR 1984 In: Verma DPS, Holm T (eds) Genes Involved in Microbe-PlantInteractions,Plant Gene Research 1, Springer, Wien New York. pp. 287-309 Horseh RB, Fry JE, Hoffmann NL, Eichholtz E, Rogers SG, Fraley RT 1985 Science 227:1229-1231 MalmbergR, Messing J, Sussex I 1984 In: MolecularBiology of Plants - A Laboratory Course Manual, Cold Spring Harbor Laboratory, New York pp. 36-37 ManiatisT, Fritsch EF, SambrookJ 1982 MolecularCloning:A laboratory manual, Cold Spring Harbor Laboratory, New York, 545 pages McCown BH 1987 In: YPS Bajaj (ed) Biotechnologyin Agricultureand Forestry, Vol 4, Springer-VerlagBerlin Heidelberg(In press) Nagata T, Takebe I 1971 Planta 99:12-20 Nonnand P and Lalonde M. 1986 Plant Soil 90:429-453 Parsons TJ, Sinkar VP, Stettler RF, Nester EW, Gordon MP 1986 Bio/Technology 4 : 533-536 Peerbolte R, Ruigrok P, Wullems G, SchilperoortR 1987 Plant Mol. Biol. 9:51-57 Ptrinet P, Vallte G, Tremblay FM 1988 P1. Cell Tissue Organ Cult. (In press) SimpsonRB, Spielman A, MargossianL, McKnightTD 1986 Plant Mol. Biol. 6:403-415 Steffen A, Eriksson T, Schieder O 1986 Theor. Appl. Genet. 72: 135-140 Stomp A-M, Loopstra C, SederoffR, Chilton S, Fillati J, Dupper G, Tedeschi P, Kinlaw C 1987 In: Hanover J, Keathley D (eds) Genetic Manipulation of Woody Plants, Plenum Press, New York, (In press) Tremblay FM, Lalonde M 1984 P1. Ceil Tissue Organ Cult. 3: 189-199. TremblayFM, Ptrinet P, Lalonde M 1986 In: YPS Bajaj (ed) Bioteclmologyin Agricultureand Forestry, Vol 1: Trees, SpringerVerlag Berlin Heidelberg,pp. 87-100 Turgeon R 1982 In: Kahl G and Schell JS (eds) MolecularBiology of Plant Tumors, Academic Press, New York, pp. 391-414 Vahala T, Stabel P, Eriksson T 1987 In: Hanover J, Keathley D (eds) Genetic Manipulationof Woody Plants, Plenum Press, New York, NY (In press)
