Search

A mutant of Arabidopsis with reduced internodal cell length, acaulis5 (acl5), has recently been shown to have reduced transcript levels of a gene for endoxyloglucan transferase, EXGT-A1 (Y. Hanzawa, T. Takahashi, Y. Komeda [1997] Plant J 12: 863-874). In the present study, we cloned genomic fragments of five members of the EXGT gene family, EXGT-A1, EXGT-A3, EXGT-A4, XTR2, and XTR3, and examined their expression in the wild type and in a series of acl mutants. In wild-type plants, the EXGT-A3 gene showed higher expression in lower internodes (internodes between nodes bearing axillary shoots) than in upper and young internodes, in which EXGT-A1 was highly expressed. EXGT-A4 was preferentially expressed in roots and XTR3 in siliques. The XTR2 gene was constitutively expressed. In acl1, acl3, and acl4 mutants, which have a severe defect in leaf expansion as well as in internode elongation, the EXGT-A1 gene showed reduced levels of expression before bolting of plants. In contrast, XTR3 was increased in these mutant seedlings. Reduction of EXGT-A1 expression was also detected after bolting of all acl mutants except acl2, whose growth defect is restricted to lower internodes. These results suggest the involvement of each EXGT in different aspects of organ development.

We previously described the preparation of a protein fraction that bound to the promoter region of the rRNA gene of Vicia faba. We now describe two additional protein fractions that interact with this DNA element. The three protein fractions have similar binding properties with the exception of their binding stabilities. DNase I footprinting experiments revealed that these protein fractions bind to a region that is located about 60 bp upstream from the site of initiation of transcription. From methylation interference experiments and gel-retardation assays with mutated oligonucleotides, it was evident that the protein fractions required the same residues for the formation of stable DNA-protein(s) complexes. However, the three protein fractions were chromatographically distinguishable and the stability of their binding to DNA was different. Although the functions of these fractions are still unknown, the fact that several protein fractions can bind to a narrow region near the site of initiation of transcription is of considerable interest in terms of the mechanism of transcription of the rRNA gene.

Male-associated DNA sequences were analyzed in a dioecious plant, Cannabis sativa L. (family: Moraceae), which is known to have sex chromosomes. DNA was isolated from male and female plants and subjected to random amplification of polymorphic DNA. Two out of 15 primers yielded fragments of 500 and 730 bp which were detected in all male plants but not in any of the female plants tested. These two DNA fragments were cloned and used as probes in gel blot analysis of genomic DNA. When the male and female DNAs were allowed to hybridize with the 500-bp probe, no differences in patterns were observed between male and female plants. By contrast, when these DNAs were allowed to hybridize with the 730-bp probe, much more intense bands specific to male plants were detected, in addition to less intense bands that were common to both sexes. The 730-bp DNA fragment was named MADC1 (male-associated DNA sequence in Cannabis sativa). The sequence of MADC1 did not include a long open reading frame and it exhibited no significant similarity to previously reported sequences.

We have isolated three HSP90-family genes from Arabidopsis: HSP81-1 which is heat-inducible, and HSP81-2 and -3 which are highly expressed under normal growth temperatures. Northern blot analysis and RNase protection analysis, using gene specific probes, showed that HSP81-2 and -3 mRNA were present in all tissues and abundant in roots, floral bud clusters, and flowers at 22 degrees C. A small amount of HSP81-1 mRNA was detected only in roots. In situ hybridization and histochemical analysis using transgenic plants carrying chimeric gene fusions, with an HSP81 promoter region fused to a beta-glucuronidase (GUS) gene, confirmed these results. At 22 degrees C, high GUS activity was observed in the root apical meristems, pollen and tapeta in HSP81-2::GUS and HSP81-3::GUS transgenic plants, while only branches of the root in HSP81-1::GUS transgenic plants expressed high GUS activity. After 2 hours of 35 degrees C treatment, extensively high GUS activity was observed in all tissues in HSP81-1::GUS transgenic plants, while elevated but tissue specific expression was observed in HSP81-2 and -3 transgenic plants. Exogenous application of various chemicals such as ABA, GA3, kinetin, IAA, NaCl, and mannitol revealed that 10 mM IAA and 0.1 M NaCl significantly enhanced the accumulation of HSP81-2 and -3 transcripts. Only a slight response to IAA was observed in HSP81-1 mRNA accumulation at 22 degrees C; the increase was possibly caused by a novel pathway other than heat-shock-response pathway.

