Replisome Localization in Vegetative and Aerial Hyphae of Streptomyces coelicolor

Abstract

Using a functional fusion of DnaN to enhanced green fluorescent protein, we examined the subcellular localization of the replisome machinery in the vegetative mycelium and aerial mycelium of the multinucleoid organism Streptomyces coelicolor. Chromosome replication took place in many compartments of both types of hypha, with the apical compartments of the aerial mycelium exhibiting the highest replication activity. Within a single compartment, the number of “current” ongoing DNA replications was lower than the expected chromosome number, and the appearance of fluorescent foci was often heterogeneous, indicating that this process is asynchronous within compartments and that only selected chromosomes undergo replication.

Streptomycetes, which are gram-positive soil bacteria known for their ability to produce many valuable antibiotics and other secondary metabolites, are among the most striking examples of multicellular bacteria. Their hyphae grow by tip extension, forming a branched vegetative mycelium in which septation occurs at some distance from the growing hyphal tips (4, 23). Compartments of the vegetative hyphae contain several uncondensed copies of the large (8- to 9-Mbp), linear chromosome (2, 10). During further development, Streptomyces colonies form an aerial mycelium, in which long chains of exospores are formed. Rapidly growing aerial hyphae may contain up to 50 uncondensed chromosomes in one tip compartment. After an aerial hypha has stopped growing, a large number of FtsZ rings form a regular ladder in the long tip compartment, giving rise to sporulation septa that delimit prespore compartments (25). At this stage the chromosomes condense and are segregated into unigenomic prespore compartments. The prespore compartments eventually metamorphose into chains of physically separate spores.

The model organism Streptomyces coelicolor A3(2) is genetically the best-studied streptomycete (1). Although the S. coelicolor life cycle and its regulators have been extensively investigated, little is known about the dynamics of chromosome replication during this complex process. So far, Streptomyces chromosome replication has been studied only in the young vegetative mycelium using pulse-labeling with [³H]thymidine (17), but the localization of replisome machinery in a single compartment has not been addressed. Visualization of a replisome(s) within single compartments of both vegetative and aerial hyphae should shed some light on important features of the multinucleoid prokaryotic cell.

The use of replication proteins tagged with green fluorescent protein (GFP) has provided an opportunity for direct observation of chromosome dynamics in vivo in a single bacterial cell (11, 19). Visualization of the replicating machinery has been achieved by fusing GFP to various DNA polymerase III holoenzyme subunits, including α (PolC) (19) and τ (DnaX) (11) in Bacillus subtilis, χ (HolB) and δ′ (HolC) in Caulobacter crescentus (13), and β (DnaN) (15) in Escherichia coli. Recently, systematic localization of over 100 proteins fused to the GFP in B. subtilis allowed identification of new proteins associated with the replication machinery (21). So far, studies of the localization of the DNA replication apparatus using GFP-tagged replication proteins have focused only on unicellular rod-shaped bacteria that divide by binary fission and have a circular chromosome.

In this study we addressed for the first time the question of the localization of the replication machinery within compartments of Streptomyces vegetative and aerial hyphae. We attempted to determine whether replication is restricted to a fixed intracellular position(s) or is randomly distributed in multinucleoid compartments at different stages of S. coelicolor growth. The basic components of the DNA replication machinery are highly conserved in bacteria. As it is in other bacteria, DNA polymerase III is essential for replication of the S. coelicolor linear chromosome (5). S. coelicolor possesses a set of genes annotated with the replisome (http://www.sanger.ac.uk/Projects/S_coelicolor/), including dnaN, which encodes the β-subunit of DNA polymerase III. This gene is located in the typical eubacterial block of genes (rnpA, rpmH, dnaA, and dnaN). The β subunits dimerize to form the sliding clamp that links the core polymerase to DNA and allows DNA replication to proceed; four β subunits are present in the DNA polymerase III holoenzyme (9, 20, 22). In this study, the GFP-tagged β subunit (DnaN-enhanced GFP [EGFP]) was used to visualize the replisome machinery in S. coelicolor.

DnaN-EGFP fusion protein is replication active in S. coelicolor.

