INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

In response to nutrient limitation, Myxococcus xanthus cells assemble a multicellular fruiting body within which myxospores are produced (3, 8, 23). Individual, rod-shaped prespores differentiate into the spherical myxospores, which are metabolically quiescent and environmentally resistant. Sporulation and aggregation are responses to the same extracellular signal, the approximately 20-kDa C-factor protein. This signal, the product of the Myxococcus csgA gene, is associated with the cell surface (10, 12, 24). Mutant csgA cells are defective in aggregation and fail to sporulate. However, if they are either mixed with wild-type cells or supplemented with purified C-factor, their abilities to aggregate and sporulate are restored (10, 24). C-factor is produced in response to starvation (10).

There is a second way for Myxococcus to sporulate, called starvation-independent sporulation. When cells are growing vigorously in a rich, aerated medium, exposure to certain reagents, such as 0.5 M glycerol (4), 0.7 M dimethyl sulfoxide (DMSO) (13), phenyl ethanol (4), or a variety of -lactam antibiotics (22), induces single cells to become spores within 2 to 3 h. From cell to cell, glycerol-induced shape changes are quite synchronous, and almost all treated cells become spores. This type of sporulation is promoted by ample nutrient and vigorous aeration, not starvation. The nature of starvation-independent spores is the same whether the inducer is glycerol or DMSO, suggesting that these reagents induce an endogenous process of sporulation, which may be otherwise induced and augmented in the process of fruiting body development.

Fruiting body sporulation, also known as starvation-dependent sporulation, is cell cooperative, depends on cell density, requires a surface like agar, and occurs subsequent to aggregation. Often no more than 1% of input cells become viable fruiting body spores, and it takes at least 3 days to complete spore maturation. In addition, there are structural differences between the two types of spores. The starvation-independent spores lack the fruiting body spore protein S (13), but they contain many more ribosomes, and their coats are thinner (28). On the one hand, many developmental mutants that cannot form mature fruiting bodies can still form viable glycerol spores, including asgA, asgB, asgC, csgA, and dev mutants (6, 19). On the other hand, many mutants that have lost glycerol inducibility still form spores in fruiting bodies (1).

Despite their differences, the starvation-dependent and starvation-independent pathways may have steps in common, steps that change the cell shape from a rod into a sphere, add thickness to the wall, and enhance resistance to injury. Both types of spores contain protein U (13); both pathways induce a -lactamase activity (22).

Here, we report a mutation that causes defects in the change in cell shape. This mutation affects fruiting body and starvation-independent sporulation in similar ways and identifies a particular step common to both.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains, media, growth, and development. The M. xanthus strains used in this study were as follows. DK 1622 (wild type) (9) and DK 5208 (Tn5-132::csgA) (15) have been described elsewhere. The origin and properties of DK 7536 (Tn5lac 7536) are presented in the text. New genotypes were constructed using the transducing phage Mx4ts18ts27 hrm (2) or Mx8clp2 (21). DK10524 (Tn5lac7536) was produced by a transductional backcross of DK 7536 into DK 1622. DK10527 (Tn5lac7536 Tn5-132::csgA) was constructed by transduction from DK10524 into DK5208 with kanamycin selection. DK10552 (Tn5lac 7536 devRS) was constructed by transduction with kanamycin selection from DK10524 into the dev mutant DK11212, produced by Bryan Julien of Kosan Biosciences.

Tn5lac mutagenesis and isolation of developmental mutants. Tn5lac was introduced into the fully motile strain DK1622 by P1 transduction (14). Km^r transductants were picked to fresh CTT-plus-kanamycin plates. Approximately 10,000 transductants were examined for increased lacZ expression during development over a 3-day period by Harvey Kimsey. Vegetatively growing cells were transferred on toothpicks from CTT agar to TPM agar plates containing 40 µg of X-Gal (5-bromo-4-chloro-3-indoyl--D-galactopyranoside)/ml or CTT agar plates containing 20 µg of X-Gal/ml. Levels of -galactosidase activity were estimated by comparing the intensity of blue dye deposited by developing cells versus vegetatively growing cells. In all, 272 strains were retained because they displayed increased lacZ expression during development.

