Gudrun Schröder

Tumour genes in plants: T-DNA encoded cytokinin biosynthesis.

Gene 4 from the T‐region of Ti plasmids is responsible for cytokinin effects in crown gall cells; we investigated whether it codes for an enzyme of hormone biosynthesis. In a first set of experiments, gene 4 from octopine plasmid pTiAch5 and nopaline plasmid pTiC58 was expressed in Escherichia coli, and the gene products were identified by reaction with antiserum raised against a decapeptide derived from the DNA sequence of the gene. Extracts from cells expressing the gene contained high isopentenyl‐transferase activity catalyzing the formation of N6‐(△2‐isopentenyl)adenosine from 5′‐AMP and △2‐isopentenylpyrophosphate. The cytokinin was identified by sequential h.p.l.c. chromatography and mass spectrometry. In a second set of experiments it was shown that crown gall cells contained isopentenyltransferase activity and a protein of mol. wt. 27 000 which was identified as the product of gene 4 by reaction with the antiserum. Isopentenyltransferase activity was specifically inhibited by the antiserum. No comparable enzyme activity or immunoreactive protein was detected in cytokinin‐autotrophic, T‐DNA free tobacco cells. The results establish that gene 4 from the T‐region of octopine and nopaline Ti plasmids codes for an enzyme of cytokinin biosynthesis.

The mRNA for lysopine dehydrogenase in plant tumor cells is complementary to a Ti-plasmid fragment

The interaction between the soil bacterium Agrobacterium tumefaciens and many dicotyledonous plants represents a unique system of naturally evolved genetic engineering [ 1,2]. During infection of wounded plants a part of the Ti-plasmid, called T-DNA, is transferred from the bacteria into the nuclei of the plant cells, and is responsible for the known phenotypes of crown gall cells: Tumorous growth and synthesis of new substances, called opines, which are used specifically by the bacteria as sources of carbon, nitrogen, and energy. Studies with tumor inducing mutants of Ti-plasmids suggest that opine synthesis is controlled by a specific region on the T-DNA [3,4]. However, such experiments cannot resolve the question whether the T-DNA operates by activation of otherwise not expressed plant genes or whether T-DNA itself contains structural genes which are active in eucaryotic cells. We present results suggesting that the octopine plasmid pTi Ach5 codes for lysopine dehydrogenase, the enzyme responsible for octopine and lysopine synthesis in transformed plant tissues [5,6].

Nuclear and Polysomal Transcripts of T-DNA in Octopine Crown Gall Suspension and Callus-Cultures

SummaryTo establish a detailed map of the transcribed parts of the T-DNA in two octopine crown gall lines grown in suspension culture, T-DNA-derived steady-state nuclear and polysomal RNA as well as RNA synthesized in isolated nuclei purified from the crown gall tissues, was analyzed by Southern blot hybridization to specific fragments of the T-region of the octopine plasmid pTi ACH5. In addition total RNA isolated from the same lines grown as callus tissue on solid agar, was analyzed for T-DNA specific transcripts. The results show that all of the T-DNA is trancribed although different segments are transcribed to significantly different extents. Roughly the same hybridization patterns was found for nuclear and polysomal poly-A+ and poly-A− RNA. The transcription pattern was found to be different for cells in the stationary phase of growth compared with actively growing cells.

The conserved part of the T-region in Ti-plasmids expresses four proteins in bacteria.

The T‐region of Ti‐plasmids expresses four proteins (mol. wts. 74,000, 49,000, 28,000 and 27,000) in Escherichia coli minicells. Promoter activities are determined by sequences within the T‐region, and the protein‐coding regions map in that part of the T‐region which is highly conserved in octopine and nopaline plasmids and which is responsible for shoot and root inhibition when expressed in plant cells. Three of the regions expressed in bacteria correlate with three regions which are transcribed in transformed plant cells; the fourth protein‐coding region has no corresponding transcript in plants. At least three of the proteins synthesized in E. coli minicells are also expressed in cell‐free systems prepared from E. coli and from Agrobacterium tumefaciens; the fourth protein (mol. wt. 49,000) is poorly expressed in both cell‐free extracts. The possibility is discussed that the same genes are expressed in Agrobacteria and in transformed plant cells and that in both cases the gene products mediate growth regulatory effects to non‐transformed plant cells.