In order to localize the replication machinery in the hyphae of S. coelicolor, we constructed strain J3337, which expresses chromosomally encoded EGFP-tagged DnaN instead of the wild-type protein. To maximize the likelihood that the fusion protein would be functional, a 10-amino-acid, flexible, proline- and glycine-rich linker was used (12). PCR targeting (7, 8) was used to construct J3337. Briefly, an egfp-aac3(IV)-oriT cassette (conferring apramycin resistance [Apra^r]) was inserted downstream of dnaN in a kanamycin resistance (Kan^r)-marked cosmid, H18. This construct was used for conjugation into S. coelicolor M145. Apra^r exconjugants were screened for the loss of Kan^r, indicating that there was double-crossover allelic exchange of the dnaN locus. The presence of DnaN-EGFP in S. coelicolor cell extracts was examined by scanning sodium dodecyl sulfate-polyacrylamide gel electrophoresis-separated total proteins with a phosphorimager (Typhoon 8600 variable mode imager) equipped with a 488-nm laser as described previously (12). A fusion protein that was the expected size (68 kDa) was identified in the cell extract (Fig. 1A). In liquid media the growth rate of the strain obtained (J3337) was similar to that of the wild type, but on solid media J3337 grew more slowly (data not shown). To examine the subcellular localization of ongoing DNA replication in the vegetative and aerial hyphae, J3337 was inoculated at the acute-angle junction of coverslips inserted at a 45° angle in MM agar containing 1% mannitol (14). After 26 h (vegetative mycelium) and 42 or 66 h (aerial mycelium), the presence of fluorescence foci was analyzed using a Nikon Eclipse 600 or Zeiss Axio Imager Z1 microscope equipped with a ×100 objective. Green fluorescent foci were seen in both vegetative and aerial hyphae (see below). No such foci were seen in control strain S. coelicolor M145(pIJ8641) (J. Sun, unpublished data) expressing EGFP (27 kDa).

FIG. 1.

Analysis of Streptomyces strains expressing DnaN-EGFP. (A) Time course of DnaN-EGFP production in S. coelicolor (left) and expression of DnaN-EGFP in S. lividans TK24 after treatment with novobiocin (right). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis-separated total proteins were scanned with a phosphorimager. Strains were grown on MM agar containing 1% mannitol (S. coelicolor) or in YEME-TSB liquid medium (S. lividans) at 30°C (14). Lane WT, wild-type strain; lane M, molecular weight markers; lane egfp, S. coelicolor M145(pIJ8641) expressing unfused EGFP (27 kDa). (B) Effect of DNA replication on the presence of replisome foci. S. lividans TK24 dnaN-egfp cultures were treated with novobiocin (final concentration, 200 μg/ml) for 120 min (middle panels) or with kanamycin (final concentration, 50 μg/ml) for 120 min (right panels). The left panels contained untreated cells. The strain was grown in 2× YT medium (13) at 30°C. Scale bars, 5 μm.

Presence of replisome foci requires ongoing DNA replication.

Fluorescent foci of DnaN-EGFP are expected to be observed only during the stages of S. coelicolor development, including germinating spores (Fig. 1B and 2 A), when DNA replication takes place. Indeed, no foci were observed in chains of spores (Fig. 2E), and the level of the DnaN-EGFP fusion protein was perceptibly decreased in older colonies (after 108 h) (Fig. 1A). To examine whether the presence of fluorescent foci requires ongoing DNA replication, we used novobiocin, a well-known inhibitor of the gyrase that introduces the negative supercoils required for DNA replication to proceed (16). Since S. coelicolor is naturally resistant to novobiocin, we chose for this purpose a closely related novobiocin-sensitive strain, Streptomyces lividans TK24 (24). Using the S. coelicolor cosmid derivative described above, we introduced the dnaN-egfp fusion into the S. lividans TK24 chromosome (much of the S. lividans chromosome is about 99% identical at the DNA sequence level to the chromosome of S. coelicolor). In the S. lividans TK24 strain producing the DnaN-EGFP fusion protein (Fig. 1A), fluorescent foci were formed in both types of hyphae with the same distribution and intensity with which they were formed in the equivalent S. coelicolor strain (data not shown). After novobiocin was added to a liquid culture of young developing mycelia (120 min, 200 μg/ml), the fluorescent foci disappeared; only diffuse, albeit quite bright, fluorescence was observed (Fig. 1B). In contrast, the presence of kanamycin (an inhibitor of protein synthesis) did not affect focus formation (Fig. 1B). This supports the idea that the formation of fluorescent foci requires ongoing DNA replication. The inhibition of chromosome replication by novobiocin was reversible; the fluorescent foci appeared again after removal of novobiocin from the medium (data not shown). Interestingly, the addition of novobiocin caused an increase in the DnaN-EGFP level (Fig. 1A), resulting in the bright, dispersed fluorescence observed in the emerging germ tubes (Fig. 1B) (the level of DnaN-EGFP remained constant in the cells not treated with novobiocin [data not shown]). This suggests that inhibition of replication may increase the level of dnaN transcription; notably, the two dnaN promoters (27) are located within the oriC region. Thus, these data show clearly that the observed fluorescent foci represent replisomes but not “storage structures” of nonactive β subunits; formation of such structures should be expected in nonreplicative cells (particularly in nonreplicative cells “overexpressing” DnaN-EGFP) rather than in replication-active cells.

