Streptothricins are known as antimicrobial agents produced by Streptomyces spp. Bacterial resistance to streptothricin is mediated by specific enzymes exhibiting an acetyltransferase activity which renders the drug non-toxic for bacteria. The nucleotide sequence of several streptothricin resistance genes from bacteria have been described. Certain cells of eukaryotic parasites (such as Ustilago maydis or Leishmania spp.) are sensitive to streptothricin and the introduction of the bacterial resistance gene sat2 renders them resistant. We show that numerous species of plants are sensitive to low concentrations of streptothricin. Moreover, introduction of the bacterial resistance gene sat3 under the control of the 35S cauliflower mosaic virus promoter protects these cells from the toxic action of streptothricin. Therefore, sat3-mediated streptothricin resistance appears to be a promising selective marker for genetic manipulation of plant cells.

We report the presence of an open reading frame, named ORF13a, encoding a putative regulatory protein on the T-DNA of Agrobacterium rhizogenes 8196 Ri plasmid. Homologous ORFs are present at the same location in two other types of Ri plasmids. We present evidence that ORF13a is transcriptionally active. Expression of ORF13a was investigated by analysis of glucuronidase (GUS) activity in transgenic tobacco containing an ORF13aGUS fusion. The gene fusion was expressed at higher level in roots than in leaves. The putative protein encoded by ORF13a has an isoelectric point of 11.55 and carries SPXX repeated motifs suggesting a possible regulatory function for this gene.

Opines consititute a class of substances that are characteristic of crown gall and hairy root, two proliferative plant diseases caused by Agrobacterium tumefaciens and A. rhizogenes. About 20 of these compounds have been described. Being produced by plant cells and used by the pathogen as growth substrates, they can be described as chemical mediators of parasitism. This plant-bacteria interaction is a natural instance of genetic manipulation since the genes encoding opine synthesis in plant cells are transferred from a bacterial plasmid into the plant genome. The current knowledge on opines is presented, as well as some biological aspects of this unique system.

Nicotiana tabacum cv. Xanthi transgenic plants expressing ORF13 of Agrobacterium rhizogenes 8196 T-DNA under the 35S RNA promoter from the cauliflower mosaic virus displayed developmental abnormalities. They were small, with short and variable internodal lengths, their root systems were poorly developed; leaves were small, asymmetric, rounded, wrinkled and dark green; flowers were short, and irregularly shaped. They exhibited reduced apical dominance and regularly produced offshoots at the base of the plant. This phenotype was also exhibited by offshoots of normal N. tabacum cv. Xanthi stock grafted with a transgenic scion indicating that expression of ORF13 influences plant development via diffusible factor(s).

Rhizopines are selective growth substrates synthesized in nodules only by strains of rhizobia capable of their catabolism. We report the isolation and study of genes for the synthesis and catabolism of a new rhizopine, scyllo-inosamine (sIa), from alfalfa nodules induced by Rhizobium meliloti Rm220-3. This compound is similar in structure to the previously described rhizopine 3-O-methyl-scyllo-inosamine from R. meliloti L5-30 (P.J. Murphy, N. Heycke, Z. Banfalvi, M.E. Tate, F.J. de Bruijn, A. Kondorosi, J. Tempé, and J. Schell, Proc. Natl. Acad. Sci. USA 84:493-497, 1987). The synthesis (mos) and catabolism (moc) genes for the Rm220-3 rhizopine are closely linked and located on the nod-nif Sym plasmid. The mos genes are directly controlled by the NifA/NtrA regulatory system. A comparison of the sequence of the 5' regions of the two mos loci shows very extensive conservation of sequence as well as strong homology to the nifH coding region. Restriction mapping and hybridization to DNA from the four open reading frames (ORFs) of the L5-30 mos locus indicate the absence of mosA and presence of the other three ORFs (ORF1 and mosB and -C) in Rm220-3. We suggest that the L5-30 mosA gene product is involved in the conversion of scyllo-inosamine to 3-O-methyl-scyllo-inosamine. Restriction fragment length polymorphism analysis of the moc regions of both strains shows that they are very similar. Regulation studies indicate that the moc region is not controlled by the common regulatory gene nifA, ntrA, and ntrC. We discuss the striking similarities in gene structure, location, and regulation between these two rhizopine loci in relation to the rhizopine concept.

Secondary metabolite production by plant cell cultures has been the subject of many investigations. Indole alkaloids such as ajmalicine, an antihypertensive drug, or vinblastine and vincristin, two antitumor compounds produced by Catharanthus roseus, are very important in the pharmaceutical industry. Cell cultures from the same plant species have been propagated in several laboratories since they are expected to produce physiologically potent indole alkaloids (Carew 1975). Undifferentiated cell lines producing high yields of ajmalicine, or serpentine, that can be easily converted into ajmalicine by reduction, have been obtained by selection (Zenk et al. 1977).

