Planta

Planta (1987) 171:393-405

9 Springer-Verlag 1987

Structure, organization and expression of transferred DNA in Nicotiana plumbaginifolia crown-gall tissues R. Peerbolte*, W. te Lintel Hekkert, D.G. Barfield, J.H.C. Hoge, G.J. Wullems and R.A. Schilperoort Department of Plant Molecular Biology, Leiden University, Wassenaarseweg 64, NL-2333 AL Leiden, The Netherlands

Abstract. Data are provided which show that

transferred D N A (T-DNA) present in Nicotiana plumbaginifolia crown-gall lines in most cases was scrambled and not intact. Both wild-type, and 'rooter'- and 'shooter'-type mutants of octopinetype Agrobacterium tumefaciens were used to infect N. plumbaginifolia plantlets, cultured in vitro. Resulting tumors were excised from the plantlets and cultured for more than three years. During subculturing the tumor lines were scored for the following phenotypic traits: phytohormone autonomous growth in vitro (Aut+), spontaneous shoot regeneration (Reg+), root deficiency of shoots (Rod+), octopine production (Ocs +) and mannopine and agropine production (Mas +Ags +). An unexpectedly large variety of phenotypes was observed. For instance, two out of three tumor lines induced on haploid plantlets by the rooter mutant LBA4210 regenerated shoots, a phenomenon which is not observed for octopine tobacco tumor lines. Fifty percent of the crown-gall lines studied did not contain octopine. Only one line out of six independent lines analyzed was found to have a 'regular' TD N A structure. Occurrence of aberrant T - D N A structures was not correlated with the ploidy level of infected plantlets, nor with the T-region structure of the inciting bacterial strain. The pattern of T L - D N A transcripts was studied for one line and correlated well with the aberrant T - D N A * To whom correspondence should be addressed Abbreviations: Ags +/- = tumor line with ( +)/without ( - ) agropine production; Aut +/- = t u m o r line with (+)/without ( - ) phytohormone autonomous growth; Aux+/Cyt + = t u m o r cells produced by active auxin/cytokinin genes from octopine TLD N A ; Mas +/- = tumor line with ( + )/without (--) mannopine production; Ocs + / - = tumor line with (+)/without ( - ) octopine production; R e g + / - = t u m o r line with (+)/without ( - ) spontaneous shoot regeneration; R o d + / - = t u m o r line with (+)/without (--) root deficiency of shoots; T - D N A = transferred D N A

structure detected. Segments of T R - D N A , having irregular structures as well, were detected in two out of the six lines studied. The scrambled nature of the T R - D N A explained the absence of mannopine and agropine in these two lines. In addition, it was observed that N. plumbaginifolia tissue lines which did not carry T-DNA, became readily phytohormone autotrophic (habituated) at an early stage in tissue culture. Key words: Agrobacterium - Crown-gall - DNA, transferred - Nicotiana (T-DNA) - T - D N A structure.

Introduction

The soil bacterium Agrobaeterium tumefaciens incites tumors, named crown galls, on many dicotyledonous plants. These tumors are the result of the action of 'onc-genes', present on a distinct piece of DNA, called transferred D N A (T-DNA), introduced into the plant cells by the bacterium. The T - D N A is integrated into the plant's nuclear D N A and expressed (Chilton et al. 1977). The one-genes code for enzymes which catalyze the synthesis of phytohormones. Genes 1 and 2, together comprising the auxin locus, are responsible for the production of indole acetic acid in tumor cells (Ooms et al. 1981 ; Inz6 et al. 1984; Schr6der et al. 1984; Thomashow et al. 1984). The product of gene 4, the cytokinin locus, catalyzes the biosynthesis of cytokinins (Ooms et al. 1981; Barry et al. 1984; Akiyoshi et al. 1984; Buchmann et al. 1985). As a consequence of one-gene expression a severe disturbance of the natural phytohormone balance at the site of infection occurs (Ooms et al. 1981 ; Akiyoshi et al. 1983), resulting in proliferation of both transformed and non-transformed cells.

394

R. Peerbolte et al. : T - D N A in Nicotiana crown-gall tissue

Table 1. Agrobaeterium strains used to induce N. plumbaginifolia crown-gall lines Strain

Plasmid

Comments

Reference

LBA4001 LBA4013 LBA4060 LBA4210 LBA1833

pTiAch5 pALl02 pALl08 pAL228 pALl833

wild-type Ach5 strain pTiAch5 with wild-type T-region gene 2 - (IS60 insertion; Aux Cyt +) gene 4 - (Tn904 insertion; Aux + C y t - ) Tk(HSV)-NPTII(Tn5) (Aux + Cyt+)"

Klapwijk et Klapwijk et Klapwijk et Klapwijk et This paper

al. al. al. al.

(1977) (1980) (1978) (1980)

" The EcoRI-19a and small omega fragments in a clone containing pTiAch5 fragment HindIII-1 were exchanged with an EcoRI fragment derived from pAG60 (Barclay and Meller 1983) containing the structural gene for kanamycin resistance from transposon Tn5, linked to the Tk promoter from Herpes Simplex Virus. The resulting plasmid was introduced via a single crossover into pAL969, a stable cointegrate of pTiAch5 and the IncP plasmid R772 (Hille et al. 1983)

Besides one-genes, other genes located on the T - D N A segment are expressed in plant cells. Some of these code for the production of opines, such as octopine. Opines are unusual amino-acid derivatives not found in normal plant cells. Octopinetype bacteria are able to introduce into plant cells two distinct pieces of T-DNA, namely T L - D N A and T R - D N A . The T L - D N A harbors, among others, the onc-genes and the octopine-synthase gene (Ocs+); T R - D N A , which is not necessarily present in tumor cells, carries genes which code for the production of two other opines, namely mannopine and agropine (Mas § Ags § ; see Fig. 1). The morphology of a tumor on a certain plant species depends on the expression of the onc-genes present in that tumor. On tobacco stems, unorganized tumors develop well when both auxin and cytokinin genes from octopine T L - D N A are active (Aux+Cyt+). An Aux+Cyt - octopine-type tobacco tumor develops slowly and produces more adventituous roots than normally found. An A u x Cyt § tobacco tumor regenerates shoots (Reg § while tumor development is retarded (Ooms et al. 1981; Garfinkel etal. 1981). The onc-genes of T - D N A render tobacco tissue independent of exogenous growth substances and thus allow rapid growth of crown-gall tissue on medium without auxins and cytokinins (phytohormone autotrophy; Aut § In addition, non-transformed cells, if present in a tumor line, will grow on medium lacking phytohormones as a result of cross-feeding by transformed cells. If T - D N A in transformed cells does not contain any active onc-gene, the phenotype of the tissue is A u t - . So far, studies on crown-gall tumor lines have been focussed primarily on Nicotiana tabacum tissues. In addition to the expression of T - D N A oncgenes, tumor morphology can also depend on the plant species involved. For example, an A. tumefaciens mutant with an inactive auxin locus induces attenuated shooty tumors on tobacco, but is hardly

oncogenic on tomato and induces normal 'wildtype' tumors on N. rustica (Ooms et al. 1981). Differences in host response demonstrate an important role of the physiological and genetic background of the host. The possibility cannot be excluded, however, that these differences can also be the result of variation in T - D N A expression or even of variation in T - D N A structure in different plant species. These considerations led us to start studies on crown-gall tissues of a species other than tobacco. The plant species we chose to study was Nicotiana plumbaginifolia since tissue-culture conditions for this species are well known (Bourgin et al. 1979; Barfield et al. 1985). Tumors were induced on haploid and di-haploid N. plumbaginifolia plantlets with wild-type and mutant A. tumefaciens strains. The phenotypes of the tumor lines were studied during subculturing, for more than three years. In order to understand their phenotypes in relation to T - D N A structure, organization and expression, the T - D N A structures of several lines were studied as well as the pattern of T - D N A transcripts for one selected line.