Measurement of -galactosidase activity from Tn5lac. Cells were induced to develop either in submerged culture or on TPM plates. If they were developed on plates, 100 µl of a cell suspension of 5 × 10⁹ cells/ml was spotted in 20-µl aliquots. At intervals, cells were harvested from these plates into 400 µl of TPM buffer and stored at 20°C until all samples were collected. If cells were developed in microtiter wells, entire dishes were removed from the 32°C incubator and stored at 20°C until all samples were collected. Samples obtained by either method were treated as follows (adapted from reference 18). The samples were dispersed by a 10-s exposure to a microtip sonicator from Heat Systems-Ultrasonics. Samples from microtiter wells were transferred to microcentrifuge tubes. The samples, now all in tubes, then underwent a second sonication for 5 min in a cup horn (Tekmar) with ice water cooling. Before this second sonication, glass beads (acid washed; diameter, 425 to 600 µm) were added to the sample tubes to disrupt spores, which otherwise resist shear breakage by sonication alone. -Galactosidase specific activity was assayed as described by Kuspa et al. (18), with two modifications: -mercaptoethanol was omitted from the Z buffer, and samples were not sedimented before they were assayed. To assay -galactosidase from cells induced to sporulate under nonstarvation conditions, 0.5 M glycerol was added to the cultures at time zero. At each time point, 0.5 ml was withdrawn, the suspension was centrifuged, and the supernatant discarded. The pellet was stored at 20°C. When all samples had been collected, the cell pellets were thawed and resuspended in 1 ml of TPM buffer. Glass beads were added to each sample, and samples were sonicated in a cup horn sonicator and immediately assayed for -galactosidase activity.

Scanning electron microscopy. To reveal the cellular structures inside a fruiting body, cells were allowed to develop on Nuclepore polycarbonate filters. Cell suspensions (20-µl droplets of 5 × 10⁹ cells/ml in TPM buffer) were placed onto the filters, and the filters were placed onto a TPM agar plate. The droplets of cells were allowed to dry into the filters before the plates were inverted to incubate at 32°C for 3 days. After development, the fruiting body-like aggregates were fixed by immersing the filters in a 2.5% glutaraldehyde-veronal acetate solution overnight. Following glutaraldehyde fixation, the filters were transferred into a 1% OsO₄-veronal acetate solution for 1 h. Specimens were then stained for 30 min with 1% uranyl acetate before dehydration in a graded series of ethanol solutions (15, 30, 45, 60, 75, and 90%) for at least 1 h each. Finally the specimens were incubated in 100% ethanol for at least 1 h, and they were chemically dryed with hexamethyldisilane. The fruiting bodies were then dry fractured with a razor blade. The fractured samples were gold coated to a thickness of 11 nm, using a Polaron SEM coating system. Specimens were examined with a Philips 505 scanning electron microscope.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Identification of sporulation mutant 7536. A collection of 272 strains with independent Tn5lac insertions in M. xanthus, having developmental defects and developmentally regulated expression of -galactosidase (described in Materials and Methods), was screened for mutants that were defective specifically in sporulation. Identified in this screen, insertion mutant 7536 produced fewer than 10⁶ as many heat- and sonication-resistant spores as the wild type, measured by the capacity to form colonies after such treatment. Actually, 2 × 10⁸ cells were plated on each of several plates, and there were no colonies on any plate. Despite this extreme sporulation defect, mutant 7536 aggregated well (Fig. 1). Its aggregates are similar in number, size, and shape to those of the wild type.

View larger version (169K): [in this window] [in a new window]   FIG. 1.   Aggregate formation. Photographs were taken at 48 h of development on TPM agar. (A) The wild-type strain DK1622; (B) DK 7536. Bar, 0.2 mm.