Native acridone synthases I and II from Ruta graveolens L. form homodimers

Acridone synthase II cDNA was cloned from irradiated cell suspension cultures of Ruta graveolens L. and expressed in Escherichia coli. The translated polypeptide of M r 42 681 revealed a high degree of similarity to heterologous chalcone and stilbene synthases (70–75%), and the sequence was 94% identical to that of acridone synthase I cloned previously from elicited Ruta cells. Highly active recombinant acridone synthases I and II were purified to apparent homogeneity by a four‐step purification protocol, and the affinities to N‐methylanthraniloyl‐CoA and malonyl‐CoA were determined. The molecular mass of acridone synthase II was estimated from size exclusion chromatography on a Fractogel EMD BioSEC (S) column at about 45 kDa, as compared to a mass of 44±3 kDa found for the acridone synthase I on Superdex 75. Nevertheless, the sedimentation analysis by ultracentrifugation revealed molecular masses of 81±4 kDa for both acridone synthases. It is proposed, therefore, that the acridone synthases of Ruta graveolens are typical homodimeric plant polyketide synthases.

Mapping of the protein-coding regions of Rhizobium meliloti common nodulation genes.

An 8.5‐kb EcoRI fragment containing the common nod region of the megaplasmid pRme41b of Rhizobium meliloti was recloned in plasmids of Escherichia coli, and a detailed restriction map was established. The region can express at least eight proteins in E. coli minicells and in an in vitro transcription/translation system, prepared from E. coli. Protein coding regions were determined by subcloning of restriction fragments, deletion mutations and by transposon mutagenesis. The coding regions for at least three polypeptide chains (mol. wts. 23 000, 28 500 and 44 000) were mapped on a 3.3‐kb nod gene cluster. The 44 000 mol. wt. protein is expressed from a nod region, which is highly conserved in two Rhizobium species. The protein map of the 8.5‐kb fragment was correlated to a map of insertion mutations with Nod‐ and Fix‐ phenotypes. The data suggest that the proteins encoded by the nod gene cluster may be involved in early steps of the nodulation process. Nod+ Fix‐ symbiotic mutations were localized in the coding region for a 33 000 mol. wt. protein, suggesting that this polypeptide might be a fix gene product.

Hybridization selection and translation of T-DNA encoded mRNAs from octopine tumors

SummaryTo characterize the functions of T-DNA derived transcripts in plant tumor cells, we isolated these RNAs by a highly sensitive hybridization selection procedure and investigated whether they are translatable into proteins in wheat germ extracts. Results with two independent tobacco cell lines transformed with octopine plasmids pTiA6 and pTiB6S3 show that the cells contained at least three translatable T-DNA derived mRNAs. Each represented 0.0001% or less of the total mRNA activity in polyribosomal RNA. All were detected in polyadenylated as well as in nonpolyadenylated RNA fractions, and translation was inhibited by the cap analogue pm7G. Two of the proteins were encoded close to the right and the left end of the T-DNA (proteins Mr 39,000 and 14,000, respectively); the third protein was derived from the middle of the T-DNA. The results indicate that genetic manipulation of plants by Agrobacteria involves transfer of several genes which are expressed into proteins in eucaryotic cells.

cDNA for S-adenosyl-L-homocysteine hydrolase from Catharanthus roseus.