Materials and methods Bacterial strains. Bacterial strains used to induce the crown-gall lines studied are listed in Table 1. Tissues. Tissues obtained from tumors induced on di-haploid plantlets are denoted " N P " , whereas tissues derived from haploid plantlets are denoted " N P h " . Differentiation forms are indicated with prefixes " c " for callus and " s " for shoots or shoot-like structures. The N. plumbaginifolia tissue lines studied, were classified into four groups. Group I contained the non-transformed line NPO, obtained from a di-haploid N. plumbaginifolia leaf explant, which had been placed on LS medium (Linsmaier and Skoog 1965) containing 0.2rag.1-1 kinetin and 2 mg-1 1 c~-naphthaleneacetic acid (NAA). Group II was induced by A. tumefaciens strains carrying intact copies of all onc-genes (Aux + Cyt+). It comprised NPI, NP2 (induced by LBA4001) and their respective cellular clones, as well as the lines NP9, NP10 and N P h l l (induced by LBA4013) and

R. Peerbolte et al. : T - D N A in Nicotiana crown-gall tissue

395

Table 2. D N A clones used as probes for Northern and Southern blots Plasmid

Vector

Cloned pTi fragment

Indicated as

pRAL3911 pRAL3252 pRAL3076 pRAL3215 pOTY8 pFK1

pBR313 pBR322 pBR322 pACYC184 pJDB207 M13

BamHI-17a BamHI-8 BamHI-28, 30a, 17a, 2, 20a EcoRI-12 BamHI-8/HindIII-I Tn5-Km

BamHI-a7a BamHI-8 ~ pRAL3076 / EcoRI-12 pOTY8 MI3-Km"

References

TI--Tr

Hoekema et al. (1984) Oomset al. (1981) Ooms et al. (1981) Peerbolte et al. (1986b) Hirsch and Beggs (1984) Krens (unpubl.)

a An EcoRI--HindIII fragment containing the Tn5 kanamycin-resistance gene, derived from Neo-delta-18 (Bevan et al. 1983), cloned into phage M13

NP16 (induced by LBAI833). Group III was induced by LBA4210, an Aux + Cyt- 'rooter mutant' strain, and contained the tissue lines NPh3, NPh4, NPh5 and NP6. Group-IV tissues (NP7 and NPh8) were induced by the A u x - Cyt + 'shooter mutant' LBA4060.

Culture conditions of N. plumbaginifolia tissues. The Aut + callus tissues and shoots were grown on LS medium lacking phytohormones whereas A u t - callus lines were grown on MS medium (Murashige and Skoog 1962) containing 2 mg.1-1 (NAA) and 0.2 mg.1-1 kinetin. The tissues were grown at 26~ in a 12-h photoperiod (2000-4000 lx) and subcultured monthly. Protoplast isolation was performed as described by Wullems et al.

(1980). Opine assays. Octopine-synthase activity was assayed as described by Otten and Schilperoort (1978). Mannopine and agropine were monitored by means of paper electrophoresis and silver-nitrate staining as described by Leemans et al. (1981) and Peerbolte et al. (1985). Isolation ofDNA andRNA. The D N A was isolated via a procedure described by Chilton et al. (1982), slightly modified according to Memelink et al. (1983). Frozen tissue was extracted twice with phenol and the D N A was precipitated with a 0.75 volume of isopropanol, redissolved and subsequently purified on a CsCl/ethidium bromide gradient. After extraction of ethidinm bromide with isoamylalcohol, the D N A solution was dialysed, and D N A was precipitated and redissolved in a 2-amino-2-(hydroxymethyl)-l,3-propanediol (Tris; 10 mM)/ethylenediaminetetraacetic acid (EDTA, I mM) buffer. Isolation of R N A was as described by Van Slogteren et al. (1983). It was extracted with a phenol/LiC1 mixture, precipitated with LiC1, dissolved in twice-distilled water. Polyadenylated R N A was obtained by means of chromatography on an oligo(dT)-cellulose column.

(1975) and Van Slogteren et al. (1983), using GeneScreen (New England Nuclear) to immobilize glyoxylated R N A which had been separated by electrophoresis on 1.5% agarose gels. The Northern blots were hybridized as described by Van Slogteren et al. (1983) and Northern autoradiograms were exposed to Fuji (Tokyo, Japan) RX X-ray films. Before rehybridization of blots, hybridized probes were removed by incubation with a mixture containing 50% formamide at 68 ~ C. The D N A clones used as probes in the hybridization experiments are listed in Table 2. The position of hybridizing T - D N A restriction fragments are indicated in Fig. 1 B and C. The entire D N A clones were used as probes since it had been established that the vector parts of the clones did not hybridize to N. plumbaginifolia D N A nor to pTi D N A at the conditions applied. In addition to D N A from transformed lines, Southern blots contained 'reconstruction lanes' of pTi D N A mixed with nontransformed N. plumbaginifolia D N A in a ratio of 1, 2 or 3 copies T-region per N. plumbaginifolia genome. A probe which covers the entire T-region reveals bands in the reconstruction lanes representing "internal fragments" (for the TLregion e.g. BamHI-17a, EcoRI-7 and 32 g, SmaI-10c and 16a; for the TR-region e.g. EcoRI-12; see Fig. 1) as well as bands representing "border fragments" which harbor the 25-bp border repeats. Left hand TL-region border fragments 'are e.g. BamHI-8 and SmaI-17. A right hand TR-region border fragment is EcoRI-20. Since T - D N A is generally believed to be integrated at or near the 25-bp border repeats, the border fragments are not detected as such in D N A from crown-gall lines. Instead, bands are recovered representing "non-internal fragments" which usually do not comigrate with fragments in the reconstruction lane. These non-internal bands often represent plant D N A / T - D N A "fusion fragments ". In some cases non-internal bands may also represent rearranged T - D N A fragments.

Results

Southern and Northern blots. Southern blots were prepared as described by Southern (1975) and Memelink et al. (1983). Restriction fragments of DNA, which had been separated by electrophoresis on a 0.7% agarose gel, were transferred and bound to GeneScreenPlus sheets (New England Nuclear, Boston, Mass., USA). Restriction enzymes were purchased from Boehringer, Mannheim, FRG). Hybridization of Southern blots was performed as described by the manufacturer using D N A probes, labelled with 32p, by nick translation (Rigby et al. 1977). Alpha-a2P-dCTP (1.13-1014 Bq-mmol 1) was obtained from Amersham (Amersham, Bucks., UK); Southern autoradiograms were exposed to Kodak X-Omat A R X-ray films at - 7 0 ~ C, using Kyokko-LHII intensifying screens. Northern blots were prepared as described by Southern

Phenotypes. Figure 2 showns tumors on stems of d i - h a p l o i d N. plumbaginifolia p l a n t s g r o w n i n t h e greenhouse, two and a half months after infection w i t h d i f f e r e n t A. tumefaciens s t r a i n s . I n f e c t i o n s w i t h s t r a i n s h a v i n g i n t a c t onc-genes ( L B A 4 0 1 3 a n d L B A 1 8 3 3 ; A u x + C y t +) r e s u l t e d in l a r g e t u m o r s without shoots or roots. Infections with strains lacking either an intact cytokinin locus (LBA4210; Aux + Cyt ) or an intact auxin locus (LBA4060; A u x - C y t +) r e s u l t e d i n a t t e n u a t e d t u m o r s . D i f f e r ences in growth rate between tumors induced by

396

R. Peerbolte et al. : T - D N A in Nicotiana crown-gall tissue

A.