Production of a functional beta-glucuronidase (GUS) protein was induced by exposure of exponentially growing yeast cells to heat shock after transformation of the GUS gene under the control of the promoter of the heat-shock gene, HSP18.2, from Arabidopsis. Yeast cyr and bcy mutations appeared to have essentially no effect.

Flagellar mutants in Escherichia coli region I were obtained by selection for resistance to the flagellotropic phage chi. F' elements carrying this region of the E. coli genome were then constructed. Stable merodiploid strains with a flagellar defect on the exogenate and another on the endogenote were prepared. These merodiploids yielded information on the complementation behavior of mutations in this region. Region I was shown to include at least six cistrons, flaV, flaK, flaL, flaM, flaS, and flaT. Mu-induced and deletion fla mutants were also isolated. By using these mutant strains, the transcriptional order was shown to be flaV-flaK-flaL-flaM-flaS-flaT. The definition of region I fla genes and their transcriptional relationships were confirmed by genetic tests with hybrid A phage carrying fla genes in this region.

Previous studies have defined 29 genes necessary for synthesis of the Escherichia coli flagellar apparatus. This study analyzed the transcriptional control of flagellar genes, using Mu d (Apr lac) phage to generate flagellar mutants by insertion. These mutants contained operon fusions of flagellar genes to the lac genes of the Mu d phage and allowed the measurement of flagellar operon expression by detection of beta-galactosidase activity. These fusion mutants expressed the enzyme activity constitutively, and an autogenous regulation mechanism was not revealed. Lambda transducing phages carrying these chromosomal fla-lac fusions were also isolated and used to examine the effect of different fla mutations on expression of each flagellar operon. The results showed that flagellar operons are divided into six classes; (class 1) the flbB operon, which controls all of the other flagellar operons; (class 2) the flaU and flbC operons, which are controlled by the flbB operon gene products and are not required for the expression of other Fla operons; (class 3) the flbA, flaG, flaD, flaN, flaB, and flaA operons, which are under flbB operon control and are required for the expression of other fla operons; (class4) the flaZ operon, which is controlled by the gene products of the group 1 and 3 operons and is required for hag transcription; (class 5) the mocha and flaS operons, which are controlled by the gene products of the group 1 and 3 operons; and (class 6) the hag operon. These results are discussed with respect to the possible assembly sequence of the fla gene products.

flaFIX, the structural gene for the periplasmic P ring of the flagellar basal body of Salmonella typhimurium, was cloned. Two gene products with apparent molecular weights of 38,000 and 40,000 were identified by minicell analysis. Data from pulse-chase and membrane fractionation experiments and data on the inhibitory effect of the proton ionophore carbonyl cyanide m-chlorophenylhydrazone all indicated that the 40-kilodalton protein was a precursor form which, after export across the cytoplasmic membrane accompanied by cleavage of a signal peptide, gave rise to the mature protein in the periplasm. The N-terminal amino acid sequence of the FlaFIX protein, predicted from the DNA sequence, conformed well to known signal peptide sequences. The results indicate that the P-ring protein of the basal body (unlike flagellin and possible some other external flagellar components) crosses the cytoplasmic membrane in a conventional signal peptide-dependent manner.

Specialized transducing lambda phages carrying the region III flagellar genes (fla) of Escherichia coli K-12 were isolated by a new method. A strain carrying both a cryptic lambda prophage near the his genes and a deletion of the attlambda gene was used as a starting strain. The lysogen of lambdacI857pga18-bio69 was isolated in which the prophage was integrated within the lambda cryptic genes by means of recombination with the residual lambda DNA. The strains with deletions starting within the prophage and ending in these fla genes were selected from among the heat-resistant survivors of the lysogen. They were then infected with heat-inducible and lysis-defective lambda phages and, thus, specialized transducing phage lines for hag and fla were obtained. High-frequency transfer lines of rare phages carrying the fla genes were isolated by inducing a strain carrying a heat-inducible lambda prophage near the his genes and selecting by transduction of a fla deletion strain. Preliminary characterization of these transducing phages is also reported.

