INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Mycoplasmas are parasitic bacteria that have extremely low G+C contents and small genomes (9). Their morphology is irregular because of the lack of a peptidoglycan layer.

In Escherichia coli, initiation of chromosomal DNA replication occurs once during the cell's replicative cycle, and the nucleoids partition before cell division (13). The chromosomal replication of E. coli initiates in a small region and proceeds in both directions. It is mainly controlled by the timing and frequency of initiation, while the velocity of replication is constant.

In mycoplasmas, chromosome replication also starts at a fixed site, followed by bidirectional progression (19-21, 25, 40). As in many eubacteria (36), the dnaA gene is expressed and plays important roles in the initiation of replication (35). These observations suggest that the outline of chromosome replication of mycoplasmas is similar to that of E. coli. However, the process of mycoplasma cell reproduction has not been clarified. Moreover, the cell division cycle of E. coli cannot be simply applied to mycoplasmas because of their irregular cell morphology (4). A model has been suggested for the cell cycle of Mycoplasma mycoides (6, 30, 31), which is closely related to Mycoplasma capricolum (39). According to this model, an elementary rounded body grows into a filamentous form and then new elementary rounded bodies are developed within this filament and released, but this model has not been adequately substantiated.

In this study, we analyzed the process of DNA replication, cell morphology, and viability under various conditions of M. capricolum and proposed a model of cellular reproduction for this bacterium.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Cultivation. M. capricolum ATCC 27343 was grown at 37°C unless otherwise specified. Modified Edward medium (MEM) (28) was used with some modifications, i.e., 5% horse serum was replaced by 2 mg of bovine serum albumin/ml, 20 µg of cholesterol/ml, 10 µg of palmitic acid/ml, and 12 µg of oleic acid/ml, according to the synthetic medium recipe (29). For supplementation with nucleosides, 40 µg (each) of adenosine, guanosine, and uridine/ml and 20 µg of thymidine/ml were added to the medium. Frozen cultures were inoculated into the medium and grown overnight to reach an optical density at 600 nm (OD₆₀₀) around 0.05. The cultures were diluted as the OD₆₀₀ became 10⁴ and were used for assays after several hours.

Titration of total DNA. DNA content in cultures was assayed by Southern hybridization (7). Cells were lysed by mixing the growing culture with a solvent composed of 50% phenol, 48% chloroform, and 2% isoamyl alcohol. To normalize the yield of DNA extraction, 1-µl aliquots of late-growth-phase culture grown in the presence of 14 µM [¹⁴C]thymine (2.1 TBq/mol) were added to each cell sample just before cell lysis. DNA was isolated by the phenol extraction method (33), and the yield was determined from the radioactivity of ¹⁴C in the trichloroacetic acid-insoluble fraction. A series of DNA dilutions were heat treated, dot blotted on uncharged nylon sheets, and subjected to hybridization analysis. The chromosomal DNA prepared from a late-growth-phase culture was used as the template for synthesis of probes. The sheets were exposed to a Fuji imaging plate, and the radioactivity was measured with a Bio Image Analyzer BAS 1000. The radioactivity of [¹⁴C]thymine incorporated into the chromosomal DNA was much less than that of ³²P-labeled probe hybridized to the blotted DNA. Standard radioactivity was determined by using a dilution set of the chromosomal DNA.

Titration of protein contents. Cells were collected by centrifugation at 15,000 × g at 4°C for 5 min and were washed once with solution A, consisting of 20 mM Tris-HCl (pH 7.6), 0.25 M NaCl, and 10 mM EDTA. Washed cells were resuspended in solution A and lysed by addition of 0.1% sodium dodecyl sulfate. The cell lysate was diluted to the appropriate concentration, and the total protein was titrated by the Bradford method.

