DNA sequence analysis of the 4.4 kilobases (kb) Eco RI fragment 14 from T-DNA of Agrobacterium tumefaciens C58 revealed three open reading frames. One of them (945 bp) was supposed to encode the transcript e, the function of which has not been identified to date. Furthermore, a so far undescribed open reading frame (1035 bp) was identified, located in the centre of the Eco RI fragment 14 and termed gene f. The third open reading frame encoded the carboxy-terminal part of the agrocinopine synthase (Acs). The gene e-encoded protein showed significant homologies to the gene products of the Agrobacterium rhizogenes rolB gene and the Agrobacterium tumefaciens gene 5. Both gene products are supposed to regulate the plant's reaction on auxin. Depending on the plant species tested, Agrobacterium strains carrying mutations in gene e induced only small or almost no detectable crown gall tumours. According to these mutational studies and the protein homologies observed, the gene e product is suggested to be involved in tumour formation. Infection of several plant species with Agrobacterium carrying a mutated gene f, as well as expression of the gene f in transgenic tobacco plants did not lead to visible morphological changes. Therefore, in contrast to gene e, the gene f seems not to be essential for tumour formation. In order to study whether gene f is an active gene, its expression in agrobacteria and plants was monitored by translational lacZ fusion. In planta, the putative gene f-promoter mediates a tissue-specific expression pattern. Although gene f was expressed in free-living agrobacteria as well as in transgenic plants, the function of the f locus remained unclear. DNA homology studies with the f gene region revealed a mosaic-like DNA structure, indicating that this locus might be the result of genetic exchanges between different Agrobacterium strains during evolution.

Strains of Rhizobium leguminosarum (R. l.) biovar viciae containing pss mutations fail to make the acidic exopolysaccharides (EPS) and are unable to nodulate peas. It was found that they also failed to nodulate Vicia hirsuta, another host of this biovar. When peas were co-inoculated with pss mutant derivatives of a strain of R.l. bv viciae containing a sym plasmid plus a cured strain lacking a sym plasmid (and which is thus Nod-, but for different reasons) but which makes the acidic EPS, normal numbers of nodules were formed, the majority of which failed to fix nitrogen (the occasional Fix+ nodules were pressumably induced by strains that arose as a result of genetic exchange between cells of the two inoculants in the rhizosphere). Bacteria from the Fix- nodules contained, exclusively, the strain lacking its sym plasmid. When pss mutant strains were co-inoculated with a Nod- strain with a mutation in the regulatory gene nodD (which is on the sym plasmid pRL1JI), normal numbers of Fix+ nodules were formed, all of which were occupiced solely by the nodD mutant strain. Since a mutation in nodD abolishes activation of other nod genes required for early stages of infection, these nod genes appear to be dispensable for subsequent stages in nodule development. Recombinant plasmids, containing cloned pss genes, overcame the inhibitory effects of psi, a gene which when cloned in the plasmid vector pKT230, inhibits both EPS production and nodulation ability. Determination of the sequence of the pss DNA showed that one, or perhaps two, genes are required for correcting strains that either carry pss mutations or contain multi-copy psi. The predicted polypeptide product of one of the pss genes had a hydrophobic aminoterminal region, suggesting that it may be located in the membrane. Since the psi gene product may also be associated with the bacterial membrane, the products of psi and pss may interact with each other.

The complete nucleotide sequence of the intergenic region between the 25 S and 18 S wheat ribosomal RNA genes has been determined from a 4.6 kb EcoRI-BamHI fragment (1 kb = 103 bases or base-pairs) subcloned from the plasmid pTa71. Within this subclone the intergenic DNA is flanked by the 3′ end of the 25 S and the 5′ end of the 18 S ribosomal RNA sequences. Four repeat families are present within the intergenic region. The major repeat family A, consists of 12 direct repeat units of 135 or 136 base-pairs (bp) flanked by diverged truncated copies. Within each A repeat a subrepeat structure has been revealed. Family B, which is localized to the 5′ side of the A repeats, contains three repeat units, one of 152 bp, the second of 150 bp and a truncated unit of 107 bp. Family C, which is localized in the transcribed rRNA precursor, consists of two direct repeat units of 172 and 174 bp and possesses some short subrepeat motifs. The C repeats may have evolved by and diverged from one another by the insertion of short transposable sequences. Family D consists of two direct repeat units of 30 bp located 5′ to the start of transcription. Statistical analysis of repeat family A showed that there is a significant association between the similarity of any two repeat units and their distance apart in the array. The near identity of members of the A family is maintained presumably by processes such as unequal crossing over and gene conversion, but the members at each end of the array show more divergence. Sequence motifs in the A and C repeat families and in other regions including the 5′ end of 18 S RNA are related, implying that much of the intergenic DNA may have evolved from a few short ancestral sequences. The B and D repeats or their equivalent are not found in a maize ribosomal DNA repeat unit. The DNA in the external transcribed spacer DNA 5′ to the 18 S RNA sequence is longer in wheat than in maize. This is due principally to two duplications and insertion of a sequence with dyad symmetry in the wheat gene.