The Activator (Ac) transposable element has been previously introduced in Arabidopsis thaliana using Agrobacterium vectors. Mobility of the element was assayed in root-derived calli where it was shown to transpose. However here we present evidences suggesting that in the selfed progeny obtained from the regenerated plants the element is inactive. Working with restriction enzymes isoschizomers we have detected an overall increase in the methylation state in both Ac-T-DNA and Ri-T-DNA. Our results also show that the 5'-region of the Ac mRNA is methylated. The high level of methylation detected may be responsible for the inactivation of the transposable element.

To investigate the effect of plant transformation on opine catabolic (and pathogenic) Agrobacterium, we designed an experimental model in which the changes of a bacterial population associated with the root system of plants could be easily monitored. The bacterial population was composed of opine catabolic and noncatabolic strains of Agrobacterium. Bacteria were cocultivated with either transformed or normal Lotus plants. The composition of the bacterial population was estimated by dilution and plating of the growth medium on various media. We demonstrated that growth of bacteria was stimulated when they were associated with transformed plants. Furthermore, growth of opine-utilizing bacteria was specifically favored when bacteria were associated with transformed plants, but not when bacteria were associated with normal plants. This work indicates that transformation of plant cells by agrobacteria indeed can favor growth of the catabolic strain. This finding may lead to creation of engineered plant-bacteria interactions in which both partners will benefit from their association.

An 8 bp sequence repeated 6 times is present to the right of the mannopine type pRi8196 T-DNA right-border sequence. Experiments were designed to test whether these repeats have a role in T-DNA transfer. Several constructs in which different lengths of pRi8196 right-border region were linked to the cucumopine synthesis gene on an Agrobacterium-Escherichia coli shuttle vector were made. The recombinant plasmids were tested for their efficiency to act as a source of T-DNA in a binary system in which a wild-type Ri plasmid provided virulence and root-inducing functions. The T-DNA transfer efficiency of the constructs was assessed by computing the relative frequency of roots containing cucumopine. Depending on the Ri plasmid used as source of virulence functions, a high level of T-DNA transfer was observed only if 6 (pRi8196) or 5 (pRiA4) repeats were present. These results were confirmed by looking for single-stranded T-DNA molecules (T-strands) in bacteria induced for virulence. The repetition of the 8 bp unit was named 'T-DNA transfer stimulator sequence' (TSS).

The sensitivity to auxin of mesophyll protoplasts isolated from Agrobacterium rhizogenes transformed tobacco plants was shown to be higher than that of untransformed ones, by studying the action of auxin on the protoplast transmembrane electrical potential difference. This membrane response allowed us to show that single T-DNA genes, namely the rolA, B and C genes, were also able to confer an increased sensitivity to auxin on transformed protoplasts, rolB being the most powerful with a 10,000-fold increase. Plants transformed with rolB were thus chosen as a model system for further studies and the regulation of the rolB promoter in tobacco was studied by using the ß-glucuronidase (GUS) reporter gene. Tissue-specific expression of our rolB :GUS chimeric gene in root meristems and vascular tissues was strongly modified by the addition of exogenous auxin. Furthermore, the slight GUS activity detected in mesophyll protoplasts isolated from rolB:GUS plants could be increased 20 to 100 times by adding auxin. These results suggest that auxin plays a central role in the regulation of the rolB promoter in tobacco. Consequently, the sensitivity to auxin of rolB-transformed protoplasts could be modulated by auxin itself, in good correlation with the activation of the rolB gene by auxin. These interactions between the auxin signal and the rolB gene might explain some aspects of their cooperative effects for root induction on leaf fragments.

This paper presents the map and DNA sequence analysis of pRi8196 transferred DNA (T-DNA) genes encoding root-inducing and mannopine synthesis functions. A canonical 24-base-pair border repeat as well as two "pseudoborders" are present at the functional right T-DNA border. To the left of this border are homologs of the mas1' and mas2' genes of TR pRiA4. Next to these are five open reading frames (ORFs) homologous to ORFs 10-14 of TL of pRiA4. ORFs 10-12 (rolA, rolB, and rolC) are less related to their pRiA4 homologs than are the other large ORFs analyzed here. In contrast to T-DNA genes of pRiA4, pRi8196 T-DNA ORFs 11 and 12 (rolB and rolC) are sufficient to induce hairy roots on carrot disks.