The Escherichia coli operon designated flaA contains seven flagellar genes; among them are two switch protein genes whose products are believed to interface with the motility and chemotaxis machinery of the cell. Complementation analysis using several plasmids carrying different portions of the flaA operon and analysis of expression of these plasmids in minicells allowed the identification of two flagellar gene products. The MotD (now called FliN) protein, a flagellar switch protein, was determined to have an apparent molecular weight of 16,000, and the FlaAI (FliL) protein, encoded by a previously unidentified gene, had an apparent molecular weight of 17,000. DNA sequence analysis of the motD gene revealed an open reading frame of 414 base pairs. There were two possible initiation codons (ATG) for motD translation, the first of which overlapped with the termination codon of the upstream gene, flaAII (fliN). The wild-type flaAI gene on the chromosome was replaced with a flaAI gene mutated in vitro. Loss of the flaAI gene product resulted in a nonmotile and nonflagellated phenotype. The subcellular location for both the MotD and FlaAI proteins was determined; the FlaAI protein partitioned exclusively in the insoluble fraction of a whole minicell sonic extract, whereas the MotD protein remained in both the soluble and insoluble fractions. In addition, we subcloned a 2.2-kilobase-pair DNA fragment capable of complementing the remaining four genes of the flaA operon (flbD [fliO], flaR [fliP], flaQ [fliQ], and flaP [fliR]).

A series of fusions of flagellar genes to the lacZ gene was generated by insertion of Mu dII301 (Apr lac) bacteriophage into the genome of Escherichia coli. The beta-galactosidase activity in each resulting mutant was measured, and the location of the activity in the membrane, periplasmic, or cytoplasmic fraction of the cell was determined. There were three classes of mutants: those which had beta-galactosidase activity mainly in the membrane fraction, those which had it distributed in the soluble and membrane fractions, and those which had it in the cytoplasmic fraction only. The last, soluble-fraction-only, class was predominant in fla-lac gene fusions. In particular, the following mutants were shown to have beta-galactosidase activity in the membrane fractions: on the inner membrane, mutants with flaB fusions, and on the inner and outer membranes, mutants with flaA4850, flaM, and flaU4849 fusions. These results suggest that fla-lacZ gene fusions produce proteins which are able to detect the signals of the leader sequence and the membrane-anchoring region of the flagellar system.

Autoregulation of the expression of flagellar genes was investigated by the technique of operon fusion. The results suggested that the flaU gene is a repressor and the flaD gene is an activator of transcription of the hag, flaS, and Mocha operons. The action of the putative flaU repressor appears to be masked by its interaction with other flagellar proteins during assembly; thus, repression is apparent only when the interacting proteins are absent. This hypothesis is supported by the phenotype of an unusual flaU mutant, which represses even though it is unable to promote flagellar assembly. Presumably, the mutant synthesizes a repressor whose activity is no longer masked by interaction with other flagellar proteins.

The production of hook protein and flagellin in 29 Fla- mutants of Escherichia coli K-12 was determined by the complement fixation assay. Six mutants produced hook protein, and four of them also produced flagellin. A flaE mutation was introduced into these fla mutants carrying the hook structure. All of these mutants made polyhooks and were used as hosts for a newly isolated host-range mutant of chi phage that has a high affinity for the hook structure. All except one mutant produced significant amounts of progeny phages. A flaD flaE double mutant was that exception which did not yield significant amounts of progeny by the phage propagation method. All of the flaE double mutants produced comparable amounts of polyhooks, and no qualitative difference was detected between chi-sensitive and chi-insensitive mutants by the complement fixation assay. Accordingly, it was thought that the polyhook of the flaD flaE mutant had a mechanical defect for chi phage infection. This assumption was confirmed by tethered-cell experiments; the flaD flaE mutant did not rotate. These results are well explained by a proposed regulation pathway of flagellar genes. flaE mutants can express other genes which govern the final step of the flagellar morphogenesis, whereas flaD mutants cannot rotate, possibly because the mocha operon is not expressed. The results obtained in E. coli were also found to be applicable to Salmonella typhimurium.

