Isolation of a cDNA encoding a granule-bound 152-kilodalton starch-branching enzyme in wheat.

Screening of a wheat (Triticum aestivum) cDNA library for starch-branching enzyme I (SBEI) genes combined with 5'-rapid amplification of cDNA ends resulted in isolation of a 4,563-bp composite cDNA, Sbe1c. Based on sequence alignment to characterized SBEI cDNA clones isolated from plants, the SBEIc predicted from the cDNA sequence was produced with a transit peptide directing the polypeptide into plastids. Furthermore, the predicted mature form of SBEIc was much larger (152 kD) than previously characterized plant SBEI (80-100 kD) and contained a partial duplication of SBEI sequences. The first SBEI domain showed high amino acid similarity to a 74-kD wheat SBEI-like protein that is inactive as a branching enzyme when expressed in Escherichia coli. The second SBEI domain on SBEIc was identical in sequence to a functional 87-kD SBEI produced in the wheat endosperm. Immunoblot analysis of proteins produced in developing wheat kernels demonstrated that the 152-kD SBEIc was, in contrast to the 87- to 88-kD SBEI, preferentially associated with the starch granules. Proteins similar in size and recognized by wheat SBEI antibodies were also present in Triticum monococcum, Triticum tauschii, and Triticum turgidum subsp. durum.

Starch-branching enzymes preferentially associated with A-type starch granules in wheat endosperm.

Two starch granule-bound proteins (SGP), SGP-140 and SGP-145, were preferentially associated with A-type starch granules (>10 microm) in developing and mature wheat (Triticum aestivum) kernels. Immunoblotting and N-terminal sequencing suggested that the two proteins were different variants of SBEIc, a 152-kD isoform of wheat starch-branching enzyme. Both SGP-140 and SGP-145 were localized to the endosperm starch granules but were not found in the endosperm soluble fraction or pericarp starch granules younger than 15 d post anthesis (DPA). Small-size starch granules (<10 microm) initiated before 15 DPA incorporated SGP-140 and SGP-145 throughout endosperm development and grew into full-size A-type starch granules (>10 microm). In contrast, small-size starch granules harvested after 15 DPA contained only low amounts of SGP-140 and SGP-145 and developed mainly into B-type starch granules (<10 microm). Polypeptides of similar mass and immunologically related to SGP-140 and/or SGP-145 were also preferentially incorporated into A-type starch granules of barley (Hordeum vulgare), rye (Secale cereale), and triticale (x Triticosecale Wittmack) endosperm, which like wheat endosperm have a bimodal starch granule size distribution.

A starch-branching enzyme gene in wheat produces alternatively spliced transcripts.

A wheat gene, denoted Sbe1, encoding a type I starch-branching enzyme (SBEI) was isolated from a genomic library and shown to comprise 14 exons distributed over a 5.7 kb DNA region. Analyses of kernel RNA by 5' rapid amplification of cDNA ends (5'-RACE) and reverse transcription-polymerase chain reaction (RT-PCR) demonstrated a considerable sequence variation at the 5' ends of SBEI gene transcripts. DNA sequence alignments between the 5'-RACE products and the Sbe1 genomic DNA indicated that the first two exons and first intron were differentially processed to generate three classes of the mature transcript. One form of the SBEI gene transcript in 12-day old kernels contained the exon I+II+III combination at the 5' end, whereas other forms differed by inclusion of intron 1 or exclusion of exon II sequences. RT-PCR analysis of Sbe1-uidA::nptII chimeric mRNA produced in transgenic wheat cultured cells confirmed that the isolated Sbe1 was able to produce all three forms of SBEI gene transcripts by alternative splicing of the primary mRNA. The variants of processed Sbe1 mRNA were potentially translated into N-terminal variants of the SBEI precursor with different transit peptide sequences.

Molecular cloning and expression analysis of peroxidase genes from wheat.