B.

I

TL - DNA 0

2

4

6

8

10

12

IL

I

I

I

I

I

I

I

I

18C

122e

8

I

_

1

l#,l ~ I

116o

I 17 I

e

I

~

III

7

I

loc

I

I 14

lalbl

16

18

I

l

I

20

I

22

I

24

I

I

kb

2

Barn H I

l l

I

3 ....

T.-oNA

I

II 32g

3

17o

I

19a

l

12 3b I

13

Hindl]I

4 l

20

l

eco RI Srno I Hpo I Bgl

c I

g

C. pOTY 8 I

TI-

Tr

I 5

D.

.

.

.

7

2

1

4

6b

3

.

&' .

I

E.

6o

auxin ,ocus

/

~

] [cytokinin l,ocus i i

/

%

I I

I I

Auf

.

2'

I

0'

I

I

mannopine oeus I ~[ locus i I

i I

I I I

1'

.

I ~

Aut+Reg+[ 'I F.

.

3'

I 0

I.o;I I I

4-

Ags+

Fig. 1 A-F. Physical map of the T-region of the Ti plasmids of wild-type octopine A. tumefaciens strain Ach5. A Extent of wild-type T L - D N A and T R - D N A (border repeats not indicated). B Restriction map of the T-region of pTiAch5. Numbering of restriction fragments according to De Vos et al. (1981). Insertion sites of Tn904 (A, inactivating gene 4) in LBA4210 (Klapwijk et al. 1980) and IS60 (zx, inactivating gene 2) in LBA4060 (Klapwijk et al. 1978) are indicated. Basepair numbering according to Barker et al. (1983). C Extent of the T-region clones pOTY8 (Hirsch and Beggs 1984) and pRAL3252+pRAL3076 (T1-Tr; Ooms et al. 1981). D Open reading frames on T L - D N A and T R - D N A according to Barker et al. (1983). Genes are numbered according to Willmitzer et al. (1982; TL) and Salomon et al. (1984; TR). Genes which have a clear phenotypic effect, are hatched. E T - D N A loci. F Phenotypes related to T - D N A loci: A u t + = phytohormone antotrophy in tissue culture; Reg + =regeneration of shoots; Ocs + = octopine synthesis; Mas + = mannopine synthesis; Ags + agropine synthesis

wild-type and mutant strains on axenic plantlets grown in vitro were less pronounced. Tumors obtained from the latter infections were excised from the plantlets and put into tissue culture. Names and classification of these lines have been given in Materi'als and methods. As a control, the nontransformed callus line c N P O was induced from a di-haploid N. plumbaginifolia leaf explant. After two months in culture, c N P O callus was transferred to medium lacking phytohormones on which it continued to grow as green friable callus (Fig. 3A). This non-transformed line therefore showed an Aut+ phenotype. The capability of c N P O to regenerate shoots decreased over time (Reg + > -). Figure 3 B-I shows the morphology of most of the N. plumbaginifolia crown-gall lines studied.

The crown-gall line cNP1, induced by wild-type strain LBA4001 grew as a compact Aut+ callus which unexpectedly regenerated shoots (Reg+). The callus produced octopine (Ocs +), but no mannopine or agropine ( M a s - Ags-), and the shoots were O c s - M a s - A g s - . Cellular clones of cNP1, obtained via protoplast isolation, grew as green/ white compact calli. Sometimes, differentiation into small leaf-like structures was observed. Particularly cNP1-4 showed differentiation into teratomata (Reg +) which were Ocs +. These shoot-like structures showed a morphology similar to tobacco 'crown-gall shoots': thick fleshy leaves (Tfl +), reduced apical dominance (Rad +) and root deficiency (Rod+). Their lack o f roots prevented these shoots from growing in soil. Unfortunately they remained too small to be grafted. The Reg + trait

R. Peerbolte et al. : T - D N A in Nicotiana crown-gall tissue

Fig. 2 A - D . Crown-gall tumors on stems of N. plumbaginifolia plants in the greenhouse. Wounds on stems of di-haploid plants were inoculated with the following octopine type strains: A LBA4013, having a wild-type Ach5 T-region with intact auxin and cytokinin loci (Aux + Cyt+); B LBAI833, having a manipulated Aux + Cyt + T-region (this paper); C LBA4210, an Ach5 derivative which has a Tn904 insertion inactivating gene 4 (the cytokinin locus; Aux + C y t - ) ; D LBA4060, an Ach5 derivative which has an IS60 insertion inactivating gene 2 of the auxin locus (Aux Cyt +)

of the cNP1 clones declined in the course of time. Figure 3 B shows the cNP1 clone cNP1-4. Cellular clones from shoots derived from the uncloned cNP1 tissue all had an A u t - phenotype and were probably not transformed. NP2, also induced by LBA4001, was cloned as well. Phenotypically the cNP2 clones very much resembled cNPO (Aut + Reg + > - Ocs-). N P I 6 was induced by an Aux + Cyt + A. tumefaciens strain carrying an engineered T-region (LBA1833). It was our intention to introduce into N. plumbaginifolia the structural gene for kanamycin/G418 resistance from the transposon Tn5, linked to the Tk promoter from HSV (Barclay et al. 1983), in order to obtain kanamycin/G418resistant tissue. In c N P I 6 no opines were detected. Seven months after the tumor had been put into tissue culture, the tissue regenerated some shoots. When cultured separately from the callus, these shoots formed a rather distorted root system. In soil the shoots matured into plants with a very aberrant phenotype. These plants did form flower buds, but flowers did not develop and seed could not be obtained. Callus derived from the shoots proved to be Aut + Ocs- M a s - A g s - . cNP16 grew well on medium lacking phytohormones and containing 80 mg.1-1 G418, whereas control tissues died. The tissue however, did not survive 100 mg. 1-1 kanamycin sulphate. When the NPTII enzyme assay (Paszkowski et al. 1984) was performed on extracts from cNP16, positive results could not be obtained (not shown). This indicates that kanamy-

397

cin sulphate is a more reliable selection drug than G418 in N. plumbaginifolia. The NPh3 tumor was induced by the Aux + C y t - strain LBA4210. In tissue culture on LS-H medium, the cNPh3 callus surprisingly started to produce shoots (sNPh3) instead of roots. Callus and shoots were Ocs M a s - A g s - . After having grown on LS medium containing 3.10 -5 mol.1-1 of the hypomethylating agent 5-azacytidine for a period of three to twelve weeks, mannopine and agropine, and sometimes also a weak octopine signal, were detected in callus tissue but not in shoots. The shoots produced both callus and roots at the base but were unable to grow in soil and could not be grafted. Table 3 summarizes the phenotypes of all tissues studied. Some general conclusions can be drawn from the data presented in this Table. (i) All crown-gall tissues are Aut +. This is not surprising since all tissues have been selected for this trait. (ii) Six out of twelve crown-gall lines studied (50%) unexpectedly were O c s - . (iii) Five out of the six lines (83%) induced by Aux + Cyt + strains unexpectedly were Reg +, either stably or only at a certain stage during tissue culture. (iv)Two out of four lines (50%) induced by an Aux + C y t - strain instead of roots surprisingly produced shoots. (v) Eight out of twelve lines (67%) showed a phenotypic ' switch' at a certain stage of subculturing. (vi) Taken together, nine out of twelve crown-gall lines (75%) exhibited at least one phenotypic trait which deviated from the phenotype to be expected from the A. tumefaciens strain used to induce the line. (vii) In five out of twelve lines (42%) the TRD N A opines mannopine and agropine were detected.