Radiolabeling of DNA and analysis of replication intermediates. [³²P]dAMP was prepared from [-³²P]dATP as described previously (21). Radiolabeling of mycoplasmal DNA was carried out by addition of 92.5 KBq of [³²P]dAMP per ml (16 nM) to each culture. Two minutes later, 1 mM cold dAMP was added to each culture. For analysis on alkaline agarose gels, DNA synthesis was stopped by mixing aliquots of cultures with a solvent composed of phenol, chloroform, and isoamyl alcohol. The chromosomal DNA was prepared by the phenol method and subjected to 1% alkaline agarose gel electrophoresis (33). The fractionated DNA was transferred onto charged nylon sheets, and each sheet, with the half-dried gel attached, was exposed to the imaging plate, followed by analysis with the image analyzer. For analysis by field inversion gel electrophoresis (FIGE), DNA synthesis was stopped by mixing aliquots of cultures with an equal volume of fixing solution composed of 75% ethanol, 2% phenol, 21 mM sodium acetate (pH 5.3), and 2 mM EDTA (12). The chromosomal DNA was isolated by the agarose block method as previously described (22, 26). The following procedures were performed as described previously (21).

Measurement of origin/terminus ratio. Cells were lysed by mixing the cultures with a solvent composed of phenol, chloroform, and isoamyl alcohol. DNA was isolated by the phenol method. Dilution sets of the chromosomal DNA were heat treated, dot blotted on uncharged nylon sheets, and subjected to hybridization analysis (7). The plasmid pUNH119 was used as the standard for the origin titration. This plasmid harbors a 1,949-bp fragment extending in the origin region from nucleotide 2691 to 4639 (numbered according to previous reports [19, 20]). The plasmid used as the standard for terminus titration was clone 4 of the gyrase gene reported previously (34). These plasmids were digested by single-cutter endonucleases, diluted appropriately, heat treated, and dot blotted onto the sheets. Radiolabeled probes were made by using DNA fragments of about 500 bp complementary to the insertion sequences of the standard plasmids. The amount of plasmid on each sheet was normalized by hybridization using a probe complementary to the ampicillin resistance gene, which is carried by both of the standard plasmids. Hybridization was performed sequentially with the same sheets. The results did not depend on the probing order.

Microscopic observation. M. capricolum cells were collected by centrifugation at 10,000 × g at 4°C for 3 min, suspended in phosphate-buffered saline (PBS) containing 3% glutaraldehyde and 10 mM EDTA, and incubated for 30 min at room temperature for fixation. The fixed cells were collected, washed with PBS, and resuspended in PBS. For light microscopic observation, cell suspensions were dropped onto glass slides and covered with coverslips. For observation by fluorescence microscopy, an equal volume of 20-µg/ml 4',6-diamidino-2-phenylidole (DAPI) solution was mixed with the fixed cell suspension. Fluorescence microscopic images were photographed with Fuji super G 400 or TriX pan 400 (Kodak) film, captured by using Quickscan 35 (Minolta), and analyzed with NIH-Image. The deviation in image intensity among films was confirmed to be less than 10% with the fluorescence intensity of calibration beads for a flow cytometer (Bio-Rad). For electron microscopic observation, fixed cells were placed on a 180-Å grid covered by a collodion membrane. Grids were allowed to dry, negatively stained with 2% ammonium molybdate for 1.5 min, and observed with a transmission electron microscope (8).

Examination of cell viability. The viability of individual cells was examined by using thin-layer solid medium (37). The cultures were mixed with an equal volume of fresh medium not supplemented with nucleotides containing 2% low-melting-temperature agarose at 37°C. Aliquots of 100 µl were spread on slide glasses. The cell mixtures on the slide glasses were incubated at 37°C in a moist chamber. For microscopic observation, cells were fixed with 100% ethanol, followed by two washes with PBS. For fluorescence observation, cells were stained by addition of 10 µl of DAPI solution and were covered with a coverslip. Total CFUs in each culture were counted by inoculating cultures onto MEM plates as previously described (35).