Previous studies showed that the structural gene for the flagellar hook subunit protein (molecular weight 42,000) was one of a group of flagellar genes located on the Escherichia coli genome near pyrC. Several lines of evidence indicate that the flaK gene is the structural gene for the hook subunit protein. Fla+ strains that were insensitive to chi infection could be isolated as revertants of an FlaK- amber mutant strain but from no other Fla- strain. The hook subunit proteins isolated from such chi-sensitive revertants of the FlaK- strain were shown to be antigenically and electrophoretically different from the hook protein isolated from the wild-type strain. Thus, reversion of a mutation in the flaK gene resulted in alteration of the structure of the hook protein. Furthermore, in programming experiments with hybrid lambda containing flagellar genes, lambdafla with flaK genetic activity programmed the synthesis of a 42,000-molecular weight protein, whereas lambdafla without flaK genetic activity did not.

Escherichia coli mutants with defects in 29 flagellar genes identified so far were examined by electron microscopy for possession of incomplete flagellar structures in membrane-associated fractions. The results are discussed in consideration of the known transcriptional interaction of flagellar genes. Hook-basal body structures were detected in flaD, flaS, flaT, flbC, and hag mutants. The flaE mutant had a polyhook-basal body structure. An intact basal body appeared in flaK mutants. Putative precursors of the basal body were detected in mutants with defects in flaM, flaU, flaV, and flaY. No structures homologous to the flagellar basal body or its parts were detected in mutants with defects in flaA, flaB, flaC, flaG, flaH, flaI, flaL, flaN, flaO, flaP, flaQ, flaR, flaW, flaX, flbA, flbB, and flbD. One flaZ mutant had an incomplete flagellar basal body structure and another formed no significant structure, suggesting that flaZ is responsible for both basal body assembly and the transcription of the hag gene.

Previous studies have defined 28 genes necessary for the synthesis of the flagellar apparatus of Escherichia coli K-12. This study analyzed the influence of the flagellar genes on the expression of the hag gene (structural gene for flagellin). To this end, a hag::Mu d(Apr lac) mutant which had the lac genes fused to the promoter of the hag gene was constructed. This allowed the measurement of hag gene expression by detection of beta-galactosidase activity. The following observations were made. (i) The hag gene was expressed constitutively in Fla+ cells. (ii) hag gene expression was positively regulated by flaA, FLAB, flaC, flaD, flaE, flaG, flaH, flaI, flaK, flaL, flaM, flaN, flaO, flaP, flaQ, flaR, flaV, flaW, flaX, flaY, flaZ, flbA, and flbB genes.hag-lac expression was not observed in strains with these fla mutations. (iii) The hag gene was expressed in mutants with flaS, flaT, flaU, and flbC defects. Therefore, these genes were not involved in regulation of hag gene transcription.

Of the more than 30 genes required for flagellar function, 6 are located between pyrC and ptsG on the Escherichia coli genetic man. This cluster of genes is called flagellar region I. Four-point transductional crosses were used to establish the position and order of the region I flagellar genes with respect to the outside markers ptsG and pyrC. Bacteriophage lambda-E. coli hybrids that contained most of the genes necessary for flagellar formation were constructed. The properties of specific hybrids that carried the region I fla genes were examined by genetic complementation and by measuring the capacity of the hybrids to direct the synthesis of specific polypeptides. The results of these tests with lambda hybrids and with a series of deletion mutations derived from the lambda hybrids demonstrated the existence of at least six flagellar-specific cistrons. These directed the synthesis of polypeptides with the following apparent molecular weights: flaV, 11,000; flaK, 42,000; flaL, 30,000 and 27,000; flaM, 38,000; flS, 60,000; and flaT, 35,000. Plasmid ColE1-E. coli hybrids with region I flagellar genes were also used to program the synthesis of polypeptides in minicell-producing strains. The polypeptides synthesized in these experiments were identical to polypeptides of the hook-basal body structure and helped to confirm the assignment of genes to specific polypeptides. The synthesis of all of these polypeptides was regulated by the same mechanism that regulates the synthesis of other flagellar-related structural components.

Two mutants of Escherichia coli strictly deficient in uridine-diphosphoglucose pyrophosphorylase activity (galU) were found to have very small numbers of flagellar filaments and hooks. In these mutants, both the rate of flagellin (flagellar protein) synthesis and the amount of messenger ribonucleic acid specific for flagellin were considerably lower than in the parental strains. Motile revertants from the galU mutants were isolated and were found to carry a suppressor mutation, which was mapped in the flaH cistron. These strains formed swarms under conditions of catabolite repression; their intracellular concentration of cyclic adenosine 5'-monophosphate was the same as that in the parental strains. These results suggest that the outer membrane affects flagellar formation through the flaH gene product.