Transferred DNA. The high percentage of lines with aberrant phenotypes raises the question of whether these deviations are caused by lack of expression of certain T - D N A genes which are further intact, or by deletions due to irregular T - D N A structures. The lack of opine gene expression in cNPh3 is probably an example of the first possibility since in this case gene expression can be induced by the demethylating agent 5-azacytidine. In all other cases this compound could not restore the expected phenotype. In order to find out whether the aberrant phenotypes could be correlated with irregular T - D N A structures, the T - D N A of most of the lines listed in Table 3 was analyzed with the Southern blotting technique. The results of these analyses have been summarized in the columns " T L " and " T R " of Table 3. The lines marked with an asterisk have been analyzed in de-

398

R. Peerbolte et al. : T - D N A in Nicotiana crown-gall tissue

Fig. 3A-I, Phenotypes of N. plumbaginifolia crown-gall tissues. A cNPO : non-transformed Aut + line of green friable callus, The picture was taken after the tissue had been in culture for about three years. B cNP1-4: Aut + Reg + cellular clone from NP1 tumor, induced by the wild-type octopine A. tumefaciens strain LBA4001 on a di-haploid N. plumbaginifolia plantlet. The picture was taken after the clone had been in tissue culture for four years. C cNPh3: tumor on haploid N. plumbaginifolia plantlet induced by the rooter mutant LBA4210. The picture was taken three weeks after infection, one week before the tumor was put into tissue culture. D sNPh5: rooting-deficient (Rod +) shoot-like structures from an Aut + tumor induced by the rooter mutant LBA4210 on a haploid N. plumbaginifolia plantlet. The picture was taken after three years of tissue culture. E cNP6: Aut + callus, induced by the rooter mutant LBA4210 on a di-haploid N. plumbaginifolia plantlet. The picture was taken four months after infection and three months after the tissue had been put into tissue culture. Some rootlike structures emerging from the callus are visible. Roots were not formed anymore during later stages of tissue culture, F cNP7: Aut + Reg + tumor line induced by the shooter mutant LBA4060 on a di-haploid N. plumbaginifolia plantlet. Some shoots emerging from the callus are visible. The picture was taken four months after tumor induction and three months after the tissue was excised from the infected N. plumbaginifotia plantlet. G cNPh8 : tumor line induced by the shooter mutant LBA4060 on a haploid N. plurnbaginifolia plantlet. The picture was taken six weeks after infection. Leaf-like structures developed from the callus at that time (Reg+). Later this line ceased to regenerate and grew as green callus. H cNP9: Aut + Reg + tumor line, induced on a di-haploid N. plumbaginifolia planttet by LBA4013, which has a wild-type T-region. The picture was taken after three years o f tissue culture. I sNP10: Rod + shoot-like structures derived from the Aut + Reg + cNP10 tumor line, induced by LBA4013 on a di-haploid N. plumbaginifolia plantlet. The picture was taken after three years of tissue culture

R. P e e r b o l t e et al. : T - D N A in Nicotiana c r o w n - g a l l tissue

399

Table 3. P h e n o t y p e s o f N. plumbaginifolia c r o w n - g a l l tissues. T L = T L - D N A sequences de t e c t e d ( + ) or n o t ( - ) ; T R = T R - D N A sequences d etected ( + ) or n o t ( - - ) ; * = T - D N A s t u d i e d in d e t a i l ; 9 = n o t s t u d i e d ; =t= = 4 c e l l ul a r clones o f NP1 ( N P I - 1 , NP1-2, NP1-3 a n d N P I - 4 ) a n d 3 o f N P 2 (NP2-1, NP2-2, NP2-3) were s t u d i e d ; ? = T R - D N A is n o t necessarily pre s e nt in c r o w n - g a l l lines; + = (weak) p h e n o t y p i c e x p r e s s i o n ; + > - = o r i g i n a l l y p o s i t i v e b u t l a t e r on n e g a t i v e ; + / - - t h e line p r o d u c e d a p o s i t i v e sector a n d a n e g a t i v e one, w h i c h were f u r t h e r s u b c u l t u r e d s e p a r a t e l y ; - / a = onl y pos i t i ve a ft e r i n d u c t i o n w i t h 5-azacytidine. U n e x p e c t e d p h e n o t y p e s are p r i n t e d in b o l d face Aut I

Aux- Cyt-

II

A u x + Cy t +

*NPO

*NP1 *NP2 *NP9 NPI0 NPhl 1 * NP16 III

(--) expected (LBA4001) + (LBA4001) + (LBA4013) (LBA4013) ( LBA4013) (LBA1833)

Aux + Cyt*NPh3 NPh4 NPh5 NP6

IV

expected

expected (LBA4210) ( LBA4210) (LBA4210) (LBA4210)

A u x - Cyt + * NP7 NPh8

expected (LBA4060) (LBA4060)

.

Reg .

Ocs .

TL

Mas

.

.

TR

9

?

.

+

+ ) .

.

+

-

+

+ + + + + +

+ [+ ) . -+ + /+ )-

+ . + ) -+ -

+

-

+

+

?

?

?

+ + + +

+ + -

- [a + + ) +

+

- [a + + + ) -

-- ]a + + + ) -

+

+

+

+

+

?

?

?

+ +

+ + ) -

+ +

+ +

+

+

-

tail. From the data obtained it became clear that in addition to the five lines which already had been concluded to have T R - D N A sequences according to opine production, one more line (NP9) was detected with T R - D N A sequences. This means that 50% of the lines contained T R - D N A . As expected no T - D N A sequences were detected in D N A from the non-transformed line N P O (not shown). Surprisingly in NP2 and its clones, putatively transformed by LBA4001, no TD N A sequences were detected either (not shown). Apparently, the Aut + character of these lines was not due to the action of T - D N A onc-genes, but rather to a feature which the tissues had acquired in the early course of tissue culture and which is known as 'habituation' (Sacristan and Melchers 1977). The T - D N A patterns of the Ach5 wild-typeinduced cNP1 clones cNPI-1, cNP1-2, cNP1-3 and cNP1-4 were identical and irregular. Figure 4 A shows BamHI digests of D N A from the four cNP1 clones. Regular internal BamHI-fragments are BamHI-30b usually too small to be detected , BamHI-28 (both located between BamHI-8 and BamHI-I 7a) and BamHI-17a (see Fig. 1 B). Indeed BamHI-28 was detected as an internal fragment in the cNPI clones. BamHI-i 7a, however, was not observed. Four non-internal bands were detected.

.

Ags

. + +

.

.

.

+ + +

+

.

?

-.