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

DNA and protein content of culture. The coupling of DNA and protein syntheses was examined by monitoring the net contents. The protein contents of cultures grown in MEM followed the OD₆₀₀ through growth phase (Fig. 1C). However, the DNA content did not increase after the OD₆₀₀ reached 0.1, while it increased in parallel with OD₆₀₀ in the early-growth stage (Fig. 1A). We searched for factors capable of preventing the reduction of DNA synthesis and found that the nucleoside mix used for a synthetic medium (29) was effective. The DNA content in the supplemented cultures increased almost in parallel with OD₆₀₀, even in the later-growth stage (Fig. 1B). The protein synthesis and increase in OD₆₀₀ were not affected by this supplementation (Fig. 1D). These results showed that DNA replication is coupled with protein synthesis, if DNA replication is not inhibited by nucleoside starvation.

View larger version (24K): [in this window] [in a new window]   FIG. 1.   DNA and protein contents of batch cultures. Mycoplasma cells were cultured without (A and C) or with (B and D) nucleoside supplementation. DNA (A and B) and protein (C and D) contents in the cultures are shown by open circles. OD600 is shown by solid circles.

Replication fork velocity. To determine the time for one round of chromosome replication, we assayed the time required for completion of replication of a pulse-labeled endonuclease fragment. We used [³²P]dAMP, which can be incorporated into the chromosome (21, 24). We tested whether [³²P]dAMP taken up into a mycoplasma cell can be rapidly diluted by cold dAMP by monitoring the polymerization process of Okazaki fragments (Fig. 2). Mycoplasma cultures were labeled with 16 nM [³²P]dAMP for 2 min; then 1 mM cold dAMP was added, and DNA was isolated after various incubation periods and analyzed by denaturing gel electrophoresis. Okazaki fragments were widely distributed in size. The labeling of nascent Okazaki fragments was stopped at 10 s after the addition of cold dAMP, and the sizes of labeled small fragments started to shift. This result showed that [³²P]dAMP was available for the pulse-labeling of the chromosomal DNA.

View larger version (76K): [in this window] [in a new window]   FIG. 2.   Polymerization of labeled Okazaki fragments. Mycoplasma cells were labeled with [32P]dAMP. After 2 min, labeled dAMP was diluted with 1 mM cold dAMP. DNA was isolated at 0, 10, 20, 30, 60, 90, and 120 s after dilution and was analyzed by alkaline agarose gel electrophoresis in lanes 1 through 7, respectively. Fragment sizes are indicated on the left.

View larger version (37K): [in this window] [in a new window]   FIG. 3.   Completion of replication of the chromosomal fragments. Mycoplasma cells were labeled with [32P]dAMP for 2 min, and 1 mM cold dAMP was added. The chromosomal DNA isolated at each time point was digested with BamHI and subjected to FIGE followed by autoradiography. (A) The band intensities of Bm1 and Bm2 fragments are shown as the saturation extent by solid and open circles, respectively. The saturation extents until 40 min were fitted with the dashed line. (B) Autoradiogram of Bm1 and Bm2 fragments at 0, 5, 10, 15, 20, 25, 30, 40, 50, 60, and 70 min (lanes 1 through 11, respectively).

Origin/terminus ratio. The ratio of replicating intermediates in the chromosome molecules was monitored by titrating the origin/terminus copy number ratio. The origin/terminus ratio was estimated by the hybridization method. The origin region has been defined (19), and a DNA fragment expanding in this region was used for titrating the origin. The region, which is replicated last, has not been identified. Therefore, we used a DNA region expanding on the gyrase gene which was reported to be in the position opposite the origin on the chromosome map (34). The origin/terminus ratio was estimated to be 2.0. This value did not change through the growth phase, and it did not depend on nucleoside supplementation.

Definition of cell types. Transmission electron microscopy revealed that the morphology of M. capricolum cells was classified into two types, i.e., rod and branched types (Fig. 4A and B, respectively). Most cells in the ordinary-growth phase had a comparatively regular rod-like structure. A small fraction of cells had an irregular branched structure, with tubes radiating out from the center of the cell body. These cell types could also be distinguished by light microscopy (Fig. 4C and D).

