Using a phase-locked mutant of Myxococcus xanthus to study the role of phase variation in development.

The bacterium Myxococcus xanthus undergoes a primitive developmental cycle in response to nutrient deprivation. The cells aggregate to form fruiting bodies in which a portion of the cells differentiate into environmentally resistant myxospores. During the growth portion of the M. xanthus life cycle, the organism also undergoes a phase variation, in which cells alternate between yellow and tan colony-forming variants. Phase variation occurs in our laboratory strain (M102, a derivative of DK1622) at a frequency high enough that a single colony of either the yellow or the tan phase already contains cells of the alternate phase. In this study we demonstrate that tan cells within a predominantly yellow population of phase variation-proficient cells are preferentially recovered as heat- and sonication-resistant spores. To further investigate the possibility of a differential role of tan and yellow cells during development, a tan-phase-locked mutant was used to compare the developmental phenotypes of a pure tan population with a predominantly yellow, phase variation-proficient population. Pure tan-phase populations did not produce fruiting bodies or mature spores under conditions in which predominantly yellow wild-type populations did so efficiently. Pure populations of tan-phase cells responded to developmental induction by changing from vegetative rod-shaped cells to round forms but were unable to complete the maturation to heat- and sonication-resistant, refractile spores. The developmental defect of a tan-phase-locked mutant was rescued by the addition of phase variation-proficient cells from a predominantly yellow culture. In such mixtures the tan-phase-locked mutant not only completed the process of forming spores but also was again preferentially represented among the viable spores. These findings suggest the intriguing possibility that the tan-phase cells within the vegetative population entering development are the progenitors of spores and implicate a requirement for yellow-phase cells in spore maturation.     

Myxospore induction in a nondispersed growing mutant of Myxococcus xanthus.

Myxococcus xanthus RB5, a rough-colony-forming, nondispersed growing mutant of strain FBt, forms macroscopic, multicellular masses of radially oriented cells in shake cultures. The cells appear to be held together by slime fibrils. Physical and enzymatic methods to disrupt the spheres were unsuccessful as were attempts to isolate dispersed growing mutants. During incubation of the spheres in starvation medium, the cells within convert to myxospores, indistinguishable from those formed in fruiting bodies. Myxospores were also induced in artifically constructed, dense masses of cells of a nonmotile strain.     

Site-specific integration and expression of a developmental promoter in Myxococcus xanthus.

A series of intercellular signals are involved in the regulation of gene expression during fruiting body formation of Myxococcus xanthus. Mutations which block cell interactions, such as csgA (formerly known as spoC), also prevent expression of certain developmentally regulated promoters. csgA+ cells containing Tn5 lac omega DK4435, a developmentally regulated promoter fused to lacZ, began synthesizing lacZ mRNA 12 to 18 h into the developmental cycle. beta-Galactosidase specific activity increased about 12 h later. Neither lacZ mRNA nor beta-galactosidase activity was detected in a developing csgA mutant containing omega DK4435. The developmental promoter and its fused lacZ reporter gene were cloned into a pBR322-derived plasmid vector containing a portion of bacteriophage Mx8. These plasmids preferentially integrated into the M. xanthus chromosome by site-specific recombination at the bacteriophage Mx8 attachment site and maintained a copy number of 1 per chromosome. The integrated plasmids were relatively stable, segregating at a frequency of 0.0007% per generation in the absence of selection. The cloned and integrated promoter behaved like the native promoter, expressing beta-galactosidase at the proper time during wild-type development and failing to express the enzyme during development of a csgA mutant. The overall level of beta-galactosidase expression in merodiploid cells containing one native promoter and one promoter fused to lacZ was about half that of cells containing a single promoter fused to lacZ. These results suggest that the timing of developmentally regulated gene expression is largely independent of the location of this gene within the chromosome. Furthermore, they show that site-specific recombination can be a useful tool for establishing assays for promoter or gene function in M. xanthus.     

The strategy of Myxococcus xanthus for group cooperative behavior

New evidence has been presented from our laboratory that the gliding bacterium, Myxococcus xanthus, does not home by chemotaxis toward a nutrient source. Our experiments, those of others, and the theory presented here combine to suggest a model, called the Pied Piper model. It hypothesizes a gene that has a high mutation rate forward and back (say something in excess 10^-4mutations per cell generation) which leads to switching between two motility states. Occasionally rare organisms become genetically, but reversibly, changed so that they move unidirectionally instead of mostly forward and back as do the bulk of the cells. When such a leader cell arises, it continues to move in its original orientation, and causes a cohort of cells to move together away from the bulk of the cells. That is, in the less common mutational state it counteracts the usual tendency to just move forward and backward achieving little net movement. The assumption of a genetic element that mutates in a reversible way is suggested by numerous cases of reversible switches now known in a wide range of bacteria serving a variety of functions. A second aspect of the model is that mechanisms exist that cause cells to move in the same direction as their nearby neighbors. This process results in a regular spacing of bands of cells to form mounds in the absence of a leader. The action of C-factor, a factor secreted by the cells which has been largely studied in the laboratory of Dale Kaiser, and extracellular fibrils, (rod-shaped protein and carbohydrate bodies) largely studied in the laboratory of Martin Dworkin, may be key elements in coordinating (or linking) the movements of neighboring cells. Based on the assumption of the absence of chemotaxis, computer simulations of pattern formation for gliding bacterial swarms and flares are consistent with observed behaviors and thus are additional evidence that chemotactic motility of the type exhibited by Escherichia coli, is not necessary for the group movements of M. xanthus. Some tests for this model are suggested.     