--

. --

+

-

+

The two largest bands (-17 kb and ~ 13 kb) hybridized with BamHI-8, whereas the other bands did not (Fig. 4B). Therefore the -17-kb and -13-kb non-internal BamHI fragments were probably left hand TL-DNA/plant D N A fusion fragments indicating that the cNPI clones harbored at least two copies of TL-DNA. Figure 4 C shows EcoRI digests of c N P I - I and cNP1-4, probed with BamHI-8. Three bands were detected: EcoRI-a (6.9 kb), EcoRI-b (5.9 kb), and EcoRI-c (1.0 kb). EcoRI-7, which is normally detected as an internal EcoRI T - D N A fragment, was not present in the cNPI clones. The band EcoRI-c represented a fragment which was somewhat smaller than the internal fragment EcoRI-32g (located between EcoRI-3 and EcoRI-7). When a probe was used which covered the entire TL-region, an additional non-internal EcoRI band of 1.8 kb was observed (not shown). The reconstruction of an EcoRI/ BamHI double digest, probed with BamHI-8, showed three fragments: fragment x, starting at the left BamHI-8 site and ending at the left EcoRI32g site; fragment y, extending from the right EcoRI-32g site to the right BamHI-8 site, and EcoRI-32g (Fig. 4D, see Fig. I B). When an E c o R I / B a m H I double digest of cNPI-1 and cNP14 D N A ' s was probed with BamHI-8, the fragments EcoRI-a and EcoRI-c were detected again, as well

400

R. Peerbolte et al. : T - D N A in Nicotiana crown-gall tissue

Fig. 4 A - D . T - D N A in cNP1 clones (induced by LBA4001, Aux + Cyt+). A Southern blot of D N A from cNPI-1 (lane 2), cNPI-2 (lane 3), cNP1-3 (lane 4) and cNP1-4 (lane 5) f r a g m e n t e d with BamHI and probed with p R A L 3 2 5 2 + p R A L 3 0 7 6 (T1--Tr, see Fig. l C), together covering the entire TL + T R region. Lane 1 shows a reconstruction experiment (R) containing non-transformed eNPO D N A mixed with three copies of wild-type pTiAch5 per diploid N. plumbaginifolia genome (3 x reconstruction). B Southern blot of D N A from cNPI-1 (lane 1) and c N P l - 4 (lane 2) fragmented with BamHI and probed with BamHI fragment 8. Lane 3 shows a 1 x reconstruction. C Southern blot of D N A from c N P l - i (lane 1) and cNPI-4 (lane 2) fragmented with EcoRI and probed with BamHI-8. Lane 3 shows a i x reconstruction. D Southern blot of D N A from c N P I - I (lane 1) and cNPI-4 (lane 2) fragmented with b o t h EcoRI and B a m H I and probed with BamHI-8. Lane 3 shows a 1 x reconstruction

as the internal fragment y. These results imply that EcoRI-a was a T-DNA/plant D N A fusion fragment, containing EcoRI-3 sequences and the left T - D N A boundary. EcoRI-b hybridized to a clone overlapping the right part of the TL-region (not shown) and probably was an EcoRI-7 derivative. When other restriction enzymes were used, unexpected banding pattern were observed as well. Fragment SmaI-10c was not detected as an internal fragment in c N P I clones (not shown), which is in accordance with the absence of EcoRI-7 and BamHI-17a. In addition, BglII fragments a and b were not detected as internal T L - D N A fragments (not shown), indicating that the EcoRI-c band most likely did not represent the internal fragment EcoRI-32g. T R - D N A sequences were not detected in cNPI clones. Studies of possible effects of the aberrant TD N A structure of the c N P I - I clones on expression of T - D N A genes did not show differences between the cNP1 clones at the transcriptional level, not even between the R e g - c N P I - I clone and the Reg § cNP1-4 clone. It appeared that not all T - D N A genes were expressed (not shown). Transcript 2 of

the auxin locus was not detected. This correlated with the absence of BglII fragments a and b and explained, together with the presence of transcript 4, the tendency o f c N P 1 clones to regenerate shootlike structures. Transcript 4 also accounted for the Aut + phenotype of the cNPI clones. Transcripts 6a/b were absent. This correlated with the absence of the internal fragments EcoRI-7, BamHI-17a and SmaI-10c. The presence of transcript 3 explained the Ocs § phenotype. Transcript I was expressed more strongly than usually found in tobacco tissues. Moreover, a transcript comigrating with transcript 3 faintly hybridized to BamHI-8. Since this D N A clone does not harbor gene-3 sequences, this transcript should have come from some other, truncated T - D N A gene. When pOTY8 was used as a probe to hybridize to EcoRI-digested D N A from cNP9 (induced by the Aux § Cyt + strain LBA4013), bands were detected comigrating with the internal fragments EcoRI-7 (TL-DNA) and EcoRI-12 ( T R - D N A ) (Fig. 5, lane 2). In addition five non-internal bands were observed. Three of these bands hybridized with BamHI-8 (Fig. 5, lane 4). These bands prob-

R. Peerbolte et al. : T-DNA in Nicotiana crown-gall tissue

401

Fig. 5. T-DNA in cNP9 (induced by LBA4013, Aux + Cyt+). Southern blots with DNA isolated from cNPg, Aut § Reg- Ocs- Mas- Ags- callus, induced on a di-haploid N. plurnbaginifolia plantlet. Lane 1 : 1 x reconstruction of cNPO DNA and pTiAch5 DNA, fragmented with EcoRI and probed with pOTY8 (see Fig. 1 C). L a n e 2 : c N P 9 DNA fragmented with EcoRI and probed with pOTY8. Lane 3.: 1 x reconstruction of cNPO DNA and pTiAch5 DNA, fragmented with EcoRI and probed with BamHI-8 after removal of the pOTY8 probe. Lane 4 : c N P 9 DNA fragmented with EcoRI and probed with BamHI-8 after removal of the pOTY8 probe. Lane 5: 1 x reconstruction of cNPO DNA and pTiAch5 DNA, fragmented with EcoRI and probed with EcoRI-12 after removal of the BamHI-8 probe. Lane 6 : c N P 9 DNA fragmented with EcoRI and probed with EcoRI-12 after removal of the BamHI-8 probe

ably represented T-DNA/plant D N A fusion fragments, harboring the left T L - D N A boundary. This indicated that at least three T L - D N A segments should be present in cNP9. The 5.4-kb band which comigrated with EcoRI-12, did hybridize to an EcoRI-12 probe indeed (Fig. 5, lane 6). However, it also hybridized to BamHI-8 (Fig. 5, lane 4). This indicated that T L - D N A and T R - D N A might exist in a tandem arrangement which would imply that the 5.4-kb EcoRI band did not represent the internal fragment EcoRI-~2. If so, the absence of EcoRI-12 meant that cNP9 did not contain an intact copy of T R - D N A . This conclusion was in accordance with the Mas Ags- phenotype of the tissue. In addition to the 5.4-kb fragment, three non-internal EcoRI fragments hybridized to EcoRI-12, which demonstrated that even more truncated T R - D N A segments were present in cNP9. BamHI fragment 17a was observed as an internal T L - D N A fragment in cNP9 (not shown). Nevertheless the tissue was found to be O c s - . This might be the result of minor changes in the TDNA, impairing expression of gene 3 without affecting the BamHI-17a restriction sites. The presence of EcoRI-7 indicated that gene 4 was present in cNP9, accounting for the Aut + phenotype. For cNP16 (induced by the Aux + Cyt + TkN P T I I strain LBA1833) Southern blotting experiments revealed that a restriction fragment containing the NPTII gene construct was present in this line. Additionally it was observed that cNP16 con-