View larger version (71K): [in this window] [in a new window]   FIG. 4.   Images of M. capricolum cells. The left and right columns show images of cells at OD600s of 0.05 and 0.4, respectively. (A and B) Electron microscopy. (C and D) Phase-contrast microscopy. (E and F) DAPI-stained cell images. Bars below panels B and F represent 2 and 5 µm, respectively. Arrowheads point to the position of constriction.

Cell type conversion. The occurrence of branched cells was examined by phase-contrast microscopy (Fig. 5A). In cultures supplemented with nucleosides, the proportion of branched cells did not change significantly until the OD₆₀₀ reached 0.17, after which it started to increase, and the branched cells finally accounted for 37% of the total cell number. In cultures without nucleoside supplementation, the proportion increased in the earlier stage. The proportion started to increase after the OD₆₀₀ reached 0.10, and the branched cells finally accounted for 90% of the total cell number. The starting point of the increase in the proportion of branched cells corresponded to the cessation of DNA synthesis due to nucleoside starvation, with subsequent protein synthesis (Fig. 1). These results suggest that nucleoside starvation induced cell type conversion.

View larger version (15K): [in this window] [in a new window]   FIG. 5.   Cell type conversion. (A) Proportions of branched cells for cultures without and with nucleoside supplementation are shown by solid and open circles, respectively. OD600s for cultures without and with nucleoside supplementation are shown by solid and open squares, respectively. (B) Proportions of branched cells in temperature-sensitive-mutant and wild-type cultures are shown by open and solid circles, respectively. The shift to the restrictive temperature was performed at time zero. OD600s for temperature-sensitive mutant and wild-type cultures are shown by open and solid squares, respectively.

Cell type conversion in a temperature-sensitive mutant deficient in DNA replication. We examined the cell type conversion of a temperature-sensitive mutant strain in which the elongation reaction of DNA replication is specifically inhibited at the restrictive temperature (21). Mutant and wild-type cells were grown at 33°C until the OD₆₀₀ reached 0.05, and then the incubation temperature was shifted to the nonpermissive temperature of 41°C (Fig. 5B). The OD₆₀₀ of mutant cells continued to increase for more than 180 min after the temperature shift. The proportion of branched cells started to increase just after the temperature shift, so that they accounted for 54% of the total cell population at 120 min after the temperature shift, while it did not increase significantly in wild-type cultures.

Characterization of rod and branched cells. The rod type cells in cultures supplemented with the nucleosides at an OD₆₀₀ of 0.05 were analyzed by electron microscopy. The average length of rod cells was 0.941 µm, with a standard deviation (SD) of 0.231 µm, and constriction sites were found around the middle in 21.8% (113 of 519) of the rod cells (Fig. 4). The average ratio of the distance from the constriction site to the furthest cell pole, relative to the cell length, was 0.576, with an SD of 0.051. The cells stained with DAPI were analyzed by fluorescence microscopy, and the DNA contents in individual cells were estimated from fluorescence intensity (Fig. 6). The average DNA content at an OD₆₀₀ of 0.05 was normalized to 1 U. The SD of DNA content in rod cells was 0.397 U. These results were not considerably different in the other growth stages or in cultures without nucleoside supplementation (data not shown).

View larger version (26K): [in this window] [in a new window]   FIG. 6.   DNA contents of individual cells. DNA contents of rod cells at an OD600 of 0.05 (A) and of branched cells at an OD600 of 0.4 (B) are shown. The amount of individual fluorescence was measured as DNA content in a cell. The average of the values shown in panel A was normalized to 1 U.