FIG. 2.

Localization of replicative DNA polymerase in the vegetative mycelium and aerial mycelium of S. coelicolor. S. coelicolor dnaN-egfp was grown for 26 h (vegetative mycelium) or for 42 to 66 h (aerial mycelium). The images show overlays of the DnaN-EGFP (green) fluorescence and cell walls stained with WGA (red) or DNA stained with 7-aminoactinomycin D (only in the case of spore chains). Scale bars, 5 μm. The diagram in the center shows the positions of the different types of compartments analyzed in the vegetative mycelium and aerial mycelium. (A) Germinating spores; (B) vegetative hyphae; (C) aerial hyphae; (D) apical compartments of aerial hyphae; (E) prespores and spores. The numbered arrowheads indicate examples of different types of fluorescence, as follows: arrowhead 1, bright foci; arrowhead 2, diffuse foci; arrowhead 3, dispersed fluorescence; arrowhead 4, lack of fluorescence. Red and yellow arrowheads indicate cross walls and tips of hyphae, respectively.

Comprehensive analysis of the S. coelicolor chromosome sequence revealed an additional gene (SCO1180) that encodes a putative DNA polymerase III beta chain (with a C terminus similar to Mycobacterium leprae DNA polymerase III beta chain; 30.7% identity in 358-amino-acid overlap). SCO1180 is located in the left arm of the linear chromosome (outside the typical eubacterial block of replication genes). Deletion of SCO1180 had no effect on J3337 growth or the formation of fluorescent foci, suggesting that this gene is not directly involved in chromosome replication (data not shown).

Chromosome replication is not uniform in the vegetative mycelium.

We observed different types of fluorescence in different hyphal compartments, from bright foci to diffuse foci (often both were present in the same compartment) and from dispersed fluorescence to a lack of fluorescence (Fig. 2 shows examples of different types of fluorescence). We assumed that the diffuse or dispersed fluorescence was a result of the different stages of assembly and disassembly of the replication machinery. To relate the positions of the different forms of DnaN-EGFP fluorescence to the positions of compartments in the mycelium, we stained the cell wall with wheat germ agglutinin-tetramethylrhodamine conjugate (WGA) as described previously (25).

In the vegetative mycelium (Fig. 2B), only 22% of the compartments analyzed had compact bright foci (8%) or mixed foci (bright and diffuse) (14%) (Fig. 3), indicating that these compartments were replication active. In 60% of the vegetative hyphal compartments, the foci were diffuse (27%) or fluorescence was evenly dispersed through the entire compartment (33%), indicating that replisomes were at various stages of disassembly, while 18% of the vegetative compartments analyzed lacked any green fluorescence, suggesting that DNA replication was not taking place in them and indicating that there was rather rapid turnover of the fluorescent protein. In contrast, Kummer and Kretschmer (17) reported that vegetative mycelium did not contain compartment-sized regions without DNA synthesis. However, it should be noted that the cell walls were not visualized in their study; therefore, it was impossible to examine DNA synthesis in relation to the single compartment. Moreover, the mycelia were labeled with [³H]thymidine for up to 2 hours, while the replication of the entire S. coelicolor chromosome lasts approximately 90 to 100 min (18). If during the experiment the replication forks were disassembled, the newly synthesized DNA strands would have already been labeled. Thus, the study of Kummer and Kretschmer did not reflect the “current” replication activity.

FIG. 3.

Types of DnaN-EGFP fluorescence in the vegetative and aerial hyphae of S. coelicolor. In the analysis 256 and 183 compartments of the vegetative mycelium and the aerial mycelium, respectively, were examined.

Cell wall staining (Fig. 2) did not indicate any reproducible pattern of distribution of foci within the compartments analyzed. Similarly, Yang and Losick (26) observed no regular distribution of oriC foci within the vegetative hyphae. Vegetative compartments containing bright green fluorescence foci usually had from three to five foci with variable spacing that ranged from 0.5 to 7 μm but sometimes was even greater. In vegetative hyphae, cell wall growth occurs mostly at hyphal tips (6), suggesting that the tip region might also be the major site of DNA replication (26). However, we observed no difference in the distribution or number of foci between the tips and the rest of the vegetative hyphae (including branches). The spacing between the apical replisome and the tip was quite variable; most apical foci (75%) were more than 1.5 μm from the tip. We showed previously that in vegetative hyphae, the origin of replication (oriC) of apical chromosomes in most hyphal tip compartments is bound by ParB protein, which is involved in chromosome segregation, the distance between the apical ParB complex and the tip being fairly constant (1.4 μm) (12). Other ParB foci occurred irregularly in vegetative hyphae. This is consistent with (but does not prove) the hypothesis that ParB association prevents replisome formation and arrests further replication from the origin to which it is attached.