A PCR-based screening approach was used to isolate genomic clones from wheat encoding peroxidase isozymes. Three complete genes (pox1, pox2 and pox4) and one truncated gene (pox3) were characterized. The nucleotide sequences predicted mature proteins of 31 kDa, in which all the highly conserved motifs of secreted plant peroxidases were preserved. The coding regions showed 73-83% DNA sequence identity, with the highest level of similarity noted for the tandemly oriented pox2 and pox3. Expression of respective pox genes in various tissues of wheat was assessed by the RT-PCR technique, which showed that all four genes are active. The primary pox1 mRNA was spliced to remove three introns, whereas processing of the other pox transcripts involved only two intervening sequences. Splicing occurred at consensus GU/AG splice sites except for the first introns of pox1, pox2 and pox4 transcripts, where processing took place at unusual GC donor sites. The RNA analysis suggested that the pox1, pox2 and pox4 genes are predominantly expressed in roots. Lower levels of expression were found for pox4 and pox3 in leaves. Infection of wheat by the powdery mildew fungus selectively induced expression of pox2 in leaves.

Analysis of a desiccation and ABA-responsive promoter isolated from the resurrection plant Craterostigma plantagineum.

The resurrection plant Craterostigma plantagineum can recover from severe desiccation within 24 h of contact with water, and it is used as a model system to analyse desiccation tolerance in higher plants. During drying or ABA treatment a specific set of transcripts accumulates rapidly in leaves and other tissues. In order to study transcriptional mechanisms of stress-induced gene expression one gene (CDeT27-45) was selected for promoter analysis. Chimeric gene fusions were constructed of the CDeT27-45 promoter and beta-glucuronidase or luciferase. These constructs were tested in a homologous transient expression system which allowed the identification of promoter elements conferring ABA inducibility. By introducing the chimeric gene fusions into tobacco via Agrobacterium-mediated transformation we found that the promoter activity is under strict tissue-specific and developmental control. In tobacco the promoter was only active in developing embryos and in mature pollen grains-two tissues which are naturally desiccation tolerant in tobacco. The specific temporal expression pattern was attributed to particular 5' upstream sequences. The promoter analysis presented here should allow the separation of important regulatory components as a first step in dissecting events in the signal transduction chain.

Horizontal gene transfer of the Escherichia coli pap and prs pili operons as a mechanism for the development of tissue-specific adhesive properties.

Escherichia coli strains bind to Gal alpha 1-4Gal-containing glycolipids via P pili-associated G-adhesins. Three functional classes of adhesins with different binding specificities are encoded by conserved G-alleles. We suggest that the Class I papG-allele of strain J96 is a novel acquisition possibly introduced via horizontal gene transfer into one of the two P pili gene clusters carried by this strain. Closely related strains in the ECOR collection of natural E. coli isolates carry either a Class II or a Class III G-adhesin. Data indicate that genetic exchanges involving either entire pap or prs gene clusters or individual pap/prs genes have occurred. We propose that the retention and spread of pap/prs DNA among E. coli is the result of selection pressure exerted by mammalian intestinal isoreceptors.

Processed mRNA with differential stability in the regulation of E. coli pilin gene expression.

E. coli expressing the papA-I genes produce pili that mediate specific adhesion to mammalian cells. We show that the major pilus subunit gene, papA, is part of a polycistronic transcriptional unit subject to specific posttranscriptional processing. A primary transcript also encoding the papB regulatory gene product is endonucleolytically cleaved, resulting in the rapid decay of the papB-encoding 5' half of the mRNA, whereas the papA-encoding 3' half remains as a quite stable transcript. Processing and differential mRNA stability thereby result in accumulation of mRNAs encoding only the major pilus subunit. A sequence immediately downstream of the papA coding region may serve as a stability determinant for the papA transcript and concomitantly attenuate read-through transcription into the minor pilus subunit gene papH. This suggests that differential expression of genes within an operon may include endo- and exonucleolytic processing of the mRNA.

Nucleotide sequence, regulation and functional analysis of the papC gene required for cell surface localization of Pap pili of uropathogenic Escherichia coli.

The papC gene of uropathogenic Escherichia coli is required for the formation of digalactoside-binding Pap pili. papC forms part of an operon wherein the regulatory gene papB, the major pilin gene papA, a minor pilin-like gene papH, and papC are co-transcribed. Furthermore, the extent of PapC synthesis was found to affect the number of pili expressed on the cell surface. The DNA sequence of the papC gene is presented and its deduced amino acid sequence is compared to that of the FaeD protein encoded by the K88 pili gene cluster. The PapC protein was localized to the E. coli outer membrane where it may form a trans-membrane channel through which pilin subunits are surface localized.