An earlier report (Baulcombe, D. C., and Buffard, D. (1983) Planta 157, 493-501) described the isolation of cDNA clones from mRNAs which are produced in increased amounts when aleurone layers of wheat are treated with gibberellic acid. It is shown here that for one of those cDNAs (2473) the change in level of mRNA in aleurone parallels the change in level of alpha-amylase mRNA. This result was obtained in experiments where the level of gibberellic acid was varied and also when the mRNA was isolated from wheat genotypes which varied in ability to respond to gibberellic acid. In contrast to this, the pattern of 2437 mRNA accumulation in immature grains and in leaf tissue was quite distinct from the pattern of alpha-amylase mRNA accumulation. Analysis of wheat DNA showed that the 2437 mRNA is encoded by a small family of genes located on the short arm of the group 6 chromosomes. One member of this gene family was cloned and sequenced. The coding sequence is interrupted by eight introns and encodes a protein of Mr 55,433. By using hybridization probes from the 5' exon in an S1 nuclease protection assay it was shown that the 2437 mRNA was produced in aleurones (coordinately with alpha-amylase) and in immature grains (not coordinately with alpha-amylase). However, sequence comparison of 1 kilobase of the 5'-flanking region with the sequence of alpha-amylase genes provided no indication of the regulatory elements which would be active in aleurone cells. The protein sequence deduced from the gene sequence has extensive homology with the yeast carboxypeptidase Y, especially in the active site and substrate binding regions. This homology is greater than with the carboxypeptidase I from barley. It is suggested therefore that there are several types of carboxypeptidase encoded in the cereal genome. The sequence of the 2437 protein would represent one of these types and the barley carboxypeptidase I, another.

We have isolated and characterized a gene encoding the phytochrome polypeptide of Avena. Based on nucleotide sequence identity with previously sequenced cDNA clones this gene is designated as type 3 (phy3). The gene is about 5.9 kb long with six exons and five introns, one each of the latter in the 5 ' and 3 ' -untranslated regions. The largest exon encodes the entire 74-kDa, chromophore-bearing, N-terminal domain of the photoreceptor postulated to be directly involved in its mechanism of action. The transcription start point, identified by mung-bean nuclease digestion, is located 24 to 35 bp downstream from a tandem TATA box. Sequence elements homologous to a number of motifs implicated as upstream regulatory elements in other genes are present in the 5'-flanking DNA of phy3. Particularly intriguing are three elements at positions -140, -470 and -650. These elements share homology with the ‘GT’ motif postulated to be a component of the light-regulatory element of genes encoding the small subunit of ribulose bisphosphate carboxylase.

Ribosomal RNA genes are localized at chromosomal sites termed nucleolus organizers because nucleoli form around transcribed ribosomal RNA genes. The relative activities of arrays of ribosomal RNA genes can be estimated cytologically by comparing the sizes of nucleoli in the same cell. Also, active nucleolus organizers give rise to visible constrictions in metaphase chromosomes whereas inactive nucleolus organizers do not. With these assays the differential expression of nucleolus organizers and ribosomal RNA genes has been observed frequently, especially in interspecies hybrids. Studies on wheat have revealed that differences in gene expression are associated with differences in chromatin structure and cytosine methylation. Active loci have higher proportions of their genes decondensed and accessible to proteins and also higher proportions with a non-methylated cytosine residue at a CCGG site in the region of the promoter. Short, related sequences with dyad symmetry have been noted between —140 and —70 base pairs from where transcription is initiated in a wheat ribosomal RNA gene. Similar sequences are reiterated upstream of the promoter over 2000 base pairs. From comparison of this gene structure with that of Xenopus ribosomal RNA genes it can be concluded that these short sequences are likely to act as enhancers of transcription by binding to specific regulatory proteins that function to stimulate the attachment of polymerase I complexes. Differential expression of arrays of ribosomal RNA genes results when genes have different numbers of enhancer repeats or a higher affinity for the regulatory protein(s). This model to explain differential gene expression and the origins of genetic variation affecting ribosomal RNA gene expression are discussed.