Fruiting body cell morphology. The screen utilized to isolate the 7536 insertion mutant would not have distinguished between a defect in sporulation and one in germination. To address this issue, the cells within wild-type and 7536 fruiting bodies were examined by scanning electron microscopy, which, unlike a colony assay, does not rely on germination (Fig. 2). After 3 days of development, the wild type forms spores that pack very regularly within a fruiting body because of their uniform size, spherical shape, and high density. Within the 7536 mutant fruiting body-like aggregates, however, a heterogeneous mixture of cell sizes and shapes is seen. The shapes range from shortened rods to bean shapes to ovoids of various proportions. Because these shapes could be intermediates in the morphogenesis of spores, their accumulation would imply that 7536 has a defect in spore morphogenesis. Also, because of their variety of shapes, the cells within an 7536 fruiting body are not packed in an orderly fashion, in striking contrast to the organization of cells in a wild-type fruiting body (Fig. 2).

View larger version (185K): [in this window] [in a new window]   FIG. 2.   Scanning electron micrographs of aggregates and their contents. Photographs were taken at 48 h of development on TPM agar. (a and b) The wild-type strain, DK1622, at low and high magnifications, respectively. (c and d) DK 7536 at low and high magnifications, respectively. Bars, 5 µm.

Gene expression from the 7536 insertion. Since the mutation arises from an insertion of the Tn5lac transposable element (14), which is properly oriented to fuse the affected locus transcriptionally to lacZ, expression of the locus can be monitored by measuring -galactosidase specific activity (Fig. 3). The 7536 insertion is not expressed vegetatively; it is induced after 17 h of development in submerged culture, a time that matches the normal beginning of sporulation after the onset of starvation. Moreover, in the wild type, sporulation is completely dependent on the extracellular C-signal (20, 26), so that csgA mutants form fewer than 10⁶ spores of the signal-competent strain. To test the C-signal dependence of the 7536 gene, we transduced the insertion into a strain that cannot make C-signal. The csgA-deficient mutant of 7536 does not express its -galactosidase (Fig. 3).

View larger version (19K): [in this window] [in a new window]   FIG. 3.   Expression of -galactosidase, during development, from the Tn5lac promoter fusion 7536. Cells were sampled immediately after transfer to starvation buffer for the 0-h point. Expression was measured in a wild-type (DK1622) background (squares), in a devRS mutant background (circles), and in a csgA mutant background (diamonds).

Starvation-independent sporulation of 7536. If wild-type M. xanthus cells are exposed in a rich, well-aerated growth medium to any one of the set of reagents or antibiotics described above, then starvation-independent spores are produced. Within a few hours of exposure to any one of these substances, the rod-shaped cells shorten, become spherical, and finally turn bright in a phase-contrast optical system (Fig. 4). Within a few minutes of addition of glycerol, for example, growth ceases, as measured by culture turbidity (4). Such spores are resistant to detergents, sonication, and mild heating (sporulation of wild-type cells carried out in parallel with mutant cells; see all rows in Table 1). Many mutants which cannot form fruiting bodies or starvation-dependent spores due to a developmental defect can still be induced by glycerol, such as the asg, csg, and dev mutants mentioned above. Since the 7536 mutant failed to complete the starvation-dependent shape change, the question arose whether glycerol might also elicit an incomplete shape change. As shown in Fig. 4, by 3 h after the addition of glycerol, both the wild type and the mutant have responded by changing their shape, although differently. While the wild-type cells become spherical and phase bright, many of the mutant cells become ovoid, not round, and never phase bright. After 24 h of exposure to 0.5 M glycerol, the wild type will have formed many phase-bright spores. These spores tend to stick to each other and to the culture flask (the images for the wild type in Fig. 4 were made from scrapings from the flask surface). While the wild type retained no more than a small percentage of rod-shaped cells at 24 h, the 7536 mutant had only rods in the culture. No ring of mutant culture material adhering to the flask was visible.

View larger version (127K): [in this window] [in a new window]   FIG. 4.   Phase-contrast visible-light micrographs of cells exposed to 0.5 M glycerol. (A) The wild-type strain DK1622 (A and B) and the mutant strain DK 7536 (C and D) are shown at 3 h (A and C) and 24 h (B and D). Bar, 10 µm.