Viability of branched cells. To determine the correlation between cell division and cell type, CFU was monitored with respect to culture OD₆₀₀ (Fig. 7). The rate of increase through time paralleled that of the OD₆₀₀ until the latter reached 0.1. Thereafter, the relative rate of increase diverged and the viability as measured by CFUs was no longer reflected by the OD₆₀₀ unless the medium was supplemented with nucleosides. The point at which the rates of increase in CFUs and the OD₆₀₀ diverged corresponded to the starting point of the increase in the population of branched cells. This uncoupling was not observed in the nucleoside-supplemented cultures. These results indicate that cells ceased division upon cell type conversion. To examine the viability of branched cells, cultures at an OD₆₀₀ of 0.4 were inoculated onto a thin solid medium on glass slides and were incubated at 37°C (Fig. 8). Light and fluorescence microscopic observation at 3 and 6 h after the inoculation revealed that all cells formed microcolonies, indicating the ability of branched cells to reproduce.

View larger version (17K): [in this window] [in a new window]   FIG. 7.   CFUs in cell type conversion. Solid and open circles, CFU numbers for cultures without and with nucleoside supplementation, respectively. Solid and open squares, OD600s of cultures without and with supplementation, respectively.

View larger version (105K): [in this window] [in a new window]   FIG. 8.   Growth of branched cells in thin-layer solid medium. Cultures at an OD600 of 0.4 were mixed with low-melt agarose, spread on glass slides, and analyzed after 0 (A), 3 (B), and 6 (C) h of incubation at 37°C. In the left and right columns are images of DAPI-stained cells observed by phase-contrast and fluorescence microscopy, respectively. Bar, 5 µm.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

We showed that M. capricolum cells were starved of nucleosides in MEM, which is widely used for cultivation of mycoplasmas (28). MEM contains animal DNA, which is thought to be used as the source of nucleotides essential for growth (18). We examined the effects of DNA in the medium on the growth and incorporation of dAMP into the chromosomal DNA but did not find any effects (data not shown). Further studies are necessary to determine if this starvation is specific to M. capricolum and if there is a need for the specific DNA supplementation.

The DNA content of M. capricolum in cultures increased with the protein content without nucleoside starvation, and the DNA contents in individual rod cells did not change throughout the growth phase. These observations suggest that the initiation frequency of DNA replication agrees with the rate of increase of protein content, although it is unclear if the frequency can be modified according to the growth conditions, as has been shown for E. coli (13).

The polymerization process of Okazaki fragments in mycoplasma cells was examined in this study. We found a signal that migrated rapidly in the denaturing gel, the intensity of which increased with the chase time. It is unlikely that this signal was caused by host exonuclease activity present at the time of lysis, because the signal intensity did not depend on the time from cell lysis to the first centrifugation (data not shown). The signal was presumably derived from pseudo-Okazaki fragments caused by excision of uracil incorporated into DNA (17). A homolog of uracil N-glycosylase has also been found in genome analyses of mycoplasmas (10, 14).

Monitoring of the progression of DNA replication in mycoplasmas has been reported previously (19-21, 25). However, the replication fork velocity has not been studied, because of the difficulty of achieving a synchronous reproductive cycle in mycoplasmas. We used pulse-labeling coupled with FIGE and examined the fork velocity without synchronization (Fig. 3). The absolute velocity of the replication fork in M. capricolum was about 10 times slower than that in E. coli (13) (Table 1). This slow progression of DNA replication may be related to the slow growth of mycoplasmas. An absence of replication machinery has not been observed in the genetic components of mycoplasmas (10, 14). The time for one round of chromosome replication was estimated to be around 94 min. This value roughly corresponded to the doubling time, suggesting that DNA replication occurs in an interval between two cell divisions. This assumption was supported by the origin/terminus ratio. The value of 2.0 can be explained by the assumption that the replication procedure takes most of the time of one division interval, and consequently most DNA molecules in the culture are replicating intermediates.