A cDNA clone (pTag1436) carrying a complete coding sequence for a γ-gliadin polypeptide has been identified and sequenced. By hybridisation to size fractionated poly A(+) RNA from wheat nullisomic-tetrasomic lines, homologous transcripts from the Gli-A1, Gli-B1 and Gli-D1 loci were identified. These mRNAs differed from those complementary to a low molecular weight (LMW) glutenin cDNA clone. Hybridization of pTag1436 to digested wheat DNA produced a pattern of fragments unrelated to that obtained using a LMW glutenin cDNA probe. These results indicate that the γ-gliadin and LMW glutenin families, although both located at the Gli-1 loci, are distinct by hybridisation.

Bacillus thuringiensis subsp. kurstaki HD-73 produces a crystal protein which is lethal to many lepidopteran larvae. The gene encoding this crystal protein has been isolated from a 75-kb plasmid and engineered into a recombinant Escherichia coli plasmid for analysis. The complete nucleotide sequences of the coding region and 387-bp 5' and 376-bp 3' to the coding region have been determined. The 3537-bp of the coding region specify a protein of Mr 133 330. The full-length gene and several 3' -truncated derivatives of the gene were examined in both E. coli and in an E. coli minicell-expression system to determine if the carboxy end of the protein is essential for toxicity. The results presented here provide the primary structure of the crystal protein gene and show that the N-terminal 68-kDal peptide is toxic, but at a lower level than the full-length gene product.

Cloned cDNA and genomic sequences have been analyzed to deduce the amino acid sequence of phytochrome from etiolated Avena. Restriction endonuclease site polymorphism between clones indicates that at least four phytochrome genes are expressed in this tissue. Sequence analysis of two complete and one partial coding region shows approximately 98% homology at both the nucleotide and amino acid levels, with the majority of amino acid changes being conservative. High sequence homology is also found in the 5'-untranslated region but significant divergence occurs in the 3'-untranslated region. The phytochrome polypeptides are 1128 amino acid residues long corresponding to a molecular mass of 125 kdaltons. The known protein sequence at the chromophore attachment site occurs only once in the polypeptide, establishing that phytochrome has a single chromophore per monomer covalently linked to Cys-321. Computer analyses of the amino acid sequences have provided predictions regarding a number of structural features of the phytochrome molecule.

Recombinant phage lambda177.4 contains a gene for beta phaseolin, a major storage glycoprotein of French bean seed. A 3.8-kilobase Bgl II-BamHI fragment containing the entire 1700-base-pair coding region, together with 863 base pairs of 5' and 1226 base pairs of 3' flanking sequence, was inserted into the A66 Ti plasmid of Agrobacterium tumefaciens and used to transform tobacco. The level of phaseolin in the seeds of plants regenerated from cloned tissue was 1000-fold higher than in other tissues. The molecular weight of the phaseolin RNA transcript in tobacco seeds was identical to that found in bean seeds. The phaseolin protein in tobacco seed was glycosylated and appeared to undergo removal of the signal peptide. However, a large proportion of the phaseolin was cleaved into discrete peptides. These same peptides were formed as phaseolin was degraded during tobacco seed germination. The phaseolin gene appeared to be inserted as a single copy, and the proportion of phaseolin per genome copy in tobacco seeds (up to 3% of the total embryo proteins) resembled that in the bean seeds (40% of total seed protein, expressed from about 14 copies per diploid genome). Furthermore, the transplanted gene was turned on during tobacco seed development, and its protein product, phaseolin, was localized in the embryonic tissues. Finally, the phaseolin gene was inherited as a Mendelian dominant trait in tobacco.

A cloned DNA fragment from the maize allele Adhl-S3034 contains all of Mul, an insertion element involved in Robertson's Mutator activity. The element is 1367 base pairs (bp) long and is flanked by nine bp direct repeats of insertion site DNA. It has inverted terminal repeats of 215 and 213 bp showing 95% homology. Within the element are two direct repeats of 104 bp showing 96% homology. Four open reading frames (ORFs) were found, two in each DNA strand. Mul can be divided into two halves, each containing one terminal inverted repeat, an internal direct repeat, and two overlapping ORFs. The GC content of each half is high (70%), while that of a central 60 base portion of the element is low (26%). The central region contains the only sequence resembling the TAATA Goldberg and Hogness eukaryotic promoter signal. Multiple copies of DNA sequences related to Mul found in Mutator maize plants are generally similar in organization to the cloned element. A larger version containing a discrete 300 to 400 base pair insertion was found in some Mutator lines.