The majority of the Met synthesized in plants is utilized for methylations performed with S-adenosyl-L-Met. These reactions play a major role in the modification of a large variety of acceptor molecules, such as lipids, polysaccharides, nucleic acids, proteins, and secondary plant products (reviewed by Giovanelli, 1987). The hydrolysis of SAH to adenosine and L-homocysteine by SAH hydrolase is one of the steps in the regeneration of S-adenosyl-L-Met. In cell cultures of Catharanthus roseus, a change from MX growth medium to a solution of 8% Suc induces the expression of many proteins (Vetter et al., 1992; our Ünpublished results). In a differential screen for induced mRNAs we identified a cDNA coding for a protein of more than 90% identity with the two SAH hydrolases known from plants (Table I). The similarity to the enzymes from vertebrates (human, AC P23526; rat, AC M15185), Caenorhabditis elegans (AC P27604), Dictyostelium discoideum (AC P10819), and the bacterium Rhodobacter capsulatus (AC M80630) was less pronounced (60-63%). SAH hydrolases are highly conserved enzymes (Sganga et al., 1992), and two consensus patterns have been defined previously (Bairoch, 1992). The sequences of the plant SAH hydrolases suggest that the consensus patterns should be extended to allow additional conservative amino acid replacements in the two motifs. The C. roseus protein deviates by one exchange from consensus pattern 1 (C-N-I-F-S-T-Q-EXaa-A-A-A-A-I-A instead of C-N-I-F-S-T-Q-@ or NI-XaaA-A-A-A-I-A), and the parsley protein differs in two amino acids from consensus pattern 2 (G-K-V-A-L-1-Xaa-G-Y-G-DV-G-K-G instead of G-K-V-A-y-y-Xaa-G-Y-G-D-V-G-K-G). In view of the high overall conservation of SAH hydrolases, it is interesting that there are distinct size differences. The three plant enzymes contain 485 amino acids, whereas the proteins from vertebrates (man, rat), D. discoideum, or C. elegans are smaller by about 50 amino acids (430-437 residues). Apart from a shorter NH2-terminal region, they lack a specific stretch of 40 amino acids (positions 150-190 in the plant enzymes). It seems possible that the different subunit composition of the enzymes may be one of the reasons (see Ogawa et al., 1987, for overview). The enzyme from the bacterium R. capsulatus appears to be more closely related to Table 1. Characteristics of CRSAHH from C. meus

Ti-Plasmids: Genetic Engineering of Plants

Crown gall is a neoplastic disease of most dicotyledonous plants and is caused by the soil bacterium Agrobacterium tumefaciens. A large extra-chromosomal plasmid in these bacteria was found to be responsible for its tumor-inducing capacity and was, therefore, called Ti-plasmid (1). Bacteria-free crown gall cells can be cultured in the absence of phytohormones and this hormone-independent growth defines tumor cells in plants (2). Sterile tumor tissues have been shown to contain a DNA segment (called T-DNA) which is homologous and colinear with a defined fragment of the Ti-plasmid, and it is covalently linked to plant DNA (3–9). The T-DNA has been localized in the nucleus (10, 11) and is directly responsible for the hormone-independent growth of the tumor cells. It is also responsible for the synthesis of low molecular weight compounds, called opines, which are not found in normal plant tissue. The opine produced defines crown galls as octopine, nopaline or agropine type tumors (12). Opines can be utilized by A. tumefaciens selectively as sources for carbon, nitrogen and energy, and, thus, the interaction between these bacteria and plants can be seen as a special parasitic relationship which benefits the bacteria (4).

The use of Ti-plasmids for the genetic engineering of plants

Crown gall is a neoplastic disease of many dicotyledonous plants and is caused by the soil bacterium Agrobacterium tumefaciens. A large extrachromosomal plasmid in these bacteria was found to be responsible for its tumor-inducing capacity and was, therefore, called Ti-plasmid [36]. Bacteria-free crown gall cells can be cultured in the absence of phytohormones like auxines and cytokinines, and this hormone-independent growth defines tumorous growth in plants [1]. Sterile tumor tissues have been shown to contain a DNA segment (called T-DNA) which is homologous and colinear with a defined fragment of the Ti-plasmid, and it is covalently linked to plant DNA [4,23,29,33,34,44,46]. The T-DNA has been localized in the nucleus [5,37] and is directly responsible for the hormone-independent growth of the cells. It is also responsible for the synthesis of a number of low molecular weight compounds, called opines, which are not found in normal plant tissues. The opine produced defines crown galls as octopine, nopaline, or agropine type tumors [15](Fig. 1). Opines can be utilized by Agrobacterium tumefaciens selectively as sources for carbon, nitrogen, and energy; thus, the interaction between these bacteria and plants can be seen as a special parasitic relationship which benefits the bacteria [29].