M. capricolum cells were morphologically classified into two types, i.e., rod and branched (Fig. 4). The rod cells are assumed to be the reproductive form under normal-growth conditions because (i) the majority of cells in normal-growth cultures were rod type, (ii) constrictions were found in 22% of the rod cells, (iii) the rod cells had cell length and DNA content distributions suitable for reproduction by division, (iv) the DNA contents of rod cells did not change during the normal-growth phase, which agrees with the constant increase in the DNA contents of the cultures, and (v) the branched cells cannot be a stage of the ordinary cell division cycle, because the DNA contents of branched cells were not significantly larger than those of rod cells (Fig. 6). If the rod cells became branched before division, the DNA contents of branched cells should be significantly larger than that of the rod cells.

Cell type conversion was induced by starvation of nucleosides and by transfer of a temperature-sensitive DNA replication mutant to the nonpermissive temperature. These results suggest that the inhibition of DNA replication with subsequent protein synthesis induced the cell type conversion. In the conversion, cells did not divide; i.e., the increase in CFUs was extensively reduced at the beginning of the conversion, and no anucleate minicells were observed in the microscopic field (data not shown). It is likely that mycoplasma cells whose division system is ready convert to the branched type when division is inhibited by the nonreplicated chromosomal DNA.

In ordinary growing cultures, a small proportion of cells was found as the branched type, and the proportion did not depend on the growth stage. Presumably, a small fraction of rod cells occasionally converts to the branched type and then returns to the rod type. The reproductive capability of branched-type cells was confirmed by microplate observation (Fig. 8).

We propose a model for the reproductive cycle of M. capricolum (Fig. 9). In ordinary growth, the rod cells divide into two nascent cells. DNA replication occurs in a cell division interval. On the other hand, a cell whose DNA replication is inhibited by nucleoside starvation cannot undergo cell division, and it makes new projections due to the excess potential of cell division. In ordinary growth, a small fraction of rod cells which have some delay in completion of chromosome replication also convert to the branched type.

View larger version (18K): [in this window] [in a new window]   FIG. 9.   Model for reproduction of M. capricolum cells. The state of chromosomal DNA replication is indicated by a bar(s) in the center of a cell.

M. mycoides is closely related to M. capricolum (39), and the appearance of this species is also similar to that of M. capricolum (30). Buxton and Fraser (6) reported that M. mycoides develops from an elementary rounded body into a branched and filamentous form, and then new rounded elementary bodies develop within these filaments and are released when they are mature. Our conclusion is not in agreement with this hypothesis. Models of cell division into two equivalent cells were also proposed for Mycoplasma gallisepticum (23), Mycoplasma mobile (32), and Spiroplasma citri (11), which are also related to M. capricolum (39). The growth phase dependency of the morphological change of Mycoplasma pneumoniae is similar to that in M. capricolum. Cells of M. pneumoniae are spherical in young cultures, become branched and filamentous with the growth phase, and then change to asymmetrical rounded forms in declining cultures (3, 16). The morphological change in M. pneumoniae was suggested to respond to the change in growth medium components. The cell type conversion of M. pneumoniae may be caused by a process similar to that in M. capricolum.

A morphological change induced by nucleoside starvation, whereby cells became filamentous with thymine starvation, was also reported for E. coli. This process was independent of the SOS response (15). However, no branching was seen in this case. Akerlund et al. (2) showed that 5% of E. coli cells formed branches in nutrient-poor medium when chromosome replication or nucleoid segregation was genetically disturbed. These phenomena may be related to branch formation in mycoplasmas.

Since mycoplasmas lack the peptidoglycan layer, changes in the cytoplasm and cell membrane can be directly reflected in the appearance of the cells. Therefore, branch formation by mycoplasmas is probably coupled with abnormal assembly of proteins that play roles in cell division or maintenance of cell shape. This is supported by the observation that cell type conversion required subsequent protein synthesis (data not shown). FtsZ protein is known to form a Z ring at the position of septation prior to cell division and to play key roles in cell division in E. coli and other walled bacteria (1). Homologs of the ftsZ gene have been identified in some mycoplasmas, including M. capricolum (5, 10, 14, 38). FtsZ may be related to the branch formation of mycoplasmas. It has been reported that overexpression of FtsZ protein induces a branch at the stalk in Caulobacter crescentus (27).