tained no regular T - D N A but rather a number of truncated T - D N A segments (not shown). The D N A from the sNPh3 shoots (induced by the rooter mutant LBA4210) was fragmented with BamHI. When the T-region clone pOTY8 was used as a probe many bands were detected which hybridized with high intensity, indicating a high copy number (Fig. 6A, lanes 3-5). This was confirmed by a BamHI-digest of sNPh3 D N A , probed with the border fragment BamHI-8. Many non-internal left TL-DNA/ptant D N A fusion fragments were detected, hybridizing with high intensity (Fig. 6 A, lanes 1 and 2). When the internal fragment BamHI-17a was used as a probe a number of bands were detected (Fig. 6A, lane 6), whereas only one BamHI-band was expected to hybridize to BamHI-17a: the one representing the 11.7-kb fragment BamHI-17a* (i.e. BamHI-17a::Tn904). Indeed this fragment was detected in sNPh3 DNA, but in addition at least five, weaker hybridizing bands were observed as well. This indicated that truncated T L - D N A fragments were present in sNPh3. In addition to BamHI-17a*, also EcoRI-7* and EcoRI-32g were detected as internal fragments (not shown). From these data it could not be concluded that an intact auxin locus was present in sNPh3. The absence of octopine in sNPh3 implied that none of the T L - D N A segments present in sNPh3 expressed the octopine-synthase gene 3 correctly. In Fig. 6A, lanes 7 and 8 show that a 5.4-kb band comigrating with the internal T R - D N A frag-

402

R. Peerbolte et al. : T-DNA in Nicotiana crown-gall tissue

Fig. 6A, B. T-DNA in sNPh3 (induced by LBA4210, Aux + Cyt-) and sNP7 (induced by LBA4060, Aux- Cyt+). A Southern blots containing DNA from sNPh3, R o d - Ocs- shoots from Aut + Reg + Ocs- Mas- Ags- cNPh3, induced by the rooter mutant LBA4210 on a haploid N. plumbaginifolia plantlet. Lane 1 : 1 x reconstruction of untransformed cNPO DNA mixed with pTi DNA from LBA4210, fragmented with BamHI and probed with BamHI-8. Lane 2 : s N P h 3 DNA, fragmented with BamHI and probed with BarnHI-8. Lane 3: sNPh3 DNA fragmented with BamHI and probed with pOTY8 (see Fig. 1 C); exposure time of autoradiogram to X-ray film=2 weeks. L a n e 4 : 1 x reconstruction of untransformed cNPO DNA and the pTi plasmid of LBA4210, fragmented with BamHI and probed with pOTY8; exposure time=2 weeks. Lane 5 : s N P h 3 DNA fragmented with BamHI and probed with pOTY8; exposure time= 16 h. Lane 6 : s N P h 3 DNA fragment with BamHI and probed with BamHI fragment 17a, after removal of the pOTY8 probe; exposure time=16 h. Lane 7 : 1 x reconstruction of untransformed cNPO DNA and LBA4210 pTi DNA, fragmented with EcoRI and probed with EcoRI-12. L a n e 8 : s N P h 3 DNA fragmented with EcoRI and probed with EcoRI-12. B Southern blot containing DNA from sNP7, Rod + Ocs + shoots from Aut + Reg + Ocs + Mas- Ags- callus cNP7, induced by the shooter mutant LBA4060 on a di-haploid N. plumbaginifolia plantlet. Lane 1: I x reconstruction of cNPO DNA and LBA4060 pTi DNA, fragmented with SmaI and probed with pOTY8. Lane 2 : s N P 7 DNA fragmented with SmaI and probed with pOTY8

ment EcoRI-12 was present in sNPh3 D N A , indicating that intact copies of T R - D N A might be present. The M a s - A g s - phenotype, however, was not in favor of this assumption. On the other hand, T R - D N A opine genes in sNPh3 might have been inactivated by D N A methylation, as was the case in cNPh3 which could be induced to produce mannopine and agropine by the hypomethylating agent 5-azacytidine. For some unknown reason, induction of T R - D N A opine genes in sNPh3 by 5-azacytidine might not have been strong enough to be detected. In addition to the 5.4-kb band, at least five non-internal EcoRI bands were detected which hybridized to EcoRI-12. This indicated that, next to truncated T L - D N A segments, sNPh3 harbored truncated T R - D N A segments as well. The IS60 insertion in the pTi plasmid of LBA4060 is located in the internal fragments

EcoRI-32g and BglII-b and in the non-internal fragment BamHI-8 (Ooms et al. 1981). Southern blots revealed that D N A from sNP7 (induced by LBA4060) contained the following internal fragments: EcoRI fragments 32g* (i.e. EcoRI32g::IS60) and 7; SmaI fragments 16a and 10c* (Fig. 6B) and BamHI-17a. These results indicated that sNP7 contained a normal T L - D N A segment, extending from the right T L - D N A 25-bp border repeat in EcoRI-19a to the left repeat in Sinai-17. The number of non-internal T-DNA/plant D N A fusion fragments indicated a copy number of one T L - D N A copy per diploid N. plurnbaginifolia genome. These findings were in accordance with the Aut + Reg + Ocs + phenotype of the callus and the R o d + Ocs + phenotype of the shoots. N o TRD N A sequences were detected, which agreed with the fact that sNP7 was found to be M a s - Ags-.

R. Peerbolte et al. : T - D N A in Nicotiana crown-gall tissue

Summarizing the T - D N A data it can be concluded that of the six lines studied in detail, only one line, which also happened to exhibit the expected phenotype, appeared to have a regular TD N A structure. One other line surprisingly did not have any T - D N A sequences at all and the remaining four lines, all showing aberrant phenotypes, were found to have irregular T - D N A structures. Discussion

This paper deals with N. plurnbaginifolia tissues, both of haploid and di-haploid origin, which have been transformed with wild-type and T-region mutant strains of Agrobacterium tumefaciens. There are a number of other reports on transformed N. plumbaginifolia lines (Bevan etal. 1983; De Block et al. 1984; Horsch et al. 1984), which, however, mainly deal with the introduction of foreign genes (not derived from the Ti-plasmid) into N. plumbaginifolia. N o t much is known about TD N A structure, organization and expression in N. plumbaginifolia crown-gall lines. N. plumbaginifolia callus tissue was found to habituate rather easily. Phytohormone-independent growth in vitro was observed for several lines lacking T-DNA, indicating that the Aut + trait should be regarded as a less reliable marker in N. plumbaginifolia to select for transformed tissues. Whereas tumor development on the plant appeared to be quite similar for tobacco and N. plumbaginifolia, the development of the N. plumbaginifolia crown-gall lines in vitro was found to be quite different. The most striking result of the present study is the high incidence of aberrant T-DNAlinked phenotypes of cultured stem tumors. Aberrant phenotypes were found in 75% of the N. plumbaginifolia crown-gall lines studied. After cocultivation of N. plumbaginifolia protoplasts with LBA4013, Czak6 and Marton (1986) found that about 50% of the transformed lines exhibited aberrant phenotypes. Similar results were obtained by Wullems et al. (1981) with tobacco protoplasts. Van Slogteren et al. (1983) reported that 40-50% of the transformed clones obtained from tobacco stem tumors showed aberrant phenotypes. In all these cases aberrant phenotypes were explained by the presence of irregular T - D N A structures resulting in loss or inactivation of T - D N A genes. Recently, Simpson et al. (1986) gave an overwiev of T - D N A structures in crown-gall lines. A b o u t 35% of tobacco crown-gall lines had irregular T - D N A structures, a percentage also found approximately for tobacco lines which had been transformed with binary vector systems (Simpson et al. 1986; Spiel-