Resistance properties. Table 1 shows results of tests of the resistance properties of the 7536 cells after glycerol induction. Glycerol-induced spore-like cells, rather than fruiting body contents, were tested because they can be obtained in higher yields, in more pure form, and with fewer potential viability-damaging manipulations. Mutant-cell survival is compared in the table to that of similarly treated wild-type cells, with the mutant/wild-type fraction expressed as a percentage. The standard spore assay uses three 10-s cycles of sonication and a 2-h heat treatment at 50°C to distinguish glycerol spores from vegetative cells. No resistance to this treatment was developed by the mutant; nor was there resistance to sonication alone, or to heat alone, or to 0.01% SDS exposure. Compared with the survival of glycerol-treated wild-type cells (taken as 100%), less than 0.01% survival was seen after incubation of the glycerol-treated mutant spores at 40 or 50°C.

Growth in the presence of glycerol. Closer examination of the mutant revealed that between 9 and 11 h of incubation in CTT plus 0.5 M glycerol, the ovoid cells induced in the first few hours began to elongate, eventually reaching full rod length. This recovery, by the 7536 mutant, of sustained growth in the presence of glycerol was confirmed by measurements of the culture turbidity. Upon the addition of glycerol, the turbidity of the mutant culture begins to rise while that of the wild type decreases, never rising above its initial value during the 24-h period of observation (Fig. 5A). The rise in turbidity in the 7536 mutant is continuous after 7 to 8 h of incubation in glycerol, about the time that the spheres, seen under the microscope, were reverting to rods. Since starvation-independent sporulation is induced in a complete medium (CTT plus glycerol), there are ample nutrients for growth. The final turbidity of the mutant culture at 24 h is like that of a saturated wild-type culture. Reversion to the rod shape implies that the spore structure has not been completed in the 7536 mutant and that the intermediate cell is able to grow in complete medium, even in the presence of remaining glycerol.

View larger version (14K): [in this window] [in a new window]   FIG. 5.   (A) Cell density after addition of glycerol to a CTT culture. Squares, wild-type (DK1622) culture; circles, a culture of cells carrying the 7536 insertion (DK10524). Glycerol was added to the cultures at 0 h, and cell density was measured in a Klett colorimeter. (B) Expression of -galactosidase from the Tn5lac promoter fusion 7536 after induction by 0.5 M glycerol.

7536 and the C-signal. As shown above, 7536 is not expressed in a csgA mutant, which cannot make the C-signal (Fig. 3). In light of this dependence, the substantial induction of -galactosidase by starvation in the 7536 strain implies that it is capable of producing the C-signal. To assess the level of C-factor production of the mutant in vivo, it was cocultured with csgA cells. When a C-signal-deficient mutant, csgA, is codeveloped with wild-type cells, the mutant can be fully rescued (11): it aggregatesthe aggregates containing both mutant and wild type cells; it produces a wild-type complement of spores within fruiting bodies; the mutant spores are recognized by their oxytetracycline resistance. Table 2 shows the results of coculture assays of C-factor production by 7536, and they lead to several conclusions. On the one hand, the insertion mutant (unlike a csgA strain) cannot be complemented, or rescued, when mixed with wild-type cells. This indicates that the mutant defect in 7536 is cell autonomous relative to C-signaling-induced sporulation. On the other hand, when 7536 was mixed with a csgA mutant, the sporulation defect of the csgA mutant was rescued, and the spore levels were wild type. Since spore levels depend on the level of C-factor (12), the 7536 mutant thus appears to make wild-type levels of C-signal. Since 7536 is dependent on the C-signal for its transcription, 7536 is expected to be part of the C-signal response pathway; the question is, which part? Since the 7536 mutant can aggregate but cannot complete spore morphogenesis, it would be expected to lie in the sporulation branch of the C-signal response pathway. The dev operon has been identified as a component of that branch (25, 26). The Tn5lac of 7536 was transduced into a devRS mutant background, and one of the curves in Fig. 3 shows that 7536 is not expressed in the dev mutant background. Thus, 7536 depends on dev function, and consequently 7536 can be consistently placed just after dev in the sporulation branch of the C-signal response pathway, as represented in Fig. 6.