Gliding movements in Myxococcus xanthus.

Prokaryotic gliding motility is described as the movement of a cell on a solid surface in the direction of the cell's long axis, but its mechanics are unknown. To investigate the basis of gliding, movements of individual Myxococcus xanthus cells were monitored by employing a video microscopy method by which displacements as small as 0.03 micron could be detected and speeds as low as 1 micron/min could be resolved. Single cells were observed to glide with speeds varying between 1 and 20 microns/min. We found that speed variation was due to differences in distance between the moving cell and the nearest cell. Cells separated by less than one cell diameter (0.5 micron) moved with an average speed of 5.0 micron/min, whereas cells separated by more than 0.5 micron glided with an average speed of 3.8 microns/min. The power to glide was found to be carried separately at both ends of a cell.     

Chromosome replication in Myxococcus xanthus.

The rates of DNA synthesis during the cell-division cycle were measured in Myxococcus xanthus growing in three different media permitting a twofold variation in doubling time. In all three media, simple DNA cycles were observed. Synthesis of DNA occurred during 85% of the cell-division cycle, independent of generation time, from 5 to 11 h. Cells were observed to contain one bacterial nucleoid at birth that later divided synchronously midway through the cell cycle. Nucleoid segregation appeared to begin before chromosome replication was completed. The DNA content of exponential-phase bacteria was determined to be about 20 +/- 3 X 10(-9) microgram per cell; newborn bacteria contained about 14 +/- 2 X 10(-9) microgram of DNA per cell. Exponential-phase bacteria showed about a 50% increase in DNA in the presence of chloramphenicol (50 microgram/ml). The number of randomly segregating chromosomes present in exponential-phase bacteria was determined by following the fate of prelabeled DNA during outgrowth in nonradioactive media. The results are consistent with a model in which cells are born with exactly one complete unreplicated chromosome. The molecular weight of such a chromosome is about 8.4 +/- 1.2 X 10(9).     

Identification of a vegetative promoter in Myxococcus xanthus. A protein that has homology to histones

A physical map of 330 x 10(3) base-pairs near the replication origin of Myxococcus xanthus chromosome has been established already. Using DNA fragments from this region, Northern blot hybridization analysis was carried out in order to identify the genes expressed during vegetative growth. One of the genes, tentatively designated as vegA, was cloned and its entire DNA sequence was determined. The amino acid sequence of the gene product deduced from the DNA sequence reveals that the VegA protein is a very basic protein with a molecular weight of 18,700. The gene was expressed in Escherichia coli using an expression vector, and its gene product was identified using SDS/polyacrylamide gel electrophoresis. From the results of S1 nuclease mapping, the vegA promoter was found to contain the sequence TAGACA at the -35 region and the sequence AAGGGT at the -10 region. These two regions are separated by 18 nucleotides. Genetic analysis suggests that the vegA gene may be essential for the growth of M. xanthus. From a computer-aided search for homologies to know protein structures, it was found that the VegA protein has homologies to histone H4 of Tetrahymena thermophila and histone H2B of sea urchin.     

Tactic behavior of Myxococcus xanthus.

With time-lapse videomicroscopy it was demonstrated that cells of Myxococcus xanthus are capable of directed (tactic) movement toward appropriate targets. Mutants that had lost A motility (J. Hodgkin and D. Kaiser, Mol. Gen. Genet. 171:177-191, 1979) were unable to show directed movement. Cells showed directed movement to polystyrene latex beads and to glass beads, as well as to clumps of Micrococcus luteus. This is consistent with other observations in an accompanying paper (M. Dworkin and D. Eide, J. Bacteriol. 154:437-442, 1983) that indicate that M. xanthus does not perceive chemical gradients.     

Isolated fibrils rescue cohesion and development in the Dsp mutant of Myxococcus xanthus.

Extracellular fibrils are involved in cell cohesion and cell development in Myxococcus xanthus. One group of social motility mutants, Dsp, is unable to produce extracellular fibrils; these mutants also lose the abilities to cohere and to develop. Extracellular fibrils isolated from vegetative wild-type cells and added to Dsp cells fully restored the abilities of these cells to cohere and to undergo normal morphological development. The fibrils thus mimic the ability of intact, wild-type cells to carry out the same rescue. Optimal cohesion rescue by fibrils required calcium and magnesium ions, did not require protein synthesis, but was energy dependent, i.e., sodium azide and sodium cyanide blocked rescue. Cohesion rescue was also blocked by the diazo dye Congo red. Cohesion rescue is genus specific, i.e., isolated fibrils did not cause the cohesion of Pseudomonas aeruginosa, Bacillus subtilis, Proteus mirabilis, Escherichia coli, or the related myxobacterium Stigmatella aurantiaca. Developmental rescue of Dsp by isolated fibrils included aggregation, fruiting body formation, and myxospore morphogenesis. Developmental gene expression in the Dsp mutant was only partially rescued by the isolated fibrils.     