403

mann and Simpson 1986). In most of these cases the irregularity of the T - D N A structures was not very severe, only affecting the T - D N A termini, and even without phenotypic implications in several cases. For tobacco, really complex T - D N A structures and organizations have been reported only for a relatively small number of crown-gall lines (Kwok et al. 1985; Peerbolte et al. 1986a; Van Lijsebettens et al. 1986). Moreover, any aberrant phenotype resulting from an irregular T - D N A structure is usually masked in uncloned tobacco tumor lines. In the case of N. plumbaginifolia, aberrant phenotypes are even easily detected in uncloned tumor lines. In addition the complexity of irregular T - D N A structures and organizations is greater than usually found for tobacco. Moreover, in line with the high incidence of aberrant phenotypes, the frequency of occurrence of irregular T - D N A structures was found to be very high: only one of six lines studied in detail showed a regular T - D N A structure. For tobacco it has been described that T - D N A organizations consisting of many copies of (truncated) T - D N A segments may be accompanied by gene inactivation caused by methylation of cytosine residues in the D N A (Amasino et al. 1984; Peerbolte et al. 1986b). These silent genes can be activated by means of the hypomethylating agent 5-azacytidine. In the present study a similar result was obtained: in cNPh3, which produced shoots with a highly aberrant T - D N A architecture, opine genes could be induced by 3 . 1 0 - s mol. 1-1 5-azacytidine in the culture medium. For tobacco we previously discussed the possibilities of irregular T - D N A structures being caused either inside the bacterium before transfer and integration, or inside the plant cell after integration (Peerbolte et al. 1986a). For N. plurnbaginifolia in several cases, for instance in sNPh3, non-internal fragments were present at high copy numbers. From the data obtained it was concluded that these fragments most likely represent plant D N A / T D N A fusion fragments. This would mean that in these cases amplification events have taken place after T - D N A integration in the genome. Such amplifications after integration seem to be the main cause for irregular T - D N A structures in N. plumbaginifolia crown-gall tissues. Based on opine biosynthesis and T - D N A studies it was concluded that in 50% of the lines studied T R - D N A was present. This is considerably more than the 10% (based on opine synthesis) found by Czak6 and Marton (1986) for clones obtained via the cocultivation method. O f the six lines studied in detail for T - D N A structure and organi-

404

zation, two lines were found to have copies of truncated TR-DNA. This indicates that the percentage of lines with TR-DNA sequences might be considerably higher than the percentage of lines producing mannopine and-or agropine. Irregular T-DNA structures often go together with the loss of non-selected genes. The frequent occurrence of irregular T-DNA structures might render N. plumbaginifolia a less suitable host for the introduction of 'desired' (usually non-selective) genes. If the incidence of irregular T-DNA structures is an important general criterion for dividing plant species into groups of good and bad host suitability, this might complicate the genetic manipulation of plant species of interest. We thank Drs. M.W. Bevan, M.J.J. van Haaren and P.J.J. Hooykaas for providing us with several bacterial strains and T-DNA clones and Mr. Rolf de Kam for assistance in RNA isolations and Northern blot analysis. Dr. Eltjo Meijer is acknowledged for critically reading the manuscript.

References Akiyoshi, D.E., Klee, H., Amasino, R.M., Nester, E.W., Gordon, M.P. (1984) T-DNA of Agrobacterium tumefaciens encodes an enzyme of cytokinin biosynthesis. Proc. Natl. Acad. Sci. USA 81, 5994-5998 Akiyoshi, D.E., Morris, R.O., Hinz, R., Mischke, B.S., Kosuge, T., Garfinkel, D.J., Gordon, M.P., Nester, E.W. (1983) Cytokinin/anxin balance in crown gall tumors is regulated by specific loci in the T-DNA. Proc. Natl. Acad. Sci. USA 80, 407-411 Amasino, R.M., Powell, A.L.T., Gordon, M.P. (1984) Changes in T-DNA methylation and expression are associated with phenotypic variation and plant regeneration in a crown gall tumor line. Mol. Gen. Genet. 197, 437-446 Barclay, S.L., Meller, E. (1983) Efficient transformation of Dictyostelium discoideum amoebae. Mol. Cell Biol. 3, 211%2130 Barfield, D.G., Robinson, S.J., Shields, R. (1985) Plant regeneration from protoplasts of long term haploid suspension cultures of N. plumbaginifolia. Plant Cell Rep. 4, 104-107 Barker, R.F., Idler, K.B., Thompson, D.V., Kemp, J.D. (1983) Nucleotide sequence of the T-DNA region from the Agrobacterium tumefaciens octopine Ti plasmid pTi 15955. Plant Mol. Biol. 2, 335-350 Barry, G.F., Rogers, S.G., Eraley, R.T., Brand, L. (1984) Identification of a cloned cytokinin biosynthetic gene. Proc. Natl. Acad. Sci. USA 81, 4776M780 Bevan, M.W., FlaveIl, R.B., Chilton, M.-D. (1983) A chimaeric antibiotic resistance gene as a selectable marker for plant cell transformation. Nature 304, 184-187 Bourgin, J.-P., Chupeau, Y., Missonier, C. (1979) Plant regeneration from mesophyll protoplasts of several Nicotiana species. Physiol. Plant. 45, 288-292 Buchmann, I., Marner, F.-J., Schr6der, G., Waffenschmidt, S., Schr6der, J. (1985) Tumor genes in plants: T-DNA encoded cytokinin biosynthesis. EMBO J, 4, 853-859 Chilton, M.-D., Drummond, M.H., Merlo, D.J., Sciaky, D., Montoya, A.L., Gordon, M.P., Nester, E.W. (1977) Stable incorporation of plasmid DNA into higher plant cells: the