View larger version (16K): [in this window] [in a new window]   FIG. 6.   C-signal control of sporulation. 7536 in the diagram refers to the function normally performed by those genes. The point at which the starvation-independent inducers act is also indicated by an upward-pointing arrow. The C-signal control of FruA activation (5) and the feedback loop from C-signal reception to csgA expression (6) have been described previously.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

The starvation-dependent and starvation-independent pathways of sporulation have different time scales, different nutritional requirements, and different inducers. As well, they produce structurally different spores, since fruiting body spores are coated with protein S, which is completely lacking in glycerol spores (13, 28). Nevertheless, the two sporulation pathways share the step indicated by 7536 in Fig. 6, a step that is necessary to transform a vegetative rod into a spherical spore. The locus identified by insertion 7536 is transcribed in response to starvation as well as to glycerol. In fruiting body development, that locus begins to be expressed around 17 h, coordinately with the time of sporulation.

The C-signal triggers first the aggregation and then the sporulation process, as befits fruiting body development (20, 25, 26). These processes are coupled as branches 2 and 3 of Fig. 6. As implied by that circuit, the 7536 insertion is not expressed during fruiting body development in a mutant that cannot make C-signal (Fig. 3). Moreover, the very severe sporulation defect of 7536, decreasing the spore number to 10⁶ of the wild-type spore number, is like that of a csgA mutant (Table 2). The full rescue of csgA mutant sporulation (Table 2 shows that this mutant is rescued) implies that the 7536 mutant (the rescuer) produces normal levels of C-factor and thus that the positive feedback on csgA expression (branch 1 in Fig. 6) is fully active (6). Based on the data of Ellehauge et al. (5), this branch is drawn ahead of FruA because Act1 and not FruA sets the level of C-factor (6). The normal size and shape of 7536 aggregates (Fig. 1) implies that branch 2 also functions normally. Expression of -galactosidase from the lacZ insertion in 7536 implies that branch 3 in Fig. 6 is normal up to and including the expression of the 7536 locus. Based on these data, uniquely consistent positions can be assigned to the 7536 function and to the inducing action of glycerol when it acts as a starvation-independent inducer. These two processes are positioned on the sporulation branch of the C-signaling pathway adjacent to each other. Since 7536 expression is blocked in a dev mutant (Fig. 3), and since dev insertion mutations severely reduce sporulation (27), dev must lie upstream of 7536 function in the sporulation branch of the pathway. Glycerol induces 7536 expression but does not induce dev (4414) expression (16), although dev mutants can be induced to make viable glycerol spores. As demonstrated by Ellehauge et al. (5), dev follows activated FruA, as illustrated by branch 3 of the C-signaling pathway (Fig. 6).

For the same reasons, the glycerol stimulus that induces sporulation must feed into the C-signal response pathway between dev and the 7536 function, which is shown as branch 4 in Fig. 6. Since branch 4 enters beyond dev, its position would explain first why glycerol spores lack protein S, which is encoded by tps, because tps has to be induced by A-signaling (18). It would also explain the observation that most developmental mutant defects, including dev and those that precede it, such as asg and csg, do not eliminate glycerol spore inducibility. On the one hand, the nearly 100% induction of sporulation by glycerol implies that every growing cell is equipped to become a spore. On the other hand, fruiting body sporulation, in which only 1% of the input cells become viable spores (Table 1), may be limited by the residual capacity for protein synthesis in starving cells and by Dev induction of sporulation functions. As yet, not much is known about this induction.

The 7536 insertion mutant begins, but is unable to complete, the normally observed change to a stable spherical shape in both fruiting body and glycerol sporulation. Within the mutant aggregates (Fig. 3) are a variety of cellular forms: short rods, bean shapes, and various ovoidsall plausibly intermediates in the normal process of shape change. The virtually continuous spectrum of intermediate forms inside these mutant aggregates suggests a normally continuous transition of shape, instead of a few discrete intermediate shapes. The cellular forms induced by exposing the mutant to glycerol are generally ovoid (Fig. 4), are more sensitive to heat, sonication, and detergents than the wild type (Table 1), and begin to grow by rod elongation, even in the presence of glycerol. (Nutrients sufficient for rapid growth are present in the glycerol sporulation medium.) The turbidity increase that accompanies the growth of the 7536 mutant, after it first becomes ovoid, implies that glycerol has induced only a temporary halt to growth and to cell elongation in the mutant. This temporary growth arrest may open a window on the mechanism of cell shape change in myxosporulation.