Twokinds ofcellular responses toauxin, thehyperpolarization ofprotoplasts andthedivision ofprotoplast-derived cells, were compared inNicotiana tabacumplants transformed bydifferent T-DNAfragments ofAgrobacterium rhizogenes strain A4.Using transmembrane potential difference measurements tocharacterizehormonal sensitivity ofmesophyll protoplasts, wefound a30fold increase insensitivity toauxin inprotoplasts transformed by thewholeRiA4T-DNA.Furthermore, therolgenesoftheRiA4 TL-DNA, together orassingle genes, wereabletoincrease the sensitivity toauxinbyfactors upto104. Thedifferent effects of thesingle rolgenesonthesensitivity ofmesophyll protoplasts toauxin, rolBbeing themostpowerful, wereconsistent withtheir respective rhizogenic effects onleaffragments (ASpena, T Schmulling, CKoncz, JSchell [1987] EMBOJ6:3891-3899). No difference wasseenconcerning theeffects ofauxin ondivision ofcells derived fromnormal ortransformed protoplasts. These results suggest that onlysomecellular responses toauxin could beselectively altered byrolgenes.Theyalsoshowthatroltransformed tobaccos canbea modelsystemtostudy auxin action inplants. hairy roottobacco regenerants (23) wereshowntobemore sensitive toauxin thantheir normal counterparts. This feature suggests thatRiT-DNAgenesinduce theproliferation of transformed cells byaunique mechanism, ascompared toA. tumefaciens oncogenes, which cause disease byencoding enzymesforhormone biosynthesis (30). Recently, itwasshownthat asetofonlyafewgenes ofthe pRiA4TL-DNA, namely thethree genes rolA, B,andC(28), isable toinduce thefull hairy rootsyndrome intobacco (4, 10,24,27). Moreover, eachofthese genes isable onits own tomodify tobacco plant development (17,18,22,24), rolA androlB being able toinduce transformed rootformation (4, 24,27). Inthis paper wecompared theauxin sensitivity ofnormal tobacco plants withsensitivity ofplants transformed with the wholeT-DNAofpRiA4orthethree genes rolA, B,andC. Auxineffects onthetransmembrane potential difference of protoplasts andoncell division wereinvestigated. According totheformer test, single rol genes werefound toincrease the sensitivity ofmesophyll protoplasts toauxin. However, the proliferation ofbothnormal andtransformed protoplastsderived cells exhibited asimilar dependence onauxin.

Hairy root formation was induced by inoculating epicotyl, cotyledon and hypocotyl of Allocasuarina verticillata with two strains of Agrobacterium rhizogenes, A4 (agropine type) or 2659 (cucumopine type). Shoot regeneration from hairy roots induced by strain A4, and from control roots excised from untreated seedlings, required a hormone-enriched medium, whereas roots induced by strain 2659 spontaneously regenerated shoots on hormone-free medium. Newly formed shoots were multiplied and rooted to produce transgenic plants that developed unusually extensive, ageotropic root systems. The transformation was demonstrated by the presence of specific opines and by Southern blot analysis.

Two kinds of cellular responses to auxin, the hyperpolarization of protoplasts and the division of protoplast-derived cells, were compared in Nicotiana tabacum plants transformed by different T-DNA fragments of Agrobacterium rhizogenes strain A4. Using transmembrane potential difference measurements to characterize hormonal sensitivity of mesophyll protoplasts, we found a 30-fold increase in sensitivity to auxin in protoplasts transformed by the whole Ri A4 T-DNA. Furthermore, the rol genes of the Ri A4 T(L)-DNA, together or as single genes, were able to increase the sensitivity to auxin by factors up to 10(4). The different effects of the single rol genes on the sensitivity of mesophyll protoplasts to auxin, rolB being the most powerful, were consistent with their respective rhizogenic effects on leaf fragments (A Spena, T Schmülling, C Koncz, J Schell [1987] EMBO J 6: 3891-3899). No difference was seen concerning the effects of auxin on division of cells derived from normal or transformed protoplasts. These results suggest that only some cellular responses to auxin could be selectively altered by rol genes. They also show that rol-transformed tobaccos can be a model system to study auxin action in plants.

The responses to auxin of Lycopersicon esculentum roots transformed by (T(l)+T(r))-DNA of the Ri plasmid of agropine-type Agrobacterium rhizogenes strain 15834 and Catharanthus trichophyllus roots transformed by the (T(l)+T(r))-DNA, and by T(l)- or T(r)- DNA alone of the same bacterial strain were compared to that of their normal counterparts. The transmembrane electrical potential difference of root protoplasts was measured as a function of the concentration of exogenous naphthalene acetic acid. The sensitivity to auxin expressed by this response was shown to be independent of the measurement conditions and of the basal polarization of isolated protoplasts. According to this electrical response, as well as to the modulation by auxin of proton excretion by root tips and root tip elongation, roots transformed by (T(l)+T(r)) DNA are 100 to 1000 times more sensitive to exogenous auxin than normal roots, as is the case with normal and transformed roots from Lotus corniculatus (WH Shen, A Petit, J Guern, J Tempé [1988] Proc Natl Acad Sci USA 85: 3417-3421). Further-more, transformed roots of C. trichophyllus are not modified in their sensitivity to fusicoccin, illustrating the specificity of the modification of the auxin sensitivity. Roots transformed by the T(r)-DNA alone showed the same sensitivity to auxin as normal roots, whereas the roots transformed by the T(l)-DNA alone exhibited an auxin sensitivity as high as the roots transformed by (T(l)+T(r))-DNA. It was concluded that the high sensitivity to auxin is controlled by the T(l)-DNA in agropine type Ri plasmids.