Different pattern of replisome formation in aerial hyphae.

In older cultures (66 h) we could clearly distinguish regions of very weak replication activity in what we believe were older parts of the vegetative hyphae, as well as regions where most of the hyphal compartments contained bright foci. The latter represented intensively growing unbranched aerial hyphae that were easily recognized by their strong staining with WGA (12). Apical compartments of the aerial mycelium (Fig. 2D) always contained fluorescent foci; 23% of them contained only bright foci, 45% contained mixed bright and diffuse foci, and the rest contained diffuse foci (Fig. 3). Thus, the aerial apical compartments were much more replication active than other compartments of the mycelium (Fig. 3). Interestingly, 31% of subapical “second” compartments showed no fluorescence (Fig. 2E). Generally, prespores and spores in chains did not exhibit any fluorescence foci (Fig. 2E); only diffuse fluorescence was observed, presumably the result of disassembly of the replication machinery, and the intensity diminished as the spore chains matured. Foci reappeared in the initial stage of spore germination; at this stage germinating spores contained double or even triple bright fluorescent foci, suggesting that in these spores DNA replication had already started (Fig. 2A). As germination proceeded, emerging germ tubes possessed increasing numbers of foci (up to five or six foci) (Fig. 1B and 2A).

The aerial apical compartments contained more foci (on average, six foci) than any other compartments of the mycelium. The number of foci in these compartments varied, and some apical had more than 10 bright foci (one example had 14 foci within a compartment about 35 μm long). In shorter and younger compartments (10 to 20 μm) the foci were more densely spaced than they were in longer and older hyphae (20 to 40 μm), as if the latter hyphae were reaching the growth limit and replication was stopping. Statistically, the distance between an apical focus and the tip in the aerial hyphae, particularly in younger compartments, was shorter than the corresponding distance in the vegetative mycelium; in the majority of apical foci (64%) the distances ranged from 0 to 1.5 μm. Most likely the fast growth of apical compartments of aerial hyphae and the associated demand for increased chromosome number require an increase in replisome number within one compartment. As in the case of vegetative hyphae, the formation and distribution of DnaN-EGFP foci in developing aerial hyphae seemed to complement, rather than echo, the distribution of ParB complexes (12); ParB-EGFP foci formed in apical compartments only when replication of chromosomes was completed and were associated with every copy of the chromosome. Again, this suggests that the formation of ParB complexes may compete with the assembly of replisomes. The distances between DnaN-EGFP foci were irregular in aerial compartments, whether they were apical or subapical.

Asynchrony of replication within compartments.

The mixture of bright and diffuse DnaN-EGFP foci seen in a substantial proportion (45%) of the apical compartments analyzed, and to a lesser extent in other compartments, suggests that the replication of chromosomes was asynchronous within these compartments. Moreover, within a single compartment consisting of both vegetative and aerial hyphae the number of bright compact fluorescent foci was substantially less than the chromosome number; a single aerial apical compartment could contain up to 50 chromosomes, while we observed a maximum of 14 bright foci per compartment. Only 10% of compartments contained more than 10 foci. Since each focus should be associated with the formation of two daughter chromosomes from one parent, it is expected that the number of replisomes should not exceed one-half the final number of chromosomes and should therefore seldom exceed about 25. Although it is possible that each fluorescent focus represents the replication of two or more chromosomes, the low apparent ratio of replisomes to chromosomes, coupled with the asynchrony of replication indicated by the heterogeneous appearance of foci in many compartments, leads us to think that not all chromosomes in actively growing compartments are associated with replisomes at any particular time.

Conclusions.

Streptomyces chromosome replication takes place along much of the vegetative and aerial hyphae, although the rapidly extending apical compartments of the aerial mycelium have more active replication than any other compartments, consistent with the particularly rapid rate of extension of these compartments suggested by Chater and Losick (3). Within a single compartment, the number of ongoing DNA replications seems to be lower than the chromosome number, indicating that this process is asynchronous and that only selected chromosomes undergo replication at any one time. Thus, replication appears to follow the Jesuit dictum “many are called, but few are chosen.” Clarification of the mechanism of origin selection in this multinucleoid organism would be of considerable interest.