R. Peerbolte et al. : T-DNA in Nicotiana crown-gall tissue molecular basis of crown gall tumorigenesis. Cell 11, 263 271 Chilton, M.-D., Tepfer, D.A., Petit, A., David, C., Casse-Delbart, F., Temp6, J. (1982) Agrobacterium rhizogenes inserts T-DNA into the genomes of host plant root cells. Nature 295, 432-434 Czak6, M., Marton, L. (1986) Independent integration and seed-transmission of the TR-DNA of the octopine Ti plasmid pTi Ach5 in Nicotiana plumbaginifolia. Plant Mol. Biol. 6, 101-109 De Block, M., Herrera-Estrella, L., Van Montagu, M., Schell, J., Zambryski, P. (1984) Expression of foreign genes in regenerated plants and their progeny. EMBO J. 3, 1681-1689 De Vos, G., De Beuckeleer, M., Van Montagu, M., Schell, J. (1981) Restriction endonuclease mapping of the octopine tumor-inducing plasmid pTiAch5 of Agrobacterium tumefaciens. Plasmid 6, 249-253 Garfinkel, D.J., Simpson, R.B., Ream, L.W., White, F.F., Gordon, M.P., Nester, E.W. (1981) Genetic analysis of crown gall: fine structure map of the T-DNA by site-directed mutagenesis. Cell 27, 143-153 Hille, J., Van Kan, J., Klasen, I., Schilperoort, R. (1983) Sitedirected mutagenesis in Eseheriehia coli of a stable R772: : Ti cointegrate plasmid from Agrobacterium tumefaciens. J. Bacteriol. 154, 693 701 Hirsch, P.R., Beggs, J.D. (1984) Agrobacterium tumefaeiens TDNA in the yeast Saeeharomyces cerevisiae. Mol. Gen. Genet. 195, 209-214 Hoekema, A., Roelvink, P.W., Hooykaas, P.J.J., Schilperoort, R.A. (1984) Delivery of T-DNA from the Agrobacterium tumefaciens chromosome into plant cells. EMBO J. 3, 2485-2490 Horsch, R.B., Eraley, R.T., Rogers, S.G., Sanders, P.R., Lloyd, A., Hoffmann, N. (1984) Inheritance of functional foreign genes in plants. Science 223, 496-498 Inz6, D., Follin, A., Van Lijsebettens, M., Simoens, C., Genetello, C., Van Montagu, M., Schell, J. (1984) Genetic analysis of the individual T-DNA genes of Agrobacterium tumefaciens; further evidence that two genes are involved in indole3-acetic acid synthesis. Mol. Gen. Genet. 194, 265-274 Klapwijk, P.M., Oudshoorn, M., Schilperoort, R.A. (1977) Inducible permease involved in the uptake of octopine, lysopine and octopine acid by Agrobacterium tumefaciens strains carrying virulence-associated plasmids. J. Gen. Microbiol. 102, 1-11 Klapwijk, P.M., Scheulderman, T., Schilperoort, R.A. (1978) Coordinated regulation of octopine degradation and conjugative transfer of Ti plasmids in Agrobacterium tumefaciens: evidence for a common regulatory gene and separate operons. J. Bacteriol. 136, 775-785 Klapwijk, P.M., Van Breukelen, J., Korevaar, K. Ooms, G., Schilperoort, R.A. (1980) Transposition of Tn904 encoding streptomycin resistance into the octopine Ti plasmid of Agrobacterium tumefaciens. J. Bacteriol. 141, 129-136 Kwok, W.W., Nester, E.W., Gordon, M.P. (1985) Unusual plasmid DNA organization is an octopine crown gall tumor. Nucleic Acids Res. 13, 459-471 Leemans, J., Shaw, Ch., Deblaere, R., De Greve, H., Hernalsteens, J.P., Maes, M., Van Montagu, M., Schell, J. (1981) Site-specific mutagenesis of Agrobacterium Ti plasmids and transfer of genes to plant cells. J. Mol. Appl. Genet. l, 149-164 Linsmaier, E.M., Skoog, F. (1965) Organic growth factor requirements of tobacco tissue cultures. Physiol. Plant. 18, 100-127 Memelink, J., Wullems, G.J., Schilperoort, R.A. (1983) Nopaline T-DNA is maintained during regeneration and genera-

R. Peerbolte et al. : T-DNA in Nicotiana crown-gall tissue tive propagation of transformed tobacco plants. Mol. Gen. Geuet. 190, 516-522 Murashige, T., Skoog, F. (1962) A revised method for rapid growth and bioassays with tobacco tissue cultures. Physiol. Plant. 15, 473-479 Ooms, G., Hooykaas, P.J.J., Moolenaar, G., Schilperoort, R.A. (1981) Crown gall plant tumors of abnormal morphology, induced by Agrobacterium tumefaciens carrying mutated octopine Ti-plasmids; analysis of T-DNA functions. Gene 14, 33-50 Otten, L.A.B.M., Schilperoort, R.A. (1978) A rapid microscale method for the detection of lysopine and nopaline dehydrogenase activities. Biochim. Biophys. Acta 527, 497-500 Paszkowski, J., Shillito, R.D., Saul, M., Mandak, V., Hohn, T, Hohn, B., Potrykus, I. (1984) Direct gene transfer to plants. EMBO J. 3, 2717-2722 Peerbolte, R., Krens, F.A., Mans, R.M.W., Floor, M., Hoge, J.H.C., Wullems, G.J., Schilperoort, R.A. (1985) Transformation of plant protoplasts with DNA: cotransformation of non-selected calf thymus carrier DNA and meiotic segregation of transforming DNA sequences. Plant Mol. Biol. 5, 235-246 Peerbolte, R., Leenhouts, K., Hooykaas-Van Slogteren, G.M.S., Hoge, J.H.C., Wullems, G.J., Schilperoort, R.A. (1986a) Clones from a shooty tobacco crown gall tumor I: deletions, rearrangements and amplifications resulting in irregular T-DNA structures and organizations. Plant Mol. Biol. 7, 265-284 Peerbolte, R., Leenhouts, K., Hooykaas-Van Slogteren, G.M.S., Wullems, G.J., Schilperoort R.A. (1986b) Clones from a shooty tobacco crown gall tumor II: irregular TDNA structures and organization, T-DNA methylation and conditional expression of opine genes. Plant Mol. Biol. 7, 285-299 Rigby, P.W.J., Dieckmann, M., Rhodes, C., Berg, P. (1977) Labelling deoxyribonucleic acid to high specific activity in vitro by nick translation with DNA polymerase I. J. Mol. Biol. 113, 237-251 Sacrist~m, M.D., Melcbers, G. (1977) Regeneration of plants from "habituated" and "Agrobacterium-transformed" single-cell clones of tobacco. Mol. Gen. Genet. 152, 111-117 Salomon, F., Deblaere, R., Leemans, J., Hernalsteens, J.-P., Van Montagu, M., Scbell, J. (1984) Genetic identification of functions of TR-DNA transcripts in octopine crown galls. EMBO J. 3, 141 146

405 Schr6der, G., Waffenschmidt, S., Weiler, E.W., Schr6der, J. (1984) The T-region of Ti plasmids codes for an enzyme synthesizing indole-3-acetic acid. Eur. J. Biochem. 138, 387-391 Simpson, R.B., Spielmann, A., Margossian, L., McKnight, D. (1986) A disarmed binary vector from Agrobacterium tumefaciens functions in Agrobacterium rhizogenes. Plant Mol. Biol. 6, 403-415 Spielmann, A., Simpson, R.B. (1986) T-DNA structure in transgenic tobacco plants with multiple independent integration sites. Mol. Gen. Genet. 205, 34-41 Southern, E.M. (1975) Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98, 503-517 Thomashow, L.S., Reeves, S., Thomashow, M.F. (1984) Crown gall oncogenesis: evidence that a T-DNA gene from the Agrobacterium Ti plasmid pTiA6 encodes an enzyme that catalyzes synthesis of indoleacetic acid. Proc. Natl. Acad. Sci. USA 81, 5071 5075 Van Lijsebettens, M., Inz6, D., Schell, J., Van Montagu, M. (1986) Transformed cell clones as a tool to study T-DNA integration mediated by Agrobacteriurn tumefaciens. J. Mol. Biol. 188, 129-145 Van Slogteren, G.M.S., Hoge, J.H.C., Hooykaas, P.J.J., Schilperoort, R.A. (1983) Clonal analysis of heterogeneous crown gall tumor tissues induced by wild-type and shooter mutant strains of Agrobacterium tumefaciens - expression of T-DNA genes. Plant Mol. Biol. 2, 321 333 Willmitzer, L., Simons, G., Schell, J. (1982) The TL-DNA in octopine crown gall tumours codes for seven well defined polyadenylated transcripts. EMBO J. 1, 13%146 Wullems, G.J., Molendijk, L., Ooms, G., Schilperoort, R.A. (1981) Differential expression of crown gall tumor markers in transformants obtained after in vitro Agrobacterium tumefaciens-induced transformation of cell wall regenerating protoplasts derived from Nicotiana tabacum. Proc. Natl. Acad. Sci. USA 78, 4344-4348 Wullems, G.J., Molendijk, L., Schilperoort, R.A. (1980) The expression of tumor markers in intraspecific somatic hybrids of normal and crown gall cells from Nicotiana tabacum. Theor. Appl. Gen. 56, 203-208

Received 20 December 1986; accepted 5 February 1987