Biogenesis of E. coli Pap pili: papH, a minor pilin subunit involved in cell anchoring and length modulation.

The biogenesis of Escherichia coli Pap pili, encoded by the pap gene cluster, was studied. A novel gene, papH, was identified and found to encode a weakly expressed pilin-like protein. PapH was dispensable for digalactoside-specific binding and for formation of Pap pili. However, in papH deletion mutants 50%-70% of total pilus antigen was found free of the cells. We present evidence showing coregulation of papH and the adjacent gene, papA, which encodes the major pilin subunit. A decrease in the PapA to PapH ratio resulted in a large fraction of cells producing shortened pili, whereas overproduction of PapA relative to PapH resulted in cells with lengthened pili. The data show that PapH has roles in anchoring the pilus to the cell and in modulating pilus length.

Fimbriation and P-antigen recognition of Escherichia coli strains harbouring mutated recombinant plasmids encoding fimbrial adhesins of the uropathogenic E. coli strain KS71.

Deletion mutants of recombinant plasmids encoding the KS71B fimbrial antigens of the uropathogenic Escherichia coli strain KS71 (O4:K12) were constructed. The effects of these mutations were tested by transforming the mutated plasmids into non-fimbriated E. coli HB101 cells and testing the transformants for fimbriation and haemagglutination. A deletion transcriptionally upstream from the fimbrial subunit gene increased the expression of KS71B fimbriae. Deletion of the fimbrial subunit gene resulted in non-fimbriated but haemagglutinating transformants, whereas a deletion 6 kb transcriptionally downstream from the subunit gene resulted in non-haemagglutinating but fimbriate transformants, indicating that fimbriation and haemagglutination were genetically separable. We also present evidence suggesting that the fimbrillin and haemagglutinin are physically associated in the wild-type KS71 strain.

Transcriptional activation of a pap pilus virulence operon from uropathogenic Escherichia coli.

A gene cluster mediating production of pili in uropathogenic Escherichia coli was analysed with respect to regulation of pili synthesis. Two cistrons, papB and papI, were localized upstream of the major pilus subunit gene, papA. The papI-papB-papA region was characterized by nucleotide sequencing and by transcriptional analysis. The papA gene was primarily represented by an 800 nucleotide long transcript but was also co-transcribed with papB as a less abundant 1300 nucleotide long mRNA. Both transcripts presumably terminated at the same site downstream of the papA coding sequence. The weakly expressed papI gene was transcribed in the opposite direction to that of papB and papA. Studies with lacZ operon fusions showed that the papB gene encoded a trans-active effector required for papA transcription. Similarly, the papI gene stimulated papB transcription in trans. Furthermore, full expression of papA was cis dependent upon the papI-papB region. Transcription of the papB gene was shown to be dependent upon cAMP and its receptor protein. A binding site for the cAMP-CRP complex was postulated in the DNA sequence upstream of the papB promoter.

Globoside-specific adhesins of uropathogenic Escherichia coli are encoded by similar trans-complementable gene clusters.

Uropathogenic Escherichia coli frequently express globoside-specific adhesins, shown to mediate binding to uroepithelial cells. For one gene cluster pap, it recently has been demonstrated that globoside binding is not dependent on expression of the pilus subunit gene papA. Instead, two other pap genes papF and papG are specifically required for globoside binding (F. P. Lindberg et al., EMBO J. 3:1167-1173, 1984). By restriction enzyme mapping, DNA hybridization, DNA sequencing, and protein expression in minicells, we show that three gene clusters encoding globoside binding have a very similar structure and gene organization, although they were cloned from different E. coli isolates. Major differences between the adhesin clones were restricted to the central part of the pilin gene (papA) and to one of the two adhesin gene (papG). The three functional units required for biogenesis of globoside-binding pili, i.e., pilin synthesis, pilin export, and pilin assembly, as well as expression of adhesion function, were all trans complementable among the gene clusters.