Pigmentation phenotype instability in Myxococcus xanthus.

Cells of Myxococcus xanthus FB2 produce tan or yellow colonies. Subcultures of tan colonies yielded tan and yellow colonies and subcultures of most yellow colonies yielded only yellow colonies. Strain FB2 variants in which the color type is more stable were obtained. Yellow cells were distinguishable from tan by the presence of pigment(s) with an absorption maximum at 379 nm. Fluctuation Test experiments and the presence of this pigment(s) in liquid cultures of FB2 indicated that tan phenotype cells spontaneously became or segregated yellow cells in liquid culture. The frequency of appearance of yellow cells was increased in low density cultures (less than 10(6)/ml). The increase cannot be explained by differences in growth rates of the two phenotypes. No evidence that cell-cell contact or culture medium constituents affect the appearance of the yellow phenotype was found. Ultraviolet irradiation of FB2 resulted in an increased proportion of cells producing yellow colonies among the survivors. Greater UV resistance of yellow cells and UV-induced conversion of tan to yellow accounts for this increase. Low level photoreactivation of viability and of the tan phenotype occurred. Incubation of FB2 in medium containing mitomycin C, nalidixic acid, phenethyl alcohol, or at 36.5 degrees C also resulted in conversion of tan to yellow cells.     

DNA replication during aggregation phase is essential for Myxococcus xanthus development

Previous studies have demonstrated that fruiting body-derived Myxococcus xanthus myxospores contain two fully replicated copies of its genome, implying developmental control of chromosome replication and septation. In this study, we employ DNA replication inhibitors to determine if chromosome replication is essential to development and the exact time frame in which chromosome replication occurs within the developmental cycle. Our results show that DNA replication during the aggregation phase is essential for developmental progression, implying the existence of a checkpoint that monitors chromosome integrity at the end of the aggregation phase.     

Isolation of a surface glycoprotein from Myxococcus xanthus.

The isolation of a glycoprotein from vegetative cells of Myxococcus xanthus is reported. The protein, abbreviated VGP, was first identified during a survey of surface proteins as a major protein that could be radioiodinated in vegetative, but not developing, cells (P.Y. Maeba, J. Bacteriol. 155:1033-1041, 1983). The protein was extracted from membranes with Triton X-100 and subsequently purified by DEAE-cellulose chromatography, chromatofocusing, and gel filtration. The protein has an Mr of approximately 74,000 as determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and an isoelectric point of 3.2 to 3.3. The carbohydrate moiety which made up approximately 13.5% of the weight of the VGP comprised primarily neutral sugars and smaller amounts of hexosamines and uronic acids. The amino acid content revealed no unusual features, but analysis by the method of Barrantes (F. Barrantes, Biochem. Biophys. Res. Commun. 62:407-414, 1975) indicated it is likely a peripheral membrane protein. The protein makes up approximately 1% of the total cell protein and is a prominent surface structure. Because glycoproteins have been implicated in cellular interactions in a number of systems, the VGP may play an important role in the social behavior exhibited by M. xanthus.     

Two recA genes in Myxococcus xanthus.

Two recA genes, recA1 and recA2, in Myxococcus xanthus were cloned by using the recA gene of Escherichia coli, and their DNA sequences were determined. On the basis of deduced amino acid sequences, RecA1 and RecA2 have 67.0% identity to each other and 60.5 and 60.9% identities to E. coli RecA, respectively. Expression of recA2 was detected in both vegetative and developmental cells by Northern blot (RNA) analysis, and a threefold induction was observed when cells were treated with nalidixic acid. Repeated attempts to isolate a recA2 disruption mutant have failed, while a recA1 disruption mutant was readily isolated. Both the recA1 and recA2 genes expressed in E. coli complement the UV sensitivity of an E. coli recA strain.     

Nutrition of Myxococcus xanthus, a fruiting myxobacterium.

The minimal requirements for vegetative growth of Myxococcus xanthus have been sought. Isoleucine, leucine, and valine were required, and vitamin B12 was needed for the synthesis of methionine. Pyruvate was an excellent energy source and an efficient source of cellular carbon. Acetate, aspartate, glutamate, and most tricarboxylic acid cycle intermediates could also be utilized, but were less efficient sources of carbon and energy than was pyruvate. Many mono- and disaccharides were tested, but, in agreement with earlier results, none served as carbon-energy sources. A minimal medium (A1) has been devised that includes the essential amino acids and vitamin B12, with pyruvate and aspartate as carbon-energy sources. In this medium, M. xanthus could propagate indefinitely, and on it vegetative cells formed colonies with greater than 75% efficiency; hence, it is likely that no organic cofactors other than those present in A1 are required in more than trace amounts.