A recombinant plasmid (pcPvNGS-01) containing sequences related to glutamine synthetase (GS) has been identified from a cDNA library constructed from poly (A)+ RNA isolated from root nodules of Phaseolus vulgaris L. The identification of this recombinant relied on the observations that: (a) the clone hybridized strongly to purified GS mRNA; (b) in hybrid-select translation experiments, the clone selected mRNA that produced a polypeptide identical in molecular weight to purified GS subunits which was immunoprecipitated with anti-GS-antiserum; and (c) the translated nucleotide sequence of the cloned cDNA was homologous to a partial amino acid sequence of higher plant GS. The cloned cDNA hybridized to poly (A)+ RNA of different mobilities from leaves, roots, and nodules of P. vulgaris. In RNA "dot" blots washed at different stringencies, differences were observed both in the relative amounts of GS mRNA in different tissues and in the strength of their hybridization to the cDNA probe. The cloned probe hybridized to several fragments of restricted P. vulgaris DNA but not to DNA from Rhizobium phaseoli. These results suggest that GS is coded for by a small multigene family showing organ-specific expression.

A full-length cDNA clone of alfalfa mosaic virus (AMV) RNA3 was prepared and sequenced. The 2,037 base sequence contains two open reading frames of 903 and 666 nucleotides that code for a 32,400 dalton protein (32.4K protein) and the 24,380 dalton coat protein, respectively. A 5'-noncoding sequence of 240 bases preceeding the 32.4K protein contains homologous regions that may have a function in its translation. The intercistronic junction is 49 bases long, the last 36 bases representing the 5'-end of the subgenomic RNA4. The remaining 179 bases comprise the 3'-terminal noncoding sequence.

Sequences coding for the bean seed protein phaseolin were inserted into transferred DNA regions of tumor-inducing plasmids. Constructions were devised in which the coding region of phaseolin was fused in the correct reading frame with the coding region of octopine synthase and placed under the transcriptional control of the octopine synthase promoter. Other plasmids were prepared to permit expression of the phaseolin-encoding sequences from the flanking phaseolin promoter region. The RNA transcribed in sunflower cells transformed with these constructions was characterized by hybridization procedures, SI nuclease mapping, and by translation in vitro of extracted RNA. These tests showed that the genomic intervening sequences were correctly excised. Immunoreactive phaseolin polypeptides were detected by enzyme-linked immunosorbent assay and by antibody hybridization to electrophoretically separated protein extracts of sunflower tissues isolated from crown gall tumors and of transformed sunflower cells grown in tissue culture. These results demonstrate the expression of a plant gene after transfer to a taxonomically distinct botanical family.

The complete nucleotide sequence of the transferred region (T-DNA) of an octopine tumor inducing (Ti) plasmid fromAgrobacterium tumefaciens (pTi15955) has been determined. A total of 24 595 nucleotides extending approximately 900 bases to either side of the outermost, T-DNA boundaries was sequenced. Computer analysis of the sequenced portion of the Ti plasmid revealed that recognition sites for 72 restriction endonucleases are present in the DNA sequence at least once; no site forEcoK exists in this DNA sequence. Two imperfect 24 base repeats border the T-DNA sequence; the left starts at position 909 and the right ends at position 23 782, giving the T-DNA region a total length, of 22 874 nucleotides. Another two similar 24 base repeats lie within T-DNA and divide it, into three distinct domains: T-left (TL-DNA) 13 175 bp of apparently eukaryotic origin; T-center (TC-DNA) 1816 bp of prokaryotic origin; and T-right (TR-DNA) 7 883 bp of eukaryotic origin. The T-DNA contains nine reported transcripts, however, 26 open reading frames longer than 300 bases that start with an ATG initiation codon were found. Fourteen open reading frames are bounded by putative eukaryotic promoters, ribosome binding sites, and poly(A) addition sites and occur only in TL-and TR-DNAs. No open reading frames showing eukaryotic promoter sequences are located within the TC-DNA.