Adhesion to human cells by Escherichia coli lacking the major subunit of a digalactoside-specific pilus-adhesin.

Pathogenic bacteria frequently possess pili with specific binding properties that allow them to attach to epithelial tissue. In Escherichia coli, the pili associated with pyelonephritis (Pap pili) bind to digalactoside-containing glycolipids on the uroepithelium. Transposon-insertion mutants and deletion mutants of the cloned genetic determinant encoding synthesis of such digalactoside-binding Pap pili have been studied in E. coli K-12. Mutants that completely lack synthesis of the major Pap pili subunit protein, the papA gene product, and thereby no longer produce pili were shown to retain the binding specificity of intact Pap pili. Reduced expression of some of the remaining pap genes, presumably due to polarity effects from papA::Tn5 insertions, was circumvented by the use of a copy-number mutant plasmid vector. Derivatives carrying the papA-D genes produced Pap pili but did not bind to human cells. The products of the genes papE-G are essential for digalactoside-specific hemagglutination and for attachment to urinary bladder cells. The papC and papD genes presumably aid in surface localization and/or polymerization of the pili-adhesin subunits and are required for expression of pili as well as of the binding properties. Serological evidence is presented that suggests that a minor pilus component(s), presumably produced by the papE, -F, or -G gene, is the actual binding moiety in the digalactoside-specific interaction of Pap pilus-adhesin.

Mutations in E coli cistrons affecting adhesion to human cells do not abolish Pap pili fiber formation.

A chromosomal DNA fragment which mediates Pap (pili associated with pyelonephritis) pili formation, mannose-resistant hemagglutination ( MRHA ) and binding to uroepithelial cells has been isolated from the uropathogenic Escherichia coli clinical isolate J96 , and genetically studied. Analysis of polypeptides expressed by the Pap DNA led to detection of a number of polypeptides ranging in mol. wt. from 13 000 to 81 000 daltons. The gene order and transcriptional orientation for four of the corresponding cistrons was: 13 000 ( papB ) 19 500 ( papA , structural gene for the Pap pilus subunit), 81 000 ( papC ) and 28 500 ( papD ). Analyses of a lacZ- papA gene fusion located a promoter upstream from papA within the cloned DNA. Transposon Tn5 insertions in any of these four cistrons decreased or eliminated Pap pili formation. A number of transposon Tn5 mutations were identified in a region distal to papD that expressed normal levels of the papA protein on the cell surface in the form of recognizable pili structures but did not agglutinate human erythrocytes or adhere to uroepithelial cells. This region expressed polypeptides of 15 000, 24 000, 26 000 and 35 000 daltons. This finding shows that Pap pili formation and binding properties can be genetically dissociated.

Nucleotide sequence of the papA gene encoding the Pap pilus subunit of human uropathogenic Escherichia coli.

The papA gene of the uropathogenic strain Escherichia coli J96, coding for the Pap pili subunit, was subjected to DNA sequencing, and found to code for an 185-amino acid-long polypeptide with a 22-amino acid-long signal peptide. Here we present the primary sequence, the hydrophilicity profile, and the predicted polypeptide secondary structure of the Pap pili subunit.

Genetics of digalactoside-binding adhesin from a uropathogenic Escherichia coli strain.

The uropathogenic strain Escherichia coli J96 mediates mannose-resistant hemagglutination owing to production of a digalactoside-binding adhesin. A cosmid clone from this strain has been isolated that, when harbored in E. coli K-12, expressed Pap pili and this adhesin (R. Hull et al., Infect. Immun. 33:933-938, 1981). By transposon mutagenesis and by the construction of a number of hybrid plasmid derivatives, we have demonstrated that about 8.5 kilobases of DNA is required to generate a mannose-resistant hemagglutination-positive phenotype in E. coli K-12 strain P678-54. The structural gene for the Pap pili monomer, papA, has been identified and mapped close to the promotor-proximal end of the Pap operon. Although strain P678-54 that harbored a Tn5 insertion within papA showed a mannose-resistant hemagglutination-positive phenotype, it was negative in a competitive enzyme-linked immunosorbent assay with anti-Pap pilus serum. This could mean that a Pap adhesin is encoded by a region on the Pap operon that is distinct from papA.